Osteoporosis, 2nd Edition (2-Volume Set)

xvii

CHAPTER 1 Chapter Title

Contributors

Numbers in parentheses indicate the volume number and pages on which the authors’ contributions begin.

John P. Bilezikian (2:71) Departments of Medicine and Pharmacology, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Peter V. N. Bodine (1:305) Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087 Henry Bone (2:533) Michigan Bone and Mineral Clinic, Detroit, Michigan 48236 Jean-Philippe Bonjour (1:621) Department of Internal Medicine, Division of Bone Diseases (World Health Organization Collaborating Center for Osteoporosis and Bone Diseases), University Hospital, Geneva CH-1211, Switzerland Adele L. Boskey (1:107) Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021 Mary L. Bouxsein (1:509) Orthopedic Biomechanics Laboratory, Department of Orthopedic Surgery, Beth Israel Deaconess Medical Center, and Harvard Medical School, Boston, Massachusetts 02215 Alan L. Burshell (2:195) Department of Internal Medicine, Alton Ochsner Medical Foundation, Ochsner Clinic, New Orleans, Louisiana 70131 Elizabeth Capezuti (1:795) Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, Georgia 30322 Dennis R. Carter (1:471) Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304 Jane A. Cauley (1:741) University of Pittsburgh, Pittsburgh, Pennsylvania 15261

M. Arlot (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Laura K. Bachrach (2:151) Department of Pediatrics, Stanford University School of Medicine, Stanford, California 94305 Daniel Baran (2:229) Departments of Medicine, Orthopedics, and Cell Biology, University of Massachusetts Medical Center, and Merck & Company, Inc., Worcester, Massachusetts 01655 M. Janet Barger-Lux (2:59) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 Elizabeth Barrett-Conner (1:819) Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California 90293 David J. Baylink (1:405; 2:675) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Gary S. Beaupré (1:471) Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304 Belinda R. Beck (1:701) Griffith University, School of Physiotherapy and Exercise Science, Queensland 9726, Australia Daniel D. Bikle (2:237) Department of Medicine, University of California, San Francisco, School of Medicine, and Department of Veterans Affairs, San Francisco Veterans Affairs Medical Center, San Francisco, California 94121

xvii

xviii

CONTRIBUTORS

Jacqueline R. Center (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia P. Chavassieux (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Di Chen (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Michael Chorov (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 Roberto Civitelli (2:651) Division of Bone and Mineral Diseases, Washington University School of Medicine, and Barnes-Jewish Hospital, St. Louis, Missouri 63110 Cyrus Cooper (1:557) MRC Environmental Epidemiology Unit, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, England Felicia Cosman (2:577) Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993 Gilbert J. Cote (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 Ann B. Cranney (2:539) Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada K1Y 1J8 Sarah Dallas (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Bess Dawson-Hughes (2:545) Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, Massachusetts 02111 Chris de Laet (1:585) Institute for Medical Technology Assessment, Erasmus University, Rotterdam 3000 DR, The Netherlands Arthur A. DeCarlo (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Pierre D. Delmas (2:459) INSERM Research Unit 403, Hôpital E. Herriot, and Claude Bernard University, Lyon 69003, France Marc K. Drezner (2:479) Department of Medicine, University of Wisconsin Medical School, Madison, Wisconsin 53792 Thomas A. Einhorn (1:3) Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118

John A. Eisman (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Sol Epstein (2:327) Roche Laboratories, Nutley, New Jersey 07110 Kenneth G. Faulkner (2:433) GE Medical Systems, Madison, Wisconsin 53717 David Feldman (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305 Lorraine A. Fitzpatrick (2:259) Division of Endocrinology and Metabolism and Internal Medicine, Mayo Clinic and Foundation, Rochester, Minnesota 55905 H. Fleisch (1:449) University of Berne, CH-3008 Berne, Switzerland Robert F. Gagel (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 Patrick Garnero (2:459) INSERM Research Unit 403, Hôpital E. Herriot, and Synarc, Lyon 69003, France Harry K. Genant (2:411) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 Jayashree A. Gokhale (1:107) Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021 Deborah T. Gold (2:479) Departments of Psychiatry and Behavioral Sciences, Sociology, Psychology, and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Steven R. Goldring (2:351) Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, and New England Baptist Bone and Joint Institute, Harvard Institutes of Medicine, Boston, Massachusetts 02215 Gail A. Greendale (1:819) Department of Medicine, Division of Geriatrics, University of California, Los Angeles, School of Medicine, Los Angeles, California 90095 Jeane Ann Grisso (1:795) Center for Clinical Epidemiology and Biostatistics, Division of General Internal Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104 Coleman Gross (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305

xviii

xix

CONTRIBUTORS

Gordon H. Guyatt (2:539) Department of Medicine and Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada L8N 3Z5 Reinhard Gysin (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Robert P. Heaney (1:669; 2:513) Creighton University, Omaha, Nebraska 68178 Hunter Heath III (2:259) U.S. Medical Division, Eli Lilly and Company, Indianapolis, Indiana 46285 Michael H. Heggeness (2:485) Department of Orthopaedic Surgery, Center for Spinal Disorders, Baylor College of Medicine, Houston, Texas 77030 N. Kathryn Henderson (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Ana O. Hoff (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030 Siu L. Hui (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Marjorie K. Jeffcoat (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Michael Jergas (2:411) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 C. Conrad Johnston (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Stefan Judex (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 Gerard Karsenty (1:213) Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030 Jennifer L. Kelsey (1:535) Department of Health, Research, and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Sundeep Khosla (2:49; 2:709) Department of Internal Medicine, Division of Endocrinology and Metabolism, Mayo Clinic and Foundation, Rochester, Minnesota 55905

Donald B. Kimmel (2:29) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 B. Jenny Kiratli (2:207) Spinal Cord Injury Center, Palo Alto Veterans Affairs Health Care System, Palo Alto, California 94304 Michael Kleerekoper (2:403) Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201 Robert F. Klein (2:103) Bone and Mineral Research Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Lynn Kohlmeier (2:341) Spokane Osteoporosis Center, Endocrine Associates of Spokane, Spokane, Washington 99204 Barry S. Komm (1:305) Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087 K.-H. William Lau (2:675) Departments of Medicine and Biochemistry, Jerry L. Pettis Veterans Affairs Medical Center, and Loma Linda University, Loma Linda, California 92357 Cassandra A. Lee (1:3) Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118 Gary M. Leong (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Jane B. Lian (1:21) Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655 Robert Lindsay (2:577) Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993 Kenneth W. Lyles (2:479) Departments of Medicine and Aging Center, Sarah W. Stedman Center for Nutritional Studies, GRECC, Veterans Affairs Medical Center, Duke University Medical Center, Durham, North Carolina 27710 Sharmilla Majumdar (2:3) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 Peter J. Malloy (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305 W. J. Maloney (2:385) Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis, Missouri 63110 Robert Marcus (2:3; 2:341) Veterans Affairs Medical Center, Palo Alto, California 94304; and Department of Medicine, Stanford

xix

xx

CONTRIBUTORS

University School of Medicine, Stanford, California 94305 Thomas J. Martin (1:361) St. Vincent’s Institute of Medical Research, Melbourne 3065, Australia Kenneth B. Mathis (2:485) Department of Orthopaedic Surgery, Center for Spinal Disorders, Baylor College of Medicine, Houston, Texas 77030 Kenneth J. McLeod (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 L. Joseph Melton III (1:557; 2:49) Department of Internal Medicine, Division of Endocrinology and Metabolism, and Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905 P. J. Meunier (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Subburaman Mohan (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Lis Mosekilde (2:725) Department of Cell Biology, Institute of Anatomy, University of Aarhus, DK-8000, Aarhus C, Denmark Douglas B. Muchmore (2:603) Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285 Gregory R. Mundy (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Ran Namgung (1:599) Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea Dorothy Nelson (1:569) Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201 Lorene Nelson (1:569) Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Robert A. Nissenson (1:221) Endocrine Unit, San Francisco Veterans Affairs Medical Center, and Departments of Medicine and Physiology, University of California, San Francisco, San Francisco, California 94121 Bjorn R. Olsen (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115

Eric S. Orwoll (1:339; 2:103) Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Babatunde Oyajobi (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Roberto Pacifici (2:85) Division of Bone and Mineral Diseases, Washington University, St. Louis, Missouri 63110 Charles Y. C. Pak (2:699) Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center, Dallas, Texas 75390 Socrates E. Papapoulos (2:631) Department of Endocrinology and Metabolic Diseases, University of Leiden Medical Center, 2333 ZA Leiden, The Netherlands A. Michael Parfitt (1:433) Division of Endocrinology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205 Millan S. Patel (1:213) Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030 Huibert A. P. Pols (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Richard Prince (2:621) Department of Medicine, University of Western Australia, Sir Charles Gairdner Hospital, Perth 6009, Western Australia Xuezhong Qin (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Yi-Xian Qin (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 Lawrence G. Raisz (2:19) Department of Endocrinology and Metabolism, University of Connecticut Health Center, Farmington, Connecticut 06032 Robert R. Recker (2:59) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 Michael S. Reddy (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Jonathan Reeve (1:585; 2:725) Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Cambridge CB2 2QQ, United Kingdom

xx

xxi

CONTRIBUTORS

Anthony M. Reginato (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115; and Arthritis Unit, Massachusetts General Hospital, Boston, Massachusetts 02114 Ian R. Reid (2:553) Department of Medicine, The University of Auckland, Auckland 1, New Zealand B. Lawrence Riggs (2:49) Department of Internal Medicine, Division of Endocrinology and Metabolism, Mayo Clinic and Foundation, Rochester, Minnesota 55905 Rene Rizzoli (1:621) Department of Internal Medicine, Division of Bone Diseases (World Health Organization Collaborating Center for Osteoporosis and Bone Diseases), University Hospital, Geneva CH-1211, Switzerland Pamela Gehron Robey (1:107) Bone Research Branch, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892 Gideon A. Rodan (1:361) Department of Bone Biology and Osteoporosis Research, Merck Sharp & Dohme Research Laboratories, West Point, Pennsylvania 19486 Clifford Rosen (2:747) Maine Center for Osteoporosis Research and Education, St. Joseph Hospital, Bangor, Maine 04401 Michael Rosenblatt (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 F. Patrick Ross (1:73) Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110 Clinton T. Rubin (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, and the Center for Biotechnology, State University of New York, Stony Brook, New York 11794 Loran M. Salamone (1:741) University of Pittsburgh, Pittsburgh, Pennsylvania 15261 Adina Schneider (2:303) Mount Sinai Hospital, New York, New York 10029 D. J. Schurman (2:385) Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305 Ann V. Schwartz (1:795) Department of Epidemiology and Biostatistics, University of California, San Francisco, School of Medicine, San Francisco, California 94143 Ego Seeman (1:771) Austin and Repatriation Medical Centre, University of Melbourne, Heidelberg, Melbourne 3084, Australia

Elizabeth Shane (2:303; 2:327) College of Physicians and Surgeons of Columbia University, New York, New York 10032 Jay R. Shapiro (2:271) Kennedy Krieger Institute, Baltimore, Maryland 21224; and Uniformed Services University, Bethesda, Maryland 20814 Janet Shaw (1:701) Departments of Exercise and Sport Science, University of Utah, Salt Lake City, Utah 84112 Kathy M. Shipp (2:479) Departments of Physical Therapy and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Shonni J. Silverberg (2:71) Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Ethel S. Siris (2:603) Columbia University College of Physicians and Surgeons, and Toni Stabile Center for the Prevention and Treatment of Osteoporosis, Columbia – Presbyterian Medical Center, New York, New York 10032 Charles W. Slemenda† (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Steven R. Smith (2:195) Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana 70808 R. L. Smith (2:385) Orthopedic Research Laboratory, Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305; and Veterans Affairs Medical Center, Palo Alto, California 94304 Christine M. Snow (1:701) Bone Research Laboratory, Oregon State University, Corvallis, Oregon 97331 MaryFran Sowers (1:535; 1:721) Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan 48109 Bonny L. Specker (1:599) Ethel Austin Martin Program in Human Nutrition, South Dakota State University, Brookings, South Dakota 57007 Gary S. Stein (1:21) Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655 S. Aubrey Stoch (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 †

Deceased.

xxi

xxii

CONTRIBUTORS

Steven L. Teitelbaum (1:73) Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110 Kathy Traianedes (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Reginald C. Tsang (1:599) Department of Pediatrics, University of Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio 45229 André G. Uitterlinden (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Marjolein C. H. van der Meulen (1:471) Sibley School of Mechanical and Aerospace Engineering, Cornell University, Ithaca, New York 14853; and Biomechanics and Biomaterials, Hospital for Special Surgery, New York, New York 10021

Johannes P. T. M. van Leeuwen (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Marie Luz Villa (1:569) Department of Medicine, Division of Gerontology and Geriatrics, University of Washington School of Medicine, Seattle, Washington 98104 WenFang Wang (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115 Kristine M. Wiren (1: 339) Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Christian Wüster (2:747) Department of Medicine, Novo Nordisk Pharma, 55127 Mainz, Germany

xxii

xxiii

CHAPTER 1 Chapter Title

Preface

It will come as no surprise to the reader that this second edition of Osteoporosis substantially outweighs its predecessor. The reason for this increase is the astonishing progress in osteoporosis medicine that has taken place during the 5 years since publication of the first edition. Major advances have occurred at both the basic science and clinical levels, and we have sought to incorporate them into the second edition. Some illustrative areas where new insights have emerged are skeletal differentiation and regulation, particularly regarding complementary actions of Hedgehog and PTH-related proteins; the RANK/RANK-ligand osteoprotegerin system as a central feature of osteoclast biology; the skeletal phenotypes of mouse gene knockout models; the discovery of a second estradiol receptor which is relatively enriched in bone; and the molecular basis of bisphosphonate action. These developing areas represent new and important conceptual themes that were unknown in 1995, but which have been accorded detailed consideration in the second edition. Many of the chapters devoted to basic science emphasize new concepts that offer novel molecular targets for future therapeutics. Great progress has also been made in clinical practice. When the first edition was published, therapeutic choices were confined to several antiresorptive drugs. Few controlled clinical trials had been conducted for any agent, and those that had been published relied on the surrogate end point of bone mineral density, rather than on fracture incidence. Large randomized controlled trials using fracture incidence as the primary outcome have now become the industry standard, and results of such trials have led to the registration of new agents for the prevention and treatment of postmenopausal osteoporosis, glucocorticoid-associated osteoporosis, and osteoporosis in men. Perhaps most exciting, the recent demonstration that parathyroid hormone substantially increases bone mineral density and reduces fracture incidence validates the concept of skeletal anabolic

therapy and offers for the first time a potential approach to the eventual cure of osteoporosis. In response to these and other developments, new chapters have been introduced and others have been expanded, updated, or divided. New chapters include contributions on the developmental biology of bone, gene knockout models, major European epidemiological fracture studies, micro-CT assessment of bone architecture, evidence-based analysis of osteoporosis therapy, regulatory considerations for design of osteoporosis trials, skeletal effects of phytoestrogens and selective estradiol receptor modulators, and novel approaches to osteoporosis therapeutics. In the interest of maintaining freshness of ideas, we initiated our own remodeling process by creating a modest degree of authorship turnover. We thank previous contributors for their excellent work and forewarn them that we may call on them again for future editions. We note with sadness the passing of two prominent members of the osteoporosis community who were involved with the first edition of this book, Charles W. Slemenda and Louis V. Avioli. Despite his young age, Charlie provided many valuable insights into multiple aspects of osteoporosis. Lou Avioli elected not to write a chapter, but we benefited greatly from his advice regarding content and authors as well as from his enthusiasm for the project. We are pleased to acknowledge once again the efforts of Dr. Jasna Markovac of Academic Press for her unflagging interest and inspiration for this project. We also greatly appreciate the Academic Press editorial staff, particularly Mica Haley and Jenny Wrenn, for their cheerful hard work and enthusiasm no matter how short the deadline or late the chapter. ROBERT MARCUS DAVID FELDMAN JENNIFER KELSEY Stanford, California

xxiii

CHAPTER 1

The Bone Organ System Form and Function CASSANDRA A. LEE AND THOMAS A. EINHORN Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118

I. II. III. IV.

Introduction Composition and Organization of Bone Cellular Control of Bone Homeostasis Bone Modeling and Remodeling

V. Bone Biomechanics VI. Summary References

I. INTRODUCTION

its simplest sense, Wolff’s law suggests that “form follows function’’ [3]. To fulfill these structure/function relationships adequately, bone is constantly being broken down and rebuilt in a process called remodeling. The cellular link between bone resorbing cells or osteoclasts, and bone forming cells or osteoblasts, is known as coupling. How bone resorption and bone formation are linked is not entirely understood (see also Chapter 12), but the consequences of accentuating one or the other preferentially leads to disease. Too much bone resorption at the expense of formation results in osteoporosis, a loss of bone strength and integrity, resulting in fractures after minimal trauma. However, under normal states of bone homeostasis, the remodeling activities in bone serve to remove bone mass where the mechanical demands of the skeleton are low and form bone at those sites where mechanical loads are transmitted repeatedly. Hence, bone is a well-designed organ system whose ability to maintain itself depends on the integrated processing of external mechanical input and physiological signals from the systemic environment and the transduction of these demands into cellular and chemical events.

Bone is a vital, dynamic connective tissue that has evolved to reflect a balance between its two major functions, provision of mechanical integrity for locomotion and protection and involvement in the metabolic pathways associated with mineral homeostasis. In addition, bone is the primary site of hemopoiesis, and recent findings support its important role as a component of the immune system [1,2]. Since the observations of Galileo, it has been assumed that the inherent architecture of bone is influenced by the mechanical stresses associated with normal function. A more formal definition of these structure/function relationships was provided by German anatomists and engineers during the late 19th century in what has since been known as Wolff’s law [3]. The tenets of Wolff’s law were based on a recognition of the correlation between the patterns of trabecular alignment in bone and the directions of the principal stresses, which were estimated to occur during normal skeletal function. Under these physiological conditions, the structure/function relationships observed in bone, coupled with its role in maintaining mineral homeostasis, strongly suggest that it is an organ of optimum structural design. In

OSTEOPOROSIS, SECOND EDITION VOLUME 1

3

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

4

LEE AND EINHORN

II. COMPOSITION AND ORGANIZATION OF BONE Bone is a composite material composed of an organic and an inorganic phase. By weight, approximately 70% of the tissue is mineral or inorganic matter, water comprises 5 to 8%, and the organic or extracellular matrix makes up the remainder. Approximately 95% of the mineral phase is composed of a specific crystalline hydroxyapatite, and this is impregnated with impurities, which make up the remaining 5% of the inorganic phase. Ninety-eight percent of the organic phase is composed of type I collagen and a variety of noncollagenous proteins; cells make up the remaining 2% of this phase [4].

A. Organic Phase The organic phase of bone plays a wide variety of roles, determining the structure and the mechanical and biochemical properties of the tissue. Growth factors and cytokines, and extracellular matrix proteins such as osteonectin, osteopontin, bone sialoprotein, osteocalcin, proteoglycans, and other phosphoproteins and proteolipids, make small contributions to the overall volume of bone and major contributions to its biologic function (see also Chapter 4). Collagen is a ubiquitous protein of extremely low solubility, which consists of three polypeptide chains composed of approximately 1000 amino acids each. It is the major structural component of the bone matrix; constructed in the form of a triple helix of two identical 1(I) chains and one unique 2(I) chain cross-linked by hydrogen bonding between hydroxyproline and other charged residues. This produces a very rigid linear molecule 300 nm in length. Each molecule is aligned with the next in a parallel fashion in a quarter-staggered array to produce a collagen fibril. The collagen fibrils are then grouped in bundles to form the collagen fiber. Within the collagen fibril, gaps, called “hole zones,” exist between the ends of the molecules. In addition, “pores’’ exist between the sides of parallel molecules (Fig. 1). Noncollagenous proteins or mineral deposits can be found within these spaces, and mineralization of the matrix is thought to be initiated in the hole zones. Several noncollagenous proteins have been described in bone. One of the more extensively studied of these in bone is osteocalcin or bone-carboxyglutamic acid-containing protein (bone Gla protein). This is a small (5.8 kDa) protein in which three glutamic acid residues are carboxylated as a result of a vitamin K-dependent post translational modification. The carboxylation of these residues confers on this protein calcium and mineral binding properties. Osteocalcin accounts for 10 to 20% of the noncollagenous protein present in bone and is closely associated with the mineral phase. While the function of this bone-specific protein is

FIGURE 1 Collagen fiber and fibril structure showing putative locations of pores and hole zones. Reprinted with permission from T. A. Einhorn, Bone metabolism and metabolic bone disease, In “Orthopaedic Knowledge Update 4” (J. W. Frymoyer, ed.), pp. 69 – 88. Amer. Acad. Orthop. Surg., Rosemont, IL (1993).

not known, it is thought to play some role in attracting osteoclasts to sites of bone resorption. It may also regulate the rate of mineralization or the final shape assumed by mineral crystals. Other noncollagenous proteins found in bone may also be important to their calcium and mineral-binding properties. In addition, several of the bone matrix proteins, such as osteopontin, bone sialoprotein, bone acidic glycoprotein, thrombospondin, and fibronectin, contain arginine – glycine – aspartic acid (RGD) sequences. Such amino acid sequences, characteristic of cell-binding proteins, are recognized by a family of cell membrane proteins known as integrins. The integrins span the cell membrane and provide a link between the extracellular matrix and the cytoskeleton of the cell. Integrins on osteoblasts, osteoclasts, and fibroblasts provide a means for anchoring these cells to the extracellular matrix. Once anchored, the cells are then enabled to express their phenotype and conduct the type of activities that characterize their funcions [5]. Growth factors and cytokines such as transforming growth factor- (TGF-), insulin-like growth factor (IGF), osteoprotegerin (OPG), the tumor necrosis factors (TNFs), the interleukins, and the bone morphogenetic proteins (BMPs 2 – 10) are present in very small quantities in bone matrix. Such proteins have important effects regulating bone cell differentiation, activation, growth, and turnover (see Chapter 14). It is also likely that these growth factors serve as coupling factors that link the processes of bone formation and bone remodeling (Table 1).

5

CHAPTER 1 The Bone Organ System

TABLE 1

Noncollagenous Proteins of the Extracellular Matrix

Structural matrix molecules Osteocalcin

Vitamin K dependent. Made by osteoblasts. May inhibit mineral deposition. Regulates activity of osteoclasts and their precursors. May mark turning point between bone formation and bone resorption. Restricted to the osteogenic lineage

Osteopontin

Contains RGD site. Anchors osteoclasts to bone. Supports cell attachments. Possibly inhibits mineralization. May regulate tissue repair and proliferation. Highly expressed in bone and inflammatory tissue

Bone Sialoprotein

Made by osteoblasts and hypertrophic chondrocytes. May initiate mineralization. Supports cell attachment. Binds Ca2 with high affinity. Restricted to skeletal lineage

Matrix Gla protein

Vitamin K dependent. May inhibit mineralization. May function in cartilage metabolism. Expressed in a variety of connective tissues

Chondroiton sulfate proteoglycan I (decorin)

Regulation of collagen fibrillogenesis. Found in extracellular matrix space

Chondroiton sulfate proteoglycan II (biglycan)

May bind to collagen. Regulates mineralization in vitro. Pericellular location

Fibronectin

Contains RGD site. Binds to numerous cell types, fibrin, heparin, gelatin, and collagen. Expressed in variety of connective tissues

Thrombospondin

Contains RGD site. Functions in cell attachment. Binds heparin, platelets, type I and V collagen, thrombin, fibrinogen, laminin, plasminogen, plasminogen activator inhibitor, histidine-rich glycoprotein. Expressed in variety of connective tissues

Osteonectin

Strong affinity for Ca2 and hydroxyapatite. Binds to growth factors. Potential association with osteoblast growth and/or proliferation. May play role in matrix mineralization. Expressed in a variety of connective tissues

Enzymatic matrix modifiers Matrix metalloproteinase (MMP) 9

Regulates growth plate angiogenesis and apoptosis. Degrades components of the ECM. Cleaves type IV, V, and XI collagens, elastin

Lysyl oxidase

Copper-dependent extracellular enzyme that catalyzes oxidative deamination of elastin and collagen precursors, which in turn spontaneously cross-link collagen and elastin, thereby forming a mature and functional ECM

Stromelysin

Is MMP3. Degrades most constituents of the ECM. Activates other MMPs

Latent morphogens and cytokines Bone morphogenetic protein family

Osteoinductive effects, promotes osteogenesis, chondrogenesis, and some induce other tissue types

Tissue growth factor  family

Mitogen for periosteal osteoprogenitors and marrow stromal cells. Direct stimulatory effect on bone collagen synthesis. Decreases bone resorption by inducing apoptosis of osteoclasts. Inhibits osteoblast differentiation in vitro

Fibroblast growth factor

Angiogenic properties. Important with neovascularization and wound healing. Important in healing and bone repair. Promotes bone cell replication

B. Inorganic Phase

C. Organization of Bone

The inorganic component of bone is composed mainly of a calcium phosphate mineral analogous to crystalline calcium hydroxyapatite. This apatite is present as a platelike crystal, which is 20 to 80 nm long and 2 to 5 nm thick. The small amounts of impurities in hydroxyapatite, such as carbonate, which can replace the phosphate groups, or chloride and fluoride, which can replace the hydroxyl groups, may alter certain physical properties of the crystal, such as its solubility [6]. These altered properties may impart important biologic effects that are critical to normal function.

The skeleton is composed of two parts: an axial skeleton, which includes the vertebrae, pelvis, and other flat bones such as the skull and sternum, and an appendicular skeleton, which includes all of the long bones. The long bones are divided into three parts: the epiphysis, metaphysis, and diaphysis. The epiphysis is that portion of the long bone found at either end and develops from a center of ossification that is distinct from the rest of the long bone shaft. It is separated from the rest of the bone by a layer of growth cartilage. This growth cartilage is known as the physis. The metaphysis is the zone between the physis and the

6

LEE AND EINHORN

FIGURE 2

Woven bone. Note arrangement of this bone tissue in which there is active bone formation (top), no particular lamellar organization of the tissue, and a high degree of cellularity.

central portion of the long bone shaft (known as the diaphysis). The metaphysis is the region where remodeling of the bone takes place during growth and development. The diaphysis comprises the majority of the length of a long bone. At the microscopic level, bone appears as an extremely well-organized tissue whereby mineral impregnates the collagen fibril array in such a way as to provide mechanical properties, which, in tension, are nearly as strong as cast iron. There are two types of bone tissue: woven bone and lamellar bone. Woven bone is considered immature or primitive bone and is normally found in the embryo, the newborn, in fracture callus, and in certain metaphyseal regions of the growing skeleton. It is also found in certain bone tumors, in patients with osteogenesis imperfecta, and in patients with Paget disease. Lamellar bone, however, is a more mature bone that results from the remodeling of woven bone or preexisting bone tissue. Woven, or primary bone, is a coarse-fibered tissue that does not show any uniform orientation of the collagen fibers (Fig. 2). It has more cells per unit volume than lamellar bone, and its mineral content and cells are randomly arranged. Newly formed woven bone, which is not as well mineralized as mature lamellar bone, contains particles with a smaller average crystal size. Moreover, the relative disorientation of the collagen fibers gives it isotropic mechanical characteristics; i.e., when tested, the mechanical behavior of woven bone is similar regardless of the orientation of the applied forces. At the time of birth, all bone in the body is woven. However, beginning at approximately 1 month of age, lamellar bone begins to develop. By 1 year

of age, lamellar bone will have effectively replaced much of the woven bone, as the latter is resorbed. By the age of 4, most of the bone in the body will be lamellar. Lamellar bone is a highly organized material with respect to the stress orientation of the collagen fibers (Fig. 3). As a result of this structural organization, lamellar bone exhibits anisotropic properties; i.e., the mechanical behavior of lamellar bone differs depending on the orientation of the applied forces. Moreover, its ability to resist loads is greatest when the forces are directed in a parallel fashion to the longitudinal axis of the collagen fibers. Anatomically, woven and lamellar bone are organized into trabecular (spongy or cancellous) and cortical (dense or compact) bone compartments. Cortical bone has four times the mass of trabecular bone, although the metabolic turnover rate of trabecular bone is much higher than that of cortical bone (bone turnover is a surface event, and trabecular bone has a greater surface area than cortical bone). Trabecular bone is found principally at the ends of long bones and in cuboid bones, such as the vertebrae (Fig. 4). The internal beams or plates of trabecular bone form a three-dimensional branching lattice, which is oriented along lines of stress. Trabecular bone is subject to a complex set of stresses and strains, although it is best designed for resisting of compressive loads. Cortical bone appears as an entirely different structure. It is solid and arranged not as interconnecting plates, but rather as cylinders, which are ellipsoid in cross section (Fig. 5). Cortical bone is usually subject to bending and torsional forces, as well as compressive loads.

7

CHAPTER 1 The Bone Organ System

FIGURE 3

Lamellar bone. Note organization of this tissue in which there is a well-delineated orientation of the collagen fibers and a coordinated alignment of the cells.

Haversian bone is the most complex type of cortical bone. It is composed of vascular channels surrounded circumferentially by lamellae of bone (Fig. 6). This complex arrangement of bone around the vascular channel is called

FIGURE 4

the osteon. The osteon is an irregular, branching, and anastomosing cylinder composed of a neurovascular canal surrounded by cell-permeated layers of bone matrix. Osteons are usually oriented in the long axis of the bone and are the

Trabecular bone. Note the perpendicular orientation between the horizontal and the vertical trabeculae.

8

LEE AND EINHORN

FIGURE 5 Cross-section of cortical bone from the middiaphysis of the tibia. Note that cortical bone is a solid structure arranged not as interconnecting plates, but as ellipsoid cylinders. On the inner surface (endosteum), there is a structure that resembles that of trabecular bone.

FIGURE 6

Close-up view of a section of cortical bone. Note the distribution of vascular channels forming the osteons.

9

CHAPTER 1 The Bone Organ System

FIGURE 7 Scanning electron micrograph of cortical bone showing individual osteons, surrounded by lamellar bone, which is impregnated with well-aligned cellular lacunae.

major structural units of cortical bone. They are connected to one another by Volkmanns canals, which are oriented perpendicularly to the osteon. Thus, cortical bone is a complex structure composed of many adjacent osteons and their interstitial and circumferential lamellae (Fig. 7). Most vessels in haversian canals have the ultrastructural features of capillaries, although some smaller-sized vessels may resemble lymphatic vessels. When examined histologically, these small vessels contain only precipitated protein; their endothelial walls are not surrounded by a basement membrane. The basement membrane of capillary walls may function as a rate-limiting or selective ion-limiting transport barrier because all material traversing the vessel wall must go through the basement membrane. The presence of this barrier is particularly important in calcium and phosphorous ion transport to and from bone. It is also important in explaining the response of bone to mechanical loads. The capillaries in the central canals are derived from the principal nutrient arteries of the bone: the epiphyseal and metaphyseal arteries [6]. The periosteum lines the outer surface of bone. It is composed of two layers. The outer fibrous layer is in direct contact with muscle and other soft tissue elements and is populated by undifferentiated fibroblast-like cells. The inner layer is known as the cambium layer and it is populated by fibroblast-appearing cells, many of which are committed

progenitors of chondrocytes and osteoblasts (Fig. 8).This layer contributes to appositional bone growth during bone development and is responsible for the expansion of the diameters of the long bones with aging.

III. CELLULAR CONTROL OF BONE HOMEOSTASIS A. Bone Cells Bone metabolism is regulated by bone cells, which respond to various environmental signals, including chemical, mechanical, electrical, and magnetic stimuli. In general, specific responses are governed by cellular receptors found on the membrane of the cell or intracellularly. Cell membrane receptors bind the exogenous signal and transfer the information across the cell’s cytoplasm to the nucleus through a series of interactions that involve a complex set of transduction mechanisms. Intracellular receptors (cytoplasmic or nuclear) bind the stimulus (usually a steroid hormone, which has crossed the cell membrane and entered the cell) and then translocate that effector to the nucleus where the steroid – receptor complex binds to a specific DNA promotor sequence of a gene. Three cell types are found in bone, osteoblasts, osteoclasts, and osteocytes.

10

LEE AND EINHORN

FIGURE 8

Close-up view of the periosteum of a long bone. The darker staining material at the lower portion of the figure is mineralized cortical bone. Above this is the periosteum, which consists of two layers. The outer layer shows elongated fibroblast-like cells embedded in a fibrous-like tissue. The inner layer, known as the cambium layer, is composed of a higher density of cells, which are slightly more plump and are embedded in a loose connective tissue.

Osteoblasts, generally regarded as bone-forming cells, govern bone metabolism (see Chapter 2). Their most obvious function is to synthesize osteoid, the protein component of bone tissue, but they also initiate bone resorption by

FIGURE 9

elaborating various neutral proteases (Fig. 9) [8]. The proteases remove surface osteoid, after which other cells participate in bone resorption. Because osteoblasts contain the receptors for most chemical mediators of bone metabolism,

Low-power view of osteoblasts lining the bone surface.

11

CHAPTER 1 The Bone Organ System

FIGURE 10

including bone resorption, they play a critical role in the regulation of bone turnover. Osteocytes are abundant in mineralized bone matrix, but their function is poorly understood. Evidence suggests that they may receive mechanical input signals and transmit these stimuli to other cells in bone (Fig. 10). Osteoblasts and osteocytes are both derived from the same mesenchymal stem cell precursor (found in bone marrow stroma, periosteum, soft tissues, and possibly peripheral blood vessel endothelium). Once an osteoblast has synthesized osteoid and mineralized it, the osteoblast becomes an osteocyte. Evidence suggests that the lineage of the cell type can be identified by the expression of specific cell surface antigens, the expression of alkaline phosphatase when the cell is in its secretory osteoblastic phase, and the loss of alkaline phosphatase activity when the cell evolves to an osteocyte [8]. Osteoclasts, the active agents in bone resorption, are ultimately responsible for the remodeling of bone (see Chapter 3). In cortical bone, they are found at the apex of the classical “cutting cone’’ (Fig. 11, see also color plate). In trabecular bone they create the resorptive cavities, known as Howship’s lacunae (Fig. 12), seen on bone surfaces undergoing active remodeling. Osteoclasts are multinucleated, but their progenitors are hemopoietic mononuclear cells.

Differentiation toward an osteoclastic phenotype occurs early in the development of these cells.

B. Cellular Mechanisms in Bone All bone surfaces are continuous and lined by resting osteoblastic lining cells. On close inspection, it is possible to observe small intercellular gaps between the cells and their cytoplasmic processes. The lining osteoblasts are in communication with osteocytes through cell processes within the canaliculi that form gap junctions. (Fig. 13) Rapid fluxes of bone calcium across these junctions may be involved in the transmission of information between osteoblasts on the bone surface and osteocytes within the structure of bone itself. Although the cellular layer protects the bone from the extracellular fluid space, the osteoblasts on the bone surface are in direct chemical contact with the osteocytes within the mineralized bone by virtue of these cellular processes within the canaliculi. This organizational structure is consistent with the concept that bone cells are in intimate communication with each other and that osteoblasts receive the majority of local and systemic signals and then transmit them to other cells in bone. Conversely, strain-generated signals such as streaming potentials could

12

LEE AND EINHORN

FIGURE 11

A cutting cone in cortical bone. Note that it is composed of an osteoclast, which appears to be tunneling through bone, and this cell is followed closely by a group of endothelial cells that ultimately transition to osteoblasts. (See also color plate.)

be perceived by osteocytes, and their regulatory information can be passed on to the osteoblasts. Osteoclasts at specific bone sites are activated only after disruption of the osteoid layer that covers the bone sur-

FIGURE 12

faces, a bone lining osteoblast-mediated effect. This exposure of the underlying mineralized matrix may be caused by the degradation of surface osteoid by the neutral proteases elaborated by flat, elongated osteoblasts or by the

Low-power view of osteoclasts forming Howship’s lacunae.

13

CHAPTER 1 The Bone Organ System

This molecule, related to the TNF family, upregulates osteoclast development and increases the activity of mature osteoclasts [9]. It may also play a role in immunologic tissue by regulating the interaction between T-cells and dendritic cells in vitro [10]. In vivo, RANKL exists in the cell as a ligand, anchored to the surface of an osteoblast or a bone marrow stromal cell and interacts with RANK on a preosteoclast. RANKL also exists in a secreted form that binds directly to the preosteoclast [11]. It is here that it promotes osteoclast differentiation from hemopoietic precursors. Once this reaction occurs, the preosteoclast will differentiate to a mature osteoclast only if the macrophage colony-stimulating factor (M-CSF) is present. Various other factors, such as PTH, interleukin-11, prostaglandin E2 (PGE2), and 1,25(OH)2D, upregulate RANKL on the surface of osteoblasts, hence increasing the development and activity of osteoclasts. The interaction between RANKL and its preosteoclast receptor is controlled by osteoprotegerin (OPG), a secreted protein. OPG is a soluble decoy protein that may modulate communication between osteoblasts and osteoclasts, thus playing a major role in bone homeostasis. Its main function is to halt bone resorption by inhibiting osteoclast formation via interruption of RANKL. In vitro, studies have shown that OPG can induce apoptosis of osteoclast-like cells (see Chapter 3). The proposed mechanism for OPG function is via reduction/disruption of formation of the F-actin ring in isolated osteoclasts. The F-actin ring is a cytoskeletal structure correlated with bone resorption [12].

IV. BONE MODELING AND REMODELING A. Bone Modeling FIGURE 13

Osteocyte (OC) cytoplasmic process in contact with the secreting surface of an osteoblast (OB). TEM  16,125. Reprinted with permission from C. Palumbo, Morphological study of intercellular junctions during osteocyte differentiation, In “Bone” (R. Baron, ed.), pp. 401 – 409. Pergamon Press, Elmsford, NY.

contraction of osteoblasts in response to stimulation by parathyroid hormone (PTH), 1,25-dihydroxyvitamin D, or prostaglandins of the E series. This contraction allows osteoclasts to gain access to the mineralized bone. Three recently discovered molecules that possibly play a key role in the coordinated maintenance of bone homeostasis are osteoprotegerin (OPG), RANK, (receptor activator of NF-B), and its ligand, RANKL, (see Chapter 3). The ligand (RANKL) has proven to be an essential factor in osteoclast differentiation by promoting osteoclast formation.

Bone formation begins in utero and continues throughout adolescence until skeletal maturity (see Chapter 5). Long bones form by two mechanisms. Endochondral ossification occurs in the long bones or the appendicular skeleton. It involves the differentiation of mesenchymal lineage cells to chondroblasts and then chondrocytes with the synthesis of a proteoglycan-rich, type II collagen-based extracellular matrix. This matrix is then modified biochemically by enzymes elaborated by hypertrophic chondrocytes to produce an environment that will permit the deposition of calcium. Once the extracellular matrix is calcified, it becomes a target for blood vessel invasion, and with this angiogenic response comes osteoclasts (which degrade the calcified cartilage) and osteoblast precursors. The calcified cartilage first formed is known as the primary spongiosum, and the bone that is laid down upon this tissue is known as secondary spongiosum. This secondary spongiosum is in the form of woven bone.

14

LEE AND EINHORN

Intramembranous bone formation occurs in the flat bones of the skeleton such as the skull and pelvis. It involves the direct formation of bone tissue by cells of the mesenchymal lineage, which have already undergone a biological commitment to the formation of osteoblastic cells. Bones also grow in length by endochondral ossification and in width by intramembranous or subperiosteal new bone formation.

B. Bone Remodeling Following skeletal maturity, bone continues to remodel throughout life and adapt its material properties to the mechanical demands placed upon it (see Chapters 12 and 15). Remodeling also has the function of maintaining the biomechanical competence of the skeleton by preventing the accumulation of fatigue damage and maintaining a tissue whose components are available for mineral homeostasis. Remodeling is thus a continuous process whereby there is a constant removal and replacement of whole volumes of bone tissue, a process conducted by osteoclasts and osteoblasts on bone surfaces, which, together with their precursor cells, form the bone remodeling system. A knowledge of bone remodeling is essential to the understanding of the pathophysiology of osteoporosis. The bone remodeling process governs the way bone is replaced, gained, or lost at specific sites and the way cumulative effects determine the three-dimensional structure of bone. The rate of turnover also determines the age of the bone tissue, and various physical and chemical properties of the bone are dependent on age and function [13]. According to bone histologists (see Chapter 15), the skeleton is composed of individual structural units or bone metabolic units (BMU) [6]. Cortical bone constitutes approximately 80% of the skeletal mass and trabecular bone approximately 20%. Bone surfaces may be undergoing formation or resorption, or they may be inactive. These processes occur throughout life in both cortical and trabecular bone. Bone remodeling is a surface phenomenon and it occurs on periosteal, endosteal, haversian canal, and trabecular surfaces. The rate of cortical bone remodeling, which may be as high as 50% per year in the midshaft of the femur during the first 2 years of life, eventually declines to a rate of 2 to 5% per year in the elderly. Rates of remodeling in trabecular bone are proportionally higher throughout life and may normally be 5 to 10 times higher than cortical bone remodeling rates in the adult [13]. The BMU of cortical bone is the osteon or haversian system, a cylinder of about 200 to 250 m in diameter running parallel to the long axis of the bone. As described earlier, the canals are connected to each other by transverse Volkmann’s canals and periodically either divide or reunite to form a branching network. Osteons form approximately two-thirds of cortical bone volume, a

proportion that falls with age, with the remainder consisting of interstitial bone representing the previous generations of osteons. There are also subperiosteal and subendosteal circumferential lamellae. In trabecular bone, the BMU is constructed differently. In two-dimensional sections, the trabecular surfaces are shaped like thin crescents about 600 m long and about 60 m in depth. Three-dimensionally, these BMUs are actually larger than they appear in two-dimensional histological sections with prolongations in different directions that interlock with adjacent BMUs [14]. These BMUs follow the same shape as the trabecular surface, most of which are concave toward the marrow. Under normal conditions, the remodeling process of resorption followed by formation is closely coupled and results in no net change in bone mass. As such, the BMU consists of a group of cells that participate in remodeling in a concerted and coordinated fashion. Cortical bone remodeling proceeds via cutting cones (Fig. 11) and is similar to processes in other hard biological tissues. Cuttings cones, or sheets of osteoclasts, bore holes through the hard bone, leaving tunnels, which appear in cross section as cavities. The head of the cutting cone consists of osteoclasts that resorb the bone. Following closely behind the osteoclast is a capillary loop and a population of endothelial cells and perivascular mesenchymal cells that are progenitors for osteoblasts and soon begin to lay down osteoid and refill the resorption cavity. By the end of the process, a new osteon will have been formed. Trabecular bone remodeling proceeds on the surface of bone whereby osteoclasts resorb bone at specific sites (Fig. 14). These areas are then filled in with newly formed osteoid. The mechanisms that control the activity and site specificity of this process are unknown. Cortical bone remodeling occurs in discrete temporal foci that are active for about 4 to 8 months [13]. Mononuclear precursor cells proliferate into a team of new osteoclasts that initiate the cutting cone. The osteoclasts then are removed (or disappear), and there is a quiescent interval or “reversal phase’’ during which time the newly resorbed area of bone is smoothed and a layer of cement substance (osteoid) is deposited. The team of osteoblasts that follow the osteoclast attempts to replace exactly as much bone as has been removed. If it is successful in this venture, bone homeostasis will have been maintained. If not, osteoporosis may result. One of the conditions that affects the ability of the bone-resorbing and bone-forming cells to be coupled and result in no net change in bone mass is aging. During the process of remodeling, changes in the amount of bone are slow and changes in the shape of bone are barely perceptible. According to the model proposed by Parfitt, the normal remodeling sequence in bone follows a scheme of quiescence, activation, resorption, reversal, formation, and return

15

CHAPTER 1 The Bone Organ System

FIGURE 14

High-power scanning electron micrograph of an intersection of bone trabeculae showing multiple resorption sites. (Courtesy of S. Goldstein, University of Michigan.)

to quiescence (see Chapter 15). In the adult, approximately 80% of trabecular and approximately 95% of intracortical bone surfaces are inactive with respect to bone remodeling [15,16]. The surface of bone is covered by a layer of thin flattened lining cells approximately 15 m in diameter, which arise by terminal transformation of osteoblasts. Between these lining cells and bone is a layer of unmineralized osteoid. These lining cells have receptors for a variety of substances, which are important for initiating bone resorption (PTH, PGE2), and may respond to such substances by resorbing this surface osteoid, which is covering the bone. In doing so, mineralized bone will be exposed and the activation sequence of bone remodeling may be initiated. The conversion of a small area of bone surface from quiescence to activity is referred to as activation. The cycle of this response begins with the recruitment of osteoclasts followed by the initiation of mechanisms for their attraction (chemotaxis) and attachment to the bone surfaces. Several known growth factors may be active in promoting chemotaxis. In addition, several proteins are known to be attachment factors for osteoclasts, such as those that contain the RGD amino acid sequences as noted earlier. Osteopontin, osteocalcin, and osteonectin

may be important proteins in this process (see Chapters 3 and 4). In the adult skeleton, activation occurs about every 10 s. For intracortical remodeling, osteoclast precursors travel to the site of activation via the circulation, gaining access to the site by either a Volkmann or a haversian canal. In trabecular remodeling, activation occurs at sites that are apposed to bone marrow cells. The osteoclast is a very mobile cell that can resorb bone over an area approximately two to three times the area with which it is in direct contact. In cortical bone, the osteoclast and the cutting cone travel at a speed of about 20 or 40 m per day, roughly parallel to the long axis of the bone and about 5 to 10 m per day perpendicular to the main direction of advance [7]. In trabecular bone, osteoblasts erode to a depth of about two-thirds of the final cavity; the remainder of the cavity is eroded more slowly by mononuclear cells [17]. The reversal phase is a time interval between the completion of resorption and the initiation of bone formation at a particular skeletal site. Under normal conditions, it lasts about 1 to 2 weeks. The appearance of new osteoblasts at the base of the resorption cavity depends on chemotaxis for these osteoblasts and their progenitors, as

16

LEE AND EINHORN

well as conditions that stimulate proliferation. Hence chemotaxis, attachment, proliferation, and differentiation occur in a stepwise and concerted fashion in order for bone formation ultimately to take place. Electron microscope studies of the reversal phase have shown the release of osteocytes and the appearance of mononuclear phagocytes, some of which appear to be smoothing over the ragged surface left by the resorption process [18]. Because the coupling mechanism appears to be important to this bone remodeling process, the release of proteins from the bone matrix may be important in coordinating the activity between osteoclasts and osteoblasts. Unlike bone resorption, bone formation is a two-step process. First, the osteoid is synthesized and laid down at specific sites. Following this, the osteoblast must mineralize this newly formed protein matrix. The new matrix begins to mineralize after about 5 to 10 days from the time of deposition and, as a result, matrix apposition and mineralization are systematically out of step as the osteoid seam first increases and then decreases in thickness. The rate of mineral apposition can be measured directly in vivo after double tetracycline labeling; the mean distance between fluorescent bands divided by the time interval between the midpoints of the tetracycline labels can be used to calculate the mineral apposition rate [15].

V. BONE BIOMECHANICS A. Basic Concepts A basic knowledge of biomechanics must begin with an understanding of the terms stress and strain. These terms are used to describe the phenomenon whereby, when a force is applied to bone, the bone will not only be deformed from its original shape, but an internal resistance will be generated to counter the applied force. This internal resistance is known as stress and it is equal in magnitude but opposite in direction to the applied force. Because stress is distributed over the entire cross-sectional area of a section of bone, it is expressed as units of force per unit area. Strain is a term used to describe the changes in shape that bone experiences when it is subjected to an applied force. Strain is dimensionless and is therefore reported as a fraction or a percentage. It is equivalent to the change in length divided by the original length of the section of bone. Although an applied force can be directed at bone from any angle producing any set of complex stress patterns, all stresses can be resolved into three types: tension, compression, and shear Tension is produced in bone when two forces are directed away from each other along the same straight line. The resistance to a loading situation of this type is produced by intermolecular attractive forces, which

prevent the bone from being pulled apart. An example of a tensile force producing a failure in bone occurs when a tendon or ligament that is inserted into bone undergoes acute loading and, instead of failing within its own substance (tearing), it detaches itself from the bone by actually pulling a piece of bone off with it. Compression results from two forces that are directed toward each other along the same straight line. The common vertebral compression fracture sustained in osteoporotic patients is an example of the failure of bone as a result of this type of loading configuration (Fig. 15). Finally, when two forces are directed parallel to each other but not along the same line, shear stresses are produced. In nature, the three basic stress types can combine as a result of a variety of complex loading configurations and lead to different fracture patterns (Fig. 16). Bending results from a combination of tensile and compressive forces and leads to clinical fractures that show a predominantly transverse fracture pattern. Torsion or twisting produces shear stresses along the entire length of a bone and can result in spiral fractures. Comminuted fractures appear as shattered bones and this occurs because the amount of energy transmitted to the bone is so great that a variety of fracture lines have to be propagated in order for it to be dissipated. Because stress and strain are properties related to the quality of the tissue experiencing the load, the quality of bone can influence the magnitude of the stresses and strains generated. In normal, well-mineralized bone, physiological stresses will generally result in small strains. In poorly mineralized tissue, such as osteomalacic bone, bone will experience larger strains in response to the same stress. Moreover, because in nature, forces are applied to bone not only from perpendicular and horizontal directions but also from oblique angles, conditions will arise in which a variety of complex mechanical relationships will be generated. When bones show different mechanical properties if loads are transmitted from different directions, they express a property known as anisotropy [22]. In general, bone resists loads best when the loads are oriented in the direction of customary loading. For example, the femur is much better adapted to resisting compressive loads than bending loads [20]. Thus, the same stresses generated in the femur if one were to jump from a 4foot wall and land on his feet (compressive stresses) may fracture the femur if they were oriented from a transverse direction(bending stresses). In the proximal femur (hip joint area), loads are best resisted when they are transmitted along lines that are parallel to the trabecular systems [21].

B. Biomechanical Properties The biomechanical properties of bone can be described at two levels. First, the material properties of bone are

17

CHAPTER 1 The Bone Organ System

FIGURE 15

Sagittal section of a vertebral compression fracture. Note collapse of the vertebra through the central trabecular

section.

FIGURE 16

Fracture patterns in a cylindrical section of bone subjected to different complex loading configurations. (a) Pure tensile loading produces a purely transverse fracture. (b) Pure compressive loading produces an oblique fracture. (c) Torsional loading produces a spiral fracture. (d) Bending, a combination of compressive and tensile loading, produces an essentially transverse fracture with a small fragment on the concave side. (e) Bending in which compressive forces make a greater contribution to the complex loading configuration, leading to a transverse fracture with a larger fragment on the concave side.

defined by the tissue level qualities of the tissue, which are independent of its structure or geometry. Second, there are structural properties in bone, which are manifest when bone functions as a whole anatomical unit. Classically, the material properties of bone are defined by performing standardized mechanical tests on uniform, machined specimens of bone. Structural properties are determined on whole sections of bone whose normal geometry has been maintained. It is important to recognize that when a patient sustains a fracture, that event most likely represents the failure of bone at either the material or the structural level or both. When a uniform section of bone is tested under controlled laboratory conditions, and the applied forces and deformations are known, four basic mechanical properties of bone can be derived from a plot of the stress – strain relationship. These properties are stress, modulus, energy absorptive capacity, and deformation. By convention, stress is plotted on the ordinate (y axis) and strain on the abscissa (x axis). Figure 17 shows a stress – strain plot of an idealized material. Considering that combinations of the three types of stress can produce different stress patterns as a result of different types of externally applied loads (tension, compression, bending, and torsion), the terms used to define the parameters of the y axis can include any of these loading conditions. Under these circumstances, the stress – strain

18

FIGURE 17 A standard stress/strain curve of bone loaded in bending. The linear portion of the curve represents the elastic region and the slope of this part of the curve is used to derive the stiffness of the bone. Loading in this region will result in nonpermanent deformation, and the energy returned to the bone when the load is removed is known as resilience. The nonlinear portion of the curve represents the plastic region in which the bone will be permanently deformed by the load. The junction of these two regions defines the yield point and the stress here is known as the elastic limit. The maximum stress at the point of failure is known as the ultimate strength of the bone. The maximum strain at this point represents the bone’s ductility. The area under the curve is known as the strain energy, and the total energy stored at the point of fracture defines the toughness of the material. Reprinted with permission from T. A. Einhorn, Calcif. Tissue Int. 51, 333 – 339 (1992).

curve is actually a load versus deformation relationship whereby the y axis could be labeled as torque, compressive load, tensile force, or shear. At low levels of stress there is a linear relationship between the applied load and the resultant deformation. This proportionality is known as the modulus of elasticity or Young’s modulus. It is a measure of the rigidity of the bone tissue and is equivalent to the slope of the linear part of the curve. It is calculated by dividing the stress by the strain at any point along this straight line. This linear part of the curve is also known as the elastic region. The physiological significance of this property relates to the fact that forces applied to bone at any point along this line will only deform the bone temporarily. After the load is removed, it will return to its original shape. At the point where the curve becomes nonlinear, the elastic region ends and the stress at this point is known as the elastic limit. Further loading beyond this point will result in a permanent deformation in the material and this property is known as plasticity. This part of the curve is known as the plastic region. Not all materials (and, for that matter, not all bones) have significant plastic properties. Instead, a material may exhibit elastic deformation but, upon reaching its yield point, will fail. This type of material is considered to be brittle.

LEE AND EINHORN

The maximum height of the curve defines the maximum stress in the bone tissue and is the point at which the material fails. The strain at the point of failure is known as the ductility. The area under the curve is a measure of the energy absorptive capacity or strain energy. This energy is dissipated when the bone fractures and is lost at the point of failure. Note that energy stored by the bone up to the point where it reaches its elastic limit is known as resilience. This energy is recovered if the bone returns to its original shape after the load is removed. When whole bones are subjected to experimental or physiological loading conditions, their mechanical behavior is dependent not only on the mass of the tissue and its material properties, but also on its geometry and architecture. When whole bones are loaded to failure, they, too, produce curves that are configured like the curve shown in Fig. 21. However, under these conditions, the slope of the linear portion defines the stiffness of bone. Strength is determined by extrapolating the height of the curve to the y axis. Depending on the loading conditions, strength can be expressed in terms of compressive, bending, tensile, or torsional strength. The area under the curve now defines the toughness of the bone. Of particular interest in the study of metabolic bone diseases are fractures of the vertebral bodies. Here, the vertebrae fracture as a result of predominantly axially directed compressive loads. The reduced load-bearing capacity of each vertebra is related to the material properties of the bone as well as the way in which the vertebral trabeculation is altered through the processes of postmenopausal and agerelated bone loss. Studies have shown that it is the removal of the horizontal trabeculae or lateral support cross-ties that alter the architectural arrangement and lead to reduced load-bearing capacity [22]. As a result, the vertical trabeculae begin to behave as columns and, as such, are subjected to critical buckling loads [23]. A 50% reduction in the cross-sectional area contributed by these horizontal trabeculae will be associated with a 75% reduction in the loadbearing capacity of the vertebral body [24]. Most fractures occurring in nature result from a combination of axial compression, bending, and torsion. In bending or torsion, the cross-sectional area of a structure is more important in resisting loads than its mass or density [25]. Ideally, in bending or torsion, bone should be distributed as far away from the neutral axis of the load as possible. The geometric parameter used to describe this phenomenon in bending is the “areal moment of inertia’’ [25]. Similarly, in torsion, deformation would be resisted more efficiently if bone were distributed further away from the neutral axis of torsion. This property is known as the “polar moment of inertia’’ [25]. During aging, the outer cortical diameter of bone increases and the cortical wall diameter becomes thinner [26,27]. This results from the combined effects of increased endosteal resorption and periosteal bone formation. Although the net effect may be cortical thinning, the

CHAPTER 1 The Bone Organ System

increased diameter of the bone distributes the material further from the neutral axis and improves its resistance to bending and torsional loads.

VI. SUMMARY Bone is a mechanically optimized organ system whose composition and organization reflect the functional demands made upon it. Far from being an inert substance, it is also a living tissue that serves several important functions in the organism. As a biological entity, bone tissue is a composite material composed of a proteinaceous extracellular matrix or ground substance that has been impregnated by an inorganic calcium phosphate mineral phase. In this sense, it can be likened to a material such as fiberglass with flexibilities, rigidities, and other mechanical properties related to the composite nature of its components. However, what distinguishes a material like bone from other composite tissues and materials is the fact that it is constantly being broken down and rebuilt in the process known as remodeling. The cellular link between bone resorbing cells, osteoclasts, and bone-forming cells, osteoblasts, may be regulated by the release of small molecules from the extracellular matrix during bone resorption. The complexities of the remodeling process require further investigation, but current knowledge suggests that it is composed of several phases, including quiescence, activation, resorption, reversal, and formation. Although the transduction of mechanical signals through bone and the stimulation of bone cells by hormonal agents have been studied extensively, more information is needed before a truly comprehensive understanding of the bone organ system can be developed. Ultimately, it would be important to know how tissue development, cellular function, mechanical input, and the material properties of the tissue are organized in time, space, and function in order to maintain bone homeostasis. The information and concepts described in this chapter form the basis for understanding bone as a tissue as well as an organ system. The ability to maintain homeostasis, prevent bone resorption, and, perhaps more importantly, enhance bone formation will be essential to the management of osteoporosis now and in the future.

References 1. I. A. Bab and T. A. Einhorn, Polypeptide factors regulating osteogenesis and bone marrow repair. J. Cell Biochem. 55, 358 – 365 (1994). 2. M. Horowitz and R. L. Jilka, Colony stimulating factors in bone remodeling, In “Cytokines and Bone Metabolism” (M. Gowen, ed.), pp. 185 – 227. CRC Press, Boca Raton, FL, 1992. 3. J. Wolff, In “Das Gesetz der Transformation der Knochen” (A. Hirchwald, ed.). Berlin, 1892. 4. T. A. Einhorn, Bone metabolism and metabolic bone disease. In “Orthopaedic Knowledge Update 4 Home Study Syllabus” (J. W. Frymoyer, ed.), pp. 69 – 88. Am. Acad. Orthop. Surg., Rosemont, 1994.

19 5. E. Ruoslahti, Integrins. J. Clin. Invest. 87, 1 – 5 (1991). 6. F. S. Kaplan, W. C. Hayes, T. M. Keaveny, A. L. Boskey, T. A. Einhorn, and J. P. Iannotti, Form and function of bone, In “Orthopaedic Basic Science” (S. R. Simon, ed.), pp. 127 – 184. Am. Acad. Orthop. Surg., Rosemont, 1994. 7. T. A. Einhorn and R. J. Majeska, Neutral proteases in regenerating bone. Clin. Orthop. 262, 286 – 297 (1991). 8. S. P. Bruder and A. I. Caplan, Terminal differences of osteogenic cells in the embryonic chick tibia is revealed by a monoclonal antibody against osteocytes. Bone 11, 189 – 198 (1990). 9. T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentation. Crit Rev. Eukaryot. Gene Expr. 8, 107 – 23 (1998) 10. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morrony, A. J. Oliveirados-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature (England) 397, 315 – 323, (1999). 11. N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shimma, H. Yasuda, K. Yano, T. Morinaga, and K. Higasio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem Biophys Res. Commun. 253, 395 – 400 (1998). 12. Y. Hakeda, Y. Kobayashi, K. Yamaguchi, H. Yasuda, E. Tsuda, K. Higashio, T. Miyata, and M. Kumegawa, Osteoclastogenesis inhibitory factor (OCIF) directly inhibits bone-rresorbing activity of isolated mature osteoclasts. Biochem Biophys Res Commun. 251, 796 – 801 (1998). 13. A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures, In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), pp. 45 – 93. Raven Press, New York, 1988. 14. J. Kragstrup and F. Melsen, Three-dimensional morphology of trabecular bone osteons reconstructed from serial sections. Metab. Bone Dis. Relat. Res. 5, 127 – 130 (1983). 15. A. M. Parfitt, The psychological and clinical significance of bone histomorphometric data. In “Bone Histomorphometry. Techniques and Interpretations” (R. Recker, ed.) pp. 143 – 223. CRC Press, Boca Raton, FL, 1983 16. A. M. Parfitt, The cellular basis of bone remodeling: The quantum concept reexamined in light of recent advances in cell biology of bone. Calcif. Tissue Int. 36, S37 – S45 (1984). 17. E. F. Eriksen, F. Melsen, and L. Mosekilde, Reconstruction of the resorptive site in iliac trabecular bone: A kinetic model for bone resorption in 20 normal individuals. Metab. Bone Dis. Rel. Res. 5, 235 – 242 (1984). 18. A. M. Parfitt, A. R. Villaneuva, M. M. Crouch, C. H. E. Mathews, and H. Duncan, Classification of osteoid seams by combined use of cell morphology and tetracycline labeling: Evidence for intermittency of mineralization. In “Bone Histomorphometry: Second International Workshop” (P. J. Meunier, ed.), pp. 299 – 310. Armour Montagu, Paris, 1977. 19. L. J. Melton, E. Y. S. Chao, and J. M. Lane, Biomechanical aspects of fractures. In “Osteoporosis: Etiology, Diagnosis and Management” (B. L. Riggs and L. J. Melton, eds.), pp. 111 – 131. Raven Press, New York, 1988. 20. A. H. Burstein, D. T. Reilly, and M. J. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt. Surg. 58A, 82 – 86 (1976). 21. T. D. Brown and A. B. Ferguson, Jr., The development of a computational stress analysis of the femoral head. J. Bone Jt. Surg. 60A, 619 – 629 (1978). 22. Li, Mosekilde, A. Viidik, and L. E. Mosekilde, Correlation between the compressive strength of iliac and vertebral trabecular bone in normal individuals. Bone 6, 291 – 925 (1985).

20 23. P. R. Townsend, Buckling studies of single human trabeculae. J. Biomech. 8, 199 – 201 (1975). 24. H. Yamada, “Strength of Biological Materials” (F. G. Evans, ed.). Williams and Williams, Baltimore, MD, 1970. 25. T. A. Einhorn, Bone strength: The bottom line. Calcif. Tissue Int. 51, 333 – 339 (1992). 26. R. W. Smith and R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Henry Ford Hosp. Med. J. 28, 168 – 170 (1980).

LEE AND EINHORN 27. C. B. Ruff and W. C. Hayes, Superiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 948 (1982). 28. D. B. Phemister, The pathology of ununited fractures of the neck of the femur with special reference to the head. J. Bone Jt. Surg. 21, 681 – 693 (1939). 29. A. M. Pankovich, Primary internal fixation of femoral neck fractures. Arch. Surg. 110, 20 – 26 (1975).

CHAPTER 2

Osteoblast Biology JANE B. LIAN AND GARY S. STEIN Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655

I. Overview II. Embryonic Origins and Signaling Cascades for Osteogenesis III. In Vivo Tissue Level Organization of Osteoblasts: Phenotypic Features and Function IV. Cellular Cross-Talk of Osteoblast Lineage Cells: Functions in Calcium Homeostasis, Bone Turnover, and Hematopoiesis

V. Osteoblasts in Vitro: Stages in Development of the Osteoblast Phenotype VI. Molecular Mechanisms Mediating Progression of Osteoblast Differentiation VII. Concluding Remarks References

I. OVERVIEW

of genes associated with the biosynthesis, organization, and mineralization of the bone extracellular matrix. This chapter provides a perspective of the signaling pathways and molecular mechanisms mediating osteoblast growth, differentiation, and activity that serve as a basis for understanding the factors regulating bone development and growth, the continued remodeling of bone, and the regeneration of injured tissue. Recent advances in the identification of obligatory factors for development of the skeleton that also contribute to osteoblast growth and differentiation in the adult skeleton will be presented. During the past decade, numerous model systems have been developed to study the cell biology of bone and gain insight into various cell types important for bone formation and function. The advantages and limitations of recently described in vivo and cellular models to address physiological and molecular mechanisms regulating osteoblast growth and differentiation will be discussed. Knowledge of unique properties and definition of the mechanisms that control progression through the osteoblast cell lineage will allow a rational intervention for abnormalities in skeletal development, fracture repair, pathologies of metabolic bone diseases, and implant stability. These are basic biological questions and

Bone formation takes place not only during embryonic development and growth but throughout life to support normal bone remodeling and fracture repair. The requirement for continuous renewal of bone, through the remodeling process involving resorption and formation, necessitates recruitment, proliferation, and differentiation of osteoblast-lineage cells. Subpopulations of osteoblasts are recognizable in vivo morphologically in relation to tissue organization and exhibit subtle phenotypic differences with respect to the expression of genes and responses to physiologic mediators of bone formation. This chapter presents current understanding of the phenotypic definition of the spectrum of bone-forming cells with respect to their functional properties and responses. It is now apparent that growth factor, cytokine, and hormone responsive regulatory signals mediate competency for the expression of genes associated with metabolic responses as a function of the stages of osteoblast growth and differentiation. Osteoblast differentiation is a multistep series of events modulated by an integrated cascade of gene expression that initially supports proliferation and the sequential expression

OSTEOPOROSIS, SECOND EDITION VOLUME 1

21

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

22

LIAN AND STEIN

concerns of today’s clinician for the treatment of bonerelated disorders.

II. EMBRYONIC ORIGINS AND SIGNALING CASCADES FOR OSTEOGENESIS The complexities of bone formation are immediately apparent in the embryo where different regions of the skeleton arise from specific primordial structures and skeletogenesis involves two different processes. Intramembranous bone formation, as occurs in the development of the flat bones of the skull, results from the differentiation of mesenchymal cell condensations directly to osteoblasts. The endochondral sequence of bone formation, as occurs for all long bones, involves the differentiation of mesenchymal progenitors first to form a cartilage template of the bone, which is then replaced by bone. Progenitors of the boneforming cells for all osseous tissues derive from the mesodermal germ cell layer. The dorsal paraxial mesoderm gives rise to somites and the sclerotome, which is a source of cells for most of the axial skeleton (vertebrae, skull, ribs, sternum), the lateral plate mesoderm gives rise to the appendicular skeleton (limbs), and the cephalic mesoderm gives rise to the neural crest, which provides progenitor cells for facial skeletal structures. Thus, different regions of

FIGURE 1

the skeleton have distinct embryonic lineages reflecting origins from these specific primordial structures. Considering these different embryonic developmental programs of the mesoderm to form intramembranous bone and subtypes of endochondral bone (e.g., limbs and vertebrae), an early osteoprogenitor may divert from a stem cell at these specific skeletal sites. It has been shown that axial and appendicular-derived osteoblasts exhibit different responses to hormones [e.g., 1,2]. It remains to be determined whether this selective activity reflects the tissue environment or inherent properties of the cells selected at an early stage during osteoblast differentiation. Our understanding of skeletal patterning and limb development has been expanded significantly by characterization of the signaling factors and transcription factors that serve as morphogenic determinants of bone formation [3 – 8]. Well-documented interactions between epithelium and mesenchyme are first necessary for tissue differentiation [9] and are reviewed elsewhere [10]. Signals essential for limb patterning arise from the zone of polarizing activity, which resides in the posterior limb mesoderm, underlying the epithelial apical ectodermal ridge (AER) of the developing limb bud. Gradients of morphogenic factors and integration of the various regulatory cascades reflect the complexity of skeletal development (Fig. 1). Growth factors and their receptors in the transforming growth factor- (TGF-) superfamily, epidermal growth factor (EGF), and the fibroblast

Signaling molecules regulating position and pattern formation and osteoblast differentiation during development of the skeleton. Encircled regulators, the FGF family, BMPs, Hedgehogs, and HOX genes, can be considered major signaling “centers” in that they induce (arrows) the expression of several different  classes of genes essential for skeletal development. Inhibitory controls are indicated ( ). Interrelationships are evident by feedback loops (e.g., FGFs and Hedgehogs), coordinate positive and negative regulation of common genes, and potential interaction of factors from one center with factors regulated by a different center. Relevant references include 21,22,27,28,47,67 – 70,75 – 77,698 – 705, as well as those described in the text.

23

CHAPTER 2 Osteoblast Biology

growth factor (FGF) family are implicated in mesenchymal condensations [11,12]. Basic fibroblast growth factor (FGF-2) and retinoic acid have been characterized as soluble mediators of communication between the AER and mesodermal progress zone [10,13 – 17]. FGF-2 activates several signaling pathways [18], including, for example, Wnt genes [19], Notch ligand expression [20,21], Hedgehog factors, and transcriptional regulators, such as helix – loop – helix proteins [22] (Fig. 1). Indian hedgehog (Ihh) is a key factor in the normal development of endochondral bone formation [23]. It is expressed abundantly in the developing growth plate in the mature and hypertrophic chondrocyte zones [24] is a key regulator of chondrocyte maturation. Ihh signaling is mediated through its receptor patch, and this signaling can be antagonized by a cell surface protein [25]. Several studies indicate that the rate of cartilage differentiation by Ihh is mediated by the parathyroid hormone-related protein (PTHrP) and its receptor [24,26 – 28]. PTHrP supports chondrocyte proliferation. Thus together Ihh and PTHrP regulate the proportions of proliferating and hypertrophic chondrocytes. Indeed, the PTHrP receptor will rescue the Ihh null mouse [27]. Knowledge of how mesenchymal condensations are initiated and grow and how their sizes and boundaries are regulated is being accrued through genetic studies in mice and the characterization of molecular defects in skeletal development [7] (Table 1). These studies have identified several specific players. Extracellular matrix molecules, cell surface receptors, and cell adhesion molecules, such as fibronectin, tenascin, syndecan, and N-CAM, initiate condensation and

TABLE 1

set boundaries for the forming mesenchyme, whereas Hox genes modulate the proliferation of cells within condensations [7]. Significant contributions to these earlier stages of skeletal development are provided by the TGF- superfamily of regulatory factors [reviewed in 29 – 32]. Bone morphogenetic proteins (BMPs) are actively involved in determining parameters of size and shape during mesenchymal cell condensation [33,34]. Selective expression of BMPs regulate mesenchymal condensations [35] and contribute to restriction of options for lineages. BMP-2, BMP-4, and BMP-7 (also designated OP-1) are potent inducers of osteogenesis in vivo and cell differentiation in vitro [36 – 41]. Specificity of the activities of TGF-s with target cells is regulated by their activation of distinct factors. TGF- and BMP bind to distinct receptors; each has associated kinase activity that phosphorylates Smad proteins. Smads 2 and 3 mediate TGF- responses, whereas Smads 1, 5, and 8 are activated by BMP receptors and transduce BMP signals. Interactions between receptor-activated Smads and Smad 4, a DNA-binding Smad, result in translocation of the complex to the nucleus for the transcription of target genes [29,42 – 44]. BMPs regulate several classes of genes that are key factors in normal bone development influencing spatial and temporal events (Fig. 1). These include fibroblast growth factors [45], homeobox containing genes [46 – 49], which contribute to position and pattern formation, cell adhesion proteins [50,51], and transcription factors such as helix – loop – helix [22,52], winged helix [53 – 55], SOX9, which is essential for chondrocyte differentiation [56,57], and RUNX2/core binding factor

Representative Mouse Models Involving Factors Related to Skeletal Development and Bone Formation

Homeobox proteins BAPX1 [648,649] (null)

Expressed in prechondrogenic cells, perinatal lethal skeletal dysplasia; malformations and absence of specific bones

Goosecoid [650] (null)

Craniofacial due to condensation defects

DLX-5 [651,652] (null)

Craniofacial and bone defects, die at birth

Msx-2 [653,654] (null)

Skull ossification defects

Msx-2 [655] (gain of function)

Craniosynostasis

TGF- superfamily TGF- receptor type II (transgenic dominant negative) a. Expressed in osteoblasts [656]

a. Decreased remodeling; increased trabecular bone

b. Expressed in skeleton [657]

b. Osteoarthritis

TGF-2 [658] (transgenic)

Development normal; osteoporosis postnatal

BMP-2 [659] (null)

Failure of mesoderm induction

BMP-4 [660,661] (null)

Die early in gastrulation

BMP-5 [34] (mutant)

Short-ear mouse gene (continues)

24

LIAN AND STEIN

TABLE 1 TGF- superfamily GDF-5 [662] (mutant)

(continued)

Brachypodism; bone and joint development problems

BMP-7 [663,664] (null)

Die after birth; polydactly; rib abnormalities

BMP receptor type I [665] (null)

Die 9.5 days post coital

BMP receptor type IB [666] (null)

Appendicular skeletal defects related to reduced chondrocyte proliferation and differentiation

Bone matrix proteins Osteopontin [667] (null)

Development is normal; increased osteoclastogenesis

Osteocalcin [277,278] (null)

Thickened bone increased mineral, but decreased crystal size

Osteonectin [668,669] (null)

Osteopenia; mesenchymal cell proliferation increased

Thrombospondin [670,671] (null)

Increased vascular density; increased marrow osteoprogenitor; increased cortical density

Biglycan [280] (null)

Thin bones, expression in preosteogenic cells

Decorin [672] (null)

Skin fragility

Alkaline phosphatase [673 – 675] (null)

Arrest of chondrocyte differentiation; defective matrix mineralization

Gelatinase [676] (null)

-3 Integrin [677] (null) BMP-5 [34] (mutant)

Transient aberrant growth plate vascularization and ossification Increased bone mass, but osteosclerotic; defective osteoclasts Short-ear mouse

Growth factors and receptors FGF-2 [261] (null)

Trabecular bone mass decreased; decreased mineralization of bone marrow stromal ex vivo cultures

FGF-2 [260] (transgene)

Increased apoptosis in calvaria

FGF-3 receptors [678,679] (null)

Skeletal overgrowth

(mutations) [680]

Achondroplasias dwarfism

FGF-1 receptor [681,682]

Embryonic growth and mesodermal patterning; axial organization and limb development

PTH/PTHrP receptor [80] (transgene)

Delays endochondral bone formation (EBF)

PTH/PTHrP receptor [28] (null)

Accelerates EBF; increase in osteoblast number and matrix; delay in vascular invasion; decrease in trabecular bone

PTHrP [74] (null)

Die at birth; premature chondrocyte differentiation and accelerated EBF

PTHrP x PTH/PTHrPR [78] (double knockout)

Accelerates differentiation of growth plate chondrocytes and EBFs; no delay in vascular invasion

PTHrP [683] (transgene)

In cartilage, delays chondrocyte maturation; mice are born with cartilaginous skeletons

IGF-1 [684] (transgene)

In bone, increases trabecular bone volume without proliferation

Other regulatory factors Indian Hedgehog [23] (null)

Severe dwarfism in the null mouse

Sonic Hedgehog [685]

A segment polarity gene; absence of distal limbs, most of the ribs and spinal column

Noggin [63] (null)

A BMP antagonist; cartilage hyperplasia

C-Abl [686] (null)

A nonreceptor tyrosine kinase; osteoporotic and osteoblast maturation defect

p27kip1 [253,687,688] (null)

A cyclin-dependent kinase inhibitor; larger size animals and bones; increased osteoprogenitors in bone marrow ex vivo cultures

TWRY [689] (mutant)

Nucleotide pyrophosphatase (NPPS) gene mutation; abnormal calcification of cartilage, spine, tendons, limbs

Transcriptional regulators RAR [690] (transgene)

Interferes with chondrogenesis, appendicular skeletal defect

SOX-9 (null)

An HMG domain TF expressed in cartilage lethal malformation syndrome XY reversal

c-Fos [691] (transgene)

Tumors in cartilage and bone

c-Fos [692] (null)

Osteopetrosis

Fra-1 [693] (transgene)

Progressive increase in bone mass; osteosclerotic

Fra-1 [693,694] (null)

Embryonic lethal from a placental defect

MFH-1 [695]

A forkhead or winged helix transcription factor; embryonic lethal and perinatal with skeletal defects

Plzf [696]

Promyelocytic leukemia zinc finger proteins; a growth inhibitory and proapoptotic factor essential for skeletal patterning

ATF-2 [697]

Chondrodysplasia

CBFA1 [58,59] (null)

A runt domain protein, absence of mineralized tissue essential for osteoblast differentiation

25

CHAPTER 2 Osteoblast Biology

A1 (CBFA1), which is essential for skeletal formation [58,59]. When growth of condensation ceases, the next stage of skeletal development, cellular differentiation, initiates. Here, Chordin [60], Noggin, and potentially Gremlin [61] play a critical role in skeletal development by inhibiting BMP2 signaling [62 – 65]. Feedback loops and complementary regulation of many of these factors (Fig. 1) ensure the progression of cellular differentiation, cartilage development, endochondral bone formation, and intramembranous osteogenesis [66 – 70]. Systemic factors are also critical for the early stages of limb formation and influence the renewal of osteogenesis throughout adult life. While parathyroid hormone (PTH) stimulates the growth of osteoprogenitor populations [71], PTHrP functions as a cellular cytokine regulating cell growth for differentiation in development [72,73] and is a key regulator of chondrocyte maturation [74]. PTHrP influences Hox gene expression and both peptide and its receptor are upregulated by TGF- [75 – 77]. Ablation of the PTH/PTHrP receptor in mice [28,74,78] and mutations in the receptor in human [79] reveal its central role in the regulation of endochondral bone formation (EBF). The disorganized growth plate of the receptor knockout mouse can be rescued by expression of the receptor; however, endochondral bone formation becomes delayed [80,81] (Table 1). Commitment of stem cells to specific mesenchymal lineages occurs early in development of the limb. Transcription

FIGURE 2

factors which function as “master switches” mediate cell differentiation by induction of a set of phenotypic genes that characterize the muscle, adipocyte, chondrocyte, or osteoblast cells (Fig. 2). Such regulatory proteins have also been implicated in key roles during the progressive differentiation of osteoblasts through specific stages of maturation (Fig. 2, lower panel). Among the most notable are cfos, a protooncogene, the helix – loop – helix proteins (Twist, Id, Scleraxis), leucine zipper proteins (hXBP-1), zinc finger proteins (zif268), homeodomain proteins (Msx-2, Dlx-5), steroid receptors, and runt domain transcription factors (RUNX2/Cbfa1/AML/PEBP2), which are obligatory for osteoblast differentiation. Modifications in the representation of classes of transcription factors at specific sites during embryogenesis and at different stages of osteoblast differentiation (Fig. 2, bottom) reflect linkage to the transcriptional control of osteoblast phenotype development. Helix – loop – helix factors, negative regulators of osteogenesis, illustrate this point. Id (inhibitor of differentiation), twist, and scleraxis are expressed in mesoderm of developing embryo [82,83]. Scleraxis is expressed in cells that form the skeleton and is not detected at the onset of ossification [83]. Id and Twist expression must be downregulated for osteoblast differentiation to proceed [84,85]; overexpression of these factors inhibits osteogenesis in vitro [86]. Msx-2 and Dlx-5 are members of the homeodomain gene family of transcription factors [87]. Both factors are expressed in mesenchymal cells at sites that will undergo skeletogenesis

Transcriptional control of phenotype lineages from mesenchymal stem cells. (Top) Activation of gene transcription for commitment to the muscle cell phenotype by MyoD to adipocytes by PPAR2 and C/EBP, chondrocytes by SOX9, and osteoblasts by BMP-2 and CBFA1 is illustrated. Several in vitro studies demonstrate that phenotype commitment can be blocked or switched by forced expression of the regulatory factor characteristic of a different lineage (designated ) [175,706 – 710]. (Bottom) Developmental expression of several transcription factors influencing osteoblast differentiation. Thick lines represent highest cellular levels.

26 [88 – 90]. Their importance for normal bone formation is realized by skeletal abnormalities that result from mutations or misexpression of these factors [91; and see Table 1]. Msx-2 must be down regulated for progression of osteoblast differentiation [92,57,8]. The runt homology domain-related core-binding factor RUNX2 has been shown to play an essential role in bone formation in the embryo, demonstrated by the inhibition of mineralized tissue formation in the CBFA1/ RUNX2 null mutation mouse model [58,59]. The family of RUNX/CBFA transcription factors comprises three related genes that each support tissue specification and organogenesis [93– 95]. RUNX1 (CBFA2/AML-1B/PEBP2B) and its partner proteins CBF are critical for hematopoietic cell differentiation [96 – 98]; RUNX3 (CBFA3/AML-2/PEBP2C) is required for gut development (Dr. Y. Ito, Kyoto University, Japan, personal communication); RUNX2 (CBFA1/ AML-3/ PEBP2A) is essential for the differentiation of osteoblastic cells for formation of the mineralized skeleton. In human, these factors were first designated as AML because they were identified as the protein encoded by a gene locus rearranged in acute myelogenous leukemia (AML). Other names include core-binding factor  (CBFA) and polyoma enhancer-binding protein (PEBP2). The Human Genome Nomenclature Committee has now referred to this family as RUNX. CBFA1 was initially identified in bone as an osteoblastspecific DNA-binding protein that activated transcription of the tissue-specific osteocalcin gene [99 – 103] (also see Chapter 6). The obligatory role of CBFA1 for the formation of mature bone in the developing skeleton has been shown by the absence of a calcified skeleton and bone formation in CBFA1 null mutant mice [58,59]. The heterozygous mice exhibited phenotypic features akin to mouse models [104] and human cleidocranial dysplasia (CCD) abnormalities. Various mutations in CBFA1 were then identified in patients with CCD [105 – 108]. CBFA1 is expressed not only in abundance in osteoblasts and hypertrophic chondrocytes, but in cartilage [109 – 111], thymus [112], and testis [113], as well as early stage osseous and chondroprogenitor cells Marrow mesenchymal stromal cells lacking overt expression of chondrocyte or osteoblast markers and several nonosseous mesenchymal cell lines also have significant CBFA1 expression [114 – 116]. The transient expression of CBFA1 in early embryogenesis, followed by an upregulation in late stages of bone development [58,102,103], suggests that CBFA1 may be important in both early specification of the mesenchymal stromal phenotype and in supporting the final stages of osteoblast differentiation [103,117]. Evidence is also provided by the observations that RUNX antisense blocks in vitro differentiation of osteoblasts to the final mineralization stage [103]. Furthermore, CBFA1 can induce the expression of bone-related genes in nonosseous

LIAN AND STEIN

cells [102,103,118]. Finally, a dominant-negative CBFA1 mutant protein, expressed only in mature osteoblasts by the osteocalcin promoter, resulted in osteopenic bone due to decreased osteoblast activity [117]. The activities of growth regulators of skeletal development, FGFs, TGF-1, or BMP and CBFA1, appear to be closely linked. Activating mutations in FGF receptors cause premature fusion of the calvarial sutures in human disorders [119] and in a recently described mouse model [120]. FGF expression in this disorder is linked to expression of CBFA1. FGFR1 and ligands FGF2 and FGF8 induce CBFA1 in vivo and in C3H10T1/2 cells. Several reports have shown that CBFA1 is a downstream target of BMP-2 in osteoblastic and nonosseous cell lines [114,118,121,122] and BMP-7 [102]. Evidence has been presented demonstrating a protein – protein interaction of the CBFA2 factor with Smad1 or Smad 5 for functional cooperativity of TGF-mediated gene transcription [123]. Furthermore, Smad2 and CBFA1 cooperativity in osteoblasts has been shown [124]. These studies reflect cross-talk between two classes of signaling factors essential for osteogenesis, the Smad proteins, which mediate TGF-1 and BMP activities, and CBFA1 factors. Other studies suggest that CBFA1 induction of osteogenesis may require a BMP responsive factor. For example, while both TGF and BMP-2 induce CBFA1 expression [121] in the premyoblastic C2C12 cell line, only BMP-2 leads to expression of the osteoblast phenotype. Thus, BMP-2, but not TGF-1 signaling, leads to a factor that, together with CBFA1, may be required to mediate osteogenesis. This concept is further supported by the ability of BMP-2 to activate osteocalcin in calvarial cells from CBFA1 null mice [58]. In addition, subclones of the osteoblastic MC3T3-E1 cell line, which expressed CBFA1, were found incompetent for the synthesis of a mineralized bone ECM [125]. Thus, while CBFA1 gene ablation studies in mice revealed that CBFA1 is necessary for bone formation and can activate some osteogenic genes in nonosseous cells, CBFA1 may not be sufficient for de novo skeletal formation. We still do not understand if CBFA1 can provide a cell with the same cascade of signals induced by the bone morphogenetic proteins necessary for bone formation. In summary, our present knowledge of signaling factors and transcriptional regulators of bone development is growing. The significance of many of these growth factors and morphogens in regulating the progression of the pluripotent stem cell and multipotential mesenchymal cell to the committed osteoprogenitor and finally to recognizable osteoblasts is appreciated from mouse models in which these genes or receptors for these proteins have been expressed in transgenic mouse models or ablated (null mutations) with consequences on formation of the skeleton (summarized in Table 1). However, identification of the sequelae of events and integration of their activities can only be postulated at the present time. It is apparent that the convergence of multiple pathways and the

27

CHAPTER 2 Osteoblast Biology

coordination of activities are operative through complex feedback loops as illustrated above (Fig. 1), not only in development, but in renewal of osteogenesis in the adult [126,127].

III. IN VIVO TISSUE LEVEL ORGANIZATION OF OSTEOBLASTS: PHENOTYPIC FEATURES AND FUNCTION Based on morphological and histological studies, osteoblastic cells are categorized in a presumed linear sequence progressing from osteoprogenitor cells to preosteoblasts, which mature to osteoblasts and then to lining cells or osteocytes [128 – 134] (Fig. 3). There is a gradient of differentiation that can be observed morphologically either in the periosteum or in the marrow as the osteoprogenitor cell reaches the bone surface and the osteoblast phenotype becomes fully expressed. Osteoblasts that are derived from proliferating osteoprogenitors can be observed in clusters at the bone surface (Fig. 4, see also color plate). These cells synthesize the bone extracellular matrix, designated osteoid (Fig. 5, see also color plate). In metabolic bone disorders leading to decreased calcium or phosphate deposition in bone, as in vitamin D deficiency, wide osteoid seams are evident (Fig. 5B). Mineralization leads to the

FIGURE 3

final stage of osteoblast differentiation. When the boneforming osteoblast becomes encased in its own mineralized matrix, they are osteocytes. On a quiescent bone surface, the osteoblast flattens to a lining cell, forming an endosteum. Four forms of the osteoblast cell lineage are thus recognized in vivo. They are the committed progenitors (preosteoblasts), mature osteoblasts, osteocytes, and the bone-lining cell (Figs. 4 and 5). Particularly important in defining phenotypic differences is an understanding of gene expression and protein localization at the single cell level in relation to the development of bone tissue organization. During the past, high-resolution in situ hybridization methods have been developed for bone sections that effectively support the assessment of mRNA levels in specific cells and with respect to intracellular localization [135 – 139]. The application of immunological detection methods [140 – 142] to sections of intact bone has supported the establishment of linkage between patterns of gene expression observed in culture with those that occur developmentally in tissue. Distinct gradients of particular markers are evident, a phenomenon that may be related to a coupling of cell polarity with physiological function [143]. Polymerase chain reaction (PCR) methods have been applied successfully to determinations of cellular RNAs in single cells [144], and in situ PCR protocols offer the potential for extending this approach [145,146]. Variable levels of expressed genes are observed in neighboring cells.

Growth and differentiation of osteoblast lineage cells. Progression through the osteoblast lineage from a pluripotent mesenchymal stem cell to a mature osteocyte is regulated by numerous physiologic mediators, including, but not limited to, transforming growth factor- (TGF-), superfamily members, steroid hormones, growth regulators, and transcription factors. Each of these subpopulations of osteoblasts have many common features, but exhibit distinct different properties/functions and express osteoblast phenotypic genes to varying extents (see text for details).

28

FIGURE 4 Organization of osteoblasts in bone tissue. (A) Intramembranous bone formation (40). (B) Bone-forming surface (200), Goldner and von Kossa stain shows (A) osteoblast clusters and (B) surface osteoblasts, preosteocytes below the osteoid (blue), and mineralized tissue (black) with an osteocyte in its lacunae. Osteoprogenitor cells are evident at higher magnification behind the large surface cuboidal osteoblasts. Osteocyte cellular processes that extend through canaliculi cannot be seen with light level microscopy. (See also color plate.)

Viable approaches are thereby provided for investigating protein functions within multiple contexts that range from transactivation in the nucleus to support of extracellular matrix mineralization. The following summarizes the morphologic and phenotypic functional features of each osteoblast population. For a more detailed description of bone cell ultrastructure, the reader is referred to Refs. 130,132, 134, and 147.

A. Osteoprogenitors and Preosteoblasts: More Cells Build Bigger Bones Progression of the most primitive pluripotent cell to the undifferentiated multipotential mesenchymal cell and presumed osteoprogenitor is not understood. The expression and subsequent differentiation of the early osteoprogenitor

LIAN AND STEIN

FIGURE 5 Mineralization defects due to vitamin D deficiency and low serum calcium results in osteomalacia. Cortical bone sections from a 6week control rat (A) and a rat maintained on a vitamin D-deficient diet from weanling to 6 weeks. (B) Osteoblast differentiation from the periosteal and endosteal surfaces is seen, as well as a wide osteoid due to a much greater decrease in the rate of mineral deposition. Toluidine blue/von Kossa stain (200). (See also color plate.)

cells should be considered within the context of embryologic development, bone formation during growth, and bone tissue remodeling in the adult skeleton. Under each of these circumstances, progenitor cells must be responsive to a broad spectrum of regulatory signals that mediate their proliferation, commitment, and progression of phenotype development, as well as sustaining their structural and functional properties. In fully developed bone, there is a requirement for utilization of the same factors that can mediate the growth and differentiation of osteoprogenitor cells during skeletal development, as well as for osteoblast differentiation during bone remodeling and fracture healing in the adult [127,148]. From a developmental perspective, mesenchymalderived osteoprogenitor cells arise/reside in the periosteal tissue or the bone marrow stroma. The marrow and its stromal “bedding” give rise to multipotential cells of both

CHAPTER 2 Osteoblast Biology

hematopoietic lineage (origin of osteoclasts) and nonhematopoietic lineage cells from which many tissuespecific cells derive, such as chondrocytes, myoblasts, and adipocytes. When suspensions of marrow cells are plated in vitro, clonal colonies of adherent fibroblasts are formed; each derived from the single cell that has been designated as the colony-forming fibroblastic unit or CFU/F [149,150]. Formation of CFUs requires the presence of hematopoietic cells [151]. A proportion of these cells have high proliferative and differentiation capacity and exhibit characteristics of stem cells when transplanted in the closed environment of a diffusion chamber [152] or transplanted into the circulation [153]. Studies have demonstrated that marrow stromal cells can be transfected with reporter genes and then transplanted either into specific tissue sites (Subcutaneous, bone tumors, fracture sites) or systemically (e.g., by tail vein infusion) and they maintain competency for differentiation in vivo [154]. For clinical applications of genetically engineered cells for the local reconstruction of bone tissue [155 – 159] or to treat systemic bone diseases [153,160 – 163], in vitro expansion and modification of the cells forming adherent marrow colonies are important considerations. However, mechanisms related to the retention of in vivo properties, homing, engraftment, and differentiation after transplantation must be addressed [159,164 – 166]. Many groups have shown the adherent marrow population differentiates in vivo and in vitro to several mesenchymal lineage cells: adipocytes, chondrocytes, osteoblasts, and myoblasts [reviewed in 167,116,168 – 174]. The plasticity of these lineages is indicated by several lines of evidence. Forced expression of the transcription factors that function as “master switches” (Fig. 2) in phenotype commitment can transdifferentiate a cell to a different phenotype. The reciprocal relationship between adipocyte and osteoblast differentiation is suggested by such studies [115,175]. Forced expression of PPAR2 in marrow stromal cell lines results in the inhibition of terminal osteoblast differentiation with concomitant downregulation of CBFA1 [115,175]. The bipotential property of the late stage osteoprogenitor or preadipocyte [176 – 183] is markedly sensitive to biological regulatory signals influencing “master switch” transcription factor expression. Regulatory signals influencing osteogenesis in preference to adipogenesis can include BMP-2 and the BMP receptor [115,179,182, 184 – 187] and TGF-1 [188]. Retinoid signaling pathways [189] and leptin signaling (through the central nervous system) [181,190] favor adipogenesis. In contrast, 1,25(OH)2D3 inhibits adipogenesis [191]. One molecular mechanism for commitment to the osteogenic or adipogenic lineage which has been identified involves the role of mitogen-activated protein kinase family members (ERK, JNK, and p38). ERK functions as positive regulators of osteogenic differentiation. Inhibiting ERK activation blocked

29 the bone phenotype and resulted in the adipogenic differentiation of human mesenchymal stem cells [192]. Implications for the plasticity of osteoprogenitor and adipocyte cells in relation to osteoporosis and the aging skeleton have been reviewed in detail [180,193]. Presently, a key obstacle in understanding the origin of osteoblast lineage cells is the inability to identify an osteoprogenitor cell prior to the expression of bone phenotypic properties. Using characterization of hematopoietic stem cells as a paradigm, several groups have developed antibodies to cell surface proteins using presumptive marrow stromal cell populations. These reagents have the potential for both recognition and purification of skeletal stem cells [172,194 – 202]. STRO-1 positive cells [203] are well documented with respect to their pluripotential and osteoprogenitor properties [196,204 – 206]. Antigens for other antibodies have been characterized. Interestingly, an antigen to a cell surface marker antibody (SB-10) produced in response to mesenchymal stem cells is the activated leukocyte cell adhesion molecule ALCAM [195]. Expression of ALCAM becomes downregulated in concert with changes in morphology and detection of alkaline phosphatase activity of the periosteal osteoprogenitors as they migrate and develop into osteoblasts. There still remains considerable debate as to whether pluripotential osteoprogenitors (marked, for example, as STRO-1 or ALCAM positive cells) have the capacity to function as a “stem cell.” Proliferation and differentiation of the osteoprogenitor pool is influenced by several regulatory factors. Plateletderived growth factor and epidermal growth factor [207] have been identified as important in stimulating expansion of the CFU/F [151]. The leukemia inhibitory factor (LIF) maintains stem cell populations and osteoprogenitors and inhibits their differentiation in vitro [208,209], but has also been reported to have osteogenic activity in vivo [210]. The fibroblast growth factor-2 [211] and TGF-1 are potent mitogens for periosteal osteoprogenitors and marrow stromal cells [212 – 214]. The osteoinductive effects of bone morphogenetic proteins are complex, modulating growth, osteoinduction, and even apoptosis, depending on the specific BMP, concentration dependency, and the progenitor cell phenotype [36,185,215,216]. BMP-2 rapidly induces osteoblast differentiation in marrow stromal cells [37,183, 217,218], but the effect is slower in a number of pluripotent cell lines [219,220] and in the mouse myogenic C2C12 cell line [221]. These growth factors are expressed and produced by osteoblast lineage cells and are stored in the bone extracellular matrix. A local mechanism for stimulating the proliferation of progenitors in the bone microenvironment is thereby provided [222,223]. The osteoprogenitor appears to have limited selfrenewal capacity compared to the stem cell [224]. In contrast, a key feature of the osteoprogenitor/preosteoblast population is its capacity to divide and increase the size of

30

LIAN AND STEIN

bone. Labeling studies ([3H]thymidine and autoradiography) indicate that the proliferating cells are principally confined to progenitor cells and preosteoblasts with very few osteoblasts labeled [225 – 227]. The determined osteoprogenitor is recognizable in bone as a preosteoblast. Preosteoblasts are usually observed as one or two layers of cells behind the osteoblast near bone-forming surfaces [130 – 134,147,228]; i.e., they are usually present where active mature osteoblasts are laying down a bone matrix. These cells appear elongated, fibroblastic, or spindle shaped with an oval or elongated nucleus and with notable glycogen content. However, some of the cells in the layer found directly behind the active osteoblasts may appear morphologically closer to the cuboidal osteoblast and can clearly be defined as preosteoblasts (Fig. 4). Preosteoblasts may express a few phenotypic markers of the osteoblast, e.g., alkaline phosphatase activity, but less than mature osteoblasts [147,229]. The preosteoblast, however, has not yet acquired many of the differentiated characteristics of mature osteoblasts; for example, there is no evidence of a developed rough endoplasmic reticulum [130].

B. Cell Cycle Regulatory Parameters: Control Mechanisms Supporting Osteoblast Growth With recognition of decreased osteoblast surfaces in osteoporotic bone [230] and reports of decreased marrow

osteoprogenitors with age [231 – 235], defining mechanisms contributing to the regulation of proliferative activity in osteoblast lineage cells is increasing in importance. To understand regulatory parameters of proliferation, one must consider mechanisms that support the requisite responsiveness to growth factors through signaling pathways and the consequent induction of proliferation. To explain the induction, synthesis, activation, and suppression of the complex and interrelated regulatory factors associated with the growth control of osteoprogenitor cell proliferation in vivo, an understanding of mechanisms that control cell proliferation is required. Proliferation is controlled through the cell cycle by the activity of regulatory proteins which support progression of cells that have responded to a mitogenic stimulus through DNA replication and cell division. The cell cycle is a stringent growth-regulated series of sequential biochemical and molecular events that support genome replication and mitotic division [reviewed in 236]. The stages of the cell cycle and checkpoints that monitor competency of cells to progress through DNA replication and mitotic division are illustrated in Fig. 6. Suppression of certain cell cycle regulated genes is requisite for the cessation of proliferation and upregulation of phenotypic genes. When quiescent cells (G0) are stimulated to proliferate and divide, they enter G1, the first phase of the cell cycle where the enzymes required for DNA replication are synthesized. Before a cell can progress through G1 and begin DNA synthesis (S phase), it must pass through a checkpoint

FIGURE 6 Control of cell cycle progression in bone cells. The cell cycle is regulated by several critical cell cycle checkpoints (indicated by check marks), at which competency for cell cycle progression is monitored. Entry into an exit from the cell cycle is controlled by growth-regulatory factors (e.g., cytokines, growth factors, cell adhesion, and/or cell – cell contact) that determine the self-renewal of stem cells and expansion of precommitted progenitor cells. The biochemical parameters associated with each cell cycle checkpoint are indicated. Options for defaulting to apoptosis during G1 and G2 are evaluated by surveillance mechanisms that assess the fidelity of structural and regulatory parameters of cell cycle control. Apoptosis also occurs in mature differentiated bone cells.

CHAPTER 2 Osteoblast Biology

in late G1, which is known as the restriction point [237]. At this cell cycle restriction point, both positive and negative external growth signals are integrated. If conditions are appropriate, the cell proceeds through the remainder of G1 and enters the S phase. Once the cell passes the restriction point, it is refractory to withdrawal of mitogens or to growth inhibitory signals and is committed to progressing through the remainder of the cell cycle unless it is subjected to DNA damage or metabolic disturbance [237]. In mammalian cells, progression through the cell cycle is regulated by a cascade of cyclin-containing growth regulatory factors that transduce growth factor-mediated signals into discrete phosphorylation events, controlling the expression of genes responsible for both initiation of proliferation and competency for cell cycle progression. Cyclin activity is modulated by the formation of complexes with a family of threonine/serine kinases designated cyclin-dependent kinases (cdk) [238,239]. Cdks are regulated by both positive and negative phosphorylation, as well as by their reversible association with specific cyclins during defined phases of the cell cycle [240]. In general, the levels of cdk proteins remain relatively constant during the cell cycle, whereas the expression of specific cyclins is confined to distinct phases of the cell cycle where they are degraded quickly after having completed their function. An emerging concept is that the cyclins and cdks are responsive to regulation by the phosphorylation-dependent signaling pathways associated with activities of the early response genes, which are upregulated following the mitogen stimulation of proliferation [reviewed in 240 – 243]. Cyclindependent phosphorylation activity is functionally linked to activation and suppression of both p53 and RB-related tumor suppressor genes [244,245]. p53 accumulates in response to stress, inducing arrest at G1 or G2. The retinoblastoma protein (Rb), a tumor suppressor, is a member of a family of related proteins that include p105, p107, and p130. Rb has been shown to have a critical role in the regulation of cell proliferation, particularly in progression through G1 [reviewed in 244]. Rb functions as a signal transducer, receiving both growth-promoting and -inhibitory signals and linking them to the transcriptional machinery required for cell cycle progression or cell cycle arrest. In quiescent cells or cells reentering G1 from mitosis, Rb exists in an underphosphorylated or dephosphorylated state. Phosphorylation of Rb occurs late in G1 and modifies the activities of regulatory complexes that are required for gene expression linked to the onset of S phase [246]. The activities of the cdk are downregulated by a series of inhibitors (designated CDIs) and mediators of ubiquitination, which signal destabilization and/or destruction of these regulatory complexes in a cell cycle-dependent manner [247]. The cyclin inhibitory protein (CIP) class of CDI includes the proteins p21, p27, and p57. Growth arrest is, in part, due to induction of the cyclin-dependent

31 kinase inhibitor protein p21, which can interact with multiple cyclin – cdk complexes. The INK class is represented by proteins p15, p16, p18, and p19, which are linked to apoptosis control mechanisms. Expression of cell cycle regulatory proteins, cyclins, and cyclin-dependent kinases appears not to be solely confined to control of proliferation but, for example, associated differentiation in bone osteoblasts and nonosseous cells [248 – 250]. Cell cycle regulatory factors, particularly cyclin E, have been noted in several systems involved in the regulation of differentiation, in myoblasts [251], in osteoblasts [249], and in promyeloid cell differentiation into macrophages [252]. During osteoblast differentiation, cyclin-dependent kinase inhibitors (cdki) are also developmentally expressed. The cdki p21 (CIP/WAF1) is expressed in the growth period. In contrast, p27 (KIP-1) is expressed in the immediate postproliferative period and is upregulated again during differentiation [253]. Studies characterizing bone abnormalities associated with null mutations of cell cycle and cell growth regulatory factors have revealed their significance in providing signals for the control of both the number and the differentiation of bone-related cells. For example, marrow harvested from p27-/- mice shows a three- to fourfold increase in osteogenic nodule formation compared to wild type. Thus absence of this cdk inhibitor allows the marrow population to extend their growth phase, increasing the cell numbers. This expansion of the osteoprogenitor population is consistent with the larger size of the animals and the proportionally increased cortical width of the long bones [253]. Investigations of the effects of growth factors and osteogenic hormones on cell cycle target genes are increasing our understanding of their precise molecular mechanisms in the regulation of growth, differentiation, and apoptosis of osteoprogenitor cells and osteoblasts (Fig. 6). Several studies have reported BMP-2 and BMP-4 induction of cell cycle arrest in the G1 phase that is mediated by enhanced expression of the p21 cyclin inhibitor [254] and rapid induction of cyclin G, a cyclin that is increased after the induction of p53 by DNA damage [255]. Both of these events are linked to the induction of apoptosis, and in the developing tooth, p21 and BMP-4 are coexpressed in cells destined to undergo apoptosis in a transitional epithelial structure known as the enamel knot [256]. The apoptotic-promoting effects of BMP-2 have been reported to oppose the estradiol-induced growth of human breast cancer cells. Where estradiol stimulates cyclins and cyclin-dependent kinases, the BMP induction of the cyclin kinase inhibitor p21 leads to the inactivation of cyclin D1 [257]. The abundance of TGF- and BMPs in the early stages of osteoblast maturation and the targeting of BMP action to p21 may provide a mechanism not only for promoting osteogenic differentiation, but for apoptosis of proliferating cells that are recruited to the bone surface and may not progress to the mature osteocyte.

32

LIAN AND STEIN

Other cytokines and growth factors that target the proliferation phase have their effects coupled through p21. IL-6 promotes differentiation and exhibits antiapoptotic effects on human osteoblasts [258]. The effects of IL-6 on the p21 promoter are mediated by STAT-binding proteins and a STAT response element in the p21 promoter. Fibroblast growth factors are classic mitogens of the osteoprogenitor pool, as well as modulators of osteoblast differentiation [259 – 263]. FGF signaling also activates STAT1 and p21 [264], a mechanism that accounts for the ability of FGF-2 to induce both mitogenic responses and growth arrest in cancer cells [264,265]. TGF- also inhibits cell cycle progression in part through the upregulation of p21 gene expression [253,266]. Regulation of the p21 promoter is mediated by the TGF- induction of Smad3 and Smad4 [266,267]. The steroid hormone 1,25(OH)2D3 exerts antiproliferative effects in undifferentiated cells also mediated by the enhanced expression of p21 [268] and p27 [253]. This finding is consistent with the high levels of p27 in mature osteoblasts and 1,25(OH)2D3 induction of markers of the mature osteoblast phenotype. It is becoming increasingly evident that each step in the regulatory cycles (cell cycle, cyclin/cdk cycle, cdki cycle) governing proliferation is responsive to multiple signaling pathways and has multiple regulatory options. The diversity in cyclin – cyclin-dependent kinase complexes accommodates the control of proliferation under multiple biological circumstances and provides functional redundancy as a compensatory mechanism. Similarly, the inhibitors of cyclin – cdk complexes bind to and regulate multiple cyclin – cdk-containing complexes at several checkpoints [240,269,270]. The regulatory events associated with these proliferation-related cycles support control within the contexts of (a) responsiveness to a broad spectrum of positive and negative mitogenic factors, (b) cell – cell and cell – extracellular matrix interactions, (c) monitoring genome integrity and invoking DNA repair and/or apoptotic mechanisms if required, and (d) competency for differentiation. Perturbation of any of these cell cycle regulatory mechanisms can result in unregulated or neoplastic growth.

C. Osteoblasts: Producer of the Bone Extracellular Matrix When the preosteoblast ceases to proliferate, a key signaling event occurs for development of the mature osteoblast from the spindle-shaped osteoprogenitor. The osteoblast expresses all of the differentiated functions required to synthesize bone. Osteoblasts are defined in vivo by their appearance along the bone surface as large cuboidal cells actively producing matrix (Figs. 4 and 5), which is not yet calcified (osteoid tissue). Several structural features characterize this osteoblast, including its size and

cuboidal morphology, a round distinguishing nucleus at the base of the cell (opposite to the bone surface), a strongly basophilic cytoplasm, and a prominent Golgi complex located between the nucleus and the apex of the cell [271]. At the ultrastructural level, one observes an extremely welldeveloped rough endoplasmic reticulum with dilated cisternae and a dense granular content, and a large circular Golgi complex consisting of multiple Golgi stacks. These are typical characteristics of a secretory cell. The osteoblast synthesizes and vectorially secretes most of the bone ECM protein; others are accumulated. Fetal bone is enriched in type III collagen, whereas type I collagen accounts for 80 – 90% of the osteoid in the adult with minor collagens type V. Specialized noncollagenous proteins for cell adhesion and with calcium and phosphatebinding properties are found in variable amounts changing with age. The most abundant noncollagenous proteins include osteonectin, osteocalcin, bone sialoprotein, and osteopontin; and based on many in vitro studies, it is assumed that they participate in specialized functions necessary for both the structural integrity of bone tissue and bone turnover. The protein properties and gene regulation of bone extracellular matrix constituents are detailed elsewhere in this volume (see Chapter 4). The primary functional activity of the active surface osteoblast is production of an extracellular matrix with competency for mineralization. In this regard, the high level of the tissue-nonspecific alkaline phosphatase (TNAP) (bone, kidney, liver isoform) and the ability to synthesize a number of noncollagenous proteins that are in either representative or restricted abundance in mineralized tissues are important features. For example, hypertrophic chondrocytes and odontoblasts share some phenotypic features with osteoblasts, as osteocalcin expression [272,273] and high alkaline phosphatase activity, sufficient to allow for histochemical detection in cells associated with the active formation of mineralized matrix. Alkaline phosphatase activity, a hallmark of the osteoblast phenotype, is a widely accepted marker of new bone formation and early osteoblast activity. Gradations of enzyme intensity and mRNA expression are found in bone with lowest levels (or absence) in osteocytes and osteoprogenitors and maximal levels in surface osteoblasts and hypertrophic chondrocytes at the mineralization front [134,229]. Alkaline phosphatase activity is still considered critical to the initiation of mineralization, a concept supported by characterization of the genetic defect in hypophosphatasia [274]. Several of the noncollagenous bone-related proteins are implicated in initiating or regulating the mineral phase of bone from studies of the biochemical, molecular, and functional properties [275,276] (Table 1). However, ablation of the genes encoding some of the more abundant and bone restricted noncollagenous proteins (osteocalcin [277,278], osteopontin [279], biglycan [280]) have resulted in subtle changes in

CHAPTER 2 Osteoblast Biology

the mineral phase of bone. Future directions in which double mutants are derived may reveal severe mineralization defects. The anabolic activities of osteoblasts are regulated in large part by the insulin-like growth factor (IGF)/IGFbinding protein (IGFBP) system [281; reviewed in 282]. IGF-I and IGF-II are synthesized in may tissues and both are highly expressed in active osteoblasts [283]. IGFs stimulate cell proliferation and collagen synthesis [284 – 286] and, at the same time, inhibit matrix collagen degradation by decreasing collagenase 3 transcription [287]. The synthesis of IGF-I is regulated by physiologic mediators of bone formation, PTH stimulates [288], whereas glucocorticoids [289] and growth factors (e.g., FGF-2, PDGF-2) are inhibitory [290]. The activities of IGF-I and IGF-II are regulated by a family IGFBPs, designated IGFBP-1 through IGFBP-6 [291]. These binding proteins have either stimulatory effects (e.g. IGFBP-5) [286,292 – 294] or inhibitory activity (e.g., IGFBP4) [294] and appear to be expressed at different levels in subpopulations of osteoblasts [295]. Several clinical studies are revealing associations of IGF-1 and IGFBPs with metabolic bone diseases [296 – 298, and reviewed by 299]. Regulation of bone development and cellular differentiation by the extracellular matrix (ECM) is well established [300 – 302] (see Chapter 4). Osteoblast differentiation and functional activity are supported by cell – matrix interactions [303,304]. Functional studies establishing requirement of the type I collagen and other ECM components in promoting osteoblast/osteocyte differentiation have been carried out by modifying the production of osteoblastsecreted products or culturing osteoblasts on various matrices [305 – 308]. The molecular mechanisms mediating osteoprogenitor/osteoblast – matrix interactions are being identified. A spectrum of integrins have been shown to be expressed by osteoblasts [309]. Interactions between integrin receptors and fibronectin are required for both osteoblast differentiation [310,311] and cell survival [312]. Osteoblasts appear to use 1 integrins to adhere to the full range of RGD-containing bone matrix proteins. Disruption of the collagen – 21 interaction suppresses expression of the osteoblast phenotype [313,314]. The noncollagenous RGD containing bone matrix protein, bone sialoprotein (BSP), was also shown to mediate the collagen-induced osteoblast differentiation of bone marrow cells [315]. This activity of BSP is consistent with the reported coordinated expression of 3 integrin and BSP during osteoblast differentiation in vitro [316,317]. Based on early findings of Vukicevic et al. [306], who reported laminin-1 stimulates osteoblast differentiation, and recent observations that osteoprogenitors selectively attach to laminin-1 [318], it will be instructive to address the specificity of cell – matrix interactions with respect to subpopulations of osteoblast lineage cells. Studies are identifying integrin receptors in the

33 regulation of BMP-2-mediated differentiation of mesenchymal cells [319]. In addition to cell – matrix interactions, cell – cell communication is important for the differentiation and maturation of osteoblasts. Cytoplasmic processes on the secreting side of the surface osteoblast extend deep into the osteoid matrix and are in contact with the extended cellular processes of osteocytes. Junctional complexes (gap junctions) are often found between the osteoblasts on the surface as well as between cellular processes. In this manner, surface osteoblasts establish cell – cell communication with neighboring cells in the mineralized matrix. Gap junctions are a structure of six multiple protein units (connexins) that couple with an identical unit in a neighboring cell to form a channel connecting the two cytoplasms. Studies in osteoblasts [320,321] suggest that the selective utilization of connexin proteins contributes to the modulation of molecular permeability. Several members of the cadherin family of cell – cell adhesion proteins are expressed in osteoblasts, including cadherin-11, cadherin-4, N-cadherin, and OB-cadherin [50,51]. N-cadherin is present on proliferative preosteoblastic cells and may support osteoblast differentiation [322], but is lost as they become osteocytic [323]. In contrast, OB-cadherin is barely detected in osteoprogenitor cells and is upregulated in alkaline phosphatase-expressing cells [324]. Indeed, the relative abundance of cadherin defines the differentiation pathway of mesenchymal precursors to specific lineages [51,325]. Expression of R-cadherin, N-cadherin, and cadherin-11, present in progenitor cell lines, is modified in response to differentiation, e.g., R-cadherin is downregulated and cadherin-11 upregulated in response to BMP2-induced osteogenesis. In addition, ICAM-1 and VCAM-1 have been reported on the osteoblast surface, thereby providing a potential mechanism for T-cell interactions that contribute to the regulation of bone turnover [326]. Signaling pathways from the extracellular matrix through the cytoskeleton and finally to the nucleus, which allow expression and upregulation of bone-specific and bone-related genes, need to be investigated. For example, -catenin, which colocalizes and coprecipitates with cadherins [50], is a potential candidate. CD44, the hyaluronate receptor, is a nonintegrin adhesion receptor that is linked to the cytoskeleton. CD44 has been identified as a useful marker for osteocyte differentiation [200,201] and is also expressed in osteoclasts [327].

D. Osteocytes and Bone-Lining Cells: Gatekeepers of the Structural Integrity of Bone As the active matrix-forming osteoblast becomes encased in the mineralized matrix, the cell differentiates further into osteocytes. Labeling studies suggest that the

34 transition from an osteoblast to an osteocyte is approximately 3 – 5 days [169,328]. The osteocyte is considered the most mature or terminally differentiated cell of the osteoblast lineage. Osteocytes are embedded in bone matrix occupying spaces (lacunae) in the interior of bone and are connected to adjacent cells by long cytoplasmic projections, enriched in microfilaments, and lie within channels (canaliculi) through the mineralized matrix. These cell processes maintain contact with other osteocytes or with processes from the cells lining the bone surface [134,329]. It is estimated that 25,000/mm3 osteocytes are found embedded within the bone, contributing to the metabolic functions of bone [330]. Some, but not all, of the biochemical features of the osteoblast are expressed in the osteocyte. The morphologic properties of osteocytes change as the surrounding osteoid mineralizes and reflects their functional activity (Fig. 3). Being derived from osteoblasts, a young osteocyte has many of the ultrastructural characteristics of a cell involved in protein synthesis (rough endoplasmic reticulum, large Golgi), except that there is a decrease in the volume of the cell. An older osteocyte, located deeper within the calcified bone, shows less of these features; and in addition, glycogen stores become evident in its cytoplasm. Osteocytes have been shown to synthesize new bone matrix at the surface of the lacunae, and there is some evidence for their ability to resorb calcified bone from the same surface [331]. The osteocyte is a terminally differentiated cell, not capable of cell division even when separated from its matrix. In isolated cultures, they retain their cellular projections [332]. Osteocytes will be phagocytized and digested together with the other components of bone during osteoclastic bone resorption. On quiescent bone surfaces, the osteoblast develops into a flattened bone-lining cell of a single layer forming the endosteum against the marrow and underlying the periosteum directly on the mineralized surfaces. These osteoblasts are in direct communication with the osteocytes within the mineralized matrix through cellular processes that lie within the canaliculi. In the adult, the majority of bone surfaces are occupied by another osteoblast subtype with distinct phenotypic features. The bone-lining cell displays a flat and highly elongated cell shape with a spindle-shaped nucleus and fewer organelles than active osteoblasts [130, 271,329,333]. They are considered to provide a selective barrier between bone and other extracellular fluid compartments and contribute to mineral homeostasis by regulating the fluxes of calcium and phosphate in and out of bone fluids [333]. The organization of osteocytes in bone reflects their function and their capacity to respond, allowing these cells to communicate and transmit regulatory signals. The osteocytes and surface-lining cells form a continuum, or syncytium, by connection of their cytoplasmic projections through gap junctions that facilitate the exchange of both mechanical and metabolic signals [334,335] for

LIAN AND STEIN

responsiveness to physiologic demands on the skeleton. Osteoblasts and osteocytes are coupled metabolically and electrically through different gap junction proteins called connexins, described earlier [321,336,337]. Rapid fluxes of bone calcium across these junctions are thought to facilitate the transmission of information between osteoblasts on the bone surface and osteocytes within the structure of bone [338]. This structural organization and the direct contact of the active osteoblast or surface lining cells with the osteocytes is consistent with the concept that bone cells, responding to varying physiological signals, can communicate their responses. Bone-lining cells receive the majority of systemic and local signals and can transmit these to osteocytes. Reciprocally, mechanical forces on the bone produce stress-generated signals that are perceived by osteocytes, which then transmit the regulatory information to surface osteoblasts. Stress-generated electric potentials experienced by bone are either produced by strain in the organic components (piezo electric potential) or result from electrolyte fluid flow produced by deformation of the bone (streaming potential) [339 and reviewed in 340]. Thus, the ability of bone to act as a tissue responding to physiological homeostatic demands and functioning as a structural connective tissue organ to meet physical demands depends on communication among its resident cells. Studies in bone tissue and isolated cells following applied stress have advanced our understanding of osteocyte functions and responses [341]. Osteocytes produce IGF-1 and release prostaglandins in response to stress [342]. Direct evidence that osteocytes sense mechanical loading [343] has been demonstrated by rapid changes in metabolic activity by [3H]-uridine uptake [344], increased metabolic activity (e.g., glucose-6-phosphate dehydrogenase) [345], increased gene expression [346 – 348], and activation of a volume-sensitive Ca2 influx pathway potentiated by PTH [349]. Signals induced by fluid flow that have been reported include prostaglandin PGE-2, cAMP, and nitrous oxide [350 – 352]. In addition, extracellular matrix receptors, such as the integrins and CD44 receptors, are thought to mediate cellular sensing of mechanical loads [353]. The disruption of cell – matrix interactions by loading could induce a mechanical twisting of the integrins [354]. Integrins are tightly coupled to the cytoskeleton [355] and together the integrin – cytoskeleton complex facilitates the transduction of mechanical signals that may ultimately lead to modifications in gene expression [356]. Thickening of actin stress fibers and increased synthesis of cytoskeleton components in osteoblasts in response to mechanical strain have been documented [357]. The majority of the evidence to date suggests that mechanical tension can trigger bone remodeling and may favor bone formation. Increased osteopontin expression and synthesis may facilitate bone remodeling by osteoclasts [348,358,359]. However, it has been reported that mechanical strain inhibits expression of the osteoclast

CHAPTER 2 Osteoblast Biology

differentiation factor TRANCE [360]. Related to bone formation, several studies have demonstrated significant increases in PGE-2, which stimulates IRF-1 in osteocytes in response to mechanical strain [361 and reviewed in 282]. BMP-2 and BMP-4 (but not other BMPs) were induced in a model of distraction osteogenesis [362]. Thus, mechanical strain induces factors for the proliferation, differentiation, and anabolic activities of osteoblasts [363]. The life span of osteoblast lineage cells is dependent on several factors. Because more osteoblasts are recruited to bone remodeling sites than can be organized on the bone surface for further differentiation by mineralizing osteoid, a high percentage of surface preosteoblasts will die [364]. Apoptosis of preosteoblast clusters may be triggered by the lack of an adequate ECM and appropriate cell – matrix interactions for survival [312]. In vitro, osteoblast apoptosis is particularly evident in cells on the surface of multilayered bone nodules formed in primary calvarial cell cultures [365]. Apoptosis is a general mechanism for limiting organ size in embryonic development [366] and in the adult when there is a need to regenerate tissue. Here growth factor and cytokine effects on apoptosis of specific cell subpopulations in bone are likely to be contributing to tissue turnover [258,364,367]. In contrast to osteoblasts, osteocytes are very long lived in their lacunae, but will undergo apoptosis when their structural integrity is compromised. Microfracture in bones [368] and disruption of cell – cell contacts with the consequent inability to receive stimulatory signals and cell nutrients will lead to apoptosis. Increased empty lacunae and apoptotic cells (detected by DNA fragmentation using the TUNEL assay) are observed during bone turnover in aged human bone [369,370], in glucocorticoidtreated mouse models [371], and following estrogen withdrawal [372]. Parathyroid hormone [373], bisphosphonates [374], and estrogen [375,376] have been reported to prevent apoptosis. Of significance, calbindin-D (28k), which is expressed in osteoblasts, suppresses apoptosis by reducing caspase-3 activity through protein – protein interactions [377]. More studies are required to address intracellular apoptotic control mechanisms and pathways operative in the environment of the osteoblast [365].

IV. CELLULAR CROSS-TALK OF OSTEOBLAST LINEAGE CELLS: FUNCTIONS IN CALCIUM HOMEOSTASIS, BONE TURNOVER, AND HEMATOPOIESIS The biological significance and unanswered questions of the interrelationship of bone tissue cells with the hematopoietic and immune systems were highlighted at a National Institute of Health conference [378]. Osteoblasts

35 control osteoclast differentiation from hematopoietic precursors. They also support long-term bone marrow cultures and regulate hematopoiesis by the production of stimulatory factors (e.g., GM-CSF) [379,380], as well as by cell – cell interactions between early hematopoietic cells and osteoblasts via 1 integrins on CD34 cell and various cell adhesion on bone marrow stromal cells (e.g., VCAM1) [381] [reviewed in 382]. The immune and bone organ systems are linked by the production of multiple cytokines from T lymphocytes regulating bone turnover by the modulation of both osteoblast and osteoclast activities. Cross-talk between osteoblasts and other cellular systems is beginning to be investigated. Endothelial cell (EC) and osteoblast cross-talk is likely to be important for vascular invasion into the bone matrix. Osteoblasts secrete paracrine factors that regulate endothelial cell function [383], including vascular endothelial growth factor (VEGF) and its receptors [384]. VEGF secreted by ECs has been reported to enhance the anabolic effects of vitamin D3 on osteoblasts [385] and to be necessary for angiogenesis during endochondral bone formation in vivo [386]. Of note, bone sialoprotein, which is upregulated in osteogenic tumors and mediates cell attachment via V3 integrins, can promote adhesion of endothelial cells [387]. An important function of osteoblast lineage cells is their response to endocrine factors and the production of paracrine and autocrine factors for the sequelae of events mediating bone turnover. At all stages, from the initial activation of bone resorption to formation of new bone at the resorption site in the adult (bone remodeling unit), crosstalk between osteoblast lineage cells and other cell phenotypes is necessitated for the regulation of bone remodeling. The coupling of osteoblast and osteoclast activities through cross-talk is mediated by several mechanisms, which are detailed later. Direct interactions between a ligand on osteoblast lineage cells and its receptor on mononuclear preosteoclasts activate signaling cascades for osteoclast differentiation. Numerous indirect pathways are operative in which calciotrophic hormones and cytokines stimulate the secretion of factors from one cell type that activate or suppress activities of the other cell phenotype. Osteoprogenitors in the marrow, surface osteoblasts, and osteocytes have receptors for cytokines [reviewed in 388,389], parathyroid hormone, 1,25(OH)2D3 [390,391], and estrogen [392], which are key regulators of osteoclast activities. In addition to cytokine and hormone feedback loops, the transcriptional control of genes involved in the activation or suppression of osteoclast and osteoblast functions are coordinated. Two observations suggest a potential interrelationship between osteoblasts and osteoclasts by an unknown mechanism. Annexin II was identified as a vitamin D3 receptor [393] mediating the nongenomic rapid effects of 1,25(OH)2D3, which increase intracellular calcium in osteoblasts [394 – 396]. Annexin II was also reported to be

36

LIAN AND STEIN

secreted by osteoclasts and stimulate osteoclast differentiation and GM-CSF [397], an autocrine and paracrine product of osteoblasts [398]. These features provide mechanisms for regulating osteoblast and osteoclast activities to support calcium homeostasis and bone turnover. Osteoblasts and stromal osteoprogenitors primarily support osteoclastogenesis by the secretion of soluble cytokines and through cell – cell interactions with osteoclast precursors. Stromal osteoblastic cells are the major source of the macrophage colony-stimulating factor (M-CSF/CSF1) and the IL-6 family of cytokines, which are potent stimulators of bone resorption and participate in osteoclastogenesis at early and later stages. Other cytokines, such as TNF- and IL-1, which are predominantly derived from monocytes and have mitogenic effects on the mononuclear osteoclast precursor, feed back on stromal cells to regulate their production of M-CSF, as well as IL-11, another key regulator of osteoclast formation [reviewed in 399; 400,401]. IL-6, IL-11, and other bone-resorbing factors, such as LIF and oncostatin M, transduce their regulatory signals through the gp130 signal transduction pathway. Cytokine production by human bone marrow stromal cells can be effected by age and estrogen status [402]. PTH and vitamin D stimulate osteoblast production of IL-6, as do IL-1,

TNF- and TGF-. The mechanisms by which these factors affect the generation of early osteoclast precursors from CFU-GM colonies, osteoclast differentiation, and activity have been well documented (in several reviews [403 – 406]) (see Chapter 3). Two discoveries provide insight into mechanisms for the essential role of osteoblasts in mediating osteoclastogenesis directly through cell – cell interactions (Fig. 7). One is the characterization of osteoprotegerin (OPG), also designated osteoclastogenesis inhibitory factor (OCIF), a secreted protein with strong homology to the TNF receptor family. OPG is expressed in several tissues, including bone, cartilage, kidney, and blood vessels [407,408]. Several experimental approaches established OPG as a soluble factor competent to inhibit osteoclast differentiation [409 – 411]. Expression of the gene in osteoblast lineage cells is upregulated by calcium and is downregulated by the glucocorticoid, dexamethasone [411]. These findings are consistent with the conditions required for osteoclast differentiation in cocultures of bone marrow stromal cells and spleen cells [412]. Overexpression of OPG in transgenic mice resulted in severe osteopetrosis [407]. The presence of F4/80 positive osteoclast precursors in these mice suggested that OPG inhibits terminal stages of osteoclast differentiation.

FIGURE 7 Osteoblast – osteoclast cross-talk in the regulation of bone turnover. M-CSF and OPGL/RANKL are osteoblast factors that promote osteoclast differentiation: M-CSF binds to the c-fos receptor on the mononuclear preosteoclast, whereas RANKL mediates cell – cell interaction with the RANK receptor to promote fusion of the preosteoclast to the differentiated multinucleated resorbing cell. Other regulatory hormones, cytokines, and bone matrix proteins are shown. A soluble “decoy” OPG receptor is also produced by osteoblasts that can block osteoclast precursor interactions with stromal cells.

37

CHAPTER 2 Osteoblast Biology

In search of the specific ligand for OPG, several groups cloned a novel member of the transmembrane TNF ligand family, leading to the discovery of an osteoclast differentiation factor [411,415,416]. The ligand is identical to a TNF cytokine family member designated TRANCE (TNF-related activation induced cytokine) [417], a cytokine that regulates T-cell-dependent immune responses and to RANK (receptor activation of NF-,) a TNF receptor family member cloned from T cells [418]. A standard nomenclature has been proposed for the TNF factors related to bone resorption, designating the ligand which is expressed at high levels in osteoblasts as RANKL [419]. Importantly, RANKL was demonstrated to have competency for inducing osteoclast formation from hematopoietic cells in the absence of stromal cells [415,416]. The studies define RANKL as acting directly with RANK on osteoclast precursors. The soluble decoy receptor which blocks this interaction remains designated OPG (Fig. 7). Mice with a disrupted RANKL gene completely lack osteoclasts because of the inability of osteoblasts to support their differentiation [420]. Of interest, activating mutations in RANK have been identified as the cause of the bone disorder familial expansile osteolysis [421]. Thus, RANKL is essential for the differentiation of osteoclasts from the early stages of mononuclear cell fusion, activation, and survival. Together RANK, RANKL, and M-CSF represent essential factors required for coupling stromal/osteoblastic cells to the formation of osteoclasts and are approximately controlled by cytokine and hormonal mediators of bone resorption for regulated bone turnover.

It is instructive to consider regulated expression of RANK and RANKL in osteoblasts in an environment poised for bone resorption or bone formation. Expression of the RANKL ligand in osteoblasts is upregulated by stimulator of bone resorption while OPG is downregulated ([411]; and reviewed in Hofbauer et al. [422]). Notably, regulatory elements for the osteoblast differentiation transcription factor CBFA1 occur in both the OPF and OPG-L gene promoters [423,424]. CBFA1 upregulates OPG in osteoblasts [423], thereby suppressing osteoclastogenesis when bone formation is needed. Regulation of the RANKL promoter by CBFA1 was not examined; however, CBFA1 regulates RANKL mRNA, while vitamin D decreases CBFA1 cellular levels [425,426] and increases RANKL expression. These changes are consistent with vitamin D mediating osteoclast differentiation. Thus, through transcriptional control of the CBFA1 promoter by the same steroid hormone that regulates osteoclastogenesis, as well as transcriptional regulation of OPG and OPG-L by CBFA1, feedback loops can contribute to a balance between bone resorption and bone formation (illustrated in Fig. 8). The OPG system is discussed extensively in Chapters 3, 6, 12, and 14. Crosstalk between osteoclast activity and osteoblast recruitment maintains the fidelity of bone tissue organization. Following the activation and resorption phases of the bone remodeling sequence, the recruitment, proliferation, and differentiation of osteoprogenitors and osteoblasts on the resorbed surface is accomplished in part by the resorbed bone microenvironment. Stored growth factors in the bone matrix provide a local concentration to initiate the formation phase by recruitment of osteoprogenitors to the

FIGURE 8 CBFA and vitamin D3 regulate bone turnover by coordination of gene transcription. Vitamin D promotes osteoclast differentiation in part by increasing RANKL. Modest downregulation of the mouse CBFA1 promoter by 1,25(OH)2D3 [425] may ensure physiologic levels of this factor. CBFA1 is a positive regulator of OPG expressing bone resorption while promoting bone formation. Thus, OPG (RANK) /RANKL ratios may be regulated through the combined activities of vitamin D and CBFA.

38

LIAN AND STEIN

resorbed bone surface. Here the role of TGF-1 is central to increasing osteoblast differentiation at sites of bone resorption [427].

V. OSTEOBLASTS IN VITRO: STAGES IN DEVELOPMENT OF THE OSTEOBLAST PHENOTYPE A. Cell Culture Models 1. PRIMARY CELL CULTURES Valuable insight into the biology and pathology of bone has been obtained over the past several years from cultured mammalian and avian osteoblasts that undergo developmental expression of genes and establishment of bone tissue-like organization, analogous to osteoblast differentiation in vivo. The clinical relevance of culture models should not be overlooked. Primary cell cultures offer

TABLE 2 Cell Culture Models for Studying Commitment, Growth, Differentiation, and Physiological Responses of Osteoblastic Cells Primary cell cultures for osteoblast differentiation Periosteum [441] Calvariae [432] Trabeculae [434] Cortical bone [435] Marrow stroma [437,439] Tumor-derived cell lines [477] ROS 17/2.8; ROS 24/1 UMR106 SaOS-2 U2-OS MG-63 Clonal and transformed cell lines i. Pluripotential/bipotential C3H10T1/2 [462] ROB-C26 [459] 2T3 [182] MLB13 Myc [183] ii. Monopotential MC3T3-E1 (preosteoblast) [125] C2C12 (premyoblast) [461] MLO-Y4 (osteocytic) [473] iii. Marrow stromal origin ST-2 [463] W20-17 [37]

several advantages, particularly for studying cell growth control mechanisms and differentiation in the context of a mineralizing matrix. Primary cultures often reflect the in vivo status of the osteoblast and phenotypic properties of a mouse model [253,261] (Table 2). Cultured calvarial osteoblasts from osteopetrotic rats that exhibit precocious and intensified mineralization within the intact animal during development retain this defect in vitro [428,429]. The pathology is reflected by parallel in vitro and in vivo modification in the temporal sequence of cell growth and tissue-specific gene expression as well as by features of cell organization and extracellular matrix mineral deposition that are revealed at the histochemical and ultrastructural levels. In several mouse models, marrow cultures supporting osteoblast differentiation and osteoclastogenesis have revealed cell type-specific defects and provided insight into molecular mechanisms [430]. The method of releasing populations of cells from a bone extracellular matrix by outgrowth cultures of bone fragments or by enzymatic digestion with trypsin and collagenase was developed several decades ago [431 – 435]. Calvarial or trabeculae-derived osteoblasts support the exploration of mechanisms associated with the differentiation of cells committed to the osteoblast phenotype. Marrow cell cultures permit the evaluation of gene expression related to the commitment of stem cells to the osteoblast lineage, initial stages of differentiation, and the limits of plasticity that characterize pluripotent osteoprogenitor cells [133,171,202, 436 – 440]. Osteoprogenitor cells that differentiate have also been isolated from the periosteum [441,442]. Primary cultures of normal diploid fetal rat calvarialderived osteoblasts form multilayered cellular nodules having bone tissue-like properties [143,443 – 447], whereas adherent marrow-derived cells give rise to osteoblastic colonies. Osteoblasts isolated from more mature bone as described for chick [446], bovine, and human [447] form a uniform layer of mineralized matrix. The isolated cells from bone tissue are not a heterogeneous mixture, but will become heterogeneous in long-term culture as the bone matrix is synthesized and mineralized osteoblasts result from outgrowth or released cells, particularly from fetal tissues. Although primary cell cultures are a mixed population of cells, ultrastructural, histochemical, single cell gene analyses of the cell layers and bone nodule colonies revealed that preosteoblastic, osteoblast, and osteocyte-like cells can be identified in these different models. Primary cell cultures may not always be appropriate or practical for some lines of experimentation. They are limited in their ability to maintain phenotypic properties with passaging. The cumulative observations from primary cultures of fetal and adult bone, marrow, and periosteum reported over the last decade reveal considerations for the following in interpretation of these studies. The age of isolation influences the growth properties and representations of subpopulations of bone-forming cells. The expression of

CHAPTER 2 Osteoblast Biology

osteogenic and other phenotypic responses appears to also be related to bone sites and cell passages. Thus, studies of osteoblast activities must be controlled carefully. Cells closer to the progenitor/preosteoblast stage are differentiated more readily in vitro. It should additionally be noted that protocols have been developed for the culture of human bone cells from multiple sites offering viable options for the application of culture techniques to the evaluation of skeletal diseases or for the evaluation of selective responses of osteoblasts that have been observed in vivo [435,448,449]. Although limitations must be acknowledged, the effective use of cultured human osteoblasts in assessing functional activity related to disease has been validated, e.g., with cells from patients with Paget’s disease [450], osteogenesis imperfecta [451], and other skeletal disorders [452]. Several studies have indicated that osteoblasts from osteoporotic patients and control subjects exhibit few differences when compared in vitro [453,454]. Caution must be exercised in the interpretation of osteoblast responses to agents in vitro relative to in vivo effects. For example, osteoblasts harvested from bone biopsies of osteoporotic patients treated with fluoride exhibit increased proliferative activity [455]. In contrast, fluoride is not mitogenic to osteoblasts in culture [456]. 2. TUMOR-DERIVED, TRANSFORMED, AND IMMORTALIZED CELL LINES Several classes of nonmalignant, clonal murine and rat cell lines have facilitated the investigation of biochemical and molecular parameters of osteoblast differentiation (Table 2). Calvarial-derived cells that undergo spontaneous immortalization during passage in cultures [433,457] exhibit selective, and often unstable [458], expression of bone cell phenotypic properties and responses to steroid hormones and growth factors. Pluripotent clonal cell lines are being used frequently to examine molecular mechanisms of cell phenotype commitment. The ROB-C26 [459], isolated from neonatal rat calvaria, has tripotential (osteogenic, myogenic, and adipogenic) properties. A myogenic cell line (C2C12 cells) has been well characterized with respect to its differentiation potential in response to TGF-1 and BMP-2/4 [221]. TGF-1 blocks myotube formation only, whereas BMP-2/4 induces the osteoblast phenotype. The expression of BMPs and their receptors [460] and Smad signaling factors, as well as responses to different BMPs [461], have been studied [43,408]. The pluripotential C3H10T/2 cell line, which differentiates into adipocytes, chondrocytes, and myotubes when treated with azacytidine, will also differentiate to osteoblasts in response to BMP-2 [219,462]. The stromal cell line W20-17 [37], which was subcloned (by limiting dilution from bone marrow), and other similar mouse stromal lines (ST-2 cells) [463] have bipotential properties. These cells, which are widely used to support osteoclast formation, can

39 be differentiated to mature osteoblasts by BMP-2. The MC3T3-E1 monopotential preosteoblastic cell line [464] that was obtained by subjecting calvarial-derived osteoblasts (from C57B1/6 mice) to scheduled passaging retained the ability to undergo a development sequence of gene expression [465]. These cells establish a mineralized bone extracellular matrix [466] and, like primary rat calvarial cells, differentiation can be modulated by various stimuli. In recent years, subclones from these cell lines have been isolated with distinct properties, likely reflecting stable stages of osteoblast maturation [125]. Viral transformation of calvarial-derived rodent [467], marrow stromal cells, or human osteoblasts has also provided model systems for addressing regulatory mechanisms operative in bone cells [183,467 – 469]. The introduction of a temperature-sensitive viral gene, which at permissive temperatures selectively supports proliferation or postproliferative phenotypic gene expression, offers new options in the pursuit of skeletal regulatory mechanisms in both murine [182,183] and human [470,471] cells that are involved in development and bone disease. Characterized properties reveal that viral transformation appears to restrict their properties to mono- or bipotential lineages at an early stage of commitment, e.g., mouse limb bud-derived cell lines (MLB13 Myc clones 14 and 17) [183] and human marrow stromal cells [472]. The 2T3 cell line from mouse calvaria [182] and the MLO-Y4 osteocyte-like cell line from long bone [473] have been established from transgenic mice harboring the SV40 large T antigen. Stable cell lines permit conditional and reversible expression of osteoprogenitor or osteoblast phenotypic properties that are developmentally regulated. The limitation is that these cell lines may not reproducibly support formation of a mineralized matrix. Osteosarcoma cell lines, which typically express a limited gene characteristic of bone cells in vivo, have been utilized by many investigators to support studies directed to the control of genes expressed during osteoblast proliferation and differentiation. A series of rat osteosarcomaderived cell lines (ROS) exhibit a wide range of phenotypic responses. The ROS 17/2.8 [474] exhibits most properties of mature osteoblasts, including high levels of osteocalcin [475]; the ROS 24/1 cell line lacks the vitamin D receptors [396]. Several human osteosarcoma cell lines have been widely used for the characterization of hormonal responses. These include the rat UMR-106 and human MG-63, SAOS-2, U2-OS, and TE-85 cell lines. The large quantity of cells available permits isolation and characterization of gene regulatory molecules that control transcription at the level of the gene, as well as the signaling pathways that mediate responsiveness to growth and phenotypic cues [476]. However, caution must be exercised in assuming that identical regulatory parameters are operative in normal diploid osteoblasts and osteosarcoma cells

40 [477 – 482]. The abrogation of key components of growth control and proliferation – differentiation interrelationships in transformed and tumor cells results in a consequential coexpression of cell growth and tissue-specific genes that are, in normal diploid cells, sequentially expressed in response to stringently regulated physiological signaling mechanisms.

B. Developmental Sequence of Gene Expression Characterizes Stages in the Osteoblast Differentiation Pathway Primary cell cultures and cultures of established lines that produce an organized bone-like matrix provided a basis for studies that have mapped the temporal expression of cell growth and tissue-specific genes during the progressive establishment of the osteoblast phenotype [446,477, 483 – 486]. Profiles of gene expression defined developmental stages of osteoblast maturation and allowed for investigating regulatory mechanisms that support the progression of osteoblast growth and differentiation and developmental stage-specific responses to physiological mediators of bone formation and remodeling that include growth factors and steroid hormones [486 – 497]. These characterizations have facilitated the investigation of gene regulatory signaling pathways and control mechanisms associated with development and maintenance of bone tissuelike organization. The sequential expression of cell growth and tissue-specific genes presented in Fig. 9 (see also color plate) has been mapped during progressive development of the bone cell phenotype in cultured osteoblasts and marrow stromal cells from several species and several sites [128,167,320, 483,486,498,499] by the combined application of Northern blot analysis, in situ hybridization, nuclear run-on transcription and histochemistry. Four principal developmental periods can be defined [499] by expression of the major functional bone matrix proteins, often designated “phenotypic markers”. Initially, proliferation supports expansion of the osteoblast cell population to form a multilayered cellular nodule and biosynthesis of the type I collagen bone extracellular matrix (ECM). At this time, genes requisite for the activation of proliferation (e.g., c-myc, c-fos, c-jun) and cell cycle progression (e.g., histones, cyclins) are expressed together with the expression of genes encoding growth factors (e.g., FGF, IGF-I), cell adhesion proteins (e.g., fibronectin), and others associated with the regulation of ECM biosynthesis (e.g., TGF, type I collagen). However, several of the BMPs reach peak expression in immediate postproliferative osteoblasts [500] and may function in the regulation of related BMPs supporting osteoblast growth and differentiation [501,502]. For example, BMP-2 is not only autoregulated but additionally downregulated by

LIAN AND STEIN

BMP-4 and BMP-6 in osteoblasts [503]. However, BMP-2 enhances BMP-3 and BMP-4 expression during the mineralization stage [504]. Following the initial proliferation period, a second stage of gene expression is associated with the maturation and organization of the bone ECM. Genes that contribute to rendering the extracellular matrix competent for mineralization (e.g., alkaline phosphatase) are upregulated. Collagen synthesis continues and undergoes cross-link maturation [505]. Two principal transition points are key elements of this temporal expression of genes that support the progression of differentiation (Fig. 9B, bottom). These transitions have been established experimentally and defined functionally as restriction points during osteoblast differentiation to which developmental expression of genes can proceed but cannot pass without additional cellular signaling [483]. The first transition point is at the completion of the proliferation period when genes for cell cycle and cell growth control are downregulated, and expression of genes encoding proteins for extracellular matrix maturation and organization is initiated. The absence during the proliferation period of gene expression observed in postproliferative mature osteoblasts is called “phenotype suppression” [506]. The model is supported by the binding of regulatory factors abundant in proliferating osteoblasts (e.g., oncogene-encoded [507] or helix – loop – helix proteins [84,85,508,509]) to regulatory elements in postproliferative expressed genes, which results in the suppression of transcription of the genes until later in development. The second transition is at the onset of extracellular matrix mineralization. Signals for the third developmental period involve gene expression related to the accumulation of hydroxyapatite in the ECM. Genes encoding several proteins with mineral binding proteins (e.g. osteopontin, osteocalcin and bone sialoprotein) exhibit maximal expression at this time when mineralization of the bone tissue-like organized matrix is ongoing. This profile suggests functional roles for these proteins in regulation of the ordered deposition of hydroxyapatite. A fourth developmental period follows in mature mineralized cultures during which time collagenase is elevated, apoptotic activity occurs, and compensatory proliferative activity is evident [510,511]. Although the biological significance of gene expression during the fourth developmental stage remains to be formally established, it appears to serve an editing/remodeling function for modifications in the bone extracellular matrix that sustain the structural and functional properties of the tissue. This working model of osteoblast differentiation has been supported by studies that demonstrate the stage-specific expression of family members of numerous classes of proteins that regulate development of the bone cell phenotype. These include, for example, the IGF family of binding proteins [512], cyclin and cyclin-dependent kinases [249], bone morphogenic proteins [492,500], oncogene-encoded proteins

CHAPTER 2 Osteoblast Biology

41

FIGURE 9 Stages of maturation in primary cultures of bone-derived cells. (A) Morphology and histochemical staining of fetal rat calvaria-derived osteoblasts at three stages: proliferation (P) with toluidine blue, matrix maturation (MM) with alkaline phosphatase staining, and mineralization (M) detection by the von Kossa stain. Multilayers of osteoblasts form a typical bone nodule having a mineralized extracellular matrix. (See also color plate.) (B) Northern blot analyses of marker genes expressed maximally at each stage. (See also color plate.) (C) Schematic illustration of several genes temporally expressed during 35 days of culture. Increases in mRNA transcripts of OC and OP during the culture period parallel calcium (Ca2 ) deposition. The increase and peak levels of collagenase are related to remodeling or editing of the extracellular matrix. Induction of genes reflecting the apoptosis of cells associated with the mineralized nodules is shown [365,483,494]. (D) The reciprocal and functionally coupled relationship between cell growth and differentiation-related gene expression is illustrated by arrows. Broken vertical lines between the three principal periods indicate experimentally established transition or restriction points requiring the downregulation of cell growth and ECM signaling events for the progression of differentiation.

42 [507], heat shock proteins [513], homeodomain proteins [514], and TGF- receptors [515]. Notably, several genes associated with skeletal cells, osteonectin [305,485] and matrix Gla protein [516], are expressed constitutively. The sequential and stringently regulated expression of genes that defines periods of osteoblast phenotype development is illustrated schematically as a reciprocal and functionally coupled relationship between proliferation and differentiation (Fig. 9B, bottom). The extent to which developmental expression of genes is sequential and mutually exclusive rather than concomitant is in part dependent on the position in the osteoblast lineage. For example, while in calvarial-derived osteoblasts, alkaline phosphatase expression is only observed in postproliferative cells, and alkaline phosphatase is compatible with proliferation in earlier stage bone marrow-derived osteoprogenitor cells. The interrelationship between proliferation and differentiation often times provides explanations for discordance between in vivo and in vitro effects. FGF, for example, is a potent mitogen for mesenchymal cells, chondrocytes, and osteoblasts and stimulates endosteal bone formation [212]. FGF-2 is useful for fracture repair in vivo [517]. However, FGF effects in vitro are stage specific. If added to a proliferation stage, mineralization and subsequent maturation of osteoblasts are inhibited [260,263,518]. In part, the results are mediated by continued proliferation modifying the differentiated phenotype and induced levels of collagenase modifying the ECM in a manner incompatible with ECM mineralization. In postproliferative cultures, FGF-2 promotes differentiation [263].

C. Differential Responsiveness of Hormones and Growth Factors as a Function of Stage of Osteoblast Maturation Based on the in vitro models of osteoblast differentiation, we can better understand the properties and physiologic responses of the cells at their individual stages of differentiation. This is best exemplified by selective responses in bone marrow stromal cell cultures and calvarial-derived osteoblasts to growth factors [263,491,518] and hormones [487,519 – 525]. The parathyroid hormone will promote the differentiation of preosteoblasts but suppress late stages of maturation [526,527]. Although caution should be exercised in translation from in vitro to in vivo effects of PTH on bone formation, these studies indicate that even pulsed PTH administration may increase bone formation by stimulating the proliferation of progenitors, and not by anabolic effects of differentiated osteoblasts [528]. It is established that TGF- stimulates the replication of progenitor cells and directly stimulates collagen synthesis. When proliferating calvarial osteoblasts are exposed to TGF-, a block in differentiation is observed

LIAN AND STEIN

[490 – 492,529]. The mitogenic effects of TGF- are not apparent on mature postproliferative osteoblasts. The insulin-like growth factors [284] and fibroblast growth factors (FGF-1, FGF-2) [530], which have distinct roles in development [11], also have selective effects on osteoblast subpopulation in vitro. The various steroid hormones, including glucocorticoids [519,531], 1,25(OH)2D3 [487,489,520], and estrogen [521], also have selective effects, either promoting differentiation of the cells at early stages of maturation or inhibiting anabolic activities and promoting resorptive properties of the osteoblast at later stages. In general, growth factors and steroid hormones have the most robust responses in immature osteoblasts and can radically modify their program of differentiation when added to proliferating cells. Glucocorticoids regulate numerous osteoblast-expressed genes, which contribute to bone formation, including cytokines, growth factors, and bone matrix proteins [reviewed in 193,531,532] (see Chapter 44). The molecular mechanisms by which glucocorticoids exert selective effects on a particular gene are complex, but numerous examples have been documented. Increases in alkaline phosphatase, osteocalcin, and collagen are observed at early maturation stages, but inhibition of these genes occurs in differentiated osteoblasts [519,533,534]. Both transcriptional and posttranscriptional mRNA stability contribute to positive and negative regulation of a gene, as shown for osteocalcin [519] and collagen [535]. Glucocorticoids promote osteoblast colony formation in human and rat marrow-derived cells and accelerate osteoblast differentiation in proliferating calvarial derived cells, reflected by both increased numbers and size of the bone nodules and early mineralization [439,493,494]. Because postproliferative cultures cannot be stimulated to produce more mineralizing nodules, the mature osteoblast is refractory to growth stimulation by dexamethasone [129,493,494,536]. This may be a consequence of glucocorticoids pushing cells to terminal differentiation or apoptosis, thereby depleting the pool of preosteoblasts capable of nodule formation [129,371,493]. It is noteworthy that dexamethasone exerts an antiproliferative effect on mouse osteoblasts [525,537] and blocks their maturation. These findings, together with glucocorticoid effects on osteoclast activity (reviewed in [193], contribute to glucocorticoid-induced osteopenia observed in vivo when pharmacologic doses of glucocorticoids are required [290,531, 538 – 541] (see Chapter 44). The active metabolite of vitamin D, 1,25(OH)2D3, has complex effects on the skeletal system related to targeting of many cell types, dose, and timing [542 – 544] (see Chapter 9). Vitamin D is a biphasic regulator of osteoblast activity for bone formation and bone resorption. Vitamin D regulates the expression of genes in osteoblasts that form the bone ECM or provide signals for osteoclast differentiation. The up-

CHAPTER 2 Osteoblast Biology

regulation by 1,25(OH)2D3 of numerous osteoblast parameters related to bone matrix formation and mineralization (e.g., collagen, alkaline phosphatase, osteopontin, osteocalcin, matrix Gla protein), and bone resorption by cytokines (e.g., osteoprotegerin [545]), reflects influences of the hormone on osteoblast function and regulation of bone turnover. However, pharmacological doses and long-term exposure of this hormone to rats can result in abnormalities of bone formation [544,546,547]. In vitro analysis of 1,25(OH)2D3 in primary cultures of rat osteoblasts show stage-dependent effects. The steroid is antiproliferative in the growth period and can block formation of the mineralized nodule when introduced during the growth period [487 –489,524,548] or stimulate differentiation-related gene expression in mature osteoblasts. Because of these properties, acute versus continuous exposure of cells to 1,25(OH)2D3 can lead to opposing results [487,489]. The selective effects based on the stage of osteoblast maturation are evident by in situ hybridization studies, in vivo and in vitro, and accompanying changes in cell morphology in vitro [136,137,143,536,549]. These changes in morphology and gene expression may relate to bone-resorbing effects of 1,25(OH)2D3 on surface-lining osteoblasts, which must (a) retract to allow for osteoclast interaction with the bone mineral and (b) provide local factors for the induction of osteoclast activity. Studies from many groups using different osteoblast models have reinforced two important concepts: (1) that the stage of osteoblast maturation influences the selective responsiveness of specific genes to hormones or growth factors and (2) that there is a window of responsiveness of a cell during which the factor can alter development and maintenance of the bone cell phenotype. These analyses have provided clinically relevant information toward an understanding of the consequences incurred by the osteoblast when exposed to therapeutic agents that may stimulate or inhibit cell proliferation or differentiation.

VI. MOLECULAR MECHANISMS MEDIATING PROGRESSION OF OSTEOBLAST DIFFERENTIATION A. Stage-Specific Expression of Transcription Factors and Their Contribution to OsteoblastSpecific Gene Expression The progression of osteoblast differentiation requires the sequential activation and suppression of genes that encode phenotypic and regulatory proteins. Key molecular events initiate from the extracellular matrix and signaling molecules, such as BMPs and TGF-s, that can indirectly result in a cascade of gene expression (see earlier discussion). In addition, regulatory factors that directly engage in protein – DNA, as well as protein – protein interactions, are

43 important mechanisms for both the activation and the suppression of genes reflecting stages of osteoblast maturation [550]. Transcription factors described in Section I, which contribute to position and pattern formation in the developing embryo, provide mechanisms for regulating the progression of osteoblast differentiation in the adult. The selective representation of these factors during osteoblast differentiation and family members within a class of transcription factors (Fig. 2), as well as evidence for their functional consequences (e.g., by forced expression, antisense or antibody blocking studies) on osteoblasts, provides compelling evidence for their regulatory effects in driving osteoblast maturation. The focus on transcription factors that regulated the developmental expression of osteocalcin [reviewed in 551,552] has been further supported by the characterization of mice carrying null mutations of several transcription factors (Table 1). Homeodomain proteins, Fos/Jun family members, helix – loop – helix factors, and RUNX2/CBFA1 proteins are among the well-characterized transcription factors (see Fig. 2). The homeodomain protein-binding sites contribute to bone-specific expression of several genes, collagen type I [553,554], osteocalcin [514,555], and, most recently, regulated expression of bone sialoprotein by Dlx5 [556]. During osteoblast differentiation in vitro, Msx-2 is expressed maximally in the preosteoblast and is subsequently downregulated [514] with the onset of osteocalcin expression. In situ hybridization studies confirm the reciprocal expression of osteocalcin and Msx-2 in cells of developing bone [557]. In a reciprocal fashion, Dlx-5 appears in the postproliferative osteoblast and increases during mineralization [558]. It appears that there is selectivity for the expression of homeodomain factors during osteoblast differentiation as well as developmental variations in activities [557]. Consistent with the developmental expression of Msx-2 in early osteoprogenitors and Dlx-5 in mature osteoblasts, Msx-2 downregulates OC, whereas Dlx-5 can positively regulate OC and collagen as well. Protein – protein interactions between Msx-2 and Dlx-5 are also determinants for developmental activities of these factors, as demonstrated by Zhang et al. [559]. The molecular mechanisms of this heterodimer interaction in alleviating suppression of the osteocalcin gene were demonstrated [555]. These activities may in part reflect the sequence content of homeodomain responsive promoter elements in genes that are expressed developmentally in osteoblasts [514,553,558,560,561]. Helix – loop – helix transcription factors are expressed at high levels in proliferating osteoblasts, thereby facilitating downregulation of the gene in these immature cells [85,509]. This complexity ensures developmental, tissuespecific regulated expression of the postproliferative bonespecific genes in osteoblasts. The fos and jun family members also exhibit developmental stage-specific expression and activities during osteoblast

44

LIAN AND STEIN

differentiation in vivo [562] and in vitro [507]. Studies of c-fos expression in transgenic mouse models reveal the importance of c-fos in establishing the osteoblast phenotype [563]. In vivo immunohistochemical staining reveals that c-fos is expressed in osteoprogenitor cells, in the perichondrium and periosteal tissues, but not in mature osteoblasts [562]. During osteoblast differentiation in vitro, c-fos and jun expression is expressed maximally in proliferating preosteoblasts. However, retention and upregulation of fra2 and jun-D protein levels and mRNA expression were observed postproliferatively in differentiated osteoblasts [564]. Confirmation of the functional activities of these factors in regulating progressive maturation of the osteoblast phenotype has been demonstrated by several experimental approaches, including specific repressor and enhancer activities of c-fos/ c-jun and fra2/jun-D, respectively, on the osteocalcin promoter, as well as inhibition of osteoblast differentiation resulting from antisense inhibition of fra2 expression [564]. RUNX2/CBFA1, described in Section I as an essential transcription factor for osteogenesis, regulates osteoblast differentiation. It is noteworthy that CBFA1 is present in the osteoprogenitors where osteoblast-specific genes are not expressed. CBFA1 mRNA, cellular protein levels, and DNA-binding activity are increased during osteoblast differentiation [103] and may contribute to specific CBFA1 activities in cells at distinct stages of osteoblast differentiation. In consideration of the precise functional activities of CBFA1 in the regulation of osteoblastic genes, multiple molecular mechanisms are operative. Oftentimes expression of a gene can be regulated during development (embryonic versus postnatal) or in relation to cell-type specificity by the production of protein isoforms resulting from either alternative usage of ATG initiation sites or splicing variant. Several examples of alternatively spliced products imparting tissue specificity have been recorded in genes. The Fibronectin gene gives rise to multiple splice variants, and the splicing process is cell type, developmental stage, and age regulated [565,566]. In 2(XI) collagen, alternative

FIGURE 10

splicing gives rise to proteins that are expressed in a tissuespecific manner. Isoforms lacking exons 6 – 8 of the 2(XI) collagen gene code for a protein that is found in cartilaginous tissue of developing limbs and axial skeleton, whereas the transcript containing exons 6 – 8 is found in nonchondrogenic tissue such as calvaria and periosteum [567]. These mechanisms may contribute to the timing of expression or the protein isoforms may exhibit specialized functions. Several NH2 and COOH-terminal CBFA1 isoforms have been reported (Fig. 10) [93,568,569]. Both the aminoterminal isoforms, type I (MRIPV) and type II (MASNS) isoforms, have nearly equivalent transactivation activity on different promoters and in various cell lines [118, 570 – 572]. To date, no studies have systematically addressed the relative expression of these CBFA1 isoforms and other isoforms resulting from C-terminal splicing events. Noteworthy in the regulation of CBFA1 activities are posttranslational modifications, including phosphorylation by MAPK [573]. In addition, CBFA1 interacts with several coregulatory proteins, both activators and suppressors of CBFA1 activity [95,96,574,575,575,576], which also appear to be developmentally expressed during osteoblast differentiation [574]. CBFA1 also has properties that can mediate modifications in chromatin, an important component of tissue-specific transcriptional control that is described in the following section.

B. Contributions of Nuclear Architecture to Transcriptional Control during Osteoblast Differentiation It is becoming increasingly apparent that nuclear architecture provides a basis for support of the stringently regulated modulation of cell growth and tissue-specific transcription necessary for the onset and progression of osteoblast differentiation. Here, multiple lines of evidence point to contributions by three levels of nuclear organization to in

CBFA1 NH2 terminal isoforms present in osteoblast lineage cells. Schematic illustration of the origin of CBFA1 regulated by an upstream (PU) promoter is transcribed at MET-1 in exon 1 (MASNS isoform or type II designation by Harada et al. [118]). The type I isoform (MRIPV), regulated by the downstream promoter (PD), was the first CBFA1 protein to be cloned from T cells [94], which is also expressed in mesenchymal lineage cell lines, chondrocytes, and osteoblasts [114 – 118]. The MASNS isoform was initially identified by Stewart et al. [96a] and originates at MET 69 regulated by the same downstream promoter (PU) from which OSF-2 was first characterized [102]. OSF-2 is poorly transcribed from the MET-1, which does not reside in a Kozak sequence [96b].

CHAPTER 2 Osteoblast Biology

vivo transcriptional control where structural parameters of the genome are functionally coupled to regulatory events. The primary level of gene organization establishes a linear ordering of promoter regulatory elements. This representation of regulatory sequences reflects competency for the responsiveness to physiological regulatory signals as discussed earlier. The osteocalcin gene provides a paradigm for the involvement of nuclear organization in transcriptional control that is linked to bone formation, homeostatic regulation, and bone remodeling. The regulatory elements of the bone-specific osteocalcin gene are organized in a manner that supports responsiveness to physiologic mediators and developmental expression in relation to bone cell differentiation. Characterized regulatory elements and cognate transcription factors can support osteocalcin suppression in nonosseous cells, immature osteoblasts, and growthstimulated osteoprogenitors/osteoblasts, transcriptional activation in postproliferative cells, and steroid hormone enhancement (Fig. 11). For example, the OC box (99 to 76 bp) is a multipartite element that binds several classes of transcription factors. The OC box binds homeodomain proteins (e.g., Msx-2, Dlx-5 [514,577]) and a bone tissuespecific complex of unknown origin, designated the OC box-binding protein (OCBP), which contributes to the activation of gene transcription [578,579]. The minimal OC promoter containing the OC box is sufficient to support osteoblast-specific expression in vitro. A bipartite element in the proximal promoter confers FGF-2 and cAMP responsiveness [580]. This element regulates the suppression of osteocalcin synthesis in response to numerous modulators of cell growth, including IGF-1, bFGF, cAMP, and PTH [581]. A series of AP-1 sites occur in the OC gene promoter [506,582,583], including one that is the TGRE [584]. Fos (c-fos, fral, fra2) and jun (c-jun, jun-D, jun-B) oncogeneencoded families of transcription factors form homo or heterodimeric complexes that regulate gene expression at AP-1 motifs. These AP-1 sites provide another example of regulatory sequences that contribute to positive and negative regulation dependent on the biological circumstances. The c-fos/c-jun heterodimer represses, whereas the fra2/jun-D heterodimer activates transcription. The latter is more abundant in postproliferative osteoblasts. The vitamin D responsive element (VDRE) functions as an enhancer [585 – 588], binding the transcriptionally active VDR/RXR heterodimer complex (see Chapter 9). The core motif, two steroid half elements with a three nucleotide space, is highly conserved [589 – 591]. However, subtle variations, both within the core domain and in the flanking sequences, render VDRE promoter elements of various genes selectively ligand responsive in a developmental and tissue-specific manner. Specificity of VDRE utilization is further conferred by protein – DNA and/or protein – protein interactions in addition to the VDR/RXR complexes

45 [479,591 – 595]. The osteocalcin VDRE transcription factor complex appears to be a target for modifications in vitamin D-mediated transcription by other physiologic factors, including glucocorticoids [592], TNF-a [596], and retinoic acid [597 – 601]. It is evident from available findings that the linear organization of gene regulatory sequences is necessary but insufficient to accommodate the requirements for physiological responsiveness to homeostatic, developmental, and tissue-related regulatory signals. Parameters of chromatin structure and nucleosome organization are a second level of genome architecture. There is a requirement to render promoter regulatory elements competent for protein – DNA and protein – protein interactions that mediate positive and negative controls. Additionally, activities of regulatory complexes at the proximal and distal promoter must be integrated. Modifications in chromatin reduce the distance between promoter elements, thereby supporting interactions between the modular components of transcriptional control. Each nucleosome (approximately 140 nucleotide base pairs wound around a core complex of two each of H3, H4, H2, and H2B histone proteins) contracts linear spacing by sevenfold. Folding of nucleosome arrays into solenoid-type structures provides a potential for interactions that support synergism between promoter elements and responsiveness to multiple signaling pathways. The molecular mechanisms that mediate chromatin remodeling are being defined [reviewed in 602 – 605]. A family of proteins comprising multimeric protein complexes has been described in yeast (SWI/SNF complex) and in mammalian cells that promote transcription by altering chromatin structure [604 – 608]. These alterations render DNA sequences containing regulatory elements accessible for binding cognate transcription factors and mediate protein – protein interactions that influence the structural and functional properties of chromatin. Multiple lines of evidence suggest that the remodeling of nucleosomal structure involves alterations in histone – DNA and/or histone – histone interactions. Histone acetylation and phosphorylation posttranslational modifications have been functionally linked with changes in nucleosomal structure that alter the accessibility to specific regulatory elements [605]. Core histone hyperacetylation enhances the binding of most transcription factors to nucleosomes [609 – 611]. Alterations in the chromatin organization of the osteocalcin (OC) gene promoter during osteoblast differentiation provide a paradigm for remodeling chromatin structure and nucleosome organization that is linked to a long-term commitment to phenotype-specific gene expression [612 – 615]. In nonosseous cells, the packing of chromatin contributes to the extent that promoter elements are accessible to transcriptional activation complexes. An array of nucleosomes on the OC promoter in nonosseous cells contributes to maintaining the suppression of gene transcription. Figure 11 schematically depicts modifications in chromatin structure and

46

LIAN AND STEIN

FIGURE 11

Developmental remodeling of nucleosome organization in the osteocalcin gene promoter correlates with transcriptional activity during the differentiation of normal diploid osteoblasts. The positioning of nucleosomes in the osteocalcin gene promoter was determined by combining DNase I, micrococcal nuclease, and restriction endonuclease digestions with indirect end labeling [272]. Filled circles represent the placement of nucleosomes with the shadows indicating movement within nuclease-protected segments. Note the differences in nucleosome organization of the gene between nonosseous cells when the OC gene is not transcribed (line 1) and in osteoblasts expressing osteocalcin (lines 2 and 3). Vertical arrows correspond to the limits of the distal (600 to 400) and proximal (170 to 70) sites of DNase I hypersensitivity (DHS), which are observed in osteoblasts when the OC is expressed. DHS is increased in response to vitamin D (represented by larger arrows). Positions of CBFA1 elements (sites A, B, C), the VDRE (465 to 437), the osteocalcin box (OC box 99 to 77), and TATA element (31 to 28) are designated.

nucleosome organization that parallel competency for transcription and the extent to which the osteocalcin gene is activated and transcribed in bone cells. Basal expression and enhancement of osteocalcin gene transcription are accompanied by two changes in the structural properties of chromatin when the gene is activated in a bone cell. When the gene is activated in osteoblasts, there is a rearrangement in nucleosome placement [613]. A single nucleosome becomes positioned between proximal regulatory elements, and distal steroid hormone-dependent regulatory sequences provide a basis for accessibility of transactivation factors to cognate sequences (Fig. 11). This model is derived from experimen-

tal evidence. DNase I hypersensitivity is detected in two regulatory domains of the promoter encompassing the VDRE and CBFA sequences in the proximal and distal promoter [612 – 614]. Vitamin D treatment enhances DNase I hypersensitivity. These domains contribute to tissue-specific and vitamin D enhancer activity. A third level of nuclear architecture that contributes to transcriptional control is provided by the nuclear matrix [616]. The anastomosing network of fibers and filaments that constitute the nuclear matrix supports the structural properties of the nucleus as a cellular organelle and accommodates structural modifications associated with

47

CHAPTER 2 Osteoblast Biology

proliferation, differentiation, and changes necessary to sustain phenotypic requirements of specialized cells. As the intricacies of gene organization and regulation are elucidated, the implications of a fundamental biological paradox become strikingly evident. With a limited representation of gene-specific regulatory elements and a low abundance of cognate transactivation factors, how can sequence-specific interactions occur to support a threshold for the initiation of transcription within nuclei of intact cells? Viewed from a quantitative perspective, the in vivo regulatory challenge is to account for the formation of functional transcription initiation complexes with a nuclear concentration of regulatory sequences that is approximately 20 nucleotides per 2.5 yards of DNA and a similarly restricted level of DNAbinding proteins. Regulatory functions of the nuclear matrix include, but are by no means restricted to, DNA replication [617], gene location [618,619], imposition of physical constraints on the chromatin structure that support the formation of loop domains, concentration, and targeting of transcription factors [620 – 623], RNA processing and transport of gene transcripts [624 – 628], and posttranslational modifications of chromosomal proteins, as well as imprinting and modifications of chromatin structure [629]. Chromatin loop domains (10 – 100 kb) are tethered to components of the nuclear matrix through MAR (matrix attachment regions) sequences.

The initial indication of linkage between the nuclear matrix and control of osteoblast differentiation was first provided by the observation of a parallel representation of nuclear matrix proteins and developmental stage-specific gene expression [630]. Direct involvement of the nuclear matrix in the control of bone-specific gene transcription is provided by several lines of evidence. Two nuclear matrix DNA-binding proteins regulate the activation of osteocalcin gene transcription, identified as NMP1. YY1 modulates vitamin D enhancer activity and NMP2 characterized as RUNX/CBFA1 is a transcriptional activator protein and contributes to promoter organization [99]. The mechanism by which transcription factors contribute to overall promoter architecture is becoming clear. Transcription factors and chromatin-remodeling factors associated with the nuclear matrix, such as CBFA factors [631] and the YY1 transcription factor [632], which regulate osteocalcin gene activity, can contribute to conformational modifications in the promoter structure. Through protein – DNA interactions of the CBFA element with nuclear matrix-associated CBFA factors, local chromatin changes are induced by coactivator/corepressor proteins associated with the DNAbinding complex, and the OC promoter becomes poised for transcription. Figure 12 extends the model presented in Fig. 11 by illustrating the three-dimensional organization of the promoter that is permissive for the binding of

FIGURE 12 Role of the nuclear matrix-associated CBFA1 factor in organizing promoter architecture. A three-dimensional representation of the osteocalcin promoter is shown. The interaction of OC CBFA recognition motifs with CBFA1 provides a conformation stability mediated by the tight association of CBFA1 with the nuclear matrix scaffold. This conformation, together with the positioned nucleosome, facilitates the integration of regulatory signals between the proximal (TATA) and the distal (VDRE) elements necessary for basal and vitamin D enhancer activity. Mutations of three CBFA elements in the rat OC gene resulting in loss of DNase I hypersensitivity and vitamin D responsiveness support this model [647].

48

LIAN AND STEIN

FIGURE 13

CBFA-interacting proteins regulating CBFA transactivation functions. Coregulator proteins interact directly with the indicated domains of CBFA transcription factors. Structurally and functionally homologous segments include the conserved DNA-binding runt homology domain, transcriptional activation, and suppression domains [643,711,712], as well as subcellular targeting signals [631,713]. The promoter organizing functions of CBFA factors may involve interacting proteins, including CBF [714], necessary for DNA binding of CBFA to regulatory elements, coactivators ALY [633] and p300, the latter having histone acetylase (HAT) activity [634], and Groucho/TLE [574,575,639], a negative regulator of CBFA enhancer activity. In addition, other transcription factors [635], Smad signal transduction proteins [123], and steroid receptors [637] contribute to synergistic gene expression with CBFA1 (shown in Fig. 14) [reviewed in 94,95].

transactivation factors and protein – protein interactions between distal and proximal elements (e.g., TFIIB TATA box-binding protein and the VDR/RXR). We postulate that this promoter architecture is mediated and stabilized by transcription factors that are targeted to subnuclear domains, such as CBFA1. Among the numerous CBFA-interacting proteins that have been identified (Fig. 13), several have enzyme activities for chromatin remodeling (Fig. 14). Coactivators include ALY [633] and p300, which has histone acetylase (HAT) activity [634], as well as interacting transcriptional regulators [635], ear-2 [636], signal transduction proteins [123], and steroid receptors [637,638]. Corepressor modulators include HES [575] and the negative regulatory factor Groucho/TLE [576,639], which is also associated with the nuclear matrix [640]. This protein has histone deacetylase activity and a histone H3-interacting domain [641,642]. We have shown that the highest levels of Groucho at the mRNA and cellular protein levels are observed in early stage osteoblasts [574]. Groucho/TLE is also abundant in muscle tissue and may contribute to the suppression of osteocalcin in nonosseous cells and proliferating osteoblasts. As indicated in Fig. 13 (bottom), cooperative interactions between CBFA and other transcription factors, through as yet undefined mechanisms, have been reported. The synergistic transactivation of gene transcription between CBFA

and the Ets factors is well documented [95,96], as well as CBFA with C/EBP factors [643] and Smad factors. These interactions provide options for positive and negative regulation of a spectrum of CBFA-regulated genes within the cell or related to a cellular phenotype. Of interest, osteoblast-specific synergistic interactions between an estrogen response element and CBFA elements have been reported [638]. These respective activities contribute to the enhancer and suppressor transcriptional properties of CBFA factors as illustrated in Fig. 14 [574,575,644]. Several reports have indicted that not all osteoblast-expressed genes containing CBFA regulatory sequences are activated by Cbfa1, some are repressed; e.g., BSP [645] and the CBFA1 gene are autoregulated by CBFA1 [646]. The significance of the localization of these transcription factors in specific subnuclear domains, as well as their ability to interact with chromatin modifying proteins, is related to tissue-specific, as well as steroid hormone-dependent gene control of transcription. Colocalization of these regulatory components that are functionally linked to activation and suppression have been established in situ [574]. Validation of these models is further provided by studies in which the CBFA sites in the rat OC promoter were mutated. Three CBFA motifs are strategically positioned in the bone-specific rat osteocalcin promoter. Sites A and B flank the vitamin D response element in the distal

49

CHAPTER 2 Osteoblast Biology

FIGURE 14

Model of CBFA transcriptional regulation of gene activation and suppression mediated by modifications in chromatin architecture. CBFA factors interact with histone acetylases (HATs) (e.g., p300) to open chromatin domains for the binding of regulatory factors that lead to transcriptional activation. CBFA interactions with coregulators having associated histone deacetylase activity (HDAC) result in the suppression of transcription.

promoter, and sites B and C define the orders of a positioned nucleosome in the proximal promoter. The functional significance of multiple CBFA elements in contributing to promoter structure was addressed by mutating individual or multiple AML sites within the context of the native rat ( 1.1 kb) osteocalcin promoterCAT [647]. All three sites are required for maximal basal expression of the rOC promoter. Strikingly, mutation of the three CBFA1 sites leads to abrogation of responsiveness to vitamin D, glucocorticoids, and TGF-. The mechanism of this loss of promoter responsiveness was related to an absence of DNase I hypersensitive sites at the vitamin D response element and over the proximal tissue-specific basal promoter. These findings strongly support a multifunctional role for AML factors in regulating gene expression, not only as a simple transactivator, but also by facilitating modification in promoter architecture and chromatin organization. It is already apparent that local nuclear environments generated by the multiple aspects of nuclear structure are intimately tied to the developmental expression of cell growth and tissue-specific genes. Membrane-mediated initiation of signaling pathways that ultimately influence transcription has been recognized for some time. Here, the mechanisms that sense, amplify, dampen, and/or integrate regulatory signals involve structural as well as functional

components of cellular membranes. Extending the structure – regulation paradigm to nuclear architecture expands the cellular context in which cell structure – gene expression interrelationships are operative. The nuclear structure is a primary determinant of transcriptional control. Thus, the power of addressing gene expression within the threedimensional context of nuclear structure would be difficult to overestimate.

VII. CONCLUDING REMARKS This chapter has presented the cell biology of osteoblasts within the content of our current understanding of the regulatory controls operative in promoting osteoblast differentiation. We have attempted to address how physiologic parameters of gene expression are integrated to support the requirements of bone development and functional integrity of the tissue. During osteoblast phenotype development and bone formation, stages of maturation are defined by levels of expression of subsets of osteoblast genes. A cohort of tissue-specific, developmental, steroid hormone and growth factor-related transcription factor complexes impinge on gene transcription, providing a complex and integrated series of regulatory signals for the selective activation and repression of genes related to osteoblast activity. Additionally,

50

LIAN AND STEIN

the effects of a hormone or growth factor on expression of a specific gene are related to the phenotype as well as the stage of cellular maturation because of the different representation of proteins that contribute to gene regulation. Thus clinical consideration for treatment and therapeutic regimens can be approached with greater knowledge of the consequential effects at the level of gene-regulating responses. We have presented a growing body of evidence for cross-talk at all levels of bone biology, including among the subpopulations of bone cells, between the extracellular matrix and intracellular signaling factors, and finally at the DNA level between transcription factor complexes at multiple elements. Such interactions and complexities need to be considered in future applications of therapeutic strategies.

Acknowledgments The authors gratefully appreciate preparation of the manuscript by Judy Rask and thank colleagues Janet Stein and André van Wijnen for helpful discussions and members of our research group: Chaitali Banerjee, Amjad Javed, Hicham Drissi, Kaleem Zaidi, and Eva Balint. The National Institutes of Health grants supporting the research program related to this chapter include AR45688, AR45689, AR39588, and DE12528. The contents of this chapter are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.

References 1. S. Suwanwalaikorn, M. van Auken, M. I. Kang, S. Alex, L. E. Braverman, and D. T. Baran, Site selectivity of osteoblast gene expression response to thyroid hormone localized by in situ hybridization. Am. J. Physiol. 272, E212 – E217 (1997). 2. C. Kasperk, J. Wergedal, D. Strong, J. Farley, K. Wangerin, H. Gropp, R. Ziegler, and D. J. Baylink, Human bone cell phenotypes differ depending on their skeletal site of origin. J. Clin. Endocrinol. Metab. 80, 2511 – 2517 (1995). 3. A. P. Davis, D. P. Witte, H. M. Hsieh-Li, S. S. Potter, and M. R. Capecchi, Absence of radius and ulna in mice lacking hoxa-11 and hoxd-11. Nature 375, 791 – 795 (1995). 4. T. Lufkin, M. Mark, C. P. Hart, P. Dolle, M. LeMeur, and P. Chambon, Homeotic transformation of the occipital bones of the skull by ectopic expression of a homeobox gene. Nature 359, 835 – 841 (1992). 5. E. Laufer, C. E. Nelson, R. L. Johnson, B. A. Morgan, and C. Tabin, Sonic hedgehog and Fgf-4 act through a signaling cascade and feedback loop to integrate growth and patterning of the developing limb bud. Cell 79, 993 – 1003 (1994). 6. F. Rousseau, J. Bonaventure, L. Legeai-Mallet, A. Pelet, J. M. Rozet, P. Maroteaux, M. Le Merrer, and A. Munnich, Mutations in the gene encoding fibroblast growth factor receptor-3 in achondroplasia. Nature 371, 252 – 254 (1994). 7. B. K. Hall and T. Miyake, All for one and one for all: Condensations and the initiation of skeletal development. BioEssays 22, 138–147 (2000). 8. J. Jernvall and I. Thesleff, Reiterative signaling and patterning during mammalian tooth morphogenesis. Mech. Dev. 92, 19 – 29 (2000). 9. B. K. Hall, The induction of neural crest-derived cartilage and bone by embryonic epithelia: An analysis of the mode of action of an epithelial-mesenchymal interaction. J. Embryol. Exp. Morphol. 64, 305 – 320 (1981).

10. R. L. Johnson, R. D. Riddle, and C. J. Tabin, Mechanisms of limb patterning. Curr. Opin. Genet. Dev. 4, 535 – 542 (1994). 11. S. Iseki, A. O. Wilkie, and G. M. Morriss-Kay, Fgfr1 and Fgfr2 have distinct differentiation- and proliferation-related roles in the developing mouse skull vault. Development 126, 5611 – 5620 (1999). 12. B. L. Hogan, Morphogenesis. Cell 96, 225 – 233 (1999). 13. A. D. Weston, V. Rosen, R. A. Chandraratna, and T. M. Underhill, Regulation of skeletal progenitor differentiation by the BMP and retinoid signaling pathways. J. Cell Biol. 148, 679 – 690 (2000). 14. E. Koyama, E. B. Golden, T. Kirsch, S. L. Adams, R. A. Chandraratna, J. J. Michaille, and M. Pacifici, Retinoid signaling is required for chondrocyte maturation and endochondral bone formation during limb skeletogenesis. Dev. Biol. 208, 375 – 391 (1999). 15. L. Niswander, C. Tickle, A. Vogel, I. Booth, and G. R. Martin, FGF4 replaces the apical ectodermal ridge and directs outgrowth and patterning of the limb. Cell 75, 579 – 587 (1993). 16. N. Wanek, D. M. Gardiner, K. Muneoka, and S. V. Bryant, Conversion by retinoic acid of anterior cells into ZPA cells in the chick wing bud. Nature 350, 81 – 83 (1991). 17. C. Thaller and G. Eichele, Isolation of 3,4-didehydroretinoic acid, a novel morphogenetic signal in the chick wing bud. Nature 345, 815 – 819 (1990). 18. H. J. Kim, D. P. Rice, P. J. Kettunen, and I. Thesleff, FGF-, BMPand Shh-mediated signalling pathways in the regulation of cranial suture morphogenesis and calvarial bone development. Development 125, 1241 – 1251 (1998). 19. C. Hartmann and C. J. Tabin, Dual roles of wnt signaling during chondrogenesis in the chicken limb. Development 127, 3141 – 3159 (2000). 20. A. B. Zimrin, M. S. Pepper, G. A. McMahon, F. Nguyen, R. Montesano, and T. Maciag, An antisense oligonucleotide to the notch ligand jagged enhances fibroblast growth factor-induced angiogenesis in vitro. J. Biol. Chem. 271, 32499 – 32502 (1996). 21. S. Artavanis-Tsakonas, K. Matsuno, and M. E. Fortini, Notch signaling. Science 268, 225 – 232 (1995). 22. D. P. Rice, T. Aberg, Y. Chan, Z. Tang, P. J. Kettunen, L. Pakarinen, R. E. Maxson, and I. Thesleff, Integration of FGF and TWIST in calvarial bone and suture development. Development 127, 1845 – 1855 (2000). 23. B. St Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). 24. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 25. P. T. Chuang and A. P. McMahon, Vertebrate Hedgehog signalling modulated by induction of a Hedgehog- binding protein. Nature 397, 617 – 621 (1999). 26. U. I. Chung, B. Lanske, K. Lee, E. Li, and H. Kronenberg, The parathyroid hormone/parathyroid hormone-related peptide receptor coordinates endochondral bone development by directly controlling chondrocyte differentiation. Proc. Natl. Acad. Sci. USA 95, 13030 – 13035 (1998). 27. S. J. Karp, E. Schipani, B. St Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 28. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Juppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog- regulated bone growth. Science 273, 663 – 666 (1996). 29. J. Massague, TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998).

51

CHAPTER 2 Osteoblast Biology 30. A. Erlebacher, E. H. Filvaroff, S. E. Gitelman, and R. Derynck, Toward a molecular understanding of skeletal development. Cell 80, 371 – 378 (1995). 31. N. A. Wall and B. L. Hogan, TGF-beta related genes in development. Curr. Opin. Genet. Dev. 4, 517 – 522 (1994). 32. D. M. Kingsley, The TGF-beta superfamily: New members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8, 133 – 146 (1994). 33. K. M. Lyons, R. W. Pelton, and B. L. Hogan, Patterns of expression of murine Vgr-1 and BMP-2a RNA suggest that transforming growth factor-beta-like genes coordinately regulate aspects of embryonic development. Genes Dev. 3, 1657 – 1668 (1989). 34. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 35. S. Vainio, I. Karavanova, A. Jowett, and I. Thesleff, Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75, 45 – 58 (1993). 36. A. Yamaguchi, T. Katagiri, T. Ikeda, J. Wozney, V. Rosen, E. Wang, A. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell Biol. 113, 681 – 687 (1991). 37. R. S. Thies, M. Bauduy, B. A. Ashton, L. Kurtzberg, J. M. Wozney, and V. Rosen, Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130, 1318 – 1324 (1992). 38. E. A. Wang, V. Rosen, J. S. D’Alessandro, M. Bauduy, P. Cordes, T. Harada, D. I. Israel, R. M. Hewick, K. M. Kerns, P. LaPan, D. P. Luxenberg, D. McQuaid, I. K. Moutsatsos, J. Nove, and J. M. Wozney, Recombinant human bone morphogenetic protein induces bone formation. Proc. Natl. Acad. Sci. USA 87, 2220 – 2224 (1990). 39. T. K. Sampath, J. C. Maliakal, P. V. Hauschka, W. K. Jones, H. Sasak, R. F. Tucker, K. H. White, J. E. Coughlin, M. M. Tucker, and R. H. Pang, Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol. Chem. 267, 20352 – 20362 (1992). 40. V. Rosen and R. S. Thies, The BMP proteins in bone formation and repair. Trends. Genet. 8, 97 – 102 (1992). 41. I. Asahina, T. K. Sampath, I. Nishimura, and P. V. Hauschka, Human osteogenic protein-1 induces both chondroblastic and osteoblastic differentiation of osteoprogenitor cells derived from newborn rat calvaria. J. Cell Biol. 123, 921 – 933 (1993). 42. C. H. Heldin, K. Miyazono, and P. ten Dijke, TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 390, 465 – 471 (1997). 43. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy, and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP-2induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 44. Y. Zhang and R. Derynck, Regulation of Smad signalling by protein associations and signalling crosstalk. Trends Cell Biol. 9, 274 – 279 (1999). 45. G. R. Martin, The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571 – 1586 (1998). 46. K. Miyama, G. Yamada, T. S. Yamamoto, C. Takagi, K. Miyado, M. Sakai, N. Ueno, and H. Shibuya, A BMP-inducible gene, dlx5, regulates osteoblast differentiation and mesoderm induction. Dev. Biol. 208, 123 – 133 (1999). 47. Y. S. Hu, H. Zhou, V. Kartsogiannis, J. A. Eisman, T. J. Martin, and K. W. Ng, Expression of rat homeobox gene, rHOX, in developing

48.

49.

50.

51.

52.

53.

54.

55.

56. 57.

58.

59.

60.

61.

62.

63.

64.

and adult tissues in mice and regulation of its mRNA expression in osteoblasts by bone morphogenetic protein 2 and parathyroid hormone-related protein. Mol. Endocrinol. 12, 1721 – 1732 (1998). S. J. Tang, P. A. Hoodless, Z. Lu, M. L. Breitman, R. R. McInnes, J. L. Wrana, and M. Buchwald, The Tlx-2 homeobox gene is a downstream target of BMP signalling and is required for mouse mesoderm development. Development 125, 1877 – 1887 (1998). S. Rastegar, H. Friedle, G. Frommer, and W. Knochel, Transcriptional regulation of Xvent homeobox genes. Mech. Dev. 81, 139 – 149 (1999). S. L. Cheng, F. Lecanda, M. K. Davidson, P. M. Warlow, S. F. Zhang, L. Zhang, S. Suzuki, T. St. John, and R. Civitelli, Human osteoblasts express a repertoire of cadherins, which are critical for BMP-2-induced osteogenic differentiation. J. Bone Miner. Res. 13, 633 – 644 (1998). E. Hay, J. Lemonnier, D. Modrowski, A. Lomri, F. Lasmoles, and P. J. Marie, N- and E-cadherin mediate early human calvaria osteoblast differentiation promoted by bone morphogenetic protein-2. J. Cell. Physiol. 183, 117 – 128 (2000). X. L. Yang, H. Matsuura, Y. Fu, T. Sugiyama, and N. Miura, MFH1 is required for bone morphogenetic protein-2-induced osteoblastic differentiation of C2C12 myoblasts. FEBS Lett. 470, 29 – 34 (2000). M. Yamagata, and M. Noda, The winged-helix transcription factor CWH-3 is expressed in developing neural crest cells. Neurosci. Lett. 249, 33 – 36 (1998). A. Buchberger, M. Schwarzer, T. Brand, O. Pabst, K. Seidl, and H. H. Arnold, Chicken winged-helix transcription factor cFKH-1 prefigures axial and appendicular skeletal structures during chicken embryogenesis. Dev. Dyn. 212, 94 – 101 (1998). D. M. Dean, R. R. Berger, and M. M. Sanders, A winged-helix family member is involved in a steroid hormone-triggered regulatory circuit. Endocrinology 139, 4967 – 4975 (1998). C. Healy, D. Uwanogho, and P. T. Sharpe, Regulation and role of Sox9 in cartilage formation. Dev. Dyn. 215, 69 – 78 (1999). W. Bi, J. M. Deng, Z. Zhang, R. R. Behringer, and B. de Crombrugghe, Sox9 is required for cartilage formation. Nature Genet. 22, 85 – 89 (1999). T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y.-H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). F. Otto, A. P. Thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). S. Piccolo, Y. Sasai, B. Lu, and E. M. De Robertis, Dorsoventral patterning in Xenopus: Inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86, 589 – 598 (1996). D. R. Hsu, A. N. Economides, X. Wang, P. M. Eimon, and R. M. Harland, The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol. Cell 1, 673 – 683 (1998). L. B. Zimmerman, J. M. Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992).

52 65. E. Gazzerro, V. Gangji, and E. Canalis, Bone morphogenetic proteins induce the expression of noggin, which limits their activity in cultured rat osteoblasts. J. Clin. Invest. 102, 2106 – 2114 (1998). 66. E. Abe, M. Yamamoto, Y. Taguchi, B. Lecka-Czernik, C. A. O’Brien, A. N. Economides, N. Stahl, R. L. Jilka, and S. C. Manolagas, Essential requirement of BMPs-2/4 for both osteoblast and osteoclast formation in murine bone marrow cultures from adult mice: antagonism by noggin. J. Bone Miner. Res. 15, 663 – 673 (2000). 67. T. Kameda, C. Koike, K. Saitoh, A. Kuroiwa, and H. Iba, Developmental patterning in chondrocytic cultures by morphogenic gradients: BMP induces expression of indian hedgehog and noggin. Genes Cells 4, 175 – 184 (1999). 68. C. Marcelle, M. R. Stark, and M. Bronner-Fraser, Coordinate actions of BMPs, Wnts, Shh and noggin mediate patterning of the dorsal somite. Development 124, 3955 – 3963 (1997). 69. E. Hirsinger, D. Duprez, C. Jouve, P. Malapert, J. Cooke, and O. Pourquie, Noggin acts downstream of Wnt and Sonic Hedgehog to antagonize BMP4 in avian somite patterning. Development 124, 4605 – 4614 (1997). 70. S. Pathi, J. B. Rutenberg, R. L. Johnson, and A. Vortkamp, Interaction of Ihh and BMP/Noggin signaling during cartilage differentiation. Dev. Biol. 209, 239 – 253 (1999). 71. J. M. Hock, J. Fonseca, M. Gunness-Hey, B. E. Kemp, and T. J. Martin, Comparison of the anabolic effects of synthetic parathyroid hormone- related protein (PTHrP) 1 – 34 and PTH 1 – 34 on bone in rats. Endocrinology 125, 2022 – 2027 (1989). 72. M. Mannstadt, H. Juppner, and T. J. Gardella, Receptors for PTH and PTHrP: Their biological importance and functional properties. Am. J. Physiol. 277, F665 – F675 (1999). 73. J. M. Moseley and T. J. Martin, Parathyroid hormone-related protein: Physiological actions. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 363 – 376. Academic Press, San Diego, 1996. 74. A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormonerelated peptide gene. Genes Dev. 8, 277 – 289 (1994). 75. D. B. Pateder, R. N. Rosier, E. M. Schwarz, P. R. Reynolds, J. E. Puzas, M. D’Souza, and R. J. O’Keefe, PTHrP expression in chondrocytes, regulation by TGF-beta, and interactions between epiphyseal and growth plate chondrocytes. Exp. Cell Res. 256, 555 – 562 (2000). 76. R. Serra, A. Karaplis, and P. Sohn, Parathyroid hormone-related peptide (PTHrP)-dependent and -independent effects of transforming growth factor beta (TGF-beta) on endochondral bone formation. J. Cell Biol. 145, 783 – 794 (1999). 77. L. K. McCauley, A. J. Koh, C. A. Beecher, Y. Cui, J. D. Decker, and R. T. Franceschi, Effects of differentiation and transforming growth factor beta 1 on PTH/PTHrP receptor mRNA levels in MC3T3-E1 cells. J. Bone Miner. Res. 10, 1243 – 1255 (1995). 78. B. Lanske, M. Amling, L. Neff, J. Guiducci, R. Baron, and H. M. Kronenberg, Ablation of the PTHrP gene or the PTH/PTHrP receptor gene leads to distinct abnormalities in bone development. J. Clin. Invest. 104, 399 – 407 (1999). 79. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 80. E. Schipani, B. Lanske, J. Hunzelman, A. Luz, C. S. Kovacs, K. Lee, A. Pirro, H. M. Kronenberg, and H. Juppner, Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide. Proc. Natl. Acad. Sci. USA 94, 13689 – 13694 (1997). 81. B. Zerega, S. Cermelli, P. Bianco, R. Cancedda, and F. D. Cancedda, Parathyroid hormone [PTH(1-34)] and parathyroid hormone-related

LIAN AND STEIN

82.

83.

84.

85.

86.

87. 88.

89.

90.

91.

92.

93.

94.

95.

96.

protein [PTHrP(1-34)] promote reversion of hypertrophic chondrocytes to a prehypertrophic proliferating phenotype and prevent terminal differentiation of osteoblast-like cells. J. Bone Miner. Res. 14, 1281 – 1289 (1999). Z. F. Chen and R. R. Behringer, Twist is required in head mesenchyme for cranial neural tube morphogenesis. Genes Dev. 9, 686 – 699 (1995). P. Cserjesi, D. Brown, K. L. Ligon, G. E. Lyons, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, and E. N. Olson, Scleraxis: A basic helix-loop-helix protein that prefigures skeletal formation during mouse embryogenesis. Development 121, 1099 – 1110 (1995). T. Ogata and M. Noda, Expression of Id, a negative regulator of helix-loop-helix DNA binding proteins, is down-regulated at confluence and enhanced by dexamethasone in a mouse osteoblastic cell line, MC3T3E1. Biochem. Biophys. Res. Commun. 180, 1194 – 1199 (1991). S. S. Murray, C. A. Glackin, K. A. Winters, D. Gazit, A. J. Kahn, and E. J. Murray, Expression of helix-loop-helix regulatory genes during differentiation of mouse osteoblastic cells. J. Bone Miner. Res. 7, 1131 – 1138 (1992). C. A. Glackin, M. Lee, G. Lowe, S. Morales, J. Wergedal, and D. Strong, Overexpressing human TWIST in high SaOS (HSaOS) cells results in major differences in cellular morphology, proliferation rates, and ALP activities: Knocking out human Id-2 greatly reduces the expression of all TWIST family members, indicating that Id-2 may regulate TWIST during osteoblast differentiation. J. Bone Miner. Res. 12, S154 – S154 (1997). [Abstract] W. J. Gehring, M. Affolter, and T. Burglin, Homeodomain proteins. Annu. Rev. Biochem. 63, 487 – 526 (1994). M. Mina, J. Gluhak, W. B. Upholt, E. J. Kollar, and B. Rogers, Experimental analysis of Msx-1 and Msx-2 gene expression during chick mandibular morphogenesis. Dev. Dyn. 202, 195 – 214 (1995). G. Q. Zhao, S. Zhao, X. Zhou, H. Eberspaecher, M. Solursh, and B. de Crombrugghe, rDlx, a novel distal-less-like homeoprotein is expressed in developing cartilages and discrete neuronal tissues. Dev. Biol. 164, 37 – 51 (1994). D. Ferrari, L. Sumoy, J. Gannon, H. Sun, A. M. Brown, W. B. Upholt, and R. A. Kosher, The expression pattern of the Distal-less homeobox-containing gene Dlx-5 in the developing chick limb bud suggests its involvement in apical ectodermal ridge activity, pattern formation, and cartilage differentiation. Mech. Dev. 52, 257 – 264 (1995). E. W. Jabs, U. Muller, X. Li, L. Ma, W. Luo, I. S. Haworth, I. Klisak, R. Sparkes, M. L. Warman, J. B. Mulliken, M. L. Snead, and R. Maxson, A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 75, 443 – 450 (1993). M. Dodig, T. Tadic, M. S. Kronenberg, S. Dacic, Y. H. Liu, R. Maxson, D. W. Rowe, and A. C. Lichtler, Ectopic Msx2 overexpression inhibits and Msx2 antisense stimulates calvarial osteoblast differentiation. Dev. Biol. 209, 298–307 (1999). N. A. Speck and T. Stacy, A new transcription factor family associated with human leukemias. Crit. Rev. Eukary. Gene Expr. 5, 337 – 364 (1995). J. J. Westendorf and S. W. Hiebert, Mammalian runt-domain proteins and their roles in hematopoiesis, osteogenesis and leukemia. J. Cell. Boichem. (32 – 33) Suppl., 51 – 58 (1999). Y. Ito, Molecular basis of tissue-specific gene expression mediated by the runt domain transcription factor PEBP2/CBF. Genes Cells 4, 685 – 696 (1999). E. Ogawa, M. Maruyama, H. Kagoshima, M. Inuzuka, J. Lu, M. Satake, K. Shigesada, and Y. Ito, PEBP2/PEA2 represents a family of transcription factors homologous to the products of the Drosophila runt gene and the human AML1 gene. Proc. Natl. Acad. Sci. USA 90, 6859 – 6863 (1993).

53

CHAPTER 2 Osteoblast Biology 96a. M. Stewart, A. Terry, M. Hu, M. O’Hara, K. Blyth, E. Baxter, E. Cameron, D. E. Onions, and J. C. Neil, Proviral insertions induce the expression of bone-specific isoforms of PEBP2alphaA (CBFA1): Evidence for a new myc collaborating oncogene. Proc. Natl. Acad. Sci. USA 94, 8646 – 8651 (1997). 96b. K. Thirunavukkarasu, M. Mahajan, K. W. McLarren, S. Stifani, and G. Karsenty, Two domains unique to osteoblast-specific transcription factor Osf2/Cbfa1 contribute to its transactivation function and its inability to heterodimerize with Cbf beta, Mol. Cell Biol. 18, 4197–4208 (1998). 97. M. Niki, H. Okada, H. Takano, J. Kuno, K. Tani, H. Hibino, S. Asano, Y. Ito, M. Satake, and T. Noda, Hematopoiesis in the fetal liver is impaired by targeted mutagenesis of a gene encoding a non-DNA binding subunit of the transcription factor, polyomavirus enhancer binding protein 2/core binding factor. Proc. Natl. Acad. Sci. USA 94, 5697 – 5702 (1997). 98. T. Okuda, J. van Deursen, S. W. Hiebert, G. Grosveld, and J. R. Downing, AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell 84, 321 – 330 (1996). 99. H. L. Merriman, A. J. van Wijnen, S. Hiebert, J. P. Bidwell, E. Fey, J. Lian, J. Stein, and G. S. Stein, The tissue-specific nuclear matrix protein, NMP-2, is a member of the AML/CBF/PEBP2/runt domain transcription factor family: Interactions with the osteocalcin gene promoter. Biochemistry 34, 13125 – 13132 (1995). 100. C. Banerjee, S. W. Hiebert, J. L. Stein, J. B. Lian, and G. S. Stein, An AML-1 consensus sequence binds an osteoblast-specific complex and transcriptionally activates the osteocalcin gene. Proc. Natl. Acad. Sci. USA 93, 4968 – 4973 (1996). 101. P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). 102. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 103. C. Banerjee, L. R. McCabe, J.-Y. Choi, S. W. Hiebert, J. L. Stein, G. S. Stein, and J. B. Lian, Runt homology domain proteins in osteoblast differentiation: AML-3/CBFA1 is a major component of a bone specific complex. J. Cell. Biochem. 66, 1 – 8 (1997). 104. D. O. Sillence, H. E. Ritchie, and P. B. Selby, Animal model: Skeletal anomalies in mice with cleidocranial dysplasia. Am. J. Med. Genet. 27, 75 – 85 (1987). 105. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 106. S. Mundlos, Cleidocranial dysplasia: Clinical and molecular genetics. J. Med. Genet. 36, 177 – 182 (1999). 107. B. Lee, K. Thirunavukkarasu, L. Zhou, L. Pastore, A. Baldini, J. Hecht, V. Geoffroy, P. Ducy, and G. Karsenty, Missense mutations abolishing DNA binding of the osteoblast- specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nature Genet. 16, 307 – 310 (1997). 108. Y. Zhang, N. Yasui, N. Kakazu, T. Abe, K. Takada, S. Imai, M. Sato, S. Nomura, T. Ochi, S. Okuzumi, H. Nogami, T. Nagai, H. Ohashi, and Y. Ito, PEBP2alphaA/CBFA1 mutations in Japanese cleidocranial dysplasia patients. Gene 244, 21 – 28 (2000). 109. H. Enomoto, M. Enomoto-Iwamoto, M. Iwamoto, S. Nomura, M. Himeno, Y. Kitamura, T. Kishimoto, and T. Komori, Cbfa1 is a positive regulatory factor in chondrocyte maturation. J. Biol. Chem. 275, 8695 – 8702 (2000). 110. M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of

111. 112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). M. Satake, S. Nomura, Y. Yamaguchi-Iwai, Y. Takahama, Y. Hashimoto, M. Niki, Y. Kitamura, and Y. Ito, Expression of the Runt domain-encoding PEBP2 alpha genes in T cells during thymic development. Mol. Cell. Biol. 15, 1662 – 1670 (1995). S. Ogawa, H. Harada, M. Fujiwara, S. Tagashira, T. Katsumata, and H. Takada, Cbfa1, an essential transcription factor for bone formation, is expressed in testis from the same promoter used in bone. DNA Res. 7, 181 – 185 (2000). K. Tsuji, Y. Ito, and M. Noda, Expression of the PEBP2A/ AML3/CBFA1 gene is regulated by BMP4/7 heterodimer and its overexpression suppresses type I collagen and osteocalcin gene expression in osteoblastic and nonosteoblastic mesenchymal cells. Bone 22, 87 – 92 (1998). F. Gori, T. Thomas, K. C. Hicok, T. C. Spelsberg, and B. L. Riggs, Differentiation of human marrow stromal precursor cells: Bone morphogenetic protein-2 increases OSF2/CBFA1, enhances osteoblast commitment, and inhibits late adipocyte maturation. J. Bone Miner. Res. 14, 1522 – 1535 (1999). K. Satomura, P. Krebsbach, P. Bianco, and R. P. Gehron, Osteogenic imprinting upstream of marrow stromal cell differentiation. J. Cell. Biochem. 78, 391 – 403 (2000). P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). H. Harada, S. Tagashira, M. Fujiwara, S. Ogawa, T. Katsumata, A. Yamaguchi, T. Komori, and M. Nakatsuka, Cbfa1 isoforms exert functional differences in osteoblast differentiation. J. Biol. Chem. 274, 6972 – 6978 (1999). M. Muenke and A. Wilkie, Craniosynostosis syndromes. In “The Metabolic and Molecular Bases of Inherited Disease” (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, and B. Vogelstein, eds.). McGraw-Hill, New York, 2000. Y. X. Zhou, X. Xu, L. Chen, C. Li, S. G. Brodie, and C. X. Deng, A Pro250Arg substitution in mouse Fgfr1 causes increased expression of Cbfa1 and premature fusion of calvarial sutures. Hum. Mol. Genet. 9, 2001 – 2008 (2000). M. H. Lee, A. Javed, H. J. Kim, H. I. Shin, S. Gutierrez, J. Y. Choi, V. Rosen, J. L. Stein, A. J. van Wijnen, G. S. Stein, J. B. Lian, and H. M. Ryoo, Transient upregulation of CBFA1 in response to bone morphogenetic protein-2 and transforming growth factor 1 in C2C12 myogenic cells coincides with suppression of the myogenic phenotype but is not sufficient for osteoblast differentiation. J. Cell. Biochem. 73, 114 – 125 (1999). Y. Takazawa, K. Tsuji, A. Nifuji, H. Kurosawa, Y. Ito, and M. Noda, An osteogenesis-related transcription factor, core-binding factor A1, is constitutively expressed in the chondrocytic cell line TC6, and its expression is upregulated by bone morphogenetic protein-2. J. Endocrinol. 165, 579 – 586 (2000). J. Hanai, L. F. Chen, T. Kanno, N. Ohtani-Fujita, W. Y. Kim, W. H. Guo, T. Imamura, Y. Ishidou, M. Fukuchi, M. J. Shi, J. Stavnezer, M. Kawabata, K. Miyazono, and Y. Ito, Interaction and functional cooperation of PEBP2/CBF with Smads: Synergistic induction of the immunoglobulin germline Calpha promoter. J. Biol. Chem. 274, 31577 – 31582 (1999). K. S. Lee, H. J. Kim, Q. L. Li, X. Z. Chi, C. Ueta, T. Komori, J. M. Wozney, E. G. Kim, J. Y. Choi, H. M. Ryoo, and S. C. Bae, Runx2 is a common target of transforming growth factor beta1 and bone morphogenetic protein 2, and cooperation between runx2 and smad5 induces osteoblast-specific gene expression in the pluripotent mesenchymal precursor cell line C2C12. Mol. Cell Biol. 20, 8783–8792 (2000).

54 125. D. Wang, K. Christensen, K. Chawla, G. Xiao, P. H. Krebsbach, and R. T. Franceschi, Isolation and characterization of MC3T3-E1 preosteoblast subclones with distinct in vitro and in vivo differentiation/mineralization potential. J. Bone Miner. Res. 14, 893 – 903 (1999). 126. A. Vortkamp, S. Pathi, G. M. Peretti, E. M. Caruso, D. J. Zaleske, and C. J. Tabin, Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech. Dev. 71, 65 – 76 (1998). 127. C. Ferguson, E. Alpern, T. Miclau, and J. A. Helms, Does adult fracture repair recapitulate embryonic skeletal formation? Mech. Dev. 87, 57 – 66 (1999). 128. G. S. Stein and J. B. Lian, Molecular mechanisms mediating developmental and hormone regulated expression of genes in osteoblasts: an integrated relationship of cell growth and differentiation. In “Cellular and Molecular Biology of Bone” (A. Noda, ed.), pp. 47 – 95. Academic Press, San Diego, 1993. 129. J. E. Aubin, K. Turksen, and J. N. M. Heersche, Osteoblastic cell lineage. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 1 – 45. Academic Press, San Diego, 1993. 130. S. C. Marks, Jr. and S. N. Popoff, Bone cell biology: The regulation of development, structure, and function in the skeleton. Am. J. Anat. 183, 1 – 44 (1988). 131. P. J. Nijweide, A. van der Plas, and A. A. Othof, Osteoblastic differentiation. In “Cell and Molecular Biology of Vertebrate Hard Tissue” (D. Evered and S. Harnett, eds.), pp. 61 – 72. Wiley, New York, 1988. 132. E. Bonucci, The histology, histochemistry and ultrastructure of bone. In “Bone Regulatory Factors” (A. Pecile and B. de Bernard, eds.), pp. 15 – 39. Plenum Press, New York, 1990. 133. M. Owen and A. J. Friedenstein, Stromal stem cells: Marrow-derived osteogenic precursors. In “Cell and Molecular Biology of Vertebrate Hard Tissues” (D. Evered and S. Harnett, eds.), pp. 42 – 60. Wiley, Chichester, England, 1988. 134. M. E. Holtrop, The ultrastructure of bone. Ann. Clin. Lab. Sci. 5, 264 – 271 (1975). 135. S. Nomura, A. J. Wills, D. R. Edwards, J. K. Heath, and B. L. Hogan, Developmental expression of 2ar (osteopontin) and SPARC (osteonectin) RNA as revealed by in situ hybridization. J. Cell Biol. 106, 441 – 450 (1988). 136. M. Weinreb, D. Shinar, and G. A. Rodan, Different pattern of alkaline phosphatase, osteopontin, and osteocalcin expression in developing rat bone visualized by an in situ hybridization. J. Bone Miner. Res. 5, 831 – 842 (1990). 137. J. N. M. Heersche, S. M. Reimers, J. L. Wrana, M. M. Y. Waye, and A. K. Gupta, Changes in expression of alpha 1 type 1 collagen and osteocalcin mRNA in osteoblasts and adontoblasts at different stages of maturity as shown by in situ hybridization. Proc. Finn. Dent. Soc. 88, 173 – 182 (1992). 138. M. Sandberg, H. Autio-Harmainen, and E. Vuorio, Localization of the expression of types I, III, and IV collagen, TGF-beta 1 and c-fos genes in developing human calvarial bones. Dev. Biol. 130, 324 – 334 (1988). 139. K. Lee, J. D. Deeds, S. Chiba, M. Un-No, A. T. Bond, and G. V. Segre, Parathyroid hormone induces sequential c-fos expression in bone cells in vivo: In situ localization of its receptor and c-fos messenger ribonucleic acids. Endocrinology 134, 441 – 450 (1994). 140. M. P. Mark, C. W. Prince, S. Gay, R. L. Austin, M. Bhown, R. D. Finkelman, and W. T. Butler, A comparative immunocytochemical study on the subcellular distributions of 44 kDa bone phosphoprotein and bone gamma- carboxyglutamic acid (Gla)- containing protein in osteoblasts. J. Bone Miner. Res. 2, 337 – 346 (1987). 141. M. D. McKee and A. Nanci, Ultrastructural, cytochemical and immunocytochemical studies of bone and its interfaces. Cells Mater. 3, 219 – 243 (1993). 142. C. S. Carlson, H. M. Tulli, M. J. Jayo, R. F. Loeser, R. P. Tracy, K. G. Mann, and M. R. Adams, Immunolocalization of noncollage-

LIAN AND STEIN

143.

144.

145.

146.

147. 148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

nous bone matrix proteins in lumbar vertebrae from intact and surgically menopausal cynomolgus monkeys. J. Bone Miner. Res. 8, 71 – 81 (1993). S. M. Pockwinse, L. G. Wilming, D. M. Conlon, G. S. Stein, and J. B. Lian, Expression of cell growth and bone specific genes at single cell resolution during development of bone tissue-like organization in primary osteoblast cultures. J. Cell. Biochem. 49, 310 – 323 (1992). F. Liu, L. Malaval, A. K. Gupta, and J. E. Aubin, Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast differentiation: Polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol. 166, 220 – 234 (1994). C. Zeng, S. McNeil, S. Pockwinse, J. A. Nickerson, L. Shopland, J. B. Lawrence, S. Penman, S. W. Hiebert, J. B. Lian, A. J. van Wijnen, J. L. Stein, and G. S. Stein, Intranuclear targeting of AML/CBF regulatory factors to nuclear matrix-associated transcriptional domains. Proc. Natl. Acad. Sci. USA 95, 1585 – 1589 (1998). G. S. Stein, A. J. van Wijnen, J. L. Stein, J. B. Lian, S. Pockwinse, and S. McNeil, Interrelationships of nuclear structure and transcriptional control: Functional consequences of being in the right place at the right time. J. Cell Biochem. 70, 200 – 212 (1998). S. B. Doty and B. H. Schofield, Enzyme histochemistry of bone and cartilage cells. Prog. Histochem. Cytochem. 8, 1 – 38 (1976). T. Nakase, S. Nomura, H. Yoshikawa, J. Hashimoto, S. Hirota, Y. Kitamura, S. Oikawa, K. Ono, and K. Takaoka, Transient and localized expression of bone morphogenetic protein 4 messenger RNA during fracture healing. J. Bone Miner. Res. 9, 651 – 659 (1994). J. E. Aubin, Osteoprogenitor cell frequency in rat bone marrow stromal populations: Role for heterotypic cell-cell interactions in osteoblast differentiation. J. Cell Biochem. 72, 396 – 410 (1999). A. J. Friedenstein, R. K. Chailakhjan, and K. S. Lalykina, The development of fibroblast colonies in monolayer cultures of guinea- pig bone marrow and spleen cells. Cell Tissue Kinet. 3, 393 – 403 (1970). A. J. Friedenstein, N. V. Latzinik, Gorskaya YuF, E. A. Luria, and I. L. Moskvina, Bone marrow stromal colony formation requires stimulation by haemopoietic cells. Bone Miner. 18, 199 – 213 (1992). A. J. Friedenstein, R. K. Chailakhyan, and U. V. Gerasimov, Bone marrow osteogenic stem cells: In vitro cultivation and transplantation in diffusion chambers. Cell Tissue Kinet. 20, 263 – 272 (1987). Z. Hou, Q. Nguyen, B. Frenkel, S. K. Nilsson, M. Milne, A. J. van Wijnen, J. L. Stein, P. Quesenberry, J. B. Lian, and G. S. Stein, Osteoblast-specific gene expression after transplantation of marrow cells: Implications for skeletal gene therapy. Proc. Natl. Acad. Sci. USA 96, 7294 – 7299 (1999). P. H. Krebsbach, S. A. Kuznetsov, K. Satomura, R. V. Emmons, D. W. Rowe, and P. G. Robey, Bone formation in vivo: Comparison of osteogenesis by transplanted mouse and human marrow stromal fibroblasts. Transplantation 63, 1059 – 1069 (1997). S. D. Boden, L. Titus, G. Hair, Y. Liu, M. Viggeswarapu, M. S. Nanes, and C. Baranowski, Lumbar spine fusion by local gene therapy with a cDNA encoding a novel osteoinductive protein (LMP-1). Spine 23, 2486 – 2492 (1998). D. Gazit, G. Turgeman, P. Kelley, E. Wang, M. Jalenak, Y. Zilberman, and I. Moutsatsos, Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone: a novel cell-mediated gene therapy. J. Gene Med. 1, 121 – 133 (1999). T. D. Alden, D. D. Pittman, G. R. Hankins, E. J. Beres, J. A. Engh, S. Das, S. B. Hudson, K. M. Kerns, D. F. Kallmes, and G. A. Helm, In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum. Gene Ther. 10, 2245 – 2253 (1999). R. T. Franceschi, D. Wang, P. H. Krebsbach, and R. B. Rutherford, Gene therapy for bone formation: In vitro and in vivo osteogenic activity of an adenovirus expressing BMP7. J. Cell Biochem. 78, 476 – 486 (2000).

55

CHAPTER 2 Osteoblast Biology 159. S. P. Bruder and B. S. Fox, Tissue engineering of bone: Cell based strategies. Clin. Orthop. S68 – S83 (1999). 160. J. C. Marx, J. A. Allay, D. A. Persons, S. A. Nooner, P. W. Hargrove, P. F. Kelly, E. F. Vanin, and E. M. Horwitz, High-efficiency transduction and long-term gene expression with a murine stem cell retroviral vector encoding the green fluorescent protein in human marrow stromal cells. Hum. Gene Ther. 10, 1163 – 1173 (1999). 161. A. I. Caplan, D. Reuben, and S. E. Haynesworth, Cell-based tissue engineering therapies: the influence of whole body physiology. Adv. Drug Deliv. Rev. 33, 3 – 14 (1998). 162. K. D. Riew, N. M. Wright, S. Cheng, L. V. Avioli, and J. Lou, Induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene in a rabbit spinal fusion model. Calcif. Tissue Int. 63, 357 – 360 (1998). 163. J. A. Allay, J. E. Dennis, S. E. Haynesworth, M. K. Majumdar, D. W. Clapp, L. D. Shultz, A. I. Caplan, and S. L. Gerson, LacZ and interleukin-3 expression in vivo after retroviral transduction of marrowderived human osteogenic mesenchymal progenitors. Hum. Gene Ther. 8, 1417 – 1427 (1997). 164. D. J. Prockop, Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 276, 71 – 74 (1997). 165. J. B. Lian, G. S. Stein, J. L. Stein, and A. J. van Wijnen, Marrow transplantation and targeted gene therapy to the skeleton. Clin. Orthop. Rel. Res. 3795, 5146–5155 (2000). 166. J. B. Lian and C. Niyibizi, Breakout Session 4: Gene therapy for genetic diseases of bone. Clin. Orthop. Rel. Res. 3795, 5159–5163 (2000). 167. J. E. Aubin, Bone stem cells. J. Cell Biochem. Suppl. 30 – 31, 73 – 82 (1998). 168. P. Bianco, M. Riminucci, S. Kuznetsov, and P. G. Robey, Multipotential cells in the bone marrow stroma: Regulation in the context of organ physiology. Crit. Rev. Eukaryot. Gene Expr. 9, 159 – 173 (1999). 169. M. Owen, Marrow stromal stem cells. J. Cell Sci. 10, 63 – 76 (1988). 170. M. F. Pittenger, A. M. Mackay, S. C. Beck, R. K. Jaiswal, R. Douglas, J. D. Mosca, M. A. Moorman, D. W. Simonetti, S. Craig, and D. R. Marshak, Multilineage potential of adult human mesenchymal stem cells. Science 284, 143 – 147 (1999). 171. S. E. Haynesworth, J. Goshima, V. M. Goldberg, and A. I. Caplan, Characterization of cells with osteogenic potential from human marrow. Bone 13, 81 – 88 (1992). 172. S. E. Haynesworth, M. A. Baber, and A. I. Caplan, Cell surface antigens on human marrow-derived mesenchymal cells are detected by monoclonal antibodies. Bone 13, 69 – 80 (1992). 173. A. Muraglia, R. Cancedda, and R. Quarto, Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model. J. Cell Sci. 113(Pt. 7), 1161 – 1166 (2000). 174. Y. Negishi, A. Kudo, A. Obinata, K. Kawashima, H. Hirano, N. Yanai, M. Obinata, and H. Endo, Multipotency of a bone marrow stromal cell line, TBR31-2, established from ts-SV40 T antigen gene transgenic mice. Biochem. Biophys. Res. Commun. 268, 450 – 455 (2000). 175. B. Lecka-Czernik, I. Gubrij, E. J. Moerman, O. Kajkenova, D. A. Lipschitz, S. C. Manolagas, and R. L. Jilka, Inhibition of Osf2/Cbfal expression and terminal osteoblast differentiation by PPARgamma2. J. Cell Biochem. 74, 357 – 371 (1999). 176. H. A. Kodama, Y. Amagai, H. Koyama, and S. Kasai, Hormonal responsiveness of a preadipose cell line derived from newborn mouse calvaria. J. Cell. Physiol. 112, 83 – 88 (1982). 177. J. N. Beresford, J. H. Bennett, C. Devlin, P. S. Leboy, and M. E. Owen, Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures. J. Cell Sci. 102, 341 – 351 (1992). 178. J. M. Bennett, D. Catovsky, M. T. Daniel, G. Flandrin, D. A. Galton, H. R. Gralnick, and C. Sultan, Proposal for the recognition of

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

minimally differentiated acute myeloid leukaemia (AML-MO). Br. J Haematol. 78, 325 – 329 (1991). X. Ji, D. Chen, C. Xu, S. E. Harris, G. R. Mundy, and T. Yoneda, Patterns of gene expression associated with BMP-2-induced osteoblast and adipocyte differentiation of mesenchymal progenitor cell 3T3-F442A. J. Bone Miner. Metab. 18, 132 – 139 (2000). M. E. Nuttall, A. J. Patton, D. L. Olivera, D. P. Nadeau, and M. Gowen, Human trabecular bone cells are able to express both osteoblastic and adipocytic phenotype: Implications for osteopenic disorders. J. Bone Miner. Res. 13, 371 – 382 (1998). T. Thomas, F. Gori, S. Khosla, M. D. Jensen, B. Burguera, and B. L. Riggs, Leptin acts on human marrow stromal cells to enhance differentiation to osteoblasts and to inhibit differentiation to adipocytes. Endocrinology 140, 1630 – 1638 (1999). D. Chen, X. Ji, M. A. Harris, J. Q. Feng, G. Karsenty, A. J. Celeste, V. Rosen, G. R. Mundy, and S. E. Harris, Differential roles for bone morphogenetic protein (BMP) receptor type IB and IA in differentiation and specification of mesenchymal precursor cells to osteoblast and adipocyte lineages. J. Cell Biol. 142, 295 – 305 (1998). V. Rosen, J. Nove, J. J. Song, R. S. Thies, K. Cox, and J. M. Wozney, Responsiveness of clonal limb bud cell lines to bone morphogenetic protein 2 reveals a sequential relationship between cartilage and bone cell phenotypes. J. Bone Miner. Res. 9, 1759 – 1768 (1994). S. Nishida, N. Endo, H. Yamagiwa, T. Tanizawa, and H. E. Takahashi, Number of osteoprogenitor cells in human bone marrow markedly decreases after skeletal maturation. J. Bone Miner. Metab. 17, 171 – 177 (1999). D. A. Puleo, Dependence of mesenchymal cell responses on duration of exposure to bone morphogenetic protein-2 in vitro. J. Cell. Physiol 173, 93 – 101 (1997). M. E. Nuttall and J. M. Gimble, Is there a therapeutic opportunity to either prevent or treat osteopenic disorders by inhibiting marrow adipogenesis? Bone 27, 177 – 184 (2000). J. M. Gimble, C. Morgan, K. Kelly, X. Wu, V. Dandapani, C. S. Wang, and V. Rosen, Bone morphogenetic proteins inhibit adipocyte differentiation by bone marrow stromal cells. J. Cell Biochem. 58, 393 – 402 (1995). R. A. Ignotz and J. Massague, Type beta transforming growth factor controls the adipogenic differentiation of 3T3 fibroblasts. Proc. Natl.Acad. Sci.USA 82, 8530 – 8534 (1985). A. Chawla and M. A. Lazar, Peroxisome proliferator and retinoid signaling pathways co-regulate preadipocyte phenotype and survival. Proc. Natl. Acad. Sci. USA 91, 1786 – 1790 (1994). P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). K. A. Kelly and J. M. Gimble, 1,25-Dihydroxy vitamin D3 inhibits adipocyte differentiation and gene expression in murine bone marrow stromal cell clones and primary cultures. Endocrinolgy 139, 2622 – 2628 (1998). R. K. Jaiswal, N. Jaiswal, S. P. Bruder, G. Mbalaviele, D. R. Marshak, and M. F. Pittenger, Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogenactivated protein kinase. J. Biol. Chem. 275, 9645 – 9652 (2000). S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocr. Rev. 21, 115 – 137 (2000). S. P. Bruder, D. J. Fink, and A. I. Caplan, Mesenchymal stem cells in bone development, bone repair, and skeletal regeneration therapy. J. Cell. Biochem. 56, 283 – 294 (1994). S. P. Bruder, N. S. Ricalton, R. E. Boynton, T. J. Connolly, N. Jaiswal, J. Zaia, and F. P. Barry, Mesenchymal stem cell surface antigen SB-10 corresponds to activated leukocyte cell adhesion

56

196.

197.

198.

199.

200.

201. 202.

203.

204.

205.

206.

207.

208.

209.

210.

211.

212.

213.

214.

LIAN AND STEIN molecule and is involved in osteogenic differentiation. J. Bone Miner. Res. 13, 655 – 663 (1998). S. Gronthos, S. E. Graves, S. Ohta, and P. J. Simmons, The STRO-1  fraction of adult human bone marrow contains the osteogenic precursors. Blood 84, 4164 – 4173 (1994). A. Wetterwald, W. Hoffstetter, M. G. Cecchini, B. Lanske, C. Wagner, H. Fleisch, and M. Atkinson, Characterization and cloning of the E11 antigen, a marker expressed by rat osteoblasts and osteocytes. Bone 18, 125 – 132 (1996). C. J. Joyner, A. Bennett, and J. T. Triffitt, Identification and enrichment of human osteoprogenitor cells by using differentiation stagespecific monoclonal antibodies. Bone 21, 1 – 6 (1997). P. J. Nijweide and R. J. Mulder, Identification of osteocytes in osteoblast-like cell cultures using a monoclonal antibody specifically directed against osteocytes. Histochemistry 84, 342 – 347 (1986). D. E. Hughes, D. M. Salter, and R. Simpson, CD44 expression in human bone: a novel marker of osteocytic differentiation. J. Bone Miner. Res. 9, 39 – 44 (1994). H. H. Jamal and J. E. Aubin, CD44 expression in fetal rat bone: In vivo and in vitro analysis. Exp. Cell Res. 223, 467 – 477 (1996). K. Turksen and J. E. Aubin, Positive and negative immunoselection for enrichment of two classes of osteoprogenitor cells. J. Cell Biol. 114, 373 – 384 (1991). P. J. Simmons and B. Torok-Storb, Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1. Blood 78, 55 – 62 (1991). R. J. Byers, J. Brown, C. Brandwood, P. Wood, W. Staley, L. Hainey, A. J. Freemont, and J. A. Hoyland, Osteoblastic differentiation and mRNA analysis of STRO-1-positive human bone marrow stromal cells using primary in vitro culture and poly (A) PCR. J. Pathol. 187, 374 – 381 (1999). K. Stewart, S. Walsh, J. Screen, C. M. Jefferiss, J. Chainey, G. R. Jordan, and J. N. Beresford, Further characterization of cells expressing STRO-1 in cultures of adult human bone marrow stromal cells. J. Bone Miner. Res. 14, 1345 – 1356 (1999). B. O. Oyajobi, A. Lomri, M. Hott, and P. J. Marie, Isolation and characterization of human clonogenic osteoblast progenitors immunoselected from fetal bone marrow stroma using STRO-1 monoclonal antibody. J. Bone Miner. Res. 14, 351 – 361 (1999). M. E. Owen, J. Cave, and C. J. Joyner, Clonal analysis in vitro of osteogenic differentiation of marrow CFU-F. J. Cell. Sci. 87, 731 – 738 (1987). L. Malaval, A. K. Gupta, and J. E. Aubin, Leukemia inhibitory factor inhibits osteogenic differentiation in rat calvaria cell cultures. Endocrinology 136, 1411 – 1418 (1995). L. Malaval, A. K. Gupta, F. Liu, P. D. Delmas, and J. E. Aubin, LIF, but not IL-6, regulates osteoprogenitor differentiation in rat calvaria cell cultures: Modulation by dexamethasone. J. Bone Miner. Res. 13, 175 – 184 (1998). J. Cornish, K. Callon, A. King, S. Edgar, and I. R. Reid, The effect of leukemia inhibitory factor on bone in vivo. Endocrinology 132, 1359 – 1366 (1993). I. Martin, A. Muraglia, G. Campanile, R. Cancedda, and R. Quarto, Fibroblast growth factor-2 supports ex vivo expansion and maintenance of osteogenic precursors from human bone marrow. Endocrinology 138, 4456 – 4462 (1997). T. Nakamura, K. Hanada, M. Tamura, T. Shibanushi, H. Nigi, M. Tagawa, S. Fukumoto, and T. Matsumoto, Stimulation of endosteal bone formation by systemic injections of recombinant basic fibroblast growth factor in rats. Endocrinology 136, 1276 – 1284 (1995). M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). J. M. Hock, E. Canalis, and M. Centrella, Transforming growth factor-beta stimulates bone matrix apposition and bone cell

215.

216.

217.

218.

219.

220.

221.

222.

223.

224.

225.

226. 227. 228.

229.

230.

231.

replication in cultured fetal rat calvariae. Endocrinology 126, 421 – 426 (1990). T. Onishi, Y. Ishidou, T. Nagamine, K. Yone, T. Imamura, M. Kato, T. K. Sampath, P. T. Dijke, and T. Sakou, Distinct and overlapping patterns of localization of bone morphogenetic protein (BMP) family members and a BMP type II receptor during fracture healing in rats. Bone 22, 605 – 612 (1998). M. Komaki, T. Katagiri, and T. Suda, Bone morphogenetic protein-2 does not alter the differentiation pathway of committed progenitors of osteoblasts and chondroblasts. Cell Tissue Res. 284, 9 – 17 (1996). D. J. Rickard, T. A. Sullivan, B. J. Shenker, P. S. Leboy, and I. Kazhdan, Induction of rapid osteoblast differentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev. Biol. 161, 218 – 228 (1994). A. Yamaguchi, T. Ishizuya, N. Kintou, Y. Wada, T. Katagiri, J. M. Wozney, V. Rosen, and S. Yoshiki, Effects of BMP-2, BMP-4, and BMP-6 on osteoblastic differentiation of bone marrow-derived stromal cell lines, ST2 and MC3T3-G2/PA6. Biochem. Biophys. Res. Commun. 220, 366 – 371 (1996). E. A. Wang, D. I. Israel, S. Kelly, and D. P. Luxenberg, Bone morphogenetic protein-2 causes commitment and differentiation in C3H10T1/2 and 3T3 cells. Growth Factors 9, 57 – 71 (1993). T. Katagiri, A. Yamaguchi, T. Ikeda, S. Yoshiki, J. M. Wozney, V. Rosen, E. A. Wang, H. Tanaka, S. Omura, and T. Suda, The non-osteogenic mouse pluripotent cell line, C3H10T1/2, is induced to differentiate into osteoblastic cells by recombinant human bone morphogenetic protein-2. Biochem Biophys Res. Commun. 172, 295 – 299 (1990). T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J. M. Wozney, A. Fujisawa-Sehara, and T. Suda, Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J. Cell Biol. 127, 1755 – 1766 (1994). P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix. J. Biol. Chem. 261, 12665 – 12674 (1986). M. W. Long, J. A. Robinson, E. A. Ashcraft, and K. G. Mann, Regulation of human bone marrow-derived osteoprogenitor cells by osteogenic growth factors. J. Clin. Invest. 95, 881 – 887 (1995). L. Malaval, F. Liu, P. Roche, and J. E. Aubin, Kinetics of osteoprogenitor proliferation and osteoblast differentiation in vitro. J. Cell Biochem. 74, 616 – 627 (1999). E. A. Tonna and E. P. Cronkite, An autoradiographic study of periosteal cell proliferation with tritiated thymidine. Lab. Invest. 11, 455 – 461 (1967). B. L. Scott, Thymidine-3H electron microscope radioautography of osteogenic cells in the fetal rat. J. Cell Biol. 35, 115 – 126 (1967). M. Owen, Cell population kinetics of an osteogenic tissue. J. Cell Biol. 19, 19 – 32 (1963). S. C. Luk, C. Nopajaroonsri, and G. T. Simon, The ultrastructure of endosteum: A topographic study in young adult rabbits. J. Ultrastruct. Res. 46, 165 – 183 (1974). S. P. Bruder and A. I. Caplan, Terminal differentiation of osteogenic cells in the embryonic chick tibia is revealed by a monoclonal antibody against osteocytes. Bone 11, 189 – 198 (1990). E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodeling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption, and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990). A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, and M. Kleerekoper, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and osteoporosis. J. Clin. Invest. 72, 1396 – 1409 (1983).

57

CHAPTER 2 Osteoblast Biology 232. J. Glowacki, Influence of age on human marrow. Calcif. Tissue Int. 56, S50 – S51 (1995). 233. R. J. Bergman, D. Gazit, A. J. Kahn, H. Gruber, S. McDougall, and T. J. Hahn, Age-related changes in osteogenic stem cells in mice. J. Bone Miner. Res. 11, 568 – 577 (1996). 234. P. J. Roholl, E. Blauw, C. Zurcher, J. A. Dormans, and H. M. Theuns, Evidence for a diminished maturation of preosteoblasts into osteoblasts during aging in rats; an ultrastructural analysis. J. Bone Miner. Res. 9, 355 – 366 (1994). 235. G. D’Ippolito, P. C. Schiller, C. Ricordi, B. A. Roos, and G. A. Howard, Age-related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J. Bone Miner. Res. 14, 1115 – 1122 (1999). 236. W. E. Mercer, Checking on the cell cycle. J. Cell. Biochem. Suppl. 30 – 31, 50 – 54 (1998). 237. A. B. Pardee, G1 events and regulation of cell proliferation. Science 246, 603 – 608 (1989). 238. D. O. Morgan, Principles of CDK regulation. Nature 374, 131 – 134 (1995). 239. S. S. Millard and A. Koff, Cyclin-dependent kinase inhibitors in restriction point control, genomic stability, and tumorigenesis. J. Cell. Biochem. Suppl 30 – 31, 37 – 42 (1998). 240. G. S. Stein, T. A. Owen, A. J. van Wijnen, B. Cho, J. L. Stein, and J. B. Lian, Mechanisms of action of skeletal growth factors in osteoblasts. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 51 – 65. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 241. T. Hunter and J. Pines, Cyclins and cancer II: Cyclin D and cdk inhibitors come of age. Cell 79, 573 – 582 (1994). 242. T. K. MacLachlan, N. Sang, and A. Giordano, Cyclins, cyclin-dependent kinases and cdk inhibitors: Implications in cell cycle control and cancer. Crit. Rev. Eukaryot. Gene Expr. 5, 127 – 156 (1995). 243. L. H. Hartwell and T. A. Weinert, Checkpoints: Controls that ensure the order of cell cycle events. Science 246, 629 – 634 (1989). 244. P. Stiegler, M. Kasten, and A. Giordano, The RB family of cell cycle regulatory factors. J. Cell. Biochem. Suppl 30 – 31, 30 – 36 (1998). 245. C. Cadwell and G. P. Zambetti, Regulators and mediators of the p53 tumor suppressor. J. Cell. Biochem. Suppl 30 – 31, 43 – 49 (1998). 246. R. A. Weinberg, The retinoblastoma protein and cell cycle control. Cell 81, 323 – 330 (1995). 247. D. M. Koepp, J. W. Harper, and S. J. Elledge, How the cyclin became a cyclin: regulated proteolysis in the cell cycle. Cell 97, 431 – 434 (1999). 248. H. Kiyokawa, V. M. Richon, R. A. Rifkind, and P. A. Marks, Suppression of cyclin-dependent kinase 4 during induced differentiation of erythroleukemia cells. Mol. Cell. Biol. 14, 7195 – 7203 (1994). 249. E. Smith, B. Frenkel, R. Schlegel, A. Giordano, J. B. Lian, J. L. Stein, and G. S. Stein, Expression of cell cycle regulatory factors in differentiating osteoblasts: Postproliferative up-regulation of cyclins B and E. Cancer Res. 55, 5019 – 5024 (1995). 250. R. F. Morrison and S. R. Farmer, Role of PPARgamma in regulating a cascade expression of cyclin- dependent kinase inhibitors, p18(INK4c) and p21(Waf1/Cip1), during adipogenesis. J Biol. Chem. 274, 17088 – 17097 (1999). 251. E. G. Reynaud, K. Pelpel, M. Guillier, M. P. Leibovitch, and S. A. Leibovitch, p57(Kip2) stabilizes the MyoD protein by inhibiting cyclin E-Cdk2 kinase activity in growing myoblasts. Mol. Cell. Biol. 19, 7621 – 7629 (1999). 252. Q. Liu, R. W. VanHoy, J. H. Zhou, R. Dantzer, G. G. Freund, and K. W. Kelley, Elevated cyclin E levels, inactive retinoblastoma protein, and suppression of the p27(KIP1) inhibitor characterize early development of promyeloid cells into macrophages. Mol. Cell. Biol. 19, 6229 – 6239 (1999). 253. H. Drissi, D. Hushka, F. Aslam, Q. Nguyen, E. Buffone, A. Koff, A. van Wijnen, J. B. Lian, J. L. Stein, and G. S. Stein, The cell cycle

254.

255. 256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

267.

268.

269.

regulator p27kip 1 contributes to growth and differentiation of osteoblasts. Cancer Res. 59, 3705 – 3711 (1999). K. Yamato, S. Hashimoto, N. Okahashi, A. Ishisaki, K. Nonaka, T. Koseki, M. Kizaki, Y. Ikeda, and T. Nishihara, Dissociation of bone morphogenetic protein-mediated growth arrest and apoptosis of mouse B cells by HPV-16 E6/E7. Exp. Cell Res. 257, 198 – 205 (2000). K. Okamoto and C. Prives, A role of cyclin G in the process of apoptosis. Oncogene 18, 4606 – 4615 (1999). J. Jernvall, T. Aberg, P. Kettunen, S. Keranen, and I. Thesleff, The life history of an embryonic signaling center: BMP-4 induces p21 and is associated with apoptosis in the mouse tooth enamel knot. Development 125, 161 – 169 (1998). N. Ghosh-Choudhury, G. Ghosh-Choudhury, A. Celeste, P. M. Ghosh, M. Moyer, S. L. Abboud, and J. Kreisberg, Bone morphogenetic protein-2 induces cyclin kinase inhibitor p21 and hypophosphory lation of retinoblastoma protein in estradiol-treated MCF-7 human breast cancer cells. Biochim. Biophys. Acta 1497, 186 – 196 (2000). T. Bellido, C. A. O’Brien, P. K. Roberson, and S. C. Manolagas, Transcriptional activation of the p21(WAF1, CIP1, SD11) gene by interleukin-6 type cytokines: A prerequisite for their pro- differentiating and ant-apoptotic effects on human osteoblastic cells. J Biol. Chem. 273, 21137 – 21144 (1998). S. Walsh, C. Jefferiss, K. Stewart, G. R. Jordan, J. Screen, and J. N. Beresford, Expression of the developmental markers STRO-1 and alkaline phosphatase in cultures of human marrow stromal cells: Regulation by fibroblast growth factor (FGF)-2 and relationship to the expression of FGF receptors 1 – 4. Bone 27, 185 – 195 (2000). A. Mansukhani, P. Bellosta, M. Sahni, and C. Basilico, Signaling by fibroblast growth factors (FGF) and fibroblast growth factro receptor 2 (FGFR2)- activating mutations blocks mineralization and induces apoptosis in osteoblasts. J. Cell Biol. 149, 1297 – 1308 (2000). A. Montero, Y. Okada, M. Tomita, M. Ito, H. Tsurukami, T. Nakamura, T. Doestschman, J. D. Coffin, and M. M. Hurley, Disruption of the fibroblast growth factor-2 gene results in decreased bone mass and bone formation. J. Clin. Invest. 105, 1085 – 1093 (2000). H. Tanaka, H. Ogasa, J. Barnes, and C. T. Liang, Actions of bFGF on mitogenic activity and lineage expression in rat osteoprogenitor cells: Effect of age. Mol. Cell. Endocrinol. 150, 1 – 10 (1999). F. Debiais, M. Hott, A. M. Graulet, and P. J. Marie, The effects of fibroblast growth factor-2 on human neonatal calvaria osteoblastic cells are differentiation stage specific. J. Bone Miner. Res. 13, 645 – 654 (1998). M. R. Johnson, C. Valentine, C. Basilico, and A. Mansukhani, FGF signaling activates STAT1 and p21 and inhibits the estrogen response and proliferation of MCF-7 cells. Oncogene 16, 2647 – 2656 (1998). H. Wang, M. Rubin, E. Fenig, A. DeBlasio, J. Mendelsohn, J. Yahalom, and R. Wieder, Basic fibroblast growth factor causes growth arrest in MCF-7 human breast cancer cells while inducing both mitogenic and inhibitory G1 events. Cancer Res. 57, 1750 – 1757 (1997). K. Pardali, A. Kurisaki, A. Moren, P. T. Dijke, D. Kardassis, and A. Moustakas, Role of Smad proteins and transcription factor Sp1 in p21WAF1/Cip1 regulation by transforming growth factor-beta. J. Biol. Chem. 275, 29244 – 29256 (2000). M. B. Datto, Y. Yu, and X. F. Wang, Functional analysis of the transforming growth factor beta responsive elements in the WAF1/ Cip1/p21 promoter. J. Biol. Chem. 270, 28623 – 28628 (1995). M. Liu, M. H. Lee, M. Cohen, M. Bommakanti, and L. P. Freedman, Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev. 10, 142 – 153 (1996). J. W. Harper, G. R. Adami, N. Wei, K. Keyomarsi, and S. J. Elledge, The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin- dependent kinases. Cell 75, 805 – 816 (1993).

58 270. Y. Xiong, G. J. Hannon, H. Zhang, D. Casso, R. Kobayashi, and D. Beach, p21 is a universal inhibitor of cyclin kinases. Nature 366, 701 – 704 (1993). 271. D. A. Cameron, The Golgi apparatus in bone and cartilage cells. Clin. Orthop. 58, 191 – 211 (1968). 272. C. F. Descalzi, C. Gentili, P. Manduca, and R. Cancedda, Hypertrophic chondrocytes undergo further differentiation in culture. J. Cell Biol. 117, 427 – 435 (1992). 273. J. B. Lian, M. D. McKee, A. M. Todd, and L. C. Gerstenfeld, Induction of bone-related proteins, osteocalcin and osteopontin, and their matrix ultrastructural localization with development of chondrocyte hypertrophy in vitro. J. Cell. Biochem. 52, 206 – 219 (1993). 274. P. S. Henthorn, M. Raducha, K. N. Fedde, M. A. Lafferty, and M. P. Whyte, Different missense mutations at the tissue-nonspecific alkaline phosphatase gene locus in autosomal recessively inherited forms of mild and severe hypophosphatasia. Proc. Natl. Acad. Sci. USA 89, 9924 – 9928 (1992). 275. A. L. Boskey, Matrix proteins and mineralization: an overview. Connect. Tissue Res. 35, 357 – 363 (1996). 276. A. L. Boskey, Biomineralization: Conflicts, challenges, and opportunities. J. Cell Biochem. Suppl 30 – 31, 83 – 91 (1998). 277. A. L. Boskey, S. Gadaleta, C. Gundberg, S. B. Doty, P. Ducy, and G. Karsenty, Fourier transformed infrared microspectroscopic analysis of bones of osteocalcin deficient mice provides insight into the function of osteocalcin. Bone 23, 187 (1998). 278. P. Ducy, C. Desbois, B. Boyce, G. Pinero, B. Story, C. Dunstan, E. Smith, J. Bonadio, S. Goldstein, C. Gundberg, A. Bradley, and G. Karsenty, Increased bone formation in osteocalcin-deficient mice. Nature 382, 448 – 452 (1996). 279. H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontin-deficient mice are resistant to ovariectomy-induced bone resorption [published erratum appears in Proc. Natl. Acad. Sci. USA 96(19), 1094 (1999)]. Proc. Natl. Acad. Sci. USA 96, 8156 – 8160 (1999). 280. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 281. J. M. Hayden, S. Mohan, and D. J. Baylink, The insulin-like growth factor system and the coupling of formation to resorption. Bone 17, 93S – 98S (1995). 282. C. A. Conover, Insulin-like growth factors and the skeleton. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 101 – 116. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 283. D. M. Shinar, N. Endo, D. Halperin, G. A. Rodan, and M. Weinreb, Differential expression of insulin-like growth factor-I (IGF-I) and IFG- II messenger ribonucleic acid in growing rat bone. Endocrinology 132, 1158 – 1167 (1993). 284. T. Thomas, F. Gori, T. C. Spelsberg, S. Khosla, B. L. Riggs, and C. A. Conover, Response of bipotential human marrow stromal cells to insulin-like growth factors: Effect on binding protein production, proliferation, and commitment to osteoblasts and adipocytes. Endocrinology 140, 5036 – 5044 (1999). 285. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122, 254 – 260 (1988). 286. V. Gangji, S. Rydziel, B. Gabbitas, and E. Canalis, Insulin-like growth factor II promoter expression in cultured rodent osteoblasts and adult rat bone. Endocrinology 139, 2287 – 2292 (1998). 287. E. Canalis, S. Rydziel, A. M. Delany, S. Varghese, and J. J. Jeffrey, Insulin-like growth factors inhibit interstitial collagenase synthesis in bone cell cultures. Endocrinology 136, 1348 – 1354 (1995). 288. R. C. Pereira and E. Canalis, Parathyroid hormone increases mac25/insulin-like growth factor-binding protein-related protein-1 expression in cultured osteoblasts. Endocrinology 140, 1998 – 2003 (1999).

LIAN AND STEIN 289. R. C. Pereira, F. Blanquaert, and E. Canalis, Cortisol enhances the expression of mac25/insulin-like growth factor- binding protein-related protein-1 in cultured osteoblasts. Endocrinology 140, 228 – 232 (1999). 290. A. M. Delany, Y. Dong, and E. Canalis, Mechanisms of glucocorticoid action in bone cells. J. Cell. Biochem. 56, 295 – 302 (1994). 291. V. Hwa, Y. Oh, and R. G. Rosenfeld, The insulin-like growth factorbinding protein (IGFBP) superfamily. Endocr. Rev. 20, 761 – 787 (1999). 292. C. Richman, D. J. Baylink, K. Lang, C. Dony, and S. Mohan, Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140, 4699 – 4705 (1999). 293. B. Gabbitas and E. Canalis, Insulin-like growth factors sustain insulin-like growth factor-binding protein-5 expression in osteoblasts. Am. J. Physiol. 275, E222 – E228 (1998). 294. X. Qin, D. Byun, K. W. Lau, D. J. Baylink, and S. Mohan, Evidence that the interaction between insulin-like growth factor (IGF)- II and IGF binding protein (IGFBP)-4 is essential for the action of the IGFII-dependent IGFBP-4 protease. Arch. Biochem. Biophys. 379, 209 – 216 (2000). 295. Y. Chen, H. Shu, C. Ji, S. Casinghino, K. Kim, C. M. Gundberg, M. Centrella, and T. L. McCarthy, Insulin-like growth factor binding proteins localize to discrete cell culture compartments in periosteal and osteoblast cultures from fetal rat bone. J. Cell. Biochem. 71, 351 – 362 (1998). 296. P. Gamero, E. Sornay-Rendu, and P. D. Delmas, Low serum IGF-1 and occurrence of osteoporotic fractures in postmenopausal women. Lancet 355, 898 – 899 (2000). 297. P. Lakatos, J. Foldes, Z. Nagy, I. Takacs, G. Speer, C. Horvath, S. Mohan, D. J. Baylink, and P. H. Stern, Serum insulin-like growth factor-I, insulin-like growth factor binding proteins, and bone mineral content in hyperthyroidism. Thyroid 10, 417 – 423 (2000). 298. J. G. Kim, C. S. Shin, Y. M. Choi, S. Y. Moon, S. Y. Kim, and J. Y. Lee, The relationship among circulating insulin-like growth factor components, biochemical markers of bone turnover and bone mineral density in postmenopausal women under the age of 60. Clin. Endocrinol. (Oxford) 51, 301 – 307 (1999). 299. C. J. Rosen, Clinical aspects of the insulin-like growth factors. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 117 – 128. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 300. J. C. Adams and F. M. Watt, Regulation of development and differentiation by the extracellular matrix. Development 117, 1183 –1198 (1993). 301. C. H. Damsky, Extracellular matrix-integrin interactions in osteoblast function and tissue remodeling. Bone 25, 95 – 96 (1999). 302. M. A. Horton Interactions of connective tissue cells with the extracellular matrix. Bone 17, 51S – 53S (1995). 303. W. J. Grzesik and P. G. Robey, Bone matrix RGD glycoproteins: Immunolocalization and interaction with human primary osteoblastic bone cells in vitro. J. Bone Miner. Res. 9, 487 – 496 (1994). 304. R. J. Majeska, M. Port, and T. A. Einhorn, Attachment to extracellular matrix molecules by cells differing in the expression of osteoblastic traits. J. Bone Miner. Res. 8, 277 – 289 (1993). 305. M. P. Lynch, J. L. Stein, G. S. Stein, and J. B. Lian, The influence of type I collagen on the development and maintenance of the osteoblast phenotype in primary and passaged rat calvarial osteoblasts: Modification of expression of genes supporting cell growth, adhesion, and extracellular matrix mineralization. Exp. Cell Res. 216, 35 – 45 (1995). 306. S. Vukicevic, F. P. Luyten, H. K. Kleinman, and A. H. Reddi, Differentiation of canalicular cell processes in bone cells by basement membrane matrix components: Regulation by discrete domains of laminin. Cell 63, 437 – 445 (1990). 307. A. G. Andrianarivo, J. A. Robinson, K. G. Mann, and R. P. Tracy, Growth on type I collagen promotes expression of the osteoblastic

CHAPTER 2 Osteoblast Biology

308.

309.

310.

311.

312.

313.

314.

315.

316.

317.

318. 319.

320.

321.

322.

323.

324.

325.

phenotype in human osteosarcoma MG-63 cells. J Cell Physiol. 153, 256 – 265 (1992). R. T. Franceschi and B. S. Iyer, Relationship between collagen synthesis and expression of the osteoblast phenotype in MC3T3-E1 cells. J. Bone Miner. Res. 7, 235 – 246 (1992). S. Gronthos K. Stewart, S. E. Graves, S. Hay, and P. J. Simmons, Integrin expression and function on human osteoblast-like cells. J. Bone Miner. Res. 12, 1189 – 1197 (1997). A. M. Moursi, R. K. Globus, and C. H. Damsky, Interactions between integrin receptors and fibronectin are required for calvarial osteoblast differentiation in vitro. J. Cell Sci. 110(Pt 18), 2187 – 2196 (1997). A. M. Moursi, C. H. Damsky, J. Lull, D. Zimmerman, S. B. Doty, S. Aota, and R. K. Globus, Fibronectin regulates calvarial osteoblast differentiation. J. Cell Sci. 109(Pt 6), 1369 – 1380 (1996). R. K. Globus, S. B. Doty, J. C. Lull, E. Holmuhamedov, M. J. Humphries, and C. H. Damsky, Fibronectin is a survival factor for differentiated osteoblasts. J. Cell Sci. 111, 1385 – 1393. G. A. Gronowicz and M. E. Derome, Synthetic peptide containing Arg-Gly-Asp inhibits bone formation and resorption in a mineralizing organ culture system of fetal rat parietal bones. J. Bone Miner. Res. 9, 193 – 201 (1994). M. Mizuno, R. Fujisawa, and Y. Kuboki, Type I collagen-induced osteoblastic differentiation of bone-marrow cells mediated by collagen-alpha2beta1 integrin interaction. J. Cell. Physiol. 184, 207 – 213 (2000). M. Mizuno, T. Imai, R. Fujisawa, H. Tani, and Y. Kuboki, Bone sialoprotein (BSP) is a crucial factor for the expression of osteoblastic phenotypes of bone marrow cells cultured on type I collagen matrix. Calcif. Tissue Int. 66, 388 – 396 (2000). G. B. Schneider, S. W. Whitson, and L. F. Cooper, Restricted and coordinated expression of beta3-integrin and bone sialoprotein during cultured osteoblast differentiation. Bone 24, 321 – 327 (1999). T. V. Byzova, W. Kim, R. J. Midura, and E. F. Plow, Activation of integrin alpha(V)beta(3) regulates cell adhesion and migration to bone sialoprotein. Exp. Cell Res. 254, 299 – 308 (2000). P. Roche, H. A. Goldberg, P. D. Delmas, and L. Malaval, Selective attachment of osteoprogenitors to laminin. Bone 24, 329 – 336 (1999). A. Jikko, S. E. Harris, D. Chen, D. L. Mendrick, and C. H. Damsky, Collagen integrin receptors regulate early osteoblast differentiation induced by BMP-2. J. Bone Miner. Res. 14, 1075 – 1083 (1999). T. H. Steinberg, R. Civitelli, S. T. Geist, A. J. Robertson, E. Hick, R. D. Veenstra, H. Z. Wang, P. M. Warlow, E. M. Westphale, and J. G. Laing, Connexin43 and connexin45 form gap junctions with different molecular permeabilities in osteoblastic cells. EMBO J. 13, 744 – 750 (1994). D. T. Yamaguchi, D. Ma, A. Lee, J. Huang, and H. E. Gruber, Isolation and characterization of gap junctions in the osteoblastic MC3T3-E1 cell line. J. Bone Miner. Res. 9, 791 – 803 (1994). S. L. Ferrari, K. Traianedes, M. Thorne, M. H. Lafage-Proust, P. Genever, M. G. Cecchini, V. Behar, A. Bisello, M. Chorev, M. Rosenblatt, and L. J. Suva, A role for N-cadherin in the development of the differentiated osteoblastic phenotype. J. Bone Miner. Res. 15, 198 – 208 (2000). Y. S. Lee and C. M. Chuong, Adhesion molecules in skeletogenesis. I. Transient expression of neural cell adhesion molecules (NCAM) in osteoblasts during endochondral and intramembranous ossification. J. Bone Miner. Res. 7, 1435 – 1446 (1992). M. Okazaki, S. Takeshita, S. Kawai, R. Kikuno, A. Tsujimura, A. Kudo, and E. Amann, Molecular cloning and characterization of OB-cadherin, a new member of cadherin family expressed in osteoblasts. J Biol. Chem. 269, 12092 – 12098 (1994). C. S. Shin, F. Lecanda, S. Sheikh, L. Weitzmann, S. L. Cheng, and R. Civitelli, Relative abundance of different cadherins defines differentiation of mesenchymal precursors into osteogenic, myogenic, or adipogenic pathways. J. Cell Biochem. 78, 566 – 577 (2000).

59 326. Y. Tanaka, I. Morimoto, Y. Nakano, Y. Okada, S. Hirota, S. Nomura, T. Nakamura, and S. Eto, Osteoblasts are regulated by the cellular adhesion through ICAM-1 and VCAM-1, J. Bone Miner. Res. 10, 1462 – 1469 (1995). 327. H. Nakamura, S. Kenmotsu, H. Sakai, and H. Ozawa, Localization of CD44, the hyaluronate receptor, on the plasma membrane of osteocytes and osteoclasts in rat tibiae. Cell Tissue Res. 280, 225 – 233 (1995). 328. R. W. Young, Cell proliferation and specialization during endochondral osteogenesis in young rats. J. Cell Biol. 14, 357 (1962). 329. D. N. Menton, D. J. Simmons, S. L. Chang, and B. Y. Orr, From bone lining cell to osteocyte—an SEM study. Anat. Rec. 209, 29 – 39 (1984). 330. J. E. Puzas, The osteoblast. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), pp. 15 – 21. Raven Press, New York, 1993. 331. E. M. Aarden, E. H. Burger, and P. J. Nijweide, Function of osteocytes in bone. J. Cell. Biochem. 55, 287 – 299 (1994). 332. A. van der Plas and P. J. Nijweide, Isolation and purification of osteocytes. J. Bone Miner. Res. 7, 389 – 396 (1992). 333. S. C. Miller and W. S. Jee, The bone lining cell: A distinct phenotype? Calcif. Tissue Int. 41, 1 – 5 (1987). 334. S. B. Doty, Morphological evidence of gap junctions between bone cells. Calcif. Tissue Int. 33, 509 – 512 (1981). 335. S. C. Rawlinson, J. R. Mosley, R. F. Suswillo, A. A. Pitsillides, and L. E. Lanyon, Calvarial and limb bone cells in organ and monolayer culture do not show the same early responses to dynamic mechanical strain. J. Bone Miner. Res. 10, 1225 – 1232 (1995). 336. R. Civitelli, E. C. Beyer, P. M. Warlow, A. J. Robertson, S. T. Geist, and T. H. Steinberg, Connexin43 mediates direct intercellular communication in human osteoblastic cell networks. J. Clin. Invest. 91, 1888 – 1896 (1993). 337. H. J. Donahue, K. J. McLeod, C. T. Rubin, J. Andersen, E. A. Grine, E. L. Hertzberg, and P. R. Brink, Cell-to-cell communication in osteoblastic networks: Cell line- dependent hormonal regulation of gap junction function. J. Bone Miner. Res. 10, 881 – 889 (1995). 338. C. T. Rubin and L. E. Lanyon, Osteoregulatory nature of mechanical stimuli: Function as a determinant for adaptive bone remodeling. J. Orthop. Res. 5, 300 – 310 (1987). 339. S. Weinbaum, S. C. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading- induced bone fluid shear stresses. J. Biomech. 27, 339 – 360 (1994). 340. T. A. Einhorn, Biomechanics of bone. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 25 – 38. Academic Press, San Diego, 1996. 341. K. J. McLeod, C. T. Rubin, M. W. Otter, and Y. X. Qin, Skeletal cell stresses and bone adaptation. Am. J. Med. Sci. 316, 176 – 183 (1998). 342. J. Klein-Nulend, A. van der Plas, C. M. Semeins, N. E. Ajubi, J. A. Frangos, P. J. Nijweide, and E. H. Burger, Sensitivity of osteocytes to biomechanical stress in vitro. FASEB J. 9, 441 – 445 (1995). 343. C. H. Turner, M. R. Forwood, and M. W. Otter, Mechanotransduction in bone: Do bone cells act as sensors of fluid flow? FASEB J. 8, 875 – 878 (1994). 344. A. J. el Haj, S. L. Minter, S. C. Rawlinson, R. Suswillo, and L. E. Lanyon, Cellular responses to mechanical loading in vitro. J. Bone Miner. Res. 5, 923 – 932 (1990). 345. T. M. Skerry, L. Bitensky, J. Chayen, and L. E. Lanyon, Early strainrelated changes in enzyme activity in osteocytes following bone loading in vivo. J. Bone Miner. Res. 4, 783 – 788 (1989). 346. S. Nomura and T. Takano-Yamamoto, Molecular events caused by mechanical stress in bone. Matrix Biol. 19, 91 – 96 (2000). 347. J. M. Lean, C. J. Jagger, T. J. Chambers, and J. W. Chow, Increased insulin-like growth factor I mRNA expression in rat osteocytes in response to mechanical stimulation. Am. J. Physiol. 268, E318 – E327 (1995).

60 348. C. D. Toma, S. Ashkar, M. L. Gray, J. L. Schaffer, and L. C. Gerstenfeld, Signal transduction of mechanical stimuli is dependent on microfilament integrity: Identification of osteopontin as a mechanically induced gene in osteoblasts. J. Bone Miner. Res. 12, 1626 – 1636 (1997). 349. A. Miyauchi, K. Notoya, Y. Mikuni-Takagaki, Y. Takagi, M. Goto, Y. Miki, T. Takano-Yamamoto, K. Jinnai, K. Takahashi, M. Kumegawa, K. Chihara, and T. Fujita, Parathyroid hormone-activated volumesensitive calcium influx pathways in mechanically loaded osteocytes. J Biol. Chem. 275, 3335 – 3342 (2000). 350. T. N. McAllister, T. Du, and J. A. Frangos, Fluid shear stress stimulates prostaglandin and nitric oxide release in bone marrow-derived preosteoclast-like cells. Biochem. Biophys. Res. Commun. 270, 643 – 648 (2000). 351. S. W. Fox, T. J. Chambers, and J. W. M. Chow, Nitric oxide is an early mediator of the induction of bone formation by mechanical stimulation. J. Bone Miner. Res. 10, S201 – S201 (1995). [Abstract] 352. E. H. Burger, J. Klein-Nulend, C. M. Semeins, N. E. Ajubi, and P. J. Nijweide, Osteocytes but not periosteal fibroblasts produce nitric oxide (NO) in response to pulsatile fluid flow. Trans. Orthop. Res. Soc. 21, 531 – 531 (1996). [Abstract] 353. N. Wang, J. P. Butler, and D. E. Ingber, Mechanotransduction across the cell surface and through the cytoskeleton. Science 260, 1124 – 1127 (1993). 354. D. M. Salter, J. E. Robb, and M. O. Wright, Electrophysiological responses of human bone cells to mechanical stimulation: Evidence for specific integrin function in mechanotransduction. J. Bone Miner. Res. 12, 1133 – 1141 (1997). 355. A. L. Hitt and E. J. Luna, Membrane interactions with the actin cytoskeleton. Curr. Opin. Cell Biol. 6, 120 – 130 (1994). 356. F. M. Pavalko, N. X. Chen, C. H. Turner, D. B. Burr, S. Atkinson, Y. F. Hsieh, J. Qiu, and R. L. Duncan, Fluid shear-induced mechanical signaling in MC3T3-E1 osteoblasts requires cytoskeletonintegrin interactions. Am. J Physiol 275, C1591 – C1601 (1998). 357. M. C. Meazzini, C. D. Toma, J. L. Schaffer, M. L. Gray, and L. C. Gerstenfeld, Osteoblast cytoskeletal modulation in response to mechanical strain in vitro. J. Orthop. Res. 16, 170 – 180 (1998). 358. K. Terai, T. Takano-Yamamoto, Y. Ohba, K. Hiura, M. Sugimoto, M. Sato, H. Kawahata, N. Inaguma, Y. Kitamura, and S. Nomura, Role of osteopontin in bone remodeling caused by mechanical stress. J. Bone Miner. Res. 14, 839 – 849 (1999). 359. R. R. Miles, C. H. Turner, R. Santerre, Y. Tu, P. McClelland, J. Argot, B. S. DeHoff, C. W. Mundy, P. R. Rosteck, Jr., J. Bidwell, J. P. Sluka, J. Hock, and J. E. Onyia, Analysis of differential gene expression in rat tibia after an osteogenic stimulus in vivo: Mechanical loading regulates osteopontin and myeloperoxidase. J. Cell Biochem. 68, 355 – 365 (1998). 360. J. Rubin, T. Murphy, M. S. Nanes, and X. Fan, Mechanical strain inhibits expression of osteoclast differentiation factor by murine stromal cells. Am. J. Physiol. Cell. Physiol. 278, C1126 – C1132 (2000). 361. H. Drissi, A. Lomri, F. Lasmoles, X. Holy, E. Zerath, and P. J. Marie, Skeletal unloading induces biphasic changes in insulin-like growth factor-I mRNA levels and osteoblast activity. Exp. Cell Res. 251, 275 – 284 (1999). 362. M. Sato, T. Ochi, T. Nakase, S. Hirota, Y. Kitamura, S. Nomura, and N. Yasui, Mechanical tension-stress induces expression of bone morphogenetic protein (BMP)-2 and BMP-4, but not BMP-6, BMP-7, and GDF-5 mRNA, during distraction osteogenesis. J. Bone Miner. Res. 14, 1084 – 1095 (1999). 363. T. S. Gross, J. L. Edwards, K. J. McLeod, and C. T. Rubin, Strain gradients correlate with sites of periosteal bone formation. J. Bone Miner. Res. 12, 982 – 988 (1997). 364. R. L. Jilka, R. S. Weinstein, T. Bellido, A. M. Parfitt, and S. C. Manolagas, Osteoblast programmed cell death (apoptosis):

LIAN AND STEIN

365.

366. 367. 368.

369.

370.

371.

372.

373.

374.

375.

376.

377.

378.

379.

380.

381.

382. 383.

modulation by growth factors and cytokines. J. Bone Miner. Res. 13, 793 – 802 (1998). M. Lynch, C. Capparelli, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during bone-like tissue development in vitro. J. Cell. Biochem. 68, 31 – 49 (1998). D. P. Rice, H. J. Kim, and I. Thesleff, Apoptosis in murine calvarial bone and suture development. Eur. J Oral Sci. 107, 265 – 275 (1999). D. E. Hughes and B. F. Boyce, Apoptosis in bone physiology and disease. Mol. Pathol. 50, 132 – 137 (1997). O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Miner. Res. 15, 60 – 67 (2000). A. L. J. J. Bronckers, W. Goei, G. Luo, G. Karsenty, R. N. D’Souza, D. M. Lyaruu, and E. H. Burger, DNA fragmentation during bone formation in neonatal rodents assessed by transferase-mediated end labeling. J. Bone Miner. Res. 11, 1281 – 1291 (1996). B. S. Noble, H. Stevens, N. Loveridge, and J. Reeve, Identification of apoptotic changes in osteocytes in normal and pathological human bone. Bone 20, 273 – 282 (1997). R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Roberson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999). A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 13, 1243 – 1250 (1998). D. G. Monroe, D. F. Jin, and M. M. Sanders, Estrogen opposes the apoptotic effects of bone morphogenetic protein 7 on tissue remodeling. Mol. Cell. Biol. 20, 4626 – 4634 (2000). T. Bellido, M. Huening, M. Raval-Pandya, S. C. Manolagas, and S. Christakos, Calbindin-D28K is expressed in osteoblastic cells and suppresses their apoptosis by inhibiting caspase-3 activity. J Biol. Chem. 275, 26328 – 26332 (2000). W. J. Sharrock, Bone and the hematopoietic and immune systems: A report of the proceedings of a scientific workshop. J. Bone Miner. Res. 13, 537 – 543 (1998). R. S. Taichman and S. G. Emerson, Human osteoblasts support hematopoiesis through the production of granulocyte colonystimulating factor. J. Exp. Med. 179, 1677 – 1682 (1994). H. A. Kodama, Y. Amagai, H. Koyama, and S. Kasai, A new preadipose cell line derived from newborn mouse calvaria can promote the proliferation of pluripotent hemopoietic stem cells in vitro. J. Cell. Physiol. 112, 89 – 95 (1982). P. J. Simmons, B. Masinovsky, B. M. Longenecker, R. Berenson, B. Torok-Storb, and W. M. Gallatin, Vascular cell adhesion molecule-1 expressed by bone marrow stromal cells mediates the binding of hematopoietic progenitor cells. Blood 80, 388 – 395 (1992). R. S. Taichman and S. G. Emerson, The role of osteoblasts in the hematopoietic microenvironment. Stem Cells 16, 7 – 15 (1998). G. Fiorelli, C. Orlando, S. Benvenuti, F. Franceschelli, S. Bianchi, P. Pioli, A. Tanini, M. Serio, F. Bartucci, and M. L. Brandi, Characterization, regulation, and function of specific cell membrane receptors for insulin-like growth factor I on bone endothelial cells. J. Bone Miner. Res. 9, 329 – 337 (1994).

CHAPTER 2 Osteoblast Biology 384. M. M. Deckers, M. Karperien, B. C. van der, T. Yamashita, S. E. Papapoulos, and C. W. Lowik, Expression of vascular endothelial growth factors and their receptors during osteoblast differentiation. Endocrinology 141, 1667 – 1674 (2000). 385. D. S. Wang, M. Miura, H. Demura, and K. Sato, Anabolic effects of 1,25-dihydroxyvitamin D3 on osteoblasts are enhanced by vascular endothelial growth factor produced by osteoblasts and by growth factors produced by endothelial cells. Endocrinology 138, 2953 – 2962 (1997). 386. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999). 387. A. Bellahcene, K. Bonjean, B. Fohr, N. S. Fedarko, F. A. Robey, M. F. Young, L. W. Fisher, and V. Castronovo, Bone sialoprotein mediates human endothelial cell attachment and migration and promotes angiogenesis. Circ. Res. 86, 885 – 891 (2000). 388. T. Suda, I. Nakamura, E. Jimi, and N. Takahashi, Regulation of osteoclast function. J. Bone Miner. Res. 12, 869 – 879 (1997). 389. G. D. Roodman, Advances in bone biology: The osteoclast. Endocr. Rev. 17, 308 – 332 (1996). 390. A. van der Plas, E. M. Aarden, J. H. Feijen, A. H. de Boer, A. Wiltink, M. J. Alblas, L. de Leij, and P. J. Nijweide, Characteristics and properties of osteocytes in culture. J. Bone Miner. Res. 9, 1697 – 1704 (1994). 391. G. Boivin, P. Mesguish, J. W. Pike, R. Bouillon, P. J. Meunier, M. R. Haussler, P. M. Dubois, and G. Morel, Ultrastructural immunolocalization of endogenous 1,25-dihydroxyvitamin D3 and its receptor in osteoblasts and osteocytes from neonatal mouse and rat calvaria. Bone Miner. 3, 125 – 136 (1987). 392. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 393. D. T. Baran, J. M. Quail, R. Ray, J. Leszyk, and T. Honeyman, Annexin II is the membrane receptor that mediates the rapid actions of 1 alpha, 25-dihydroxyvitamin D3. J. Cell. Biochem. 78, 34 – 46 (2000). 394. M. Lieberherr, Effects of vitamin D3 metabolites on cytosolic free calcium in confluent mouse osteoblasts. J. Biol. Chem. 262, 13168 – 13173 (1987). 395. R. Civitelli, Y. S. Kim, S. L. Gunsten, A. Fujimori, M. Huskey, L. V. Avioli, and K. A. Hruska, Nongenomic activation of the calcium message system by vitamin D metabolites in osteoblast-like cells. Endocrinology 127, 2253 – 2262 (1990). 396. D. T. Baran, A. M. Sorensen, V. Shalhoub, T. Owen, A. Oberdorf, G. Stein, and J. Lian, 1 Alpha, 25-dihydroxyvitamin D3 rapidly increases cytosolic calcium in clonal rat osteosarcoma cells lacking the vitamin D receptor. J. Bone Miner. Res. 6, 1269 – 1275 (1991). 397. C. Menaa, R. D. Devlin, S. V. Reddy, Y. Gazitt, S. J. Choi, and G. D. Roodman, Annexin II increases osteoclast formation by stimulating the proliferation of osteoclast precursors in human marrow cultures. J. Clin. Invest. 103, 1605 – 1613 (1999). 398. D. Modrowski, A. Lomri, and P. J. Marie, Endogenous GM-CSF is involved as an autocrine growth factor for human osteoblastic cells. J. Cell. Physiol. 170, 35 – 46 (1997). 399. M. Alsina and G. D. Roodman, Nonskeletal effects of cytokines secreted by bone cells. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 375 – 384. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 400. S. L. Teitelbaum, M. M. Tondravi, and F. P. Ross, Osteoclast biology. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.), pp. 61 – 94. Academic Press, San Diego, 1996. 401. P. A. Hill, A. Tumber, S. Papaioannou, and M. C. Meikle, The cellular actions of interleukin-11 on bone resorption in vitro. Endocrinology 139, 1564 – 1572 (1998).

61 402. D. Cheleuitte, S. Mizuno, and J. Glowacki, In vitro secretion of cytokines by human bone marrow: Effects of age and estrogen status. J. Clin. Endocrinol. Metab. 83, 2043 – 2051 (1998). 403. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 404. S. V. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 1 – 17 (1998). 405. S. L. Teitelbaum, M. M. Tondravi, and F. P. Ross, Osteoclasts, macrophages, and the molecular mechanisms of bone resorption. J. Leukocyle. Biol. 61, 381 – 388 (1997). 406. K. Insogna, S. Tanaka, L. Neff, W. Horne, J. Levy, and R. Baron, Role of c-Src in cellular events associated with colony-stimulating factor-1-induced spreading in osteoclasts. Mol. Reprod. Dev. 46, 104 – 108 (1997). 407 .W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 408. N. Yamamoto, S. Akiyama, T. Katagiri, M. Namiki, T. Kurokawa, and T. Suda, Smad1 and smad5 act downstream of intracellular signalings of BMP-2 that inhibits myogenic differentiation and induces osteoblast differentiation in C2C12 myoblasts. Biochem. Biophys. Res. Commun. 238, 574 – 580 (1997). 409. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 410. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 411. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 412. N. Udagawa, N. Takahashi, T. Akatsu, H. Tanaka, T. Sasaki, T. Nishihara, T. Koga, T. J. Martin, and T. Suda, Origin of osteoclasts: Mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc. Natl. Acad. Sci. USA 87, 7260 – 7264 (1990). 413. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 414. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 415. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998).

62 416. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 417. B. R. Wong, R. Josien, S. Y. Lee, B. Sauter, H. L. Li, R. M. Steinman, and Y. Choi, TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J. Exp. Med. 186, 2075 – 2080 (1997). 418. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tometsko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390, 175 – 179 (1997). 419. Proposed standard nomenclature for new tumor necrosis factor family members involved in the regulation of bone resorption. The American Society for Bone and Mineral Research President’s Committee on Nomenclature. J. Bone Miner. Res. 15, 2293 – 2296 (2000). 420. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 421. A. E. Hughes, S. H. Ralston, J. Marken, C. Bell, H. MacPherson, R. G. Wallace, W. van Hul, M. P. Whyte, K. Nakatsuka, L. Hovy, and D. M. Anderson, Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nature Genet. 24, 45 – 48 (2000). 422. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 423. K. Thirunavukkarasu, D. L. Halladay, R. R. Miles, X. Yang, R. J. Galvin, S. Chandrasekhar, T. J. Martin, and J. E. Onyia, The osteoblast-specific transcription factor Cbfa1 regulates the expression of osteoprotegerin (OPG), a potent inhibitor of osteoclast differentiation and function. J. Biol. Chem. 275, 25163 – 25172 (2000). 424. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1999). 425. H. Drissi, Q. Luc, R. Shakoori, S. Chuva de Sousa Lopes, J.-Y. Choi, A. Terry, M. Hu, S. Jones, J. C. Neil, J. B. Lian, J. L. Stein, A. J. Van Wijnen, and G. S. Stein, Transcriptional autoregulation of the bone related CBFA1/RUNX2 gene. J. Cell. Physiol. 184, 341 – 350 (2000). 426. Y. H. Gao, T. Shinki, T. Yuasa, H. Kataoka-Enomoto, T. Komori, T. Suda, and A. Yamaguchi, Potential role of Cbfa1, an essential transcriptional factor for osteoblast differentiation, in osteoclastogenesis: regulation of mRNA expression of osteoclast differentiation factor (ODF). Biochem. Biophys. Res. Commun. 252, 697 – 702 (1998). 427. A. Erlebacher, E. H. Filvaroff, J. Q. Ye, and R. Derynck, Osteoblastic responses to TGF-beta during bone remodeling. Mol. Biol. Cell 9, 1903 – 1918 (1998). 428. M. E. Jackson, V. Shalhoub, J. B. Lian, G. S. Stein, and S. C. Marks, Jr., Aberrant gene expression in cultured mammalian bone cells demonstrates an osteoblast defect in osteopetrosis. J. Cell. Biochem. 55, 366 – 372 (1994). 429. V. Shalhoub, M. E. Jackson, C. Paradise, G. S. Stein, J. B. Lian, and S. C. Marks, Jr., Heterogeneity of colony stimulating factor-1 gene expression in the skeleton of four osteopetrotic mutations in rats and mice. J. Cell. Physiol. 166, 340 – 350 (1996).

LIAN AND STEIN 430. K. Aoki, E. Didomenico, N. A. Sims, K. Mukhopadhyay, L. Neff, A. Houghton, M. Amling, J. B. Levy, W. C. Horne, and R. Baron, The tyrosine phosphatase SHP-1 is a negative regulator of osteoclastogenesis and osteoclast resorbing activity: Increased resorption and osteopenia in me(v)/me(v) mutant mice. Bone 25, 261 – 267 (1999). 431. W. A. Peck, S. J. Birge, and S. A. Fedak, Bone cells: Biochemical and biological studies after enzymatic isolation. Science 146, 1476 – 1477 (1964). 432. G. L. Wong and D. V. Cohn, Target cells in bone for parathormone and calcitonin are different: enrichment for each cell type by sequential digestion of mouse calvaria and selective adhesion to polymeric surfaces. Proc. Natl. Acad. Sci. USA 72, 3167 – 3171 (1975). 433. J. E. Aubin, J. N. M. Heersche, M. J. Merriless, and J. Sodek, Isolation of bone cell clones with differences in growth, hormone responses and extracellular matrix production. J. Cell Biol. 92, 452 – 461 (1982). 434. P. G. Robey and J. D. Termine, Human bone cells in vitro. Calcif. Tissue Int. 37, 453 – 460 (1985). 435. P. J. Marie, Human osteoblastic cells: A potential tool to assess the etiology of pathologic bone formation. J. Bone Miner. Res. 9, 1847 – 1850 (1994). 436. L. Malaval, D. Modrowski, A. K. Gupta, and J. E. Aubin, Cellular expression of bone-related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures. J. Cell Physiol. 158, 555 – 572 (1994). 437. S.-L. Cheng, J. W. Yang, L. Rifas, S.-F. Zhang, and L. V. Avioli, Differentiation of human bone marrow osteogenic stromal cell in vitro: Induction of the osteoblast phenotype by dexamethasone. Endocrinology 134, 277 – 286 (1994). 438. B. A. Ashton, C. C. Eaglesom, I. Bab, and M. E. Owen, Distribution of fibroblastic colony-forming cells in rabbit bone marrow and assay of their osteogenic potential by an in vivo diffusion chamber method. Calcif. Tissue Int. 36, 83 – 86 (1984). 439. P. S. Leboy, J. N. Beresford, C. Devlin, and M. E. Owen, Dexamethasone induction of osteoblast mRNAs in rat marrow stromal cell cultures. J. Cell. Physiol. 146, 370 – 378 (1991). 440. G. E. Schoeters, L. Saint-Georges, H. R. Van den, and O. Vanderborght, Mineralization of adult mouse bone marrow in vitro. Cell Tissue Kinet. 21, 363 – 374 (1988). 441. H. Nakahara, J. E. Dennis, S. P. Bruder, S. E. Haynesworth, D. P. Lennon, and A. I. Caplan, In vitro differentiation of bone and hypertrophic cartilage from periosteal-derived cells. Exp. Cell Res. 195, 492 – 503 (1991). 442. Y. Koshihara, M. Hirano, M. Kawamura, H. Oda, and S. Higaki, Mineralization ability of cultured human osteoblast-like periosteal cells does not decline with aging. J. Gerontol. 46, B201 – B206 (1991). 443. B. Ecarot-Charrier, F. H. Glorieux, M. van der Rest, and G. Pereira, Osteoblasts isolated from mouse calvaria initiate matrix mineralization in culture. J. Cell. Biol. 96, 639 – 643 (1983). 444. J. R. Nefussi, M. L. Boy-Lefevre, H. Boulekbache, and N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion. Differentiation 29, 160 – 168 (1985). 445. U. Bhargava, M. Bar-Lev, C. G. Bellows, and J. E. Aubin, Ultrastructural analysis of bone nodules formed in vitro by isolated fetal rat calvaria cells. Bone 9, 155 – 163 (1988). 446. L. C. Gerstenfeld, S. D. Chipman, J. Glowacki, and J. B. Lian, Expression of differentiated function by mineralizing cultures of chicken osteoblasts. Dev. Biol. 122, 49 – 60 (1987). 447. S. W. Whitson, W. Harrison, M. K. Dunlap, D. E. Bowers, Jr., L. W. Fisher, P. G. Robey, and J. D. Termine, Fetal bovine cells synthesize bone-specific matrix proteins. J. Cell Biol. 99, 607 – 614 (1984). 448. L. G. Jenis, B. Ongphiphadhanakul, L. E. Braverman, G. S. Stein, J. B. Lian, R. Lew, and D. T. Baran, Responsiveness of gene expression markers of osteoblastic and osteoclastic activity to calcitonin in

CHAPTER 2 Osteoblast Biology

449.

450.

451.

452.

453.

454.

455.

456.

457.

458.

459.

460.

461.

462.

463.

464.

the appendicular and axial skeleton of the rat in vivo. Calcif. Tissue Int. 54, 511 – 515 (1994). P. J. Marie, M. C. de Vernejoul, D. Connes, and M. Hott, Decreased DNA synthesis by cultured osteoblastic cells in eugonadal osteoporotic men with defective bone formation. J. Clin. Invest 88, 1167 – 1172 (1991). B. G. Mills, F. R. Singer, L. P. Weiner, and P. A. Holst, Long-term culture of cells from bone affected by Paget’s disease. Calcif. Tissue Int. 29, 79 – 87 (1979). J. E. Puzas, J. S. Brand, and K. V. Jackman, Isolation and characterization of bone cells from a patient with osteogenesis imperfecta type 1b. Bone Miner. 1, 373 – 382 (1986). C. Silve, B. Grosse, C. Tau, M. Garabedian, J. Fritsch, P. D. Delmas, G. Cournot-Witmer, and S. Balsan, Response to parathyroid hormone and 1,25-dihydroxyvitamin D3 of bone- derived cells isolated from normal children and children with abnormalities in skeletal development. J Clin. Endocrinol. Metab. 62, 583 – 590 (1986). M. Kassem, K. Brixen, L. Mosekilde, and E. F. Eriksen, Human marrow stromal osteoblast-like cells do not show reduced responsiveness to in vitro stimulation with growth hormone in patients with postmenopausal osteoporosis. Calcif. Tissue Int. 54, 1 – 6 (1994). M. M. Wong, L. G. Rao, H. Ly, L. Hamilton, S. Ish-Shalom, W. Sturtridge, J. Tong, R. McBroom, R. G. Josse, and T. M. Murray, In vitro study of osteoblastic cells from patients with idiopathic osteoporosis and comparison with cells from nonosteoporotic controls. Osteoporos. Int. 4, 21 – 31 (1994). P. J. Marie, M. C. de Vernejoul, and A. Lomri, Stimulation of bone formation in osteoporosis patients treated with fluoride associated with increased DNA synthesis by osteoblastic cells in vitro. J Bone Miner. Res. 7, 103 – 113 (1992). J. B. Kopp and P. G. Robey, Sodium fluoride does not increase human bone cell proliferation or protein synthesis in vitro. Calcif. Tissue Int. 47, 221 – 229 (1990). K. W. Ng, P. R. Gummer, V. P. Michelangeli, J. F. Bateman, T. Mascara, W. G. Cole, and T. J. Martin, Regulation of alkaline phosphatase expression in a neonatal rat clonal calvarial cell strain by retinoic acid. J. Bone Miner. Res. 3, 53 – 61 (1988). J. E. Aubin, I. Tertinegg, R. Ber, and J. N. Heersche, Consistent patterns of changing hormone responsiveness during continuous culture of cloned rat calvaria cells. J. Bone Miner. Res. 3, 333 – 339 (1988). A. Yamaguchi and A. J. Kahn, Clonal osteogenic cell lines express myogenic and adipocytic developmental potential. Calcif. Tissue Int. 49, 221 – 225 (1991). M. Namiki, S. Akiyama, T. Katagiri, A. Suzuki, N. Ueno, N. Yamaji, V. Rosen, J. M. Wozney, and T. Suda, A kinase domain-truncated type I receptor blocks bone morphogenetic protein-2-induced signal transduction in C2C12 myoblasts. J. Biol. Chem. 272, 22046 – 22052 (1997). M. Inada, T. Katagiri, S. Akiyama, M. Namika, M. Komaki, A. Yamaguchi, K. Kamoi, V. Rosen, and T. Suda, Bone morphogenetic protein-12 and -13 inhibit terminal differentiation of myoblasts, but do not induce their differentiation into osteoblasts. Biochem. Biophys. Res. Commun. 222, 317 – 322 (1996). D. Gazit, R. Ebner, A. J. Kahn, and R. Derynck, Modulation of expression and cell surface binding of members of the transforming growth factor-beta superfamily during retinoic acid-induced osteoblastic differentiation of multipotential mesenchymal cells. Mol. Endocrinol. 7, 189 – 198 (1993). N. Udagawa, N. Takahashi, T. Akatsu, T. Sasaki, A. Yamaguchi, H. Kodama, T. J. Martin, and T. Suda, The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with mouse spleen cells. Endocrinology 125, 1805 – 1813 (1989). H. Kodama, Y. Amagai, H. Sudo, S. Kasai, and S. Yamamoto, Establishment of a clonal osteogenic cell line from newborn mouse calvaria. Jpn. J. Oral Biol. 23, 899 – 901 (1981).

63 465. J.-Y. Choi, B.-H. Lee, K.-B. Song, R.-W. Park, I.-S. Kim, K.-Y. Sohn, J.-S. Jo, and H.-M. Ryoo, Expression patterns of bone-related proteins during osteoblastic differentiation in MC3T3-E1 cells. J. Cell. Biochem. 61, 609 – 618 (1996). 466. H. Sudo, H.-A. Kodama, Y. Amagai, S. Yamamoto, and S. Kasai, In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96, 191 – 198 (1983). 467. J. K. Heath, S. B. Rodan, K. Yoon, and G. A. Rodan, Rat calvarial cell lines immortalized with SV-40 large T antigen: Constitutive and retinoic acid-inducible expression of osteoblastic features. Endocrinology 124, 3060 – 3068 (1989). 468. E. Mathieu, G. Schoeters, P. F. vander, and J. Merregaert, Establishment of an osteogenic cell line derived from adult mouse bone marrow stroma by use of a recombinant retrovirus. Calcif. Tissue Int. 50, 362 – 371 (1992). 469. A. Houghton, B. O. Oyajobi, G. A. Foster, R. G. Russell, and B. M. Stringer, Immortalization of human marrow stromal cells by retroviral transduction with a temperature sensitive oncogene: Identification of bipotential precursor cells capable of directed differentiation to either an osteoblast or adipocyte phenotype. Bone 22, 7 – 16 (1998). 470. B. S. Komm, S. Wang, J. G. Haddad, and C. R. Lyttle, Characterization of conditionally transformed osteoblast-like cells derived from rat calvaria. J. Bone Miner. Res. 8, S359 – S359 (1993). [Abstract] 471. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7, 127 – 136 (1992). 472. K. C. Hicok, T. Thomas, F. Gori, D. J. Rickard, T. C. Spelsberg, and B. L. Riggs, Development and characterization of conditionally immortalized osteoblast precursor cell lines from human bone marrow stroma. J. Bone Miner. Res. 13, 205 – 217 (1998). 473. Y. Kato, J. J. Windle, B. A. Koop, G. R. Mundy, and L. F. Bonewald, Establishment of an osteocyte-like cell line, MLO-Y4. J. Bone Miner. Res. 12, 2014 – 2023 (1997). 474. R. J. Majeska, S. B. Rodan, and G. A. Rodan, Maintenance of parathyroid hormone response in clonal rat osteosarcoma lines. Exp. Cell Res. 111, 465 – 468 (1978). 475. S. K. Nishimoto and P. A. Price, Secretion of the vitamin K-dependent protein of bone rat osteosarcoma cells: Evidence for an intracellular precursor. J. Biol. Chem. 255, 6579 – 6583 (1980). 476. J. M. Nelissen, R. Torensma, M. Pluyter, G. J. Adema, R. A. Raymarkers, Y. van Kooyk, and C. G. Figdor, Molecular analysis of the hematopoiesis supporting osteoblastic cell line U2-OS. Exp. Hematol. 28, 422 – 432 (2000). 477. G. A. Rodan and M. Noda, Gene expression in osteoblastic cells. Crit. Rev. Eukary. Gene Expr. 1, 85 – 98 (1990). 478. G. S. Stein, J. B. Lian, T. A. Owen-Hughes, J. Holthuis, R. Bortell, and A. J. van Wijnen, Molecular mechanisms that mediate a functional relationship between proliferation and differentiation. In “Molecular and Cellular Approaches to the Control of Proliferation and Differentiation” (G. S. Stein and J. B. Lian, eds.), pp. 299 – 341. Academic Press, Orlando, 1992. 479. R. Bortell, T. A. Owen, V. Shalhoub, A. Heinrichs, M. A. Aronow, C. Rochette-Egly, Y. Lutz, J. L. Stein, J. B. Lian, and G. S. Stein, Constitutive transcription of the osteocalcin gene in osteosarcoma cells is reflected by altered protein-DNA interactions at promoter regulatory elements. Proc. Natl. Acad. Sci. USA 90, 2300 – 2304 (1993). 480. J. Holthuis, T. A. Owen, A. J. van Wijnen, K. L. Wright, A. RamseyEwing, M. B. Kennedy, R. Carter, S. C. Cosenza, K. J. Soprano, J. B. Lian et al., Tumor cells exhibit deregulation of the cell cycle histone gene promoter factor HiNF-D. Science 247, 1454 – 1457 (1990).

64 481. F. M. van den Ent, A. J. van Wijnen, T. J. Last, R. Bortell, J. L. Stein, J. B. Lian, and G. S. Stein, Concerted control of multiple histone promoter factors during cell density inhibition of proliferation in osteosarcoma cells: Reciprocal regulation of cell cycle-controlled and bone-related genes. Cancer Res. 53, 2399 – 2409 (1993). 482. J. B. Lian, G. S. Stein, J. L. Stein, R. Bortell, T. A. Owen, V. Shalhoub, A. Heinrichs, C. W. van Oordt, and A. J. van Wijnen, Implications of abrogated growth control on expression of a cell cycleregulated histone and the osteoblast-specific osteocalcin genes in osteosarcoma cells. In “Frontiers of Osteosarcoma Research: Interdisciplinary Survey of Clinical and Research Advances” (J. F. Novak and J. H. McMaster, eds.), pp. 227 – 287. Hogrefe & Huber, Seattle, WA, 1993. 483. T. A. Owen, M. Aronow, V. Shalhoub, L. M. Barone, L. Wilming, M. S. Tassinari, M. B. Kennedy, S. Pockwinse, J. B. Lian, and G. S. Stein, Progressive development of the rat osteoblast phenotype in vitro: Reciprocal relationships in expression of genes associated with osteoblast proliferation and differentiation during formation of the bone extracellular matrix. J. Cell. Physiol. 143, 420 – 430 (1990). 484. L. D. Quarles, D. A. Yohay, L. W. Lever, R. Caton, and R. J. Wenstrup, Distinct proliferative and differentiated stages of murine MC3T3-E1 cells in culture: An in vitro model of osteoblast development. J. Bone Miner. Res. 7, 683 – 692 (1992). 485. K. Ibaraki, J. D. Termine, S. W. Whitson, and M. F. Young, Bone matrix mRNA expression in differentiating fetal bovine osteoblasts. J. Bone Miner. Res. 7, 743 – 754 (1992). 486. G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation/differentiation interrelationships during progressive development of the osteoblast phenotype. Endocr. Rev. 14, 424 – 442 (1993). 487. T. A. Owen, M. S. Aronow, L. M. Barone, B. Bettencourt, G. S. Stein, and J. B. Lian, Pleiotropic effects of vitamin D on osteoblast gene expression are related to the proliferative and differentiated state of the bone cell phenotype: Dependency upon basal levels of gene expression, duration of exposure, and bone matrix competency in normal rat osteoblast cultures. Endocrinology 128, 1496 – 1504 (1991). 488. H. Ishida, C. G. Bellows, J. E. Aubin, and J. N. Heersche, Characterization of the 1,25-(OH)2D3-induced inhibition of bone nodule formation in long-term cultures of fetal rat calvaria cells. Endocrinology 132, 61 – 66 (1993). 489. M. Broess, A. Riva, and L. C. Gerstenfeld, Inhibitory effects of 1,25(OH)2 vitamin D3 on collagen type I, osteopontin, and osteocalcin gene expression in chicken osteoblasts. J Cell. Biochem. 57, 440 – 451 (1995). 490. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Effects of transforming growth factor beta and epidermal growth factor on cell proliferation and the formation of bone nodules in isolated fetal rat calvaria cells. J. Cell. Physiol. 140, 386 – 395 (1989). 491. E. C. Breen, R. A. Ignotz, L. McCabe, J. L. Stein, G. S. Stein, and J. B. Lian, TGF beta alters growth and differentiation related gene expression in proliferating osteoblasts in vitro, preventing development of the mature bone phenotype. J. Cell. Physiol. 160, 323 – 335 (1994). 492. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. Dallas, J. Q. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of transforming growth factor beta on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, alkaline phosphatase, and type I collagen mRNA in long-term cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 493. C. G. Bellows, J. E. Aubin, and J. N. M. Heersche, Physiological concentrations of glucocorticoids stimulate formation of bone nodules from isolated rat calvaria cells in vitro. Endocrinology 121, 1985 – 1992 (1987).

LIAN AND STEIN 494. V. Shalhoub, D. Conlon, M. Tassinari, C. Quinn, N. Partridge, G. S. Stein, and J. B. Lian, Glucocorticoids promote development of the osteoblast phenotype by selectively modulating expression of cell growth and differentiation associated genes. J. Cell. Biochem. 50, 425 – 440 (1992). 495. E. Canalis and J. B. Lian, 1,25-Dihydroxyvitamin D3 effects on collagen and DNA synthesis in periosteum and periosteum-free calvariae. Bone 6, 457 – 460 (1985). 496. J. R. Harrison, D. N. Petersen, A. C. Lichtler, A. T. Mador, D. W. Rowe, and B. E. Kream, 1,25-Dihydroxyvitamin D3 inhibits transcription of type I collagen genes in the rat osteosarcoma cell line ROS 17/2.8. Endocrinology 125, 327 – 333 (1986). 497. M. A. Aronow, L. C. Gerstenfeld, T. A. Owen, M. S. Tassinari, G. S. Stein, and J. B. Lian, Factors that promote progressive development of the osteoblast phenotype in cultured fetal rat calvaria cells. J. Cell. Physiol. 143, 213 – 221 (1990). 498. G. S. Stein, J. B. Lian, and T. A. Owen, Relationship of cell growth to the regulation of tissue- specific gene expression during osteoblast differentiation. FASEB J. 4, 3111 – 3123 (1990). 499. G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation-differentiation interrelationships during progressive development of the osteoblast phenotype: Update 1995. Endocr. Rev. 4, 290 – 297 (1995). 500. N. Ghosh-Choudhury, M. A. Harris, J. Q. Feng, G. R. Mundy, and S. E. Harris, Expression of the BMP 2 gene during bone cell differentiation. Crit. Rev. Eukaryot. Gene Expr. 4, 345 – 355 (1994). 501. F. J. Hughes, J. Collyer, M. Stanfield, and S. A. Goodman, The effects of bone morphogenetic protein-2, -4, and -6 on differentiation of rat osteoblast cells in vitro. Endocrinology 136, 2671 – 2677 (1995). 502. S. D. Boden, G. Hair, L. Titus, M. Racine, K. McCuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 503. R. C. Pereira, S. Rydziel, and E. Canalis, Bone morphogenetic protein-4 regulates its own expression in cultured osteoblasts. J. Cell Physiol. 182, 239 – 246 (2000). 504. D. Chen, M. A. Harris, G. Rossini, C. R. Dunstan, S. L. Dallas, J. Q. Feng, G. R. Mundy, and S. E. Harris, Bone morphogenetic protein 2 (BMP-2) enhances BMP-3, BMP-4, and bone cell differentiation marker gene expression during the induction of mineralized bone matrix formation in cultures of fetal rat calvarial osteoblasts. Calcif. Tissue Int. 60, 283 – 290 (1997). 505. L. C. Gerstenfeld, A. Riva, K. Hodgens, D. R. Eyre, and W. J. Landis, Posttranslational control of collagen fibrillogenesis in mineralizing cultures of chick osteoblasts. J. Bone Miner. Res. 8, 1031 – 1043 (1993). 506. J. B. Lian, G. S. Stein, R. Bortell, and T. A. Owen, Phenotype suppression: A postulated molecular mechanism for mediating the relationship of proliferation and differentiation by Fos/Jun interactions at AP-1 sites in steroid responsive promoter elements of tissuespecific genes. J. Cell. Biochem. 45, 9 – 14 (1991). 507. L. R. McCabe, M. Kockx, J. Lian, J. Stein, and G. Stein, Selective expression of fos- and jun-related genes during osteoblast proliferation and differentiation. Exp. Cell Res. 218, 255 – 262 (1995). 508. S. R. Siddhanti and L. D. Quarles, Molecular to pharmacologic control of osteoblast proliferation and differentiation. J. Cell. Biochem. 55, 310 – 320 (1994). 509. M. Tamura and M. Noda, Identification of a DNA sequence involved in osteoblast-specific gene expression via interaction with helixloop-helix (HLH)-type transcription factors. J. Cell Biol. 126, 773 – 782 (1994). 510. M. P. Lynch, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during in vitro bone formation. J. Bone Miner. Res. 9, S352 – S352 (1994). [Abstract]

65

CHAPTER 2 Osteoblast Biology 511. S. K. Winchester, S. R. Bloch, G. J. Fiacco, and N. C. Partridge, Regulation of expression of collagenase-3 in normal, differentiating rat osteoblasts. J. Cell Physiol. 181, 479 – 488 (1999). 512. R. S. Birnbaum and D. M. Wiren, Changes in insulin-like growth factor-binding protein expression and secretion during the proliferation, differentiation, and mineralization of primary cultures of rat osteoblasts. Endocrinology 135, 223 – 230 (1994). 513. A. R. Shakoori, A. M. Oberdorf, T. A. Owen, L. A. Weber, E. Hickey, J. L. Stein, J. B. Lian, and G. S. Stein, Expression of heat shock genes during differentiation of mammalian osteoblasts and promyelocytic leukemia cells. J. Cell. Biochem. 48, 277 – 287 (1992). 514. H. M. Hoffmann, K. M. Catron, A. J. van Wijnen, L. R. McCabe, J. B. Lian, G. S. Stein, and J. L. Stein, Transcriptional control of the tissue-specific, developmentally regulated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc. Natl. Acad. Sci. USA 91, 12887 – 12891 (1994). 515. Y. Takeuchi, S. Fukumoto, K. Nakayama, and T. Matsumoto, Decrease in transforming growth factor- receptors by differentiation of osteoblastic cells is dependent upon collagen production. J. Bone Miner. Res. 9, S126 – S126 (1994). [Abstract] 516. L. M. Barone, T. A. Owen, M. S. Tassinari, R. Bortell, G. S. Stein, and J. B. Lian, Developmental expression and hormonal regulation of the rat matrix Gla protein (MGP) gene in chondrogenesis and osteogenesis. J. Cell. Biochem. 46, 351 – 365 (1991). 517. N. Kawaguchi, H. F. DeLuca, and M. Noda, Id gene expression and its suppression by 1,25-dihydroxyvitamin D3 in rat osteoblastic osteosarcoma cells. Proc. Natl. Acad. Sci. USA 89, 4569 – 4572 (1992). 518. K.-T. Tang, C. Capparelli, J. L. Stein, G. S. Stein, J. B. Lian, A. C. Huber, L. E. Braverman, and W. J. DeVito, Acidic fibroblast growth factor inhibits osteoblast differentiation in vitro: Altered expression of collagenase, cell growth-related and mineralization-associated genes. J. Cell. Biochem. 61, 152 – 166 (1996). 519. V. Shalhoub, F. Aslam, E. Breen, A. van Wijnen, R. Bortell, G. S. Stein, J. L. Stein, and J. B. Lian, Multiple levels of steroid hormonedependent control of osteocalcin during osteoblast differentiation: Glucocorticoid regulation of basal and vitamin D stimulated gene expression. J. Cell. Biochem. 69, 154 – 168 (1998). 520. L. C. Gerstenfeld, D. Zurakowski, J. L. Schaffer, D. P. Nichols, C. D. Toma, M. Broess, S. P. Bruder, and A. I. Caplan, Variable hormone responsiveness of osteoblast populations isolated at different stages of embryogenesis and its relationship to the osteogenic lineage. Endocrinology 137, 3957 – 3968 (1996). 521. P. V. N. Bodine, R. A. Henderson, J. Green, M. Aronow, T. Owen, G. S. Stein, J. B. Lian, and B. S. Komm, Estrogen receptor- is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139, 2048 – 2057 (1998). 522. T. Ishizuya, S. Yokose, M. Hori, T. Noda, T. Suda, S. Yoshiki, and A. Yamaguchi, Parathyroid hormone exerts disparate effects on osteoblast differentiation depending on exposure time in rat osteoblastic cells. J Clin. Invest. 99, 2961 – 2970 (1997). 523. Y. Ishida and J. N. Heersche, Progesterone- and dexamethasone-dependent osteoprogenitors in bone cell populations derived from rat vertebrae are different and distinct. Endocrinology 140, 3210 – 3218 (1999). 524. C. G. Bellows, S. M. Reimers, and J. N. Heersche, Expression of mRNAs for type-I collagen, bone sialoprotein, osteocalcin, and osteopontin at different stages of osteoblastic differentiation and their regulation by 1,25dihydroxyvitamin D3. Cell Tissue Res. 297, 249 – 259 (1999). 525. C. G. Bellows, A. Ciaccia, and J. N. Heersche, Osteoprogenitor cells in cell populations derived from mouse and rat calvaria differ in

526.

527. 528. 529.

530.

531.

532.

533.

534.

535.

536.

537.

538.

539. 540.

541.

542.

their response to corticosterone, cortisol, and cortisone. Bone 23, 119 – 125 (1998). A. Hollnagel, M. Ahrens, and G. Gross, Parathyroid hormone enhances early and suppresses late stages of osteogenic and chondrogenic development in a BMP-dependent mesenchymal differentiation system (C3H10T1/2). J. Bone Miner. Res. 12, 1993 – 2004 (1997). C. Gray and S. J. Jones, PTH(1-34) suppresses appositional bone formation by cultured rat cranial osteoblasts. Bone 23, 453 – 457 (1998). M. H. Kroll, Parathyroid hormone temporal effects on bone formation and resorption. Bull. Math. Biol. 62, 163 – 188 (2000). M. Iwasaki, K. Nakata, H. Nakahara, T. Nakase, T. Kimura, K. Kimata, A. I. Caplan, and K. Ono, Transforming growth factor-beta 1 stimulates chondrogenesis and inhibits osteogenesis in high density culture of periosteum- derived cells. Endocrinology 132, 1603 – 1608 (1993). P. J. Marie, A. Lomri, F. Debiais, and J. Lemonnier, Fibroblast growth factors and osteogenesis. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 179 – 196. Lippincott Williams & Wilkins, Philadelphia, 2000. Y. Ishida and J. N. Heersche, Glucocorticoid-induced osteoporosis: Both in vivo and in vitro concentrations of glucocorticoids higher than physiological levels attenuate osteoblast differentiation. J. Bone Miner. Res. 13, 1822 – 1826 (1998). E. Canalis, Clinical review 83: Mechanisms of glucocorticoid action in bone: implications to glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 81, 3441 – 3447 (1996). C. Kasperk, U. Schneider, U. Sommer, F. Niethard, and R. Ziegler, Differential effects of glucocorticoids on human osteoblastic cell metabolism in vitro. Calcif. Tissue Int. 57, 120 – 126 (1995). H. C. Tenenbaum and J. N. Heersche, Dexamethasone stimulates osteogenesis in chick periosteum in vitro. Endocrinology 117, 2211 – 2217 (1985). A. M. Delany, B. Y. Gabbitas, and E. Canalis, Cortisol downregulates osteoblast alpha 1 (I) procollagen mRNA by transcriptional and posttranscriptional mechanisms. J. Cell Biochem. 57, 488 – 494 (1995). S. M. Pockwinse, J. L. Stein, J. B. Lian, and G. S. Stein, Developmental stage-specific cellular responses to vitamin D and glucocorticoids during differentiation of the osteoblast phenotype: Interrelationship of morphology and gene expression by in situ hybridization. Exp. Cell Res. 216, 244 – 260 (1995). J. B. Lian, V. Shalhoub, F. Aslam, B. Frenkel, J. Green, M. Hamrah, G. S. Stein, and J. L. Stein, Species-specific glucocorticoid and 1,25dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: Dexamethasone inhibits osteoblast differentiation and vitamin D downregulates osteocalcin gene expression. Endocrinology 138, 2117 – 2127 (1997). S. C. Manolagas and R. S. Weinstein, New developments in the pathogenesis and treatment of steroid-induced osteoporosis. J. Bone Miner. Res. 14, 1061 – 1066 (1999). S. F. Hodgson, Corticosteroid-induced osteoporosis. Endocrinol. Metab. Clin. NA 19, 95 – 111 (1990). D. J. Defranco, J. B. Lian, and J. Glowacki, Differential effects of glucocorticoid on recruitment and activity of osteoclasts induced by normal and osteocalcin-deficient bone implanted in rats. Endocrinology 131, 114 – 121 (1992). I. R. Reid, G. E. Chapman, T. R. C. Fraser, A. D. Davies, A. S. Surus, J. Meyer, N. L. Hyq, and H. K. Ibbertson, Low serum osteocalcin levels in glucocorticoid-treated asthmatics. J. Clin. Endocrinol. Metab. 62, 379 – 383 (1986). J. B. Lian and G. S. Stein, Vitamin D regulation of osteoblast growth and differentiation. In “Nutrition and Gene Expression” (C. D. Berdanier and J. L. Hargrove, eds.), pp. 391 – 429. CRC Press, Boca Raton, FL, 1992.

66 543. J. M. Hock, M. Gunness-Hey, J. Poser, H. Olson, N. H. Bell, and L. G. Raisz, Stimulation of undermineralized matrix formation by 1,25 dihydroxyvitamin D3 in long bones of rats. Calcif. Tissue Int. 38, 79 – 86 (1986). 544. P. J. Marie and R. Travers, Continuous infusion of 1,25-dihydroxyvitamin D3 stimulates bone turnover in the normal young mouse. Calcif. Tissue Int. 35, 418 – 425 (1983). 545. G. J. Atkins, D. R. Haynes, S. M. Geary, M. Loric, T. N. Crotti, and D. M. Findlay, Coordinated cytokine expression by stromal and hematopoietic cells during human osteoclast formation. Bone 26, 653 – 661 (2000). 546. J. K. Burmester, R. J. Wiese, N. Maeda, and H. DeLuca, Structure and regulation of the rat 1,25-dihydroxyvitamin D3 receptor. Proc. Natl. Acad. Sci. USA 85, 9499 – 9502 (1988). 547. J. B. Lian, D. L. Carnes, and M. J. Glimcher, Bone and serum concentrations of osteocalcin as a function of 1,25-dihydroxyvitamin D3 circulating levels in bone disorders in rats. Endocrinology 120, 2123 – 2130 (1987). 548. L. C. Gerstenfeld, M. Broess, S. Bruder, A. Caplan, and W. J. Landis, Regulation of osteoblast extracellular matrix formation: Posttranslational regulation of extracellular matrix deposition and relationship between embryonic development and hormonal response. In “Chemistry and Biology of Mineralized Tissues: Proceedings of the Fourth International Conference on the Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 287 – 315. Elsevier, Amsterdam, 1992. 549. H. Zhou, P. Choong, R. McCarthy, S. T. Chou, T. J. Martin, and K. W. Ng, In situ hybridization to show sequential expression of osteoblast gene markers during bone formation in vivo. J. Bone Miner. Res. 9, 1489 – 1499 (1994). 550. G. S. Stein, J. B. Lian, J. L. Stein, A. J. van Wijnen, and M. Montecino, Transcriptional control of osteoblast growth and differentiation. Physiol. Rev. 76, 593 – 629 (1996). 551. J. B. Lian and G. S. Stein, Osteoblast biology. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.), pp. 23 – 60. Academic Press, San Diego, 1996. 552. J. E. Aubin and F. Liu, The osteoblast lineage. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 51 – 68. Academic Press, San Diego, 1996. 553. M. Dodig, M. S. Kronenberg, A. Bedalov, B. E. Kream, G. Gronowicz, S. H. Clark, K. Mack, Y. H. Liu, R. Maxon, Z. Z. Pan, W. B. Upholt, D. W. Rowe, and A. C. Lichtler, Identification of a TAAT-containing motif required for high level expression of the COL1A1 promoter in differentiated osteoblasts of transgenic mice. J. Biol. Chem. 271, 16422 – 16429 (1996). 554. J. A. Rossert, S. S. Chen, H. Eberspaecher, C. N. Smith, and B. de Crombrugghe, Identification of a minimal sequence of the mouse pro-1(I) collagen promoter that confers high-level osteoblast expression in transgenic mice and that binds a protein selectively present in osteoblasts. Proc. Natl. Acad. Sci. USA 93, 1027 – 1031 (1996). 555. E. P. Newberry, T. Latifi, and D. A. Towler, Reciprocal regulation of osteocalcin transcription by the homeodomain proteins msx2 and dlx5. Biochemistry 37, 16360 – 16368 (1998). 556. M. D. Benson, J. L. Bargeon, G. Xiao, P. E. Thomas, A. Kim, Y. Cui, and R. T. Franceschi, Identification of a homeodomain binding element in the bone sialoprotein gene promoter that is required for its osteoblast- selective expression. J. Biol. Chem. 275, 13907 – 13917 (2000). 557. M. Bidder, T. Latifi, and D. A. Towler, Reciprocal temporospatial patterns of Msx2 and Osteocalcin gene expression during murine odontogenesis. J. Bone Miner. Res. 13, 609 – 619 (1998). 558. H.-M. Ryoo, H. M. Hoffmann, T. L. Beumer, B. Frenkel, D. A. Towler, G. S. Stein, J. L. Stein, A. J. van Wijnen, and J. B. Lian, Stage-specific expression of Dlx-5 during osteoblast differentiation:

LIAN AND STEIN

559.

560.

561.

562.

563.

564.

565.

566.

567.

568.

569.

570.

571.

572.

573.

574.

Involvement in regulation of osteocalcin gene expression. Mol. Endocrinol. 11, 1681 – 1694 (1997). H. Zhang, G. Hu, H. Wang, P. Sciavolino, N. Iler, M. M. Shen, and C. Abate-Shen, Heterodimerization of Msx and Dlx homeoproteins results in functional antagonism. Mol. Cell. Biol. 17, 2920 – 2932 (1997). J. Rossert, H. Eberspaecher, and B. de Crombrugghe, Separate cisacting DNA elements of the mouse pro-alpha 1 (I) collagen promoter direct expression of reporter genes to different type I collagen-producing cells in transgenic mice. J. Cell Biol. 129, 1421 – 1432 (1995). D. A. Towler, C. D. Bennett, and G. A. Rodan, Activity of the rat osteocalcin basal promoter in osteoblastic cells is dependent upon homeodomain and CP1 binding motifs. Mol. Endocrinol. 8, 614 – 624 (1994). M. Machwate, A. Jullienne, M. Moukhtar, and P. J. Marie, Temporal variation of c-fos proto-oncogene expression during osteoblast differentiation and osteogenesis in developing bone. J. Cell. Biochem. 57, 62 – 70 (1995). A. E. Grigoriadis, K. Schellander, Z.-Q. Wang, and E. F. Wagner, Osteoblasts are target cells for transformation in c-fos transgenic mice. J. Cell Biol. 122, 685 – 701 (1993). L. R. McCabe, C. Banerjee, R. Kundu, R. J. Harrison, P. R. Dobner, J. L. Stein, J. B. Lian, and G. S. Stein, Developmental expression and activities of specific fos and jun proteins are funtionally related to osteoblast maturation: Role of fra-2 and jun D during differentiation. Endocrinology 137, 4398 – 4408 (1996). A. R. Kornblihtt, C. G. Pesce, C. R. Alonso, P. Cramer, A. Srebrow, S. Werbajh, and A. F. Muro, The fibronectin gene as a model for splicing and transcription studies. FASEB J. 10, 248 – 257 (1996). A. Staffa, N. H. Acheson, and A. Cochrane, Novel exonic elements that modulate splicing of the human fibronectin EDA exon. J Biol. Chem. 272, 33394 – 33401 (1997). M. Sugimoto, T. Kimura, N. Tsumaki, Y. Matsui, K. Nakata, H. Kawahata, N. Yasui, Y. Kitamura, S. Nomura, and T. Ochi, Differential in situ expression of alpha2(XI) collagen mRNA isoforms in the developing mouse. Cell Tissue Res. 292, 325 – 332 (1998). V. Geoffroy, D. A. Corral, L. Zhou, B. Lee, and G. Karsenty, Genomic organization, expression of the human CBFA1 gene, and evidence for an alternative splicing event affecting protein function. Mamm. Genome 9, 54 – 57 (1998). Z. S. Xiao, R. Thomas, T. K. Hinson, and L. D. Quarles, Genomic structure and isoform expression of the mouse, rat and human Cbfa1/Osf2 transcription factor. Gene 214, 187 – 197 (1998). Z. S. Xiao, T. K. Hinson, and L. D. Quarles, Cbfa1 isoform overexpression upregulates osteocalcin gene expression in non-osteoblastic and pre-osteoblastic cells. J. Cell. Biochem. 74, 596 – 605 (1999). C. Banerjee, A. Javed, J. Choi, J. Green, A. van Wijnen, J. Stein, G. Stein, and J. Lian, Representation and activities of multiple Cbfa transcription factors in osteoblasts. Bone 23, S321 – S321 (1998). [Abstract] A. Javed, B. Guo, S. Hiebert, J. Choi, J. Green, S. Zhao, M. A. Osborne, S. Stifani, J. L. Stein, G. S. Stein, A. J. van Wijnen, and J. B. Lian, RUNX (CBF/AML/PEBP2) dependent repression of tissuespecific gene transcription involves in situ association with Groucho/TLE/R-esp proteins. J. Bone Miner. Res. 15, S498 – S498 (2000). [Abstract] G. Xiao, D. Jiang, P. Thomas, M. D. Benson, K. Guan, G. Karsenty, and R. T. Franceschi, MAPK pathways activate and phosphorylate the osteoblast-specific transcription factor, Cbfa1. J. Biol. Chem. 275, 4453 – 4459 (2000). A. Javed, B. Guo, S. Hiebert, J.-Y. Choi, J. Green, S.-C. Zhao, M. A. Osborne, S. Stifani, J. L. Stein, J. B. Lian, A. J. van Wijnen, and G. S. Stein, Groucho/TLE/R-Esp proteins associate with the nuclear matrix and repress RUNX (CBF/AML/PEBP2) dependent

67

CHAPTER 2 Osteoblast Biology

575.

576.

577.

578.

579.

580.

581.

582.

583.

584.

585.

586.

587.

588.

589.

activation of tissue-specific gene transcription. J. Cell Sci. 113, 2221 – 2231 (2000). K. W. McLarren, R. Lo, D. Grbavec, K. Thirunavukkarasu, G. Karsenty, and S. Stifani, The mammalian basic helix loop helix protein HES-1 binds to and modulates the transactivating function of the runt-related factor cbfa1. J. Biol. Chem. 275, 530 – 538 (2000). D. Levanon, R. E. Goldstein, Y. Bernstein, H. Tang, D. Goldenberg, S. Stifani, Z. Paroush, and Y. Groner, Transcriptional repression by AML1 and LEF-1 is mediated by the TLE/Groucho corepressors. Proc. Natl. Acad. Sci. USA 95, 11590 – 11595. D. A. Towler, S. J. Rutledge, and G. A. Rodan, Msx-2/Hox 8.1: A transcriptional regulator of the rat osteocalcin promoter. Mol. Endocrinol. 8, 1484 – 1493 (1994). H. M. Hoffmann, T. L. Beumer, S. Rahman, L. R. McCabe, C. Banerjee, F. Aslam, J. A. Tiro, A. J. Van Wijnen, J. L. Stein, G. S. Stein and J. B. Lian, Bone tissue-specific transcription of the osteocalcin gene: role of an activator osteoblast-specific complex and suppressor hox proteins that bind the OC box. J. Cell. Biochem. 61, 310 – 324 (1996). H. Hoffmann, J. Green, A. J. van Wijnen, J. L. Stein, G. S. Stein, and J. B. Lian, Expression screening of factors binding to the osteocalcin bone-specific promoter element OC Box I: Isolation of a novel osteoblast differentiation-specific factor. J. Cell. Biochem. 80, 156 – 168 (2000). J. M. Boudreaux and D. A. Towler, Synergistic induction of osteocalcin gene expression: Identification of a bipartite element conferring fibroblast growth factor 2 and cyclic AMP responsiveness in the rat osteocalcin promoter. J. Biol. Chem. 271, 7508 – 7515 (1996). G. Boguslawski, L. V. Hale, X. P. Yu, R. R. Miles, J. E. Onyia, R. F. Santerre, and S. Chandrasekhar, Activation of osteocalcin transcription involves interaction of protein kinase A- and protein kinase C-dependent pathways. J. Biol. Chem. 275, 999 – 1006 (2000). T. A. Owen, R. Bortell, S. A. Yocum, S. L. Smock, M. Zhang, C. Abate, V. Shalhoub, N. Aronin, K. L. Wright, A. J. van Wijnen, J. L. Stein, T. Curran, J. B. Lian, and G. S. Stein, Coordinate occupancy of AP-1 sites in the vitamin D responsive and CCAAT box elements by Fos-Jun in the osteocalcin gene: Model for phenotype suppression of transcription. Proc. Natl. Acad. Sci. USA 87, 9990 – 9994 (1990). K. Ozono, T. Sone, and J. W. Pike, The genomic mechanism of action of 1,25-dihydroxyvitamin D3. J. Bone Miner. Res. 6, 1021 – 1027 (1991). C. Banerjee, J. L. Stein, A. J. van Wijnen, B. Frenkel, J. B. Lian, and G. S. Stein, Transforming growth factor-beta 1 responsiveness of the rat osteocalcin gene is mediated by an activator protein-1 binding site. Endocrinology 137, 1991 – 2000 (1996). M. B. Demay, J. M. Gerardi, H. F. DeLuca, and H. M. Kronenberg, DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsive to 1,25-dihydroxyvitamin D3. Proc. Natl. Acad. Sci. USA 87, 369 – 373 (1990). E. R. Markose, J. L. Stein, G. S. Stein, and J. B. Lian, Vitamin Dmediated modifications in protein-DNA interactions at two promoter elements of the osteocalcin gene. Proc. Natl. Acad. Sci. USA 87, 1701 – 1705 (1990). S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc. Natl. Acad. Sci. USA 86, 4455 – 4459 (1989). C. M. Terpening, C. A. Haussler, P. W. Jurutka, M. A. Galligan, B. S. Komm, and M. R. Haussler, The vitamin D-responsive element in the rat bone Gla protein gene is an imperfect direct repeat that cooperates with other cis-elements in 1,25-dihydroxyvitamin D3-mediated transcriptional activation. Mol. Endocrinol. 5, 373 – 385 (1991). M. B. Demay, M. S. Kiernan, H. F. DeLuca, and H. M. Kronenberg, Characterization of 1,25-dihydroxyvitamin D3 receptor interactions

590.

591.

592.

593.

594.

595.

596.

597.

598.

599.

600.

601.

602.

603.

604.

with target sequences in the rat osteocalcin gene. Mol. Endocrinol. 6, 557 – 562 (1992). M. Schrader, S. Nayeri, J. P. Kahlen, K. M. Muller, and C. Carlberg, Natural vitamin D3 response elements formed by inverted palindromes: Polarity-directed ligand sensitivity of vitamin D3 receptorretinoid X receptor heterodimer-mediated transactivation. Mol. Cell. Biol. 15, 1154 – 1161 (1995). P. N. MacDonald, D. R. Sherman, D. R. Dowd, S. C. Jefcoat, Jr., and K. DeLisle, The vitamin D receptor interacts with general transcription factor IIB. J. Biol. Chem. 270, 4748 – 4752 (1995). T. A. Owen, R. Bortell, V. Shalhoub, A. Heinrichs, J. L. Stein, G. S. Stein, and J. B. Lian, Postproliferative transcription of the rat osteocalcin gene is reflected by vitamin D-responsive developmental modifications in protein-DNA interactions at basal and enhancer promoter elements. Proc. Natl. Acad. Sci. USA 90, 1503 – 1507 (1993). J. C. G. Blanco, I.-M. Wang, S. Y. Tsai, M.-J. Tsai, B. W. O’Malley, P. W. Jurutka, M. R. Haussler, and K. Ozato, Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc. Natl. Acad. Sci. USA 92, 1535 – 1539 (1995). A. Staal, A. J. van Wijnen, J. B. Lian, J. L. Stein, H. DeLuca, J. C. Birkenhager, H. A. P. Pols, J. P. T. M. van Leeuwen, and G. S. Stein, Different identities of nuclear protein complexes recognizing the vitamin D responsive elements of the osteocalcin and osteopontin genes. J. Bone Miner. Res. 10, S167 – S167 (1995). [Abstract] P. N. MacDonald, D. R. Dowd, S. Nakajima, M. A. Galligan, M. C. Reeder, C. A. Haussler, K. Ozato, and M. R. Haussler, Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol. Cell. Biol. 13, 5907 – 5917 (1993). H. Kuno, S. M. Kurian, G. N. Hendy, J. White, H. F. DeLuca, C. O. Evans, and M. S. Nanes, Inhibition of 1,25-dihydroxyvitamin D3 stimulated osteocalcin gene transcription by tumor necrosis factoralpha: Structural determinants within the vitamin D response element. Endocrinology 134, 2524 – 2531 (1994). R. Schule, K. Umesono, D. J. Mangelsdorf, J. Bolado, J. W. Pike, and R. M. Evans, Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61, 497 – 504 (1990). B. Cheskis and L. P. Freedman, Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDRretinoid X receptor heterodimers. Mol. Cell. Biol. 14, 3329 – 3338 (1994). M. Schrader, I. Bendik, M. Becker-Andre, and C. Carlberg, Interaction between retinoic acid and vitamin D signaling pathways. J. Biol. Chem. 268, 17830 – 17836 (1993). S. A. Kliewer, K. Umesono, D. J. Mangelsdorf, and R. M. Evans, Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D3 signalling. Nature 355, 446 – 449 (1992). R. Bortell, T. A. Owen, J. P. Bidwell, P. Gavazzo, E. Breen, A. J. van Wijnen, H. F. DeLuca, J. L. Stein, J. B. Lian, and G. S. Stein, Vitamin D-responsive protein-DNA interactions at multiple promoter regulatory elements that contribute to the level of rat osteocalcin gene expression. Proc. Natl. Acad. Sci. USA 89, 6119 – 6123 (1992). G. S. Stein, A. J. van Wijnen, M. Montecino, J. L. Stein, and J. B. Lian, Nuclear structure/gene expression interrelationships. J. Cell. Physiol. 181, 240 – 250 (1999). N. J. McKenna, Z. Nawaz, S. Y. Tsai, M. J. Tsai, and B. W. O’Malley, Distinct steady-state nuclear receptor coregulator complexes exist in vivo. Proc. Natl. Acad. Sci. USA 95, 11697 – 11702 (1998). A. N. Imbalzano, Energy-dependent chromatin remodelers: Complex complexes and their components. Crit. Rev. Eukaryot. Gene Expr. 8, 225 – 255 (1998).

68 605. J. L. Workman and R. E. Kingston, Alteration of nucleosome structure as a mechanism of transcriptional regulation. Annu. Rev. Biochem. 67, 545 – 579 (1998). 606. H. Kwon, A. N. Imbalzano, P. A. Khavari, R. E. Kingston, and M. R. Green, Nucleosome disruption and enhancement of activator binding by a human SWI/SNF complex. Nature 370, 477 – 481 (1994). 607. J. Cote, J. Quinn, J. L. Workman, and C. L. Peterson, Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265, 53 – 60 (1994). 608. C. L. Peterson, Y. Zhao, and B. T. Chait, Subunits of the yeast SWI/SNF complex are members of the actin- related protein (ARP) family. J Biol. Chem. 273, 23641 – 23644 (1998). 609. M. Vettese-Dadey, P. Walter, H. Chen, L.-J. Juan, and J. L. Workman, Role of the histone amino termini in facilitated binding of a transcription factor. Mol. Cell. Biol. 14, 970 – 981 (1994). 610. M. Vettese-Dadey, P. A. Grant, T. R. Hebbes, C. Crane-Robinson, C. D. Allis, and J. L. Workman, Acetylation of histone H4 plays a primary role in enhancing transcription factor binding to nucleosomal DNA in vitro. EMBO J. 15, 2508 – 2518 (1996). 611. K. Ura, H. Kurumizaka, S. Dimitrov, G. Almouzni, and A. P. Wolffe, Histone acetylation: Influence on transcription, nucleosome mobility and positioning, and linker histone-dependent transcriptional repression. EMBO J. 16, 2096 – 2107 (1997). 612. M. Montecino, S. Pockwinse, J. Lian, G. Stein, and J. Stein, DNase I hypersensitive sites in promoter elements associated with basal and vitamin D dependent transcription of the bone- specific osteocalcin gene. Biochemistry 33, 348 – 353 (1994). 613. M. Montecino, J. Lian, G. Stein, and J. Stein, Changes in chromatin structure support constitutive and developmentally regulated transcription of the bone-specific osteocalcin gene in osteoblastic cells. Biochemistry 35, 5093 – 5102 (1996). 614. M. Montecino, B. Frenkel, J. Lian, J. Stein, and G. Stein, Requirement of distal and proximal promoter sequences for chromatin organization of the osteocalcin gene in bone-derived cells. J. Cell. Biochem. 63, 221 – 228 (1996). 615. E. C. Breen, A. J. van Wijnen, J. B. Lian, G. S. Stein, and J. L. Stein, In vivo occupancy of the vitamin D responsive element in the osteocalcin gene supports vitamin D-dependent transcriptional upregulation in intact cells. Proc. Natl. Acad. Sci. USA 91, 12902 – 12906 (1994). 616. G. S. Stein, J. B. Lian, S. I. Dworetzky, T. A. Owen, R. Bortell, J. P. Bidwell, and A. J. van Wijnen, Regulation of transcription-factor activity during growth and differentiation: Involvement of the nuclear matrix in concentration and localization of promoter binding proteins. J. Cell. Biochem. 47, 300 – 305 (1991). 617. R. Berezney and D. S. Coffey, Nuclear protein matrix: Association with newly synthesized DNA. Science 189, 291 – 292 (1975). 618. B. D. Nelkin, D. M. Pardoll, and B. Vogelstein, Localization of SV40 genes within supercoiled loop domains. Nucleic Acids Res. 8, 5623 – 5633 (1980). 619. S. I. Robinson, B. D. Nelkin, and B. Vogelstein, The ovalbumin gene is associated with the nuclear matrix of chicken oviduct cells. Cell 28, 99 – 106 (1982). 620. S. I. Dworetzky, K. L. Wright, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Sequence-specific DNA-binding proteins are components of a nuclear matrix-attachment site. Proc. Natl. Acad. Sci. USA 89, 4178 – 4182 (1992). 621. A. J. van Wijnen, J. P. Bidwell, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1. Biochemistry 32, 8397 – 8402 (1993). 622. A. Stief, D. M. Winter, W. H. Stratling, and A. E. Sippel, A nuclear DNA attachment element mediates elevated and position- independent gene activity. Nature 341, 343 – 345 (1989). 623. J. Schaack, W. Y. Ho, P. Freimuth, and T. Shenk, Adenovirus terminal protein mediates both nuclear matrix association and

LIAN AND STEIN

624.

625. 626.

627.

628.

629.

630.

631.

632.

633.

634.

635.

636.

637.

638.

639.

640.

efficient transcription of adenovirus DNA. Genes Dev. 4, 1197 – 1208 (1990). J. B. Lawrence, R. H. Singer, and L. M. Marselle, Highly localized tracks of specific transcripts within interphase nuclei visualized by in situ hybridization. Cell 57, 493 – 502 (1989). S. Zeitlin, A. Parent, S. Silverstein, and A. Efstratiadis, Pre-mRNA splicing and the nuclear matrix. Mol. Cell Biol. 7, 111 – 120 (1987). K. C. Carter, D. Bowman, W. Carrington, K. Fogarty, J. A. McNeil, F. S. Fay, and J. B. Lawrence, A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science 259, 1330 – 1335 (1993). D. L. Spector, Higher order nuclear organization: Three-dimensional distribution of small nuclear ribonucleoprotein particles. Proc. Natl. Acad. Sci. USA 87, 147 – 151 (1990). B. J. Blencowe, J. A. Nickerson, R. Issner, S. Penman, and P. A. Sharp, Association of nuclear matrix antigens with exon-containing splicing complexes. J. Cell Biol. 127, 593 – 607 (1994). J. R. Davie, Nuclear matrix, dynamic histone acetylation and transcriptionally active chromatin. Mol. Biol. Rep. 24, 197 – 207 (1997). S. I. Dworetzky, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Progressive changes in the protein composition of the nuclear matrix during rat osteoblast differentiation. Proc. Natl. Acad. Sci. USA 87, 4605 – 4609 (1990). C. Zeng, A. J. van Wijnen, J. L. Stein, S. Meyers, W. Sun, L. Shopland, J. B. Lawrence, S. Penman, J. B. Lian, G. S. Stein, and S. W. Hiebert, Identification of a nuclear matrix targeting signal in the leukemia and bone-related AML/CBF transcription factors. Proc. Natl. Acad. Sci. USA 94, 6746 – 6751 (1997). S. McNeil, B. Guo, J. L. Stein, J. B. Lian, S. Bushmeyer, E. Seto, M. L. Atchison, S. Penman, A. J. van Wijnen, and G. S. Stein, Targeting of the YY1 transcription factor to the nucleolus and the nuclear matrix in situ: The C-terminus is a principal determinant for nuclear trafficking. J. Cell. Biochem. 68, 500 – 510 (1998). L. Bruhn, A. Munnerlyn, and R. Grosschedl, ALY, a contextdependent coactivator of LEF-1 and AML-1, is required for TCRalpha enhancer function. Genes Dev. 11, 640 – 653 (1997). I. Kitabayashi, A. Yokoyama, K. Shimizu, and M. Ohki, Interaction and functional cooperation of the leukemia-associated factors AML1 and p300 in myeloid cell differentiation. EMBO J. 17, 2994 – 3004 (1998). H. Ogihara, T. Kanno, E. Morii, D. K. Kim, Y. M. Lee, M. Sato, W. Y. Kim, S. Nomura, Y. Ito, and Y. Kitamura, Synergy of PEBP2/CBF with mitranscription factor (MITF) for transactivation of mouse mast cell protease 6 gene. Oncogene 18, 4632 – 4639 (1999). M. Y. Ahn, G. Huang, S. C. Bae, H. J. Wee, W. Y. Kim, and Y. Ito, Negative regulation of granulocytic differentiation in the myeloid precursor cell line 32Dcl3 by ear-2, a mammalian homolog of Drosophila seven-up, and a chimeric leukemogenic gene, AML1/ETO. Proc. Natl. Acad. Sci. USA 95, 1812 – 1817 (1998). Y. M. Ning and D. M. Robins, AML3/CBFalpha1 is required for androgen-specific activation of the enhancer of the mouse sex-limited protein (Slp) gene. J. Biol. Chem. 274, 30624 – 30630 (1999). H. Sasaki-Iwaoka, K. Maruyama, H. Endoh, T. Komori, S. Kato, and H. Kawashima, A trans-acting enhancer modulates estrogenmediated transcription of reporter genes in osteoblasts. J. Bone Miner. Res. 14, 248 – 255 (1999). B. D. Aronson, A. L. Fisher, K. Blechman, M. Caudy, and J. P. Gergen, Groucho-dependent and -independent repression activities of Runt domain proteins. Mol. Cell Biol. 17, 5581 – 5587 (1997). B. Guo, A. Javed, J. Green, J. L. Stein, J. B. Lian, A. J. van Wijnen, and G. S. Stein, Groucho/TLE proteins associate with the nuclear matrix and repress Cbfa/AML mediated transactivation on osteocalcin promoter. Bone 23, S184 – S184 (1998). [Abstract]

69

CHAPTER 2 Osteoblast Biology 641. G, Chen, J. Fernandez, S. Mische, and A. J. Courey, A functional interaction between the histone deacetylase Rpd3 and the corepressor groucho in Drosophila development. Genes Dev. 13, 2218 – 2230 (1999). 642. A. Palaparti, A. Baratz, and S. Stifani, The Groucho/transducin-like enhancer of split transcriptional repressors interact with the genetically defined amino-terminal silencing domain of histone H3. J. Biol. Chem. 272, 26604 – 26610 (1997). 643. S. W. Hiebert, W. Sun, J. N. Davis, T. Golub, S. Shurtleff, A. Buijs, J. R. Downing, G. Grosveld, M. F. Roussell, D. G. Gilliland, N. Lenny, and S. Meyers, The t(12;21) translocation converts AML-1B from an activator to a repressor of transcription. Mol. Cell. Biol. 16, 1349 – 1355 (1996). 644. B. Lutterbach, J. J. Westendorf, B. Linggi, S. Isaac, E. Seto, and S. W. Hiebert, A mechanism of repression by acute myeloid leukemia-1, the target of multiple chromosomal translocations in acute leukemia. J. Biol. Chem. 275, 651 – 656 (2000). 645. A. Javed, G. L. Barnes, A. Javed, B. O. Jassanya, J. L. Stein, L. Gerstenfeld, G. S. Stein, and J. B. Lian, Cbfa family members act as a negative regulator of the bone sialoprotein gene. J. Bone Miner. Res. 14, S165 – S165 (1999). [Abstract] 646. H. Drissi, Q. Luc, R. Shakoori, S. Chuva de Sousa Lopes, J.-Y. Choi, A. Terry, M. Hu, S. Jones, J. C. Neil, J. B. Lian, J. L. Stein, A. J. van Wijnen, and G. S. Stein, Transcriptional autoregulation of the bone related CBFA1/RUNX2 gene. J. Cell. Physiol. 184, 341 – 350 (2000). 647. A. Javed, S. Gutierrez, M. Montecino, A. J. van Wijnen, J. L. Stein, G. S. Stein, and J. B. Lian, Multiple Cbfa/AML sites in the rat osteocalcin promoter are required for basal and vitamin D responsive transcription and contribute to chromatin organization. Mol. Cell. Biol. 19, 7491 – 7500 (1999). 648. C. Tribioli and T. Lufkin, The murine Bapx1 homeobox gene plays a critical role in embryonic development of the axial skeleton and spleen. Development 126, 5699 – 5711 (1999). 649. L. A. Lettice, L. A. Purdie, G. J. Carlson, F. Kilanowski, J. Dorin, and R. E. Hill, The mouse bagpipe gene controls development of axial skeleton, skull, and spleen. Proc. Natl. Acad. Sci. USA 96, 9695 – 9700 (1999). 650. J. A. Rivera-Perez, M. Wakamiya, and R. R. Behringer, Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development. Development 126, 3811 – 3821 (1999). 651. M. J. Depew, J. K. Liu, J. E. Long, R. Presley, J. J. Meneses, R. A. Pedersen, and J. L. Rubenstein, Dlx5 regulates regional development of the branchial arches and sensory capsules. Development 126, 3831 – 3846 (1999). 652. D. Acampora, G. R. Merlo, L. Paleari, B. Zerega, M. P. Postiglione, S. Mantero, E. Bober, O. Barbieri, A. Simeone, and G. Levi Craniofacial, vestibular and bone defects in mice lacking the distalless-related gene dlx5. Development 126, 3795 – 3809 (1999). 653. I. Satokata, L. Ma, H. Ohshima, M. Bei, I. Woo, K. Nishizawa, T. Maeda, Y. Takano, M. Uchiyama, S. Heaney, H. Peters, Z. Tang, R. Maxson, and R. Maas, Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation. Nature Genet. 24, 391 – 395 (2000). 654. A. O. Wilkie, Z. Tang, N. Elanko, S. Walsh, S. R. Twigg, J. A. Hurst, S. A. Wall, K. H. Chrzanowska, and R. E. Maxson, Jr., Functional haploinsufficiency of the human homeobox gene MSX2 causes defects in skull ossification. Nature Genet. 24, 387 – 390 (2000). 655. Y. H. Liu, Z. Tang, R. K. Kundu, L. Wu, W. Luo, D. Zhu, F. Sangiorgi, M. L. Snead, and R. E. Maxson, Msx2 gene dosage influences the number of proliferative osteogenic cells in growth centers of the developing murine skull: A possible mechanism for MSX2mediated craniosynostosis in humans. Dev. Biol. 205, 260 – 274 (1999). 656. E. Filvaroff, A. Erlebacher, J. Ye, S. E. Gitelman, J. Lotz, M. Heillman, and R. Derynck, Inhibition of TGF-beta receptor signaling in

657.

658.

659.

660.

661.

662.

663.

664.

665.

666.

667.

668.

669.

670.

671.

672.

osteoblasts leads to decreased bone remodeling and increased trabecular bone mass. Development 126, 4267 – 4279 (1999). R. Serra, M. Johnson, E. H. Filvaroff, J. LaBorde, D. M. Sheehan, R. Derynck, and H. L. Moses, Expression of a truncated, kinasedefective TGF-beta type II receptor in mouse skeletal tissue promotes terminal chondrocyte differentiation and osteoarthritis. J. Cell Biol. 139, 541 – 552 (1997). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). K. A. Lawson, N. R. Dunn, B. A. Roelen, L. M. Zeinstra, A. M. Davis, C. V. Wright, J. P. Korving, and B. L. Hogan, BMP4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev. 13, 424 – 436 (1999). G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF beta-superfamily. Nature 368, 639 – 643 (1994). A. T. Dudley, K. M. Lyons, and E. J. Robertson, A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev. 9, 2795 – 2807 (1995). G. Luo, C. Hofmann, A. L. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). Y. Mishina, A. Suzuki, N. Ueno, and R. R. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). S. E. Yi, A. Daluiski, R. Pederson, V. Rosen, and K. M. Lyons, The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621 – 630 (2000). S. R. Rittling, H. N. Matsumoto, M. D. McKee, A. Nanci, X. R. An, K. E. Novick, A. J. Kowalski, M. Noda, and D. T. Denhardt, Mice lacking osteopontin show normal development and bone structure but display altered osteoclast formation in vitro. J. Bone Miner. Res. 13, 1101 – 1111 (1998). A. D. Bradshaw, A. Francki, K. Motamed, C. Howe, and E. H. Sage, Primary mesenchymal cells isolated from SPARC-null mice exhibit altered morphology and rates of proliferation. Mol. Biol. Cell 10, 1569 – 1579 (1999). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectindeficient mice. J. Clin. Invest. 105, 915 – 923 (2000). T. R. Kyriakides, Y. H. Zhu, L. T. Smith, S. D. Bain, Z. Yang, M. T. Lin, K. G. Danielson, R. V. Iozzo, M. LaMarca, C. E. McKinney, E. I. Ginns, and P. Bornstein, Mice that lack thrombospondin 2 display connective tissue abnormalities that are associated with disordered collagen fibrillogenesis, an increased vascular density, and a bleeding diathesis. J. Cell Biol. 140, 419 – 430 (1998). K. D. Hankenson, S. D. Bain, T. R. Kyriakides, E. A. Smith, S. A. Goldstein, and P. Bornstein, Increased marrow-derived osteoprogenitor cells and endosteal bone formation in mice lacking thrombospondin 2. J. Bone Miner. Res. 15, 851 – 862 (2000). K. G. Danielson, H. Baribault, D. F. Holmes, H. Graham, K. E. Kadler, and R. V. Iozzo, Targeted disruption of decorin leads to abnormal collagen fibril morphology and skin fragility. J. Cell Biol. 136, 729 – 743 (1997).

70 673. K. G. Waymire, J. D. Mahuren, J. M. Jaje, T. R. Guilarte, S. P. Coburn, and G. R. MacGregor, Mice lacking tissue non-specific alkaline phosphatase die from seizures due to defective metabolism of vitamin B-6. Nature Genet. 11, 45 – 51 (1995). 674. S. Narisawa, N. Frohlander, and J. L. Millan, Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dyn. 208, 432 – 446 (1997). 675. K. N. Fedde, L. Blair, J. Silverstein, S. P. Coburn, L. M. Ryan, R. S. Weinstein, K. Waymire, S. Narisawa, J. L. Millan, G. R. MacGregor, and M. P. Whyte, Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J. Bone Miner. Res. 14, 2015 – 2026 (1999). 676. T. H. Vu, J. M. Shipley, G. Bergers, J. E. Berger, J. A. Helms, D. Hanahan, S. D. Shapiro, R. M. Senior, and Z. Werb, MMP9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93, 411 – 422 (1998). 677. K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking beta3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest 105, 433 – 440 (2000). 678. C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). 679. J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M. Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). 680. R. Shiang, L. M. Thompson, Y. Z. Zhu, D. M. Church, T. J. Fielder, M. Bocian, S. T. Winokur, and J. J. Wasmuth, Mutations in the transmembrane domain of FGFR3 cause the most common genetic form of dwarfism, achondroplasia. Cell 78, 335 – 342 (1994). 681. T. P. Yamaguchi, K. Harpal, M. Henkemeyer, and J. Rossant, fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 8, 3032 – 3044 (1994). 682. C. X. Deng, A. Wynshaw-Boris, M. M. Shen, C. Daugherty, D. M. Ornitz, and P. Leder, Murine FGFR-1 is required for early postimplantation growth and axial organization. Genes Dev. 8, 3045 – 3057 (1994). 683. E. C. Weir, W. M. Philbrick, M. Amling, L. A. Neff, R. Baron, and A. E. Broadus, Targeted overexpression of parathyroid hormone-related peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc. Natl. Acad. Sci. USA 93, 10240 – 10245 (1996). 684. G. Zhao, M. C. Monier-Faugere, M. C. Langub, Z. Geng, T. Nakayama, J. W. Pike, S. D. Chernausek, C. J. Rosen, L. R. Donahue, H. H. Malluche, J. A. Fagin, and T. L. Clemens, Targeted overexpression of insulin-like growth factor I to osteoblasts of transgenic mice: Increased trabecular bone volume without increased osteoblast proliferation. Endocrinology 141, 2674 – 2682 (2000). 685. C. Chiang, Y. Litingtung, E. Lee, K. E. Young, J. L. Corden, H. Westphal, and P. A. Beachy, Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407 – 413 (1996). 686. B. Li, S. Boast, S. K. de los, I. Schieren, M. Quiroz, S. L. Teitelbaum, M. M. Tondravi, and S. P. Goff, Mice deficient in Abl are osteoporotic and have defects in osteoblast maturation. Nature Genet. 24, 304 – 308 (2000). 687. M. L. Fero, M. Rivkin, M. Tasch, P. Porter, C. E. Carow, E. Firpo, K. Polyak, L. H. Tsai, V. Broudy, R. M. Perlmutter, K. Kaushansky, and J. M. Roberts, A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)deficient mice. Cell 85, 733 – 744 (1996). 688. H. Kiyokawa, R. D. Kineman, K. O. Manova-Todorova, V. C. Soares, E. S. Hoffman, M. Ono, D. Khanam, A. C. Hayday, L. A. Frohman, and A. Koff, Enhanced growth of mice lacking the cyclindependent kinase inhibitor function of p27(Kip1). Cell 85, 721 – 732 (1996).

LIAN AND STEIN 689. A. Okawa, I. Nakamura, S. Goto, H. Moriya, Y. Nakamura, and S. Ikegawa, Mutation in Npps in a mouse model of ossification of the posterior longitudinal ligament of the spine. Nature Genet. 19, 271 – 273 (1998). 690. D. E. Cash, C. B. Bock, K. Schughart, E. Linney, and T. M. Underhill, Retinoic acid receptor alpha function in vertebrate limb skeletogenesis: a modulator of chondrogenesis. J Cell Biol. 136, 445 – 457. 691. A. E. Grigoriadis, Z. Q. Wang, and E. F. Wagner, Fos and bone cell development: Lessons from a nuclear oncogene. Trends Genet. 11, 436 – 441 (1995). 692. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 693. K. Matsuo, W. Jochum, J. M. Owens, T. J. Chambers, and E. F. Wagner, Function of Fos proteins in bone cell differentiation. Bone 25, 141 (1999). 694. M. Schreiber, C. Poirier, A. Franchi, R. Kurzbauer, J. L. Guenet, G. F. Carle, and E. F. Wagner, Structure and chromosomal assignment of the mouse fra-1 gene, and its exclusion as a candidate gene for oc (osteosclerosis). Oncogene 15, 1171 – 1178 (1997). 695. K. Iida, H. Koseki, H. Kakinuma, N. Kato, Y. Mizutani-Koseki, H. Ohuchi, H. Yoshioka, S. Noji, K. Kawamura, Y. Kataoka, F. Ueno, M. Taniguchi, N. Yoshida, T. Sugiyama, and N. Miura, Essential roles of the winged helix transcription factor MFH-1 in aortic arch patterning and skeletogenesis. Development 124, 4627 – 4638 (1997). 696. M. Barna, N. Hawe, L. Niswander, and P. P. Pandolfi, Plzf regulates limb and axial skeletal patterning. Nature Genet. 25, 166 – 172 (2000). 697. A. M. Reimold, M. J. Grusby, B. Kosaras, J. W. Fries, R. Mori, S. Maniwa, I. M. Clauss, T. Collins, R. L. Sidman, M. J. Glimcher, and L. H. Glimcher, Chondrodysplasia and neurological abnormalities in ATF-2-deficient mice. Nature 379, 262 – 265 (1996). 698. S. Murakami, M. Kan, W. L. McKeehan, and B. de Crombrugghe, Up-regulation of the chondrogenic Sox9 gene by fibroblast growth factors is mediated by the mitogen-activated protein kinase pathway. Proc. Natl. Acad. Sci. USA 97, 1113 – 1118 (2000). 699. B. K. Zehentner, C. Dony, and H. Burtscher, The transcription factor Sox9 is involved in BMP-2 signaling. J. Bone Miner. Res. 14, 1734 – 1741 (1999). 700. M. M. Hurley, S. Tetradis, Y. F. Huang, J. Hock, B. E. Kream, L. G. Raisz, and M. G. Sabbieti, Parathyroid hormone regulates the expression of fibroblast growth factor-2 mRNA and fibroblast growth factor receptor mRNA in osteoblastic cells. J. Bone Miner. Res. 14, 776 – 783 (1999). 701. M. J. Bitgood and A. P. McMahon, Hedgehog and BMP genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172, 126 – 138 (1995). 702. M. Iwasaki, A. X. Le, and J. A. Helms, Expression of indian hedgehog, bone morphogenetic protein 6 and gli during skeletal morphogenesis. Mech. Dev. 69, 197 – 202 (1997). 703. Y. Zhang, X. Zhao, Y. Hu, T. St. Amand, M. Zhang, R. Ramamurthy, M. Qiu, and Y. Chen, Msx1 is required for the induction of Patched by Sonic hedgehog in the mammalian tooth germ. Dev. Dyn. 215, 45 – 53 (1999). 704. K. Zhao, W. Wang, O. J. Rando, Y. Xue, K. Swiderek, A. Kuo, and G. R. Crabtree, Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95, 625 – 636 (1998). 705. X. Xu, M. Weinstein, C. Li, M. Naski, R. I. Cohen, D. M. Ornitz, P. Leder, and C. Deng, Fibroblast growth factor receptor 2 (FGFR2)mediated reciprocal regulation loop between FGF8 and FGF10 is essential for limb induction. Development 125, 753 – 765 (1998).

71

CHAPTER 2 Osteoblast Biology 706. W. C. Yeh, Z. Cao, M. Classon, and S. L. McKnight, Cascade regulation of terminal adipocyte differentiation by three members of the C/EBP family of leucin zipper proteins. Genes Dev. 9, 168 – 181 (1995). 707. S. O. Freytag, D. L. Paielli, and J. D. Gilbert, Ectopic expression of the CCAAT/enhancer-binding protein alpha promotes the adipogenic program in a variety of mouse fibroblastic cells. Genes Dev. 8, 1654 – 1663 (1994). 708. P. Tontonoz, E. Hu, and B. M. Spiegelman, Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79, 1147 – 1156 (1994). 709. E. Hu, P. Tontonoz, and B. M. Spiegelman, Transdifferentiation of myoblasts by the adipogenic transcription factors PPAR gamma and C/EBP alpha. Proc. Natl. Acad. Sci. USA 92, 9856 – 9860 (1995). 710. D. J. Rickard, M. Kassem, T. E. Hefferan, G. Sarkar, T. C. Spelsberg, and B. L. Riggs, Isolation and characterization of osteoblast

711. 712.

713.

714.

precursor cells from human bone marrow. J. Bone Miner. Res. 11, 312 – 324 (1996). N. Lenny, S. Meyers, and S. Hiebert, Functional domains of the t(8;21) fusion protein, AML-1/ETO. Oncogene 11, 1761 – 1769 (1995). S. Meyers, N. Lenny, W.-H. Sun, and S. W. Hiebert, AML-2 is a potential target for transcriptional regulation by the t(8;21) and t(12;21) fusion proteins in acute leukemia. Oncogene 13, 303 – 312 (1996). Y. Ito, and S. C. Bae, The runt domain transcription factor, PEBP2/CBF, and its involvement in human leukemia. In “Oncogenes as Transcriptional Regulators” (M. Yaniv and J. Ghysdael, eds.), Vol. 2, pp. 107 – 132. Birkhauser Verlag, Basel, Switzerland. T. Kanno, Y. Kanno, L. F. Chen, E. Ogawa, W. Y. Kim, and Y. Ito, Intrinsic transcriptional activation-inhibition domains of the polyomavirus enhancer binding protein 2/core binding factor alpha subunit revealed in the presence of the beta subunit. Mol. Cell Biol. 18, 2444 – 2454 (1998).

CHAPTER 3

Osteoclast Biology F. PATRICK ROSS AND STEVEN L. TEITELBAUM Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110

I. II. III. IV. V.

VI. Mechanisms of Bone Resorption VII. Humoral Regulation of Osteoclastic Bone Resorption VIII. Diseases of the Osteoclast References

Introduction Osteoclast Morphology Origin of Osteoclasts Models of Osteoclast Function Osteoclast Attachment and Polarization

I. INTRODUCTION

polykaryons generally juxtaposed to bony surfaces and often in resorption pits (Howship’s lacunae) (Fig. 1). These excavations are the products of the capacity of the osteoclast to degrade both organic and inorganic matrices of bone, an event in which mineral removal precedes collagenolysis [1]. Consequently, resorption pits are lined by demineralized collagen, which has not yet been degraded, a phenomenon best appreciated by scanning electron microscopy [2] (Fig. 2). All forms of adult osteoporosis reflect enhanced bone resorption relative to formation and should be viewed in the context of the remodeling cycle. Bone remodeling, a process characterized by the coupling of osteoclast and osteoblast recruitment, occurs throughout life in thousands of sites within the human skeleton. While the fundamental purpose of bone remodeling is unknown, it probably serves to replace effete, aged bone with that which is newly synthesized. Remodeling is initiated by osteoclasts, or their precursors, attaching to trabecular or endosteal bone surfaces. The mechanism by which the osteoclast binds to bone has been a focus of intense investigation and its recent unraveling promises to yield novel approaches to osteoporosis prevention. In this regard, it has been hypothesized that a thin layer of unmineralized type I collagen covering the surface of bone must be removed prior to osteoclast attachment [3]. Degradation of this surface collagen is proposed to be the

Osteoporosis, regardless of etiology, always represents enhanced bone resorption relative to formation. Thus, insights into the pathogenesis of this disease, and progress in its prevention and/or cure, depend on understanding the mechanisms by which bone is degraded. The osteoclast is the principal, if not exclusive, resorptive cell of bone. It is a member of the monocyte/ macrophage family, and most successful treatments of osteoporosis, to date, target osteoclastic bone resorption. Major advances have been witnessed in delineating the mechanisms by which osteoclasts are generated from their precursors and stimulated to degrade bone once they are fully differentiated. In fact, the essential osteoclastogenic molecules are now in hand and, most importantly, are the targets of rational antiosteoporosis drug design. These potential therapeutic agents reflect advances made in understanding osteoclast biology.

II. OSTEOCLAST MORPHOLOGY The osteoclast is a multinucleated cell whose capacity to degrade hard tissues depends on cell/matrix contact. Thus, when viewed in histological sections, osteoclasts appear as

OSTEOPOROSIS, SECOND EDITION VOLUME 1

73

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

74

ROSS AND TEITELBAUM

FIGURE 1 Osteoclasts (arrows) in resorption bay (Howship’s lacuna). Resorptive cells are juxtaposed to bone, and their nuclei are polarized toward the antiresorptive plasma membrane (nondecalcified, Goldner stain). consequence of secretion of neutral collagenase by osteoblasts [4]. While provocative, this hypothesis is unproven and, in fact, the putative layer of lining collagen may be an artifact of tissue fixation.

FIGURE 2

Upon attaching to a nascent remodeling site, osteoclasts polarize their resorptive machinery (infra vide) toward the cell/bone interface and generate a Howship’s lacuna approximately 50 m deep. Once the degradative phase of

Scanning electron micrograph of resorption pit formed on bone slice by isolated avian osteoclast. Bundles in the bottom of the pit represent bone collagen that the cell has demineralized in preparation for degradation.

CHAPTER 3 Osteoclast Biology

the remodeling cycle is complete, osteoclasts are replaced in the resorptive bay by mononuclear cells of unknown origin, which in turn are replaced by cells of osteoblastic lineage. Histological sections of normal human osteoclasts contain varying numbers of nuclei generally maximizing at 10. Nuclear number may mirror osteoclast activity. For example, osteoclasts in Paget’s disease, a disorder of accelerated resorption, are enormous and often contain as many as 100 nuclei [5]. Alternatively, hypernucleated osteoclasts are also encountered in circumstances of resorptive dysfunction, such as various forms of osteopetrosis. Whether the abundant osteoclast nuclei seen in some osteopetrotic patients reflect increased levels of osteoclastogenic agents, such as parathyroid hormone, or an abnormality intrinsic to the osteoclast per se is not known. Osteoclast nuclei are distinct from one another, a feature useful in distinguishing the cell from the megakaryocyte. Moreover, osteoclast nuclei are typically polarized away from the plasma membrane juxtaposed to bone, residing closest to the antiresorptive surface of the cell. Formation of the osteoclast polykaryon probably depends on matrix attachment. Furthermore, osteoclast multinucleation is not the product of endomitosis, as mitotic figures are rarely encountered. However, when placed in culture with mononuclear phagocytes, osteoclasts incorporate new nuclei and expel others [6]. Thus, the most compelling hypothesis holds that once an osteoclast, or its progenitor, attaches to a putative resorptive site, additional mononuclear precursors that fuse with the immobilized cell are recruited. While the life span of osteoclasts is unknown, they clearly undergo programmed cell death. Increased numbers of apoptotic osteoclasts appear when exposed to specific families of bisphosphonates, particularly those that modulate the mevalonate pathway [7]. Alternatively, agents such as interleukin-1 (IL-1), M-CSF, and RANK-ligand (RANKL) extend the cell’s survival (vide infra). Osteoclasts are rich in lysosomal enzymes, with the most studied being tartrate-resistant acid phosphatase (TRAP). While presence of this enzyme in histological sections serves to identify osteoclasts, the bone isoform of TRAP is also expressed by macrophages derived from other organs, most notably lung and spleen [8,9]. At first glance these data would suggest that TRAP is of little use in identifying osteoclasts formed in culture. Alternatively, Suda and colleagues generated bona fide osteoclasts in vitro from alveolar and splenic macrophages, and the same is true regarding circulating human monocytes [10], suggesting that TRAP-expressing cells resident in these tissues are, in fact, capable of osteoclastogenesis [11]. Similarly, the antiosteoclastogenic cytokine interleukin-4 (IL-4) prompts appearance of TRAP-negative polykaryons in culture, thereby associating resorptive activity with expression of

75 the enzyme [12]. Thus, while osteoclasts generated in vitro from macrophages invariably express TRAP, confirmation of their identity requires demonstration of other phenotypic features, preferably the capacity to resorb bone. Because osteoclasts are enormous, often larger than 100 m in diameter, histological sections, typically 5 to 10 m thick, accommodate only a small proportion of each cell. Reflecting the size and complexity of the osteoclast, a given section may contain apparently separate fragments of the same cell, generally considered distinct osteoclasts by bone histomorphometrists. Thus, so-called mononuclear osteoclasts may represent a portion of a multinucleated cell. The likelihood of multiple fragments of the same cell in a single slide underscores the limitations of attempting to distinguish altered numbers of osteoclasts from changes in size or plasma membrane complexity. Being a member of the monocyte/macrophage family, osteoclasts share a number of morphological features with foreign body giant cells, such as abundant lysosomes. However, the unique capacity to resorb bone endows the osteoclast with morphological features distinct from those of other related macrophage-like cells. For example, unlike foreign body giant cells, osteoclasts are rich in mitochondria and, in fact, probably represent the cell containing the greatest density of these organelles [13]. Clearly, however, the morphological feature of the osteoclast that most dramatically distinguishes it from related cells is its ruffled membrane (Fig. 3). This complex infolding of plasma membrane, polarized to the cell/bone interface, is best appreciated by transmission electron microscopy. The ruffled membrane is unique to the osteoclast and matrix specific, appearing only when the cell is in contact with bone. Because of the enormity of an osteoclast, its plasma membrane may touch bone in more than one location and thus separate sites of ruffling will develop within the same cell. Moreover, the extent of ruffled membrane formation appears to parallel the degradative activity of the cell and its exposure to resorptive agonists or antagonists [14]. In fact, as osteoclasts alternate between their resorptive/adherent and mobile/detached states, they respectively express and lose their ruffled membranes. While yet to be visualized in situ, ruffled membrane formation probably represents the insertion of proton pumpbearing vesicles into the plasma membrane, a process similar to exocytosis (Fig. 4). In fact, the small GTPase, Rab-3, which modulates the fusion of exocytic vesicles to the plasma membrane, may also regulate ruffled membrane formation [15]. The ruffled membrane is structurally distinct from the nonresorptive plasma membrane in that it contains abundant, “spike-like” structures shown by light [16] (Fig. 5, see also color plate) and ultrastructural immunocytochemistry to represent vesicular proton pumps. These same projections are present in acidifying vesicles within the cytoplasm. This

76

ROSS AND TEITELBAUM

FIGURE 3

Transmission electron micrograph of osteoclast attached to bone (B). The convoluted infolding of plasmalemma is the ruffled membrane (R). The sealing or “clear” zone (C) is rich in actin filaments oriented perpendicular to the bone surface. Note the abudance of mitochondria (arrows).

observation supports the hypothesis that these proton pumpbearing vacuoles are precursors of the ruffled membrane. While its role, if any, in the resorptive process is not clear, TRAP activity also polarizes to the ruffled membrane and, like other lysosomal enzymes, is secreted into the resorptive microenvironment [17], where it dephosphorylates osteopontin, thereby altering the adhesive properties of this important bone matrix protein [18]. Osteoclasts contain a prominent cytoskeleton, components of which are critical to resorptive function. Microtubules extend from organizing centers to the periphery of the cell. These structures polymerize on cell/substrate adherence and, in so doing, acquire the capacity to transport vesicles [19]. Thus, microtubules are likely to play a central role in osteoclast polarization, delivering proton pump-containing vacuoles for insertion into the resorptive membrane [20] (Fig. 6). The intimacy between osteoclasts and bone required for resorption is reflected, in the cell, by the matrix attachment or “sealing” zone (Fig. 3). This distinct morphological entity, organized as a ring, completely surrounds the ruffled membrane. While organelle free, and thus also referred to as the “clear zone,” this attachment area is, in actuality, rich in microfilaments polarized perpendicularly to the bone surface [21]. The sealing zone is a well-demarcated structure, not unique to the osteoclast, and is also formed

when other members of the macrophage/monocyte family attach to the matrix. Like microtubules, microfilaments organize upon adherence to bone, and the filamentous distribution of F-actin directly correlates with the resorptive activity of osteoclasts [22] (Fig. 7, see also color plate). In fact, when osteoclasts are in the process of degrading bone, Factin localizes in a ring-like structure of punctate plasma membrane protrusions known as focal adhesions or podosomes [23] (Fig. 8). In addition to F-actin, these structures contain matrix-recognizing integrins and proteins such as vinculin and talin, which link these attachment heterodimers to the cytoskeleton [24]. Thus, actin filaments in osteoclasts probably anchor the plasma membrane to the cytoskeleton and are central to the means by which osteoclasts recognize bone. Following formation of a resorptive lacuna, the cell becomes motile, an event attended by dissolution of the actin ring. Evidence in hand indicates that the cyclical mobilization of the actin cytoskeleton in osteoclasts is under the aegis of the Rho family of GTPases [25]. While the magnitude of podosome expression appears to mirror the degradative activity of the cell [26], the punctate appearance of these attachment structures suggests that they do not serve as the osteoclast’s “tight seal.” Unraveling the molecular mechanisms serving to isolate the resorptive microenvironment

77

CHAPTER 3 Osteoclast Biology

FIGURE 4

Model of ruffled membrane formation. In the nonattached state the osteoclast is unpolarized, with acidifying vesicles distributed throughout its cytoplasm. Once in contact with bone, matrix-derived signals, probably mediated via integrins such as v3, prompt targeting of acidifying vesicles to the apical (resorptive) surface of the cell. Insertion of these vesicles into the bone-adjacent plasma membrane yields the characteristic, highly ruffled resorptive surface of the cell. The “spikes” within the vesicles represent the H-ATPase (proton pump).

of the osteoclast is a challenge likely to yield important insights into regulating its activity.

III. ORIGIN OF OSTEOCLASTS Avian osteoclasts (OC) immunostained for vacuolar HATPase (proton pump). (A) Cells are juxtaposed to bone, and the pump (brown-staining reaction product, arrows) is polarized toward the cell – bone interface. (B) Osteoclasts are unattached to bone, and HATPase is distributed diffusely throughout the cytoplasm. Reprinted with permission from Blair et al., Science 245, 855 – 856. Copyright 1989 American Association for the Advancement of Science. (See also color plates).

FIGURE 5

The hematopoietic origin of osteoclasts is now in hand. Its recognition, however, follows a contentious history with initial debate focusing on whether osteoclasts and osteoblasts derive from a common precursor [27]. Early attempts to resolve this controversy involved experiments in which the circulations of two rats were joined. Using this parabiotic approach, Gothlin and Ericsson (reviewed in Ref. 28) established that osteoclasts migrating to a fracture are derived from the blood of its partner. In contrast, osteoblasts are not donor derived, suggesting that their ontogeny differs from that of bone-resorbing cells.

The hematopoietic origin of osteoclasts is also suggested by experiments in which quail cartilaginous limb buds were transplanted onto chick chorioallantoic membranes [29]. The host circulation vascularized the rudiments, which

78

ROSS AND TEITELBAUM

FIGURE 6 Model of osteoclast polarization. In the nonattached state the osteoclast is unpolarized, with acidifying vesicles distributed throughout its cytoplasm. Once in contact with bone, matrix-derived signals, probably mediated via integrins such as v3, prompt targeting by trafficking along microtubules of acidifying vesicles to the apical (resorptive) surface of the cell. Insertion of these vesicles into the bone-adjacent plasma membrane yields the characteristic, highly ruffled resorptive surface of the cell.

FIGURE 8 Isolated murine osteoclast stained with rhodamine phalloidin to delineate F-actin. The fluorescent rings are well-organized sealing zones containing several layers of punctate, focal adhesions, or podosomes.

FIGURE 7 Varying organization of the actin cytoskeleton during the different stages of osteoclast activity. Prior to resorptive phase (stages 1 – 3), actin (red) and vinculin (green) are found in punctate, podosomal structures. During resorption (stage 4), podosomes are lost and actin forms a dense continuous band surrounded by a double ring of vinculin and talin (not shown). During the postresorptive period (stage 5), the system reverts toward its original state. (See also color plate.)

79

CHAPTER 3 Osteoclast Biology

ultimately developed into bone. Most of the osteoclasts present in the neovascularized graft were of chick origin, indicating that they were derived hematopoietically. The hematopoietic origin of osteoclasts, and its distinction from that of osteoblasts, is also established by studies of murine and human osteopetrosis, a disease discussed in detail later (Section VIII). Given the hematopoietic derivation of the osteoclast, efforts turned to identifying its precursor. Because of phenotypic similarities, the cell was suspected of membership in the monocyte/macrophage family. Initial experiments aimed at determining if such is the case involved the injection of thorium dioxide-labeled macrophages into sibling rats who immediately thereafter had a femur fractured. Initially, only labeled monocytes and macrophages were present in the fracture but, ultimately, thorium dioxide-bearing osteoclasts appeared. These observations prompted the search for functional similarities between osteoclasts and macrophages [28]. Early in vitro studies suggested that peripheral blood monocytes and elicited rodent peritoneal macrophages are capable of degrading bone [30 – 32], but it is unlikely that the models used in these experiments are reflective of bona fide resorption. Moreover, despite formation of TRAP-expressing polykaryons by monocytic leukemic lines placed in osteoclastogenic conditions [33], these cells are not authentic osteoclasts. The first successful attempt at generating osteoclasts in vitro utilized long-bone primordia [34]. Culture of these rudiments with embryonic liver and marrow mononuclear cells, but not peripheral blood monocytes or peritoneal macrophages, propagates osteoclasts. Colony-forming unitgranulocyte – macrophage (CFU-GM) or more primitive stem cells are also capable of differentiating into bone resorptive polykaryons. The suggestion that osteoclasts are derived from pluripotent hematopoietic stem cells [35] is supported by experiments in which spleen cells taken from 5-fluorouracil-treated mice were cultured with IL-3 [36]. Blast cell colonies so formed mature into TRAP-expressing polykaryons responsive to calcitonin and are capable of resorbing bone. In contrast, mature macrophages are unable, in this system, to differentiate into osteoclasts. The experiments discussed thus far indicate that isolated, primitive hematopoietic precursors, but not more differentiated macrophages, are capable of osteoclast differentiation in vitro. The studies of Takahashi et al. [37] provided insight into the apparent osteoclastogenic capacity of immature and mature macrophages. In these experiments, mouse marrow cells in the presence of 1,25(OH)2D3 or parathyroid hormone (PTH) formed osteoclasts [38]. Interestingly, most TRAPexpressing polykaryons appeared near alkaline phosphataseexpressing cells, suggesting that osteoblasts play a role in osteoclast differentiation. This group next demonstrated that osteoclasts are induced by coculture of mouse spleen cells and osteoblasts [39]. Furthermore, marrow-derived stromal

cell lines may substitute for osteoblasts in inducing osteoclastogenesis [38,40]. Most importantly, in this coculture system, mature monocytes and macrophages differentiate into osteoclasts [11]. As will be discussed, the means by which stromal cells and osteoblasts promote osteoclastogenesis are in hand. Colony-forming assays, along with osteoclast and macrophage markers, serve to define further the lineage of human osteoclasts. CD34 hematopoietic stem cells, maintained with granulocyte – macrophage colony-stimulating factor (GM-CSF), form CFU-GM colonies [41]. Treatment of these colonies with 1,25(OH)2D3 generates granulocyte, macrophage, and mixed colonies, as well as those consisting of polygonal cells. In defined conditions, only CFUGM and polygonal cell colonies yield osteoclast-like cells, a few of which are capable of forming small resorptive pits. More recent studies, discussed in Section VII, provide further evidence that cells in the myeloid lineage contain osteoclast precursors. Thus, the hematopoietic origin of osteoclasts is established as is the cells’ membership in the monocyte/ macrophage family. Moreover, it is likely that macrophages at various maturational states may differentiate into osteoclasts.

IV. MODELS OF OSTEOCLAST FUNCTION The majority of resorptive experiments in whole animals involve the administration of drugs, systemic hormones, and cytokines (reviewed in Ref. 42). Additional studies include those exploring the effects of weightlessness [43,44], a relevant question in view of the possibility of long-term space travel. Interpretation of these experiments has depended on the quantitation of osteoclast number or net resorptive activity. The advent of densitometric techniques applicable to small animals [45] and of assays for products of rodent bone degradation [46] has increased the usefulness of these models. This approach has yielded important phamacological information, but the possibility that the target for the intervention is not the osteoclast itself, but rather an intermediary cell which modulates the bone resorptive polykaryon, is a major problem confounding whole animal experiments. However, gene deletion techniques have led to invaluable information regarding the molecular mechanisms of osteoclastogenesis in vivo. Fetal mouse [34] and rat bone cultures [47] utilizing intact explanted rudiments facilitate the exploration of paracrine, as opposed to endocrine factors in the resorptive process. The presence of an infinite variety of other cells confounds the direct assessment of osteoclast function. None the less, these systems continue to yield important data regarding specific aspects of osteoclast biology. Bone biopsy has been used as a means of estimating osteoclast activity in humans [48]. In addition to its invasive

80 nature, this approach is limited in that it does not actually measure resorptive activity but provides an estimate of osteoclast number. Noninvasive determinants of human osteoclast function include urinary markers such as hydroxyproline [49] and the dehydropyridinoline moiety [50]. This last approach utilizes a sensitive enzyme-linked immunosorbent assay (ELISA) to detect fragments arising directly from the degradation of mature bone collagen and is probably the current method of choice. Highly enriched or pure populations of osteoclasts would obviate many of the difficulties attending whole animal or organ culture systems. Early attempts utilized enzymatic digestion of bone and/or bone marrow suspensions, coupled with rapid sedimentation yielding small numbers of both mammalian osteoclasts [51 – 54]. The discovery that hens fed a low-calcium diet mobilize abundant osteoclasts [55] permits isolation of these cells in numbers sufficient for biochemical studies, including characterization of an acidic, cathepsin B-like collagenolytic enzyme [56] and a reconstitutible proton pump complex [57]. The same cells, following extensive purification, were used to confirm the presence of pp60c-src in osteoclasts [58], a protein essential to the cells’ polarization [59]. Because avian osteoclasts probably lack the calcitonin receptor, their relationship to mammalian osteoclasts is tenuous. Large numbers of rabbit osteoclasts have been obtained by simple adhesion techniques [60]. Sequencing of the cDNA library generated from these cells confirmed the presence of mRNA for osteopontin and cathepsin K [60,61]. Finally, freshly isolated rat osteoclasts lend themselves to the exploration of intracellular signaling [62,63]. A related strategy, applied to both human and avian systems, is based on immunoisolation of osteoclasts using monoclonal antibodies to suitable surface markers [64 – 66]. Such antibodies were raised to human osteoclastomas and screened for their ability to inhibit bone resorption in vitro. A single clone, 23C6, recognizes the integrin receptor, v3 [67]. This reagent has been used as a blocking antibody and to obtain highly purified polykaryons from which cDNA libraries have been generated [68]. Expression cloning of such a library led to the discovery that annexin II, a putative plasma membrane calcium channel, is secreted by osteoclasts and stimulates their activity [69]. In analogous studies, monoclonal antibodies were raised to purified avian osteoclasts [64]. Such an antibody, 121F, identifies a cell surface protein related structurally to superoxide dismutase [70]. The potential importance of this finding is underscored by the fact that oxygen-free radicals inhibit osteoclastic bone resorption [71]. While the ability to isolate and study mature osteoclasts is essential to understanding the mechanisms by which they degrade bone, the use of differentiated resorptive polykaryons precludes the identification of osteoclast precursors and the biochemical signals regulating osteoclastogenesis. Thus,

ROSS AND TEITELBAUM

efforts have focused on the generation of osteoclasts in vitro. The first successful attempts utilized mononuclear cells derived from calcium-deficient hens known to contain many mature osteoclasts [55]. When these TRAP-negative monocytic cells are cultured, at high density, they fuse within 5 – 6 days, yielding an almost homogeneous population of bone resorptive polykaryons with many, but not all, components of the osteoclast phenotype [2]. The large numbers of both precursors and fused cells allow for performance of a range of biochemical and cell biological studies. Thus, mature cells were characterized with respect to the integrins present and, most importantly, the ability of an antibody against the integrin v3 to block bone resorption [72]. Given that 1,25(OH)2D3 and retinoic acid stimulate osteoclastogenesis in vivo, the precursors served as a means to examine regulation, by these hormones, of this functionally important integrin [73,74]. In fact, the promoter of the 3 gene contains both vitamin D and retinoic acid responsive elements [75,76]. The same generated avian polykaryons have been used to demonstrate a decrease in intracellular calcium concentrations following liganding of the integrin v3 with soluble peptides [77]. Attempts to generate mammalian osteoclasts in vitro have also been rewarding. In early studies, unfractionated marrow cultures, treated with 1,25(OH)2D3, led to the production of small numbers of rabbit, primate, and human osteoclast-like cells [78 – 80]. The cells, most notably those of human origin, exhibit a limited ability to excavate characteristic resorption pits. In contrast, primary murine osteoblasts [39] and several clonal stromal cells lines [11,40], when cocultured with purified murine monocytic precursors, engender greater numbers of multinucleated cells, which resorb bone avidly and express the complete osteoclast phenotype. This system, which depends on contact between stromal cells/osteoblasts and osteoclast precursors, has been used to demonstrate that prostaglandin E2 and IL-4 have opposite effects on osteoclastogenesis [81]. Using the same assay and a function-blocking antibody to macrophage colony-stimulating factor (M-CSF), it was possible to reproduce, in vitro [82], in vivo data concerning the essential nature of this cytokine (M-CSF) for osteoclast production [83,84]. Studies have elucidated the minimal requirements for the generation of murine and human osteoclasts. In short, three proteins are sufficient to induce the differentiation of macrophages into osteoclasts. These osteoclastogenic proteins are M-CSF, receptor for activation of nuclear factor-kB (RANK), and RANK ligand (RANK-L).1 Interestingly, the soluble protein, osteoprotegerin (OPG), competes with RANK as a decoy receptor for RANK-L and thus attenuates osteoclast differentiation. Human bone marrow-derived CD34 cells [85,86] and murine 1 RANK-L is synonymous with osteoprotegerin ligand (OPGL), TRANCE, and osteoclast differentiation factor.

81

CHAPTER 3 Osteoclast Biology

spleen- or bone marrow-derived macrophages serve as osteoclast precursors, targeted by RANK-L and M-CSF [87,88]. RANK-L exists as soluble and membrane-bound forms and is produced by vascular endothelium [89], activated T cells [90], and, probably of most significance to osteoclast biology, those of the osteoblast/stromal lineage. This observation explains the mechanism by which osteoblasts and marrow stromal cells prompt macrophage differentiation into osteoclasts. OPG is also a product of osteoblastic/stromal cells [91], whereas RANK is found on cells of the osteoclast lineage [92,93]. With these proteins in hand, it is now feasible to generate large numbers of highly purified osteoclasts, using as targets, cells of the macrophage/monocyte lineage [88], now known to be osteoclast precursors. This approach has facilitated the execution of a wide range of biochemical and molecular experiments, adding substantially to our understanding of bone biology, a subject discussed further in Section VII.

V. OSTEOCLAST ATTACHMENT AND POLARIZATION As a prerequisite to initiating resorption, the osteoclast must attach to bone and create a functional, polarized plasma membrane. This binding event involves interactions between membrane-bound receptors on the cell surface and matrix proteins in bone. A number of surface markers have been identified on osteoclasts and/or their precursors [94], but, with the exception of integrins, few have been characterized functionally with respect to attachment. Antibody-blocking experiments suggest that members of the 1 family of integrins may participate in the resorptive process, presumably by binding to type I collagen [95]. However, the most compelling data indicate that v3 is the integrin most essential to osteoclast activity. Initial studies implicating this heterodimer in bone degradation were performed in vitro using antibodies to v3 [72] or peptides that block its function [96]. These experiments indicate that the integrin, on osteoclasts, binds one or more bone matrix proteins containing the motif Arg-Gly-Asp (RGD) [72,97 – 100]. The most convincing evidence that v3 is essential to osteoclast function comes, however, from the 3 knockout mouse [101]. This animal has dysfunctional osteoclasts as evidenced by their failure to form actin rings, a normal ruffled membrane or to resorb bone in vitro. Most importantly, these mice, which fail to express v3, are hypocalcemic and become progressively osteosclerotic with age. These studies suggest that v3 is a promising therapeutic target to inhibit accelerated bone resorption as occurs in postmenopausal osteoporosis. In fact, an organic molecule mimicking RGD prevents the profound bone loss rapidly following oophorectomy in the rat [102].

While the precise bone matrix protein(s) recognized by v3 the integrin in vivo is not known, osteopontin is a reasonable candidate. For example, immunoelectron microscopy suggests that v3 and OPN colocalize in bone [103], and an immunopurified antibody to OPN inhibits osteoclast/bone interaction [72]. Perhaps related more directly to polarization, the avian osteoclast cytoskeleton reorganizes upon v3 occupancy by soluble osteopontin [104]. In this instance, the mechanism involves activation of the important intracellular enzyme phosphoinositol-3 kinase, inhibition of which arrests bone resorption in vitro [105]. Activated phosphoinositol-3 kinase generates phosphotidylinositol-3,4 bisphosphate. This phospholipid, by binding to the actin-capping protein gelsolin, causes actin depolymerization and subsequent reelongation [106]. The most compelling evidence that osteopontin is pivotal to osteoclast function is the fact that the osteopontin knockout mouse is protected from oophorectomy-induced osteoporosis [107]. While the v3 integrin is pivotal to the resorptive process, its most important function is probably not formation of the tight seal, but transmission of bone-derived, intracellular signals. Thus, plating of mononuclear cells committed to osteoclast differentiation on a matrix recognized by v3 activates c-src [108], a plasma membrane protooncogene essential for osteoclast polarization and bone resorptive activity [109]. While it has been hypothesized that activated c-src associates with the tyrosine kinase pyk-2, which prompts a cascade of events probably responsible for organizing the osteoclast cytoskeleton during bone resorption [108], this remains to be proven. The dramatic polarization of the osteoclast, a phenomenon probably mediated by v3, distinguishes the cell from other members of the monocyte/macrophage family. It is clear that osteoclast polarization requires bone matrix recognition. The means by which the most dramatic manifestation of such polarization, namely the ruffled membrane, is generated are beginning to emerge. The information in hand suggests that the process involves migration, via microtubules, of acidifying vesicles from the trans-Golgi network to the bone-apposed plasma membrane into which they insert under the aegis of a Rab GTPase [15]. Thus, ruffled membrane formation is a process akin to regulated exocytosis, eventuating in focal complexity of the plasma membrane due to vesicular incorporation. A provocative alternative hypothesis holds that it is the endosomal, rather than exocytic, pathway that contributes to the ruffled membrane [110].

VI. MECHANISMS OF BONE RESORPTION The functional role of the osteoclast is to resorb bone, a composite matrix consisting of both inorganic and organic elements. The inorganic component is largely substituted

82

ROSS AND TEITELBAUM

FIGURE 9

Essential components of osteoclast-mediated bone degradation. See text for details.

hydroxyapatite, whereas the organic phase contains 20 or more proteins, with type 1 collagen the single major species (90% of total protein by weight) [111]. The initial step in bone resorption is attachment of the cell to the matrix, followed by the creation of an isolated, extracellular, resorptive microenvironment bordered by the ruffled membrane. Events leading to attachment and formation of the resorptive space have been discussed earlier. Studies with isolated osteoclasts reveal that dissolution of the inorganic phase of bone precedes that of protein [1]. Demineralization involves acidification of the extracellular microenvironment [112] (Fig. 9), a process mediated by a vacuolar H-ATPase in the ruffled membrane of the polarized cell [113] (Fig. 5). The structure and functional activity of this multienzyme complex are very similar, if not identical, to that of the analogous proton pump in the intercalated cell of the kidney [114,115]. This pump is a multimer, in which only some of the units are intrinsic membrane proteins. Others, including the 70-kDa protein containing ATPase activity, are attached, noncovalently, to subunits buried in the membrane. It is possible that one or more subunits may be present as an isotype, producing an osteoclast-specific form of the pump [116]. In support of this complex being the critical moiety in osteoclast acidification, the fungal metabolite bafilomycin A, a potent and specific inhibitor of all vacuolar proton

pumps, blocks bone resorption [117]. The intact proton pump has been isolated from avian osteoclasts, and the identity of several subunits to those present in other vacuolar pumps was established by Western analysis [57]. Importantly, the activity of the isolated complex is restored by the incorporation into lipid vesicles [57]. As with all members of the vacuolar pump family, extrusion of a proton through the plasma membrane is accompanied by the hydrolysis of one molecule of ATP. Given the fact the osteoclast transports protons extracellularly by an electrogenic mechanism raises the issue of maintenance of intracellular pH. Turning to the paradigm of the renal intercalated cell, Teti et al. [118] found that osteoclasts express, on their antiresorptive border, an energy-independent Cl/HCO3 exchanger similar to band 3 of the erythrocyte. Finally, electroneutrality is preserved by a plasma membrane Cl channel, charge coupled to the HATPase, resulting in the secretion of HCl into the resorptive microenvironment [119,120]. Because blockade of the Cl channel arrests H secretion, it is likely that impaired anion transport also impedes bone resorption. Acidification of the isolated resorptive environment prompts mineral mobilization as well as subsequent solubilization of the organic phase of bone [1], the products of which are endocytosed by the osteoclast, and transported to, and released at its antiresorptive surface [121,122].

83

CHAPTER 3 Osteoclast Biology

Given the low pH at the osteoclast/bone interface, the lysosomal family of acidic hydrolases, delivered to the resorptive microenvironment via the mannose-6-phosphate receptor [123,124], presented themselves as candidate molecules to degrade bone collagen and noncollagenous proteins. In fact, there is compelling evidence that, in the mammalian osteoclast, cathepsin K fulfills this role. Cathepsins have the capacity to fragment bone organic matrix [56] and, most importantly, mice deleted of the cathepsin K gene have increased skeletal mass [125]. Thus, cathepsin K is a potential antiosteoporosis target. Several studies reveal a possible role for neutral collagenases in the early phase of bone resorption. Bone is believed to be covered by a thin layer of osteoid (unmineralized matrix), which, some posit, needs be removed prior to osteoclastic degradation [4]. This hypothesis is, however, controversial, and the osteoblast is proposed to be the source of the putative neutral collagenase in this circumstance [126 – 129]. The most compelling evidence that these enzymes participate in the resorptive process comes from the demonstration that mutation in vivo of the site in type 1 collagen targeted by neutral collagenases attenuates bone resorption [130]. In addition to degrading bone, the osteoclast also resorbs mineralized cartilage in the hypertrophic zone of the growth plate. There is little evidence that the so-called “chondroclast” is distinct from the osteoclast; in fact, cartilage resorption by the cell is mediated via matrix metalloproteinase-9 [131]. While the means by which osteoclasts degrade bone are reasonably well defined, less is known regarding the mechanisms terminating their activity. The most provocative argument holds that sensing a high ambient Ca2 within the resorptive space, by a plasma membrane cation receptor, prompts withdrawal of the osteoclast from the bone surface and terminates resorption [132].

osteoclast recruitment. Injection of RANK-L into mice results in a rapid increase in serum calcium, indicating that the molecule also stimulates the resorptive activity of mature osteoclasts [134]. Targeted deletion of the RANK-L, OPG, and RANK genes prompts phenotypes consistent with the role of each protein as described in the current model. Thus, mice lacking either RANK-L [135] or RANK [136] are severely osteopetrotic secondary to lack osteoclasts, while the absence of OPG results in profound osteoporosis [137]. Binding of M-CSF and RANK-L to their respective receptors, c-fms and RANK, is the necessary and sufficient event to initiate osteoclastogenesis. Discovery of these molecules has yielded insights into the interplay between regulatory humoral and local factors and the intracellular signals controlling formation of the osteoclast (Fig. 10, see also color plate). In brief, a range of hormones, cytokines, and growth factors, targeting primarily to mesenchymederived cells in the bone microenvironment, control the expression of M-CSF, RANK-L, and OPG, with the overall impact on osteoclast recruitment dependent mainly on the ratio of OPG to RANK-L. Finally, the macrophage products, IL-1, TNF, and IL-6, regulate the capacity of stromal cells to promote osteoclast precursor differentiation (reviewed in Ref. 138). These proinflammatory cytokines also directly target myeloid precursors and mature osteoclasts, initiating signals prompting cell fusion and altered function and survival. Treatment of osteoclast precursors with IL-1 accelerates fusion and generates bone-resorbing polykaryons [139]. The cytokine also increases the life span of mature osteoclasts by a mechanism involving the activation of NF-B [140]. A more controversial

VII. HUMORAL REGULATION OF OSTEOCLASTIC BONE RESORPTION As discussed earlier, four molecules, M-CSF, RANK, RANK-L, and OPG, acting in concert, are major regulators of osteoclast formation and function. OPG, a member of the TNF receptor family, inhibits osteoclast formation. RANK-L, a novel member of the TNF superfamily, is the OPG ligand. Whereas RANK-L is primarily membrane bound, it also exists as a soluble form. Not surprisingly, RANK, the receptor for RANK-L, expressed on osteoclasts and their precursors, is a TNF receptor-related protein. The fact that purified macrophages give rise to numerous functional osteoclasts, when treated only with M-CSF and RANK-L [88] and that osteoclastogenesis is blocked by OPG [133], establishes these molecules as central to

FIGURE 10

Regulation of osteoclast formation. A range of hormones and cytokines, targeting to stromal cells, enhance expression of M-CSF and RANK-L, the pivotal osteoclastogenic molecules, and the proinflammatory cytokines IL-1, IL-6, and TNF. These proteins act to stimulate both formation and activity of mature polykaryons. HSC, hematopoietic stem cell. (See also color plate.)

84

FIGURE 11

Key elements of RANK signaling. Receptor ligation is followed by the recruitment of adaptor molecules, including TRAF6, which, interacting with c-src, stimulates the PI-3 K/Akt pathway. Additionally, the NF-B and AP-1 families of transcription factors, key elements in osteoclast formation and function, are activated. Akt plays a role in phosphorylation of the IKK complex [367]. (See also color plate.)

hypothesis holds that TNF induces osteoclast formation in a manner not requiring RANK-L [141]. While the extracellular domain of RANK identifies it as a member of the TNF receptor superfamily, the intracellular amino acid sequence is unlike that of all orthologs. Despite these differences, the known proximal and distal signaling events arising from ligation of RANK and other members of the family are similar (Fig. 11, see also color plate). Two critical distal events in RANK signaling are the activation of JNK and the NF-B transcription complex. Mice lacking both the p50 and p52 NF-B subunits fail to generate osteoclasts [142]. Similarly, the AP-1 family proteins, c-fos and c-jun, are targets of JNK and play a role in osteoclast formation. Thus, animals lacking c-fos generate no multinucleated, bone-resorbing cells [143], nor do precursors which overexpress a form of c-jun containing a mutation that renders the transcription factor nonactivatable by JNK [144]. The proximal RANK-initiated signals are more complex and may reflect those following ligation of other members of the TNF receptor superfamily. For example, activation of the type 1 TNF receptor, which promotes osteoclast formation [145], is followed by binding of TRAFs, adaptor proteins capable of driving a variety of downstream signals [146]. Similarly, overexpressed RANK in embryonic human kidney cells recruits TRAFs 1, 2, 3, and 5 [147]. Because mice lacking TRAF 2, 3, or 5 have normal bones [148 – 150], these molecules may not be necessary for RANK-induced osteoclastogenesis. However, these three TRAFs bind to overlapping sites in the RANK cytoplasmic tail [146] and thus it may be necessary to generate mice lacking combinations of TRAFs 2, 3, and 5 to establish the role of these adaptor proteins in osteoclast function.

ROSS AND TEITELBAUM

Absence of TRAF6 results in osteopetrosis. While this observation raises the possibility that TRAF6 links RANK-L to osteoclastogenesis, TRAF6-/- osteopetrotic mice have numerous osteoclasts, which, however, are unable to attach to bone, form a ruffled membrane, or degrade the matrix [151]. These animals bear a striking resemblance to those lacking c-src [59], a fact explained by the constitutive binding of the protooncogene to RANK and its interaction with TRAF6 following receptor occupancy [152]. The subsequent downstream events (Fig. 11) include stimulation of phosphatidylinositol-3 kinase (which by phosphorylating Akt, initiates cytoskeletal reorganization), antiapoptotic signals, and activation of NF-B [153], all key components of osteoclast function. In murine osteoclasts, RANK-derived signals stimulate JNK1 (but not JNK2) via the MEKMEKK axis (Srivastava et al., unpublished data). In short, the link between RANK activation and the AP-1 and NF-B families of transcription factors suggests these signals represent key components of RANK-induced osteoclast formation. However, the presence of numerous, albeit nonfunctional, osteoclasts in TRAF6 null mice indicates that generation of these cells does not require TRAF6-initiated signals. As will become evident, the majority of the osteoclasttargeting cytokines/hormones/steroids exert their influence, wholly or in part, by their ability to increase or decrease M-CSF, OPG, and RANK-L expression. Thus, if M-CSF production is enhanced, the number of precursors, and their ability to differentiate, is augmented. Similarly, a net excess of RANK-L activity (as a result of decreased OPG secretion, increased RANK-L synthesis, or both) stimulates osteoclast recruitment and function. A net deficiency of RANK-L activity (arising from the reciprocal set of circumstances) will arrests osteoclast differentiation and activity. Finally, a number of cytokines, in addition to controlling formation of osteoclasts, through their effects on OPG and RANK-L expression, also target the mature polykaryon, altering its resorptive capacity. Table 1, which summarizes the role of the various cytokines regulating production of M-CSF, OPG, and RANKL, reveals an apparent conundrum, namely that a given osteoclast agonist can increase both OPG and RANK-L. The net effect for any single molecule is explained by assuming greater potency for the induction of either the osteoclaststimulating or inhibiting moiety. It must be recognized that in vivo, the processes of osteoclast formation and subsequent function are controlled in a continuous and rapid manner, reflecting the net effect of the numerous regulators discussed next.

A. Parathyroid Hormone Parathyroid hormone is a major accelerator of bone remodeling. Because the remodeling process is initiated by

85

CHAPTER 3 Osteoclast Biology

TABLE 1 Summary of Factors that Modulate Expression of RANK-L, OPG, and M-CSFa RANK-L

OPG

M-CSF

q

q

q

q

p

Hormones 1,25(OH)2vitamin D3 Estrogen q

p

TNF

q

q

IL-1

q

q

IL-6

q

IL-11

q

PTH Cytokines

q

modulating cells and osteoclasts is necessary [11,82]. The capacity of PTH to increase OPGL and decrease OPG [167] explains the osteoclastogenic properties of the hormone. Intact PTH and parathyroid hormone-related protein (PTHrP) target the same receptor [168] and exert similar biological effects on bone cells [169]. In contrast, the carboxyl-terminal portion of PTHrP may uniquely dampen osteoclastic bone resorption [170]. It should be pointed out that the biological implications of this phenomenon are controversial, as it occurs only in selected systems [171]. Whether carboxyl-terminal PTHrP targets osteoclasts or exerts its inhibitory effects via intermediary cells is not clear [172].

Growth factors TGF-

p

q q

BMP-2 Others Prostaglandin E2

B. Cytokines and Chemokines

q

p

a Adapted, with additions, from Hofbauer et al. [138]. Data on the role of PTH are found in Lee and Lorenzo [167].

osteoclast recruitment, states of PTH excess are typically associated with morphological and biochemical evidence of increased resorptive activity. Importantly, PTH sensitizes bone to 1,25(OH)2D3 [154]. Thus, it is not surprising that 1,25(OH)2D3, in the parathyroprivic state, does not mobilize osteoclasts [155]. Osteoclasts, in vivo or in organ culture, undergo profound changes when exposed to PTH. The population of these cells grows proportional to PTH concentration and the hormone augments the ruffled membrane area and the number of cells exhibiting this resorptive organelle [21,156]. Clear zone and overall cell size are also enhanced [157]. The fact that PTH administration induces c-fos mRNA expression in osteoclasts [158] indicates that the hormone prompts the differentiation of precursors along an osteoclastic pathway [143]. Taken together, these observations suggest that PTH stimulates osteoclast recruitment and activates differentiated polykaryons. While it had been proposed that the hormone targets osteoclasts directly, high-affinity PTH receptors have not been demonstrated in these cells [159,160], nor is there is convincing evidence that the hormone exerts a direct biological effect on isolated osteoclasts [161,162]. However, osteoblasts [158] and marrow stromal cells [163] from which they are derived are sensitive to PTH. The fact that isolated osteoclasts, cocultured with these cells, respond to PTH indicates that osteoblasts and marrow stromal cells serve as intermediaries in PTH-stimulated, osteoclastic bone resorption [164]. PTH-treated osteoblasts or stromal cells secrete resorptive activity [165,166], but contact between

Cytokines regulating osteoclast biology can be divided broadly into three groups: those that facilitate differentiation, survival, and proliferation of precursors; those that promote osteoclast precursor differentiation, and those that alter the resorptive capacity of the mature polykaryon. As discussed later, a number of cytokines exert more than one function. Some act at both proliferative and differentiation stages, whereas others also target mature osteoclasts, modulating bone degradation. Osteoclast formation and activity are controlled by both systemic hormones and locally produced cytokines/chemokines. The latter molecules are themselves regulated in a complex manner in which paracrine and/or autocrine events play a central role. Proteins such as stem cell factor, IL-1, IL-3, IL-6, GMCSF, M-CSF, erythropoietin, epidermal growth factor, and basic fibroblast growth factor are required for the optimal differentiation of stem cells to CFU-GM [85,173], an early osteoclast precursor [41]. As discussed elsewhere in this chapter, GM-CSF and M-CSF play overlapping/redundant roles as regarding the formation of precursors capable of being induced to form bone-resorbing polykarons. Members of the hematopoietic family of cytokines target osteoclast precursors, stimulating both proliferation and differentiation. Other molecules, produced in the bone microenvironment by marrow stromal cells, immature monocytes, and osteoblasts and their precursors, as well as T lymphocytes [88,174,175], are also capable of modulating osteoclastogenesis by acting in a paracrine and/or autocrine manner to regulate the production of M-CSF, RANK-L, and OPG. Whereas most cytokines promote osteoclast formation by stimulating proliferation and/or differentiation of osteoclast precursors, several, including M-CSF, RANK-L, IL-1, IL-6, and TNF, also directly modulate mature osteoclasts. When binding to early osteoclast precursors, M-CSF provides a signal required for survival, proliferation, and maturation [176,177]. In contrast, the same protein

86 increases the motility of mature osteoclasts, with consequent diminution in bone resorption [178,179]. The cytokines IL-6, IL-11, oncostatin M, and leukocyte inhibitory factor share a common signaling pathway. The heterodimeric signaling complex comprises a common subunit, gp130, linked to a second protein, specific for each cytokine [180]. Whereas IL-6 and IL-11 [181,182] are important regulators of osteoclast function, the roles of oncostatin M and leukocyte inhibitory factor are less clear [183,184]. Mice unable to transduce signals from any member of the gp130 family as a result of targeted deletion of the common subunit have normal bones [185]. This result is not surprising given the multiplicity of pathways whereby control of the critical regulator of osteoclast formation and function, namely RANK-L, can be exercised (Fig. 10). Ligation of IL-6 to its receptor, in the mature osteoclast, promotes bone resorption [68], a process accentuated in Paget’s disease [186], in which the cytokine appears to play an important pathogenetic role [187]. The resorptive properties of IL-6 have been documented with osteoclastomaderived, osteoclast-like cells treated with antisense IL-6 oligonucleotides [188]. An antibody that blocks IL-6 function arrests bone loss in vivo [189,190] and blunts oophorectomy-induced osteoclastogenesis [190,191]. Mice in which the IL-6 gene has been removed by targeted deletion do not lose bone following oophorectomy [192]. Finally, IL-6 has been shown to play a role in PTH-induced bone resorption in vivo [193]. Taken together, these results provide strong support for the hypothesis that IL-6 appears to modulate, in part, the effects of estrogen deprivation on bone resorption (see Chapter 41). A molecular explanation for this is provided by the observation that IL-6, targeting to mature osteoclasts, blunts the inhibitory role of high extracellular calcium by a pathway that involves induction, on the surface of the cell, of ADP-ribosyl cyclase [194,195]. In contrast, Kitazawa et al. [189] found decreased numbers of osteoclasts in both oophorectomized and sham-operated mice administered the anti-IL-6 antibody. Furthermore, this antibody, administered in vivo, decreases the histological evidence of bone resorption but not urinary excretion of pyrodinoline cross-links. A naturally occurring inhibitor of IL-1 action, the IL-1 receptor antagonist is produced by monocytes [196] and dampens osteoclast formation and bone resorption in oophorectomized rats and mice [189,197]. Similarly, TNFbinding protein, a soluble form of the TNF receptor that blunts TNF activity, decreases the number of osteoclasts produced by murine cells in vivo and in vitro [189,198]. Furthermore, mice transgenic for the TNF-binding protein, a soluble TNF receptor, are resistant to oophorectomy-induced bone loss, indicating a central role for this cytokine in the most significant form of osteoporosis [199]. Regulation of cytokine expression as relating to osteoclastogenesis is incompletely understood. As regarding

ROSS AND TEITELBAUM

inductive proteins, IL-1 and TNF stimulate the synthesis of M-CSF and IL-11 by marrow stromal cells [200,201]. The same agonists, as well as IL-6, regulate their own production by monocytes via a complex set of paracrine and autocrine signals [202]. The release of IL-6 by stromal cells is stimulated by PTH [203]. As discussed later, estrogen controls the production of several cytokines central to different aspects of osteoclast biology. There are relatively few reports examining the role of chemokines in bone biology. Macrophage inflammatory protein-1, a product of osteoblasts, stimulates motility but decreases the resorptive activity of isolated osteoclasts [204]. Similarly, IL-8, secreted by osteoclasts [205], acts via an autocrine loop to inhibit bone resorption [206]. The latter finding is in conflict with an earlier report that osteoclast recruitment and matrix degradation increase in parallel with IL-8-induced neoangiogenesis [207].

C. Estrogen Estradiol is critical to skeletal preservation in both women [208] and men [209 – 211], as its denial prompts rapid loss, especially of cancellous bone. The estrogendeprived skeleton, particularly as seen shortly after menopause, exhibits histological features of accelerated remodeling, including abundant osteoclasts and resorption bays [3]. Biochemical evidence of enhanced osteoclastic activity in estrogen-deficient patients includes increased hydroxyproline excretion [212]. Interestingly, tamoxifen, which antagonizes estrogen in other organs, is an estrogen agonist in bone, wherein the effects of both steroids are additive [213,214]. Given the identification of molecular targets of estrogenic activity, the means by which the steroid dampens osteoclast function are now more clearly understood. While the role of estrogen is discussed in greater detail in Chapters 10 and 41, it is pertinent that the sex steroid not only modulates M-CSF, but, by controlling monocytic secretion of IL-1 and TNF, the ratio of OPG to RANK-L as well. Likewise, there are reports that the steroid either does [190] or does not [189,215,216] block the secretion of IL-6 from the same cells. There is, however, agreement that estrogen inhibits the production of IL-6 by monocytes [217 – 220]. Additionally, while the sex steroid dampens the secretion of GM-CSF and IL-1 receptor antagonist from cultured human monocytes [220,221], it does not alter the production of IL-11 by stromal cells [181]. Estrogen also diminishes expression of the type I IL-1 receptor on osteoclast and their precursors, thereby limiting the capacity of the cytokine to target these cells [222]. Finally, in addition to modulating the expression of factors that stimulate osteoclast formation and function, estrogen also induces osteoclast apoptosis, a subject discussed in Section VIII.

87

CHAPTER 3 Osteoclast Biology

D. Vitamin D 1,25(OH)2D3, the biologically active form of vitamin D, is a potent inducer of resorption whether administered in vivo [223,224] or added to bone organ culture [22,225]. The molecular basis of this activity is thought to lie largely in the ability of this secosteroid to stimulate, by osteoblastic/stromal cells, a net increase in RANK-L as opposed to OPG. The molecular basis for this result has been defined by the work of Kitazawa and colleagues [226], who demonstrated a functional vitamin D response element in the murine RANK-L promoter. A secondary component of the activity of 1,25(OH)2D3 is that it increases the production of M-CSF, the only cytokine, besides RANK-L, necessary to produce osteoclasts in vitro from both murine and human precursors [88]. A number of structural modifications of the vitamin D molecule have yielded analogues of 1,25(OH)2D3 with distinct biological properties of potential therapeutic importance. For example, 22-oxacalcitriol, while being an effective inducer of cell maturation, is only 1% as effective as 1,25(OH)2D3 in stimulating osteoclastic bone resorption [227]. The molecular basis of this observation is not clear at this time. Vitamin D analogues are discussed in detail in Chapter 9. The fact that vitamin D receptor-ablated mice in whom normal calcium and phosphate levels had been preserved by dietary means exhibited no differences from their wild-type littermates in bone histomorphometry suggests that the major role of the steroid in bone function is to maintain mineral homeostasis [228]. Finally, by promoting expression of type 1 IL-1 and E2 receptors in marrow-derived stromal cells [229,230], the secosteroid primes them to respond to these agonists.

E. Glucocorticoids Gluocorticoid-induced osteoporosis is among the most devastating forms of osteopenia. It predominantly affects trabecular bone [231,232], being manifest most profoundly in the axial skeleton. The subject is discussed extensively in Chapter 44. While glucocorticoids appear to increase osteoclastic activity in vivo, their effect on resorbing cells is complex. Patients chronically treated with glucocorticoids have increased numbers of osteoclasts [233] and resorption bays [234]. There is also biochemical evidence of accelerated skeletal degradation such as enhanced hydroxyproline excretion [235]. Given their osteoclastogenic properties, it is surprising that the resorptive activity and survival of isolated osteoclasts exposed to these agents are attenuated [236]. This conundrum prompted a hypothesis that glucocorticoids, in low dose, permit osteoclast precursor differentiation, whereas higher concentrations blunt resorption [237]. A more likely explanation

of the apparent discrepancy between in vivo and in vitro effects of glucocorticoids on osteoclasts relates to the steroid’s actions on mesenchymal marrow cells. By simultaneously increasing RANK-L, and decreasing OPG, the net effect is an increase in osteoclast number and resorptive capacity [138]. In addition to these direct effects, glucocorticoids suppress intestinal absorption of calcium [238]. This event, in conjunction with augmented renal calcium loss [239], provokes secondary hyperparathyroidism [233,240]. Furthermore, the resorptive impact of such relatively mild hyperparathyroidism is amplified as glucocorticoids sensitize bone cells to PTH [237] and 1,25(OH)2D3 [241], agents known to stimulate osteoblasts to accelerate osteoclast activity. Given these data, it is not surprising that osteoclasts do not proliferate in parathyroidectomized animals receiving glucocorticoids [242]. Finally, while their effects on PTH responsivity are in keeping with stimulated osteoclastic bone degradation, glucocorticoids also inhibit the production of other molecules known to enhance resorption. For example, the steroid suppresses the expression of IL-1 and IL-6 [243] and blunts the synthesis of prostaglandins [244].

F. Thyroid Hormone Thyroid hormone excess, whether endogenous [232,245] or iatrogenic [246-248], prompts bone loss in humans and experimental animals. The skeletal changes of hyperthyroidism are those of accelerated remodeling with an abundance of osteoclasts, osteoblasts, and osteoid [246,247]. In some instances, one encounters peritrabecular marrow fibrosis (osteitis fibrosa) [246,249]. The effects of thyroid on bone are discussed in Chapter 47. Thyroid hormone stimulates bone resorption in vivo [250,251], in organ culture [252,253], and in isolated cell systems consisting of osteoclasts and osteoblasts [254]. The urine of hyperthyroid patients contains increased amounts of the products of bone degradation, including hydroxyproline [250,255] and pyrodinoline cross-links [256,257]. However, the effect of thyroid hormone on osteoclasts is indirect and, like other stimulatory agents, is mediated by osteoblasts [254]. At this time there are no reports as to the impact of thyroid hormone treatment of osteoblastic/stromal cells on the OPG/RANK-L ratio.

G. Prostaglandins Prostaglandins exert dramatic and diverse effects on osteoclast function. While most is known about the resorptive impact of prostaglandins of the E series (PGE), thromboxane [258] and products of the lipooxygenase pathway [259,260] also modulate the osteoclast (see Chapter 13).

88

ROSS AND TEITELBAUM

The predominant effect of a relatively prolonged exposure of the osteoclast to PGE is enhanced resorption. Interestingly, the resorptive effects of intense mechanical force on osteoclast function may also be prostaglandin mediated [261]. While administration of prostaglandins in vivo has not been shown to accelerate resorption, cyclooxygenase inhibitors dampen bone remodeling [262]. Moreover, PGE potently induces osteoclast activity in organ culture [263]. Once again the ability to increase RANK-L and concomitantly decrease OPG explains the resorptive impact of PGE [264,265]. In contrast to the ability of PGE to stimulate osteoclast formation via upregulation of RANK-L on stromal cells, the eicosinoid inhibits the activity of preformed osteoclasts [81,266]. Isolated osteoclasts, when exposed to PGE, cease resorbing bone [266]. The cells transiently contract in a manner similar to that attending calcitonin exposure [267], a finding consistent with the fact that, like calcitonin, prostaglandins express their biological properties through cAMP generation. In fact, many of the resorptive effects of prostaglandins are mimicked by cell-permeant analogues of the nucleotide [81,268]. Interestingly, while prostaglandins and calcitonin both activate adenylate cyclase, the two agents appear to impact the enzyme differently [269].

H. Calcitonin Calcitonin, when administered to vertebrates, dampens bone resorption by directly targeting osteoclasts [161,270]. Thus, the hormone prompts a rapid decline in circulating Ca2 activity [271] and urinary hydroxyproline excretion [272], phenomena indicative of suppressed bone degradation. Furthermore, addition of calcitonin to any in vitro mammalian osteoclast model blunts the capacity of the cell to degrade bone. In actuality, the failure of calcitonin to suppress resorption in a mammalian system raises concern as to whether the model contains bona fide osteoclasts. However, there is no convincing evidence that endogenously secreted calcitonin directly modulates osteoclast activity (see Chapter 73). Reflecting the rapid biological effects of the hormone, a characteristic feature of mammalian osteoclasts is the expression of calcitonin receptors [273]. In fact, calcitoninbinding sites are the only peptide hormone receptors unequivocally resident in these cells. While avian osteoclasts probably lack calcitonin receptors [268], mammalian polykaryons failing to express them are not osteoclasts. However, calcitonin receptors are not unique to the osteoclast, as they also appear in macrophages derived from nonskeletal sources [274], such as lung [275]. This finding is in keeping with the capacity of various macrophage precursors, including those of alveolar origin, to differentiate

in vitro into polykaryons with the complete osteoclast phenotype [11]. When added to organ culture, calcitonin decreases the number of osteoclasts [276] and nuclei per cell [277]. The ruffled membrane disappears, a process reflecting its internalization and vacuolar conversion [123]. Proteins destined for the ruffled membrane, such as the mannose-6-phosphate receptor, are rerouted to intracellular vesicles [123]. TRAP secretion is altered by calcitonin, transiently increasing but ultimately declining [278]. Similarly, osteopontin mRNA expression by osteoclasts is blunted by the hormone [279]. Within seconds of calcitonin exposure, osteoclasts become immobile and contract [51,280,281], a phenomenon associated with changes in the microtubule [282] and microfilament [267] cytoskeleton (Fig. 12). The arrested motility of the cell has been attributed to two distinct events termed quiescence and retraction, each of which may be mediated by distinct signaling pathways [283]. Prolonged exposure to calcitonin results in a declining sensitivity of osteoclasts to the hormone [284,285]. The escape phenomenon, which reduces the therapeutic efficacy of the hormone, has been attributed to induced PTH secretion [286] and anticalcitonin antibody generation [287]. Takahashi and co-workers [288] have shown, however, that calcitonin downregulates its own receptor in osteoclasts. Calcitonin exposure is followed in the osteoclast by an immediate increase in intracellular Ca2 activity and generation of cyclic AMP [289]. The hormone also activates protein kinase C [287]. These accumulated data suggest that calcitonin mediates a number of intraosteoclastic events via distinct signaling pathways. The effects of PTH and calcitonin are discussed further in Chapter 12.

I. Superoxide and Nitric Oxide Oxygen-derived free radicals and nitric oxide (NO) alter osteoclastic bone resorption dramatically [290]. Superoxide is produced by osteoclasts resorbing bone [71,291,292] and therefore may function in an autocrine manner. These anions are located in the ruffled membrane [290] and are induced when osteoclasts contact skeletal matrix [71]. Osteoclasts generate superoxide via NADPH oxidase [292]. The cells also contain superoxide dismutase [70,293], which, when added to osteoclast cultures, decreases resorptive activity [294]. Superoxide anion generation may also play a role in regulated bone resorption. For example, IL-1 and PTH-induced osteoclast activity is blunted by superoxide dismutase [71], and calcitonin diminishes superoxide generation [71,291]. In contrast to superoxide, hitric oxide is antiresorptive [295,296], and the two molecules may act together to regulate osteoclast function [290]. NO targets directly to osteoclasts [295,296] and, like calcitonin, inhibits spreading

89

CHAPTER 3 Osteoclast Biology

FIGURE 12 Osteoclasts exposed to cyclic AMP-inducing agents, including calcitonin and prostaglandin E, undergo rapid contraction and motility arrest. Here, an untreated, isolated rat osteoclast (A) was exposed for 2 h to 108M dibutyryl cyclic AMP (B), inducing immediate contraction and immobility (phase contrast).

[295]. The molecular mechanisms responsible for the osteoclast-inhibiting effects of NO and CT differ [295]. NO is produced by osteoclasts via NO synthase activity [290]. The enzyme in this cell is both constitutively present and regulated [290]. Thus, like superoxide, NO may autoregulate osteoclast function. However, this short-lived reactive radical is also synthesized in abundance by endothelial cells that are in close proximity to, and may govern the activity of, osteoclasts [295]. Perhaps reflecting their known capacity to modulate osteoclasts, osteoblasts also secrete NO in response to inflammatory cytokines [297,298]. Interestingly, osteopontin dampens NO synthesis by renal cells, an event perhaps mediated through the integrin v3 [299]. Whether the bone matrix protein also has an impact on osteoclastic production of NO is unknown.

J. Osteoclast Apoptosis Net bone resorption represents the sum of osteoclast activity and number, with the latter parameter itself dependent on the rates of cell formation and death. Treatment of mature osteoclasts with estrogen results in the rapid initiation of apoptosis, an event mediated by transforming growth factor  [300]. Nitrogen-containing bisphosphonates induce osteoclast death, an event resulting from the blunted synthesis

of farnesyl and geranylgeranyl pyrophosphate, two moieties central to the acylation, and thus activation, of members of the Ras superfamily [301 – 303]. Signaling through Ras proteins modulates cell survival. Separately, bisphosphonates lacking the nitrogen atom stimulate osteoclast death by an undetermined mechanism, but one that does not involve small GTPases [303]. Caspases, a family of cysteine proteases [304], are involved in both the initiation and the execution phases of apoptosis, and these molecules modulate the life span of the osteoclast. Thus, treatment of osteoclast cultures with a pan inhibitor of caspase action, Z-VAD-Fmk, decreases the number of apoptotic cells [305]. Preliminary data indicate the likely involvement of caspases 3 and/or 7, both executor, or downstream caspases [306]. The proximal signals leading to the activation of these distally acting proteases, in osteoclasts, are unknown.

VIII. DISEASES OF THE OSTEOCLAST A. Osteopetrosis The osteopetroses are a group of congenital diseases characterized by increased skeletal mass due to dysfunctional osteoclasts. Despite pathogenetic heterogeneity, all

90

FIGURE 13

Molecular mechanisms of osteopetrosis. See text for details. (See also color plate.)

forms of osteopetrosis are distinguished by radiopaque bones, with loss of differentiation between the cortex and the marrow. The molecular defect of a number of osteopetrotic mutants, particularly of murine origin, are now defined and serve as a rich source of information regarding the origin and the function of osteoclasts (Fig. 13 see also color plate). 1. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLAST PROGENITORS Initial evidence regarding the origin of osteoclasts was provided by Walker [307,308] wherein the circulation of a normal mouse was joined to that of an osteopetrotic, microphthalmic (mi/mi), or gray-lethal (gl/gl) littermate. By 6 weeks of age all of the excess mutant skeleton was resorbed. Walker concluded that mature osteoclasts, or their progenitors, were introduced from the normal mouse into the mutant’s marrow and subsequent experiments support this hypothesis. For example, the skeletal lesion of irradiated gl/gl or mi/mi mice is cured by infused, normal marrow or spleen cells. Conversely, mi/mi or gl/gl spleen cells induce osteopetrosis in irradiated normal mice [309]. Together, these experiments provide direct evidence for the hematopoietic origin of osteoclasts. These studies also establish that ultimate proof of the mutated gene of interest is, in fact, essential for osteoclast differentiation and requires rescue of osteopetrotic mice with normal osteoclast precursors, or administration of the gene product. The technique of homologous recombination has yielded a plethora of osteopetrotic knockout mice and major insights into the genes regulating osteoclastogenesis. For example, deletion of the macrophage and B-cell-specific PU.1 gene generates mice completely devoid of osteoclasts and their precursors [310]. These animals represent the earliest known developmental defect in osteoclastogenesis. c-fos, deletion of which prompts osteopetrosis [143], is a widely expressed nuclear protooncogene belonging to a multigene family including fosB, fra-1, and fra-2 [311]. While the fos mutant lacks osteoclasts, marrow macrophages are actually increased relative to wild type. Thus, c-fos may exert its osteoclast differentiating effect distal to PU.1. Similarly, deletion of the p50 and p52

ROSS AND TEITELBAUM

subunits of NF-B prompts osteoclast-deficient osteopetrosis in which macrophage differentiation is extant [312]. NF-B, therfore, impacts osteoclastogenesis in a manner reminiscent of c-fos. The osteopetrotic phenotype of the NF-B knockout indicates cell surface receptors activating this transcription complex may also participate in the osteoclastogenic process. Thus, deletion of RANK-L [135] or its functional blockade by the overexpression of osteoprotegerin [91] each results in osteopetrotic mice devoid of osteoclasts. 2. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLAST FUNCTION Among the most interesting of the osteopetrotic mutants is the c-src -/- mouse [59]. This animal, produced by homologous recombination, exhibits no detectable abnormalities in brain or platelets where pp60c-src is normally abundant, but is affected by a form of osteopetrosis in which osteoclasts are numerous. This latter finding suggests that the osteoclast lineage is intact but that the cell is functionally impaired [109]. In fact, electron microscopic examination reveals that src -/- osteoclasts contain no ruffled membranes (Fig. 14). Moreover, osteoclasts generated in vitro from src mutant marrow cells fail to pit bone. These observations signify that lack of bone resorption by src -/- osteoclasts is intrinsic to this cell and not reflective of the bone microenvironment [313]. Similar results were obtained with deletion of TRAF6 [151], a protein associating with the intracellular domain of RANK and which signals following association with c-src [152]. The absence of ruffling in src mutant osteoclasts suggests that pp60c-src is fundamental to osteoclast polarization and may participate in transporting vacuolar H -ATPase from internal pools of acidic vesicles to the plasma membrane. Although this hypothesis lacks direct proof, it is supported by several lines of circumstantial evidence. For example, pp60c-src is a specific marker of macrophages committed to the osteoclast as compared to the host defense phenotype [145]. Both pp60c-src and the vacuolar H -ATPase associate with microtubules in the putative pathway toward the formation of the ruffled membrane [20]. The osteoclast src protein is increased and decreased, respectively, in response to parathyroid hormone and calcitonin [314]. Moreover, bovine adrenal gland chromaffin granules are highly enriched in pp60c-src [315 – 317]. Because these structures are secretory organelles, they provide indirect evidence that the tyrosine kinase has a role in vesicle transport and/or secretion, events akin to ruffled membrane formation. Interestingly, the role of pp60c-src in osteoclast function appears to involve its capacity both as a tyrosine kinase and docking protein, as a kinase-inactive construct substantially, but not completely, rescues the c-src-/- osteoclast [318]. With the realization that the v3 integrin and cathepsin K may be critical to optimal osteoclast function, mice

CHAPTER 3 Osteoclast Biology

FIGURE 14 Osteoclasts from normal and c-src-/- mice. Wild-type cells (A) contain a well-developed ruffled membrane (r), which is lacking in the osteopetrotic mutant (B). Reproduced from J. Clin. Invest. 90, 1622 – 1627 (1992). Copyright permission of The Society for Clinical Investigation.

91

92 devoid of one or the other molecule have been generated [101,125]. While each lack the characteristic pathological features of osteopetrosis, they both contain dysfunctional osteoclasts eventuating in failure to normally resorb bone and, hence, enhanced bone mass. Two general phenotypes of osteopetrosis have been described in humans. One is relatively asymptomatic, whereas the other, usually present at birth, is fatal in infancy or early childhood. While the molecular mechanisms underlying the malignant forms of osteopetrosis are not yet resolved, there is information regarding the “benign’’ phenotype. One such example is characterized by renal tubular acidosis. Consistent with the capacity of the isoform’s inhibitors to also block parathyroid hormone-induced bone resorption, mutation of the carbonic anhydrase II gene is responsible for this form of benign osteopetrosis [319]. Carbonic anhydrase II, in normal osteoclasts, catalyzes the hydration of CO2 to carbonic acid. Dissociation of carbonic acid into bicarbonate ions and protons permits formation of the resorptive microenvironment, which is probably defective in carbonic anhydrase II-deficient patients. Considering the importance of proton secretion in bone degradation, one might expect inactivating mutations of the osteoclast H -ATPase to impair the resorptive process. Indeed, osteoclasts generated in vitro from marrow cells of a patient with craniometaphyseal dysplasia, a rare genetic sclerosing bone disease, fail to resorb bone [320]. These osteoclasts do not express the H -ATPase as determined by immunohistochemistry. 3. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLASTOGENIC MICROENVIRONMENT The murine recessive op mutation prompts an osteopetrotic phenotype characterized by a failure to generate monocytes, macrophages, and osteoclasts [321,322]. These animals are not cured by marrow transplantation. Alternatively, op/op hematopoietic progenitors, administered to wild-type mice, differentiate into normal osteoclasts. These experiments prove that the op mutation does not target macrophage and osteoclast progenitors but rather that the microenvironment in which these cells develop is defective [323,324]. In fact, op/op bone cells and fibroblasts fail to produce competent M-CSF [84,325]. Furthermore, the op mutation maps to a single base pair insertion in the coding region of the M-CSF gene, resulting in a translation frame shift, and insertion of a stop codon 21 bp downstream, producing a truncated, nonfunctional M-CSF protein [326]. The observation that the op/op mouse is rescued by subcutaneous injections of recombinant M-CSF provides additional evidence that lack of this protein is the cause of its osteopetrosis [83,327,328]. The op/op mouse underscores the importance of the hematopoietic microenvironment in osteoclast development and points to the fact that osteoclasts and macrophages share a common lineage.

ROSS AND TEITELBAUM

Surprisingly, op/op mice progressively recover from lack of both macrophages and osteoclasts with age and, by 22 weeks, the marrow cavity size and cellularity appear unremarkable [329]. The numbers of mononuclear phagocytes, macrophages, and osteoclast precursors progressively increase to normal. Thus, the animal is able to compensate for a lack of M-CSF, presumably by other cytokines with whose function it overlaps. The spontaneous rescue of the op/op mouse has been shown to reflect expression of GM-CSF, a protein with many activities shared, in vitro, with M-CSF [330]. Supporting this contention, GM-CSF administered to op/op mice during their osteopetrotic phase cures the bone phenotype. This observation, taken with the normal skeletons of GM-CSF knockout mice [331], indicates that either M-CSF or GM-CSF, in the absence of the other, will promote osteoclastogenesis. 4. CURE OF HUMAN OSTEOPETROSIS Based on the animal models described earlier, a female patient with autosomal-recessive osteopetrosis was cured after receiving her brother’s marrow. In this instance, the donor origin of osteoclasts was established by following the Y chromosome [332]. The pretransplantation abundance of osteoclasts, albeit dysfunctional, established the patient’s capacity to provide an osteoclastogenic environment for normal marrow. Given an appropriate donor, marrow transplantation is the treatment of choice for malignant osteopetrosis.

B. Paget’s Disease Paget’s disease of bone is a paradigm of remodeling gone awry. As such, the initiator of bone remodeling, namely the osteoclast, is the pathogenetic cell. In fact, despite the array of abnormal histological features attending Paget’s disease, it is the appearance of the osteoclast that is pathognomonic of this condition [333]. The number of osteoclasts, and their nuclei, are increased greatly in Paget’s disease and the cells are enormous (Fig. 15). In contrast to normal osteoclasts in which nuclei polarize toward the nonresorptive surface, those in pagetic cells are distributed diffusely throughout the cytoplasm. Reflecting osteoclast size and number, bone resorption is accelerated greatly. Resorption bay volume and prevalence are increased manyfold [333,334]. Indeed, the initial cellular event in the pagetic skeleton is rapid resorption, often resulting in large, radiographically evident, lytic lesions [335]. Interestingly, one finds histological evidence of enhanced resorption in radiographically uninvolved bone. It is unresolved whether this phenomenon represents subclinical Paget’s disease or, as has been suggested [333,336], mild

93

CHAPTER 3 Osteoclast Biology

FIGURE 15

Osteoclasts (arrows) in bone biopsy taken from a patient with Paget’s disease. The cells are enormous and contain many nuclei, which, in contrast to normal osteoclasts (see Fig. 1), are nonpolarized (nondecalcified, toluidine blue stain).

hyperparathyroidism. In any event, given the central role that the osteoclast plays in Paget’s disease, it is not surprising that successful treatment entails arresting bone resorption [337,338]. While the precise etiology of Paget’s disease is not yet understood, the measles virus appears to be the most likely candidate. For example, osteoclasts in pagetic bone contain filamentous nuclear [339 – 341] and cytoplasmic inclusions [342] typical of the paramyxovirus family. Most importantly, experiments performed in Roodman’s laboratory demonstrate that targeting this virus to osteoclast precursors in vivo yields a mouse with osteoclasts mimicking those of pagetic patients [343]. This laboratory also found that similar to the in vivo situation, bone marrow derived from patients with Paget’s disease gives rise to numerous, large, hypernucleated osteoclast-like cells [344]. In contrast to polykaryons generated from normal marrow, these multinucleated cells exhibit increased sensitivity to 1,25-dihydroxyvitamin D and produce the osteoclastogenic cytokine IL-6, suggesting that an autocrine event mediates osteoclast proliferation in the disease [345]. Supporting this hypothesis is the fact that IL-6 circulates in increased amounts in pagetic patients [186]. The abundance of osteoclasts in Paget’s disease reflects the proliferative capacity of precursor cells. The number of CFU-GM colonies formed from pagetic marrow is increased, as is replication of CD34 cells, the earliest osteoclast progenitor yet identified in human [41]. It is of interest that pagetic marrow cells not in the osteoclast lineage (i.e., CD34 cells) may serve an accessory function in the

generation of resorptive polykaryons [41]. In this regard, pagetic stromal cells overexpress the RANK ligand, and marrow isolated from affected patients generates increased numbers of osteoclasts in response to the cytokine [346].

C. Cancer Osteolysis is the product of many malignant neoplasms either resident in or distant from bone. In most instances, it appears that tumor-induced osteolysis reflects the recruitment of osteoclasts, by-products of the neoplasm or bone matrix per se to the site of potential bone destruction. Such recruitment may be (i) the result of intimate interactions of a localized solid tumor, in bone, with osteoclast progenitors, (ii) humoral factors secreted by a tumor absent of skeletal metastases, or (iii) bidirectional paracrine stimulation of the tumor and bone marrow cells by cytokines. 1. SOLID TUMORS RESIDENT IN BONE Tumors commonly metastatic to the skeleton include those primary in lung, prostate, thyroid, and kidney [347]. Breast cancer is, however, by far the predominant source of osteolytic lesions. Reflecting excessive bone resorption, carcinoma of the breast is frequently complicated by hypercalcemia and fracture. The evidence that osteoclasts are pivotal to development of metastatic tumor-induced osteolysis includes: (i) in contrast to their ability to degrade soft tissues, cancer cells appear to have a limited, if any, capacity to resorb bone [334];

94

ROSS AND TEITELBAUM

FIGURE 16 Osteoclastic bone resorption in skeletal metastasis. Osteoclasts recruited by tumor resorb bone. Breast cancer cells (arrows) are in close proximity to osteoclasts (arrowheads) and follow the polykaryons into resorption bays (hematoxylin and eosin). (ii) breast cancer-induced osteolysis is blunted by osteoclast-inhibiting bisphosphonates [348]; (iii) osteoclasts are abundant in foci of metastatic breast cancer (Fig. 16); and (iv) breast cancer cells secrete potential osteoclastogenic agents [349 – 351]. In fact, evidence shows that parathyroid hormone-related protein is pivotal to the osteolytic properties of metastatic breast cancer. This protein, particularly produced by those breast cancers with a predisposition for skeletal metastasis, prompts marrow stromal cells, or osteoblasts, to synthesize osteoclast-recruiting cytokines [352]. The tumor-recruited osteoclasts resorb bone, releasing stored growth factors, notably transforming growth factor , which induces the neoplasm to proliferate and synthesize additional PTHrP [353]. This being the case, one would expect inhibition of tumor-induced osteoclastic bone resorption to decrease the tumor burden; in fact, bisphosphonates have such an effect [354]. 2. HUMORAL HYPERCALCEMIA OF MALIGNANCY PTHrP is also the active agent secreted by most tumors inducing humoral hypercalcemia of malignancy [355,356]. The amino terminus of the molecule is recognized by the PTH receptor. As this receptor has not been shown

convincingly to function in osteoclasts, the osteoclastogenic, and consequent, hypercalcemic effects of PTHrP are probably mediated by osteoblasts or marrow stromal cells in which the PTH/PTHrP receptor is abundant. 3. MULTIPLE MYELOMA In contrast to other B-cell malignancies, lytic bone lesions are common in patients with multiple myeloma. In fact, myeloma represents one of few malignancies almost always associated with osteolysis without an apparent osteoblastic component [357]. The central role osteoclasts play in the generation of myeloma bone lesions is underscored by the dramatic impact of bisphosphonates [358]. Bone biopsies from myeloma patients contain abundant osteoclasts in close proximity to the tumor [359,360]. Thus, myeloma cells alone, or in combination with surrounding stromal cells, likely produce factors capable of activating or recruiting osteoclasts. Medium conditioned by extirpated myeloma contains candidate molecules, including IL-6, IL-1, TNF, and macrophage inflammatory protein [361,362]. IL-6, which impacts osteoclasts profoundly, is also critical to myeloma proliferation [363,364]. Interestingly, and in keeping with the role of the

95

CHAPTER 3 Osteoclast Biology

cytokine in osteoclastogenesis attending postmenopausal osteoporosis, IL-6 mRNA is expressed by stromal and not myeloma cells. These observations raise the issue as to how myeloma cells induce cytokine production by accessory cells. Evidence indicates that the osteoclastogenic properties of myeloma cells depend on their contact, via the 41 integrin, with VCAM-1 on stromal cells [365]. 4. INFLAMMATORY OSTEOLYSIS Inflammatory osteolysis is a common form of clinically significant bone loss. This family of conditions includes alveolar bone loss attending periodontal disease, periprosthetic osteolysis, which frequently follows orthopedic implant surgery, and osteopenia accompanying rheumatoid arthritis. Each of these conditions is characterized by abundant osteoclast proliferation, eventuating in dramatic, localized bone destruction. In the case of rheumatoid arthritis, the osteopenia may also be systemic. TNF, interacting with its p55 (type 1) receptor, appears to be central to the osteoclastogenesis of inflammatory osteolysis [145,366] and, hence, newly available drugs inhibiting this cytokine hold therapeutic promise in these disorders. 5. POSTMENOPAUSAL OSTEOPOROSIS Postmenopausal (type 1) osteoporosis is due to an absolute increase bone resorption relative to the premenopausal state [190]. This enhancement of resorptive activity reflects the absence of estrogen’s antiosteoclastogenic effects. Thus, bone biopsies taken following estrogen withdrawal are rich in osteoclasts [190]. The means by which estrogen suppresses osteoclastogenesis was discussed earlier and in Chapter 41.

References 1. H. C. Blair, A. J. Kahn, E. C. Crouch, J. J. Jeffrey, and S. L. Teitelbaum, Isolated osteoclasts resorb the organic and inorganic components of bone. J. Cell Biol. 102, 1164 – 1172 (1986). 2. J. I. Alvarez, S. L. Teitelbaum, H. C. Blair, E. M. Greenfield, N. A. Athanasou, and F. P. Ross, Generation of avian cells resembling osteoclasts from mononuclear phagocytes. Endocrinology 128, 2324 – 2335 (1991). 3. H. M. Frost, Tetracycline-based histological analysis of bone remodeling. Calcif. Tissue Int. 3, 211 – 237 (1969). 4. T. J. Chambers and K. Fuller, Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic attack. J. Cell Sci. 76, 155 – 165 (1985). 5. M. D. Fallon, H. A. Schwamm, An update on the pathogenesis, pathophysiology, and treatment of osteitis deformans. Path. Annu. 24, 115 – 159 (1989). 6. A. Teti, G. Volleth, A. Carano, and A. Zambonin-Zallone, The effects of parathyroid hormone on 1,25-dihydroxyvitamin D3 on monocyteosteoclast fusion. Calcif. Tissue Int. 42, 302 – 308 (1988). 7. J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J. Masarachia, G. Wesolowski, R. G. Russell, G. A. Rodan, and A. A. Reszka, Alendronate mechanism of action:

8. 9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc. Natl. Acad. Sci. USA 96, 133 – 138 (1999). T. Efstratiadis and D. W. Moss, Tartrate-resistant acid phosphatase in human alveolar macrophages. Enzyme 34, 140 – 143 (1985). J. Schindelmeiser, D. Munstermann, and H. Witzel, Histochemical investigations on the localization of the purple acid phosphatase in the bovine spleen. Histochemistry 87, 13 – 19 (1987). H. M. Massey and A. M. Flanagan, Human osteoclasts derive from CD14-positive monocytes. Br. J. Haematol. 106, 167 – 170 (1999). N. Udagawa, N. Takahashi, T. Akatsu, H. Tanaka, T. Sasaki, T. Nishihara, and T. Suda, Origin of osteoclasts: Mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc. Natl. Acad. Sci. USA 87, 7260 – 7264 (1990). A. McInnes and D. M. Rennick, Interleukin 4 induces cultured monocytes/macrophages to form giant multinucleated cells. J. Exp. Med. 167, 598 – 611 (1988). M. E. Holtrop, Light and electronmicroscopic structure of osteoclasts. In “Bone” (B. K. Hall, ed.), pp. 1 – 29. CRC Press, Boca Raton, FL, 1991. M. D. Fallon, S. L. Teitelbaum, and A. J. Kahn, Multinucleation enhances macrophage-mediated bone resorption. Lab. Invest. 49, 159 – 164 (1983). Y. Abu-Amer, S. L. Teitelbaum, P. Schlesinger, and F. P. Ross, Expression and regulation of RAB3 proteins in osteoclasts and their precursors. J. Bone Miner. Res. 14, 1855 – 1860 (1999). D. M. Kallio, P. R. Garant, and C. Minkin, Evidence of coated membranes in the ruffled border of osteoclasts. J. Ultrastruct. Res. 37, 169 – 177 (1971). G. Andersson, B. Ek-Rylander, and C. Minkin, Acid phosphatases. In “Biology and Physiology of the Osteoclast” (B. R. Rifkin and C. V. Gay, eds.), pp. 55 – 80. CRC Press, Boca Raton, FL, 1992. B. Ek-Rylander, M. Flores, M. Wendel, D. Heinegard, and G. Andersson, Dephosphorylation of osteopontin and bone sialoprotein by osteoclastic tartrate-resistant acid phosphatase: Modulation of osteoclast adhesion in vitro. J. Biol. Chem. 269, 14853 – 14856 (1994). O. Fukushima, P. J. Bekker, and C. V. Gay, Characterization of the functional stages of osteoclasts by enzyme histochemistry and electron microscopy. Anat. Rec. 231, 298 – 315 (1991). Y. Abu-Amer, F. P. Ross, P. Schlesinger, M. M. Tondravi, and S. L. Teitelbaum, Substrate recognition by osteoclast precursors induces s-crc/microtubule association. J. Cell Biol. 137, 247 – 258 (1997). M. E. Holtrop and L. G. Raisz, Comparison of the effects of 1,25dihydroxycholecalciferol, prostaglandin E2, and osteoclast-activating factor with parathyroid hormone on the ultrastructure of osteoclasts in cultured long bones of fetal rats. Calcif. Tissue Int. 29, 201 (1979). M. E. Holtrop, K. A. Cox, M. B. Clark, M. F. Holick, and C. S. Anast, 1,25-Dihydroxycholecalciferol stimulates osteoclasts in rat bones in the absence of parathyroid hormone. Endocrinology 108, 2293 – 2301 (1981). A. Zambonin-Zallone, A. Teti, A. Carano, and P. C. Marchisio, The distribution of podosomes in osteoclasts cultured on bone laminae: Effect of retinol. J. Bone Miner. Res. 3, 517 – 523 (1988). A. Zambonin-Zallone, A. Teti, M. Grano, A. Rubinacci, M. Abbadini, M. Gaboli, and P. C. Marchisio, Immunocytochemical distribution of extracellular matrix receptors in human osteoclasts: A 3 integrin is colocalized with vinculin and talin in the podosomes of osteoclastoma giant cells. Exp. Cell Res. 182, 645 – 652 (1989). D. Zhang, N. Udagawa, I. Nakamura, H. Murakami, S. Saito, K. Yamasaki, Y. Shibasaki, N. Morii, S. Narumiya, N. Takahashi, and T. Suda, The small GTP-binding protein, rho p21, is involved in bone resorption by regulating cytokskeletal organization in osteoclasts. J. Cell Sci. 108, 2285 – 2292 (1995).

96 26. A. Miyauchi, K. A. Hruska, E. M. Greenfield, R. Barattolo, S. Colucci, A. Zambonin-Zallone, S. L. Teitelbaum, and A. Teti, Osteoclast cytosolic calcium, regulated by voltage operated calcium channels and extracellular calcium, controls podosome assembly and bone resorption. J. Cell Biol. 111, 2543 – 2552 (1990). 27. H. Rasmussen and P. Bordier, The cellular basis of metabolic bone disease. N. Engl. J. Med. 289, 25 – 32 (1973). 28. G. Gothlin and J. L. E. Ericsson, The osteoclast: Review of ultrastructure, origin, and structure-function relationship. Clin. Orthop. Rel. Res. 120, 201 – 231 (1976). 29. A. J. Kahn and D. J. Simmons, Investigation of cell lineage in bone using a chimaera of chick and quail embryonic tissue. Nature 258, 325 – 327 (1975). 30. A. J. Kahn, C. C. Stewart, and S. L. Teitelbaum, Contact-mediated bone resorption by human monocytes in vitro. Science 199, 988 – 990 (1978). 31. S. L. Teitelbaum, C. C. Stewart, and A. J. Kahn, Rodent peritoneal macrophages as bone resorbing cells. Calcif. Tissue Int. 27, 255 – 261 (1979). 32. A. J. Kahn, J. D. Malone, and S. L. Teitelbaum, Osteoclast precursors, mononuclear phagocytes, and bone resorption. Trans. Assoc. Am. Phys. 94, 267 – 278 (1981). 33. M. Nagai and N. Sato, Reciprocal gene expression of osteoclastogenesis inhibitory factor and osteoclast differentiation factor regulates osteoclast formation. Biochem. Biophys. Res. Commun. 257, 719 – 723 (1999). 34. E. H. Burger, J. W. M. Van Der Meer, J. K. Van De Gevel, J. C. Gribnau, C. W. Thesingh, and R. Van Furth, In vitro formation of osteoclasts from long-term cultures of bone marrow mononuclear phagocytes. J. Exp. Med. 156, 1604 – 1614 (1982). 35. B. A. A. Scheven, J. W. M. Visser, and P. J. Nijweide, In vitro osteoclast generation from different bone marrow fractions, including a highly enriched haematopoietic stem cell population. Nature 321, 79 – 81 (1986). 36. N. Kurihara, T. Suda, Y. Miura, H. Nakauchi, H. Kodama, K. Hiura, Y. Hakeda, and M. Kumegawa, Generation of osteoclasts from isolated hematopoietic progenitor cells. Blood 74, 1295 – 1302 (1989). 37. N. Takahashi, H. Yamana, S. Yoshiki, G. D. Roodman, G. R. Mundy, S. J. Jones, A. Boyde, and T. Suda, Osteoclast-like cell formation and its regulation by osteotropic hormones in mouse bone marrow cultures. Endocrinology 122, 1373 – 1382 (1988). 38. N. Takahashi, T. Akatsu, T. Sasaki, G. C. Nicholson, J. M. Moseley, T. J. Martin, and T. Suda, Induction of calcitonin receptors by 1 alpha, 25-dihydroxyvitamin D3 in osteoclast-like multinucleated cells formed from mouse bone marrow cells. Endocrinology 123, 1504 – 1510 (1988). 39. N. Takahashi, T. Akatsu, N. Udagawa, T. Sasaki, A. Yamaguchi, J. M. Moseley, T. J. Martin, and T. Suda, Osteoblastic cells are involved in osteoclast formation. Endocrinology 123, 2600 – 2602 (1988). 40. N. Udagawa, N. Takahashi, T. Akatsu, T. Sasaki, A. Yamaguchi, H. Kodama, T. J. Martin, and T. Suda, The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with mouse spleen cells. Endocrinology 125, 1805 – 1813 (1989). 41. N. Kurihara, C. Chenu, M. Miller, C. Civin, and G. D. Roodman, Identification of committed mononuclear precursors for osteoclastlike cells formed in long term human marrow cultures. Endocrinology 126, 2733 – 2741 (1991). 42. T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endocr. Rev. 13, 66 – 80 (1992). 43. R. K. Globus, D. D. Bikle, B. Halloran, and E. Morey-Holton, Skeletal response to dietary calcium in a rat model simulating weightlessness. J. Bone Miner. Res. 1, 191 – 197 (1986).

ROSS AND TEITELBAUM 44. F. Tilton, J. Degioanni, and V. Schneider, Long-term follow-up of skylab bone demineralization. Aviat. Space Environ. Med. 51, 1209 – 1216 (1980). 45. M. G. Griffin, R. Kimble, W. Hopfer, and R. Pacifici, Dual-energy X-ray absorptiometry of the rat: Accuracy, precision, and measurement of bone loss. J. Bone Miner. Res. 8, 795 – 800 (1993). 46. S. P. Robins, H. Woitge, R. Hesley, J. Ju, S. Seyedin, and M. J. Seibel, Direct, enzyme-linked immunoassay for urinary deoxypyridinoline as a specific marker for measuring bone resorption. J. Bone Miner. Res. 9, 1643 – 1649 (1994). 47. L. G. Raisz, Stimulation of bone resorption by parathyroid hormone in tissue culture. Nature 197, 1015 – 1017 (1963). 48. S. L. Teitelbaum, J. E. Russell, J. M. Bone, J. J. Gilden, and L. V. Avioli, The relationship of biochemical and histometric determinants of uremic bone. Arch. Pathol. Lab. Med. 103, 228 – 230 (1979). 49. P. D. Delmas, Biochemical markers of bone turnover in osteoporosis. In “Osteoporosis” (B. L. Riggs and L. J. Melton, eds.), pp. 297 – 316. Raven Press, New York, 1988. 50. M. J. Seibel, S. P. Robins, and J. P. Bilezikian, Urinary pyridinium crosslinks of collagen: Specific markers of bone resorption in metabolic bone disease. Trends Endocrinol. Metab. 3, 263 – 270 (1992). 51. T. J. Chambers and C. J. Magnus, Calcitonin alters behavior of isolated osteoclasts. J. Pathol. 136, 27 – 39 (1982). 52. R. Nelson and G. Bauer, Isolation of osteoclasts by velocity sedimentation at unit gravity. Calcif. Tissue Res. 22, 303 – 309 (1977). 53. P. Osdoby, M. Martini, and A. Caplan, Isolated osteoclasts and their presumed progenitor cells, the monocyte, in culture. J. Exp. Zool. 244, 331 – 339 (1982). 54. T. Hefley and P. Stern, Isolation of osteoclasts from fetal rat long bones. Calcif. Tissue Int. 34, 480 – 487 (1982). 55. A. Zambonin-Zallone, A. Teti, and M. Primavera, Isolated osteoclasts in primary culture: First observation on structure and survival in culture media. Anat. Embryol. 165, 405 – 413 (1982). 56. H. C. Blair, S. L. Teitelbaum, L. E. Grosso, D. L. Lacey, H.-L. Tan, D. W. McCort and J. J. Jeffrey, Extracellular matrix degradation at acid pH: Avian osteoclast acid collagenase isolation and characterization. Biochem. J. 290, 873 – 884 (1993). 57. J. P. Mattsson, P. H. Schlesinger, D. J. Keeling, S. L. Teitelbaum, D. K. Stone, and X.-S. Xie, Isolation and reconstitution of a vacuolar-type proton pump of osteoclast membranes. J. Biol. Chem. 269, 24979 – 24982 (1994). 58. W. C. Horne, L. Neff, D. Chatterjee, A. Lomri, J. B. Levy, and R. Baron, Osteoclasts express high levels of pp60c-src in association with intracellular membranes. J. Cell Biol. 119, 1003 – 1013 (1992). 59. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 60. K. Tezuka, T. Sato, H. Kamioka, P. J. Nijweide, K. Tanaka, T. Matsuo, M. Ohta, N. Kurihara, Y. Hakeda, and M. Kumegawa, Identification of osteopontin in isolated rabbit osteoclasts. Biochem. Biophys. Res. Commun. 186, 911 – 917 (1992). 61. K. Tezuka, Y. Tezuka, A. Maejima, T. Sato, K. Nemoto, H. Kamioka, Y. Hakeda and M. Kumegawa, Molecular cloning of a possible cysteine proteinase predominantly expressed in osteoclasts. J. Biol. Chem. 269, 1106 – 1109 (1994). 62. Z. Zimolo, G. Wesolowski, H. Tanaka, J. L. Hyman, J. R. Hoyer, and G. A. Rodan, Soluble v3-integrin ligands raise [Ca2]i in rat osteoclasts and mouse-derived osteoclast-like cells. Am. J. Physiol. 266, C376 – C381 (1994). 63. G. Shankar, I. Davison, M. H. Helfrich, W. T. Mason, and M. A. Horton, Integrin receptor-mediated mobilisation of intranuclear calcium in rat osteoclasts. J. Cell Sci. 105, 61 – 68 (1993). 64. M. J. Oursler, L. V. Bell, B. Clevinger, and P. Osdoby, Identification of osteoclast-specific monoclonal antibodies. J. Cell Biol. 100, 1592 – 1600 (1985).

97

CHAPTER 3 Osteoclast Biology 65. T. J. Chambers, K. Fuller, J. A. Darby, J. A. S. Pringle, and M. A. Horton, Monoclonal antibodies against osteoclasts inhibit bone resorption in vitro. Bone Miner. 1, 127 – 135 (1986). 66. P. Collin-Osdoby, M. J. Oursler, D. Webber, and P. Osdoby, Osteoclast-specific monoclonal antibodies coupled to magnetic beads provide a rapid and efficient method of purifying avian osteoclasts. J. Bone Miner. Res. 6, 1353 – 1365 (1991). 67. J. Davies, J. Warwick, N. Totty, R. Philp, M. Helfrich, and M. Horton, The osteoclast functional antigen, implicated in the regulation of bone resorption, is biochemically related to the vitronectin receptor. J. Cell Biol. 109, 1817 – 1826 (1989). 68. Y. Ohsaki, S. Takahashi, T. Scarcez, A. Demulder, T. Nishihara, R. Williams, and G. D. Roodman, Evidence for an autocrine/paracrine role for interleukin-6 in bone resorption by giant cells from giant cell tumors of bone. Endocrinology 131, 2229 – 2234 (1992). 69. S. Takahashi, S. V. Reddy, J. M. Chirgwin, R. Devlin, C. Haipek, J. Anderson, and G. D. Roodman, Cloning and identification of annexin II as an autocrine/paracrine factor that increases osteoclast formation and bone resorption. J. Biol. Chem. 269, 28696 – 28701 (1994). 70. M. J. Oursler, P. Collin-Osdoby, L. Li, E. Schmitt, and P. Osdoby, Evidence for an immunological and functional relationship between superoxide dismutase and a high molecular weight osteoclast plasma membrane glycoprotein. J. Cell. Biochem. 46, 331 – 344 (1991). 71. I. R. Garrett, B. F. Boyce, R. O. Oreffo, L. Bonewald, J. Poser, and G. R. Mundy, Oxygen-derived free radicals stimulate osteoclastic bone resorption in rodent bone in vitro and in vivo. J. Clin. Invest. 85, 632 – 639 (1990). 72. F. P. Ross, J. I. Alvarez, J. Chappel, D. Sander, W. T. Butler, M. C. Farach-Carson, K. A. Mintz, P. G. Robey, S. L. Teitelbaum, and D. A. Cheresh, Interactions between the bone matrix proteins osteopontin and bone sialoprotein and the osetoclast integrin v3 potentiate bone resorption. J. Biol. Chem. 268, 9901 – 9907 (1993). 73. M. Chiba, S. L. Teitelbaum, and F. P. Ross, Treatment of avian osteoclast precursors with retinoic acid alters the rate of synthesis of 3 but not v mRNA and increases association of the heterodimer, by stimulating the synthesis of 3 protein. J. Bone Miner. Res. 8, S144 (1993). 74. H. Mimura, X. Cao, F. P. Ross, M. Chiba, and S. L. Teitelbaum, 1,25(OH)2D3 vitamin D3 transcriptionally activates the 3-integrin subunit gene in avian osteoclast precursors. Endocrinology 134, 1061 – 1066 (1994). 75. X. Cao, S. L. Teitelbaum, S. Dedhar, L. Zhang, and F. P. Ross, Retinoic acid and calreticulin, a novel ER-derived transcription factor, inhibit 1,25(OH)2D3-induced gene transcription. J. Bone Miner. Res. 9, S144 (1994). 76. X. Cao, F. P. Ross, L. Zhang, P. N. MacDonald, J. Chappel, and S. L. Teitelbaum, Cloning of the promoter for the avian integrin 3 subunit gene and its regulation by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 268, 27371 – 27380 (1993). 77. A. Miyauchi, J. Alvarez, E. M. Greenfield, A. Teti, M. Grano, S. Colucci, A. Zambonin-Zallone, F. P. Ross, S. L. Teitelbaum, and D. Cheresh, Recognition of osteopontin and related peptides by an v3 integrin stimulates immediate cell signals in osteoclasts. J. Biol. Chem. 266, 20369 – 20374 (1991). 78. K. Fuller and T. J. Chambers, Generation of osteoclasts in cultures of rabbit bone marrow and spleen cells. J. Cell. Physiol. 132, 441 – 452 (1987). 79. B. R. MacDonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long-term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 233 (1986). 80. G. D. Roodman, K. J. Ibbotson, B. R. MacDonald, T. J. Kuehl, and G. R. Mundy, 1,25-dihydroxyvitamin D3 causes formation of

81.

82. 83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

multinucleated cells with several osteoclast characteristics in cultures of primate marrow. Proc. Natl. Acad. Sci. USA 82, 8213 – 8217 (1985). D. L. Lacey, J. M. Erdmann, S. L. Teitelbaum, H.-L. Tan, J. Ohara, and A. Shioi, Interleukin-4, interferon gamma and prostaglandin E impact the osteoclast forming potential of murine bone marrow macrophages in vitro. Endocrinology 136, 2367 – 2376 (1995). A. Shioi, F. P. Ross, and S. L. Teitelbaum, Enrichment of generated murine osteoclasts. Calcif. Tissue Int. 55, 387 – 394 (1994). H. Kodama, M. Nose, S. Niida, and A. Yamasaki, Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J. Exp. Med. 173, 1291 – 1294 (1991). W. Wiktor-Jedrzejczak, A. Bartocci, A. W. Jr. Ferrante, A. AhmedAnsari, K. W. Sell, J. W. Pollard, and E. R. Stanley, Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic (op/op) mouse. Proc. Natl. Acad. Sci. USA 87, 4828 – 4832 (1990). D. N. Haylock, L. B. To, T. L. Dowse, C. A. Juttner, and P. J. Simmons, Ex vivo expansion and maturation of peripheral blood CD34 cells into the myeloid lineage. Blood 80, 1405 – 1412 (1992). S. Huang and W. M. M. Terstappen, Lymphoid and myeloid differentiation of single human CD34, HLA-DR, CD38 hematopoietic stem cells. Blood 83, 1515 – 1526 (1994). D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). P. Collin-Osdoby, L. Rothe, J. Weihe, T. Sunyer, and P. Osdoby, Osteoprotegerin ligand (OPGL) is expressed by human vascular endothelial cells and upregulated in response to the calciotropic agents TNF, PTH, or vitamin D3. J. Bone Miner. Res. 14, S232 (2000). R. Josien, B. R. Wong, H. L. Li, R. M. Steinman, and Y. Choi, TRANCE, a TNF family member, is differentially expressed on T cell subsets and induces cytokine production in dendritic cells. J. Immunol. 162, 2562 – 2568 (1999). W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, K. Yano, T. Morinaga, and K. Higashio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400 (1998). H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H.-L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). M. A. Horton and M. H. Helfrich, Antigenic markers of osteoclasts. In “Biology and Physiology of the Osteoclast.” (B. R. Rifkin and C. V. Gay, eds.), pp. 33 – 54. CRC Press, Boca Raton, FL, 1992. S. Nesbitt, A. Nesbit, M. Helfrich, and M. Horton, Biochemical characterization of human osteoclast integrins: Osteoclasts express v3, 21 and v1 integrins. J. Cell Biol. 268, 16737 – 16745 (1993). J. E. Fisher, M. P. Caulfield, M. Sato, H. A. Quartuccio, R. J. Gould, V. M. Garsky, G. A. Rodan, and M. Rosenblatt, Inhibition of osteoclastic

98

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111. 112.

113.

114.

ROSS AND TEITELBAUM bone resorption in vivo by echistatin, an “arginyl-glycyl-aspartyl” (RGD)-containing protein. Endocrinology 132, 1411 – 1413 (1993). M. Grano, P. Zigrino, S. Colucci, G. Zambonin, L. Trusolino, M. Serra, N. Baldini, A. Teti, P. C. Machisio, and A. Z. Zallone, Adhesion properties and integrin expression of cultured human osteoclastlike cells. Exp. Cell Res. 212, 209 – 218 (1994). G. van der Pluijm, H. Mouthaan, C. Baas, H. de Groot, S. Papapoulos, and C. Lowik, Integrins and osteoclastic resorption in three bone organ cultures: Differential sensitivity to synthetic Arg-Gly-Asp peptides during osteoclast formation. J. Bone Miner. Res. 9, 1021 – 1028 (1994). M. Flores, M. Norgard, D. Heinegard, F. P. Reinholt, and G. Andersson, RGD-directed attachment of isolated rat osteoclasts to osteopontin, bone sialoprotein, and fibronectin. Exp. Cell Res. 201, 526 – 530 (1992). M. A. Horton, M. L. Taylor, T. R. Arnett, and M. H. Helfrich, Arggly-asp (RGD) peptides and the anti-vitronectin receptor antibody 23C6 inhibit dentine resorption and cell spreading by osteoclasts. Exp. Cell Res. 195, 368 – 375 (1991). K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking 3 integrins are osteosclerotic due to dysfunctional osteoclasts. J. Clin. Invest. 105, 433 – 440 (2000). V. W. Engleman, G. A. Nickols, F. P. Ross, M. A. Horton, S. L. Settle, P. G. Ruminski, and S. L. Teitelbaum, A peptidomimetic antagonist of the v3 integrin inhibits bone resorption in vitro and prevents osteoporosis in vivo. J. Clin. Invest. 99, 2284 – 2292 (1997). F. P. Reinholt, K. Hultenby, A. Oldberg, and D. Heinegard, Osteopontin: A possible anchor of osteoclasts to bone. Proc. Natl. Acad. Sci. USA 87, 4473 – 4475 (1990). K. A. Hruska, F. Rolnick, M. Huskey, U. Alvarez, and D. Cheresh, Engagement of the osteoclast integrin alpha v beta 3 by osteopontin stimulates phosphatidylinositol 3-hydroxyl kinase activity. Endocrinology 136, 2984 – 2992 (1995). I. Nakamura, T. Sasaki, S. Tanaka, N. Takahashi, E. Jimi, T. Kurokawa, Y. Kita, S. Ihara, T. Suda, and Y. Fukui, Phosphatidylinositol-3 kinase is involved in ruffled border formation in osteoclasts. J. Cell Physiol. 172, 230 – 239 (1997). M. Chellaiah and K. Hruska, Osteopontin stimulates gelsolin-associated phosphoinositide levels and phosphatidylinositol triphosphatehydroxyl kinase. Mol. Biol. Cell 7, 743 – 753 (1996). H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontindeficient mice are resistant to ovariectomy-induced bone resorption. Proc. Natl. Acad. Sci. USA 96, 8156 – 8160 (1999). L. T. Duong, P. T. Lakkakorpi, I. Nakamura, M. Machwate, R. M. Nagy, and G. A. Rodan, PYK2 in osteoclasts is an adhesion kinase, localized in the sealing zone, activated by ligation of alpha(v)beta3 integrin, and phosphorylated by src kinase. J. Clin. Invest. 102, 881 – 892 (1998). B. F. Boyce, T. Yoneda, C. Lowe, P. Soriano, and G. R. Mundy, Requirement of pp60c-src expression for osteoclasts to form ruffled borders and resorb bone in mice. J. Clin. Invest. 90, 1622 – 1627 (1992). H. Palokangas, M. Mulari, and H. K. Vaananen, Endocytic pathway from the basal plasma membrane to the ruffled border membrane in bone-resorbing osteoclasts. J. Cell Sci. 110, 1767 – 1780 (1997). A. L. Boskey, Noncollagenous matrix proteins and their role in mineralization. Bone Miner. 6, 111 – 123 (1989). I. A. Silver, R. J. Murrills, and D. J. Etherington, Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp. Cell Res. 175, 266 – 276 (1988). H. C. Blair, S. L. Teitelbaum, R. Ghiselli, and S. Gluck, Osteoclastic bone resorption by a polarized vacuolar proton pump. Science 245, 855 – 857 (1989). Z. Q. Wang and S. Gluck, Isolation and properties of bovine kidney brush border vacuolar H()-ATPase: A proton pump with enzymatic

115.

116.

117.

118.

119.

120.

121. 122.

123.

124.

125.

126.

127.

128.

129.

130.

and structural differences from kidney microsomal H()-ATPase. J. Biol. Chem. 265, 21957 – 21965 (1990). S. L. Alper, J. Natale, S. Gluck, H. F. Lodish, and D. Brown, Subtypes of intercalated cells in rat kidney collecting duct defined by antibodies against erythroid band 3 and renal vacuolar H-ATPase. Proc. Natl. Acad. Sci. USA 86, 5429 – 5433 (1989). D. Chatterjee, M. Chakraborty, M. Leit, L. Neff, S. JamsaKellokumpu, R. Fuchs, and R. Baron, Sensitivity to vanadate and isoforms of subunits A and B distinguish the osteoclast proton pump from other vacuolar H ATPases. Proc. Natl. Acad. Sci. USA 89, 6257 – 6261 (1992). J. P. Mattsson, K. Vaananen, B. Wallmark, and P. Lorentzon, Omeprazole and bafilomycin, two proton pump inhibitors: differentiation of their effects on gastric, kidney and bone H-translocating ATPases. Biochim. Biophys. Acta 1065, 261 – 268 (1991). A. Teti, H. C. Blair, P. Schlesinger, M. Grano, A. Zambonin-Zallone, A. J. Kahn, S. L. Teitelbaum, and K. A. Hruska, Extracellular protons acidify osteoclasts, reduce cytosolic calcium, and promote expression of cell-matrix attachment structures. J. Clin. Invest. 84, 773 – 780 (1989). H. C. Blair, S. L. Teitelbaum, H.-L. Tan, C. M. Koziol, and P. H. Schlesinger, Passive chloride permeability charge coupled to HATPase of avian osteoclast ruffled membrane. Am. J. Physiol. 260, C1315 – C1324 (1991). P. H. Schlesinger, H. C. Blair, S. L. Teitelbaum, and J. C. Edwards, Characterization of the osteoclast ruffled border chloride channel and its role in bone resorption. J. Biol. Chem. 272, 18636 – 18643 (1997). S. A. Nesbitt and M. A. Horton, Trafficking of matrix collagens through bone-resorbing osteoclasts. Science 276, 266 – 269 (1997). J. Salo, P. Lehenkari, M. Mulari, K. Metsikko, and H. K. Vaananen, Removal of osteoclast bone resorption products by transcytosis. Science 276, 270 – 273 (1997). R. Baron, L. Neff, W. Brown, D. Louvard, and P. J. Courtoy, Selective internalization of the apical plasma membrane and rapid redistribution of lysosomal enzymes and mannose-6-phosphate receptors during osteoclast inactivation by calcitonin. J. Cell Sci. 97, 439 – 447 (1990). H. C. Blair, S. L. Teitelbaum, P. A. Schimke, J. D. Konsek, C. M. Koziol, and P. H. Schlesinger, Receptor-mediated uptake of mannose-6-phosphate bearing glycoprotein by isolated chicken osteoclasts. J. Cell. Physiol. 137, 476 – 482 (1988). M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzog, C. Debouck, and I. Kola, Cathepsin K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J. Bone Miner. Res. 14, 1654 – 1663 (1999). J. A. Hamilton, S. Lingelbach, N. C. Partridge, and T. J. Martin, Regulation of plasminogen activator by bone-resorbing hormones in normal and malignant osteoblasts. Endocrinology 116, 2186 – 2191 (1985). M. C. Meikle, S. Bond, R. M. Hembry, J. Compston, P. I. Croucher, and J. J. Reynolds, Human osteoblasts in culture synthesize collagenase and other matrix metalloproteinases in response to osteotropic hormones and cytokines. J. Cell Sci. 103, 1093 – 1099 (1992). N. C. Partridge, B. E. Kemp, M. C. Veroni, and T. J. Martin, Activation of adenosine 3’,5’-monophosphate-dependent protein kinase in normal and malignant bone cells by parathyroid hormone, prostaglandin E2, and prostacyclin. Endocrinology 108, 220 – 225 (1981). J. Pfeilschifter, J. Erdmann, W. Schmidt, A. Naumann, H. W. Minne, and R. Ziegler, Differential regulation of plasminogen activator and plasminogen activator inhibitor by osteotropic factors in primary cultures of mature osteoblasts and osteoblast precursors. Endocrinology 126, 703 – 711 (1990). W. G. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999).

99

CHAPTER 3 Osteoclast Biology 131. T. H. Vu, J. M. Shipley, G. Bergers, J. E. Berger, J. A. Helms, D. Hanahan, S. D. Shapiro, R. M. Senior, and Z. Werb, MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93, 411 – 422 (1998). 132. M. Zaidi, B. S. Moonga, and O. A. Adebanjo, Novel mechanisms of calcium handling by the osteoclast: A review-hypothesis. Proc. Assoc. Am. Physicians 111, 319 – 327 (1999). 133. S. Morony, C. Capparelli, R. Lee, G. Shimamoto, T. Boone, D. L. Lacey, and C. R. Dunstan, A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1 beta, TNF-alpha, PTH, PTHrP, and 1,25(OH)(2)D-3. J. Bone Miner. Res. 14, 1478 – 1485 (1999). 134. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145, 527 – 538 (1999). 135. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, D. S. A. Oliveira, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymphnode organogenesis. Nature 397, 315 – 323 (1999). 136. J. Li, I. Sarosi, X.-Q. Yan, S. Morony, C. Capparelli, H.-L. Tan, S. McCabe, R. Elliott, S. Scully, G. Van, S. Kaufman, S.-C. Juan, Y. Sun, J. Tarpley, L. Martin, K. Christensen, J. McCabe, P. Kostenuik, H. Hsu, F. Fletcher, C. R. Dunstan, D. L. Lacey, and W. J. Boyle, RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Natl. Acad. Sci. USA 97, 1566 – 1571 (2000). 137. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 138. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 139. E. Jimi, I. Nakamura, L. T. Duong, T. Ikebe, N. Takahashi, G. A. Rodan, and T. Suda, Interleukin 1 induces multinucleation and boneresorbing activity of osteoclasts in the absence of osteoblasts/stromal cells. Exp. Cell Res. 247, 84 – 93 (1999). 140. E. Jimi, I. Nakamura, T. Ikebe, S. Akiyama, N. Takahashi, and T. Suda, Activation of NF-kappaB is involved in the survival of osteoclasts promoted by interleukin-1. J. Biol. Chem. 273, 8799 – 8805 (1998). 141. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor  stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J. Exp. Med. 191, 275 – 285 (2000). 142. G. Franzoso, L. Carlson, L. Xing, L. Poljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997). 143. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 144. J.-P. David, K. Sabapathy, A. Behrens, B. Berger, K. Hochedlinger, O. Hoffmann, and E. F. Wagner, c-Jun phosphorylation by JNK1 but not by JNK2 is necessary for efficient osteoclastogenesis in response to TRANCE stimulation. J. Bone Miner. Res. 14, S149 (1999). 145. Y. Abu-Amer, F. P. Ross, J. Edwards, and S. L. Teitelbaum, Lipopolysaccharide-stimulated osteoclastogenesis is mediated by

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

160.

tumor necrosis factor via its p55 receptor. J. Clin. Invest. 100, 1557 – 1565 (1997). R. H. Arch, R. W. Gedrich, and C. B. Thompson, Tumor necrosis factor receptor-associated factors (TRAFs): A family of adapter proteins that regulates life and death. Genes Dev. 12, 2821 – 2830 (1998). L. Galibert, M. E. Tometsko, D. M. Anderson, D. Cosman, and W. C. Dougall, The involvement of multiple tumor necrosis factor receptor (TNFR)-associated factors in the signaling mechanisms of receptor activator of NF-kappaB, a member of the TNFR superfamily. J. Biol. Chem. 273, 34120 – 34127 (1998). H. Nakano, S. Sakon, H. Koseki, T. Takemori, K. Tada, M. Matsumoto, E. Munechika, T. Sakai, T. Shirasawa, H. Akiba, T. Kobata, S. M. Santee, C. F. Ware, P. D. Rennert, M. Taniguchi, H. Yagita, and K. Okumura, Targeted disruption of Traf5 gene causes defects in CD40- and CD27-mediated lymphocyte activation. Proc. Natl. Acad. Sci. USA 96, 9803 – 9808 (1999). W. C. Yeh, A. Shahinian, D. Speiser, J. Kraunus, F. Billia, A. Wakeham, J. L. de la Pompa, D. Ferrick, B. Hum, N. Iscove, P. Ohashi, M. Rothe, P. V. Goeddel, and T. W. Mak, Early lethality, functional NF-kappaB activation, and increased sensitivity to TNF-induced cell death in TRAF2-deficient mice. Immunity 7, 715 – 725 (1997). Y. Xu, G. Cheng, and D. Baltimore, Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses. Immunity 5, 407 – 415 (1996). M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, H. A. van der, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. D. Cao, J. M. Penninger, C. J. Paige, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). B. R. Wong, D. Besser, N. Kim, J. R. Arron, M. Vologodskaia, H. Hanafusa, and Y. Choi, TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-src. Mol. Cell 4, 1041 – 1049 (1999). P. J. Coffer, J. Jin, and J. R. Woodgett, Protein kinase B (c-Akt): A multifunctional mediator of phosphatidylinositol 3-kinase activation. Biochem. J. 335, 1 – 13 (1998). J. P. T. M. Van Leeuwen, J. C. Birkenhager, M. P. Bos, G. J. Van Der Bemd, M. P. Herrmann-Erlee, and H. A. P. Pols, Parathyroid hormone sensitizes long bones to the stimulation of bone resorption by 1,25-dihydroxyvitamin D3. J. Bone Miner. Res. 7, 303 – 309 (1992). J. A. Delmez, A. S. Dusso, E. Slatopolsky, and S. L. Teitelbaum, Modulation of renal osteodystrophy by extrarenal production of calcitriol. Am. J. Nephrol. 15, 85 – 89 (1995). G. J. King, M. E. Holtrop, and L. G. Raisz, The relation of ultrastructural changes in osteoclasts to resorption in bone cultures stimulated with parathyroid hormone. Metabol. Bone Dis. Rel. Res. 1, 67 (1978). E. Barengolts, R. Buschmann, D. H. Shevrin, E. C. Abramson, and S. C. Kukreja, Effects of hypercalcemia-producing tumor extract and parathyroid hormone on osteoclast ultrastructure. Acta Anat. 137, 160 – 164 (1991). K. Lee, J. D. Deeds, S. Chiba, M. Un-No, A. T. Bond, and G. V. Segre, Parathyroid hormone induces sequential c-fos expression in bone cells in vivo: In situ localization of its receptor and c-fos messenger ribonucleic acids. Endocrinology 134, 441 – 450 (1994). C. M. Silve, G. T. Hradek, A. L. Jones, and C. D. Arnaud, Parathyroid hormone receptor in intact embryonic chicken bone: Characterization and cellular localization. J. Cell Biol. 94, 379 – 386 (1982). M. F. Rouleau, H. Warshawsky, and D. Goltzman, Parathyroid hormone binding in vivo to renal, hepatic, and skeletal tissues of the rat using a radiographic approach. Endocrinology 118, 919 – 931 (1986).

100 161. T. J. Chambers, N. A. Athanasou, and K. Fuller, Effect of parathyroid hormone and calcitonin on the cytoplasmic spreading of isolated osteoclasts. J. Endocrinol. 102, 281 – 286 (1984). 162. T. J. Chambers, P. M. J. McSheehy, B. M. Thomson, and K. Fuller, The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116, 234 – 239 (1985). 163. J. Vilamitjana-Amedee, R. Bareille, F. Rouais, A. I. Caplan, and M. F. Harmand, Human bone marrow stromal cells express an osteoblastic phenotype in culture. In Vitro Cell Dev. Biol. Anim. 29A, 699 – 707 (1993). 164. N. Shevde, P. Anklesaria, J. S. Greenberger, I. Bleiberg, and J. Glowacki, Stromal cell-mediated stimulation of osteoclastogenesis. Proc. Natl. Acad. Sci. USA 205, 306 – 315 (1994). 165. H. M. I. Perry, W. Skogen, J. C. Chappel, G. D. Wilner, A. J. Kahn, and S. L. Teitelbaum, Conditioned medium from osteoblast-like cells mediate parathyroid hormone induced bone resorption. Calcif. Tissue Int. 40, 298 – 300 (1987). 166. H. M. I. Perry, W. Skogen, J. Chappel, A. J. Kahn, G. D. Wilner, and S. L. Teitelbaum, Partial characterization of a parathyroid hormonestimulated factor(s) from osteoblast-like cells. Endocrinology 125, 2075 – 2082 (1989). 167. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 168. G. V. Segre, A. B. Abou-Samra, H. Juppner, E. Schipani, T. Force, P. Urena, M. Freeman, X. F. Kong, L. F. Jr. Kolakowski, and J. Hock, Characterization of cloned PTH/PTHrP receptors. J. Endocrinol. Invest. 15, 11 – 17 (1992). 169. E. Blind, F. Raue, V. Knappe, J. Schroth, and R. Siegler, Cyclic AMP formation in rat bone and kidney cells is stimulated equally by parathyroid hormone-related protein (PTHrP) 1 – 34 and PTH 1 – 34. Exp. Clin. Endocrinol. 101, 150 – 155 (1993). 170. A. J. Fenton, T. J. Martin, and G. C. Nicholson, Carboxyl-terminal parathyroid hormone-related protein inhibits bone resorption by isolated chicken osteoclasts. J. Bone Miner. Res. 9, 515 – 519 (1994). 171. T. Sone, H. Kohno, H. Kikuchi, T. Ikeda, R. Kasai, Y. Kikuchi, R. Takeuchi, J. Konishi, and C. Shigeno, Human parathyroid hormonerelated peptide(107 – 111) does not inhibit bone resorption in neonatal mouse calvariae. Endocrinology 131, 2742 – 2746 (1992). 172. H. Kaji, T. Sugimoto, M. Kanatani, A. Miyauchi, T. Kimura, S. Sakakibara, M. Fukase, and K. Chihara, Carboxyl-terminal parathyroid hormone fragments stimulate osteoclast-like cell formation and osteoclastic activity. Endocrinology 134, 1897 – 1904 (1994). 173. T. Tamura, N. Udagawa, N. Takahashi, C. Miyaura, S. Tanaka, Y. Yamada, Y. Koishihara, Y. Ohsugi, K. Kumaki, and T. Taga, Soluble interleukin-6 receptor triggers osteoclast formation by interleukin 6. Proc. Natl. Acad. Sci. USA 90, 11924 – 11928 (1993). 174. N. J. Horwood, V. Kartsogiannis, J. M. Quinn, E. Romas, T. J. Martin, and M. T. Gillespie, Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144 – 150 (1990). 175. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 176. N. Takahashi, N. Udagawa, S. Tanaka, H. Murakami, I. Owan, T. Tamura, and T. Suda, Postmitotic osteoclast precursors are mononuclear cells which express macrophage-associated phenotypes. Dev. Biol. 163, 212 – 221 (1994).

ROSS AND TEITELBAUM 177. S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J. Clin. Invest. 91, 257 – 263 (1993). 178. J. Owens and T. J. Chambers, Macrophage colony-stimulating factor (M-CSF) induces migration in osteoclasts in vitro. Biochem. Biophys. Res. Commun. 195, 1401 – 1407 (1993). 179. G. Hattersley, E. Dorey, M. A. Horton, and T. J. Chambers, Human macrophage colony-stimulating factor inhibits bone resorption by osteoclasts disaggregated from rat bone. J. Cell. Physiol. 137, 199 – 203 (1988). 180. P. C. Heinrich, I. Behrmann, G. Muller-Newen, F. Schaper, and L. Graeve, Interleukin-6-type cytokine signalling through the gp130/ Jak/STAT pathway. Biochem. J. 334, 297 – 314 (1998). 181. G. Girasole, G. Passeri, R. L. Jilka, and S. C. Manolagas, Interleukin-11: A new cytokine critical for osteoclast development. J. Clin. Invest. 93, 1516 – 1524 (1994). 182. E. Romas, N. Udagawa, H. Zhou, T. Tamura, M. Saito, T. Taga, D. J. Hilton, T. Suda, K. W. Ng, and T. J. Martin, The role of gp130-mediated signals in osteoclast development: regulation of interleukin 11 production by osteoblasts and distribution of its receptor in bone marrow cultures. J. Exp. Med. 183, 2581 – 2591 (1996). 183. C. B. Ware, M. C. Horowitz, B. R. Renshaw, J. S. Hunt, D. Liggitt, S. A. Koblar, B. C. Gliniak, H. J. McKenna, T. Papayannopoulou, B. Thoma, et al. Targeted disruption of the low-affinity leukemia inhibitory factor receptor gene causes placental, skeletal, neural and metabolic defects and results in perinatal death. Development 121, 1283 – 1299 (1995). 184. N. Malik, H. S. Haugen, B. Modrell, M. Shoyab, and C. H. Clegg, Developmental abnormalities in mice transgenic for bovine oncostatin M. Mol. Cell. Biol. 15, 2349 – 2358 (1995). 185. K. Kawasaki, Y. H. Gao, S. Yokose, Y. Kaji, T. Nakamura, T. Suda, K. Yoshida, T. Taga, T. Kishimoto, H. Kataoka, T. Yuasa, H. Norimatsu, and A. Yamaguchi, Osteoclasts are present in gp130-deficient mice. Endocrinology 138, 4959 – 4965 (1997). 186. G. D. Roodman, N. Kurihara, Y. Ohsaki, A. Kukita, D. Hosking, A. Demulder, J. F. Smith, and F. R. Singer, Interleukin 6. A potential autorcrine/paracrine factor in Paget’s disease of bone. J. Clin. Invest. 89, 46 – 52 (1992). 187. N. Kurihara, S. V. Reddy, C. Menaa, D. Anderson, and G. D. Roodman, Osteoclasts formed by normal human bone marrow cells transduced with the measles virus nucleocapsid gene express a pagetic phenotype. J. Clin. Invest. 105, 607– 614 (2000). 188. S. V. Reddy, S. Takahashi, M. Dallas, R. E. Williams, L. Neckers, and G. D. Roodman, Interleukin-6 antisense deoxyoligonucleotides inhibit bone resorption by giant cells from human giant cell tumors of bone. J. Bone Miner. Res. 9, 753 – 757 (1994). 189. R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). 190. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin6. Science 257, 88 – 91 (1992). 191. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17 beta-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblasts in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 192. V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Ciliberto, G. A. Rodan, and F. Costantini, Interleukin-6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994).

CHAPTER 3 Osteoclast Biology 193. A. Grey, M. A. Mitnick, U. Masiukiewicz, B. H. Sun, S. Rudikoff, R. L. Jilka, S. C. Manolagas, and K. Insogna, A role for interleukin-6 in parathyroid hormone-induced bone resorption in vivo. Endocrinology 140, 4683 – 4690 (1999). 194. L. Sun, O. A. Adebanjo, B. S. Moonga, S. Corisdeo, H. K. Anandatheerthavarada, G. Biswas, T. Arakawa, Y. Hakeda, A. Koval, B. Sodam, P. J. Bevis, A. J. Moser, F. A. Lai, S. Epstein, B. R. Troen, M. Kumegawa, and M. Zaidi, CD38/ADP-ribosyl cyclase: A new role in the regulation of osteoclastic bone resorption. J. Cell Biol. 146, 1161 – 1172 (1999). 195. O. A. Adebanjo, B. S. Moonga, T. Yamate, L. Sun, C. Minkin, E. Abe, and M. Zaidi, Mode of action of interleukin-6 on mature osteoclasts: Novel interactions with extracellular Ca2 sensing in the regulation of osteoclastic bone resorption. J. Cell Biol. 142, 1347 – 1356 (1998). 196. W. P. Arend, H. G. Welgus, R. C. Thompson, and S. P. Eisenberg, Biological properties of recombinant human monocyte-derived interleukin 1 receptor antagonist. J. Clin. Invest. 85, 1694 – 1697 (1990). 197. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. T. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 198. M. E. Choen-Solal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. de Vernejoul Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. Metab. 77, 1648 – 1653 (1993). 199. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vassalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1703 (1997). 200. R. Felix, H. Fleisch, and P. R. Elford, Bone-resorbing cytokines enhance release of macrophage colony-stimulating activity by osteoblastic cell MC3T3-E1. Calcif. Tissue Int. 44, 356 – 360 (1989). 201. J. Glowacki, N. Shevde, and J. S. Greenberger, Distinct roles of interleukins-1,-6, and-11 in osteoclastic differentiation of hemopoietic progenitors. J. Bone Miner. Res. 8, A190 (1993). 202. D. D. Chaplin and K. Hogqist, Interactions between TNF and interleukin-1. In “Tumor Necrosis Factors.” The Molecules and Their Emerging Role in Medicine” (B. Beutler, ed.), pp. 197 – 220. Raven Press, New York. 203. Y. Sakagami, G. Girasole, X. P. Yu, H. S. Boswell, and S. C. Manolagas, Stimulation of interleukin-6 production by either calcitonin gene-related peptide or parathyroid hormone in two phenotypically distinct bone marrow-derived murine stromal cell lines. J. Bone Miner. Res. 8, 811 – 816 (1993). 204. K. Fuller, J. M. Owens, and T. J. Chambers, Macrophage inflammatory protein-1 alpha and IL-8 stimulate the motility but suppress the resorption of isolated rat osteoclasts. J. Immunol. 154, 6065 – 6072 (1995). 205. L. Rothe, P. Collin-Osdoby, Y. Chen, T. Sunyer, L. Chaudhary, A. Tsay, S. Goldring, L. Avioli, and P. Osdoby, Human osteoclasts and osteoclast-like cells synthesize and release high basal and inflammatory stimulated levels of the potent chemokine interleukin-8. Endocrinology 139, 4353 – 4363 (1998). 206. P. Collin-Osdoby, D. Kirsch, F. Anderson, O. Joost, A. Dean and P. Osdoby, The chemokine IL-8 as an autocrine inhibitor of osteoclast bone resorptive activity via IL-8 receptors expressed by avian osteoclasts and human osteoclast-like cells. J. Bone Miner. Res. 11, S179 (1996). 207. P. A. Collin-Osdoby, L. Rothe, S. Bekker, and F. Anderson, Osteoclast recruitment, formation, and bone resorptive activity are increased in parallel with neoangiogenesis on the chick chorioallantoic membrane in response to the angiogenic peptides bFGF or IL-8. Bone 23, S550 (1998).

101 208. C. Christiansen, M. S. Christensen, and I. Transbol, Bone mass in postmenopausal women after withdrawal of oestrogen/gestagen replacement therapy. Lancet 1, 459 – 461 (1981). 209. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 210. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 211. E. R. Simpson, Genetic mutations resulting in estrogen insufficiency in the male. Mol. Cell. Endocrinol. 145, 55 – 59 (1998). 212. R. Civitelli, S. Gonnelli, F. Zacchei, S. Bigazzi, A. Vattimo, L. V. Avioli, and C. Gennari, Bone turnover in postmenopausal osteoporosis: Effect of calcitonin treatment. J. Clin. Invest. 82, 1268 – 1274 (1988). 213. R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen prevents the skeletal effects of ovarian hormone deficiency in rats. J. Bone Miner. Res. 2, 449 – 456 (1987). 214. R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen inhibits osteoclast-mediated resorption of trabecular bone in ovarian hormone-deficient rats. Endocrinology 122, 1146 – 1150 (1988). 215. D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumour necrosis factor but not interleukin 6 from adult human osteoblasts in vitro. Osteoporos. Int. 2, 94 – 102 (1992). 216. L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 and tumor necrosis factor-: Lack of regulation by 17-estradiol. Endocrinology 130, 2528 – 2534 (1992). 217. G. Pioli, G. Basini, M. Pedrazzoni, G. Musetti, V. Ulietti, D. Bresciani, P. Villa, A. Bacci, G. Hughes, G. Russell, and M. Passeri, Spontaneous release of Interleukin-1 and interleukin-6 by peripheral blood monocytes after ovariectomy. Clin. Sci. 83, 503 – 507 (1992). 218. S. H. Ralston, R. G. G. Russell, and M. Gowen, Estrogen inhibits release of tumor necrosis factor from peripheral blood mononuclear cells in post-menopausal women. J. Bone Miner. Res. 5, 983 – 988 (1990). 219. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurty, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). 220. R. Pacifici, C. Brown, E. Puscheck, E. Friedrich, E. Slatopolsky, D. Maggio, R. McCracken, and L. V. Avioli, Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proc. Natl. Acad. Sci. USA 88, 5134 – 5138 (1991). 221. R. Pacifici, J. L. Vannice, L. Rifas, and R. D. Kimble, Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). 222. T. Sunyer, J. Lewis, P. Collin-Osdoby, and P. Osdoby, Estrogen’s boneprotective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). 223. Y. Tanaka and H. F. DeLuca, Bone mineral mobilization of 1,25(OH)2D3, a metabolite of vitamin D. Arch. Biochem. Biophys. 146, 574 – 578 (1971). 224. J. J. Reynolds, H. Pavlovitch, and S. Falsan, 1,25-Dihydroxycholecalciferol increases bone resorption in thyroparathyroidectomized mice. Calcif. Tissue Res. 21, 207 – 212 (1976). 225. L. G. Raisz, C. L. Trummel, M. F. Holick, and H. F. DeLuca, 1,25Dihydroxycholecalciferol: A potent stimulator of bone resorption in tissue culture. Science 175, 768 – 769 (1972). 226. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1990).

102 227. J. Abe, M. Morikawa, and K. Miyamoto, Synthetic analogues of vitamin D3 with an oxygen atom in the side chain skeleton: A trial of the development of vitamin D compounds which exhibit potent differentiation-inducing activity without inducing hypercalcemia. FEBS Lett. 226, 58 – 62 (1987). 228. M. Amling, M. Priemel, T. Holzmann, K. Chapin, J. M. Rueger, R. Baron, and M. B. Demay, Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: Formal histomorphometric and biomechanical analyses. Endocrinology 140, 4982 – 4987 (1999). 229. D. L. Lacey, L. E. Grosso, S. A. Moser, J. Erdmann, H. L. Tan, R. Pacifici and D. T. Villareal, IL-1-induced murine osteoblast IL-6production is mediated by the type 1 IL-1 receptor and is increased by 1,25-dihydroxyvitamin D3. J. Clin. Invest. 91, 1731 – 1742 (1993). 230. T. Bellido, G. Girasole, G. Passeri, X. P. Yu, H. Mocharla, R. L. Jilka, A. Notides, and S. C. Manolagas, Demonstration of estrogen and vitamin D receptors in bone marrow-derived stromal cells: Upregulation of the estrogen receptor by 1,25-dihydroxyvitamin D3. Endocrinology 133, 553 – 562 (1993). 231. E. Seeman, H. W. Wagner, K. P. Offord, R. Kumar, W. J. Johnson, and B. L. Riggs, Differential effects of endocrine dysfunction on the axial and appendicular skeleton. J. Clin. Invest. 69, 1302 – 1309 (1982). 232. J. C. Gallagher, J. Aaron, A. Horsman, R. Wilkinson, and B. E. Nordin, Corticosteroid osteoporosis. Clin. Endocrinol. Metab. 2, 255 – 368 (1973). 233. T. J. Hahn, L. R. Halstead, S. L. Teitelbaum, and B. H. Hahn, Altered mineral metabolism in glucocorticoid-induced osteopenia. J. Clin. Invest. 64, 655 – 665 (1979). 234. D. W. Dempster, M. A. Arlot, and P. J. Meunier, Mean wall thickness and formation periods of trabecular bone packets in corticosteroidinduced osteoporosis. Calcif. Tissue Int. 35, 410 – 417 (1983). 235. A. G. Need, J. C. Philcox, J. C. Hartley, and B. E. Nordin, Calcium metabolism and osteoporosis in corticsteroid-treated postmenopausal women. Aust. N. Z. J. Med. 16, 341 – 346 (1986). 236. J. Tobias and T. J. Chambers, Glucocorticoids impair bone resorptive activity and viability of osteoclasts disaggregated from neonatal rat long bones. Endocrinology 125, 1290 – 1295 (1989). 237. G. L. Wong, Basal activities and hormone responsiveness of osteoclast-like and osteoblast-like bone cells are regulated by glucocorticoids. J. Biol. Chem. 254, 6337 – 6340 (1979). 238. C. Colette, L. Monnier, N. Pares Herbute, F. Blotman, and J. Mirouze, Calcium absorption in corticoid treated subjects: Effects of a single oral dose of calcitriol. Horm. Metab. Res. 19, 335 – 338 (1987). 239. H. K. Nielsen, K. Thomsen, E. F. Eriksen, P. Charles, T. Storm, and L. Mosekilde, The effect of high-dose glucocorticoid administration on serum bone gamma carboxyglutamic acid-containing protein, serum alkaline phosphatase and vitamin D metabolites in normal subjects. Bone Miner. 4, 105 – 113 (1988). 240. Y. Suzuki, Y. Ichikawa, E. Saito, and M. Homma, Importance of increased urinary calcium excretion in the development of secondary hyperparathyroidism of patients under glucocorticoid therapy. Metab. Clin. Exp. 32, 151 – 156 (1983). 241. G. L. Wong, B. P. Lukert, and J. S. Adams, Glucocorticoids increase osteoblast-like bone cell response to 1,25(OH)2D3. Nature 285, 254 – 257 (1980). 242. S. C. Kukreja, E. N. Bowser, G. K. Hargis, W. J. Henderson, and G. A. Williams, Mechanism of glucocorticoid-induced osteopenia: Role of parathyroid glands. Proc. Soc. Exp. Biol. Med. 152, 358 – 361 (1976). 243. H. Ishikawa, H. Tanaka, K. Iwato, O. Tanabe, H. Asaoku, M. Nobuyoshi, I. Yamamoto, M. Kawano, and A. Kuramoto, Effect of glucocorticoids on the biologic activities of myeloma cells: Inhibition of interleukin-1 osteoclast activating factor-induced bone resorption. Blood 75, 715 – 720 (1990).

ROSS AND TEITELBAUM 244. J. L. Masferrer, S. T. Reddy, B. S. Zweifel, K. Seibert, P. Needleman, R. S. Gilbert, and H. R. Herschman, In vivo glucocorticoids regulate cyclooxygenase-2 but not cyclooxygenase-1 in peritoneal macrophages. J. Pharmacol. Exp. Ther, 270, 1340 – 1344 (1994). 245. S. A. Fraser, J. B. Anderson, D. A. Smith, and G. M. Wilson, Osteoporosis and fractures following thyrotoxicosis. Lancet 1, 981 – 983 (1971). 246. M. F. Fallon, H. M. I. Perry, M. Bergfeld, D. Droke, L. V. Avioli, and S. L. Teitelbaum, Exogenous hyperthyroidism with osteoporosis. Arch. Intern. Med. 143, 442 – 444 (1983). 247. M. Yamamoto, A. Markatos, J. G. Seedor, P. Masarachia, M. Gentile, G. A. Rodan, and R. Balena, The effects of the amino bisphosphonate alendronate on thyroid hormone-induced osteopenia in rats. Calcif. Tissue Int. 53, 278 – 282 (1993). 248. J. J. Stepan and Z. Limanova, Biochemical assessment of bone loss in patients on long-term thyroid hormone treatment. Bone Miner. 17, 377 – 388 (1992). 249. R. H. Follis, Skeletal changes associated with hyperthyroidism. Bull. Johns Hopkins Hosp. 92, 405 (1953). 250. J. C. Krakauer and M. Kleerekoper, Borderline-low serum thyrotropin level is correlated with increased fasting urinary hydroxyproline excretion. Arch. Intern. Med. 152, 360 – 364 (1992). 251. E. DeMenis, G. Da Rin, I. Roiter, P. Legovini, G. Foscolo, and N. Conte, Bone turnover in overt and subclinical hyperthyroidism due to autonomous thyroid hormone. Horm. Res. 37, 217 – 220 (1992). 252. H. Kawaguchi, C. C. Pilbeam, F. N. Woodiel, and L. G. Raisz, Comparison of the effects of 3,5,3’-triiodothyroacetic acid and triiodothyronine on bone resorption in cultured fetal rat long bones and neonetal mouse calvariae. J. Bone Miner. Res. 9, 247 – 253 (1994). 253. G. R. Mundy, J. L. Shapiro, J. G. Bandelin, E. M. Canalis, and L. G. Raisz, Direct stimulation of bone resorption by thyroid hormones. J. Clin. Invest. 58, 529 – 534 (1976). 254. J. M. Britto, A. J. Fenton, W. R. Holloway, and G. C. Nicholson, Osteoblasts mediate thyroid hormone stimulation of osteoclastic bone resorption. Endocrinology 134, 169 – 176 (1994). 255. P. Charles, J. W. Poser, and L. Mosekilde, Estimation of bone turnover evaluated by 47Ca-kinetics: Efficiency of serum bone gamma-carboxyglutamic acid-containing protein, serum alkaline phosphatase, and urinary hydroxyproline excretion. J. Clin. Invest. 76, 2254 – 2258 (1985). 256. P. Garnero, V. Vassy, A. Bertholin, J. P. Riou, and P. D. Delmas, Markers of bone turnover in hyperthyroidism and the effects of treatment. J. Clin. Endocrinol. Metab. 78, 955 – 959 (1994). 257. R. D. Harvey, K. C. McHardy, I. W. Reid, F. Paterson, P. D. Bewsher, A. Duncan, and S. P. Robins, Measurement of bone collagen degradation in hyperthyroidism and during thyroxine replacement therapy using pyridinium crosslinks as specific urinary markers. J. Clin. Endocrinol. Metab. 72, 1189 – 1194 (1991). 258. S. Saito, K. Yamasaki, S. Yamada, A. Matsumoto, T. Akatsu, N. Takahashi, Y. Shibasaki, T. Suda, and T. Fukuhara, A stable analogue of thromboxane A2, 9,11-epithio-11,12-methanothromboxane A2, stimulates bone resorption in vitro and osteoclast-like cell formation in mouse marrow culture. Bone Miner. 12, 15 – 23 (1991). 259. W. E. Gallwitz, G. R. Mundy, C. H. Lee, M. Qiao, G. D. Roodman, M. Raftery, S. J. Gaskell, and L. F. Bonewald, 5-lipoxygenase metabolites of arachidonic acid stimulate isolated osteoclasts to resorb calcified matrices. J. Biol. Chem. 268, 10087 – 10094 (1993). 260. S. Meghji, J. R. Sandy, A. M. Scutt, W. Harvey, and M. Harris, Stimulation of bone resorption by lipoxygenase metabolites of arachidonic acid. Prostaglandins 36, 139 – 149 (1988). 261. K. Imamura, H. Ozawa, T. Hiraide, N. Takahashi, Y. Shibasaki, T. Fukuhara, and T. Suda, Continuously applied compressive pressure induces bone resorption by a mechanism involving prostaglandin E2 synthesis. J. Cell. Physiol. 144, 222 – 228 (1990).

CHAPTER 3 Osteoclast Biology 262. J. L. Saffar and P. Leroux, Role of prostaglandins in bone resorption in a synchronized remodeling in the rat. Bone 9, 141 – 145 (1988). 263. J. R. Nefussi and R. Baron, PGE2 stimulates both resorption and formation of bone in vitro: Differential responses of the periosteum and the endosteum in fetal rat long bone cultures. Anat. Rec. 211, 9 – 16 (1985). 264. M. R. Wani, K. Fuller, N. S. Kim, Y. Choi, and T. Chambers, Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: Synergistic activation of differentiation, cell spreading, and fusion. Endocrinology 140, 1927 – 1935 (1999). 265. C. S. Lader and A. M. Flanagan, Prostaglandin E2, interleukin 1 alpha, and tumor necrosis factor-alpha increase human osteoclast formation and bone resorption in vitro. Endocrinology 139, 3157 – 3164 (1998). 266. T. J. Chambers and N. N. Ali, Inhibition of osteoclastic motility by prostaglandins I2, E2 and 6-oxoE1/. J. Pathol. 139, 383 – 397 (1983). 267. P. T. Lakkakorpi and H. K. Vaananen, Calcitonin, prostaglandin E2 and dibutyrl cyclic adenosine 3’,5’-monophosphate disperse the specific microfilament structure in resorbing osteoclasts. J. Histochem. Cytochem. 38, 1487 – 1493 (1990). 268. G. C. Nicholson, S. A. Livesey, J. M. Moseley, and T. J. Martin, Actions of calcitonin, parathyroid hormone, and prostaglandin E2 on cyclic AMP formation in chicken and rat osteoclasts. J. Cell. Biochem. 31, 229 – 241 (1986). 269. G. C. Nicholson, J. M. Moseley, A. J. P. Yates, and T. J. Martin, Control of cyclic adenosine 3’,5’-monophosphate production in osteoclasts: Calcitonin-induced persistent activation and homologous densensitization of adenylate cyclase. Endocrinology 120, 1902 – 1908 (1987). 270. J. Kaneshisa, Time course of “escape” from calcitonin-induced inhibition of motility and resorption of disaggregated osteoclasts. Bone 10, 125 – 129 (1989). 271. D. H. Copp, A. G. F. Davidson, and B. Cheney, Evidence for a new parathyroid hormone which lowers blood calcium. Proc. Can. Fed. Biol. Sci. 4, 17 (1961). 272. D. Gonzalez, G. Ghiringhelli, and C. Mautalen, Acute antiosteoclastic effect of salmon calcitonin in osteoporotic women. Calcif. Tissue Int. 38, 71 – 75 (1986). 273. G. C. Nicholson, J. M. Moseley, P. M. Sexton, F. A. O. Mendelsohn, and T. J. Martin, Abundant calcitonin receptors in isolated rat osteoclasts. J. Clin. Invest. 78, 355 – 360 (1986). 274. S. R. Goldring, M. Roelke, and J. Glowacki, Multinucleated cells elicited in response to implants of devitalized bone particles possess receptors for calcitonin. J. Bone Miner. Res. 3, 117 – 120 (1988). 275. A. Vignery, M. J. Raymond, H. Y. Qian, F. Wang, and S. A. Rosenzweig, Multinucleated rat alveolar macrophages express functional receptors for calcitonin. Am. J. Physiol. 261, F1026 – F1032 (1991). 276. A. J. Fenton, T. J. Martin, and G. C. Nicholson, Long-term culture of disaggregated rat osteoclasts: Inhibition of bone resorption and reduction of osteoclast-like cell number by calcitonin and PTHrP[107 – 139]. J. Cell. Physiol. 155, 1 – 7 (1993). 277. D. M. Kallio, P. R. Garant, and C. Minkin, Ultrastructural effects of calcitonin on osteoclasts in tissue cultures. J. Ultrastruct. Res. 39, 205 – 216 (1972). 278. S. Yumita, G. C. Nicholson, D. J. Rowe, G. N. Kent, and T. J. Martin, Biphasic effect of calcitonin on tartrate-resistant acid phosphatase activity in isolated rat osteoclasts. J. Bone Miner. Res. 6, 591 – 597 (1991). 279. H. Kaji, T. Sugimoto, A. Miyauchi, M. Fukase, K. Tezuka, Y. Hakeda, M. Kumegawa, and K. Chihara, Calcitonin inhibits osteopontin mRNA expression in isolated rabbit osteoclasts. Endocrinology 135, 484 – 487 (1994). 280. S. Pandalai, and C. V. Gay, Effects of parathyroid hormone, calcitonin, and dibutyrl-cyclic AMP on osteoclast area in cultured chick tibia. J. Bone Miner. Res. 5, 701 – 705 (1990).

103 281. T. J. Chambers, and A. Moore, The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J. Clin. Endocrinol. Metab. 57, 819 – 824 (1983). 282. M. H. Zheng, J. M. Papadimitriou, and G. S. Nicholson, A quantitative cytochemical investigation of osteoclasts and multinucleated giant cells. Histochem. J. 23, 180 – 188 (1991). 283. M. Zaidi, H. K. Datta, B. S. Moonga, and I. Maclntyre, Evidence that the action of calcitonin on rat osteoclasts is mediated by two G proteins acting via separate post-receptor pathways. J. Endocrinol. 126, 473 – 481 (1990). 284. H. H. Messer and D. H. Copp, Changes in response to calcitonin following prolonged administration to intact rats. Proc. Soc. Exp. Biol. Med. 146, 643 – 647 (1974). 285. J. F. Obie and C. W. Cooper, Loss of calcemic effects of calcitonin and parathyroid hormone infused continuously into rats using the alzet osmotic minipumps. J. Pharmacol. Exp. Ther. 209, 422 – 428 (1979). 286. N. Glajchen, S. Thomas, P. Jowell, S. Epstein, F. Ismail, and M. Fallon, The effect of high dose salmon calcitonin on bone mineral metabolism in the normal rat. Calcif. Tissue Int. 46, 28 – 32 (1990). 287. Y. Su, M. Chakraborty, M. H. Nathanson, and R. Baron, Differential effects of the 3’,5’-cyclic adenosine monophosphate and protein kinase C pathways on the response of isolated rat osteoclasts to calcitonin. Endocrinology 131, 1497 – 1502 (1992). 288. S. Takahashi, S. Goldring, M. Katz, S. Hilsenbeck, R. Williams, and G. D. Roodman, Downregulation of calcitonin receptor mRNA expression by calcitonin during human osteoclast-like cell differentiation. J. Clin. Invest. 95, 167 – 171 (1995). 289. B. S. Moonga, A. S. Alam, P. J. Bevis, F. Avaldi, C. L. Soncini, C. L. Huang, and M. Zaidi, Regulation of cytosolic free calcium in isolated rat osteoclasts by calcitonin. J. Endocrinol. 132, 241 – 249 (1992). 290. P. Collin-Osdoby, G. A. Nickols, and P. Osdoby, Bone cell function, regulation, and communication: Role for nitric oxide. J. Cell. Biochem. 56, 1 – 10 (1994). 291. L. L. Key, Jr., W. L. Ries, R. G. Taylor, B. D. Hays, and B. L. Pitzer, Oxygen derived free radicals in osteoclasts: The specificity and location of the nitroblue tetrazolium reaction. Bone 11, 115 – 119 (1990). 292. M. J. Steinbeck, W. H. Appel, Jr., A. J. Verhoeven, and M. J. Karnovsky, NADPH-oxidase expression and in situ production of superoxide by osteoclasts actively resorbing bone. J. Cell Biol. 126, 765 – 772 (1994). 293. S. T. Deahl, L. W. Oberley, T. D. Oberley, and J. H. Elwell, Immunohistochemical identification of superoxide dismutases, catalase, and glutathione-S-transferases in rat femora. J. Bone Miner. Res. 7, 187 – 198 (1992). 294. W. L. Ries, L. L. Key, Jr., and R. M. Rodriguiz, Nitroblue tetrazolium reduction and bone resorption by osteoclasts in vitro inhibited by a manganese-based superoxide dismutase mimic. J. Bone Miner. Res. 7, 931 – 939 (1992). 295. I. MacIntyre, M. Zaidi, A. S. Alam, H. K. Datta, B. S. Moonga, P. S. Lidbury, M. Hecker, and J. R. Vane, Osteoclastic inhibition: An action of nitric oxide not mediated by cyclic GMP. Proc. Natl. Acad. Sci. USA 88, 2936 – 2940 (1991). 296. T. P. Kasten, P. Collin-Osdoby, N. Patel, P. Osdoby, M. Krukowski, T. P. Misko, S. L. Settle, M. G. Currie, and G. A. Nickols, Potentiation of osteoclast bone-resorption activity by inhibition of nitric oxide synthase. Proc. Natl. Acad. Sci. USA 91, 3569 – 3573 (1994). 297. C. W. Lowik, P. H. Nibbering, M. van de Ruit, and S. E. Papapoulos, Inducible production of nitric oxide in osteoblast-like cells and in fetal mouse bone explants is associated with suppression of osteoclastic bone resorption. J. Clin. Invest. 93, 1465 – 1472 (1994). 298. P. D. Damoulis and P. V. Hauschka, Cytokines induce nitric oxide production in mouse osteoblasts. Biochem. Biophys. Res. Commun. 201, 924 – 931 (1994).

104 299. S. M. Hwang, C. A. Lopez, D. E. Heck, C. R. Gardner, D. L. Laskin, J. D. Laskin, and D. T. Denhardt, Osteopontin inhibits induction of nitric oxide synthase gene expression by inflammatory mediators in mouse kidney epithelial cells. J. Biol. Chem. 269, 711 – 715 (1994). 300. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nature Med. 2, 1132 – 1136 (1996). 301. S. P. Luckman, D. E. Hughes, F. P. Coxon, R. Graham, G. Russell, and M. J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including Ras. J. Bone Miner. Res. 13, 581 – 589 (1998). 302. S. P. Luckman, F. P. Coxon, F. H. Ebetino, R. G. Russell, and M. J. Rogers, Heterocycle-containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: Evidence from structure-activity relationships in J774 macrophages. J. Bone Miner. Res. 13, 1668 – 1678 (1998). 303. R. G. Russell, M. J. Rogers, J. C. Frith, S. P. Luckman, F. P. Coxon, H. L. Benford, P. I. Croucher, C. Shipman, and H. A. Fleisch, The pharmacology of bisphosphonates and new insights into their mechanisms of action. J. Bone Miner. Res. 14(Suppl. 2), 53 – 65 (1999). 304. D. W. Nicholson, Caspase structure, proteolytic substrates, and function during apoptotic cell death. Cell Death Differ. 6, 1028 – 1042 (1999). 305. N. Okahashi, M. Koide, E. Jimi, T. Suda, and T. Nishihara, Caspases (interleukin-1-beta-converting enzyme family proteases) are involved in the regulation of the survival of osteoclasts. Bone 23, 33 – 41 (1998). 306. N. Namba, X. Feng, J. C. Chappel, S. F. Dowdy, S. L. Teitelbaum, and F. P. Ross, A dominant negative form of caspase-7, the major downstream caspase in osteoclasts, inhibits endogenous apoptosis. J. Bone Miner. Res. 14, S148 (1999). 307. D. G. Walker, Congenital osteopetrosis in mice cured by parabiotic union with normal siblings. Endocrinology 91, 916 – 920 (1972). 308. D. G. Walker, Osteopetrosis cured by temporary parabiosis. Science 180, 275 (1973). 309. D. G. Walker, Spleen cells transmit osteopetrosis in mice. Science 190, 785 – 787 (1975). 310. M. M. Tondravi, S. McKercher, K. Anderson, J. M. Erdmann, M. Quiroz, R. Maki, and S. L. Teitelbaum, Targeted deletion of PU.1 induces osteopetrosis rescuable by bone marrow transplantation. Nature 386, 81 – 84 (1997). 311. R. Muller, Role of Fos proteins in the induction of transformation and DNA synthesis. In “The Fos and Jun Families of Transcription Factors” (P. Angel and P. Herrlich, eds.), pp. 189 – 197. CRC Press, Boca Raton, FL, 1994. 312. V. Iotsova, J. Caamano, J. Loy, Y. Young, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NFkB1 and NFkB2. Nature Med. 3, 1285 (1997). 313. C. Lowe, T. Yoneda, B. F. Boyce, H. Chen, G. R. Mundy, and P. Soriano, Osteopetrosis in src-deficient mice is due to an autonomous defect of osteoclasts. Proc. Natl. Acad. Sci. USA 90, 4485 – 4489 (1993). 314. T. Yoneda, M. Niewolna, C. Lowe, E. Izbicka, and G. R. Mundy, Hormonal regulation of pp60c-src expression during osteoclast formation in vitro. Mol. Endocrinol. 7, 1313 – 1318 (1993). 315. C. Grandori and H. Hanafusa, pp60c-src is complexed with a cellular protein in subcellular compartments involved in exocytosis. J. Cell Biol. 107, 2125 – 2135 (1988). 316. S. J. Parsons and C. E. Creutz, pp60c-src activity detected in the chromaffin granule membrane. Biochem. Biophys. Res. Commun. 34, 736 – 742 (1986). 317. K. M. Oddie, J. S. Litz, J. C. Balserak, D. M. Payne, C. E. Creutz, and S. J. Parsons, Modulation of pp60c-src tyrosine kinase activity during secretion in stimulated bovine adrenal chromaffin cells. J. Neurosci, Res. 24, 38 – 48 (1989).

ROSS AND TEITELBAUM 318. P. Schwartzberg, L. Xing, C. A. Lowell, E. Lee, L. Garrett, S. Reddy, G. D. Roodman, B. Boyce, and H. E. Varmus, Complementation of osteopetrosis in src-/- mice does not require src kinase activity. J. Bone Miner. Res. 11, S135 (1996). 319. W. S. Sly, D. Hewett-Emmett, M. P. Whyte, Y.-S. Yu, and R. E. Tashian, Carbonic anhydrase II deficiency identified as the primary defect in the autosomal recessive syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification. Proc. Natl. Acad. Sci. USA 80, 2752 – 2756 (1983). 320. T. Yamamoto, N. Kurihara, K. Yamaoka, K. Ozono, M. Okada, K. Yamamoto, S. Matsumoto, T. Michigami, J. Ono, and S. Okada, Bone marrow-derived osteoclast-like cells from a patient with craniometaphyseal dysplasia lack expression of osteoclast-reactive vacuolar proton pump. J. Clin. Invest. 91, 362 – 367 (1993). 321. S. C. Marks, Jr. and P. W. Lane, Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 67, 11 – 18 (1976). 322. W. W. Wiktor-Jedrzejczak, A. Ahmed, C. Szczylik, and R. R. Skelly, Hematological characterization of congenital osteopetrosis in op/op mouse: Possible mechanism for abnormal macrophage differentiation. J. Exp. Med. 156, 1516 – 1527 (1982). 323. N. Takahashi, N. Udagawa, T. Akatsu, H. Tanaka, Y. Isogai, and T. Suda, Deficiency of osteoclasts in osteopetrotic mice is due to a defect in the local microenvironment provided by osteoblastic cells. Endocrinology 128, 1792 – 1796 (1991). 324. S. C. Marks, Jr., M. F. Seifert, and J. L. McGuire, Congenitally osteopetrotic (op/op) mice are not cured by transplants of spleen or bone marrow cells from normal littermates. Metabol. Bone Dis. Rel. Res. 5, 183 – 186 (1984). 325. R. Felix, M. G. Cecchini, W. Hofstetter, P. R. Elford, A. Stutzer, and H. Fleisch, Impairment of macrophage colony-stimulating factor production and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse. J. Bone Miner. Res. 5, 781 – 789 (1990). 326. H. Yoshida, S.-I. Hayashi, T. Kunisada, M. Ogawa, S. Nishikawa, H. Okamura, T. Sudo, L. D. Shultz, and S.-I. Nishikawa, The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442 – 443 (1990). 327. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage colony stimulating factor restores in vivo bone resorption in the op/op mouse. Endocrinology 127, 2592 – 2594 (1990). 328. H. Kodama, A. Yamasaki, M. Nose, S. Niida, Y. Ohgama, M. Abe, M. Kumegawa, and T. Suda, Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony-stimulating factor. J. Exp. Med. 173, 269 – 272 (1991). 329. S. J. Begg, J. M. Radley, J. W. Pollard, O. T. Chisholm, E. R. Stanley, and I. Bertoncello, Delayed hematopoietic development in osteopetrotic (op/op) mice. J. Exp. Med. 177, 237 – 242 (1993). 330. Y. Y. Myint, K. Miyakawa, M. Naito, L. D. Shultz, Y. Oike, K. Yamamura, and K. Takahashi, Granulocyte/macrophage colonystimulating factor and interleukin-3 correct osteopetrosis in mice with osteopetrosis mutation. Am. J. Pathol. 154, 553 – 566 (1999). 331. E. Stanley, G. J. Lieschke, D. Grail, D. Metcalf, G. Hodgson, J. A. M. Gall, D. W. Maher, J. Cebon, V. Sinickas, and A. R. Dunn, Granulocyte/macrophage colony-stimulating factor-deficient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl. Acad. Sci. USA 91, 5592 – 5596 (1994). 332. P. F. Coccia, W. Krivit, J. Cervenka, C. Clawson, J. H. Kersey, T. H. Kim, M. E. Nesbit, N. K. Ramsay, P. I. Warkentin, S. L. Teitelbaum, A. J. Kahn, and D. M. Brown, Successful bone-marrow transplantation for infantile malignant osteopetrosis. N. Engl. J. Med. 302, 701 – 708 (1980). 333. P. J. Meunier, J. M. Coindre, C. M. Edouard, and M. E. Arlot, Bone histomorphometry in Paget’s disease. Arthritis Rheum. 23, 1095 – 1103 (1980).

CHAPTER 3 Osteoclast Biology 334. A. Boyde, Scanning electron microscope studies of bone. In “The Biochemistry and Physiology of Bone” (G. H. Bourne, ed.), Vol. 2, p. 259. Academic Press, New York, 1972. 335. B. Maldague and J. Malghem, Dynamic radiologic patterns of Paget’s disease of bone. Clin. Orthop. Rel. Res. 217, 126 – 151 (1987). 336. E. S. Siris, T. P. Clemens, D. McMahon, A. Gordon, T. P. Jacobs, and R. E. Canfield, Parathyroid function in Paget’s disease of bone. J. Bone Miner. Res. 4, 75 – 79 (1989). 337. R. G. G. Russell, R. Smith, C. Preston, R. J. Walton, and C. G. Woods, Diphosphonates in Paget’s disease. Lancet 1, 894 (1974). 338. M. R. A. Khairi, P. Meunier, C. Edouard, P. Courpron, J. Bernard, G. P. DeRosa, and C. C. Johnston, Quantitative bone histology in Paget’s disease of bone: Influence of sodium etidronate therapy. Calcif. Tissue Res. 22, 355 (1977). 339. A. Rebel, K. Malkani, M. Basle, and C. H. Bregeon, Osteoclast ultrastructure in Paget’s disease. Calcif. Tissue Res. 20, 187 (1976). 340. B. G. Mills, and F. R. Singer, Nuclear inclusions in Paget’s disease of bone. Science 194, 201 – 202 (1976). 341. G. Gherardi, V. L. O. Cascio, and E. Bonucci, Fine structure of nuclei and cytoplasm of osteoclasts in Paget’s disease of bone. Histopathology 4, 63 – 74 (1980). 342. L. Harvey, T. Gray, M. N. C. Beneton, D. L. Douglas, J. A. Kanis, and R. G. G. Russell, Ultrastructural features of the osteoclasts from Paget’s disease of bone in relation to a viral aetiology. J. Clin. Pathol. 35, 771 – 779 (1982). 343. S. V. Reddy, C. Menaa, N. Kurihara, G. Landucci, B. A. Koop, J. J. Windle, D. N. Forthal, and D. G. Roodman, Toward an in vivo model of Paget’s disease. J. Bone Miner. Res. 14, S143 (1999). 344. A. Kukita, C. Chenu, L. M. McManus, G. R. Mundy, and G. D. Roodman, Atypical multinucleated cells form in long term marrow cultures from patients with Paget’s disease. J. Clin. Invest. 85, 1280 – 1286 (1990). 345. J. A. Hoyland, A. J. Freemont, and P. T. Sharpe, Interleukin-6, IL-6 receptor, and IL-6 nuclear factor gene expression in Paget’s disease. J. Bone Miner. Res. 9, 75 – 80 (1994). 346. C. Menaa, S. V. Reddy, J. Bruder, D. Anderson, J. Cornish, T. Cundy, N. Kurihara, and G. D. Roodman, Enhanced RANK ligand expression and responsivity in Paget’s disease. J. Bone Miner. Res. 14, S143 (1999). 347. M. Tubiana-Hulin, Incidence, prevalence, and distribution of bone metastases. Bone 12, S9 – S10 (1991). 348. S. D. Averbuch, New bisphosphonates in the treatment of bone metastases. Cancer 72, 3443 – 3452 (1993). 349. A. Valentin-Opran, G. Elison, S. Saez, and G. R. Mundy, Estrogens and antiestrogens stimulate release of bone resorbing activity by cultured human breast cancer cells. J. Clin. Invest. 75, 726 – 731 (1985). 350. W. J. Kort, A. M. Bijma, J. J. van Dam, A. C. van der Ham, J. M. Hekking, H. F. van der Ingh, W. S. Meijer, M. G. Wilgenburg, and F. J. Zijlstra, Eicosanoids in breast cancer patients before and after mastectomy. Prostagl. Leuko. Ess. Fatty Acids 45, 319 – 327 (1992). 351. S. M. Scholl, P. Crocker, R. Tang, P. Pouillart, and J. W. Pollard, Is colony-stimulating factor-1 a key media of breast cancer invasion and metastasis? Mol. Carcinogen. 7, 207 – 211 (1993).

105 352. T. A. Guise, and G. R. Mundy, Cancer and bone. Endocr. Rev. 19, 18 – 54 (1998). 353. J. J. Yin, K. Selander, J. M. Chirgwin, M. Dallas, B. G. Grubbs, R. Wieser, J. Massague, G. R. Mundy, and T. A. Guise, TGF-beta signaling blockade inhibits PTHrP secretion by breast cancer cells and bone metastases development. J. Clin. Invest. 103, 197 – 206 (1999). 354. I. J. Diel, E. F. Solomayer, S. D. Costa, C. Gollan, R. Goerner, D. Wallwiener, M. Kaufmann, and G. Bastert, Reduction in new metastases in breast cancer with adjuvant clodronate treatment. N. Engl. J. Med. 339, 357 – 363 (1998). 355. A. A. Budayr, R. A. Nissenson, R. F. Klein, K. K. Pun, O. H. Clark, D. Diep, C. D. Arnaud, and G. J. Strewler, Increased serum levels of a parathyroid hormone-like protein in malignancy-associated hypercalcemia. Ann. Intern. Med 111, 807 – 812 (1989). 356. N. Horiuchi, M. P. Caulfield, J. E. Fisher, M. E. Goldman, R. L. McKee, J. E. Reagan, J. J. Levy, R. F. Nutt, S. B. Rodan, T. L. Schofield, T. L. Clemens, and M. Rosenblatt, Similarity of synthetic peptide from human tumor to parathyroid hormone in vivo and in vitro. Science 238, 1566 – 1568 (1987). 357. J. F. Rossi, R. Bataille, D. Chappard, C. Alexandre, and C. Janbon, B cell malignancies presenting with unusual bone involvement and mimicking multiple myeloma: Study of nine cases. Am. J. Med. 83, 10 – 16 (1987). 358. R. A. Kyle, Maintenance therapy and supportive care for patients with multiple myeloma. Semin. Oncol. 26, 35 – 42 (1999). 359. G. R. Mundy, L. G. Raisz, R. A. Cooper, G. P. Schechter, and S. E. Salmon, Evidence for the secretion of an osteoclast stimulating factor in myeloma. N. Engl. J. Med. 291, 1041 – 1046 (1974). 360. A. Valentin-Opran, S. A. Charhon, P. J. Meunier, C. M. Edouard, and M. E. Arlot, Quantitative histology of myeloma-induced bone changes. Br. J. Haematol. 52, 601 – 610 (1982). 361. X. G. Zhang, R. Bataille, M. Jourdan, S. Saeland, J. Banchereau, P. Mannoni, and B. Klein, Granulocyte-macrophage colony-stimulating factor synergizes with interleukin-6 in supporting the proliferation of human myeloma cells. Blood 76, 2599 – 2605 (1990). 362. M. Kawano, I. Yamamoto, K. Iwato, H. Tanaka, H. Asaoku, O. Tanabe, H. Ishikawa, M. Nobuyoshi, Y. Ohmoto, and Y. Hirai, Interleukin-1 beta rather than lymphotoxin as the major bone resorbing activity in human multiple myeloma. Blood 73, 1646 – 1649 (1989). 363. M. Kawano, T. Hirano, T. Matsuda, T. Taga, Y. Horii, K. Iwato, H. Asaoku, B. Tang, O. Tanabe, and H. Tanaka, Autocrine generation and requirement of BSF-2/IL-6 for human multiple myeloma. Nature 332, 83 – 85 (1988). 364. X. G. Zhang, B. Klein, and R. Bataille, Interleukin-6 is a potent myeloma-cell growth factor in patients with aggressive multiple myeloma. Blood 74, 11 – 13 (1989). 365. Y. Mori, T. Michigami, M. Dallas, M. Niewolna, B. Story, R. Lobb, G. R. Mundy, and T. Yoneda, Anti-4 integrin antibody suppresses the bone disease of myeloma and disrupts myeloma-marrow stromal cell interactions. J. Bone Miner. Res. 14, S173 (1999). 366. K. D. Merkel, J. M. Erdmann, K. P. McHugh, Y. Abu-Amer, F. P. Ross, and S. L. Teitelbaum, Tumor necrosis factor- mediates orthopedic implant osteolysis. Am. J. Pathol. 154, 203 – 210 (1999).

CHAPTER 4

The Biochemistry of Bone JAYASHREE A. GOKHALE AND ADELE L. BOSKEY Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021

PAMELA GEHRON ROBEY Bone Research Branch, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892

I. II. III. IV. V.

VI. Other Proteins VII. Requirements for Matrix Mineralization VIII. Pathways of Matrix Mineralization References

Introduction Collagen Glycosaminoglycan-Containing Proteins Glycoproteins Gla-Containing Proteins

in the cells that destroy bone (osteoclasts) and, in addition, the osteoblastic precursors that will replace the calcified cartilage with bona fide bone. The initial bone formed, woven bone, is a rather unorganized conglomeration of collagenous and noncollagenous proteins that induce the precipitation of mineral. Through modeling by osteoclasts, this primordial bone is removed and replaced by the formation of lamellar bone, a more highly organized structure with alternating layers of mineralized extracellular matrix, whose plywood-like structure provides bone with its mechanical strength. This structure is further reinforced by the development of the Haversian canal system centered around a blood vessel and a nerve, providing nutrients and signals to the cells entombed in bone (osteocytes), while maintaining communication through osteocytic cell processes in canaliculae. Initially it was hypothesized that mineralized matrices were composed of a unique set of matrix proteins, the combination of which would initiate the precipitation of hydroxapatite. However, with the development of techniques for isolation of the components without degradation [2,3] and cell culture systems that faithfully recapitulate the osteoblastic phenotype [4 – 9], it became apparent that most,

I. INTRODUCTION A. Bone: The Tissue The skeleton is essentially responsible for providing not only structural support and protection to the body’s organs, but also for serving as a reservoir for calcium, magnesium, and phosphate, ions that are of critical importance in physiology. The fabric of bone is a unique composite of living cells embedded in a remarkable threedimensional structure of extracellular matrix that is stabilized by a mineral, a carbonate-rich analogue of the geologic mineral hydroxyapatite. During development, mesenchymal cells form the skeleton via two basic pathways [reviewed in 1]. Endochondral bone is formed by an initial condensation of mesenchymal cells that induces the expression of the chondrocytic phenotype. The cartilaginous structure that is formed through the sequential expression of a number of genes that regulate morphogenesis serves as a temporary model. As part of that developmental sequence, the cartilage becomes calcified. The provisional calcified cartilagenous precursor is subsequently replaced by bone. Invasion by blood vessels brings

OSTEOPOROSIS, SECOND EDITION VOLUME 1

107

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

108

GOKHALE, BOSKEY, AND ROBEY

if not all, of the proteins synthesized by bone-forming cells are also synthesized by nonskeletal cells. There are few, if any, truly bone-specific proteins. However, it is now clear that the composition of bone is quite different: 70 – 90% of bone is composed of mineral with only 10 – 30% represented by protein. The ratio of collagenous to noncollagenous proteins differs greatly from that of other tissues, with collagenous protein composing 90% of the organic matrix (compared to 10 – 20% in other tissues) and noncollagenous proteins accounting for 10% (compared to as much as 80– 90% in soft tissues with the exception of tendon). In addition, all of the known collagenous and noncollagenous proteins in bone studied to date differ from those in other tissues in their chemical nature. These diverse forms result from alternative splicing of mRNA and different posttranslational modifications such as glycosylation, phosphorylation, and sulfation. These chemical differences most likely influence the function of these proteins, and the appropriate mixture provides the extracellular matrix with the ability to calcify. While the mineralized matrix is viewed by many as mere cement, it is actually a fairly dynamic aggregate structure. Although the most abundant proteins in bone matrix play structural roles in providing the scaffolding and binding sites for the regulation of mineral deposition and turnover, these proteins also function in modulating cellular activity. Cell – matrix interactions have become increasingly recognized as important determinants during all stages of development and tissue homeostasis. In addition, extracellular matrix proteins are the secretory products of cells in the osteoblastic lineage and, as such, they represent biochemical markers of the formation process (either as complete forms or precursor molecules) or the resorption process (in their degraded form). Therefore, it is necessary to have an indepth understanding of the biosynthetic process by which these proteins are formed. In addition, in order to decipher intelligently the clues provided by their assay ofin body fluids for determination of the status of the skeleton, it is important to understand their function in bone homeostasis.

B. Bone Matrix Formation: The Role of Maturational Stage It is now well accepted that bone formation is accomplished by cells in the osteoblastic lineage that pass through a series of maturational stages [10 – 13; see also Chapter 2]. The lineage is composed of cells that start off as uncommitted precursors. These precursors may be highly proliferative during development, but most likely at maturity, these stem cells are quiescent, thus preserving a reservoir of cells that have gone through a limited number of mitoses. Upon command, by signals that have yet to be fully identified, they become committed to the osteoblastic lineage and are

recognizable as fibroblast-like, proliferative osteoprogenitors. At some point, their rate of proliferation slows, and they are more aptly termed preosteoblasts, mainly by virtue of their location immediately adjacent to the real workhorse of the lineage, the osteoblast, on the opposite side of where mineralization will occur or is occurring. The term osteoblast is best thought of as a histological definition that describes a particular cell that has a large nucleus with prominent nucleoli indicative of a high rate of genetic expression. Additionally, it has a greatly expanded rough endoplasmic reticulum that is somewhat polarized such that the cell secretes enormous amounts of matrix toward the mineralization front, creating a layer of unmineralized osteoid. For reasons that are not yet known, a limited number of cells disengage themselves from the osteoblastic layer and are left behind as apposition proceeds. As these cells become buried in matrix that becomes mineralized (through a somewhat nebulous stage termed an osteoid osteocyte or an osteoblastic osteocyte), the cell maintains its communications with cells above it in the osteoblastic layer. This occurs via retention of cell contacts as cell processes become surrounded by mineralized matrix. Consequently, the most differentiated member of the lineage, the osteocyte, is still in communication with the osteoblastic layer above it. It is very likely that these embedded osteocytes serve a mechanoreceptor function. Through the cell processes in the canaliculae, they constantly monitor the surrounding environment and signal to the osteoblastic layer or lining cell layer when and where resorption is needed to remove and refurbish fatigued matrix or to respond to alterations in mechanical load. The expression of bone matrix proteins is somewhat stage specific (Fig. 1). However, currently it is still not perfectly clear whether one cell at one particular stage of maturation (the osteoblast) is capable of making all the components of the mineralized matrix or whether the process requires cells at different stages of maturation simultaneously to be present. Considering the amount of coupling that occurs between cells at different stages of maturation through gap and tight junctions, a great deal of cooperativity in the bone-forming process is likely. In vitro model systems have not really sorted out this issue, as none of the cell culture systems currently available go through the different stages of maturation in a synchronous fashion. Uncommitted progenitors, preosteoblasts, osteoblasts, and sometimes even osteocytes can be identified in populations of cells that are in the process of forming bone nodules in vitro (the current state of the art in vitro) [14]. It has been demonstrated that proliferation of a single cell gives rise to the whole spectrum of maturational stages, and nature must have a reason for reestablishing this continuum. Although identification of the various maturational stages and the characterization of the biological properties of the cells within each stage have advanced our under-

109

CHAPTER 4 The Biochemistry of Bone

FIGURE 1

Maturational stage and bone matrix gene expression. Osteoblastic cells pass through a series of maturational stages, each of which can be partially characterized by the bone matrix proteins that they produce. In addition, osteoclasts also secrete proteins that become incorporated into mineralized matrix.

standing of bone formation greatly, the mechanism of how and when various factors influence bone matrix protein expression is not clear. The literature is cluttered with numerous reports describing the effects of hormones, growth factors, cytokines, chemokines, and so on osteoblastic metabolism, and the results of these studies are extremely variable. This variability may reflect the fact that the effects that these factors exert most likely depend on animal species, donor age and sex, stage of maturation of the cells, state of transformation, and certain cell culture parameters, such as length of time in culture, density, and nutrient conditions. This chapter focuses on the structural aspects of bone matrix proteins and their genes and only highlights what factors are known to influence gene expression. The topic of what factors influence expression at a particular stage of maturation is addressed in more detail in Chapter 2. Only a few years ago, relatively few human bone matrix genes had been isolated and characterized. The current list is quite lengthy, and most of the information provided in this chapter pertains to human genes and proteins unless information is available only from another animal species.

C. Mineral The mechanical strength of bone is attributable to the presence of mineral, which converts the pliable organic

matrix into a more rigid structure [15,16]. In the composite structure of cells, protein and mineral, the bone mineral (apatite) crystals, approximately 300 Å in their longest dimension, are aligned along the collagen fibril axis [17 – 19]. A variety of structural analyses, including X-ray and electron diffraction [20 – 22], infrared spectroscopy [23], high-voltage electron microscopy [24], nuclear magnetic resonance (NMR), and X-ray absorption fine structure analysis (EXAFS) [25 – 27] have shown that mineral crystals are analogous to the naturally occurring geologic mineral, hydroxyapatite (Ca10(PO4)6(OH)2) (Fig. 2). In bone, mineral includes numerous ions not found in pure hydroxyapatite. For example, HPO42, CO32, Mg2, Na, F, and citrate are adsorbed on the crystal surfaces and/or substituted in the lattice for the constituent Ca2, PO43, and OH ions [28 – 34]. This poorly crystalline apatite, because of its small size and large number of latticesubstituted and surface-adsorbed ion impurities, goes into solution more readily than the larger, more perfect crystals of geologic hydroxyapatite. For example, some ions (HPO42, CO32, Na2, and F) can be in the lattice and on the surface, whereas others (Mg2 and citrate) prefer surface locations. This altered solubility allows bone mineral to play an important role in Ca2, Mg2, and PO43 ion homeostasis [35]. Despite claims of the presence of other mineral phases in bone, e.g., brushite [36], octacalcium phosphate [37],

110

GOKHALE, BOSKEY, AND ROBEY

FIGURE 2 Crystal lattice structure. A portion of the apatite structure is depicated as it would be viewed along the length (c axis) of the hydroxyapatite crystal, showing the hexagonal arrangement of the Ca2 and PO43ions about the OH position.

amorphous calcium phosphate [38], and whitlockite [39], current evidence supports the viewpoint that bone mineral is predominantly apatitic, with numerous, perhaps unique, impurities [40]. This viewpoint is maintained for the rest of the chapter.

gene activation and regulation, (ii) RNA transcription and processing, (iii) translation into protein, posttranslational modification, and secretion from the cell, and finally (iv) deposition in the extracellular matrix. The following description pertains specifically to the individual  chains of type I collagen. However, virtually all of these processes apply to the other proteins found in bone.

II. COLLAGEN In the vertebrate body, the major structural protein is collagenous in nature. Collagen is defined as a trimeric molecule composed of  chain subunits that are wound together to form a triple helix [41]. A significant feature of component  chains is that their primary sequence is composed almost entirely of a repeating triplet sequence, GlyX-Y, where X is often proline and Y is often hydroxyproline [42]. Collagenous proteins are either homotrimeric, composed of three identical  chains, or heterotrimeric, with two or three different  chains. Currently there are over 23 different  chains that associate to yield 13 different types of collagens, with 6 more potential types that to date have only been identified at the cDNA level [43 – 45]. The various forms of collagen are as diverse as the tissues in which they are found. These variations on a theme convey distinct features such that the collagen type is uniquely suited to carry out particular functions in a given tissue. In each of these tissues, collagen most likely serves a mechanical function. For example, in mineralized tissues, collagen fibrils provide tensile strength [15,46]. Given the complex nature of collagen expression, it serves as a useful example for describing the pathway of (i)

A. Gene Structure The structure of mammalian genes is as varied as the proteins for which they code. An example of a prototype gene structure is shown in Fig. 3. In the 5 region of the gene, there is usually a sequence that is not transcribed from DNA into RNA. As described later, this untranslated region (UTR), the promoter, is the primary site for the regulation of gene activity. Structures can also form due to the presence of palindromic type sequences (mirror image and complimentary) that allow the DNA to take on conformations other than the typical Watson – Crick double helix. These unusual conformations (such as z DNA) may serve as recognition sites for certain factors and enzymes that regulate gene activity [47 – 50]. Following this 5 UTR is the portion of the gene that will be transcribed into an RNA sequence through the action of RNA polymerases. Genes contain sequences that are found in mature RNA and dictate the amino acid sequence of protein (exons), interspersed with sequences that are removed and do not end up in mature mRNA or in the translated protein (introns). The exon – intron structure of a gene

111

CHAPTER 4 The Biochemistry of Bone

FIGURE 3 Basic structure of prototype gene. In addition to exons (E), introns (I), and the transcription start site (toward the 3 end), the gene contains a promoter toward the 5 end, upstream of the transcription start site. It includes defined sequences or cis-acting elements such as CCAAT and TATA boxes. Trans-acting factors (proteins) can bind to cis elements to regulate transcription.

can be simple, with only a single exon, or highly complex, with multiple exons as exemplified by the genes for type I collagen. The gene for the constituent  chains of type I collagen, COL1A1, located on chromosome 17q21.3-q22, is 18 kb in length and contains 51 exons [51,52], whereas COL1A2, on chromosome 7q21.3-q22, is 35 kb and has 52 exons [52 – 54]. It is often the case that exons code for functional regions of the resulting protein, and may thereby represent an ancestral gene. In many of the genes for collagen  chains, the exons are often 54 bases, or multiples thereof, that code for six [GLY-X-Y]n triplets. Following the open reading frame that designates the protein sequence is another stretch of DNA sequence that codes for mRNA that again does not end up in protein. This 3-untranslated region (3 UTR) contains sequences that may also influence gene activity. Signals for polyadenylation, another posttranscriptional modification, are also located within this region [55,56].

B. Gene Activation and Transcription The process of gene activation is a complex series of events that are mediated by cis and trans-acting factors (see Fig. 3). Cis-acting elements are defined as sequences present in DNA that are required for the binding of factors resident within the nuclear environment that influence (either positively or negatively) gene transcription. They include the binding sites for enzymes that are required for the synthesis of RNA complementary to the DNA template, such as polymerases. These types of cis-acting elements are exemplified by the sequences TATA and CAAT, which have been postulated to serve as the binding/orientation sites for

polymerases. In addition, GC-rich regions may also regulate polymerase activity, perhaps through the formation of three-dimensional structures. Cis-acting elements are found most commonly upstream (5) from the RNA transcription start site (the promoter region), where polymerase initiates the synthesis of mRNA. However, cis elements can also be located in sites quite distant from the transcription start site and may also reside in intronic sequences, or in the 3 regions of the gene. Cis-acting elements are also the binding sites for transacting factors, which are defined as proteins that either singly or in combination have the ability to up- or downregulate gene activity. Trans-acting elements are extremely diverse, and many are the gene products of protooncogenes or receptors bound to their ligands. These DNA-binding proteins have a number of structural motifs that allow them to bind to the double helix [reviewed in 57], including helix – loop – helix [58,59], and leucine zipper proteins, which can also have important functions in the cytoplasm [60]. Other motifs include HMG-1 box-binding proteins [61] and zinc-binding proteins (fingers, twists, and clusters) [62]. By binding to specific sequences, these complexes can either enhance the activity of the polymerizing enzymes, leading to high levels of a particular mRNA species, or suppress their activity, such that there are very few or virtually no copies produced. The promoters for COL1A1 and COL1A2 have been characterized in detail and contain similar but not identical promoter elements [63 – 67]. At 29 bp from the transcription start site, the COL1A1 promoter contains a TATA box, whereas it is absent in the COL1A2 promoter. Further upstream, both contain a CCAAT sequence (100 bp in COL1A1 and 82 bp in COL1A2), as well as a long stretch of Cs and Ts, which confer S1 nuclease and DNAse

112

GOKHALE, BOSKEY, AND ROBEY

hypersensitivity, implying a relatively open structure. The COL1A1 sequence then becomes A-G rich, which could indicate another potential change in tertiary structure due to base pairing between these two regions. An SP1 element (the binding site of a constitutive transcription factor) is also present. Studies utilizing rather extended regions of this COL1A1 promoter have demonstrated that it is active. However, maximal activity may require sequences present in the first intron, which also contains SP1 and DNAse hypersensitive regions [68]. This concept remains controversial because both stimulatory [69,70] and inhibitory [70] elements have been described in this region. The 2(I) contains a positive enhancer in the first intron and a CAATbinding protein (CTF/NF1)-binding site [71]. Other elements include a vitamin D response element (VDRE) in COL1A1 [72] and a CAAT-like region that binds to nuclear factor 1 (NF1) in the COL1A2 promoter [73]. It is of interest that the amount of mRNA for COL1A1 is twice the amount of COL1A2, a ratio that is reflected in the final triple helical molecule. Consequently, there must be factors (e.g., ascorbic acid) that regulate the production of these two gene products such that there are never excessive amounts of COL1A2 mRNA [74 – 77].

C. Gene Regulation While there is only one copy of the genes that code for the COL1A1 and COL1A2 in mammalian genomes, the regulation of type I collagen production in bone is somewhat different from that in soft connective tissues. It has been recently found that different parts of the COL1A1 promoter are required to maintain production by bone-forming cells in vitro [78 – 79]. In bone cell and organ cultures, collagen synthesis increased by heparin [80], organic phosphate [81], interleukin (IL)-4 [82], and gallium [83]. In contrast, collagen synthesis is decreased by prostaglandin E2 [84] 1,25-dihydroxyvitamin D3 [85], cortisol [86], parathyroid hormone (PTH) [87], epidermal growth factor (EGF) [88], basic fibroblast growth factor (bFGF) [89], IL10 [90], and lead [91]. Although the COL1A1 promoter contains a VDRE-like element, binding of this element by the VDR that has bound to its ligand inhibits expression. In addition, removing this element from the promoter does not totally abolish the inhibitory effect of 1,25-dihydroxyvitamin D3, indicating that other cis- and/or trans-acting factors are involved [85]. Depending on the concentration and the stage of the cell culture, dexamethasone, can either increase or decrease collagen synthesis [92,93].

D. RNA Processing Once transcription is initiated, a precursor form of RNA (often termed Hn or heteronuclear RNA) is transcribed

from the DNA template. This precursor form contains the intronic sequences that must be removed prior to translation to yield mRNA with only exonic, 5 UTR, and 3 UTR sequences. This process is accomplished through the action of a number of enzymes that associate to form what is termed a spliceosome [94,95]. Several types of splicing reactions can occur based on the sequences that bridge the exon and intron. Splicing occurs by bringing together the junction consensus sequence (GT ........ AT) at both the 5 and the 3 end of the intron and cleavage by enzymes via formation of a lariat-like structure [96]. Differential splicing can also occur via exon skipping, or exon splitting due to the presence of splice junctions that are buried within the exonic sequences. The factors that regulate differential splicing are not well understood but are believed to provide a particular cell type with the ability to generate different mRNA species. The mRNA is further modified by the addition of a methyl cap at the 5 end. The addition of a polyadenyl tail in the noncoding region at the 3 end completes the formation of the mature mRNA. The regulation of polyadenylation (choice between different polyadenylation sites and length of polyadenylation) is not well known but is implicated in regulating the half-life of the mRNA. At this point, the fully processed mRNA species are transported to the cytoplasm (Fig. 4). In the case of COL1A1, the sizes are 7.2 and 5.9 kb, and for COL1A2, the sizes are 6.5 and 5.5 kb.

E. Translation and Secretion Once in the cytoplasm, the mRNA species associate with ribosomes, possibly through the formation of stem loops that have been described in some mRNA species, including the mRNA for COL1A1 [97]. The initiation codon in most proteins is AUG (or CUG) for methionine. The initial translation of secreted proteins produces a signal peptide that allows binding of the ribosome and nascent polypeptide chain to the cytoplasmic surface of the endoplasmic reticulum (ER). As translation proceeds, the protein is extruded into the lumen of the ER and the signal peptide is cleaved. However, it should be noted that, like many other proteins, collagen  chains are synthesized in a precursor form. Noncollagenous sequences are found at both amino (pN) and the carboxyl (pC) termini, resulting in a chain termed a pro chain [reviewed in 98]. During passage of the nascent polypeptides through the ER, posttranslational processing begins by the action of certain enzymes that are resident along the secretory pathway. Prolyl hydroxylases will form hydroxyprolyl residues immediately preceding glycyl residues in the triplet sequence. In most cases, proline is hydroxylated in the 4 position; however, in certain tissues, another enzyme will hydroxylate in the 3 position. Prolyl hydroxylation is required

113

CHAPTER 4 The Biochemistry of Bone

FIGURE 4 Gene activation and RNA transcription. Gene activity is regulated by the interaction of cis-acting (DNA sequences) and trans-acting (nuclear factors) elements that either initiate or repress transcription. Once a gene is activated, polymerases translate the antisense genomic sequence, including the exons (which ultimately code for the protein) and the introns (intervening sequences). The initial transcription product, HnRNA, is processed through the action of spliceosomes to remove the intervening sequences. RNA is further modified by methylation of the 5 end and by the addition of poly(A) to the 3 end prior to transport to the cytoplasm.

for the formation of a stable helical structure because the hydroxyl groups participate in intrahelix hydrogen bonding. Certain lysyl residues will also be hydroxylated by lysyl hydroxylase in the ER. Hydroxylysyl residues can be further modified through the action of sugar transferases that add first galactosyl residues, and sometimes an additional glucosyl residue to form galactosyl-hydroxylysine and glucosyl-galactosyl-hydroxylysine, respectively. The function of these sugar modifications is unknown. Once translation of the pro chains has been completed, the noncollagenous precursor sequences at the carboxy termini of two pro1 and one pro2 chains associate and bond through disulfide bridges. This aggregation is then followed by triple helix formation progressing from the C terminus moving toward the N terminus as the protein passes through the ER. Posttranslational modifications continue until the target residues are no longer accessible due to triple helix formation [reviewed in 98]. The formation of this unique conformation is dependent on the GLY-X-Y sequence throughout the helical domain of the molecule. This sequence allows the formation of rodlike triple helices [41], as glycine is the only amino acid small enough to fit within the center of the triple helix. The X and Y amino acids occur on the surface of the triple helix and are arranged in charged clusters, which facilitate the interaction of the individual molecules in the formation of

fibrils [99]. It is currently believed that there are specific domains within the fibrils that interact with cells, fibronectin, decorin, and other matrix molecules, but these have not been well identified. Individual  chains of the collagen molecule coil about one another in an extended rigid helix. The structure is stabilized by hydrogen bonding between OH groups on hydroxyproline and intrachain water [100] and by aldehyde derived cross-links [101]. More details of the collagen structure have bene reviewed by Kuhn [102]. In addition, the crystal structure of a small triple helical peptide [103] has provided confirmation of earlier structure predictions. With the passage of the hydroxylated and glycosylated precursor to the Golgi apparatus, the procollagen molecule is further modified by the addition of N-linked oligosaccharides to certain asparaginyl residues in the carboxy-terminal precursor region. Initially, a high mannose structure is transferred to the acceptor residue, and the oligosaccharide is trimmed and rebuilt by the addition of N-acetylgalactosamine and N-acetylglucosamine to form a complex type oligosaccharide. Again, the function of these sugar modifications is unknown, although they are thought to be important for transport of the molecule from the cell [104]. In the Golgi apparatus, O-linked glycosylation reactions also occur; however, collagen  chains are not modified in this fashion.

114

GOKHALE, BOSKEY, AND ROBEY

The process by which collagen molecules are moved from one cellular compartment to another is not completely known, but it has been speculated to be mediated by proteins termed chaperonins, some of which are heat shock proteins [105]. Due to the major role that collagen plays in all tissues, and the complicated nature of its biosynthesis, its processing and secretion may be tightly regulated by such proteins. It is likely that a derangement in the secretory apparatus that affects collagen secretion would result in disease. The triple helical and posttranslationally modified procollagen molecules within secretory granules are then transported to the cell surface and extruded from the cell (Fig. 5).

F. Deposition and Fibril Formation Although the exact process of fibril formation is not fully characterized, it has been speculated that it can occur either prior to or during the process of secretion by the formation of deep invaginations of the cell membrane, which provide a secluded environment [106]. It is clear that the nontriple helical precursor extensions must be removed in

FIGURE 5

an orderly fashion in order for normal fibers to form. Specific peptidases remove the N-terminal extension (yielding pC collagen) or the C-terminal extension (yielding pN collagen). When both extensions are removed, the fully mature triple helical collagen molecule still contains short nontriple helical telopeptides at both termini. Interestingly, it has been suggested that both pN and pC propeptides participate in feedback inhibition of collagen synthesis [107 – 111]. In addition, the pN peptide remains, at least in part, within bone matrix and was identified as the 24-kDa phosphoprotein [112,113]. The pC propeptide escapes into the circulation and has been used as a measure of collagen biosynthesis [114]. However, because pC can be contributed from the synthesis by soft tissues as well, it is not totally reflective of bone formation. The mature collagen molecules then form head to tail and lateral associations, and it is thought that the lateral associations are inhibited by the presence of the pN collagen. The addition of pN collagen or pC collagen molecules to the outermost layer of fibers may in fact dictate when fibril growth stops. Small fibers tend to be coated with pN collagen, whereas larger fibers are associated with pC collagen [115,116]. Fiber diameter growth may also be regulated by

Protein translation, modification, and secretion. Once in the cytoplasm, mRNA is translated into protein by ribosomes. In the case of bone matrix proteins (and the majority of secreted proteins), mRNA codes for a signal peptide that allows the ribosome to attach to the endoplasmic reticulum, allowing the nascent peptide to be extruded. In the lumen, hydroxylases modify certain prolyl and lysyl residues. Once the C-terminal portion is completed, the C propeptides of three chains associate and triple helix formation proceeds to the amino terminus. The pathway to secretion proceeds through the Golgi apparatus where the molecule is further modified by the addition of oligosaccharides to the C propeptide, and galactosyl and glucosyl-galactosyl residues to lysine and hydroxylysine residues by sugar transferases. Upon secretion, propeptides at N and C termini are cleaved, and the molecules are deposited in a four-dimensional staggered array and are further stabilized by covalent cross-links.

115

CHAPTER 4 The Biochemistry of Bone

the binding of other collagenous or noncollagenous proteins to the outer fibril surface [116 – 119]. This is an important point because it may be that the surface of type I collagen fibrils is not directly accessible for binding, and that collagen may be functioning in a purely structural fashion when intact within bone collagen fibers. Other activities may occur once fibril-associated proteins are removed or when the collagen fibril is degraded. The structure of fibrillar type I collagen has been described in detail [120]. Katz’s structural models, based on low angle X-ray scattering, confirm the structural pattern predicted by Hodge and Petrushka [121]. According to this model, individual collagen fibrils are aligned in a quarterstaggered array, with a 280-nm periodicity. As a result of the quarter stagger, there are gaps (holes) within the fibrillar structures, and it is in these gaps and in the overlapping regions adjacent to them that bone mineral first appears [18,122 – 125]. Specifically, Traub and co-workers [125] have shown that the first mineral crystals appear in a specific region of the collagen fibril known as the e band. However, it should be noted that the majority of these studies have been performed in mineralizing turkey tendon, which may not be identical to other patterns of matrix mineralization. Once fibers have been formed in the extracellular environment, they are further stabilized by the formation of inter- and intramolecular cross-links. This process occurs through the action of lysyl oxidase, which converts lysyl and hydroxylysyl residues in the telopeptide regions to allysine and hydroxyallysine, which are aldehydic in nature. Immature cross-links that are reducible by sodium borohydride are formed by the condensation of the aldehydes with other lysyl and hydroxylysyl residues within the helical region of a neighboring molecule. With time, these crosslinks become more insoluble by condensation with histidinyl residues or its aldehydic derivatives. Through this time-dependent process, as reviewed (126), up to four 

TABLE 1 Collagen

chains can ultimately become cross-linked, leading to great stability and insolubility of the collagenous scaffolding (Fig. 5).

G. Collagen Types A broad range in the collagen pattern and molar mixture is displayed from one tissue to another and certain types are concentrated in specific tissues [reviewed in 43,44]. In comparison, bone matrix proper contains a rather limited array of collagen types (Table 1), which will be discussed in detail in this chapter. Based on their structural features, collagens can be roughly divided into two groups: fibrillar and nonfibrillar. Fibrillar collagens (types I, II, III, and V) are by far the most abundant forms and are found in the interstitial spaces of connective tissues throughout the body. Type I collagen, the predominant collagen of skin, tendon, and bone, is composed of two 1(I) and one 2(1) chains and forms the major scaffolding of virtually all the connective tissues (with the exception of cartilage). Cartilage contains predominantly type II collagen ([1(II)]3) with limited amounts of other collagen types as described later. Type III, composed of three identical 1(III) chains, and type V, composed of a combination of 1(V), 2(V), and 3(V) chains, are often codistributed with type I. Fetal tissue contains proportionally more type III collagen, which has been reported to coat the surfaces of type I fibrils or to form thin reticulin-like fibrils. Type V collagen is often pericellular and is also enriched in particular tissues such as smooth muscle and blood vessels. The nonfibrillar collagens (types IV, VI, VII, VIII, IX, X, and XI) are characterized by triple helical domains that are either shorter or longer than those of the fibrillar types and may contain stretches of noncollagenous sequences. Type IV collagen, with the composition [1(IV)22(IV)], is

Collagen Types Found in Bone Matrix Location/function

Molecular structure

Type I:[1(1)2 (1)] and [1(1)3]

Constitutes 90% of matrix in the bone matrix. Acts as scaffolding and binds to other proteins that initiate hydroxyapatite deposition

67-nm banded fibrils

Type III:[1(III)3]

Present only in trace amounts and can regulate collagen fiber thickness

67-nm banded, coats type I fibrils

Type :[1(V)22(V)] and [1(V)2(V)

Their absence can result in collagen fibrils of larger diameter

67-nm banded, coats type I fibrils in some tissues

Type X:[1(X)3]

Present in hypertrophic cartilage and can be involved in matrix organization via formation of the template for type I collagen

Probably fishnet-like lattice

116

GOKHALE, BOSKEY, AND ROBEY

found in basement membranes, including those that surround vascular endothelial cells that invade bone during osteogenesis. Type VI collagen is significantly shorter than other collagen types and is composed of three distinct  chains. These molecules form rope-like microfibrillar structures. Anchoring fibrils are composed of type VII collagen, which is 1.5 times longer than type I. Another short chain collagen, type VIII, is found in Descemet’s membrane of the eye, is synthesized by endothelial cells in culture, and may be related to type X collagen found in calcified cartilage. The localization of type X to hypertrophic chondrocytes is highly specific, but it does not appear to have a major role in cartilage calcification. Type IX and type XI are homologous to type V and are minor constituents in cartilage. Type IX belongs to the socalled FACIT class (fibril associated collagen with interrupted triplex). Structures of the isolated FACIT collagenous proteins, as distinct from those of the fibrillar collagens, include nontriple helical domains, as predicted from the non-GLY-X-Y repeats. Type IX is composed of three different types of  chains, 1(IX), 2(IX), and 3(IX), that form a short and a long triple helix joined by a flexible hinge region. A glycosaminoglycan chain is also attached to one of the  chains at the amino terminus, making this collagen a proteoglycan as well. Type IX has been found covalently attached to and as a coating to type II collagen fibrils and covalently attached to it. Type XII is similar to type IX but has three projections extending from the triple helix. This type may also be associated with type I fibrils in tendon. Type XIV (as well as XII) is structurally related to the type IX collagen fibrils, which trim type II collagen in cartilage.

H. Bone Collagen(s) Although bone matrix proper has been reported to contain only type I collagen, other types are certainly present but not at the levels found in soft connective tissues. Several FACITs have been detected in bone [127,128], and there are occasional reports of low levels of type III and type V [129,130] molecules as well. The FACIT collagens found associated with type I collagen in other tissues are types XII and XIV. Given the potential role of these low abundance collagens in regulating fibril diameter, it is possible that the collagen fibrils in bone grow to much larger diameters than in soft tissues due to the reduced proportion of these diameter-regulating types. However, it may be that these other collagen forms may originate from the vasculature that infiltrates bone to a great extent and may persist in the bone during chemical extraction prior to demineralization. Based on sequence analysis [131], type XII collagen is predicted to be a Mr 340,000 protein with repeating

domains consisting of regions homologous to the type III motif of fibronectin and a von Willebrand factor domain [132]. In addition, there is a type IX collagen-like domain, and several RGD cell-binding domains [132]. The presence of collagen, cell binding, and matrix-binding domains implies that these FACIT collagens may have novel functions. Type XIV collagen’s sequence indicates that it has a similar series of fibronectin and von Willebrand factor domains [132]. Type XII collagen is a homotrimer with two triple helical domains and a large (Mr 190,000) N-terminal globular domain. The helical domains appear to form arms stretching out from the globular domain [129], resulting in a cruciform appearance [133] with one thin and three thick arms. Like type IX collagen, type XII collagen made by fibroblasts, but not by all cells, contains CS  chains [133]. In bone, type XII collagen is seen in the periosteum [134] and is made by periosteal cells in culture [135]. Type XII-like collagens also trim the surface of type I fibrils [136], as does a type XIV variant [137]. The FACIT collagens seem to have a fundamental role in determining matrix structure, as demonstrated by animals lacking or containing mutated forms of the FACIT collagens [138]. These animals exhibit a spectrum of bone and cartilage disorders, presumably due to abnormal fibril formation. In bone, the proteoglycan decorin, as well as types XII and XIV collagens, may be important for regulating in vitro and perhaps in situ type I collagen fibrillogenesis. The observation that a type XII molecule trims the fibrils at specific sites, while proteoglycans are found at other sites along the collagen fibrils, supports this view. This will be discussed in more detail later. Type I collagen is found not only in bone, tendon, and dentin, but also in the sclera of the eyes, the lungs, the cornea, and the skin. These type I collagens are not identical but differ in the extent of all the posttranslational modifications [102]. The predominant form of glycosylation in bone is galactosyl-hydroxylylysine, whereas in soft tissue it is glucosyl-galactosyl hydroxylysine [139,140]. In bone, the cross-linking pattern is also different and originates from the hydroxyallysine pathway, resulting in the formation of a lysyl-pyridinoline cross-link, as opposed to soft connective tissue where cross-links originate predominantly from the allysine pathway leading to hydroyxlysylpyridinoline [141]. This modification is now the basis of a clinical assay that can measure specifically the degradation of type I collagen from bone [126]. The altered crosslinking pattern in bone has been speculated to be due to the deposition of mineral within collagen fibrils [101]. These altered cross-links may be important for the mechanical properties of the tissue. In all connective tissues, the collagens serve mechanical functions, providing elasticity and structure for the component tissues. Although type I collagen is widely distributed,

117

CHAPTER 4 The Biochemistry of Bone

the tissue most affected when type I collagen metabolism is abnormal is bone. In bone, extensive data indicate a link between type 1 collagen and bone strength [142]. Insight into these functions comes from detailed analysis of human and other animal tissues in which collagen synthesis is altered. Classic examples are the various forms of osteogenesis imperfecta (brittle bone disease) in which bone fragility has been associated with spontaneous mutations in the type I collagen genes [99,143 – 147] and the Mov-13 mouse in which a viral insertion within the first intron totally silences the 1(I) gene [148]. The remaining 2(I) chains are unable to form a stable triple helical molecule [although 1(I) can form homotrimers]. These mice die during gestation, indicating a critical role for type I collagen during development [149,150]. Transgenic mice in which other mutations are introduced [151,152] or in which the entire 2(I) chain is ablated [153] provide animal models for various types of osteogenesis imperfecta (OI) [154]. The bone changes in such animal models of OI show features similar to those seen in humans. Specifically, when glycine is substituted by aspartate or cysteine at positions close to the C terminus, the disease mimics the perinatal lethal phenotype [155]. Similar substitutions midmolecule generally result in a less severe phenotype [145,146,154] characterized by the presence of thinner than normal collagen fibrils and active osteoblasts containing numerous collagen fibrils. When the entire 2(I) chain is absent, animals that make an 1(I) trimer show even fewer bone abnormalities [153], similar to reported cases in humans [156]. However, there are many exceptions to this general rule of mutation location within the molecule as being predictive of severity of the phenotype. The mineral crystals in the bones of patients and transgenic animals with OI tend to be smaller than those in agematched control bones [157,158]. In the OI mouse (oim) that lacks the 2(I) chain [153], tendon [159] and bone 160 mineralization is aberrant. In the oim tendon, the crystals occasionally appear outside the collagen matrix, a feature never noted when collagen production is normal [161]. Similarly, in oim bones, the pattern of initial mineral deposition and crystal growth along the collagen differs from normal: the crystals appear outside the collagen matrix and with regions of collagen that are less mineralized than those in normal controls [158]. In addition, the presence of thinner fibrils in OI patients may be insufficient to provide nucleation and scaffolding sites, which can potentially translate into fragile bones [162]. It is not known at this time whether mineral seen away from the collagen fibrils was formed in the absence of a collagen backbone or whether it “broke away’’ and was later seen out in the matrix because the collagen structure was not sufficient to support it. Collagen per se does not cause mineral deposition; i.e., it is not a mineral nucleator, as it lacks the appropriate conformation that matches the ion surface of the deposited mineral surface [163]. Nonetheless, data from OI tissues clearly

demonstrate the importance of collagen for providing a scaffold to organize the mineral. As discussed later, noncollagenous matrix proteins appear to initiate and regulate mineral deposition [reviewed in 164,165]. Thus, an additional function of type I collagen is to provide a site for the retention of noncollagenous proteins, some of which appear to be covalently bound [166] whereas others are more loosely associated through specific collagen-binding domains. In patients with OI, decreases in some matrix proteins [167,168] may be due to the deficit of collagen to which they bind or to absolute decreases in protein production [169,170] associated with “protein suicide” resulting from the destruction of abnormal collagen within the cell [99]. Decreased type I collagen production as seen in these OI models demonstrates the template-like and mechanical functions of collagen. It is likely that animals null for 1(I) are not compatible with life because of the role type I collagen plays in the development of lung, blood vessels, and mesenchyme. However, overexpression of type I collagen also provides insight into collagen function. While finding an increase in type I collagen production in bone is relatively rare, it has been reported that MAV.2-0 (myeloblastosis-associated retrovirus) causes an osteoblastic hyperplasia and a relative increase in cortical bone thickness [171], providing another animal model for the characterization of collagen function.

III. GLYCOSAMINOGLYCANCONTAINING PROTEINS Proteoglycans are a class of macromolecules characterized by the covalent attachment of long  chains of repeating disaccharides that are often sulfated, termed glycosaminoglycans (GAGs). The different types of GAGs, chondroitin sulfate (CS), dermatan sulfate (DS), keratan sulfate (KS), heparan sulfate (HS), and hyaluronan (HA, which is unsulfated), are named based on the sugar composition of the repeating disaccharides (Fig. 6). GAGs, with the exception of HA, are synthesized in the Golgi apparatus by the sequential addition of sugar residues to a sugar transporter molecule, dolichol phosphate. Certain sugar residues become sulfated through the action of sulfotransferases; however, the level of sulfation can be extremely variable even within one GAG chain. Ultimately, the growing GAG is transferred from the dolichol phosphate to an acceptor site on a protein core. These sites are most frequently specific seryl or threonyl residues, but in the case of KS, the acceptor site is an asparaginyl residue. There are many different families of core proteins but the factors that dictate which residues become gagosylated and by what type of GAG are not well understood. In contrast to the other GAG chains, HA has not been found covalently attached to a core

118

GOKHALE, BOSKEY, AND ROBEY

FIGURE 6 Disaccharide composition of glycosaminoglycans (GAGs). The GAG side chains that are covalently attached to proteoglycan core proteins are composed of repeating disaccharide units. The composition of the disaccharides, along with modifications by acetylation, results in the formation of chondroitin sulfate, which is epimerized to form dermatan sulfate, heparan sulfate, and keratan sulfate. Hyaluronan is the sole GAG that remains unsulfated and is not covalently linked to core proteins.

protein. HA does associate with certain cartilage proteoglycans by a noncovalent linkage. Proteoglycans can be glycosylated by the addition of N- and O-linked oligosaccharides. In addition to sulfation of the GAG  chains, the oligosaccharides can also be phosphorylated and/or sulfated, generating a class of macromolecules with multiple posttranslational modifications.

A. Aggrecan During endochondral bone formation, cartilage first hypertrophies and forms a temporary mineralized tissue, calcified cartilage, which is then invaded by blood vessels. The invading vasculature brings with it (1) cells that destroy the calcified matrix (osteoclasts or a related cell type) and (2) osteoprogenitor cells. Because cartilage macromolecules can be in close proximity to forming bone and may actually be incorporated into the initial bony tissue, it is im-

portant to describe its extracellular matrix constituents. It is also not clear whether all of the large CS-proteoglycans isolated from bones are remnants of those in calcified cartilage or specific bone products. The presence of elevated amounts of CS-proteoglycans in the bones of osteopetrotic animals with defective osteoclasts was linked to the inability of these animals to resorb calcified cartilage [172]. The basic scaffolding upon which cartilage matrix is built is type II collagen. In addition, a number of proteoglycans, including (i) a large CS molecule, termed aggrecan reviewed in 173, have the ability to form aggregates with HA and (ii) two small proteoglycans, decorin and biglycan [reviewed in 174 – 177]. Other proteins, including COMP, CD-RAP, chondroadherin, and matrilin-1, are present, but at lower levels [178,179]. Intact aggrecan has a Mr 2.5 million kDa, with a core protein ranging in apparent Mr between 180,000 and 370,000 with just over 100 GAG chains (mostly CS, but with some KS) of about Mr 25,000 (Fig. 7). Based on enzymatic cleavage and sequence homology, five

119

CHAPTER 4 The Biochemistry of Bone

FIGURE 7

A representation of the chemical features of the large hyaluronic acid-binding proteoglycan, aggrecan. GAG, glycosaminoglycan; CS, chondroitin sulfate; KS, keratan sulfate; G1, G2, G3, globular domains (see text for description); EGF, epidermal growth factor; CRP, C-reactive protein.

domains have been defined in the core protein: three globular domains, two of which bind to HA (G1 and G2), an interglobular domain, a domain rich in serine-glycine repeats to which the CS and KS glycosaminoglycan  chains are attached, and the third globular domain, G3, at the C terminus. Each of the globular domains are cysteine rich and stabilized by disulfide bonds [180]. The G1 domain in the N terminus is structurally homologous [181] to “link protein” [reviewed in 182], which stabilizes the interaction between the proteoglycan and HA in cartilage [183]. The adjacent G2 domain provides a flexible hinge, whereas the more linear central domain is the site of KS and CS covalent attachment. The C-terminal G3 domain is also predicted from sequence analysis to be globular and contains a set of epidermal growth factor (EGF)-like and complement regulatory protein (CRP)-like sequences at the carboxy terminus [184,185]. The individual GAG chains form extended flexible structures, whereas the serines in the central domains have -D-xylose attachments with restricted orientation [186]. Whereas NMR analysis of isolated CS chains has provided insight into their conformation in different solutions [187], their actual conformation when attached to the core protein is not known. The human aggrecan gene is located on chromosome 15q26 [188]. However, the complete genomic sequence has been reported only in the rat and is 63 kb in length. There are 18 exons with 30 kb of intronic sequence separating the first and the second exon. The intermediate exons roughly encode for each of the structural domains of the molecule with the exception of G1-B (which contains the hyaluronicbinding domain) and G2-B (which contains link protein-

like sequences), which are split between two exons, and G3, the lectin-binding domain, which is coded for by three exons. The splice junctions are primarily symmetrical phase 1 type (in frame) [189,190]. The rat gene promoter lacks a TATA box, and the major transcription start site is located in close proximity with a number of SP1 sites. In addition, there are four AP2 sites located 120 kb upstream in a GC-rich region, and two of the SP1 sites overlap. A GC-rich region is also found in the first exon, which also has four AP2 sites. A potential NF-b (nuclear factor) site is located between 123 and 103 bp, and another AP2 site is located between 81 and 37 bp [190]. The resulting mRNA species of 8.2 and 8.9 kb predict a 2124 amino acid residue protein, including a 19 residue signal peptide. A stretch of 1164 residues contain Ser-Gly repeats, the CS attachment site [189,190]. Structures of the core proteins, individual domains, and segments of these domains have been determined based on NMR and molecular modeling [186,191] and neutron and X-ray scattering of molecules in solution [192]. These studies reveal that the core protein has a fairly linear structure. The protein core of the aggrecan-like proteoglycans (CS/KS-containing) is fairly homologous in a wide variety of tissues, ranging from tadpole tails to human articular cartilage [118]. The structures of isolated individual large aggregating proteoglycans (aggrecan) from hyaline cartilage and the aggregates which they form have been visualized at the electron microscopic (EM) level by rotary shadowing [193,194]. While EM studies of the cartilage molecule support a bottle brush-like conformation predicted earlier from physical and chemical analysis [183],

120 NMR and light-scattering studies of these molecules in solution indicate that they have a more compact form [191,192]. Whether the aggrecan found in bone is there as a bone product and not as a cartilage remnant has yet to be determined. The function of these large proteoglycans in bone is also unknown. In other tissues, the CS-proteoglycans serve a hydrodynamic function, aiding in the retention of both water and cations, and the exclusion of anions [195]. The large proteoglycans form gels, which can both swell and retain water, contributing to the osmotic pressure of the tissue [196]. These proteoglycans in cartilage are also responsible for matrix organization [185], in part through interactions with the glycosaminoglycan chain of type IX collagen that trims the type II collagen fibrils [197]. Each of these properties, as reviewed elsewhere [198], depend on the arrangement, size, and number of the constituent CS chains. In cartilage, proteoglycans are believed to play a major role in the regulation of calcification [for review see 199]. The large aggregating cartilage proteoglycans, in solution can inhibit hydroxyapatite formation and growth [200 – 203]. They can also chelate calcium [204] and serve as a source of calcium ions for mineralization if they are degraded into non-Ca2-binding fragments. Although there is some debate as to whether this chelation is involved in the inhibition of mineralization, it is clear that proteoglycans and their component GAGs stearically block hydroxyapatite formation and growth [201]. It has also been shown that when proteoglycans are degraded, their inhibitory ability is lost or diminished both in vivo [205] and in vitro [206]. However, because the relative amount of large CS-proteoglycans in bone is low, it seems unlikely that they have a critical role in preventing osteoid mineralization. Unfortunately, this cannot be proven based on existing data because none of the mutant animals in which the large CS-proteoglycan is altered in cartilage were reported to show bone abnormalities. This is the case for the cartilage matrix deficiency (cmd) mouse, which lacks a core protein [207], and the brachymorphic mouse (bm), which has a deficiency of the enzyme required for the sulfation of glycosaminoglycan side chains [208]. Only in the case of the bm was the mineral examined, and no abnormalities in the bone were detected by X-ray diffraction, infrared, or EM, although cartilage calcification was altered [208]. In the nanomelic chick, which contains a mutated core protein [209], studies show that although there is no overt bone defect, certain bones are abnormally shaped, perhaps due to the altered distribution of load applied to cartilage-deficient tissues during development. The likely functions for the large CS proteoglycans in cartilage are maintenance of tissue hydration and the regulation of mechanical [210] and hormonal signal transduction. However, it is not known if aggrecan or a related molecule provides these functions in bone.

GOKHALE, BOSKEY, AND ROBEY

B. Versican Another large CS proteoglycan related to aggrecan is found in cartilage at low levels. It has been termed versican, as it is found in variable forms in a large number of extracellular matrices. The protein core, with a Mr 360,000, has a structure similar to that of aggrecan with the exception that it lacks the G2 domain. In addition, versican contains only 12 – 15 CS side chains (Mr 45,000) in contrast to 100 in aggrecan [184] (Fig. 8). The distribution of versicans is somewhat ubiquitous as it is found in smooth muscle and aorta [211], glomeruli [212], brain [213,214], and epidermis [215]. A versican-like molecule is highly enriched in developing mesenchyme destined to become bone and may serve to capture space [216]. Versicans from soft connective tissues are capable of aggregating with HA [185]. In addition, it is now apparent that PG-M, identified as a constituent of the matrix surrounding condensing limb bud mesenchyme, is a splice variant of versican [217]. However, versican isolated from bone cell cultures apparently does not aggregate with HA (Fedarko and Gehron Robey, unpublished results). The versican gene localizes to human chromosome 5q12 – q14 [218,219]. The human gene has been isolated [220], is over 90 kb in length, and is composed of 15 exons with a splice variant that utilizes an additional exon [221]. The sequence predicts a 20 residue signal peptide and a 2389 residue mature protein [222]. Exon 1 codes for the signal peptide and a few amino acids found in the mature molecule, and the HA-binding region (G1) is in exons 3 – 6. These exons share homology with the other HA-binding protein, the link protein. This region also contains an Ig-like protein conformation whose function is unknown. Subsequently, exons 6 and 7 are differentially utilized and contain GAG attachment sites. The carboxy-terminal domain (G3), which contains homology to selectins, EGF, and CRP, is contained within exons 9 – 14. The 3-untranslated region contains three different potential polyadenylation sites [220]. The promoter region has a TATA box located 16 bp upstream from the transcription start site. Transfection analysis indicates that there is a positive enhancer between 209 and 445 bp and a negative element between 445 and 632. Other elements present include an XRE (xenobiotic responsive element that may downregulate P450 levels), SP1-binding sites, CRE (cyclic AMP responsive element), and a CCAAT transcription factor-binding site. Based on differential splicing and polyadenylation, three mRNA species of 10, 9, and 8 kb are produced [217]. Versican expression in other tissues can be modulated by a number of factors, including transforming growth factor (TGF-) [223,224], PDGF [223], and interleukin-1 [225]. There is little information on the modulation of its expression during bone development, but it appears to be

121

CHAPTER 4 The Biochemistry of Bone

FIGURE 8

A representation of the chemical features of the widely distributed proteoglycan that is related to, but not identical with, aggrecan. CS, chondroitin sulfate; G1 and G3, globular domains (see text for description); EGF, epidermal growth factor; CRP, C-reactive protein.

upregulated by TGF- in adult human bone cells and fetal bovine long bone cells (Heegaard and Gehron Robey, unpublished data). A potential function for this molecule is to serve as a bridge between the extracellular environment and the cell by binding to HA via the amino-terminal-binding region and to molecules that have yet to be identified on the cell via the carboxy-terminal domain [184]. Versican may also be involved in cell motility and growth [226]. Proteoglycans are known to interact with growth factors, providing a storage site, or facilitating interactions between the extracellular matrix and the cell. Some structural elements within the molecule may also function directly as signals for cellular growth or differentiation [227]. For example, versican contains EGF-like sequences such that upon its destruction, it may serve to stimulate proliferation of osteoprogenitors. EGF has been reported to stimulate proliferation of osteoblastic cells in vitro [228].

C. Small Leucine-Rich Repeat Progteoglycans Another family of proteoglycans is represented by a group whose protein core is characterized by a smaller size and a leucine-rich repeat sequence that is approximately 24 amino acids in length [reviewed in 176,229]. These small proteoglycans are broadly distributed and are found in the

extracellular matrices of many tissues [230]. In cartilage and bone, there are several members of this family, including decorin and biglycan. While they are highly homologous, they exhibit distinctly different patterns of expression and tissue localization, indicative of divergent functions within these tissues. It has been reported that proteoglycans must be removed prior to matrix mineralization and that they inhibit hydroxyapatite crystal growth in test tube experiments, although these two statements do not really comment on the role of a particular species of proteoglycan during matrix mineralization in vivo. It is apparent that there is a net loss of sulfate during matrix mineralization [231,232]. However, using animals and organ cultures, it has been found that a species of sulfate-labeled material accumulates rapidly at the mineralization front [233,234]. One possibility is that the versicanlike molecule is being removed and replaced by the smaller proteoglycans found in bone [235]. Considering the fact that versican has more GAGs per protein core than the small proteoglycans, this would account for the net decrease in sulfate content; however, these data do not rule out the presence of small proteoglycans during matrix mineralization. The leucine-rich core proteins of decorin and biglycan are quite similar in sequence to that of the ribonuclease inhibitor protein. Their structure (in the absence of Ca2 ions) has been determined using X-ray crystallography to 2.3 Å resolution. The ribonuclease inhibitor protein structure

122 consists of 15 leucine-rich repeats, alternatively 28 and 29 residues long, each forming right-handed -- structural motifs resulting in a horseshoe shape as opposed to a globular structure, with the parallel  sheets in the horseshoe circumference [236]. The -- motif is frequently seen in proteins as a way of connecting two antiparallel  strands; however, it can also be used to connect parallel  strands, resulting in a more open structure [237]. This interesting sequence has also been identified in two morphogenetic proteins in Drosophila, chaoptin and toll, in von Willebrand factor-binding protein (GP 1) and in the leucine-rich protein in plasma [238], and in a number of other proteoglycans, which have formed a family, SLRPs, some of which are also found in bone. 1. DECORIN (PG-II, PG-40) Decorin, so named for its ability to bind to and decorate collagen fibrils [118,119,239,240], has also been termed PG-II and PG-40. In soft connective tissues and bone cell cultures, decorin (and biglycan) has DS side chains, whereas in bone, decorin (and biglycan) bears CS [reviewed in 176,241,242]. This difference is of interest because DS is formed from CS by the action of a specific epimerase, and this epimerization can occur uniformly or focally within the GAG chain. The factors that regulate whether a proteoglycan will bear CS as opposed to DS are not known. As one would expect from its proposed function of binding to collagen fibrils, decorin is fairly widely distributed and is found virtually coincident with type I collagen, although the timing of its appearance may be somewhat different. Histochemical studies show the presence of proteoglycans with low molecular weight in the d and e bands of type I collagen fibrils, which disappeared when mineralization occurred [116,117,239]. These histochemical data first suggested that a small proteoglycan might also play a role in mineralization. In the developing skeleton, it decorin is also distributed more uniformly than in the mature animal. In cartilage, decorin is present in very low levels and is restricted to the interterritorial matrix [243]. As bone is formed, it is produced by preosteoblasts and osteoblasts, but its synthesis is not maintained by osteocytes [243]. Nonglycanated forms of decorin (and biglycan) have been found in the intervertebral disk [244]. Decorin has a core protein of approximately Mr 38,000, which includes 10 of the leucine-rich repeat sequences. Although there are three potential GAG attachment sites, only one is utilized with the attachment of a single GAG chain of Mr 40,000, resulting in a molecule with an apparent Mr 130,000 as determined by sodium dodecyl sulfate –polyacrylamide gel electrophoresis (SDS-PAGE) [245] (Fig. 9). The decorin molecule was predicted to contain loops of  strands [246]. It has been demonstrated by far-UV CD spectroscopy that both decorin and biglycan exist as predominantly

GOKHALE, BOSKEY, AND ROBEY

 sheets, and biglycan has a significantly higher  helical structure and assumes different structures in the solution 247. The human gene for decorin has been localized to 12q23 [248 – 250]. In mouse, it is located on chromosome 10, just proximal to the steel locus [251]. A restriction fragment length polymorphism (RFLP ) is also present in humans [252]. The gene is over 38 kb in length and contains 9 exons. The gene shares 55% homology to and is organized in a similar fashion to the biglycan gene (described later) except that the intronic sequences are much longer (two of which are 5.4 kb and greater than 13.2 kb) [249]. The leucine-rich repeat sequences are not organized specifically within the exons. Although the 5 end of the gene eluded characterization for some time, it is now evident that there are two alternatively spliced leader exons (exons 1a and 1b) [249,250]. Exon 1a contains two GC-rich sequences, whereas 1b contains two TATA boxes and one CAAT box that are in close proximity to the transcriptional start site [253]. In addition, AP1, AP5, and NF-b sites were also identified along with a homopurine/homopyrimidine mirror repeat sequence. This region has been postulated to take on a hairpin triplex structure that is unique to decorin compared to the promoters of the other SLRPs. Gene transcription results in a major mRNA species of 1.6 kb and a minor species of 1.9 kb [254,255]. The sequence predicts a 359 residue protein that includes a 30 residue prepro peptide. The synthesis of decorin can be modulated by TGF- [256 – 258]. Dexamethasone upregulates decorin, but downregulates biglycan [259]. The phytoestrogen ipriflavonemetabolite III has been shown to upregulate decorin [260]. Mechanical loading appears to stimulate the synthesis of decorin, but not of biglycan in cartilage [261]. While it would appear that the propeptide is cleaved from the mature decorin in bone, evidence shows that it is maintained in other tissues such as cartilage [262]. Decorin has been shown to bind to and regulate the fibrillogenesis of type I, type II, and type VI collagens [263,264]. It also has a high affinity for thrombospondin [265], TGF- [266], other growth factors, and fibronectin [267,268]. In bone, the proposed functions of decorin are the regulation of collagen fibril diameter and fibril orientation, and possibly the prevention of premature osteoid calcification. However, it is not clear if decorin within the tissue is actually inhibitory to matrix mineralization. Decorin isolated from skin was initially shown to regulate the rate of collagen fibrillogenesis in vitro [269]. Whether bone decorin has the same effect has not yet been demonstrated. Interestingly, bone decorin must be chemically different from tendon decorin, as a different peptide map can be generated by V8 protease [270]. Bone collagen fibrils are a composite of type I with trace amounts of types III, V, and XI and are also trimmed by decorin. Therefore, it is difficult to determine which of these are essential to the

123

CHAPTER 4 The Biochemistry of Bone

FIGURE 9 The two most abundant proteoglycans present in bone matrix are the small chondroitin sulfate/dermatan sulfate proteoglycans, decorin and biglycan. The core protein of each is highly homologous to a number of proteins due to the presence of a leucine-rich repeat sequence. CS, chondroitin sulfate; DS, dermatan sulfate; C – C, disulfide bonding.

regulation of collagen fibrillogenesis and maturation in this tissue. Decorin has been reported to bind with high avidity to type VI collagen [264] and hence may be more concentrated in areas where type VI collagen is more abundant, e.g., in the upper zones of epiphyseal growth plate [271]. Decorin, in solution, has a high affinity for type I collagen (Kd  M10) [272] and, in contrast, has a low affinity for hydroxyapatite (Kd  13 g/ mol). When bound to a solid surface, decorin binds rather nonspecifically to apatite relative to other GAG-containing molecules (N, the number of binding sites  333) [273]. Further, it has no detectable direct effect on solution-mediated hydroxyapatite formation or growth [273]. With a low affinity for Ca2 (0.001 g/mg) [267] and a higher affinity for other divalent cations [275], the disappearance of decorin from the collagen fibrils in bone [117] indicates that it is unlikely that it has a direct role in mineralization. However, once removed, other possible nucleators may be exposed, which in turn would facilitate mineralization. Whether this removal also

affects collagen cross-linking and fibril spacing to facilitate mineralization, as well as unmasking nucleators localized under the decorin, remains to be determined, as does the mechanism responsible for the removal of decorin. It is interesting to note that collagen fibril diameter in bone is somewhat reduced in the decorin knockout mouse (M. Young, personal communication), and decorin expression is reduced in certain skin diseases characterized by excessive keratinization [276], stressing the physiologic importance of decorin in regulating fibril formation and collagen – matrix interactions. It is also possible that a key function of decorin is to serve as a binding site for growth factors in these tissues [277]. This may be relevant in the keratinization diseases, in the kidney [276], and in cartilage repair, as well as in the bone where there are other binding sites for these growth factors (see later). It is also of interest to note that there is a decreased expression of decorin in some patients, with OI [169,170,278]. Decorin has also been proposed to play a role in matrix organization by

124

GOKHALE, BOSKEY, AND ROBEY

FIGURE 10

Biglycan immunolocalization in newly formed trabecular bone. Biglycan was found predominantly in osteoid (Ost) and, to a lesser extent, in mineralized matrix (MM) in human bone. Osteocytic lacunae (OCy) also contain high levels of biglycan. Courtesy of Dr. Paolo Bianco.

binding to other matrix proteins, such as fibronectin [267,268]. Interestingly, it has been found that decorin (and biglycan) inhibits the attachment of bone cells to fibronectin, perhaps by blocking the fibronectin cell attachment site that mediates binding. Decorin also binds to thrombospondin, another matrix protein that is speculated to serve as a matrix organizer. It has been reported that decorin binds to TGF- and blocks its activity in fibroblasts [266]; however, it may also activate TGF- in bone cells [279]. 2. BIGLYCAN (PG-I, PG-SI) Biglycan, also known as PG-I and PG-S, is highly related to decorin and exhibits 55% homology at the protein level [238,280]. Despite its high homology to decorin, it is evident that biglycan is uniquely distributed and present in a temporal and spatial pattern not coincident with that of decorin, indicative of a distinctly different function. Biglycan has been localized to the pericellular environment of endothelial, epithelial, and muscle cells [243]. However, in skeletal tissues, biglycan does not appear to be associated with the collagen fibrils, although the biglycan mouse has highly abnormal collagen fibrils in boneskin (M. Young, personal communication). In contrast, it is more abundant in the growth plate [243] and is concentrated in the intraterritorial matrix and in preosteogenic cells, implying a role in early bone development. It is highly upregulated in differentiating bone cell cultures [281] and, interestingly, it is maintained (at the mRNA and protein levels) in osteocytic lacunae and in the canaliculae, which contain

the osteocytic cell processes (Fig. 10). This implies that biglycan may be important in osteocytic cell metabolism. In mineralizing osteoblast cultures, biglycan expression begins during the preosteoblastic phase of the culture and is commensurate with calcium uptake in contrast to decorin, which increases gradually and remains high as mineralization progressed [282]. Biglycan has a protein core of Mr 37,000 to which (in most forms) two GAG side chains are attached. The aminoterminal domain contains the GAG attachment sites, followed by 12 of the leucine-rich repeat sequences (Fig. 9). The first and the last repeat contain a characteristic pattern of cysteinyl residues that result in a particular pattern of intramolecular disulfide bonding [238,283]. The carboxy domain has a sequence that is unique to biglycan (and differs from decorin and other leucine-rich repeat sequence containing proteins). Biglycan has the propensity to self-aggregate in solution [284], although the physiological relevance of these aggregated structures is not known. It has also been determined that nonglycanated biglycan can be secreted into the matrix and that the level of the core protein alone increases with age in human articular cartilage [174]. Other than predicted structures from the amino acid sequences of the biglycan core protein [238], little is known about the tertiary and quaternary structure of biglycan. The gene for the biglycan core protein is localized to Xq27-ter in humans, the only matrix protein that is not on an autosomal chromosome [285]. The gene is 7 kb in size and codes for a 368 residue proform that is processed to form a

125

CHAPTER 4 The Biochemistry of Bone

mature core with 331 residues [238,285]. There are eight exons with relatively small introns compared to those in decorin. Exon 2 codes for the signal peptide, propeptide, and GAG attachment sites. The bulk of the core, the leucine-rich repeat sequences, are not obviously distributed between the remaining exons. The 3 portion of the gene is rich in CT and CA repeats [286] and may result in the Z DNA conformation. The promoter does not contain a TATA box or a CCAAT box, but has a number of cis-acting elements, including SP1, AP1, AP2, NF1, and NF-b-binding sites [285,287]. A number of in vitro transfection experiments have shown that the SP1 site is active, but that it may be binding to a factor other than those described previously [287]. The final mRNA species are 2.1 and 2.6 kb in size. A number of factors have been reported to regulate biglycan synthesis in bone cell cultures and often the pattern is distinctly different from those affecting decorin. TGF- increases biglycan (Gehron Robey, unpublished results) in normal human bone cells and in MC3T3 [258]. Other cell lines were not stimulated by TGF-, but IGF-I and IGF-II increased biglycan levels [288]. Retinoic acid suppresses biglycan in chondrocytes [289]. Dexamethasone and 1,25-dihydroxyvitamin D3 have been reported to decrease its expression in human bone and marrow cell cultures [290,291]. Fluoride, given to cultures of rat osteoblasts in clinically relevant concentrations, decreased GAG chain length and composition [292]. Biglycan (and decorin) binds to type V collagen [293], a collagen abundant in blood vessels. They both also inhibit the activity of antithrombin by interacting with heparin cofactor II. Thus, it has been suggested that one of the functions of biglycan is to create a thromboresistant vascular surface [293]. Whether this is the same for the chondroitin sulfate containing biglycan of bone remains to be determined. In solution, biglycan binds only low amounts of calcium (0.012 mM/ g) but appreciable amounts of small divalent cations such as zinc [294]. Its in vitro interaction with type I collagen can be blocked by increasing phosphate concentrations, implying that biglycan, as distinct from decorin, which is not affected by solution phosphate, interacts through its anionic residues [295]. Also, in solution, biglycan at low concentrations can promote apatite formation, whereas at higher concentrations it inhibits the growth and proliferation of mineral crystals [273]. These effects appear to be due to the highly specific high-affinity binding of biglycan for apatite (Kd 294 g/ mol). The relative extent of apatite formation induced by biglycan compared to other mineral nucleators and its absence from bone collagen fibrils suggest that its primary function may not be related to mineralization of bone. In calcifying cartilage, however, biglycan may play some role in the mineralization process. Biglycan appears to have a regulatory role in bone development. This suggestion is based on the observation that

patients with Turner’s syndrome (genotype 45, XO; i.e., females that are missing one or part of one X chromosome) have decreased biglycan levels, short stature, and other skeletal deformities, including early onset osteoporosis [296]. The biglycan knockout mouse also has a short stature, altered mechanical properties, and altered mineral distribution 297. In contrast, there is overexpression of biglycan in patients with Klinefelter’s disease (genotype 47, XXX) in which patients are excessively tall [296]. Biglycan, like decorin, also binds to a variety of growth factors in solution [266]; thus, it may have functions similar to those of decorin regarding the storage of these growth factors. Biglycan interacts with a number of extracellular components, including the cell-binding domain of fibronectin (thereby inhibiting attachment mediated by fibronectin) [268,298 – 300], TGF- [266], which may alter its activity, and heparin cofactor II [293]. Nonglycanated biglycan has been found to bind to colony-stimulating factor (CSF), which then stimulates the proliferation of nonadherent thymic and peritoneal exudate cells [301]. 3. FIBROMODULIN Fibromodulin is another SLRP that contains keratan sulfate found predominantly in articular cartilage, but also exists in bone [302,303]. The amount of fibromodulin correlates with the size of collagen fibrils in cartilage [304]. In developing bone induced by demineralized bone matrix, fibromodulin is heavily localized to fibrillar bundles [305]. Observations from the fibromodulin knockout mouse have indicated that in the absence of functional fibromodulin, collagen fibrils in tail tendon are abnormal. In these mice, fibrils are significantly thinner, indicating a role for fibromodulin in collagen fibrillogenesis 306. Similarly, a decrease in fibromodulin content and a alteration in structure have been shown to correlate with the aging-related degeneration of vertebral disks 307. The intact protein is approximately 59 kDa, and the protein core shares a great deal of homology with decorin and biglycan, but bears keratan sulfate GAG chains linked to asparaginyl residues rather than chondroitin or dermatan sulfate linked to seryl/threonyl residues [308]. There is also a considerable amount of tyrosine sulfation at the aminoterminal portion of the molecule. The human gene, which is at least 8.5 kb, has been isolated and partially characterized. It has an intron – exon organization, which differs markedly from that of decorin and biglycan. The first exon contains the untranslated region, and most of the coding sequence is located in the second exon, with the remainder residing in the third exon [308]. Fibromodulin interacts with triple helical types I and II collagen [309] and is capable of binding TGF- [310]. Decorin and fibromodulin are the most active collagenbinding proteins, which bind to distinctly different regions on collagen fibrils [309].

126

GOKHALE, BOSKEY, AND ROBEY

D. Heparan Sulfate Proteoglycans Although heparan sulfate proteoglycans are not usually found within the extracellular matrix of bone, they are cell surface associated, either by phosphoinositol linkages that are cleavable by phospholipase C or by transmembrane insertions. There are at least two heparan sulfate proteoglycans present on cells in the osteoblastic lineage, and most likely there are more. The most predominant type is similar to the syndecan family, first identified on mammary epithelial cells [311]. The intact syndecan molecule is Mr 400,000 and contains a core protein of Mr 80,000 and several heparan sulfate side chains of Mr 60,000 [312]. Glycan, a helper receptor for the type I and type II TGF- receptors, has also been identified in most connective tissue cells [313,314]. It has also been determined that bFGF binding to its receptor is facilitated by heparan sulfate on the cell surface [315], although the identity of this molecule has not been determined in bone cells.

E. Other SRLPs, Proteoglycans, and LeucineRich Repeat Proteins The structures and functions of the less abundant proteoglycans have been reviewed by Hardingham and Fosang [316]. In addition to decorin and biglycan, which are class 1 SRLPs [229], a third small proteoglycan whose core protein structure has not been described is the hydroxyapatite (HA)-binding proteoglycan, HA-PGIII, isolated from porcine bone by Nagata et al. [317]. An analogous hydroxyapatite binding protein from bovine bone has been isolated by Hashimoto et al. [318] and was shown to be covalently cross-linked to type I collagen. The localization of this protein in bone has not been described nor are there data on its function, its ability to regulate fibrillogenesis, or bind growth factors. A protein that contains the leucine-rich repeat sequence was originally identified as osteoinductive factor (OIF) [319,320]. However, the osteoinduction component of this preparation is in fact TGF-, and the other protein has now been renamed osteoglycin. It is not known whether this protein exists as a proteoglycan. However, it is of interest that all of the three proteins found in bone with the leucinerich repeat sequences have the ability to bind to TGF-. Another proteoglycan, PG-Lb, also known as epiphycan, has been isolated from calcifying cartilage and is very homologous to osteoglycin [321 – 324]. PG-100 was found in fibroblasts and preosteoblasts [325], and HA-PGII has been found to be associated with forming mineral crystals [317]. Given the low abundance and the variability of expression between one animal species and another, it is not clear what role they play in bone formation and/or mineralization. Until details of the structures and distributions of these

proteins in bone are determined, it is difficult to speculate on their functions. However, by analogy with the other small proteoglycans, it is likely that they will play some role in regulating matrix structure, collagen fibril diameter, and interaction with other matrix molecules.

F. Hyaluronan Hyaluronan is a ubiquitous component of connective tissue matrices [326] consisting of repeating sequences of glucuronic acid and N-acetylglucosamine linked by 1-3 and 1-4 glycosidic linkages. Chains may be several thousand residues in length, placing hyaluronan among the highest molecular weight glycosaminoglycans. Comper and Laurent [195] described the structure of hyaluronan based on models of fluid flow in the 1970s, showing that the hyaluronan formed space-filling gels. In mature bone, hyaluronan is only a small percentage ( 5% on a weight basis) of the total glycosaminoglycans present [327]. The properties of hyaluronan have been reviewed by Knudson and Knudson [328]. Unlike other GAGs, the synthesis of hyaluronan occurs outside of the cell by a large complex of sugar transferases that have yet to be fully characterized. Consequently, the synthesis of hyaluronan is most likely regulated by factors that are quite distinct from those that regulate other GAGs. Hyaluronan is produced at least during early phases of osteogenesis in vitro; however, very little is known about its metabolism in bone. Large amounts are produced by bone cells in culture [329]. To date, there have been no reports on the modulation of hyaluronan production by osteoblastic cells in response to growth factors and hormones that are known to affect bone. This paucity of information stems from the fact that there is no available antibody that binds to it specifically, and because it is synthesized by a multienzyme complex, there is no single probe for measuring mRNA. However, there is an indirect histochemical assay that relies on a peroxidase-conjugated link protein that will bind to hyaluronan in a specific fashion [330], and methods for blocking hyaluronan binding to the hyaluronan receptor (CD44) are available [331]. In order to understand the role of hyaluronan in bone formation, it would be of interest to apply these techniques to bone Hyaluronan binds weakly and nonspecifically to apatite, with affinity constants varying as a function of hyaluronan chain length [332,333]. There is some discrepancy in reports of the effects of hyaluronan on apatite formation and growth. However, these discrepancies may be related to the different systems used for the studies. Thus, where only pH was kept constant, hyaluronan of molecular weight 104 was shown to have no effect on apatite growth at low Ca  P products [333]. With higher Ca  P products maintained at constant composition, hyaluronan of molecular weight 108

127

CHAPTER 4 The Biochemistry of Bone

inhibited apatite growth slightly [332]. This difference is not attributable to hydrodynamic or size effects, as it has been shown that smaller molecular weight hyaluronans have a similar action [334]. Because the inhibition was small relative to other macromolecules studied, one may conclude that hyaluronan in bone is not apt to have a role in the control of mineralization. The structure and regulation of hyaluronan-binding proteins have been reviewed 335. Hyaluronan forms stable complexes with the large aggregating proteoglycans (aggrecan and versican) of cartilage [336] and with collagen-binding heparan sulfates [337]. The interaction of hyaluronan with the heparan sulfate of cell membranes is believed to be one of the ties that hold cells in place in a variety of tissues and this may also be the case in bone. Hyaluronan has been reported to play a role in cell migration and cell adhesion and invasion, as well as differentiation and proliferation of connective tissue cells [338,339]. An additional function of hyaluronan is the maintenance of tissue hydration, a function that it fulfills in many connective tissues. Removal of this molecule by the enzyme hyaluronidase results in tissue desiccation and facilitates tissue mineralization in cartilage [340]. Because of its abundance in early development, hyaluronan is thought to play a role in embryogenesis. Hyaluronan has been shown to facilitate mesenchymal cell movement [341] and cell adhesion [342] and it appears to bind via CD44 as the receptor [343]. It has been suggested that a major function of hyaluronan in mesenchyme-derived tissues is to act as a space filler, holding water and thereby increasing tissue volume [344]. In chondrocyte cultures, Knudson and Toole [345] have shown that hyaluronan promotes cell proliferation, probably because it provides a space/volume where cells can be anchored and provided with nutrition. Studies also show that hyaluronan can alter the extent of collagen production and proliferation of fibroblasts in culture [346]. Hyaluronan may also function as an organizer of the bone marrow [347] and may regulate hematopoiesis that is supported by steroid [348]. Serum hyaluronan concentrations are elevated in rheumatoid arthritis, osteoarthritis, liver cirrhosis, Werner syndrome, renal failure, psoriasis, and various malignancies; consequently, it may be a useful marker for measuring disease activity [349]. It is also of interest that intercellular adhesion molecule-1 (ICAM-1), a cell surface receptor for hyaluronan found on endothelial cells, is upregulated in inflamed tissue [350]. Consequently, hyaluronan may be involved in this tissue reaction.

IV. GLYCOPROTEINS This class of proteins is characterized by the covalent linkage of sugar moieties attached via asparaginyl or seryl residues as described earlier. Collagen also contains another

form of glycosylation (galactosyl and glucosyl-galactosylhydroxylysine), which is virtually specific to collagen. In bone, the most relevant and abundant glycoproteins are represented by alkaline phosphatase, osteonectin, and the cell attachment proteins, which include, but are not limited to, sialoproteins. These glycoproteins may also be further modified by posttranslational sulfation and phosphorylation.

A. Alkaline Phosphatase Alkaline phosphatase is not typically thought of as a matrix protein. However, studies indicate that under normal conditions, alkaline phosphatase is shed by cells of the osteoblastic lineage in culture as they pass through the G2/M phase of the cell cycle [351], and a Ca2-binding glycoprotein with alkaline phosphatase activity has been isolated from mineralized matrix extracts [352]. It is possible that this protein is indeed alkaline phosphatase that has been shed from the cell surface by cleavage of its phosphoinositol type linkage with phospholipase C or in a membranebound form (matrix vesicles). Although this enzymic activity is shared by many types of tissues, there is no doubt that the induction of alkaline phosphatase activity in uncommitted progenitors marks the entry of a cell into the osteoblastic lineage and is a hallmark in bone formation. However, it should be noted that many studies utilize the enzymatic activity alone to measure the presence or absence of alkaline phosphatase. This may be somewhat misleading, as it is possible that the protein can be present but inactive or, conversely, that the protein can be partially degraded and still maintain enzymatic activity. As in all studies where the most reliable information is obtained by measuring both mRNA levels and protein levels, in studies involving alkaline phosphatase, a clearer picture of the metabolism of this protein may be best obtained by measuring the mRNA and the enzymatic activity, as well as the status of the protein (intact versus degraded). Histological localization of alkaline phosphatase through utilization of a chromophore-generating enzyme substrate marks very specific sites within bone marrow and sites of new bone formation [353 – 356] (Fig. 11). Developmental studies in vivo and in vitro have shown that expression precedes mineralization and is maintained during early stages of hydroxyapatite deposition [10 – 12,357 – 359]. Within marrow, cells that are supportive of hematopoiesis are alkaline phosphatase positive (Westen – Bainton cells) and represent, at least in part, members of the bone marrow stromal stem cell population [360]. It has been reported that these cells are related to preadipocytes. In certain pathological conditions, when hematopoiesis is halted, these cells lose their alkaline phosphatase activity and proceed to form fat cells [361]. During osteogenesis, the alkaline phosphatase reaction product very

128

FIGURE 11

Alkaline phosphatase in developing bone. By histochemical staining for alkaline phosphatase activity during development, areas that are destined to become bone, as shown here in developing human subperiosteal bone, can be clearly illustrated. The fibrous layer (F) of the periosteum is negative, whereas preosteoblasts (POb) and osteoblasts (Ob) produce high levels of activity. Although a glycoprotein with alkaline phosphatase activity has been isolated from the bone matrix, it is not easily detected in mineralized matrix (MM) by this histochemical assay. Courtesy of Dr. Paolo Bianco.

clearly demarcates areas that will become bone from those that will not. Initially, the cells are all rather flat and spindle shaped, indicative of preosteoblasts, but as osteogenesis continues, some cells (usually those in close proximity to blood vessels) take on the morphology of osteoblasts (plump, highly biosynthetically active cells). As cells mature further to become osteocytes, alkaline phosphatase activity is lost, presumably due to the fact that its function is no longer required [355]. The enzyme exists as a dimer and the identical monomers have an Mr 50,000 – 85,000 depending on animal species and degree of posttranslational modification. Each monomer binds to two zinc and one magnesium ion in each active site. Although the bone/liver/kidney isozyme is the product of the same gene, there are tissue-specific posttranslational modifications that can be detected by monoclonal antibodies [362 – 364]. The enzyme is glycosylated

GOKHALE, BOSKEY, AND ROBEY

and attached to the cytoplasmic membrane on the external surface through a phosphatidyl-inositol-glycan group, which can be cleaved by phospholipase C, thereby releasing it from the cell surface [365,366]. Based on sequence and structural analyses, the E. Scherichia coli enzyme shows extensive homology with mammalian alkaline phosphatase, the major differences being on the enzyme surface [367]. The structure of the enzyme isolated from E. coli has been determined by X-ray crystallographic analyses of single crystals of the (a) Zn2-containing enzyme [368], (b) enzyme with inorganic phosphate bound to the active site [369], and (c) cadmium salt of the enzyme [369]. From E. coli single crystal data, the enzyme takes the form of a twofold symmetrical dimer (100  50  50 Å in dimensions), with the two active sites separated 30 Å from each other. Each subunit consists of a central 10-stranded  sheet surrounded by 15 -helices of various lengths. The active site is in the carboxyl end of the central  sheet, and all the ligands for the three metal ions come from this structure. One serine has been shown to be involved in phosphorylation and dephosphorylation reactions [369]. The phosphate is closely associated with all three metal ions, and the guanidinium group from an arginyl residue. The coordination of the metal ions in the active site is reported to be very similar to the active site of phospholipase C [370], which shows no other structural homology with alkaline phosphatase. The human gene for alkaline phosphatase is located on chromosome 1 [371]. It contains 12 exons, is over 50 kb in length [372], and has an RFLP [373]. The rat gene is at least 49 kb with 13 exons and has a similar gene organization [374]. The gene predicts a protein with 524 amino acids that includes a 17 amino acid signal peptide. The potential for heterogeneity between the bone/liver/kidney isozymes may reflect the fact that there are five potential glycosylation sites. The C-terminal region is hydrophobic, as would be expected for a protein that is linked to the cell membrane via phosphoinositol. It has been determined that regulation of the bone/liver/kidney isozyme is controlled by two leader exons, 1A and 1B [375], with alternative promoters separated by 25 kb. The upstream promoter is used preferentially by bone cells and facilitates its high level of expression in this cell type [376]. The transcription start site is preceded by a GC-rich region, a TATA box, and three SP1 sites [377,378]. The downstream promoter is constitutively active, produces low levels of activity, and is used in the kidney. Three mRNA species of 2.5, 4.1, and 4.7 kb are produced as the result of differential splicing [379]. The list of factors that regulate alkaline phosphatase is lengthy and the results are extremely variable. Expression can be affected by oxygen tension [380] and, in general, requires the presence of serum [381]. In human bone cells,

129

CHAPTER 4 The Biochemistry of Bone

1,25-dihydroxyvitamin D3 stimulates alkaline phosphatase [382]. In immortalized rat osteoblastic cells, retinoic acid also stimulated the enzyme [383], and the region 108 to 45 from the transcription start site was required for this stimulation [384]. Pretreatment of fibroblastic cells with substances that induce cAMP followed by retinoic acid treatment causes an elevation in the mRNA coded for by the exon 1B promoter [385]. IL-4 causes an increase in alkaline phosphatase mRNA in proliferating osteoblastic cells [82]. Dexamethasone increases the utilization of the upstream promoter, resulting in an increase in the tissue-specific expression of alkaline phosphatase [386]. Calcitonin increases alkaline phosphatase expression in osteosarcoma cells [387]. In deer antlers, an interesting model of bone formation, alkaline phosphatase was found to be increased by IGF-I [388]. Ascorbate has also been reported to upregulate alkaline phosphatase expression in chrondrocyte cultures [389]. Alkaline phosphatase activity has been found to be decreased by IL-10 [90] and by lead [91]. Reaction mechanisms for the E. coli enzyme have been predicted from kinetic labeling studies [390], as well as from the single crystal structure [369]. These studies show that the enzyme has phosphate transfer activity and hydrolase activity, transferring a phosphate to or from the serine in the active site from or to the substrate, respectively. The predicted structural identity between mammalian and E. coli enzymes [367] implies a conserved mechanism and suggests that the bone (and cartilage), liver, and kidney isozymes have similar nonspecific phosphomonoesterase activities. The specific function of alkaline phosphatase in bone has been debated for more than 60 years. In the 1920s, Robison, before the existence of the multiple isozymes was known, suggested that the function of this enzyme was to hydrolyze phosphate esters, providing a source of inorganic phosphate [391]. Human studies have shown that different isoform structures are likely to exist due to differences in glycosylation. These studies have also shown that higher alkaline phosphatase activity was present in trabecular bone than in cortical bone [392]. In diseases such as rickets, with impaired mineralization, activity of the bone isozyme is increased [393]. It has also been observed that the commencement of mineralization in culture coincides with increased alkaline phoasphatase activity, indicating that this enzyme was necessary for proper mineralization [see reviews in 394,395]. This hypothesis was confirmed by the discovery of mutations in the alkaline phosphatase gene in hypophosphatasia, a disease characterized by improper mineral deposition [396], and by the observation that cells that do not normally mineralize will form a mineralized matrix when transfected with the alkaline phosphatase gene [397]. Because mineral deposition in vitro will occur (i) when alkaline phosphatase is added to cell-free solutions of calcium ions and phosphate esters [398], (ii) in cell cultures

in which alkaline phosphatase activity is inhibited by levamisole [399,400], and (iii) in the absence of alkaline phosphatase substrates [401,402], the transfection studies do not necessarily identify a definitive function. Mice with null mutations in either the tissue-nonspecific alkaline phosphatase [403] or the bone-specific alkaline phosphatase also provide evidence of the importance of alkaline phosphatase for mineralization [404]. Even the tissue-nonspecific alkaline phosphatase knockout shows increased osteoid and defective growth plate development, leading one to question what the specific mechanism of action of alkaline phosphatase might be. It had been suggested that alkaline phosphatase could also function as a protein phosphate transferase in bone [405] and such transferase activity was noted at physiologic rather than basic pH (required for optimal hydrolytic activity). The transferase activity is in line with the mechanistic studies, and the abundance of phosphorylated proteins implies the need for such an activity. In this light, it should be noted that using non hydrolyzable adenosine triphosphate derivatives retards mineralization in culture [406], perhaps because matrix protein phosphorylation is impaired. A role for alkaline phosphatase in signal transduction may also be speculated based on its analogy to other glycan-linked proteins [365,407]. However, few experimental data support this function. Although a precise function for bone alkaline phosphatase is not known, it seems apparent from hypophosphatasia defects [396] that the enzyme plays a crucial role in mineralization. Because its phosphohydrolase activity is optimal at pH 10 [394], a less direct role than hydrolyzing phosphate esters to produce elevated phosphate concentrations is likely. Part of the function of alkaline phosphatase in mineralization may be associated with its abundance in the membrane-bound bodies, matrix vesicles, believed to be the foci of initial mineralization (vide infra). Another postulated function is the dissolution of calcium pyrophosphate dihydrate crystals in cartilage [408].

B. Osteonectin (SPARC, BM-40) As a result of the development of novel techniques for the extraction of bone matrix proteins in a nondegraded form [2,3], one of the first noncollagenous bone matrix proteins to be isolated and characterized was osteonectin [409]. Osteonectin can constitute up to 15% of the noncollagenous proteins in bone depending on the developmental age and the animal species [410]. Although it was initially thought to be bone specific, with the advent of sensitive antibodies and in situ hybridization, it became apparent that osteonectin is expressed in a number of tissues during development and by many cell types in vitro. In fact, osteonectin was independently identified as SPARC (secreted phosphoprotein acidic

130

GOKHALE, BOSKEY, AND ROBEY

and rich in cysteine) in parietal endoderm [411], culture shock protein in endothelial cells [412], and BM-40 from the EHS mouse basement membrane tumor [413,414]. By in situ hybridization in mouse and human tissue [415,416], osteonectin was found in skin, tendon, developing whiskers, certain nerve trunks, and parietal endoderm. Steroid-producing tissues such as the adrenals, testis, and ovary also produced osteonectin [417]. Immunohistochemical analysis localized transient osteonectin protein to decidua [418] and developing spermatozoa [419], whereas low, but constant, levels were found in renal distal tubule epithelial cells [420,421] and duct cells of salivary gland [420]. In the skeleton, osteonectin mRNA is found in odontoblasts, periosteal cells, osteoblasts, and, to a lesser extent, in osteocytes and hypertrophic chondrocytes [415,416]. By immunohistochemistry (Fig. 12), osteoblasts and osteoprogenitor cells contain osteonectin [422,423], as do odontoblasts [424], periodontal ligament, and gingival cells [425]. Again, low levels are found in chondrocytes and hypertrophic chondrocytes in the growth plate [422,423,426]. During subperiosteal bone formation, there is no detectable osteonectin antigen in the cells or matrix in the fibrous outer layer. Preosteoblasts have a small amount of cytoplasmic reactivity, but osteonectin cannot be detected by immunostaining. In the osteoblastic layer, there is a dramatic

FIGURE 12

upregulation of expression with intense staining of both the cytoplasm and the osteoid, but expression is much lower in osteocytic cells [12,422]. Woven bone contains less osteonectin than more mature, lamellar bone [410], which may explain the differences in concentration between bones of rats (less lamellar in character) and human (more lamellar and osteonal in nature). By radioimmunoassay (RIA), it is noted that bone contains 1000 – 10,000 times more osteonectin than any other connective tissue, such as skin and tendon [427]. Osteonectin is produced transiently in many tissues undergoing differentiation, and it is produced constitutively only by cell types that appear to have ion transport as a common feature. It is associated with development, remodeling, cell turnover, and tissue repair. In vitro studies have revealed that it is involved in counteradhesion and antiproliferation of cells 428. It accumulates in mineralized matrices (dentin, bone, calcified cartilage) and in the  granules of platelets [429]. It has also been found to be associated with prostate tumors [430]. The identification of a bone matrix protein in platelet  granules was the first in a series of discoveries that many bone matrix proteins are found in platelets and megakaryocytes. The reason for the localization of these proteins in cells of this distinctly different lineage is not known. Osteonectin has an apparent Mr 35,000 without reduction of disulfide bonds and appears to increase in size up to

Osteonectin immunolocalization in trabecular bone. (A) Osteonectin is highly enriched in the osteoid and in osteoblasts (Ob) and, to a lesser extent, in osteocytes (open arrow). In addition, osteonectin is also enriched between lamellae (small arrows), which is better demonstrated by immunofluorescence (B). Courtesy of Dr. Paolo Bianco.

CHAPTER 4 The Biochemistry of Bone

Mr 40,000 – 46,000 following reduction, indicative of intrachain disulfide bonds (Fig. 13). The estimated size is Mr 32,000 by molecular sieve chromatography and Mr 29,000 by equilibrium sedimentation, suggesting that it takes on a rather compact hydrodynamic conformation [429]. Osteonectin is also phosphorylated and glycosylated, two posttranslational modifications that may be regulated differentially depending on the tissue of origin. Due to the nature of the amino acid composition and the nature of the posttranslational modifications, osteonectin is acidic with a pI of 5 [431]. There is a single gene (20,000 kb) for osteonectin located on human chromosome 5 [432] at 5q31 – q33, and with one RFLP in the 5 region [433]. The gene contains 10 exons, several of which code for potentially functional regions. While the exons are relatively small (130 bases), the first intron is approximately 10 kb in length. The coding sequence predicts a 17 residue signal peptide and a 286 residue mature protein. The signal peptide is contained in exon 2, whereas exon 3 contains the amino terminus. This region varies from one animal species to another; however, it contains a preponderance of acidic residues. If this region takes on an -helical conformation, the orientation of the carboxy side chains away from the helix would result in the generation of 12 low-affinity calcium-binding sites. Subse-

FIGURE 13

The chemical characteristics of osteonectin indicate the presence of two -helical regions at the amino terminus, along with an ovomucoid-like sequence with extensive disulfide bonding and two EF hand structures.

131 quent exons code for a cysteine-rich region with homology to an ovomucoid-like (serine protease inhibitor) sequence. A structure that is highly homologous to the EF hand (highaffinity Ca2-binding structure) found in calmodulin is encoded by exon 10 and another one, although not as highly homologous, is found in exon 9 [434 – 436]. Sequence predictions indicate that the EF hand domain is in the C terminus. Several other domains, in addition to the EF hand region, have been defined, including a disulfide-rich domain and a pentapeptide KKGHK domain [437]. Three additional genes show high homology with osteonectin. One gives rise to a synaptic junction glycoprotein in rat, SCI [438], and hevin in human [439], a neural retina protein, QR1 [440], testican [441], and FRP [442]. These molecules contain the carboxy-terminal portion of osteonectin (three-fourths of the molecule) and substitute different amino-terminal ends. The role of these homologous forms of osteonectin in their respective tissues is not known. The promoter region of the gene has been isolated and shown to be active by transfection experiments [430,435, 443 – 445]. The promoter does not contain TATA or CCAAT sequences but contains a purine-rich region with GA repeats between 55 and 126. The bovine promoter has a complimentary CCCT-rich sequence further upstream that is believed to cooperate with the GGA-rich region to control transcription. This information, along with S1 nuclease sensitivity, suggests that this region may become hinged or form a triplex conformation. A novel nuclear factor from osteoblastic cells that binds to this region has been identified [445]. There are also numerous other sites, including multiple SP1 sites, one AP1, one CRE, a growth hormone element (GHE), a heat shock element (HSE), and a metal responsiveness element (MRE). The first exon also contains four CCTG repeats, sequences that have been implicated in transcriptional control. In connective tissues, the gene is transcribed to form mRNA species of 2.2 and 3.0 kb. The cDNA codes for a 17 residue signal peptide and a 303 residue protein with a 17 residue signal peptide [411,432,434,443], but with no evidence to suggest the existence of a propeptide [245]. Osteonectin may be differentially glycosylated and/or phosphorylated [446], a possible explanation for differences noted in molecular weights and reactivity with different antibody preparations. There are at least two potential Nglycosylation sites that bear diantenary oligosaccharides (an intermediate between high mannose and complex type oligosaccharides that contains variable amounts of sialic acid and fucose) [447]. It has been found that intracellular forms of osteonectin found in megakaryocytes and certain osteosarcomas have a different carbohydrate structure from secreted forms of osteonectin found in bone [448]. These structural differences may also be reflected in functional differences [449] and could explain differences in the reactivity of osteonectin with different monoclonal antibodies [450].

132

GOKHALE, BOSKEY, AND ROBEY

Factors that regulate the synthesis of osteonectin are not well understood. It appears that the control of osteonectin synthesis is deregulated rapidly, as cells of various tissues begin to produce osteonectin in vitro when they do not normally synthesize it in vivo. Biosynthesis has been studied in a number of cell culture systems, and the regulation and association of osteonectin with mineral formation appear to be dependent on the animal species and culture conditions. In cartilage, IL-1 decreases osteonectin synthesis, whereas TNF- and IL-6 have no effect. TGF-, PDGF, and IGF-I are stimulatory and able to reverse the effect of IL-1 [451]. In bovine bone cell cultures that exhibit extensive mineralization, osteonectin appeared at early stages and remained high thereafter [282]. The effect of TGF- is variable and a stimulation [452] as well as a lack of effect [453] have been reported. The expression of osteonectin by normal human bone cells is not altered dramatically by any treatment (Gehron Robey, unpublished results), although very modest increases with dexamethasone, retinoic acid, IGF-I and dibutyryl cAMP have been reported in other systems [288,411,454]. Interestingly, osteonectin is expressed at constant levels even after heat shock, whereas other proteins decrease after this form of stress [455]. This resistance to heat shock may be related to the presence of a potential heat shock element in the promoter. The only structural studies of osteonectin/SPARC/BM40 other than those predicted from cDNA sequences [e.g., 434,437,456] are NMR evaluations [457] and circular dichroism (CD) studies [71]. The CD measurements were interpreted as showing that the protein conformation had 6%  helix, 26%  sheet, with the remainder nonordered in the absence of Ca2, and less than 3%  helix, 21%  sheet, and 82% nonordered structure in its presence. Analysis of the crystal structure has shown that a follistatin-like

FIGURE 14

domain is related to the serine-protease domain of the Kazal family. There is an insertion into the inhibitory loop in BM-40 and a protruding N-terminal  hairpin with striking similarities to epidermal growth factor 458. NMR evaluations showed the presence of a typical EF hand [457], which in other systems is involved in calcium chelation and calcium transport (Fig. 14). The initial investigations of the function of osteonectin demonstrated that when bound to denatured collagen, osteonectin bound calcium and phosphate ions, suggesting that it was promoting mineral deposition [409]. Osteonectin and its metalloprotease-cleaved fragments also bind to type I collagen [409,459], to types III and V collagens [448,460], and to thrombospondin [461]. Such interactions are likely to be important in determining the organization of the osteoid in bone. Later studies showed that osteonectin was an efficient inhibitor of hydroxyapatite formation in solution [462]. As discussed later, these concepts are not mutually exclusive. Because macromolecules that in low concentration act as nucleators by binding to matrix to provide suitable substrate surface for nucleation can in higher concentrations bind directly to one or more faces on the growing crystal. This can block further growth of the crystal. The tissue distribution of osteonectin within bone suggests, however, that it is not involved in the initiation of mineralization [463]. Expressed by cells in the unmineralized, mineralizing, and mineralized bone, osteonectin accumulates only within the mineralized matrix. Whether it has a specific function in further regulating growth and proliferation of mineral or simply accumulates within the mineralized tissue because of its affinity for hydroxyapatite (Kd  8 108, 11.3 mg osteonectin/gm apatite [460]) is not yet known. A role in the regulation of mineralization does not appear to be a principal function for osteonectin because it is

Structure of an EF hand high-affinity Ca2-binding site. Depiction of the theoretical structure and the amino acid sequence for the EF hand, which has an extremely high affinity for ionized calcium. Courtesy of Dr. Neal S. Fedarko.

133

CHAPTER 4 The Biochemistry of Bone

a widely expressed protein and, in other tissues, it appears to be more involved in regulating cell shape and cell matrix interactions [437] via calcium binding. For example, a fragment of the molecule has been shown to regulate the proliferation of endothelial cells [464]; the intact molecule acts to regulate cell shape in an ion-independent fashion [465,466]. Similar to thrombospondin, it modulates cell – matrix interactions [467]. The original description of the young osteonectin knockout mouse focused on the formation development of cataracts [468]. More recent studies of older mice indicate that the mice develop osteopenia with a significant loss of trabecular bone associated withdue to a decreased rate in bone formation [469]. A number of functions have been suggested by are derived from experiments utilizing synthetic peptides from different parts of the molecule and determining their effect on endothelial cell cultures. While this is an extremely useful approach that provides much intriguing information, it is not clear whether the intact molecule has the same functions or if the effects are cell specific. A synthetic peptide influenced endothelial cultures undergoing tube formation by decreasing the synthesis of fibronectin and thrombospondin-1 and by increasing the synthesis of plasminogen activator inhibitor-1 (PAI-1) [470]. The intact protein decreased bFGF-induced endothelial cell migration [471]. The binding of osteonectin to endothelial cells similarly may be through one of the EF hand structures at the carboxy terminus [472]. Osteonectin and a peptide derived from a noncalcium-binding region have the ability to stop endothelial cell cycle progression, although it does not have the same effect on fibroblasts [473]. In endothelial cells, it may play a role in barrier function by regulating endothelial cell shape [474]. In addition a peptide that becomes conformationally constrained by binding to Ca2 inhibits the proliferation of endothelial cells [475]. Overexpression of osteonectin causes attachment and spreading of endothelial cells in calcium-deficient medium [476,477]. The various activities of osteonectin in embryogenesis and during wound repair have been reviewed extensively by Sage and co-workers [428,478]. Osteonectin may also bind to copper and serve as a source of this ion during angiogenesis [479]. Osteonectin is produced during mid- and late stages of wound repair [480] and interacts with PDGF (specifically with the  chain) and may modulate its activity by inhibiting binding to its receptor [481]. Treatment of fibroblasts with osteonectin induces metalloproteinase activity [482] and may anchor plasminogen and increase its activation, pointing to a function in tissue remodeling [483]. The disassembly of focal adhesion may be through a follistatin-like region that contains one of the EF hand structures [484]. Another interesting study utilized antiosteonectin in developing frog embryos and found that neurulation and myotome development were impaired [485]. Overexpression of osteonectin in C. elegans caused devel-

opmental abnormalities as well as paralysis [486]. Each of these studies demonstrate that osteonectin can influence development in a number of connective tissues. However, while it is known to affect bone development in the mouse, a its specific role in bone has yet to be established.

C. Tetranectin Tetranectin is a tetrameric protein with a Mr of 21,000 (subunits with Mr of 5800) that was first isolated from serum and found to bind to the kringle 4 domain of plasminogen [487]. It is immunolocalized in developing woven bone [488] and is expressed by osteoblastic cultures undergoing matrix mineralization [488]. The cDNA has been cloned [489] and the gene has been isolated [490]. The gene is 12 kb in length and contains three exons. The cDNA predicts for a 21 residue signal peptide and a 181 residue mature protein. It has sequence homology with asialoprotein receptor and the G3 domain of aggrecan and versican core proteins [491]. Overexpression of tetranectin by tumor cells caused an increase in matrix mineralization upon implantation into nude mice [488]. These data suggest that tetranectin may play a role in mineral deposition.

D. RGD-Containing Glycoproteins One of the major breakthroughs in the field of cell – matrix interactions was the discovery of a sequence contained within matrix proteins that would bind to cell surface receptors, thereby mediating attachment [492]. These cell surface receptors belong, for the most part, to the family of integrins. These receptors are formed by one  subunit and one  subunit, each of which have a cytoplasmic extension that may associate with intracellular signaling pathways, a transmembrane domain and an extracellular domain. The extracellular domains of the  and  subunits form a binding pocket that recognizes the cell attachment sequence in the extracellular matrix protein [reviewed in 493,494]. While there are a number of amino acid sequences that bind to integrins, the most frequently utilized sequence is ArgGly-Asp (RGD) [492]. In bone there are at least seven matrix proteins that contain RGD: thrombospondin, fibronectin, vitronectin, collagen (described earlier), osteopontin, bone sialoprotein, and fibrillin. The reason for this redundancy is not entirely clear, although a clue may lie in the fact that these proteins exhibit different patterns of expression during bone formation. Consequently, the interaction of cells with an ever-changing matrix may mediate, in part, changes in cellular metabolism. What follows is a description of the RGD-containing proteins identified in bone, presented roughly in the order of their appearance during developmental bone formation.

134

GOKHALE, BOSKEY, AND ROBEY

1. THROMBOSPONDIN(S) This complex, modular glycoprotein [495 – 498] was first isolated and characterized from the granules of platelets [499]. Since its initial discovery, it has been found that there are at least five members of this family. The various forms are found in a large variety of connective tissues, in particular, in areas of demarcation such as at the dermal – epidermal junction in the skin, surrounding muscle fibers, separating glandular epithelium, and in the lung in peritubular spaces [500,501]. Thrombospondin is relatively less abundant in mineralized matrix relative to other glycoproteins; however, due to its complex chemical nature, it is most likely active in modulating cell metabolism. The expression of thrombospondin during development is sporadic and usually coincides with morphogenetic events such as cell proliferation, migration, and commitment, followed by its disappearance as differentiation continues [502]. In bone, it is found during early stages of osteogenesis. Immunohistochemical localization indicated low levels of expression in the periosteum, with primary localization in developing osteoid [503] (Fig. 15). Osteoblastic cells are stained intensely. There is a moderate accumulation of thrombospondin in mineralized matrix [504] and, by Western blotting, the protein can be detected in bone matrix extracts [503]. Thrombospondin is a highly complex molecule with a Mr 450,000 [reviewed in 505,506] (Fig. 16), composed of three identical subunits ranging in Mr from 150,000 to

180,000 that are held together by disulfide bonds. Each monomer has a number of intramolecular disulfide bonds that give rise to a molecule with a roughly dumbbell shape with distinct functional domains. The small amino-terminal globular domain contains a fibrinogen-like sequence along with a region that may have cell-binding activities [507] and heparin- and platelet-binding sites. In addition to homologies to the propeptide of the (1)I chains of types I and III collagen, von Willebrand factor, and the circumsporozoite protein from Plasmodium falciparum, this small globular domain is attached to an extended stalk region that contains three properdin-like (type I) and three EGF-like (type II) repeat sequences. There is a cluster of cysteine residues in the stalk region that participate in the cross-linking of the monomers and binding sites for types I and V collagens, thrombin, fibrinogen, laminin, plasminogen, and plasminogen activator. A large disulfide-bonded domain makes up the carboxy-terminal region of the molecule and contains sequence homologies to parvalbumin and fibrinogen, with seven calmodulin-like (type III) repeat sequences, although this sequence does not take on the EF hand structure. This region binds to the histidine-rich glycoprotein of serum, activates platelet aggregation, and has multiple Ca2-binding sites. Binding of Ca2 participates in the conformation of the globular domain. The RGD sequence is within the Ca2-binding region; consequently, it is not clear whether the RGD actually is active in mediating cell attachment under normal physiological conditions [508].

FIGURE 15 Thrombospondin immunolocalization in developing bone. Thrombospondin is first expressed in the preosteoblastic layer (POb) in the periosteum of human developing long bone and is highly concentrated in osteoid and osteoblasts (Ob). Somewhat lower levels are maintained in the mineralized matrix (MM). Courtesy of Dr. Wojciech J. Grzesik.

135

CHAPTER 4 The Biochemistry of Bone

FIGURE 16 Thrombospondin is a disulfide-linked trimer that has globular domains at the amino and carboxy terminus interconnected by a stalk region. Each of these domains has a number of binding sites for other proteins, suggesting numerous potential functions in cell – matrix interactions. The cell attachment consensus sequence, RGD, is in the carboxy-terminal domain; however, its availability depends on the calcium ion concentration, which is known to affect the conformation of this region.

Initially it was thought that there was only one gene for thrombospondin. However, it is now apparent that in humans there are at least five (TSP-5 being equivalent to the cartilage molecule COMP), located on chromosomes 1 (TSP-3), 5 (TSP-4), 6 (TSP-2), 15 (TSP-1), and 19 (COMP) [509 – 514], all at least 16 kb in length. While the coding sequences are all highly homologous and differ only in the number of times that the type I, II, and III sequences are repeated, they utilize different promoters [515]. The complete pattern of expression of the different thrombospondin genes is not complete to date [516], although it is known that TSP-1, TSP-2, and TSP-3 are expressed in bone [517,518]. A promoter from the thrombospondin 1 gene has been isolated and characterized [519,520]. It contains a TATA box and an Egr1 site that is flanked by overlapping GC boxes, followed by a GC-rich region. Binding sites for NFY, AP2, SP1, and an SRE have also been identified. Based on the inhibition of TSP-1 transcription by c-jun, an AP1 site may also be present [521]. The resulting mRNA is 6.1 kb [495]. The organization of the TSP-2 and TSP-3 promoters is similar [522 – 524]. Thrombospondin synthesis has been demonstrated in several cell culture systems, including those of adult human

trabecular bone cells [503], in cultures of rat marrow stromal fibroblasts undergoing nodule formation [525], MC3T3 [526], and MG-63 cells [527]. Synthesis appears to be inhibited by dexamethasone [527]. While reducing the overall net synthesis of thrombospondin, TGF-, increases the amount of thrombospondin retained by the cell layer/matrix fraction [503]. Mutant (Tsp-2 null) mice have abnormalities in connective tissue structure and function, including fragile skin [528] and bone (Bornstein, personal communication). Mutant mice have increased cortical bone thickness and density, which has been explained in terms of effects on cell adhesion and differentiation, and collagen fibrillogenesis [529]. While the precise functions of the thrombospondins are not known, they have been postulated to play a role in development, angiogenesis, tumorogenesis, and wound healing [516,530 – 536], and bone remodeling 537. Thrombospondins bind to the small proteoglycan, decorin [265], and along with this molecule may bind to growth factors such as TGF-, which later serve as cell signals. Thrombospondins also bind to osteonectin [461], and considering their colocalization within the  granules of platelets, this complex, along with PDGF and TGF-, may have a

136

GOKHALE, BOSKEY, AND ROBEY

function in repair processes in soft connective tissues as well as in fracture healing. The functions of the various isoforms of thrombospondin have been indicated by several in vitro studies. In an osteoblast-like cell culture (MC3T3 cells), mRNA for one form of thrombospondin increased during differentiation [526], suggesting that thrombospondin has a role in this process. Thrombospondin has been shown to increase osteoclast resorption of dentin slices independent of increases in osteoclast attachment [537]. Thus it might similarly affect differentiation in bone, either by regulating the latter processes or by directly causing the activation of cell signals. Thrombospondin has been shown to bind to TGF- [525]. In vitro, thrombospondin has been found to be active in the attachment of osteoblastic cells. However, cell spreading required the synthesis of other molecules [503]. Thrombospondin has been reported to bind the vitronectin receptor, v3, in an RGD-dependent fashion. However, it was found that attachment and spreading of osteoblastic cells were not inhibited by GRGDS, a competitive inhibitor used routinely to demonstrate the RGD dependency of cell attachment [504,538]. Thrombospondin may also bind to decorin, which is known to interfere with the cell attachment to fibronectin [539]. The precise cell surface receptor responsible for mediating cell adhesion to thrombospondin is not known, although a cell surface receptor that recognizes the sequence VVM in the carboxy terminus of the molecule has been identified [540]. In properdin-like repeats, the VTCG sequence has been found to mediate the

FIGURE 17

attachment of platelets, monocytes, endothelial cells, and certain tumor cells via cell surface CD36 [541 – 543]. Thrombospondin has been shown to interact with fibrinogen/fibrin but not with heparin [544]. Mice that lack thrombospondin have disordered collagen in their fragile soft tissues, increased bone density, and altered fibroblast cell attachment [528]. Bone mineral properties have not yet been determined. However, the properties of these mutant animals confirm the importance of thrombospondin in bone development. Interestingly, unlike other RGD proteins in bone, thrombospondin does not mediate osteoclast cell attachment [545,546]. 2. FIBRONECTIN Fibronectin, one of the most abundant extracellular matrix proteins, is also a major constituent of serum [reviewed in 547,548]. It is produced by virtually all connective tissue cells at some stage of development and accumulates in extracellular matrices throughout the body. Fibronectin appears at an early stage of bone formation during development [504] or during induction by demineralized bone matrix [549,550]. Osteoblasts and osteocytes stain intensely for fibronectin and it is also accumulated in mineralized matrix [504] (Fig. 17). Western blotting analysis of bone extracts indicates that it is relatively abundant. Although fibronectin is a product of osteoblastic cells, it could also be adsorbed from the circulation; consequently, there may be multiple forms from different sources entombed within mineralized matrix.

Fibronectin immunolocalization in developing bone. There is virtually no expression of fibronectin in the periosteum and preosteoblasts (POb) of developing long bone, but there are high levels of expression in the osteoblastic layer (Ob). In addition, high levels of fibronectin are maintained in the mineralized matrix (MM). Courtesy of Dr. Wojciech J. Grzesik.

137

CHAPTER 4 The Biochemistry of Bone

Fibronectin is a dimeric protein with a Mr 400,000, composed of two subunits of Mr 250,000 that are highly homologous, but variable depending on the cell source, held together by two disulfide bonds near the carboxy termini (Fig. 18). Each of the subunits has multiple domains that bind to fibrin (domains I and VIII), heparin, and certain bacteria (domain I), gelatin and collagen (domain II), DNA (domain IV), cell surfaces, including the RGD site (domain VI), and another heparin-binding site (domain VII). Each of these functional domains is composed of different combinations of homologous repeats (types I, II, and III). Type I and II repeats are formed of sequences 45 – 50 amino acids in length that contain a disulfide loop. The type III repeat is twice as long but does not contain a loop structure. The RGD sequence is located in a type III sequence in domain VI approximately one-third of the way from the carboxy terminus. Details of the fibronectin structure have been evaluated from combinations of predictions from amino acid sequence, single crystal X-ray [551], and NMR studies of isolated domains [552,553]. Fibronectin is considered to have three types of domains, two unique  sheets containing type I moieties at the N terminus, 12 type II domains each with hydrophobic pockets, and 17 – 19 type III domains. The overall structure consists of 35% anti parallel  sheets and no  helices. There is a uniform distribution of these pleated sheets along a long chain, resulting in independent

FIGURE 18

structural domains [554]. Most of the type III domains, one of which contains the cell-binding RGD sequence, are in a flexible region of the molecule. The NMR structure [552] suggests that the abundant  sheets form stacks separated by linker regions, which stabilize the molecule. Each of these domains has a distinct function [555]. The N-terminal domain seems to be required for extracellular matrix deposition [556], another domain is required for binding chondroitin sulfate [557], and the C terminus appears to be needed to stimulate its own synthesis [558]. Interactions with cells are modulated by a type III domain containing the RGD [559,560]. The linking region between these domains allows the molecule to stretch to up to seven times its native length [554]. The detailed functions of each of these domains, and the way that they interact with the cytoskeletons of connective tissue cells, have been reviewed [548]. What is apparent from each of these structure – function studies is that fibronectin is essential for matrix deposition. The fibronectin gene is located on chromosome 7 and is very complex, with up to 50 exons [561]. The chicken gene is 50 kb [562] and six RFLPs have been identified [563]. The functional domains, composed of type I, II, and III repeat sequences, are each coded for by an exon. Consequently, it would appear that fibronectin arose from the duplication of multiple genes. Nucleic acid sequence analysis indicates that approximately 90% of the entire coding

Fibronectin is composed of nonidentical subunits that are disulfide bonded at their carboxy termini. The molecule is composed of a series of repeating units (types I, II, and III) that give rise to domains with affinities for other proteins. There are several known splice variants (with or without EIIIB, EIIIA, and V; see text for description). The splice variant present in bone is not known. The cell attachment consensus sequence in a type III unit is RGD; however, other sequences that participate in cell attachment have been identified.

138 sequence is composed of these three repeats. The fibrinbinding domains are composed mainly of type I repeats, whereas the collagen domain is mainly type III repeats. The human gene promoter has been identified and it contains TATA and CCAAT boxes, is GC rich, and has an SP1 and a CRE-binding site [564,565]. Promoter analysis indicates that the CCAAT and the CRE located between164 and - 90 are essential for gene activity. However, gelshift analysis indicates that there are different complements of proteins that bind to this region depending on the tissue source [566,567]. Other studies have shown that the factor (a heterodimer of 43 kDa and a 73-kDa protein) termed ATF2, which binds to the CRE, facilitates the recruitment of factors that bind to the CCAAT element [568]. The gene is transcribed to form mRNA of 7.5 kb, but as might be anticipated, there is a great deal of heterogeneity based on differential splicing and up to 20 different mRNA species have been identified [569,570]. Within the mRNA sequences there are three regions, EIIIA, EIIIB, and V, that can be inserted or deleted depending on the tissue. An example is seen in the differences between plasma (void of E, but containing V regions) and tissue fibronectins (which contain various combinations of Es), which are the result of exon skipping. Differences between fibronectin produced by different cell types have also been found to be the result of exon subdivision (splicing within an exon). Factors that regulate differential splicing are not known, nor is the nature of the splice variant produced by bone cells. Fibronectin synthesis has been demonstrated in many osteoblastic cell culture systems. However, there is not much information on the nature of factors that regulate its production by bone cells. In adult human trabecular bone cell cultures, TGF- increases fibronectin synthesis [453]. In rat and human cells, PTH and TGF- increased fibronectin up to 11-fold. Estrogen caused a decrease in PTH-stimulated levels but had no effect on TGF--stimulated levels [571]. Gallium nitrate, currently under investigation as a therapeutic compound for increasing bone mass, also stimulates fibronectin synthesis in rat calvarial cells and ROS 17/2.8 cells [83]. The elimination of the fibronectin gene in transgenic animals (and all its multisphere variants) was lethal in utero; connective tissues did not form, indicating that fibronectin is a component essential for development of these tissues [572]. Fibronectin is capable of mediating cell attachment and spreading of rat osteosarcoma cells and normal human trabecular bone cells in vitro [504,573]. Numerous integrins mediate cell attachment to fibronectin, including the v3 vitronectin receptor. Interestingly, the attachment of normal cells to fibronectin was RGD independent, indicating that another receptor may be responsible. In fact, it was found that the osteoblastic layer in developing bone, which is

GOKHALE, BOSKEY, AND ROBEY

actively producing fibronectin, is also positive for the 4 integrin subunit [504]. This subunit can combine with 1 to form a fibronectin receptor that is RGD independent. The sequence that mediates attachment to 41 is not yet known. These results are somewhat different from what has been reported for rat osteoblastic cells, which were reported to attach to fibronectin in an RGD-dependent fashion. However, examination of data indicates that incubation with extremely high concentrations of GRGDS (200 M) decreased attachment by only 45% [574]. Fibronectin also mediates attachment of osteoclasts; however, it has been shown that osteoclasts utilize the v3 integrin [545,546, 575]. High-resolution electron microscopy studies have demonstrated that fibronectin can play a role in early biological crystal nucleation, which may be of significance in ectopic calcification, primary nucleation in calcified tissue, and bone ingrowth on ceramic implants [576]. Fibronectin has also been shown to cause apatite formation in solution [577]. 3. VITRONECTIN Vitronectin, also termed the S-protein of the complement system, is produced predominantly by the liver. It is found in serum at concentrations of 200 – 400 g/ml and in extracellular matrices [578]. In fact, it was first identified as the “serum-spreading” factor [579]. It is also found in basement membranes, but generally appears in most matrices containing the fibrillar collagens. It is detectable in developing bone by immunohistochemistry and is found in a very limited number of cells lying on the surface of newly formed bone. However, it is not clear that these cells are in fact osteoblasts. In addition, bone matrix is only faintly stained by immunological techniques, indicating accumulation in bone matrix at very low levels [504]. However, prior to mineral deposition, vitronectin is increased in concentration in the unmineralized osteoid 580, implying that it may be involved in preparing the matrix for mineralization. Although vitronectin is synthesized primarily in the liver, it is also produced by mesenchymal cells at low levels [581,582]. It may also be a biosynthetic product of osteoblastic cells [580]. The protein has a Mr 70,000, and the primary structure of human vitronectin was predicted from cDNA analysis by Oldberg et al. [583] and Jenne and Stanley [584]. An RGD sequence close to the amino terminus and a heparin-binding site in the carboxyl terminus were predicted. Ehrlich’s study [585] has defined several additional homologous domains in the mammalian vitronectin sequences obtained from different sources. From the amino to the carboxy terminus there is a “somatomedin B” domain that is rich in cysteines, followed by an RGD cell attachment site, a collagen-binding domain, a cross-linking site for transglutaminase, a plasminogen-binding site, a heparin-binding site, a PAI-binding site, and an endogenous cleavage site. Sites for sulfation and

CHAPTER 4 The Biochemistry of Bone

cAMP-dependent phosphorylation are also present. The human gene is located on chromosome 17q [586]. In vitro, vitronectin is very active in mediating attachment of all cell types. This in vitro finding may be somewhat misleading, as of all the serum proteins, vitronectin is the most active in binding to standard tissue culture plastic. Consequently, when cells are cultured in vitro, there may be a selection process whereby cells that bear the vitronectin receptor v3 have a distinct advantage over cells that do not have this receptor. Integrins other than the v3 may also bind to vitronectin [587]. Bone cells attach very strongly to vitronectin [504,588]. However, it is not clear what role this interaction plays in bone formation. Because osteoclasts have integrins that are vitronectin receptors [589], it would seem likely that the vitronectin in bone is needed for osteoclast adherence. Further, antibodies to the vitronectin receptor inhibit in vitro osteoclast action [589]. However, the vitronectin receptor distribution on the osteoclasts changes depending on whether the cells are attached or motile [590]. In addition, other RGD-containing proteins such as osteopontin or bone sialoprotein may be utilized by the osteoclast’s v3 receptor. Vitronectin has been shown to inhibit secondary nucleation of apatite crystals in vitro [591]; however, a direct effect on mineral deposition has not been established. 4. OSTEOPONTIN (BSP-1, SPP, pp66, Eta-1) This acidic glycoprotein was identified independently in several cell systems. In bone, it was termed BSP-1, now known as osteopontin [592 – 594]; however, it was also described as a secreted phosphoprotein, SPP, and pp66, a protein that is dramatically upregulated by cell transformation [595 – 597] and in association with tumor progression [598 – 599]. In addition to bone where osteopontin is largely accumulated, it is found in the kidney tubule epithelium, especially in the loop of Henle and the distal convoluted tubule [600], and it is secreted into forming urine (perhaps to inhibit crystal formation) [601 – 603]. It is also produced by mammary epithelium, as it is found at high concentrations in milk [604]. Immunolocalization and in situ hybridization have identified osteopontin in the uterus and placental membranes and in metrial gland cells [440,605]. In adult human tissue, its expression is found in epithelial cells of the gastrointestinal tract, gall bladder, pancreas, urinary and reproductive tracts, lung bronchi, mammary and salivary glands, sweat ducts, and smooth muscle [606], as well as in a variety of tumors [607]. Osteopontin is also found within neuronal cells in the brain and in the inner ear [440,608 – 610]. Platelets and megakaryocytes also contain mRNA for osteopontin, although the level of actual protein appears to be very low [611]. In general, data that have emerged in recent years indicate that osteopontin mediates autocrine – paracrine func-

139 tions in the regulation of tissue formation and plays a role in tissue homeostasis [reviewed in 612]. Inspection of osteopontin production at the cellular level during subperiosteal bone formation indicates that it is produced by osteoblasts and, to a lesser extent, by osteocytes, making it a late marker of osteoblastic differentiation and an early marker of matrix mineralization [11,12,613 – 616]. Marrow also contains cells that contain both osteopontin message and protein. In bone marrow ablation studies in which gene expression was measured as a function of new bone formation, osteopontin exhibited a biphasic pattern of expression. Levels were found to be high during the early phase of the process, most likely corresponding to the initial phase of cell proliferation [617,618]. Subsequently, levels fell to baseline, but began to climb to maximal levels just prior to or coincident with matrix mineralization [463,619]. In addition to being produced by osteoblasts, osteopontin is produced by hypertrophic chondrocytes [620 – 622] and osteoclasts [601,609,610,623,624]. Osteopontin has been localized in bone matrix at the EM level and is highly enriched at cement lines [613,625,626]. Osteopontin is also found in dentin [609]. Some studies have reported that osteopontin is not prominent in osteoid and is most prominent in cells close to the metaphyseal/diaphyseal border where there is active resorption, although the cells surrounding the osteoclasts contained the mRNA for osteopontin whereas the osteoclasts did not [627]. The molecular weight of osteopontin is highly variable depending on the method of analysis. By SDS-PAGE, the Mr varies from 44,000 to 75,000 depending on the percentage of acrylamide in the gel [245,582]. By equilibrium sedimentation analysis, it has a Mr of 44,000 (Fig. 19). Due to the nature of posttranslational modifications, it does not stain well with Coomassie brilliant blue, but becomes blue with Stains All [239,628] in agreement with its acidic pI of 5.0. The structure of osteopontin was predicted by Prince from the primary sequence of bovine osteopontin [629,630]. There is an RGD cell-binding domain, and a single poly-aspartyl repeat sequence. This poly-aspartyl sequence is highly conserved in all species, implying a functional importance for this domain. Direct analysis of the protein indicates that the bone form has an N-linked oligosaccharide, 5 – 6 O-linked chains, 12 phosphoserine residues, and 1 phosphothreonine [592]. Chick, rat, mouse, and human proteins show considerable homology, although potential phosphorylation sites vary [631]. In a posttranslational modification, osteopontin becomes cross-linked to fibronectin through the action of transglutaminase [632], which may further stabilize its deposition in bone matrix. The structure of a segment of the osteopontin molecule has been determined by NMR methods [633]. This structural study examined small peptides in the cell-binding (RGD) domain and defined the stereospecificity of the

140

GOKHALE, BOSKEY, AND ROBEY

FIGURE 19 The osteopontin molecule is composed of numerous stretches of  helix (depicted as cylinders) interconnected in several cases by -pleated sheets, one of which contains the cell attachment consensus sequence (RGD). A stretch of polyaspartic acid (Poly Asp), along with phosphorylated residues (PO4), makes osteopontin a highly acidic molecule. Adapted from Denhardt and Guo, FASEB J. 7, 1475 – 1482 (1993).

serine peptide, GRGDSL. Both the RGD cell-binding domain and a RGD-free cell binding domain in the N terminus have the structures required for integrin interactions needed for cell attachment [634]. Although to date there have been no conformational studies on intact osteopontin, elegant NMR studies of the dentin phosphoprotein’s (phosphophoryn) conformation at different pHs reveal some general features of phosphorylated protein structures that might appear in osteopontin [635,636]. These NMR studies showed that dentin phosphophoryn consisted of structurally organized repeating charged clusters separated by neutral amino acids that were postulated to serve as flexible hinges. The charge clusters are those involved in calcium chelation, and most likely in interactions with the mineral. The availability of expression systems for the native protein [e.g., 637] can facilitate determination of function based on extents of phosphorylation and varying sialic acid contents. Sequence analysis demonstrates that osteopontin and the other phosphorylated sialoproteins have structural features (-pleated sheets containing anionic and phosphorylated residues) that make them well suited for interactions with hydroxyapatite [638]. The osteopontin gene has been isolated in many animal species [639 – 643] and is localized to 4q13 – 21 in humans. The gene contains seven exons and one RFLP [644], and several different alleles [645] have been reported. While the amino acid sequence is highly conserved, there are signifi-

cant differences that appear to be the result of differential splicing of certain exons in different tissues [640,644,646]. In bone, the mRNA predicts a 301 residue protein that includes a 16 residue signal peptide [630,644], whereas osteopontin from osteosarcoma appears to have an insertion due to alternative splicing [646]. The osteopontin promoter is highly complex as would be expected, given the ranges of tissues in which it is synthesized at very precise times and locations. The first kilobase of the mouse osteopontin promoter has been studied intensely. It contains a TATA box, an inverted CCAAT, and a GC box going from 3 to 5 upstream from the transcription start site. There is a positive enhancer between 543 and 253 bp and a negative element between 777 and 543 [642]. There are five PEA-3 (polyoma enhancer activator) sites, multiple TPA sites, SP1, thyroid hormone response (THR), growth hormone factor (GHF), AP4, AP5, AP1, ras activation element (RAE) sites, and a vitamin D response element (VDRE) site [647]. Transcription in bone gives rise to a 1.6-kb mRNA. Due to the correlation of osteopontin production with initial matrix mineralization [648], there have been many studies on the effect of growth factors and hormones on osteopontin synthesis [649]. In ROS 17/2.8 cells, osteopontin is stimulated by 1,25-dihydroxyvitamin D3 [650,651] and TGF- [652]. Long-term treatment with TGF- caused a decrease in expression of osteopontin, indicating a decrease

CHAPTER 4 The Biochemistry of Bone

in osteoblastic phenotype [653]. Osteopontin synthesis is also decreased by dexamethasone [654] and PTH, perhaps due to the induction of cAMP [655]. Endothelin-1, a product of endothelial cells, stimulates osteopontin in osteosarcoma cells [656]. Intermittent compressive forces also stimulate osteopontin [657]. In osteoclast cultures, it has been found that calcitonin inhibits osteopontin mRNA [658]. Phosphoproteins in general have long been linked to the mineralization process based on their accumulation at the mineralization front [106,659] and on the inability of dephosphorylated bone matrices to support mineralization in metastable calcium phosphate solutions [660,661]. The two major non collagenous matrix proteins osteopontin and bone sialoprotein, which accumulate in cement lines and in spaces among mineralized fibrils, show different densities and distribution throughout the tissue. Furthermore, the amount of these proteins generally correlates with the type of bone, speed of tissue formation, and packing density of collagen fibrils [662]. At the molecular level there is indirect evidence from cell cultures deprived of ATP that matrix protein phosphorylation is an essential step in the formation of a mineralizable matrix [406]. Osteopontin is an inhibitor of hydroxyapatite formation and growth in a variety of in vitro systems [663 – 665]. Dephosphorylated osteopontin lacks the ability to inhibit hydroxyapatite formation or growth [663,664], indicating the importance of the phosphate residues and explaining, in part, why osteopontin from different tissues with varying degrees of phosphorylation [e.g., 666] may have diverse effects on mineral formation and growth. Identification of the phosphorylated residues and protein domains required for this inhibition remain to be determined, both in solution and when the protein is bound to a matrix, thereby retaining the conformation it would have in situ. Based on the EM appearance of apatite crystals grown in the presence of 0 – 100 g/ml osteopontin, it appears that this protein blocks crystal elongation [663] rather than secondary nucleation, as is the case for dentin phosphophoryn [667]. This implied that osteopontin binds with high affinity to one or more apatite crystal faces. In fact, it has been shown [613] that osteopontin binds to hydroxyapatite with both high specificity (N  0.026 mol/m2) and high affinity (Kd  1087 g/ mol). It is of interest to note that osteopontin inhibits elastin calcification [668] and in vitro calcification mediated by vascular smooth muscle cells [669]. Thus osteopontin may be important in the prevention of atherosclerotic plaque development. In urine, an osteopontin homologue, uropontin, is similarly believed to prevent the formation and growth of kidney stones based on the in vitro effects of uropontin on oxalate crystal formation and agglomeration [601,602]. Antibodies to this protein localize osteopontin in layers on the kidney stone, suggesting binding to non calcium phosphate minerals. Although it was initially thought that osteopontin might be the phosphorylated

141 protein that was responsible for the nucleation of bone mineral, in solution, osteopontin does not appear to act as a nucleator in any of the in vitro systems studied to date [663,665]. These differences could be due to whether osteopontin is free (in vitro) or bound to the matrix (in vivo). In the latter case, it could be acting as a nucleator. In a number of connective tissue cells in culture, osteopontin appears to be an early marker of cell adherence, being visible immunohistochemically as soon as cells attach to the substratum. Osteopontin promotes osteoblastic cell attachment in vitro [504,573] and therefore may be important in determining the arrangement of cells in the matrix. It has been reported that osteopontin utilizes the vitronectin receptor, v3; however, the evidence is indirect [670]. A fragment of osteopontin also promotes osteoblastic attachment and, interestingly, it does not contain RGD [634]. The evidence for the utilization of the v3 by osteoclasts is much more direct, as attachment could be blocked with a monoclonal specific for the vitronectin receptor [575]. It is possible that this is the initial function of the phosphorylated bone sialoproteins (BSPs) and that later they play different roles by regulating (by osteopontin) and initiating (by BSP) mineralization. Likewise, osteopontin may also play a part later in recruiting osteoclasts to the mineral surface [671] and mediating their activity, as it has been reported that binding of osteopontin by osteoclasts can trigger intracellular signaling [672]. It also appears that osteoclasts dephosphorylate osteopontin, which may in turn alter its activity following bone resorption [673]. It has been shown that osteopontin inhibits the production of nitric oxide synthase (NOS) stimulated by cytokines in kidney cells [674]. This may be of interest in light of the potential role for NOS in osteoblast activity [675]. Another function is in association with osteopontin’s identity to an early T lymphocyte activation-1 (Eta-1) gene. This gene regulates resistance to different strains of viruses, and this resistance has been shown to correlate to different osteopontin alleles [645]. With respect to bone, studies from osteopontin knockout mice indicate that it is not necessary for normal bone development [676]. However, these animals have larger crystals in their bones than age-matched controls [677] and defective osteoclast function [676]. Other functions of osteopontin include enhancement of cell survival by inhibiting apoptosis [676]. This may explain why increased osteopontin expression is associated with metastatic tumor cells 678. In general, whenever ectopic mineral is formed in disease states, many of the bone matrix proteins are found because their expression has been induced by factors that are not yet well understood, or alternatively, due to their affinity for hydroxyapatite. For example, osteopontin has been found in diffuse calcification sites in association with atherosclerosis [679], in mineralized foci within certain mammary tumors [680,681], granulomas

142 of various etiologies [682], and pristane-induced calcified deposits in mice (Gokhale, unpublished data). 5. BONE SIALOPROTEIN (FORMERLY BSP-II) Another glycoprotein, somewhat more bonespecific than osteopontin, is the heavily sialylated glycoprotein, bone sialoprotein (formerly known as BSP-II). A fragment of the protein was first isolated by Andrews and co-workers [683], who initially termed it bone sialoprotein. Subsequently Fisher and co-workers isolated the intact molecule [684] and Oldberg and co-workers later determined the sequence [685]. It can comprise up to 10% of the noncollagenous protein of bone, depending on the animal species. Outside of the skeleton under normal circumstances, examples of BSP expression are very limited. BSP is present in the circulation and may derive in part from platelets [686]. It is possible that BSP is a product of megakaryocytes [687]; however, studies utilizing gray platelets (which lack granules) indicate that BSP may be adsorbed from the serum [686]. BSP is very specific to mineralized tissues and is found in dentin, cementum, and certain regions of hypertrophic chondrocytes [245,628,684, 688 – 691]. In normal healthy tissue, there may be low levels of BSP expression in mammary epithelial cells [692; Van der Pluijm and Gehron Robey, unpublished data] and it is associated with microcalcifications in breast carcinoma [693]. Within the skeleton, BSP expression is also quite limited. During subperiosteal bone formation, the fibrous periosteum and preosteoblastic layers are devoid of expression. Cells in the osteoblastic layer contain BSP, which appears just before or coincident with mineralization. By immunohistochemical staining, BSP localization is not uniform, with only groups of well-defined osteoblasts staining intensely for BSP. The pattern of localization is somewhat atypical in that the cytoplasm is not uniformly stained as seen with other bone matrix proteins. Localization is enriched in the Golgi apparatus. This unique localization was verified at the EM level by the immunogold technique in calcified and undecalcified sections. In tracing the biosynthesis through the secretory pathway, it was noted that BSP, translated and transported through the RER, is concentrated in the Golgi apparatus [353]. Subsequently, it is packaged into secretory vesicles that contain a discrete packet of material that appears to have a honeycomblike substructure. These secretory packets move to the cell surface where their contents are deposited in toto and mineralize immediately in a pericellular area that is relatively devoid of collagen fibrils. These packets (which do not appear to be enclosed within a membrane and are smaller than what has been called an extracellular matrix vesicle) appear to be the first detectable mineralized structures in new bone [353]. Despite this precise localization of BSP with initial matrix mineralization, it is not produced continuously in subse-

GOKHALE, BOSKEY, AND ROBEY

quent stages. After the initial deposition of mineral, the same cells that were previously BSP positive become devoid of BSP, despite the fact that they are identical positionally and morphologically to their previous state of development. These data suggest that the secretion of BSP is not constitutive, but regulated [353]. As shown by in situ hybridization and immunohistochemistry, BSP is also found as chondrocytes begin to hypertrophy during endochondral bone formation, suggesting another example of the osteoblast-like character of hypertrophic chondrocytes [622,689]. The region that separates the developing bone from the cartilage, the lamina limitans, and developing osteoid in this area are also intensely stained [627,689]. In adult trabecular bone, BSP is distributed in a lamellar pattern, and osteoid osteocytes and osteocytes contain the protein (Fig. 20). BSP mRNA and protein were also localized to morphologically well-defined osteoclasts sitting within Howship’s lacunae (Fig. 21). This was the first demonstration of a bone matrix protein being synthesized by both osteoblasts and osteoclasts [689]. BSP has an apparent Mr of approximately 75,000 as judged by SDS-PAGE and is composed of 50% carbohydrate (12% sialic acid, 7% glucosamine, 6% galactosamine) (Fig. 22). It is also rich in aspartic acid, glutamic acid, and glycine, and due to this unique composition, it does not stain well with Coomassie brilliant blue, but is stained by Stains All [684]. BSP, distinct from osteopontin, has two or three sets of polyglutamic acid stretches, each starting with a serine/phosphoserine, and tends to be more highly glycosylated and less phosphorylated [648]. Structure prediction [694] places the polyglutamate stretches in an  helical domain, whereas the proline rich cell-binding RGD containing domain would occur at a V-shaped segment, with the arms of the V highly anionic. The low content of hydrophobic amino acids predicts an open, extended structure [694] analogous to that observed for the bovine BSP in rotary shadowed micrographs [628]. In addition to glycosylation, BSP can also be phosphorylated and sulfated [695]. The sulfate can be localized to either carbohydrate side chains or tyrosine residues [696]. From sequence homologies, a region for such tyrosine sulfation was noted to be between the postulated apatite and the RGD cell-binding sites [694]. Interestingly, in rabbits, BSP is also a proteoglycan. Rabbit bone extracts revealed a protein with keratan sulfate side chains that was identified as BSP after removal of the side chains by keratanase [697]. This keratan sulfate-BSP may also be related to a similar molecule that was identified as a constituent of medullary bone in chickens, which is elevated and decreased during the egg-laying cycle [698]. The functional significance of this posttranslational modification is not known; however, it is an example of how a bone matrix protein can take on different forms, and presumably different functional properties, depending on the animal species. There is also an RGD cell attachment domain in

143

CHAPTER 4 The Biochemistry of Bone

FIGURE 20

Bone sialoprotein immunolocalization in trabecular bone. While highly expressed within the osteoblasts in developing bone, BSP in adult bone is found primarily in the matrix between lamellae (small arrows) and is often enriched at cement or reversal lines (not shown). In addition, some, but not all, osteoid osteocytes (Ost Ocy) and osteocytes (Ocy) contain BSP. Courtesy of Dr. Paolo Bianco.

BSP, located near the carboxy terminus, which recognizes the vitronectin receptor [670] and facilitates the in vitro attachment of fibroblasts [699], osteoblastic cells [504,588], and osteoclasts [575]. A S. aureus-binding site is located

near the amino terminus [700], a factor that implicates BSP in the etiology of osteomyelitis [701]. Although NMR data on RGD peptides exist [553], detailed conformational studies of intact BSP are not available.

FIGURE 21 Bone sialoprotein immunolocalization and in situ hybridization in osteoclasts. Osteoclasts contain both BSP (A) and its mRNA (B and C, Nomarski optics), the first demonstration of the dual source of BSP in bone. It is speculated that osteoclasts utilize endogenous BSP to attach to the bone matrix in preparation for forming the sealing zone. It is not yet known if BSP mRNA is transcribed within osteoclastic nuclei or if mRNA is translated. It is possible that osteoclasts gain these moieties by fusing with osteoblastic cells. Courtesy of Dr. Paolo Bianco.

144

GOKHALE, BOSKEY, AND ROBEY

FIGURE 22 Sequence analysis predicts the presence of multiple stretches of polyglutamic acid (Poly-Glu) in the first half of the molecule and tyrosine-rich regions in the amino- and carboxy-terminal domains. In the carboxy-terminal region, many of these tyrosines are sulfated. The cell attachment consensus sequence (RGD) is flanked by such regions at the carboxy terminus of the molecule. The molecule is composed of 50% carbohydrate, including a high concentration of sialic acid residues. Glycosylation is somewhat restricted to the amino-terminal 50% of the molecule. Adapted from Fisher, McBride, Termine, and Young, J. Biol. Chem. 265, 2347 – 2351 (1990).

However, NMR analysis of a recombinant C-terminal region that contains the RGD indicated the presence of a random coil. It is likely, however, that similar to phosphophoryn, BSP chelation of calcium will involve phosphorylated clusters, and perhaps the sulfated tyrosines are unique to this form of BSP, as contrasted with the other two phosphorylated sialoproteins [317,695,696]. The human gene for bone sialoprotein was initially localized to 4q28-q32 [702 – 704], but now appears to be linked to the osteopontin gene and dentin matrix protein-1 at 4q1321 which have a similar organization and may have arisen by gene duplication (Fisher and Young, personal communication). It is approximately 15 kb in length, containing seven exons, the first six of which are small, with most of the coding sequence located in the last exon [703,704]. The signal peptide and first two amino acids of the mature protein are coded for in exon 2, and exons 3 and 4 contain regions that are rich in tyrosine and phenylalanine. The polyglutamyl acid stretches are contained in exons 5 and 6, and the RGD region is in exon 7. The splice junctions are all type 0, which means that differential splicing would leave the codon intact. Consequently, any splice variant results in an mRNA that will remain within the reading frame and not change the coding sequence. The cDNA codes for a 320 residue protein that includes a 16 residue prepeptide such that the mature protein (unglycosylated) has a predicted molecular weight of 33,600 [702].

The promoter region contains some unusual characteristics [704,705]. There is an inverted TATA and CCAAT box in close proximity to an AP1 site (148 to 142 bp), a CRE (122 to 116 bp), and a homeobox-binding site (200 to 191 bp). A retinoic acid response element (RARE) is present and overlaps with a glucocorticoid response element (GRE, 1038 to 1022 bp). A VDRE overlapping the inverted TATA has also been identified [703]. There is a polypurine (CT rich) stretch that is also found in the osteopontin promoter [643], which can possibly take on DNA triplex conformation [706]. An AC-rich region is also present, which may take on a left-handed helical configuration. This type of structure can either stimulate or inhibit transcription of the gene [reviewed in 707]. A functional YY-1 site has been identified in intron 1 [704,707]. However, the elements that convey tissue specificity to the expression of this gene have not yet been determined. Transcription of the gene results in a mRNA of 2.0 kb, although higher molecular forms have been described [702, Beresford, personal communication]. Studies on the regulation of BSP in vitro have been hampered somewhat by the fact that it is only reproducibly produced in cultures that are actively mineralizing. Because most in vitro model systems are not mineralizing, expression levels are usually low, but detectable. An exception is the cell line UMR-106-BSP, a variant that constitutively produces high levels of BSP [696]. Biosynthesis of BSP is

CHAPTER 4 The Biochemistry of Bone

tightly coordinated with the maturational stage of osteoblastic cells. Using fetal bovine bone cell cultures, BSP mRNA was increased rapidly at the stage of matrix mineralization. These data suggest that the production of BSP is regulated by autocrine factors that are produced by the cells as they go through different stages of maturation. Studies utilizing 1,25-dihydroxyvitamin D3 have shown that unlike osteopontin, BSP synthesis is decreased. A derivative of ipriflavone (metabolite III) has been reported to increase the synthesis of bone sialoprotein [260]. BSP is relatively unique to mineralizing tissues [702,709], expressed only by those cells forming a mineralized matrix [648]. These observations, combined with solution studies showing that BSP acts as a hydroxyapatite nucleator [665] and immunohistochemical data showing that BSP expression in culture is maximal during early matrix mineralization [282,648], suggest that BSP is involved in bone mineralization [317,710,711]. The solution studies of the effect of BSP on apatite formation suggest which structural features are important in this process. When the effect of BSP on mineralization is monitored in an agar gel or at constant pH in solution, it facilitates hydroxyapatite deposition [665,712], although BSP can also block seed growth [713]. Blocking the carboxylic groups, presumably those in the polyglutamyl domains, destroys the nucleation abilities of BSP, while dephosphorylating the molecule has less of an effect. This suggests that apatite – BSP interactions occur predominantly through the polyglutamyl repeats; however, other portions of the molecule are also involved [713]. Although solution data do not prove that BSP has this same function in situ, it does demonstrate the nature of the interaction between BSP and hydroxyapatite. Effects of the variable phosphorylation of BSP remain to be determined. Numerous studies have shown that BSP is active in mediating cell attachment of a number of different cell types that is dependent on RGD [504,573,575,588,589,714]. In studies where it was noted that endogenously produced enzymes could cleave BSP into more or less defined fragments, it was noted that there is a cleavage site just upstream from the RGD that results in a large fragment that lacks RGD and a small fragment that contains RGD at the amino terminus. Surprisingly, the large fragment, which lacks RGD, was able to support cell attachment, whereas the small fragment, which contains the RGD, was relatively inactive [588]. These data suggest that there is a sequence upstream from the RGD that is capable of supporting the attachment and spreading of bone cells, whereas RGD without this upstream region is not active. Although it was initially thought that these data suggested binding of BSP to an integrin other than the v3, current data suggest that the fragments associate with different regions within the same binding site on the integrin, albeit with different affinity and efficacy 713. Because the RGD region is clustered with tyrosine phosphorylation consensus sequences (tyrosyl

145 residues preceded by an acidic residue and a  turn), it was thought that the degree of sulfation may also influence cell attachment. However, unsulfated BSP (prepared from UMR-106-BSP cells treated with the sulfation inhibitor chlorate) was able to support attachment and spreading [715]. It should be noted, however, that cell attachment is an in vitro assay that may not totally reflect the subtleties of a process that occurs in vivo. Although data are not definitive, it appears that osteoblastic cells attach to BSP via the v3 integrin (vitronectin) receptor. Data that osteoclasts use this receptor are much more definitive and several studies have demonstrated that the attachment of osteoclasts can be blocked with antibodies against it [575,589]. In addition, changes in intracellular Ca2 and pH levels in osteoclasts have been shown to be altered, indicating that the binding of BSP (and osteopontin as well) may initiate a signaling pathway and ultimately cause a change in cellular metabolism. It is not yet known what stage of osteoclast formation and/or activation these two molecules are influencing, as there is some controversy as to whether they mediate formation of the sealing zone, which is required for the formation of the extracellular lysosome-like compartment. Other potential functions include mediating interactions between cells and collagen fibrils [716]. Early studies that utilized the injection of isotopic sulfate into animals along with autoradiography and biochemical isolation techniques described a population of sulfated molecules that moved rapidly from the osteoblastic cells to the mineralization front [233,234]. In addition, microprobe analysis indicated that newly formed hydroxyapatite was in close proximity to sulfate [717]. Because proteoglycans account for virtually all of the free sulfate that is incorporated into macromolecules, it was thought that this population of sulfated macromolecules was proteoglycan(s) [233,234]. However, new data that BSP can also be sulfated may provide strong evidence for the role of BSP in matrix mineralization. There are a few suggestions that abnormal BSP metabolism may result in abnormal bone formation and/or disease. In one case, it was demonstrated that BSP was lacking in a cow with an osteopetrosis-like syndrome (Weintroub, personal communication). Given the potential role of BSP in osteoclastic resorption, this may be an indication that in some cases of osteopetrosis, the matrix required for optimal osteoclastic activity is lacking. It has been demonstrated that the production of BSP by certain primary cancers (breast, prostate, and thryoid, to name a few) coincides with the invasion of the tumor into the surrounding tissue and the appearance of mineralized foci within the primary tumor. Furthermore, the levels of BSP in the primary tumors are prognostic of long-term outcome [718,719]. Studies with carcinoma cell lines have shown that metastatic cells do in fact preferentially attach to BSP [720], and it has been

146 demonstrated that cell-bound BSP binds to factor H, a mediator in the alternate complement pathway, and may be involved in the escape of tumor cells from immunosurveillance [721] BSP knockout mice [722], which have a totally nonfunctional BSP gene, were reported to be indistinguishable from wild-type mice at birth, 8.5 days, and 1 month. However, at 1 year they were 25% smaller than the wild type. Histologically, the predominant observation in long bones and in the calvaria was a decrease in marrow space. Incisors were also elongated. X-ray diffraction of the homogenized bones of the knockout animals revealed no differences in mineral crystal relative to controls (Aubin, personal communication). Detailed analyses of spatial changes in mineral properties have not yet been reported. 6. BONE ACIDIC GLYCOPROTEIN-75 (BAG-75) Another sialoprotein originally isolated from rat bone has an apparent Mr of 75,000 and hence is called bone acidic glycoprotein-75 (BAG-75) [723 – 725]. This protein is heavily glycosylated and contains 7% sialic acid and 8% phosphate. Thirty percent of the residues are acidic in nature. It is found only in bone, dentin, and growth plate cartilage, but in culture, cells from soft connective tissues have also been found to synthesize low levels of this protein [724]. It is not known what percentage of this protein exists in bone. BAG-75 shares sequence homology with the dentin phosphoprotein, phosphophoryn, osteopontin, BSP, and dentin matrix protein and therefore may have similar structural features [723,724,726]. The structures of phosphophoryn fragments, predicted from NMR data, were discussed earlier with osteopontin. The cDNA and the gene have not yet been cloned for this molecule. However, some data are available from direct amino acid sequencing. The amino terminus is about 30% homologous with osteopontin. In addition, it does contain polyacid stretches as do osteopontin and bone sialoprotein [723,727]. However, BAG-75 contains both polyaspartate and polyglutamate domains, as well as several phosphorylation sites and an RGD cell binding site [638]. The protein binds with high affinity to both hydroxyapatite and Ca2 ions [723], as well as to collagen [727]. Immunolocalized next to cells in bone and concentrated in newly formed osteoid, this protein may combine the properties of osteopontin (a mineralization inhibitor) and bone sialoprotein (a nucleator). Preliminary studies with its homologue dentin matrix protein demonstrate that it may have both functions (Boskey, unpublished data). BAG-75 also inhibits the resorptive activity of osteoclasts, presumably by blocking access to bone mineral [728]. Related to BAG-75 is dentin matrix protein-1 [729,730] originally isolated from teeth, but now shown to be expressed by bone marrow stromal cells (Fisher, personal

GOKHALE, BOSKEY, AND ROBEY

communication). The nonphosphorylated recombinant DMP-1 [731] is a potent apatite nucleator in solution (Boskey, unpublished data). 7. MICROFIBRILLAR PROTEINS As in many connective tissues, bone matrix also contains microfibrillar structures [732]. Although the protein constituents of the microfibrills have not been completely identified, by analogy to other tissues they are likely composed of a family of proteins that include type VI collagen, fibrillins, fibulins, latent TGF-binding proteins (LTBPs), the MAGP proteins, and Big-h3 [733]. Fibrillins and LTGP are cysteine-rich glycoproteins that are related by their composition of repeating EGF motifs. Fibrillin-1 and 2 are similar but not identical components of the microfibrillar aggregates in the skeleton, as well as in eyes and vascular tissues [734,735]. The structure of fibrillin-1 was predicted by Sakai’s group [736,737] to consist of eight repeated EGF-like motifs, an RGD cellbinding region, several cysteine motifs, and a cysteine-poor COOH terminus. The structure as deduced from primary sequence is thought to be stabilized by calcium-binding EGF domains [736,738]. The proteins associate with glycoproteins, lysyloxidase, a 58-kDa microfibril associated protein, and other protein species forming beaded domains [739]. Fibrillin mutations have been found to cause Marfan’s syndrome, a disease characterized by skeletal, ocular and cardiovascular abnormalities [740]. The association of fibrillins with microfilaments and elastin implies an anchoring function. Fibrillins only form microfibrils in the presence of Ca2; the recently reported mutations in Marfan’s patients who lacked this Ca2-binding domain [738] speak to the importance of these microfibrillar structures. How the microfibrils regulate the growth of long bones is not known; however, immunolabeling in the chick embryo demonstrated labeling in the primary axis and the ventral surface of the notochord [741], suggesting a role in early development. Related to the fibrillins are the latent TGF--binding proteins (LTBPs) and, as their name implies, their proposed function includes binding to latent TGF-. However, they also appear to be microfibrillar proteins based on immunohistochemical analysis colocalizing LTPB-1 with fibrillin at light microscopic and EM levels. These studies show that in addition to regulating TGF-1 activity, LTBP1 may function as a structural component of connective tissue microfibrils. LTBP1 may therefore be a candidate gene for Marfan-related connective tissue disorders in which linkage to fibrillins has been excluded [742]. 8. OTHER CELL-BINDING PROTEINS As microanalytical techniques become more refined, both at the protein level by immunohistochemistry, radioimmunoassays, and enzyme-linked immunosorbent

147

CHAPTER 4 The Biochemistry of Bone

assays (ELISA) and at the nucleic acid level by reverse transcriptase polymerase chain reaction (RT-PCR), it is certain that the list of glycoproteins, including RGD-containing proteins, will continue to grow. These proteins, while not as abundant as those described earlier, most likely make significant contributions to the development of bone. As transgenic animals are generated that either lack or contain a mutated form of these molecules, and/or as chemical amounts of purified proteins are generated through the use of recombinant techniques for use in cell culture and other model systems, more information will become available on the precise roles of all the glycoproteins in bone development. Tenascin (neuronectin), originally identified in muscle, is also made in bone cells [743,744], bone marrow cells [745], and cartilage cells [746]. Tenascin has a hexameric quaternary structure, with six arms attached to a central globular domain [747]. It has been found to be associated with mesenchyme, which is differentiating into bone or cartilage [748,749], but was absent from mature tissue. This indicates that tenascin may be important for early matrix development, where it acts as a modulator of cell – extracellular matrix interactions [747]. A 36-kDa cartilage matrix protein with an RGD sequence [750], this protein has the highest affinity for fibroblasts of all the bone matrix proteins tested to date [751]. Based on protein and DNA sequencing, the protein consists of a 337 amino acid peptide chain, with 10 leucine-rich repeats surrounded by disulfide-bonded loops at each end. Based on its strong cell-binding ability and its distribution, it is most likely that this protein is involved in regulating cell adherence in a variety of connective tissues. Another proteoglycan, osteoadherin, a member of the leucin-rich repeat family, has been purified from bovine bone [752]. Biochemical analysis has shown that it has a molecular mass of 85 da and contains RGD sequences and keratan sulfate chains. It is synthesized primarily by osteoblasts and can bind to hydroxyapatite with efficiency similar to that of fibronectin. This binding appears to be mediated by integrin V3. It has been demonstrated that depending on the species, osteoadherin and osteomodulin are the same protein/gene [753]. A 90 Da protein, periostin (previously called OSF-2), is expressed preferentially in periosteum and periodontal ligaments 754, suggesting a potential role in bone and tooth formation as well as in maintenance of structure. It is secreted by osteoblasts and osteoblast-like cell lines and exists as several isoforms. Distinct from the other cell-binding proteins discussed here, it does not contain a RGD sequence (L. Bonewald, personal communication). Biochemical analysis has shown that glycosylation is not a major component of this protein. In vitro functional studies have shown that the antiperiostin antibody inhibits attachment and spreading of MC3T3-E1 cells, indicating that periostin

is involved in cell adhesion. In addition, TGF- increased the expression of periostin in primary osteoblastic cells significantly. These results indicated that periostin may play a role in recruitment and attachment of osteoblast precursors in the periosteum.

V. GLA-CONTAINING PROTEINS Bone contains a number of proteins that are modified posttranslationally by vitamin K-dependent enzymes to form the amino acid -carboxy glutamic acid (gla). Due to the sequence requirements of the carboxylating enzymes, the gla proteins of bone share some sequence homology with certain blood coagulation factors that require -carboxylation to maintain their activity.

A. Osteocalcin Osteocalcin was the first bone matrix protein to be isolated by the use of nondegradative techniques from acid demineralized bone [755,756]. It comprises up to 15% of the noncollagenous protein, although the level varies depending on the animal species [410] and accounts for up to 80% of the total gla content of mature bone [757]. It was initially reported to be virtually exclusive to bone and was considered the only bone-specific protein. Consequently, a great deal of effort has gone into trying to determine when and where osteocalcin is synthesized and how it functions in bone metabolism. Extensive screening of protein and RNA extracts [758,759] and tissue sections by immunohistochemistry [613,760 – 762] from virtually all tissues has failed to detect osteocalcin in any tissue other than dentin and bone. The one exception was marrow, which led to the discovery that megakaryocytes and platelets contain mRNA for osteocalcin (along with the mRNAs for many of the other bone matrix proteins). It is still not clear how much osteocalcin protein is actually synthesized by platelets and megakaryocytes or what its function is in these cells [611,763]. During bone development, osteocalcin production is very low and does not reach maximal levels until late stages [764 – 767]. By immunohistochemistry, mineralization fronts stain intensely for osteocalcin, but it has been difficult to demonstrate osteocalcin in osteoid and in cells. However, using an antibody against the precursor form of osteocalcin, the primary cell type that stains in developing human subperiosteal bone is the osteocyte. This antibody intensely stained the cell processes in canaliculae and suggests that perhaps osteocalcin bypasses the osteoid layer by being secreted directly at the mineralization front through the osteocytic cell processes [768] (Fig. 23). However, it should be noted that in other animal species, particularly in

148

GOKHALE, BOSKEY, AND ROBEY

FIGURE 23

Osteocalcin immunolocalization in developing bone. Localization of osteocalcin using an antibody against the mature secreted form of the protein sharply demarcates the mineralization front (MF) in developing bone (A). Note, however, the lack of localization within cells that should be synthesizing this molecule. However, when utilizing an antibody raised against the precursor peptide (which is not maintained within mineralized matrix), it can be seen that osteoid osteocytes and osteocytes contain high levels of the proform of the molecule (B and C). Courtesy of Dr. Paolo Bianco.

rat, endosteal osteoblasts are positive for osteocalcin [758,760,761]. The protein has a molecular weight of 5300 but migrates with an apparent Mr of 14,000 Da on SDS-PAGE. The entire sequence has been determined by direct protein sequencing [769,770]. Depending on the animal species, there is one intramolecular disulfide bond and three to five residues of -carboxy glutamic acid [756] (Fig. 24). The original structural predictions by Hauschka and Carr [771] based on circular dichroism suggested that osteocalcin had a structure with an extensive (40%)  helix in the presence of calcium ions. As detailed elsewhere [772], the predicted structure of osteocalcin in the presence of Ca2 consists of two antiparallel -helical domains, one containing -carboxyglutamic acid residues and one rich in acidic amino acids. Both these domains were proposed as sites for calcium chelation. The -carboxyglutamic acids were calculated to be 0.5 nm apart, corresponding to the 0.55-nm interatomic spacing of Ca2 ions in the 001 plane of the apatite lattice, suggesting that this domain might be involved in binding to the mineral. A -pleated sheet in the C terminus was suggested as a cell-binding site. More recent insight into the osteocalcin structure comes from comparisons of NMR data for Ca2 and Pb2 salts [773] and Ca2 and Lu3 (lutecium) salts [774]. These NMR studies show that Pb and Lu3 compete for Ca2 binding sites. Because Pb2 blocks the binding of osteocalcin to hydroxyapatite in

FIGURE 24 This small molecule contains two stretches of  helix (depicted as cylinders) and two regions of -pleated sheet (arrows). The carboxylated residues of glutamic acid in the amino-terminal helix orient the carboxy groups to the exterior, thereby conferring calcium ion binding with relatively high affinity. There is one intramolecular disulfide bridge (C – C) in the middle region of the molecule. Adapted from Hauschka and Carr, Biochemistry 21, 258 – 272 (1985).

CHAPTER 4 The Biochemistry of Bone

solution, such data imply that the osteocalcin – apatite interaction occurs through the same domain as Ca2 chelation in solution. Comparison of Lu3 and Ca2 data for the dog apoprotein demonstrate, the presence of two high-affinitybinding sites for Ca2 and the conformational changes that occur when Ca2 is present. The human gene is localized on chromosome 1 [775] and the structure has been determined [776]. The gene is 1.2 kb with four exons that predict a protein of 125 amino acids. The signal peptide contains 26 amino acids in exon 1, a propeptide of 49 amino acids in exon 2 along with the -carboxylation recognition sequence, two stretches that become -carboxylated in exon 3, and the remainder of the molecule and untranslated region in exon 4 [777,778]. Interestingly, the mouse genome contains three osteocalcin genes, two of which are activated in bone and one activated in the kidney [754]. Although some of the basic elements have been determined in the human promoter, most of the extensive characterization has been done primarily in rodent promoters. It contains a TATA box and a CCAAT box. In addition there is one NF1-binding site and one AP2-binding site, a viral core enhancer, and a CRE. There is a VDRE at 463 to 437 bp [779] that is flanked by other nuclear-binding sites [780 – 783]. A GRE is found in close proximity to the TATA (16 to 1 bp) [784]. A vitamin A response element has also been identified [785]. Apparent AP1 binding sites are also located in close proximity to, or overlapping with, the VDRE and the CAAT box [786]. Because of the highly specific nature of osteocalcin expression, the promoter has been scrutinized intensely to determine what properties convey tissue specificity. This has led to the characterization of the “osteocalcin box” [787,788], located between 99 and 76 bp, which is functionally active [789,790] and contains a binding site for Msx-1 or Msx-2 (homeodomain proteins). Msx-related factors have been found to be synthesized by osteoblastic cells [791]. The first intron has also been characterized and found to have a potential silencer [792,793]. Further characterization of this promoter led to the identification of a binding site, OSE2, located between bp 146 and 132, which binds the transcription factor, cbfa1, the so-called osteogenic “master gene”[794]. The effects of growth factors and hormones on osteocalcin vary somewhat, which most likely reflects differences in the culture systems under investigation (stage of maturity, length of time in culture, etc.). Osteocalcin synthesis is upregulated by 1,25-dihydroxyvitamin D3 [781 – 783,788, 795 – 797] and by 22-oxacalcitriol [798]. In general, most factors decrease osteocalcin expression, such as PTH [799], glucocorticoids [800,801], TGF- [802,803], PGE2 [799], IL-1 [804,805], TNF- [804], IL-10 [90], and lead [91]. Mechanical loading has also been reported to have a negative effect [806].

149 The proposed functions for osteocalcin have been reviewed extensively [807,808]. Because osteocalcin has a high and relatively specific affinity for apatite, probably due to the binding of the gla domain to the 100 (a axis) face of the apatite crystal, the protein has been proposed as a specific regulator of the length of the mineral crystals in bone. Osteocalcin is not expressed in culture until mineralization starts [808,809], which fits the model that it is a regulator of the size and habit of the mineral crystals rather than a promoter of mineral crystal formation. In solution, osteocalcin is an effective crystal growth inhibitor [429,810], whereas -carboxyglutamate-free osteocalcin has no such effect, which is in good agreement with the proposed function. Similarly, during new bone formation, osteocalcin staining and expression occur after mineralization starts [811,812], demonstrating its function in later stages of bone formation and remodeling. Osteocalcin is made by osteoblasts/osteocytes and is considered to be a marker of osteoblastic function [813,814] as well as a coupler of osteoblast – osteoclast action. It also appears to be important for induction of the osteoclast phenotype [815]. Studies probing the function of osteocalcin have shown that osteocalcin-deficient bone particles are not resorbed readily [816], in good agreement with the report that osteocalcin is involved in osteoclast recruitment [817,818]. This concept is also supported by the defective osteocalcin production noted in some humans and animals with osteopetrosis, a severely deforming disease characterized by the failure to remodel bone and calcified cartilage [819 – 821]. Osteocalcin may have other functions. The osteoclastrecruiting function has been questioned based on earlier observations that animals with osteocalcin-deficient bones due to warfarin treatment fail to show significant abnormalities in bone growth and fracture healing [755]. These animals did show excessive growth plate mineralization [822], which could be interpreted as a failure to remodel calcified cartilage. Whether this remodeling effect was due to the failure of osteocalcin to bind to the mineral in calcified cartilage or to a defect in the actions of matrix-gla protein (see earlier discussion), which is produced in the cartilage, is not known. As reviewed here, sufficient data link osteocalcin to a regulatory role in bone mineral maturation. It is likely, however, that several proteins have a similar function, thus it will be difficult to prove this function in vivo. Lambs that lack -carboxylated osteocalcin show decrease in both new bone formation and bone resorption [823], supporting a role for osteocalcin in remodeling. Mice with a null allele (both bone genes deleted) appear normal at birth and at 4 weeks, but by 24 weeks, they have increased bone mass. Preliminary analysis suggested a role for osteocalcin in regulating bone growth in vivo 794. Mineral crystals in the bones of osteocalcin-null animals fail to mature [824], supporting its role in bone remodeling.

150

GOKHALE, BOSKEY, AND ROBEY

B. Matrix Gla Protein Studies that measured the content of gla residues in bone during development indicated that gla-containing proteins were present at earlier stages of development than could be accounted for by the expression of osteocalcin, the major gla-containing protein in bone. Consequently, the presence of matrix gla protein, MGP, was predicted prior to its actual isolation and purification. It was first isolated from bone due to its copurification with BMP [825,826]. While there is little information on the developmental expression of MGP throughout the body, it is known that it is expressed in lung, kidney, heart, cartilage [827], and smooth muscle cells [828]. MGP is more abundant in cartilage than in bone [829]. In the skeleton, MGP expression appears early and remains at the same level at all subsequent stages of development [765]. However, close examination of early developmental stages indicates that MGP expression is much more limited, found only in lung bud and at the limb bud mesenchyme – epithelial interface [830]. MGP has a Mr 15,000 although it migrates as a substantially larger molecule on SDS-PAGE. The secreted form contains five residues of gla and one disulfide bridge in a 77 to 79 amino acid residue protein. A distinct physical property of matrix gla protein is its insolubility in physiologic solutions (10 g/ml) and its tendency to selfassociate via hydrophobic interactions. It also appears that a propeptide present at the carboxy terminus is removed to form the mature protein [831]. Matrix gla proteins from five different species have been shown to have phosphorylated serine residues [832,833]. Thus the protein is a phosphorylated gla protein. Due to its insolubility, along with difficulties in isolating a purified protein, it has been difficult to predict the potential structure of MGP, and detailed conformational studies are not yet available. However, cDNA sequences for several species allowed the primary structure to be predicted [832,833]. The MGP gene has been localized to 12p [834]. The gene is approximately 3.9 kb long and contains four exons. The signal peptide is coded for by exon 1 and an helical region by exon 2. The recognition sequence for the carboxylating enzymes is found in exon 3, and a sequence that actually becomes -carboxylated is in exon 4. There are a series of AluI repeats in the 3 -untranslated region of the gene. The promoter has been characterized and found to contain a TATA box and a CAAT box, along with a perfect palindromic sequence that is similar to a RARE [835]. Little is known about the factors that regulate the synthesis of MGP. In ROS 17/2.8 cells, 1,25-dihydroxyvitamin D3 was found to stimulate its synthesis, although higher levels were needed than for the induction of osteocalcin. It is also apparent that retinoic acid induces MGP expression greatly, which may be highly significant for cartilage development [835].

The matrix gla protein (along with osteocalcin) was initially suggested to be important for the process of endochondral ossification because warfarin-treated rats showed premature epiphyseal closure [822], indicative of impaired remodeling of calcified cartilage. At the time of the report, the matrix gla protein had not been identified, but because cartilage chondrocytes were not thought to produce osteocalcin, a second gla protein was postulated. When the matrix gla protein was identified as a chondrocyte product [827], warfarin data were reinterpreted to consider a potential role for the matrix gla protein. The findings that retinoic acid, a known teratogen, stimulated matrix gla protein gene expression, coupled with the observation that matrix gla protein copurifies with the bone morphogenetic proteins, led Price to suggest that the matrix gla protein may mediate effects of retinoic acid during cellular differentiation, perhaps by serving as a natural carrier for BMP. A recent discovery indicates that the matrix gla protein can be found in a variety of soft tissues and that its mRNA is abundant in lung tissue, although the gla protein content was low [836]. This suggests that the protein has a more general secretory function. Mice that have their MGP gene deleted die prematurely because of massive tracheal cartilage and blood vessels calcification [828]. The endochondral cartilage in these animals is also excessively mineralized, but trabecular and cortical bones appear comparable in mineral properties to age-matched controls [837]. This is convincing evidence that MGP is an in vivo inhibitor of mineralization. This has further been shown in cell culture studies in which ablation of the MGP to sternal chondrocyte cultures resulted in dystrophic mineralization, whereas the addition of exogenous MGP prevented calcification under conditions in which mineralization is normally observed [838]. Additional insights into MGP function may come from reports that in breast cancer cells there is an increase in the expression of MGP [839]. Interestingly, in prostate cells undergoing apoptosis, there is also an increase in MGP [840]. In light of data in knockout animals and in cell culture, it seems likely that expression of this protein may be a protective action by the cell against unwanted calcification. Expression of MGP in cells from non-onion fractures, but not healing fractures, indicates its importance in bone healing [841].

C. Protein S This gla-containing protein is synthesized primarily in the liver, but has been isolated from bone matrix [842]. Although synthesis was demonstrated in osteosarcoma cells [768,842], it has not been reported to be synthesized endogenously in bone. However, it should be noted that patients with protein S deficiency suffer from osteopenia, and

151

CHAPTER 4 The Biochemistry of Bone

consequently this protein, irrespective of its origin, most likely plays a role in bone metabolism.

VI. OTHER PROTEINS A. Proteolipids Proteolipids, as a general class of macromolecules, are membrane components, consisting of a hydrophobic protein component and covalently bound lipid [843,844]. Proteolipids have been isolated from a variety of connective tissues, including bone [845 – 848] and calcified cartilage [849,850], where there are cell and matrix vesicle membrane components [851]. Based on analyses of the apoprotein amino acid compositions, it is clear that there may be more than one type of proteolipid component in bone. Although the primary structures of several of the bone and/or cartilage proteolipid apoproteins have been determined [852 – 854], their detailed structure is not known. Although an oral bacterial proteolipid associated with calcification has been sequenced [855], structures of the bone and cartilage proteolipids have not yet been described. However, the structures of several other proteolipids that appear to have common features have been described in detail. The protein conformations of a variety of proteolipid proteins have been determined using NMR [856], fluorescent labeling [857,858], electron microscopy and quasi-elastic light scattering [859], and Fourier transform infrared spectroscopy [860]. In general, these transmembrane proteins have hydrophobic domains throughout the molecule, including N and C termini. In many cases, the transmembrane domain  helices are highly ordered, although they have abundant hydrophobic residues (e.g., polyvaline). These hydrophobic domains facilitate interactions with the lipids in the membranes in which the proteolipids are contained. In contrast, the N- and C-termini are generally flexibly disordered and contain covalently linked lipids, such as palmitoyl-cysteine or acidic phospholipids. Bone and related cartilage proteolipids have several functions. Based on studies with the bacterial analogue of the chondrocyte proteolipid [848,861], a role in proton transport has been postulated [862,863]. This activity is quite distinct from that of another chondrocyte proteolipid, which is a phosphate ion transporter, taking phosphate ions into and out of the cell [864]. They have also been demonstrated to act as hydroxyapatite nucleators both in vitro [845,847,850], in a gelatin gel [865], and when implanted in a millipore chamber in vivo [866]. The calcifiable proteolipids are associated [867,868] with a complex consisting of decreasing molar amounts of calcium, the acidic phospholipids, and inorganic phosphate [869]. These complexes are capable of acting as hydroxyapatite nucleators in the presence, as well as in the absence, of the proteolipid

apoprotein [867,870,871]. Such complexes were also shown to be components of the membranes of extracellular matrix vesicles [872] and to be able to act as ion transporters [873]. Proteolipids isolated from cartilage matrix vesicles have been shown to include the annexins [874]. Annexins are Ca2 and phospholipid-binding proteins [875] previously known as lipocortin, calpactin, endonexin, chromobindin, anchorin, and so on [876]. Annexins have also been shown to have ion transporter roles [877], but other functions are also known. Lipocortin I is a phospholipase A2 inhibitor [878], and anchorin is a collagen-and cytoskeletal-binding protein [879]. The annexins are synthesized by osteoblasts [880] as well as by chondrocytes [855,881] and are abundant in matrix vesicle membranes where they are involved in the initiation of calcification [882]. Annexins share a 17 amino acid residue homology, which is probably important for the Ca2-dependent phospholipid binding [879]. Wu et al. [851] have reported that the “nucleational core complex,” i.e., the essential structure needed to cause hydroxyapatite formation in matrix vesicles, consists of annexins (i.e., proteolipid), Ca2, Pi2, and phosphatidylserine. This is in good agreement with earlier reports indicating that cartilage proteolipids and their associated complexed acidic phospholipids were hydroxyapatite nucleators. Because proteolipids are concentrated in matrix vesicle membranes [872] and because most are involved in ion transport, these proteolipids seem to be of general importance in accumulating ions within the cell and/or extracellular matrix vesicles. As ions accumulate within vesicles, in the presence of proteolipids and phosphatidylserine, mineral crystals form associated with the membranes. Thus, the two functions of the proteolipids may be closely related.

B. Enzymes and Inhibitors It is apparent from the slow destruction of bone matrix proteins that occurs with increasing age [883,884] that enzymes are present within the mineralized matrix. The origin of these enzymes can be from the circulation due to their affinity for hydroxyapatite or from neighboring cells that are highly associated in bone such as endothelial cells, hematopoietic cells, and cartilage cells. In addition, cells in the osteoblastic lineage also synthesize a number of enzymes and their inhibitors. A source of enzymatic activity that has often been overlooked is the osteoclasts themselves, which secrete lysosomal types of enzymes into the extracellular compartment formed by the sealing zone. 1. METALLOPROTEINASES The matrix metalloproteinases (MMPs) [reviewed in 885] are a family of enzymes that can be roughly divided into collagenases, which produce the initial cleavage of

152 native, triple helical collagen; gelatinases, which further degrade the structurally compromised collagen, and stromelysins, which degrade proteoglycans. However, there is some degree of overlap among these groups. It has long been known that connective tissue cells that are synthesizing extracellular matrix proteins also have the capacity to synthesize proteins that destroy the very same proteins. Osteoblastic cells are no exception, and it has been reported that they synthesize collagenase-1 (MMP-1) [886], collagenase-3 (MMP-13) [887], gelatinase A (MMP-2) [888] and gelatinase B (MMP-9), [889 – 892], and MT1-MMP (vide infra). Some osteoblastic cells have been found to bear a cell surface receptor for collagenase. In addition, matrix vesicles contain MMPs [893] along with a stromelysin-like activity that has the ability to degrade proteoglycans [894,895]. Osteoclasts also have been reported to contain collagenases [896,897]. To date, a number of transgenic animals have been generated that are deficient in an MMP, but generally did not display a skeletal defect. The one exception is the recently reported MT1-MMP knockout mouse, which while normal at birth quickly develops a severe skeletal phenotype characterized by dwarfism, osteopenia, and arthritis [898]. Tissue inhibitors (TIMPs, forms 1 and 2) inhibit the activity of metalloproteinases [899]. However, it is not clear to date at which point this activity is produced. Depending on the cell culture system, it is apparent that there is a great deal of variability in the ability to produce these enzymes and inhibitors. In those cell culture systems producing MMPs, it has been found that MMP activity is stimulated by PTH [891,900], TNF- [901], and retinoic acid [902]. The crystal structures of two of the matrix metalloproteases [903] and one form of collagenase [904] have been resolved. These are both members of the stromelysin III family (to date, 11 such enzymes have been identified). Structural features include a histidine-rich -helical catalytic domain that chelates the metal ion required for activity [905], a C-terminal domain that is involved in linking the enzyme to the membrane, and pre- and pro-N-terminal domains that are thought to be responsible for maintaining the enzyme in inactive form via interaction with TIMP [906]. The catalytic domain is a spherical region with a methionine-based turn containing Zn2 and Ca2-binding sites, with proline-86, phenylalanine-79, and aspartate-232 forming the base of the active site residues [904]. The crystal structure of elastase [906], the enzyme responsible for the degradation of elastin, has similar features. The crystal structure of the active site of collagenase [905,907] shows the presence of a five-stranded -pleated sheet, three  helices, and binding sites for Ca2 and Zn2, but here the zinc is chelated only by histidines. Collagenase cleaves at a unique site in the collagen triple helix [908] and at a minor site in the nonhelical N-terminal

GOKHALE, BOSKEY, AND ROBEY

region. Activated by tyrosine kinase-dependent phosphorylation, collagenase-mediated turnover of the bone matrix seems essential during growth and repair, but not during early development. Thus, homozygous transgenic mice whose type I collagen does not contain the unique cleavage site appear normal at birth, but develop thickened skin, uteri, and bone during growth and have impaired fracture healing [909]. In bone, osteoclasts produce several cysteine proteases as well as metalloproteases. It is believed that during osteoclastic resorption the cysteine proteases that have acidic optimal pHs act first [910,911]. Then, as the mineral is dissolved in the acidic environment and the acidity is neutralized, the matrix metalloproteases function. Specific inhibitors of the cysteine proteases have been used effectively to inhibit osteoclastic resorption [912]. Such inhibitors, while blocking the actions of the cysteine proteases, increase the activities of some lysosomal enzymes [913]. As reviewed in detail elsewhere [911], metalloproteases degrade proteoglycans, collagens, and matrix proteins [914 – 916]. One of the most important of the cysteine protease degradative enzymes in bone may be cathepsin K, as this enzyme is expressed mainly by osteoclasts and appears to initiate the bone degradation process. In this light, cathepsin K knockout animals have osteopetrosis associated with abnormal matrix degradation but normal mineral resorption [917,918]. 2. PLASMINOGEN ACTIVATOR (PA) AND PLASMINOGEN ACTIVATOR INHIBITOR (PAI) Another enzyme system that is represented in bone matrix is plasminogen activator (both urinary and tissue types, uPA and tPA) and PAI-1 [919, reviewed in 920]. PA and PAI have been identified in both transformed and nontransformed cell culture systems [921], (Kopp and Gehron Robey, unpublished results). However, it is also possible that enzymic activity detected in bone matrix can also originate from other neighboring cell types. The proposed functions of these proteins are varied. They range from the controlled activation of other enzymes, such as matrix metalloproteinases, to the activation of growth factors, such as TGF- and the IGFs that are bound and modulated by the IGF-binding proteins. The synthesis of enzymes and inhibitors is regulated by a number of growth factors and hormones, and in general they are modulated in different directions; i.e., if the enzyme goes up the inhibitor goes down, and vice versa. PA is upregulated by PTH [921 – 924], 1,25-dihydroxyvitamin D3 [925], and acidic and basic FGF, EGF, and PDGF [926]. PAI is concurrently decreased by these factors. Activity is either upregulated or downregulated by TGF-, depending on the cell type [927], although there is generally an increase in PAI-1 [926,928]. PA activity is decreased and PAI activity is increased by glucocorticoids [929].

153

CHAPTER 4 The Biochemistry of Bone

3. MATRIX PHOSPHOPROTEIN KINASES A third category of enzymes that appear to be critical for the formation of mineralized connective tissues are the extracellular matrix phosphoprotein kinases. Matrix protein kinases isolated from bone and dentin [930 – 934] are casein II kinases, whose activities can be inhibited by heparin and 2,3-diphosphoglycerate. Analogous to some tyrosine kinases found in the extracellular matrix [935], these phosphoprotein kinases are responsible for the extracellular addition of phosphate to noncollagenous matrix proteins [936]. Deficient phosphorylation due to altered casein kinase II activity has been reported in the hypophosphatemic mouse, an animal model of human hypophosphatemic rickets, that is resistant to phosphate and vitamin D treatment [937]. Further, because there are protein kinases in the extracellular matrix, it is likely that phosphoprotein phosphatases are also present.

C. Bone Morphogenetic Proteins (BMPs) The existence of BMPs (as a concept rather than as isolated proteins) was first suggested by Huggins [938] and later by Urist [939] who observed that the demineralized bone matrix, when implanted in an ectopic site, would cause cartilage and bone formation. Their observations led to an intense search to purify and characterize the factor present in bone. What followed was the identification of a class of proteins that is now recognized as the TGF- superfamily, which include the BMPs, and are capable of inducing bone when implanted into various sites within test animals. The BMPs, with or without carriers, alone or in combination, all seem to have the ability to cause mesenchymal cells to differentiate, and hence to induce bone formation (see Chapter 5 by Olsen). Effects of BMPs on cell morphology and function, proliferation, and differentiation have been reviewed in detail elsewhere [940 – 942]. However, it is now clear that there is a great deal of redundancy and that these proteins are not specific to bone. It is also not yet known how the temporal and spatial pattern of expression of these proteins controls developmental processes. The recognition that the response to these proteins is mediated by cell surface receptors [943] has opened the door to a rapidly expanding field, which will provide a great deal of insight into how this class of proteins functions. Six of the seven well-characterized BMPs are members of the TGF- superfamily whose primary structures have been elucidated by molecular cloning [941]. Of these, only BMP-1 does not have the TGF- structure. The carboxy-terminal regions of BMP-2 – BMP-7 have high structural homology with seven cysteine residues at conserved locations. The mature protein (lacking the propeptides) forms dimers. As reviewed by Wozney and co-workers [941,944,945],

these may exist as hetero-or homodimers, each with different activities. All of the mature BMPs have basic amino termini, which persist following the removal of long propeptides. The mature chains are also glycosylated. BMP-1, in contrast, has domains resembling those in metallo endoproteases and the C1 and C1q of components of complement. The crystal structure of one of the TGF-s (TGF-2) has been determined [946]. It has been shown by NMR [947] that it has an extended rather than a globular structure without a hydrophobic core due to the arrangement of the four disulfide bonds, which form the so-called “TGF- knot’’ in one section of the molecule. Such extended structures common to the TGF- superfamily may be important for interaction with the receptors of the TGF s, which are transmembrane serine-threonine kinases [948]. Based on sequence analyses, it has been found that all of the BMPs have hydrophobic leader sequences, suggesting that they are secreted proteins. This is in line with their proposed functions as growth factor bone-inducing agents. The identification of the defect in the short-ear mutant mouse [949] as an abnormality in the BMP5 gene led to the suggestion that BMPs are involved in the formation, patterning, and repair of specific morphological features. BMP has also been linked to the patterning in the developing chick limb [950]. Another illustration of effects of BMPs is seen in mouse model of fibrodysplasia ossificans progressive, which lacks this protein [951]. Studies have identified BMP receptors on mesenchymal cells, providing a clue as to how these soluble bone derived proteins may function as growth factors in situ. These types of studies have shown that the BMP family is functionally redundant in that phenotypic changes often are not apparent unless more than one of these factors is knocked out [952] (see Chapter 5).

D. Growth Factors The skeletal matrix and mineral are repositories for growth factors [953 – 956], such as the transforming growth factor- family [941], insulin-like growth factors [957], platelet-derived growth factor [958], basic and acidic fibroblast growth factors [953], interleukins [959,960], granulocyte – macrophage colony-stimulating factor [961], tumor necrosis factor [962], and the bone morphogenetic proteins (see earlier discussion). Many of these growth factors are products of bone cells as well as other cells and have specific receptors on osteoblasts, osteoclasts, and osteocytes [963]. These receptors include cell surface receptors that are linked to G-proteins, have directly linked kinase activities [964], and/or that are channel linked [965]. Thus they serve autocrine (self-regulating) and paracrine (regulating other cells) functions [966]. Their actions are regulated by factors such as systemic hormones and by mechanical stress.

154

GOKHALE, BOSKEY, AND ROBEY

The actions of these growth factors in bone are being reported at an exponential rate and have been reviewed by Gowen et al. [962], Rosen et al. [957], and Yoneda [967] (see also Chapter 14). In general these growth factors have mitogenic and proliferative activities in bone, and many influence the expression of specific phenotypic proteins, thereby affecting both bone formation and resorption. The ability to affect these processes depends on the presence of receptors for the particular growth factor or cytokine. The mechanism of action of cytokine binding to such receptors generally involves action of a transmembrane protein kinase, which in turn results in phosphorylation or dephosphorylation of multiple proteins in the signal transduction cascade [948].

E. Serum Proteins (2-HS-Glycoprotein, Fetuin, Albumin, etc.) The list of nonstructural proteins that have been identified in bone that originate from serum and become entrapped in bone is quite lengthy [968]. This ability of bone to adsorb such a broad spectrum of proteins makes it uniquely suited to be the reservoir of multiple growth factors or proteins that may be required and liberated at least in part by bone resorption. Albumin, 2-HS-glycoprotein, fetuin, transferrin, 1-antitrypsin, 1-antichymotrypsin, IgG, haptoglobulin, hemopexin, serum cholinesterase, and soluble fibronectin are among the plasma proteins that accumulate in bone in detectable amounts [969 – 971]. Their accumulation is most likely due to their binding to hydroxyapatite. This affinity for hydroxyapatite is frequently the basis of the separation of different classes of the immunoglobulins [972], removal of interfering plasma constituents [973] and other purification schemes. Human 2-HS-glycoprotein is produced in the liver and circulates in the bloodstream [974]. There is some debate as to whether connective tissue cells can produce the protein, but it is apparent that it specifically accumulates in mineralized tissues. The high concentration of 2-HS-glycoprotein in bone cannot be due entirely to the presence of blood in the tissue, as the relative amount is too high and the ratio of the protein to albumin is enhanced 100-fold in bone relative to blood [970]. Chondrocytes have been shown to express 2-HSglycoprotein where the protein appears to enhance endochondral ossification. The message is absent from normal human bone cell lines [975], indicating that it probably accumulates in human bone by adsorption from the circulation. In rodents a bone sialoprotein has been shown to have high homology with the 2-HS-glycoprotein [976,977]. However, while Ohnishi et al. [977] found that the protein was expressed by osteoblasts, immunolocalization of the analogous protein by Mizuno et al. [978] failed to show its

presence within the cell per se, implying that like the human homologue it is deposited in bone from other sources. 2-HS-Glycoprotein consists of two nonidentical glycosylated peptide chains (chains A and B) that are held together by disulfide bonds [979,980]. These subunits are characterized by repeating Ala-Ala and Pro-Pro sequences. The initial biosynthetic product is a single polypeptide that is cleaved to form the A and B chains of the mature molecule [981]. The major glycoside is sialic acid [982]. Prediction of the primary structure and chemical digestion studies have led to descriptions of the domain structure of the protein [983,984]. In addition to the single disulfide bond linking the two individual chains by their extreme NH2 and COOH ends, there are five intradisulfide bonds in the A (heavy) chain. The light B chain has no intrachain S – S bonds. The A chain is composed of three domains, consisting of S – S linked loops. Of the five loops that span 4- to 19-amino-acid residues, two highly homologous loops form one domain, flanked on either side by the other tandem repeats. This structure is typical of that of bovine fetuin [985], a member of the cystatin superfamily, and of the histidine-rich glycoprotein family [984]. The cystatin family proteins all have “cystatin-like” domains of linearly arranged disulfide linked loops [984,986] and a variety of N-glycosides [987] that vary within the family, and for the same protein they vary with species. Members of the cystatin family are inhibitors of cysteine proteases. The 2HS-glycoprotein is also homologous to a nonphosphorylated sialoprotein found in rodent bone [976,977]; however, the rodent counterpart of 2-HS-glycoprotein consists of one rather than two chains. 2-HS-Glycoprotein can bind TGF-/BMP cytokines and block their osteogenic activity in culture. Studies with dexamethasone-treated rat bone marrow cell cultures have shown that recombinant fetuin as well as native serum protein can inhibit osteogenesis by similar efficiency. Northern blots showed that both 2-HS-glycoprotein and high doses of TGF- suppressed transcripts of alkaline phosphatase, osteopontin, collagen type I, and bone sialoprotein. These data suggest that 2-HS-glycoprotein (together with TGF) is involved in regulation of osteogenesis in remodeling bone [988]. The human gene sequence for 2-HS-glycoprotein is known [989]. The gene is on chromosome 3, and two RFLPs have been identified. The single mRNA species predicts an 18 residue signal peptide, followed by a sequence that codes for the A and B peptides with an intervening sequence between them. This sequence is presumably lost during cleavage of the precursor to form the mature molecule. In tissues other than bone, 2-HS-glycoprotein has several functions. It has been shown to have opsonic properties [971], to promote endocytosis, and to bind DNA [990,991].

155

CHAPTER 4 The Biochemistry of Bone

In liver, 2-HS-glycoprotein (pp63) plays an essential role in insulin signaling, inhibiting the insulin-dependent tyrosine kinase activity of the insulin receptor [992] without competing with insulin binding. It is also believed to be involved in the immune response [993,994]. 2-HS-Glycoprotein has been suggested to play a variety of roles in bone, but none of these has been proven conclusively. It appears at selected times during development and may be involved in differentiation [995]. This idea is supported by earlier studies demonstrating that 2-HS-glycoprotein is found in higher concentrations in young bone as opposed to mature bone [996,997]. The protein appears to have a higher affinity for calcium phosphates than other serum proteins, as the addition of calcium and phosphate to serum led to the removal of all the 2-HS-glycoprotein but removed less than 1% of the albumin [970]. 2-HS-Glycoprotein is also believed to be involved in bone mineralization [998], as decreased levels of this protein are seen in bones with abnormal mineralization pattern due to malnourishment, certain malignancies [993], and Paget’s disease [970]. In contrast, 2-HS-glycoprotein levels are high in some patients with OI [996]. The suggestion that this glycoprotein might affect mineralization is also supported by solution studies. In fact, the ability of serum to inhibit the solution-mediated conversion of amorphous calcium phosphate to hydroxyapatite [999] was attributed to the presence of this high-affinity glycoprotein. This was later confirmed in unpublished in vitro studies where purified 2-HS-glycoprotein was seen to be a much better inhibitor than other serum proteins. The 2-HS-glycoprotein-deficient mouse did not show skeletal abnormalities [1000]; however, the serum from these animals did not inhibit apatite formation as efficiently as that from wildtype animals. The mutant animals also developed ectopic calcifications, confirming the role of this protein as a serum inhibitor of calcification. Other studies suggest that the protein is involved in the recruitment of osteoclast precursors to resorptive sites [817,1001]. Because of its homology to cysteine protease inhibitors, a role in regulating bone resorption seems likely, and it may be that the high concentrations in very young bone may be related to the need to limit remodeling during development. Whole serum [1002] and albumin have been shown to inhibit hydroxyapatite growth in solution [1002 – 1004]. The ability of albumin to inhibit apatite growth is attributed to the affinity of albumin for apatite [973,1005 – 1007]. Specifically, albumin at 50 – 250 g/ml alters the linear rate of growth of apatite seed crystals by binding to the mineral on several faces [1003] and blocking the growth of crystal agglomerates [1004]. Although the primary function of albumin in bone is not apt to be one of regulation of mineralization, the extent of inhibition of hydroxyapatite growth in solution indicated that phosphoproteins were more effective

inhibitors than albumin. Ions such as citrate and magnesium were also less effective in retarding apatite growth in solution [1003]. Transferrin [1008], IgG, IgE, and the other serum proteins also bind to apatite. Studies from the Laboratories of Borkey and Heywood indicated that IgG had no effect on hydroxyapatite formation, morphology, or growth. None of the other serum proteins has been reported to have any effect on either inhibition or promotion of mineralization. Effects on other events in bone matrix deposition are not known; however, because these serum proteins increase in concentration as mineralization progresses [970], it is unlikely that they play significant roles in the bone deposition process.

VII. REQUIREMENTS FOR MATRIX MINERALIZATION Analyses of diseased tissues and tissues from transgenic animals indicate that there are a number of cellular and extracellular factors essential for physiologic mineral deposition. Physiologic mineral deposition is distinguished here from dystrophic or pathologic mineral deposition by its ordered, characteristic appearance on an oriented collagenous matrix. In contrast, in dystrophic calcium phosphate deposition in general, and dystrophic apatite deposition in particular, the mineral crystals are often not collagen associated. They surround and engulf cells and may be longer in size and different in included ion composition from physiologic mineral [1009,1010]. From the just described definition, it is apparent that for physiologic mineral deposition, there must be an appropriate collagen-based matrix. This is emphasized by (a) the smaller size of hydroxyapatite crystals in OI bone [157] and (b) the relative abundance of mineral that is not associated with collagen in these bones with deficient and/or impaired collagen production [158]. Although in some cases the defective mineralization in the OI bones may also be attributed to altered matrix protein production [167] or retention, collagen is clearly an absolute requirement for physiologic bone mineralization. Similarly, because fibronectin forms the basis on which collagen is deposited, it must be also a requirement. Equally apparent from this definition is the essential presence of Ca2+ and Pi ions for mineralization. Calcium ions may be supplied from the cells or from circulating or localized calcium-binding proteins. Phosphate ions may be derived from the breakdown of pyrophosphate, an abundant metabolic product, from hydrolysis of phosphoesters or phosphoproteins, or from circulating Pi ions. The exact Ca2 and Pi content of the extracellular fluid of bone is not known, but in cartilage, micropuncture studies showed the pH to be 7.58, and total Ca2 and total Pi to be 1 – 12 and 3 – 12 mg/dl, respectively [231]. For the formation of

156

GOKHALE, BOSKEY, AND ROBEY

apatite, a basic environment is also required, and many of the highly anionic matrix proteins probably contribute to creating this environment. Which of these matrix proteins is truly essential for mineralization of bone can only be guessed until the sequence of protein expression is determined precisely and appropriate knockout and transgenic models are developed. Even in these cases it may be difficult to prove an essential role for mineralization, as it is already apparent that there are redundant controls of this critical process. It is certain that the cells are required for the production of a physiologic matrix, synthesizing and exporting necessary enzymes, growth factors, and matrix molecules. As discussed later, the formation of extracellular matrix vesicles is also apt to prove critical for the initiation of mineralization in some cases.

VIII. PATHWAYS OF MATRIX MINERALIZATION Apatite crystals found in bone are distinct from most naturally occurring geological crystals, being smaller and containing more impurities. In addition, bone apatite crystals have a plate-like habit [1011], are arranged in an oriented fashion on a collagen-based matrix, and have a very limited size distribution [19]. In general, the mineral crystals in bone (and dentin) are smaller than those in enamel [1012] and in dystrophic deposits in severely atherosclerotic plaques [1013] or other soft tissue calcifications [202]. The bone mineral crystals do vary in size with tissue site, age, and disease [157,1014], but the range in the lengths of the smallest bone mineral crystals and their orientation implies that their growth must be regulated. Bone mineralization is thus distinct from solution-mediated Ca2 phosphate precipitation where similarly sized, nonoriented small crystals that are formed can ripen to appreciably larger sizes [1015]. This is also distinct from geologic apatite formation where high temperatures and pressures yield extremely large single crystals.

A. Physical Chemistry of Mineralization Calcium phosphate precipitation from solution can yield a variety of phases, depending on the pH, Ca to Pi ratio, and solution supersaturation [21,1015,1016]. When the pH is in the physiologic range (7.4 – 7.8), apatite formation occurs with solution Ca:P molar ratios as high as 2 to 1 and as low as 1 to 1 as long as the solution is supersaturated with respect to apatite (has a Ca P product that exceeds the solubility product for apatite). Depending on the supersaturation, intermediate phases such as amorphous calcium phosphate [1017], octacalcium phosphate [1018 – 1020], or

other intermediates may form [21,1021], but in all these cases, the final product is apatitic. Apatite crystals develop in solution when individual ions or ion clusters associate in the same orientation as in the crystal lattice that they are trying to form. When sufficient ion clusters are oriented correctly, they can persist in solution and can serve as a “critical nucleus” for further crystal growth. Homogeneous nucleation, in which crystals form de novo, is a rare process [1015]. Thus it is likely that in most instances of solution-mediated apatite deposition, nucleation occurs on foreign materials such as dust, scratches on the container, and buret tips. Such heterogeneous nucleation yields the initial crystals, which then facilitate additional growth by the process of secondary nucleation. In secondary nucleation, growth sites on the preformed apatite crystals serve as branch points for the formation of new crystals, analogous in many ways to the branching of the growing glycogen molecule during glycogenesis. Proliferation by secondary nucleation results in numerous small crystals. Crystal growth, in the absence of secondary nucleation, would result in fewer, but larger crystals. This suggests that most of the crystals in bone form by a secondary nucleation-like process or by growth from individual nuclei. Unfortunately, what regulates crystal size in bone cannot be determined from studies of proteinfree solutions. The mineral in bone, as in other physiologically calcified tissues, is associated with an organic matrix (Fig. 25). The protein(s) within such matrices can alter the nucleation and growth of mineral crystals in several ways. When isolated in an environment relatively free of body fluids, the protein(s) can chelate Ca2 or Pi ions, reducing the fluid supersaturations, which in turn would prevent crystal nucleation and/or growth. The protein(s) can form a protected environment around the crystal nucleus, sequestering it and thus preventing crystal growth, or stabilizing the nucleus, protecting it from the external environment. The protein(s) may bind Ca2 and/or Pi ions, forming a surface that resembles the apatite surface. In this manner, the protein serves as an epitaxial (similar surface) nucleator, thereby providing a surface for the start of nucleation. The protein(s) may also bind to one or more faces of the growing crystal because its side chains match positions in the lattice, thereby blocking growth in one or more directions or even blocking growth beyond a specific size. The protein(s) might also bind to other proteins, changing their conformation and their ability to affect crystal nucleation and growth according to the pathways described earlier, or might associate with cells resulting in a change in the extracellular Ca P concentration, or pH. The elegant ultrastructural studies of Addadi and colleagues that combined X-ray crystallographic and electron microscopic techniques provide illustrations for each of these mechanisms for the formation of larger crystals of

157

CHAPTER 4 The Biochemistry of Bone

FIGURE 25

Cell-mediated matrix mineralization in developing bone. Early mineralization in chick bone. Electron micrograph showing a 17-day-old embryonic tibia, stained with uranyl acetate and lead citrate. Mineral clusters (C) outside the osteoblast (OB) are associated with collagen (thin arrows) and extracellular matrix vesicles (inset). Empty vesicles (thick arrows) as well as vesicles with mineral are seen. Courtesy of Dr. Steven B. Doty.

calcium carbonates, calcium sulfates, brushite, and octacalcium phosphate [163,1022,1023]. For example, fibronectin has been shown to bind to the ionic surfaces of calcite that did not include water molecules, but does not bind at all to brushite whose surfaces all have bound water [1024]. The acidic macromolecules from sea animals have been shown to determine the shape of calcite crystals [1025]. Cells have been shown to interact with specific faces on such crystals in the presence and absence of RGD-containing macromolecules [1026]. Scanning electron micrographs have similarly been used to identify the binding sites for polyaspartic acid, mollusk shell proteins, and rat dentin phosphoprotein on the surface of octacalcium phosphate [1024]. There are also examples of each of these mechanisms from solution studies of apatite formation. For these, to date, there is no direct evidence of the exact nature of the protein – mineral interaction. Studies of the effects of bone matrix proteins on apatite formation include studies in which preformed seed crystals are added to Ca P solutions, and the rate of crystal growth is determined at fixed Ca P and fixed pH [21] or variable Ca P OH [1021].

Other studies have looked at the formation (nucleation and growth) of apatite from solutions in the presence of insoluble proteins, proteins immobilized on polyanionic beads [1027], or proteins in solution [462]. Diffusion studies in which the protein is held within an agarose [664], silicate [712], or denatured collagen [202] gel have also provided insight into apatite nucleation and growth. From such studies, one can find examples of the mechanisms listed earlier. It should be emphasized, however, that a protein, because of its affinity for apatite, may, in low concentrations, act as a nucleator and in higher concentrations serve to regulate crystal growth. The multifunctional roles of each of the bone matrix proteins should be apparent from the previous sections of this review. The extracellular matrix vesicles (Fig. 25, inset, discussed later) and their component lipids may facilitate Ca and P accumulation, while shielding the apatite nucleus. As reviewed in detail elsewhere [1028], illustration of this behavior in vesicles is seen in the iron oxide forming bacteria and in model liposomes. In model liposomes, where Ca2 accumulation is facilitated by an ionophore, initial mineral

158

GOKHALE, BOSKEY, AND ROBEY

crystals form inside the liposomes in association with the liposome membrane where they eventually grow and puncture the liposome membrane and become exposed to the external solution [1029 – 1032]. Proteins that can prevent apatite nucleation by chelating Ca2 include aggrecan and its associated proteoglycan aggregates [200]. This may be more complicated in vivo where free Ca2 levels may be affected by Donnan equilibrium effects. However, it has been shown that phosphate can then cause the release of this Ca2, facilitating the mineralization process [664,1028]. An example of a protein that appears to block crystal growth by binding to a specific surface or surfaces is osteopontin, which binds with high specificity and decreases the length of the crystals formed in the collagen-gel system [663]. This is in contrast to the dentin phosphoprotein phosphophoryn, which blocks secondary nucleation by binding to growth sites [667]. Immobilized albumin, osteocalcin, and osteonectin appear to act as nucleators [1027], although evidence shows that the beads on which these proteins are immobilized are themselves nucleators. Fibronectin has also been shown to be capable of nucleating apatite in solution [577]. The only bone/dentin specific matrix proteins shown to act as nucleators are bone sialoprotein [665,712,1033], biglycan [1034], and dentin matrix protein 1 (Boskey, unpublished data). When stripped of its associated matrix proteins, collagen itself, does not nucleate apatite, but when its associated phosphoproteins are present, apatite nucleation on collagen occurs in a dose-dependent manner directly related to the extent of protein phosphorylation [124,660]. Collagen peptides containing the carboxylate region, but not the Nterminal domain can also nucleate apatite in solution [1035].The complexed acidic phospholipids and their associated proteolipids [848] and lipoproteins [851] also nucleate apatite formation in solution [851,1036] and when implanted in a cell-free environment in vivo [866].

B. Cell-Regulated Mineralization In each of the just mentioned solution-mediated studies, hydroxyapatite formation was performed in the absence of cells. However, it is the cells that regulate bone mineralization by producing and secreting the appropriate extracellular matrix components, regulating their interactions, and controlling the flux of Ca, P, and OH ions in their extracellular environments. One must look to cell-mediated studies to understand the pathways of matrix mineralization in vitro [reviewed in 1037]. Unfortunately, most of these cellmediated studies have been concerned with understanding the origin and differentiation of the bone cells rather than with the mineralization process per se. However, from analyses of the cell culture systems, one can learn much about mineralization. Another complicating issue with the

cell culture studies is that most rely on 10 mM -glycerolphosphate, an excellent substrate for alkaline phosphatase, as a source of inorganic phosphate. The hydrolysis of 10 mM -glycerophosphate results in inorganic phosphate concentrations two to five times higher than physiologic levels, in turn producing a medium Ca P product much higher (10 mM2) than physiologic (2 – 5 mM2). In such high Ca P medium, larger, nonoriented mineral crystals distinct from those that exist in situ may be formed [1010,1038]. Organ culture for the study of bone formation was first described by Dame Honor Fell in 1925 [1039]. Similar studies of cultured limb rudiments have been used more recently to show stages of differentiation and the effects of a variety of hormones and growth factors on bone formation [1010,1040]. Tenenbaum et al. used an inverted folded chick periosteum model as their organ culture system [1041,1042] with similar results. The limitation of the organ culture systems for the study of the mineralization process is the variety of cell types present. For that reason, isolated osteoblast cultures or cultures of cells that differentiate into osteoblasts and form a mineralized matrix may be preferable. Based on the system described by Friedenstein et al. [1043], Owen [1044], Ashton et al. [1045], and Benayahu et al. [1046], many investigators have stimulated marrow stem cell cultures to differentiate into osteoblasts [reviewed in 1047]. In these, and in isolated osteoblast systems, steroids such as dexamethasone [1048 – 1050], retinoic acid [383], BMP [1051], and -glycerophosphate [1052] are thought to enhance this differentiation. Osteoblasts isolated from fetal tissues, e.g., fetal rat calvariae, are commonly used to study osteoblasts, as are tumor-derived osteoblast-like cells. Examples of these include ROS17/2.8 [6], MG-63 [1053], SaOS [1054], and MC3T3-E1 [1055] cell lines. Methods have also been developed for isolating and culturing cells from more mature human (and animal) bones [9] and immortalized cell lines [1056 – 1058]. As techniques emerge, there will undoubtedly be an increase in the number of cell culture model systems that will exhibit several phenotypic traits. However, it will have to be determined by stringent criteria (physiological matrix mineralization verified at the EM level) whether these model systems will be useful in studying late stages of bone formation (i.e., matrix mineralization). Rat calvarial cells plated at low density yield isolated colonies, some of which are “osteoblast like” based on morphology, matrix composition, alkaline phosphatase expression, and mineralization [525,1059,1060]. In these osteoblast nodules/colonies, active type I collagen expression follows cell proliferation and then is reduced to baseline levels, whereas alkaline phosphatase activity and expression, osteopontin, collagenase, and osteocalcin expression commence after this. In fact, in such cultures, osteocalcin expression is highly correlated with mineral accretion

CHAPTER 4 The Biochemistry of Bone

[814]. BSP, osteopontin, and osteocalcin expression continue to increase as mineralization proceeds [10]. Mineral deposition in such cultures has been characterized by electron microscopy and wide angle X-ray diffraction [1061 – 1063] and shown to contain hydroxyapatite crystals, although not always oriented in the same direction as the collagen fibrils. These types of cell and organ culture systems have allowed the definition of the proposed sequential events involved in the recruitment and proliferation of osteoprogenitor cells and their differentiation into osteoblasts, and the temporal changes in the expression of bone matrix proteins [10]. Consistently, in each of the different model systems [9,11,608,814,1064,1065], osteopontin, collagen, osteonectin, and BSP expression have been shown to precede mineral deposition, whereas osteocalcin expression occurs after mineral deposition has commenced [10]. This pattern is consistent with the functions suggested for these proteins in this chapter, and although such cultures have not yet defined which of these proteins are necessary for mineralization, they did shed some light on the debate concerning the site of initial calcification in bone. In the 1960s the identification of extracellular matrix vesicles (MV) [1066,1067] as the location of the first mineral deposits in a variety of mineralizing tissues, including membranous bone [1068] and calcifying cartilage [1066,1067,1069], led to the suggestion that these membrane-bound bodies were the site of initial calcification in these tissues. Later studies, reviewed by Anderson [1070,1071] and Wuthier et al. [1072] revealed that these bodies were enriched in enzymes associated with mineralization, among them alkaline phosphatase, ATPase, and proteoglycan-degrading enzymes. The MV membranes contained proteolipids capable of facilitating Ca ion transport into the vesicles, and also had matrix collagens (type II, IX, and X) associated with them [1073]. Tissue slices, suspensions of isolated vesicles, and liposome models of matrix vesicles, in solutions of calcium and phosphate, each led to the deposition of apatite. The mechanism seemed to involve the transport of Ca ions into a vesicle that already contained a high concentration of phosphate ions [1073], formation and growth of vesicle associated mineral crystals, and rupture of the vesicle and the spread of larger mineral crystals into the extravesicular environment. As discussed earlier, the vesicle in solution served the purpose of providing a protected/stabilizing environment for the formation of the first crystals. When inhibitors of mineralization were included with the vesicles, mineralization was facilitated over vesicle-free controls [1074]. Vesicles also serve to provide enzymes that can modify the extracellular matrix, facilitating the proliferation of crystals. Extracellular matrix vesicles have been reported to be the site of initial mineralization in bone and cartilage and organ culture systems as described earlier. In such systems,

159 the mineral-associated vesicles are seen prior to the appearance of bulk mineral. One of the major issues in the debate between those that believe that collagen-based matrix is the initial site of mineralization and those who believe mineralization starts in vesicles concerns how the mineral would travel from the vesicles to its highly aligned/oriented sites on the collagen. Culture data support the matrix vesicle theory but do not prove a link between vesicle and collagenmediated mineralization. An association between vesicles and collagen suggests such a link may exist [1075]. The application of computerized axial tomographs to high-resolution electron micrographs by Landis has provided some explanations. Looking at the highly oriented calcified turkey tendon [159], mineral was seen in vesicles at sites different from mineral associated with the collagen. However, at certain sites, the mineral could be seen branching (showing the link between vesicle and collagen associated mineral). Thus when viewed in three dimensions, it appeared that the mineral does not “swim” but rather grows out radially, extending to the collagen fibril. The question remains whether this mineral associates with mineral already in the collagen or whether mineral from the vesicle manages to grow out to specific sites along the collagen fibril. From the simplistic point of view, the former seems more likely, as annealing of crystals does take place. To date, there is no evidence that collagen-based mineralization cannot occur in the absence of vesicles. Further, because mineralization starts at numerous discrete sites within the collagen fibrils, it has been argued that there would not be sufficient space for matrix vesicles to serve as so many distinct nuclei [19]. In contrast, from solution-based data, it is clear that there are matrix proteins associated with the collagen that can act as nucleators, implying that the latter mechanism may predominate. Collagen fibrils depleted of matrix proteins are not mineralized readily in solution. Collagen mineralization as seen at the EM level starts at specific sites, the so-called e bands, where several matrix proteins are known to localize [125]. Some of the matrix proteins in these sites may shield others, preventing nucleation. Others associated with the collagen may serve as nucleators, as described earlier, allowing the oriented deposition of mineral crystals. The growth of these crystals may also become limited by the space available in the collagen fibrils and by the adsorption of other matrix proteins. From the material presented in this review, we can suggest that (a) decorin, which trims the collagen fibrils and disappears from the e bands in mineralized fibrils [117], matrix gla-protein, and 2-HS glycoprotein are among the shielding agents, (b) bone sialoprotein and other phosphorylated proteins, which are closely associated with the collagen hole zones, serve as nucleators, and (c) several of the other matrix proteins produced by osteoblasts may play a role in regulating crystal size.

160

GOKHALE, BOSKEY, AND ROBEY

Acknowledgments The authors thank Drs. Paolo Bianco, Wojciech J. Grzesik, Neal S. Fedarko, and Steven B. Doty for providing photographic materials and Luz Ingles for secretarial assistance. Dr. Boskey’s research as discussed in this review was supported by NIH Grants DE04141, AR037661, and AR41325.

References 1. P. Gehron Robey, Normal bone formation – structure. In “Bone Formation and Repair” (C. T. Brighton, G. Friedlander, and J. M. Lan, eds.), pp. 3 – 12, AAOS, Rosemont, IL, 1994. 2. J. D. Termine, A. B. Belcourt, P. J. Christner, K. M. Conn, and M. U. Nylen, Properties of dissociatively extracted fetal tooth matrix proteins. I. Principal molecular species in developing bovine enamel. J. Biol. Chem. 255, 9760 (1980). 3. J. D. Termine, A. B. Belcourt, K. M. Conn, and H. K. Kleinman, Mineral and collagen-binding proteins of fetal calf bone. J. Biol. Chem. 256, 10403 (1981). 4. W. Peck, S. Birge, and C. Fedak, Bone cells: Biochemical and biological studies after enzymatic isolation. Science 146, 1476 (1964). 5. G. L. Wong and P. V. Cohn, Target cells in bone for parahormone and calcitonin are different: Enrichment for each cell type by sequential digestion of mouse calvaria and selective adhesion to polymeric surfaces. Proc. Natl. Acad. Sci. USA. 72,(8); 3167 – 3171 (1975). 6. R. J. Majeska, S. B. Rodan, and G. A. Rodan, Parathyroid hormone-responsive clonal cell lines from rat osteosarcoma. Endocrinology 107, 1494 (1980). 7. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 (1987). 8. J. N. Beresford, J. A. Gallagher, J. W. Poser, and R. G. G. Russell, Production of osteocalcin by human bone cells in vitro. Effects of 1,25(OH)2D3, 24,25(OH)2D3, parathyroid hormone, and glucocorticoids. Metab. Bone Dis. Relat. Res. 5, 229 (1984). 9. P. Gehron Robey and J. D. Termine, Human bone cells in vitro. Calcif. Tissue Int. 37, 453 – 460 (1985). 10. G. S. Stein and J. B. Lian, Molecular mechanisms mediating developmental and hormone-regulated expression of genes in osteoblasts: An integrated relationship of cell growth and development, In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 47 – 95, Academic Press, New York, 1993. 11. P. G. Strauss, E. I. Closs, J. Schmidt, and V. Erfle, Gene expression during osteogenic differentiation in mandibular condyles in vitro. J. Cell Biol. 110, 1369 – 1378 (1990). 12. P. Gehron Robey, P. Bianco, and J. D. Termine, The cellular biology and molecular biochemistry of bone formation. In “Disorders of Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 241 – 263. Raven Press, New York, 1992. 13. A. Caplan, The mesengenic process. Clin. Plast. Surg. 21, 429 – 435 (1994). 14. F. Liu, L. Malaval, A. K. Gupta, and J. E. Aubin, Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast differentiation: Polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol. 166, 220 – 234 (1994). 15. S. Lee and C. L. Davidson, The role of collagen in the elastic properties of calcified tissues. J. Biomech. 10, 473 – 486 (1977). 16. A. L. Boskey, Mineralization, structure, and function of bone. In “Dynamics of Bone and Cartilage Metabolism” (M. J. Seibel, S. P.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28. 29. 30.

31.

32.

33.

Robins, and J. P. Bilzekian, eds.), pp.153 – 162. Academic Press, 1999. R. A. Robinson and M. L.Watson, Crystal-collagen relationships in the bone as observed in the electron microscope. III. Crystal and collagen morphology as a function of age. Ann. NY. Acad. Sci. 60, 596 – 628 (1955). A. L. Arsenault and M. D. Grynpas, Crystals in calcified epiphyseal cartilage and cortical bone of the rat. Calcif. Tissue Int. 43, 219 – 225 (1988). J. Christoffersen and W. J. Landis, A contribution with review to the description of mineralization of bone and other calcified tissues in vivo. Anat. Rec. 230, 435 – 450 (1991). A. S. Posner and F. Betts, Molecular control of tissue mineralization. In “The Chemistry and Biology of Mineralized Connective Tissues” (A. Veis, ed.), pp. 257 – 271, Elsevier, Amsterdam, 1981. E. D. Eanes, Dynamic aspects of apatite phases of mineralized tissues, model studies. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 213 – 221, Ebsco Media, Birmingham, AL, 1985. J. Moradian-Oldak, F. Frolow, L. Addadi, and S. Weiner, Interactions between acidic matrix macromolecules and calcium phosphate ester crystals: Relevance to carbonate apatite formation in biomineralization, Proc. Roy. Soc. Lond. B. 247, 47 – 55 (1992). R. T. Bailey and C. Holt, Fourier transform infrared spectroscopy and characterisation of biological calcium phosphates. In “Calcified Tissue,” (D. W. L. Hukins, ed.), chapt. 5. Macmillan Press, Houndsmills, England, 1989. D. D. Lee, W. J. Landis, and M. J. Glimcher, The solid, calciumphopshate mineral phases in embryonic chick bone characterized by high-voltage electron microscopy. J Bone Mineral Res. 1, 425 – 432 (1986). E. D. Eanes, L. Powers, and J. L. Costa, Extended X-ray absorption fine structure (EXAFS) studies on calcium in crystalline and amorphous solids of biological interest. Cell Calcium 2, 251 – 262 (1981). W. P. Aue, A. H. Roufosse, M. J. Glimcher, and R. G. Griffin, Solid state phosphorus nuclear magnetic resonance study of synthetic solid phase of calcium phosphate; potential models of bone mineral. Biochemistry 23, 6110 – 6114 (1984). D. W. L. Hukins, A. J. Cox, and J. E. Harries, EXAFS characterisation of poorly crystalline deposits from biological systems in the presence of highly crystalline material. J. Physique 48, 1181 – 1184 (1986). A. S. Posner, Crystal chemistry of bone mineral. Physiol. Rev. 40, 760 – 792 (1969). R. A. Young, Biological apatite vs hydroxyapatite at the atomic level. Clin. Orthop. Rel. Res. 113, 249 – 262 (1975). L. Cohen and R. Kitzes, Infrared spectroscopy and magnesium content of bone mineral in osteoporotic women. Israel J. Med. Sci. 17, 1123 – 1125 (1981). C. Rey, M. Shimizu, B. Collins, and M. J. Glimcher, Resolution-enhanced fourier transform infrared spectroscopy study of the environment of phosphate ions in early deposits of a solid phase of calcium-phosphate in bone and enamel, and their evolution with age. I. Investigations in the nu 4 PO4 domain. Calcif. Tissue Int. 46, 384 – 394 (1990). C. Rey, M. Shimizu, B. Collins, and M. J. Glimcher, Resolution-enhanced Fourier transform infrared spectroscopy study of the environment of phosphate ion in the early deposits of a solid phase of calcium phosphate in bone and enamel and their evolution with age. 2. Investigations in the nu3PO4 domain. Calcif. Tissue Int. 49, 383 – 388 (1991). C. Rey, V. Renugopalakrishnan, M. Shimizu, B. Collins, and M. J. Glimcher, A resolution-enhanced Fourier transform infrared

161

CHAPTER 4 The Biochemistry of Bone

34.

35.

36.

37. 38.

39.

40.

41.

42. 43. 44. 45. 46.

47.

48.

49.

50.

51.

52.

53.

54.

spectroscopic study of the environment of the CO32 ion in the mineral phase of enamel during its formation and maturation. Calcif. Tissue Int. 49, 2259 – 2268 (1991). J. D. Termine and D. R. Lundy, Hydroxide and carbonate in rat bone mineral and its synthetic analogues. Calcif. Tissue Res. 13, 73 – 82 (1973). A. J. Kahn and N. C. Partridge, Bone resorption in vivo. In “Bone” (B. K. Hall, ed.), Vol. 2, pp. 199 – 140. CRC Press, Boca Raton, FL, 1990. A. H. Roufosse, W. J. Landis, W. K. Sabine, and M. G. Glimcher, Identification of brushite in newly deposited bone mineral from embryonic chiecks. J. Ultrastruct. Res. 68, 235 – 255 (1979). W. E. Brown, Crystal growth of bone mineral. Clin. Orthop. Rel. Res. 44, 205 – 220 (1966). A. S. Posner and F. Betts, Synthetic amorphous calcium phosphate and its relation to bone mineral structure. Accts. Chem. Res. 8, 273 – 281 (1975). F. C. M. Driessens, J. W. E. Van Dijk, and J. M. P. M. Borggreven, Biological calcium phosphates and their role in the physiology of bone and dental tissues. Calcif. Tissue Res. 26, 127 – 136 (1978). Y. Wu, M. J. Glimcher, C. Rey, and J. L. Ackerman, A unique protomate phosphate group in bone mineral not present in synthetic calcium phosphates: Identification by phosphorus-31 solid state NMR spectroscopy. J. Mol. Biol. 244, 423 – 435 (1994). K. A. Piez, Molecular and aggregate structures of collagens. In “Extracellular Matrix Biochemistry” (K. A. Piez and A. H. Reddi, eds.), pp. 1 – 40. Elsevier, New York, 1984. P. P. Fietzek and K. Kuhn, The primary structure of collagen, Int. Rev. Connect. Tissue Res. 7, 1 – 60 (1976). G. R. Martin, R. Timpl, P. K. Muller, and K. Kuhn, The genetically distinct collagens. Trends Biochem. Sci. 251, 285 – 287 (1985). K. S. E. Cheah, Collagen genes and inherited connective tissue disease. Biochem. J. 229, 287 – 303 (1985). J. C. Brown and R. Timpl, The collagen superfamily. Int. Arch. Allergy. Immunol. 107(4), 484 – 490 (1995). D. A. D. Parry, The molecular and fibrillar structure of collagen and its relationship to the mechanical properties of connective tissue. Biophys. Chem. 29, 195 – 209 (1988). D. B. Haniford and D. E. Pulleybank, Transition of a cloned d(AT)n-d(AT)n tract to a cruciform in vivo. Nucleic Acid Res. 13, 4343 – 4363 (1985). M. J. McLean and R. D. Wells, The role of DNA sequence in the formation of Z-DNA versus cruciforms in plasmids. J. Biol. Chem. 263, 7370 – 7377 (1988). C. V. Heath, R. M. Denome, and C. N. Cole, Spatial constraints on polyadenylation signal function. J. Biol. Chem. 265, 9098 – 9104 (1990). L. H. Naylor and E. M. Clark, d(TG)n-d(CA)n sequences upstream of the rat prolactin gene form Z-DNA and inhibit gene transcription. Nucleic Acids Res. 18, 1595 – 1601 (1990). M.-L. Chu, W. De Wet, M. Bernard, J. F. Ding, M. Morabito, J. Myers, C. Williams, and F. Ramirez, Human pro allpha 1(I) collagen gene structure reveals evolutionary conservation of a pattern of introns and exons. Nature 310, 337 – 340 (1984). F. Ramirez, M. Bernard, M.-L. Chu, L. Dickson, F. Sangiorgi, D. Weil, W. de Wet, C. Junien, and M. Sobel, Isolation and characterization of the human fibrillar collagen genes. Ann. N.Y. Acad. Sci. 460, 117 – 129 (1985). W. deWet, M. Bernard, V. Benson-Chanda, M.-L. Chu, L. Dickson, D. Weil, and F. Ramirez, Organization of the human pro-alpha 2(I) collagen gene. J. Biol. Chem. 262, 16032 – 16036 (1987). H. Kuivaniemi, G. Tromp, M. L. Chu, and D. J. Prockop, Structure of a full-length cDNA clone for the prepro2(I) chain of human type I procollagen. Comparison with the chicken gene confirms

55.

56. 57. 58.

59. 60.

61. 62.

63.

64.

65.

66.

67.

68.

69.

70.

71. 72.

73.

74.

unusual patterns of gene conservation. Biochem. J. 252, 633 – 640 (1988). S. Strickland, J. Huarte, D. Belin, A. Vassalli, R. J. Rickles, and J. D. Vassalli, Antisense RNA directed against the 3 noncoding region prevents dormant mRNA activation in mouse oocytes. Science 241, 680 – 684 (1988). J. L. Manley, Polyadenylation of mRNA precursors. Biochim. Biophys. Acta 6950(1), 1 – 12 (1988). W. M. Krojewska, Regulation of transcription in eukaryotes by DNA-binding proteins. Int. J. Biochem. 24, 1885 – 1899 (1992). C. Murre, G. Bain, M. A. van Dijk , I. Engel, B. A. Furnari, M. E. Massari, J. R. Matthews, M. W. Quong, R. R. Rivera, and M. H. Stuiver, Structure and function of helix-loop-helix proteins. Biochim. Biophys. Acta. 1218, 129 – 135 (1994). S. E. Phillips, Built by association: Structue and funtion of helixloop-helix DNA-binding proteins. Structure 2, 1 – 4 (1994). V. J. Dwarki, M. Montminy, and I. M. Verma, Both the basic region and the “leucine ziper” domain of the cyclic AMP response element binding (CREB) protein are essential for transcriptional activation. EMBO J. 9, 225 – 232 (1990). D. Landsman and M. Bustin, A signature for the HMG-1 box DNAbinding proteins. Bioessays 15, 539 – 546 (1993). B. L. Vallee and D. S. Auld, Functional zinc-binding motifs in enzymes and DNA-binding proteins. Faraday Discuss 93, 47 – 65 (1992). M. L. Chu, W. deWet, M. Bernard, and F. Ramirez, Fine structural analysis of the human pro collagen gene: Promoter structure, AluI repeats and polymorphism transcripts. J. Biol. Chem. 260, 2315 – 2320 (1985). M. H. Finer, S. Aho, and L. Gerstenfeld, Unusual DNA sequences located within the promoter region and the first intron of the chicken proalpha1(I) collagen gene. J. Biol. Chem. 262, 13323 (1987). A. Lichtler, M. L. Stover, and J. Angilly, Isolation and characterization of the rat 1(I) collagen promoter: Regulation by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 264, 3072 (1989). A. Schmidt, P. Rossi, and B. de Crombrugghe, Transcriptional control of the mouse 2(I) collagen gene: Functional deletion analysis of the promoter and evidence for cell-specific expression. Mol. Cell. Biol. 6, 347 – 354 (1986). G. Vogeli, H. Ohkubo, M. E. Sobel, Y. Yamada, I. Pastan, and B. de Crombrugghe, Structure of the promoter for chicken 2 type I collagen gene, Proc. Natl. Acad. Sci. USA 78, 5534 – 5538 (1981). G. T. Merlino, C. McKeon, B. de Crombrugghe et al., Regulation of the expression of genes encoding types I, II and III collagen during chick embryonic development. J. Biol. Chem. 258, 10041 (1983). C. M. S. Rossouw, W. P. Vergeer, S. J. du Plooy, M. P. Bernard, F. Ramirez, and W. J. de Wet, DNA sequences in the first intron of the human pro-1(I) collagen gene enhance transcription. J. Biol. Chem. 262, 15151 – 15157 (1987). P. Bornstein, J. McKay, and J. K. Morishima, Regulatory elements in the first intron contribute to transcriptional control of the human 1(I) collagen gene. Proc. Natl. Acad. Sci. USA 82, 1321 (1987). E. Vuorio and B. de Crombrugghe, The family of collagen genes. Annu. Rev. Biochm. 59, 837 – 872 (1990). A. Lichtler, M. L. Stover, J. Angilly, B. Kream, and D. W. Rowe, Isolation and characterization of the rat alpha1(I) collagen promoter. J. Biol. Chem. 269, 16518 (1994). R. D’Souza, K. Niederreither, and B. de Crombrugghe, Osteoblastspecific expression of the alpha 2(I) collagen promoter in transgenic mice: correlation with the distribution of TGF-1. J. Bone Miner. Res. 8, 1127 – 1136 (1993). W. deWet, M. L. Chu, and D. J. Prockop, The mRNAs for the pro(I) and pro2(I) chains of type I procollagen are translated at the same

162

75.

76. 77. 78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

GOKHALE, BOSKEY, AND ROBEY rate in normal human fibroblasts and fibroblasts from two variants of osteogenesis imperfecta with altered steady state rations of the two mRNAs. J. Biol. Chem. 258, 14385 – 14389 (1983). J. Vuust, M. E. Sobel, and G. R. Martin, Regulation of type I collagen synthesis: total pro1(I) and pro2(I) mRNAs are maintained in a 2:1 ratio under varying rates of collagen synthesis. Eur. J. Biochem. 151, 449 – 453 (1985). P. Bornstein, D. Horlein, and J. McPherson, “Myelofibrosis and the Biology of Connective Tissue,” p. 61 A. R. Liss, New York, 1984. P. Bornstein and H. Sage, Structurally distinct collagen types. Annu. Rev. Biochem. 49, 957 – 1003 (1980). P. H. Krebsbach, J. R. Harrison, A. C. Lichtler, C. O. Woody, D. W. Rowe, and B. E. Kream, Transgenic expression of COL1A1-chloramphenicol acetyltransferase fusion genes in bone: Differential utilization of promoter elements in vivo and in cultured cells. Mol. Cell. Biol. 13, 5168 – 5174 (1993). D. Pavlin, A. C. Lichtler, A. Bedalov, B. E. Kream, J. R. Harrison, H. F. Thomas, G. A. Gronowicz, S. H. Clark, C. O. Woody, and D. W. Rowe, Differential utilization of regulatory domains within the 1(I) collagen promoter in osseous and fibroblastic cells. J. Cell Biol. 116, 227 – 236 (1992). Y. Shibata, Y. Abiko, K. Goto, Y. Moriya, and H. Takiguchi, Heparin stimulates the collagen synthesis in mineralized cultures of the osteoblast-like cell line, MC3T3-E1. Biochem. Int. 28, 335 – 344 (1992). H. C. Tenenbaum, H. Limeback, C. A. McCulloch, H. Mamujee, B. Sukhu, and M. Torontali, Osteogenic phase-specific co-regulation of collagen synthesis and mineralization by beta-glycerophosphate in chick periosteal cultures. Bone 13, 129 – 138 (1992). K. Nohtomi, K. Sato, K. Shizume, K. Yamazaki, H. Demura, K. Hosada, Y. Murata, and H. Seo, Stimulation of interleukin-4 of cell proliferation and mRNA expression of alkaline phosphatase and collagen type I in human osteoblast-like cells of trabecular bone. Bone Miner. 27, 69 – 79 (1994). R. S. Bockman, P. T. Guidon, Jr., L. C. Pan, R. Salvatori, and A. Kawaguchi, Gallium nitrate increases type I collagen and fibronectin mRNA and collagen protein levels in bone and fibroblast cells. J. Cell Biochem. 52, 396 – 403 (1993). L. G. Raisz, P. M. Fall, D. N. Petersen, A. Lichtler, and B. E. Kream, Prostaglandin E2 inhibits alpha 1(I)procollagen gene transcription and promoter activity in the immortalized rat osteoblastic clonal cell line. Mol. Endocrinol. 7, 17 – 22 (1993). D. Pavlin, A. Bedalov, M. S. Kronenberg, B. E. Kream, D. W. Rowe, C. L. Smith, J. W. Pike, and A. C. Lichtler, Analysis of regulatory regions in the COL1A1 gene responsible for 1,25-dihydroxyvitamin D3-mediated transcriptional repression in osteoblastic cells. J. Cell Biochem. 56, 490 – 501 (1994). L. G. Raisz, P. M. Fall, B. Y. Gabbitas, T. L. McCarthy, B. E. Cream, and E. Canalis, Effects of prostaglandin E2 on bone formation in cultured fetal rat calvariae: Role of insulin-like growth factor-I. Endocrinology 133, 1504 – 1510 (1993). B. E. Kream, D. LaFrancis, D. N. Peterson, C. Woody, S. Clark, D. W. Rowe, and A. Lichtler, Parathyroid hormone represses ? 1(I) collagen promoter activity in cultured calvariae from neonatal transgenic mice. Mol. Endocrinol. 7, 399 – 408 (1993). R. Hata, H. Hori, Y. Nagai, S. Tanaka, M. Kondo, M. Hiramatsu, N. Utsumi, and M. Kumegawa, Selective inhibition of type I collagen synthesis in osteoblastic cells by epidermal growth factor. Endocrinology 115, 867 – 876 (1984). M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type I collagen gene expressionin osteblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). P. Van Vlasselaer, B. Borremans, U. van Gorp, J. R. Dasch, and R. De Waal-Malefyt, Interleukin 10 inhibits transforming growth

91. 92.

93.

94. 95. 96.

97.

98. 99. 100.

101.

102.

103.

104. 105.

106.

107.

108.

109. 110.

111.

112.

factor-beta (TGF-beta) synthesis required for osteogenic commitment of mouse bone marrow cells. J. Cell Biol. 124, 569 – 577 (1994). R. F. Klein and K. M. Wiren, Regulation of osteoblastic gene expression by lead. Endocrinology 132, 2531 – 2537 (1993). J. W. Dietrich, E. M. Canalis, D. M. Maina, and L. G. Raisz, Effects of glucocorticoids on fetal rat bone collagen synthesis in vitro. Endocrinology 104, 715 – 721 (1979). E. Canalis, Effect of glucocorticoids on type I collagen synthesis, alkaline phosphatase activity, and deoxyribonucleic acid content in cultured rat calvariae, Endocrinology 112, 931 – 939 (1983). A. Newman, RNA splicing: Activity in the spliceosome. Curr. Biol. 4, 462 – 464 (1994). S. M. Berget, Exon recognition in vertebrate splicing. J. Biol. Chem. 270, 2411 – 2414 (1995). M. J. Moore and P. A. Sharp, Evidence for two active sites in the spliceosome provided by stereochemistry of pre-mRNA splicing. Nature 365, 364 – 368 (1993). Y. Yamada, M. Mudryj, and B. de Crombrugghe, A uniquely conserved regulatory signal is found around the translation initiation site in three different collagen genes. J. Biol. Chem. 258, 14914 (1983). J. H. Fessler and L. I. Fessler, Biosynthesis of procollagen. Annu. Rev. Biochem. 47, 129 – 162 (1978). D. J. Prockop, Mutations in collagen genes as a cause of connective-tissue diseases. N. Engl. J. Med. 326, 540 – 546 (1992). G. N. Ramachandran and C. Ramakrishnan, Molecular structure. In “Biochemistry of Collagen” (G. N. Ramachandran and A. H. Reddi, eds.), pp. 45 – 84. Plenum Press, New York, 1976. M. Yamauchi, E. P. Katz, K. Otsubo, K. Teraoka, and G. L. Mechanic, Cross-linking and stereospecific structure of collagen in mineralized skeletal tissues. Connect. Tissue Res. 21, 159 – 169 (1989). K. Kuhn, The classical collagens: Type I, II and III. In “Structure and Function of Collagen Types” (R. Mayne and R. E. Burgeson, eds.), pp. 1 – 42, Academic Press, Orlando, 1987. J. Bella, M. Eaton, B. Brodsky, and H. M. Berman, Crystal and molecular structure of a collagen-like peptide at 1.9 angstroms. Science 266, 75 – 81 (1994). R. Kornfeld and S. Kornfeld, Comparative aspects of glycoprotein structure. Annu. Rev. Biochem. 45, 217 – 237 (1976). J. J. Sauk, K. Norris, J. M. Kerr, M. J. Sommerman, and M. F. Young, Diverse forms of stress result in changes in celular levels of osteonectin/SPARC witout altering mRNA levels in osteoligament cells. Calicif. Tissue Int. 49, 58 – 62 (1991). M. Weinstock and C. P. Leblond, Radioautographic visualization of the deposition of a phosphoprotein at the mineralization front in the dentin of the rat incisor. J. Cell Biol. 56, 838 – 845 (1973). L. Paglia, J. Wilczek, L. Diaz de Leon, G. R. Martin, D. Horlein, and P. Muller, Inhibition of procollagen cell-free synthesis by amino-terminal extension peptides. Biochemistry 18, 5030 – 5034 (1979). W. Schlumberger, M. Thie, H. Volmer et al., Binding and uptake of Col 1(I), a peptide capable of inhibiting collagen synthesis in fibroblasts. J. Cell Biol. 46, 244 (1988). M. Wiestner, S. Fischer, W. Dessau et al., Collagen types synthesized by isolated calvarium cells. Exp. Cell Res. 133, 115 (1981). C. H. Wu, C. B. Donovan, and G. Y. Wu, Evidence for pretranslational regulation of collagen synthesis by procollagen propeptides. J. Biol. Chem. 261, 10482 – 10484 (1989). R. S. Ayacock, R. Raghow, G. P. Striklin, J. M. Seyer, and A. H. Kang, Post transcriptional inhibition of collagen and fibronectin synthesis by a synthetic homolog of a portion of the carboxy-terminal propeptide of human type I collagen. J. Biol. Chem. 261, 14355 – 14360 (1986). J. D. Termine, A. B. Belcourt, K. M. Conn et al., Mineral and collagen-binding proteins of fetal calf bone. J. Biol. Chem. 256, 10403 (1981).

CHAPTER 4 The Biochemistry of Bone 113. L. W. Fisher, P. Gehron Robey, N. Tuross, A. S. Otsuka, D. A. Tepen, F. S. Esch, S. Shimasaki, and J. D. Termine, The Mr 24,000 phosphoprotein from developing bone is the N-terminal propeptide of the 1 chain of type I collagen. J. Biol. Chem. 262, 13457 – 13462 (1987). 114. D. R. Eyre, New biomarkers of bone resorption. J. Clin. Endocrinol. Metab. 74, 470A – 470C (1992). 115. R. A. Fleischmajer and J. S. Perlish, The role of the amino propeptide in collagen fibrillogenesis. Rheumatology 10, 91 (1986). 116. R. A. Fleischmajer, L. W. Fisher, E. D. MacDonald, L. Jacobs Jr., J. S. Perlish, and J. D. Termine, Decorin interacts with fibrillar collagen of embryonic and adult human skin. J. Struct. Biol. 106, 82 – 90 (1991). 117. J. E. Scott and M. Haigh, Proteoglycan type I collagen fibril interactions in bone and non-calcifying connective tissues. Biosci. Rep. 5, 71 – 81 (1985). 118. J. E. Scott, Proteoglycan:collagen interactions in connective tissues: Ultrastructural, biochemical, functional and evolutionary aspects. Int. J. Biol. Macromol. 13, 157 – 161 (1991). 119. J. E. Scott, Proteoglycan-fibrillar collagen interactions. Biochem. J. 252, 313 – 323 (1988). 120. E. P. Katz and S. T. Li, The intermolecular space of reconstituted collagen fibrils. J. Mol. Biol. 73, 351 – 369 (1973). 121. A. J. Hodge and J. A. Petruska, Recent studies with electron microscope on ordered aggregates of the tropocollagen macromolecule. In “Aspects of Protein Structure” (G. N. Ramachandran, ed.), pp. 289 – 300. Academic Press, New York, 1963. 122. A. L. Arsenault, Crystal-collagen relationships in calcified turkey leg tendons visualized by selected-area dark field electron microscopy. Calcif. Tissue Int. 43, 202 – 212 (1988). 123. M. E. Maitland and A. L. Arsenault, A correlation between the distribution of biological apatite and amino acid sequence of type I collagen. Calcif. Tissue Int. 48, 341 – 352 (1991). 124. M. Glimcher, Recent studies of the mineral phase in bone and its possible linkage to the organic matrix by protein-bound phosphate bonds. Phil. Trans. R. Soc. Lond. B. 304, 479 – 508 (1984). 125. W. Traub, A. Jodaikin, T. Arad, A. Veis, and B. Sabsay, Dentin phosphophoryn binding to collagen fibrils. Matrix 12, 197 – 201 (1992). 126. D. R. Eyre, M. A. Paz, and P. M. Gallop, Cross-linking in collagen and elastin. Annu. Rev. Biochem. 53, 717 – 748 (1984). 127. C. Niyibizi and D. R. Eyre, Identification of the cartilage alpha1(IX) chain in type V collagen from bovine bone. FEBS Lett. 242, 314 – 318 (1989). 128. T. Pihlajaniemi and M. Tammnen, The alpha1 chain of type XIII collagen: Polypeptide structure, alternative splicing and tissue distribution. In Structure, “Molecular Biology and Pathology of Collagen.” (R. Fleishmajer, B. R. Olsen, and K. Kuhn, eds.). New York Academy of Science. 129. M. van der Rest, Collagen of bone. In “Bone Matrix and Bone Specific Products” (B. K. Hall, ed.), pp. 187 – 239. CRC Press, Boca Raton, FL, 1991. 130. C. Niyibizi and D. R. Eyre, Bone type V collagen: Chain composition and location of a trypsin cleavage site. Connect. Tissue Res. 20, 247 – 250 (1989). 131. M. Yamagata, K. Yamada, S. S. Yamada, and T. Shinomura, The complete primary structure of type XII collagen shows a chimeric molecule with reiterated fibronectin type III motifs, von Willebrand factor A motifs, a domain homologous to a non-collagenous region of type IX collagen, and a short collagenous domain. J. Cell Biol. 115, 209 – 221 (1991). 132. S. J. Perkins, K. F. Smith, S. C. Williams, and P. I. Haris, The secondary structure of von Willebrand factor type A domains in factor B of human complement by Fourier transform infrared spectroscopy: Is occurrence in collagen types VI, VII, XII and XIV, the integrins and other proteins? J. Mol. Biol. 238, 104 – 119 (1994).

163 133. M. Koch, C. Bernasconi, and M. Chiquet, A major oligomeric fibroblast proteoglycan identified as a novel large form of type-XII collagen. Eur. J. Biochem. 207, 847 – 856 (1992). 134. S. P. Oh, C. M. Griffith, E. D. Hay, and B. R. Olsen, Tissue-specific expression of type XII collagen during mouse embryonic development. Dev. Dyn. 196, 37 – 46 (1993). 135. H. Nakahara, K. Watanabe, S. P. Sugrue, B. R. Olsen, and A. I. Caplan, Temporal and spatial distribution of type XII collagen in high cell density culture of periosteal-derived cells. Dev. Biol. 142, 481 – 485 (1991). 136. D. R. Keene, G. P. Lundstrum, M. P. Morris, D. W. Stoddard, and R. E. Burgeson, Two type XII-like collagens localize in the surface of banded collagen fibrils. J. Cell Biol. 113, 971 – 978 (1991). 137. J. Trueb and B. Trueb, Type XIV collagen is a variant of indulin. Eur. J. Biochem. 207, 549 – 557 (1993). 138. B. R. Olsen, The roles of collagen genes in skeletal development and morphogenesis. Experientia 51, 194 – 195 (1995). 139. S. A. Pinnell, R. Fox, and S. M. Krane, Human collagens, differences in glycosylated hydroxylysines in skin and bone. Biochim. Biophys. Acta 229, 119 (1971). 140. J. R. Segrest and L. W. Cunningham, Variations in human urinary O-hydroxylysyl glucoside levels and their relationship to collagen metabolism. J. Clin. Invest. 49, 1497 (1970). 141. L. Knott and A. J. Bailey, Collagen cross-links in mineralizing tissues: A review of their chemistry, function, and clinical relevance. Bone 22(3), 181 – 187 (1998). 142. A. L. Boskey, T. L. Wright, and R. D. Blank, Collagen and bone strength, J. Bone Miner. Res. 3, 330 – 335 (1999). 143. D. J. Prockop, Mutations that alter the primary structure of type I collagen: The periols of a system for generation large structures by the principle of nucleated growth. J. Biol. Chem. 265, 15349 – 15352 (1990). 144. R. J. Wenstrup, M. C. Willing, B. J. Starman, and P. H. Byers, Distinct biochemical phenotypes predict clinical severity in nonlethal variants of osteogenesis imperfecta. Am. J. Hum. Genet. 46, 975 – 982 (1990). 145. P. H. Byers, Inherited disorders of collagen gene structure and expression. Am. J. Hum. Genet. 34, 72 – 80 (1989). 146. P. H. Byers, Brittle bones – fragile molecules: Disorders of collagen gene structure and expression. Trends Genet. 6, 293 – 300 (1990). 147. H. Kuivaniemi, G. Tromp, and D. J. Prockop, Mutations in collagen genes: Causes of rare and some common diseases in humans. FASEB J. 5, 2052 – 2060 (1991). 148. J. Lohler, R. Timpl, and R. Jaenisch, Embryonic lethal mutation in mouse collagen I gene causes rupture of blood vessels and is associated with erythropoietic and mesenchymal cell death. Cell 38(2), 597 – 607 (1984). 149. K. Harbers, M. Juehan, H. Delius, and R. Jaenisch, Insertion of retrovirus into the first intron of 1(I) collagen gene leads to embryonic fetal mutation in mice. Proc. Natl. Acad. Sci. USA 81, 1504 – 1508 (1984). 150. A. Schniecke, K. Harbers, and R. Jaenisch, Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha1(I) collagen gene. Nature 304, 315 – 320 (1983). 151. A. Stacey, J. Bateman, T. Choi, T. Mascara, W. Cole, and R. Jaenisch, Perinatal lethal osteogenesis imperfecta in transgenic mice bearing an engineered mutant pro-1(I) collagen gene. Nature 332, 131 – 136 (1988). 152. J. Bonadio, T. L. Saunders, E. Tsai, S. A. Goldstein, J. MorrisWiman, L. Brinkley, D. F. Dolan, R. A. Altschuler, J. E. Hawkins, J. F. Bateman, T. Maresca, and R. Jaenisch, Transgenic mouse model of the mild dominant form of osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 87, 7145 – 7149 (1990). 153. S. D. Chipman, H. O. Sweet, D. J. McBride, M. T. Davidson, S. C. Marks Jr, A. R. Schuldiner, D. W. Rowe, R. J. Wenstrup, and J. R.

164

154.

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165. 166. 167.

168.

169.

170.

171.

172.

GOKHALE, BOSKEY, AND ROBEY Shapiro, Defective pro a 2 (I) collagen synthesis in a new recessive mutation in mice: A potential model for human osteogenesis imperfecta, Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993). J. Bonadio and S. A. Goldstein, Our understanding of inherited skeletal fragility and what this has taught us about bone structure and function. In “Cellular and Molecular Biology of Bone” (M Noda, ed.), pp. 169 – 189. Academic Press, San Diego, 1993. J. S. Khillan, A. E. Olsen, S. Kontusaari, B. Solov, and D. J. Prockop, Trangenic mice that express a mini-gene version of the human gene for type I procollagen (COLA1) developed a phenotype resembling a lethal form of osteogenesis imperfecta. J. Biol. Chem. 266(34), 23373 – 23379 (1991). W. de Wet, T. Pihlajaniemi, J. C. Myers, T. E. Kelly, and D. J. Prockop, Synthesis of a shortened pro-2 chain, and decreased synthesis of pro-a2 (I) chains in a proband with osteogenesis imperfecta. J. Biol. Chem. 258, 7721 – 7728 (1983). U. Vetter, E. D. Eanes, J. B. Kopp, J. D. Termine, and P. Gehron Robey, Changes in apatite crystal size in bones of patients with osteogenesis imperfecta. Calcif. Tissue Int. 49, 248 – 250 (1991). W. Traub, T. Arad, U. Vetter, and S. Weiner, Ultrastructural studies of bones from patients with osteogenesis imperfecta. Matrix Biol. 14, 337 – 345 (1994). P. Fratzl, O. Paris, K. Klaushofer, and W. J. Landis, Bone mineralization in an osteogenesis imperfecta mouse model studied by small-angle x-ray scattering. J. Clin. Invest. 97(2) 396 – 402 (1996). N. P. Camacho, L. Hou, T. R. Toledano, W. A. Ilg, C. F. Brayton, C. L. Raggio, L. Root, and A. L. Boskey, The material basis for reduced mechanical properties in oim mice bones. J. Bone. Miner. Res. 14(2), 264 – 272 (1999). W. Traub, T. Arad, and S. Weiner, Three-dimensional ordered distribution of crystals in turkey tendon collagen fibers. Proc. Natl. Acad. Sci. USA 86, 9822 – 9826 (1989). P. Sarathchandra, F. M. Pope, and S. Y. Ali, Morphometric analysis of type I collagen fibrils in the osteoid of osteogenesis imperfecta. Calcif. Tissue Int. 65(5), 390 – 395 (1999). L. Addadi and S. Weiner, Interaction between acidic proteins and crystals: Stereochemical requirements in biomineralization. Proc. Natl. Acad. Sci. USA 82, 4110 – 4114 (1985). L. W. Fisher and J. D. Termine, Noncollagenous proteins influencing the local mechansims of calcification. Clin. Orthop. 200, 362 – 365 (1985). A. L. Boskey, Mineral-matrix interactions in bone and cartilage. Clin. Orthopaed. 281, 244 – 274 (1992). W. Traub, T. Arad, and S. Weiner, Origin of mineral crystal growth in collagen fibrils. Matrix 12, 251 – 255 (1992). U. Vetter, L. W. Fisher, K. P. Mintz, J. B. Kopp, N. Tuross, J. D. Termine, and P. Gehron Robey, Osteogenesis imperfecta: Changes in noncollagenous proteins in bone. J. Bone Miner. Res. 6, 501 – 505 (1991). L. W. Fisher, M. A. Drum, P. Gehron Robey, K. M. Conn, and J. D. Termine, Osteonectin content in human osteogenesis imperfecta bone shows a range similar to that of two bovine models of OI. Calcif. Tissue Int. 40, 260 – 264 (1987). N. S. Fedarko, M. Moerike, R. Brenner, P. Gehron Robey, and U. Vetter, Extracellular matrix formation by osteoblasts from patients with osteogenesis imperfecta. J. Bone Min. Res. 7, 921 – 930 (1992). N. S. Fedarko, P. Gehron Robey, and U. K. Vetter, Extracellular matrix stoichiometry in osteoblasts from patients with osteogenesis imperfecta. J. Bone Miner. Res. 10, 1122 – 1129 (1995). A. Boyde, A. J. Banes, and R. M. Dullamans, A morphological study of an avian bone disorder caused by myeloblastosis-associated virus. Metab. Bone Dis. Rel. Res. 1, 235 (1978). A. L. Boskey and S. C. Marks, Mineral and matrix alterations in the bones of incisors-absent (ia/ia) osteopetrotic rat. Calcif. Tissue Int. 37, 287 – 292 (1985).

173. N. B. Schwartz, E. W. Pirok, III, J. R. Mensch, Jr., and M. S. Domowicz, Domain organization, genomic structure, evolution, and regulation of expression of the aggrecan gene family. Prog. Nucleic Acid Res. Mol. Biol. 62, 177 – 225 (1999). 174. K. M. Dyne, M. Valli, A. Forlino, M. Mottes, H. Kress, and G. Cetta, Deficient expression of the small proteoglycan decorin in a case of severe/lethal osteogenesis imperfecta. Am. J. Med. Genet. 53, 161 – 166 (1996). 175. N. S. Fedarko, Isolation and purification of proteoglycans. EXS 70, 9 – 35 (1994). 176. H. Kresse, H. Hausser, and E. Schinherr, Small proteoglycans. In “Proteoglycans” (P. Jolläs, ed.), pp. 73 – 100. Birkhauser Verlag, Basel, Switzerland, 1994. 177. M. Yanagishita, Function of proteoglycans in the extracellular matrix. Acta Pathol. Japonica 43, 283 – 293 (1993). 178. P. J. Neame, H. Tapp, and A. Azizan, Noncollagenous, nonproteoglycan macromolecules of cartilage. Cell. Mol. Life Sci. 55, 1327 – 1340 (1999). 179. F. Deak, D. Piecha, C. Bachrati, M. Paulsson, and I. Kiss, Primary structure and expression of matrilin-2, the closest relative of cartilage matrix protein within the von Willebrand factor type A-like module superfamily. J. Biol. Chem. 272, 9268 – 9274 (1997). 180. J. D. Sandy, C. R. Flannery, R. E. Boynton, and P. J. Neame, Isolation and characterization of disulfide-bonded peptides from the three globular domains of aggregating cartilage proteoglycan. J. Biol. Chem. 265, 21108 – 21113 (1990). 181. M. Morgelin, M. Paulsson, T. E. Hardingham, D. Heinegard, and J. Engel, Cartilage proteoglycans: Assembly with hyaluronate and link protein as studied by electron microscopy. Biochem J. 253, 175 – 185 (1988). 182. P. J. Neame and F. P. Barry, The link proteins. Experientia 49, 393 – 402 (1993). 183. T. E. Hardingham, The role of link protein in the structure of cartilage proteoglycan aggregates. Biochem. J. 177, 237 – 247 (1979). 184. L. Kjellän and U. Lindahl, Proteoglycans: Structures and interactions. Annu. Rev. Biochem. 60, 443 – 475 (1991). 185. D. Heinegard and A. Oldberg, Structure and biology of cartilage and bone matrix noncollagenous macromolecules. FASEB J. 3, 2042 – 3053 (1989). 186. H. Paulssen, R. Busch, V. Sinnwell, and A. Poller-Kruger, Conformation analysis of D-xylose-containing O-glycopeptide sequences. Carbohydr. Res. 214, 227 – 234 (1991). 187. J. E. Scott, Supramolecular organization of extracellular matrix glycosaminoglycans in vitro and in the tissues. FASEB J. 6, 2639 – 2645 (1992). 188. J. R. Korenberg, X. N. Chen, K. Doege, J. Grover, and P. J. Roughley, Assignment of the human aggrecan gene (AGC1) to 15q26 using fluorescence in situ hybridization analysis. Genomics 16(2), 546 – 548 (1993). 189. K. J. Doege, M. Sasaki, E. Horigan, J. R. Hassell, and Y. Yamada, Complete primary structure of the rat cartilage proteoglycan core protein deduced from cDNA clones. J. Biol. Chem. 262, 17757 – 17767 (1987). 190. K. J. Doege, K. Garrison, S. N. Coulter, and Y. Yamada, The structure of the rat aggrecan gene and preliminary characterization of its promoter. J. Biol. Chem. 269, 29232 – 29240 (1994). 191. D. A. Horita, P. J. Hajduk, P. F. Goetinck, and L. E. Lerner, NMR studies of peptides derived from the putative binding regions of cartilage proteins: No evidence for binding to hyaluronan. J. Biol. Chem. 269, 1699 – 1704 (1994). 192. S. J. Perkins, A. S. Nealis, D. G. Dunham, T. E. Hardingham, and H. Muir, Neutron and X-ray solution-scattering studies of the ternary complex between proteoglycan-binding region, link protein and hyaluronan. Biochem. J. 285, 263 – 268 (1992).

CHAPTER 4 The Biochemistry of Bone 193. J. A. Buckwalter and L. C. Rosenberg, Structural changes during development in bovine fetal epiphyseal cartilage. Coll. Rel. Res. 3, 489 – 504 (1983). 194. J. A. Buckwalter, P. J. Roughley, and L. C. Rosenberg, Age-related changes in cartilage proteoglycans: Quantitative electron miscroscopic studies. Microsc. Res. Tech. 28, 398 – 408 (1994). 195. W. D. Comper, and T. C. Laurent, Physiological function of connective tissue polysaccharides. Physiol. Rev. 58, 255 – 315 (1978). 196. W. D. Comper and R. P. W. Williams, Hydrodynamics of concentration proteoglycan solutions. J. Biol. Chem. 262, 13464 – 13471 (1987). 197. Nishimura, Y. Muragaki, and B. R. Olsen, Tissue-specific forms of type IX collagen-proteoglycan arise from the use of two widely separated promoters. J. Biol. Chem. 264, 20033 – 20041 (1989). 198. W. D. Comper, Physicochemical aspects of cartilage extracellular matrix. In “Cartilage: Molecular Aspects” (B. Hall and S. Newman, eds.), pp. 59 – 96. CRC Press, Boca Raton, FL, 1991. 199. G. H. Hunter, Role of proteoglycan in the provisional calcification of cartilage. Clin. Orthop. Rel. Res. 262, 256 – 263 (1991). 200. C. C. Chen, A. L. Boskey, and L. C. Rosenberg, The inhibitory effect of cartilage proteoglycans on hydroxyapatite growth. Calcif. Tissue Int. 36, 285 – 290 (1984). 201. C. C. Chen, and A. L. Boskey, Mechanisms of proteoglycan inhibition of hydroxyapatite growth. Calcif. Tissue Int. 37, 395 – 400 (1985). 202. A. L. Boskey, Hydroxyapatite formation in a dynamic gel system: Effects of type I collagen, lipids, and proteoglycans. J. Phys. Chem. 93, 1628 – 1633 (1989). 203. D. E. Eanes, Mixed phospholipid liposome calcification. Bone Miner. 17, 269 – 272 (1992). 204. G. K. Hunter , An ion-exchange mechanism of cartilage calcification. Connect. Tissue Res. 16, 111 – 120 (1987). 205. A. L. Boskey, M. Maresca, A. L. Armstrong, and M. G. Ehrlich, Treatment of proteoglycan aggregates with growth plate neutral metalloprotease reduces their ability to inhibit hydroxyapatite proliferation in a gelatin gel. J. Orthopaed. Res. 10, 313 – 319 (1992). 206. D. E. Eanes and A. W. Hailer, Effect of ultrafilterable fragments from chondroitinase and protease-treated aggrecan on calcium phosphate precipitation in liposomal suspensions. Calcif. Tissue Int. 55, 176 – 179 (1994). 207. K. Kimata, H. Barrach, K. Brown, and J. P. Pennypacker, Absence of proteoglycan core protein in cartilage from the cmd/cmd (cartilage matrix deficiency) mouse, J. Biol. Chem. 256, 6961 – 6968 (1981). 208. A. L. Boskey, M. Maresca, and A. Hjerpe, Hydroxyapatite formation in the presence of proteoglycans of reduced sulfate content: Studies in the brachymorphic mouse. Calcif. Tissue Int. 49, 389 – 393 (1991). 209. W. S. Argraves, P. J. McKeown-Longo, and P. F. Goetinck, Absence of proteoglycan core protein in the cartilage mutant nanomelia. FEBS Lett. 131, 265 – 268 (1981). 210. S. Weinbaum, S. C. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. J. Biomech. 27, 339 – 360 (1994). 211. L. Y. Yao, C. Moody, E. Schonherr, T. N. Wight, and L. J. Sandell, Identification of the proteoglycan versican in aorta and smooth muscle cells by DNA sequence analysis, in situ hybridization and immunohistochemistry. Matrix Biol. 14, 213 – 225 (1994). 212. G. J. Thomas, M. T. Bayliss, K. Harper, R. M. Mason, and M. Davies, Glomerular mesangial cells in vitro synthesize an aggregating proteoglycan immunologically related to versican, Biochem. J. 302, 49 – 56 (1994). 213. T. J. Crawford, I. G. Melhado, and F. R. Jirik, Expression of versican mRNA is developmentally regulated in the brain of the embryonic chick and the developing rat. Brain Res. Dev. Brain Res. 76, 264 – 267 (1993).

165 214. A. Bignami, G. Perides, and F. Rahemtulla, Versican, a hyaluronate-binding proteoglycan of embryonal precartilaginous mesenchyma, is mainly expressed postnatally in rat brain. J. Neurosci. Res. 34, 97 – 106 (1993). 215. D. R. Zimmerman, M. T. Dours-Zimmerman, M. Schubert, and L. Bruckner-Tuderman, Versican is expressed in the proliferating zone in the epidermis and in association with the elastic network of the dermis. J. Cell Biol. 124, 817 – 825 (1994). 216. L. W. Fisher, The nature of the proteoglycans of bone. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 188 – 196. EBSCO Media, Birmingham, AL, 1985. 217. K. Ito, T. Shinomura, M. Zako, M. Ujita, and K. Kimata, Multiple forms of mouse PG-M, a large chondroitin sulfate proteoglycan generated by alternative splicing. J. Biol. Chem. 270, 958 – 965 (1995). 218. M. Yamagata, T. Shinomura, and K. Kimata, Tissue variation of two large chondroitin sulfate proteoglycans (PG-M/versican and PG-H/aggrecan) in chick embryos. Anat. Embryol. 187, 433 – 444 (1993). 219. R. V. Iozzo, M. F. Naso, L. A. Cannizzaro, J. J. Wasmuth, and J. D. McPherson, Mapping of the versican proteoglycan gene (CSPG2) to the long arm of human chromosome 5 (5q12 – 5q14). Genomics 14, 845 – 851 (1992). 220. M. F. Naso, D. R. Zimmermann, and R. V. Iozzo, Characterization of the complete genomic structure of the human versican gene and functional analysis of its promoter. J. Biol. Chem. 269, 32999 – 33008 (1994). 221. M. T. Dours-Zimmermann and D. R. Zimmermann, A novel glycosaminoglycan attachment domain identified in two alternative splice variants of human versican. J. Biol. Chem. 269, 32992 – 32998 (1994). 222. D. R. Zimmermann and E. Ruoslahti, Multiple domains of the large fibroblast proteoglycan, versican, EMBO J. 8; 2575 – 2588 (1990). 223. E. Schonherr, H. T. Jarvelainen, L. J. Sandell, and T. Wight, Effects of platelet-derived growth factor and transforming growth factorbeta 1 on the synthesis of a large versican-like chondroitin sulfate proteoglycan by arterial smooth muscle cells. J. Biol. Chem. 266, 17640 – 17647 (1991). 224. V. M. Kahari, H. Larjava, and J. Uitto, Differential regulation of extracelular matrix proteoglycan (PG) gene expression: Transforming growth factor-beta 1 up-regulates biglycan (PGI), and versican (large fibroblast PG) but down-regulates decorin (PGII) mRNA levels in human fibroblasts in culture. J. Biol. Chem. 266, 10608 – 10615 (1991). 225. E. E. Qwarnstrom, H. T. Jarvelainen, M. G. Kinsella, C. O. Ostberg, L. J. Sandell, R. C. Page, and T. N. Wight, Interleukin-1 beta regulation of fibroblast proteoglycan synthesis involves a decrease in versican steady-state mRNA levels. Biochem. J. 294, 613 – 620 (1993). 226. G. Perides, F. Rahemtulla, W. S. Lane, R. A. Asher, and A. Bignami, Isolation of a large aggregating proteoglycan from human brain. J. Biol. Chem. 267, 23883 – 23887 (1992). 227. J. Engel, Domains in proteins and proteoglycans of the extracellular matrix with functions in assembly and cellular activities. Int. J. Biol. Macromol. 13, 147 – 151 (1991). 228. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Biphasic effects of epidermal growth factor on bone nodule formation by isolated rat calvaria cells in vitro. J. Bone Miner. Res. 2, 385 – 393 (1987). 229. N. Matsushima, T. Ohyanagi, T. Tanaka, and R. H. Kretsinger, Super-motifs and evolution of tandem leucine-rich repeats within the small proteoglycans: Biglycan, decorin, lumican, fibromodulin, PRELP, keratocan, osteoadherin, epiphycan, and osteoglycin. Proteins 38, 210 – 225 (2000). 230. H. Kresse, H. Hausser, E. Schonherr, and K. Bittner, Biosynthesis and interactions of small chondroitin/dermatan sulphate proteoglycans. Eur. J. Clin. Chem. Clin. Biochem. 32, 259 – 254 (1994).

166 231. D. S. Howell, J. C. Pita, J. F. Marquez, and J. E. Madruga, Partition of calcium, phosphate and protein in the fluid phase aspirated at calcifying sites in epiphyseal cartilage. J. Clin. Invest. 47, 1121 – 1132 (1968). 232. J. Klein-Nulend, P. Veldhuijzen, R. J. van de Stadt et al., Influence of intermittent compressive force on proteoglycan content in calcifying growth plate cartilage in vitro. J. Biol. Chem. 262, 15490 (1987). 233. C. W. Prince, F. Rahemtulla, and W. T. Butler, Metabolism of rat bone proteoglycans in vivo. Biochem. J. 216, 589 – 596 (1983). 234. C. W. Prince, F. Rahemtulla, and W. T. Butler, Incorporation of [35S]sulphate into glycosaminoglycans by mineralized tissues in vivo. Biochem. J. 224, 941 – 945 (1984). 235. L. W. Fisher, The nature of proteoglycans of bone. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 188 – 196, EBSCO Media, Birmingham, AL, 1985. 236. B. Kobe and J. Deisnhofer, Crystal structure of porcine ribonuclease inhibitor, a protein with leucine-rich repeats. Nature 366, 751 – 756 (1993). 237. C. Branden and J. Tooze , Motifs of protein structure. In “Introduction to Protein Structure,” Chap. 2, pp. 11 – 32. Garland Publishing, New York, 1991. 238. L. W. Fisher, J. D. Termine, and M. F. Young, Deduced protein sequence of bone small proteoglcyan I (biglycan) shows homology with proteoglcyan II (decorin) and several non-connective tissue proteins in a variety of species. J. Biol. Chem. 264, 4571 – 4576 (1989). 239. G. A. Pringle and C. M. Dodd, Immunoelectron microscopy localization of the core protein of decorin near the d and e bands of tendon collagen fibrils by use of monoclonal antibodies. J. Histochem. Cytochem. 38, 1405 – 1411 (1990). 240. D. Simonescu, R. V. Iozzo, and N. A. Kefalides, Bovine pericardial proteoglycan: Biochemical, immunochemical and ultrastructural studies. Matrix 9, 301 – 310 (1989). 241. L. W. Fisher, J. D. Termine, S. W. Dejter, S. W. Whitson, M. Yanagishita, J. H. Kimura, V. C. Hascall, H. K. Kleinman, J. R. Hassell, and B. Nilson, Proteoglycans of developing bone. J. Biol. Chem. 258, 6588 – 6594 (1983). 242. H. U. Choi, T. L. Johnson, P. Subhash, L.-H. Tang, and L. Rosenberg, Characterization of the dermatan sulfate proteoglycans, DSPGI and DS-PGII, from bovine articular cartilage and skin isolated by octyl-spharose chromatography. J. Biol. Chem. 264, 2876 – 2884 (1989). 243. P. Bianco , L. W. Fisher, M. F. Young, J. D. Termine, and P. Gehron Robey, Expression and localization of the two small proteoglycans biglycan and decorin in developing human skeletal and non-skeletal tissues. J. Histochem. Cytochem. 38, 1549 – 1563 (1990). 244. B. Johnstone, M. Markopoulos, P. Neame, and B. Caterson, Identification and characterization of glycanated and non-glycanated forms of biglycan and decorin in the human intervertebral disc. Biochem. J. 292, 661 – 666 (1993). 245. L. W. Fisher, G. R. Hawkins, N. Tuross, and J. D. Termine, Purification and partial characterization of small proteoglycans I and II, bone sialoproteins I and II, and osteonectin from the mineral compartment of developing human bone. J. Biol. Chem. 262, 9702 – 9708 (1987). 246. D. D. Krantz, R. Ziovetski, B. L. Kagan, and S. L. Ziporsky, Ampipathic -structure of a leucine-rich repeat peptide. J. Biol. Chem. 266, 16901 – 16907 (1991). 247. P. Krishnan, A. M. Hocking, J. M. Scholtz, C. N. Pace, K. K. Holik, and D. J. McQuillan, Distinct secondary structures of the leucinerich repeat proteoglycans decorin and biglycan: Glycosylation-dependent conformational stability. J. Biol. Chem. 274(16), 10945 – 10950 (1999). 248. O. W. McBride, L. W. Fisher, and M. F. Young, Localization of PGI (biglycan, BGN) and PGII (decorin, DCN, PG-40) genes on human

GOKHALE, BOSKEY, AND ROBEY

249.

250.

251.

252.

253.

254.

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

chromosomes Xq13-qter and 12q, respectively. Genomics 6(2); 219 – 225 (1990). K. G. Danielson, A. Fazzio, I. Cohen, L. A. Cannizzaro, I. Eichstetter, and R. V. Iozzo, The human decorin gene: Intron-exon organization, discovery of two alternatively spliced exons in the 5 untranslated region, and mapping of the gene to chromosome 12q23. Genomics 15, 146 – 160 (1993). U. Vetter, W. Vogel, W. Just, M. F. Young, and L. W. Fisher, Human decorin gene: Intron-exon junctions and chromosomal localization, Genomics 15, 161 – 168 (1993). T. Scholzen, M. Solursh, S. Suzuki, R. Reiter, J. L. Morgan, A. M. Buchberg, L. D. Siracusa, and R. V. Iozzo, The murine decorin, complete cDNA cloning genomic organization, chromosomal assignment, and expression during organogenesis and tissue differentiation. J. Biol. Chem. 269, 28270 – 28281 (1994). M. D. Briggs and D. H. Cohn, Length polymorphism within a complex dinucleotide repeat in the human decorin (DCN) gene. Hum. Mol. Genet. 2, 1087 (1993). M. Santra, K. G. Danielson, and R. V. Iozzo, Structural and functional characterization of the human decorin gene promoter: A homopurine-homopyrimidine S1 nuclease-sensitive region is involved in transcriptional control. J. Biol. Chem. 269, 579 – 587 (1994). T. Krusius, K. R. Gehlsen, and E. A. Ruoslahti, fibroblast chondroitin sulfate proteoglycan core proteincontains lectin-like and growth factor-like sequences. J. Biol. Chem. 262; 13120 – 13125 (1987). A. A. Day, C. I. McQuillan, J. D. Termine, and M. R. Young, Molecular cloning and sequence analysis of the cDNA for small proteoglycan II of bovine bone. Biochem. J. 248, 801 – 805 (1987). A. Bassols and J. Massague, Transforming growth factor- regulates the expression and structure of extracellular matrix chondroitin/dermatan sulfate proteoglycans. J. Biol. Chem. 263, 3039 – 3045 (1988). D. M. Rosen, S. A. Stempien, A. Y. Thompson, and S. M. Seyedin, Transforming growth factor-beta modulates the expression of osteoblast and chondroblast phenotypes in vitro. J. Cell Physiol. 134, 337 – 346 (1988). Y. Takeuchi, T. Matsumoto, E. Ogata, and Y. Shishiba, Effects of transforming growth factor 1 and L-ascorbate on synthesis and distribution of proteoglycans in murine osteoblast-like cells, J. Bone Miner. Res. 8, 823 – 829 (1993). S. Kimoto, S. L. Cheng, S. F. Zhang, and L. V. Avioli, The effect of glucocorticoid on the synthesis of biglycan and decorin in human osteoblasts and bone marrow stromal cells. Endocrinology 135(6); 2423 – 2431 (1994). S. L. Cheng, S. F. Zhang, T. L. Nelson, P. M. Warlow, and R. Civitelli, Stimulation of human osteoblast differentiation and function by ipriflavone and its metabolites. Calcif. Tissue Int. 55, 356 – 362 (1994). N. A. Visser, G. P. Vankampen, M. H. Dekoning, and J. K. Vanderkorst, The effects of loading on the synthesis of biglycan and decorin in intact mature articular cartilage in vitro. Connect. Tissue Res. 30, 241 – 250 (1994). R. S. Sawhney, T. M. Hering, and L. J. Sandell, Biosynthesis of small proteoglycan II (decorin) by chondrocytes and evidence for a procore protein. J. Biol. Chem. 266, 9231 – 9240 (1991). K. G. Vogel and J. A. Trotter, The effect of proteoglycans on the morphology of collagen fibrils formed in vitro. Collagen Rel. Res. 7, 105 – 114 (1987). D. J. Bidanset, C. Guidry, L. C. Rosenberg, H. Choi, R. Timpl, and M. Hook, Binding of the proteoglycan decorin to collagen type VI. J. Biol. Chem. 267, 5250 – 5256 (1992). B. Merle, H. V. Choi, L. Rosenberg, J. Lawler, P. D. Delmas, and P. Clezardin, In vitro interaction between decorin and thrombospondin. In “FECTS XIV Meeting,” p. F13. Lyon, France, 1994.

CHAPTER 4 The Biochemistry of Bone 266. Y. Yamaguchi, D. M. Mann, and E. Rouslahti, Negative regulation of transforming growth factor- by the proteoglycan decorin. Nature 346, 281 – 284 (1990). 267. G. Schmidt, H. Hausser, and H. Kresse, Interaction of the small proteoglycan decorin with fibronectin. Biochem. J. 266, 591 – 595 (1990). 268. M. Winnermoller, G. Schmidt, and H. Kresse, Influence of decorin on fibroblast adhesion to fibronectin. Eur J Cell Biol. 54, 10 – 17 (1991). 269. K. G. Vogel, M. Paulsson, and D. Heinegard, Specific inhibition of type I and type II collagen fibrillogenesis by the small proteoglycan of tendon. Biochem. J. 223, 587 – 597 (1984). 270. K. G. Vogel and L. W. Fisher, Comparisons of antibody reactivity and enzyme sensitivity between small proteoglycans from bovine tendon, bone, and cartilage. J. Biol. Chem. 261, 11334 – 11340 (1986). 271. D. E. Ashurst and Y. S. Bland, The collagens of the soft tissues during development in the knee and hip joints of the rabbit: Their synthesis and distribution. In “FECTS XIV Meeting,” p. 4H. Lyon, France, 1994. 272. D. V. Brown and K. C. Vogel, Characteristics of the in vitro interaction of a small proteoglycan (PGII) of bovine tendon with type I collagen. Matrix 9, 468 – 478 (1989). 273. A. L. Boskey, L. Spevak, and L. C. Rosenberg, Effects of bone CSproteoglycans, DS-decorin, and DS-biglycan on hydroxyapatite formation in a gelatin gel, Calci. Tissue Int. 61(4); 298 – 305 (1997). 274. A. L. Boskey and L. Spevak, The effects of decorin and biglycan on in vitro apatite formation in a gelatin gel. In “FECTS XIV Meeting,” p. N6. Lyon, France (1994). 275. L. Rosenberg, H. U. Choi, L. Tang, T. Johnson, J. Liu, and T. M. Lave, Metal-binding properties of biglycan. Trans 40th ORS 19, 5 (1994). 276. P. M. Steinlen, E. Maessen, J. Kresse, I. M. Van Vlijmen et al., Expression of tenascin, biglycan and decorin in disorders of keratinization. Br. J. Dermatol. 130, 564 – 568 (1994). 277. E. Rouslahti and Y. Yamaguchi, Proteoglycans as modulators of growth factor activities. Cell 64, 867 – 869 (1991). 278. K. M. Dyne, M. Valli, M. Forlino, M. Mottes, and G. Cetta, Decreased decorin expression in patients with osteogenesis imperfecta. In “FECTS Meeting,” p. E6. Lyon, France, 1994. 279. Y. Takeuchi, Y. Kodama, and T. Matsumoto, Bone matrix decorin binds transforming growth factor- and enhances its bioactivity. J. Biol. Chem. 51, 32634 – 32638 (1994). 280. T. Krusius and E. Ruoslahti, Primary structure of an extracellular matrix proteoglycan coreprotein deduced from cloned cDNA. Proc. Natl. Acad. Sci. USA. 83, 7683 – 7687 (1986). 281. N. S. Fedarko, J. D. Termine, and P. Gehron Robey, Temporal sequence of proteoglycan biosynthesis by normal human bone cell in vitro. J. Biol. Chem. 265, 12200 – 12209 (1990). 282. K. Ibaraki, S. W. Whitson, J. D. Termine, and M. F. Young, Bone matrix mRNA expression in differentiating fetal bovine osteoblasts. J. Bone Miner. Res. 7, 743 – 754 (1992). 283. P. G. Neame, H. C. Choi, and L. C. Rosenberg, The primary structure of the core protein of the small leucine-rich proteoglycan (PGI) from bovine articular cartilage. J. Biol. Chem. 264, 8653 – 8661 (1989). 284. J. Liu, T. M. Laue, H. U. Choi, L.-H. Tang, and L. Rosenberg, The self-association of biglycan from bovine articular cartilage. J. Biol. Chem. 269, 28366 – 28373 (1994). 285. L. W. Fisher, A. M. Heegaard, U. Vetter, W. Vogel, W. Just, J. D. Termine, and M. F. Young, Human biglycan gene. J. Biol. Chem. 266, 14371 – 14377 (1991). 286. V. Asundi, K. Cowan, D. Matzura, W. Wagner, and K. L. Dreher, Characterization of extracellular matrix proteoglycan transcripts expressed by vascular smooth muscle cells. Eur. J. Cell Biol. 52, 98 – 104 (1990).

167 287. A. M. Heegaard, P. Gehron Robey, W. Vogel, W. Just, R. L. Widom, J. Scholler, J. Fisher, and M. F. Young, Functional characterization of the human biglycan 5-flanking DNA and binding of the transcription factor c-Krox. J. Bone Miner. Res. 12, 2050 – 2060 (1997). 288. D. Thiebaud, H. L. Guenther, A. Porret, P. Burckhardt, H. Fleisch, and W. Hofstetter, Regulation of collagen type I and biglycan mRNA levels by hormones and growth factors in normal and immortalized osteoblastic cell lines. J. Bone Miner. Res. 9, 1347 – 1354 (1994). 289. D. Pearson and J. Sasse, Differential regulation of biglycan and decorin by retinoic acidin bovine chondrocytes. J. Biol. Chem. 267, 25364 – 25370 (1992). 290. P. Gehron Robey, N. S. Fedarko, W. Lindner et al., Biosynthesis of a small protoeglycan (PG-I) by human bone cells. J. Cell Biol. 105, 299a (1987). 291. S. Kimoto, S.-L. Cheng, S.-F. Zhang, and L. V. Avioli, The effect of glucocorticoid on the synthesis of biglycan and decorin in human osteoblasts and bone marrow stromal cells. Endocrinology 135, 2423 – 2431 (1994). 292. R. J. Waddington, and M. S. Langley, Structural analysis of proteoglycans synthesized by mineraling bone cells in vitro in the presence of fluoride. Matrix Biol. 17, 255 – 268 (1998). 293. H. C. Whinna, H. U. Choi, L. C. Rosenberg, and F. C. Church, Interaction of heparin cofactor II with biglycan and decorin. J. Biol. Chem. 268, 3920 – 3924 (1993). 294. L. Rosenberg and E. B. Hunziker, Induction of repair in partial thickness articular cartilage lesions by timed release of TGF. In “ORS 40th Annual Mtg,” pp. 236 – 241, 1994. 295. G. Pogany, D. J. Hernandez, and K. G. Vogel, The in vitro interaction of proteoglycans with type I collagen is modulated by phosphate. Arch. Biochem. Biophys. 313, 102 – 111 (1994). 296. U. Vetter, N. S. Fedarko, M. F. Young, J. D. Termine, W. Just, W. Vogel, and P. Gehron Robey, Biglycan synthesis in fibroblasts of patients with Turner’s syndrome and other X-chromosome abnormalities. J. Bone Min. Res. 6, S133 (1991). 297. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 298. G. Schmidt, H. Hausser, and H. Kresse, Interaction of the small proteoglycan decorin with fibronectin: Involvement of the sequence NKISK of the core protein. Biochem. J. 280, 411 – 414 (1991). 299. D. J. Bidanset, R. LeBaron, L. Rosenberg, J. E. Murphy-Ullrich, and M. Hook, Regulation of cell substrate adhesion: Effects of small galactosaminoglycan-containing proteoglycans. J. Cell Biol. 118, 1523 – 1531 (1992). 300. P. Gehron Robey, W. J. Grzesik, The biochemistry of bone forming cells: Cell matrix interactions. In “Proceedings of the International Conference on Tooth Eruption, Resorption, and Replacement by Implants” (Z. Davidovitch, ed.), pp. 167 – 172. EBSCO Press, Birmingham, AL, 1994. 301. I. Kamo, A. Kikuchi, I. Nonaka, E. Yamada, and J. Kondo, Haemopoietic activity associated with biglycan like proteoglycan. Biochem. Biophys. Res. Commun. 195, 1119 – 1126 (1993). 302. H. Greiling, Structure and biological functions of keratan sulfate proteoglycans. EXS 70, 101 – 102 (1994). 303. S. Ekman, and D. Heinegard, Immunohistochemical localization of matrix proteins in the femoral joint cartilage of growing commercial pigs. Vet. Pathol. 29(6), 514 – 520 (1992). 304. H. Hedlund, S. Mengarelli-Widholm, D. Heinegard, F. P. Reinhold, and O. Svensson, Fibromodulin distribution and association with collagen. Matrix Biol. 14, 227 – 232 (1994).

168 305. A. Hulth, O. Johnell, L. Lindberg, and D. Heinegard, Sequential appearance of macromolecules in bone induction in the rat. J. Orthop. Res. 11, 367 – 378 (1993). 306. L. Svensson, A. Aszodi, F. P. Reinholt, R. Fassler, D. Heinegard, and A. Oldberg, Fibromodulin-null mice have abnormal collagen fibrils, tissue organization, and altered lumican deposition in tendon. J. Biol. Chem. 274, 9636 – 9647 (1999). 307. R. Sztrolovics, M. Alini, J. S. Mort, and P. J. Roughley, Age-related changes in fibromodulin and lumican in human intervertebral discs. Spine 24, 1765 – 1771 (1999). 308. P. Antonsson, D. Heinegard, and A. Oldberg, Structure and deduced amino acid sequence of the human fibromodulin gene. Biochim. Biophys. Acta. 1174, 204 – 206 (1993). 309. E. Hedbom and D. Heinegård, Binding of fibromodulin and decorin to separate sites on fibrillar collagens. J. Biol. Chem. 268, 27307 – 27312 (1993). 310. D. Fukushima, R. Batzow, A. Hildebrand, and E. Ruoslahti, Localization of transforming growth factor beta binding site in betaglycan: Comparison with small extracellular matrix proteoglycans. J. Biol. Chem. 28, 22710 – 22715 (1993). 311. M. Bernfield, R. Kokenyesi, M. Kato, M. T. Hinkes, J. Spring, R. L. Gallo, and E. J. Lose, Biology of the syndecans: A family of transmembrane heparan sulfate proteoglycans. Annu. Rev. Cell Biol. 8, 365 – 393 (1992). 312. J. N. Beresford, N. S. Fedarko, L. W. Fisher, R. J. Midura, M. Yanagishita, J. D. Termaine, and P. Gehron Robey, Analysis of the proteoglycans synthesized by human bone cells in vitro. J. Biol. Chem. 262, 17164 – 17172 (1987). 313. J. L. Andres, K. Stanley, S. Cheifetz, and J. Massague, Membraneanchored and soluble forms of betaglycan, a polymorphic proteoglycan that binds transforming growth factor-beta. J. Cell Biol. 109(6 Pt 1), 3137 – 3145 (1989). 314. P. R. Segarini and S. M. Seyedin, The high molecular weight receptor to transforming growth factor-beta contains glycosaminoglycan chains. J. Biol. Chem. 263(17), 8366 – 8370 (1988). 315. M. Klagsbrun, The affinity of fibroblast growth factors (FGFs) for heparin; FGF-heparan sulfate interactions in cells and extracellular matrix. Curr. Opin. Cell Biol. 2(5), 857 – 863 (1990). 316. T. E. Hardingham and A. J. Fosang, Proteoglycans: Many forms and many functions. FASEB J. 6, 861 – 870 (1992). 317. T. Nagata, H. A. Goldberg, Q. Zhang, C. Domenicucci, and J. Sodek, Biosynthesis of bone proteins by fetal porcine calvariae in vitro: Rapid association of sulfated sialoproteins (secreted phosphoprotein-1 and bone sialoprotein) and chondroitin sulfate proteoglycan (CS-PGIII) with bone mineral. Matrix 11, 86 – 100 (1991). 318. L. G. Hashimoto, B. Caterson, and M. Yamauchi, EDTA-insoluble, calcium-binding proteoglycans in bone. In “FECTS XIV Meeting,” p. N10. Lyon, France, 1994. 319. H. Bentz, A. Y. Thompson, R. Armstrong, R. J. Chang, K. A. Piez, and D. M. Rosen, Transforming growth factor-beta 2 enhances the osteoinductive activity of a bovine bone-derived fraction containing bone matrix morphogenetic protein 2 and 3. Matrix 11, 269 – 275 (1991). 320. M. Ujita, T. Shinomura, and K. Kimata, Molecular cloning of the mouse osteoglycin-encoding gene. Gene 158, 237 – 240 (1995). 321. T. Shinomura, and K. Kimata, Proteoglycan-Lb, a small dermatan sulfate proteoglycan expressed in embryonic chick epiphyseal cartilage, is structurally related to osteoinductive factor. J. Biol. Chem. 267, 1265 – 1270 (1992). 322. K. Kurita, T. Shinomura, M. Ujita, M. Zako, D. Kida, H. Iwata, and K. Kimata, Occurrence of PG-Lb, a leucine-rich small chondroitin/dermatan sulphate proteoglycan in mammalian epiphyseal cartilage: molecular cloning and sequence analysis of the mouse cDNA. Biochem. J. 318, 909 – 914 (1996). 323. H. J. Johnson, L. Rosenberg, H. U. Choi, S. Garza, M. Hook, and P. J. Neame, Characterization of epiphycan, a small proteoglycan with

GOKHALE, BOSKEY, AND ROBEY

324.

325.

326. 327.

328.

329.

330.

331. 332.

333.

334.

335. 336. 337.

338. 339.

340.

341. 342.

343.

344.

a leucine-rich repeat core protein. J. Biol. Chem. 272, 18709 – 18717 (1997). Y. Iwata, T. Shinomura, K. Kurita, M. Zako, and K. Kimata, The gene structure and organization of mouse PG-Lb, a small chondroitin/dermatan sulphate proteoglycan. Biochem. J. 331, 959 – 964 (1998). A. Bosse, H. Kresse, K. Schwarz, and K. M. Muller, Immunohistochemical characterization of the small proteoglycans decorin and proteoglycan-100 in heterotopic ossification. Calcif. Tissue Int. 54, 119 – 124 (1994). B. P. Toole, “Cell Biology of Extracellular Matrix,” (E. D. Hay, ed.), New York, 1991. S. O. Hjertquist and L. Veljens, The glycosaminoglycans of dog compact bone and epiphyseal cartilage in the normal state and in hyper parathyroidism. Calcif. Tissue Res. 2, 314 – 333 (1968). C. B. Knudson and W. Knudson, Hyaluronan-binding proteins in development, tissue homeostasis, and disease. FASEB J. 7, 1233 – 1241 (1993). N. S. Fedarko, J. D. Termine, and P. Gehron Robey, High-performance liquid chromatographic separation of hyaluronan and four proteoglycans produced by human bone cell cultures. Anal Biochem. 188, 398 – 407 (1990). S. J. Green, G. Tarone, and C. B. Underhill, Distribution of hyaluronate and hyaluronate receptors in the adult lung. J. Cell Sci. 89, 145 – 156 (1988). C. B. Knudson, Hyaluronan receptor-directed assembly of chondrocyte pericellular matrix. J. Cell Biol. 120, 825 – 834 (1993). P. Paskalakis, D. H. Vynios, C. P. Tsiganos, E. Dalas, C. Maniatis, and P. G. Koutsoukos, Effect of proteoglycans on hydroxyapatite growth in vitro: The role of hyaluronan. Biochim. Biophys. Acta. 1158, 129 – 136 (1993). A. L. Boskey and B. L. Dick. Hyaluronan interactions with hydroxyapatite do not alter in vitro hydroxyapatite crystal proliferation and growth, Matrix 11, 442 – 446 (1991). P. Paschalakis, D. H. Vynios, C. P. Tsiganos, E. Dalas, C. Maniatis, and P. G. Koutsoukos, Effect of proteoglycans on hydroxyapatite growth in vitro: The role of hyaluronan. Biochim. Biophys. Acta. 1158, 129 – 136 (1993). A. J. Day, The structure and regulation of hyaluronan-binding proteins. Biochem. Soc. Trans. 27, 115 – 121 (1999). V. C. Hascall, Interaction of cartilage proteoglycans with hyaluronic acid. J. Supramol. Struct. 7, 101 – 120 (1977). J. Koda, A. Rapraeger, and M. Bernfield, Heparan sulfate proteoglycans from mouse mammary epithelial cells: Cell surface proteoglycan as a receptor for interstitial collagens. J. Biol. Chem. 260, 8157 – 8162 (1985). B. Toole, Developmental role of hyaluronate. Connect. Tissue Res. 10, 83 – 100 (1982). G. Oster, J. Murray, and P. Maini, A model for chondrogenic condensations in the developing limb: The role of extracellular matrix and cell tractions. J. Embryol. Exp. Morphol. 89, 93 – 112 (1985). I. Binderman, R. M. Green, and J. P. Pennypacker, Calcification of differentiating skeletal mesenchyme in vitro. Science 206, 222 – 225 (1979). B. P. Toole, G. Jackson, and J. Gross, Hyaluronate in morphogenesis in vitro. Proc. Natl. Acad. Sci. USA 69, 1384 – 1389 (1972). L. A. Goldstein, D. F. H. Zhou, L. J. Picker, C. N. Minty, R. F. Bargatze, J. F. Ding, and E. L. Butcher, A human lymphocyte homing receptor, the hermes antigen, is related to cartilage proteoglycan and link proteins. Cell 56, 1063 – 1072 (1984). J. Lesley and R. Hyman, CD44 can be activated to function as an hyaluronic acid receptor in normal murine T cells. Eur. J. Immunol. 22, 2719 – 2723 (1992). P. H. Weigel, G. M. Fuller, and R. D. Le Beof, A model for the role of the hyaluronic acid and fibrin in the early events during the

CHAPTER 4 The Biochemistry of Bone

345.

346.

347. 348.

349.

350.

351.

352.

353.

354.

355.

356.

357.

358.

359.

360.

361.

inflammatory response and wound healing. J. Theor. Biol. 119, 219 – 228 (1986). C. Knudson and B. Toole, Changes in the pericellular matrix during differentiation of limb bud mesoderm. Dev. Biol. 112, 308 – 318 (1985). B. A. Mast, R. F. Diegelmann, T. M. Krummel, and K. Cohen, Hyaluronic acid modulates proliferation, collagen and protein synthesis of cultured fetal fibroblasts. Matrix 13, 441 – 446 (1993). J. J. Minguell, Is hyaluronic acid the organizer of the extracellular matrix in marrow stroma? Exp. Hematol. 21, 7 – 8 (1993). M. Siczkowski, T. Andrew, S. Amos, and M. Y. Gordon, Hyaluronic acid regulates the function and distribution of sulfated glycosaminoglycans in bone marrow stromal cultures. Exp. Hematol. 21, 126 – 130 (1993). A. Engstrom-Laurent, Changes in hyaluronan concentrationin tissues and body fluids in disease states. Ciba Found. Symp. 143, 233 – 247 (1989). P. A. G. McCourt, B. Ek, N. Forsberg, and S. Gustafson, Intercellular adhesion molecule-1 is a cell surface receptor for hyaluronan. J. Biol. Chem. 269, 30081 – 30084 (1994). N. S. Fedarko, P. Bianco, U. K. Vetter, and P. Gehron Robey, Human bone cell enzyme expression and cellular heterogeneity: Correlation of alkaline phosphatase enzyme activity with cell cycle. J. Cell Physiol. 144, 115 – 121 (1990). B. de Bernard, P. Bianco, E. Bonucci, M. Costantini, G. C. Lunazzi, P. Martinuzzi, C. Modricky, L. Moro, E. Panfili, and P. Pollesello, Biochemical and immunohistochemical evidence that in cartilage an alkaline phosphatase is a Ca-binding glycoprotein, J. Cell Biol. 103, 1615 – 1623 (1986). P. Bianco, M. Riminucci, G. Silvestrini, E. Bonucci, J. D. Termine, L. W. Fisher, and P. Gehron Robey, Localization of bone sialoprotein (BSP) to golgi and post-golgi secretory structures in osteoblasts and to discrete sites in early bone matrix. J. Histochem. Cytochem. 41, 193 – 203 (1993). P. Bianco and A. Boyde, Alkaline phosphatase cytochemistry in confocal scanning light microcopy for imagining the bone marrow stroma. Basic Appl. Histochem. 33, 17 – 23 (1989). P. Bianco, N. S. Fedarko, E. Bonucci, J. D. Termine, and P. Gehron Robey, Acid and alkaline phosphatase activities of bone cells revisited. In “Proceedings of the Int. Soc. Bioanaloging Skeletal Implants,” pp. 17 – 27, 1991. P. Bianco, P. Ballanti, S. Mazzaferro, G. Coen, and E. Bonucci, Combined use of tetracycline fluorescence and alkaline phosphatase cytochemisry in the evaluation of osteoblastic activity in bone biopsies. Basic Appl. Histochem. 31(Suppl.), 29 (1987). J. Zernik, K. Twarog, and W. B. Upholt, Regulation of alkaline phosphatase and ? 2(I) procollagen synthesis during early intramembranous bone formation in the rat mandible. Differentiation 44, 207 – 215 (1990). S. Sakano, Y. Murata, T. Miura, H. Iwata, K. Sato, N. Matsui, and H. Seo, Collagen and alkaline phosphatase gene expression during bone morphogenetic protein (BMP-induced cartilage and bone differentiation. Clin. Orthop. 292, 337 – 344 (1993). P. Collin, J. R. Nefussi, A. Wetterwald, V. Nicolas, M. L. BoyLefever, H. Fleisch, and N. Forest, Expression of collagen, osteocalcin and bone alkaline phosphatase in a mineralizing rat osteoblastic cell culture. Calcif. Tissue Int. 50, 175 – 183 (1992). P. Bianco, and M. Riminucci, The bone marrow stroma in vivo: Ontogeny, structure, cellular composition and changes in disease. In “Marrow Stromal Cell Culture” (J. N. Beresford and M. E. Owen, eds.), pp. 10 – 25. Cambridge Univ. Press, Cambridge, UK. P. Bianco, M. Costantini, and L. C. Dearden, Alkaline phosphatase positive precursors of adipocytes in the human bone marrow. Br. J. Haematol. 88, 401 – 403 (1988).

169 362. B. C. Nair, R. J. Majeska, and G. A. Rodan, Rat alkaline phosphatase. I. Purification and characterization of the enzyme from osteosarcoma: Generation of monoclonal and polyclonal antibodies. Arch. Biochem. Biophys. 254, 18 – 27 (1987). 363. B. C. Nair, D. E. Johnson, R. J. Majeska et al., Rat alkaline phosphatase. II. Structural similarities between the osteosarcoma, bone, kidney, and placental isoenzymes. Arch. Biochem. Biophys. 254, 28 – 34 (1987). 364. S. Nagoya, T. Uede, T. Wada, S. Ishii, S. Yamawaki, and K. Kikuchi, Detection of bone-type alkaline phsophatase by monoclonal antibodies reacting with human osteosarcoma-associated antigen. Jpn. J. Cancer Res. 82, 862 – 870 (1991). 365. M. G. Low, The glycosyl-phosphatidylinositol anchor of membrane proteins. Biochim. Biophys. Acta. 988, 427 – 454 (1989). 366. M. Noda, K. Yoon, G. A. Rodan, and D. E. Koppel, High lateral mobility of endogenous and transfected alkaline phosphatase: A phosphitdylinositol anchored membrane protein. J. Cell Biol. 105, 1671 – 1677 (1987). 367. E. E. Kim and H. W. Wyckoff, Structure of alkaline phosphatases. Clin. Chem. Acta 186, 175 – 187 (1989). 368. H. W. Wyckoff, M. D. Handschumacher, H. M. K. Murthy, and J. M. Sowadski, The three-dimensional structure of alkaline phosphatase from E coli. Adv. Enzymol. Rel. Areas Mol. Biol. 55, 453 – 480 (1983). 369. E. E. Kim and H. W. Wyckoff, Reaction mechanisms of alkaline phosphatase based on crystal structures: Two-metal ion catalysis. J. Mol. Bio. (England) 218, 449 – 464 (1991). 370. S. Hansen, L. K. Hansen, and E. Hough, The crystal structure of tris-inhibited phospholipase C from Bacillus cereus at 1.9 A resolution: The nature of the metal ion in site 2. J. Mol. Biol. 231(3), 870 – 876 (1993). 371. D. M. Swallow, Mapping of the gene coding for the human liver/bone/kidney isozyme of alkaline phosphatase to chromosome 1. Ann. Hum. Genet. 60, 229 – 235 (1986). 372. H. H. Hsu, The partial sequencing of human nonspecific (bone/liver/kidney) alkaline phosphatase gene. Int. J. Biochem. 23, 391 – 393 (1991). 373. J. E. Beever and H. A. Lewin, RFLP at the bovine liver, bone and kidney alkaline phosphatase (ALPL) locus. Anim. Genet. 23(6), 577 (1992). 374. M. Terao, M. Studer, M. Gianni, and E. Garattini, Isolation and characterization of the mouse liver/bone/kidney-type alkaline phosphatase gene. Biochem. J. 268(3), 641 – 648 (1990). 375. M. Studer, M. Terao, M. Gianni, and E. Garattini, Characterization of a second promoter for the mouse liver/bone/kidney-type alkaline phosphatase gene: Cell and tissue specific expression. Biochem. Biophys. Res. Commun. 179(3), 1352 – 1360 (1991). 376. J. Zernik, B. Kream, and K. Twarog, Tissue-specific and dexamethasone-inducible expression of an alkaline phosphatase from alternative promoters of the rat bone/liver/kidney/placenta gene. Biochem. Biophys. Res. Commun. 176, 1149 – 1156 (1991). 377. M. Kiledjian and T. Kadesch, Analysis of the human liver/bone/kidney alkaline phosphatase promoter in vivo and in vitro. Nuclei Acids Res. 18, 957 – 961 (1990). 378. J. Zernik, M. A. Thiede, K. Twarog, M. L. Stover, G. A. Rodan, W. B. Upholt, and D. W. Rowe, Cloning and analysis of the 5 region of the rat bone/liver/kidney/placenta alkaline phosphatase gene: A dual function promter. Matrix 10, 38 – 47 (1990). 379. S. Matsuura, F. Kishi, and T. Kajii, Characterization of a 5-flanking region of the human liver/bone/kidney alkaline phosphatase gene: Two kinds of mRNA from a single gene. Biochem. Biophys. Res. Commun. 168(3), 993 – 1000 (1990). 380. J. A. Deren, F. S. Kaplan, and C. T. Brighton, Alkaline phosphatase production by periosteal cells at various oxygen tensions in vitro. Clin. Orthop. 252, 307 – 312 (1990).

170 381. L. Y. Yohay, J. Zhang, K. M. Thraikill, J. M. Arthur, and L. D. Quarles, Role of serum in the developmental expression of alkaline phosphatase in MC3T3-E1 osteoblasts. J. Cell Physiol. 158, 467 – 475 (1994). 382. E. Kyeyune-Nyombi, K. H. Lau, D. J. Baylink, and D. D. Strong, 1,25-Dihydroxyvitamin D3 stimulates both alkaline phosphatase gene transcription and mRNA stability in human bone cells. Arch. Biochem. Biophys. 291, 316 – 325 (1991). 383. K. W. Ng, P. R. Gummer, V. P. Michelangeli, J. E. Baterman, T. Mascara, W. G. Cole, and T. J. Martin, Regulation of alkaline phosphatase expression in a neonatal rat calvarial cell strain by retinoic acid. J. Bone Miner. Res. 3, 43 – 61 (1988). 384. J. K. Heath, L. J. Suva, K. Yoon, M. Kiledjian, T. J. Martin, and G. A. Rodan, Retinoic acid stimulates transcriptional activity from the alkaline phosphatase promoter in the immortalized rat calvarial cell line, RCT-1. Mol. Endocrinol. 6, 636 – 646 (1992). 385. M. Gianni, M. Terao, S. Sozzani, and E. Garattini, Retinoic acid and cyclic AMP synergistically induce the expression of liver/bone/kidney-type alkaline phosphatase gene in L929 fibroblastic cells. Biochem. J. 296, 67 – 77 (1993). 386. M. Subramaniam, D. Colvard, P. E. Keeting, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Glucocorticoid regulation of alkaline phosphatase, osteocalcin, and proto-oncogenes in normal human osteoblast-like cells. J. Cell. Biochem. 50, 411 – 424 (1992). 387. J. R. Farley, S. L. Hall, S. Herring, Calcitonin acutely increases net 45Ca uptake and alters alkline phosphatase specific activity in human osteosarcoma cells. Metabolism 42(1); 97 – 104 (1993). 388. P. J. Price, B. O. Oyajobi, R. O. Oreffo, and R. G. G. Russell, Cells cultured from the growing tip of red dear antler express alkaline phosphatase and proliferate in response to insulin-like growth factor-I. J. Endocrionol. 143, R9 – R16 (1994). 389. P. S. Leboy, L. Vaias, B. Uschman, E. Golub, and S. L. Adams, Pacific M’ ascorbic acid induces alkaline phosphatase, type X collgen, and calcium deposition in cultured chick chondrocytes. J. Biol. Chem. 264, 17281 – 17286 (1989). 390. J. F. Chlebowski and J. E. Coleman, Pre-steady state kinetics of phosphorothioate hydrolysis by alkaline phosphatase. J. Biol. Chem. 247, 6007 – 6010 (1989). 391. R. Robison, The possible significance of hexose phosphoric esters in ossification. Biochem. J. 17, 286 – 293 (1923). 392. P. Magnusson, L. Larsson, M. Magnusson, M. W. Davie, and C. A. Sharp, Isoforms of bone alkaline phosphatase: Characterization and origin in human trabecular and cortical bone. J. Bone Miner. Res. 14, 1926 – 1933 (1999). 393. G. I. Baroncelli, S. Bertelloni, C. Ceccarelli, V. Amato, and G. Saggese, Bone turnover in children with vitamin D deficiency rickets before and during treatment. Acta. Paediatr. 89, 513 – 518 (2000). 394. M. P. Whyte, Alkaline phosphatase: Physiological role explored in hypophosphatasia. In “Bone and Mineral Research” (W. A. Peck, ed.), Vol. 6, pp. 175 – 218. Elsevier, Amsterdam, 1989. 395. H. Harris, The human alkaline phosphatases: What we know and what we don’t know. Clin. Chim. Acta 186, 133 – 150 (1989). 396. M. J. Weiss, D. E. C. Cole, K. Ray, M. P. Whyte, M. A. Lafferty, R. A. Mulivor, and H. Harris, A misuse mutation in the human liver/bone/kidney alkaline phosphatase causing a lethal form of hyphophosphatasia. Proc. Natl. Acad. Sci. USA 85, 7666 – 7669 (1989). 397. K. Yoon, E. Golub, and G. A. Rodan, Alkaline phosphatase transfected cells promote calcium and phosphate deposition. Connect. Tissue Res. 22, 17 – 25 (1989). 398. H. I. Khouja, A. Bevington, G. J. Kemp, and R. G. Russell, Calcium and orthophosphate deposits in vitro do not imply osteoblastmediated mineralization: Mineralization by betaglycerophosphate in the absence of osteoblasts. Bone 11, 385 – 391 (1990).

GOKHALE, BOSKEY, AND ROBEY 399. M. D. Fallon, M. P. Whyte, and S. L. Teitelbaum, Stereospecific inhibition of alkaline phosphatase by L-tetramisole prevents in vitro cartilage calcification. Lab. Invest. 43, 489 – 494 (1980). 400. B. Zimmerman, H. C. Wachtel, and J. Vormann, Kinetics of ßglycerophosphate-induced endochondral mineralization in vitro: Calcium accumulation, alkaline phosphatase activity, and effects of levamisole. Calcif. Tissue Int. 51, 54 – 61 (1992). 401. A. L. Boskey, W. Ziecheck, P. Guidon, and S. B. Doty, Gallium nitrate inhibits alkaline phosphatase activity in differentiating mesenchymal cell culture. Bone Min. 20, 179 – 192 (1993). 402. Y. Ishikawa and R. E. Wuthier, Development of an in vitro mineralization model with growth plate chondrocytes that does not require ß-glycerophosphate. Bone Miner. 17, 152 – 157 (1992). 403. K. N. Fedde, L. Blair, J. Silverstein, S. P. Coburn, L. M. Ryan, R. S. Weinstein, K. Waymire, S. Narisawa, J. L. Millan , G. R. MacGregor, and M. P. Whyte, Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophopshatasia. J. Bone Miner. Res. 14, 2015 – 2026 (1999). 404. S. Narisawa, N. Frohlander, and J. L. Millan, Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dyn. 208; 432 – 446 (1997). 405. W. M. Burch, G. Hammer, and R. E. Wuthier, Phosphotyrosine and phosphoprotein phosphatase activity of alkaline phosphatase in mineralizing cartilage. Metabolism 34, 169 – 175 (1985). 406. A. L. Boskey, S. B. Doty, R. Mendelsohn, and I. Binderman, ATP promotes mineralization in differentiating chick limb-bud mesenchymal cell cultures. Micros. Res. Techniq. 28, 492 – 504 (1994). 407. L. N. Wu, W. B. Valhmu, G. C. Lloyd, B. R. Genge, and R. E. Wuthier, Isolation of two glycosylated forms of membrane-bound alkaline phosphatase from avian growth plate cartilage matrix vesicle-enriched microsomes. Bone Miner. 7, 113 – 125 (1989). 408. Y. Xu, K. P. Pritzer, and T. F. Cruz, Characteriztion of chondrocyte alkaline phosphatase as a potential mediator in the dissolution of calcium pyrophosphate dihydrate crystals. J. Rheumatol. 21, 912 – 919 (1994). 409. J. D. Termine, H. K. Kleinman, S. W. Whitson, K. M. Conn, M. L. McGarvey, and G. R. Martin, Osteonectin, a bone-specific protein linking mineral to collagen. Cell 26, 99 – 105 (1981). 410. K. M. Conn and J. D. Termine, Matrix protein profiles in calf bone development. Bone 6, 33 – 36 (1985). 411. I. J. Mason, A. Taylor, J. G. Williams, H. Sage, and B. L. M. Hogan, Evidence from molecular cloning that SPARC, a major product of mouse embryo parietal endoderm, is related to endothelial cell “culture shock” glycoprotein of MR-43000. EMBO J. 5, 1465 – 1472 (1986). 412. H. Sage, C. Johnson, and P. Bornstein, Characterization of a novel serum albumin binding glycoprotein secreted by endothelial cells in culture. J. Biol. Chem. 259, 3993 – 4007 (1984). 413. M. Dziadek, M. Paulsson, M. Aumailley and R. Timpl, Purification and tissue distribution of a small protein (BM-40) extracted from a basement membrane tumor. Eur. J. Biochem. 161, 455 – 464 (1986). 414. K. Mann, R. Deutzmann, M. Paulsson, and R. Timpl, Solubilization of protein BM-40 from a basement membrane tumor with chelating agents and evidence for its identity with osteonectin and SPARC. FEBS Lett. 218, 167 – 172 (1987). 415. P. W. H. Holland, S. J. Harper, and J. H. McVey, In vivo expression of mRNA for the Ca-binding protein SPARC (osteonectin) revealed by in situ hybridization. J. Cell Biol. 105, 473 – 482 (1987). 416. M. Metsaranta, M. F. Young, M. Sandberg, J. Termine, and E. Vuorio, Localization of osteonectin expression in human fetal skeletal tissues by in situ hybridization. Calcif. Tissue Int. 45, 146 – 152 (1989). 417. P. L. Porter, E. H. Sage, T. F. Lane, S. E. Funk, and A. M. Gown, Distribbution of SPARC in normal and neoplastic human tissue. J. Histochem. Cytochem. 43, 791 – 800 (1995).

171

CHAPTER 4 The Biochemistry of Bone 418. W. M. Wewer, R. Albrechtsen, L. W. Fisher et al., Osteonectin/ SPARC/BM-40 in human decidua and carcinoma, tissues characterized by de novo formation of basement membrane. Am. J. Pathol. 132, 345 (1988). 419. H. Sage, A secreted protein associated with cellular proliferation inhibits cell spreading in vitro. J. Cell Biol. 107, 587a (1989). 420. P. Bianco, J. B. Kopp, and L. W. Fisher, New data on the distribution of osteonectin/SPARC in human and mouse tissues and their potential implications. J. Bone Miner. Res. 4(Suppl.), S245 (1989). 421. J. B. Kopp, P. Bianco, M. F. Young, J. D. Termine, and P. Gehron Robey, Differential expression of osteopontin and osteonectin by cultured renal tubular epithelial cells. Kidney Int. 41, 56 – 64 (1992). 422. P. Bianco, G. Silverstrini, J. D. Termine, and E. Bonucci, Immunohistochemical localization of osteonectin in developing human bone using monoclonal antibodies. Calcif. Tissue Int. 43, 155 – 161 (1988). 423. G. Jundt, L. H. Berghauser, J. D. Termine et al., Osteonectin: A differentiation marker of bone cells. Cell Tissue Res. 248, 409 (1987). 424. P. S. Tung, C. Domenicucci, S. Wasi et al., Specific immunohistochemical localization of osteonectin and collagen types I and III in fetal and adult porcine dental tissues. J. Histochem. Cytochem. 33, 531 (1985). 425. S. Kasugai, T. Nagata, and J. Sodek, Temporal studies on the tissue compartmentalization of bone sialoprotein (BSP), osteopontin (OPN), and SPARC protein during bone formation in vitro. J. Cell Physiol. 152, 467 – 77 (1992). 426. P. Gehron Robey, L. W. Fisher, J. T. Stubbs, and J. D. Termine, Biosynthesis of osteonectin and a small proteoglycan (PG-II) by connective tissue cells in vitro. In “Development and Diseases of Cartilage and Bone Matrix” (A. Sen and T. Thornhill, eds.), pp. 115 – 125. A. R. Liss, New York, 1986. 427. D. D. Stenner, R. P. Tracy, B. L. Riggs, and K. G. Mann, Human platelets contain and secrete osteonectin, a major protein of mineralized bone. Proc. Natl. Acad, Sci. USA 83, 6892 – 6896 (1986). 428. Q. Yan, and E. H. Sage, SPARC, a matricellular glycoprotein with important biological functions. J. Histochem. Cytochem. 47, 1495 – 1506 (1999). 429. R. W. Romberg, P. G. Werness, P. Lollar, B. L. Riggs, and K. G. Mann, Isolation and characterization of native adult osteonectin. J. Biol. Chem. 260, 2728 – 2736 (1985). 430. R. Thomas, L. D. True, J. A. Bassuk, P. H. Lange, and R. L. Vessella, Differential expression of osteonectin/SPARC during human prostate cancer progression. Clin. Cancer Res. 6, 1140 – 1149 (2000). 431. R. W. Romberg, P. G. Werness, P. Lollar, B. L. Riggs, and K. G. Mann, Isolation and characterization of native adult osteonection. J. Biol. Chem. 260, 2728 – 2736 (1985). 432. A. Swaroop, B. L. M. Hogan, and U. Francke, Molecular analysis of the cDNA for human SPARC/Osteonectin/BM-40: Sequence, expression and localization of the gene to chromosome 5q31 – q33. Genomics 2, 37 (1988). 433. R. C. Schwartz, M. F. Young, and P. Tsipouras, Two RFLPs in the 5 end of the human osteonectin (ON) gene. Nucleic Acids Res. 16(18), 9076 (1988). 434. M. E. Bolander, M. F. Young, L. W. Fisher, Y. Yamada, and J. D. Termine, Osteonectin cDNA sequence reveals potential binding regions for calcium and hydroxyapatite and shows homologies with both a basement membrane protein (SPARC) and a serine-protease inhibitor (ovomucoid). Proc. Natl. Acad. Sci. USA 85, 2919 – 2923 (1988). 435. D. M. Findlay, L. W. Fisher, C. I. McQuillan, J. D. Termine, and M. F. Young, Isolation of the osteonectin gene: Evidence that a variable region of the osteonectin molecule is encoded within one exon. Biochemistry 27, 1483 – 1489 (1988). 436. H. H. McVey, S. Nomura, P. Kelly, I. J. Mason, and B. L. Hogan,

437.

438.

439.

440.

441.

442.

443.

444.

445.

446.

447.

448.

449.

450.

451.

452.

453.

Characterization of the mouse SPARC/Osteonectin gene: Intron/exon organization and an unusual promoter region. J. Biol. Chem. 263, 11111 – 11116 (1988). J. A. Bassuk, M. L. Iruela-Arispa, T. F. Lane, J. M. Benson, R. A. Berg, and E. H. Sage, Molecular analyses of chicken embryo sparc (osteonectin). Eur. J. Biochem. 218, 117 – 127 (1993). I. G. Johnston, T. Paladin, J. W. Curd, and I. R. Brown, Molecular cloning of SC1: A putative brain extracellular matrix glyprotein showing partial similarity to osteonectin/BM40/SPARC. Neuron 2, 165 – 176 (1990). P. Maurer, C. Hohenadl, E. Hohenester, W. Gohring, R. Timpl, and J. Engel, The C-terminal portion of BM-40 (SPARC/osteonectin) is an autonomously folding and crystallisable domain that binds calcium and collagen IV. J. Mol. Biol. 253, 347 – 357 (1995). M. Guermah, P. Crisanti, D. Laugier, L. Bidou, B. Pessac, and G. Calothy, Transcription of quail gene expressed in embryonic retinal cells is shut off sharply at hatching. Proc. Natl. Acad. Sci. USA 88, 4503 – 4507 (1991). P. M. Alliel, J. P. Perin, P. Jolles, and F. J. Bonnet, Testican, a multidomain testicular proteoglycan resembling modulators of cell social behaviour. Eur. J. Biochem. 214, 347 – 350 (1993). A. Zwijsen, H. Blockx, W. Van Arnhem, J. Willems, L. Fransen, K. Devos, J. Raymackers, A. Van de Voorde, and H. Slegers, Characterization of a rat C6 glioma-secreted follistatin-related protein (FRP): Cloning and sequence of the human homologue. Eur. J. Biochem. 225, 937 – 946 (1994). M. F. Young, D. M. Findlay, P. Dominguez et al., Osteonectin promoter: DNA sequence analysis and S1 endonuclease site, potentially associated with transcriptional control in bone cells. J. Biol. Chem. 264, 450 (1989). P. Dominguez, K. Ibaraki, P. Gehron Robey, T. Hefferan, J. D. Termine, and M. F. Young, The expression of the osteonectin gene is potentially controlled by multiple cis and trans acting elements. J. Bone Mines. Res. 6, 1127 – 1136 (1991). K. Ibaraki, P. Gehron Robey, and M. F. Young, The transcription of the osteonectin gene is potentially controlled by AP1, AP2, AP3 and a novel GAGA binding factor. Connect Tissue Res. 27, 130 (1992). L. Malaval, B. Darbouret, C. Preaudat, J. P. Jolu, and P. D. Delmas, Intertissular variations in osteonectin: A monoclonal antibody directed to bone osteonectin shows reduced affinity for platelet osteonectin. J. Bone Miner. Res. 6(4), 315 – 323 (1991). R. C. Hughes, A. Taylor, H. Sage, and B. L. M. Hogan, Distinct patterns of glycosylation of colligin, a collagen-binding glycoprotein, and SPARC (osteonectin), a secreted Ca-binding glycoprotein: Evidence for the localization of colligin in the endoplasmic reticulum. Eur. J. Biochem. 163, 57 – 65 (1987). R. J. Kelm, Jr., and K. G. Mann, The collagen binding specificity of bone and platelet osteonectin is related to differences in glycosylation. J. Biol. Chem. 266, 9632 – 9639 (1991). R. J. Kelm, Jr., G. A. Hair, K. G. Mann, and B. W. Grant, Characterization of human osteoblast and megakaryocyte-derived osteonectin (SPARC). Blood 80, 3112 – 3119 (1992). X. C. Villarreal, L. Malaval, K. G. Mann, P. Delmas, and G. L. Long, Epitope mapping of two monoclonal antibodies to the central portion of human osteonectin. Calcif. Tissue Int. 48, 138 – 141 (1991). S. Chandrasekhar, A. K. Harvey, M. G. Johnson, and G. W. Becker, Osteonectin/SPARC is a product of articular chondrocytes/cartilage and is regulated by cytokines and growth factors. Biochim. Biophys. Acta. 1221(1), 7 – 14 (1994). M. Noda, and G. A. Rodan, Type beta transforming growth factor regulates expression of genes encoding bone matrix proteins. Connect. Tissue Res. 21(1 – 4), 71 – 75 (1989). J. L. Wrana, M. Maeno, B. Hawrylyshyn, K. L. Yao, C. Domenicucci, and J. Sodek, Differential effects of transforming

172

454.

455.

456.

457.

458.

459.

460.

461.

462.

463.

464.

465.

466.

467.

468.

469.

469a.

470.

GOKHALE, BOSKEY, AND ROBEY growth factor-? on the synthesis of extracellular matrix proteins by normal fetal rat calvarial bone cell populations. J. Cell Biol. 106, 915 – 924 (1988). H. Zhou, R. G. Hammonds, Jr., D. M. Findlay, P. J. Fuller, T. J. Martin, and K. W. Ng, Retinoic acid modulation of mRNA levels in malignant, nontransformed, and immortalized osteoblasts. J. Bone Miner. Res. 6(7), 767 – 777 (1991). M. Neri, F. Descalzi-Cancedda, and R. Cancedda, Heat-shock response in cultured chick embryo chondrocytes: Osteonectin is a secreted heat-shock protein. Eur. J. Biochem. 205, 569 – 574 (1992). M. F. Young, A. A. Day, P. Dominguez, C. I. McQuillan, L. W. Fisher, and J. D. Termine, Structure and expression of osteonectin mRNA in human tissue. Connect. Tissue Res. 24, 17 – 28 (1990). J. Engel, W. Taylor, M. Paulsson, H. Sage, and B. Hogan, Calcium binding domains and calcium-induced conformational transition of SPARC/BM-40/osteonectin, an extracellular glycoprotein expressed in mineralized and nonmineralized tissues. Biochemistry 26, 6958 – 6965 (1987). E. Hohenester, P. Maurer, and R. Timpl, Crystal structure of a pair of follistatin-like and EF-hand calcium-binding domains in BM-40. EMBO J. 16(13), 3778 – 3786 (1997). B. Tyree, The partial degradation of osteonectin by a bone-derived metalloprotease enhances binding to type I collagen. J. Bone Miner. Res. 4, 877 – 883 (1989). H. Sage, R. B. Vernon, S. E. Funk, E. A. Everitt, and J. Angello, SPARC, a secreted protein associated with cellular proliferation, inhibits cell spreading in vitro and exhibits Ca2 dependent binding to the extracellular matrix. J. Cell Biol. 109, 341 – 356 (1989). P. Clezardin, L. Malavel, A.-S. Ehrensperger, P. Delmas, M. Dechavanne, and J. L. McGregor, Complex formation of human thrombospondin with osteonectin. Eur. J. Biochem. 175, 275 – 284 (1988). R. W. Romberg, P. G. Werness, B. L. Riggs, and K. G. Mann, Inhibition of hydroxyapatite crystal growth by bone-specific and other calcium-binding proteins. Biochemistry 25, 1176 – 1180 (1986). H. J. Roach, Why does bone matrix contain non-collagenous proteins? The possible roles of osteocalcin, osteonectin, osteopontin, and bone sialoprotein in mineralisation and resorption. Cell. Biol. Int. 18, 617 – 628 (1994). S. E. Funk and E. H. Sage, The Ca2 binding glycoprotein SPARC modulates cell cycle progression in bovine aortic endothelial cells. Proc. Natl. Acad. Sci. USA 88, 2648 – 2652 (1991). E. H. Sage, Modulation of endothelial cell shape by SPARC does not involve chelation of extracellular Ca2 and Mg2. Biochem. Cell. Biol. 70, 56 – 62 (1992). T. F. Lane and H. Sage, Functional mapping of SPARC: Peptides from two distinct Ca(2) binding sites modulates cell shape. J. Cell Biol. 111, 3065 – 3076 (1990). H. Sage and P. Bornstein, Extracellular proteins that modulate cellmatrix interations: SPARC, tenascin, and thrombospondin. J. Biol. Chem. 266, 14831 – 14834. D. T. Gilmour, G. J. Lyon, M. B. Carlton, J. R. Sanes, J. M. Cunningham, J. R. Anderson, B. L. Hogan, M. J. Evans, and W. H. Colledge, Mice deficient for the secreted glycoprotein SPARC/osteonectin/BM40 develop normally but show severe age-onset cataract formation and disruption of the lens. EMBO J. 17, 1860 – 1870 (1998). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectin-deficient mice [published erratum appears in J. Clin. Invest. 105(9), 1325 (2000)]. J. Clin. Invest. 105, 915 – 923 (2000). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectin-deficient mice. J. Clin. Invest. 105, 1325 (2000). T. F. Lane, M. L. Iruela-Arispe, and E. H. Sage, Regulation of gene expression by SPARC during angiogenesis in vitro: Changes in

471.

472.

473.

474.

475.

476.

477.

478. 479.

480.

481.

482.

483.

484.

485.

486.

487.

488.

fibronectin, thrombospondin-1, and plasminogen activator inhibitor-1. J. Biol. Chem. 267, 16736 – 16745 (1992). P. Hasselaar and E. H. Sage, SPARC antagonizes the effect of basic fibroblast growth factor on the migration of bovine aortic endothelial cells. J. Cell. Biochem. 49, 272 – 283 (1992). J. C. Yost and E. H. Sage, Specific interaction of SPARC with endothelial cells is mediated through a carboxyl-terminal sequence containing a calcium-binding EF hand. J. Biol. Chem. 268, 25790 – 25796 (1993). S. E. Funk and E. H. Sage, Differential effects of SPARC and cationic SPARC peptides on DNA synthesis by endothelial cells and fibroblasts. J. Cell. Physiol. 154, 53 – 63 (1993). S. E. Goldblum, X. Ding, S. E. Funk, and E. H. Sage, SPARC (secreted protein acidic and rich in cysteine) regulates endothelial cell shape and barrier function. Proc. Natl. Acad. Sci. USA 91, 3448 –3452 (1994). E. H. Sage, J. A. Bassuk, J. C. Yost, M. J. Folkman, and T. F. Lane, Inhibition of endothelial cell proliferation by SPARC is mediated through a Ca2-binding EF-hand sequence. J. Cell. Biochem. 57, 127 – 140 (1995). E. A. Everitt and E. H. Sage, Expression of SPARC is correlated with altered morphologies in transfected F9 embryonal carcinoma cells. Exp. Cell Res. 199, 134 – 146 (1992). E. A. Everitt and E. H. Sage, Overexpression of SPARC in stably transfected F9 cells mediates attachment and spreading in Ca(2)deficient medium. Biochem. Cell. Biol. 70, 1368 – 1379 (1992). T. F. Lane and E. H. Sage, The biology of SPARC, a protein that modulates cell-matrix interactions. FASEB J. 8, 163 – 173 (1994). T. F. Lane, M. L. Iruela-Arispe, R. S. Johnson, and E. H. Sage, SPARC is a source of copper-binding peptides that stimulate angiogenesis. J. Cell Biol. 125, 929 – 943 (1994). M. J. Reed, P. Puolakkainen, T. F. Lane, D. Dickerson, P. Bornstein, and E. H. Sage, Differential expression of SPARC and thrombospondin 1 in wound repair: Immunolocalization and in situ hybridization. J. Histochem. Cytochem. 41, 1467 – 1477 (1993). E. W. Raines, T. F. Lane, M. L. Ireula-Arispe, R. Ross, and E. H. Sage, The extracellular glycoprotein SPARC interacts with plateletderived growth factor (PDGF-AB and -BB and inhibits the binding of PDGF to its receptors. Proc. Natl. Acad. Sci. USA 89, 1281 – 1285 (1992). P. M. Tremble, T. F. Lane, E. H. Sage, and Z. Werb, SPARC, a secreted protein associated with morphogenesis and tissue remodeling, induces expression of metalloproteinases in fibroblasts through a novel extracellular matrix-dependent pathway. J. Cell Biol. 121, 1433 – 1444 (1993). R. J. Kelm, Jr., N. A. Swords, T. Orfeo, and K.G. Mann, Osteonectin in matrix remodeling, a plasminogen-osteonectin-collagen complex. J. Biol. Chem. 269, 30147 – 30153 (1994). J. E. Murphy-Ullrich, T. F. Lane, M. A. Pallero, and E. H. Sage, SPARC mediates focal adhesion disassembly in endothelial cells through a follistatin-like region and the Ca2-binding EF-hand. J. Cell Biochem. 57, 341 – 350 (1995). L. Purcell, J. Guia-Gray, S. Scanga, and M. Ringuette, Developmental anomalies of xenopus embroys following microinjection of SPARC antibodies. J. Exp. Zool. 265, 153 – 164 (1993). J. E. Schwarzbauer and C. S. Spencer, The Caenorhabditis elegans homologue of the extracellular calcium binding protein SPARC/osteonectin affects nematode body morphology and mobility. Mol. Biol. Cell 4, 941 – 952 (1993). I. Clemmensen, L. C. Petersen, and C. Kluft, Purification and characterization of a novel, oligomeric, plasminogen kringle 4 binding protein from human plasma: Tetranectin. Eur. J. Biochem. 156, 327 – 333 (1986). U. M. Wewer, K. Ibaraki, P. Schjorring, M. E. Durkin, M. F. Young, and R. Albrechtsen, A potential role for tetranectin in mineralization during osteogenesis. J. Cell Biol. 127, 1767 – 1775 (1994).

CHAPTER 4 The Biochemistry of Bone 489. U. M. Wewer and R. Albrechtsen, Tetranectin, a plasminogen kringle 4-binding protein. Cloning and gene expression pattern in human colon cancer. Lab. Invest. 67, 253 – 262 (1992). 490. L. Berglund and T. E. Petersen, The gene structure of tetranectin, a plasminogen binding protein. FEBS Lett. 309, 15 – 19 (1992). 491. Fuhlendorff, I. Clemmensen, and S. Magnusson, Primary structure of tetranectin, a plasminogen kringle 4 binding plasma protein: Homology with a sialoglycoprotein receptors and cartilage proteoglycan core protein. Biochemistry 26, 6757 – 6764 (1987). 492. E. Ruoslahti and M. D. Pierschbacher, New perspectives in cell adhesion: rgd and integrins. Science 238, 491 (1987). 493. R. O. Hynes, Integrins: A family of cell surface receptors. Cell 48, 548 – 554 (1987). 494. R. O. Hynes, Integrins: Versatility modulation and signalling in cells adhesion. Cell 69, 11 – 25 (1992). 495. D. B. Donoviel, P. Framson, C. F. Eldridge, M. Cooke, S. Kobayashi, and P. Bornstein, Structural analysis and expression of the human thrombospondin gene promoter. J. Biol. Chem. 263, 18590 – 18593 (1988). 496. J. W. Lawler, L. H. Derick, J. E. Connolly, J. H. Chen, and F. C. Chao, The structure of human platelet thrombospondin. J. Biol. Chem. 260, 3762 – 3772 (1985). 497. S. S. Margossian, J. W. Lawler, and H. S. Slayter, Physical characterization of platelet thrombospondin. J. Biol. Chem. 256(14), 7495 – 7500 (1981). 498. J. Lawler, F. C. Chao, and C. M. Cohen, Evidence for calciumsensitive structure in platelet thrombospondin: Isolation and partial characterization of thrombospondin in the presence of calcium. J. Biol. Chem. 257(20), 12257 – 12265 (1982). 499. N. L. Baenziger, G. N. Brodie, and P. W. Majerus, Isolation and properties of a thrombin-sensitive protein of human platelets. J. Biol. Chem. 248, 2723 – 2731 (1972). 500. T. N. Wight, G. J. Raugi, S. M. Mumby et al., Light microscopic immunolocalization of thrombospondin in human tissues. J. Histochem. Cytochem. 33, 295 (1985). 501. K. S. O’Shea, L. H. Liu, L. H. Kinnunen, and Y. M. Dixit, Role of the extracellular matrix protein thrombospondin in the early development of the mouse embryo. J. Cell Biol. 111(6 Pt 1), 2713 – 2723 (1990). 502. K. S. O’Shea, L. H. J. Liu, and V. M. Dixit, Thrombospondin and a 140 kd fragment promote adhesion and neurite outgrowth from embryonic central and peripheral neurons and from PC12 cells. Neuron 7, 231 – 237 (1991). 503. P. Gehron Robey, M. F. Young, L. W. Fisher, and T. D. McClain, Thrombospondin is an osteoblast-derived component of mineralized extracellular matrix. J. Cell Biol. 108, 719 – 727 (1989). 504. W. J. Grzesik and P. Gehron Robey, Bone matrix RGD glycoproteins: Immunolocalization and interaction with human primary osteoblastic bone cells in vitro. J. Bone Miner. Res. 9, 487 – 496 (1994). 505. W. A. Frazier, Thrombospondin: A modular adhesive glycoprotein of platelets and nucleated cells. J. Cell Biol. 105, 625 – 632 (1987). 506. P. Bornstein, Thrombospondins: Structure and regulation of expression. FASEB J. 6, 3290 – 3299 (1992). 507. D. D. Roberts, D. M. Haverstick. V. M. Dixit, W. A. Frazier, S. A. Santoro, and V. Ginsburg, The platelet glycoprotein thrombospondin binds specifically to sulfated glycolipids. J. Biol. Chem. 260, 9405 – 9411 (1985). 508. X. Sun, D. F. Mosher, and A. Rapraeger, Heparan sulfate-mediated binding of epithelial cell surface proteoglycan to thrombospondin. J. Biol. Chem. 264, 2885 – 2889 (1989). 509. F. W. Wolf, R. L. Eddy, T. B. Shows, and V. M. Dixit, Structure and chromosomal localization of the human thrombospondin gene. Genomics 6(4), 685 – 691 (1990). 510. E. Jaffe, P. Bornstein, and C. M. Disteche, Mapping of the thrombospondin gene to human chromosome 15 and mouse chromosome 2 by in situ hybridization. Genomics 7(1), 123 – 126 (1990).

173 511. T. L. LaBell and P. H. Byers, Sequence and characterization of the complete human thrombospondin 2 cDNA: Potential regulatory role for the 3 untranslated region. Genomics 17(1), 225 – 229 (1993). 512. H. L. Vos, S. Devarayalu, Y. de Vries, and P. Bornstein, Thrombospondin 3 (Thbs3), a new member of the thrombospondin gene family. J. Biol. Chem. 267(17), 12192 – 12196 (1992). 513. G. Newton, S. Weremowicz, C. C. Morton, N. A. Jenkins, D. J. Gilbert, N. G. Copeland, and J. Lawler, The thrombospondin-4 gene. Mamm. Genome 10, 1010 – 1016 (1999). 514. J. T. Hecht, M. Deere, E. Putnam, W. Cole, B. Vertel, H. Chen, and J. Lawler, Characterization of cartilage oligomeric matrix protein (COMP) in human normal and pseudoachondroplasia musculoskeletal tissues. Matrix Biol. 17, 269 – 278 (1998). 515. T. Shingu and P. Bornstein, Characterization of the mouse thrombospondin 2 gene. Genomics 16, 78 – 84 (1993). 516. J. Adams and J. Lawler, Curr. Biol. 3, 188 – 190 (1993). 517. J. A. Carron, W. B. Bowler, S. C. Wagstaff, and J. A. Gallagher, Expression of members of the thrombospondin family by human skeletal tissues and cultured cells. Biochem. Biophys. Res. Commun. 263, 389 – 391 (1999). 518. M. Melnick, H. Chen, Y. Zhou, and T. Jaskoll, Thrombospondin-2 gene expression and protein localization during embryonic mouse palate development. Arch. Oral Biol. 45, 19 – 25 (2000). 519. P. Framson and P. Bornstein, A serum response element and a binding site for NF-Y mediate the serum response of the human thrombospondin 1 gene. J. Biol. Chem. 268, 4989 – 4996 (1993). 520. T. Shingu and P. Bornstein, Overlapping Egr-1 and Sp1 sites function in the regulation of transcription of the mouse thrombospondin 1 gene. J. Biol. Chem. 269, 32551 – 32557 (1994). 521. A. Mettouchi, F. Cabon, N. Montreau, P. Vernier, G. Mercier, D. Blangy, H. Tricoire, P. Vigier, and B. Binetruy, SPARC and thrombospondin genes are repressed by the c-jun oncogene in rat embryo fibroblasts. EMBO J. 13, 5668 – 5678 (1994). 522. P. Bornstein, S. Devarayalu, P. Li, C. M. Disteche, and P. Framson, A second thrombospondin gene in the mouse is similar in organization to thrombospondin 1 but does not respond to serum. Proc. Natl. Acad. Sci. USA 88, 8636 – 8640 (1991). 523. K. W. Adolph, G. L. Long, S. Winfield, E. I. Ginns, and P. Bornstein, Structure and organization of the human thrombospondin 3 gene (THBS3). Genomics 27, 329 – 336 (1995). 524. K. W. Adolph, D. J. Liska, and P. Bornstein, Analysis of the promoter and transcription start sites of the human thrombospondin 2 gene (THBS2). Gene 193, 5 – 11 (1997). 525. L. Malaval, D. Modrowski, A. K. Gupta, and J. E. Aubin, Cellular expression of bone related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures. J. Cell Physiol. 158, 555 – 572 (1994). 526. N. V. Sherbina and P. Bornstein, Modulation of thrombospondin gene expression during osteoblast differentiation in MC3T3-E1 cells. Bone 13, 197 – 201 (1992). 527. Z. Abbadia, J. Amiral, M. C. Trzeciak, P. D. Delmas, and P. Clezardin, The growth-supportive effect of thromobospondin (TSP1) and the expression of TSP1 by human MG-63 osteoblastic cells are both inhibited by dexamethasone. FEBS Lett. 335, 161 – 166 (1993). 528. T. R. Kyriakides, Y. H. Zhu, L. T. Smith, S. D. Bain, Z. Yang, M. T. Lin, K. G. Danielson, R. V. Iozzo, M. LaMarca , C. E. McKinney, E. I. Ginns, and P. Bornstein, Mice that lack thrombospondin 2 display connective tissue abnormalities that are associated with disordered collagen fibrillogenesis, an increased vascular density, and a bleeding diathesis. J. Cell Biol. 140, 419 – 430 (1998). 529. T. R. Kyriakides, Y. H. Zhu, Z. Yang, and P. Bornstein, The distribution of the matricellular protein thrombospondin 2 in tissues of embryonic and adult mice. J. Histochem. Cytochem. 46, 1007 – 1015 (1998).

174 530. M. L. Iruela-Arispe, P. Bornstein, and H. Sage,Thrombospondin exerts an antiangiogenic effect on cord formation by endothelial cells in vitro. Proc. Natl. Acad. Sci. USA 88, 5026 – 5030 (1991). 531. M. L. Iruela-Arispe, D. J. Liska, E. H. Sage, and P. Bornstein, Differential expression of thrombospondin 1, 2, and 3 during murine development. Dev. Dynam. 197, 40 – 56 (1993). 532. G. J. Raugi, J. E. Oleud, and A. M. Gown,Thrombospondin in early human wound tissue. J. Invest. Dermatol. 89, 551 – 554 (1987). 533. M. J. Reed, P. Puolakkainen, T. F. Lane, D. Dickerson et al., Differential expression of SPARC and thrombospondin 1 in wound repair: Immunolocalization and in situ hybridization. J. Histochem. Cytochem. 41, 1467 – 1477 (1993). 534. V. Castle, J. Varani, S. Fligiel, E. V. Prochownik, and V. Dixit, J. Clin Invest. 87, 1883 – 1888 (1991). 535. D. J. Good, P. J. Poiverini, F. Rastinejad, M. M. Le Beau, R. S. Lemons, W. Frazier, and N. P. Bouck, A tumor suppression dependent-inhibitors of angiogeneses in immunologically and functionally indistinguishable from a fragment of thrombospondin. Proc. Natl. Acad. Sci. USA 87, 6624 – 6628 (1990). 536. S. S. Tolsma, O. V. Volpert, D. J. Good, W. A. Frazier, P. J. Polverini, and N. Bouck, Peptides derived from two spearate domains of the matrix protein thrombospondin-1 have anti-antiogenic activity. J. Cell Biol. 122, 497 – 511 (1993). 537. J. A. Carron, W. D. Fraser, and J. A. Gallagher, Thrombospondin promotes resorption by osteoclasts in vitro. Biochem. Biophys. Res. Commun. 213, 1017 – 1025 (1995). 538. S. Schultz-Cherry, H. Chen, D. F. Mosher, T. M. Misenheimer, H. C. Krutzsch, D. D. Roberts, and J. E. Murphy-Ullrich, Regulation of transforming growth factor- activation by discrete sequences of thrombospondin 1. J. Biol. Chem. 270, 7304 – 7310 (1995). 539. M. Winnemoller, P. Schîn, P. Vischer, and H. Kresse, Interactions between thrombospondin and the small proteoglycan decorin: Interference with cell attachment. Eur. J. Biol. 59, 47 – 55 (1992). 540. G. Gao and W. A. Frazier, Identification of a receptor candidate for the carboxyl-terminal cell binding domain of thrombospondins. J. Biol. Chem. 269, 29650 – 29657 (1994). 541. C. A. Prater, J. Plotkin, D. Jaye, and W. A. Frazier, The properdinlike type I repeats of human thrombospondin contain a cell attachment site. J. Cell Biol. 112, 1031 – 1040 (1991). 542. W. X. Li, R. J. Howard, and L. L. K. Leung, Identification of SVTCG in thrombospondin as the conformation-dependent, high affinity binding site for its receptor, CD36. J. Biol. Chem. 268, 16179 – 16184 (1993). 543. A. S. Asch, I. Liu, F. M. Briccetti, J. W. Barnwell, F. KwakyeBerko, A. Dokun, J. Goldberger, and M. Pernambuco, Analysis of CD36 binding domains: Ligand specificity controlled by dephosphorylation of an ectodomain. Science 262, 1436 – 1440 (1993). 544. T. S. Panetti, B. J. Kudryk, and D. F. Mosher, Interaction of recombinant procollagen and properdin modules of thrombospondin-1 with heparin and fibrinogen/fibrin. J. Biol. Chem. 274, 430 – 437 (1999). 545. M. Grano, P. Zigrino, S. Colucci, G. Zambonin et al., Adhesion properties and integrin expression of cultured human osteoclast-like cells. Exp. Cell Res. 212, 209 – 218 (1994). 546. M. A. Horton, E. L. Dorey, S. A. Nesbitt, J. Samanen, F. E. Ali, J. M. Stadel, A. Nichols, R. Greig, and M. H. Helfrich, Modulation of vitronectin receptor-mediated osteoclast adhesion by Arg-GlyAsp peptide analogs: A structure-function analysis. J. Bone Miner. Res. 8(2), 239 – 247 (1993). 547. R. Hynes, Molecular biology of fibronectin, Annu. Rev. Cell Biol. 1, 67 – 90 (1985). 548. D. F. Mosher (ed.), “Fibronectin,” Academic Press, New York, 1989. 549. R. E. Weiss and A. H. Reddi, Appearance of fibronectin during the differentiation of cartilage, bone and bone marrow. J. Cell Biol. 88, 630 (1981).

GOKHALE, BOSKEY, AND ROBEY 550. M. J. Weiss, P. S. Henthorn, M. A. Lafferty, C. Slaughter, M. Raducha, and H. Harris, Isolation and characterization of a cDNA encoding a human liver/bone/kidney-type alkaline phosphatase. Proc. Natl. Acad. Sci. USA 83, 7182 – 7186 (1986). 551. D. J. Leahy, H. P. Erickson, I. Aukhil, P. Joshi, and W. A. Hendricson, Crystallization of a fragment of human fibronectin: Introduction of methionine by site-directed mutagenesis to allow phasing via selenomethionine. Proteins 19, 48 – 54 (1994). 552. A. M. Krezel, G. Wagner, J. Seymour- Ulmer, and R. A. Lazarus, Structure of the RGD protein decorsin: Conserved motif and distinct function in leech proteins that affect blood clotting. Science 264, 1944 – 1947 (1994). 553. R. F. Cachau, E. H. Serpersu, A. S. Mild van, T. August, and L. M. Amzel, Recognition in cell adhesion: A compulsive study of RGDcontaining peptides by Monte Carlo and NMR methods. J. Mol. Recog. 2, 179 – 186 (1989). 554. H. P. Erickson, Reversible unfolding of fibronectin type III and immunoglobulin domains provides the structural basis for stretch and elasticity of titin and fibronectin. Proc. Natl. Acad. Sci. USA 21, 10114 – 10118 (1994). 555. J. E. Schwarzbauer, Alternatives splicing of fibronectin: Three variants, three functions. Bioessays 13, 527 – 533 (1991). 556. J. E. Schwarzbauer, Identification of the fibronectin sequences required for assembly of a fibrillar matrix. J. Cell Biol. 113, 463 – 473 (1991). 557. F. J. Barkalow and J. E. Schwarzbauer, Interactions between fibronectin and chondroitin sulfate are modulated by molecular context. J. Biol. Chem. 269, 3957 – 3962 (1994). 558. K. Ichihara-Tanaka, T. Maeda, K. Titani, and K. Sekiguchi, Matrix assembly of recombinant fibronectin polypeptide consisting of amino-terminal 70 kDa and carboxyl-terminal 37 kDa regions. FEBS Lett. 299(2), 155 – 158 (1992). 559. J. Sottile, J. Schwarzbauer, J. Seleque, and D. F. Mosher, Five type I modules of fibronectin form a functional unit that binds to fibroblasts and Staphylococcus aureus. J. Biol. Chem. 266, 12840 – 12843 (1991). 560. T. Darribere, V. E. Koteliansky, M. A. Chernousov, and S. K. Akiyama, Distinct regions of human fibronectin are essential for fibril assembly in an in vivo developing system. Dev. Dyn. 194, 63 – 70 (1992). 561. K. Vibe-Pedersen, S. Magnusson, and F. E. Baralle, Donor and acceptor splice signals within an exon of the human fibronectin gene: A new type of differential splicing. FEBS Lett. 207, 287 – 291 (1986). 562. H. Hirano, Y. Yamada, M. Sullivan, B. de Crombrugghe, I. Pastan, and K. M. Yamada, Isolation of genomic DNA clones spanning the entire fibronectin gene. Proc. Natl. Acad. Sci. USA 80, 46 – 50 (1983). 563. R. Gardella, M. Colombi, and S. Barlati, Human fibronectin gene (FN1) RFLPs: Mapping and linkage disequilibrium analysis. Hum. Genet. 92, 639 – 641 (1993). 564. D. C. Dean, C. L. Bowlus, and S. Bourgeois, Cloning and analysis of the promoter region of the human fibronectin gene. Proc. Natl. Acad. Sci. USA 84, 1876 – 1880 (1987). 565. D. D. Dean, Z. Schwartz, L. Bonewald, O. E. Muniz, S. Morales, R. Gomez, B. P. Brooks, M. Qiao, D. S. Howell, and B. D. Boyan, Matrix vesicles produced by osteoblast-like cells in culture become significantly enriched in proteoglycan-degrading metalloproteinases after addition of beta-glycerophosphate and ascorbic acid. Calcif. Tissue Int. 54, 399 – 408 (1994). 566. S. Miao, P. K. Suri, L. Shu-Ling, A. Abraham, N. Cook, P. Milos, and M. A. Zern, Role of the cyclic AMP response element in rat fibronectin gene expression. Hepatology 17, 882 – 890 (1993). 567. A. F. Muro, V. A. Bernath, and A. R. Kornblihtt, Interaction of the 170 cyclic AmP response element with the adjacent CCAAT box

CHAPTER 4 The Biochemistry of Bone

568.

569.

570.

571.

572.

573.

574.

575.

576.

577.

578.

579.

580.

581. 582. 583.

584.

585.

in the human fibronectin gene promoter. J. Biol. Chem. 267, 12767 – 12774 (1992). A. Srebrow, A. F. Muro, S. Wergajh, P. A. Sharp, and A. R. Hornblihtt, The CRE-binding factor ATF-2 facilitates the occupation of the CAAT box in the fibronectin gene promoter. FEBS Lett. 327, 25 – 28 (1993). A. R. Kornblihtt, K. Umezawa, K. Vibe-Pedersen, and F. E. Baralle, Primary structure of human fibronectin: Differential splicing may generate at lease 10 polypeptides from a single gene. EMBO J. 4, 1755 – 1759 (1985). R. S. Patel, E. Odermatt, J. E. Schwarzbauer, and R. O. Hynes, Organization of the fibronectin gene provides evidence for exon shuffling during evolution. EMBO J. 6(9), 2565 – 2572 (1987). C. Eielson, D. Kaplan, M. A. Mitnick, I. Paliwal, and K. Insogna, Estrogen modulates parathyroid hormone-induced fibronectin productcion in human and rat osteoblast-like cells. Endocrinology 135, 1639 – 1644 (1994). E. L. George, E. N. Georges-Labouesse, R. S. Patel-King, H. Rayburn, and R. O. Hynes, Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development 119, 1079 – 1091 (1993). R. J. Majeska, M. Port, and T. A. Einhorn, Attachment to extracellular matrix molecules by cells differing in the expression of osteoblastic traits. J. Bone Miner. Res. 8(3), 277 – 289 (1993). D. A. Puleo and R. Bizios, Mechanisms of fibronectin-mediated attachment of osteoblasts to substrates in vitro. Bone Miner. 18, 215 – 226 (1992). F. P. Ross, J. Chappel, J. I. Alvarez, D. Sander, W. T. Butler, M. C. Farach-Carson, K. A. Mintz, P. G. Robey, S. L. Teitelbaum, and D. A. Cheresh, Interactions between the bone matrix proteins osteopontin and bone sialoprotein and the osteoclast integrin alpha v beta 3 potentiate bone resorption. J. Biol. Chem. 268, 9901 – 9907 (1993). G. Daculsi, P. Pilet, M. Cottrel, and G. Guicheux, Role of fibronectin during biological apatite crystal nucleation: Ultrastructural characterization. J. Biomed. Mater. Res. 47(2), 228 – 233 (1999) G. Daculsi, P. Pilet, M. Cottrel, and G. Guicheux, Role of fibronectin during biological apatite crystal nucleation: Ultrastructural characterization. J. Biomed. Mater. Res. 47, 228 – 233 (1997). D. Seiffert, T. J. Podor, and D. J. Loskutoff, Distribution of Vitronectin. In “Biology of Vitronectins and Their Receptors” (K. T. Preissner, S. Rosenblatt, C. Kost, J. Wegerhoff, and D. F. Mosher, eds.), pp. 75 – 81. Excerpta Medica, Amsterdam, 1993. E. G. Hayman, M. D. Pierschbacher, S. Suzuki, and E. Ruoslahti, Serum spreading factor (vitronectin) is present at the cell surface and in tissues. Proc. Natl. Acad. Sci. USA 80, 4003 – 4007 (1983). T. Kumagai, I. Lee, Y. Ono, M. Maeno, and M. Takagi, Ultrastructural localization and biochemical characterization of vitronectin in developing rat bone. Histochem. J. 30, 111 – 119 (1998). K. T. Preissner, Structure and biological role of vitronectin. Annu. Rev. Cell. Biol. 7, 275 – 310 (1991). K. T. Preissner and D. Jenne, Structure of vitronectin and its biological role in haemostasis. Throm. Haemost. 66, 123 – 132 (1991). A. Oldberg, E. G. Hayman, M. D. Pierschbacher, and E. Ruoslahti, Complete amino acid sequence of human vitronectin deduced from cDNA: Similarity of cell attachment sites in vitronectin and fibronectin. EMBO J. 4, 2519 – 2524 (1985). D. Jenne and K. K. Stanley, Molecular cloning of S-protein, a link between complement coagulation and cell-substrate adhesion. EMBO J. 4, 3153 – 3157 (1985). H. J. Ehrlich, B. Richter, D. von der Ahe, and K. T. Preissner, Primary structure of vitronectins and homology with other proteins. In “Biology of Vitronectins and their Receptors” (K. T. Preissner, S. Rosenblatt, C. Kost, J. Wegerhoff, and D. F. Mosher, eds.), pp. 59 – 66. Excerpta Medica, Amsterdam, 1993.

175 586. K. T. Preissner and D. Jenne, Vitronectin: A new molecular connection in haemostasis. Thromb. Haemost. 66(2), 189 – 194 (1991). 587. J. W. Smith, D. J. Vestal, S. V. Irwin, T. A. Burke, and D. A. Cheresh, Purification and functional characterization of integrin alpha v beta 5. An adhesion receptor for vitronectin. J. Biol. Chem. 265(19), 11008 – 11013 (1990). 588. K. P. Mintz, W. J. Grzesik, R. J. Midura, P. G. Robey, J. D. Termine, and L. W. Fisher, Purification and fragmentation of nondenatured bone sialoprotein: Evidence for a cryptic, RGD-resistant cell attachment domain. J. Bone Miner. Res. 8, 985 – 995 (1993). 589. J. Davies, J. Warwick, N. Totty, R. Philip, M. Helfrich, and M. Horton, The osteoclast functional antigen, implicated in the regulation of bone resorption, is biochemically related to the vitronectin receptor. J. Cell Biol. 109, 1817 – 1826 (1980). 590. R. Baron, J.-H. Ravesloot, L. Neff, M. Chakraborty, D. Chatterjee, A. Lomri, and W. Horne, Cellular and molecular biology of the osteoclast. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), p. 445 – 495. Academic Press, San Diego, 1993. 591. R. P. M. Rohanizadeh, J. M. Bouler, D. Couchourel, Y. Fortun, and G. Daculsi, Apatite precipitation after incubation of biphasic calcium-phosphate ceramic in various solutions: Influence of seed species and proteins. J. Biomed. Mater. Res. 42, 530 – 539 (1998). 592. C. W. Prince, T. Oosawa, W. T. Butler, M. Tomana, A. S. Bhon, and R. E. Schrohenloher, Isolation, characterization and biosynthesis of a phosphorylated glycoprotein formed at bone. J. Biol. Chem. 262, 2900 – 2907 (1987). 593. A. Franzen and D. Heinegard, Isolation of new phosphorylated glycoprotein from mineralized phase of bone that exhibits limited homology to adhesive protein osteopontin. J. Biol. Chem. 252, 15938 – 15945 (1988). 594. A. Oldberg, A. Franzen, and D. Heinegard, The primary structure of a cell-binding bone sialoprotein. J. Biol. Chem. 263, 19430 – 19432 (1988). 595. J. H. Smith and D. T. Denhardt, Molecular cloning of a tumor promoter-inducible mRNA found in JB6 mouse epidermal cells. J. Cell Biochem. 34, 13 – 22 (1987). 596. G. R. Laverdure, D. Banerjee, I. Chackalaparampil, and B. B. Mukerjee, FEBS Lett. 222, 261 – 265 (1987). 597. D. R. Senger, C. A. Perruzzi, and A. Papadopoulos, Elevated expression of secreted phosphoprotein 1 (osteopontin, 2ar) as a consequence of neoplastic transformation. Anticancer Res. 9, 1291 – 1300 (1989). 598. A. M. Craig, M. Nemir, B. B. Mukerjee, A. F. Chambers, and D. T. Denhardt, Identification of the major phosphoprotein secreted by many rodent cell lines as 2ar/osteopontin. Biochem. Biophys. Res. Commun. 157, 166 – 173 (1988). 599. A. F. Chambers, E. I. Behrend, S. M. Wilson, and D. T. Denhardt, Induction of expression of osteopontin (OPN; secreted phosphoprotein) in metastatic, ras-transformed NIH 3T3 cells. Anticancer Res. 12(1), 43 – 47 (1992). 600. C. A. Lopez, J. R. Hoyer, P. D. Wilson, P. Waterhouse, and D. T. Denhardt, Heterogeneity of osteopontin expression among nephrons in mouse kidneys and enhanced expressionin sclerotic glomeruli. Lab. Invest. 69, 355 – 363 (1993). 601. H. Shiraga, W. Min, W. J. van Dusen, M. D. Clayman, D. Miner, C. L. T. Terrell, J. R. Sherbotie, J. W. Foreman, C. Pryzysiecki, E. G. Nelson, and J. R. Hoyer, Inhibition of calcium oxalate crystal growth in vitro by uropontin: Another member of the aspartic acid-rich protein superfamily, Proc. Natl. Acad. Sci. USA 89, 426 – 430 (1992). 602. E. M. Worcester, S. S. Blumenthal, A. M. Beshensky, and D. L. Lewand, The calcium oxalate crystal growth inhibitor protein produced by mouse kidney cortical cells in culture is osteopontin. JBMR 7, 1029 – 1036 (1992). 603. K. Kohr, Y. Suzuki, K. Yoshida, and K. Yamamoto, Molecular cloning and sequencing of cDNA encoding urinary stone protein

176

604.

605.

606.

607.

608.

609.

610.

611.

612.

613. 614.

615.

616.

617.

618.

619.

620.

GOKHALE, BOSKEY, AND ROBEY which is identical to osteopontin. Biochem. Biophys. Res. Commun. 184, 859 – 864 (1992). D. R. Senger, C. A. Perruzzi, A. Papadopoulos, and D. G. Tenen, Purification of a human milk protein closely similar to tumor-secreted phosphoproteins and osteopontin. Biochim. Biophys. Acta 996(1 – 2), 43 – 48 (1989). P. Waterhouse, R. S. Parhar, X. Guo, R. K. Lala, and D. T. Denhardt, Regulated temporal and spatial expression of the calciumbinding proteins calcyclin and OPN (osteopontin) in mouse tissues during pregnancy. Mol. Reprod. Dev. 32, 315 – 323 (1992). L. F. Brown, B. Berse, L. Vande Water, A. Papadopoulos-Sergion, C. A. Perruzzi, E. J. Manseau, H. F. Dvorak, and D. R. Senger, Expression and distribution of osteopontin in human tissues; widespread association with luminal epithelial surfaces. Mol. Biol. Cell. 3, 1169 – 1180 (1992). X. Guo, Y. P. Zhang, D. A. Mitchell, D. T. Denhardt, and A. F. Chambers, Identification of a ras-activated enhancer in the mouse osteopontin promoter and its interaction with a putative ETS-related transcription factor whose activity correlates with the metastatic potential of the cell. Mol. Cell Biol. 15(1), 476 – 487 (1995). K. Yoon, R. Buenaga, and G. A. Rodan, Tissue specificity and developmental expression of rat osteopontin. Biochem. Biophys. Res. Commun. 148, 1129 – 1136 (1987). M. P. Mark, W. T. Butler, C. W. Prince et al., Developmental expression of 44-kDa bone phosphoprotein (osteopontin) and bone gamma-carboxyglutamic acid (Gla)-containing protein (osteocalcin) in calcifying tissues fo rat. Differentiation 37, 123 (1988). M. F. P. Mark, C. W. Prince, S. Gay et al., 44-kDal bone phosphoprotein (osteopontin) antigenicity at ectopic sites in newborn rats: Kidney and nervous tissues. Cell Tissue Res. 251, 23 (1988). S. L. Smock, D. N. Petersen, W. A. Grasser, D. T. Denhardt, and M. A. Thiede, The osteoblastic phenotype of rat bone marrow megakaryocytes and peripheral blood platelets. J. Bone Miner. Res. 9(Suppl. 1), S126 (1994). M. P. Mark, C. W. Prince, S. Gay, R. L. Austin, M. Bnown, R. D. Finkelman, and W. T. Butler, A comparative immunocytochemical study on the subcellular distributions of 44 kDa bone phosphoprotein and bone gamma-carboxy-glutamic acid (Gla)-containing protein in osteoblasts. J. Bone Miner. Res. 2(4), 337 – 346 (1987). Deleted at proof. M. P. Mark, C. W. Prince, T. Oosawa, S. Gay, A. L. Bronkers, and W. T. Butler, Immunohistochemical demonstration of a 44-kD phosphoprotein in developing rat bones. J. Histochem. Cytochem. 35, 707 – 715 (1987). J. Chen, K. Singh, B. B. Mukherjee, and J. Sodek, Developmental expression of osteopontin (OPN) mRNA in rat tissues: Evidence for a role for OPN in bone formation and resorption. Matrix 13, 113 – 123 (1993). J. Chen, Q. Zhang, C. A. G. McCulloch, and J. Sodek, Immunohistochemical localization of bone sialoprotein in fetal porcine bone tissues; comparison with secreted phosphoprotein 1 (SPP-1, osteopontin) and SPARC (osteonectin). Histochem. J. 23, 281 – 289 (1991). L. J. Suva, J. B. Seedor, N. Endo, H. A. Quartuccio, D. D. Thompson, I. Bab, and G. A. Rodan, Pattern of gene expression following rat tibial marrow ablation. J. Bone Miner. Res. 8, 379 – 388 (1993). R. Quarto, D. Thomas, and C. T. Liang, Bone progenitor cell deficits and the age-associated decline in bone repair capacity. Calcif. Tissue Int. 56, 123 – 129 (1995). M. D. McKee, M. C. Farach-Carson, W. T. Butler, P. V. Hauschka, and A. Nanci, Ultrastructural immunolocalization of noncollagenous (osteopontin and osteocalcin) and plasma (albumin and alpha2 HS-glycoprotein) proteins in rat bone. J. Bone Miner. Res. 8, 485 – 496 (1993). J. B. Lian, M. D. McKee, A. M. Todd, and L. C. Gerstenfeld, Induction of bone-related proteins, osteocalcin and osteopontin, and

621.

622.

623.

624.

625.

626.

627.

628.

629. 630.

631. 632.

633.

634.

635.

636.

637.

638.

their matrix ultrastructural localization with development of chondrocyte hypertrophy in vitro. J. Cell Biochem. 52, 206 – 219 (1993). C. S. Carlson, H. M. Tulli, M. J. Jayo, R. F. Loeser, R. P. Tracy, K. G. Mann, and M. R. Adams, Immunolocalization of noncollagenous bone matrix proteins in lumbar vertebrae from intact and surgically menopausal cynomogous monkeys. J. Bone Miner. Res. 8, 71 – 81 (1993). M. Galotto, G. Campanile, G. Robino, F. D. Cancedda, P. Bianco, and R. Cancedda, Hypertrophic chondrocytes undergo further differentiation to osteoblast-like cells and participate in the initial bone formation in developing chick embryo. J. Bone Miner. Res. 9(8), 1239 – 1249 (1994). K. Merry, R. Dodds, A. Littlewood, and M. Gowen, Expression of osteopontin mRNA by osteoclasts and osteoblasts in modelling adult human bone. J. Cell Sci. 104, 1013 – 1020 (1993). K. Tezuka, T. Sato, H. Kamioka, P. J. Nijweide et al., Identification of osteopontin in isolated rabbit osteoclasts. Biochem. Biophys. Res. Commun. 186, 911 – 917 (1992). M. D. McKee and A. Nanci, Osteopontin and the bone remodeling sequence: Colloidal-gold immunocytochemistry of an interfacial extracellular matrix protein. Ann. N. Y. Acad. Sci. 760, 177 – 189 (1995). M. P. Mark, C. W. Prince, T. Oosawa, S. Gay, A. L. Bronckers, and W. T. Butler, Immunohistochemical demonstration of a 44-KD phosphoprotein in developing rat bones. J. Histochem. Cytochem. 35, 707 – 715 (1987). K. Hultenby, F. P. Reinholt, M. Norgard, A. Oldberg, M. Wendel, and D. Heinegard, Distribution and synthesis of bone sialoprotein in metaphyseal bone of young rats show a distinctly different pattern from that of osteopontin. Eur. J. Cell. Biol. 63, 230 – 239 (1994). A. Franzen and D. Heinegard, Isolation and characterization of two sialoproteins present only in bone calcified matrix. Biochem. J. 232, 715 – 724 (1985). C. W. Prince, Secondary structure predictions for rat osteopontin. Connect. Tissue Res. 21, 15 – 20 (1989). J. M. Kerr, L. W. Fisher, J. D. Termine, and M. F. Young, The cDNA cloning and RNA distribution of bovine osteopontin. Gene 108(2), 237 – 243 (1991). W. T. Butler, The nature and significance of osteopontin. Connect. Tissue Res. 23, 123 – 126 (1989). S. Beninati, D. R. Senger, E. Cordella-Miele, A. B. Mukherjee, I. Chackalaparampil, V. Shanmugam, K. Singh, and B. B. Mukherjee, Osteopontin: Its transglutaminase-catalyzed posttranslational modifications and cross-linking to fibronectin. J. Biochem. (Tokyo) 115, 675 – 682 (1994). H. Mickos, K. Sunderg, and B. Luning, Synthesis, NMR spectra and function of peptides with alpha-methylserine attached to the RGD sequence of osteopontin. Acta Chem. Scand. 46, 989 – 993 (1992). S. Van Dijk, J. A. D’Errico, M. J. Somerman, M. C. Farach-Carson, and W. T. Butler, Evidence that a non-RGD domain in rat osteopontin is involved in cell attachment. J. Bone Miner. Res. 8, 1499 – 1506 (1993). J. A. Evans and S. I. Chan, Phosphophoryn, a biomineralization template protein: pH-dependent protein folding experiments. Biopolymers 34, 507 – 527 (1994). J. S. Evans, T. Chio, and S. I. Chan, Phosphophoryn, an “acidic” biomineralization regulatory protein: Conformational folding in the presence of Cd (II). Biopolymers 34, 1359 – 1375 (1994). S. Ashkar, D. B. Teplow, M. J. Glimcher, and R. A. Saavedra, In vitro phosphorylation of mouse osteopontin expressed in E coli. Biochem. Biophys. Res. Commun. 191, 126 – 133 (1993). J. P. Gorski, Acidic phosphoproteins from bone matrix: A structural rationalization of their role in biomineralization. Calcif. Tissue Int. 50, 391 – 396 (1992).

177

CHAPTER 4 The Biochemistry of Bone 639. N. Hijiya, M. Setoguchi, K. Matsuura, Y. Higuchi, S. Akizuki, and S. Yamamoto, Cloning and characterization of the human osteopontin gene and its promoter. Biochem. J. 303, 255 – 262 (1994). 640. K. Rafidi, I. Simkina, E. Johnson, M. A. Moore, and L. C. Gerstenfeld, Characterization of the chicken osteopontin-encoding gene. Gene 140, 163 – 169 (1994). 641. E. I. Behrend, A. F. Chambers, S. M. Wilson, and D. T. Denhardt, Comparative analysis of two alternative first exons reported for the mouse osteopontin gene. J. Biol. Chem. 268, 11172 – 11175 (1993). 642. A. M. and Craig, D. T. Denhardt, The murine gene encoding secreted phosphoprotein 1 (osteopontin): Promoter structure, activity, and induction in vivo by estrogen and progesterone. Gene 100, 163–171 (1991). 643. Y. Miyazaki, M. Setoguchi, S. Yoshida, Y. Higuchi, S. Akizuki, and S. Yamamoto, The mouse osteopontin gene: Expression in monocytic lineages and complete nucleotide sequence. J. Biol. Chem. 265, 14432 – 14438 (1990). 644. M. F. Young, J. M. Kerr, J. D. Termine, U. M. Wewer, M. G. Wang, O. W. McBride, and L. W. Fisher, cDNA cloning, mRNA distribution and heterogeneity, chromosomal location, and RFLP analysis of human osteopontin (OPN). Genomics 7, 491 – 502 (1990). 645. R. Patarca, R. A. Saavedra, and H. Cantor, Molecular and cellular basis of genetic resistance to bacterial infection: The role of the early T-lymphocyte activation-1/osteopontin gene. Crit. Rev. Immunol. 13, 225 – 246 (1993). 646. M. C. Kieffer, D. M. Aauer, and P. J. Barr, The cDNA and derived amino acid sequence for human osteopontin. Nucleic Acids Res. 17, 3306 (1989). 647. D. T. Denhardt and X. Guo, Osteopontin: A protein with diverse functions. FASEB J. 7, 1475 – 1482 (1993). 648. J. Sodek, J. Chen, S. Kasugai, T. Nagata, Q. Zhang, M. D. McKee, and A. Nanci, Elucidating the functions of bone sialoprotein and osteopontin in bone formation. In “The Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 297 – 306. Elsevier, Amsterdam, 1992. 649. M. Noda and D. T. Denhardt, Regulation of osteopontin gene expression in osteoblasts. Ann. N. Y. Acad. Sci. 760, 242 – 248 (1995). 650. C. W. Prince and W. T. Butler, 1,25-Dihydroxyvitamin D3 regulates the biosynthesis of osteopontin, a bone-derived cell attachment protein, in clonal osteoblast-like osteosarcoma cells. Coll. Relat. Res. 7, 305 (1987). 651. P. L. Chang and C. W. Prince, 1 alpha, 25-Dihydroxyvitamin D3 enhances 12-O-tetradecanoylphorbol-13 acetate-induced tumorigenic transformation and osteopontin expression in mouse JB6 epidermal cells. Cancer Res. 53, 2217 – 2220 (1993). 652. M. Noda, K. Yoon, C. W. Prince, W. T. Butler, andd G. A. Rodan, Transcriptional regulation of osteopontin production in rat osteosarcoma cells by type  transforming growth factor. J. Biol. Chem. 263, 13916 – 13921 (1988). 653. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini et al., Effects of transforming growth factor beta on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, osteopontin, alkaline phosphatase, and type I collagen mRNA in longterm cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 654. K. Yoon, R. Buenaga, and G. A. Rodan, Tissue specificity and developmental expression of rat osteopontin. Biochem. Biophys. Res. Commun. 148(3), 1129 – 1136 (1987). 655. M. Noda and G. A. Rodan, Transcriptional regulation of osteopontin production in rat osteoblast-like cells by parathyroid hormone. J. Cell. Biol. 108, 713 (1989). 656. M. Shioide and M. Noda, Endothelin modulates osteopontin and osteocalcin messenger ribonucleic acid expression in rat osteoblastic osteosarcoma cells. J. Cell. Biochem. 53, 176 – 180 (1993). 657. T. Kubota, M. Yamauchi, J. Onozaki, S. Sato et al., Influence of an intermittent compressive force on matrix protein expression by

658.

659. 660.

661.

662.

663.

664.

665.

666.

667.

668.

669.

670.

671.

672.

673.

674.

675.

ROS 17/2.8 cells, with selective stimulation of osteopontin. Arch. Oral Biol. 38, 23 – 30 (1993). Kaji, T. Sugimoto, A. Miyauchi, M. Fukase et al., Calcitonin inhibits osteopontin mRNA expression in isolated rabbit osteoclasts. Endocrinology 135, 484 – 487 (1994). S. T. Li and E. Katz, On the state of demineralized matrices of bone and dentin. Calcif. Tissue Res. 22, 275 – 284 (1977). A. Endo, Potential role of phosphoprotein in collagen mineralization: An experimental study in vitro. J. Orthop. Assoc. 61, 563 – 569 (1987). M. J. Glimcher, Mechanisms of calcification in bone: Role of collagen fibrils and collagen-phosphoprotein complexes in vitro and in vivo. Anat. Rec. 224, 139 – 153 (1989). A. Nanci, Content and distribution of noncollagenous matrix proteins in bone and cementum: Relationship to speed of formation and collagen packing density. J. Struct. Biol. 126, 256 – 269 (1999). A. L. Boskey, M. Maresca, W. Ullrich, S. B. Doty, W. T. Butler, and C. W. Prince, Osteopontin-hydroxyapatite interactions in vitro: Inhibition of hydroxyapatite formation and growth in a gelatin-gel. Bone Miner. 22, 147 – 159 (1993). G. K. Hunter, C. L. Kyle, and H. A. Goldberg, Modulation of crystal formation by bone phosphoproteins: Structural specificity of the osteopontin-mediated inhibition of hydroxyapatite formation. Biochem. J. 300, 723 – 728 (1994). G. K. Hunter and H. A. Goldberg, Nucleation of hydroxyapatite by bone sialoprotein. Proc. Natl. Acad. Sci. USA 90, 8562 – 8565 (1993). K. Singh, A. B. Mukherjee, and M. W. DeVouge, Physiological properties and differential glycosylation of phosphorylated and non-phosphorylated forms of osteopontin secreted by normal rat kidney cells. J. Biol. Chem. 265, 18696 – 18701 (1992). A. L. Boskey, M. Maresca, S. Doty, B. Sabsay, and A. Veis, Concentration dependent effect of dentin-phosphophoryn in the regulation of in vitro hydroxyapatite formation and growth. Bone Miner. 11, 55 – 65 (1990). M. Bacarani-Contri, F. Taparelli, and I. Pasquali-Ronchetti, Osteopontin is a constitutive component of normal elastin fibers in human skin and aorta. Matrix Biol., in press. T. Wada, M. D. McKee, S. Steitz, and C. M. Giachelli, Calcification of vascular smooth muscle cell cultures: Inhibition by osteopontin. Circ. Res. 84(2), 166 – 178 (1999). A. Oldberg, A. Franzen, D. Heinegard, M. Pierschbacher, and E. Ruoslahti, Identification of a bone sialoprotein receptor in ostesarcoma cells. J. Biol. Chem. 263, 19433 – 19436 (1988). F. P. Reinholt, K. Hultenby, A. Oldberg, and D. Heinegard, Osteopontin: A possible anchor of osteoclasts to bone surface. Proc. Natl. Acad. Sci. USA 87, 4473 – 4475 (1990). A. Miyauchi, J. Alvarez, E. M. Greenfield, A. Teti, M. Grano, S. Colucci, A. Zambonin-Zallone, F. P. Ross, S. L. Teitelbaum, and D. Cheresh, Binding of osteopontin to the osteoclast integrin alpha v beta 3. Osteoporos Int. 3(Suppl. 1,) 132 – 135 (1993). B. Ek-Rylander, M. Flores, W. Wendel, D. Heinegard, and G. Andersson, Dephosphorylation of osteopontin and bone sialoprotein by osteoclastic tartrate-resistant acid phosphatase: Modulation of osteoclast adhesion in vitro. J. Biol. Chem. 269, 14853 – 14856 (1994). S. Hwang, C. A. Lopez, D. E. Heck, C. R. Gardner, D. L. Laskin, J. D. Laskin, and D. T. Denhardt, Osteopontin inhibits induction of nitric oxide synthase gene expression by inflammatory mediators in mouse kidney epithelial cells. J. Biol. Chem. 269, 711 – 715 (1994). S. H. Ralston, D. Todd, M. Helfrich, N. Benjamin, and P. S. Grabowski, Human osteoblast-like cells produce nitric oxide and express inducible nitric oxide synthase. Endocrinology 135(1), 330 – 336 (1994).

178 676. S. R. Rittling, H. N. Matsumoto, M. D. McKee, A. Nanci, X. R. An, K. E. Novick, A. J. Kowalski, M. Noda, and D. T. Denhardt, Mice lacking osteopontin show normal development and bone structure but display altered osteoclast formation in vitro. J. Bone Miner. Res. 13, 1101 – 1111 (1998). 677. A. L. Boskey and E. P. Paschalis, Matrix proteins and biomineralization. In “Engineering Bone” (J. E. Davies, ed.), in press. 678. D. T. Denhardt and M. Noda, Osteopontin expression and function: Role in bone remodeling. J. Cell. Biochem. Suppl. 30-31, 92 – 102 (1998). 679. L. A. Fitzpatrick, A. Severson, W. D. Edwards, and R. T. Ingram, Diffuse calcification in human coronary arteries: Association of osteopontin with atherosclerosis. J. Clin. Invest. 94, 1597 – 1604 (1994). 680. A. Bellahcene and V. Castronovo, Increased expression of osteonectin and osteopontin, two bone matrix proteins, in human breast cancer. Am. J. Pathol. 146, 95 – 100 (1995). 681. S. Hirota, A. Ito, J. Nagoshi, M. Takeda, A. Kurata, Y. Takatsuka, K. Kohri, S. Nomura, and Y. Kitamura, Expression of bone matrix protein messenger ribonucleic acids in human breast cancers: Possible involvement of osteopontin in development of calcifying foci. Lab. Invest. 72, 64 – 69 (1995). 682. I. Carlson, K. Tognazzi, E. J. Manseau, H. F. Dvorak, and L. F. Brown, Osteopontin is strongly expressed by histiocytes in granulomas of diverse etiology. Lab. Invest. 77,103 – 108 (1997). 683. A. T. Andrews, G. M. Herring, and P. W. Kent, Some studies on the composition of bovine cortical-bone sialoprotein. Biochem J. 104(3), 705 – 715 (1967). 684. L. W. Fisher, S. W. Whitson, L. V. Avioli, and J. D. Termine, Matrix sialoprotein of developing bone. J. Biol. Chem. 258, 12723 – 12727 (1983). 685. A. Oldberg, A. Franzen, and D. Heinegard, Primary structure of bone sialoprotein. J. Biol. Chem. 263, 19430 – 19432 (1988). 686. C. Chenu and P. D. Delmas, Platelets contribute to circulating levels of bone sialoprotein in humans. J. Bone Miner. Res. 7, 47 – 54 (1992). 687. M. A. Thiede, S. L. Smock, D. N. Petersen, W. A. Grasser, D. D. Thompson, and S. K. Nishimoto, Presence of messenger ribonucleic acid encoding osteocalcin, a marker of bone turnover, in bone marrow megakaryocytes and peripheral blood platelets. Endocrinology 135, 929 – 937 (1994). 688. Q. Zhang, C. Domenicucci, H. A. Goldberg, J. L. Wrana, and J. Sodek, J. Biol. Chem. 262, 9702 – 9708 (1990). 689. P. Bianco, L. W. Fisher, M. F. Young, J. D. Termine, and P. Gehron Robey, Expression of bone sialoprotein (BSP) in developing human tissues. Calcif. Tissue Int. 49, 421 – 426 (1991). 690. J. Chen, H. S. Shapiro, and J. Sodek, Developmental expression of bone sialoprotein mRNA in rat mineralized connective tissues. J. Bone Miner. Res. 7, 987 – 997 (1992). 691. J. Chen, M. D. McKee, A. Nanci, and J. Sodek, Bone sialoprotein mRNA expression and ultrastructural localization in fetal porcine calvarial bone: Comparisons with osteopontin. Histochem. J. 26, 67 – 78 (1994). 692. A. Bellahcene, N. Antoine, N. Clausse, E. Tagliabue, L. W. Fisher, J. M. Kerr, P. Jares, and V. Castronovo, Detection of bone sialoprotein in human breast cancer tissue and cell lines at both protein and messenger ribonucleic acid levels. Lab. Invest. 75, 203 – 210 (1996). 693. V. Castronovo and A. Bellahcene, Evidence that breast cancer associated microcalcifications are mineralized malignant cells. Int. J. Oncol. 12, 305 – 308 (1998). 694. B. Ecarot-Charrier, F. Bouchard, and C. Delloye, Bone sialoprotein II synthesized by cultured osteoblasts contains tyrosine sulfate. J. Biol. Chem. 264, 20049 – 20053 (1989). 695. R. J. Midura, D. J. McQuillan, K. J. Benham, L. W. Fisher, and V. C. Hascall, A rat osteogenic cell line (UMR106-01) synthesizes a

GOKHALE, BOSKEY, AND ROBEY

696.

697.

698.

699.

700.

701.

702.

703. 704.

705.

706.

707.

708.

709.

710.

711.

712.

713.

highly sulfated form of bone sialoprotein. J. Biol. Chem. 265, 5285 – 5291 (1990). H. S. Shapiro, J. Chen, J. L. Wrana, Q. Zhang, M. Blum, and J. Sodek, Characterization of porcine bone sialoprotein: Primary structure and cellular expression. Matrix 13, 431 – 440 (1993) R. W. Kinne and L. W. Fisher, Keratan sulfate proteoglycan in rabbit compact bone is bone sialoprotein II. J. Biol. Chem. 262, 10206 – 10211 (1987). L. W. Fisher and H. Schraer, The glycosaminoglycans of estrogeninduced medullary bone in Japanese quail. Arch. Biochem. Biophys. 205, 396 – 403 (1980). M. J. Somerman, L. W. Fisher, R. A. Foster, and J. J. Sauk, Human bone silaoproteins I and II enhance fibroblast attachment in vitro. Calcif. Tissue Int. 43, 50 – 53 (1988). L. W. Fisher, Structure/function studies of the sialoproteins and proteoglycans of bone: It is still the early days. In “The Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 177 – 186. Elsevier, Amsterdam, 1992. A. Oldberg, C. Ryden, A. I. Yacoub, I. Maxe, and D. Heinegard, Specific binding of bone sialoprotein to Staphylococcus aureus isolated from patients with osteomyelitis. Eur. J. Biochem. 184(2), 331 – 336. L. W. Fisher, O. W. McBride, J. D. Termine, and M. F. Young, Human bone sialoprotein: Deduced protein sequence and chromosomal localization. J. Biol. Chem. 265, 2347–2351 (1990). J. J. Li and J. Sodek, Cloning and characterization of the rat bone sialoprotein gene promoter. Biochem. J. 289(Pt 3), 625 – 629 (1993). J. M. Kerr, L. W. Fisher, J. D. Termine, M. G. Wang, O. W. McBride, and M. F. Young, The human bone sialoprotein gene (IBSP): Genomic localization and characterization. Genomics 17, 408 – 415 (1993). R. H. Kim, H. S. Shapiro, J. J. Li, J. L. Wrana, and J. Sodek, Characterization of the human bone sialoprotein (BSP) gene and its promoter sequence. Matrix Biol. 14, 31 – 40 (1994). A. Bianchi, R. D. Wells, N. H. Heintz, and M. S. Caddle, Sequences near the origin of replication of the DHFR locus of Chinese hamster ovary cells adopt left-handed Z-DNA and triplex structures. J. Biol. Chem. 265, 21789 – 21796 (1990). M. F. Young, K. Ibaraki, J. M. Kerr, and A. M. Heegaard, Molecular and cellular biology of the major noncollagenous proteins in bone. In (M. Noda, ed.), pp. 191 – 234. J. M. Kerr, D. R. Hiscock, W. Grzesik, P. G. Robey, and M. F. Young, The human bone sialoprotein gene contains an NF-E1/YY1 cis-acting sequence with putative regulatory activity. Calcif. Tissue Int. 60, 276 – 282 (1997). A. Oldberg, B. Jirskog-Hed, S. Axelsson, and D. Heinegard, Regulation of bone sialoprotein mRNA by steroid hormones. J. Cell Biol. 109, 3183 – 3186 (1989). J. Chen, H. S. Shapiro, J. L. Wrana, S. Reimers, J. N. M. Heersche, and J. Sodek, Localization of bone sialoprotein (BSP) expression to sites of mineralized tissue formation in fetal rat tissues by in situ hybridization. Matrix 11, 133 – 143 (1991). J. Chen, C. A. G. McCulloch, and J. Sodek, Bone sialoprotein in developing porcine dental tissues: Cellular expression and comparison of tissue localization with osteopontin and osteonectin. Arch. Oral Biol. 38, 241 – 249 (1993). G. K. Hunter and H. A. Goldberg, Modulation of crystal formation by bone phosphoproteins: Role of glutamic acid-rich sequences in the nucleation of hydroxyapatite by bone sialoprotein. Biochem. J. 302, 175 – 179 (1994). J. T. Stubbs, Jr., K. P. Mintz, E. D. Eanes, D. A. Torchia, and L. W. Fisher, Characterization of native and recombinant bone sialoprotein: Delineation of the mineral-binding and cell adhesion domains and structural analysis of the RGD domain. J. Bone Miner. Res. 12, 1210 – 1222 (1997).

CHAPTER 4 The Biochemistry of Bone 714. M. E. Flores, M. Norgard, D. Heinegard, F. P. Reinholt, and G. Andersson, RGD-directed attachment of isolated rat osteoclasts to osteopontin, bone sialoprotein, and fibronectin. Exp. Cell Res. 201, 526 – 530 (1992). 715. K. P. Mintz, L. W. Fisher, W. J. Grzesik, V. C. Hascall, and R. J. Midura, Chlorate-induced inhibition of tyrosine sulfation on bone sialoprotein synthesized by a rat osteoblast-like cell line (UMR 106-01 BSP). J. Biol. Chem. 269, 4845 – 4852 (1994). 716. R. Fujisawa and Y. Kuboki, Affinity of bone sialoprotein and several other bone and dentin acidic proteins to collagen fibrils. Calcif. Tissue Int. 51, 438 – 442 (1992). 717. A. L. Arsenault and F. P. Ottensmeyer, Quantitative spatal destributions of calcium, phosphorus and sulfur in calcifying epiphysis by high resolution electron spectroscopic imaging. Proc. Natl. Acad. Sci. USA 80, 1322 – 1326 (1983). 718. A. Bellahcene, M. P. Merville, and V. Castronovo, Expression of bone sialoprotein, a bone matrix protein, in human breast cancer. Anticancer Res. 54, 2823 – 2826 (1994). 719. D. Waltregny, A. Bellahcene, X. de Leval, B. Florkin, U. Weidle, and V. Castronovo, Increased expression of bone sialoprotein in bone metastases compared with visceral metastases in human breast and prostate cancers. J. Bone Miner. Res. 15, 834 – 843 (2000). 720. G. van der Pluijm, H. J. M. Vloedgraven, S. Papapoulos, C. W. G. M. Lowik, W. J. Grzesik, J. M. Kerr, P. Gehron Robey, Attachment characteristics and involvement of integrins in adhesion of breast cancer cell lines to extracellular bone matrix constituents. Lab. Invest. 77(6), 665 – 675 (1997). 721. N. S. Fedarko, B. Fohr, P. G. Robey, M. F. Young, and L. W. Fisher, Factor H binding to bone sialoprotein and osteopontin enables tumor cell evasion of complement-mediated attack. J. Biol. Chem. 275, 16666 – 16672 (2000). 722. J. E. Aubin, A. K. Gupta, R. Zirngbl, and J. Rossant, Knockout mice lacking bone sialoprotein expression have bone abnormalities. J. Bone Miner. Res. 11S, S29 (1996). 723. J. F. Gorski and K. Shimizu, Isolation of new phosphorylated glycoprotein from mineralized phase of bone that exhibits limited homology to adhesive protein osteopontin. J. Biol. Chem. 263, 15938 – 15945 (1988). 724. J. P. Gorski, D. Griffin, G. Dudley, C. Stanford, R. Thomas, C. Huang, E. Lai, B. Karr, and M. Solursh, Bone acidic glycoprotein75 is a major synthetic product of osteoblastic cells and localized as 75- and/or 50-kDa forms in mineralized phases of bone and growth plate and in serum. J. Biol. Chem. 265, 14956 – 14963 (1990). 725. J. P. Gorski, Is all bone the same? Distinctive distributions and properties of non-collagenous matrix proteins in lamellar vs. woven bone imply the existence of different underlying osteogenic mechanisms. Crit. Rev. Oral. Biol. Med. 9, 201 – 223 (1998). 726. A. George, J. Gui , N. A. Jenkins, D. J. Gilbert, N. G. Copeland, and A. Veis , In situ localization and chromosomal mapping of the AG1 (Dmp1) gene. J. Histochem. Cytochem. 42, 1527 – 1531 (1994). 727. Y. Chen, B. S. Bal, and J. P. Gorski, Calcium and collagen binding properties of osteopontin, bone sialoprotein, and bone acidic glycoprotein-75 from bone. J. Biol. Chem. 267, 24871 – 24878 (1992). 728. M. Sato, W. Grasser, S. Harm, C. Fullenkamp, and J. P. Gorski, Bone acidic glycoprotein 75 inhibits resorption activity of isolated rat and chicken osteoclasts. FASEB J. 6, 2966 – 2976 (1992). 729. A. George, B. Sabsay, P. A. L. Simonian, and A. Veis, Characterization of a novel dentin matrix acidic phosphoprotein. J. Biol. Chem. 268, 12624 – 12630 (1993). 730. A. George, R. Silbertsein, and A. Veis, In situ hybridization shows Dmp1 (AG1) to be a developmentally regulated dentin-specific protein produced by mature odontoblasts. Connect. Tissue 43s 67 – 92 (1995).

179 731. R. Srinivasan, B. Chen, J. P. Gorski, and A. George, Recombinant expression and characterization of dentin matrix protein-1. Connect. Tissue Res. 40, 251 – 258 (1999). 732. D. R. Keene, L. Y. Sakai, and R. E. Burgeson, Human bone contains type III collagen, type VI collagen, and fibrillin: Type III collagen is present on specific fibers that may mediate attachment of tendons, ligaments, and periosteum to calcified bone cortex. J. Histochem. Cytochem. 39, 59 – 69 (1991). 733. S. Kitahama, M. A. Gibson, G. Hatzinikolas, S. Hay, J. L. Kuliwaba, A. Evdokiou, G. J. Atkins, and D. M. Findlay, Expression of fibrillins and other microfibril-associated proteins in human bone and osteoblast-like cells. Bone 27, 61 – 67 (2000). 734. K. Kainulainen, L. Karttunen, L. Puhakka, L. Sakai, and L. Peltonen, Mutations in the fibrillin gene responsible for dominant ectopia lentis and neonatal Marfan syndrome. Nature Genet. 6, 64 – 69 (1994). 735. F. Ramirez and L. Pereira, The fibrillins. Int. J. Biochem. Cell. Biol. 31, 255 – 259 (1999). 736. G. M. Corson, S. C. Chalberg, H. C. Dietz, N. L. Charbonneau, and L. Y. Sakai, Fibrillin binds calcium and is coded by cDNAS that reveal a multidomain structure and alternatively spliced exons at the 5 end. Genomics 17, 476 – 484 (1993). 737. D. P. Reinhardt, G. M. Corson, and L. Y. Sakai, Expression and characterization of recombinant fibrillin peptides. In “FECTS Lyon, p. F17, 1994. 738. C. M. Kielty and C. A. Shuttleworth, The role of calcium in the organization of fibrillin microfibrils. FEBS Lett. 336, 323 – 326 (1993). 739. M. Raghunath, C. M. Kielty, K. Kainulainen, A. Child, L. Peltonen, and B. Steinmann, Analyses of truncated fibrillin caused by a 366 bp deletion in the FBN1 gene resulting in Marfan syndrome. Biochem. J. 302 889 – 896 (1994). 740. L. Peltonen and K. Kainulainen, Elucidation of the gene defect in Marfan syndrome: Success by two complementary research strategies. FEBS Lett. 307, 116 – 121 (1992). 741. B. C. Gallagher, L. Y. Sakai, and C. D. Little, Fibrillin delineates the primary axis of the early avian embryo. Dev. Dyn. 196, 70 – 78 (1993). 742. S. L. Dallas, D. R. Keene, S. P. Bruder, J. Saharinen, L. Y. Sakai, G. R. Mundy, and L. F. Bonewald, Role of the latent transforming growth factor beta binding protein 1 in fibrillin-containing microfibrils in bone cells in vitro and in vivo. J. Bone Miner. Res. 15, 68 – 81 (2000). 743. E. J. Mackie, I. Thesleff, and R. Chiquet-Ehrismann, Tenascin is associated with chondrogenic and osteogenic differentiation in vivo and promotes chondrogenesis in vitro. J. Cell Biol. 105, 2569 – 2579 (1987). 744. M. Ekblom, R. Fassler, B. Tomasini-Johansson, K. Nilsson, and P. Ekblom, Downregulation of tenascin expression by glucocorticoids in bone marrow stromal cells and in fibroblasts. J. Cell Biol. 123, 1037 – 1045 (1993). 745. Y. Soini, D. Kamel, M. Apaja-Sakkinen, I. Virtanen, and V. P. Lehto, Tenascin immunoreactivity in normal and pathological bone marrow. J. Clin. Pathol. 46, 218 – 221 (1993). 746. X. Chevalier, N. Groult, B. Larget-Piet, L. Zardi, and W. Hornebeck, Tenascin distribution in articular cartilage from normal subjects and from patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum. 7, 1013 – 1022 (1994). 747. J. F. Riou, M. Umbhauer, D. L. Shi, and J. C. Boucaut, Tenascin: A potential modulator of cell-extracellular matrix interactions during vertebrate embryogenesis. Biol. Cell 75, 1 – 9 (1992). 748. E. J. Mackie and R. P. Tucker, Tenascin in bone morphogenesis: Expression by osteoblasts and cell type-specific expression of splice variants. J. Cell Sci. 3, 765 – 771 (1992). 749. L. Vakeva, E. Mackie, T. Kantomaa, and I. Thesleff, Comparison of the distribution patterns of tenascin and alkaline phosphatase in

180

750.

751.

752.

753.

754.

755.

756.

757.

758.

759.

760.

761.

762.

763.

764.

765.

766.

GOKHALE, BOSKEY, AND ROBEY developing teeth, cartilage, and bone of rats and mice. Anat. Rec. 228, 69 – 76 (1990). T. Larsson, Y. Sommarin, M. Paulsson, and P. Antonsson, Cartilage matrix proteins: A basic 36-kDa protein with a restricted distribution to cartilage and bone. J. Biol. Chem. 266, 20428 – 20433 (1991). P. J. Neame, Y. Sommarin, R. E. Boynton, and D. Heinegard, The structure of a 38-kDa leucine-rich protein (chondroadherin) isolated from bovine cartilage. J. Biol. Chem. 269(34), 21547 – 21554 (1994). M. Wendel, Y. Sommarin, and D. Heinegard, Bone matrix proteins: Isolation and characterization of a novel cell-binding keratan sulfate proteoglycan (osteoadherin) from bovine bone. J. Cell Biol. 141, 839 – 847 (1998). R. Buchaille, M. L. Couble, H. Macgloire, and F. Bleicher, Expression of the small leucine-rich proteoglycan osteoadherin/osteomodulin in human dental pulp and developing rat teeth. Bone 27, 265 – 276 (2000). K. Horiuchi, N. Amizuka, S. Takeshita, H. Takamatsu, M. Katsuura, H. Ozawa, Y. Toyama, L. F. Bonewald, and A. Kudo, A Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J. Bone Miner. Res. 14, 1239 – 1249 (1999). P. A. Price, A. S. Otsuka, J. W. Poser et al., Characterization of gamma-carboxyglutamic acid containing protein from bone. Proc. Natl. Acad. Sci. USA 73, 1447 (1976). P. V. Hauschka, J. B. Lian, and P. M. Gallop, Direct identification of the calcium-binding amino acid, gamma-carboxyglutamic acid in mineralized tissue. Proc. Natl. Acad. Sci. USA 72(10), 3925 – 3929. P. A. Price, A. S. Otsuka, J. W. Poser, J. Kristaponis, and N. Raman, Characterization of gamma-carboxyglutamic-acid containing protein from bone. Proc. Natl. Acad. Sci. USA 73, 1447 – 1451 (1976). J. D. Fraser and P. A. Price, Lung, heart and kidney express high levels of mRNA for the vitamin K-dependent matrix Gla protein; implications for the possible functions of matrix gla protein and for the tisssue distribution of the -carboxylase, J. Biol. Chem. 263, 11033 – 11036 (1988). P. A. Price, Gla-containing proteins of mineralized tissues. In “Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 169 – 176. Elsevier, Amsterdam, 1992. A. L. J. J. Bronckers, S. Gay, M. T. DiMuzio, and W. T. Butler, Immunolocalization of carboxyglutamic acid containing proteins in developing rat bones. Coll. Relat. Res. 5, 273 – 281 (1985). A. L. J. J. Bronckers, S. Gay, R. D. Finkelman, and W. T. Butler, Developmental appearance of Gla proteins (osteocalcin) and alkaline phosphatase in tooth germs and bone of the rat. Bone Miner. 2, 361 – 373 (1987). A. J. Camarda, W. T. Butler, R. D. Finkelman, and A. Nanci, Immunocytochemical localization of carboxyglutamic acid-containing proteins and osteocalcin in rat bone and dentin. Calcif. Tissue Int. 40, 349 – 355 (1987). J. C. Fleet and J. M. Hock, Identification of osteocalcin mRNA in nonosteoid tissue of rats and humans by reverse transcription-polymerase chain reaction. J. Bone Miner. Res. 9, 1565 – 1573 (1994). P. V. Hauschka and P. V. Reid, Timed appearance of a calciumbinding protein containing carboxyyglutamic acid in developing chick bone. Dev. Biol. 65, 426. Y. Otawara and P. A. Price, Developmental appearance of matrix Gla protein during calcification in the rat. J. Biol. Chem. 261, 10828 (1986). P. A. Price, J. W. Lothringer, and S. K. Nishimoto, Absence of the vitamin K-dependent bone protein in fetal rat mineral: Evidence for another carboxyglutamic acid-containing component in bone. J. Biol. Chem. 255, 2938 – 2942 (1980).

767. P. A. Price, J. W. Lothringer, S. A. Baukol, and A. H. Reddi, Developmental appearance of the vitamin K-dependent proteins in bone metabolism. Annu. Rev. Nutr. 8, 865 (1988). 768. R. Kasai, P. Bianco, P. Gehron Robey, and A. J. Kahn, Production and characterization of an antibody against the human bone GLA protein (BGP/osteocalcin) propeptide and its use in immunocytochemistry of bone cells. Bone Miner. 25, 167 – 182 (1994). 769. L. C. Pan and P. A. Price, The propeptide of rat bone gamma-carboxyglutamic acid protein share homology with other vitamin Kdependent protein precursors. Proc. Natl. Acad. Sci. USA 82(18), 6109 – 6113. 770. J. W. Poser, F. S. Esch, N. C. Ling et al., Isolation and sequence of the vitamin K-dependent protein from human bone: Undercarboxylation of the first glutamic acid residue. J. Biol. Chem. 255, 8685 (1980). 771. P. V. Hauschka and S. A. Carr, Calcium-dependent alpha-helical structure in osteocalcin. Biochemistry 21, 2538 – 2547 (1982). 772. P. V. Hauschka and F. H. Wians, Jr., Osteocalcin-hydroxyapatite interactions in the extracellular organic matrix of bone. Anat. Rec. 224, 180 – 188 (1988). 773. T. L. Dowd, J. F. Rosen, C. M. Gundberg, and R. K. Gupta, The displacement of calcium from osteocalcin at submicrolar concentration of free lead. Biochim. Biophys. Acta 1226, 131 – 137 (1994). 774. D. T. Isbell, S. Du, A. G. Schroering, G. Colombo, and J. G. Shelling, Metal ion binding to dog osteocalcin studied by 1H NMR spectroscopy. Biochemistry 32, 11352 – 11362 (1993). 775. E. Puchacz, J. G. Lian, G. S. Stein, J. Wozney, K. Heubner, and C. Croce, Chromosomal localization of the human osteocalcin gene. Endocrinology 124, 2648 (1989). 776. A. J. Celeste, V. Rosen, and J. L. Buecker, Isolation of the human gene for bone Gla protein. EMBO J. 5, 1885 – 1889 (1986). 777. L. C. Pan and P. A. Price, The propeptide of rat bone gammacarboxyglutamic acid protein shares homology with other vitamin K-dependent protein precursors. Proc. Natl. Acad. Sci. USA 82, 6109 – 6113 (1985). 778. L. C. Pan, M. K. Williamson, and P. A. Price, Sequence of the precursor to rat bone gamma-carboxyglutamic acid protein that assimulates in warfarin-treated osteosarcoma cells. J. Biol. Chem. 260, 13398 – 13401 (1985). 779. S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basalactivation and inducible response to hormonal vitamin D3. Proc. Natl. Acad. Sci. USA 84, 4455 – 4459 (1989). 780. J. P. Bidwell, A. J. van Wijnen, E. G. Fey, S. Dworetzky, Penman, J. L. Stein, J. B. Lian, and G. S. Stein, Osteocalcin gene promoter binding factors are tissue-specific nuclear matrix components. Proc. Natl. Acad. Sci. USA 90, 3162 – 3166 (1993). 781. M. D. Demay, D. A. Roth, and H. M. Kronenberg, Regions of the rat osteocalcin gene which mediate the effect of 1,25-dihydroxyvitamin D3 on gene transcription. J. Biol. Chem. 264, 2279 – 2282 (1989). 782. M. D. Demay, J. M. Gerardi, H. F. DeLuca, and H. M. Kronenberg, DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsive to 1,25-dihydroxy vitamin D3. Proc. Natl. Acad. Sci. USA 87, 369 – 373 (1990). 783. K. Yoon, S. J. C. Rutledge, R. F. Buenaga, and G. A. Rodan, Characterization of the rat osteocalcin gene: Stimulation of promoter activity by 1,25-dihydroxy vitamn D3. Biochemistry 27, 8521 – 8526 (1988). 784. A. A. J. Heinrichs, R. Bortell, S. Rahman, J. L. Stein, E. S. Alnemri, G. Litwack, J. B. Lian, and SG. S. Stein, Identification of multiple glucocorticoid receptor binding sites in the rat osteocalcin gene promoter. Biochemistry 32, 11436 – 11444 (1993). 785. R. Schule, K. Umesono, D. J. Magelsdorf, J. Bolado, J. W. Pike, and R. M. Evans, Jun-Fos and receptors for vitamins A and D

CHAPTER 4 The Biochemistry of Bone

786.

787.

788.

789.

790.

791.

792.

793.

794.

795.

796.

797.

798.

799.

800.

recognize a common response element in the human osteocalcin gene. Cell 61, 497 – 504 (1990). T. A. Owen, M. A. Aronow, L. M. Baron, B. Bettencourt, G. S. Stein, and J. B. Lian, Pleiotropic effecs of vitamin D on osteoblast gene expresion are related to the prolifrative and differentiatied state of the bone cell phenotype: Dependency upon basal levels of gene expression, duration of exposure and bone matrix competency in noral rat osteoblast cultures. Endocrinology 128, 1496 – 1504 (1991). J. Lian, C. Stewart, E. Puchacz, S. Mackowiak, V. Shalhoub, D. Collart, G. Zambetti, and G. S. Stein, Structure of the rat osteocalcin gene and regulation of vitamin D-dependent expression. Proc. Natl. Acad. Sci. USA 86, 1143 – 1147 (1989). E. R. Markose, J. L. Stein, G. S. Stein, and J. B. Lian, Vitamin Dmediated modifications in protein-DNA interactions a two promoter elements of the osteocalcin gen. Proc. Natl. Acad. Sci. USA 87, 1701 – 1705 (1990). A. A. J. Heinrichs, C. Banerjee, R. Bortell, T. A. Owen, J. L. Stein, G. S. Stein, and J. B. Lian, Identification and characterization of two proximal elements in the rat osteocacin gene promoter that may confer species-specific regulation. J. Cell Biochem. 53, 240 – 250 (1993). A. J. J. Heinrichs, R. Bortell, M. Bourke, J. B. Lian, G. S. Stein, and J. L. Stein, Proximal promoter binding protein contributes to developmental, tissue-restricted expression of the rat osteocalcin gene. J. Cell Biochem. 57, 90 – 100 (1995). H. M. Hoffmann, K. M. Catron, A. J. van Wijnen, L. R. McCabe, J. B. Lian, G. S. Stein, and J. L. Stein, Transcriptional control of the tissue-speciic, developmentally regulated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc. Natl. Acad. Sci. USA 91, 12887 – 12891 (1994). R. Frenkel, J. Mijnes, M. A. Aronow, G. Zambetti, C. Banerjee, J. L. Stein, J. B. Lian, and G. S. Stein, Position and orientation-selective silencer in protein-coding sequences of the rat osteocalcin gene. Biochemistry 32, 13636 – 13643 (1993). B. Frenkel, M. Menetcino, J. L. Stein, J. B. Lian, and G. S. Stein, A composite intragenic silencer domain exhibits negative and positive transcriptional control of the bone-specific osteocalcin gene: Promoter and cell type requirements. Proc. Natl. Acad. Sci. USA 91, 10923 – 10927 (1994). P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). J. Fraser, Y. Otawara, and P.A. Price, 1,25-dihydroxyvitaine D3 stimulates the synthesis of matrix -carboxyglutaic acid protein by osteosarcoma cells: Mutually exclusive expression of vitamin K-dependent bone proteins by clonal osteoblastic cell lines. J. Biol. Chem. 263, 911 – 916 (1988). C. M. Terpening, C. A. Haussler, P. W. Jurutka, M. A. Galliggan, B. S. Komm, and M. R. Haussler, The vitamin D-responsive element in the rat bone gla protein gene is an imperfect direct repeat that cooperates with other cis-elements in 1,25-dihydroxyvitamin D3-mediated transcriptional activation. Mol. Endocrinol. 5, 373 – 385 (1991). P. Maenpaa, A. Mahonen, and A. Pirskanen, Hormonal regulation of vitamin D receptor levels and osteocalcin synthesis in human osteosarcoma cells. Calcif. Tissue Int. 49, S85 – S86 (1991). N. Pernalete, T. Mori, Y. Nishii, E. Slatopolsky, and A. J. Brown, The activity of 22-oxacalcitriol in osteoblast-like (ROS 17/2.8) cells. Endocrinology 129, 778 – 784 (1991). D. Lajeunesse, G. M. Kiebzak, C. Frondoza, and B. Sacktor, Regulation of osteocalcin secretion by human primary bone cells and by the human osteosarcoma cell line MG-63. Bone Miner. 14, 237 – 250 (1991). A. M. Delany, Y. Dong, and E. Canalis, Mechanisms of glucocorticoid action in bone cells. J. Cell Biochem. 56, 295 – 302 (1994).

181 801. G. Schepmoes, E. Breen, T. A. Owen, M. A. Aronow, G. S. Stein, and J. B. Lian, Infuence of dexamethasone on the vitamin D mediated regulation of the osteocalcin gene expression. J. Cell Biochem. 47, 184 – 196 (1991). 802. A. Pirskanen, T. Jaaskelainen, and P. H. Maenpaa, Effects of transforming growth factor beta 1 on the regulation of osteocalcin synthesis in human MG-63 osteosarcoma cells. J. Bone Miner. Res. 9, 1635 – 1642 (1994). 803. R. T. Ingram, S. K. Bonde, B. L. Riggs, and L. A. Fitzpatrick, Effects of transforming growth factor beta (TGF beta) and 1,25 dihydroxyvitamin D3 on the function, cytochemistry and morphology of normal human osteoblast-like cells. Differentiation 55, 153 – 163 (1994). 804. R. S. Taichman and P. V. Hauschka, Effects of interleukin-1 beta and tumor necrosis factor-alpha on osteoblastic expression of osteocalcin and mineralized extracellular matrix in vitro. Inflammation 16, 587 – 601 (1992). 805. D. B. Evans, R. A. Bunning, and R. G. Russell, The effects of recombinant human interleukin-1 beta on cellular proliferation and the production of prostaglandin E2, plasminogen activator, osteocalcin and alkaline phosphatase by osteoblast-like cells derived from human bone. Biochem. Biophys. Res. Commun. 166, 208 – 216 (1990). 806. D. M. Raab-Cullen, M. A. Thiede, D. N. Petersen, D. B. Kimmel, and R. R. Recker, Mechanical loading stimulates rapid changes in periosteal gene expression. Calcif. Tissue Int. 55, 473 – 478 (1994). 807. J. B. Lian and C. M. Gundberg, Osteocalcin-biochemical considerations and clinical applications. Clin. Orthop. 226, 267 – 291 (1988). 808. P. V. Hauschka, J. B. Lian, D. E. C. Cole, and C. Gundberg, Osteocalin and matrix gla protein: Vitamin K-dependent proteins in bone. Physiol. Rev. 69, 990 – 1047 (1989). 809. L. C. Gerstenfeld, S. D. Chipman, J. Glowacki, and J. B. Lian, Expression to differentiate function by mineralizing cultures of chicken osteoblasts. Dev. Biol. 122, 49 – 60 (1986). 810. A. L. Boskey, F. H. Wians., Jr., and P. V. Hauschka, The effect of osteocalcin on in vitro lipid-induced hydroxyapatite formation and seeded hydroxyapatite growth. Calcif. Tissue Int. 37, 75 (1985). 811. C. G. Groot, J. K. Danes, J. Blok, A. Hoogendijk, and P. V. Hauschka, Light and electron microscopic demonstration of osteocalcin antigenicity in embryonic and adult rat bone. Bone 7, 379 – 385 (1986). 812. J. N. M. Heersche, S. Reimers, J. L. Wrana, M. M. Y. Waye, and A. K. Gupta, Changes in expression of alpha I collagen and osteocalcin mRNA in osteoblasts and odontoblasts at different stages of maturity as shown by in situ hybridization. Proc. Finn. Dent. Soc. 88, 173 – 182 (1992). 813. G. A. Rodan and M. Noda, Gene expression in osteoblastic cells. Crit. Rev. Eukaryot. Gene Expr. 1(2), 85 – 98 (1991). 814. M. A. Aronow, L. C. Gerstenfeld, T. A. Owen, M. Tassinari, G. S. Stein, and J. B. Lian, Factors that promote progressive development of the osteoclast phenotype in cultured rat calvarial cells. J. Cell Physiol. 143, 213 – 221 (1990). 815. W. H. Liggett, J. B. Lian, J. S. Greenberger, and J. Glowacki, Osteocalcin promotes differentiation of osteoclast progenitors from murine long-term bone marrow cultures. J. Cell. Biochem. 55, 190 – 199 (1994). 816. D. J. DeFranco, J. Glowaki, K. A. Cox, and J. B. Lian, Normal bone particles are preferentially resorbed in the presence of osteocalcindeficient bone particles in vitro. Calcif. Tissue Int. 49, 43 (1991). 817. J. D. Malone, S. L. Teitelbaum, G. L. Griffin, R. M. Senior, and A. J. Kahn, Recruitment of osteoclast precursors by purified bone matrix constituents. J. Cell Biol. 92, 227 – 230 (1982). 818. J. Glowacki and J. B. Lian, Impaired recruitment and differentiation of osteoclast progenitors by osteocalcin-depleted bone implants. Cell Differ. 21, 247 (1987).

182 819. J. B. Lian and S. C. Marks, Jr., Osteopetrosis in the rat: Coexistence of reductions in osteocalcin and bone resorption. Endocrinology 126, 955 – 962 (1990). 820. V. Shalhoub, B. Betterncourt, M. E. Jackson, C. A. Mackay, J. Lian, and G. S. Stein, Abnormalities of phosphoprotein gene expression in three osteopetrotic rat mutations: Elevated mRNA transcripts, protein synthesis, and accumulation in bone of mutant animals. J. Cell Physiol. 158, 110 – 120 (1994). 821. J. Bollerslev, S. C. Marks, Jr, L. Mosekilde, and J. B. Lian, Cortical bone osteocalcin content and matrix composition in autosomal dominant osteopetrosis type I. Eur. J. Endocrinol. 130, 592 – 594 (1994). 822. P. A. Price, M. K. Williamson, T. Haba, R. B. Dell, and W. S. S. Jee, Excessive mineralization with growth plate closure in rats on chronic warfarin treatment. Proc. Natl. Acad. Sci. USA 79, 7734 – 7738 (1982). 823. P. Pastoureau, P. Vergnaud, P. J. Meunier, and P. D. Delmas, Osteopenia and bone-remodeling abnormalitities in warfarin-treated lambs. J. Bone Miner. Res. 8, 1417 – 1426 (1992). 824. A. L. Boskey, S. Gadaleta, C. Gundberg, S. B. Doty, P. Ducy, and G. Karsenty, Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 23, 187 – 196 (1998). 825. P. A. Price, M. R. Urist, and Y. Otawara, Matrix gla protein, a new g-carboxyglutamic acid-containing protein which is associated with the organic matrix of bone. Biochem. Biophys. Res. Commun. 117, 765 – 771 (1983). 826. P. A. Price and M. K. Williamson, Primary structure of bovine matrix gla protein, a new vitamin K-dependent bone protein. J. Biol. Chem. 260, 14791 – 14975 (1985). 827. J. E. Hale, J. D. Fraser, and P. A. Price, The identification of matrix gla protein in cartilage. J. Biol. Chem. 263, 5820 – 5824 (1988). 828. G. Luo, P. Ducy, M. D. McKee, G. J. Pinero, E., Loyer, R. R. Behringer, and G. Karsenty, Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 386, 78 – 81 (1997). 829. J. S. Rice, M. K. Williamson, and P. A. Price. Isolation and sequence of the vitamin K-dependent matrix Gla protein from the calcified cartilage of the soupfin shark. J. Bone Miner. Res. 9, 567 – 576 (1994). 830. G. Luo, R. D’Souza, D. Hogue, and G. Karsenty, The matrix glaprotein gene is a marker of the chondrogenesis cell lineage during mouse development. J. Bone Miner. Res. 10, 325 – 334 (1995). 831. J. E. Hale, M. K. Williamson, and P. A. Price, Carboxyl-terminal proteolytic processing of matrix gla protein. J. Biol. Chem. 266, 21145 – 21149 (1991). 832. P. A. Price, J. S. Rice, and M. K. Williamson, Conserved phosphorylation of serines in the Ser-x-glu/Ser (P) sequences of the vitamin K-dependent matrix gla protein from shark, lamb, rat, cow, and human. Protein Sci. 5, 822 – 830 (1994). 833. M. T. B. Wiedemann and D. Belluoccio, Molecular cloning of avian matrix Gla protein. Biochim. Biophys. Acta 1395, 47 – 49 (1998). 834. M. L. Cancela, C.-L. Hsieh, U. Francke, and P. A. Price, Molecular structure, chromosome assignment and promoter organization of the human matrix Gla protein gene. J. Biol. Chem. 265, 15040 – 15048 (1990). 835. M. L. Cancela and P. A. Price, Retinoic acid induces matrix gla protein gene expression in human cells. Endocrinology 130, 102–108 (1992). 836. S. R. Rannels, M. L. Cancela, E. B. Wolpert, and P. A. Price, Matrix gla protein mRNA expression in cultured type II pneuomocytes. Am. J. Physiol. 265, L270 – L278 (1993). 837. A. L. Boskey, M. D. McKee, and G. P. Karsenty, Mineral characterization of bones and soft tissues in matrix gla protein-deficient mice. In “Chemistry and Biology of Mineralized Tissues” (M. Goldberg, A. Boskey, and C. Robinson, eds.), pp. 63 – 67 (2000).

GOKHALE, BOSKEY, AND ROBEY 838. K. Yagami, J. Y. Suh, M. Enomoto-Iwamoto, E. Koyama, W. R. Abrams, I. M. Shapiro, M. Pacifici, and M. Iwamoto, Matrix GLA protein is a developmental regulator of chondrocyte mineralization and, when constitutively expressed, blocks endochondral and intramembranous ossification in the limb. J. Cell. Biol. 147, 1097 – 1108 (1999). 839. L. Chen, J. P. O’Bryan, H. S. Smith, and E. Liu, Overexpression of matrix Gla protein mRNA in malignant human breast cells: Isolation by differential cDNA hybridization. Oncogene 5, 1391 – 1395 (1990). 840. M. M. Briehl and R. L. Misefeld, Isolation and characterization of transcripts induced by androgen withdrawal and apoptotic cell death in the rat ventral prostate. Mol. Endocrinol. 5, 1381 – 1388 (1991). 841. D. M. Lawton, J. G. Andrew, D. R. Marsh, J. A. Hoyland, and A. J. Freemont, Expression of the gene encoding the matrix gla protein by mature osteoblasts in human fracture non-unions. Mol. Pathol. 52, 92 – 96 (1999). 842. C. Maillard, M. Berruyer, C. M. Serre, M. Dechavane, and P. D. Delmas, Protein S, a vitamin K-dependent protein, is a bone matrix component synthesized and secreted by osteoblasts. Endocrinology 130, 1599 – 1604 (1992). 843. J. Folch-Pi and P. Stoffyn, Proteolipids from membrane systems. Ann. N. Y. Acad. Sci. 195, 86 – 107 (1972). 844. J. Folch-Pi and J. Sakura, Preparation of the proteolipid apo-protein from bovine heart, liver and kidney. Biochim. Biophys. Acta 427, 401 – 427 (1977). 845. J. Ennever, B. Boyan-Salyers, and L. J. Riggan, Proteolipid and bone matrix calcification in vitro. J. Dent. Res. 56, 967 – 970 (1977). 846. J. Ennever, J. J. Vogel, and L. J. Riggan, Phospholipids of a bone matrix calcification nucleator. J. Dent. Res. 57, 731 – 734 (1978). 847. A. L. Boskey, Phospholipids and calcification: An overview. In “Cell-Mediated Calcification and Matrix Vesicles” (S. Yousuf Ali, ed.), pp. 175 – 179. Elsevier, Amsterdam, 1986. 848. B. D. Boyan, Proteolipid-dependent calcification. In “The Chemistry and Biology of Mineralized Tissues” (W. F. Butler, ed.), pp. 125 – 131. EBSCO Media, Birmingham, 1985. 849. B. D. Boyan and N. M. Ritter, Proteolipid-lipid relationships in normal and vitamin D-deficient chick cartilage. Calcif. Tissue Int. 36, 332 – 337 (1984). 850. B. D. Boyan, Z. Schwartz, L. D. Swain, and A. Khare, Role of lipids in calcification of cartilage. Anat. Rec. 224, 211 – 219 (1989). 851. L. N. Wu, T. Yoshimori, B. R. Genge, G. R. Sauer, and R. E. Wuthier, Characterization of the nucleational core complex responsible for mineral induction by growth plate cartilage matrix vesicles. J. Biol. Chem. 268, 25084 – 25094 (1993). 852. R. E. Wuthier, Mechanism of de novo mineral formation by matrix vesicles. Connect. Tissue Res. 22, 27 – 32 (1989). 853. G. J. Barton, R. H. Newman, P. S. Freemont, and M. J. Crumpton, Amino acid sequence analysis of the annexin super-gene family. Eur. J. Biochem. 198, 749 – 760 (1991). 854. J. Johansson, T. Szyperski, T. Curstedt, and K. Wuthrich, The NMR structure of the pulmonary surfactant-associated polypeptide SP-C in an apolar solvent contains a valyl-rich alpha-helix. Biochemistry 33, 6015 – 6023 (1994). 855. S. van Dijk, D. D. Dean, Y. Liu, Y. Zhao, J. M. Chirgwin, Z. Schwartz, and B. D. Boyan, Purification, amino acid sequence, and cDNA sequence of a novel calcium-precipitating proteolipid involved in calcification of corynebacterium matruchotii. Calc. Tissue Int. 62, 350 – 358 (1998). 856. K. Jung, H. Jung, and H. R. Kaback, Dynamics of lactose permease of Escherichia coli determined by site-directed fluorescence labeling. Biochemistry 33, 3980 – 3985 (1994).

CHAPTER 4 The Biochemistry of Bone 857. J. Stim, A. A. Bernardo, F. T. Kear, Y. Y. Qiu, and J. A. Arruda, Renal cortical basolateral Na/HCO3  co-transporter. II. Detection of conformational changes with fluorescein isothiocyanate labeling. J. Memb. Biol. 140, 39 – 46 (1994). 858. M. B. Beest, K. Hoekstra, A. Sein, and D. Hoekstra, Reconstruction of proteolipid protein: Some properties and its role in interlamellar attachment. Biochem. J. 300, 545 – 552 (1994). 859. G. Fernandez-Ballester, J. Castesana, A. M. Fernandez, and J. L. Arrondo, A role for cholesterol as a structural effector of the nicotinic acetylcholine receptor. Biochemistry 33, 4065 – 4071 (1994). 860. B. D. Boyan, W. J. Landis, J. Knight, G. Dereszewski, and J. Zeagler, Microbial hydroxyapatite formation as a model of proteolipiddependent membrane-mediated calcification. Scan. Electr. Microsc. 4, 1793 – 1800 (1984). 861. L. D. Swain, R. D. Renthal, and B. D. Boyan, Resolution of ion translocating proteolipid subclasses active in bacterial calcification. J. Dent. Res. 68, 1094 – 1097 (1989). 862. J. D. Swain and B. D. Boyan, Ion-translocating properties of calcifiable proteolipids. J. Dent. Res. 67, 526 – 530 (1988). 863. C. Montessuit, J. Caverzassio, and J. P. Bonjour, Mechanism of Pi accumulation in maxrip vesecles and its potential role in the calcification process. Bone Miner. 17, 386 (1992). 864. A. L. Boskey, Hydroxyapatite formation in a dynamic gelatin gel system: Effects of type I collagen, lipids, and proteoglycans. J. Phys. Chem. 93, 1628 – 1633 (1989). 865. C. L. Raggio, B. D. Boyan, and A. L. Boskey, In vivo hydroxyapatite formation induced by lipids. J. Bone Miner. Res. 1, 409 – 415 (1986). 866. B. D. Boyan-Salyers and A. L. Boskey, Relationship between protolipids and calcium-phospholipid-phosphate complexes in Bacterionema matruchotii calcification. Calcif. Tissue Int. 30, 167 – 174 (1980). 867. A. L. Boskey and B. D. Boyan-Salyers, Relationship between proteolipids and calcium-phospholipid-phosphate complexes in bacterionema matruchotti calcification. Calcif. Tissue Int. 30, 167 – 174 (1978). 868. A. L. Boskey and A. S. Posner, The role of synthetic and bone extracted calcium-phospholipid-phosphate complexes in hydroxyapatite formation. Calcif. Tissue Res. 23, 251 – 258 (1977). 869. B. D. Boyan and A. L. Boskey, Co-isolation of proteoplipids and calcium-phospholipid-phosphate complexes. Calcif. Tissue Int. 36, 214 – 218 (1984). 870. A. L. Boskey, The role of Ca-phospholipid-PO4 complexes, membrane lipids and matrix vesicles in calcification. In “Matrix Vesicles” (A. Ascenzi, E. Bonnucci, and B. de Bernard, eds.), pp. 161 – 167. Wichtig Editore, Milan Italy, 1981. 871. A. L. Boskey, Phospholipids and biological calcification. In “Proc. First Int. Conf. Chem. Biol. Mineralized Connective Tissues” (A. Veis, ed.), pp. 531 – 537. Elsevier/North Holland, New York, 1981. 872. R. E. Wuthier and S. Gore, Partition of inorganic ion and phospholipids in isolated cell, membrane and matrix vesicle fractions: Evidence for Ca:Pi acidic phospholipid complexes. Calcif. Tissue Res. 24, 163 – 171 (1977). 873. A. Yaari, I. Shapiro, and A. L. Boskey, Ionophoretic activity of a Ca-PL-PO4 complex. Calcif. Tissues Int. 36, 317 – 319 (1984). 874. B. R. Genge, L. N. Wu, H. D. Adkinsson, IV, and R. E. Wuthier, Matrix vesicle annexins exhibit proteolipid-like properties: Seclective partitioning into lipophilic solvents under acidic conditions. J. Biol. Chem. 226, 10678 – 10685 (1991). 875. R. B. Pepinsky, R. Tizard, R. J. Mattaliano, K. L. Sinclair, G. T. Miller, J. L. Browning, E. P. Chow, C. Burne, K. S. Huang, D. Pratt, L. Wachter, C. Hession, A. Z. Frey, and B. P. Wallner, Five distinct calcium and phospholipid binding proteins share homology with lipocortin I. J. Biol. Chem. 263, 10799 – 10811 (1988). 876. M. J. Crumpton and J. R. Dedman, Protein terminology tangle. Nature 345, 212 (1990).

183 877. E. Rojas, H. B. Pollard, H. T. Haigler, C. Parra, and A. L. Burns, Calcium-activated endonexin II forms calcium channels across acidic phospholipids bilayer membranes. J. Biol. Chem. 265, 21207 – 21215 (1990). 878. G. Weber and E. Ferber, Selective inhibition of phospholipase A2 by different lipocortins. Biol. Chem. Hoppe Seyler 371, 725 – 731 (1990). 879. M. J. Geisow, U. Fritsche, J. M. Hexham, B. Dash, and T. Johnson, A consensus amino-acid sequence repeat in Torpedo and mammalian Ca-dependent membrane-binding proteins. Nature 320, 636 – 638 (1986). 880. F. Suarez, B. Rothhut, C. Comera, L. Touqui et al., Expression of annexin I, II, V and VI by rat osteoblasts in primary culture: Stimulation of annexin I expression by dexamethasone. J. Bone Miner. Res. 8, 1201 – 1210 (1993). 881. B. R. Genge, L. N. Wu, and R. E. Wuthier, Differential fractionation of matrix vesicle proteins: Further characterization of the acidic phospholipid-dependent Ca-binding proteins. J. Biol. Chem. 265, 10678 – 10685 (1990). 882. R. E. Wuthier, Mechanism of de novo mineral formation by matrix vesicles. Connect. Tissue Res. 22, 27 – 33 (1989). 883. L. W. Fisher and J. D. Termine, Purification of the noncollagenous proteins from bone: Technical pitfalls and how to avoid them. In “Current Advances in Skeletogenesis” (A. Ornoy, A. Harell, and J. Sela, eds.), pp. 467 – 472. Elsevier, Amsterdam, 1985. 884. I. R. Dickson and M. K. Bagga, Changes with age in the noncollagenous proteins of human bone. Connect. Tissue Res. 14, 77 – 85 (1985). 885. H. Nagase and J. F. J. Woessner, Matrix metalloproteinases. J. Biol. Chem. 274, 21491 – 21494 (1999). 886. S. Gack, R. Vallon, J. Schmidt, A. Grigoriadis, J. Tuckermann, J. Schenkel, H. Weiher, E. F. Wagner, and P. Angel, Expression of interstitial collagenase during skeletal development of the mouse is restricted to osteoblast-like cells and hypertrophic chondrocytes. Cell Growth Differ. 6, 759 – 767 (1995). 887. N. Johansson, U. Saarialho-Kere, K. Airola, R. Herva, L. Nissinen, J. Westermarck, E. Vuorio, J. Heino, and V. M. Kahari, Collagenase-3 (MMP-13) is expressed by hypertrophic chondrocytes, periosteal cells, and osteoblasts during human fetal bone development. Dev. Dyn. 208, 387 – 397 (1997). 888. M. C. Meikle, S. Bord, R. M. Hembry, and J. J. Reynolds, The synthesis of collagenase, gelatinase-A (72 kDa) and -B (95 kDa), and TIMP-1 and -2 by human osteoblasts from normal and arthritic bone. Bone 17, 255 – 260 (1995). 889. J. K. Heath, S. J. Atkinson, M. C. Meikle et al., Mouse osteoblasts synthesize collagenase in response to bone resorbing agents. Biochim. Biophys. Acta 802, 151 (1984). 890. K. Otsuka, J. Sodek, and H. Limeback, Synthesis of collagenase and collagenase inhibitors by osteoblast-like cells in culture. Eur. J. Biochem. 145, 123 (198). 891. D. K. Scott, K. D. Brinckerhoff, J. C. Clohisy, C. O. Quinn, and N. C. Partridge, Parathyroid hormone induces transcription of collagenase in rat osteoblastic cells by a mechanism using cyclic adenosine 3,5-monophosphate and requiring protein synthesis. Mol. Endocrinol. 6, 2153 – 2159 (1992). 892. J. A. Lorenzo, C. C. Pilbeam, J. F. Kalinowski, and M. S. Hibbs, Production of both 92- and 72-kDa gelatinases by bone cells. Matrix 12, 282 – 290 (1992). 893. T. H. Omura, A. Noguchi, C. A. Johanns, J. J. Jeffrey, and N. C. Partridge, Identification of a specific receptor for interstitial collagenase on osteoblastic cells. J. Biol. Chem. 269, 24994 – 24998 (1994). 894. N. Katsura and K. Yamada, Isolation and characterization of a metalloprotease associated with chicken epiphyseal cartilage matrix vesicles. Bone 7, 137 – 143 (1986).

184 895. S. Bord, A. Horner, C. A. Beeton, R. M. Hembry, and J. E. Compston, Tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) distribution in normal and pathological human bone. Bone 24, 229 – 235 (1999). 896. J. M. Dellaise, Y. Eeckout, L. Neff, C. Francois-Gillet, A. Herriet, Sutt, G. Vaes, and R. Baron, (Pro)collagenase (matrix metalloproteinases-1) is present in rodent osteoclasts and in the underlying bone-resorbing compartment. J. Cell Sci. 106, 1071 – 1082 (1993). 897. G. Leloup, J. M. Delaisse, and G. Vaes, Relationship of the plasminogen activator/plasmin cascade to osteoclast invasion and mineral resorption in explanted fetal metatarsal bones. J. Bone Miner. Res. 891 – 902 (1994). 898. K. Holmbeck, P. Bianco, J. Caterina, S. Yamada, M. Kromer, S. A. Kuznetsov, M. Mankani, P. G. Robey, A. R. Poole, I. Pidoux, J. M. Ward, and H. Birkedal-Hansen, MT1-MMP-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen turnover. Cell 99, 81 – 92 (1999). 899. M. C. Meikle, S. Bord, R. M. Hembry, and J. J. Reynolds, Rabbit calvarial osteoblasts in culture constitutively synthesize progelatinase-A, and TIMP1 and TIMP-2. Biochim. Biophys. Acta 1224, 99 – 102 (1994). 900. A. T. Vernillo, N. S. Ramamurthy, H. M. Lee, and B. R. Rifkin, Parathyroid hormone regulation of matrix degra 75, 3 ding enzymes in rat osteoblastic osteosarcoma 17/2.8 cells. Res. Commun. Chem. Pathol. Pharmacol. 323 – 339 (1992). 901. F. S. Panagakos and S. Kumar, Modulation of proteases and their inhibitors in immortal human osteoblast-like cells by tumor necrosis factor-alpha in vitro. Inflammation 18, 243 – 265 (1994). 902. S. Varghese, S. Rydziel, J. J. Jeffrey, and E. Canalis, Regulation of interstitital collagenase expression and collagen degradation by retinoic acid in bone cells. Endocrinology 134, 2438 – 2444 (1994). 903. W. Bode, F. X. Gomis-Ruth, and W. Stockler, Astacins serralysins, snake venom and matrix metalloproteinases exhibit identical zincbinding environments (HEXXHSSGXXH and Met-turn) and topologies and should be grouped into a common family, the “metzincins,” FEBS Lett. 331, 134 – 140 (1993). 904. P. Reinemer, F. Grams, R. Huber, and T. Kleine, Structural implications for the role of the N-terminus in the “Superactivatin” of collagenases. A crystallographic study. FEBS Lett. 338, 227 – 233 (1994). 905. W. Bode, P. Reinemer, R. Huber, and T. Kleine, The X-ray crystal structure of the catalytic domain of human neutrophil collagenase inhibited by a substrate analogue reveals the essentials for catalysis and specificity. EMBO J. 13, 1263 – 1269 (1994). 906. D. B. McKay, M. M. Thayer, K. M. Flaherty, H. Pley, and D. Benvegnu, Crystallograpic structures of the elastase of Pseudomonas aeruginosa. Matrix Suppl. 1, 112 – 115 (1992). 907. J. C. Spurlino, A. M. Smallwood, D. D. Carlton, T. M. Banks, K. J. Vavra, J. S. Johnson, E. R. Cook, J. Falvo, R. C. Wahl, and T. A. Pulvino, 1.56 A structure of mature truncated human fibroblast collagenase. Proteins 19(2), 98 – 109 (1994). 908. M. F. French, A. Bhown, and H. E. Van Wart, Identification of Clostridium collagenase hyperactive sites in type I, II, and III collagens: Lack of correlation with local triple helical stability. J. Protein Chem. 11, 83 – 97 (1992). 909. H. Wu, X. Liu, M. Bynne, Y. Harada, S. M. Krane, and R. Jaenisch, Collagenases are critical for skeletal and soft connective tissue remodeling : Abnormalities in mice that express, in the germline, an engineered mutation in the Col 1a1 gene that encodes collagenase resistance. J. Bone Miner Res. 9, S141(1994). 910. V. Everts, J. M. Delaisse, W. Korper, A. Niehof, G. Vaes, and W. Beertsen, Degradation of collagen in the bone-resorbing compartment underlying the osteoclast involves both cysteineproteinases and matrix metalloproteases. J. Cell. Physiol. 150, 221 – 231 (1992).

GOKHALE, BOSKEY, AND ROBEY 911. G. Vaes, J. M. Delaisse, and Y. Eeckhout, Relative roles of collagenase and lysosomal cysteine-proteinases in bone resorption. Matrix Suppl. 1, 383 – 388 (1992). 912. P. A. Hill, D. J. Buttle, S. J. Jones, A. Boyde, M. Murata, J. J. Reynolds, and M. C. Meikle, Inhibition of bone resorption by selective inactivators of cysteine proteinases. J. Cell Biochem. 45, 118–130 (1994). 913. J. P. Montenez, J. M. Delaisse, P. M. Tulkens, and B. K. Kishore, Increased activities of cathepsin B and other lysosomal hydrolases in fibroblasts and bone tissue cultured in the presence of cysteine proteinases inhibitors. Life Sci. 55, 1199 – 1208 (1994). 914. J. J. Wu, M. W. Lark, L. E. Chun, and D. R. Eyre, Sites of stromelysin cleavage in collagen types II, IX, X, and XI of cartilage. J. Biol. Chem. 266, 5625 – 5628 (1991). 915. C. Niyibizi, R. Chan, J. J. Wu, and D. Eyre, A 92 kDA gelatinase (MMP-9) cleavage site in native type V collagen. Biochem. Biophys. Res. Commun. 202, 328 – 333 (1994). 916. A. M. Bollen and D. R. Eyre, Direct extraction of gelatinases from rat bone. Connect. Tissue Res. 29, 223 – 230 (1993). 917. M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzogm, C. Debouck, and I. Kola, Cathepsin K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J. Bone Miner. Res. 14, 1654 – 1663 (1999). 918. P. Saftig, E. Hunziker, V. Everts, S. Jones, A. Boyde, O. Wehmeyer, A. Suter, and K. von Figura, Functions of cathepsin K in bone resorption: Lessons from cathepsin K deficient mice. Adv. Exp. Med. Biol. 477, 293 – 303 (2000). 919. K. Hoekman, C. W. Lîwik, M. van de Ruit, O. L. Bijvoet, J. H. Verheijen, and S. E. Papapoulos, The effect of tissue type plasminogen activator (tPA) on osteoclastic resorption in embryonic mouse long bone explants: A possible role for the growth factor domain of tPA. Bone Miner. 17, 1 – 13 (1992). 920. T. J. Martin, E. H. Allan, and S. Fukumoto, The plasminogen activator and inhibitor system in bone remodelling. Growth Regul. 3, 209 – 214 (1993). 921. J. A. Hamilton, S. R. Lingelbach, N. C. Partridge et al., Stimulation of plasminogen activator in osteoblast like cells by bone-resorbing hormones. Biochem. Biophys. Res. Commun. 122, 230 (1984). 922. B. D. Catherwood, L. Titus, C. O. Evans, J. Rubin, S. D. Boden, and M. S. Nanes, Increased expression of tissue plasminogen activator messenger ribonucleic acid is an immediate esponse to parathyroid hormone in neonatal rat osteoblasts. Endocrinology 134, 1429 – 1436 (1994). 923. S. Fukumoto, E. H. Allan, J. A. Yee, T. D. Gelehrter, and T. J. Martin, Plasminogen activator regulation in osteblasts: Parathyroid hormone inhibition of type-1 plasminogen activator inhibitor and its mRNA. J. Cell. Physiol. 152, 346 – 355 (1992). 924. J. Pfeilschifter, J. Erdmann, W. Schmidt, A. Naumann, H. W. Minne, and R. Ziegler, Differential regulation of plasminogen activator and plasminogen activator inhibitor by osteotropic factors in primary cultures of mature osteoblasts and osteoblast precursors. Endocrinology 126, 703 – 711 (1990). 925. S. Fukumoto, E. H. Allan, and T. J. Martin, Regulation of plasminogen activator inhibitor-1 (PAI-1) expression by 1,25-dihydroxyvitamin D-3 in normal and malignant rat osteoblasts. Biochim. Biophys. Acta 1201, 223 – 228 (1994). 926. S. L. Cheng, V. Shen, and W. A. Peck, Regulation of plasminogen activator and plasminogen activator inhibitor production by growth factors and cytokines in rat calvarial cells. Calcif. Tissue Int. 49, 321 – 327 (1991). 927. F. W. Fawthrop, B. O. Oyajobi, R. A. Bunning, and R. G. Russell, The effect of transforming growth factor beta on the plasminogen activator activity of normal human osteoblast-like cells and a human osteosarcoma cell line MG-63. J. Bone Miner. Res. 7, 1363 – 1371 (1992).

CHAPTER 4 The Biochemistry of Bone 928. E. H. Allan, R. Zeheb, T. D. Gelehrter, J. H. Heaton, S. Fukumoto, J. A. Yee, and T. J. Martin, Transforming growth factor beta inhibits plasminogen activator (PA) activity and stimulates production of urokinase-type PA, PA inhibitor-1 mRNA, and protein in rat osteoblast-like cells. J. Cell. Physiol. 149, 34 – 43 (1991). 929. S. Fukumoto, E. H. Allan, R. Zeheb, T. D. Gelehrter, and T. J. Martin, Glucocorticoid regulation of plasminogen activator inhibitor-1 messenger ribonucleic acid and protein in normal and malignant rat osteoblasts. Endocrinology 130, 797 – 804 (1992). 930. Y. Mikuni-Takagaki and M. J. Glimcher, Post-translational processing of chicken bone phosphoproteins: Identification of the bone phosphoproteins of embryonic tibia. Biochem. J. 268, 585 – 591 (1990b). 931. Y. Mikuni-Takagaki and M. J. Glimcher, Post-translational processing of chicken bone phosphoproteins. Identification of bone (phospho)protein kinase. Biochem. J. 268, 593 – 597 (1990a). 932. C. Sfeir and A. Veis, Casein kinase localization in the enodplasmic reticulum of the RO”S 17/2.8 cell line. J. Bone Miner. Res. 10, 607 – 615 (1995). 933. Y. Suzuki, A. Yamaguchi, T. Ikeda, T. Kawase, S. Saito, and Y. Mikuni-Takagaki, In situ phosphorylation of bone and dentin proteins by the casein kinase II-like enzyme. J. Dent. Res. 77, 1799 – 1806 (1998). 934. A. Veis, C. Sfeir, and C. B. Wu, Phosphorylation of the proteins of the extracellular matrix of mineralized tissues by casein kinase-like activity. Crit. Rev. Oral. Biol. Med. 8, 360 – 379 (1997). 935. E. H. Fischer, H. Charbonneau, and N. K. Tonks, Protein tyrosine phosphatases: A diverse family of intracellular and transmembrane enzymes. Science 253, 401 – 420 (1991). 936. E. Salih, S. Ashkar, L. C. Gerstenfeld, and M. J. Glimcher, Identification of the phosphorylated sites of metabolically 32P-labeled osteopontin from cultured chicken osteoblasts. J. Biol. Chem. 272, 13966 – 13973 (1997). 937. L. Rifas, S. Cheng, L. R. Halstead, A. Gupta, K. A. Hruska, and L. V. Avioli, Skeletal casein kinase activity defect in the HYP mouse. Calcif. Tissue Int. 61, 256 – 259 (1997). 938. C. B. Huggins, The formation of bone under the influence of epithelium of the urinary tract. Arch. Surg. 22, 377 – 408 (1931). 939. M. R. Urist, Bone: Formation by autoinduction, Science 150, 893 – 899 (1965). 940. M. R. Urist, R. J. De Lange, and A. M. Finerman, Bone cell differentiation and growth factors. Science 220, 680 – 686 (1983). 941. J. M. Wozney, Bone morphogenetic proteins and their gene expression. In “Cellular Molecular Biology of Bone,” pp. 131 – 168. Academic Press, New York, 1993. 942. A. H. Reddi and N. S. Cunningham, Initiation and promotion of bone differentiation by bone morphogenetic proteins. J. Bone Miner. Res. 8, S499 – S502 (1993). 943. S. Cheifetz, BMP receptors in limb and tooth formation. Crit. Rev. Oral. Biol. Med. 10, 182 – 198 (1999). 944. J. M. Wozney, The bone morphogenetic protein family: Multifunctional cellular regulators in the embryo and adult. Eur. J. Oral. Sci. 106, 160 – 166 (1998). 945. J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). 946. S. Daopin, M. Li, and D. R. Davies, Crystal structure of TGF-beta 2 refined at 1.8 A resolution. Proteins 17, 176 – 192 (1993). 947. S. J. Archer, A. Bax, A. B. Roberts, M. B. Sporn, Y. Ogawa, K. A. Piez, J. A. Weatherbee, M. L. Tsang, R. Lucas, and B. L. Zheng, Transforming growth factor beta 1: Secondary structure as determined by heteronuclear magnetic resonance spectroscopy. Biochemistry 32(4), 1164 – 1171 (1993). 948. J. L. Wrana, L. Attisano, R. Wieser, F. Ventura, and J. Massague, Mechanism of activation of the TGF-ß receptor. Nature 370, 341 – 346 (1994).

185 949. D. M. Kingsley, What do BMPs do in mammals? Clues from the mouse short-ear mutation. Trends Genet. 10, 16 – 21 (1994). 950. P. H. Frances, M. K. Richardson, P. M. Brickell, and C. Ticle, Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209 – 218 (1994). 951. F. S. Kaplan, J. A. Tabas, and M. A. Zasloff, Fibrodysplasia ossificans proressiva: A clue from the fly? Calcif. Tissue Int. 47, 117 – 125 (1989). 952. J. L. Christian, BMP, Wnt and Hedgehog signals: How far can they go? Curr. Opin. Cell. Biol. 12, 244 – 249 (2000). 953. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin-sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 954. E. Canalis, T. McCarthy, and M. Centrella, Isolation of growth factors from adult bovine bone. Calcif. Tissue Int. 43, 346 – 351 (1988). 955. D. J. Baylink, R. D. Finkelman, and S. Mohan, Growth factors to stimulate bone formation. J. Bone Miner. Res. 8, S565 – S572 (1993). 956. E. Solheim, Growth factors in bone. Int. Orthop. 22, 410 – 416 (1998). 957. C. J. Rosen, L. R. Donahue, and S. J. Hunter, Insulin-like growth factors and bone: The osteoporosis connection. Proc. Soc. Exp. Biol. Med. 206, 83 – 102 (1994). 958. D. T. Graves, A. J. Owen, R. K. Barth, P. Tempst, A. Winoto, L. Fors, and L. E. Hood, Detection of c-sis transcripts and synthesis of PDGF-like proteins by human osteosarcoma cells. Science 265, 972 (1984). 959. Y. Ishimi, C. Miyaura, C. H. Jin, T. Akatsu, T. Abe, Y. Nakamura, A. Yamaguichi, S. Yoshiki, T. Matsuda, T. Hirano, T. Kishimoto, and T. Suda, IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297 – 3303 (1990). 960. P. E. Keeting, L. Rifas, S. A. Harris, D. S. Colvard, T. C. Spelsberg, W. A. Peck, and B. L. Riggs, Evidence for interleukin-1ß production by cultured normal human osteoblast-like cells. J. Bone Miner. Res. 6, 827 – 833 (1991). 961. E. C. Weir, K. L. Insogna, and M. C. Horowitz, Osteoblast-like cells secrete granulocyte-macrophage colony-stimulating factor in response to parathyroid hormone and lipopolysaccaride. Endocrinology 124, 899 – 904 (1989). 962. M. Gowen, K. Chapman, A. Littlewood, D. Hughes, D. Evans, and G. Russell, Production of tumor necrosis factor by human osteoblasts is modulated by other cytokines, but not by osteotropic hormones. Endocrinology 126, 1250 – 1255 (1990). 963. G. R. Mundy, Immune system and bone remodeling. Trends Endocrin. Metab. 1, 307 – 311 (1990). 964. J. D. Bjorge, T. O. Chan, M. Antcza, H. J. Kung, and D. J. Fujita, Activated type I phosphoinostol kinase is associated with the epidermal growth factor (EGF) receptor following EGF stimulation. Proc. Natl. Acad. Sci. USA 87, 3816 – 3820 (1990). 965. M. Berridge, The molecular basis of communication within the cell. Sci. AM 253, 142 – 152 (1985). 966. M. Centrella, T. L. McCarthy, and E. Canalis, Growth factors and cytokines. In “Bone” (B. Hall, ed.), Vol. 4, pp. 47 – 73. CRC Press, Boca Raton, FL, 1992. 967. T. Yoneda, Cytokines in bone: Local translators in cell-to-cell communications. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 375 – 412. Academic Press, New York, 1993. 968. P. D. Delmas, R. P. Tracy, B. L. Riggs et al., Identification of the noncollagenous proteins of bovine bone by two-dimensional gel electrophoresis. Calcif. Tissue Int. 36, 308 (1984). 969. I. R. Dickson, A. R. Poole, and A. Veis, Localization of plasma 2HS-glycoprotein in mineralizing human bone. Nature 256, 430 – 432 (1975).

186 970. B. A. Ashton, H.-J. Hohling, and J. T. Triffit, Plasma proteins present in human cortical bone: Enrichment of the 2HS-glycoprotein. Calcif. Tissue Res. 22, 27 – 33 (1976). 971. R. A. Terkeltaub, D. A. Santoro, G. Mandel, and N. Mandel, Serum and plasma inhibit neutrophil stimulation by hydroxyapatite crystals: Evidence that serum alpha 2-HS glycoprotein is a potent and specific crystal-bound inhibitor. Arthritis Rheum. 31, 1081 – 1089 (1988). 972. K. Aoyama and J. Chiba, Separation of different molecular forms of mouse IgA and IgM monoclonal antibodies by high-performance liquid chromatography on spherical hydroxyapatite beads. J. Immunol. Methods 162, 201 – 210 (1993). 973. H. Iwase, K. Gondo, T. Koike, and I. Ono, Novel precolumn deproteinization method using a hydroxyapatite cartridge for the determination of theophylline and diazepam in human plasma by highperformance liquid chromatography with ultraviolet detection. J. Chromatogr. B. Biomed. Appl. 655, 73 – 81 (1994). 974. H. E. Schultze, K. Heide, and H. Haupt, Charterisierung eines niedermolekularen a2 mukoids aus human serum. Natur-wissenschaften. 49, 440–453 (1962). 975. F. Yang, Z. Schwartz, L. Swain, C.-C. Lee, B. Bowman, and B. Boyan, 2-HS-glycoprotein: Expression in chondrocytes and augmentation of alkaline phosphatase and phospholipase A2 activity. Bone 12, 7 – 15 (1991). 976. T. Ohnishi, N. Arakaki, O. Nakamura, S. Hirono, and Y. Daikuhara, Purification, characterization, and studies on biosynthesis of a 59-kDA bone sialic acid-containing protein (BSP) from rat mandible using a monoclonal antibody: Evidence from rat mandible using a monoclonal antibody that 59-kDa BSP may be the rat counterpart of human 2-HS glycoprotein and is synthesized by both hepatocytes and osteoblasts. J. Biol. Chem. 266, 14636 – 14645 (1991). 977. T. Ohnishi, O. Nakamura, M. Ozawa, N. Arakaki, T. Muramatsu, and Y. Daikuhara, Molecular cloning and sequence analysis of cDNA for a 59kD bone sialoprotein of the rat: Demonstration that it is a counterpart of human a2-HS glycoprotein and bovine fetuin. J. Bone Miner Res. 8, 367 – 377 (1993). 978. M. Mizuno, M. C. Farach-Carson, G. J. Pinero, R. Fujisawa, J. C. Brunn, J. M. Seyer, G. R. Bousfield, M. P. Mark, and W. T. Butler, Identification of the rat bone 60K acidic glycoprotein as 2HSglycoprotein. Bone Miner. 13, 1 – 12 (1991). 979. F. Gejyo, J. L. Chang, W. Burgi, K. Schmid, G. D. Offner, and R. H. Troxler, Characterization of the ß-chain of human plasma a2HS-glycoprotein. J. Biol. Chem. 258, 4966 – 4971 (1983). 980. R. H. Troxler and K. Schmid, The complete aminoacid sequence of the A-chain of human plasma 2-HS glycoprotoein. J. Biol. Chem. 261, 1665 – 1676 (1986). 981. C. C. Lee, B. H. Bowman, and F. Yang, Human 2-HS-glycoprotein: The A and B chains with a connecting sequence are encoded by a single mRNA transcript. Proc. Natl. Acad. Sci. USA 84, 4403 – 4407 (1987). 982. K. Matsushima, M. Cheng, and S. Migita, Purification and physicochemical characterization of human alpha 2-HS-glycoprotein. Biochim. Biophys. Acta 701, 200 – 205 (1982). 983. T. Araki, Y. Yoshioka, and K. Schmid , The position of the disulfide bonds in human plasma ? 2 HS-glycoprotein and the repeating double disulfide bonds in the domain structure. Biochim. Biophys. Acta 994, 195 – 199 (1989). 984. J. Kellerman, H. Haupt, E.-A. Auerswald, and W. Muller-Esterl, The arrangement of disulfide loops in human 2-HS glycoprotein: Similarity to the disulfide bridge structures of cystatins and kininogens. J. Biol. Chem. 264, 14121 – 14128 (1989). 985. D. L. Christie, K. M. Dziegielewska, R. M. Hill, and N. R. Sanders, Fetuin: The bovine homologue of human 2HS glycoprotein. FEBS Lett. 214, 226 – 227 (1987).

GOKHALE, BOSKEY, AND ROBEY 986. M. Haasemann, P. Nawratil, and W. Muller-Esterl, Rat tyrosine kinase inhibitor shows sequence similarity to human alpha 2-HS glycoprotein and bovine fetuin. Biochem. J. 274, 899 – 902 (1991). 987. T. Hayase, K. G. Rice, K. M. Dziegielewska, M. Kuhlenschmidt et al., Comparison of N-glycosides of fetuins from different species and human alpha 2-HS-glycoprotein. Biochemistry 31, 4915 – 4921 (1992). 988. C. Binkert, M. Demetriou, B. Sukhu, M. Szweras, H. C. Tenenbaum, and J. W. Dennis, Regulation of osteogenesis by fetuin. J. Biol. Chem. 274, 28514 – 28520 (1999). 989. C. C. Lee, B. H. Bowman, and F. M. Yang, Human alpha 2-HS-glycoprotein: The A and B chains with a connecting sequence are encoded by a single mRNA transcript. Proc. Natl. Acad. Sci. USA 84(13), 4403 – 4407 (1987). 990. J. G. Lewis and C. M. Andre, A serum DNA-binding protein in malignant disease. FEBS Lett. 92, 211 (1978). 991. J. G. Lewis and C. M. Andre, Effect of human a2HS-glycoprotein on mouse macrophage function. Immunology 3, 329 (1980). 992. P. R. Srinivas, A. S. Wagner, L. V. Reddy, D. D. Deutsch et al., Serum alpha 2-HS-glycoprotein is an inhibitor of the human insulin receptor at the tyrosine kinase level. Mol. Endocrinol. 7, 1145 – 1155 (1993). 993. W. P. Bradley, A. P. Blasco, J. F. Weiss, J. C. Alexander, N. A. Silverman, and P. B. Chretien, Correlations among serum proteinbound carbohydrates, serum glycoprotein, lymphocyte reactivity and tumor burden in cancer patients. Cancer 40, 2264 – 2275 (1977). 994. C. J. van Oss, C. F. Gillman, P. M. Bronson, and J. R. Border, Opsonic-properties of human serum 2-HS-glycoprotein. Immun. Commun. 3, 329 – 335 (1974). 995. F. Yang, Z. L. Chen, J. M. Bergeron, R. L. Cupples, and W. E. Friedrichs, Human 2-HS-glycoprotein/bovine fetuin homoluge in mice: Identification and developmental retulation of the gene. Biochim. Biophys. Acta 1130, 149 – 156 (1992). 996. I. R. Dickson, M. K. Bagga, and C. R. Paterson, Variations in the serum concentration and urine excretion of 2-HS-glycoprotein, a bone-related protein, in normal individuals and in patients with osteogenesis imperfecta. Calcif. Tissue Int. 35, 16 – 20 (1983). 997. Quelch, W. G. Cole, and R. A. Melick, Noncollagenous proteins in normal and pathological human bone. Calcif. Tissue Int. 36, 545 – 549 (1985). 998. J. T. Triffit, U. Gehauer, B. A. Ashton, M. H. Owen, and J. J. Reynolds, Origin of plasma 2-HS-glycoprotein and its accumulation in bone. Nature 262, 226 – 227 (1976). 999. N. C. Blumenthal, F. Betts, and A. S. Posner, Effect of carbonate and biological macromolecules on formation and properties of hydroxyapatite. Calcif. Tissue Res. 18, 81 – 90 (1975). 1000. W. Jahnen-Dechent, T. Schinke, A. Trindl, W. Muller-Esterl, F. Sablitzky, S. Kaiser, and M. Blessing, Cloning and targeted deletion of the mouse fetuin gene. J. Biol. Chem. 272, 31496 – 31503 (1997). 1001. G. C. Colclasure, W. S. Lloyd, M. Lamkin, W. Gonnerman, R. F. Troxler, G. D. Offner, W. Burgi, K. Schmid, and R. B. Nimberg, Human serum 2HS-glycoprotein modulates in vitro bone resorption. J. Clin. Endocrinol. Metab. 66, 187 – 192 (1988). 1002. J. Garnett and P. Dieppe, The effects of serum and human albumin on calcium hydroxyapatite crystal growth. Biochem. J. 266, 863 – 868 (1990). 1003. H. Gilman and D. W. Hukins, Seeded growth of hydroxyapatite in the presence of dissolved albumin. J. Inorg. Biochem. 55, 21 – 30 (1994). 1004. H. Gilman and D. W. Hukins, Seeded growth of hydroxyapatite in the presence of dissolved albumin at constant composition. J. Inorg. Biochem. 55, 31 – 39 (1994). 1005. H. Furedi-Milhofer, Adsorption of human serum albumin on precipitated hydroxyapatite. J. Colloid. Interface Sci. 69, 460 – 480 (1979).

CHAPTER 4 The Biochemistry of Bone 1006. M. J. Mura-Galelli, J. C. Voegel, S. Behr, E. F. Bres, and P. Schaaf, Adsorption/desorption of human serum albumin on hydroxyapatite: A critical analysis of the Langmuir model. Proc. Natl. Acad. Sci. USA 88, 5557 – 5561 (1991). 1007. Y. Yoshiuka, F. Gejyo, T. Marti, E. E. Ricki, W. Burgi, G. D. Offner, S. Shimabayashi, Y. Tanizawa, and K. Ishida, Effect of phosphorylated organic compound on the adsorption of bovine serum albumin by hydroxyapatite. Chem. Pharm. Bull. 39, 2183 – 2188 (1991). 1008. H. Iwase, K. Gondo, T. Koike, and I. Ono, Novel precolumn deproteinization method using a hydroxyapatite cartridge for the determination of theophylline and diazepam in human plasma by high-performance liquid hromotography with ultraviolet detection. J. Chromotogr. B Biomed. Appl. 665, 73 – 81 (1994). 1009. A. L. Boskey, V. Vigorita, and P. G. Bullough, Calcium-acidic phospholipid-phosphate complexes: Promoter of mineralization common to pathologic hydroxyapatite-containing calcifications. Am. J. Pathol. 133, 22 – 29 (1988). 1010. G. Gronowicz, F. N. Woodiel, M. B. McCarthy, and L. G. Raisz, In vitro mineralization of fetal rat parietal bones in defined serum-free medium: Effect of ß-glyceral phosphate. J. Bone Miner. Res. 4, 313 – 324 (1989). 1011. D. D. Lee and M. J. Glimcher, Three-dimensional spatial relationship between the collagen fibrils and the inorganic calcium phosphate crystals of pickeral (Americanus americanus) and herring (Clupea harengus) bone. J. Mol. Biol. 217, 487 – 501 (1991). 1012. S. Weiner, Organization of extracellularly mineralized tissues: A comparative study of biological crystal growth. CRC Crit. Rev. Biochem. 20, 365 – 408 (1986). 1013. E. Dmitrovsky and A. L. Boskey, Calcium acidic phospholipid phosphate complexes in human atherosclerotic aorta. Calcif. Tissue Int. 37, 121 – 125 (1985). 1014. J. M. Burnell, D. J. Baylink, C. H. Chestnut, III, and E. J. Teubner, The role of skeletal calcium deficiency in postmenopausal osteoporosis. Calcif. Tissue Int. 38, 187 – 192 (1986). 1015. A. E. N. Nielsen, “Kinetics of Precipitation.” Pergammon Press, Oxford, 1964. 1016. G. H. Nancollas and B. Tomazic, Growth of calcium phosphate on hydroxyapatite crystals: Effect of supersaturation and ionic medium. J. Phys. Chem. 78, 2218 – 2225 (1974). 1017. A. L. Boskey and A. S. Posner, Conversion of amorphous calcium phosphate to microcrystalline hydroxyapatite: A pH-dependent, solution-mediated, solid-solid conversion. J. Phys. Chem. 77, 2313 – 2317 (1973). 1018. W. T. Brown and L. C. Chow, Chemical properties of bone mineral. Annu. Rev. Mater. Sci. 6, 213 – 262 (1976). 1019. E. D. Eanes and J. L. Meyer, The maturation of crystalline calcium phosphates in adqueous suspensions at physiologic pH. Calcif. Tissue Res. 23, 259 – 269 (1977). 1020. W. E. Brown, N. Eidelman, and B. Tomazic, Octacalcium phosphate as a precursor in biomineral formation. Adv. Dent. Res. 1, 306 – 313 (1987). 1021. L. J. Bréceviç and H. Füredi-Milhofer, Precipitation of calcium phosphates from electrolyte solutions. II. The formation and transformation of precipitates. Calcif. Tissue Res. 10, 82 – 90 (1972). 1022. J. Moradian-Oldak, S. Weiner, L. Addadi, W. J. Landis, and W. Traub, Electron imaging and diffraction study of individual crystals of bone, mineralized tendon and synthetic carbonate apatite. Connect. Tissue Res. 25, 219 – 228 (1991). 1023. H. Füredi-Milhofer, J. Moradian-Oldak, S. Weiner, A. Veis, K. P. Mintz, and L. Addadi, Interactions of matrix proteins from mineralized tissues with octacalcium phosphate. Connect. Tissue Res. 30, 251 – 264 (1994). 1024. D. Hanein, B. Geiger, and L. Addadi, Fibronectin adsorption to surfaces of hydrated crystals: An analysis of the importance of bound water in protein– substrate interactions. Langmuir 9, 1058 – 1065 (1993a).

187 1025. J. Aizenberg, S. Albeck, S. Weiner, and L. Addadi, Crystal – protein interactions studied by overgrowth of calcite on biogenic skeletal elements. J. Crystal Growth 142, 156 – 164 (1994). 1026. D. Hanein, H. Sabanay, L. Addadi, and B. Geiger, Selective interactions of cells with crystal surfaces: Implications for the mechanism of cell adhesion. J. Cell Sci. 104, 275 – 288 (1993). 1027. A. Linde, A. Lussi, and M. A. Crenshaw, A mineral induction by immobilized polyanionic proteins. Calcif. Tissue Int. 44, 286 – 295 (1989). 1028. G. K. Huneter, In vitro studies on matrix-mediated mineralization. In “Bone Metabolism and Mineralization” (B. K. Hall, ed.), pp. 225 – 247. CRC Press, Boca Raton, FL, 1992. 1029. B. R. Heywood and E. D. Eanes, An ultrastructural study of the effects of acidic phospholipid substitutions on calcium phosphate precipitation in anionic liposomes. Calcif. Tissue Int. 50, 149 – 156 (1992). 1030. E. D. Eanes, A. W. Hailer, R. J. Midura, and V. C. Hascall, Proteoglycan inhibition of calcium phosphate precipitation in liposomal suspensions. Glycobiology 2, 571 – 578 (1992). 1031. D. Skrtic and E. D. Eanes, Effect of different phospholipid-cholesterol membrane compositions on liposome-mediated formation of calcium phosphates. Calcif. Tissue Int. 50, 253 – 260 (1992a). 1032. D. Skrtic and E. D. Eanes, Membrane-mediated precipitation of calcium phosphate in model liposomes with matrix vesicle-like lipid composition. Bone Miner. 16, 109 – 119 (1992). 1033. B. Ganss, R. H. Kim, and J. Sodek, Bone sialoprotein. J. Crit. Rev. Oral Biol. Med. 10, 79 – 98 (1999). 1034. A. L. Boskey, L. Spevak, S. B. Doty, and L. Rosenberg, Effects of bone CS-proteoglycans, DS-decorin, and DS-biglycan on hydroxyapatite formation in a gelatin gel. Calc. Tissue Int. 61, 298 – 305 (1997). 1035. K. Sato, Y. Kumagai, and J. Tanaka, Apatite formation on organic monolayers in simulated body environment. J. Biomed. Mater. Res. 50, 16 – 20 (2000). 1036. A. L. Boskey, B. D. Boyan, S. B. Doty, A. Feliciano, K. Greer, D. Weiland, L. D. Swain, and Z. Schwartz, Studies of matrix vesicleinduced mineraliztion in a gelatin gel. Bone Miner. 17, 257 – 263 (1992). 1037. P. Gehron Robey, Cell-mediated mineralization in vitro. In “Mineralization in Biological Systems” (E. Bonucci, ed.), pp. 107 – 127. CRC Press, Boca Raton, FL, 1992. 1038. H. I. Roach, Induction of normal and dystrophic mineralization by glycero-phosphates in long-term bone organ culture. Calcif. Tissue Int. 50, 553 – 562 (1992). 1039. H. Fell, The histogenesis of cartilage and bone in the long bones of the embryonic fowl. J. Morpol. 40, 417 – 459 (1925). 1040. H. I. Roach, Long-term organ culture of embryonic chick femurs: A system for investigating bone and cartilage formation at an intermediate level of organization. J. Bone Miner. Res. 5, 85 – 100 (1990). 1041. H. C. Tenenbaum and J. N. M. Heersche, Differentiation of osteoblasts and formation of mineralized bone in vitro. Calcif. Tissue Int. 34, 76 – 79 (1982). 1042. H. C. Tenenbaum and J. N. M. Heersche, Dexamethasone stimulates osteogenesis in chick periosteum in vitro. Endocrinology 117, 2211 – 2217 (1985). 1043. A. J. Friedenstein, K. V. Petrakova, A. I. Kurolesova, and G. P. Frolova, Heterotopic transplants of bone marrow: Analysis of precursor cells for osteogenic and hematopoietic tissues. Transplant 6, 230 – 247 (1968). 1044. M. Owen, Marrow stromal stem cells. J. Cell. Sci. S10, 63 – 76 (1988). 1045. B. A. Ashton, T. D. Allen, C. R. Howlett, C. C. Eaglesom, A. Hattori, and M. Owen, Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo. Clin. Orthop. Relat. Res. 151, 294 – 307 (1980).

188 1046. D. Benayahu, A. Fried, D. Zipori, and S. Weintroub, Subpopulations of marrow stromal cells share a variety of osteoblastic markers. Calcif. Tissue Int. 49, 202 – 207 (1991). 1047. J. N. Beresford, Osteogenic stem cells and the stromal system of bone and marrow. Clin. Orthop. Relat. Res. 240, 270 – 280 (1989). 1048. A. E. Grigoriadis, J. N. M. Heersche, and J. E. Aubin, Differentation of muscle fat cartilage and bone from progenitor cells present in a bone derived clonal cell population: Effects of dexamethasone. J. Cell. Biol. 106, 2139 – 2151 (1988). 1049. C. G. Bellows and J. E. Aubin, Determination of numbers of osteoprogenitors present in isolated fetal rat calvaria cells in vitro. Dev. Biol. 133, 8 – 13 (1989). 1050. C. G. Bellows, J. E. Aubin, and J. N. M. Heersche, Physiological concentrations of glucocorticoids timulate formation of bone nodules from isolated rat calvaria cells in vitro. Endocrinology 121, 1985 – 1992 (1989). 1051. A. Yamaguichi, T. Katagiri, T. Ikeda, J. M. Wozney, V. Rosen, E. A. Want, A. J. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell Biol. 113, 681 – 687 (1991). 1052. S. Sakamoto, M. Sakamoto, L. Goldberg, L. Colarusso, and Y. Gotoh, Mineralization induced by beta-glycerophosphate in cultures leads to a marked incrased in collagenase synthesis by mouse osteogenic MC3T3-E1 cells under subsequent stimulation with heparin. Biochem. Biophys. Res. Commun. 162, 773 – 780 (1989). 1053. R. T. Franceschi, C. J. Linson, T. C. Peter, and P. R. Romano, Regulation of cellular adhesion and fibronectin synthesis by 1 alpha,25dihydroxyvitamin D3. J. Biol. Chem. 262(9), 4165 – 4171 (1987). 1054. S. B. Rodan, Y. Imai, M. A. Thiede, G. Wesolowski, D. Thompson, Z. Bar-Shavit, S. Shull, K. Mann, and G. A. Rodan, Characterization of a human osteosarcoma cell line (Saos-2) with osteoblastic properties. Cancer Res. 47(18), 4961 – 4966 (1987). 1055. H. Sudo, H. A. Kodama, Y. Amagai, S. Yamamoto, and S. Kasai, In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96(1), 191 – 198 (1983). 1056. J. K. Heath, S. B. Rodan, K. G. Yoon, and G. A. Rodan, Rat calvarial cell lines immortalized with SV-40 large T-antigen: Constitutive and retinoic acid-inducible expression of osteoblastic features. Endocrinology 124, 3060 – 3068 (1989). 1057. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7(2), 127 – 136 (1992). 1058. S. A. Harris, R. J. Enger, B. L. Riggs, and T. C. Spelsberg, Development and characterization of a conditionally immortalized human fetal osteoblastic cell line. J. Bone Miner. Res. 10(2), 178 – 186 (1995). 1059. T. S. Owen, M. S. Aronow, V. Shaloub, D. M. Barone, L. Wiling, M. Tassinari, M. B. Kennedy, S. Pockwinse, J. B. Lian, and G. S.

GOKHALE, BOSKEY, AND ROBEY

1060.

1061.

1062.

1063.

1064.

1065.

1066. 1067. 1068.

1069. 1070. 1071. 1072.

1073.

1074.

1075.

Stein, Progressive development of the rat osteoblast phenotype in vitro: Reciprocal relationships in expression of genes association with osteoblast proliferation and differentiation during formation of the bone extraellular matris. J. Cell. Physiol. 143, 420 – 430 (1990). K. Turksen and J. E. Aubin, Positive and negative immunoselection for enrichment of two classes of osteoprogenitor cells. J. Cell Biol. 114(2), 373 – 384 (1991). B. Ecarot-Charrier, N. Shepard, G. Charette, M. Grynpas, and F. H. Glorieux, Mineralization in osteoblast cultures: A light and electron microscopic study. Bone 9, 147 – 154 (1988). L. C. Gerstenfeld, S. D. Chipman, C. M. Kelly, K. J. Hodgens, and D. D. Lee, Collagen expression, ultrastructural assembly, and mineralization in cultures of chick embryo osteoblasts. J. Cell Biol. 106, 979 – 989 (1988). J. R. Nefussi, M. L. Boy-Lefebre, H. Boulekbache, and N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion. Differentiation 29, 160 – 168 (1985). M. Weinreb, D. Shinar, and G. A. Rodan, Different pattern of alkaline phosphatase, osteopontin, and osteocalcin expression in developing rat bone visualized by in situ hybridization. J. Bone Miner. Res. 5, 831 – 842 (1990). S. W. Whitson, M. A. Whitson, D. E. Bowers, Jr., and M. C. Falk, Factors influencing synthesis and mineralization of bone matrix from fetal bovine bone cells grown in vitro. J. Bone Miner. Res. 7(7), 727 – 741 (1992). H. C. Anderson, Electron microscopic studies of induced cartilage development and calcification. J. Cell Biol. 35, 81 (1967). E. Bonnuci, Fine structure of early cartilage calcification. J. Ultrastruct. Res. 201, 33 (1967). G. W. Bernard and D. C. Pease, An electron microscopic study of initial intramembranous osteogenesis. Am. J. Anat. 125, 271 – 290 (1969). S. Y. Ali, Analysis of matrix vesicles and their role in the calcification of epiphyseal cartilage. Fed. Proc. 35, 135 – 142 (1976). H. C. Anderson, Matrix vesicle calcification. Metab. Bone Dis. Rel. Res. 1, 83 (1978). H. C. Anderson, Biology of disease: Mechanism of mineral formation in bone. Lab. Invest. 60, 320 – 331 (1989). R. E. Wuthier, L. N. Wu, G. R. Sauer, and B. R. Genge, Mechanism of matrix vesicle calcification: Characterization of ion channels and the nucleational core of growth plate vesicles. Bone Miner. 17, 290 – 295 (1992). R. E. Wuthier, Electrolytes of isolated epiphyseal chondrocytes, matrix vesicles, and extracellular fluid. Calcif. Tissue Res. 23, 125 – 133 (1977). T. Kirsch and R. E. Wuthier, Stimulation of calcification of growth plate cartilage matrix vesicles by binding to type II and X collagens. J. Biol. Chem. 269, 11462 – 11469 (1994). L. N. Wu, B. R. Genge, G. C. Lloyd, and R. E. Wuthier, Collagenbinding proteins in collagenase-released matrix proteins and different types of collagen. J. Biol. Chem. 266, 195 – 203 (1991).

CHAPTER 5

Developmental Biology of Bone ANTHONY M. REGINATO,*† WENFANG WANG,* AND BJORN R. OLSEN* *Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115, and † Arthritis Unit, Massachusetts General Hospital, Boston, Massachusetts 02114

I. II. III. IV. V.

Introduction Craniofacial Bone Development Axial Bone Development Limb Bone Development Sonic Hedgehog and Indian Hedgehog in Skeletal Development VI. Bone Morphogenetic Proteins, Homologues, and Their Antagonists in Skeletal Development

VII. Fibroblast Growth Factors and Their Receptors in Skeletal Development VIII. Chondrocyte Differentiation and Endochondral Bone Formation IX. Osteoblast Differentiation and Function X. Cell – Matrix Interactions in Skeletal Development XI. Summary and Perspectives References

I. INTRODUCTION

initial German descriptions) of the future bones (Fig. 1, see also color plate). The cartilage anlagen is replaced by bone during endochondral ossification. During this process, chondrocytes in the center of cartilage undergo a program of hypertrophy and apoptosis and the matrix calcifies and is invaded by blood vessels, osteoblasts, and osteoclasts. Gradually, the cartilage is replaced by bone and bone marrow. At the same time, differentiation of cells to osteoblasts in the surrounding perichondrium leads to the formation of a sleeve of bone that surrounds the developing bone marrow space. As the bone marrow space expands toward the ends of the anlagen, the process of chondrocyte hypertrophy, apoptosis, and vessel ingrowth continues at the ends, defining what becomes the epiphyseal growth plate (Fig. 2, see also color plate). Growth plates are the centers for longitudinal bone growth and their cartilage is composed of at least three types of cells: (1) resting chondrocytes, (2) proliferating

The development of the vertebrate skeleton depends on the regulated differentiation, function, and interactions of its cellular components. The cells are derived from three sources: cranial neural crest, paraxial mesoderm, and lateral mesoderm. The cranial neural crest gives rise to most of the craniofacial skeleton, the paraxial mesoderm forms the axial skeleton, and the lateral plate mesoderm generates the appendicular skeleton. Mesenchymal cells from these three compartments undergo condensation at the future sites of cartilage or bone and differentiate into chondrocytes or osteoblasts depending on the location within the developing skeleton. In membranous bones, such as the calvaria and the mandible, mesenchymal cells differentiate directly into osteoblasts, whereas in endochondral bones, mesenchymal cells differentiate into chondrocytes that produce cartilage models (frequently called anlagen, based on

OSTEOPOROSIS, SECOND EDITION VOLUME 1

189

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

190

REGINATO, WANG, AND OLSEN

FIGURE 1 Diagram illustrating how mesenchymal cell condensations give rise to membranous bones (e.g., in calvaria) by direct differentiation of osteoblasts, bone matrix production, and bone growth and remodeling, or to endochondral bone (e.g., in long, tubular bones) by differentiation of chondrocytes, production of cartilage models (anlagen), followed by replacement of the cartilage by bone, and bone growth and remodeling. In some cases (e.g., the mandible) membranous bone is formed around a cartilage model (in the case of the mandible, this cartilage is the Meckel’s cartilage), but does not replace the cartilage as in endochondral ossification. (See also color plate.)

FIGURE 2

Proximal end (left) and proximal growth plate region (right) of tibia from a 17-day old mouse. The growth plate, with layers of proliferating and hypertrophic cartilage, separates the bone and marrow of the secondary ossification center in the epiphysis from the trabecular bone and marrow in the diaphysis of the tibia. Courtesy of Dr. D. Glotzer. (See also color plate.)

191

CHAPTER 5 Developmental Biology of Bone

FIGURE 3 In situ hybridization showing expression of Col10a1 in hypertrophic chondrocytes of growth plates of tibia and metatarsal bones from a mouse embryo at day 17.5 of development. Courtesy of Dr. E. Zelzer. (See also color plate.)

chondrocytes, and (3) hypertrophic chondrocytes, which are larger and surrounded by a mineralized matrix containing type X collagen as a specific marker (Fig. 3, see also color plate). Hypertrophic chondrocytes die through a process of apoptosis, and osteoblasts, which are brought in with invading blood vessels, form trabecular bone under the growth plate. Ultimately, all cartilage in anlagen is replaced by bone with the exception of the articular surfaces of joints and costal cartilage. Current understanding of the molecular pathways that control patterning, growth, and differentiation of the various skeletal elements has been derived from a combination of genetics, experimental developmental biology, and in vitro cellular and biochemical studies. Knowledge of embryonic skeletal development will provide a better basis for generating strategies to repair cartilage and bone in patients with skeletal diseases such as osteoporosis.

II. CRANIOFACIAL BONE DEVELOPMENT Craniofacial bones arise from two distinct lineages of skeletogenic mesenchyme: neural crest and paraxial mesoderm. Neural crest cells from the caudal midbrain and rhombomeres 1, 2, and 4 produce cartilage and bones of the face, jaws, and the front and sides of the brain case (frontal, temporal and parietal bones) [1 – 3]. The paraxial mesoderm gives rise to the posterior parts of the head skeleton and the skull base. Most of the craniofacial bones, such as the calvaria and jaws, are formed by membranous bone formation. During this process, osteoblasts are formed

by direct differentiation from mesenchymal cells (Fig. 1). Condensed mesenchymal cells form a multilayered membrane and osteogenesis begins within the core of the membrane and radiates outward. In some bones, such as the mandible, membranous bone formation occurs around a core of cartilage (Meckel’s cartilage) (Fig. 1). The transcription factor Cbfa1 is required for osteoblast differentiation and is thus essential for both membranous and endochondral bone formation [4 – 6]. Disruption of Cbfa1 in mice leads to a lack of osteoblast differentiation and complete absence of bone [5,6]. Craniofacial bone formation is dynamic and complex and no simple molecular pattern has emerged from studies of gene knockouts in mice. Inactivation of genes for transcription factors such as gsc (murine homologue of Drosophila goosecoid [7]), Otx2 (homologue of Drosophila orthodenticle – otd [8,9]), Mf1 (forkhead winged-helix transcription factor [10]), Dlx1, 2, and 5 [11,12], and Pax3 [13] affects proliferation, differentiation, and/or migration of neural crest cells and cause malformation or absence of certain craniofacial bones. In humans, abnormalities that affect neural crest cells, and consequently patterning of craniofacial bones, include several types of Waardenburg syndrome (Table 1) [14]. Types 1 and 3 of the syndrome, caused by mutations in the transcription factor PAX3, are characterized by a combination of craniofacial bone and soft tissue abnormalities, partial albinism, hearing loss, spina bifida, cleft lip/palate, and scapular anomalies [15]. In the calvarium, osteogenic fronts from neighboring skeletal elements meet at the cranial sutures. During early postnatal life these sutures remain open to allow the cranial vault to grow and expand to accommodate the enlarging

192

REGINATO, WANG, AND OLSEN

TABLE 1

Human Genetic Disorders and Their Associated Genes

Human syndrome Waardenburg syndrome type 1 (WS1)

Gene(s) affecteda PAX3

Waardenburg syndrome type 3 (WS3)

PAX3

Spondylocostal dysostosis (SCDO1)

DLL3

Synpolydactyly (SPD)

HOXD13

Pallister – Hall syndrome (PHS)

GLI3

Postaxial polydactyly type A (PAP-A)

GLI3

Greig cephalopolysyndactyly syndrome (GCPS)

GLI3

Brachydactyly (BD)

CDMP1/GDF5

Acromesomelic dysplasia Hunter – Thompson type (AMDH)

CDMP1/GDF5

Acromesomelic dysplasia Grebe type (AMDG)

CDMP1/GDF5

Proximal symphalangism (SYM1)

NOG

Multiple synostosis syndrome 1 (SYNS1)

NOG

Pfeiffer syndrome

FGFR1, FGFR2

Crouzon syndrome (CS)

FGFR2

Apert syndrome

FGFR2

Jackson – Weiss syndrome (JWS)

FGFR2

Beare – Stevenson syndrome

FGFR2

Saethre – Chotzen syndrome (SCS)

TWIST

Muenke syndrome

FGFR3

Crouzon syndrome with acanthosis nigricans

FGFR3

Achondroplasia (ACH)

FGFR3

Hypochondroplasia (HCH)

FGFR3

Thanatophoric dysplasia (TD)

FGFR3

Campomelic dysplasia (CMD1)

SOX9

Jansen metaphyseal chondrodysplasia

PTHrPR

Blomstrand chondrodysplasia

PTHrPR

Schmid metaphyseal chondrodysplasia (MCDS)

COL10A1

Cleidocranial dysplasia (CCD)

CBFA1

Osteogenesis imperfecta (OI)

COL1A1, COL1A2

Achondrogenesis type II (ACGII)

COL2A1

Hypochondrogenesis

COL2A1

Spondyloepiphyseal dysplasia congenita (SEDC)

COL2A1

Kniest dysplasia

COL2A1

brain. Premature fusion of the sutures results in craniosynostosis. Mutations in the transcription factors TWIST [16,17] and MSX2 [18] and in the fibroblast growth factor receptors FGFR1, 2, and 3 [19] cause various types of human craniosynostoses, including Saethre – Chotzen, Pfeiffer, Apert, and Crouzon syndromes (see later).

III. AXIAL BONE DEVELOPMENT The axial skeleton of vertebrates is composed of segmental units of vertebrae, intervertebral discs, and ribs. This metameric pattern can be traced back to the somites of early embryos. Somites, paired and morphologically distinct segmental units, are formed on either side of the neural tube and notochord [20]. The process of somitogenesis begins at the head level and continues in a craniocaudal sequence. The segmentation is accompanied by regionalization along the craniocaudal axis, leading to specific segmental identities (e.g., a thoracic vertebra can be distinguished easily from a lumbar vertebra). The dorsolateral part of somites gives rise to the epithelial dermomyotome, whereas the ventromedial part forms the sclerotome [21]. Sclerotomal cells differentiate into chondrocytes that make the vertebral bodies (Fig. 4, see also color plate). A vertebra is formed by the combination of the caudal half of one sclerotome and the rostral half of the next sclerotome. This process is called resegmentation [21]. In N-cadherin-null mutant mice, each epithelial somite is split into rostral and caudal halves, whereas in N-cadherin/cad11 double mutants, more fragmented somites are observed [22]. This suggests that cells in the rostral and caudal compartments of a somite have distinct adhesive affinities and that both N-cadherin and cad11 play roles in holding cells in these two compartments together before resegmentation.

Stickler syndrome Type 1 (STL1)

COL2A1

Type 2 (STL2)

COL11A1

Type 3 (STL3)

COL11A2

Marshall syndrome

COL11A1

Multiple epiphyseal dysplasia (MED)

COL9A2, COL9A3, COMP, DTDST

a For gene notations we have followed the standard of using capital letters for human genes and a capital letter followed by lowercase letters for mouse genes.

FIGURE 4

Diagram illustrating migration of sclerotomal cells from their somitic origin toward the notochord. Modified from Mundlos and Olsen [34]. (See also color plate.)

193

CHAPTER 5 Developmental Biology of Bone

The segmentation of paraxial mesoderm into somites is regulated by Notch signaling pathways. The transmembrane receptor Notch1 is strongly expressed in presomitic mesoderm, and mouse embryos carrying null alleles of Notch1 die during gestation with a delay in the transition from presomitic mesoderm to epithelial somites [23,24]. Expression of the Notch ligand Delta-like (Dll1) starts in the paraxial mesoderm, in presomitic mesoderm and posterior halves of somites [25]. In Dll1-deficient mouse embryos, the primary metameric presomites are established, but the segments have no craniocaudal polarity and no epithelial somites form [26]. Mutations in a downstream effector of Notch signaling, lunatic fringe, also affect somite formation. In lunatic fringe-deficient mice, the boundaries between individual somites fail to form, resulting in a severely disorganized axial skeleton [27,28]. The Mesp2 gene [29] and presenilin genes (PS1 and PS2) have been shown to regulate Notch signaling [30]. Disruption of Mesp2 alters the expression of Dll1 and abolishes the expression of Notch1 and Notch2 [30]. Presenilin genes are required for the spatiotemporal expression of Notch1 and Dll1 [31]. Mice with disrupted Dll3 alleles have vertebral and rib defects that are similar to many human abnormalities of the axial skeleton [32], and mutations in DLL3 have been described in patients with spondylocostal dysostosis [33]. The signaling molecule sonic hedgehog (Shh) controls proliferation and differentiation of cells in the sclerotomes. Shh is expressed in the notochord and the floor plate of the neural tube (Fig. 5A, see also color plate). It promotes sclerotome formation and represses the formation of the dermomyotome, as shown by induction of the sclerotomal marker Pax1 and repression of the dermomyotomal marker Pax3 in vitro and in vivo [35]. Shh-null mice lack most of the sclerotomal derivatives: the vertebral column and the posterior portion of the ribs [36]. Therefore, Shh is critical for development of the axial skeleton. More detailed studies indicate that Shh is not required for the initial induction of the sclerotome, as Shh-null mice display close to normal expression of molecular markers (Pax1, Myf5, and Pax3) for sclerotome, myotome, and dermomyotome [36]. Instead, Shh signaling appears to be required for maintenance of these markers [35,37]. One of the genes that Shh induces and maintains the expression of is the paired-box gene Pax1 [35]. In mice, expression of Pax1 can be detected in sclerotomal cells from embryonic day 8.5 onward [38] (Fig. 5B, see also color plate). These Pax1-positive cells are committed to chondrogenesis. Mice that are homozygous for either mutated (undulated) or inactivated Pax1 are defective in sclerotome differentiation and formation of the vertebral column [38,39]. In the absence of Pax1, the gradual loss of Sox9 and Col2a1 expression in the segmented mesenchyme prevents the sclerotomes from undergoing chondrogenesis [40].

FIGURE 5

(A) Whole mount in situ hybridization showing expression of Sonic hedgehog in the notochord of a mouse at day 9.5 of development. (B) Whole mount in situ hybridization showing expression of the transcription factor Pax1 in the sclerotomes of a mouse embryo at day 9.5 of development. Courtesy of Dr. S. Mundlos. (See also color plates.)

Early craniocaudal identities of the axial skeleton are determined by the expression of different Hox genes [41 – 43]. Almost all segments, such as vertebrae and their attached muscles, differ in size, shape, and structure. This regionalization can be visualized by specific patterns of Hox gene expression or “Hox codes.” Changes in the Hox code lead to a shifting of the regional borders and axial identities called homeotic transformations [44,45].

IV. LIMB BONE DEVELOPMENT Forelimbs and hindlimbs have distinct morphology; however, they share a basic skeletal pattern. A single, proximal long bone (humerus in forelimb and femur in hindlimb) articulates distally with a pair of long bones (ulna and radius in the forelimb and tibia and fibula in the hindlimb). The long bones are followed by a series of

194

FIGURE 6

Diagram illustrating migration of cells into limb bud mesenchyme from somites and lateral plate mesoderm. Modified from Mundlos and Olsen [34]. (See also color plate.)

carpals (forelimb) or tarsals (hindlimb) and digits. This complex three-dimensional structure is patterned along three main axes: proximal – distal (shoulder to digits), anterior – posterior (thumb to small finger), and dorsal – ventral (back of hand to palm). During embryonic development, paired limb buds appear along the lateral body wall, covered at the distal end by a layer of tall epithelial cells called the apical ectodermal ridge (AER). Mesenchymal cells underneath the AER have two separate origins: somites and lateral plate mesoderm (Fig. 6, see also color plate). Lateral plate mesodermal cells give rise to cartilage, tendons, and other connective tissues, whereas cells that originate from the somites migrate into limb buds and form myogenic cells of the muscles [46,47]. The limb axes are established and maintained by different sets of signaling molecules. Factors involved along one axis may sometimes regulate genes responsible for a different axis, thereby making processes along the three axes intimately linked. The proximal – distal axis is primarily controlled by fibroblast growth factors (FGFs) (such as Fgf4 and Fgf8). These factors, released by cells of the AER, stimulate the proliferation of cells in the underlying mesenchyme, the progress zone [48]. The ventral – dorsal axis is controlled by Wnt7a [49] and En-1 [50], and the anterior – posterior axis is regulated by Shh [51]. Shh is expressed by mesenchymal cells in the posterior region of the limb bud and defines the zone of polarizing activity (ZPA) [51]. A positive feedback loop is established between the AER and the ZPA. Fgf activity is required to maintain Shh expression by the ZPA, and Shh induces Fgf expression in the AER [52]. The cross-talk between Shh and Fgfs may be mediated by the Bmp antagonist gremlin (see later) [53]. An upstream regulator of Shh, the basic helix – loop – helix transcription factor dHand, may control Shh expression and establishment of the ZPA [54,55].

REGINATO, WANG, AND OLSEN

Mice lacking the Shh gene have no distal skeletal elements in the limbs; nevertheless, the more proximal skeletal elements (humerus in the forelimbs and femur in the hindlimbs) do develop [36]. The homeobox genes Meis1 and Meis2 have been shown to be expressed in proximal regions of the limb bud and both play an important role in specifying cell fates and differentiation patterns along the proximal – distal axis of the limbs [56,57]. The Brachyury-related (Tbx) genes, Tbx4 and Tbx5, are important for establishing the identities of the limbs [58]. Tbx5 expression is restricted to forelimb mesenchyme buds and Tbx4 to the hindlimbs [59]. Misexpression of Tbx4 in chick forelimb buds leads to the transformation of forelimb structures to those of the hindlimb [60,61]. Misexpression of Pitx1, a homeobox-containing transcription factor that regulates Tbx4 expression in the wing buds of chick embryos, induces a leg-like phenotype [62]. Pitx1-null mice have abnormalities in tibia, fibula, and tarsal bones and they resemble the bones of the forelimb [63,64]. All the major downstream mediators of limb bud patterning genes are not known, but homeobox genes (members of the Hoxa and Hoxd clusters), zinc finger transcription factors (Gli genes), and members of the transforming growth factor (TGF-) superfamily play essential roles. Genes of the Hoxd cluster are expressed in overlapping regions in the posterior and distal zones of the limb bud [65,66]. In humans, expansions of a polyalanine stretch in the N-terminal region of Hoxd13 cause synpolydactyly (SPD), a dominantly inherited human malformation of hands and feet. In heterozygous patients, digits are shorter than normal and central syndactyly is seen [67,68]. A phenocopy of SPD arises when Hoxd11, 12, and 13 gene functions are all eliminated in mice, indicating that the SPD protein interferes with the function of the three Hox proteins by a dominant-negative mechanism [69]. Homozygous mice carrying the same mutation as that found in the human SPD syndrome have the same phenotype as human patients [70]. A positive feedback exists between Hoxd11 or 12 and Shh. Hoxd12 misexpression in transgenic mice produces the transformation of anterior digits to digits of posterior morphology and digit duplications [71]. Hoxd11 or 12 genes can directly amplify the posterior Shh polarizing signal to reinforce the positive feedback loop during limb bud outgrowth [71].

V. SONIC HEDGEHOG AND INDIAN HEDGEHOG IN SKELETAL DEVELOPMENT The Hedgehog (Hh) gene was first identified in Drosophila where it is involved in the patterning of larval segments and adult appendages [72]. Sonic hedgehog (Shh)

195

CHAPTER 5 Developmental Biology of Bone

and Indian hedgehog (Ihh) are vertebrate homologues of Hh [73]. They are secreted molecules that act as extracellular signals to regulate cell proliferation and differentiation. First synthesized as large precursors, they are processed by an autoproteolytic activity of the C-terminal domain [74]. The resulting N-terminal product (the signaling part of the molecule) is further modified by the addition of a cholesterol moiety, which plays a role in facilitating the appropriate spatial distribution [75]. Therefore, these autocatalytic reactions not only release the active form but also affect the distribution of the patterning signal in tissues. The current understanding of the mechanism of Hh signaling in Drosophila is as follows [76]: Upon Hh binding to its transmembrane receptor Patched (Ptc) [77,78], another transmembrane protein, Smoothened (Smo) [79,80], is released from the Ptc – Smo complex and triggers downstream signals via Cubitus interruptus (Ci), a five zinc finger-containing transcription factor [81]. In the presence of Hh signaling, the full-length Ci protein (Ci155) is stabilized and translocates into cell nuclei where it mediates the transcriptional activation of Hh target genes [82]. In the absence of Hh binding, Ptc forms a complex with Smo and inhibits Smo action. This results in proteolysis of Ci155 by a cAMP-activated protein kinase (PKA)-dependent mechanism to generate a 75-kDa N-terminal fragment (Ci75), which enters the nucleus and acts as a repressor of Hh targets [83]. In vertebrates, a similar mechanism has been demonstrated, although it is more complex. Instead of one Ci gene, three Ci homologues have been identified, Gli1, Gli2, and Gli3 [84]. Mammalian Gli proteins display extensive homology to each other but only limited homology to the Drosophila Ci protein outside the zinc finger regions. Gli3 is processed proteolytically in a PKA-dependent manner similar to Ci in flies in the absence of Shh signaling (Gli3 – 190 to Gli3 – 83) [85]. Overexpression of Gli3 represses Gli1 expression in cell culture [86,87]. Expression of Gli proteins is associated with activation of the Shh pathway [88]. Despite the finding that Gli1 is dispensable for normal mouse development [89], the loss of Gli2 or Gli3 function affects many processes of Shh-dependent organogenesis [90 – 94]. Gli3 is particularly important in vertebrate limb patterning. Mutations in GLI3 have been identified in Pallister – Hall syndrome (PHS), postaxial polydactyly type A (PAP-A), and Greig cephalopolysyndactyly syndrome (GCPS). PHS and PAP-A are dominant human syndromes associated with postaxial polydactyly [95]. Several genetic lesions identified in PHS [96,97] result in truncations that cause an increased level of GLI3 fragments with strong repressor activities [85]. Unlike PHS, GCPS in humans and extra-toes (Xt) and polydactyly Nagoya (Pdm) (Table 2) in mice are caused by mutations that abolish Gli3 function [98 – 100]; in these disorders the levels of Gli3 repressor activity are therefore decreased. In mice, this derepression leads to ectopic Shh expression in

TABLE 2

Mouse Genetic Disorders and Their Associated Genes

Mouse syndrome

Gene affecteda

Extra toes (xt)

Gli3

Polydactyly Nagoya (Pdm)

Gli3

Short ear (se)

Bmp5

Brachypodism (bp)

Gdf5

Osteogenesis imperfecta (oim)

Col1a2

Disproportionate micromelia (dmm)

Col2a1

Chondrodysplasia (cho)

Col11a1

Cartilage matrix deficiency (cmd)

Agc

a For gene notations we have followed the standard of using capital letters for human genes and a capital letter followed by lowercase letters for mouse genes

the anterior mesenchyme of the developing limb bud and, therefore, extra digits. The role of Ihh in growth plate function is discussed in the next section.

VI. BONE MORPHOGENETIC PROTEINS, HOMOLOGUES, AND THEIR ANTAGONISTS IN SKELETAL DEVELOPMENT Bone morphogenetic proteins (BMPs) are members of the transforming growth factor- (TGF-) superfamily, which regulate a number of embryonic events [101 – 103]. Major subdivisions within the superfamily include the TGF-s, BMPs (excluding BMP1), growth/differentiation factors 1 – 10 (called GDFs or CDMPs, a subclass of BMPs), inhibins, activins, Vg-related genes, nodal-related genes, and glial-derived neurotropic factor. BMP genes are vertebrate homologues of the Drosophila decapentaplegic, a target of the Hedgehog pathway. A large number of processes are known to require BMPs, including the proliferation of mesodermal cells [104]; growth and regression of the AER at the distal tip of limb buds [105,106]; formation of the antero – posterior limb axis and Hox gene expression [107,108]; initiation of chondrogenesis and cartilage differentiation [109 – 113]; myogenesis [110,114]; and interdigital apoptosis during limb development [115 – 117]. BMPs owe their name to the early discovery that demineralized bone fragments implanted subcutaneously or intramuscularly in animals induce bone formation [118]. A search for responsible factors resulted in the identification of the family of bone morphogenetic proteins [119]. To date, 15 BMPs have been identified and are further divided into groups according to their amino acid sequence

196

REGINATO, WANG, AND OLSEN

similarities [120]. Structural studies of BMPs reveal that they contain a mature domain that is cleaved proteolytically, allowing monomeric units to become dimers that are stabilized by disulfide bridges. Expression of different BMP genes in a cell can produce homodimers and heterodimers with variable degrees of glycosylation; these variations can influence the activities and effects of BMPs [121]. BMPs elicit their effects on target cells through binding to specific cell surface type I (BMPR-IA or BMPR-IB) and type II receptors (BMPR-II) [122]. The activated receptors (serine-threonine kinases) in turn activate SMAD proteins (homologues of mothers against decapentaplegic in Drosophila), which relay the signal from the cell cytoplasm to the nucleus [123]. Individual receptor molecules have low affinity for BMPs; however, with the formation of heterotetrameric complexes (two type I receptor molecules and two type II receptor molecules held together by the BMP ligand), high affinity is achieved [124 – 126]. Type II receptors are active kinases that function upstream of the type I receptors but can independently initiate cell signals [127]. On binding to BMP2, 4, or 7, the type II receptor kinase transphosphorylates the type I receptor [127,128].

FIGURE 7

Specific signals appear to be determined primarily by the type I receptor [129]. The type I receptor phosphorylates a serine residue in SMAD1, 5, and possibly 8. After phosphorylation, these SMADs associate with SMAD4 as heterooligomers and translocate to the nucleus where they accumulate rapidly. The SMAD1 signaling pathway appears to be regulated negatively by SMAD6, which inhibits SMAD1 signaling through binding to the type I receptor and competing with SMAD4 for binding to the activated SMAD1. This produces an inactive complex of SMAD1 and SMAD6 [130,131]. The C-terminal domain of SMAD1 is required for DNA binding and subsequent transcriptional activation [124,132]. In a similar manner, SMAD7 acts as an inhibitor of SMAD2 – SMAD4 complexes downstream of TGF- signaling [133]. Regulation of BMP effects depends on the distribution of specific signaling receptors, their functional states, and secreted protein antagonists (Fig. 7, see also color plate). BMP antagonists share the functional property of binding specifically to BMPs, thus preventing their interaction with the receptors. Antagonists include molecules such as noggin [134], DAN [135], Drm [136], chordin [137], and

Diagram illustrating how different BMP antagonists (chordin, noggin, follistatin, and members of the DAN family) can bind BMP dimers and thereby control the amount of BMPs available for binding to the signaling receptor complex (consisting of type I and type II receptors) on the cell surface. The antagonists are thought to be inactivated by proteolytic fragmentation. Following ligand binding, the type II receptor transphosphorylates (orange arrow) and activates the type I receptor; this initiates signaling to the nucleus. Modified from Cho and Blitz [123]. (See also color plate.)

197

CHAPTER 5 Developmental Biology of Bone

follistatin [138]. Noggin binds several BMPs with high affinities, with marked preference for BMP2 and BMP4 over BMP7 [139]. Noggin is expressed in chondrogenic condensations and appears to regulate the shape and size of the cartilaginous anlagen by controlling the effects of BMPs [140]. Gremlin, a member of the DAN family, appears to be important for establishing prechondrogenic condensations. It is highly conserved through evolution and can bind to and block BMP2, 4, and 7 [141]. At early stages of limb development, gremlin is expressed under the control of the AER and ZPA in a pattern that is complementary to these BMPs. The function of gremlin may be to confine chondrogenesis to the central core of the limb. Analysis of different mouse mutations suggests that there is considerable overlap between the functions of various BMPs. Targeted inactivation of the genes for Bmp2, 4, and 7 in mice leads to embryonic or perinatal lethality; these Bmps must therefore have important roles in early developmental processes. Bmp2 mutations result in malformations of the amnion, chorion, and the heart [142]. Bmp4 and BmprIA null mice have early lethal defects in mesoderm formation [143,144]. Alterations in kidney, ears, eyes, ribs, sternum, and skull and polydactyly are seen in Bmp7 knockout mice [145,146]. Homozygous mutant mice carrying a targeted deletion of Gdf11/Bmp11 exhibit anterior directed homeotic transformations throughout the axial skeleton and posterior displacement of the hindlimbs, suggesting that Gdf11 acts globally to specify positional identity along the anterior-posterior axis [147]. In addition to developmental abnormalities seen in knock – out mice, several skeletal disorders have been demonstrated to be caused by mutations in BMP genes. The short ear (se) and the brachypodism (bp) mutations in mice are caused by mutations in Bmp5 [148] and Gdf5 [149], respectively. Brachydactyly type C is the result of haploinsufficiency mutations in CDMP1 (cartilage-derived morphogenetic protein 1; also called GDF5) [150]. Homozygosity for CDMP1 mutations in humans causes acromesomelic dysplasia Hunter – Thompson type [151,152]. This disorder is characterized by short stature and shortening of forearms and lower legs, as well as the long bones of hands and feet. In the more severe Grebe type, affected individuals have a null allele of CDMP1 on one chromosome and an allele on the other chromosome with a dominant-negative mutation. The abnormal protein probably forms nonfunctional heterodimers with other BMPs, preventing their secretion [152]. Interference with BMP signaling causes the absence of proximal interphalangeal joints, fusion of wrist and ankle joints, and conductive deafness. This disorder, called proximal symphalangism (SYM1), has been shown to be caused by mutations in the gene for the BMP-binding molecule noggin [153]. Noggin mutations are also associated with multiple synostosis (SYNS1) syndromes, characterized by fusion in several joints (elbows, hips, intervertebral joints)

in addition to joints in the hands and feet [153]. Not surprisingly, in mice with inactivated noggin alleles, limb cartilage becomes hypoplastic and joints fail to form [140]. Based on studies of mice with inactivated BmprIB or Gdf5 alleles, it has been suggested that BmprIB regulates distal limb chondrogenesis through both Gdf5-dependent and -independent processes, and that, reciprocally, Gdf5 acts through both IB and other type I receptors [154,155].

VII. FIBROBLAST GROWTH FACTORS AND THEIR RECEPTORS IN SKELETAL DEVELOPMENT Fibroblast growth factors are essential for several aspects of bone development, not only in the ossification of cranial sutures and limb bud outgrowth, but also in growth plate function, i.e., longitudinal bone growth. Since their first discovery as a mitogenic activity in pituitary extracts [156,157], FGFs have been shown to support the proliferation of a variety of mesenchymal and epithelial cells, regulate cell migration, differentiation, and chemotaxis [158], and be involved in a variety of nonskeletal and skeletal developmental processes [159,160]. Proteins encoded by different FGF genes (Fgf1 – Fgf19) range in size from 155 to 268 amino acid residues. Each polypeptide contains a conserved region of approximately 120 amino acids that confers a common tertiary structure and the ability to bind heparin or heparan sulfate proteoglycans (HSPGs) [161]. FGFs are released into the extracellular matrix and bind to two classes of cell surface receptors: low-affinity and high-affinity receptors. The low-affinity FGF receptors are heparan sulfate-containing proteoglycans such as syndecan and glypican [162]. These proteoglycans may restrict the ability of FGFs to diffuse far from the cells that release them [163] and may facilitate signal transduction by oligomerizing and presenting the FGF ligands to high-affinity receptors (FGFRs) [164]. Thus, treating cells with heparin-degrading enzymes or with inhibitors of glycosaminoglycan sulfation inhibits the binding of and response to FGFs [165]. FGF receptors constitute a family of four transmembrane receptor tyrosine kinases encoded by four distinct genes (Fgfr1 – Fgfr4) [166]. These proteins consist of an extracellular ligand binding domain, a single transmembrane domain, and an intracellular tyrosine kinase domain (Fig. 8, see also color plate). The extracellular domain contains two or three immunoglobulin-like (Ig-like) domains, depending on alternate RNA splicing. One splicing event involves the N-terminal domain (domain I), leading to a form of the receptor with only two Ig-like domains. The ligand-binding properties of the alternately spliced receptor are similar to those of the full-length receptor, suggesting

198

FIGURE 8

Diagram showing the various domains in an FGFR3 receptor molecule and different mutations associated with four osteochondrodysplasias. The amino acid residues are numbered from the N terminus (left) of the molecule, with glycine at position 380 being localized in the transmembrane-spanning region of the molecule. Standard single letter names are used for amino acid residues; TER indicates termination codon. (See also color plate.)

that domain I is not critical for ligand binding. In contrast, the alternative splicing of exons encompassing domain III results in either IIIb or IIIc isoforms of the FGFR1, 2, and 3 receptors and affects their ligand specificity dramatically [167,168]. Interestingly, the IIIb isoform appears to be expressed in epithelial lineages, whereas the IIIc isoform is expressed primarily in mesenchymal cells [169 – 172]. After ligand binding, the receptor molecules dimerize, followed by autophosphorylation of tyrosine residues in the intracellular domain. Both homodimer and heterodimer interactions can occur between different FGF receptors within a cell [173]. The activated receptor in turn activates downstream signaling targets, including the RAS/MAP kinase and phosphatidylinositol pathways, ultimately influencing mitogenesis and differentiation [174 – 176]. Insights into the role of FGF receptor signaling in skeletal development have been gained from studies of human genetic diseases and mice that carry mutations in the receptors. Mutations in different domains of FGFR1, 2, and 3 cause a number of human craniosynostosis and dwarfism syndromes [177,178]. All the mutations are dominantly inherited, with the vast majority representing missense mutations that result in “gain of function.” The mutations cause excessive activation of the receptor, sometimes in a ligandindependent fashion, resulting in altered cell proliferation and/or differentiation [179].

REGINATO, WANG, AND OLSEN

Mutations in FGFR1 cause Pfeiffer syndrome, whereas mutations in FGFR2 can result in Apert syndrome, Crouzon’s syndrome, Jackson – Weiss syndrome, or Beare – Stevenson syndrome, depending on the location and nature of the mutation and genetic background effects. Mutations in FGFR3 are also associated with several distinct clinical syndromes [178]. Muenke syndrome (MS), or Muenke nonsyndromic coronal craniosynostosis, is characterized by incomplete penetrance and a wide phenotypic spectrum. MS patients may present with uni- or bilateral coronal craniosynostosis and midface hypoplasia, a downslanted palpebral fissure, or ptosis [180]. Crouzon syndrome with acanthosis nigricans is a rare condition seen in a subgroup of Crouzon patients with cutaneous manifestations of hyperpigmentation, hyperkeratosis, melanocytic nevi, and verrous hyperplasia. The craniofacial abnormalities are similar to those seen in patients with Crouzon syndrome [181,182]. Finally, different mutations in FGFR3 cause the dwarfism conditions hypochondroplasia (HCH), achondroplasia ACH), and thanatophoric dysplasia (TD) (Fig. 8) [183, 184]. Achondroplasia is among the most common and well known of the skeletal dysplasias and is characterized by short stature with rhizomelic shortening of the limbs, reduced elbow extension, genu varum, trident hand, exaggerated lumbar lordosis, frontal bossing, and midface deficiencies. Hypochondroplasia has similar limb and spinal features, but they are less severe. The face is normal, but the head circumference is larger than normal. Thanatophoric dysplasia, a neonatal lethal dwarfism, is clinically divided into two types: type 1 and type 2 thanatophoric dysplasia. In type 1, the long tubular bones are curved and the vertebral bodies are flat. In type 2, the long tubular bones are straight and the vertebral bodies are not as flat as in type 1, and craniosynostosis (cloverleaf skull) is present in most cases. Type 1 is the result of one of several mutations, both in extracellular and intracellular domains of FGFR3. Type 2 is caused by a mutation in the second kinase domain of FGFR3. The underlying defect in HCH, ACH, and TD is the disruption of normal, regulated proliferation and differentiation of chondrocytes in the epiphyseal growth plates of long bones, caused by gain-of-function mutations in FGFR3 [185 – 188]. A hypothesis for the phenotypic differences between the two types of thanatophoric dysplasia and achondroplasia is based on the dual developmental origin of the skull, in that it is formed by both endochondral and membranous ossification. Mutations that affect endochondral bone would cause deficiencies primarily of the cranial base and the nasal capsule, with a secondary effect on membranous bone, resulting in a shortened anterior cranial fossa and midface hypoplasia. This would lead to an abnormal fusion of sutures between membranous bones of the skull (frontal and parietal) during intrauterine development

CHAPTER 5 Developmental Biology of Bone

and sagittal synostosis. In this view, the severity of the cranial phenotype would depend on the degree of hypoplasia of the nasal capsule and the cranial base, and a spectrum of craniofacial deformities ranging from achondroplasia to thanatophoric dysplasia would be the consequence [184]. Although attractive, the hypothesis does not explain why the limbs and vertebrae are affected more severely in type 1 than type 2 thanatophoric dysplasia and why the skull is involved more severely in type 2 than in type 1. Part of the explanation may be that the common mutation in type 1, R248C, is known to completely activate the receptor in a ligand-independent manner, whereas the type 2 mutation, K650E, activates the receptor to a lesser degree and receptor activation is still partially ligand dependent [see 184]. In addition, the type 2 mutant FGFR3 receptor may dimerize with FGFR2 in the skull, and give rise to a more severe skull phenotype than the mutant homodimer of FGFR3 in type 1. Alternatively, craniosynostosis may be the result of heterodimerization of the mutant FGFR3 receptor with FGFR2 at the edges of membranous bones in the calvaria rather than cartilage hypoplasia [184]. Confirmation that constitutive receptor activation is involved in FGFR-based skeletal dysplasias has come from studies of knockout mice. Fgfr3 knockout mice display a phenotype with overgrowth of the vertebral column, long bones, and deafness due to an expansion of the zone of proliferating and hypertrophic chondrocytes in growth plate cartilages. In many ways, this phenotype is the opposite of the phenotypes of human achondroplasia caused by activating mutations in FGFR3 [189,190] and suggests a role for FGFR3 in restraining chondrocyte proliferation and maturation during endochondral bone formation. Transgenic mice that overexpress the Fgfr3 ligand Fgf2 exhibit skeletal malformations that include long bone shortening, and microcephaly, again suggesting that enhanced signaling through FGFR3 underlies FGFR3-based dysplasias [191]. Targeted disruption of the mouse Fgf2 gene leads to a phenotype with multiple abnormalities, including decreased bone formation [192].

VIII. CHONDROCYTE DIFFERENTIATION AND ENDOCHONDRAL BONE FORMATION The process of mesenchymal cell condensation, proliferation, and differentiation of chondrocytes is called chondrogenesis. Mesenchymal cells present in the anlagen initially express both type I collagen and a splice variant of type II collagen, type IIa, which is not chondrocyte specific [193]. Chondrocytes later switch on the specific isoform of type IIb collagen. Sox9, a member of the HMG box family of

199 transcription factor genes, is expressed in mesenchymal condensations before and during chondrogenesis [194,195]. The expression of Col2A1 is dependent on Sox9. Sox9 binds to the Col2A1 enhancer located in the first intron of the gene and activates transcription of the gene in vivo and in vitro [196 – 199]. Embryonic stem cells carrying inactivated Sox9 alleles fail to undergo chondrogenesis within mesenchymal condensations when mixed with wild-type cells [200], and chondrogenic markers such as Col2a1 and aggrecan are not expressed. Heterozygous loss-of-function mutations in SOX9 lead to campomelic dysplasia, a skeletal dysplasia characterized by abnormalities in all skeletal elements that are formed by endochondral ossification [201,202]. Sox9 also plays a role in the differentiation of Sertoli cells in male gonads; this provides an explanation for the high frequency of autosomal sex reversal in campomelic dysplasia. The products of other Sox genes, L-Sox5 and Sox6, form a complex with Sox9 during the activation of Col2a1 in vitro [203,204]. Other known molecules that affect chondrogenesis are Gdf5 and BmprIB. Gdf5 is expressed in the end regions of cartilage anlagen [149,205]. The disruption of the Gdf5 gene in mice leads to abnomalities in mesenchymal condensations, and mutations in Gdf5 result in brachypodism (see earlier discussion) [149]. BmprIB expression in chick limb buds precedes the future cartilage anlagen and its activity is involved in the initial steps of chondrogenesis (see earlier discussion) [206]. The replacement of cartilage by bone and bone marrow during endochondral bone formation starts by the hypertrophy of chondrocytes in the center of cartilage anlagen. The hypertrophic extracellular matrix mineralizes, blood vessels invade the cartilage, and a primary ossification center is formed (Fig. 9, see also color plate). This results in the deposition of trabecular bone within the cartilage, as the extracellular matrix of hypertrophic cartilage is degraded and hypertrophic chondrocytes undergo apoptosis. At the same time, bone is deposited in a sleeve around the primary ossification center. As primary ossification centers expand and secondary centers are established in the end regions of cartilage anlagen, cartilaginous growth plates remain as the centers of longitudinal growth (Fig. 2). Activities of the growth plate are regulated by several known cytokines. The parathyroid hormone-related peptide (PTHrP) is expressed in the periarticular region, whereas Ihh is expressed in prehypertrophic chondrocytes [207,208]. Chondrocyte hypertrophy is accelerated in PTHrP-deficient mice [209], indicating that PTHrP promotes proliferation and inhibits hypertrophy. Ihh has also been shown to stimulate chondrocyte proliferation, and Ihh signaling is required for PTHrP function [210]. The inactivation of Ihh genes in mice confirms that PTHrP signaling requires Ihh; however, Ihh plays multiple roles in PTHrPdependent and -independent pathways [211,212]. In

200

REGINATO, WANG, AND OLSEN

protein vascular endothelial growth factor (VEGF). This is crucial for blood vessel invasion into and resorption of hypertrophic cartilage. Anti-VEGF treatment using a soluble receptor in mice prevents capillary invasion into hypertrophic cartilage and impairs trabecular bone formation [219].

IX. OSTEOBLAST DIFFERENTIATION AND FUNCTION FIGURE 9

Immunohistochemical staining of blood vessels in the leg of a mouse embryo at day 17.5 of development. Staining with anti-CD31 antibodies shows vessels in the diaphyseal region of the developing tibia, where trabecular bone and bone marrow replace cartilage. Note that no vessels penetrate into the cartilage regions of the proximal and distal ends of the tibia. ep, epidermis. Courtesy of Dr. N. Fukai. (See also color plate.)

Ihh-null mice, PTHrP is undetectable and chondrocyte differentiation is abnormal. Furthermore, Cbfa1 is not expressed in the perichondrium and no endochondral bone formation occurs. This indicates that Ihh signaling is required for the normal differentiation of chondrocytes and osteoblasts during endochondral bone formation. Signaling by PTHrP is mediated by the PTH/PTHrP receptor [208]. The receptor is a G-protein-coupled receptor with seven membrane-spanning domains and is expressed in proliferating and prehypertrophic chondrocytes, as well as in osteoblasts. Targeted disruption of the receptor gene in mice leads to acceleration of the transition from proliferative to hypertrophic chondrocytes, resulting in premature ossification and short-limbed dwarfism [208]. An activating mutation in the receptor causes a decrease in hypertrophy and results in an autosomal dominant form of dwarfism, Jansen metaphyseal chondrodysplasia [213]. A loss-offunction mutation also causes dwarfism, Blomstrand chondrodysplasia [214]. Fgfr3 is expressed in nonproliferating chondrocytes, represses Ihh signaling, and inhibits Bmp4 expression in growth plates [215]. As described earlier, activating mutations in FGFR3 cause hypochondroplasia, achondroplasia, and thanatophoric dysplasia. Consistent with this, inactivation of Fgfr3 genes results in bone overgrowth in mice. Hypertrophic chondrocytes synthesize extracellular matrix proteins that are significantly different from those that are synthesized by small chondrocytes in the growth plates. Collagen X, encoded by Col10A1, is unique to the hypertrophic matrix and serves as an excellent marker for hypertrophic chondrocytes [216]. A mutation in Col10A1 causes Schmid metaphyseal chondrodysplasia in humans [217]; transgenic mice expressing a dominant-negative form of collagen X develop spondylometaphyseal dysplasia [218]. Hypertrophic chondrocytes also synthesize the angiogenic

Osteoblasts, originating from mesenchymal cells, are responsible for bone matrix deposition in both membranous and endochondral bone formation. A key regulator of osteoblast differentiation and function is Cbfa1 [4 – 6]. Cbfa1 is a member of the Runt domain-containing transcription factor family. Members of this family share a DNA-binding domain that is homologous to Drosophila Runt protein. Cbfa1 was initially identified as an enhancer binding protein for polyoma virus and murine leukemia virus and was given various names, such as PEA2, PEBP2a, CBF (core binding factor), and AML3 (acute myeloid leukemia 3 gene) [220]. It was described independently as Osf2 for its ability to bind the cis-acting elements of osteocalcin genes [221]. Cbfa1 is highly expressed in preosteoblasts and mature osteoblasts, and many bone matrix proteins, including bone sialoprotein and collagen type I, require Cbfa1 for their expression [4]. No bone is formed in Cbfa1-deficient mice, although the cartilage skeleton is patterned normally. In null mice, osteoblasts do not differentiate [5,6]. The activity of Cbfa1 is required not only for embryonic bone development, but also for postnatal bone growth. Bone formation ceases when Cbfa1 activity is blocked by expression of a dominant-negative mutant containing only the DNA-binding Runt domain [222]. Haploinsufficiency of Cbfa1 causes Cleidocranial dysplasia (CCD) in both mice and humans. The disorder is characterized by delayed closure of sutures, hypoplastic clavicles, hypoplastic pelvis, and short stature [223,224]. Cbfa1 is also expressed in chondrocytes in that it is first turned on in prechondrocytes, turned off in proliferating chondrocytes, and then turned on again in hypertrophic chondrocytes [225,226]. Blood vessel invasion into hypertrophic cartilage does not occur in Cbfa1-null mice.

X. CELL – MATRIX INTERACTIONS IN SKELETAL DEVELOPMENT Skeletal development depends critically on the synthesis of extracellular matrix components of cartilage and bone. This is illustrated by the large number of skeletal dysplasias that are secondary to mutations in matrix molecules. In

CHAPTER 5 Developmental Biology of Bone

bone, type I collagen, a heterotrimer of two 1(I) and one 2(I) chains, represents 90% of the total protein mass. Mutations in COL1A1, the gene coding for 1(I), as well as mutations in COL1A2, the gene coding for the 2(I) chain, cause osteogenesis imperfecta (OI) [227]. Osteogenesis imperfecta comprises a heterogeneous group of inherited disorders characterized by brittle bone disease leading to increased bone fracture, hearing loss, blue sclerae, and dentinogenesis imperfecta. Patients with OI type I, a mild phenotype, fracture their bones easily, but have normal bone healing and little or no deformity. These patients have a nonfunctioning or “null” collagen gene allele, resulting in a reduced collagen content in bone [228,229]. Patients with more severe phenotypes (lethal, type II; severe deformity, type III; and moderate deformity, type IV) generally have point mutations in the triple-helical domain of either the 1(I) or the 2(I) chains, with glycine replaced by another, bulkier amino acid residue [230,231]. Large multiexon deletions or duplications are not common, but have been reported to give rise to phenotypes that range from mild to lethal [see 232]. The exact mechanism by which these different mutations cause the phenotypes is largely unknown, but general principles are emerging. Mutations that affect the C-terminal propeptide prevent the incorporation of the mutant chain into trimeric molecules and result in the mildest phenotypes. Mutations that affect the Gly-X-Y repeat domain, and allow the mutant chain to participate in triple helix formation, act in a dominant-negative manner and result in a more severe phenotype. The phenotype is largely determined by the degree of secretion of mutant and normal collagen molecules into the extracellular space and their subsequent incorporation into collagen fibrils. If a mutant collagen molecule is incorporated into the matrix, it will affect the overall stability of the fibrils. In contrast, if the mutant collagen molecule is not incorporated into the matrix, the mature matrix will be affected less severely [see 232]. Several animal models for osteogenesis imperfecta exist. One such model is the Mov13 mouse, in which a retroviral insertion results in the transcriptional block of Col1a1 [233,234]. Transgenic mice have been generated with mild to lethal phenotypes [235 – 237]. Another model is a nonlethal recessive mutation (oim) in mice with features of osteopenia, fractures, and progressive skeletal deformities. In oim mice, a point mutation leads to alteration of the last 48 amino acid residues of the pro2(I) collagen chain; this prevents the association with pro1(I) collagen chains during assembly of the heterotrimeric molecule [238]. As a result, 1(I) homotrimers are formed in both skin and bone. This closely resembles the situation in patients with type III osteogenesis imperfecta [239]. Transgenic mice expressing a partially deleted Col1a1 gene show phenotypic variability and incomplete penetrance of spontaneous fractures even

201 on an inbred background [237]. This suggests that the phenotypic variability is an inherent stochastic feature of the expression of mutated collagens. Type II collagen, a homotrimer encoded by COL2A1, is the major structural component of cartilage. In addition, it is expressed in other structures, such as the vitreous of the eye, the nucleus pulposus of intervertebral discs, and the tectorial membrane of the inner ear. Mutations in type II collagen cause a spectrum of diseases known as “type II collagenopathies.” The severity ranges from developmental lethality (achondrogenesis type II, hypochondrogenesis), to moderately severe dwarfism (spondyloepiphyseal dysplasia congenita, Kniest dysplasia), and to normal stature with premature osteoarthritis (Stickler and Marshall syndromes) [240]. In the case of lethal mutations that lead to an absence of type II collagen in embryonic cartilage, some type of cartilage develops with collagen I substituting for collagen II, chondrocytes differentiate, and bones are formed [241]. Mutations that cause a moderately severe phenotype (spondylopepiphyseal dysplasia, Kniest dysplasia) generally lead to a reduced secretion and content of type II collagen in cartilage [242 – 244]. Patients with these mutations have disproportionate dwarfism with rhizomelic shortening of the limbs, while the hands and feet are minimally affected. The mildest phenotypes (Stickler syndrome with premature osteoarthritis) are caused by premature stop codons resulting in a null allele [244,245]. Several mouse models of the human type II collagenopathies have been generated [246,247]. A naturally occurring mouse mutant, disproportionate micromelia (dmm), has been found to be caused by a three nucleotide deletion in Col2a1, leading to the substitution of Lys-Thr with Asn in the C-propeptide of the type II procollagen molecule [248]. In addition to type II collagen, a number of quantitatively minor collagens exist in cartilage. They include types IX, X, XI, and XII collagen. Type IX collagen is a member of the FACIT group of extracellular matrix proteins, which also includes types XII, XIV, XVI, and XIX. Type IX collagen is a heterotrimeric nonfibrillar collagen composed of three different  chains, 1(IX), 2(IX), and 3(IX), that are coexpressed with type II collagen in cartilage and other cartilage-like tissues. Type IX molecules are localized on the surface of the fibrils where they get cross-linked to residues within the N- and C-telopeptides of type II collagen; their function is probably to stabilize the fibril network [see 249]. Mutations in collagen IX cause autosomal-dominant multiple epiphyseal dysplasia (MED). This disorder includes the mild Ribbing type, the more severe Fairbank type, and some unclassified MED types. Splice-site mutations in COL9A2, causing skipping of exon 3 in 2(IX) transcripts, result in a mild phenotype with short stature and early-onset osteoarthritis [250,251]. Further analysis of MED has shown similar skipping of exon 3 in the a3(IX) gene transcript, caused by mutations in the intron 3

202 splice – donor site of COL9A3 [252 – 254]. Transgenic mice overexpressing a dominant-negative truncated form of the 1(IX) chain show mild chondrodysplasia and progressive osteoarthritis [255]. Mice homozygous for a null mutation in Col9a1 exhibit normal skeletal development, but develop progressive osteoarthritis-like changes in articular cartilage after birth [256]. Type X collagen, a member of the short chain collagen family, is a homotrimeric molecule that is expressed by hypertrophic chondrocytes during endochondral bone formation (Fig. 3). Transgenic mice, expressing a collagen X gene with a large in-frame deletion in the central triple helical domain, as well as mice with Col10a1-null alleles, develop variable spondylometaepiphyseal chondrodysplasia after birth [218,257]. Histological analysis shows a reduction in the height of the hypertrophic zone, severe reduction in the size and number of bone trabeculae below the growth plate cartilage, and lymphopenia in the bone marrow, the thymus, and the circulation. In humans, mutations in COL10A1 have been shown to cause the autosomaldominant disorder Schmid metaphyseal chondrodysplasia, characterized by bowing of the legs, growth retardation of the extremities, and coxa vara [217]. All the mutations identified are clustered in the C-terminal nontriple helical NC1 domain of the molecule. They are likely to prevent the formation of trimeric molecules, and the mutant chains are consequently not secreted by chondrocytes. This suggests that the phenotype seen in Schmid metaphyseal chondrodysplasia is caused by haploinsufficiency, although dominant-negative mechanisms cannot be ruled out for at least some of the mutations [258 – 260]. Type XI copolymerizes with type II collagen and appears to regulate the diameter of cartilage fibrils. It is a heterotrimer composed of products of three different genes: COL11A1, COL11A2, and COL2A1. Mutations in COL11A1 have been shown to result in Marshall or Stickler syndrome, characterized by high myopia, vitreoretinal degeneration, cleft palate, midfacial hypoplasia, premature osteoarthritis, and hearing defects [261,262]. Extensive genotype – phenotype comparisons of patients with Stickler, Stickler-like, or Marshall syndromes have suggested that null-allele mutations in COL2A1 cause the typical Stickler phenotype (vitreoretinal degeneration common and hearing loss less common), whereas splicing mutations in the COL11A1 gene are responsible for the Marshall syndrome (hearing loss common and vitreoretinal degeneration less common). Patients with glycine substitutions or small in-frame deletions in the COL11A1 gene (dominantnegative mutations) may have a mixed phenotype characteristic of both syndromes [263]. In contrast to mutations in COL2A1, mutations in COL11A2 are associated with a Stickler-like syndrome without eye involvement [264]. The explanation for this is the presence of a unique form of type XI collagen in the vitreous, which contains an 2(V)

REGINATO, WANG, AND OLSEN

chain instead of the 2(XI) chain of type XI collagen in cartilage [265]. Defects of mice with a recessively inherited chondrodysplasia (cho) provide further insights into the role of collagen type XI in skeletal development. In cho mice, a single nucleotide deletion creates a premature stop codon in the N-terminal region of 1(XI) collagen [266]. Heterozygous animals are relatively unaffected; however, with age they develop osteoarthritis. Homozygous animals die at birth with short limbs, a short snout, and a cleft palate. The vertebral column is short and the thoracic cage is small. Growth plate cartilage is disorganized with soluble proteoglycan aggregates and thicker than normal collagen fibrils. The presence of thick fibrils provides further evidence of the role of type XI collagen in regulating fibril diameters. The proteolytic processing of type XI collagen differs from that of type II collagen in that the N-propeptide domains are not cleaved after secretion. During embryogenesis, these peptide domains control fibril diameter by localizing on the fibril surface and limiting the addition of type II collagen molecules to the surface by steric hindrance [267]. cho mice have thick fibrils leading to the formation of a large-pore network of fewer and thicker fibrils instead of a meshwork of thin fibrils found in wild-type cartilage. Changes in the fibril meshwork cause the proteoglycan aggregates to be loosely entrapped within the mutant matrix and they are more soluble than in wild-type cartilage. Cartilage and bone also contain numerous noncollagenous components, of which the large cartilage proteoglycan or aggrecan is the most studied. Aggrecan contains a core protein that is modified by substitution with chondroitin and keratan sulfate chains and oligosacharides. Aggrecan binds to hyaluronic acid and forms large complexes, which are stabilized by link proteins. The proteoglycan aggregates bind water and give cartilage its ability to withstand compressive forces [268]. Within the growth plate, aggrecan and link proteins are coexpressed with type II collagen. Although aggrecan would appear to be a prime candidate for chondrodysplasias, to date no mutations in the aggrecan gene have been identified in humans. However, mutations have been identified in mice and chicken. Cartilage matrixdeficient mice (cmd) have an autosomal recessive mutation causing short limbs, tail, and snout and a cleft palate. A 7 bp deletion in the aggrecan gene has been demonstrated to result in a truncated molecule and a major reduction in the amount of aggrecan in cartilage matrix [269,270]. In chicken, the nanomelia phenotype is similar to that of the murine cmd and results from a premature stop codon in the avian aggrecan gene [271]. Link protein (LP) stabilizes the aggregates of aggrecan and hyaluronan and is important for the normal organization of hypertrophic chondrocytes in growth plate cartilage [272]. Targeted mutations in the LP gene give rise to defects in cartilage in homozygous mice and delayed bone formation with short limbs and abnormal

CHAPTER 5 Developmental Biology of Bone

craniofacial development. The animals have small epiphyses, flared metaphyses of the long bones, and flattened vertebrae, characteristic of spondyloepiphyseal dysplasia. There is a reduction in aggrecan deposition in the hypertrophic zone in growth plates and a decrease in the number of prehypertrophic and hypertrophic chondrocytes [272]. Biglycan (Bgn) is an extracellular matrix proteoglycan that is enriched in bone and other skeletal connective tissues. In vitro studies indicate that Bgn may function in connective tissue metabolism by binding to collagen fibrils and TGF-. Bgn-deficient mice show a reduced rate of growth and decreased bone mass and display an osteoporotic-like phenotype [273]. Cartilage oligomeric matrix protein (COMP) is another noncollagenous component of cartilage, composed of a pentamer of five identical subunits. COMP belongs to the thrombospondin family of matrix molecules and is localized in the pericellular, territorial matrix in cartilage. The protein contains several repeat domains, including eight calcium-binding calmodulin-like repeats. Mutations in COMP have been described in various types of multiple epiphyseal dysplasia [274 – 277]. MED is thus a heterogeneous disorder, caused by mutations in either collagen IX or COMP. Perlecan is a large heparan sulfate proteoglycan with a wide tissue distribution and multiple potential functions [278]. The perlecan core protein consists of several distinct protein modules organized in five domains, resembling pearls on a string. Domain I contains a glycosaminoglycan side chain that binds basic fibroblast growth factor FGF2 and has been shown to promote mitogenic and angiogenic activities. Other domains are capable of binding several other small and large molecules, including FGF7, fibronectin, heparin, laminin, PDGF-BB, and cell surface integrins [see 279]. Although perlecan plays a major role as a component of basement membranes, it also has an important function in cartilage. French et al. [280] demonstrated that 10T1/2 multipotential mouse embryonic fibroblasts aggregate into a dense cartilagenous nodule when cultured on perlecan, suggesting that perlecan promotes chondrogenic differentiation. Perlecan homozygous knockout mice develop severe osteochondral defects characterized by dwarfism, cleft palate, short limbs, and a short and abnormally curved vertebral column [281]. The phenotype resembles that of Col2a1-deficient [282] and dmm mice [248]. The perlecan-null bones show minor changes in epiphyseal cartilage but severe abnormalities in the growth plates. The proliferating zone is disorganized and the hypertrophic zone shows signs of increased extracellular matrix synthesis and is frequently separated from the proliferating zone [281]. It is possible that perlecan protects the extracellular matrix of cartilage by inactivating matrix proteases or masking/protecting proteins against proteolytic degradation. Also, through its ability to bind and sequester

203 Fgfs, perlecan may modulate Fgfr3 signaling pathways and thus influence both chondrocyte proliferation and differentiation. Finally, given recent evidence for the importance of heparan sulfate proteoglycans in transducing Hedgehog signals, perlecan may even be important for Ihh signaling [283]. Matrilins or cartilage matrix (Crtm) proteins are members of a novel family of extracellular matrix proteins consisting of von Willebrand factor A-like (vWFA-like) domains, epidermal growth factor (EGF)-like domains, and a coiled -helical motif. Four members of the family have been identified. Matrilin1 and matrilin3 are expressed mainly in hyaline cartilage, whereas matrilin2 and matrilin4 are expressed in a wide variety of extracellular matrices [284,285]. The roles of matrilins in cartilage and skeletal development are largely unknown. The cross-linking of matrilin1 to aggrecan core protein [286], as well as its association with type II collagen-containing fibrils [287], suggests a possible role as a structural connector. Mice carrying a null mutation in the Crtm gene coding for matrilin1 demonstrate a normal phenotype with no detectable abnormalities of matrix organization [288]. These findings suggest that matrilin1 is structurally not a critical, limiting component in cartilage structure and that other matrilins may have functionally redundant roles.

XI. SUMMARY AND PERSPECTIVES Recent advances in developmental biology, skeletal cell biology, and molecular genetics have increased our understanding of bone development significantly. We are now beginning to understand the various genetic pathways that control the patterning of mesenchymal condensations. Critical genes and regulatory cytokines involved in both chondrocyte and osteoblast differentiation have been identified. Genes and mechanisms that regulate joint formation and bone growth are becoming known. The roles of extracellular matrix proteins, not only as structural components, but also as modulators of signal transduction, are coming into sharper focus. However, much needs to be accomplished, including a better understanding of endochondral ossification and the function of cartilage in bone remodeling and the factors that control cortical and trabecular bone thickness. What mechanisms trigger the formation of primary versus secondary ossification centers in the cartilage anlagen of long bones? What distinguishes articular cartilage from epiphyseal growth plate cartilage? How is cartilage maintained throughout development and postnatal life? What determines the specific sizes and shapes of bones? These are challenging questions, but given the current interest in skeletal biology, the existence of a large number of inherited bone diseases in humans, new methodologies for genetic manipulations in various animal models, and the

204

REGINATO, WANG, AND OLSEN

rapid advances of the human and mouse genome projects, it seems likely that several of these questions will be answered relatively soon. It is hoped that the topics highlighted in this chapter will stimulate research aimed at providing answers that will benefit patients with developmental and metabolic bone diseases, including osteoporosis.

References 1. D. M. Noden, The role of the neural crest in patterning of avian cranial skeletal, connective, and muscle tissues. Dev. Biol. 96, 144 – 165 (1983). 2. D. M. Noden, Cell movements and control of patterned tissue assembly during craniofacial development. J. Craniofac. Genet. Dev. Biol. 11, 192 – 213 (1991). 3. M. Bronner-Fraser, Neural crest cell formation and migration in the developing embryo. FASEB J. 8, 699 – 706 (1994). 4. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 5. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 6. F. Otto, A. P. Thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, CBFA1, a candidate gene for the cleidocranial dysplasia syndrome, is essential for osteoblast formation and bone development. Cell 89, 765 – 771 (1997). 7. J. A. Rivera-Perez, M. Wakamiya, and R. R. Behringer, Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development. Development 126, 3811 – 2821 (1999). 8. D. Acampora, S. Mazan, Y. Lallemand, V. Avantaggiato, M. Maury, A. Simeone, and P. Brulet, Forebrain and midbrain regions are deleted in Otx2-/- mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121, 3279 – 3290 (1995). 9. I. Matsuo, S. Kuratani, C. Kimura, N. Takeda, and S. Aizawa, Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9, 2646 – 2658 (1995). 10. T. Kume, K. Y. Deng, V. Winfrey, D. B. Gould, M. A. Walter, and B. L. Hogan, The forkhead/winged helix gene Mf1 is disrupted in the pleiotropic mouse mutation congenital hydrocephalus. Cell 93, 985 – 996 (1998). 11. M. Qiu, A. Bulfone, I. Ghattas, J. J. Meneses, L. Christensen, P. T. Sharpe, R. Presley, R. A. Pedersen, and J. L. Rubenstein, Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: Mutations of Dlx-1, Dlx-2, and Dlx-1 and -2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches. Dev. Biol. 185, 165 – 184 (1997). 12. D. Acampora, G. R. Merlo, L. Paleari, B. Zerega, M. P. Postiglione, S. Mantero, E. Bober, O. Barbieri, A. Simeone, and G. Levi, Craniofacial, vestibular and bone defects in mice lacking the Distal-lessrelated gene Dlx5. Development 126, 3795 – 3809 (1999). 13. E. Dahl, H. Koseki, and R. Balling, Pax genes and organogenesis. Bioessays 19, 755 – 765 (1997). 14. A. P. Read and V. E. Newton, Waardenburg syndrome. J. Med. Genet. 34, 656 – 665 (1997). 15. M. Tassabehji, A. P. Read, V. E. Newton, R. Harris, R. Balling, P. Gruss, and T. Strachan, Waardenburg’s syndrome patients have

16.

17.

18.

19.

20. 21.

22.

23.

24.

25.

26.

27. 28.

29.

30.

31.

32.

33.

mutations in the human homologue of the Pax-3 paired box gene. Nature 355, 635 – 636 (1992). V. El, Ghouzzi, M. Le Merrer, F. Perrin-Schmitt, E. Lajeunie, P. Benit, D. Renier, P. Bourgeois, A. L. Bolcato-Bellemin, A. Munnich, and J. Bonaventure, Mutations of the TWIST gene in the Saethre-Chotzen syndrome. Nature Genet. 15, 42 – 46 (1997). T. D. Howard, W. A. Paznekas, E. D. Green, L. C. Chiang, N. Ma, R. I. Ortiz de, Luna, C. Garcia Delgado, M. Gonzalez-Ramos, A. D. Kline, and E. W. Jabs, Mutations in TWIST, a basic helix-loop-helix transcription factor, in Saethre-Chotzen syndrome. Nature Genet. 15, 36 – 41 (1997). E. W. Jabs, U. Muller, X. Li, L. Ma, W. Luo, I. S. Haworth, I. Klisak, R. Sparkes, M. L. Warman, J. B. Mulliken, et al., A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 75, 443 – 450 (1993). U. Hehr and M. Muenke, Craniosynostosis syndromes: From genes to premature fusion of skull bones. Mol. Genet. Metab. 68, 139 – 151 (1999). P. P. Tam and P. A. Trainor, Specification and segmentation of the paraxial mesoderm. Anat. Embryol. (Berlin) 189, 275 – 305. B. Christ, C. Schmidt, R. Huang, J. Wilting, and B. Brand-Saberi, Segmentation of the vertebrate body. Anat. Embryol. 197, 1 – 8 (1998). K. Horikawa, G. Radice, M. Takeichi, and O. Chisaka, Adhesive subdivisions intrinsic to the epithelial somites. Dev. Biol. 215, 182 – 189 (1999). P. J. Swiatek, C. E. Lindsell, F. F. del Amo, G. Weinmaster, and T. Gridley Notch1 is essential for postimplantation development in mice. Genes Dev. 8, 707 – 719 (1994). R. A. Conlon, A. G. Reaume, and J. Rossant, Notch1 is required for the coordinate segmentation of somites. Development 121, 1533 – 1545 (1995). B. Bettenhausen, M. Hrabe de Angelis, D. Simon, J. L. Guenet, and A. Gossler, Transient and restricted expression during mouse embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. Development 121, 2407 – 2418 (1995). M. Hrabe de Angelis, J. McIntyre, II, and A. Gossler, Maintenance of somite borders in mice requires the Delta homologue DII1. Nature 386, 717 – 721 (1997). N. Zhang and T. Gridley, Defects in somite formation in lunatic fringe-deficient mice. Nature 394, 374 – 377 (1998). Y. A. Evrard, Y. Lun, A. Aulehla, L. Gan, and R. L. Johnson, Lunatic fringe is an essential mediator of somite segmentation and patterning. Nature 394, 377 – 381 (1998). Y. Saga, N. Hata, H. Koseki, and M. M. Taketo, Mesp2: A novel mouse gene expressed in the presegmented mesoderm and essential for segmentation initiation. Genes Dev. 11, 1827 – 1839 (1997). D. B. Donoviel, A. K. Hadjantonakis, M. Ikeda, H. Zheng, P. S. Hyslop, and A. Bernstein, Mice lacking both presenilin genes exhibit early embryonic patterning defects. Genes Dev. 13, 2801 – 2810 (1999). P. C. Wong, H. Zheng, H. Chen, M. W. Becher, D. J. Sirinathsinghji, M. E. Trumbauer, H. Y. Chen, D. L. Price, L. H. Van der Ploeg, and S. S. Sisodia, Presenilin 1 is required for Notch1 and DII1 expression in the paraxial mesoderm. Nature 387, 288 – 292 (1997). K. Kusumi, E. S. Sun, A. W. Kerrebrock, R. T. Bronson, D. C. Chi, M. S. Bulotsky, J. B. Spencer, B. W. Birren, W. N. Frankel, and E. S. Lander, The mouse pudgy mutation disrupts Delta homologue Dll3 and initiation of early somite boundaries. Nature Genet. 19, 274 – 278 (1998). M. P. Bulman, K. Kusumi, T. M. Frayling, C. McKeown, C. Garrett, E. S. Lander, R. Krumlauf, A. T. Hattersley, S. Ellard, and P. D. Turnpenny, Mutations in the human Delta homologue, DLL3, cause axial skeletal defects in spondylocostal dysostosis. Nature Genet. 24, 438 – 441 (2000).

205

CHAPTER 5 Developmental Biology of Bone 34. S. Mundlos and B. R. Olsen, Defects in skeletal morphogenesis. In “Defects in Skeletal Morphogenesis” (P. Royce and B. Steinmann, eds.). Wiley, New York, in press. 35. C. Marcelle, S. Ahlgren, and M. Bronner-Fraser, In vivo regulation of somite differentiation and proliferation by Sonic Hedgehog. Dev. Biol. 214, 277 – 287 (1999). 36. C. Chiang, Y. Litingtung, E. Lee, K. E. Young, J. L. Corden, H. Westphal, and P. A. Beachy, Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407 – 413 (1996). 37. L. C. Murtaugh, J. H. Chyung, and A. B. Lassar, Sonic hedgehog promotes somitic chondrogenesis by altering the cellular response to BMP signaling. Genes Dev. 13, 225 – 237 (1999). 38. J. Wallin, J. Wilting, H. Koseki, R. Fritsch, B. Christ, and R. Balling, The role of Pax-1 in axial skeleton development. Development 120, 1109 – 1121 (1994). 39. B. Wilm, E. Dahl, H. Peters, R. Balling, and K. Imai, Targeted disruption of Pax1 defines its null phenotype and proves haploinsufficiency. Proc. Natl. Acad. Sci. USA 95, 8692 – 8697 (1998). 40. H. Peters, B. Wilm, N. Sakai, K. Imai, R. Maas, and R. Balling, Pax1 and Pax9 synergistically regulate vertebral column development. Development 126, 5399 – 5408 (1999). 41. R. Krumlauf, Hox genes in vertebrate development. Cell 78, 191 – 201 (1994). 42. A. C. Burke, C. E. Nelson, B. A. Morgan, and C. Tabin, Hox genes and the evolution of vertebrate axial morphology. Development 121, 333 – 346 (1995). 43. B. Favier, and P. Dolle, Developmental functions of mammalian Hox genes. Mol. Hum. Reprod. 3, 115 – 131 (1997). 44. F. Chen, J. Greer, and M. R. Capecchi, Analysis of Hoxa7/Hoxb7 mutants suggests periodicity in the generation of the different sets of vertebrae. Mech. Dev. 77, 49 – 57 (1998). 45. J. Zakany, M. Gerard, B. Favier, and D. Duboule, Deletion of a HoxD enhancer induces transcriptional heterochrony leading to transposition of the sacrum. EMBO J. 16, 4393 – 4402 (1997). 46. R. L. Johnson and C. J. Tabin, Molecular models for vertebrate limb development. Cell 90, 979 – 990 (1997). 47. C. Tickle and G. Eichele, Vertebrate limb development. Annu. Rev. Cell Biol. 10, 121 – 152 (1994). 48. J. K. Ng, K. Tamura, D. Buscher, and J. C. Izpisua-Belmonte, Molecular and cellular basis of pattern formation during vertebrate limb development. Curr. Top. Dev. Biol. 41, 37 – 66 (1999). 49. B. A. Parr and A. P. McMahon, Dorsalizing signal Wnt-7a required for normal polarity of D-V and A-P axes of mouse limb. Nature 374, 350 – 353 (1995). 50. C. A. Loomis, R. A. Kimmel, C. X. Tong, J. Michaud, and A. L. Joyner, Analysis of the genetic pathway leading to formation of ectopic apical ectodermal ridges in mouse Engrailed-1 mutant limbs. Development 125, 1137 – 1148 (1998). 51. R. D. Riddle, R. L. Johnson, E. Laufer, and C. Tabin, Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75, 1401 – 1416 (1993). 52. X. Sun, M. Lewandoski, E. N. Meyers, H. Y. Liu, R. E. Maxson, and G. R. Martin, Conditinal inactivation of Fgf4 reveals complexity of signalling during limb bud development. Nature Genet. 25, 83 – 86 (2000). 53. A. Zuniga, A. P. Haramis, A. P. McMahon, and R. Zeller, Signal relay by BMP antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature 401, 598 – 602 (1999). 54. J. Charite, D. G. McFadden, and E. N. Olson, The bHLH transcription factor dHAND controls Sonic hedgehog expression and establishment of the zone of polarizing activity during limb development. Development 127, 2461 – 2470 (2000). 55. M. Fernandez, Teran, M. E. Piedra, I. S. Kathiriya, D. Srivastava, J. C. Rodriguez Rey, and M. A. Ros, Role of dHAND in the anterior-posterior

56.

57.

58. 59.

60.

61.

62. 63.

64.

65.

66.

67.

68.

69. 70.

71.

72.

73.

polarization of the limb bud: Implications for the Sonic hedgehog pathway. Development 127, 2133 –2142 (2000). J. Capdevila, T. Tsukui, C. Rodriquez Esteban, V. Zappavigna, and J. C. Izpisua Belmonte, Control of vertebrate limb outgrowth by the proximal factor Meis2 and distal antagonism of BMPs by Gremlin. Mol. Cell 4, 839 – 849 (1999). N. Mercader, E. Leonardo, N. Azpiazu, A. Serrano, G. Morata, C. Martinez, and M. Torres, Conserved regulation of proximodistal limb axis development by Meis1/Hth. Nature 402, 425 – 429 (1999). H. Simon, T-box genes and the formation of vertebrate forelimb- and hindlimb specific pattern. Cell Tissue Res. 296, 57 – 66 (1999). D. L. Chapman, N. Garvey, S. Hancock, M. Alexiou, S. I. Agulnik, J. J. Gibson Brown, J. Cebra Thomas, R. J. Bollag, L. M. Silver, and V. E. Papaioannou, Expression of the T-box family genes, Tbx1-Tbx5, during early mouse development. Dev. Dyn. 206, 379 – 390 (1996). J. K. Takeuchi, K. Koshiba Takeuchi, K. Matsumoto, A. VogelHopker, M. Naitoh-Matsuo, K. Ogura, N. Takahashi, K. Yasuda, and T. Ogura, Tbx5 and Tbx4 genes determine the wing/leg identity of limb buds. Nature 398, 810 – 814 (1999). C. Rodriguez-Esteban, T. Tsukui, S. Yonei, J. Magallon, K. Tamura, and J. C. Izpisua Belmonte, The T-box genes Tbx4 and Tbx5 regulate limb outgrowth and identity. Nature 398, 814 – 818 (1999). M. Logan and C. J. Tabin, Role of Pitx1 upstream of Tbx4 in specification of hindlimb identity. Science 283, 1736 – 1739 (1999). C. Lanctot, A. Moreau, M. Chamberland, M. L. Tremblay, and J. Drouin, Hindlimb patterning and mandible development require the Ptx1 gene. Development 126, 1805 – 1810 (1999). D. P. Szeto, C. Rodriguez-Esteban, A. K. Ryan, S. M. O’Connell, F. Liu, C. Kioussi, A. S. Gleiberman, Izpisua-J. C. Belmonte, and M. G. Rosenfeld, Role of the Bicoid-related homeodomain factor Pitx1 in specifying hindlimb morphogenesis and pituitary development. Genes Dev. 13, 484 – 494 (1999). P. Dolle, J. C. Izpisua-Belmonte, E. Boncinelli, and D. Duboule, The Hox-4.8 gene is localized at the 5 extremity of the Hox-4. Mech. Dev. 36, 3 – 13 (1991). C. E. Nelson, B. A. Morgan, A. C. Burke, E. Laufer, E. DiMambro, L. C. Murtaugh, E. Gonzales, L. Tessarollo, L. F. Parada, and C. Tabin, Analysis of Hox gene expression in the chick limb bud. Development 122, 1449 – 1466 (1996). Y. Muragaki, S. Mundlos, J. Upton, and B. R. Olsen, Altered growth and branching patterns in synpolydactyly caused by mutations in HOXD13. Science 272, 548 – 551 (1996). F. R. Goodman, S. Mundlos, Y. Muragaki, D. Donnai, M. L. Giovannucci-Uzielli, E. Lapi, F. Majewski, J. McGaughran, C. McKeown, W. Reardon, J. Upton, R. M. Winter, B. R. Olsen, and P. J. Scambler, Synpolydactyly phenotypes correlate with size of expansions in HOXD13 polyalanine tract. Proc. Natl. Acad. Sci. USA 94, 7458 – 7463 (1997). J. Zakany and D. Duboule, Hox genes in digit development and evolution. Cell Tissue Res. 296, 19 – 25 (1999). K. R. Johnson, H. O. Sweet, L. R. Donahue, P. Ward-Bailey, R. T. Bronson, and M. T. Davisson, A new spontaneous mouse mutation of Hoxd13 with a polyalanine expansion and phenotype similar to human synpolydactyly. Hum. Mol. Genet. 7, 1033 – 1038 (1998). V. Knezevic, R. De Santo, K. Schughart, U. Huffstadt, C. Chiang, K. A. Mahon, and S. Mackem, Hoxd-12 differentially affects preaxial and postaxial chondrogenic branches in the limb and regulates Sonic hedgehog in a positive feedback loop. Development 124, 4523 – 4536 (1997). J. J. Lee, D. P. von Kessler, S. Parks, and P. A. Beachy, Secretion and localized transcription suggest a role in positional signaling for products of the segmentation gene hedgehog. Cell 71, 33 – 50 (1992). Y. Echelard, D. J. Epstein, B. St-Jacques, L. Shen, J. Mohler, J. A. McMahon, and A. P. McMahon, Sonic hedgehog, a member of a

206

74.

75.

76. 77. 78.

79.

80.

81. 82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

REGINATO, WANG, AND OLSEN family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75, 1417 – 1430 (1993). J. J. Lee, S. C. Ekker, D. P. von Kessler, J. A. Porter, B. I. Sun, and P. A. Beachy, Autoproteolysis in hedgehog protein biogenesis. Science 266, 1528 – 1537 (1994). J. A. Porter, K. E. Young, and P. A. Beachy, Cholesterol modification of hedgehog signaling proteins in animal development. Science 274, 255 – 259 (1996). P. W. Ingham, Transducing Hedgehog: The story so far. EMBO J. 17, 3505 – 3511 (1998). Y. Chen and G. Struhl, Dual roles for patched in sequestering and transducing Hedgehog. Cell 87, 553 – 563 (1996). D. M. Stone, M. Hynes, M. Armanini, T. A. Swanson, Q. Gu, R. L. Johnson, M. P. Scott, D. Pennica, A. Goddard, H. Phillips, M. Noll, J. E. Hooper, F. de Sauvage, and A. Rosenthal, The tumoursuppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature 384, 129 – 134 (1996). M. van den Heuvel and P. W. Ingham, Smoothened encodes a receptor-like serpentine protein required for hedgehog signalling. Nature 382, 547 – 551 (1996). J. Alcedo, M. Ayzenzon, T. Von Ohlen, M. Noll, and J. E. Hooper, The Drosophila smoothened gene encodes a seven-pass membrane protein, a putative receptor for the hedgehog signal. Cell 86, 221 – 232 (1996). P. Aza-Blanc and T. B. Kornberg, Ci: A complex transducer of the hedgehog signal. Trends Genet. 15, 458 – 462 (1999). C. H. Chen, D. P. von Kessler, W. Park, B. Wang, Y. Ma, and P. A. Beachy (1999). Nuclear trafficking of Cubitus interruptus in the transcriptional regulation of Hedgehog target gene expression. Cell 98, 305 – 316. (1999) Y. Chen, N. Gallaher, R. H. Goodman, and S. M. Smolik, Protein kinase A directly regulates the activity and proteolysis of cubitus interruptus. Proc. Natl. Acad. Sci. USA 95, 2349 – 2354 (1998). J. M. Ruppert, K. W. Kinzler, A. J. Wong, S. H. Bigner, F. T. Kao, M. L. Law, H. N. Seuanez, S. J. O’Brien, and B. Vogelstein, The GLIKrüppel family of human genes. Mol. Cell Biol. 8, 3104 – 3113 (1988). B. Wang, J. F. Fallon, and P. A. Beachy, Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate limb. Cell 100, 423 – 434 (2000). P. Dai, H. Akimaru, Y. Tanaka, T. Maekawa, M. Nakafuku, and S. Ishii, Sonic Hedgehog-induced activation of the Gli1 promoter is mediated by GLI3. J. Biol. Chem. 274, 8143 – 8152 (1999). A. Ruiz i Altaba, Combinatorial Gli gene function in floor plate and neuronal inductions by Sonic hedgehog. Development 125, 2203 – 2212 (1998). C. C. Hui, D. Slusarski, K. A. Platt, R. Holmgren, and A. L. Joyner, Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectodermand mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev. Biol. 162, 402 – 413 (1994). H. L. Park, C. Bai, K. A. Platt, M. P. Matise, A. Beeghly, C. Hui, M. Nakashima, and A. L. Joyner, Mouse gli1 mutants are viable but have defects in SHH signaling in combination with a gli2 mutation. Development 127, 1593 – 1605 (2000). J. Motoyama, J. Liu, R. Mo, Q. Ding, M. Post, and C. C. Hui, Essential function of Gli2 and Gli3 in the formation of lung, trachea and oesophagus Nature Genet. 20, 54 – 57 (1998). R. Mo, A. M. Freer, D. L. Zinyk, M. A. Crackower, J. Michaud, H. H. Heng, K. W. Chik, X. M. Shi, L. C. Tsui, S. H. Cheng, A. L. Joyner, and C. Hui, Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 124, 113 – 123 (1997). Z. Hardcastle, R. Mo, C. C. Hui, and P. T. Sharpe, The Shh signalling pathway in tooth development: Defects in Gli2 and Gli3 mutants. Development 125, 2803 – 2811 (1998).

93. Q. Ding, J. Motoyama, S. Gasca, R. Mo, H. Sasaki, J. Rossant, and C. C. Hui, Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 125, 2533 – 2543 (1998). 94. M. P. Matise, D. J. Epstein, H. L. Park, K. A. Platt, and A. L. Joyner, Gli2 is required for induction of floor plate and adjacent cells, but not most ventral neurons in the mouse central nervous system. Development 125, 2759 – 2770 (1998). 95. L. G. Biesecker, Strike three for GLI3 [published erratum appears in Nature Genet. 18, 88] (1998). Nature Genet. 17, 259 – 260 (1997). 96. S. Kang, J. M. Graham, Jr., A. H. Olney, and L. G. Biesecker, GLI3 frameshift mutations cause autosomal dominant Pallister-Hall syndrome. Nature Genet. 15, 266 – 268 (1997). 97. U. Radhakrishna, A. Wild, K. H. Grzeschik, and S. E. Antonarakis, Mutation in GLI3 in postaxial polydactyly type A. Nature Genet. 17, 269 – 271 (1997). 98. A. Vortkamp, M. Gessler, and K. H. Grzeschik, GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 352, 539 – 540 (1991). 99. C. C. Hui and A. L. Joyner, A mouse model of Greig cephalopolysyndactyly syndrome: The extra-toes J mutation contains an intragenic deletion of the Gli3 gene. Nature Genet. 3, 241 – 246. 100. H. Thien and U. Ruther, The mouse mutation Pdn (Polydactyly Nagoya) is caused by the integration of a retrotransposon into the Gli3 gene. Mamm. Genome 10, 205 – 209 (1999). 101. B. L. Hogan, Bone morphogenetic proteins: Multifunctional regulators of vertebrate development. Genes Dev. 10, 1580 – 1594 (1996). 102. D. M. Kingsley, The TGF-beta superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8, 133 – 146 (1994). 103. D. M. Kingsley, What do BMPs do in mammals? Clues from the mouse short-ear mutation. Trends Genet. 10, 16 – 21 (1994). 104. L. Niswander and G. R. Martin, FGF-4 and BMP-2 have opposite effects on limb growth. Nature 361, 68 – 71 (1993). 105. Y. Ganan, D. Macias, R. D. Basco, R. Merino, and J. M. Hurle, Morphological diversity of the avian foot is related with the pattern of msx gene expression in the developing autopod. Dev. Biol. 196, 33 – 41 (1998). 106. S. Pizette and L. Niswander, BMPs negatively regulate structure and function of the limb apical ectodermal ridge. Development 126, 883 – 894 (1999). 107. Y. Kawakami, T. Ishikawa, M. Shimabara, N. Tanda, M. EnomotoIwamoto, M. Iwamoto, T. Kuwana, A. Ueki, S. Noji, and T. Nohno, BMP signaling during bone pattern determination in the developing limb. Development 122, 3557 – 3566 (1996). 108. D. M. Duprez, K. Kostakopoulou, P. H. Francis-West, C. Tickle, and P. M. Brickell, Activation of Fgf-4 and HoxD gene expression by BMP-2 expressing cells in the developing chick limb. Development 122, 1821 – 1828 (1996). 109. D. Duprez, E. J. Bell, M. K. Richardson, C. W. Archer, L. Wolpert, P. M. Brickell, and P. H. Francis-West, Overexpression of BMP-2 and BMP-4 alters the size and shape of developing skeletal elements in the chick limb. Mech. Dev. 57, 145 – 157 (1996). 110. D. M. Duprez, M. Coltey, H. Amthor, P. M. Brickell, and C. Tickle, Bone morphogenetic protein-2 (BMP-2) inhibits muscle development and promotes cartilage formation in chick limb bud cultures. Dev. Biol. 174, 448 – 452 (1996). 111. D. Macias, Y. Ganan, T. K. Sampath, M. E. Piedra, M. A. Ros, and J. M. Hurle, Role of BMP-2 and OP-1 (BMP-7) in programmed cell death and skeletogenesis during chick limb development. Development 124, 1109 – 1117 (1997). 112. R. Merino, Y. Ganan, D. Macias, A. N. Economides, K. T. Sampath, and J. M. Hurle, Morphogenesis of digits in the avian limb is controlled by FGFs, TGFs, and noggin through BMP signaling. Dev. Biol. 200, 35 – 45 (1998).

CHAPTER 5 Developmental Biology of Bone 113. M. Enomoto-Iwamoto, M. Iwamoto, Y. Mukudai, Y. Kawakami, T. Nohno, Y. Higuchi, S. Takemoto, H. Ohuchi, S. Noji, and K. Kurisu (1998). Bone morphogenetic protein signaling is required for maintenance of differentiated phenotype, control of proliferation, and hypertrophy in chondrocytes. J. Cell Biol. 140, 409 – 418 (1998). 114. H. Amthor, B. Christ, M. Weil, and K. Patel, The importance of timing differentiation during limb muscle development. Curr. Biol. 8, 642 – 652 (1998). 115. H. Zou and L. Niswander, Requirement for BMP signaling in interdigital apoptosis and scale formation. Science 272, 738 – 741 (1996). 116. Y. Yokouchi, J. Sakiyama, T. Kameda, H. Iba, A. Suzuki, N. Ueno, and A. Kuroiwa, BMP-2/-4 mediate programmed cell death in chicken limb buds. Development 122, 3725 – 3734 (1996). 117. Y. Ganan, D. Macias, M. Duterque-Coquillaud, M. A. Ros, and J. M. Hurle, Role of TGFs and BMPs as signals controlling the position of the digits and the areas of interdigital cell death in the developing chick limb autopod. Development 122, 2349 – 2357 (1996). 118. M. R. Urist, Bone: Formation by autoinduction. Science 150, 893 – 899 (1965). 119. J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. W. Kriz, R. M. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528 – 1534 (1988). 120. A. H. Reddi, Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nature Biotechnol. 16, 247 – 252 (1998). 121. J. M. Wozney, Bone morphogenetic proteins and their expression. In “Bone Morphogenetic Proteins and Their expression” (M. Noda, ed.), pp. 131 – 165 Academic Press, San Diego, (1993). 122. S. Cheifetz, BMP receptors in limb and tooth formation. Crit. Rev. Oral Biol. Med. 10, 182 – 198 (1999). 123. K. W. Cho and I. L. Blitz, BMPs, Smads and metalloproteases: Extracellular and intracellular modes of negative regulation. Curr. Opin. Genet. Dev. 8, 443 – 449 (1998). 124. F. Liu, A. Hata, J. C. Baker, J. Doody, J. Carcamo, R. M. Harland, and J. Massague, A human Mad protein acting as a BMP-regulated transcriptional activator. Nature 381, 620 – 623 (1996). 125. T. Nohno, T. Ishikawa, T. Saito, K. Hosokawa, S. Noji, D. H. Wolsing, and J. S. Rosenbaum, Identification of a human type II receptor for bone morphogenetic protein-4 that forms differential heteromeric complexes with bone morphogenetic protein type I receptors. J. Biol. Chem. 270, 22522 – 22526 (1995). 126. B. L. Rosenzweig, T. Imamura, T. Okadome, G. N. Cox, H. Yamashita, P. ten Dijke, C. H. Heldin, and K. Miyazono, Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc. Natl. Acad. Sci. USA 92, 7632 – 7636 (1995). 127. J. L. Wrana, L. Attisano, R. Wieser, F. Ventura, and J. Massague, Mechanism of activation of the TGF- receptor. Nature 370, 341 – 347 (1994). 128. J. Massague, TGF signaling: Receptors, transducers, and Mad proteins. Cell 85, 947 – 950 (1996). 129. J. Carcamo, F. M. Weis, F. Ventura, R. Wieser, J. L. Wrana, L. Attisano, and J. Massague, Type I receptors specify growth-inhibitory and transcriptional responses to transforming growth factor  and activin. Mol. Cell Biol. 14, 3810 – 3821 (1994). 130. A. Hata, G. Lagna, J. Massague, and A. Hemmati-Brivanlou, Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev. 12, 186 – 197 (1998). 131. T. Imamura, M. Takase, A. Nishihara, E. Oeda, J. Hanai, M. Kawabata, and K. Miyazono, Smad6 inhibits signalling by the TGF- superfamily Nature 389, 622 – 626 (1997). 132. J. Kim, K. Johnson, H. J. Chen, S. Carroll, and A. Laughon, Drosophila Mad binds to DNA and directly mediates activation of vestigial by Decapentaplegic. Nature 388, 304 – 308 (1997).

207 133. H. Hayashi, S. Abdollah, Y. Qiu, J. Cai, Y. Y. Xu, B. W. Grinnell, M. A. Richardson, J. N. Topper, M. A. Gimbrone, Jr., J. L. Wrana, and D. Falb, The MAD-related protein Smad7 associates with the TGF receptor and functions as an antagonist of TGF signaling. Cell 89, 1165 – 1173 (1997). 134. W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992). 135. E. Stanley, C. Biben, S. Kotecha, L. Fabri, S. Tajbakhsh, C. C. Wang, T. Hatzistavrou, B. Roberts, C. Drinkwater, M. Lah, M. Buckingham, D. Hilton, A. Nash, T. Mohun, and R. P. Harvey, DAN is a secreted glycoprotein related to Xenopus cerberus. Mech. Dev. 77, 173 – 184 (1998). 136. J. J. Pearce, G. Penny, and J. Rossant, A mouse cerberus/Dan-related gene family. Dev. Biol. 209, 98 – 110 (1999). 137. Y. Sasai, B. Lu, H. Steinbeisser, D. Geissert, L. K. Gont, and E. M. De Robertis, Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 79, 779 – 790 (1994). 138. F. S. Esch, S. Shimasaki, M. Mercado, K. Cooksey, N. Ling, S. Ying, N. Ueno, and R. Guillemin. Structural characterization of follistatin: A novel follicle-stimulating hormone release-inhibiting polypeptide from the gonad. Mol. Endocrinol. 1, 849 – 855 (1987). 139. L. B. Zimmerman, J. M. De Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). 140. L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). 141. D. R. Hsu, A. N. Economides, X. Wang, P. M. Eimon, and R. M. Harland, The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol. Cell 1, 673 – 683 (1998). 142. H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 143. G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 144. Y. Mishina, A. Suzuki, N. Ueno, and R. R. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). 145. A. T. Dudley, K. M. Lyons, and E. J. Robertson, A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev. 9, 2795 – 2807 (1995). 146. G. Luo, C. Hofmann, A. L. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). 147. A. C. McPherron, A. M. Lawler, and S. J. Lee, Regulation of anterior/posterior patterning of the axial skeleton by growth/differentiation factor 11. Nature Genet. 22, 260 – 264 (1999). 148. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 149. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF -superfamily. Nature 368, 639 – 643 (1994). 150. A. Polinkovsky, N. H. Robin, J. T. Thomas, M. Irons, A. Lynn, F. R. Goodman, W. Reardon, S. G. Kant, H. G. Brunner, I. van der Burgt, D. Chitayat, J. McGaughran, D. Donnai, F. P. Luyten, and M. L. Warman, Mutations in CDMP1 cause autosomal dominant brachydactyly type C. Nature Genet. 17, 18 – 19 (1997).

208 151. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGFbeta superfamily member. Nature Genet. 12, 315 – 317 (1996). 152. J. T. Thomas, M. W. Kilpatrick, K. Lin, L. Erlacher, P. Lembessis, T. Costa, P. Tsipouras, and F. P. Luyten, Disruption of human limb morphogenesis by a dominant negative mutation in CDMP1. Nature Genet. 17, 58 – 64 (1997). 153. Y. Gong, D. Krakow, J. Marcelino, D. Wilkin, D. Chitayat, R. BabulHirji, L. Hudgins, C. W. Cremers, F. P. Cremers, H. G. Brunner, K. Reinker, D. L. Rimoin, D. H. Cohn, F. R. Goodman, W. Reardon, M. Patton, C. A. Francomano, and M. L. Warman, Heterozygous mutations in the gene encoding noggin affect human joint morphogenesis. Nature Genet. 21, 302 – 304 (1999). 154. S. T. Baur, J. J. Mai, and S. M. Dymecki, Combinatorial signaling through BMP receptor IB and GDF5: Shaping of the distal mouse limb and the genetics of distal limb diversity. Development 127, 605 – 619 (2000). 155. S. E. Yi, A. Daluiski, R. Pederson, V. Rosen, and K. M. Lyons, The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621 – 630 (2000). 156. H. A. Armelin, Pituitary extracts and steroid hormones in the control of 3T3 cell growth. Proc. Natl. Acad. Sci. USA 70, 2702 – 2706 (1973). 157. D. Gospodarowicz, Localisation of a fibroblast growth factor and its effect alone and with hydrocortisone on 3T3 cell growth. Nature 249, 123 – 127 (1974). 158. M. Klagsbrun, The fibroblast growth factor family: Structural and biological properties. Prog Growth Factor Res 1, 207 – 235 (1989). 159. G. R. Martin, The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571 – 1586 (1998). 160. M. C. Naski and D. M. Ornitz, FGF signaling in skeletal development. Front. Biosci. 3, 781 – 794 (1998). 161. S. Faham, R. E. Hileman, J. R. Fromm, R. J. Linhardt, and D. C. Rees, Heparin structure and interactions with basic fibroblast growth factor. Science 271, 1116 – 1120 (1996). 162. N. Perrimon and M. Bernfield, Specificities of heparan sulphate proteoglycans in developmental processes. Nature 404, 725 – 728 (2000). 163. C. Basilico and D. Moscatelli, The FGF family of growth factors and oncogenes. Adv. Cancer Res. 59, 115 – 165 (1992). 164. I. J. Mason, The ins and outs of fibroblast growth factors. Cell 78, 547 – 552 (1994). 165. B. B. Olwin and A. Rapraeger, Repression of myogenic differentiation by aFGF, bFGF, and K-FGF is dependent on cellular heparan sulfate. J. Cell Biol. 118, 631 – 639 (1992). 166. D. E. Johnson and L. T. Williams, Structural and functional diversity in the FGF receptor multigene family. Adv. Cancer Res. 60, 1 – 41 (1993). 167. T. Miki, D. P. Bottaro, T. P. Fleming, C. L. Smith, W. H. Burgess, A. M. Chan, and S. A. Aaronson, Determination of ligand-binding specificity by alternative splicing: Two distinct growth factor receptors encoded by a single gene. Proc. Natl. Acad. Sci. USA 89, 246 – 250 (1992). 168. A. T. Chellaiah, D. G. McEwen, S. Werner, J. Xu, and D. M. Ornitz, Fibroblast growth factor receptor (FGFR) 3: Alternative splicing in immunoglobulin-like domain III creates a receptor highly specific for acidic FGF/FGF-1. J. Biol. Chem. 269, 11620 – 11627 (1994). 169. A. Avivi, A. Yayon, and D. Givol, A novel form of FGF receptor-3 using an alternative exon in the immunoglobulin domain III. FEBS Lett. 330, 249 – 252 (1993). 170. G. Yan, Y. Fukabori, G. McBride, S. Nikolaropolous, and W. L. McKeehan, Exon switching and activation of stromal and embryonic fibroblast growth factor (FGF)-FGF receptor genes in prostate epithelial cells accompany stromal independence and malignancy. Mol. Cell Biol. 13, 4513 – 4522 (1993).

REGINATO, WANG, AND OLSEN 171. E. Gilbert, F. Del Gatto, P. Champion-Arnaud, M. C. Gesnel, and R. Breathnach, Control of BEK and K-SAM splice sites in alternative splicing of the fibroblast growth factor receptor 2 pre-mRNA. Mol. Cell Biol. 13, 5461 – 5468 (1993). 172. E. Scotet and E. Houssaint, The choice between alternative IIIb and IIIc exons of the FGFR-3 gene is not strictly tissue-specific. Biochim. Biophys. Acta 1264, 238 – 242 (1995). 173. F. Bellot, G. Crumley, J. M. Kaplow, J. Schlessinger, M. Jaye, and C. A. Dionne, Ligand-induced transphosphorylation between different FGF receptors. EMBO J. 10, 2849 – 2854 (1991). 174. G. Szebenyi, and J. F. Fallon, Fibroblast growth factors as multifunctional signaling factors. Int. Rev. Cytol. 185, 45 – 106 (1999). 175. H. Kouhara, Y. R. Hadari, T. Spivak-Kroizman, J. Schilling, D. Bar-Sagi, I. Lax, and J. Schlessinger, A lipid-anchored Grb2-binding protein that links FGF-receptor activation to the Ras/MAPK signaling pathway. Cell 89, 693 – 702 (1997). 176. G. Sa and T. Das, Basic fibroblast growth factor stimulates cytosolic phospholipase A2, phospholipase C-gamma1 and phospholipase D through distinguishable signaling mechanisms. Mol. Cell. Biochem. 198, 19 – 30 (1999). 177. D. Burke, D. Wilkes, T. L. Blundell, and S. Malcolm, Fibroblast growth factor receptors: lessons from the genes. Trends Biochem. Sci. 23, 59 – 62 (1998). 178. M. R. Passos-Bueno, W. R. Wilcox, E. W. Jabs, A. L. Sertie, L. G. Alonso, and H. Kitoh, Clinical spectrum of fibroblast growth factor receptor mutations. Hum. Mutat. 14, 115 – 125 (1999). 179. M. K. Webster and D. J. Donoghue, FGFR activation in skeletal disorders: Too much of a good thing. Trends Genet. 13, 178 – 182 (1997). 180. M. Muenke, K. W. Gripp, D. M. McDonald-McGinn, K. Gaudenz, L. A. Whitaker, S. P. Bartlett, R. I. Markowitz, N. H. Robin, N. Nwokoro, J. J. Mulvihill, H. W. Losken, J. B. Mulliken, A. E. Guttmacher, R. S. Wilroy, L. A. Clarke, G. Hollway, L. C. Ades, E. A. Haan, J. C. Mulley, M. M. Cohen, Jr., G. A. Bellus, C. A. Francomano, D. M. Moloney, S. A. Wall, and A. O. Wilkie, A unique point mutation in the fibroblast growth factor receptor 3 gene (FGFR3) defines a new craniosynostosis syndrome. Am. J. Hum. Genet. 60, 555 – 564 (1997). 181. D. Wilkes, P. Rutland, L. J. Pulleyn, W. Reardon, C. Moss, J. P. Ellis, R. M. Winter, and S. Malcolm, A recurrent mutation, ala391glu, in the transmembrane region of FGFR3 causes Crouzon syndrome and acanthosis nigricans. J. Med. Genet. 33, 744 – 748 (1996). 182. G. A. Meyers, S. J. Orlow, I. R. Munro, K. A. Przylepa, and E. W. Jabs, Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutation in Crouzon syndrome with acanthosis nigricans. Nature Genet. 11, 462 – 464 (1995). 183. J. Bonaventure, F. Rousseau, L. Legeai-Mallet, M. Le Merrer, A. Munnich, and P. Maroteaux, Common mutations in the fibroblast growth factor receptor 3 (FGFR 3) gene account for achondroplasia, hypochondroplasia, and thanatophoric dwarfism. Am. J. Med. Genet. 63, 148 – 154 (1996). 184. M. M. Cohen, Jr., Achondroplasia, hypochondroplasia and thanatophoric dysplasia: Clinically related skeletal dysplasias that are also related at the molecular level. Int. J. Oral Maxillofac. Surg. 27, 451 – 455 (1998). 185. R. Shiang, L. M. Thompson, Y. Z. Zhu, D. M. Church, T. J. Fielder, M. Bocian, S. T. Winokur, and J. J. Wasmuth, Mutations in the transmembrane domain of FGFR3 cause the most common genetic form of dwarfism, achondroplasia. Cell 78, 335 – 342 (1994). 186. P. L. Tavormina, R. Shiang, L. M. Thompson, Y. Z. Zhu, D. J. Wilkin, R. S. Lachman, W. R. Wilcox, D. L. Rimoin, D. H. Cohn, and J. J. Wasmuth, Thanatophoric dysplasia (types I and II) caused by distinct mutations in fibroblast growth factor receptor 3. Nature Genet. 9, 321 – 328 (1995).

CHAPTER 5 Developmental Biology of Bone 187. F. Rousseau, P. Saugier, M. Le Merrer, A. Munnich, A. L. Delezoide, P. Maroteaux, J. Bonaventure, F. Narcy, and M. Sanak, Stop codon FGFR3 mutations in thanatophoric dwarfism type 1. Nature Genet. 10, 11 – 12 (1995). 188. G. A. Bellus, I. McIntosh, E. A. Smith, A. S. Aylsworth, I. Kaitila, W. A. Horton, G. A. Greenhaw, J. T. Hecht, and C. A. Francomano, A recurrent mutation in the tyrosine kinase domain of fibroblast growth factor receptor 3 causes hypochondroplasia. Nature Genet. 10, 357 – 359 (1995). 189. C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). 190. J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M. Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). 191. J. D. Coffin, R. Z. Florkiewicz, J. Neumann, T. Mort-Hopkins, G. W. Dorn, II, P. Lightfoot, R. German, P. N. Howles, A. Kier, and B. A. O’Toole, Abnormal bone growth and selective translational regulation in basic fibroblast growth factor (FGF-2) transgenic mice. Mol. Biol. Cell 6, 1861 – 1873 (1995). 192. A. Montero, Y. Okada, M. Tomita, M. Ito, H. Tsurukami, T. Nakamura, T. Doetschman, J. D. Coffin, and M. M. Hurley, Disruption of the fibroblast growth factor-2 gene results in decreased bone mass and bone formation. J. Clin. Invest. 105, 1085 – 1093 (2000). 193. V. C. Lui, L. J. Ng, J. Nicholls, P. P. Tam, and K. S. Cheah, Tissuespecific and differential expression of alternatively spliced alpha 1(II) collagen mRNAs in early human embryos. Dev. Dyn. 203, 198 – 211 (1995). 194. L. J. Ng, S. Wheatley, G. E. Muscat, J. Conway-Campbell, J. Bowles, E. Wright, D. M. Bell, P. P. Tam, K. S. Cheah, and P. Koopman, SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev. Biol. 183, 108 – 121 (1997). 195. E. Wright, M. R. Hargrave, J. Christiansen, L. Cooper, J. Kun, T. Evans, U. Gangadharan, A. Greenfield, and P. Koopman, The Sryrelated gene Sox9 is expressed during chondrogenesis in mouse embryos. Nature Genet. 9, 15 – 20 (1995). 196. V. Lefebvre and B. de Crombrugghe, Toward understanding Sox9 function in chondrocyte differentiation. Matrix Biol. 16, 529 – 540 (1998). 197. D. M. Bell, K. K. Leung, S. C. Wheatley, L. J. Ng, S. Zhou, K. W. Ling, M. H. Sham, P. Koopman, P. P. Tam, and K. S. Cheah, Sox9 directly regulates the type-II collagen gene. Nature Genet. 16, 174 – 178 (1997). 198. K. K. Leung, L. J. Ng, K. K. Ho, P. P. Tam, and K. S. Cheah, Different cis-regulatory DNA elements mediate developmental stage- and tissue-specific expression of the human COL2A1 gene in transgenic mice. J. Cell Biol. 141, 1291 – 1300 (1998). 199. V. Lefebvre, W. Huang, V. R. Harley, P. N. Goodfellow, and B. de Crombrugghe, Sox9 is a potent activator of the chondrocyte-specific enhancer of the pro alpha1(II) collagen gene. Mol. Cell Biol. 17, 2336 – 2346 (1997). 200. W. Bi, J. M. Deng, Z. Zhang, R. R. Behringer, and B. de Crombrugghe, Sox9 is required for cartilage formation. Nat. Genet. 22, 85 – 89 (1999). 201. T. Wagner, J. Wirth, J. Meyer, B. Zabel, M. Held, J. Zimmer, J. Pasantes, F. Bricarelli, J. Keutel, E. Hustert, U. Wolf, N. Tommerup, W. Schempp, and G. Scherer, Autosomal sex reversal and campomelic dysplasia are caused by mutations in and around the SRY-related gene SOX9. Cell 79, 1111 – 1120 (1994). 202. J. W. Foster, M. A. Dominguez-Steglich, S. Guioli, G. Kowk, P. A. Weller, M. Stevanovic, J. Weissenbach, S. Mansour, I. D. Young, and P. N. Goodfellow, Campomelic dysplasia and autosomal sex reversal caused by mutations in an SRY-related gene. Nature 372, 525 – 530 (1994).

209 203. G. Zhou, V. Lefebvre, Z. Zhang, H. Eberspaecher, and B. de Crombrugghe, Three high mobility group-like sequences within a 48-base pair enhancer of the Col2a1 gene are required for cartilage-specific expression in vivo. J. Biol. Chem. 273, 14989 – 14997 (1998) 204. V. Lefebvre, P. Li, and B. de Crombrugghe, A new long form of Sox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis and cooperatively activate the type II collagen gene. EMBO J. 17, 5718 – 5733 (1998). 205. E. E. Storm and D. M. Kingsley, GDF5 coordinates bone and joint formation during digit development. Dev. Biol. 209, 11 – 27 (1999). 206. H. Zou, K. M. Choe, Y. Lu, J. Massague, and L. Niswander, BMP signaling and vertebrate limb development. Cold Spring Harb. Symp. Quant. Biol. 62, 269 – 272 (1997). 207. M. J. Bitgood and A. P. McMahon, Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172, 126 – 138 (1995). 208. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Juppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273, 663 – 666 (1996). 209. A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene. Genes Dev. 8, 277 – 289 (1994). 210. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 211. St-B. Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). 212. S. J. Karp, E. Schipani, B. St-Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian Hedgehog coordinates endochondral bone gorwth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 213. E. Schipani, K. Kruse, and H. Jüppner, A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 268, 98 – 100 (1995). 214. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 215. M. C. Naski, J. S. Colvin, J. D. Coffin, and D. M. Ornitz, Repression of hedgehog signaling and BMP4 expression in growth plate cartilage by fibroblast growth factor receptor 3. Development 125, 4977 – 4988 (1998). 216. K. Iyama, Y. Ninomiya, B. R. Olsen, T. F. Linsenmayer, R. L. Trelstad, and M. Hayashi, Spatiotemporal pattern of type X collagen gene expression and collagen deposition in embryonic chick vertebrae undergoing endochondral ossification. Anat. Rec. 229, 462 – 472 (1991). 217. M. L. Warman, M. H. Abbott, S. S. Apte, T. Hefferon, I. McIntosh, D. H. Cohn, J. T. Hecht , B. R. Olsen, and C. A. Francomano, A type X collagen mutation causes Schmid metaphyseal chondrodysplasia. Nature Genet. 5, 79 – 82 (1993). 218. O. Jacenko, P. A. LuValle, and B. R. Olsen, Spondylometaphyseal dysplasia in mice carrying a dominant negative mutation in a matrix protein specific for cartilage-to-bone transition. Nature 365, 56 – 61 (1993). 219. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999).

210 220. N. A. Speck and T. Stacy, A new transcription factor family associated with human leukemias. Crit. Rev. Eukaryot. Gene Expr. 5, 337 – 364 (1995). 221. P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). 222. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). 223. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, B. U. Zabel, R. Mertelsmann, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 224. L. F. Huang, N. Fukai, P. B. Selby, B. R. Olsen, and S. Mundlos, Mouse clavicular development: Analysis of wild-type and cleidocranial dysplasia mutant mice. Dev. Dyn. 210, 33 – 40 (1997). 225. I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). 226. M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). 227. M. S. Kocher and F. Shapiro, Osteogenesis imperfecta J. Am. Acad. Orthop. Surg. 6, 225 – 236 (1998). 228. M. C. Willing, S. P. Deschenes, R. L. Slayton, and E. J. Roberts, Premature chain termination is a unifying mechanism for COL1A1 null alleles in osteogenesis imperfecta type I cell strains. Am. J. Hum. Genet. 59, 799 – 809 (1996). 229. S. B. M. Deak, A. F. Nicholls, F. M. Pope, and D. J. Prockop, The molecular defect in a nonlethal variant of osteogenesis imperfecta. J. Biol. Chem. 258, 15192 – 15197 (1983). 230. J. F. Bateman, I. Moeller, M. Hannagan, D. Chan, and W. G. Cole, Characterization of three osteogenesis imperfecta collagen a1(I) glycine to serine mutations demonstrating a position-dependent gradient of phenotypic severity. J. Biochem. 288, 131 – 135 (1992). 231. Q. Wang, B. M. Orrison, and J. C. Marini, Two additional cases of osteogenesis imperfecta with substitutions for glycine in the alpha 2(I) collagen chain: A regional model relating mutation location with phenotype. J. Biol. Chem. 268, 25162 – 25167 (1993). 232. S. Mundlos and B. R. Olsen, Heritable diseases of the skeleton. II. Molecular insights into skeletal development-matrix components and their homeostasis. FASEB J. 11, 227 – 233 (1997). 233. A. Schnieke, K. Harbers, and R. Jaenisch, Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha 1(I) collagen gene. Nature 304, 315 – 320 (1983). 234. J. Bonadio, T. L. Saunders, E. Tsai, S. A. Goldstein, J. MorrisWiman, L. Brinkley, D. F. Dolan, R. A. Altschuler, J. E. Hawkins, Jr., J. F. Bateman, et al., Transgenic mouse model of the mild dominant form of osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 87, 7145 – 7149 (1990). 235. A. Stacey, J. Bateman, T. Choi, T. Mascara, W. Cole, and R. Jaenisch, Perinatal lethal osteogenesis imperfecta in transgenic mice bearing an engineered mutant pro-1(I) collagen gene. Nature 332, 131 – 135 (1988). 236. J. S. Khillan, A. S. Olsen, S. Kontusaari, B. Sokolov, and D. J. Prockop, Transgenic mice that express a mini-gene version of the human gene for type I procollagen (COL1A1) develop a phenotype resembling a lethal form of osteogenesis imperfecta. J. Biol. Chem. 266, 23373 – 23379 (1991). 237. R. Pereira, K. Halford, B. P. Sokolov, J. S. Khillan, and D. J. Prockop, Phenotypic variability and incomplete penetrance of spontaneous

REGINATO, WANG, AND OLSEN

238.

239.

240.

241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

251.

252.

fractures in an inbred strain of transgenic mice expressing a mutated collagen gene (COL1A1). J. Clin. Invest. 93, 1765 – 1769 (1994). S. D. Chipman, H. O. Sweet, D. J. J. McBride, M. T. Davisson, S. C. J. Marks, A. R. Shuldiner, R. J. Wenstrup, D. W. Rowe, and J. R. Shapiro, Defective pro alpha 2(I) collagen synthesis in a recessive mutation in mice: A model of human osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993). T. Pihlajaniemi, L. A. Dickson, F. M. Pope, V. R. Korhonen, A. Nicholls, D. J. Prockop, and J. C. Myers, Osteogenesis imperfecta: Cloning of a pro2(I) Collagen gene with a frameshift mutations. J. Biol. Chem. 259, 12941 – 13944 (1984). J. Spranger, A. Winterpacht, and B. Zabel, The type II collagenopathies: A spectrum of chondrodysplasias. Eur. J. Pediatr 153, 56 – 65 (1994). D. Chan, W. G. Cole, C. W. Chow, S. Mundlos, and J. F. Bateman, A COL2A1 mutation in achondrogenesis type II results in the replacement of type II collagen by type I and III collagens in cartilage. J. Biol. Chem. 270, 1747 – 1753 (1995). D. Chan and W. G. Cole, Low basal transcription of genes for tissuespecific collagens by fibroblasts and lymphoblastoid cells: Application to the characterization of a glycine 997 to serine substitution in alpha 1(II) collagen chains of a patient with spondyloepiphyseal dysplasia. J. Biol. Chem. 266, 12487 – 12494 (1991). D. Chan, T. K. Taylor, and W. G. Cole, Characterization of an arginine 789 to cysteine substitution in alpha 1 (II) collagen chains of a patient with spondyloepiphyseal dysplasia. J. Biol. Chem. 268, 15238 – 15245 (1993). A. Winterpacht, M. Hilbert, U. Schwarze, S. Mundlos, J. Spranger, and U. Zabel, Kniest and Stickler dysplasia phenotypes caused by collagen type II gene (COL2A1) defect. Nature Genet. 3, 323 – 326 (1993). N. N. Ahmad, L. Ala-Kokko, R. G. Knowlton, S. A. Jimenez, E. J. Weaver, J. I. Maguire, W. Tasman, and D. J. Prockop, Stop codon in the procollagen II gene (COL2A1) in a family with the Stickler syndrome (arthro-ophthalmopathy). Proc. Natl. Acad. Sci. USA 88, 6624 – 6627 (1991). P. Vandenberg, J. S. Khillan, D. J. Prockop, H. Helminen, S. Kontusaari, and L. Ala-Kokko, Expression of a partially deleted gene of human type II procollagen (COL2A1) in transgenic mice produces a chondrodysplasia. Proc. Natl. Acad. Sci. USA 88, 7640 – 7644 (1991). M. Rintala, M. Metsaranta, S. Garofalo, B. de Crombrugghe, E. Vuorio, and O. Ronning, Abnormal craniofacial morphology and cartilage structure in transgenic mice harboring a Gly : Cys mutation in the cartilage-specific type II collagen gene. J. Craniofac. Genet. Dev. Biol. 13, 137 – 146 (1993). J. M. Pace, Y. Li, R. E. Seegmiller, C. Teuscher, B. A. Taylor, and B. R. Olsen, Disproportionate micromelia (Dmm) in mice caused by a mutation in the C-propeptide coding region of Col2a1. Dev. Dyn. 208, 25 – 33 (1997). Y. Ninomiya and B. R. Olsen, Collagens. In “Collagens” (T. Kreis and R. Vale, eds.), pp. 380 – 404. Oxford Univ. Press, Oxford, (1999). Y. Muragaki, E. C. Mariman, S. E. van Beersum, M. Perala, J. B. van Mourik, M. L. Warman, B. R. Olsen, and B. C. Hamel, A mutation in the gene encoding the alpha 2 chain of the fibril-associated collagen IX, COL9A2, causes multiple epiphyseal dysplasia (EDM2). Nature Genet. 12, 103 – 105 (1996). E. C. Spayde, A. P. Joshi, W. R. Wilcox, M. Briggs, D. H. Cohn, and B. R. Olsen, Exon skipping mutation in the COL9A2 gene in a family with multiple epiphyseal dysplasia. Matrix Biol. 19, 121 – 128 (2000). C. G. Bonnemann, G. F. Cox, F. Shapiro, J.-J. Wu, C. A. Feener, T. G. Thompson, D. C. Anthony, D. R. Eyre, B. T. Darras, and L. M. Kunkel, A mutation in the alpha 3 chain of type IX collagen causes

CHAPTER 5 Developmental Biology of Bone

253.

254.

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

autosomal dominant multiple epiphyseal dysplasia with mild myopathy. Med. Sci. 97, 1212 – 1217 (2000). J. Lohiniva, P. Paassilta, U. Seppanen, O. Vierimaa, S. Kivirikko, and L. Ala-Kokko, Splicing mutations in the COL3 domain of collagen IX cause multiple epiphyseal dysplaisa. Am. J. Med. Genet. 90, 216 – 222 (2000). P. Paassilta, J. Lohiniva, S. Annunen, J. Bonaventure, M. Le Merrer, L. Pai, and L. Ala-Kokko, COL9A3: A third locus for multiple epiphyseal dysplasia. Am. J. Hum. Genet. 64, 1036 – 1044 (1999). K. Nakata, K. Ono, J. Miyazaki, B. R. Olsen, Y. Muragaki, E. Adachi, K. Yamamura, and T. Kimura, Osteoarthritis associated with mild chondrodysplasia in transgenic mice expressing alpha 1(IX) collagen chains with a central deletion. Proc. Natl. Acad. Sci. USA 90, 2870 – 2874 (1993). R. Faessler, P. N. J. Schnegelsberg, J. Dausman, Y. Muragaki, T. Shinya, M. T. McCarthy, B. R. Olsen, and R. Jaenisch, Mice lacking a1(IX) collagen develop noninflammatory degenerative joint disease. Proc. Natl. Acad. Sci. USA 91, 5070 – 5074 (1994). C. J. Gress and O. Jacenko, Growth plate compressions and altered hematopoiesis in collagen X null mice. J. Cell Biol. 149, 983 – 993 (2000). D. Chan, Y. M. Weng, A. M. Hocking, S. Golub, D. J. McQuillan, and J. F. Bateman, Site-directed mutagenesis of human type X collagen: Expression of alpha1(X) NC1, NC2, and helical mutations in vitro and in transfected cells. J. Biol. Chem. 271, 13566 – 13572 (1996). D. Chan, Y. M. Weng, H. K. Graham, D. O. Sillence, and J. F. Bateman, A nonsense mutation in the carboxyl-terminal domain of type X collagen causes haploinsufficiency in schmid metaphyseal chondrodysplasia. J. Clin. Invest. 101, 1490 – 1499 (1998). D. Chan, S. Freddi, Y. M. Weng, and J. F. Bateman, Interaction of collagen alpha1(X) containing engineered NC1 mutations with normal alpha1(X) in vitro: Implications for the molecular basis of schmid metaphyseal chondrodysplasia. J. Biol. Chem. 274, 13091 – 13097 (1999). A. J. Richards, J. R. Yates, R. Williams, S. J. Payne, F. M. Pope, J. D. Scott, and M. P. Snead, A family with Stickler syndrome type 2 has a mutation in the COL11A1 gene resulting in the substitution of glycine 97 by valine in alpha 1(XI) collagen. Hum. Mol. Genet. 5, 1339 – 1343 (1996). A. J. Griffith, L. K. Sprunger, D. A. Sirko-Osadsa, G. E. Tiller, M. H. Meisler, and M. L. Warman, Marshall syndrome associated with a splicing defect at the COL11A1 locus. Am. J. Hum. Genet. 62, 816 – 823 (1998). S. Annunen, J. Körkkö, M. Czarny, M. L. Warman, H. G. Brunner, H. Kääriäinen, J. B. Mulliken, L. Tranebjaerg, D. G. Brooks, G. Cox, M. A. Curtis, S. Davenport, C. Friedrich, I. Kaitila, M. Krawczynski, A. Latos-Bielenska, S. Mukai, B. R. Olsen, N. Shinno, M. Somer, M. Vikkula, J. Zlotogora, D. J. Prockop, and L. Ala-Kokko, Splicing mutations of 54-bp exons in the COL11A1 gene cause Marshall syndrome, but other mutations cause overlapping Marshall/Stickler phenotypes. Am. J. Hum. Genet. 65, 974 – 983 (1999). M. Vikkula, E. C. M. Mariman, C. H. Lui, N. Zhidkova, G. E. Tiller, M. B. Goldbring, S. E. C. van Beersum, M. Malefijt, F. H. J. van den Hoogen, H.-H. Ropers, R. Mayne, K. Cheah, B. R. Olsen, M. L. Warman, and H. G. A. Brunner, Autosomal dominant and recessive osteochondrodysplasias associated with the COL11A2 locus. Cell 80, 431 – 437 (1995). R. Mayne, R. G. Brewton, P. M. Mayne, and J. R. Baker, Isolation and characterization of the chains of type V/type XI collagen present in bovine vitreous. J. Biol. Chem. 268, 9381 – 9386 (1993). Y. Li, D. A. Lacerda, M. L. Warman, D. R. Beier, H. Yoshioka, Y. Ninomiya, J. T. Oxford, N. P. Morris, K. Andrikopoulos, F. Ramirez, et al., A fibrillar collagen gene, Col11a1, is essential for skeletal morphogenesis. Cell 80, 423 – 430 (1995).

211 267. B. R. Olsen, New insights into the function of collagens from genetic analysis. Curr. Opin. Cell Biol. 7, 720 – 727 (1995). 268. H. Watanabe, Y. Yamada, and K. Kimata, Roles of aggrecan, a large chondroitin sulfate proteoglycan, in cartilage structure and function. J. Biochem. 124, 687 – 693 (1998). 269. H. Watanabe, K. Kimata, S. Line, D. Strong, L. Y. Gao, C. A. Kozak, and Y. Yamada, Mouse cartilage matrix deficiency (cmd) caused by a 7 bp deletion in the aggrecan gene. Nature Genet. 7, 154 – 157 (1994). 270. H. Watanabe, K. Nakata, K. Kimata, I. Nakanishi, and Y. Yamada, Dwarfism and age-associated spinal degeneration of heterozygote cmd mice defective in aggrecan. Proc. Natl. Acad. Sci. USA 94, 6943 – 6947 (1997). 271. H. Li, N. B. Schwartz, and B. M. Vertel, cDNA cloning of chick cartilage chondroitin sulfate (aggrecan) core protein and identification of a stop codon in the aggrecan gene associated with the chondrodystrophy, nanomelia. J. Biol. Chem. 268, 23504 – 23511 (1993). 272. H. Watanabe and Y. Yamada, Mice lacking link protein develop dwarfism and craniofacial abnormalities. Nature Genet. 21, 225 – 229 (1999). 273. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 274. M. D. Briggs, S. M. G. Hoffman, L. M. King, A. S. Olsen, H. Mohrenweiser, J. G. Leroy, G. R. Mortier, D. L. Rimoin, R. S. Lachman, E. S. Gaines, J. A. Cekleniak, R. G. Knowlton, and D. H. Cohn, Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nature Genet. 10, 330 – 336 (1995). 275. J. Hecht, L. D. Nelson, E. Crowder, Y. Wang, F. F. B. Elder, W. R. Harrison, C. A. Francomano, C. K. Prange, G. G. Lennon, M. Deere, and J. Lawler, Mutations in exon 17B of cartilage oligomeric matrix protein (COMP) cause pseudoachondroplasia. Nature Genet. 10, 325 – 329 (1995). 276. R. Ballo, M. D. Briggs, D. H. Cohn, R. G. Knowlton, P. H. Beighton, and R. S. Ramesar, Multiple epiphyseal dysplasia, ribbing type: A novel point mutation in the COMP gene in a South African family [published erratum appears in Am. J. Med. Genet. 71, 494 (1997)]. Am. J. Med. Genet. 68, 396 – 400 (1997). 277. E. Delot, L. M. King, M. D. Briggs, W. R. Wilcox, and D. H. Cohn, Trinucleotide expansion mutations in the cartilage oligomeric matrix protein (COMP) gene. Hum. Mol. Genet. 8, 123 – 128 (1999). 278. R. V. Iozzo, Matrix proteoglycans: From molecular design to cellular function. Annu. Rev. Biochem. 67, 609 – 652 (1998). 279. B. R. Olsen, Life without perlecan has its problems. J. Cell Biol. 147, 909 – 911 (1999). 280. M. M. French, S. E. Smith, K. Akanbi, T. Sanford, J. Hecht, M. C. Farach-Carson, and D. D. Carson, Expression of the heparan sulfate proteoglycan, perlecan, during mouse embryogenesis and perlecan chondrogenic activity in vitro. J. Cell Biol. 145, 1103 – 1115 (1999). 281. M. Costell, E. Gustafsson, A. Aszodi, M. Morgelin, W. Bloch, E. Hunziker, K. Addicks, R. Timpl, and R. Fassler, Perlecan maintains the integrity of cartilage and some basement membranes. J. Cell Biol. 147, 1109 – 1122 (1999). 282. A. Aszodi, D. Chan, E. Hunziker, J. F. Bateman, and R. Fassler, Collagen II is essential for the removal of the notochord and the formation of intervertebral discs. J. Cell Biol. 143, 1399 – 1412 (1998). 283. I. The, Y. Bellaiche, and N. Perrimon, Hedgehog movement is regulated through tout velu-dependent synthesis of a heparan sulfate proteoglycan. Mol. Cell 4, 633 – 639 (1999). 284. F. Deak, D. Piecha, C. Bachrati, M. Paulsson, and I. Kiss, Primary structure and expression of matrilin-2, the closest relative of cartilage

212 matrix protein within the von Willebrand factor type A-like module superfamily. J. Biol. Chem. 272, 9268 – 9274 (1997). 285. F. Deak, R. Wagener, I. Kiss, and M. Paulsson, The matrilins: A novel family of oligomeric extracellular matrix proteins. Matrix Biol. 18, 55 – 64 (1999). 286. N. Hauser, M. Paulsson, D. Heinegard, and M. Morgelin, Interaction of cartilage matrix protein with aggrecan: Increased covalent cross-linking with tissue maturation. J. Biol. Chem. 271, 32247 – 32252 (1996).

REGINATO, WANG, AND OLSEN 287. N. Winterbottom, M. M. Tondravi, T. L. Harrington, F. G. Klier, B. M. Vertel, and P. F. Goetinck, Cartilage matrix protein is a component of the collagen fibril of cartilage. Dev. Dyn. 193, 266 – 276 (1992). 288. A. Aszodi, J. F. Bateman, E. Hirsch, M. Baranyi, E. B. Hunziker, N. Hauser, Z. Bosze, and R. Fassler, Normal skeletal development of mice lacking matrilin 1: Redundant function of matrilins in cartilage? Mol. Cell Biol. 19, 7841 – 7845 (1999).

CHAPTER 6

Mouse Genetics as a Tool to Study Bone Development and Physiology MILLAN S. PATEL AND GERARD KARSENTY Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030

I. II. III. IV.

V. Independence of Bone Resorption from Bone Formation during Bone Remodeling VI. Bone Formation Is Centrally Regulated in Vivo References

Introduction The Osteoprotegerin Pathway Vitamin D Receptor Cbfa1

I. INTRODUCTION

peutic perspective. Moreover, the ability to delete several contiguous genes or an entire cell population expands considerably the scope of the information that can be obtained and analyzed. The second reason stems from the issues that are at stake in bone biology. As with every organ, little is known about the mechanisms controlling skeletal formation during development. Mouse genetics has shown that it is, along with chick embryology, the most powerful and elegant approach to address these questions for any organ, including the skeleton. However, what is specific to bone and only very few other organs is that we still know little about the molecular mechanisms controlling the function of its constituent cells: the osteoblasts and, to a lesser degree, the osteoclasts. That such questions of physiology and pathophysiology can be addressed successfully by mouse genetics has become increasingly evident over time. Although there are physiologic differences between mice and humans, the similarities for every organ far outnumber these differences. This has now been amply demonstrated for most organ physiology

Since the early 1990s, mouse genetics, defined here as a group of techniques aimed at altering gene expression and/or function in vivo, has invaded every field of biology, including bone biology. In doing so it has revolutionized thinking in each field it has entered. There are several reasons that fully justify such prominence, however, and this point deserves emphasis: mouse genetics is like every other field of biology, by itself it does not have all the answers to the questions confronting bone biology. One reason to explain the popularity of mouse genetics comes from the versatility of the technology. Indeed, it is now possible to delete a gene, decrease or increase its expression, or express it ectopically. The fact that these different manipulations can be done in vivo, during development or after birth, and in the cell type of choice, allows two equally important questions regarding any gene product to be addressed: what it normally does and what it can do. This latter aspect may be of greater interest from a thera-

OSTEOPOROSIS, SECOND EDITION VOLUME 1

213

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

214 including bone. The demonstration is so eloquent that the argument that mouse findings cannot be used to understand human physiology has lost most of its credibility. A final reason explaining the importance of mouse genetics as a tool comes from its track record. The functions of most cloned genes have now been studied in mice. In a few instances these studies have generated such clear-cut and unexpected results that they have substantially altered the way we think. There is one of the best examples of the revolutionary power of mouse genetics. The c-src gene codes for a tyrosine kinase membrane protein and is expressed ubiquitously. Surprisingly, c-src-deficient mice have only one phenotype: osteopetrosis due primarily to a functional defect of the osteoclast [1]. Further domain function analyses of c-src in vivo showed that this phenotype was tyrosine kinase domain independent. The function of c-src and the relative importance of each of its domains could not have been obtained without in vivo studies. This example best illustrates the power of mouse genetics in skeletal biology. Recent developments that illustrate the fundamental changes in conception that mouse genetics has brought to bone biology include the discovery of osteoprotegerin (OPG) and its ligand (RANKL, receptor activator of NFB ligand/ODF, osteoclast differentiation factor), the phenotype of vitamin D receptor (VDR)-deficient mice, the dual functions of Cbfa1 during development and postnatally, the independence of bone resorption from bone formation during bone remodeling, and the regulation of osteoblast function by the hypothalamus.

II. THE OSTEOPROTEGERIN PATHWAY Osteoclast biology has been profoundly revolutionized by the identification of a group of secreted molecules that positively or negatively regulate osteoclast differentiation and function. By screening an intestinal cDNA library for novel secreted molecules, a group at Amgen, Inc. identified a new member of the TNF receptor superfamily [2]. Named osteoprotegerin, this molecule contains no hydrophobic transmembrane-spanning sequence, suggesting that it is a soluble receptor. This molecule is identical to the osteoclastogenesis inhibitory factor (OCIF) purified and subsequently cloned by a group at Snow Brand Milk Products Co., Ltd. using a biochemical approach [3]. A daily intraperitoneal injection of recombinant OPG/OCIF in large amounts, as well as a high level of overexpression in transgenic mice, resulted in osteopetrosis due to arrested osteoclast differentiation. The identification and functional study of OPG/OCIF was historically of critical importance as it demonstrated that, in addition to steroid hormones and known peptide hormones, there are novel secreted molecules able to control osteoclast differentiation. The

PATEL AND KARSENTY

specificity of OPG/OCIF function for inhibiting osteoclast differentiation was further illustrated by the phenotype of OPG/OCIF-deficient mice. These mice develop severe osteoporosis due to increased numbers of osteoclasts [4,5]. The identification of a soluble receptor with such a powerful inhibitory effect on osteoclast differentiation suggested that it may act by binding a factor with osteoclast differentiation activity. This factor was cloned by both groups at nearly the same time. The Amgen group used recombinant OPG/OCIF to screen for OPG/OCIF ligand on the surface of various cell lines in a ligand-panning assay [6]. For this screening they rightly did not limit their search solely to novel molecules. The protein they isolated and called osteoprotegerin ligand (OPGL) had in fact been cloned previously and called TRANCE or RANK ligand (RANKL). This illustrates the limits of large genomic screens, as selection of only novel genes would have missed RANKL. At the same time, Yasuda et al. [7] from Snow Brand purified to homogeneity the osteoclast differentiation factor (ODF) and showed that it was the RANK ligand. For the sake of clarity it will be referred to here as RANKL or ODF, although this nomenclature controversy has not been settled. In vitro, RANKL/ODF has all the attributes of a genuine osteoclast differentiation factor: it favors osteoclast differentiation; in conjunction with M-CSF (macrophage colony stimulating factor), it bypasses the need for stromal cells and 1,25(OH)2 vitamin D3 to induce osteoclast differentiation; and it activates mature osteoclasts to resorb mineralized bone [8]. RANKL/ODF exists either in bound form on the membrane of osteoblast progenitors and other cells or as a soluble molecule in the bone microenvironment and in blood. Systemic administration of RANKL/ODF leads to increased bone resorption in vivo, and deletion of the RANKL gene leads to mice that lack osteoclasts and develop severe osteopetrosis in addition to immunological defects [9]. Two observations are of potential interest. First, because RANKL/ODF is secreted by osteoblasts and other cells, the secreted form could conceivably be the active form while the membrane-bound RANKL/ODF could be acting as a reservoir. Second, RANKL/ODF is also expressed by T cells, cells that can in vitro induce osteoclastogenesis [10]. The production and secretion of RANKL/ODF systemically or locally by T cells may explain some of the bone abnormalities observed in inflammatory disorders [9]. As mentioned previously, ODF is also called RANK ligand because it binds to a receptor, present on T cells and bone marrow stromal cells, called RANK [11]. Transgenic mice expressing a soluble form of RANK develop an osteopetrosis similar to that observed in RANKL/ODFdeficient mice. A polyclonal antibody against the RANK extracellular domain promotes osteoclastogenesis in bone marrow cultures, suggesting that RANK activation mediates the effect of RANKL/ODF [12]. The signal transduction

215

CHAPTER 6 Mouse Genetics as a Tool

pathway initiated following binding of RANKL/ODF to RANK has also been partly elucidated, thus establishing a cascade from extracellular signals to nuclear effectors. The intracellular domain of RANK contains two binding sites for members of a family of proteins called TNF receptor-associated factors (TRAFs) [13]. TRAFs have been implicated in mediating signals induced by a subset of TNF receptor family members. RANK contains a binding site for TRAF6 in its intracellular domain [14]. Importantly, TRAF6-deficient mice exhibit an osteopetrotic phenotype due to defective osteoclast function, thus providing the beginning of a signal transduction cascade leading to osteoclast terminal differentiation [15]. This observation is even more important as TRAFs appear to enhance c-src function [16] and to control the activation of NF-B, a transcription factor required for osteoclast differentiation [17,18]. Given the information available it is now possible to illustrate this signal transduction pathway along with those transcription factors known to affect osteoclast differentiation (see Fig. 1). What has been learned from the OPG-RANK-RANKL pathway could not have been performed as quickly and as thoroughly without using a genetic approach.

III. VITAMIN D RECEPTOR Vitamin D is a pleiotropic hormone whose active form is thought to play key roles in intestinal calcium absorption, renal calcium and phosphate conservation, and osteoclastic bone resorption. Vitamin D was hypothesized to induce monocytic stem cells in the bone marrow to differentiate into osteoclasts and to indirectly regulate osteoclastic activity through its effects on osteoblasts [19]. However, when mice were generated that lacked the vitamin D receptor (VDR), normal numbers of osteoclasts were seen [20]. Because VDR-deficient mice had normal bone at 3 weeks of age but showed a 40% reduction by 7 weeks, this further suggests that osteoclast function was not severely impaired. Most importantly, the florid rickets demonstrated by these mice was rescued by calcium supplementation, as had been shown more than a decade earlier for a human patient with vitamin D-resistant rickets [21]. In fact, VDR-deficient mice confirmed that in the entire animal, even though VDR appears to have a minor direct role in bone cell biology, its major role is indirect through effects on calcium regulation.

IV. Cbfa1

FIGURE 1

(A) Genes involved in osteoclast differentiation and maturation in vivo. mi, microphthalmia. (B) Signal transduction pathway leading to a functional osteoclast.

Transcriptional hierarchies are a well-established mechanism for cell-type specification. Extensive analysis of the promoter of the most osteoblast-specific gene, osteocalcin, was used to identify the first osteoblast-specific cis-acting elements (OSEs), OSE1 and OSE2. OSE2 is required for high-level gene expression and binds to an osteoblastspecific nuclear factor [22]. A biochemical and molecular approach revealed that this factor was identified as corebinding factor A1 (Cbfa1). Cbfa proteins are mammalian homologues of the Drosophila runt protein [23]. These proteins, along with their Drosophila and Caenorhabditis elegans counterparts, form a new family of transcription factors whose hallmark is a highly conserved 128 amino acid DNA-binding domain called the runt domain [24]. To date, three distinct Cbfa proteins encoded by different genes have been identified [24]. Several lines of evidence support the role of Cbfa1 as a transcriptional activator of osteoblast differentiation. Its pattern of expression during early development is confined to the mesenchymal condensations, which model the future skeleton, and its expression is detectable as early as 10.5 days postcoitum (dpc) [25]. At this early stage of skeletogenesis, Cbfa1 expression identifies a common progenitor for the chondrocytic and osteoblastic lineages. Beginning at 14.5 dpc and thereafter throughout life, Cbfa1 expression is restricted to cells of the osteoblastic lineage and, apart from decreasing expression in prehypertrophic and hypertrophic chondrocytes as chondrogenesis proceeds is absent from any other cell type [25 – 29]. To date, Cbfa1 remains the most specific and

216 earliest marker of osteogenesis known. Moreover, its function is consistent with its osteoblast-specific expression. Cbfa1 binds to the promoters of genes expressed predominatly in osteoblasts, such as  (I) collagen, bone sialoprotein, osteopontin, and osteocalcin, as well as being able to positively regulate their expression in tissue culture and in vivo. Further experiments showed that ectopic expression of Cbfa1 in fibroblast cell lines or in primary skin fibroblasts leads to the acquisition of an osteoblastic genotype by these cells, suggesting it may be sufficient for osteoblastogenesis [26]. However, a more convincing demonstration that Cbfa1 was necessary for osteoblast formation came from mouse genetics. Cbfa1-deficient mice die immediately after birth and skeletal studies showed a complete absence of osteogenesis with a consequent lack of both endochondral and intramembranous bone formation. Cartilage formation was grossly unaffected in these mutant mice [25,27]. Detailed histological analysis of the skeleton of these mice revealed an arrest of osteoblast differentiation before 14.5 dpc and a lack of expression of the osteoblast proteins, osteopontin and osteocalcin. Taken together, these data demonstrate that the Cbfa1 gene is essential for the differentiation of osteoblasts and thus for bone formation during development of the skeleton. Analysis of mice heterozygous for the Cbfa1 deletion demonstrated specific skeletal defects that were confined to those bones that form directly from mesenchymal precursors by intramembranous ossification. The most obvious of these defects, clavicular hypoplasia and delayed cranial bone ossification resulting in widely patent fontanelles, strongly resembled the human autosomal dominant disease cleidocranial dysplasia (CCD). A mouse mutation with similar features (also named Ccd) mapped to the same location as Cbfa1, and the Cbfa1 gene was found to be partially deleted in these mice [25]. Analysis of human patients with CCD also showed mutations of Cbfa1, some in the highly conserved runt domain, demonstrating that haploinsufficiency for this gene has functional consequences [30,31]. The similarity of the mouse and human phenotypes is important as it exemplifies how useful the mouse is to uncover the pathways controlling skeletal development in humans. As seen later the same holds true when it comes to skeletal physiology. The finding that Cbfa1 is expressed postnatally at high levels in osteoblasts, and the delay in skeletal maturation shown by CCD patients, led Ducy et al. [32] to examine, through mouse genetics, the consequences of selective postnatal inhibition of Cbfa1 function in osteoblasts using a dominant-negative form of the protein. This dominantnegative form consisted of the DNA-binding domain of Cbfa1 ( Cbfa1), which had a greater binding affinity for DNA than the native protein but lacked transactivation capabilities. The osteocalcin promoter was used to direct expression of this dominant-negative protein exclusively to differentiated osteoblasts. Because osteocalcin is virtually not expressed during development, this dominant-negative form of Cbfa1 was also not expressed during development.

PATEL AND KARSENTY

As expected, mice overexpressing  Cbfa1 had no features of Ccd at birth and were indistinguishable from their wildtype littermates. Over time, however, the transgenic mice developed marked osteopenia despite normal numbers of osteoblasts. The osteopenia affected both cortical and trabecular bone, and a 70% decrease in the bone formation rate was documented, suggesting that functional inhibition of Cbfa1 postnatally markedly inhibits differentiated osteoblast function. This second function of Cbfa1, the regulation of osteoblast physiology, was accompanied by a profound decrease in transcription of the collagen genes and that of all noncollagenous protein genes studied in the transgenic mice. Endogenous Cbfa1 expression was also decreased in the transgenic mice. Investigation of this phenomenon [32] revealed that Cbfa1 is a potent stimulator of its own expression and thus identified an autoregulatory loop at the top of a gene cascade controlling postnatal osteoblast function. These results illustrate how developmentally important genes can also play later roles in physiology. As mentioned above, Cbfa1 is also expressed in prehypertrophic chondrocytes and, to a lesser extent, in hypertrophic chondrocytes at early stages of chondrogenesis. Consistent with this pattern of expression, some, but not all, skeletal elements in Cbfa1-deficient mice lack hypertrophic chondrocytes [25,27,28,29]. To address the role of Cbfa1 during chondrogenesis transgenic mice were generated in which Cbfa1 was expressed in nonhypertrophic chondrocytes [30]. Continuous expression of Cbfa1 in nonhypertrophic chondrocytes induced chondrocyte hypertrophy and endochondral ossification in locations where it normally never occurs. To determine whether this was due to transdifferentiation of chondrocytes into osteoblasts or to a specific hypertrophic chondrocyte differentiation ability of Cbfa1, the transgene was introduced into Cbfa1-deficient mice. The transgene restored chondrocyte hypertrophy and vascular invasion but did not induce osteoblast differentiation. The rescue was cell-autonomous as skeletal elements not expressing the transgene were unaffected. Despite the lack of osteoblasts in the rescued mice, there were multinucleated, TRAP-positive cells reabsorbing the hypertrophic cartilage matrix. These results identify Cbfa1 as a hypertrophic chondrocyte differentiation factor and provide a genetic argument for a common regulation of osteoblast and chondrocyte differentiation mediated by Cbfa1.

V. INDEPENDENCE OF BONE RESORPTION FROM BONE FORMATION DURING BONE REMODELING Coupling between the functions of bone resorption and bone formation was originally proposed as the mechanism by which bone mass is maintained constant throughout the

CHAPTER 6 Mouse Genetics as a Tool

reproductive years [33]. Uncoupling of necessity occurs during longitudinal bone growth, fracture repair, and secondary ossification. These processes are collectively referred to as bone modeling to distinguish them from the proposed coupled state, referred to as bone remodeling. A key prediction of the coupling hypothesis is that absence or functional deficit of one cell type should greatly inhibit the function of the other cell type. In several mouse models with genetic defects of osteoclasts, however, osteopetrosis ensues, suggesting that osteoblast activity is not constrained by a lack of osteoclast function. Mouse genetics was used by Corral et al. [34] to construct an in vivo model of inducible osteoblast ablation in order to test the reverse: whether osteoclast activity would be inhibited in the complete absence of bone formation. Transgenic mice harboring the herpes simplex virus thymidine kinase gene under control of the osteocalcin promoter were generated and, upon administration of ganciclovir, longitudinal growth as well as bone formation activity ceased. Despite this, bone was progressively lost over time. This loss was due to osteoclast activity because it could be inhibited by the classical antiresorptive agent alendronate. These results showed that, in functional terms, bone resorption is independent of bone formation in vivo. RANKL/ODF expressed by osteoblast progenitors likely plays a role in cross-regulation; however, the transgenic results described earlier indicate that this occurs at the level of osteoblast differentiation, not at the level of osteoblast function. Another mouse genetic experiment demonstrates that the converse observation is also true. Mice deficient in 3 integrin that have a defect in bone resorption develop osteosclerosis, as bone formation is not affected [35].

VI. BONE FORMATION IS CENTRALLY REGULATED IN VIVO Treatment of osteoporosis requires modulation of bone cell function per se, and to this end an unexpected but critically important finding in osteocalcin-thymidine kinase transgenic mice was that after withdrawal of ganciclovir, there was a complete reversal of the morphologic, histologic, and histomorphometric findings within 4 weeks. The striking precision of this normalization process in such a short time suggests that osteoblasts had two speeds to deposit bone matrix in this model. Upon withdrawal of ganciclovir they could quickly deposit a huge amount of bone matrix to fill up the empty bone. Once that was achieved they could “slow down” their rate of extracellular matrix production and reenter the normal cycle of resorption/formation to maintain bone mass. This versatility was then interpreted as an indication that bone formation was an endocrine function [34]. This led to the discovery of leptin as the most potent inhibitor of bone formation identified to date.

217 Two clinical observations have been made repeatedly over the past several decades regarding the pathophysiology of osteoporosis. The first is that bone loss invariably follows the cessation of gonadal function and the second is that obesity protects from osteoporosis. In endocrine terms, these observations suggest that bone mass, fat mass, and gonadal function may be regulated by common molecules. This hypothesis, involving whole organism physiology, could only be tested efficiently using mouse genetics. Given its known functions, leptin was a good candidate molecule with which to test the hypothesis. Leptin-deficient (ob/ob) mice show striking obesity and hypogonadism. These phenotypes are recessive. The latter feature should significantly predispose to low bone mass; however, these mice have an increased bone formation rate that leads to a high bone mass phenotype [36]. This high bone mass phenotype is also seen in leptin receptor-deficient mice and in leptin receptor-deficient fa/fa rats (M. Amling, personal communication). Importantly, the high bone mass phenotype precedes the appearance of obesity in ob/ob mice. In contrast, fat-free mice [37], which have very low leptin levels, do have high bone mass, demonstrating that fat is not required as an intermediary. Expression of neither leptin nor its receptor could be detected in primary osteoblasts or in bone, and leptin signaling could not be detected in cultured, primary osteoblasts, suggesting that its mechanism of action is not autocrine, paracrine, or endocrine. In contrast, intracerebroventricular (icv) infusion of leptin resulted in correction of the high bone mass phenotype in ob/ob mice as well as bone loss in wild-type mice. These effects were shown to be specifically due to a decrease in bone formation with normal numbers of osteoblasts and no alteration of osteoclast function, indicating that leptin is a specific inhibitor of osteoblast function. Because the effects of leptin on body weight are mediated by binding of the hormone to receptors in the hypothalamus and icv infusion of leptin into the third ventricle inhibits bone formation, it is likely that the hypothalamus is the major site regulating osteoblast function and thereby bone mass homeostasis (Fig. 2). This central regulation of bone remodeling is the only form of regulation yet identified that can overcome the deleterious effect of hypogonadism on bone mass. This point best underscores its physiologic relevance.

FIGURE 2 Central regulation of body weight, osteoblast function, and gonadal function by leptin and possibly other secreted molecules.

218

PATEL AND KARSENTY

How does this study relate to human physiology and the initial hypothesis that bone mass, fat mass, and gonadal function share common regulation? Both humans and mice with diet-induced obesity show increased leptin levels, proportional to their fat mass. The lack of functional inhibition by leptin in these situations has been termed “leptin resistance”; thus, the protective nature of obesity for osteoporosis can be thought of as a state of resistance to the osteopenic effects of leptin. The findings described earlier, which relied entirely on mouse and rat genetics, demonstrate unambiguously that bone formation is a centrally regulated function. The functional importance of this pathway is reinforced by the absence of low bone mass in human patients with leptin deficiency, despite their hypogonadism. This in turn exemplifies the usefulness of the mouse as a tool to study human physiology. This new view of bone remodeling, when reinforced by the central action of other hormones affecting bone metabolism, may have profound implications for our understanding and management of osteoporosis.

References 1. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 2. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, G. L. Renshaw, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 86, 309 – 319 (1997). 3. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 4. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 5. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 6. D. L. Lacey, E. Timms, H.-L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliot, A. Colombero, G. Elliot, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y.-X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 7. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/ osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc. Natl. Acad. Sci. USA. 95, 3597 – 3602 (1998).

8. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell. Biol. 145, 527 – 538 (1999). 9. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 10. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. R. Dunn, T. Martin, and M. T. Gillespie, Interleukin 18 inhibits osteoclast formation via T cell production of granulocyte macrophage colony-stimulating factor. J. Clin. Invest. 101, 595 – 603 (1998). 11. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tometsko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligend enhance T-cell growth and dendritic-cell function. Nature 390, 175 – 179 (1997). 12. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). 13. B. R. Wong, R. Josien, S. Y. Lee, M. Vologodskaia, R. M. Steinman, and Y. Choi, The TRAF family of signal transducers mediates NFkappaB activation by the TRANCE receptor. J. Biol. Chem. 273, 28355 – 28359 (1998). 14. B. G. Darnay, J. Ni, P. A. Moore, and B. B. Aggarwal, Activation of NF-kappaB by RANK requires tumor necrosis factor receptor-associated factor (TRAF) 6 and NF-kappaB-inducing kinase: Identification of a novel TRAF6 interaction motif. J. Biol. Chem. 274, 7724 – 7731 (1999). 15. M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, A. Van der Heiden, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. Cao, J. M. Penninger, C. J. Paige, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). 16. B. R. Wong, D. Besser, N. Kim, J. R. Arron, M. Vologodskaia, H. Hanafusa, and Y. Choi, TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-Src. Mol. Cell. 4, 1041 – 1049 (2000). 17. G. Franzoso, L. Carlson, L. Xing, L. Poljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997). 18. V. Iotsova, J. Caamano, J. Loy, Y. Yang, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NF-kappaB1 and NF-kappaB2. Nature Med. 3, 1285 – 1289 (1997). 19. M. F. Holick, Vitamin D: Photobiology, metabolism, mechanism of action, and clinical applications. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus et al. eds.), 4th Ed. pp. 92 – 98. Lippincott, Williams & Wilkins, Philadelphia, 1999. 20. T. Yoshizawa, Y. Handa, Y. Uematsu, S. Takeda, K. Sekine, Y. Yoshihara, T. Kawakami, K. Arioka, H. Sato, Y. Uchiyama, S. Masushige, A. Fukamizu, T. Matsumoto, and S. Kato, Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nature. Genet. 16, 391 – 396 (1997). 21. S. Balsan, M. Garabedian, M. Larchet, A. M. Gorski, G. Cournot, C. Tau, A. Bourdeau, C. Silve, and C. Ricour, Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in

CHAPTER 6 Mouse Genetics as a Tool

22.

23.

24.

25.

26.

27.

28.

29. 30.

hereditary resistance to 1,25-dihydroxyvitamin D. J. Clin. Invest. 77, 1661 – 1667 (1986). P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). J. B. Duffy, M. A. Kania, and J. P. Gergen, Expression and function of the Drosophila gene runt in early stages of neural development. Development 113, 1223 – 1230 (1991). J. J. Westendorf and S. W. Hiebert, Mammalian runt-domain proteins and their roles in hematopoiesis, osteogenesis, and leukemia. J. Cell. Biochem. Suppl. 32 – 33, 51 – 58 (1999). F. Otto, A. P. thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). P. Ducy, R. Zhang, V. Geoffroy, A. Ridall, and G. Karsenty, Osf2/Cbf1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). T. Komori, H. Yahi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Kitamura, S. Yoskiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). Deleted at proof.

219 31. B. Lee, K. Thirunabukkarasu, L. Zhou, L. Pastore, A. Baldini, J. Hecht, V. Geoffroy, P. Ducy, and G. Karsenty, Missense mutations abolishing DNA binding of the osteoblast-specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nature. Genet. 16, 307 – 310 (1997). 32. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 33. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). 34. H. M. Frost, “Bone Biodynamics,” p. 315. Little, Brown, Boston, 1964. 35. D. Corral, M. Amling, M. Prienel, E. Loyer, S. Fuchs, P. Ducy, R. Baron, and G. Karsenty, Dissociation between bone resorption and bone formation in osteopenic transgenic mice. Proc. Natl. Acad. Sci. USA 95, 13835 – 13840 (1998). 36. K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking beta3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest. 105, 433 – 440 (2000). 37. P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). 38. J. Moitra, M. M. Mason, M. Olive, D. Krylov, O. Gavrilova, B. Marcus-Samuels, L. Feigenbaum, E. Lee, T. Aoyama, M. Eckhaus, M. L. Reitman, and C. Vinson, Life without white fat: A transgenic mouse. Genes Dev. 12, 3168 – 3181.

CHAPTER 7

Parathyroid Hormone and Parathyroid HormoneRelated Protein ROBERT A. NISSENSON

Endocrine Unit, San Francisco Veterans Affairs Medical Center, and Departments of Medicine and Physiology, University of California, San Francisco, San Francisco, California 94121

I. Introduction II. Parathyroid Hormone (PTH) III. Parathyroid Hormone-Related Protein (PTHrP)

IV. Mechanism of Action of PTH and PTHrP References

I. INTRODUCTION

(paracrine) factor that controls the development, morphogenesis, and function of a variety of tissues including (but not limited to) those involved in skeletal and mineral homeostasis. PTH and PTHrP are tied together historically in that PTHrP was discovered as a result of the quest to understand the pathogenesis of malignancy-associated hypercalcemia. However, they are also related structurally and produce their major physiological effects by activating a common receptor, the PTH/PTHrP (or PTH1) receptor. This chapter focuses on our current understanding of the physiology and mechanism of action of these polypeptides.

Parathyroid hormone (PTH) and PTH-related protein (PTHrP) are major factors that regulate skeletal physiology and mineral homeostasis. The appearance of parathyroid glands during the evolution of terrestrial vertebrates underscores the primary functional role of parathyroid hormone (PTH): the maintenance of adequate levels of plasma-ionized calcium in the face of a calcium-deficient terrestrial environment. The secretion of PTH by parathyroid glands is stimulated when plasma ionized calcium levels fall. Once secreted, PTH acts to restore normal levels of ionized calcium through an integrated series of actions on bone, kidney, and (indirectly) the intestine. When present as a circulating factor, PTHrP, produces target cell effects that resemble those of PTH. This is most evident in malignancy-associated hypercalcemia where tumors elaborate sufficient quantities of PTHrP to produce biochemical abnormalities overlapping those seen in primary hyperparathyroidism. However, the major physiological function of PTHrP is to act as a local

OSTEOPOROSIS, SECOND EDITION VOLUME 1

II. PARATHYROID HORMONE (PTH) A. Secretion The parathyroid glands first appear during evolution with the movement of animals from an aquatic environment to a terrestrial environment deficient in calcium. Maintenance of

221

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

222

FIGURE 1

Relationship between plasma levels of ionized calcium and the release of PTH(1 – 84) in normal humans. Variations in plasma ionized calcium were achieved by the infusion of calcium or EDTA. Note the sigmoidal relationship, ensuring significant changes in PTH secretion with small variations in ionized calcium. Reproduced from Brown [1], with permission.

adequate levels of plasma ionized calcium (1.0 – 1.3 mM) is required for normal neuromuscular function, bone mineralization, and many other physiological processes. The parathyroid gland secretes PTH in response to very small decrements in blood-ionized calcium in order to maintain the normocalcemic state. As discussed later, PTH accomplishes this task by promoting bone resorption and releasing calcium from the skeletal reservoir; by inducing renal conservation of calcium and excretion of phosphate; and by indirectly enhancing intestinal calcium absorption by increasing the renal production of the active vitamin D metabolite [1,25(OH)2D] vitamin D. The parathyroid gland functions in essence as a “calciostat,” sensing the prevailing bloodionized calcium level and adjusting the secretion of PTH accordingly (Fig. 1) [1]. The relationship between ionized calcium and PTH secretion is a sigmoidal one, allowing significant changes in PTH secretion in response to very small changes in plasma-ionized calcium. In addition to providing acute regulation of PTH secretion, ionized calcium is also a primary factor controlling chronic secretion of the hormone. Thus, sustained hypocalcemia promotes increased expression of the PTH gene [2,3] and results in parathyroid hyperplasia. A common example of the latter is the marked parathyroid hyperplasia (secondary hyperparathyroidism) that frequently accompanies chronic renal failure. 1,25(OH)2D also serves as a negative regulator of PTH gene expression and parathyroid cell hyperplasia. In chronic renal failure, both hypocalcemia and reduced circulating levels of 1,25(OH)2D presumably contribute to the progression of secondary hyperparathyroidism. There has been great progress in our understanding of how extracellular calcium controls PTH secretion [4,5]. The

ROBERT A. NISSENSON

plasma membrane of parathyroid cells contains high levels of a calcium-sensing receptor (CaR). Unlike intracellular calcium-binding proteins, which have an affinity for free calcium in the nanomolar range (consistent with intracellular levels of free calcium), the CaR binds calcium in the millimolar range. The receptor is a member of the G-protein-coupled receptor (GPCR) superfamily. It contains calcium-binding elements in its extracellular domain and signaling determinants in its cytoplasmic regions. Calcium binding to the receptor triggers activation of the G-proteins Gq and (to a lesser extent) Gi, resulting in the stimulation of phospholipase C and inhibition of adenylyl cyclase, respectively. This results in an increase in intracellular calcium and a decrease in cyclic AMP levels in parathyroid cells. By mechanisms that are not yet clear, these signaling pathways serve to suppress the synthesis and secretion of PTH. When blood ionized calcium falls, there is less signaling by the CaRs on the parathyroid cell and PTH secretion consequently increases. The essential role of the CaR can best be seen in individuals harbouring loss-of-function mutations in the CaR gene. In the heterozygous state, such mutations result in familial hypocalciuric hypercalcemia, characterized by inappropriately high levels of PTH secretion in the face of hypercalcemia [6]. These individuals are quantitatively resistant to the suppressive effect of calcium on PTH secretion due to the reduced number of parathyroid CaRs. In the homozygous state, patients display a severe increase in PTH secretion with life-threatening hypercalcemia (neonatal severe primary hyperparathyroidism).

B. Metabolism Early studies demonstrated that PTH circulates in multiple forms that can be distinguished by radioimmunoassays specific for different regions of the PTH molecule [7 – 9]. This heterogeneity has two origins (Fig. 2). PTH(1 – 84) is subject to metabolism within the parathyroid gland, resulting in the secretion of PTH fragments as well as the intact molecule. In addition, PTH(1 – 84) is metabolized in peripheral tissues. Midregion and carboxyl-terminal fragments of PTH have a much longer half-life in the circulation than PTH(1 – 84) [10 – 13]. As a result, midregion and carboxyl-terminal fragments of PTH circulate at much higher concentrations than intact PTH(1 – 84). Rapid plasma clearance of PTH is due primarily to hepatic metabolism, with a lesser contribution by the kidneys [14 – 16]. Peripheral metabolism generates mid- and carboxylterminal fragments of PTH that resemble those secreted by the parathyroid gland. Mid- and carboxyl-terminal PTH fragments are cleared by renal excretion, and thus circulating levels of these fragments are highly dependent on renal function. Extremely high levels of PTH detected with antibodies against the mid- and carboxyl-regions of the hormone in many patients with end-stage renal disease thus

223

CHAPTER 7 PTH and PTH-Related Protein

FIGURE 2

Metabolism and clearance of PTH. PTH is subject to proteolytic cleavage in the parathyroid gland, as well as in liver and kidney, resulting in the presence of inactive midregion and carboxyl-terminal PTH fragments in the circulation. Amino-terminal PTH fragments are apparently degraded rapidly and do not accumulate in the circulation. Intact PTH has a short half-life in the circulation (2 – 4 min) due to hepatic and renal metabolism. Midregion and carboxy-terminal PTH fragments are cleared by glomerular filtration. They have a much longer half-life that is dependent on the level of renal function. Reproduced from Endres et al. [341], with permission.

reflect a combination of secondary hyperparathyroidism and reduce renal clearance of PTH fragments. Mid- and carboxyl-region PTH fragments lack the amino-terminal 1 – 34 sequence of the hormone required for binding to PTH/PTHrP receptors and producing the classical effects of PTH on kidney and bone. Metabolism of PTH could produce biologically active, amino-terminal fragments of PTH, but there is little evidence for the presence of significant levels of amino-terminal PTH fragments in the circulation [17] or for significant secretion of such fragments by the parathyroid gland [18]. Presumably, both the parathyroid gland and peripheral organs contain enzymes that degrade amino-terminal fragments of PTH. This ensures that circulating levels of biologically active PTH are derived exclusively from glandular secretion of PTH(1 – 84). A few studies have demonstrated potential biological effects of mid- or carboxyl-region fragments of PTH [19 – 21], and there is also evidence for the existence of membrane receptors for these fragments [22,23]. However, the biological role of PTH fragments remains unclear. Calcium-sensitive cathepsins are responsible for cleaving PTH (1 – 84) within the parathyroid gland. Intraglandular cleavage occurs between residues 34 and 35 or between

residues 36 and 37 [24,25], and a greater proportion of PTH is cleaved under conditions of hypercalcemia [26]. The amino-terminal fragments so produced are degraded rapidly within the parathyroid gland, and thus calcium-sensitive cleavage constitutes a mechanism for the inactivation of PTH. Therefore, the level of plasma calcium determines not only the rate of synthesis and secretion of PTH, but also the extent to which secreted PTH is biologically active.

C. Physiological Actions 1. BONE a. Bone Resorption The major physiological role of PTH is to mobilize calcium from bone in order to maintain an adequate level of plasma-ionized calcium. This is accomplished by a direct action of PTH on bone that results in increased osteoclastic bone resorption and increased flux of calcium from bone into blood. Administration of PTH produces rapid movement of calcium out of bone, an effect that is associated with structural changes in cells lining the endosteal surface [27]. It has been suggested that these lining cells form an epithelial-like barrier between the

224

ROBERT A. NISSENSON

circulation and the bone extracellular fluid [28,29], and PTH may act on these cells to promote calcium transport. PTH enhances osteoclastic bone resorption within 15 min of its administration [30] and produces a sustained increase in bone resorption that appears to require the recruitment and differentiation of new osteoclasts. PTH-induced bone resorption involves the dissolution of hydroxyapatite bone mineral in the acidic microenvironment created by the osteoclast, as well as the degradation of collagen and other matrix proteins by proteolytic enzymes. Over the years, the cellular and molecular basis of the ability of PTH to promote osteoclastic bone resorption has been a source of considerable puzzlement. PTH receptors have been localized to bone-forming osteoblasts and their precursors [31,32], but it is not clear that mature osteoclasts possess PTH receptors [31 – 35]. Indeed, PTH is not able to activate isolated osteoclasts in vitro unless osteoblast-like cells are also present [36,37]. These findings suggest that PTH may produce its actions on osteoclasts indirectly, perhaps through direct interaction with cells of the osteoblast lineage. It is possible that PTH action on osteoblast lining cells alters their attachment to the surface of bone or reduces cell – cell interactions, allowing osteoclasts to gain access to the mineralized bone surface. Indeed, PTH has dramatic effects on the morphology of isolated osteoblasts [38] and alters osteoblast expression of connexin 43, a protein involved in cell – cell communication [39 – 41]. In addition, osteoblasts are known to respond to PTH by secreting proteins such as collagenase [42 – 46] and plasminogen activators [47 – 49], which may facilitate osteoclastic bone resorption. Indeed, PTH-stimulated bone resorption is blunted in mice expressing a mutated form of type I collagen that is resistant to digestion by collagenase [50]. As mentioned earlier, studies on isolated osteoclasts indicate that these cells do not display increased bone resorption

FIGURE 3

in the presence of PTH unless accessory cells such as osteoblasts are also present [36,37,51]. Osteoblasts secrete several cytokines that could potentially influence osteoclast activity through a paracrine mechanism [52 – 54]. However, it appears that direct contact between accessory cells and osteoclasts is required for PTH-induced osteoclast activation [55]. An explanation for this derives from the discovery of the role of osteoprotegerin ligand (RANKL) and its receptor (RANK) in the regulation of osteoclast differentiation and function [56,57] (See Chapters 2, 3, and 12). RANK is a tumor necrosis factor (TNF)- receptor-related protein that is expressed in osteoclast precursors as well as in differentiated osteoclasts. RANK signaling in osteoclast precursors promotes differentiation to functional osteoclasts, and RANK signaling in differentiated osteoclasts enhances bone resorption and inhibits apoptosis [58 – 61]. In either case, RANKL binding to RANK is required for signaling. RANKL is not a secreted protein but rather is an intrinsic membrane protein expressed on the surface of cells of the osteoblast lineage. Thus, direct contact between cells of the osteoblast lineage and osteoclasts or their precursors is required for the engagement of RANKL with RANK leading to osteoclast differentiation and activation. RANKL is required for normal osteoclast development and function, and mice lacking RANKL show a loss of functional osteoclasts and osteopetrosis [62]. Cells in the microenvironment of bone also secrete a truncated TNF- receptor-like molecule termed osteoprotegerin (OPG) that binds to RANKL and prevents RANK signaling [63 – 65]. The importance of OPG as a tonic suppressor of bone turnover is evident from findings with mice lacking functional expression of OPG. These animals display increased bone resorption and osteoporosis [66,67]. Current evidence indicates that the RANKL/RANK system plays a major role in PTH-induced bone resorption and calcium mobilization (Fig. 3). Administration of exogenous,

Regulation of osteoclast differentiation and activation by PTH. Binding of PTH by receptors on osteoblasts results in increased expression of osteoprotegerin ligand (RANKL) on the cell surface. Activation of the PTH receptor can also reduce the secretion of the RANKL inhibitor osteoprotegerin (OPG), which is produced by cells in the bone microenvironment. These effects of PTH promote the action of RANKL on its receptor (RANK) on the surface of osteoclast precursors and mature osteoclasts. RANK signaling, together with the action of macrophage colony-stimulating factor, stimulates the differentiation of osteoclast precursors and promotes the activation of mature osteoclasts.

225

CHAPTER 7 PTH and PTH-Related Protein

soluble RANKL to mice elicits severe hypercalcemia within 1 day of administration, and increased osteoclast activity and bone loss are evident within 3 days [58]. Administration of OPG (RANKL antagonist) blocks the calcemic action of exogenous PTH in vivo [63]. Addition of OPG also inhibits PTH-induced osteoclast activation and bone resorption in vitro [68,69]. PTH is reported to increase the ratio of RANKL:OPG expressed by osteoblastic cells, an effect that is largely due to the ability of PTH to increase the expression of RANKL [60,70,71]. PTH also inhibits the expression of OPG [70,72]. Factors such as interleukin (IL)-6 [73] and insulin-like growth factor (IGF)-1 [74] have been suggested to play a role in mediating the actions of PTH on osteoclast formation and bone resorption. However, induction of RANKL in cells of the osteoblast lineage may well be the major mechanism underlying these effects of PTH under physiological conditions. b. Bone Formation PTH acts directly on cells of the osteoblast lineage, thereby influencing osteoblast differentiation and function, and consequently bone formation. Administration of PTH intermittently to animals or humans results in a marked anabolic response of the skeleton [75 – 82] (See also Chapter 77). PTH promotes bone formation in both trabecular and cortical bone, and these actions are associated with increased trabecular thickness and increased bone strength [83 – 89]. High levels of PTH are known to produce an increase in the number of osteoblasts, which results in part from the coupling between increased osteoclastic resorption and new bone formation. However, intermittent treatment with low doses of PTH produces a direct positive effect on bone formation that is independent of preceding bone resorption. The cellular basis for this action of PTH is not fully understood, but there are a number of potential targets for PTH (Fig. 4). PTH could promote the differentiation of stromal cell osteoblast precursors to matrix-synthesizing mature osteoblasts. The hormone could also function as an osteoblast mitogen or increase the life span of mature osteoblasts, thereby increasing their pool size. Finally, PTH could directly enhance the ability of mature osteoblasts to synthesize and secrete matrix proteins and to promote matrix mineralization. Data support a number of these possibilities. PTH receptors are present on osteoblast precursors, including bone marrow stromal cells [90 – 92]. Available evidence indicates that PTH increases the number of active, replicating osteoblasts, but its direct effect on the replication of osteoblastic cells is variable [93 – 96]. Model systems for osteoblast differentiation in vitro reveal a positive effect of PTH on differentiation, depending on the dose and mode of exposure, with intermittent treatment with low doses being most consistently effective [97 – 99]. Active osteoblasts are subject to death by apoptosis and that PTH treatment reduces the number of apoptotic osteoblasts in vivo [100].

FIGURE 4

Possible mechanisms contributing to the anabolic effect of PTH in bone. PTH may act on bone marrow stromal cell precursors to promote their differentiation to functional osteoblasts. PTH could also act directly on osteoblasts to increase their number or their functional activity. Finally, PTH could increase the life span of mature osteoblasts by inhibiting their death via apoptosis.

Taken together, available data support the notion that PTH elicits an increase in osteoblast number via actions to promote osteoblast differentiation and to inhibit osteoblast apoptosis. Direct effects of PTH on osteoblasts in vitro to promote the synthesis of matrix proteins have also been reported [101 – 103]. Some of these effects may be secondary to the release of osteoblast growth factors such as IGF-1 [104], and their relevance to the anabolic action of PTH in vivo remains to be established. Intermittent (e.g., once daily) treatment with PTH elicits skeletal effects in which increased bone formation predominates, whereas continuous treatment with high doses of PTH results in a major increase in bone resorption. Continuous treatment of target cells with high doses of PTH results in a loss of responsiveness (desensitization), and it is possible that the anabolic effects of PTH are particularly sensitive to hormone-induced desensitization. Intermittent administration of PTH could allow for resensitization of the anabolic response prior to administration of a subsequent dose of hormone. However, continuous administration of lower doses of PTH also elicits an anabolic skeletal response, suggesting that the balance between bone resorption and anabolism

226 may be related to the dose of PTH rather than to its intermittent administration. The effects of PTH also differ depending on the nature of the skeletal site, with trabecular bone displaying the greatest increase in mass in response to PTH. At doses of PTH that are anabolic in trabecular bone, cortical bone displays increased bone resorption as well as increased bone formation. The net effect of PTH treatment on cortical bone mass in thus variable. 2. KIDNEY PTH produces a series of renal actions that help ensure that calcium mobilized from bone contributes optimally for the maintenance of plasma-ionized calcium levels. The reanl actions of PTH include inhibition of renal phosphate reabsorption, stimulation of renal calcium reabsorption, and increased production of 1,25(OH)2D. The ability of PTH to inhibit renal phosphate reabsorption has been known for many years, providing the basis for the clinical Ellsworth – Howard test of renal responsiveness to the hormone [105]. Patients with primary hyperparathyroidism display hypophosphatemia and decreased renal tubular reabsorption of phosphate, whereas hypoparathyroid patients are hyperphosphatemic and have increased phosphate reabsorption. Phosphate forms a complex with free calcium in blood. Thus, for a given level of serum calcium, ionized calcium will be reduced as serum phosphate increases. Under conditions of relative hypocalcemia (e.g., during chronic dietary calcium deficiency), PTH secretion is increased, resulting in increased bone resorption. Both calcium and phosphate are released from hydroxyapitite during the process of bone resorption. By promoting renal excretion of phosphate, PTH facilitates a rise in ionized as well as total plasma calcium. Phosphate reabsorption in the proximal renal tubule is dependent in part on the activity of the type IIa sodium – phosphate cotransporter. The phosphaturic action of PTH derives from the action of the hormone to inhibit the function of this transporter [106]. The type IIa transporter is located in the apical plasma membrane and permits the coupled transport of sodium and phosphate from the tubule into the renal cell. Exposure of proximal tubular cells to PTH results in a reduced Vmax of the transporter [107, 108], and this is associated with a decrease in the amount of the transporter in the apical plasma membrane [109]. Acute exposure of the proximal tubular cells to PTH enhances the endocytosis and subsequent lysosomal degradation of the transporter, and this may be the major mechanism responsible for rapid PTH-induced inhibition of renal phosphate reabsorption [110,111]. PTH appears to regulate the type II transporter by enhancing its rate of turnover rather than by suppressing its synthesis [112]. PTH also acts to increase renal calcium reabsorption, thus ensuring that only small amounts of calcium released during PTH-induced bone resorption are lost via renal

ROBERT A. NISSENSON

excretion. The major sites for this effect of PTH are in the distal convoluted tubule and the thick ascending limb of Henle’s loop [113,114]. Evidence indicates that distal renal tubular calcium reabsorption is an active process that requires calcium influx through dihydropyridine-sensitive calcium channels located in the apical plasma membrane [115]. Drugs that inhibit these channels are effective in blocking PTH-induced renal calcium reabsorption. Unlike voltage-sensitive calcium channels in excitable tissues, PTH-responsive calcium channels in the distal nephron are activated by membrane hyperpolarization [116]. PTH appears to open calcium channels by inducing hyperpolarization of the apical plasma membrane. Calcium entering the distal renal tubular cell in this manner is transported into the extracellular compartment via a sodium – calcium exchanger present on the basolateral plasma membrane [117]. PTH promotes intestinal calcium absorption indirectly through an action to increase circulating levels of 1,25(OH)2D. This vitamin D metabolite acts directly on intestinal epithelial cells to increase the efficiency of calcium (and phosphate) absorption. Primary hyperparathyroidism is commonly associated with increased circulating levels of 1,25(OH)2D, whereas reduced levels of this metabolite are present in hypoparathyroidism [118]. PTH produces this effect by increasing the rate of production of 1,25(OH)2D through activation of the 25(OH)D-1-hydroxylase enzyme located in the proximal renal tubule [119 – 122]. The gene encoding this enzyme has been cloned in several laboratories [123 – 125]. Studies in vivo as well as in cultured renal cell lines indicate that PTH increases the expression of the 25(OH)D-1-hydroxylase gene through a transcriptional mechanism [126,127].

III. PARATHYROID HORMONERELATED PROTEIN (PTHrP) A. Role in Malignancy-Associated Hypercalcemia The frequent occurrence of hypercalcemia in individuals with a variety of malignancies has been recognized for many years. An important clue to the pathogenesis of malignancy-associated hypercalcemia (MAH) came with the recognition 20 years ago that many such individuals display an increased excretion of renal-derived (“nephrogenous”) cyclic AMP [128]. Activation of the renal PTH receptor by elevated circulating levels of PTH in hyperparathyroidism was the only known cause of increased nephrogenous cyclic AMP, and thus it was suggested that malignant tumors are capable of producing a factor that activates PTH receptors. Plasma levels of immunoreactive PTH were

227

CHAPTER 7 PTH and PTH-Related Protein

found to be low in patients with MAH, indicating that the relevant circulating factor was not PTH itself. Using the activation of PTH receptors as an assay, several groups succeeded in isolating and ultimately identifying the PTH-like etiologic factor in MAH [129 – 132]. This factor was termed PTH-related protein (1) because of its ability to bind to and activate the PTH receptor and (2) because of its limited sequence similarity to PTH [133 – 135]. The PTHrP gene is subject to alternative splicing, resulting in the production of three protein products ranging from 139 to 173 amino acids differing only in their carboxyl-terminal sequence [136,137]. PTHrP is capable of reproducing the major target cell actions of PTH and (like PTH) does so via the amino-terminal 34 amino acids or so of the protein. A comparison of the 1 – 34 sequences of PTH and PTHrP reveals significant amino acid homology, with identity in 8 of the 13 amino terminal residues. Two of the known contact sites between PTH and the PTH/PTHrP receptor are within this 13 amino-acid homologous region [138], indicating that these ligands use very similar mechanisms to activate their common receptor. The molecular mechanisms underlying the over expression of PTHrP by malignant tumors remain unclear. As the mass of PTHrP-expressing tumor cells expands, systemic levels of PTHrP eventually increase sufficiently to allow the peptide to elicit endocrine effects on PTH/PTHrP receptors in bone and kidney, resulting in MAH.

B. Physiological Roles Although PTHrP produces PTH-like target cell effects in patients with MAH, circulating levels of PTHrP are very low to undetectable in normal individuals. This, coupled with the widespread expression of the PTHrP gene in normal tissues, suggested that PTHrP was likely to have physiological functions as a local, paracrine factor rather than as a systemic hormone. Subsequent studies have confirmed that PTHrP indeed plays an important role as a paracrine factor in a wide variety of tissues (Table 1) [139 – 143], as summarized next. 1. ENDOCHONDRAL BONE DEVELOPMENT The first direct evidence concerning a physiological role for PTHrP appeared in 1994 with the report of the phenotype of mice lacking the expression of PTHrP due to targeted gene ablation [144]. These animals died shortly after birth and were found to display a form of short-limbed dwarfism with generalized chondrodysplasia. The most striking feature of mice lacking the expression of PTHrP is the disruption of normal endochondral ossification. Although the most obvious gross phenotypic abnormality is short-limbed dwarfism, the defect in endochondral bone formation is generalized. The role of PTHrP is best under-

TABLE 1 Target tissue

Target Tissue Actions of PTHrP Actions

Cartilage

Inhibits terminal chondrocyte differentiation; increases chondrocyte proliferation

Bone

Regulates bone resorption

Mammary gland

Facilitates branching morphogenesis of mammary epithelium; may play an endocrine or paracrine role in lactation

Skin

Inhibits terminal differentiation of keratinocytes; promotes normal hair follicle development

Teeth

Promotes normal tooth eruption

Extraembryonic endoderm Enhances the differentiation of primitive endoderm to parietal endoderm Smooth muscle

Serves as a general smooth muscle relaxant

Central nervous system

Inhibits neuronal L-type calcium channel activity; protects neurons from excitotoxicity

Placenta

Maintains the positive maternal – fetal transplacental calcium gradient

stood in the context of the homeostatic mechanisms regulating the differentiation of cartilage and bone during endochondral bone formation. In the long bones, chondrogenesis is initiated by the differentiation of mesenchymal cell precursors that form nodules and begin to express characteristic genes, including those encoding type II collagen and other cartilage matrix proteins [145]. These early chondrocytes are mitotically active, but the cells in the center of the nodule become hypertrophic, cease dividing, and express gene products characteristic of mature chondrocytes (such as type X collagen). Hypertrophic chondrocytes undergo programmed cell death (apoptosis), which is accompanied by vascular invasion. Subsequently, the cartilage scaffold is replaced by bone. In the growing animal, this process is continued in the growth plate where the differentiation process is subject to tight temporal and spatial control. Mesenchymal cell differentiation and early chondrocyte proliferation occur in a columnar array inward from the articular surface. This spatial profile is extended as the chondrocytes become prehypertrophic, then hypertrophic. After the hypertrophic cells undergo apoptosis, the cartilaginous scaffold is remodeled and subsequently replaced by bone. The control of endochondral bone formation is maintained by a complex series of extracellular cues and intracellular signaling pathways. One of these factors is Indian hedgehog (Ihh), a member of the ancient hedgehog family of secreted patterning molecules. Ihh functions to promote chondrocyte proliferation and to maintain the pool of proliferating chondrocytes, thus extending the length of the differentiating cartilaginous growth plate prior to terminal

228

FIGURE 5

Regulation of chondrocyte differentiation by PTHrP and Indian hedgehog (Ihh). Ihh is secreted by postmitotic prehypertrophic chondrocytes and acts as a negative feedback regulator of the differentiation of proliferating chondrocytes. This occurs, in part, via the effect of Ihh to induce the expression of PTHrP in perichondrial cells. This occurs either directly or through an unknown mediator(s). PTHrP then acts directly on PTH/PTHrP receptors in proliferating chondrocytes to inhibit their differentiation. There is also evidence that Ihh inhibits chondrocyte proliferation through a PTHrP-independent mechanism.

differentiation and ossification [146]. Ihh is produced by postmitotic prehypertrophic chondrocytes, suggesting that the factor may serve as a negative feedback signal that slows the rate of transition of chondrocytes from the proliferative to the prehypertrophic pool. Ihh also appears to directly act on cells of the osteoblast lineage to promote their differentiation to mature bone-forming cells [146]. PTHrP appears to mediate some, but not all, of the actions of Ihh on endochondral bone formation (Fig. 5). PTHrP directly inhibits the differentiation of proliferating chondrocytes to postmitotic prehypertrophic cells. Lack of PTHrP results in accelerated chondrocyte differentiation with shortened growth plates and premature ossification. The cellular composition of the growth plates of PTHrP / animals is abnormal, with a marked reduction in the number of proliferating chondrocytes. Conversely, overexpression of PTHrP in chondrocytes of mice bearing a collagen II promoter-PTHrP transgene resulted in a distinct form of chondrodysplasia, which is characterized by shortlimbed dwarfism and delayed ossification [147]. At birth, these animals displayed a cartilaginous endochondral skeleton, and histological evaluation revealed a marked suppression of the chondrocyte differentiation program. By 7 weeks of age ossification was evident, but the long bones remained foreshortened and misshapen. Similar abnormalities are seen in humans with hereditary Jansen’s metaphyseal chondrodysplasia. The latter disorder has been associated with mutations in the PTH/PTHrP receptor that result in constitutive receptor activation [148]. Ihh acts directly or indirectly on cells in the periarticular perichondrium to increase expression of the PTHrP gene [149]. The effect of Ihh (or the related protein Sonic hedgehog) to delay terminal differentiation of chondrocytes in the long bones was not seen in PTHrP / or in PTH/PTHrP

ROBERT A. NISSENSON

receptor / mice, indicating an intermediary role of PTHrP in Ihh action in endochondral bone formation [149,150]. Consistent with this conclusion, a type II collagen promoter-driven constitutively active PTH/PTHrP receptor transgene has been reported to rescue the abnormally accelerated chondrocyte differentiation program in Ihh / mice [151]. These animals nonetheless displayed short-limbed dwarfism and decreased chondrocyte proliferation, demonstrating that PTHrP is not the only mediator of the multiple actions of Ihh on endochondral ossification. This conclusion is further supported by the observation that the severity of short-limbed dwarfism is much more severe in Ihh /, PTHrP / mice than in Ihh /, PTHrP / mice [151]. Solid evidence shows that the PTH/PTHrP receptor is responsible for initiating the actions of PTHrP on the differentiation of growth plate chondrocytes. The PTH/PTHrP receptor is expressed in proliferating chondrocytes as well as in cells in the transitional zone between proliferating and hypertrophic chondrocytes, where the regulation of terminal differentiation occurs [152]. PTH/PTHrP receptor / mice display growth plate abnormalities similar to those seen in PTHrP / mice [150]. Patients with inherited mutations in the PTH/PTHrP receptor that cause constitutive (i.e., ligand-independent) signaling (Jansen’s metaphyseal chondrodysplasia) display growth plate abnormalities similar to those seen in mice overexpressing a collagen II promoter-PTHrP transgene [153]. Lack of expression of functional PTH/PTHrP receptors in humans is associated with Blomstrand chondrodysplasia [154,155], a lethal disorder characterized by premature endochondral ossification [156]. Precisely how signaling by the PTH/PTHrP receptor results in the delay of chondrocyte differentiation in the transitional zone is unclear. It is known that programmed cell death (apoptosis) occurs during the late terminal differentiation of chondrocytes. This process has been shown to be inhibited by PTHrP, which upregulates the antiapoptotic protein bcl-2 through a cyclic AMP-dependent mechanism [157]. Mice lacking expression of a functional bcl-2 gene are known to display accelerated differentiation of growth plate chondrocytes, although the severity of the phenotype is much less than that seen in PTHrP / mice. Thus, inhibition of apoptosis may be one of several pathways involved in PTHrP-induced suppression of chondrocyte differentiation. 2. MAMMARY GLAND DEVELOPMENT Targeted overexpression of PTHrP in mammary myoepithelial cells of transgenic mice provided direct evidence of a possible role for PTHrP in mammary gland development [158]. The mammary ducts of 18- to 21-day-old transgenic mice were normal in terms both in the size of the ducts and in the branching morphogenesis of the developing gland. However, by 6 weeks of age, the transgenic animal displayed a delay in the development of the mammary duct

229

CHAPTER 7 PTH and PTH-Related Protein

system and a reduction in the degree of ductal branching. Pregnant transgenic animal displayed similar defects, as well as a diminished formation of terminal ductules. Overexpression of PTH in mammary myoepithilial cells of transgenic mice produced identical morphogenetic defects, indicating that this action of PTHrP is mediated by the PTH/PTHrP receptor. The postnatal role of PTHrP in mammary gland development was studied in PTHrP / mice expressing a PTHrP transgene targeted to cartilage [159], allowing postnatal survival. At 4 months of age, female transgenic mice lack mammary glands. The mammary fat pads appear normal, but mammary epithelial ducts are missing. PTHrP / mice display arrest of mammary duct development beginning between days 15 and 18 of embryogenesis. At this time, there is a degeneration of epithelial elements within the ducts, and the initiation of normal branching morphogenesis of the mammary glands does not occur. In normal animals, PTHrP is expressed in mammary epithelial cells [159,160], whereas functional PTH/PTHrP receptors are expressed in the underlying mesenchyme [159,161]. This pattern of expression suggests that PTHrP is an epithelial signal that acts on PTH/PTHrP receptors in mesenchymal cells to promote mammary epithelial morphogenesis. Consistent with this notion, PTH/PTHrP receptor / mice display the same defects in embryonic mammary development seen in PTHrP / mice. Moreover, normal morphogenesis requires PTH/PTHrP receptor expression specifically in mammary mesenchymal cells [161]. The factors that regulate epithelial production of PTHrP, and the nature of the mesenchymal targets of PTH/PTHrP receptor signaling, are unknown. The mesenchymal genes encoding tenascin C and the androgen receptor are induced by PTHrP [162]. PTHrP / or PTH/PTHrP receptor / male mice fail to display the normal androgen-dependent apoptotic destruction of the mammary bud, suggesting that induction of the androgen receptor by PTHrP is essential for sexual dimorphism during mammary development. 3. SKIN AND TOOTH DEVELOPMENT Keratinocytes were the first normal cells shown to express PTH-like bioactivity [163] and subsequently the PTHrP gene [163]. PTHrP is expressed in the basal layer through the granulosa layer of the skin, with epidermal expression detectable as early as day 14 of embryogenesis in the rat [164,165]. PTH/PTHrP receptors are present in dermal fibroblasts [166], and novel binding sites for PTHrP are detected in keratinocytes [167]. In cultured human keratinocytes, suppression of PTHrP production resulted in increased cell proliferation [168] and decreased differentiation [169]. Thus, PTHrP may have a role in the local regulation of epidermal cell proliferation and differentiation. Targeted overexpression of PTHrP in basal keratinocytes and outerroot sheath cells of hair follicles in transgenic

mice resulted in a failure of ventral hair eruption that was evident within 6 days after birth [170]. Dorsal hair was evident, but its eruption was delayed and the hairs were shorter and thinner compared to those of normal littermates. Histological evaluation of the transgenic mice revealed thickening of the ventral epidermis and expansion and increased cellularity of the dermis. Hair follicle development was delayed substantially in both ventral and dorsal skin of transgenic mice. These effects are probably due to disruption of the normal epithelial – mesenchymal interactions required for proper hair follicle development and epidermal differentiation. PTHrP / mice that have been rescued by expression of a type II collagen-PTHrP transgene display thinning of the epidermis with hypoplastic sebaceous glands and thinning of hair [171]. These abnormalities could be reversed by targeted expression of PTHrP in skin, indicating that PTHrP expression in basal keratinocytes is necessary for maintaining normal epithelial – mesenchymal interactions during epidermal differentiation. Inhibition of PTHrP action in skin was found to produce an increase in the number of follicles involved in active hair growth, further supporting a role of PTHrP in promoting hair follicle development [172]. PTHrP apparently maintains the pool of proliferating keratinocytes by suppressing their terminal differentiation. However, the underlying mechanisms responsible for regulating expression of PTHrP and the presumed mesenchymal responses remain obscure and are not understood at the molecular level. The fibroblast growth factor (FGF)-like factor keratinocyte growth factor (KGF) has been reported to be induced by PTHrP in dermal fibroblasts [173]. Because KGF is a critical regulator of keratinocyte growth and differentiation, it could be a mediator of PTHrP action in skin. PTHrP / mice also show cranial chondrodystrophy with a failure in normal tooth eruption [174]. In normal animals, PTHrP is expressed in the enamel epithelium, whereas the PTH/PTHrP receptor is expressed in the adjacent dental mesenchyme and in alveolar bone. These findings suggest that PTHrP is a regulator of epithelial – mesenchymal interactions during tooth development, as well as promoting the resorption of alveolar bone that is required for normal tooth eruption. 5. OTHER ACTIONS OF PTHRP PTHrP is expressed in a variety of smooth muscles where it functions as a local muscle-relaxing agent. Increased intraluminal pressure (either from muscle contraction or from expanding intraluminal contents) is a known stimulus for PTHrP gene expression. Myometrial expression of PTHrP peaks just before the end of pregnancy, and this effect is specific for the pregnant uterine horn in unilaterally pregnant animals [175]. Mechanotransduction is likely to be the primary stimulus, as physical stretch induces PTHrP expression in the nonpregnant rat uterus [176].

230 Human amniotic fluid contains high levels of PTHrP [177,178], and it is possible that PTHrP produced in the amnion plays a role in suppressing myometrial contractions and/or in regulating chorionic blood flow. PTHrP is also expressed in gastric and bladder smooth muscle and promotes muscle relaxation in these tissues in response to distension [179,180]. Pharmacological doses of PTH can reproduce the relaxing effects of PTHrP, strongly indicating the involvement of the PTH/PTHrP receptor. PTHrP has effects on both the contractility and the proliferation of vascular smooth muscle. PTHrP is widely expressed in vascular smooth muscle, and administration of PTHrP in vivo and in vitro elicits a vasodilitory response [181,182]. Expression of PTHrP in vascular smooth muscle is increased in experimental models of hypertension and in response to vasoconstrictors such as angiotensin II [183,184]. Targeted overexpression of PTHrP in vascular smooth muscle of transgenic mice results in decreased baseline blood pressure as well as in a diminished hypotensive response to exogenous PTHrP, with the latter possibly due to desensitization [185]. The role of endogenous PTHrP is seen in transgenic mice overexpressing the PTH/PTHrP receptor in vascular smooth muscle [186]. These animals are hypotensive and (as expected) are hyperresponsive to exogenous PTHrP with respect to vasodilitation. PTHrP is also induced in the blood vessels bathing skeletal muscle after muscle stimulation, perhaps promoting new capillary formation in response to increased muscle contraction [187]. Taken together, these results implicate PTHrP as an important physiological regulator of static blood pressure and as a counterregulatory factor secreted in response to vasoconstriction. The genes encoding PTHrP and the PTH/PTHrP receptor are widely expressed in the central nervous system, with particularly high levels seen in cerebellar granule cells [188,189]. These cells also express high levels of L-type calcium channels, and expression of PTHrP appears to be induced by depolarization-induced calcium influx through these channels [190]. Cerebellar granule cells are subject to excitatory cell death in response to agents such as kainic acid, which trigger calcium entry through L-type calcium channels. PTHrP blocks this excitatory cell death by inhibiting L-type calcium channel activity through a mechanism that probably involves cyclic AMP signaling via the PTH/PTHrP receptor [191]. This is consistent with previous reports that exogenous PTH inhibits L-type calcium channel activity [192]. These findings suggest that PTHrP functions as a neuronal survival factor produced in response to neuroexcitatory stimuli. Addition of a blocking antibody to PTHrP prevents cerebellar granule cell survival under depolarizing conditions, strongly suggesting that PTHrP is the endogenous factor responsible for neuroprotection [193]. As discussed earlier, PTHrP is expressed in the myometrium during pregnancy in response to distension

ROBERT A. NISSENSON

produced by the growing fetus. By inducing relaxation of uterine smooth muscle, locally produced PTHrP permits progressive intrauterine growth of the fetus and may also assist in maintaining the uterus in a quiescent state until the onset of parturition. PTHrP also plays an important role in the fetal-placental unit during pregnancy. The protein is expressed in human amniotic tissue and may serve to increase chorionic blood flow [177,178]. A role for PTHrP in placental calcium transport is suggested by studies demonstrating that the loss of the positive maternal – fetal placental calcium gradient produced by parathyroidectomy of fetal sheep could be restored by perfusion of the placenta with PTHrP [194]. The fetal parathyroid gland is a site of expression of PTHrP [195], suggesting that this might be the source of PTHrP responsible for maintaining the positive maternal – fetal calcium gradient. Moreover, relative to wild type littermates, PTHrP / fetuses are hypocalcemic and have a reduced ability to accumulate calcium from the mother’s circulation [196]. Thus, fetal PTHrP appears to have an endocrine action to maintain sufficient fetal calcium levels for skeletal growth and mineralization. A role for PTHrP during lactation was first suggested by the observation that suckling is a powerful stimulator of mammary PTHrP gene expression [197]. However, the precise physiological role of PTHrP during the period of lactation remains controversial. There is evidence for the existence of a factor, distinct from PTH and vitamin D metabolites, that is capable of maintaining plasma calcium homeostasis in lactating women. Indeed, systemic maternal PTHrP levels have been reported to increase during suckling [198] and to be elevated during lactation [199]. However, others have reported that circulating levels of PTHrP are unchanged during lactation [200,201]. A further complication is the observation that extremely large quantities of PTHrP are secreted into milk rather than into the bloodstream during lactation [200]. Suckling animals and humans thus ingest large amounts of PTHrP over an extended time period. However, evidence that milk-derived PTHrP is absorbed in an active form and/or is physiologically important in suckling infants or animals is lacking. It may be that suckling-induced expression of PTHrP has a local effect (e.g., increased blood flow) that facilitates mammary function during lactation.

IV. MECHANISM OF ACTION OF PTH AND PTHrP A. Signal Transduction Many of the actions of PTH and PTHrP are initiated by binding of these proteins to the PTH/PTHrP receptor, a Gprotein-coupled receptor that activates two G-proteins and

CHAPTER 7 PTH and PTH-Related Protein

FIGURE 6

Signal transduction by the PTH/PTHrP receptor. PTH and PTHrP bind to determinants in the extracellular domain and in the body of the receptor. This leads to conformational changes in the transmembrane helices and consequent structural changes in the cytoplasmic domain. The latter permit productive interaction between the receptor and the G-proteins Gs and Gq, activating the adenylyl cyclase (AC) and phospholipase C (PL-C) signaling pathways, respectively. These pathways are thought to cooperate in determining the cellular response to the receptor activation. Most available evidence supports a primary role of the cyclic AMP/protein kinase A (PK-A) pathway in mediating the biological effects of PTH/PTHrP receptor activation, with the PL-C pathway playing a modulatory role.

thereby two major signal transduction pathways (Fig. 6). Shortly after the discovery of the cyclic AMP signaling pathway, it was found that PTH is capable of increasing levels of cyclic AMP in target cells through activation of the enzyme adenylyl cyclase [202 – 205]. Cyclic AMP is a second messenger in the cellular action of a wide variety of hormones and other extracellular regulatory molecules. It activates cyclic AMP-dependent protein kinase (PK-A), which in turns phosphorylates and thereby regulates key proteins that participate in physiological responses. Very little is known about the identity of substrates of PK-A that are phosphorylated in response to PTH/PTHrP receptor activation. These presumably include transcription factors, ion channels, transporters, and enzymes involved in cellular metabolism. PTH/PTHrP receptors also activate phospholipase C, an enzyme that hydrolyzes the plasma membrane phospholipid phosphatidylinositol-4,5-bisphosphate to produce diacylglycerol (DG) and soluble 1,4,5-inositol trisphosphate (IP3). DG and IP3 function as second messengers, the former by activating protein kinase-C (PK-C) and the

231 latter by binding to and opening calcium channels on the membrane of the endoplasmic reticulum, thereby increasing cytosolic-free calcium. The PTH/PTHrP receptor is clearly required for PTHstimulated bone resorption [206], and a number of studies have been carried out to identify the nature of the relevant signaling pathway(s). Agents that raise cellular cyclic AMP levels (e.g., analogues of cyclic AMP, forskolin) are capable of eliciting bone resorption in organ culture [207 – 211]. In addition, inhibition of cyclic AMP phosphodiesterase (thus augmenting the cellular cyclic AMP response to PTH) potentiates PTH-induced bone resorption [212]. Activation of phospholipase C-related pathways with calcium ionophores and phorbol esters also promotes bone resorption in organ culture [213 – 215], and inhibition of protein kinase C is reported to block PTH-stimulated bone resorption [216,217]. However, at least in mouse calvarial cultures, the effects of calcium ionophores and phorbol esters require the intermediary synthesis of prostaglandins, whereas PTH-induced bone resorption does not [218]. Moreover, under some circumstances, these agents can inhibit bone resorption [219-221]. Thus, available evidence indicates that the cyclic AMP pathway plays a primary second messenger role in the stimulation of bone resorption by PTH. PTH-induced differentiation of hematopoietic precursors to osteoclast-like cells also involves the cyclic AMP pathway [222 – 224], although the phospholipase C pathway may also contribute [225]. In primary cultures of human bone marrow stromal cells, the cyclic AMP pathway has been shown to downregulate expression of the OPG gene [226], which could allow for greater bone resorption in response to PTH-induced production of RANKL. From earlier studies of PTH-induced bone resorption, it is expected that the cyclic AMP pathway would promote expression of RANKL in osteoclast precursors. However, studies of the RANKL promoter failed to demonstrate a stimulatory effect of cyclic AMP [227], indicating that effects of this pathway on RANKL expression are likely to be indirect. There has been great interest in defining the signaling events that are responsible for the anabolic response of the skeleton to intermittent administration of PTH. Progress in this area has been hampered by the lack of a useful in vitro model system for the investigation of the anabolic response to PTH and the uncertainty about the cellular basis of this effect. PTH generally has been reported to have an antiproliferative effect on cultured osteoblasts, although it is reported to promote proliferation in an osteoblast precursor model [228]. PTH can also promote osteoblast differentiation in vitro, depending on the time and duration of treatment [98,99,229]. In vivo studies have demonstrated that amino-terminal fragments of both PTH and PTHrP are anabolic, implicating the PTH/PTHrP receptor as the likely initiator of this skeletal response. Interestingly, PTH(1 – 30)

232 and PTH(1 – 31), which activate adenylyl cyclase but have a greatly reduced ability to activate phospholipase C, are effective as anabolic agents in bone [230,231]. This result suggests that the cyclic AMP pathway is the major mediator of the anabolic actions of PTH. However, it should be noted that cyclic AMP signaling has generally been linked to the inhibition of osteoblast proliferation and differentiation in vitro [232 – 235]. This finding suggests that these in vitro model systems of osteoblast proliferation and differentiation may not be relevant to the mechanism of PTH-induced anabolism. Alternatively, signaling pathways other then the cyclic AMP system may participate in the anabolic response to PTH. Microdissection studies revealed the presence of PTHstimulated cyclic AMP generation in the proximal convoluted tubule where sodium-dependent phosphate cotransport occurs [236,237]. Analogues of cyclic AMP were found to be effective in reproducing the phosphaturic effect of PTH [238 – 241]. In pseudohypoparathyroidism Ia, genetic deficiency of Gs- (a protein that links receptors to the activation of adenylyl cyclase) is associated with resistance to the phosphaturic action of PTH [242 – 245]. With the discovery that an opossum kidney cell line (OK) retains PTH receptors [246] and PTH-inhibited sodium – phosphate cotransport [247], it became possible to carry out studies on the mechanisms of PTH inhibition of phosphate transport. Cyclic AMP clearly has a primary, although not exclusive, role in the negative regulation of sodium – phosphate cotransport [247 – 250] [251]. Cyclic AMP (like PTH) promotes rapid downregulation of the type II sodium – phosphate cotransporter in OK cells via enhanced transporter endocytosis and lysosomal degradation [106,109,111,252]. Activation of PK-C by the PTH/PTHrP receptor may also contribute to the inhibition of phosphate transport, as treatment of OK cells with PMA or other phorbol esters substantially inhibits sodium – phosphate cotransport and reduces the expression of the type II cotransporter in some [249,253 – 256] but not all [252] studies. The cyclic AMP pathway is known to be important in mediating the effect of PTH to increase the activity of the 25(OH)D-1-hydroxylase in the proximal renal tubule [122,257,258]. PTH has a positive effect on the renal expression of the 1-hydroxylase mRNA in vivo [126,127]. This appears to occur at the level of gene transcription, as upstream elements in the 5-region of the 1-hydroxylase gene confer transcriptional responses to PTH and forskolin in cultured kidney cells [259 – 261]. The precise elements in the promoter responsible for these effects have not been identified, but putative binding sites for the transcription factors CREB and AP-1 are present and represent possible targets. PTH-stimulated phospholipase C activation might also contribute to the 1-hydroxylase response, as the combination of a calcium ionophore and PMA was shown to promote a sustained increase in 1,25(OH)2D production in

ROBERT A. NISSENSON

perifused rat proximal tubule cells [262]. Under some circumstances, inhibitors of PK-C have been shown to suppress PTH-induced renal production of 1,25(OH)2D [263]. In light of these findings, it is possible that phospholipase C has a role in the transcriptional response of the 1-hydroxylase gene to PTH. The PTH-induced stimulation of renal calcium transport in the distal convoluted tubule appears to require activation of both PK-A and PK-C pathways [113]. Inhibition of either of these kinases suppresses PTH-induced calcium uptake by distal tubular cells [264]. Moreover, simultaneous activation of both kinases was shown to be necessary and sufficient to reproduce the effect of PTH on calcium uptake [265]. PTH does not appear to increase the activity of phospholipase C in the distal renal tubule [266], suggesting that an alternative mechanism exists for the PTH-induced generation of diacylglycerol. In this regard, PTH is capable of increasing the activity of phospholipase D, an enzyme that hydrolyzes phosphatidyl choline to produce phosphatidic acid and, indirectly, diacylglycerol [266,267]. It is possible that the activation of phospholipase D participates in the activation of PK-C that is reported to occur in response to PTH as well as amino-terminally truncated PTH fragments [268]. There is as yet very little information concerning the signaling mechanisms responsible for mediating the developmental and morphogenetic actions of PTHrP. It is likely that the cyclic AMP signaling pathway is of primary importance, and genetic deficiency of the -subunit of Gs produces a constellation of developmental abnormalities (Abright’s hereditary osteodystrophy) that overlap those seen in animals lacking PTHrP or the PTH/PTHrP receptor. However, the precise role of adenylyl cyclase and phospholipase C signaling pathways in mediating specific paracrine actions of PTHrP remains to be defined.

B. PTH/PTHrP Receptors 1. ACTIVATION OF G-PROTEINS Early studies on the PTH/PTHrP receptor demonstrated a prominent role for GTP and its analogues in regulating ligand – receptor affinity and signaling, suggesting that this receptor couples to GTP-binding (G) proteins [269 – 274]. The cloning of the cDNA encoding the PTH/PTHrP receptor [275] revealed a predicted protein sequence containing seven putative membrane-spanning domains (Fig. 7), a topology characteristic of members of the GPCR superfamily [276,277]. In the case of the PTH/PTHrP receptor, the major G-proteins that can be activated are Gs and Gq. Activation of Gs leads to increased adenylyl cyclase activity, resulting in increased cellular levels of cyclic AMP and activation of PK-A. Activation of Gq results in the stimulation of phospholipase

233

CHAPTER 7 PTH and PTH-Related Protein

FIGURE 7 Structural model of the PTH/PTHrP receptor indicating the presence of seven membrane-spanning helices, which surround a central polar cavity. The receptor contains a large, glycosylated N-terminal extracellular domain and a long C-terminal cytoplasmic tail. Agonist binding to the receptor alters the relative orientation of the transmembrane helices, promoting the activation of specific G-proteins. See text for details.

C, resulting in the mobilization of intracellular calcium and activation of PK-C. Preference of the PTH/PTHrP receptor for the cyclic AMP signaling pathway is suggested by studies on PTH target cells in vitro, where the activation of adenylyl cyclase generally occurs at lower concentrations of added PTH than activation of phospholipase C [278]. These findings are consistent with the observations that the cyclic AMP pathway is most closely associated with most of the physiological effects of PTH on bone and kidney, with activation of phospholipase C playing a modulatory role. 2. RECEPTOR ACTIVATION MECHANISMS When the cDNA sequence of the PTH/PTHrP receptor was first delineated [275], it was apparent that it encoded a protein with a predicted overall structure consistent with those of other known GPCRs. In particular, the receptor was modeled as containing seven membrane-spanning helices, with a large amino-terminal extracellular domain, three extracellular loops, three intracellular loops, and a large carboxy-terminal cytoplasmic tail (Fig. 7). However, the PTH/PTHrP receptor does not share a number of the specific sequence motifs present in the largest subfamily of GPCRs (the so-called class I family that includes receptors

for a diverse group of ligands ranging from photons to polypeptide hormones). Rather, the PTH/PTHrP receptor is a member of a second GPCR subfamily (class II) that includes receptors for calcitonin, glucagon, and a number of other polypeptide ligands. Members of the class II GPCR subfamily are presumed to share a common basic mechanism of G-protein activation, but have evolved determinants of specificity that permit binding and activation by only the appropriate peptide ligand. Mutagenesis studies have been carried out to investigate the structural features in the PTH/PTHrP receptor that are important for agonist binding and for maintaining receptor specificity. These studies have demonstrated that the large amino-terminal extracellular domain of the receptor contains critical determinants of agonist binding affinity [279 – 281]. However, the body of the receptor, which includes the extracellular loops and the transmembrane domains, also plays a role in ligand binding as well as in maintaining ligand specificity [280,282 – 284]. In a recent series of elegant biochemical studies, sites of interaction between amino-terminal PTH fragments and the PTH/PTHrP receptor have been mapped. These studies have demonstrated multiple points of contact between the 1 – 34 region of PTH/PTHrP and the receptor. Specifically, there is interaction between position 23 in the ligand and the extreme N terminus of the extracellular domain of the receptor [285]; between amino 13 of the ligand and the membrane-proximal portion of the N terminal extracellular domain of the receptor [286]; and between the N-terminus of the ligand and the extracellular end of the sixth transmembrane domain of the receptor [287]. This latter interaction is presumably required to initiate the conformational shift in the transmembrane domain of the receptor that is required for signal transduction [288]. This involves the exposure of key amino acids in the second and third cytoplasmic loops of the PTH/PTHrP receptor that are required for activation of Gs and Gq [289,290]. 3. RECEPTOR REGULATION Signal transduction by GPCRs is generally subject to strict regulatory control. This control can occur in response to agonist binding (homologous regulation) or in response to factors acting though separate pathways (heterologous regulation). Acute control of signaling is accomplished by blocking the ability of agonist-occupied receptors to sustain the activation of G-proteins (desensitization) and by physically moving the receptors into an intracellular compartment effectively separating them from G-proteins (sequestration). Chronic regulation of receptor signaling is accomplished by agonist-induced changes in steady-state levels of expression of receptors due to increased receptor catabolism following receptor internalization (downregulation) and to changes in de novo receptor synthesis. Homologous regulation commonly involves all of these mechanisms, whereas

234 heterologous regulation most often occurs through changes in steady-state levels of receptor expression. Many studies have documented homologous regulation of PTH/PTHrP receptor signaling. Treatment of cultured bone and kidney cells with PTH generally dampens the adenylyl cyclase and phospholipase C responses to a second addition of the hormone [291 – 300]. In most studies, desensitization of the PTH response occurs rapidly, within minutes of initial exposure to PTH, suggesting that the PTH/PTHrP receptor has become acutely uncoupled from its cognate G-proteins. The mechanisms underlying acute desensitization have been well studied for GPCRs such as rhodopsin and -adrenergic receptors [301 – 303]. The major mechanism underlying acute desensitization of these receptors is phosphorylation of the cytoplasmic domain of the receptor by a GPCR kinase (GRK). GRKs are serine/threonine kinases that phosphorylate only the agonist-occupied receptor, and phosphorylation facilitates the interaction of the receptor with a member of the arrestin protein family. Arrestin binding to the receptor sterically interferes with the interaction between the receptor and G-proteins, thus preventing signal transmission. There is mounting evidence that a similar mechanism applies to desensitization of PTH/PTHrP receptor signaling. The PTH/PTHrP receptor is subject to phosphorylation in response to agonist binding [304,305], and this appears to occur largely if not exclusively on serine residues in the cytoplasmic tail [305 – 307]. The kinase involved appears to be a member of the GRK family, possibly GRK2 [306,308], and a dominant-inhibitor of GRK function can suppress PTH/PTHrP receptor desensitization in human osteoblast-like cells [309]. Long-term treatment with PTH results in a loss of cellular PTH/PTHrP receptors (downregulation) and a corresponding reduction in the maximal signaling response to the hormone [299,310 – 313]. Evidence shows that this process may have pathophysiological relevance. For example, vitamin D deficiency can be associated with target cell resistance to PTH [314 – 316]. In animal studies, this resistance can be reversed by parathyroidectomy, suggesting that it is the secondary hyperparathyroidism that is responsible for target cell resistance [317]. Infusion of PTH to levels seen in severe secondary hyperparathyoidism produces downregulation of PTH/PTHrP receptors and a reduction in the adenylyl cyclase response to PTH [310]. In chronic renal failure, factors other than hyperparathyroidism may also contribute to reduced target cell expression of PTH/PTHrP receptors [318]. The initial step in downregulation of PTH/PTHrP receptors appears to be agonist-induced accumulation of the receptor in plasma membrane clathrincoated pits [31,319]. These pits are endocytic organelles that pinch off from the plasma membrane, thus becoming endocytic vesicles. Once internalized, PTH/PTHrP receptors can be recycled to the plasma membrane or can presumably progress further down the endocytic pathway to

ROBERT A. NISSENSON

the lysosomes for degradation. The molecular mechanisms underlying the agonist-induced internalization of the PTH/PTHrP receptor are not entirely clear. Agonist-stimulated receptor phosphorylation may play a role in osteoblastic cells [309], although receptor phosphorylation is not required for endocytosis in some cellular settings [307]. Arrestins have been implicated as mediators of GPCR endocytosis, and studies indicate that arrestins can become associated with the PTH/PTHrP receptor following agonist binding [320]. In addition, the cytoplasmic tail of the PTH/PTHrP receptor contains a tyrosine-based sequence that has been implicated in promoting the internalization of other membrane receptors. Mutation of this sequence markedly inhibits agonist-induced PTH/PTHrP receptor endocytosis [319]. Further studies are needed to define the relative contribution of these mechanisms to PTH/PTHrP receptor internalization and downregulation. Another mechanism for the regulation of PTH/PTHrP receptor levels is through changes in expression of the receptor gene. In osteoblastic cells, PTH is reported to decrease levels of PTH/PTHrP receptor mRNA by a mechanism involving the cyclic AMP pathway [321,322]. This may be due to direct transcriptional activation of the PTH/PTHrP receptor gene by PK-A-activated transcription factors, but the details of this pathway have yet to be elucidated. Homologous control of PTH/PTHrP receptor expression appears to be target cell specific in that PTH reportedly does not reduce expression of the PTH/PTHrP receptor gene in the kidneys of rats with secondary hyperparathyroidism [318,323]. Heterologous factors are also reported to regulate levels of PTH/PTHrP receptor expression in bone and kidney. The cytokine TGF upregulates the expression of the PTH/PTHrP receptor in osteoblastic osteosarcoma cells [324], although the opposite effect is reported in primary cultures of fetal rat osteoblasts [325] and in OK cells [326]. Dexamethasone treatment produces an increase in expression of the PTH/PTHrP receptor in osteoblastic cells, but not in kidney cells [327,328], whereas 1,25(OH)2D downregulates expression of the PTH/PTHrP receptor gene [329]. It should be pointed out that most of these studies have been carried out in cultured bone and kidney cells in vitro, and more needs to be done to establish the physiological relevance of the changes in PTH/PTHrP receptor gene expression.

C. Nontraditional Mechanisms of Action of PTHrP The discovery of PTHrP was based on the PTH-like endocrine actions of this peptide in patients with malignancyassociated hypercalcemia. The classical mechanism of action of PTHrP is thus to bind to and activate the widely expressed PTH/PTHrP receptor. The amino-terminal 1 – 34 domain of

235

CHAPTER 7 PTH and PTH-Related Protein

FIGURE 8

Potential mechanisms of action of PTHrP. The majority of the actions of PTHrP result from the binding of the amino-terminal portion of the protein to the PTH/PTHrP receptor, leading to the activation of adenylyl cyclase (AC) and phospholipase C (PLC). Activation of these effector enzymes results in increased cellular levels of cyclic AMP (cAMP), intracellular calcium, and protein kinase C (PKC). PTHrP is also processed posttranslationally, producing midregion and carboxyl-terminal fragments of the protein. These fragments have cellular effects that are presumably mediated by novel membrane receptors (Rm and Rc), acting through unknown signaling pathways. PTHrP has also been localized to the nucleus of cells (intracrine action) where it may regulate nuclear functions such as mitosis, apoptosis, and RNA processing.

PTHrP is responsible for binding to the PTH/PTHrP receptor, thus initiating signal transduction. However, it appears that the PTH/PTHrP receptor does not mediate all of the physiological actions of PTHrP. Two additional mechanisms have been identified by which PTHrP can potentially influence cellular function (Fig. 8). One involves the notion of PTHrP as a polyhormone that yields mid- and carboxyl-region fragments with distinct biological activities that are presumably mediated by novel cell surface receptors. The second mechanism relates to the ability of PTHrP to translocate to the nucleus of cells in which it is expressed, thereby altering cell proliferation and/or gene expression. 1. PTHRP AS A POLYHORMONE The PTHrP gene is subject to alternative splicing, resulting in multiple protein products (ranging from 139 to 173

amino acids) that differ only in the extent of their C termini [142]. Only the amino-terminal 34 amino acids are needed to produce all of the PTH-like actions of PTHrP on the PTH/PTHrP receptor, and several groups have been interested in assessing a possible biological role for the remainder of the molecule. Indeed, PTHrP is subject to posttranslational proteolytic processing to produce a midregion fragment (amino acids 38 – 94) and a C-terminal fragment (amino acids 107 – 139) as well as PTHrP(1 – 36) [140]. Fragments of PTHrP are secreted by some cells, at least in vitro, and thus have the potential to elicit biological responses in a paracrine or endocrine fashion. Synthetic PTHrP(107 – 139) has been reported to elicit biological effects such as inhibition of bone resorption [330], stimulation of osteoblast proliferation [331], and stimulation of interleukin-6 expression in osteoblasts [332]. The nature of the receptor and signaling pathway responsible for these actions of PTHrP is unclear, although the latter effect appeared to involve activation of protein kinase C. A physiological role for PTHrP fragments is suggested by studies of placental calcium transport. As mentioned earlier, the fetal parathyroid gland is required to maintain the normal positive maternal – fetal calcium gradient, at least in sheep. This gradient can be restored in parathyroidectomized fetuses by the administration of midregion fragments of PTHrP, but not by PTH or by amino-terminal PTHrP fragments [196,333]. This effect must therefore be initiated by a receptor distinct from the classical PTH/ PTHrP receptor. 2. INTRACRINE ACTIONS OF PTHRP Several studies have demonstrated that, once synthesized, PTHrP can localize to the nucleolus as well as being secreted. Nucleolar localization requires the presence of a targeting signal in the carboxyl region of the molecule [334] and occurs through an interaction with the targeting protein importin  [335]. Secreted PTHrP can also be taken up by cells and translocated to the nucleus, and this appears to involve a receptor distinct from the PTH/PTHrP receptor [336]. Although the functional significance of nuclear PTHrP has yet to be definitely established, a number of intriguing findings have been reported. Intracellular expression of PTHrP has been shown to protect chondrocytes from apoptosis induced by serum deprivation, and this effect was dependent on the presence of an intact nucleolar localization signal [337]. Nuclear localization of PTHrP is associated with mitogenesis in cultured vascular smooth muscle cells [338]. The mechanisms underlying this mitogenic effect are not entirely clear. In cultured keratinocytes, PTHrP is present in the nucleolus during the G1 phase of the cell cycle, but redistributes to the cytoplasm during cell division [339]. Interestingly, PTHrP is phosphorylated by the cell cycle regulatory kinase CDC2-CDK2, and this appears to promote translocation of the PTHrP from the

236

ROBERT A. NISSENSON

nucleus to the cytoplasm [340]. It is possible that PTHrP acts, at least in part, through direct interaction with ribonucleoprotein complexes, as PTHrP is capable of binding directly to RNA via a polybasic region within the nuclear localization signal [336]. Further work is needed to more clearly define the significance of this unusual mode of action of PTHrP.

Acknowledgments I am grateful to Margaret Bencsik for her skillful assistance in the preparation of this manuscript. Portions of the work discussed here were supported by NIH Grant DK35323 and by the Medical Research Service of the Department of Veterans’ Affairs. Dr. Nissenson is a Senior Research Career Scientist of the Department of Veterans’ Affairs.

References 1. E. M. Brown, Extracellular Ca2 sensing, regulation of parathyroid cell function, and role of Ca2 and other ions as extracellular (first) messengers. Physiol. Rev. 71, 371 – 411 (1991). 2. J. Russell, D. Lettieri, and L. M. Sherwood, Direct regulation by calcium of cytoplasmic messenger ribonucleic acid coding for preproparathyroid hormone in isolated bovine parathyroid cells. J. Clin. Invest. 72, 1851 – 1855 (1983). 3. T. Naveh-Many and J. Silver, Regulation of parathyroid hormone gene expression by hypocalcemia, hypercalcemia, and vitamin D in the rat. J. Clin. Invest. 86, 1313 – 1319 (1990). 4. T. Yamaguchi, N. Chattopadhyay, and E. M. Brown, G protein-coupled extracellular Ca2 (Ca2o)-sensing receptor (CaR): Roles in cell signaling and control of diverse cellular functions. Adv. Pharmacol. 47, 209 – 253 (2000). 5. E. M. Brown, Physiology and pathophysiology of the extracellular calcium-sensing receptor. Am. J. Med. 106, 238 – 253 (1999). 6. M. R. Pollak, C. E. Seidman, and E. M. Brown, Three inherited disorders of calcium sensing. Medicine 75, 115 – 123 (1996). 7. S. A. Berson and R. S. Yalow, Immunochemical heterogeneity of parathyroid hormone in plasma. J. Clin. Endocrino. Metab. 28, 1037 – 1047 (1968). 8. J. M. Canterbury and E. Reiss, Fractionation of circulating parathyroid hormone (PTH) in man. J. Lab. Clin. Med. 78, 814 (1971). 9. C. D. Arnaud, Immunochemical heterogeneity of circulating parathyroid hormone in man: Sequel to an original observation by Berson and Yalow. Mount Sinai J. Med. 40, 422 – 432 (1973). 10. S. B. Oldham, E. J. Finck, and F. R. Singer, Parathyroid hormone clearance in man. Metab. Clin. Exp. 27, 993 – 1001 (1978). 11. G. V. Segre, H. D. Niall, J. F. Habener, and J. T. Potts, Jr., Metabolism of parathyroid hormone: Physiologic and clinical significance. Am. J. Med. 56, 774 – 784 (1974). 12. J. Fox, M. Scott, R. A. Nissenson, and H. D. Heath, Effect of plasma calcium concentration on the metabolic clearance rate of parathyroid hormone in the dog. J. Lab. Clin. Med. 102, 70 – 77 (1983). 13. S. K. Libutti, H. R. Alexander, D. L. Bartlett, M. L. Sampson, M. E. Ruddel, M. Skarulis, S. J. Marx, A. M. Spiegel, W. Simmonds, and A. T. Remaley, Kinetic analysis of the rapid intraoperative parathyroid hormone assay in patients during operation for hyperparathyroidism. Surgery 126, 1145 – 1150 (1999).

14. K. J. Martin, K. A. Hruska, J. J. Freitag, S. Klahr, and E. Slatopolsky, The peripheral metabolism of parathyroid hormone. N. Engl. J. Med. 301, 1092 – 1098 (1979). 15. K. Martin, K. Hruska, A. Greenwalt, S. Klahr, and E. Slatopolsky, Selective uptake of intact parathyroid hormone by the liver: Differences between hepatic and renal uptake. J. Clin. Invest. 58, 781 – 788 (1976). 16. K. A. Hruska, A. Korkor, K. Martin, and E. Slatopolsky, Peripheral metabolism of intact parathyroid hormone: Role of liver and kidney and the effect of chronic renal failure. J. Clin. Invest. 67, 885 – 892 (1981). 17. F. R. Bringhurst, A. M. Stern, M. Yotts, N. Mizrahi, G. V. Segre, and J. T. Potts, Jr., Peripheral metabolism of PTH: Fate of biologically active amino terminus in vivo. Am. J. Physiol. 255, E886 – E893 (1988). 18. J. A. Flueck, F. P. Di Bella, A. J. Edis, J. M. Kehrwald, and C. D. Arnaud, Immunoheterogeneity of parathyroid hormone in venous effluent serum from hyperfunctioning parathyroid glands. J. Clin. Invest. 60, 1367 – 1375 (1977). 19. S. Erdmann, W. Müller, S. Bahrami, S. I. Vornehm, H. Mayer, P. Bruckner, K. von der Mark, and H. Burkhardt, Differential effects of parathyroid hormone fragments on collagen gene expression in chondrocytes. J. Cell Biol. 135, 1179 – 1191 (1996). 20. M. Nasu, T. Sugimoto, H. Kaji, J. Kano, and K. Chihara, Carboxylterminal parathyroid hormone fragments stimulate type – 1 procollagen and insulin-like growth factor-binding protein – 5 mRNA expression in osteoblastic UMR-106 cells. Endoc. J. 45, 229 – 234 (1998). 21. T. Tsuboi and A. Togari, Comparison of the effects of carboxylterminal parathyroid hormone peptide[53 – 84] and aminoterminal peptide[1-34] on mouse tooth germ in vitro. Arch. Oral Bio. 43, 335 – 339 (1998). 22. N. Inomata, M. Akiyama, N. Kubota, and H. Jüppner, Characterization of a novel parathyroid hormone (PTH) receptor with specificity for the carboxyl-terminal region of PTH-(1-84). Endocrinology 136, 4732 – 4740 (1995). 23. H. Takasu, H. Baba, N. Inomata, Y. Uchiyama, N. Kubota, K. Kumaki, A. Matsumoto, K. Nakajima, T. Kimura, S. Sakakibara, T. Fujita, K. Chihara, and I. Nagai, The 69 – 84 amino acid region of the parathyroid hormone molecule is essential for the interaction of the hormone with the binding sites with carboxyl-terminal specificity. Endocrinology 137, 5537 – 5543 (1996). 24. R. R. MacGregor, J. W. Hamilton, G. N. Kent, R. E. Shofstall, and D. V. Cohn, The degradation of proparathormone and parathormone by parathyroid and liver cathepsin B. J. Biol. Chem. 254, 4428 – 4433 (1979). 25. J. W. Hamilton, R. L. Jilka, and R. R. MacGregor, Cleavage of parathyroid hormone to the 1-34 and 35-84 fragments by cathepsin D-like activity in bovine parathyroid gland extracts. Endocrinology 113, 285 – 292 (1983). 26. G. P. Mayer, J. A. Keaton, J. G. Hurst, and J. F. Habener, Effects of plasma calcium concentration on the relative proportion of hormone and carboxyl fragments in parathyroid venous blood. Endocrinology 104, 1778 – 1784 (1979). 27. R. Talmage, The demand for bone calcium in mainenance of plasma calcium concentration. In “Mechanisms of Localized Bone Loss” (R. Horton, T. M. Tarplay, W. F. Davis, eds.), pp. 73 – 92. Info Retrieval, Washington, DC, 1978. 28. R. A. M. Talmage, Physiological role of parathyroid hormone. In “Handbook of Physiology” (R. A. A. Grup, ed.), pp. 343 – 351. American Physiological Society, Washington, DC, 1967. 29. W. F. Neuman, M. W. Neuman, and C. R. Myers, Blood:bone disequilibrium. III. Linkage between cell energetics and Ca fluxes. Am. J. Physiol. 236, C244 – C248 (1979). 30. M. E. Holtrop and G. J. King, The ultrastructure of the osteoclast and its functional implications. Clin. Orthop. Rela. Res. 475, 177 – 196 (1977). 31. C. M. Silve, G. T. Hradek, A. L. Jones, and C. D. Arnaud, Parathyroid hormone receptor in intact embryonic chicken bone: Characterization and cellular localization. J. Cell Biol. 94, 379 – 386 (1982).

CHAPTER 7 PTH and PTH-Related Protein 32. M. F. Rouleau, H. Warshawsky, and D. Goltzman, Parathyroid hormone binding in vivo to renal, hepatic, and skeletal tissues of the rat using a radioautographic approach. Endocrinology 118, 919 – 931 (1986). 33. A. Teti, R. Rizzoli, and A. Zambonin Zallone, Parathyroid hormone binding to cultured avian osteoclasts. Biochem. Biophys. Res. Commun. 174, 1217 – 1222 (1991). 34. N. Agarwala and C. V. Gay, Specific binding of parathyroid hormone to living osteoclasts. J. Bone Miner. Res. 7, 531 – 539 (1992). 35. L. G. Rao, T. M. Murray, and J. N. Heersche, Immunohistochemical demonstration of parathyroid hormone binding to specific cell types in fixed rat bone tissue. Endocrinology 113, 805 – 10 (1983). 36. T. J. Chambers, P. M. McSheehy, B. M. Thomson, and K. Fuller, The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116, 234 – 239 (1985). 37. R. L. Jilka, Are osteoblastic cells required for the control of osteoclast activity by parathyroid hormone? Bone Miner. 1, 261 – 266 (1986). 38. S. S. Miller, A. M. Wolf, and C. D. Arnaud, Bone cells in culture: Morphologic transformation by hormones. Science 192, 1340 – 1343 (1976). 39. R. Civitelli, K. Ziambaras, P. M. Warlow, F. Lecanda, T. Nelson, J. Harley, N. Atal, E. C. Beyer, and T. H. Steinberg, Regulation of connexin43 expression and function by prostaglandin E2 (PGE2) and parathyroid hormone (PTH) in osteoblastic cells. J. Cell. Biochem. 68, 8 – 21 (1998). 40. P. C. Schiller, B. A. Roos, and G. A. Howard, Parathyroid hormone up-regulation of connexin 43 gene expression in osteoblasts depends on cell phenotype. J. Bone Miner. Res. 12, 2005 – 2013 (1997). 41. H. J. Donahue, K. J. McLeod, C. T. Rubin, J. Andersen, E. A. Grine, E. L. Hertzberg, and P. R. Brink, Cell-to-cell communication in osteoblastic networks: Cell line-dependent hormonal regulation of gap junction function. J. Bone Miner. Res. 10, 881 – 889 (1995). 42. J. E. Puzas and J. S. Brand, Parathyroid hormone stimulation of collagenase secretion by isolated bone cells. Endocrinology 104, 559 – 562 (1979). 43. J. K. Heath, S. J. Atkinson, M. C. Meikle, and J. J. Reynolds, Mouse osteoblasts synthesize collagenase in response to bone resorbing agents. Biochim. Biophys. Acta 802, 151 – 154 (1984). 44. Y. Eeckhout, J. M. Delaissé, and G. Vaes, Direct extraction and assay of bone tissue collagenase and its relation to parathyroid-hormoneinduced bone resorption. Biochem. J. 239, 793 – 796, (1986). 45. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 – 1962 (1987). 46. R. L. Jilka, Procollagenase associated with the noncalcified matrix of bone and its regulation by parathyroid hormone. Bone 10, 353 – 358 (1989). 47. J. A. Hamilton, S. Lingelbach, N. C. Partridge, and T. J. Martin, Regulation of plasminogen activator production by bone-resorbing hormones in normal and malignant osteoblasts. Endocrinology 116, 2186 – 2191 (1985). 48. G. Leloup, C. Peeters-Joris, J. M. Delaisse, G. Opdenakker, and G. Vaes, Tissue and urokinase plasminogen activators in bone tissue and their regulation by parathyroid hormone. J. Bone Miner. Res. 6, 1081 – 1090 (1991). 49. T. Nonaka, H. Matsumoto, W. Shimada, K. Okada, H. Fukao, S. Ueshima, H. Kikuchi, S. Tanaka, and O. Matsuo, Effect of bone resorbing factors on u-PA and its specific receptor in osteosarcoma cell line. Clin. Chim. Acta 223, 129 – 142 (1993). 50. W. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999).

237 51. M. C. de Vernejoul, M. Horowitz, J. Demignon, L. Neff, and R. Baron, Bone resorption by isolated chick osteoclasts in culture is stimulated by murine spleen cell supernatant fluids (osteoclastactivating factor) and inhibited by calcitonin and prostaglandin E2. J. Bone Miner. Res. 3, 69 – 80 (1988). 52. S. Amano, S. Hanazawa, K. Hirose, Y. Ohmori, and S. Kitano, Stimulatory effect on bone resorption of interleukin-1-like cytokine produced by an osteoblast-rich population of mouse calvarial cells. Calcif. Tissue Int. 43, 88 – 91 (1988). 53. M. C. Horowitz, D. L. Coleman, P. M. Flood, T. S. Kupper, and R. L. Jilka, Parathyroid hormone and lipopolysaccharide induce murine osteoblast-like cells to secrete a cytokine indistinguishable from granulocyte-macrophage colony-stimulating factor. J. Clin. Invest. 83, 149 – 157 (1989). 54. D. B. Evans, R. A. Bunning, J. Van Damme, and R. G. Russell, Natural human IL-1 beta exhibits regulatory actions on human bonederived cells in vitro. Biochem. Biophys. Res. Commun. 159, 1242 – 1248 (1989). 55. E. Jimi, I. Nakamura, H. Amano, Y. Taguchi, T. Tsurukai, M. Tamura, N. Takahashi, and T. Suda, Osteoclast function is activated by osteoblastic cells through a mechanism involving cellto-cell contact [corrected and republished with original paging, article originally printed in Endocrinology 137(5), 2187 – 2190 (1996)]. Endocrinology 137, 2187 – 2190 (1996). 56. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 57. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 58. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 59. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145, 527 – 538 (1999). 60. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/ osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc. Nat. Acad. Sci. USA 95, 3597 – 3602 (1998). 61. J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 62. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 63. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Lüthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, W. J. Boyle et al., Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997).

238 64. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 65. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 66. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 67. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 68. K. Tsukii, N. Shima, S. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Yano, O. Shibata, N. Udagawa, H. Yasuda, T. Suda, and K. Higashio, Osteoclast differentiation factor mediates an essential signal for bone resorption induced by 1 alpha,25-dihydroxyvitamin D3, prostaglandin E2, or parathyroid hormone in the microenvironment of bone. Biochem. Biophys. Res. Commun. 246, 337 – 341 (1998). 69. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 70. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 71. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). 72. T. Murakami, M. Yamamoto, K. Ono, M. Nishikawa, N. Nagata, K. Motoyoshi, and T. Akatsu, Transforming growth factor-beta1 increases mRNA levels of osteoclastogenesis inhibitory factor in osteoblastic/stromal cells and inhibits the survival of murine osteoclastlike cells. Biochem. Biophys. Res. Commun. 252, 747 – 752 (1998). 73. A. Grey, M. A. Mitnick, U. Masiukiewicz, B. H. Sun, S. Rudikoff, R. L. Jilka, S. C. Manolagas, and K. Insogna, A role for interleukin-6 in parathyroid hormone-induced bone resorption in vivo. Endocrinology 140, 4683 – 4690 (1999). 74. H. Kaji, T. Sugimoto, M. Kanatani, K. Nishiyama, M. Nasu, and K. Chihara, Insulin-like growth factor-I mediates osteoclast-like cell formation stimulated by parathyroid hormone. J. Cell. Physiol. 172, 55 – 62 (1997). 75. T. Yonaga, Action of parathyroid hormone, with special reference to its anabolic effect on different kinds of tissues in rats (I). Bull. Tokyo Med. Dent. Univ. 25, 237 – 248 (1978). 76. J. Reeve, P. J. Meunier, J. A. Parsons, M. Bernat, O. L. Bijvoet, P. Courpron, C. Edouard, L. Klenerman, R. M. Neer, J. C. Renier, D. Slovik, F. J. Vismans, and J. T. Potts, Jr, Anabolic effect of human parathyroid hormone fragment on trabecular bone in involutional osteoporosis: A multicentre trial. Br. Med. J. 280, 1340 – 1344 (1980). 77. W. B. High, H. E. Black, and C. C. Capen, Histomorphometric evaluation of the effects of low dose parathyroid hormone administration on cortical bone remodeling in adult dogs. Lab. Invest. 44, 449 – 454 (1981).

ROBERT A. NISSENSON 78. M. Gunness-Hey and J. M. Hock, Increased trabecular bone mass in rats treated with human synthetic parathyroid hormone. Metab. Bone Dis. Relat. Res. 5, 177 – 181 (1984). 79. T. J. Wronski, C. F. Yen, H. Qi, and L. M. Dann, Parathyroid hormone is more effective than estrogen or bisphosphonates for restoration of lost bone mass in ovariectomized rats. Endocrinology 132, 823 – 831 (1993). 80. C. C. Liu and D. N. Kalu, Human parathyroid hormone-(1 – 34) prevents bone loss and augments bone formation in sexually mature ovariectomized rats. J. Bone Miner. Res. 5, 973 – 982 (1990). 81. J. F. Whitfield, P. Morley, and G. E. Willick, The bone-building action of the parathyroid hormone: Implications for the treatment of osteoporosis. Drugs Aging 15, 117 – 129 (1999). 82. D. W. Dempster, F. Cosman, M. Parisien, V. Shen, and R. Lindsay, Anabolic actions of parathyroid hormone on bone [published erratum appears in Endocr. Rev. 15(2), 261 (1994)]. Endocr. Rev. 14, 690 – 709 (1993). 83. T. Hirano, D. B. Burr, C. H. Turner, M. Sato, R. L. Cain, and J. M. Hock, Anabolic effects of human biosynthetic parathyroid hormone fragment (1 – 34), LY333334, on remodeling and mechanical properties of cortical bone in rabbits. J. Bone Miner. Res. 14, 536 – 545 (1999). 84. L. Zhang, N. Endo, N. Yamamoto, T. Tanizawa, and H. E. Takahashi, Effects of single and concurrent intermittent administration of human PTH (1 – 34) and incadronate on cancellous and cortical bone of femoral neck in ovariectomized rats. Tohoku J. Exp. Med. 186, 131 – 141 (1998). 85. M. Sato, G. Q. Zeng, and C. H. Turner, Biosynthetic human parathyroid hormone (1 – 34) effects on bone quality in aged ovariectomized rats. Endocrinology 138, 4330 – 4337 (1997). 86. B. D. Baumann and T. J. Wronski, Response of cortical bone to antiresorptive agents and parathyroid hormone in aged ovariectomized rats. Bone 16, 247 – 253 (1995). 87. N. E. Lane, D. B. Kimmel, M. H. Nilsson, F. E. Cohen, S. Newton, R. A. Nissenson, and G. J. Strewler, Bone-selective analogs of human PTH(1 – 34) increase bone formation in an ovariectomized rat model. J. Bone Miner. Res. 11, 614 – 625 (1996). 88. N. E. Lane, J. M. Thompson, G. J. Strewler, and J. H. Kinney, Intermittent treatment with human parathyroid hormone (hPTH[1 – 34]) increased trabecular bone volume but not connectivity in osteopenic rats. J. Bone Miner. Res. 10, 1470 – 1477 (1995). 89. L. Mosekilde, C. H. Søgaard, C. C. Danielsen and O. Tørring, The anabolic effects of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: a comparison study between hPTH-(1 – 34) and hPTH-(1 – 84). Endocrinology 129, 421 – 428 (1991). 90. K. C. Hicok, T. Thomas, F. Gori, D. J. Rickard, T. C. Spelsberg, and B. L. Riggs, Development and characterization of conditionally immortalized osteoblast precursor cell lines from human bone marrow stroma. J. Bone Miner. Res. 13, 205 – 217, (1998). 91. R. W. Zhang, S. C. Supowit, X. Xu, H. Li, M. D. Christensen, R. Lozano, and D. J. Simmons, Expression of selected osteogenic markers in the fibroblast-like cells of rat marrow stroma. Calcif. Tissue Int. 56, 283 – 291, (1995). 92. M. F. Rouleau, J. Mitchell, and D. Goltzman, Characterization of the major parathyroid hormone target cell in the endosteal metaphysis of rat long bones. J. Bone Miner. Res. 5, 1043 – 1053 (1990). 93. X. Y. Zang, Y. B. Tan, Z. L. Pang, W. Z. Zhang, and J. Zhao, Effects of parathyroid hormone and estradiol on proliferation and function of human osteoblasts from fetal long bone: An in vitro study. Chin. Med. J. 107, 600 – 603 (1994). 94. M. Sabatini, C. Lesur, M. Pacherie, P. Pastoureau, N. Kucharczyk, J. L. Fauchère, and J. Bonnet, Effects of parathyroid hormone and agonists of the adenylyl cyclase and protein kinase C pathways on bone cell proliferation. Bone 18, 59 – 65 (1996).

CHAPTER 7 PTH and PTH-Related Protein 95. B. R. MacDonald, J. A. Gallagher, and R. G. Russell, Parathyroid hormone stimulates the proliferation of cells derived from human bone. Endocrinology 118, 2445 – 2449 (1986). 96. N. C. Partridge, A. L. Opie, R. T. Opie, and T. J. Martin, Inhibitory effects of parathyroid hormone on growth of osteogenic sarcoma cells. Calcif. Tissue Int. 37, 519 – 525 (1985). 97. S. Nishida, A. Yamaguchi, T. Tanizawa, N. Endo, T. Mashiba, Y. Uchiyama, T. Suda, S. Yoshiki, and H. E. Takahashi, Increased bone formation by intermittent parathyroid hormone administration is due to the stimulation of proliferation and differentiation of osteoprogenitor cells in bone marrow. Bone 15, 717 – 723 (1994). 98. P. C. Schiller, G. D’Ippolito, B. A. Roos, and G. A. Howard, Anabolic or catabolic responses of MC3T3-E1 osteoblastic cells to parathyroid hormone depend on time and duration of treatment. J. Bone Miner. Res. 14, 1504 – 1512 (1999). 99. T. Ishizuya, S. Yokose, M. Hori, T. Noda, T. Suda, S. Yoshiki, and A. Yamaguchi, Parathyroid hormone exerts disparate effects on osteoblast differentiation depending on exposure time in rat osteoblastic cells. J. Clin. Invest. 99, 2961 – 2970 (1997). 100. R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). 101. D. Thiébaud, H. L. Guenther, A. Porret, P. Burckhardt, H. Fleisch, and W. Hofstetter, Regulation of collagen type I and biglycan mRNA levels by hormones and growth factors in normal and immortalized osteoblastic cell lines. J. Bone Miner. Res. 9, 1347 – 1354 (1994). 102. R. Yang and L. C. Gerstenfeld, Signal transduction pathways mediating parathyroid hormone stimulation of bone sialoprotein gene expression in osteoblasts. J. Biol. Chem. 271, 29839 – 29846 (1996). 103. G. Boguslawski, L. V. Hale, X. P. Yu, R. R. Miles, J. E. Onyia, R. F. Santerre, and S. Chandrasekhar, Activation of osteocalcin transcription involves interaction of protein kinase A- and protein kinase C-dependent pathways. J. Biol. Chem. 275, 999 – 1006 (2000). 104. E. Canalis, M. Centrella, W. Burch, and T. L. McCarthy, Insulin-like growth factor I mediates selective anabolic effects of parathyroid hormone in bone cultures. J. Clin. Invest. 83, 60 – 65 (1989). 105. J. H. R Ellsworth, Studies on physiology of parathyroid glands: Some responses of normal human kidneys and blood to intravenous parathyroid extract. Bull. Johns Hopkins Hosp. 55, 296 (1934). 106. H. Murer, I. Forster, H. Hilfiker, M. Pfister, B. Kaissling, M. Lötscher, and J. Biber, Cellular/molecular control of renal Na/Picotransport. Kidney Int. Suppl. 65, S2 – S10 (1998). 107. T. A. K. Berndt, Renal regulation of phosphate excretion. In “The Kidney: Physiology and Pathophysiology” (D. A. G. Seldin, ed.), pp. 2511 – 2532. Raven Press, New York, 1992. 108. H. Murer, Homer Smith Award: Cellular mechanisms in proximal tubular Pi reabsorption: Some answers and more questions. J. Am. Soc. Nephrol. 2, 1649 – 1665 (1992). 109. H. Murer, I. Forster, N. Hernando, G. Lambert, M. Traebert, and J. Biber, Posttranscriptional regulation of the proximal tubule NaPi-II transporter in response to PTH and dietary P(i). Am. J. Physiol. 277, F676 – F684 (1999). 110. I. Keusch, M. Traebert, M. Lötscher, B. Kaissling, H. Murer, and J. Biber, Parathyroid hormone and dietary phosphate provoke a lysosomal routing of the proximal tubular Na/Pi-cotransporter type II. Kidney Int. 54, 1224 – 1232 (1998). 111. Y. Zhang, J. M. Norian, C. E. Magyar, N. H. Holstein-Rathlou, A. K. Mircheff, and A. A. McDonough, In vivo PTH provokes apical NHE3 and NaPi2 redistribution and Na-K-ATPase inhibition. Am. J. Physiol. 276, F711 – F719 (1999). 112. S. A. Kempson, M. Lötscher, B. Kaissling, J. Biber, H. Murer, and M. Levi, Parathyroid hormone action on phosphate transporter mRNA and protein in rat renal proximal tubules. Am. J. Physiol. 268, F784 – F791 (1995).

239 113. P. A. Friedman and F. A. Gesek, Calcium transport in renal epithelial cells. Am. J. Physiol. 264, F181 – F198 (1993). 114. D. W. Seldin, Renal handling of calcium. Nephron 81, Suppl. 1, 2 – 7 (1999). 115. B. J. Bacskai and P. A. Friedman, Activation of latent Ca2 channels in renal epithelial cells by parathyroid hormone. Nature 347, 388 – 391 (1990). 116. F. A. Gesek and P. A. Friedman, On the mechanism of parathyroid hormone stimulation of calcium uptake by mouse distal convoluted tubule cells. J. Clin. Invest. 90, 749 – 758 (1992). 117. P. A. Friedman, Codependence of renal calcium and sodium transport. Annu. Rev. Physiol. 60, 179 – 197 (1998). 118. N. A. Breslau, Normal and abnormal regulation of 1,25-(OH)2D synthesis. Am. J. Med. Sci. 296, 417 – 425 (1988). 119. D. R. Fraser and E. Kodicek, Regulation of 25-hydroxycholecalciferol-1-hydroxylase activity in kidney by parathyroid hormone. Nature New Biol. 241, 163 – 166 (1973). 120. M. J. Favus, Factors affecting calcium metabolism in disorders of the kidney. Ann. Clin. Lab. Sci. 11, 327 – 332 (1981). 121. R. Kremer and D. Goltzman, Parathyroid hormone stimulates mammalian renal 25-hydroxyvitamin D3-1 alpha-hydroxylase in vitro. Endocrinology 110, 294 – 296 (1982). 122. C. R. Rost, D. D. Bikle, and R. A. Kaplan, In vitro stimulation of 25hydroxycholecalciferol 1 alpha-hydroxylation by parathyroid hormone in chick kidney slices: Evidence for a role for adenosine 3,5monophosphate. Endocrinology 108, 1002 – 1006 (1981). 123. T. Monkawa, T. Yoshida, S. Wakino, T. Shinki, H. Anazawa, H. F. Deluca, T. Suda, M. Hayashi, and T. Saruta, Molecular cloning of cDNA and genomic DNA for human 25-hydroxyvitamin D3 1 alphahydroxylase. Biochem. Biophys. Res. Commun. 239, 527 – 533 (1997). 124. G. K. Fu, D. Lin, M. Y. Zhang, D. D. Bikle, C. H. Shackleton, W. L. Miller, and A. A. Portale, Cloning of human 25-hydroxyvitamin D-1 alpha-hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol. Endocrinol. 11, 1961 – 1970 (1997). 125. F. H. Glorieux and R. St-Arnaud, Molecular cloning of (25-OH D)-1 alpha-hydroxylase: An approach to the understanding of vitamin D pseudo-deficiency. Recent Progr. Horm. Res. 53, 341 – 349 (1998). 126. A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, and S. Kato, Positive and negative regulations of the renal 25-hydroxyvitamin D3 1alpha-hydroxylase gene by parathyroid hormone, calcitonin, and 1alpha,25(OH)2D3 in intact animals. Endocrinology 140, 2224 – 2231 (1999). 127. T. Shinki, Y. Ueno, H. F. DeLuca, and T. Suda, Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1alphahydroxylase gene in normocalcemic rats. Proc. Natl. Acad. Sci. USA 96, 8253 – 8258 (1999). 128. A. F. Stewart, R. Horst, L. J. Deftos, E. C. Cadman, R. Lang, and A. E. Broadus, Biochemical evaluation of patients with cancerassociated hypercalcemia: Evidence for humoral and nonhumoral groups. N. Engl. J. Med. 303, 1377 – 1383 (1980). 129. S. B. Rodan, K. L. Insogna, A. M. Vignery, A. F. Stewart, A. E. Broadus, S. M. D’Souza, D. R. Bertolini, G. R. Mundy, and G. A. Rodan, Factors associated with humoral hypercalcemia of malignancy stimulate adenylate cyclase in osteoblastic cells. J. Clin. Invest. 72, 1511 – 1515 (1983). 130. G. J. Strewler, R. D. Williams, and R. A. Nissenson, Human renal carcinoma cells produce hypercalcemia in the nude mouse and a novel protein recognized by parathyroid hormone receptors. J. Clin. Invest. 71, 769 – 774 (1983). 131. S. A. Rabbani, J. Mitchell, D. R. Roy, R. Kremer, H. P. Bennett, and D. Goltzman, Purification of peptides with parathyroid hormone-like bioactivity from human and rat malignancies associated with hypercalcemia. Endocrinology 118, 1200 – 1210 (1986). 132. J. M. Moseley, M. Kubota, H. Diefenbach-Jagger, R. E. Wettenhall, B. E. Kemp, L. J. Suva, C. P. Rodda, P. R. Ebeling, P. J. Hudson,

240

133.

134.

135.

136.

137.

138.

139. 140.

141.

142.

143.

144.

145.

146.

147.

148.

ROBERT A. NISSENSON J. D. Zajac et al., Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc. Natl. Acad. Sci. USA 84, 5048 – 5052 (1987). L. J. Suva, G. A. Winslow, R. E. Wettenhall, R. G. Hammonds, J. M. Moseley, H. Diefenbach-Jagger, C. P. Rodda, B. E. Kemp, H. Rodriguez, E. Y. Chen et al., A parathyroid hormone-related protein implicated in malignant hypercalcemia: Cloning and expression. Science 237, 893 – 896 (1987). G. J. Strewler, P. H. Stern, J. W. Jacobs, J. Eveloff, R. F. Klein, S. C. Leung, M. Rosenblatt, and R. A. Nissenson, Parathyroid hormonelike protein from human renal carcinoma cells: Structural and functional homology with parathyroid hormone. J. Clin. Invest. 80, 1803 – 1807 (1987). M. Mangin, A. C. Webb, B. E. Dreyer, J. T. Posillico, K. Ikeda, E. C. Weir, A. F. Stewart, N. H. Bander, L. Milstone, D. E. Barton et al., Identification of a cDNA encoding a parathyroid hormone-like peptide from a human tumor associated with humoral hypercalcemia of malignancy. Proc. Natl. Acad. Sci. USA 85, 597 – 601 (1988). M. Mangin, K. Ikeda, B. E. Dreyer, L. Milstone, and A. E. Broadus, Two distinct tumor-derived, parathyroid hormone-like peptides result from alternative ribonucleic acid splicing. Mol. Endocrinol. 2, 1049 – 1055 (1988). M. A. Thiede, G. J. Strewler, R. A. Nissenson, M. Rosenblatt, and G. A. Rodan, Human renal carcinoma expresses two messages encoding a parathyroid hormone-like peptide: Evidence for the alternative splicing of a single-copy gene. Proc. Natl. Acad. Sci. USA 85, 4605 – 4609 (1988). M. Mannstadt, H. Jüppner, and T. J. Gardella, Receptors for PTH and PTHrP: Their biological importance and functional properties. Am. J. Physiol. 277, F665 – F675 (1999). G. J. Strewler, The physiology of parathyroid hormone-related protein. N. Engl. J. Med. 342, 177 – 185 (2000). J. J. Wysolmerski and A. F. Stewart, The physiology of parathyroid hormone-related protein: An emerging role as a developmental factor. Annu. Rev. Physiol. 60, 431 – 460 (1998). B. Lanske and H. M. Kronenberg, Parathyroid hormone-related peptide (PTHrP) and parathyroid hormone (PTH)/PTHrP receptor. Crit. Rev. Eukary. Gene Express. 8, 297 – 320 (1998). W. M. Philbrick, J. J. Wysolmerski, S. Galbraith, E. Holt, J. J. Orloff, K. H. Yang, R. C. Vasavada, E. C. Weir, A. E. Broadus, and A. F. Stewart, Defining the roles of parathyroid hormone-related protein in normal physiology. Physiol. Rev. 76, 127 – 173 (1996). M. E. Dunbar, J. J. Wysolmerski, and A. E. Broadus, Parathyroid hormone-related protein: From hypercalcemia of malignancy to developmental regulatory molecule. Am. J. Med. Sci. 312, 287 – 294 (1996). A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene. Genes Dev. 8, 277 – 289 (1994). A. Erlebacher, E. H. Filvaroff, S. E. Gitelman, and R. Derynck, Toward a molecular understanding of skeletal development. Cell 80, 371 – 378 (1995). B. St-Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). E. C. Weir, W. M. Philbrick, M. Amling, L. A. Neff, R. Baron, and A. E. Broadus, Targeted overexpression of parathyroid hormonerelated peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc. Natl. Acad. Sci. USA 93, 10240 – 10245 (1996). E. Schipani, K. Kruse, and H. Jüppner, A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 268, 98 – 100 (1995).

149. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 150. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Jüppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273, 663 – 666 (1996). 151. S. J. Karp, E. Schipani, B. St-Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 152. K. Lee, J. D. Deeds, and G. V. Segre, Expression of parathyroid hormone-related peptide and its receptor messenger ribonucleic acids during fetal development of rats. Endocrinology 136, 453 – 463 (1995). 153. M. Jansen, Uber atypische chondrodystrophie (achondroplasie) und uber eine noch nicht beschriebene angeborene wachstumsstarung des knochensystems: Metaphysare dysostosis. Z. Orthop. Chir. 61, 253 – 286 (1934). 154. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 155. A. C. Karaplis, B. He, M. T. Nguyen, I. D. Young, D. Semeraro, H. Ozawa, and N. Amizuka, Inactivating mutation in the human parathyroid hormone receptor type 1 gene in Blomstrand chondrodysplasia. Endocrinology 139, 5255 – 5258 (1998). 156. S. Blomstrand, I. Claësson and J. Säve-Söderbergh, A case of lethal congenital dwarfism with accelerated skeletal maturation. Pediatr. Radiol. 15, 141 – 143 (1985). 157. M. Amling, L. Neff, S. Tanaka, D. Inoue, K. Kuida, E. Weir, W. M. Philbrick, A. E. Broadus, and R. Baron, Bcl-2 lies downstream of parathyroid hormone-related peptide in a signaling pathway that regulates chondrocyte maturation during skeletal development. J. Cell Biol. 136, 205 – 213 (1997). 158. J. J. Wysolmerski, J. F. McCaughern-Carucci, A. G. Daifotis, A. E. Broadus, and W. M. Philbrick, Overexpression of parathyroid hormone-related protein or parathyroid hormone in transgenic mice impairs branching morphogenesis during mammary gland development. Development 121, 3539 – 3547 (1995). 159. J. J. Wysolmerski, W. M. Philbrick, M. E. Dunbar, B. Lanske, H. Kronenberg, and A. E. Broadus, Rescue of the parathyroid hormone-related protein knockout mouse demonstrates that parathyroid hormone-related protein is essential for mammary gland development. Development 125, 1285 – 1294 (1998). 160. S. L. Ferrari, R. Rizzoli, and J. P. Bonjour, Parathyroid hormonerelated protein production by primary cultures of mammary epithelial cells. J. Cell. Physiol. 150, 304 – 311 (1992). 161. M. E. Dunbar, P. Young, J. P. Zhang, J. McCaughern-Carucci, B. Lanske, J. J. Orloff, A. Karaplis, G. Cunha, and J. J. Wysolmerski, Stromal cells are critical targets in the regulation of mammary ductal morphogenesis by parathyroid hormone-related protein. Dev. Biol. 203, 75 – 89 (1998). 162. M. E. Dunbar, P. R. Dann, G. W. Robinson, L. Hennighausen, J. P. Zhang, and J. J. Wysolmerski, Parathyroid hormone-related protein signaling is necessary for sexual dimorphism during embryonic mammary development. Development 126, 3485 – 3493 (1999). 163. J. J. Merendino, Jr., K. L. Insogna, L. M. Milstone, A. E. Broadus, and A. F. Stewart, A parathyroid hormone-like protein from cultured human keratinocytes. Science 231, 388 – 390 (1986). 164. R. V. Campos, S. L. Asa, and D. J. Drucker, Immunocytochemical localization of parathyroid hormone-like peptide in the rat fetus. Cancer Res. 51, 6351 – 6357 (1991).

241

CHAPTER 7 PTH and PTH-Related Protein 165. E. J. Atillasoy, W. J. Burtis, and L. M. Milstone, Immunohistochemical localization of parathyroid hormone-related protein (PTHRP) in normal human skin. J. Invest. Dermatol. 96, 277 – 280 (1991). 166. K. K. Pun, C. D. Arnaud, and R. A. Nissenson, Parathyroid hormone receptors in human dermal fibroblasts: Structural and functional characterization. J. Bone Miner. Res. 3, 453 – 460 (1988). 167. J. J. Orloff, M. B. Ganz, A. E. Ribaudo, W. J. Burtis, M. Reiss, L. M. Milstone, and A. F. Stewart, Analysis of PTHRP binding and signal transduction mechanisms in benign and malignant squamous cells. Am. J. Physiol. 262, E599 – E607 (1992). 168. S. M. Kaiser, P. Laneuville, S. M. Bernier, J. S. Rhim, R. Kremer, and D. Goltzman, Enhanced growth of a human keratinocyte cell line induced by antisense RNA for parathyroid hormone-related peptide. J. Biol. Chem. 267, 13623 – 13628 (1992). 169. S. M. Kaiser, M. Sebag, J. S. Rhim, R. Kremer, and D. Goltzman, Antisense-mediated inhibition of parathyroid hormone-related peptide production in a keratinocyte cell line impedes differentiation. Mol. Endocrinol. 8, 139 – 147 (1994). 170. J. J. Wysolmerski, A. E. Broadus, J. Zhou, E. Fuchs, L. M. Milstone, and W. M. Philbrick, Overexpression of parathyroid hormone-related protein in the skin of transgenic mice interferes with hair follicle development. Proc. Natl. Acad. Sci. USA 91, 1133 – 1137 (1994). 171. J. Foley, B. J. Longely, J. J. Wysolmerski, B. E. Dreyer, A. E. Broadus, and W. M. Philbrick, PTHrP regulates epidermal differentiation in adult mice. J. Invest. Dermatol. 111, 1122 – 1128 (1998). 172. M. F. Holick, S. Ray, T. C. Chen, X. Tian, and K. S. Persons, A parathyroid hormone antagonist stimulates epidermal proliferation and hair growth in mice. Proc. Natl. Acad. Sci. USA 91, 8014 – 8016 (1994). 173. E. A. Blomme, Y. Sugimoto, Y. C. Lin, C. C. Capen, and T. J. Rosol, Parathyroid hormone-related protein is a positive regulator of keratinocyte growth factor expression by normal dermal fibroblasts. Mol. Cell. Endocrinol. 152, 189 – 197 (1999). 174. W. M. Philbrick, B. E. Dreyer, I. A. Nakchbandi, and A. C. Karaplis, Parathyroid hormone-related protein is required for tooth eruption. Proc. Natl. Acad. Sci. USA 95, 11846 – 11851 (1998). 175. M. A. Thiede, A. G. Daifotis, E. C. Weir, M. L. Brines, W. J. Burtis, K. Ikeda, B. E. Dreyer, R. E. Garfield, and A. E. Broadus, Intrauterine occupancy controls expression of the parathyroid hormone-related peptide gene in preterm rat myometrium. Proc. Natl. Acad. Sci. USA 87, 6969 – 6973 (1990). 176. A. G. Daifotis, E. C. Weir, B. E. Dreyer, and A. E. Broadus, Stretchinduced parathyroid hormone-related peptide gene expression in the rat uterus. J. Biol. Chem. 267, 23455 – 23458 (1992). 177. J. E. d. Ferguson, J. V. Gorman, D. E. Bruns, E. C. Weir, W. J. Burtis, T. J. Martin, and M. E. Bruns, Abundant expression of parathyroid hormone-related protein in human amnion and its association with labor. Proc. Natl. Acad. Sci. USA 89, 8384 – 8388 (1992). 178. A. M. Germain, H. Attaroglu, P. C. MacDonald, and M. L. Casey, Parathyroid hormone-related protein mRNA in avascular human amnion. J. Clin. Endocrinol. Metab. 75, 1173 – 1175 (1992). 179. M. Yamamoto, S. C. Harm, W. A. Grasser, and M. A. Thiede, Parathyroid hormone-related protein in the rat urinary bladder: A smooth muscle relaxant produced locally in response to mechanical stretch. Proc. Natl. Acad. Sci. USA 89, 5326 – 5330 (1992). 180. M. Ito, A. Ohtsuru, H. Enomoto, S. Ozeki, M. Nakashima, T. Nakayama, K. Shichijo, I. Sekine, and S. Yamashita, Expression of parathyroid hormone-related peptide in relation to perturbations of gastric motility in the rat. Endocrinology 134, 1936 – 1942 (1994). 181. G. A. Nickols, A. D. Nana, M. A. Nickols, D. J. DiPette, and G. K. Asimakis, Hypotension and cardiac stimulation due to the parathyroid hormone-related protein, humoral hypercalcemia of malignancy factor. Endocrinology 125, 834 – 841 (1989). 182. T. Massfelder, N. Fiaschi-Taesch, A. F. Stewart, and J. J. Helwig, Parathyroid hormone-related peptide: A smooth muscle tone and pro-

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

196.

liferation regulatory protein. Curr. Opin. Nephrol. Hypertension 7, 27 – 32 (1998). K. Takahashi, D. Inoue, K. Ando, T. Matsumoto, K. Ikeda, and T. Fujita, Parathyroid hormone-related peptide as a locally produced vasorelaxant: Regulation of its mRNA by hypertension in rats. Biochem. Biophys. Res. Commun. 208, 447 – 455 (1995). C. J. Pirola, H. M. Wang, A. Kamyar, S. Wu, H. Enomoto, B. Sharifi, J. S. Forrester, T. L. Clemens, and J. A. Fagin, Angiotensin II regulates parathyroid hormone-related protein expression in cultured rat aortic smooth muscle cells through transcriptional and post-transcriptional mechanisms. J. Biol. Chem. 268, 1987 – 1994 (1993). S. Maeda, R. L. Sutliff, J. Qian, J. N. Lorenz, J. Wang, H. Tang, T. Nakayama, C. Weber, D. Witte, A. R. Strauch, R. J. Paul, J. A. Fagin, and T. L. Clemens, Targeted overexpression of parathyroid hormone-related protein (PTHrP) to vascular smooth muscle in transgenic mice lowers blood pressure and alters vascular contractility. Endocrinology 140, 1815 – 1825 (1999). J. Qian, J. N. Lorenz, S. Maeda, R. L. Sutliff, C. Weber, T. Nakayama, M. C. Colbert, R. J. Paul, J. A. Fagin, and T. L. Clemens, Reduced blood pressure and increased sensitivity of the vasculature to parathyroid hormone-related protein (PTHrP) in transgenic mice overexpressing the PTH/PTHrP receptor in vascular smooth muscle. Endocrinology 140, 1826 – 1833 (1999). A. G. Schneider, K. Leuthäuser, and D. Pette, Parathyroid hormonerelated protein is rapidly up-regulated in blood vessels of rat skeletal muscle by low-frequency stimulation. Pflug. Arch. Eur. J. Physiol. 439, 167 – 173 (1999). E. C. Weir, M. L. Brines, K. Ikeda, W. J. Burtis, A. E. Broadus, and R. J. Robbins, Parathyroid hormone-related peptide gene is expressed in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA 87, 108 – 112 (1990). D. R. Weaver, J. D. Deeds, K. Lee, and G. V. Segre, Localization of parathyroid hormone-related peptide (PTHrP) and PTH/PTHrP receptor mRNAs in rat brain. Brain Res. Mol. Brain Res. 28, 296 – 310 (1995). E. H. Holt, A. E. Broadus, and M. L. Brines, Parathyroid hormonerelated peptide is produced by cultured cerebellar granule cells in response to L-type voltage-sensitive Ca2 channel flux via a Ca2/calmodulin-dependent kinase pathway. J. Biol. Chem. 271, 28105 – 28111 (1996). M. L. Brines, Z. Ling, and A. E. Broadus, Parathyroid hormone-related protein protects against kainic acid excitotoxicity in rat cerebellar granule cells by regulating L-type channel calcium flux. Neurosci. Lett. 274, 13 – 16 (1999). P. K. Pang, R. Wang, J. Shan, E. Karpinski, and C. G. Benishin, Specific inhibition of long-lasting, L-type calcium channels by synthetic parathyroid hormone. Proc. Natl. Acad. Sci. USA 87, 623 – 627 (1990). T. Ono, K. Inokuchi, A. Ogura, Y. Ikawa, Y. Kudo, and S. Kawashima, Activity-dependent expression of parathyroid hormone-related protein (PTHrP) in rat cerebellar granule neurons: Requirement of PTHrP for the activity-dependent survival of granule neurons. J. Biol. Chem. 272, 14404 – 14411 (1997). C. P. Rodda, M. Kubota, J. A. Heath, P. R. Ebeling, J. M. Moseley, A. D. Care, I. W. Caple, and T. J. Martin, Evidence for a novel parathyroid hormone-related protein in fetal lamb parathyroid glands and sheep placenta: Comparisons with a similar protein implicated in humoral hypercalcaemia of malignancy. J. Endocrinol. 117, 261 – 271 (1988). R. J. MacIsaac, I. W. Caple, J. A. Danks, H. Diefenbach-Jagger, V. Grill, J. M. Moseley, J. Southby, and T. J. Martin, Ontogeny of parathyroid hormone-related protein in the ovine parathyroid gland. Endocrinology 129, 757 – 764 (1991). C. S. Kovacs, B. Lanske, J. L. Hunzelman, J. Guo, A. C. Karaplis, and H. M. Kronenberg, Parathyroid hormone-related peptide

242

197.

198.

199.

200.

201.

202.

203.

204.

205.

206.

207.

208. 209.

210. 211.

212.

213.

214.

215.

216.

ROBERT A. NISSENSON (PTHrP) regulates fetal – placental calcium transport through a receptor distinct from the PTH/PTHrP receptor. Proc. Natl. Acad. Sci. USA 93, 15233 – 15238 (1996). M. A. Thiede and G. A. Rodan, Expression of a calcium-mobilizing parathyroid hormone-like peptide in lactating mammary tissue. Science 242, 278 – 280 (1988). M. Yamamoto, L. T. Duong, J. E. Fisher, M. A. Thiede, M. P. Caulfield, and M. Rosenblatt, Suckling-mediated increases in urinary phosphate and 3,5-cyclic adenosine monophosphate excretion in lactating rats: Possible systemic effects of parathyroid hormone-related protein. Endocrinology 129, 2614 – 2622 (1991). K. J. Mather, C. L. Chik, and B. Corenblum, Maintenance of serum calcium by parathyroid hormone-related peptide during lactation in a hypoparathyroid patient. J. Clin. Endocrinol. Metabol. 84, 424 – 427 (1999). A. A. Budayr, B. P. Halloran, J. C. King, D. Diep, R. A. Nissenson, and G. J. Strewler, High levels of a parathyroid hormone-like protein in milk. Proc. Natl. Acad. Sci. USA 86, 7183 – 7185 (1989). M. Sowers, D. Zhang, B. W. Hollis, B. Shapiro, C. A. Janney, M. Crutchfield, M. A. Schork, F. Stanczyk, and J. Randolph, Role of calciotrophic hormones in calcium mobilization of lactation. Am. J. Clin. Nutr. 67, 284 – 291 (1998). L. R. Chase and G. D. Aurbach, Renal adenyl cyclase: Anatomically separate sites for parathyroid hormone and vasopressin. Science 159, 545 – 547 (1968). L. R. Chase, S. A. Fedak, and G. D. Aurbach, Activation of skeletal adenyl cyclase by parathyroid hormone in vitro. Endocrinology 84, 761 – 768 (1969). R. Marcus and G. D. Aurbach, Bioassay of parathyroid hormone in vitro with a stable preparation of adenyl cyclase from rat kidney. Endocrinology 85, 801 – 810 (1969). T. Dousa and I. Rychlík, The effect of parathyroid hormone on adenyl cyclase in rat kidney. Biochim. Biophys. Acta 158, 484 – 486 (1968). B. Lanske, P. Divieti, C. S. Kovacs, A. Pirro, W. J. Landis, S. M. Krane, F. R. Bringhurst, and H. M. Kronenberg, The parathyroid hormone (PTH)/PTH-related peptide receptor mediates actions of both ligands in murine bone. Endocrinology 139, 5194 – 5204 (1998). G. Vaes, Parathyroid hormone-like action of N6-2-O-dibutyryladenosine-35 (cyclic)-monophosphate on bone explants in tissue culture. Nature 219, 939 – 940 (1968). L. G. Raisz, Physiologic and pharmacologic regulation of bone resorption. N. Engl. J. Med. 282, 909 – 916 (1970). M. P. Herrmann-Erlee, Studies on the role of cyclic AMP in parathyroid hormone-induced embryonic bone resorption. J. Endocrinol. 48, lix-lxi (1970). W. A. Peck and S. Klahr, Cyclic nucleotides in bone and mineral metabolism. Adv. Cyclic Nucleotide Res. 11, 89 – 130 (1979). J. A. Lorenzo, S. Sousa, and, J. Quinton, Forskolin has both stimulatory and inhibitory effects on bone resorption in fetal rat long bone cultures. J. Bone Miner. Res. 1, 313 – 317 (1986). M. P. Herrmann-Erlee and J. M. V. D. Meer, The effects of dibutyryl cyclic AMP, aminophylline and propranolol on PTE-induced bone resorption in vitro. Endocrinology 94, 424 – 434 (1974). R. Dziak and P. Stern, Parathyromimetic effects of the ionophore, A23187, on bone cells and organ cultures. Biochem. Biophys. Res. Commun. 65, 1343 – 1349 (1975). J. A. Lorenzo and S. Sousa, Phorbol esters stimulate bone resorption in fetal rat long-bone cultures by mechanisms independent of prostaglandin synthesis. J. Bone Miner. Res. 3, 63 – 67 (1988). D. C. Abraham, C. L. Wadkins, and H. H. Conaway, Enhancement of fetal rat limb bone resorption by phorbol ester (PMA) and ionophore A-23187. Calcif. Tissue Int. 42, 191 – 195 (1988). M. P. Bos, W. Most, J. P. van Leeuwen, and M. P. Herrmann-Erlee, Role of protein kinase C (PKC) in bone resorption: Effect of the

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

specific PKC inhibitor 1-alkyl-2-methylglycerol. Biochem. Biophys. Res. Commun. 184, 1317 – 1323 (1992). S. M. Sprague, M. M. Popovtzer, M. Dranitzki-Elhalel, and H. Wald, Parathyroid hormone-induced calcium efflux from cultured bone is mediated by protein kinase C translocation. Am. J. Physiol. 271, F1139 – F1146 (1996). E. Bornefalk, S. Ljunghall, and O. Ljunggren, Bone resorption induced by A23187 is abolished by indomethacin: implications for second messenger utilised by parathyroid hormone. Eur. J. Pharmacol. 345, 333 – 338 (1998). J. L. Ivey, D. R. Wright, and A. H. Tashjian, Jr., Bone resorption in organ culture: Inhibition by the divalent cation ionophores A23187 and X-537A. J. Clin. Invest. 58, 1327 – 1338 (1976). R. Dziak and P. H. Stern, Responses of fetal rat bone cells and bone organ cultures to the ionophore, A23187. Calcif. Tissue Res. 22, 137 – 147 (1976). M. Ransjö and U. H. Lerner, 12-O-tetradecanoylphorbol-13-acetate, a phorbol ester stimulating protein kinase C, inhibits bone resorption in vitro induced by parathyroid hormone and parathyroid hormonerelated peptide of malignancy. Acta Physiol. Scand. 139, 249 – 250 (1990). H. Kaji, T. Sugimoto, M. Kanatani, M. Nasu, and K. Chihara, Estrogen blocks parathyroid hormone (PTH)-stimulated osteoclast-like cell formation by selectively affecting PTH-responsive cyclic adenosine monophosphate pathway. Endocrinology 137, 2217 – 2224 (1996). M. Ransjö, A. Lie, and E. J. Mackie, Cholera toxin and forskolin stimulate formation of osteoclast-like cells in mouse marrow cultures and cultured mouse calvarial bones. Eur. J. Oral Sci. 107, 45 – 54 (1999). Y. H. Gao and M. Yamaguchi, Inhibitory effect of genistein on osteoclast-like cell formation in mouse marrow cultures. Biochem. Pharmacol. 58, 767 – 772 (1999). T. Sugimoto, M. Kanatani, H. Kaji, T. Yamaguchi, M. Fukase, and K. Chihara, Second messenger signaling of PTH- and PTHRPstimulated osteoclast-like cell formation from hemopoietic blast cells. Am. J. Physiol. 265, E367 – E373 (1993). H. Brändström, K. B. Jonsson, C. Ohlsson, O. Vidal, S. Ljunghall, and O. Ljunggren, Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 247, 338 – 341 (1998). R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1999). T. Onishi, W. Zhang, X. Cao, and K. Hruska, The mitogenic effect of parathyroid hormone is associated with E2F-dependent activation of cyclin-dependent kinase 1 (cdc2) in osteoblast precursors. J. Bone Miner. Res. 12, 1596 – 1605 (1997). Y. Isogai, T. Akatsu, T. Ishizuya, A. Yamaguchi, M. Hori, N. Takahashi, and T. Suda, Parathyroid hormone regulates osteoblast differentiation positively or negatively depending on the differentiation stages. J. Bone Miner. Res. 11, 1384 – 1393 (1996). R. H. Rixon, J. F. Whitfield, L. Gagnon, R. J. Isaacs, S. Maclean, B. Chakravarthy, J. P. Durkin, W. Neugebauer, V. Ross, W. Sung et al., Parathyroid hormone fragments may stimulate bone growth in ovariectomized rats by activating adenylyl cyclase. J. Bone Miner. Res. 9, 1179 – 1189 (1994). J. F. Whitfield, P. Morley, G. E. Willick, S. MacLean, V. Ross, J. Barbier, and R. J. Isaacs, Stimulation of femoral trabecular bone growth in ovariectomized rats by human parathyroid hormone (hPTH)(1 – 30)NH(2). Calcif. Tissue Int. 65, 143 – 147 (1999). A. Inoue, Y. Hiruma, S. Hirose, A. Yamaguchi, and H. Hagiwara, Reciprocal regulation by cyclic nucleotides of the differentiation of rat osteoblast-like cells and mineralization of nodules. Biochem. Biophys. Res. Commun. 215, 1104 – 1110 (1995).

243

CHAPTER 7 PTH and PTH-Related Protein 233. H. Kaneki, I. Takasugi, M. Fujieda, M. Kiriu, S. Mizuochi, and H. Ide, Prostaglandin E2 stimulates the formation of mineralized bone nodules by a cAMP-independent mechanism in the culture of adult rat calvarial osteoblasts. J. Cell. Biochem. 73, 36 – 48 (1999). 234. A. J. Koh, C. A. Beecher, T. J. Rosol, and L. K. McCauley, 3,5Cyclic adenosine monophosphate activation in osteoblastic cells: Effects on parathyroid hormone – 1 receptors and osteoblastic differentiation in vitro. Endocrinology 140, 3154 – 3162 (1999). 235. J. Kano, T. Sugimoto, M. Fukase, and T. Fujita, The activation of cAMP-dependent protein kinase is directly linked to the regulation of osteoblast proliferation (UMR-106) by parathyroid hormone. Biochem. Biophys. Res. Commun. 177, 365 – 369 (1991). 236. D. Chabardès, M. Imbert, A. Clique, M. Montégut, and F. Morel, PTH sensitive adenyl cyclase activity in different segments of the rabbit nephron. Pflug. Arch. Eur. J. Physiol. 354, 229 – 239 (1975). 237. S. Torikai and M. Imai, A simple method to determine adenylate cyclase activity in isolated single nephron segments by radioimmunoassay for succinyl adenosine 3,5-cyclic monophosphate. Tohoku J. Exp. Med. 129, 91 – 99 (1979). 238. C. A. Harris and J. F. Seely, Resistance to the phosphaturic effect of parathyroid hormone in the hamster. Am. J. Physiol. 237, F175 – F181 (1979). 239. T. J. Berndt, M. J. Onsgard, and F. G. Knox, Effect of cAMP analogue infusion on phosphate reabsorption in phosphate-deprived rats. Am. J. Physiol. 255, F96 – F99 (1988). 240. N. Beck and B. B. Davis, Impaired renal response to parathyroid hormone in potassium depletion. Am. J. Physiol. 228, 179 – 183 (1975). 241. T. Mano, K. Uchimura, R. Hayashi, T. Kobahashi, K. Fujiwara, M. Makino, H. Kakizawa, M. Nagata, A. Nakai, M. Wada, A. Nagasaka, and M. Itoh, Increased urinary phosphate excretion in pseudohypoparathyroidism type II with long-term treatment with phosphodiesterase inhibitor. Horm. Metab. Res. 31, 602 – 605 (1999). 242. A. M. Moses, N. Breslau, and R. Coulson, Renal responses to PTH in patients with hormone-resistant (pseudo) hypoparathyroidism. Am. J. Med. 61, 184 – 189 (1976). 243. R. M. Neer, G. W. Tregear, and J. T. Potts, Jr., Renal effects of native parathyroid hormone and synthetic biologically active fragments in pseudohypoparathyroidism and hypoparathyroidism. J. Clin. Endocrinol. Metab. 44, 420 – 423 (1977). 244. F. Albright, CH, Burnett, and P. H. Smith, Pseudohypoparathyroidism: An example of “Seabright-Bantam syndrome.” Endocrinology 30, 922 – 932 (1942). 245. M. A. Levine, Pseudohypoparathyroidism: From bedside to bench and back. J. Bone Miner. Res. 14, 1255 – 1260 (1999). 246. A. P. Teitelbaum and G. J. Strewler, Parathyroid hormone receptors coupled to cyclic adenosine monophosphate formation in an established renal cell line. Endocrinology 114, 980 – 985 (1984). 247. K. Malmström and H. Murer, Parathyroid hormone inhibits phosphate transport in OK cells but not in LLC-PK1 and JTC-12.P3 cells. Am. J. Physiol. 251, C23 – C31 (1986). 248. J. A. Cole, L. R. Forte, S. Eber, P. K. Thorne, and R. E. Poelling, Regulation of sodium-dependent phosphate transport by parathyroid hormone in opossum kidney cells: Adenosine 3,5-monophosphatedependent and -independent mechanisms. Endocrinology 122, 2981 – 2989 (1988). 249. G. Quamme, J. Pfeilschifter, and H. Murer, Parathyroid hormone inhibition of Na/phosphate cotransport in OK cells: Intracellular [Ca2] as a second messenger. Biochim. Biophys. Acta 1013, 166 – 172 (1989). 250. K. J. Martin, C. L. McConkey, A. K. Jacob, E. A. Gonzalez, M. Khan, and J. J. Baldassare, Effect of U-73,122, an inhibitor of phospholipase C, on actions of parathyroid hormone in opossum kidney cells. Am. J. Physiol. 266, F254 – F258 (1994). 251. J. H. Segal and A. S. Pollock, Transfection-mediated expression of a dominant cAMP-resistant phenotype in the opossum kidney (OK)

252.

253.

254.

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

267.

268.

cell line prevents parathyroid hormone-induced inhibition of Naphosphate cotransport: A protein kinase-A-mediated event. J. Clin. Invest. 86, 1442 – 1450 (1990). E. D. Lederer, S. S. Sohi, J. M. Mathiesen, and J. B. Klein, Regulation of expression of type II sodium-phosphate cotransporters by protein kinases A and C. Am. J. Physiol. 275, F270 – F277 (1998). M. F. Pfister, J. Forgo, U. Ziegler, J. Biber, and H. Murer, cAMPdependent and -independent downregulation of type II Na-Pi cotransporters by PTH. Am. J. Physiol. 276, F720 – F725 (1999). M. Nakai, Y. Kinoshita, M. Fukase, and T. Fujita, Phorbol esters inhibit phosphate uptake in opossum kidney cells: A model of proximal renal tubular cells. Biochem. Biophys. Res. Commun. 145, 303 – 308 (1987). J. A. Cole, S. L. Eber, R. E. Poelling, P. K. Thorne, and L. R. Forte, A dual mechanism for regulation of kidney phosphate transport by parathyroid hormone. Am. J. Physiol. 253, E221 – E227 (1987). K. Malmström, G. Stange, and H. Murer, Intracellular cascades in the parathyroid-hormone-dependent regulation of Na/phosphate cotransport in OK cells. Biochem. J. 251, 207 – 213 (1988). J. Welsh, V. Weaver, and M. Simboli-Campbell, Regulation of renal 25(OH)D3 1 alpha-hydroxylase: Signal transduction pathways. Biochem. Cell Biol. 69, 768 – 770 (1991). H. L. Henry, Parathyroid hormone modulation of 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells is mimicked and enhanced by forskolin. Endocrinology 116, 503 – 510 (1985). H. L. Brenza, C. Kimmel-Jehan, F. Jehan, T. Shinki, S. Wakino, H. Anazawa, T. Suda, and H. F. DeLuca, Parathyroid hormone activation of the 25-hydroxyvitamin D3 – 1alpha-hydroxylase gene promoter. Proc. Natl. Acad. Sci. USA 95, 1387 – 1391 (1998). A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, T. Hosoya, and S. Kato, The promoter of the human 25-hydroxyvitamin D3 1 alphahydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1 alpha,25(OH)2D3. Biochem. Biophys. Res. Commun. 249, 11 – 16 (1998). X. F. Kong, X. H. Zhu, Y. L. Pei, D. M. Jackson, and M. F. Holick, Molecular cloning, characterization, and promoter analysis of the human 25-hydroxyvitamin D3-1alpha-hydroxylase gene. Proc. Natl. Acad. Sci. USA 96, 6988 – 6993 (1999). H. K. Ro, V. Tembe, and M. J. Favus, Evidence that activation of protein kinase-C can stimulate 1,25-dihydroxyvitamin D3 secretion by rat proximal tubules. Endocrinology 131, 1424 – 1428 (1992). M. Janulis, V. Tembe, and M. J. Favus, Role of protein kinase C in parathyroid hormone stimulation of renal 1,25-dihydroxyvitamin D3 secretion. J. Clin. Invest. 90, 2278 – 2283 (1992). J. G. Hoenderop, J. J. De Pont, R. J. Bindels, and P. H. Willems, Hormone-stimulated Ca2 reabsorption in rabbit kidney cortical collecting system is cAMP-independent and involves a phorbol ester-insensitive PKC isotype. Kidney Int. 55, 225 – 233 (1999). G. Hilal, D. Claveau, M. Leclerc, and M. G. Brunette, Ca2 transport by the luminal membrane of the distal nephron: Action and interaction of protein kinases A and C. Biochem. J. 328, 371 – 375 (1997). P. A. Friedman, F. A. Gesek, P. Morley, J. F. Whitfield, and G. E. Willick, Cell-specific signaling and structure-activity relations of parathyroid hormone analogs in mouse kidney cells. Endocrinology 140, 301 – 309 (1999). A. T. Singh, J. G. Kunnel, P. J. Strieleman, and P. H. Stern, Parathyroid hormone (PTH)-(1 – 34), [Nle(8,18),Tyr34]PTH-(3 – 34) amide, PTH-(1 – 31) amide, and PTH-related peptide-(1 – 34) stimulate phosphatidylcholine hydrolysis in UMR106 osteoblastic cells: Comparison with effects of phorbol 12,13-dibutyrate. Endocrinology 140, 131 – 137 (1999). P. Morley, J. F. Whitfield, and G. E. Willick, Design and applications of parathyroid hormone analogues. Curr. Med. Chem. 6, 1095 – 1106 (1999).

244 269. H. Kather, W. Tschöpe, and B. Simon, Human fat cell adenylate cyclase: Modulation of parathyroid hormone action by guanine nucleotides. Res. Exp. Med. 171, 201 – 204 (1977). 270. V. P. Michalangeli, N. H. Hunt, and T. J. Martin, States of activation of chick kidney adenylate cyclase induced by parathyroid hormone and guanyl nucleotides. J. Endocrinol. 72, 69 – 79 (1977). 271. E. Bellorin-Font and K. J. Martin, Regulation of the PTH-receptorcyclase system of canine kidney: Effects of calcium, magnesium, and guanine nucleotides. Am. J. Physiol. 241, F364 – F373 (1981). 272. R. E. Rizzoli, T. M. Murray, S. J. Marx, and G. D. Aurbach, Binding of radioiodinated bovine parathyroid hormone-(1 – 84) to canine renal cortical membranes. Endocrinology 112, 1303 – 1312 (1983). 273. A. P. Teitelbaum, R. A. Nissenson, and C. D. Arnaud, Coupling of the canine renal parathyroid hormone receptor to adenylate cyclase: Modulation by guanyl nucleotides and N-ethylmaleimide. Endocrinology 111, 1524 – 1533 (1982). 274. R. A. Nissenson, S. R. Abbott, A. P. Teitelbaum, O. H. Clark, and C. D. Arnaud, Endogenous biologically active human parathyroid hormone: Measurement by a guanyl nucleotide-amplified renal adenylate cyclase assay. J. Clin. Endocrinol Metab. 52, 840 – 846 (1981). 275. H. Jüppner, A. B. Abou-Samra, M. Freeman, X. F. Kong, E. Schipani, J. Richards, L. F. Kolakowski, Jr., J. Hock, J. T. Potts, Jr., H. M. Kronenberg, and G. V. Segre, A G protein-linked receptor for parathyroid hormone and parathyroid hormone-related peptide. Science 254, 1024 – 1026 (1991). 276. T. Schöneberg, G. Schultz, and T. Gudermann, Structural basis of G protein-coupled receptor function. Mol. Cell. Endocrinol. 151, 181 – 193 (1999). 277. C. D. Strader, T. M. Fong, M. P. Graziano, and M. R. Tota, The family of G-protein-coupled receptors. FASEB J. 9, 745 – 754 (1995). 278. M. Babich, H. Choi, R. M. Johnson, K. L. King, G. E. Alford, and R. A. Nissenson, Thrombin and parathyroid hormone mobilize intracellular calcium in rat osteosarcoma cells by distinct pathways. Endocrinology 129, 1463 – 1470 (1991). 279. H. Jüppner, E. Schipani, F. R. Bringhurst, I. McClure, H. T. Keutmann, J. T. Potts, Jr., H. M. Kronenberg, A. B. Abou-Samra, G. V. Segre, and T. J. Gardella, The extracellular amino-terminal region of the parathyroid hormone (PTH)/PTH-related peptide receptor determines the binding affinity for carboxyl-terminal fragments of PTH(1 – 34). Endocrinology 134, 879 – 884 (1994). 280. P. R. Turner, T. Bambino, and R. A. Nissenson, A putative selectivity filter in the G-protein-coupled receptors for parathyroid hormone and secretion. J. Biol. Chem. 271, 9205 – 9208 (1996). 281. C. Bergwitz, T. J. Gardella, M. R. Flannery, J. T. Potts, Jr., H. M. Kronenberg, S. R. Goldring, and H. Jüppner, Full activation of chimeric receptors by hybrids between parathyroid hormone and calcitonin: Evidence for a common pattern of ligand-receptor interaction. J. Biol. Chem. 271, 26469 – 26472 (1996). 282. T. J. Gardella, H. Jüppner, A. K. Wilson, H. T. Keutmann, A. B. Abou-Samra, G. V. Segre, F. R. Bringhurst, J. T. Potts, Jr., S. R. Nussbaum, and H. M. Kronenberg, Determinants of [Arg2]PTH-(1 – 34) binding and signaling in the transmembrane region of the parathyroid hormone receptor. Endocrinology 135, 1186 – 1194 (1994). 283. P. R. Turner, T. Bambino, and R. A. Nissenson, Mutations of neighboring polar residues on the second transmembrane helix disrupt signaling by the parathyroid hormone receptor. Mol. Endocrinol. 10, 132 – 139 (1996). 284. T. J. Gardella, M. D. Luck, M. H. Fan, and C. Lee, Transmembrane residues of the parathyroid hormone (PTH)/PTH-related peptide receptor that specifically affect binding and signaling by agonist ligands. J. Biol. Chem. 271, 12820 – 12825 (1996). 285. M. Mannstadt, M. D. Luck, T. J. Gardella, and H. Juppner, Evidence for a ligand interaction site at the amino-terminus of the parathyroid hormone (PTH)/PTH-related protein receptor from cross-linking and mutational studies. J. Biol. Chem. 273, 16890 – 16896 (1998).

ROBERT A. NISSENSON 286. A. T. Zhou, R. Bessalle, A. Bisello, C. Nakamoto, M. Rosenblatt, L. J. Suva, and M. Chorev, Direct mapping of an agonist-binding domain within the parathyroid hormone/parathyroid hormone-related protein receptor by photoaffinity crosslinking. Proc. Natl. Acad. Sci. USA 94, 3644 – 3649. 287. A. Bisello, A. E. Adams, D. F. Mierke, M. Pellegrini, M. Rosenblatt, L. J. Suva, and M. Chorev, Parathyroid hormone-receptor interactions identified directly by photocross-linking and molecular modeling studies. J. Biol. Chem. 273, 22498 – 22505 (1998). 288. S. P. Sheikh, J. P. Vilardarga, T. J. Baranski, O. Lichtarge, T. Iiri, E. C. Meng, R. A. Nissenson, and H. R. Bourne, Similar structures and shared switch mechanisms of the beta2-adrenoceptor and the parathyroid hormone receptor: Zn(II) bridges between helices III and VI block activation. J. Biol. Chem. 274, 17033 – 17041 (1999). 289. Z. Huang, Y. Chen, S. Pratt, T. H. Chen, T. Bambino, R. A. Nissenson, and D. M. Shoback, The N-terminal region of the third intracellular loop of the parathyroid hormone (PTH)/PTH-related peptide receptor is critical for coupling to cAMP and inositol phosphate/Ca2 signal transduction pathways. J. Biol. Chem. 271, 33382 – 33389 (1996). 290. A. Iida-Klein, J. Guo, M. Takemura, M. T. Drake, J. T. Potts, Jr., A. Abou-Samra, F. R. Bringhurst, and G. V. Segre, Mutations in the second cytoplasmic loop of the rat parathyroid hormone (PTH)/PTHrelated protein receptor result in selective loss of PTH-stimulated phospholipase C activity. J. Biol. Chem. 272, 6882 – 6889 (1997). 291. G. L. Wong, Induction of metabolic changes and down regulation of bovine parathyroid hormone-responsive adenylate cyclase are dissociable in isolated osteoclastic and osteoblastic bone cells. J. Biol. Chem. 254, 34 – 37 (1979). 292. W. A. Peck and G. Kohler, Hormonal and nonhormonal desensitization in isolated bone cells. Calcif. Tissue Int. 32, 95 – 103 (1980). 293. S. R. Goldring, J. M. Dayer, and M. Rosenblatt, Factors regulating the response of cells cultured from human giant cell tumors of bone to parathyroid hormone. J. Clin. Endocrinol. Metab. 53, 295 – 300 (1981). 294. W. I. Chao and L. R. Forte, Rat kidney cells in primary culture: Hormone-mediated desensitization of the adenosine 3,5-monophosphate response to parathyroid hormone and calcitonin. Endocrinology 111, 252 – 259 (1982). 295. H. L. Henry, N. S. Cunningham, and T. A. Noland, Jr., Homologous desensitization of cultured chick kidney cells to parathyroid hormone. Endocrinology 113, 1942 – 1949 (1983). 296. C. Bergwitz, A. B. Abou-Samra, R. D. Hesch, and H. Jüppner, Rapid desensitization of parathyroid hormone dependent adenylate cyclase in perifused human osteosarcoma cells (SaOS-2). Biochim. Biophys. Acta 1222, 447 – 456 (1994). 297. S. K. Lee and P. H. Stern, Studies on the mechanism of desensitization of the parathyroid hormone-stimulated calcium signal in UMR106 cells: Reversal of desensitization by alkaline phosphatase but not by protein kinase C downregulation. J. Bone Miner. Res. 9, 781 – 789 (1994). 298. A. Fujimori, A. Miyauchi, K. A. Hruska, K. J. Martin, L. V. Avioli, and R. Civitelli, Desensitization of calcium messenger system in parathyroid hormone-stimulated opossum kidney cells. Am. J. Physiol. 264, E918 – E924 (1993). 299. S. Fukayama, A. H. Tashjian, Jr., and F. R. Bringhurst, Mechanisms of desensitization to parathyroid hormone in human osteoblast-like SaOS-2 cells. Endocrinology 131, 1757 – 1769 (1992). 300. K. K. Pun, P. W. Ho, R. A. Nissenson, and C. D. Arnaud, Desensitization of parathyroid hormone receptors on cultured bone cells. J. Bone Miner. Res. 5, 1193 – 1200 (1990). 301. M. Bünemann and M. M. Hosey, G-protein coupled receptor kinases as modulators of G-protein signalling. J. Physiol. 517, 5 – 23 (1999). 302. R. J. Lefkowitz, G protein-coupled receptors. III. New roles for receptor kinases and beta-arrestins in receptor signaling and desensitization. J. Biol. Chem. 273, 18677 – 18680 (1998).

245

CHAPTER 7 PTH and PTH-Related Protein 303. J. G. Krupnick and J. L. Benovic, The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annu. Rev. Pharmacol. Toxicol. 38, 289 – 319 (1998). 304. E. Blind, T. Bambino, and R. A. Nissenson, Agonist-stimulated phosphorylation of the G protein-coupled receptor for parathyroid hormone (PTH) and PTH-related protein. Endocrinology 136, 4271 – 4277 (1995). 305. F. Qian, A. Leung, and A. Abou-Samra, Agonist-dependent phosphorylation of the parathyroid hormone/parathyroid hormone-related peptide receptor. Biochemistry 37, 6240 – 6246 (1998). 306. E. Blind, T. Bambino, Z. Huang, M. Bliziotes, and R. A. Nissenson, Phosphorylation of the cytoplasmic tail of the PTH/PTHrP receptor. J. Bone Miner. Res. 11, 578 – 586 (1996). 307. N. Malecz, T. Bambino, M. Bencsik, R. A. Nissenson, Identification of phosphorylation sites in the G-protein coupled receptor for parathyroid hormone: Receptor phosphorylation is not required for agonist-induced internalization. Mol. Endocrinol. 12, 1846 – 1856 (1998). 308. F. Dicker, U. Quitterer, R. Winstel, K. Honold, and M. J. Lohse, Phosphorylation-independent inhibition of parathyroid hormone receptor signaling by G protein-coupled receptor kinases. Proc. Natl. Acad. Sci. USA 96, 5476 – 5481 (1999). 309. S. Fukayama, G. Kong, J. L. Benovic, E. Meurer, and A. H. Tashjian, Jr., Beta-adrenergic receptor kinase-1 acutely regulates PTH/PTHrP receptor signalling in human osteoblastlike cells. Cell. Signal. 9, 469 – 474 (1997). 310. C. A. Mahoney and R. A. Nissenson, Canine renal receptors for parathyroid hormone. Down-regulation in vivo by exogenous parathyroid hormone. J. Clin. Invest. 72, 411 – 421 (1983). 311. A. P. Teitelbaum, C. M. Silve, K. O. Nyiredy, and C. D. Arnaud, Down-regulation of parathyroid hormone (PTH) receptors in cultured bone cells is associated with agonist-specific intracellular processing of PTH-receptor complexes Endocrinology 118, 595 – 602 (1986). 312. I. Yamamoto, C. Shigeno, J. T. Potts, Jr., and G. V. Segre, Characterization and agonist-induced down-regulation of parathyroid hormone receptors in clonal rat osteosarcoma cells. Endocrinology 122, 1208 – 1217 (1988). 313. J. Mitchell and D. Goltzman, Mechanisms of homologous and heterologous regulation of parathyroid hormone receptors in the rat osteosarcoma cell line UMR-106. Endocrinology 126, 2650 – 2660 (1990). 314. L. R. Forte, G. A. Nickols, and C. S. Anast, Renal adenylate cyclase and the interrelationship between parathyroid hormone and vitamin D in the regulation of urinary phosphate and adenosine cyclic 3,5monophosphate excretion. J. Clin. Invest. 57, 559 – 568 (1976). 315. S. G. Massry, R. Stein, J. Garty, A. I. Arieff, J. W. Coburn, A. W. Norman, and R. M. Friedler, Skeletal resistance to the calcemic action of parathyroid hormone in uremia: Role of 1,25 (OH)2 D3. Kidney Int. 9, 467 – 474 (1976). 316. I. G. Lewin, S. E. Papapoulos, G. N. Hendy, S. Tomlinson, and J. L. O’Riordan, Reversible resistance to the renal action of parathyroid hormone in human vitamin D deficiency. Clin. Sci. 62, 381 – 387 (1982). 317. G. A. Nickols, D. L. Carnes, C. S. Anast, and L. R. Forte, Parathyroid hormone-mediated refractoriness of rat kidney cyclic AMP system. Am. J. Physiol. 236, E401 – E409 (1979). 318. P. Ureña, M. Mannstadt, M. Hruby, A. Ferreira, F. Schmitt, C. Silve, R. Ardaillou, B. Lacour, A. B. Abou-Samra, G. V. Segre et al., Parathyroidectomy does not prevent the renal PTH/PTHrP receptor down-regulation in uremic rats. Kidney Int. 47, 1797 – 1805 (1995). 319. Z. Huang, Y. Chen, and R. A. Nissenson, The cytoplasmic tail of the G-protein-coupled receptor for parathyroid hormone and parathyroid hormone-related protein contains positive and negative signals for endocytosis. J. Biol. Chem. 270, 151 – 156 (1995). 320. S. L. Ferrari, V. Behar, M. Chorev, M. Rosenblatt, and A. Bisello, Endocytosis of ligand-human parathyroid hormone receptor 1

321.

322.

323.

324.

325.

326.

327.

328.

329.

330.

331.

332.

333.

334.

335.

complexes is protein kinase C-dependent and involves betaarrestin2. Real-time monitoring by fluorescence microscopy. J. Biol. Chem. 274, 29968 – 29975 (1999). E. A. González and K. J. Martin, Coordinate regulation of PTH/PTHrP receptors by PTH and calcitriol in UMR 106-01 osteoblast-like cells. Kidney Inte. 50, 63 – 70 (1996). J. W. Jongen, E. C. Willemstein-van Hove, J. M. van der Meer, M. P. Bos, H. Jüppner, G. V. Segre, A. B. Abou-Samra, J. H. Feyen, and M. P. Herrmann-Erlee, Down-regulation of the receptor for parathyroid hormone (PTH) and PTH-related peptide by PTH in primary fetal rat osteoblasts. J. Bone Miner. Res. 11, 1218 – 1225 (1996). G. Turner, C. Coureau, M. R. Rabin, B. Escoubet, M. Hruby, O. Walrant, and C. Silve, Parathyroid hormone (PTH)/PTH-related protein receptor messenger ribonucleic acid expression and PTH response in a rat model of secondary hyperparathyroidism associated with vitamin D deficiency. Endocrinology 136, 3751 – 3758 (1995). L. K. McCauley, C. A. Beecher, M. E. Melton, J. R. Werkmeister, H. Jüppner, A. B. Abou-Samra, G. V. Segre, and T. J. Rosol, Transforming growth factor-beta1 regulates steady-state PTH/PTHrP receptor mRNA levels and PTHrP binding in ROS 17/2.8 osteosarcoma cells. Mol. Cell. Endocrinol. 101, 331 – 336 (1994). J. W. Jongen, E. C. Willemstein-Van Hove, J. M. Van der Meer, M. P. Bos, H. Jüppner, G. V. Segre, A. B. Abou-Samra, J. H. Feyen, and M. P. Herrmann-Erlee, Down-regulation of the receptor for parathyroid hormone (PTH) and PTH-related peptide by transforming growth factor-beta in primary fetal rat osteoblasts. Endocrinology 136, 3260 – 3266 (1995). F. Law, J. P. Bonjour, and R. Rizzoli, Transforming growth factorbeta: A down-regulator of the parathyroid hormone-related protein receptor in renal epithelial cells. Endocrinology 134, 2037 – 2043 (1994). P. Ureña, A. Iida-Klein, X. F. Kong, H. Jüppner, H. M. Kronenberg, A. B. Abou-Samra, and G. V. Segre, Regulation of parathyroid hormone (PTH)/PTH-related peptide receptor messenger ribonucleic acid by glucocorticoids and PTH in ROS 17/2.8 and OK cells. Endocrinology 134, 451 – 456 (1994). J. Yaghoobian and T. B. Drüeke, Regulation of the transcription of parathyroid-hormone/parathyroid-hormone-related peptide receptor mRNA by dexamethasone in ROS 17/2.8 osteosarcoma cells. Nephrol. Dialysis Transplant. 13, 580 – 586 (1998). H. Wald, M. Dranitzki-Elhalel, R. Backenroth, and M. M. Popovtzer, Evidence for interference of vitamin D with PTH/PTHrP receptor expression in opossum kidney cells. Pflug. Arch. Eur. J. Physiol. 436, 289 – 294 (1998). J. Cornish, K. E. Callon, G. C. Nicholson, and I. R. Reid, Parathyroid hormone-related protein-(107 – 139) inhibits bone resorption in vivo. Endocrinology 138, 1299 – 1304 (1997). J. Cornish, K. E. Callon, C. Lin, C. Xiao, J. M. Moseley, and I. R. Reid, Stimulation of osteoblast proliferation by C-terminal fragments of parathyroid hormone-related protein. J. Bone Miner. Res. 14, 915 – 922 (1999). F. De Miguel, P. Martinez-Fernandez, C. Guillen, A. Valin, A. Rodrigo, M. E. Martinez, and P. Esbrit, Parathyroid hormone-related protein (107 – 139) stimulates interleukin-6 expression in human osteoblastic cells. J. Am. Soc. Nephrol. 10, 796 – 803 (1999). A. D. Care, S. K. Abbas, D. W. Pickard, M. Barri, M. Drinkhill, J. B. Findlay, I. R. White, and I. W. Caple, Stimulation of ovine placental transport of calcium and magnesium by mid-molecule fragments of human parathyroid hormone-related protein. Exp. Physiol. 75, 605 – 608 (1990). M. T. Nguyen and A. C. Karaplis, The nucleus: A target site for parathyroid hormone-related peptide (PTHrP) action. J. Cell. Biochem. 70, 193 – 199 (1998). M. H. Lam, L. J. Briggs, W. Hu, T. J. Martin, M. T. Gillespie, and D. A. Jans, Importin beta recognizes parathyroid hormone-related

246 protein with high affinity and mediates its nuclear import in the absence of importin alpha. J. Biol. Chem. 274, 7391 – 7398 (1999). 336. M. M. Aarts, D. Levy, B. He, S. Stregger, T. Chen, S. Richard, and J. E. Henderson, Parathyroid hormone-related protein interacts with RNA. J. Biol. Chem. 274, 4832 – 4838 (1999). 337. J. E. Henderson, N. Amizuka, H. Warshawsky, D. Biasotto, B. M. Lanske, D. Goltzman, and A. C. Karaplis, Nucleolar localization of parathyroid hormone-related peptide enhances survival of chondrocytes under conditions that promote apoptotic cell death. Mol. Cell. Biol. 15, 4064 – 4075 (1995). 338. T. Massfelder, P. Dann, T. L. Wu, R. Vasavada, J. J. Helwig, and A. F. Stewart, Opposing mitogenic and anti-mitogenic actions of parathyroid hormone-related protein in vascular smooth muscle

ROBERT A. NISSENSON cells: A critical role for nuclear targeting. Proc. Natl. Acad. Sci. USA 94, 13630 – 13635 (1997). 339. M. H. Lam, S. L. Olsen, W. A. Rankin, P. W. Ho, T. J. Martin, M. T. Gillespie, and J. M. Moseley, PTHrP and cell division: Expression and localization of PTHrP in a keratinocyte cell line (HaCaT) during the cell cycle. J. Cell. Physiol. 173, 433 – 446 (1997). 340. M. H. Lam, C. M. House, T. Tiganis, K. I. Mitchelhill, B. Sarcevic, A. Cures, R. Ramsay, B. E. Kemp, T. J. Martin, and M. T. Gillespie, Phosphorylation at the cyclin-dependent kinases site (Thr85) of parathyroid hormone-related protein negatively regulates its nuclear localization. J. Biol. Chem. 274, 18559 – 18566 (1999). 341. D. Endres, R. Villanueva, C. F. Sharp, Jr., and F. R. Singer, Measurement of parathyroid hormone. Endocrinol. Metab. Clin. North Am. 18, 611 – 629 (1989).

CHAPTER 8

Calcitonin ANA O. HOFF, GILBERT J. COTE, AND ROBERT F. GAGEL Department of Endocrine Neoplasia and Hormonal Disorders, Divison of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030

I. II. III. IV.

V. Physiology and Mechanism of Action VI. The Calcitonin Receptor Superfamily VII. Summary References

Introduction Calcitonin Gene Products Calcitonin Structure/Function Relationships Secretion and Metabolism

processed to produce only CGRP. The third is a pseudogene and is not expressed. The fourth encodes the peptide amylin. In addition, a fifth gene encoding a vasodilator named adrenomedullin will be discussed because of its homology to CGRP and interactions with a family of CT-like receptors. This chapter focuses on calcitonin and therefore only the transcriptional and posttranscriptional regulation of the CALC I or CGRP gene will be discussed. The CALC I or CGRP gene is expressed in a broad spectrum of neuroendocrine cells, but is predominantly produced by the C or calcitonin-producing cell of the thyroid gland. The C cell migrates from the neural crest during embryonic life to a position in the thyroid gland. In other species (fish and birds), C cells are located in a discrete organ in the neck region called the ultimobranchial body [5]. More than 90% of the calcitonin circulating in plasma is normally produced by the thyroidal C cell; the remaining 10% or less is produced in a variety of neuroendocrine cells in the pituitary, lung, adrenal glands, prostate, and other sites. A cDNA for CT was identified in 1981 [6]; subsequent investigators identified a second mRNA with partial sequence homology to the CT cDNA and named it calcitonin gene-related peptide [7]. The 5 portions of this cDNA are identical to similar portions of the CT cDNA, whereas the 3 portion encodes CGRP. Further elucidation of the

I. INTRODUCTION In 1961 Copp made the observation that a protein extract derived from the ultimobranchial body of salmon caused a lowering of the serum calcium when injected into a rodent [1, 2]. He named this substance calcitonin (CT). In mammalian species this activity was subsequently localized to the thyroid gland [3] and shown to be produced by the C or calcitonin-producing cells of the thyroid gland [4]. The active principle was subsequently shown to be a 32 amino acid peptide, released in response to an increase in the plasma calcium concentration. Calcitonin interacts with a specific G-protein-coupled receptor on the osteoclast surface and causes inhibition of bone resorption.

II. CALCITONIN GENE PRODUCTS The calcitonin gene family includes at least four separate genes. The first (CALC I or CGRP) encodes calcitonin and a second peptide, calcitonin gene-related peptide (CGRP) (Fig. 1). The second gene (CALC II or CGRP), undoubtedly a duplication of the first, has a similar structure to the first, but the primary RNA transcript of this gene is

OSTEOPOROSIS, SECOND EDITION VOLUME 1

247

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

248

HOFF, COTE, AND GAGEL

FIGURE 1 Calcitonin gene expression. The CT/CGRP gene is transcribed in a variety of cell types, including thyroidal C cells and neurons. The primary transcript of CT/CGRP is processed to include (CT-specific pathway) or exclude (CGRP-specific pathway) exon 4 encoding CT. Following translation of either CT or CGRP mRNA, the propeptide (proCT or proCGR) undergoes additional proteolytic cleavage. CT or CGRP is cleaved from the central portion of each propeptide and generates amino-and carboxy-terminal peptides (N- and C-proCT or N- and C-proCGRP).

genomic structure of the CT/CGRP gene led to the recognition that CT or CGRP is produced as a result of alternative RNA processing (Fig. 1). Following transcription there is constitutive RNA processing that combines exons 1 – 4 of the CT transcript to produce a mRNA encoding proCT (with polyadenylation resulting in truncation of the mRNA immediately following exon 4), a processing pathway that is used predominantly in the thyroidal C cell. A second processing choice results in the combining of exons 1 – 3 and 5 and 6 to produce a mRNA encoding CGRP. This processing pathway results in the exclusion of exon 4 (containing the sequences encoding CT) from the final mRNA, a cell-specific processing choice that occurs mainly in neural types of cells in the brain, spinal cord, and various sympathetic ganglia. In the

normal thyroidal C cell, 99% of the primary RNA transcript is processed down a CT-specific pathway; in neuronal tissue, 95% of the primary transcript is processed in a CGRPspecific manner [8 – 11]. The regulatory pathways that control this alternative splice in the C cell are complex and likely involve multiple factors that enhance (CT-specific pattern) or suppress (CGRP-specific pattern) recognition of exon 4 containing the coding sequence for calcitonin. Only one set of regulatory factors has been identified, those involved in recognition and polyadenylation of exon 4 (Fig. 1) [12]. Both CT and CGRP mRNAs encode propeptides that are subsequently processed further [13]. The CT monomer (the hypocalcemic factor identified by Copp) is located centrally

249

CHAPTER 8 Calcitonin

in the proCT molecule. Cleavage of the CT monomer results in the production of two additional peptides: amino (N)-and carboxy (C)-terminal proCT (Fig. 1). Although CproCT has been shown to have some bone effects, it is unclear at present whether these effects are significant. CGRP is similarly cleaved from a central location within the proCGRP molecule, producing an amino- and carboxy-terminal fragment. There is considerable similarity between the structure of monomeric CT or CGRP. Both contain an amino-terminal ring structure defined by disulfide bonds, considerable amino-terminal homology, and both are amidated. These findings suggest that either CT or CGRP originated as a result of an intragenic duplication. The only product derived from the second CT/CGRP gene is CGRP, a peptide with a high degree of homology to CGRP. Although there is a CT-like exon (exon 4) in the second gene, it is not included in the final mRNA. Therefore, the only product from this gene is CGRP. Amylin is the secretory product of the fourth calcitonin gene. This gene is expressed predominantly in islet cells of the pancreas where amyloid composed of amylin  sheets has been implicated in the pathogenesis of diabetes mellitus. There is no known function of amylin in calcium metabolism, although its interaction with the CT and CGRP receptors is discussed later. Adrenomedullin is a 52 amino acid peptide with an amino-terminal ring structure and approximately 25% homology to CGRP. It is a potent vadodilator [14]. Although it has no known relationship to calcium metabolism and is expressed from a separate gene, it is considered a member

of the CT/CGRP superfamily because of its biological similarity to CGRP and its interaction with the CT/CGRP family of receptors [15,16].

III. CALCITONIN STRUCTURE/ FUNCTION RELATIONSHIPS Based on primary structure, the CTs can be divided into three groups: artiodactyl, including porcine, bovine, and ovine, which differ by four amino acids; primate/rodent, including human and rat calcitonins, which differ by two amino acids; and teleost/avian, including salmon, eel, goldfish, and chicken calcitonins, which differ by three amino acids (Fig. 2). In several different biological assays, the order of potency of the CTs is teleost  artiodactyl  human, although the absolute potency of each is dependent on the species studied. Fish and avian CTs are products of the ultimobranchial glands, which remain in those species as discrete organs, whereas in mammals, C cells migrate from the neural crest to be dispersed within the developing thyroid gland. Studies of substituted, deleted, and otherwise modified CTs have provided considerable information regarding structure/activity relationships of the CT molecule. For example, although the ring structure appears to protect and stabilize the molecule, linear analogues of salmon CT retain full hypocalcemic activity and adenylate cyclase activation [17,18]. Consistent with this, increased stability of the ring (as in aminosuberic 1 – 7 eel calcitonin) increases biological stability and potency. The carboxy-terminal

FIGURE 2 Amino acid alignment of calcitonins from various species. All calcitonins have a cysteine (C) at positions 1 and 7 linked by a disulfide bridge, a glycine (G) at position 28, and a proline amide (P) at position 32. Residues 4, 5, and 6 are conserved in all species. Residues identical to salmon calcitonin sCT) are boxed. eCT, eel calcitonin; gCT, goldfish calcitonin; stCT, stingray calcitonin; cCT, chicken calcitonin; pCT, porcine calcitonin; oCT, ovine calcitonin; bCT, bovine calcitonin; dCT, dog calcitonin; hCT, human calcitonin; rCT, rat calcitonin; and rbCT, rabbit calcitonin.

250 TABLE 1

HOFF, COTE, AND GAGEL

General Structure-Function Relationships for Monmeric Calcitonin

Integrity of the amino-terminal ring structure is important for the stability in circulation in some species Greater stability of the ring (as in aminosuberic 1–7 eel calcitonin) enhances potency Full biological activity can be preserved despite substitutions or deletions of one residue within the ring The sequence Leu-Ser-Thr–Cys at positions 4–7 in the ring structure is invariant Deletions in other parts of the molecule may impair biological activity, depending on spatial arrangements of hydrophobic–hydrophilic residues Single or multiple substitutions in the human CT molecule can enhance biological activity Carboxyl-terminal proline amide is essential for activity

proline amide is essential, but numerous changes are tolerated in the 8 – 22 domain. General observations regarding the biological activity of the various CT species are outlined in Table 1.

IV. SECRETION AND METABOLISM Calcitonin secretion is regulated by the extracellular calcium concentration. Increasing the extracellular calcium concentration results in increased secretion of CT [19]. This effect is mediated by the interaction of calcium with the calcium-sensing receptor (CaR), a G-protein-coupled receptor [20] expressed in high levels on the thyroidal C cell [21,22]. In contrast to the parathyroid cell, where increasing calcium concentration activates the CaR and links to pathways that inhibit parathyroid hormone secretion, activation of this same receptor in the C cell causes increased secretion of CT [23,24]. Thus the same receptor system is linked differently in two cell types, a useful physiologic adaptation that results in reduced secretion of the parathyroid hormone (a hormone that raises the serum calcium concentration) and enhanced secretion of CT (a hormone that lowers serum calcium concentration) in the presence of hypercalcemia. Other peptides, including glucagon, gastrin, cholecystokinin, secretin, and vasoactive intestinal peptide, also stimulate CT release from the C cell, although the physiologic relevance of these interactions is unclear [25]. Glucagon appears to act through a cAMP-dependent pathway; the mechanism by which gastrin stimulates CT release is less clear. Therapy with omeprazole and other proton pump inhibitors that inhibit gastric acid production increases CT secretion through increased gastrin secretion. A broad spectrum of evidence indicates that voltagegated calcium channels and the release of calcium from

intracellular pools are important in both calcium and cAMP-mediated hormone release [23,26]. Alcohol and exercise stimulate CT release [27], although the mechanisms are not understood. Pentagastrin is used routinely to detect CT abnormalities in patients with medullary thyroid carcinoma [28,29]. Calcitonin gene expression is also affected by a variety of factors. Activation of cAMP and protein kinase C pathways [30,31] and glucocorticoid treatment [32,33] increases transcription of the CT gene, whereas 1,25dihydroxy vitamin D3 inhibits transcription [34]. During the past decade a body of evidence has developed that serum CT concentrations are increased in sepsis. Although there was initial skepticism about whether the immunoreactive material was CT, studies using more specifi assays indicate that the circulating peptide is proCT rather than CT 1 – 32 (monomeric CT) and that the CT/CGRP gene is transcribed in infected tissues [35]. Even more provocative are results that show infected animals have greater survival if the proCT is neutralized by passive immunization [36]. Circulating plasma or serum CT values are higher in men than in women [37,38] and decrease with age [37,38], although these differences disappear when more specifi CT assays are utilized to measure calcitonin [37]. There are large increases in plasma immunoreactive CT during pregnancy and lactation, and it has been proposed that protection of the maternal skeleton from excessive calcium depletion may be the major physiological role of CT [39]. Estrogen therapy has been reported to increase basal CT concentrations in postmenopausal women [40], although studies utilizing a sensitive extraction assay for CT have not confirmed these observations [41].

V. PHYSIOLOGY AND MECHANISM OF ACTION Calcitonin and its related peptides (CGRP, CGRP amylin, and adrenomedullin) are small peptides that exert their biological effect by binding to and activating a family of G-protein-coupled receptors (Fig. 3). The first identified member of this family was the CT receptor (CTR) [42,43]. This receptor is expressed on osteoclasts, renal tubular cells, neurons in the central and peripheral nervous system, and other cell types [44]. Several different isoforms of the CT receptor are produced by alternative RNA processing [45,46], which couple differently to downstream transduction pathways [47,48]. The CTR is expressed on the osteoclast during the final stages of its differentiation. Binding of CT to the osteoclast CTR results in the activation of several downstream signaling pathways, leading to a decrease in bone resorption. This has been demonstrated in several ways. Addition of

251

CHAPTER 8 Calcitonin

FIGURE 3

The relationship between the calcitonin (CTR) or CT-like receptor (hCTR2 or rCTLR) and receptor activity modifying proteins (RAMP). Current information indicates that different combinations of CTR2 or CTLR and RAMPs 1 – 3 create receptors for CGRP, adrenomedullin, and amylin.

CT to in vitro bone culture assays (consisting of calvarial or long bones from newborn mice) inhibits bone resorption caused by a variety of resorptive agents, including parathyroid hormone, parathyroid hormone-related protein, interleukins, 1,25-dihydroxy vitamin D, and prostaglandins [49]. The effect of CT is of limited duration in these systems. Within 48 h following repetitive administration of CT to the bone culture system, there is a loss of the inhibitory effect of CT to prevent resorptive agent-mediated bone resorption termed “escape” [50]. This effect appears to be mediated by a decrease in CTR caused by some

combination of receptor recycling [51,52] and reduced synthesis of CTR [53 – 55]. Studies of isolated osteoclasts have also demonstrated rapid effects of CT. One model system utilizes isolated osteoclasts placed on coverslips or bone slices [56]. Within minutes after exposure to CT at concentrations as low as 3 pg/ml there is a retraction of the osteoclast ruffled border from its attachment points and decreased bone resorption [56]. Other studies have demonstrated reductions in the production of proteolytic enzymes, an increase in the pH of the contents of the resorption pit induced by CT, and

252

HOFF, COTE, AND GAGEL

alterations in the integrins, which mediate attachment of the osteoclast to the bone surface. Studies in a transgenic animal model in which the CT/CGRP gene has been deleted by homologous recombination provide additional support for the acute effect of CT to inhibit bone resorption. In mice treated with 1,25dihydroxy vitamin D there is greater hypercalcemia in CT/CGRP -/- mice than in wild-type control animals. Further evidence that the hypercalcemia is mediated by increased bone resorption is provided by the demonstration that urine deoxypyridinoline crosslinks are severalfold greater in CT/CGRP -/- mice than in wild-type control animals [57]. Collectively, data accumulated since the discovery in the 1960s argue for a role of CT to protect against hypercalcemia [58]. Acute increases in the serum calcium concentration cause release of CT from C cells. Binding of CT to the CTR on the osteoclast causes rapid inhibition of bone resorption. The effect of CT on the osteoclast is transient

with reversal of the effect within 48 h, exactly the type of response that is optimal for maintenance of the serum calcium concentration within a narrow range. Studies in CT/CGRP -/- mice hint at other effects of the CT/CGRP gene. CT/CGRP -/- mice have a greater bone mass at 1 and 3 months. Furthermore, in oophorectomized CT/CGRP -/- mice, there is preservation of bone mass, whereas wild-type control mice lose the expected amount of bone mass under the same conditions. The preservation of bone mass appears to be mediated by increased bone formation in the oophorectomized CT/ CGRP -/- mice [57]. Interestingly, there is no evidence of increased bone resorption in CT/CGRP -/- mice. These studies suggest a more complex effect of the CT/CGRP gene that includes effects on bone formation. These studies are currently in their infancy, but provide provocative new information about a broader role for this gene. CT also has effects on renal tubular function, causing natriuresis through an effect on the Na/H exchanger and

FIGURE 4 Calcitonin receptor isoforms of rat (top) and human (bottom) with extracellular domains (e1 to e4) and intracellular domains (i1 to i4) indicated. The C1b isoform contains a 37 amino acid insert in e2, compared to C1a. The human CTR cloned from BIN67 cells contains a 16 amino acid insert in i1 compared with the receptor cloned from T47D cells.

253

CHAPTER 8 Calcitonin

Na,K-ATPase [59]. Binding of CT to the renal CTR results in the activation of protein kinase A and C pathways in a cell cycle-dependent manner [60]. In addition, there is also activation of the ERK 1/2 pathways [61]. The exact physiologic role of CT in Na transport has been more difficult to establish. CT/CGRP -/- mice have no identifiable electrolyte abnormalities, although detailed studies of tubular function have not been performed in these animals.

VI. THE CALCITONIN RECEPTOR SUPERFAMILY The CTR superfamily includes receptors for CT, CGRP, amylin, and adrenomedullin (Fig. 3). The CTR was the first identified member of this receptor family [43, 62] and is unique in that it functions without an accessory protein (Fig. 3). Efforts to identify a receptor for CGRP, amylin, and adrenomedullin were unsuccessful until it was recognized that a family of receptor activity-modifying proteins (RAMPS) join with CT-like receptors (CTLR) to form receptors for these proteins. Coexpression of RAMP 1, 2, or 3 with a CT-like receptor (hCTR2 or rCTLR) creates highaffinity binding sites for CGRP [63], amylin [63], and adrenomedullin [64 – 66]. A current view of how these receptor proteins interact with each other and the four ligands is shown in Fig. 3. The CTR RNA is further modified, presumably by alternative RNA processing, to produce several unique forms of the receptor that differ by insertion or deletion of small components of the receptor (Fig. 4). The alternative RNA processing results in two major receptor modifications. The first is the insertion of a 37 amino acid in the second extracellular domain (designated C1b in Fig. 4). Cells expressing the C1b variant have a lower adenylate cyclase response [67]. In the human there is a second receptor variant with a 16 amino acid insert in the first intracellular domain (Fig. 4) [48]. The most commonly expressed form of the human CTR appears to be the variant without the 16 amino acid insert (Fig. 4, T47D). Analysis suggests that the presence of the 16 amino acid insert in the first intracellular loop of the hCTR results in greater affinity for CT binding, but lowers phospholipase D and protein kinase C activation. This effect is thought to be related to a modification of a G-protein-binding sequence by the 16 amino acid insert [68].

VII. SUMMARY In summary, the CTR is a member of the class of seven trans-membrane G-protein-coupled receptors. CTRs in osteoclasts and models of renal tubular epithelium link to both protein kinase A- and C-dependent pathways. In the

osteoclast the CTR links to cellular processes, leading to osteoclast detachment from the matrix surface and cessation of bone resorption; in the renal tubular cell, activation of the CTR affects sodium transport and cell growth.

References 1. D. H. Copp, A. G. F. Davidson, and B. A. Cheney, Evidence for a new parathyroid hormone which lowers blood calcium. Proc. Can. Fed. Biol. Soc. 4, 17 (1961). 2. D. H. Copp, E. C. Cameron, B. A. Cheney, A. G. F. Davidson, and K. G. Henze, Evidence for calcitonin-a new hormone from the parathyroid that lowers blood calcium. Endocrinology 70, 638 – 649 (1962). 3. P. F. Hirsch, G. F. Gauthier, and P. L. Munson, Thyroid hypocalcemic principle and recurrent laryngeal nerve injury as factors affecting the response to parathyroidectomy in rats. Endocrinology 73, 244 – 252 (1963). 4. G. V. Foster, A. Baghdiantz, M. A. Kumar, E. Slack, H. A. Soliman, and I. MacIntyre, Thyroid origin of calcitonin. Nature (London) 202, 1303 – 1305 (1964). 5. C. S. Le-Lievre and N. M. Le-Douarin, Mesenchymal derivatives of the neural crest: Analysis of chimaeric quail and chick embryos. J Embryol. Exp. Morphol. 34, 125 – 54 (1975). 6. J. W. Jacobs, R. H. Goodman, W. W. Chin, P. C. Dee, J. F. Habener, N. H. Bell, and J. T. Potts, Jr., Calcitonin messenger RNA encodes multiple polypeptides in a single precursor. Science 213, 457 – 459 (1981). 7. S. G. Amara, V. Jonas, M. G. Rosenfeld, E. S. Ong, and R. M. Evans, Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 298, 240 – 244 (1982). 8. H. Lou, K. M. Neugebauer, R. F. Gagel, and S. M. Berget, Regulation of alternative polyadenylation by U1 snRNPs and SRp20. Mol. Cell. Biol. 18, 4977 – 4985 (1998). 9. H. Lou and R. F. Gagel, Alternative RNA processing: Its role in regulating expression of calcitonin/calcitonin gene-related peptide. J. Endocrinol. 156, 401 – 405 (1998). 10. H. Lou, D. M. Helfman, R. F. Gagel, and S. M. Berget, Polypyrimidine tract-binding protein positively regulates inclusion of an alternative 3-terminal exon. Mol. Cell. Biol. 19, 78 – 85 (1999). 11. J. M. Yeakley, F. Hedjran, J. P. Morfin, N. Merillat, M. G. Rosenfeld, and R. B. Emeson, Control of calcitonin/calcitonin gene-related peptide pre-mRNA processing by constitutive intron and exon elements. Mol. Cell. Biol. 13, 5999 – 6011 (1993). 12. H. Lou, R. F. Gagel, and S. M. Berget, An intron enhancer recognized by splicing factors activates polyadenylation. Genes Dev. 10, 208 – 219 (1996). 13. M. Zaidi, B. S. Moonga, P. J. Bevis, Z. A. Bascal, and L. H. Breimer, The calcitonin gene peptides: Biology and clinical relevance. Crit. Rev. Clin. Lab. Sci. 28, 109 – 174 (1990). 14. K. Kitamura, J. Sakata, K. Kangawa, M. Kojima, H. Matsuo, and T. Eto, Cloning and characterization of cDNA encoding a precursor for human adrenomedullin. Biochem. Biophys. Res. Commun. 194, 720 – 725 (1993). 15. S. J. Wimalawansa, Amylin, calcitonin gene-related peptide, calcitonin, and adrenomedullin: A peptide superfamily. Crit. Rev. Neurobiol. 11, 167 – 239 (1997). 16. R. Muff, W. Born, and J. A. Fischer, Calcitonin, calcitonin generelated peptide, adrenomedullin and amylin: Homologous peptides, separate receptors and overlapping biological actions. Eur. J. Endocrinol. 133, 17 – 20 (1995).

254 17. P. M. Sexton, H. G. Schneider, C. S. DSantos, F. A. Mendelsohn, B. E. Kemp, J. M. Moseley, T. J. Martin, and D. M. Findlay, Reversible calcitonin binding to solubilized sheep brain binding sites. Biochem. J. 273, 179 – 184 (1991). 18. R. C. Orlowski, R. M. Epand, and A. R. Stafford, Biologically potent analogues of salmon calcitonin which do not contain an N-terminal disulfide-bridged ring structure. Eur. J. Biochem. 162, 399 – 402 (1987). 19. C. W. Cooper, L. J. Deftos, and J. T. Potts, Jr., Direct measurement of in vivo secretion of pig thyrocalcitonin by radioimmunoassay. Endocrinology 88, 747 – 754 (1971). 20. E. M. Brown, G. Gamba, D. Riccardi, M. Lombardi, R. Butters, O. Kifor, A. Sun, M. A. Hediger, J. Lytton, and S. C. Hebert, Cloning and characterization of an extracellular Ca(2)-sensing receptor from bovine parathyroid. Nature 366, 575 – 580 (1993). 21. J. E. Garrett, H. Tamir, O. Kifor, R. T. Simin, K. V. Rogers, A. Mithal, R. F. Gagel, and E. M. Brown, Calcitonin-secreting cells of the thyroid express an extracellular calcium receptor gene. Endocrinology 136, 5202 – 5211 (1995). 22. M. Freichel, A. Zink-Lorenz, A. Holloschi, M. Hafner, V. Flockerzi, and F. Raue, Expression of a calcium-sensing receptor in a human medullary thyroid carcinoma cell line and its contribution to calcitonin secretion. Endocrinology 137, 3842 – 3848 (1996). 23. F. Raue and H. Scherubl, Extracellular calcium sensitivity and voltage-dependent calcium channels in C cells. Endocr. Rev. 16, 752 – 764 (1995). 24. D. S. McGehee, M. Aldersberg, K. P. Liu, S. Hsuing, M. J. Heath, and H. Tamir, Mechanism of extracellular Ca2 receptor-stimulated hormone release from sheep thyroid parafollicular cells. J. Physiol. 502, 31 – 44 (1997). 25. B. A. Roos, L. J. Deftos, and P. Rorabaugh, Calcitonin secretion in vitro. II. Regulatory effects of enteric mammalian polypeptide hormones on trout C-cell cultures. Endocrinology 98, 1284 – 1288 (1976). 26. R. M. Fried and A. H. J. Tashjian, Unusual sensitivity of cytosolic free Ca2 to changes in extracellular Ca2 in rat C-cells. J. Biol. Chem. 261, 7669 – 7674 (1986). 27. J. F. Aloia, P. Rasulo, L. J. Deftos, A. Vaswani, and J. K. Yeh, Exercise-induced hypercalcemia and the calciotropic hormones. J. Lab. Clin. Med. 106, 229 – 232 (1985). 28. J. F. Hennessey, S. A. Wells, D. A. Ontjes, and C. W. Cooper, A comparison of pentagastrin injections and calcium infusion as provocative agents for the detection of medullary carcinoma of the thyroid. J. Clin. Endocr. Metab. 39, 487 – 495 (1974). 29. R. F. Gagel, A. H. Tashjian, Jr., T. Cummings, N. Papathanasopoulos, M. M. Kaplan, R. A. DeLellis, H. J. Wolfe, and S. Reichlin, The clinical outcome of prospective screening for multiple endocrine neoplasia type 2a: An 18-year experience. N. Engl. J. Med. 318, 478 – 484 (1988). 30. Y. T. Monla, S. Peleg, and R. F. Gagel, Cell type-specific regulation of transcription by cyclic adenosine 3,5-monophosphateresponsive elements within the calcitonin promoter. Mol. Endocrinol. 9, 784 – 793 (1995). 31. A. de Bustros, S. B. Baylin, C. L. Berger, B. A. Roos, S. S. Leong, and B. D. Nelkin, Phorbol esters increase calcitonin gene transcription and decrease c-myc mRNA levels in cultured human medullary thyroid carcinoma. J. Biol. Chem. 260, 98–104 (1985). 32. M. Muszynski, R. S. Birnbaum, and B. A. Roos, Glucocorticoids stimulate the production of preprocalcitonin-derived secretory peptides by a rat medullary thyroid carcinoma cell line. J. Biol. Chem. 258, 11678 – 11683 (1983). 33. G. J. Cote and R. F. Gagel, Dexamethasone differentially affects the levesl of calcitonin and calcitonin gene-related peptide mRNAs expressed in a human medullary thyroid carcinoma cell line. J. Biol. Chem. 261, 15524–15528 (1986).

HOFF, COTE, AND GAGEL 34. G. J. Cote, D. G. Rogers, E. S. Huang, and R. F. Gagel, The effect of 1,25-dihydroxyvitamin D3 treatment on calcitonin and calcitonin gene-related peptide mRNA levels in cultured human thyroid C-cells. Biochem. Biophys. Res. Commun. 149, 239 – 243 (1987). 35. R. H. Snider, Jr., E. S. Nylen, and K. L. Becker, Procalcitonin and its component peptides in systemic inflammation: Immunochemical characterization. J. Invest. Med. 45, 552 – 560 (1997). 36. E. S. Nylen, K. T. Whang, R. H. Snider, Jr., P. M. Steinwald, J. C. White, and K. L. Becker, Mortality is increased by procalcitonin and decreased by an antiserum reactive to procalcitonin in experimental sepsis. Crit. Care Med. 26, 1001 – 1006 (1998). 37. R. D. Tiegs, J. J. Body, J. M. Barta, and H. D. Heath, Secretion and metabolism of monomeric human calcitonin: Effects of age, sex, and thyroid damage. J. Bone Miner. Res. 1, 339 – 349 (1986). 38. L. J. Deftos, M. H. Weisman, G. W. Williams, D. B. Karpf, A. M. Frumar, B. J. Davidson, J. G. Parthemore, and H. L. Judd, Influence of age and sex on plasma calcitonin in human beings. N. Engl. J. Med. 302, 1351 – 1353 (1980). 39. J. C. Stevenson, C. J. Hillyard, I. MacIntyre, H. Cooper, and M. I. Whitehead, A physiological role for calcitonin: Protection of the maternal skeleton. Lancet 2, 769 – 770 (1979). 40. J. C. Stevenson, C. J. Hillyard, G. Abeyasekera, K. G. Phang, and I. MacIntyre, Calcitonin and the calcium-regulating hormones in postmenopausal women: Effect of oestrogens. Lancet 1, 693 – 695 (1981). 41. D. L. Hurley, R. D. Tiegs, J. Barta, K. Laakso, and H. Heath III, Effects of oral contraceptive and estrogen administration on plasma calcitonin in pre- and postmenopausal women. J. Bone Miner. Res. 4, 89 – 95 (1989). 42. H. Y. Lin, T. L. Harris, M. S. Flannery, A. Aruffo, E. H. Kaji, A. Gorn, L. F. Kolakowski, Jr., M. Yamin, H. F. Lodish, and S. R. Goldring, Expression cloning and characterization of a porcine renal calcitonin receptor. Trans. Assoc. Am. Phys. 104, 265 – 272 (1991). 43. H. Y. Lin, T. L. Harris, M. S. Flannery, A. Aruffo, E. H. Kaji, A. Gorn, L. F. Kolakowski, Jr., H. F. Lodish, and S. R. Goldring, Expression cloning of an adenylate cyclase-coupled calcitonin receptor. Science 254, 1022 – 1024 (1991). 44. S. R. Goldring, A. H. Gorn, M. Yamin, S. M. Krane, and J. T. Wang, Characterization of the structural and functional properties of cloned calcitonin receptor cDNAs. Horm. Metab. Res. 25, 477 – 480 (1993). 45. M. Ikegame, M. Rakopoulos, H. Zhou, S. Houssami, T. J. Martin, J. M. Moseley, and D. M. Findlay, Calcitonin receptor isoforms in mouse and rat osteoclasts. J. Bone Miner. Res. 10, 59 – 65 (1995). 46. S. K. Lee, S. R. Goldring, and J. A. Lorenzo, Expression of the calcitonin receptor in bone marrow cell cultures and in bone: A specifi marker of the differentiated osteoclast that is regulated by calcitonin. Endocrinology 136, 4572 – 4581 (1995). 47. P. Orcel, H. Tajima, Y. Murayama, T. Fujita, S. M. Krane, E. Ogata, S. R. Goldring, and I. Nishimoto, Multiple domains interacting with Gs in the porcine calcitonin receptor. Mol. Endocrinol. 14, 170 – 182 (2000). 48. E. E. Moore, R. E. Kuestner, S. D. Stroop, F. J. Grant, S. L. Matthewes, C. L. Brady, P. M. Sexton, and D. M. Findlay, Functionally different isoforms of the human calcitonin receptor result from alternative splicing of the gene transcript. Mol. Endocrinol. 9, 959 – 968 (1995). 49. J. Friedman and L. G. Raisz, Thyrocalcitonin: Inhibitor of bone resorption in tissue culture. Science 150, 1465 – 1467 (1965). 50. J. A. Wener, S. J. Gorton, and L. G. Raisz, Escape from inhibition or resorption in cultures of fetal bone treated with calcitoninand parathyroid hromone. Endocrinology 90, 752 – 759 (1972). 51. A. H. J. Tashjian, D. R. Wright, J. L. Ivey, and A. Pont, Calcitonin binding sites in bone: Relationships to biological response and “escape.” Recent Prog. Horm. Res. 34, 285 – 334 (1978).

255

CHAPTER 8 Calcitonin 52. N. S. Krieger, R. S. Feldman, and A. H. Tashjian, Jr., Parathyroid hormone and calcitonin interactions in bone: Irradiation-induced inhibition of escape in vitro. Calcif. Tissue Int. 34, 197 – 203 (1982). 53. D. Inoue, C. Shih, D. L. Galson, S. R. Goldring, W. C. Horne, and R. Baron, Calcitonin-dependent down-regulation of the mouse C1a calcitonin receptor in cells of the osteoclast lineage involves a transcriptional mechanism. Endocrinology 140, 1060 – 1068 (1999). 54. S. Wada, N. Udagawa, N. Nagata, T. J. Martin, and D. M. Findlay, Calcitonin receptor down-regulation relates to calcitonin resistance in mature mouse osteoclasts. Endocrinology 137, 1042 – 1048 (1996). 55. S. Wada, T. J. Martin, and D. M. Findlay, Homologous regulation of the calcitonin receptor in mouse osteoclast-like cells and human breast cancer T47D cells. Endocrinology 136, 2611 – 2621 (1995). 56. T. J. Chambers and A. Moore, The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J. Clin. Endocrinol. Metab. 57, 819 – 824 (1983). 57. A. O. Hoff, M. Amling, and R. F. Gagel, CT/CGRP knockout mice. Personal communication (2000). 58. B. A. Roos, M. Yoon, S. V. Cutshaw, and D. N. Kalu, Calcium regulatory action of endogenous rat calcitonin demonstrated by passive immunization with calcitonin antibodies. Endocrinology 107, 1320 – 1326 (1980). 59. M. Chakraborty, D. Chatterjee, F. S. Gorelick, and R. Baron, Cell cycle-dependent and kinase-specific regulation of the apical Na/H exchanger and the Na,K-ATPase in the kidney cell line LLC-PK1 by calcitonin. Proc. Natl. Acad. Sci. USA 91, 2115 – 2119 (1994). 60. M. Chakraborty, D. Chatterjee, S. Kellokumpu, H. Rasmussen, and R. Baron, Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 251, 1078 – 1082 (1991). 61. Y. Chen, J. F. Shyu, A. Santhanagopal, D. Inoue, J. P. David, S. J. Dixon, W. C. Horne, and R. Baron, The calcitonin receptor stimulates Shc tyrosine phosphorylation and Erk1/2 activation: Involvement of

62.

63.

64.

65.

66.

67.

68.

Gi, protein kinase C, and calcium. J. Biol. Chem. 273, 19809 – 19816 (1998). A. H. Gorn, H. Y. Lin, M. Yamin, P. E. Auron, M. R. Flannery, D. R. Tapp, C. A. Manning, H. F. Lodish, S. M. Krane, and S. R. Goldring, Cloning, characterization, and expression of a human calcitonin receptor from an ovarian carcinoma cell line. J. Clin. Invest. 90, 1726 – 1735 (1992). L. M. McLatchie, N. J. Fraser, M. J. Main, A. Wise, J. Brown, N. Thompson, R. Solari, M. G. Lee, and S. M. Foord, RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature 393, 333 – 339 (1998). N. Buhlmann, K. Leuthauser, R. Muff, J. A. Fischer, and W. Born, A receptor activity modifying protein (RAMP)2-dependent adrenomedullin receptor is a calcitonin gene-related peptide receptor when coexpressed with human RAMP1. Endocrinology 140, 2883 – 2890 (1999). K. Leuthauser, R. Gujer, A. Aldecoa, R. A. McKinney, R. Muff, J. A. Fischer, and W. Born, Receptor-activity-modifying protein 1 forms heterodimers with two G-protein-coupled receptors to define ligand recognition. Biochem. J. 351, 347 – 351 (2000). R. Muff, N. Buhlmann, J. A. Fischer, and W. Born, An amylin receptor is revealed following co-transfection of a calcitonin receptor with receptor activity modifying proteins-1 or -3. Endocrinology 140, 2924 – 2927 (1999). S. Houssami, D. M. Findlay, C. L. Brady, T. J. Martin, R. M. Epand, E. E. Moore, E. Murayama, T. Tamura, R. C. Orlowski, and P. M. Sexton, Divergent structural requirements exist for calcitonin receptor binding specificity and adenylate cyclase activation. Mol. Pharmacol. 47, 798 – 809 (1995). F. Naro, M. Perez, S. Migliaccio, D. L. Galson, P. Orcel, A. Teti, and S. R. Goldring, Phospholipase D-and protein kinase C isoenzymedependent signal transduction pathways activated by the calcitonin receptor. Endocrinology 139, 3241 – 3248 (1998).

CHAPTER 9

Vitamin D Biology, Action, and Clinical Implications DAVID FELDMAN, PETER J. MALLOY, AND COLEMAN GROSS Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305

I. II. III. IV. V. VI. VII.

VIII. 1,25(OH)2D3 Analogues with Decreased Calcemic Activity IX. Actions of Vitamin D in Classical Target Organs to Regulate Mineral Homeostasis X. Actions of 1,25(OH)2D in Nonclassical Target Organs XI. Vitamin D and Osteoporosis References

Introduction Vitamin D Metabolism Pathways of Activation and Inactivation of Vitamin D Mechanism of 1,25(OH)2D Action Nongenomic Effects of Vitamin D Physiology: Regulation of Serum Calcium Genetic Disorders and Vitamin D Receptor Polymorphisms

I. INTRODUCTION

bone or mineral metabolism, including antiproliferative, prodifferentiating, and immunosuppressive activities. This chapter describes the basic biology of vitamin D, including its metabolism, physiology, mechanism of action, and its diverse functions in the body, including those actions that relate to mineral metabolism as well as the newer actions. Several reviews of vitamin D mechanism of action and function have been published [1 – 9] as well as a comprehensive book addressing all areas of vitamin D [10]. Specific issues relating to vitamin D and osteoporosis are discussed in Chapter 68.

Vitamin D is one of the major regulators of calcium homeostasis in the body and is critically important for normal mineralization of bone. The active hormone, 1,25-dihydroxyvitamin D [1,25(OH)2D], is produced by sequential hydroxylations of vitamin D in the liver (25hydroxylation) and the kidney (1-hydroxylation). 1,25(OH)2D, acting through the vitamin D receptor (VDR), acts by a genomic mechanism identical to the classical steroid hormones to regulate target gene transcription. The traditional actions of 1,25(OH)2D are to enhance calcium and phosphate absorption from the intestine in order to maintain normal concentrations in the circulation and to provide adequate amounts of these minerals to the boneforming site for the mineralization of bone to proceed normally. However, in the past decade, it has become increasing clear that vitamin D has many additional functions that implicate the hormone in a wide array of actions relating to bone formation as well as to other areas unrelated to

OSTEOPOROSIS, SECOND EDITION VOLUME 1

A. Chemistry, Structure, and Terminology Vitamin D exists in two forms: vitamin D3 (cholecalciferol) and vitamin D2 (ergocalciferol). When written without a subscript the designation vitamin D denotes either D2 or D3. Sunlight, in the form of UV B rays, cleaves the B ring between carbon 9 and 10 to open the ring and create a

257

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

258

FELDMAN, MALLOY, AND GROSS

B. History The unfolding of the story of vitamin D from its discovery as an antirachitic factor and designation as a vitamin to its transition from being considered a vitamin to its recognition as a hormone has all occurred within the past 75 years. However, the substance appears to be evolutionarily very ancient, produced by phytoplankton exposed to sunlight approximately 750 million years ago [12]. The history of the identification of vitamin D, the beneficial effects of sunlight on rickets, the elucidation of the pathway of conversion of vitamin D to 1,25(OH)2D, and the realization that vitamin D is a steroid hormone have been detailed in multiple reviews [12 – 15].

II. VITAMIN D METABOLISM A. Dietary Sources

FIGURE 1

1,25(OH)2D metabolic pathways. UV B indicates ultraviolet radiation (wavelength 290 – 320 nm) emitted from the sun. Liver 25 refers to hepatic 25-hydroxylase, and kidney 24R and 1 are renal 24-hydroxylase and 1-hydroxylase, respectively. Reproduced with permission from M. F. Holick, In “Endocrinology” (L. J. DeGroot et al., eds.). Saunders, Philadelphia, 1995.

secosteroid structure (Fig. 1). By this process the precursor (provitamin) molecules, 7-dehydrocholesterol in animals and ergosterol in plants, are converted to the secosteroids, vitamin D3 and vitamin D2, respectively [11]. The two secosteroids differ only in the presence of a methyl group at carbon 28 and a double bond between carbon 22 and 23 on the side chain of vitamin D2. Vitamin D2 and vitamin D3 are handled identically in the body and converted, via two hydroxylation steps, to the active hormones, 1,25(OH)2D2 or 1,25(OH)2D3 (calcitriol).

There are two sources of vitamin D: dietary intake and endogenous production (Fig. 1). Endogenous vitamin D production occurs in the skin as a result of UV light exposure, and this synthetic process, which distinguishes vitamin D from the true vitamins, will be discussed later (Section IIB). Whereas only vitamin D3 is produced in the skin, both vitamin D2 from plant sources and vitamin D3 from animal sources are available in the diet. Vitamins D2 and D3 are biologically inactive and, as discussed in detail later, must be converted to hydroxylated metabolities to exhibit hormonal activity. Foods naturally containing substantial amounts of vitamin D are relatively few: egg yolks, liver, and fish liver oils (cod liver oil). In the United States the primary dietary source of vitamin D is fortified milk, which nominally contains 400 IU/quart, although this has been found to vary considerably [16]. Other supplemented sources may include cereals, breads, and fortified margarine. The recommended daily allowance for adults is 200 IU/day, for pregnant and lactating women it is 400 IU/day, for infants less than 6 months it is 300 IU/day, and for children over 6 months it is 400 IU/day. However, on average, adult intake is estimated to be less than 100 IU/day, suggesting that dietary sources of vitamin D play a minor role in vitamin D homeostasis (see Section XI,D for consequences on bone). Findings indicate that the recommended daily intake of vitamin D may be insufficient [17], especially in the elderly [18], and when sunlight exposure is limited [19,20]. The problem is magnified in the homebound elderly [21,22]. Seasonal sunlight deficiency contributes to vitamin D insufficiency [23,24]. Many studies suggest that fortification is necessary to augment daily intake and maintain baseline stores of vitamin D [25]. Even in those taking supplements, especially the elderly or individuals who are ill and

259

CHAPTER 9 Vitamin D

hospitalized, hypovitaminosis D may be common [26] and may contribute to osteoporotic fracture [27]. Evidence that vitamin D supplementation reduces fractures has been accumulating [22,28]. It is unwise to assume that vitamin D status is normal, even if patients are taking 400 IU supplementation. Many authors have concluded that 800 IU/day would be an effective intake yet still safe. This subject is further discussed in Chapter 68. Vitamin D is fat soluble and dietary sources are absorbed via the lymphatics in the proximal small bowel. Important factors for absorption include (i) gastric, pancreatic, and biliary secretions, (ii) formation of micelles, (iii) diffusion through the unstirred layer adjacent to the intestinal mucosa, (iv) brush border membrane uptake, (v) incorporation into chylomicrons, and (vi) absorption into the lymphatics. The mechanism of intestinal calcium absorption and its regulation by vitamin D has been reviewed by Wasserman [29]. Disorders that interfere with the processes just described or that disrupt the small bowel mucosa can interfere with vitamin D absorption include cystic fibrosis, chronic pancreatitis with pancreatic insufficiency, biliary obstruction, sprue (gluten enteropathy), inflammatory bowel disease involving the small bowel, and gastrointestinal surgery [30]. Assessing the absorption of vitamin D may be clinically important in patients with these or related conditions. After an oral dose of vitamin D, blood levels begin to rise at 4 h peak by 12 h, and return close to baseline by 72 h. This pharmacokinetic profile provides a useful clinical test for assessing vitamin D absorption. The serum vitamin D concentration can be measured 12 h after an oral dose of 50,000 IU of vitamin D; a value of 50 ng/ml is indicative of normal vitamin D absorption, whereas malabsorption is indicated when values are 10 ng/ml [31]. The subject of disordered vitamin D absorption is discussed more fully in Chapter 48. Although most cases of rickets are due to vitamin D deficiency, a study of rachitic children in Nigeria suggests that calcium deficiency may also contribute to this condition. The children responded better to treatment with calcium alone or calcium and vitamin D than treatment with vitamin D alone [32].

B. Photobiology of Vitamin D: Endogenous Production The photobiology of vitamin D3 has been reviewed [33,34]. Ultraviolet (UV) radiation emitted from the sun and transmitted to the earth’s surface can be broadly divided into two spectra: UV A (wavelength 320 – 400 nm) and UV B (wavelength 290 – 320 nm). Light energy is transmitted to the epidermis and dermis where stores of 7-dehydrocholesterol (provitamin D3) are located (Fig. 1). UV B radiation causes scission of the C9 – C10 bond in the steroid, yielding

the “split” or secosteroid previtamin D3. Thermal equilibration within the skin occurs over a day, converting previtamin D3 to Vitamin D3. Vitamin D3 binds to the circulating vitamin D-binding protein (DBP) and thus leaves the skin and enters the circulation. During prolonged exposure to UV B radiation, previtamin D3 synthesis plateaus at about 15% of the 7-dehydrocholesterol skin content and leads to the increasing production of the biologically inert compounds lumisterol and a small amount of tachysterol from previtamin D3. This restriction on previtamin D3 formation may serve as a mechanism to prevent the overproduction of vitamin D3. Several factors have been found to affect the cutaneous synthesis of vitamin D3, including latitude and seasonal variation, skin pigmentation, the use of topical sunscreens, and age. In addition, 1,25(OH)2D may feed back on the skin to add to the regulation since it acts on epidermal constituents [35]. In addition, UV B radiation inhibits levels of VDR, suggesting the existence of a feedback mechanism in that UV B initiates vitamin D synthesis in keratinocytes and, at the same time, limits VDR abundance [36]. 1. LATITUDE AND SEASON Because the conversion of 7-dehydrocholesterol to previtamin D3 in the skin requires UV B radiation, the amount of previtamin D3 synthesized is related to the amount of UV B radiation absorbed by the skin. The amount of solar radiation reaching the surface of the earth is limited by the changing zenith angle of the sun and decreases with increasing global latitude. Similarly, the incident radiation on the surface of the earth is diminished during the fall and winter months when the sun is lower in the sky. Therefore, the variation in cutaneous UV B radiation exposure due to seasonal variation or geographical location can influence the amount of vitamin D3 synthesized in the skin. As a result, no previtamin D3 is synthesized in Boston (42° N latitude) from November to February, and 10° farther north, in Edmonton, this period is extended from October to March. In more southerly locations, such as Los Angeles and Puerto Rico, previtamin D3 synthesis occurs year round [12]. An interesting commentary on the relative importance of sunlight was described by Holick [12] in a study of naval personnel onboard submarines. Submariners who were not exposed to sunlight for 3 months failed to maintain adequate vitamin D levels even while ingesting 600 IU/day of vitamin D, supporting the concept that 800 IU/day may be necessary to maintain normal vitamin D levels in the absence of adequate sunlight. 2. SKIN PIGMENTATION The degree of skin pigmentation (i.e., melanin content) also affects cutaneous vitamin D3 production. Melanin protects the body from excess sunlight by acting as a sink to absorb UV B rays and acts as a competitor of 7dehydrocholesterol for UV B radiation. Therefore, the more

260

FELDMAN, MALLOY, AND GROSS

melanin that is present in the skin, the less UV B radiation is available for previtamin D3 synthesis. However, individuals with high melanin levels compensate by increasing the conversion of 25OHD to 1,25(OH)2D [37]. Loomis [38] raised the hypothesis that melanin pigmentation evolved in people living near the equator to prevent the excessive production of vitamin D due to constant exposure to sunlight. As people migrated away from the equatorial regions, their sunlight exposure was shortened and, in order to allow adequate production of vitamin D and prevent rickets, the melanin levels in their skin diminished. When individuals of different skin pigmentation were exposed to the same suberythemic dose of UV radiation (27 mJ/cm2), whites showed the largest incremental rise in serum vitamin D concentrations, whereas Asians showed an intermediate increase and black individuals the smallest rise [39]. Basal concentrations of 25(OH)D are lower in young healthy blacks than in young healthy whites; however, their 1,25(OH)2D concentrations are higher than in whites, possibly due to relative secondary hyperparathyroidism [40]. Increased skin pigmentation does not limit the absolute amount of previtamin D3 made, but rather it extends the period of sunlight exposure necessary to reach the maximum production of previtamin D3 [41]. This time interval for maximum previtamin D3 production ranges from 0.5 h in lightly pigmented individuals to 3 h or more in darker pigmented subjects. 3. SUNSCREENS, SUN EXPOSURE, AND AGE Interestingly, similar to melanin, topical sunscreens act as a competitor of the photochemical production of vitamin D3 by absorbing UV radiation. Para-amino benzoic acidbased preparations with an SPF 8 rating can significantly block the cutaneous production of vitamin D3. Age is also a variable that can influence the production of vitamin D3, as the amount of 7-dehydrocholesterol in the skin and the efficiency of previtamin D3 photoproduction decreases as a consequence of advancing age [33].

C. Transport in Circulation: Vitamin D-Binding Protein (DBP) Group-specific component (Gc), a 58-kDa plasma  globulin, was originally described immunologically in 1959 [42], and approximately 16 years later Gc was identified as a vitamin D-binding protein [43]. There appears to be a single binding site for calciferols on each DBP molecule; however, only about 5% of the binding sites are normally occupied, probably due to the high concentration of DBP in the circulation [44,45]. The binding affinity of DBP for the vitamin D metabolites is as follows: 25(OH)D3  24,25(OH)2D3  vitamin D3  1,25(OH)2D3

 1,24,25(OH)3D3. The affinity of DBP for vitamin D2 and vitamin D3 is similar. DBP is primarily synthesized in the liver [45], and serum concentrations of DBP increase in pregnancy and with estrogen treatment, whereas it is decreased in liver disease, malnutrition, and nephrotic syndrome. Calcitropic hormones, however, do not appear to regulate the synthesis of DBP. In addition to liver, DBP mRNA has been detected in several rat tissues, including testis, kidney, yolk sac, placenta, and adipose tissue [45]. Using the sensitive method of reverse transcriptasepolymerase chain reaction (RT-PCR), DBP transcripts have been found additionally in lung, heart, gut, spleen, uterus, and brain [45], as well as in activated monocytes [46]. Its role when produced locally is still unclear. DBP is very polymorphic, with over 120 variants being described [47], making it useful as a population marker [48]. The gene encoding the human DBP has been cloned, its genomic structure established [44,45], and homology with both -fetoprotein and albumin recognized [45]. Interestingly, albumin binds 10% of the vitamin D sterols in the circulation. The genes encoding DBP, -fetoprotein, and albumin are all located on chromosome 4 and are considered part of the same gene family based on a conserved triple domain structure [45]. Vitamin D3 synthesized in the skin travels in plasma almost entirely bound to DBP, whereas vitamin D2 obtained in the diet is associated with both lipoproteins (chylomicrons) and DBP [49]. Like other steroid hormones in the circulation, the free or unbound 1,25(OH)2D is in equilibrium with the bound form. It is the free fraction of the 1,25(OH)2D that is hormonally active and binding to DBP inhibits accessibility of the steroid to the cell and prolongs 1,25(OH)2D halflife [50]. In serum, approximately 0.04% of 25(OH)D and 0.4% of 1,25(OH)2D are found in the free form. DBP functions as a reservoir of 25(OH)D and serves as a buffer to prevent the too rapid tissue delivery of the steroids to target cells. DBP thereby prevents vitamin D deficiency and presents 25OHD for renal activation to 1,25(OH)2D [51]. Several findings suggest that DBP may have other critical roles in the body in addition to being the vitamin D transport protein. As alluded to earlier, it circulates at micromolar concentrations, 100-fold in excess of its main ligand 25(OH)D, and is only 5% occupied with calciferols [44,45]. DBP binds monomeric G-actin molecules and is part of the extracellular actin scavenger system that protects the organism from the effects of filamentous F-actin formation when actin is released following cell lysis [52]. Additionally, DBP has been shown to be membrane associated on a number of cell types, either acquired from serum or synthesized by the cell [53]. The function of membraneassociated DBP is unclear at present but it may aid in sterol transport into the cell or it may play a role in modulating the function of 1,25(OH)2D by limiting its interaction with the cell and the VDR [45].

261

CHAPTER 9 Vitamin D

Although there are no reports of patients with DBP deficiency, a DBP knockout mouse has been described [54]. DBP null (-/-) mice are phenotypically normal and fertile. However, they have lower circulating concentrations of 25(OH)D and 1,25(OH)2D when fed a normal diet and exhibit secondary hyperparathyroidism and bone changes when fed a vitamin D-deficient diet. These findings were not seen in control normal mice and support the concept that DBP acts as a storehouse for vitamin D metabolites, thus protecting the animal in times of vitamin D deficiency. DBP markedly prolonged the serum half-life of 25(OH)D and less dramatically prolonged the half-life of vitamin D by slowing its hepatic uptake and increasing the efficiency of its conversion to 25(OH)D in the liver. After an overload of vitamin D, DBP -/- mice were unexpectedly less susceptible to hypercalcemia and its toxic effects. DBP knockout mice also show an increase in hepatic uptake and clearance of vitamin D. Thus, the role of DBP is to maintain stable serum stores of vitamin D metabolites and modulate the rates of its bioavailability, activation, and end-organ responsiveness. These properties may have evolved to stabilize and maintain serum levels of vitamin D in environments with variable vitamin D availability [54].

D. Megalin Megalin is a large multifunctional endocytic clearance receptor for circulating proteins that has been implicated in vitamin D uptake and delivery to the kidney for activation to 1,25(OH)2D [55]. Knockout of the megalin gene in mice usually is lethal, but the few survivors were characterized as having severe rickets [55]. The findings suggested that DBP may be a ligand for megalin and that megalin is critical for 25OHD uptake by the kidney. However, because megalin is located on the luminal membrane of the proximal tubule, this theory suggests that DBP carrying vitamin D is filtered by the glomerulus and reabsorbed by the tubular cells. Thus, knockout mice with the null (-/-) megalin genotype develop proteinuria [56] and lose their vitamin D – DBP complex into the urine, leading to vitamin D deficiency and rickets [55]. In the presence of megalin, the vitamin D – DBP complex would bind to megalin and be internalized by endocytic vesicle formation. Megalin may also affect vitamin D function by binding PTH and PTHrP and directing these proteins toward endocytic lysosomal catabolism. Megalin may also be involved in vitamin D uptake into other cells, facilitating its metabolism and action. Leheste et al. [56] have speculated that the Fanconi syndrome, characterized by proteinuria and osteomalacia, might be related to defects in megalin as represented by the megalin null (-/-) mouse. The role of megalin in the entry of vitamin D metabolites into metabolic or target tissues remains to be clarified.

E. Intracellular Binding Proteins Adams and colleagues [57,58] have described intracellular vitamin D-binding proteins (IDBPs), which they speculate play a role in the intracellular movement of vitamin D metabolites. The IDBPs are related to heat shock 70 (HSP 70) proteins and, as chaperones, contain intracellular organelle-targeting sequences to direct bound molecules to various intracellular destinations. Differences in IDBPs may explain the relative resistance of New World primates to vitamin D action. Gacad and Adams [58] demonstrated that this form of resistance is associated with the overexpression of an IDBP. Their data suggest that this IDBP is relatively specific for 25OHD3 and that additional HSP 70like binding proteins are present in New World primates that are specific for 1-hydroxylated-vitamin D metabolites as well as other steroid hormones. These authors hypothesize that the movement of vitamin D metabolites within cells is controlled by IDBPs, directing substrate toward metabolic pathways for activation or inactivation by enzymes or to the VDR to activate target genes [57,58].

F. Assays of Vitamin D Metabolites Assays of 25(OH)D and 1,25(OH)2D provide valuable tools to assess the vitamin D status of patients [59]. The best indicator of the overall vitamin D status of an individual, 25(OH)D, was originally measured by competitive binding assays, but is now measured by radioimmunoassay. The sensitivity of the assay was improved by the use of 125I-labeled 25(OH)D. Although measurement of 1,25(OH)2D is more difficult because it circulates at approximately 1000-fold lower concentrations than 25(OH)D, i.e., picograms per milliliter instead of nanograms per milliliter, an 125I-based radioimmunoassay is now available for determining 1,25(OH)2D concentrations. In the clinical setting, measurement of 25OHD is the most useful and measurement of 1,25(OH)2D is confirmatory. However, in cases of genetic disease, such as 1-hydroxylase deficiency (see Section VII,A) or hereditary vitamin D-resistant rickets (HVDRR) (see Section VII,B), or in cases of hypercalcemia, measurement of 1,25(OH)2D is critical to understand the pathophysiology.

III. PATHWAYS OF ACTIVATION AND INACTIVATION OF VITAMIN D A. 25-Hydroxylation The first step in activation of vitamin D to the biologically active hormone, 1,25(OH)2D, is hydroxylation at the carbon 25 position in the liver [60]. Although liver

262 parenchymal cells are the primary site for 25-hydroxylation, enzyme activity may be expressed in other tissues in other species (e.g.) in birds, kidney and intestine have 25hydroxylase activity. The 25-hydroxylase enzyme is a cytochrome P450, present in both microsomes and mitochondria in rats, but in humans, significant amounts of the 25-hydroxylase are only found in mitochondria [60, 61]. The gene encoding the human 25-hydroxylase (CYP27) has been cloned [62 – 64] and has been localized to chromosome 2q33-qter [63]. The gene product of CYP27 encodes a protein with sterol 27-hydroxylase as well as 25hydroxylase activities, the former step being important in the biosynthetic pathway of bile acids [61]. The rare genetic disease cerebrotendinous xanthomatosis is due to a deficiency of 27-hydroxylase activity [63]. This defect results in the accumulation of bile acid precursors and cholestanol, which deposit in the brain and peripheral nerves, and forms tuberous xanthomata [65]. These patients also have been reported to have low bone mineral density associated with low 25(OH)D levels and increased fracture risk [66]. A deficiency in the enzymatic activity is not clinically apparent unless severe hepatic failure develops [30]. The 25-hydroxylase is not a tightly controlled enzyme in contrast to the 1-hydroxylase (see Section IIIB2). In patients with hypervitaminosis D, 25(OH)D concentrations increase markedly (as much as 15-fold), whereas those of 1,25(OH)2D are relatively normal [67]. Production of 25(OH)D depends primarily on the concentration of vitamin D; however, higher basal vitamin D and 25(OH)D levels may diminish the production of 25(OH)D in vivo. However, 1,25(OH)2D has been shown to limit the production of 25(OH)D. Treatment with 1,25(OH)2D prevented the increase seen in circulating 25(OH)D after oral vitamin D given to volunteers [68]. This effect may be explained by the increased metabolism of 25(OH)D to 24R,25-dihydroxyvitamin D [24,25(OH)2D] due to the induction of 24hydroxylase by 1,25(OH)2D (see Section IIIC2) and therefore increased the metabolic clearance rate of 25(OH)D [69,70]. Calcium may also have a direct modulatory role on 25-hydroxylase activity [60]. However, in vivo, the role of calcium to modulate 25-hydroxylase activity is likely mediated via changes in PTH, which influences the production of 1,25(OH)2D, which in turn increases the metabolism of 25(OH)D through 24-hydroxylation. Although 25-hydroxylase activity may be decreased in cerebrotendinous xanthomatosis, abnormalities in the enzyme are otherwise very rare [60]. Severe liver disease may lead to metabolic bone disease, but etiologies are complex and include malabsorption, dietary changes, and effects of medication and alcohol, as well as defective 25hydroxylation. This subject is discussed more fully in Chapter 48.

FELDMAN, MALLOY, AND GROSS

B. 25-Hydroxyvitamin D-1-Hydroxylase 1. THE 25-HYDROXYVITAMIN D-1-Hydroxylase Enzyme Following hydroxylation in the liver, 25(OH)D is transported in the circulation bound to DBP and the kidney accomplishes the final step of vitamin D activation, namely 1-hydroxylation. This step is apparently megalin dependent. The 25-hydroxyvitamin D-1-hydroxylase (1-hydroxylase) is a mitochondrial P450 enzyme present in low abundance and localized to the proximal tubule of the nephron [71]. As a mixed function oxidase the enzyme requires NADPH, molecular oxygen, ferredoxin, and ferredoxin reductase for activity. In 1997, cDNAs for the 1-hydroxylase from the mouse, rat, and human were cloned [72 – 77]. The predicted amino acid sequence confirms that the 1-hydroxylase gene (CYP1 or CYP27B1) is a member of the cytochrome P450 enzyme superfamily. The 1-hydroxylase exhibits significant homologies to the vitamin D-25-hydroxylase (CYP27) and the 25-hydroxyvitamin D-24-hydroxylase (CYP24) enzymes. The human 1-hydroxylase gene is approximately 5 kb in length and is composed of nine exons. Fluorescent in situ hybridization analysis localized the gene to chromosome 12q13.3, confirming earlier reports that the gene defect causing 1-hydroxylase deficiency was linked to chromosome 12q14, close to the gene coding for the vitamin D receptor [78,79]. The gene is expressed in kidney epithelial cells in both proximal and distal tubules as well as selected other sites. 2. REGULATION OF 1-Hydroxylase In contradistinction to 25-hydroxylase, renal 1-hydroxylase is a tightly regulated enzyme and the critical determinant of 1,25(OH)2D synthesis (Fig. 2). The overall regulation of 1-hydroxylase is determined by the calcium and phosphorus requirements of the organism and is mediated by several bioactive substances. The principal regulator of 1-hydroxylase is parathyroid hormone (PTH) [40,80]; however, other important regulators include phosphate, 1,25(OH)2D itself, calcium, and calcitonin. The production of 1,25(OH)2D may also be modulated by other hormones, such as estrogen, prolactin, and growth hormone, but these effects in mammalian systems appear to be small. Analysis of the human 1-hydroxylase promoter has identified positive response elements for PTH and calcitonin and a negative response element for 1,25(OH)2D [81,82]. In normal calcemic states, the expression of 1hydroxylase is determined by the concentrations of calcitonin and 1,25(OH)2D [83]. In hypocalcemic states, the expression of 1-hydroxylase is determined by the levels of

CHAPTER 9 Vitamin D

263 inhibitor of PKA abrogated the PTH-mediated upregulation of 1-hydroxylase gene expression [81]. It is now well established that PTH secretion is regulated by the concentrations of both Ca2 and 1,25(OH)2D in the blood [low Ca2 increases PTH; high 1,25(OH)2D decreases PTH]. In the parathyroid gland, the circulating Ca2 concentrations are monitored by the Ca2-sensing receptor (CaR) described by Chattopadhyay and colleagues [88], whereas the VDR downregulates PTH synthesis and secretion [86].

FIGURE 2

Regulation of 1-hydroxylase and 24-hydroxylase activities

in kidney.

PTH and 1,25(OH)2D [81,82]. Positive and negative regulation of the 1-hydroxylase gene by PTH, calcitonin, or 1,25(OH)2D3 has been demonstrated at the transcriptional level in kidneys of animals and in a mouse proximal tubule cell line. Although data strongly support the role of PTH and 1,25(OH)2D in regulating renal 1-hydroxylase, the regulation of 1-hydroxylase expression by these hormones in nonrenal tissues remains to be determined. However, it is clear that the 1-hydroxylase enzyme expressed in renal and nonrenal tissues is encoded by the same gene, as mutations causing 1-hydroxylase deficiency have been found in both renal [77] and nonrenal tissues, including keratinocytes [76] and blood cells [84]. a. PTH Evidence that PTH is the primary regulator of 1-hydroxylase is substantial [85 – 87]. 1,25(OH)2D levels are increased in hyperparathyroidism and reduced in hypoparathyroidism. After parathyroidectomy, 1,25(OH)2D levels fall and are increased after administration of PTH to normal subjects and to patients with hypoparathyroidism. Moreover, substantial in vitro data indicate that PTH markedly stimulates 1-hydroxylase activity in mammalian renal slices, isolated renal tubules, and cultured renal cells. PTH has been shown to stimulate 1-hydroxylase gene promoter activity, most likely by increasing cAMP. cAMP stimulates protein kinase A (PKA) activity, which phosphorylates CRE-binding protein (CREB) and modulates transcription through cAMP-responsive elements (CRE) present in the 5’ sequence flanking the 1-hydroxylase gene. Alternatively, cAMP may modulate promoter activity by transactivation through AP-1 and AP-2 sites. In addition, an

b. Phosphate Phosphate is the second most important physiological regulator of the 1-hydroxylase with high phosphate levels suppressing and low levels stimulating enzyme activity [86,87,89]. Indeed, thyroparathyroidectomized rats maintained on high calcium, low phosphorus diets metabolize 25(OH)D to 1,25(OH)2D despite the absence of PTH, suggesting that PTH may, in part, indirectly influence 1-hydroxylase activity by way of its regulation of phosphate. In humans, phosphorus restriction increases 1,25(OH)2D concentrations to 180% of control, and phosphorus supplementation decreases circulating 1,25(OH)2D by 29% [90]. These changes reflect alterations in the synthetic rate rather than changes in the half-life of the enzyme, demonstrating the important role played by phosphate on 1-hydroxylase. The effect of elevated phosphate to inhibit 1-hydroxylation contributes to the development of renal osteodystrophy during chronic renal failure and is part of the rationale for using phosphate binder therapy to delay the onset of bone disease in these patients [86,87,89]. c. 1,25(OH)2D Interestingly, 1,25(OH)2D regulates its own production. This activity is mediated directly at the level of the 1-hydroxylase in the kidney and indirectly by inhibition of PTH (as described earlier). Low 1,25(OH)2D concentrations promote 1-hydroxylase activity and 1,25(OH)2D synthesis, whereas high 1,25(OH)2D levels inhibit enzyme activity [71,86]. The ability of 1,25(OH)2D to inhibit 1-hydroxylase activity has been demonstrated in vitro as well as in vivo [71]. This effect involves both PTHdependent and PTH-independent mechanisms; 1,25(OH)2D directly (PTH-independent) decreases 1-hydroxylase activity as well as decreasing PTH secretion (PTH-dependent). In vivo, however, it is difficult to separate the contribution of changes in calcium or PTH from direct 1,25(OH)2D actions because of the tight linkage of these systems. In VDR null (-/-) mice, 1-hydroxylase gene expression is increased (a phenomenon used to help in the cloning of this elusive gene [72]) and the upregulation of 1-hydroxylase by PTH was evident. However, a downregulation of 1-hydroxylase gene expression by 1,25(OH)2D3 was not observed, implying that the VDR is essential for the negative regulation of this gene by

264 1,25(OH)2D3, probably via an effect on PTH transcription [72,81]. Another complexity in vivo is the finding that the administration of 1,25(OH)2D chronically can regulate its serum concentration by increasing its metabolic clearance rate by induction of the 24-hydroxylase enzyme (see Section III,C) [91]. d. Calcium Although regulation of 1-hydroxylase in response to changes in serum calcium levels is mainly due to changes in PTH, calcium may act independently as well. Calcium restriction increases 1,25(OH)2D synthesis in thyroparathyroidectomized rats [92]. Additionally, an increase in calcium concentration in the media of cultured chick renal tubule cells and rat renal tissue slices leads to a decreased production of 1,25(OH)2D. The effect of calcium in the regulation of 1-hydroxylase may explain why some patients with severe hyperparathyroidism and very high serum calcium levels exhibit low 1,25(OH)2D values [93]. Although the underlying mechanism for this finding is obscure, one might speculate that the calcium-sensing receptor (CaR) originally described in parathyroid glands [94] and also found in the kidney [95] may mediate this effect [88]. However, studies in VDR null (-/-) mice indicate that calcium is likely an indirect modulator of 1-hydroxylase, as in the absence of 1,25(OH)2D action, changes in calcium did not alter the levels of 1-hydroxylase activity [81]. e. Calcitonin Calcitonin can also stimulate 1,25(OH)2D synthesis in thyroparathyroidectomized rats [96]. Similarly, 1,25(OH)2D levels increase after calcitonin administration to patients with X-linked hypophosphatemic rickets [97] as well in the HYP mouse [98] where the 1-hydroxylase response to PTH is abnormal. In normocalcemic rats where PTH concentrations are relatively low, calcitonin has been shown to be a major regulator of the renal enzyme [83]. Analysis of the human 1-hydroxylase gene promoter has demonstrated a positive regulatory region for calcitonin [81]. f. Chronic Renal Failure In a model of chronic renal failure, in 5/6ths nephectomized rats the renal 1-hydroxylase gene expression decreased and the positive effects of PTH and calcitonin were diminished [81]. This study, and others like it, also showed that PTH and calcitonin positively regulate renal 1-hydroxylase gene expression via protein kinase A (PKA)-dependent and-independent pathways, respectively, and that 1,25(OH)2D3 is a negative regulator. Furthermore, in a moderate state of chronic renal failure, renal cells expressing the 1-hydroxylase gene appear to have a diminished potential to respond to the positive regulators, PTH and calcitonin [86,87,89,99]. g. Local Production of 1,25(OH)2D The kidney is the major source of circulating detectable 1,25(OH)2D. How-

FELDMAN, MALLOY, AND GROSS

ever, humans and animals devoid of functioning renal tissue exhibit low but detectable 1,25(OH)2D concentrations in the circulation [99]. Several extrarenal tissues, including skin [76], bone [100], macrophages [84,101], colon [102], placenta [103], and prostate [104], have been shown to exhibit 1-hydroxylase activity. Macrophages can convert 25OHD to 1,25(OH)2D and, in cases of an increased macrophage pool in the body, 1,25(OH)2D production by these cells can lead to hypercalcemia [99, 105]. The usual regulators of renal 1-hydroxylase, PTH, calcitonin, and 1,25(OH)2D, apparently do not play a primary role in controlling extrarenal 1-hydroxylase activity. For example, when overexpression of macrophage 1-hydroxylase activity and hypercalcemia occurs as in sarcoid, PTH is suppressed [106,107]. Potential regulators of 1-hydroxylase in macrophages include cytokines and the nitric oxide system [99,108].

C. 25-Hydroxyvitamin D-24-Hydroxylation in Kidney and Other Sites 1. THE 25-HYDROXYVITAMIN D-24-HYDROXYLASE ENZYME 25-Hydroxyvitamin D-24-hydroxylase (24-hydroxylase) is a mitochondrial P450 enzyme, which, in general, is expressed in cells that are responsive to 1,25(OH)2D [109]. 24-Hydroxylase can convert 25(OH)D to 24,25(OH)2D, which may have some biological activity (see Section III,C,3). However, the formation of 24,25(OH)2D is generally considered to represent the first step in the degradative and excretory pathway of vitamin D (Figs. 1 and 2). The enzyme can also hydroxylate 1,25(OH)2D to 1,24,25(OH)3D, initiating the inactivation pathway of the active hormone. Thus this enzyme acts to protect the body from the overproduction of 1,25(OH)2D [110]. The intestine is a major site of hormonal inactivation by virtue of its abundant 24-hydroxylase activity [111]. In the nephron, the enzyme is distributed in the proximal and distal tubules, the glomerulus, and the mesangium [112]. Cloning of the gene encoding the rat kidney 24-hydroxylase (CYP24) [113] led to the cloning of the human gene [64,114] and characterization of its promoter region [115]. Earlier there had been some question about whether the 24-hydroxylase and 1-hydroxylase activities resided in the same protein. Data now firmly indicate that these are in fact distinct enzymes encoded by separate genes. The human 24-hydroxylase is present on chromosome 20q13 [114]. 2. REGULATION OF 24-HYDROXYLASE ACTIVITY The regulation of the 24-hydroxylase activity (see Fig. 2) has been reviewed [109]. 1,25(OH)2D is the primary regulator of 24-hydroxylase, causing a marked induction of enzymatic activity and mRNA levels via a VDR-mediated genomic pathway (see Section IV,H). Two vitamin D response

265

CHAPTER 9 Vitamin D

elements (VDREs, see Section IV,F) have been identified in the promoter of the 24-hydroxylase gene [116,117]. Because 24-hydroxylase can be induced by 1,25(OH)2D in many VDR-containing cells, regulation of this gene product has proven to be an excellent marker of 1,25(OH)2D activity. Measurement of 24-hydroxylase enzyme activity and induction of mRNA by 1,25(OH)2D have been employed extensively in studies of cultured dermal fibroblasts from HVDRR patients [7,118] (see Section VII,B) and in studies of kidney, bone, and intestinal cells, as well as in studies of new target organs and malignant cells. In the kidney, PTH stimulates 1-hydroxylase and inhibits 24-hydroxylase [119], effects that are opposite to those of 1,25(OH)2D. However, because the intestine does not respond to PTH, and 24-hydroxylase is not downregulated by PTH in cultured bone cells [120], the action of PTH on renal 24-hydroxylase appears to be most important. Calcitonin has been shown to downregulate 24-hydroxylase mRNA and enzyme activity in rat intestine in vivo [121], suggesting the presence of an intestinal calcitonin receptor and a heretofore unanticipated function for this hormone. 3. CONTROVERSY OVER WHETHER 24,25(OH)2D EXHIBITS DISTINCT BIOLOGIC ACTIVITY Controversy over the biological activity of 24,25(OH)2D has existed for a number of years and the issue still causes lively debate [122]. It is clear that 24,25(OH)2D can bind to the VDR and, at a high concentration, induce 1,25(OH)2Dlike activity [123]. Findings have shown that 24,25(OH)2D can transactivate gene expression via an osteocalcin-VDRE mediated by a VDR mechanism identical to the pathway of 1,25(OH)2D action [124]. Whether 24,25(OH)2D has a unique receptor [125,126] and unique actions independent of 1,25(OH)2D, particularly during development or in selected tissues, has been speculated on for a number of years [123,127]. Evidence supporting a role for 24,25(OH)2D has been presented [126,127] and refuted [123]. In HYP mice, administration of 24,25(OH)2D3 did not merely mimic 1,25(OH)2D3 but caused unique actions, including increasing bone size, dry weight, and bone mineral content, without causing hypercalcemia or activating bone resorption as did 1,25(OH)2D [128]. In growth plate chondrocytes where 24-hydroxylase is expressed, high dose treatment with 24,25(OH)2D has been shown to be involved in the process of regulating bone growth, development, and repair [126]. A 24-hydroxylase knockout mouse model has been generated to address the physiological role of 24,25(OH)2D [110]. 24-hydroxylase null (-/-) mice showed a reduced clearance of 1,25(OH)2D3 from the circulation and have a high degree of perinatal lethality most likely due to hypercalcemia. Histological examination of bone from these mice showed an accumulation of unmineralized osteoid matrix in calvaria, mandible, clavicle, and cortical surfaces of long bones, sites of intramembraneous ossification.

Treatment with 24,25(OH)2D rescued the bone phenotype normalizing the development of the calvaria and the accumulation of osteoid at the periosteal surface of long bones, all suggesting a role for 24,25(OH)2D [110]. However, because 24-hydroxylase initiates 1,25(OH)2D3 inactivation, these mice have high circulating 1,25(OH)2D3. To rule out the contribution of elevated 1,25(OH)2D3 concentrations acting via the VDR in this study, a subsequent study examined a double knockout mouse generated by crossing 24hydroxylase (-/-) mice with VDR (-/-) mice. The animals were fed a high calcium diet to maintain normal calcium concentrations in the serum [129]. Whereas 24-hydroxylase (-/-), VDR (/) mice showed reduced amounts of mineralized tissue in the mandible and cranial bones, 24-hydroxylase (-/-), VDR (-/-) double knockout mice showed normal bone formation at all sites. Data indicate that the impaired mineralization phenotype seen in 24-hydroxylase (-/-) mice was due to the increase in 1,25(OH)2D3 action on the bone because of loss of the 24-hydroxylase inactivation pathway. The authors conclude that 24,25(OH)2D3 is not an essential hormone for bone formation [129]. The slower rate of 1,25(OH)2D3 metabolism in VDR-ablated mice was also seen in other studies [130]. 4. OTHER METABOLITES Less research has centered around the role of 1,24,25(OH)3D as an active hormone; however, in early studies, 1,24,25(OH)3D was shown to stimulate intestinal calcium absorption, mobilize calcium from bone, and exhibit antirachitic activity in rats [131,132]. The steroid binds to the VDR, but with lower affinity than 1,25(OH)2D [123] and, as a result, 1,24,25(OH)3D has diminished potency when compared to 1,25(OH)2D. The initial step in the catabolic pathway of 1,25(OH)2D is 24-hydroxylation, which leads to the inactivation of the active hormone and the production of more polar metabolites, eventually leading to calcitroic acid [11,132,133]. The affinity of the 24-hydroxylase enzyme is 5 to 10 times greater for 1,25(OH)2D than 25(OH)D, making 1,25(OH)2D the preferred substrate. Subsequent hydroxylations and oxidations lead to the production of the C-23 carboxylic acid derivative, calcitroic acid, the major final excretory product of vitamin D metabolism (Fig. 1). The details of these complex catabolic steps have been reviewed [11,133]. Vitamin D can be metabolized to more than 35 metabolites, some of which have been shown to exhibit biological activity [11]. Using the Caco-2 human colon adenocarcinoma cell line, Bischof et al. [134] showed that in proliferating undifferentiated cells, 1,25(OH)2D3 is converted into several side chain metabolites through the C-24 oxidation pathway. In contrast, in differentiated cells where the C-24 oxidation pathway is inactive, a more polar compound produced in significant amounts was found and identified as

266

FELDMAN, MALLOY, AND GROSS

1,25-dihydroxy-3-cholecalcifierol [1,25(OH)2-3-epi-D3]. This study showed that the state of cell differentiation influenced the metabolism of 1,25(OH)2D3 and leads us to speculate about possible autocrine or intracrine effects of vitamin D metabolites to maintain differentiation and limit cell proliferation. The activity of 1,25(OH)2-3-epi-D3 is only slightly, but not significantly, less active than the native 1,25(OH)2D3 in suppressing bovine PTH gene transcription and secretion, despite having a 30-fold lower affinity for the VDR [135 – 137]. Both 1,25(OH)2D3 and 1,25(OH)2-3-epi-D3 suppress PTH secretion by 50%. In metabolism studies using bovine parathyroid cells, the concentration of 1,25(OH)2-3-epi-D3 was even higher than that of the parent substrate 1,25(OH)2D3, suggesting that this diastereomer has a slower rate of metabolism. Thus, production and accumulation of 1,25(OH)2-3-epi-D3 as a major stable metabolite of 1,25(OH)2D3 in parathyroid and other tissues may contribute to the prolonged effects of 1,25(OH)2D3 on gene transcription.

IV. MECHANISM OF 1,25(OH)2D ACTION 1,25(OH)2D regulates calcium metabolism and promotes other physiologic actions by a VDR-mediated mechanism analogous to the classical steroid hormones. The VDR, a member of the steroid – thyroid – retinoid receptor gene superfamily, acts as a regulator of target gene transcription. Several reviews of the 1,25(OH)2D-VDR system have been published [1,2,4,5,7,9,138,139], and the subject is covered extensively elsewhere [10].

A. The Vitamin D Receptor The identification of 1,25(OH)2D as the active metabolite of vitamin D led to a search for the specific protein that functioned as a receptor for this hormone. Early studies demonstrated the existence of a target tissue protein that interacted with vitamin D metabolites [140]. The protein was subsequently very well characterized in a number of laboratories and was shown to have an approximate molecular mass of 50 kDa, to exhibit saturable and specific high-affinity binding of 1,25(OH)2D, to be located in the nucleus (see later), and to bind to DNA. The binding affinity for the vitamin D metabolites roughly correlated with their potency [141], although many exceptions have been noted. Analysis of the structure of the VDR by limited proteolysis showed that the DNA-binding region was distinct from the 1,25(OH)2D-binding portion of the protein, thus dividing the protein into two functional domains [142]. 1,25(OH)2Dbinding and DNA-binding activities were shown to be sen-

sitive to sulfhydryl and metal-chelating reagents, indicating that the protein contained essential cysteine residues and required heavy metals for these activities [143]. The VDR was eventually purified to near homogeneity from chick intestinal tissue [144,145], and monoclonal antibodies to the chick VDR [146] and later to the porcine VDR [147] were developed. From biochemical and immunocytochemical studies, the VDR has been found in the nuclei of many normal human tissues, including intestine, kidney, bone, parathyroids, thyroid, skin, adrenal, liver, breast, pancreas, muscle, prostate, and lymphocytes, demonstrating that the protein is widely distributed in normal human tissues [5,148 – 152]. It has also been identified in malignant cells from most of these sites. VDR is a phosphoprotein that undergoes a hormone-dependent phosphorylation in intact cells [153]. In embryonic chick duodenal organ culture, phosphorylation of the receptor is strongly induced within 1 h by 1,25(OH)2D and this occurs prior to the initiation of calcium uptake or the induction of calcium-binding protein [154]. The VDR has been shown to be phosphorylated on serine residues in vitro by protein kinase C [155], casein kinase II [156,157], and cAMP-dependent protein kinase [158], which strongly suggests that phosphorylation of the VDR may be an essential event required for 1,25(OH)2D-induced gene activation [155 – 158]. Phosphorylation of VDR by protein kinase A (PKA) results in a reduction in the transactivation capacity of the receptor in response to 1,25(OH)2D3 [158]. In vivo, Ser208 has been shown to be the major site of phosphorylation by casein kinase II during 1,25(OH)2D3 treatment [156, 157]. In vivo, overexpression of casein kinase II causes an increase in 1,25(OH)2D-induced transcriptional activity [159]. Phosphorylation of Ser51 by protein kinase C diminishes VDR binding and nuclear localization of the VDR [155,160] so that differential phosphorylation may play a role in determining VDR activity.

B. The Gene Encoding the VDR In 1986, the chick VDR cDNA was cloned by McDonnell et al. [161], which subsequently led to the cloning of the human VDR cDNA by Baker et al. [162]. The original VDR cDNA described by Baker et al. [162] contains 4800 nucleotides and encodes a protein of 427 amino acids with a predicted molecular mass of 48,000 Da. The cDNA sequence encodes a protein containing a cysteinerich region corresponding to the DNA-binding domain (DBD) that is highly conserved among othe steroid receptors, thus confirming that the VDR is a member of the steroid hormone receptor superfamily [163]. Structure – function analyses of the VDR showed that the DBD is localized to the amino-terminal region of the polypeptide [164]. This region contains nine cysteine residues, eight of

267

CHAPTER 9 Vitamin D

FIGURE 3 Domains of the VDR. The DNA-binding domain consists of two zinc finger modules located at the amino-terminal portion of the receptor. The ligand-binding domain contains 12 -helices shown as open boxes and 3 turns shown as a filled box. E1 and AF-2 subregions of the receptor are important in transactivation. which coordinate zinc binding to form a two zinc finger structure. The ligand-binding domain (LBD) is contained in the region C-terminal to the DBD (Fig. 3). In humans, the VDR gene has been localized to chromosome 12q13-14 [79,165,166], in close proximity to the 1hydroxylase gene [78]. The structural organization of the VDR gene has been delineated by two groups [167, 168]. Miyamoto et al. [167] initially characterized the VDR gene and its promoter and showed that the gene is composed of at least 11 exons that span 75 kb of DNA (Fig. 4). The VDR protein is encoded by 8 exons. Exon 2 and 3 encode the DBD and exons 4 – 9 encode the LBD. Exon 2 contains the most proximal 3 bp of the 5-noncoding sequence, the translation initiation site, and nucleotide sequence that encodes the first zinc finger module. Exon 3 lies approximately 15 kb downstream and encodes the second zinc finger module. The 12 -helices of the ligand-binding domain are encoded by exons 4 and 6 – 9. Exons 4, 5, and 6 encode the D region or “hinge.” Exon 5 encodes a less well-conserved region of the receptor and may represent an insertion. Exons 6, 7, 8, and 9 encode a portion of the hinge and the carboxy-terminal “E” region, as well as approximately 3200 nucleotides of the 3-noncoding sequence [167]. The original human VDR cDNA sequence, which contained two potential methionine start sites, was proposed

FIGURE 4

to encode a 427 amino acid polypeptide [162]. However, subsequent DNA sequence analyses of the VDR gene from normal individuals or patients with HVDRR (see Section VII,B) identified a polymorphism in the ATG codon encoding the first methionine [169]. By not initiating at the first ATG (M1), but at an ATG three codons downstream (M4), the polymorphism lead to a VDR shortened by three amino acids (424 amino acids) (see Fig. 11). Individuals may be homozygotic or heterozygotic at the site, leading to the detection of two closely sized VDR bands on Western blots [170,171]. The numbering system employed in this chapter uses the first methionine (M1) as the starting point, as has been the convention in most papers published to date.

C. Alternate Splicing and VDR Promoters The original VDR cDNA sequence reported by Baker et al. [162] contained approximately 115 bp of the 5-noncoding sequence. Miyamoto et al. [167] showed that portions of this noncoding sequence were contained on separate exons. They subsequently identified three noncoding exons upstream of exon 2. These exons, 1A, 1B, and 1C, were shown to be spliced differentially to generate

Organization of the VDR chromosomal gene. The human VDR gene is located on chromosome 12q13-14 and spans approximately 75 k of DNA. The gene is composed of at least six 5 noncoding exons and 8 coding exons. Alternative splicing results in at least 14 types of transcripts. Locations of the start (ATG) and termination (TGA) codons are indicated.

268

FELDMAN, MALLOY, AND GROSS

three or four mRNA species each with the potential to encode a 424 or 427 amino acid protein, depending on the polymorphism site at M1. Crofts et al. [168] expanded these findings further using RT-PCR. They showed that the 5-noncoding region of the VDR gene was distributed among at least six exons (1A – 1F), which could be spliced differentially to generate 14 mRNA transcripts [168]. Most intriguingly, these authors showed that two of the transcripts have the potential to encode N-terminal variants 23 to 50 amino acids longer than the classical 427 amino acid VDR. These two alternatively spliced transcripts encode VDRs of 450 and 477 amino acids. The polymorphic N terminus of the VDR (M1 vs M4) has been shown to influence transactivation [170], possibly by modulating TFIIB interactions [172]. Whether the presence of additional amino acids upstream of the polymorphic site modulates activity of the VDR is currently under investigation. Miyamoto et al. [167] identified a putative promoter sequence upstream of exon 1A. This GC-rich sequence contains bindings sites for transcription factor SP-1, but did not contain a TATA box. Interestingly, the intron sequence between exon 1C and exon 2 was capable of responding to retinoic acid. Crofts et al. [168] subsequently showed that the expression of the VDR gene was directed by multiple promoters upstream of exons 1A, 1D, and 1F. The two alternatively spliced transcripts encoding VDRs of 450 and 477 amino acids originated from a promoter upstream of exon 1D. Especially intriguing was the finding that one subset of transcripts originating from exon 1F, the most distal exon, was expressed only in the parathyroids, kidney, and intestine, tissues involved in calcium regulation. This finding raises the possibility of differential tissue expression of variant forms of the VDR [168].

D. Three-Dimensional Structure of the VDR LBD Models of the three-dimensional structure of the LBD of the VDR have been developed from computer-modeling studies [173,174]. More recently, the crystal structure of the VDR LBD bound to 1,25(OH)2D3 was solved and reported by Rochel et al. [175]. To obtain crystals of the LBD the authors deleted amino acids 165 – 215. Although this region of the receptor is referred to as an “insertion,” it contains the site of serine phosphorylation, Ser208. However, this region is poorly conserved among nuclear receptors and has not been shown to have a direct effect on VDR transactivation [175]. The deleted VDR LBD, amino acids 118 – 425 with the insertion deleted ( 165 – 215), exhibited 1,25(OH)2D3 and analogue binding affinities similar to the wild-type VDR and was functionally active in transactivation assays when fused to the Gal-4 DBD.

Three-dimensional structure of the holo-VDR LBD. helices are shown as cylinders and three sheets located between helix 5 and helix 6 as arrows. Helix 12 is shown in black and the ligand 1,25(OH)2D3 is docked. The location of the insertion domain deleted from the LBD is shown. Reproduced with permission from D. Moras [Mol. Cell 5, 173 – 179 (2000)]. (See also color plate.)

FIGURE 5

As shown in Fig. 5 (see also color plate), the VDR LBD based on the crystal structure is composed of 12 -helices and 3 sheets [175]. The ligand-binding pocket forms a large cavity of 693 Å and is lined with hydrophobic amino acid residues. When bound to the VDR, the A ring of 1,25(OH)2D3 embraces helix H3 and orients toward the C terminus of helix H5. The 1-hydroxyl group forms hydrogen bonds with Ser237 (H3) and Arg274 (H5), and the 3-hydroxyl group forms bonds with Ser278 (H5) and Tyr143. The conjugated triene connecting the A and C rings fits into a hydrophobic channel formed between Ser275 (loop H5- ) and Trp268 ( 1) on one side and Leu233 (H3) on the other side. The C ring contacts Trp286, and the C18 methyl group is aimed at Val234 in helix H3. The 25-hydroxyl group forms hydrogen bonds with His305 (loop H6H7) and His397 (H11). The AF-2 (activation function-2) domain is contained within helix H12. From crystallographic studies of other receptors [176,177], the H12  helix is repositioned following ligand binding such that repositioning locks the ligand in the cavity of the ligandbinding pocket. The repositioning of H12 also leads to the formation of protein surfaces that allow interaction with specific comodulators and transcriptional activation.

269

CHAPTER 9 Vitamin D

Upon ligand binding, the position of helix H12 is stabilized by hydrophobic interactions involving Thr415, Leu417, Val418, Leu419, Val421, and Phe422 from helix H12 with residues Asp232, Val234, Ser235, Ile238, and Gln239 from helix H3, Ala267 and Ile268 from helix H5, and His397 and Try401 from helix H11. In addition, a salt bridge is formed between Lys264 (H4) and Gln420 (H12), and hydrogen bonding occurs between Ser235 (H3) and Thr415 (H12). Val234, Ile268, His397, and Tyr401 also interact with the ligand, indicating that the repositioning of helix H12 is controlled by 1,25(OH)2D3 [175]. Mutations have been created in several amino acids predicted to be important in ligand binding. Mutagenesis of His397 to alanine completely abolished ligand binding, whereas mutations Ser237Ala and Cys288Ala significantly reduced the binding affinity of the receptor [174]. In addition, two natural mutations found in patients with HVDRR — Arg274Leu, which also abolished ligand binding [178], and His305Gln, which reduced the binding affinity about 10-fold [179]—also showed the importance of these amino acid residues in ligand binding. Ligand-binding modeling has also been extended to docking vitamin D analogues. In some cases, such as MC903 and EB1089, which have side chain modifications, only minor adjustments to the C and D rings are required for docking. Howevers in 20-epi analogues like KH1080 and 20-epi 1,25(OH)2D3, only the low energy conformers could be docked [175]. The large volume of the binding pocket accommodates structural differences in ligand but does not as yet explain the differential activity of various vitamin D analogues (see Section VIII).

E. Heterodimerization Transcriptional activation of target genes by 1,25(OH)2D is complex and involves a sequence of events centered around the VDR. The VDR acts as a trans-acting factor that interacts with specific VDREs located in the promoter regions of 1,25(OH)2D responsive genes. The first target gene shown to have a VDRE was the osteocalcin gene [180 – 182]. In early studies examining the binding of the VDR to the osteocalcin VDRE in yeast, it was demonstrated that a protein from a nuclear extract from mammalian cells was required for the binding to occur. This factor was originally termed a nuclear accessory factor (NAF) [183 – 185] or receptor auxiliary factor (RAF) [186]. NAF/RAF was later determined to be the retinoid X receptor (RXR), a member of the steroid – thyroid – retinoid gene superfamily [187]. RXR is a 55-kDa protein that binds 9cis-retinoic acid as its ligand [188,189] and is found widely distributed in cells and tissues, including those that do not express the VDR [185]. RXR has now been shown to be the heterodimerization partner of a number of other recep-

tors in the steroid – thyroid – retinoid gene superfamily, including thyroid receptor, retinoic acid receptor, and peroxisome proliferator activating receptor [190]. Mutagenesis of the VDR has demonstrated that the E1 region (overlapping H3 – H4) and helix H10 are required for high-affinity binding to RXR. Other regions of the receptor probably contribute to the RXR interface [2,175]. Phosphorylation of RXR by mitogen-activated protein kinase (MAPK) has been shown to inhibit 1,25(OH)2D signaling [191].

F. VDREs and Target Genes 1,25(OH)2D induces a wide array of biological responses, some resulting in an upregulation of specific mRNAs and others that downregulate protein expression. Stimulatory or inhibitory actions may depend on tissue specificity or on the state of cellular differentiation. A number of proteins have been shown to be stimulated by 1,25(OH)2D, including osteocalcin, Eta-1 (for early T lymphocyte activation-1 also known as osteopontin), fibronectin, alkaline phosphatase, carbonic anhydrase, 24-hydroxylase, prolactin, c-fos, PSA, calbindin-D9K, calbindin-D28K, p21, p27, 3 integrin, sodium-dependent phosphate cotransporter 2 (NPT2), nerve growth factor (NGF), transforming growth factor 2 (TGF 2), estrogen receptor  (ER), lipoprotein lipase, and aromatase (CYP19). Proteins downregulated by 1,25(OH)2D include collagen, PTH, PTHrP, calcitonin, IL-2, atrial naturetic peptide, Bcl-2, 1hydroxylase, and c-myc. A number of these genes have been shown to have VDREs in their promoters. The human osteocalcin VDRE (GGGTGA ACG GGGGCA) is an imperfect hexanucleotide direct repeat separated by a three nucleotide spacer sequence, a so-called DR3 structure [180 – 182]. Unlike the osteocalcin VDRE, the mouse osteopontin VDRE is a perfect direct repeat of the hexanucleotide GGTTCA separated by a three nucleotide spacer. It is clear that additional aspects of the VDRE region determine whether the 1,25(OH)2D action on a target gene is stimulatory or inhibitory. The complexity in understanding how 1,25(OH)2D responsive genes are regulated in vivo is highlighted by the different VDRE structure that exist as well as the multiple possibilities for comodulator interactions.

G. Comodulator Interactions A large number of proteins, known as comodulators or coregulators, have been found to interact with the nuclear receptors and are required for gene transcription [2,8,9,139]. The particular coregulatory proteins recruited to nuclear receptors, influenced by conformational changes in the receptor – ligand complex, may contribute to the

270

FELDMAN, MALLOY, AND GROSS

specificity of transcriptional regulation [192]. Specific interactions with the VDR have been shown for a number of proteins. In addition to heterodimerizing with RXR, the VDR has been shown to bind SRC-1, a member of the p160 class of comodulators. SRC-1 has been shown to have intrinsic histone acetyltransferase (HAT) activity [193]. Interaction of the VDR with SRC-1 is ligand dependent and involves the AF-2 domain (helix H12) and helix H3 [194,195]. Mutagenesis of Tyr236 to alanine in H3 blocks SRC-1 interaction and transactivation but does not interfere with RXR or ligand binding [195]. The VDR has also been shown to interact with basal transcription factor TFIIB, a protein associated with the basal transcriptional machinery [172,196 – 199]. TFIIB interaction does not involve the AF-2 domain [198], instead it has been shown to bind to the N-terminal portion of the receptor [172, 199]. Using the VDR as bait in the yeast two-hybrid system, Baudino et al. [200] isolated a protein they termed NCoA62 (nuclear receptor coactivator; 62,000 Da). Addition of NCoA-62 to transactivation assays augmented 1,25(OH)2D3 responsiveness; however, the role of this protein in transactivation has not been defined as yet. A large class of proteins collectively called vitamin D – receptor-interacting proteins or DRIPs was isolated by affinity chromatography using glutathione S-transferase (GST)-VDR LBD immobilized to glutathione agarose [201,202]. The DRIP complex only bound to the immobilized VDR when 1,25(OH)2D3 was present. In cell-free transcription assays, DRIPs mediate ligand-dependent gene transcription by the VDR. At least 13 proteins consititute the DRIP complex, although only DRIP205 binds directly to the VDR. The other DRIPs must be recruited to the growing complex of proteins subsequent to DRIP205 binding. Smad3, a member of the Smad protein family of intracellular signal transducers of the TGF- -BMP superfamily, has also been shown to act as a coactivator of the VDR [203, 204]. Smad3 binds to the VDR LBD and enhances the transactivation function of the receptor. The Smad3– VDR complex is stabilized by SRC-1 [204].

H. Transactivation of Target Genes In the bloodstream, 1,25(OH)2D circulates mostly bound to DBP in equilibrium with a small amount of unbound or free hormone. It was generally believed that the unoccupied VDR was localized to the nucleus and that no special mechanism was required for the 1,25(OH)2D to enter the cell and make its way to the nucleus. However, 1,25(OH)2D may bind to a membrane receptor or interact with tubulin during its delivery to the VDR and translocation to the nucleus [205,206]. The role of 1,25(OH)2D in the transport of VDR from cytoplasm to nucleus has been examined

using green fluorescent protein (GFP)-tagged chimeras of full-length or truncated constructs of the VDR [207]. 1,25(OH)2D treatment showed translocation of cytoplasmic VDR to the nucleus and colocalization with RXR [208]. Truncation of either the LBD or the AF-2 region of VDR abolished hormone-dependent translocation and transactivation. The findings support the model of hormone-dependent VDR translocation and indicate that translocation from the cytoplasm to the nucleus is part of the receptor activation process. An intact AF-2 domain is required for this translocation [207]. Photobleaching data suggest that the VDR shuttles back and forth between the cytoplasm and the nucleus and that ligand increases the nuclear accumulation of VDR [208]. A simplified model of 1,25(OH)2D-regulated gene transactivation is shown in Fig. 6 (see also color plate). In the absence of 1,25(OH)2D, the VDR exhibits a low affinity for RXR or is in the cytoplasm. Upon 1,25(OH)2D binding, the VDR translocates to the nucleus and RXR proteins heterodimerize and form a high-affinity complex that acquires the ability to recognize and bind with high affinity to VDREs through their cognate DBDs. As determined from the crystal structure of the VDR LBD and models based on other steroid receptors, the VDR undergoes a major conformational change upon 1,25(OH)2D binding. The repositioning of helix H12 upon the upstream -helices in the E1 domain forms high-affinity protein surfaces capable of interacting with specific comodulator proteins required for transactivation. The VDR – RXR heterodimer attracts coactivator proteins such as SRC-1 and NCoA-62. SRC-1, through its intrinsic HAT activity, derepresses chromatin so that nucleosomes are rearranged and naked DNA is accessible, allowing for the assembly of the transcriptional apparatus. TATA-binding protein-associated factors (TAFs) are also recruited to target TATA/TBP-binding sites. Other proteins, including DRIPs and TFIIB, serve to stabilize the preiniation complex. Transcription is then initiated by RNA polymerase II. For stimulatory actions, mRNAs transcription is increased and the specifically induced mRNAs are translated into proteins that carry out the biological effects of the hormone. For inhibitory actions, mRNA transcription is suppressed, thereby reducing the expression level of selected proteins.

I. Regulation of VDR Abundance Within each target tissue, the level of VDR is not fixed but rather is regulated dynamically by a variety of physiological and developmental signals. The concentration of VDR expressed in a target cell determines the amplitude of the response evoked by 1,25(OH)2D treatment. Upregulation of VDR enhances the response to 1,25(OH)2D, and downregulation diminishes the response [209 – 213]. Of

271

CHAPTER 9 Vitamin D

FIGURE 6

Model of 1,25(OH)2D gene transactivation. Upon entering the cell, 1,25(OH)2D3 binds to the VDR, leading to the formation of a VDR:RXR heterodimer (1). The heterodimeric complex subsequently binds to vitamin D response elements (VDREs) in promoter regions of target genes through their cognate DNA-binding domains (2). Conformational changes in the VDR:RXR heterodimer initiate the recruitment of coactivating proteins, including SRC-1 and NCoA-62, to the oligomeric complex. The histone deacetylase activity of SRC-1 modifies the chromatin structure and facilitates essential contact with the general transcription apparatus (3). Additional proteins are recruited to the complex such as TBP and TAFs for targeting promoter elements (4). Binding of TFIIB and DRIPs to the complex stabilizes the preinitiation complex (5). Once the proteins have been assembled, transcription is initiated by RNA polymerase II (6). Reproduced with permission from M. R. Haussler [J. Cell. Biochem. Suppl. 32/33, 110 – 122 (1999)]. (See also color plate.)

the many factors that regulate VDR abundance, the hormone 1,25(OH)2D itself is an important modulator that increases the concentration of the receptor (homologous upregulation). Other regulators include steroid and peptide hormones, growth factors, activators of specific second messenger pathways, and intracellular calcium, which may up- or downregulate the concentration of VDR (heterologous regulation) [213]. VDR abundance has been shown to depend on the proliferation/differentiation status of the target cells, and VDR changes are also detectable during neonatal development in different tissues [213]. 1. HOMOLOGOUS REGULATION The VDR is upregulated by 1,25(OH)2D and other vitamin D metabolites that bind to the VDR itself (homologous regulation), and this has been observed both in vitro [214] and in vivo [215 – 217]. The magnitude of homologous upregulation varies from 2 to 10-fold depending on the target cell. In pig kidney cells, human skin fibroblasts, and human mammary cancer cells (MCF-7), the VDR content increases when the cells are treated with 1,25(OH)2D3, 1,24,25(OH)3D3, 24,25(OH)2D3, and 25(OH)D3, and the concentrations required for maximal upregulation closely reflect the affinities of the various metabolites for the VDR [214]. Several studies have shown that the upregulation of

the VDR is due to an increase in the transcription of the VDR gene [161,217,218]; however, other studies have found that the upregulation is mainly due to the stabilization of the ligand-occupied VDR [219 – 221]. Either one or both of these phenomena may be operative depending on the target cells under study [214,220]. When examined carefully in pig kidney cells, about two-thirds of the upregulation appeared to be due to the stabilization of the VDR and one-third due to the increased synthesis of the VDR protein [214]. Homologous upregulation of VDR appears to play a role in the treatment of psoriasis, a hyperproliferative skin disorder. Chen et al. [222] have shown that the response to 1,25(OH)2D treatment in patients with psoriasis correlated with the upregulation of VDR in psoriatic skin. In patients who showed clinical improvement with treatment, significant upregulation of VDR mRNA was observed in psoriatic lesions, whereas there was no upregulation in patients who did not respond to 1,25(OH)2D. 2. HETEROLOGOUS REGULATION Various hormones, including steroid and peptide hormones, and growth factors regulate VDR abundance (heterologous regulation) in a cell and tissue-specific manner. In cultured cells, the abundance of VDR has been shown to be closely related to the rate of cell proliferation, with VDR

272 levels being higher in proliferating cells than in quiescent cells [223,224]. Also, in some cell systems, the induction of differentiation leads to decreased VDR levels [218,225, 226]. However, activation of the protein kinase C pathway by mitogens such as basic fibroblast growth factor and phorbol esters and/or elevation of intracellular Ca2 led to a significant decrease in VDR abundance, despite stimulating cell proliferation [227]. The downregulation appears to result from a destabilization of the VDR mRNA or a decrease in VDR gene expression. In contrast, elevation of intracellular cAMP has been shown to increase VDR abundance by increasing VDR mRNA expression [211,228]. The mechanism of the cAMP-mediated upregulation may be due to an increase in the transcriptional rate of the VDR gene, as shown by a study characterizing a 1.5-kb promoter region of the mouse VDR gene [229]. Miyamoto et al. [167] cloned and characterized approximately 1.5 kb of the promoter region of the human VDR gene 5 of the transcriptional start site and demonstrated the presence of several potential binding sites for the SP-1 transcription factor and putative sites for other transcriptional activators, including cAMP response elements. Prostaglandins have also been shown to regulate VDR levels in human leukemia cells, probably through the elevation of cAMP levels in these cells [230]. While studying the regulation of the VDR gene in the intestine, Yamamoto et al. [231] demonstrated the presence of an intestine-specific cis element in the human VDR gene promoter (at 3731 to 3720), which interacts with a caudal-related homeodomain transcription factor, Cdx-2. Cdx-2 is able to activate VDR gene transcription in the intestine by binding to this element. Glucocorticoids [232 – 235], estrogens [236,237], retinoic acid [238,239], and PTH [212,228,240] have been shown to regulate VDR abundance. Changes in VDR abundance elicited by these hormones are reflected in the magnitude of 1,25(OH)2D responsiveness. However, species differences among various rodent models prevent extrapolation to humans. Even within a species, there may be tissue-specific differences. An intron fragment 3 of exon 1C of the human VDR gene appears to confer retinoic acid sensitivity, suggesting a molecular mechanism for the regulation of VDR by retinoic acid [167]. In the case of excess glucocorticoids, there is a resistance to 1,25(OH)2D, whereas in the case of excess PTH there is enhanced 1,25(OH)2D responsiveness [212,228]. Thus these hormones modulate target cell sensitivity to 1,25(OH)2D in part through the regulation of VDR levels. The abundance of VDR in parathyroid glands has been studied extensively [86]. It has been postulated that reduced VDR concentration in parathyroid glands may be related to a lack of vitamin D suppression of the gland [241]. This may contribute to the pathogenesis of secondary hyperparathyroidism in chronic renal failure by reducing the inhibition by 1,25(OH)2D of parathyroid hormone secretion

FELDMAN, MALLOY, AND GROSS

[6,242]. The low serum levels of 1,25(OH)2D in chronic renal failure may accentuate this effect [242]. Similarly, vitamin D status may alter the pattern of signs and symptoms in primary hyperparathyroidism [243 – 245].

V. NONGENOMIC EFFECTS OF VITAMIN D A. Rapid Responses to 1,25(OH)2D In addition to the classical VDR-mediated genomic pathway, 1,25(OH)2D has also been shown to elicit rapid responses [126,246,247]. The term “rapid response” is used to describe the biological effects of 1,25(OH)2D that occur within a few minutes after hormone treatment and are considered too rapid to be explained by a VDR-mediated genomic pathway. Rather, they are thought to be mediated by a direct action of 1,25(OH)2D on the plasma membrane of target cells stimulating a signal transduction pathway involving the rapid opening of voltage-sensitive Ca2 channels and activation of protein kinases. Some of the 1,25(OH)2D-induced rapid responses include changes in intracellular calcium flux, alteration in phospholipid metabolism and phosphate transport, and changes in alkaline phosphatase and adenylate cyclase activities. Also, “transcaltachia,” a process of transluminal transport of Ca2 across the intestine, has been shown to occur rapidly when vitamin D-repleted animals are treated with 1,25(OH)2D3. The rapid Ca2 transport has been proposed to be facilitated by endocytic and lysosomal vesicles, which deliver the Ca2 to the basal – lateral membrane where it is released by exocytosis into the lamina propria. However, because the transcaltachia response requires vitamin D-replete animals, a preexisting condition induced by 1,25(OH)2D may be operative, and thus transcaltachia may ultimately depend on a 1,25(OH)2D – VDR-mediated genomic pathway. One rapid effect observed in 1,25(OH)2D3-treated cultured human fibroblasts was the rapid accumulation of cGMP, which colocalizes with reorganizing VDRs within 15 s after calcitriol addition [248]. However, this rapid effect was dependent on having a functional VDR, as no increase in cGMP was seen in fibroblasts from patients with HVDRR containing defective receptors. In contrast, in rat osteosarcoma cells (ROS 24/1) that are totally devoid of detectable VDRs, 1,25(OH)2D3 induced the rapid flux of Ca2 into the cells, suggesting that a 1,25(OH)2D3 signaling system independent of the VDR exits in at least in some cells [249]. Several lines of evidence support the existence of a nongenomic 1,25(OH)2D-mediated signal transduction pathway. For instance, the antagonist 1,25(OH)2D3, which has a minimal effect on 1,25(OH)2D-induced genomic actions, blocks the effect of 1,25(OH)2D3 on transcaltachia

273

CHAPTER 9 Vitamin D

[250]. Similarly, some vitamin D analogues, such as the 6s-cis-blocked conformer that binds poorly to the VDR, are able to generate the transcaltachia response in perfused chick intestine and Ca2 influx in ROS 17/2.8 cells [247, 251]. In NB4 cells, an acute promyelocytic leukemia cell line, the 6-s-cis-blocked conformer was 20 times more effective at priming the cells for monocytic differentiation than the natural hormone. This response was attenuated by the 1,25(OH)2D3, a specific antagonist of the nongenomic response [252]. The 6-s-cis analogue, 125(OH)2lumisterol3, also induces transcaltachia and stimulates Ca2 uptake in the ROS 17/2.8 osteosarcoma cell line [251]. 125(OH)2lumisterol3 was also shown to augment glucoseinduced insulin secretion in rat pancreatic islet cells while also increasing intracellular Ca2 concentrations [253].

B. Membrane VDR One possible mechanism for these rapid, nongenomic effects is that there is a specific membrane receptor for 1,25(OH)2D that mediates the rapid responses. The presence of a specific binding protein for 1,25(OH)2D3 has been identified in plasma membrane preparations, and a protein that binds [3H]1,25(OH)2D3 was partially purified from a basal – lateral membrane preparation from chick intestine [254]. The putative membrane VDR exhibits a different ligand specificity than the classical nuclear VDR. Using antiserum to this protein, Nemere et al. [254] demonstrated that this protein was present in rat costochondral cartilage cells and that the antibody could block the 1,25(OH)2D3 increase in protein kinase C activity. Immunohistochemistry demonstrated that both resting and growth zone chondrocytes express the protein, but levels are highest in the growth zone. The binding protein is present in both plasma membranes and matrix vesicles and has a molecular mass of 66,000 Da. The putative membrane VDR was also shown to mediate the antiproliferative effects of 1,25(OH)2D3 on chondrocytes [126]. However, at the time of writing, the membrane receptor has not as yet been cloned or characterized at the molecular level.

bined excretion in feces (720 mg) and urine (280 mg) [257]. The coordinated interaction of 1,25(OH)2D and PTH to regulate 1-hydroxylase activity plays a major role in the maintenance of calcium balance (Fig. 7). Small decreases in serum calcium result in increases in PTH secretion, which stimulates the upregulation of 1-hydroxylase activity, and increased renal phosphate excretion. The combination of increased PTH and decreased phosphate leads to enhanced 1-hydroxylase activity. The augmented synthesis of 1,25(OH)2D enhances intestinal calcium absorption to restore the calcium concentration toward normal levels, which in turn feeds back to diminish PTH secretion, thereby limiting the further production of 1,25(OH)2D. In addition, 1,25(OH)2D feeds back on the kidney to inhibit the further production of 1,25(OH)2D by downregulating 1-hydroxylase gene expression while stimulating 24hydroxylase gene expression. Furthermore, serum calcium is maintained by the combined actions of PTH and 1,25(OH)2D on the bone to increase bone resorption and by the action of PTH on the kidney to increase calcium reabsorption. In hypercalcemic states, PTH is suppressed by a signal transmitted via the parathyroid CaR [94], and the entire process is reversed. In rat parathyroid glands and kidney, the expression of the CaR gene is increased by 1,25(OH)2D but not by Ca2 [258]. Upregulation of the CaR is thought to be involved in the suppressive effects of vitamin D compounds on PTH secretion. The selective action of noncalcemic vitamin D analogues that have a greater suppressive effect on PTH expression may allow for their potential use in therapeutic situations with elevated PTH concentrations [259] (see Section VIII).

VI. PHYSIOLOGY: REGULATION OF SERUM CALCIUM A. Interaction of PTH and Vitamin D to Regulate Serum Calcium The concentration of Ca2 in plasma and extracellular fluid is maintained within a narrow range, with variations up or down being associated with untoward effects [6,87,255,256]. In the balanced state, the dietary intake of approximately 1000 mg of calcium is equal to the com-

FIGURE 7 PTH.

Regulation of Ca2 levels in the blood by 1,25(OH)2D and

274

FELDMAN, MALLOY, AND GROSS

B. Extrarenal 1,25(OH)2D Synthesis and Hypercalcemic States Under normal physiological conditions the kidney is the primary site of 1,25(OH)2D formation. However, small amounts of 1,25(OH)2D are produced in various other tissues, and in selected pathological conditions the extrarenal production of 1,25(OH)2D may significantly contribute to alterations in calcium homeostasis [260]. Tissues shown to synthesize 1,25(OH)2D from 25(OH)D include human decidua and placenta, bone cells, keratinocytes, spleen, melanoma cells, hepatoma cells, and peritoneal, synovial, and pulmonary monocytes and macrophages [261]. Hypercalcemia can be expected to occur in 7 to 24% of patients with sarcoidosis [262]. Proof of the clinical significance of extrarenal production of 1,25(OH)2D was provided from studies on an anephric patient with sarcoidosis who developed hypercalcemia [263]. Cultured pulmonary alveolar macrophages from patients with sarcoidosis [101,107] and homogenized sarcoid lymph node tissue [264] have been shown to be capable of producing 1,25(OH)2D. In addition to sarcoidosis, other granulomatous disorders have been associated with hypercalcemia and elevated 1,25(OH)2D levels, including tuberculosis, leprosy, silicone-induced granulomatosis, and disseminated candidiasis [260,261]. Hypercalcemia in lymphoma patients due to elevations in 1,25(OH)2D is a well-known phenomenon and has been reviewed [265]. Both Hodgkin’s and non-Hodgkin’s lymphoma have been associated with elevated circulating 1,25(OH)2D [261,265]. Hypercalcemia in these disorders is estimated to occur in 5% of patients with Hodgkin’s disease and in 15% of patients with non-Hodgkin’s lymphoma. In one report, 1,25(OH)2D levels were elevated in 55% of a group of 22 hypercalcemic patients with non-Hodgkin’s lymphoma, and many of the normocalcemic patients with non-Hodgkin’s lymphoma had evidence of dysregulated 1,25(OH)2D synthesis [266]. Lymphocytes transformed with HTLV-1 have been shown to convert 25(OH)D to 1,25(OH)2D in vitro indicating that these lymhoma-like cells have 1-hydroxylase activity, and evidence shows that lymphomatous tissue in vitro can convert 25(OH)D to 1,25(OH)2D. However, whether the lymphoma cell itself or associated macrophages are responsible for the 1-hydroxylase activity found in lymphoma patients is still unclear at present [266]. Elevated 1,25(OH)2D concentrations are observed in pregnancy and appear to increase as gestation progresses [267]. DBP is stimulated by estrogens, and both total and free 1,25(OH)2D moieties are elevated during pregnancy and estrogen therapy [268, 269]. Only the free hormone is thought to be active [270]. The increased 1,25(OH)2D may augment the intestinal absorption of calcium that occurs during pregnancy, which is necessary to supply calcium to the developing fetal skeleton [267]. Extrarenal production of 1,25(OH)2D has been shown to occur in anephric

pregnant rats [271], in human placental tissue in vitro [103, 272], and in two pregnant women with pseudohypoparathyroidism, in whom renal production of 1,25(OH)2D was low [273]. The metabolism of vitamin D during pregnancy has been reviewed [274]. The use of the antifungal drug ketoconazole as a diagnostic test or as therapy for hypercalcemic states has been suggested [275 – 278]. Ketoconazole inhibits fungal growth by blocking the P450 enzyme 14-demethylase in the pathway to ergosterol synthesis [279]. The drug has been shown to inhibit mammalian P450 enzymes, including 24-hydroxylase [280] and 1-hydroxylase [281].

C. Local 1-Hydroxylases, Possible Autocrine/Intracrine Activity As a form of therapy in the preantibiotic era, patients with tuberculosis were often sent to sanitoria to increase their exposure to sunlight. It now seems possible that the benefits of sunlight may in part have been due to increased vitamin D synthesis. Data suggest that local 1-hydroxylase expression by macrophages leads to functionally important local production of 1,25(OH)2D, which acts in an intracrine, autocrine, and/or paracrine fashion. There is developing evidence that vitamin D deficiency is associated with an increased risk of contracting tuberculosis [282,283]. Furthermore, the synthesis of 1,25(OH)2D by macrophages may suppress Mycobacterium tuberculosis growth, perhaps by altering the local release of cytokines or nitric oxide [284]. Alternatively, 1,25(OH)2D may enhance macrophage function, perhaps by stimulating the local production of Eta-1 to enhance cell-mediated immunity [285]. In addition, analysis of polymorphisms in the VDR, especially when coupled with vitamin D deficiency, suggest that there is an increased risk of tuberculosis in certain populations [283]. On the other hand, in inflammatory arthritis, cytokines may elicit the local production of 1,25(OH)2D in the joints where 1,25(OH)2D may contribute to periarticular bone loss [286,287]. 1-Hydroxylase activity has been found in human colon [102] and prostate [104] tissue and cancer cell lines. Although the reason for its expression in these organs is unknown, it is possible that the local production of 1,25(OH)2D serves as an antiproliferative [288 – 291] and/or prodifferentiating [102,218,288,292] agent in colon and/or prostate development. Further studies in this area will help clarify the importance of 1-hydroxylase expression in these organs. Keratinocytes express 1-hydroxylase activity, and cultured keratinocytes from skin biopsies have been used to examine mutations that are associated with 1-hydroxylase deficiency [76,293]. 1,25(OH)2D acts as as antiproliferative and prodifferentiating agent in keratinocytes [35,294] and

275

CHAPTER 9 Vitamin D

is useful in treating psoriasis [295]. Recent data suggest that the beneficial effects of 1,25(OH)2D on psoriasis are due to its antiproliferative/prodifferentiating actions, as well as its ability to inhibit inflammation in dermal cells [296,297].

VII. GENETIC DISORDERS AND VITAMIN D RECEPTOR POLYMORPHISMS Examples of both over- and underproduction of the 1hydroxylated vitamin D sterols are not uncommon. Disorders associated with increased renal production of 1,25(OH)2D include hyperparathyroidism and tumoral calcinosis. Conditions that have decreased production of 1,25(OH)2D as part of their clinical picture include hypoparathyroidism and pseudohypoparathyroidism, renal failure, X-linked hypophosphatemic rickets, oncogenic osteomalacia, and hereditary 1-hydroxylase deficiency [298 – 303].

A. 1-Hydroxylase Deficiency (VDDR-I, PDDR) The clinical findings of hereditary complete deficiency of renal 1-hydroxylase were first described in 1961 by Prader et al. [304]. This disease was known as vitamin D-dependent rickets type I (VDDR-I) or pseudo vitamin D deficiency type I and, more recently, as pseudo vitamin D deficiency rickets (PDDR). This chapter refers to this genetic disorder as 1-hydroxylase deficiency now that the disease is proven to be caused by mutations in the cytochrome P450 1-hydroxylase gene (refered to as either CYP27B1 or CYP1). 1-Hydroxylase deficiency is a rare autosomal recessive disease that is manifested at an early age [76,84,302,305,306]. Hypocalcemia, elevated PTH levels, increased alkaline phosphatase, and low urine calcium are found. Affected children present with hypotonia, muscle

weakness, growth failure, and rickets. Tetany and convulsions may occur with severe hypocalcemia. As expected, patients with 1-hydroxylase deficiency have normal 25(OH)D concentrations and low levels of 1,25(OH)2D. Circulating 1,25(OH)2D does not increase after PTH infusion, consistent with defective 1-hydroxylase activity. Very large doses of vitamin D or 25(OH)D are required for adequate treatment of 1-hydroxylase deficiency; often 20,000 to over 100,000 IU of vitamin D daily is needed. However, modest doses of 1,25(OH)2D (0.25 – 2 g/day), which bypass the deficient enzyme, tend to be sufficient to restore calcium to normal and heal the rickets [307]. 1-Hydroxylase deficiency was presumed to be due to mutations in the gene encoding the 1-hydroxylase enzyme, and family linkage studies localized the defect to chromosome 12q14, close to the locus for the VDR [78, 79]. Since the cloning of the 1-hydroxylase gene, several groups have demonstrated that this disease is caused by mutations in the 1-hydroxylase gene. A number of missense mutations scattered throughout the entire region of the 1-hydroxylase gene have been identified that disrupt the enzyme activity. In one particularly interesting case, a patient with 1-hydroxylase deficiency was shown to have two mutations: a frameshift and a deletion on separate alleles, which together resulted in two null mutations [76]. Figure 8 illustrates the location of 1-hydroxylase mutations thus far reported [72,76,77,293,302,305,308].

B. Hereditary 1,25-Dihydroxyvitamin D-Resistant Rickets (HVDRR) Hereditary 1,25-dihydroxyvitamin D-resistant rickets (HVDRR), also known as vitamin D-dependent rickets type II (VDDR-II) or pseudo-vitamin D deficiency type II, is a rare genetic disease that arises as a result of mutations in the gene encoding the VDR [7,118,309 – 312]. This autosomal recessive disease is characterized by early onset rickets, hypocalcemia, secondary hyperparathyroidism, and, in

FIGURE 8 Mutations in the CYP27B1 gene causing 1-hydroxylase deficiency. Boxes represent exon sequences with intron sequences illustrated by connecting lines. Shaded boxes include untranslated sequences. Reproduced with permission from S. Kato [Mol. Cell. Endocrinol. 156, 7 – 12 (1999)].

276 contrast to 1-hydroxylase deficiency, normal or elevated serum 1,25(OH)2D levels. The heterozygotic parents have no evidence of bone disease, but a history of consanguinity is usually present. In many cases, total body alopecia accompanies the disease and provides initial evidence of the HVDRR syndrome [7]. The first description of a mutation in any member of the steroid receptor gene superfamily causing a hormone resistant phenotype was presented by Hughes et al. [313], who identified missense mutations in two unrelated families whose VDRs exhibited abnormal DNA binding [314,315]. In one family, a Gly33Asp mutation was identified in the first zinc finger module of the DBD, whereas in the second family, a Arg73Gly mutation was identified in the second zinc finger of the DBD [313]. The mutations were recreated by site-directed mutagenesis of the wild-type VDR cDNA and were shown to exhibit the properties of the native

FELDMAN, MALLOY, AND GROSS

mutant VDR, as well as to be defective in 1,25(OH)2D3induced gene transactivation [316]. These results clearly established that defects in the VDR gene were the etiology of 1,25(OH)2D resistance and the origin of the HVDRR phenotype. Several other mutations in the DBD have since been described [169,317 – 319] and all of the mutations reported are summarized elsewhere [7] (see Fig. 9). The first LBD mutation to be described was a nonsense mutation (TAC to TAA) that resulted in the premature termination of the VDR at amino acid 295 (Tyr295stop) [320]. The same mutation was subsequently found in multiple interrelated families comprising a large kindred [321]. Several other mutations that result in premature termination of the VDR have now been described [178,322]. Two premature termination signals occurred as a result of frameshifts. In one case, a 5 donor splice site was mutated, which caused exon 4 to be skipped in the processing of the VDR

FIGURE 9 Mutations in the VDR causing hereditary vitamin D-resistant rickets (HVDRR). (A) The two zinc finger modules and the amino acid composition of the DBD. Conserved amino acids are depicted as shaded circles. Mutations are indicated by large arrows and circles. The location of the intron separating exon 2 and exon 3, which encode the separate zinc finger modules, is indicated by an arrow. Numbers specify amino acid number. (B) Location of -helices (H1-H12) of the VDR LBD. -helices are depicted as open boxes, and the region containing  turns is drawn as a shaded box. E1 and AF-2 regions are shown above the -helices. The location of the mutations is indicated by arrows. fs, frameshift mutation.

CHAPTER 9 Vitamin D

transcript. In the other case, a cryptic 5 splice site was generated by a mutation in exon 6 (Fig. 9). The first missense mutation identified in the VDR LBD was an Arg274Leu. The mutant receptor exhibited a decreased binding affinity for [3H]1,25(OH)2D3 of about 1000-fold [178]. Crystallographic studies of the VDR have shown that Arg274 is a contact point for the 1-hydroxyl group of 1,25(OH)2D3 [175]. Several other missense mutations have now been identified in the VDR LBD (Fig. 9). A His305Gln mutation caused a 5- to 10-fold reduction in [3H]1,25(OH)2D3 binding and a similar reduction in gene transactivation [179]. Crystallographic studies of the VDR have shown that His305 makes contact with the 25hydroxyl group of 1,25(OH)2D3 [175]. An Arg391Cys mutation was shown not to affect ligand binding but interaction with RXR [323]. In one special case, a patient who exhibited all the signs of HVDRR, including alopecia, no mutation was found in the VDR [324]. This case illustrates the fact that additional proteins are involved in 1,25(OH)2D3 signaling and that disruption of these proteins may cause the HVDRR syndrome. Indeed, a recently published report suggests that genetic defects in coactivator molecules are associated with steroid hormone-resistant syndromes [325]. A prenatal test for HVDRR has been described in a study using amniotic fluid cells or chorionic villus samples [326]. HVDRR was confirmed using assays to determine the level of [3H]1,25(OH)2D3 binding and 1,25(OH)2D3induced 24-hydroxylase activity, as well as by restriction fragment length polymorphism (RFLP) analysis [327]. The successful treatment of children with HVDRR who are unresponsive to large doses of vitamin D derivatives or oral calcium supplements has been achieved by the chronic intravenous administration of calcium [328 – 330]. The intravenous calcium infusions were given nightly over a period of many months and, by bypassing the intestinal defect in calcium absorption, over time were able to correct the hypocalcemia. The treatment eventually resulted in the normalization of serum calcium levels, correction of secondary hyperparathyroidism, and normal mineralization of bone and healing of rickets on X-ray. The clinical improvement can be sustained if adequate serum calcium and phosphorus concentrations are maintained. Despite healing of the rickets, the alopecia does not improve as a consequence of the treatment. In 1997, VDR knockout mouse models were generated by two groups. Yoshizawa et al. [331] disrupted exon 2 to generate the VDR null (-/-)genotype, whereas Demay and colleagues [332, 333] deleted exon 3. In both models, VDR (-/-) mice were phenotypically normal at birth, suggesting that 1,25(OH)2D3 actions are not necessary for normal embryogenesis. After weaning, the mice became hypocalcemic and developed rickets similar to patients with HVDRR. Alopecia also appeared progressively as

277 the mice aged. Most of the VDR (-/-) mice generated by Yoshizawa et al. [331] were infertile and died by 15 weeks after birth. Uterine hypoplasia and impaired folliculogenesis were also noted. Mice generated by Li et al. [332] survived at least 6 months. In both mouse models, the survival of the knockout mice was enhanced by a high calcium diet supplemented with lactose [333]. Many, but not all, of the abnormalities in the reproductive organs were eliminated by the maintenance of normal calcium levels [334]. A role for 1,25(OH)2D3 in regulating estrogen levels via its regulation of aromatase gene expression [335] was not corrected by calcium repletion [334]. As in the human disease HVDRR, normalization of calcium eliminated many of the abnormalities, but not alopecia.

C. X-Linked Hypophosphatemic Rickets (XLH) X-linked hypophosphatemia (XLH) is an X-linked dominant disorder causing phosphate wasting and hypophosphatemia [299,303]. The primary defect is expressed as an abnormality of the renal proximal tubule that impairs phosphate reabsorption. Patients with XLH develop progressively severe skeletal abnormalities and growth retardation. The clinical presentation ranges from mild abnormalities to severe rickets and osteomalacia. Children with the disease exhibit rachitic bone deformities, including enlargement of the wrists and knees and bowing of the lower extremities secondary to rickets. The clinical features of XLH are not apparent until 6 – 12 months of age and may also include defects in tooth development and premature cranial synostosis. The disorder is also associated with low or inappropriately normal circulating levels of 1,25(OH)2D in the presence of low serum phosphate, which would normally increase 1-hydroxylase activity, suggesting the defective regulation of 1-hydroxylase. The gene causing XLH has been cloned and identified as PHEX, a PHosphate-regulating gene with homologies to Endopeptidases located on the X chromosome [336,337]. The PHEX gene is homologous to a family of endopeptidase genes, which includes endothelin-converting enzyme1 and neutral endopeptidase. The PHEX gene encodes a 749 amino acid membrane-bound protein that is expressed in bone, adult ovary, lung, and fetal liver. A number of genetic defects in the PHEX gene have been identified in XLH patients, including deletions, insertions, duplications, splice site, and missense and nonsense mutations [337]. The X-linked dominant expression of the disorder is likely due to haploinsufficiency rather than dominant-negative effects, as many of the mutations are inactivating mutations [303]. A possible role for PHEX in the pathophysiology of XLH is illustrated in Fig. 10 [303]. As seen in the model,

278

FELDMAN, MALLOY, AND GROSS

FIGURE 10 Model for X-linked hypophosphatemia. In the osteoblast, both the PHEX protein and phosphatonin are made. Phosphatonin is secreted from the cell in an active state (PTNa). Under normal conditions, the PHEX protein degrades some of the phosphatonin to an inactive polypeptide (PTNi). The remaining active phosphatonin is then able to bind to a membrane receptor on the renal tubule cell surface. A signal is then transmitted to the cell to downregulate NPT2 activity and regulate phosphate reabsorption. In XLH, defects in the PHEX protein prevent the degradation of phosphatonin, allowing for greater circulating levels of the hormone. The signal transmitted by the higher phosphatonin levels is amplified, resulting in a greater inhibition of NPT2 activity, leading to phosphate wasting. Reproduced with permission from M. K. Drezner [Kidney Int. 57, 9 – 18 (2000)].

normal osteoblasts express both PHEX and “phosphatonin,” a putative phosphate-lowering peptide hormone. The membrane-bound PHEX degrades some of the active phosphatonin to an inactive metabolite. The remaining phosphatonin in the circulation then interacts with a renal tubule cell receptor. Through an unknown mechanism, a signal is sent to downregulate to a small degree the sodium-dependent phosphate cotransporter (NPT2) in the kidney. In XLH, defective PHEX proteins are unable to degrade phosphatonin, leading to excessive amounts of this protein in the circulation. Consequently, the signal to downregulate the NPT2 is magnified, leading to phosphate wasting [303]. The pathophysiology of XLH is similar to oncogenic osteomalacia or tumor-induced osteomalacia (TIO) in which phosphate depletion predominates [338]. However, in TIO, the acquired hypophosphatemic state is secondary to a tumor, which is generally a small benign lesion of mesenchymal origin. Removal of the tumor leads to the remission of the clinical abnormalities, suggesting that a putative circulating factor, “phosphatonin,” is produced by the tumor and causes phosphate wasting by the kidney [303]. As in XLH, phosphatonin in TIO has been postulated to act on signaling pathways that affect the NPT2 and clearance of phosphate [303].

D. VDR Polymorphisms The role of VDR polymorphisms in osteoporosis is covered extensively in several reviews [339, 340] and in Chapter 26. Here we briefly discuss the VDR polymorphisms and their potential role in predicting the risk of developing osteoporosis. Whereas mutations in the VDR gene can cause dramatic changes in gene function [7], polymorphisms may have no effect on function, or they may be functional, but the functional difference between the variants is usually subtle [171]. Osteoporosis has strong polygenic influences, and variance in bone mineral density (BMD) is estimated to be 50 – 80% heritable [339 – 341]. The potential for polymorphic variants affecting BMD and contributing substantially to this heritable risk was originally raised by Morrison et al. [342]. They identified several polymorphisms in the VDR gene and focused on three sites in the 3 region of the gene. The sites, denoted by the restriction enzyme that identified the variants by RFLP, were BsmI and ApaI located in the intron between exon, 8 and 9, and TaqI, located in exon 9 (see Fig. 11). A capital letter (e.g., B) denotes the absence of the restriction site, whereas a lowercase letter (e.g., b) denotes the presence of the restriction site in RFLP analysis. Morrison et al. [342] hypothesized that these VDR polymorphisms were associated with approximately a full standard deviation difference in BMD and

CHAPTER 9 Vitamin D

279

FIGURE 11 Polymorphisms in the human VDR chromosomal gene. The six 5 noncoding exons and 8 coding exons are depicted as shaded boxes. The location of the start codon polymorphism (SCP) is shown above exon 2 and the FokI polymorphism it generates is shown below the exon. The lowercase f is used to indicate the absence of the FokI restriction site and a 427 amino acid protein. The uppercase F is used to indicate the presence of the FokI site and a 424 amino acid protein. The location of the B, A, and T polymorphisms comprising the BsmI, ApaI, and TaqI restriction sites are shown. A variable length poly(A) microsatellite in exon 9 is also shown.

approximately 70% of the genetic variation in BMD. The possibility that a single gene, even one as central to bone metabolism as the VDR, contributed a major portion of the genetic basis of osteoporosis was an exciting but controversial hypothesis. However, multiple studies that followed in the wake of this paper either could not confirm the effect or found smaller differences in BMD [343 – 345]. A meta-analysis of 16 studies concluded that these polymorphisms contributed only a small effect on BMD, in the range of 1 – 2% [344]. Cooper and Umbach [344] summarized these findings by saying that the VDR polymorphisms represented one genetic factor affecting BMD, but further research into the mechanisms, clinical significance, and its relation between other genetic and environmental factors was needed. Although Howard et al. [346] eventually retracted some of their findings, their paper ignited great interest in the genetic basis of osteoporosis. Whereas the 3 polymorphisms did not change the structure or function of the VDR [343], interest soon developed in a polymorphism at the translation start site (ATG → ACG), which changed the amino-terminal end of the VDR [347]. The polymorphic variants detected by FokI RFLP cause translation to start at either the first methionine (denoted f or M1), generating a 427 amino acid protein, or a second methionine (denoted F or M4) three amino acids downstream of the first methionine generating a 424 amino acid protein. The F/M4 short VDR was found to have significantly more functional activity than the f/M1 long VDR [170]. Jurutka et al. [172] showed that the decrease in f/M1 activity was due to a decrease in binding of the transcription factor TFIIB. Several population studies have shown that f/M1 homozygotic individuals have decreased BMD [170, 347 – 350], but not all populations showed this effect [351]. Importantly, Ames et al. [352] showed that f/M1

polymorphism was associated with a decrease in BMD in healthy children aged 7 – 12. Moreover, ff homozygotes show a decrease in intestinal calcium absorption. FF homozygotes had a mean calcium absorption that was 41.5% greater than ff homozygotes and 17% greater than Ff heterozygotes. These results suggest a substantial relationship between FokI VDR polymorphism and bone metabolism, both at the level of intestinal calcium absorption and BMD. Other studies also emphasize the importance of the interaction of environmental factors, such as dietary calcium intake, with genetic factors, such as VDR polymorphisms [350]. It seems clear that osteoporosis is a polygenic disease and that polymorphisms in the VDR do not fully explain all of the heritable aspects. As discussed in Chapter 26, full genome scans and other approaches are actively being pursued to discover additional loci that contribute to the inheritance of traits that are associated with low BMD and increased fracture risk. It seems likely that an interaction of multiple polymorphisms and environmental factors will cumulatively determine osteoporotic risk. Interestingly, polymorphisms in the VDR considered to be associated with osteoporosis are also associated with the risk of developing prostate cancer [353]. Ingles et al. [354] and Taylor et al. [355] showed that polymorphisms at the 3 end of the gene, and a poly (A) track, 16 – 22 nucleotides in length, in the 3-untranslated region of the VDR mRNA, are associated with an approximate four fold increase in prostate cancer risk. However, as in the studies of osteoporotic risk, apparently not all populations show this association [356]. A connection between VDR polymorphisms and the risk of other malignancies, including breast and colon cancers, is underway. Initial studies in breast cancer suggest that VDR polymorphisms are associated with

280

FELDMAN, MALLOY, AND GROSS

breast cancer risk and progression [357 – 360]. These VDR polymorphisms have also been associated with a number of other conditions, including osteoarthritis and hyperparathyroidism [343], which are discussed further in Chapter 26.

VIII. 1,25(OH)2D3 ANALOGUES WITH DECREASED CALCEMIC ACTIVITY A. Agonists In addition to being a major regulator of calcium metabolism, 1,25(OH)2D exhibits many nonclassical actions in the body, including affecting cell growth, promoting cell differentiation, and suppressing the immune response (see Section IX). These properties make 1,25(OH)2D3 an attractive candidate for treating a number of serious diseases. However, to treat these diseases effectively, the dose of 1,25(OH)2D3 might well be in the range that would induce hypercalciuria, hypercalcemia, and renal stones; therefore, these unfavorable side effects limit its clinical utility. However, structural analogues of 1,25(OH)2D3 have been developed that exhibit a reduced calcemic response compared to 1,25(OH)2D3, yet retain many of the other therapeutically useful properties of the hormone, thus increasing their therapeutic potential [10,361 – 367]. Multiple analogues have been developed by the Roche company [368], the Leo Company [369], and the Chugai Co. [370], as well as by various investigators [135,371,372]. Although scores of analogues have been studied, the more effective analogues thus far have been the ones with sidechain modifications [371]. Changes in the 1,25(OH)2D3 molecule include the insertion of extra carbons, oxygen or

FIGURE 12

unsaturation in the carbon side chain, 16-ene derivatives, 19-nor derivatives, 20-epi derivatives, and 3-epi derivatives. The structures of a few of the more prominently studied analogues are depicted in Fig. 12. Many analogues have been shown to have a reduced calcemic response and a greater growth inhibitory potency compared to 1,25(OH)2D3. The mechanism for the differential activity displayed by the analogues is not totally clear but may be related to a number of properties: (a) decreased binding to DBP [373,374], (b) altered metabolic clearance and/or production of metabolites that retain significant biological activity [133,137,375], (c) increased ability to induce dimerization with RXR [376] or recruit coregulatory proteins [377], (d) increased ability to act preferentially to maintain an active conformation of the VDR within selected target tissues or on a limited number of target genes [378,379] and (e) ability to prevent degradation of the VDR [380]. There has also been progress in developing nonsteroidal molecules that mimic 1,25(OH)2D binding to the VDR [381,382]. This approach may make it possible to employ a much larger array of analogues and possibly lead to the concept of designer vitamin D drugs that exhibit specific target gene activation and thus have minimal side effects. The vitamin D analogues MC903, EB1089, and KH1060 have been developed by the Leo Pharmaceutical Company [369]. MC903, also known as calcipotriol, is less potent than 1,25(OH)2D3 in causing hypercalcemia, yet it is equivalent to 1,25(OH)2D3 in inhibiting the growth and in inducing differentiation of the human histiocytic lymphoma cell line U937. Calcipotriol has a low affinity for DBP and is metabolized rapidly to inactive metabolites when administered systemically. The drug is currently marketed for the topical treatment of psoriasis [295]. EB1089 exhibits approximately the same

Structure of 1,25(OH)2D3 and side chain modifications in four analogues with reduced calcemic activity.

281

CHAPTER 9 Vitamin D

affinity for the VDR as 1,25(OH)2D3, but is more potent than 1,25(OH)2D3 in inhibiting the growth of multiple cancer cell lines [291,383 – 386]. This drug is in clinical trials for the treatment of breast, pancreatic, and other cancers [387,388]. KH1060, also less calcemic than 1,25(OH)2D3, affects the immune function, including inhibiting IL-2-induced T lymphocyte differentiation [389]. This property may make KH1060 useful in treating autoimmune diseases, organ transplantation, and other conditions requiring immunosuppression. The vitamin D analogue 22-oxa-1,25(OH)2D3 (OCT), developed by Chugai Pharmaceuticals [370], has a lower affinity for VDR than 1,25(OH)2D3 but is 10 times more potent than 1,25(OH)2D3 in differentiating the myeloid leukemia cell line HL-60 and 100-fold less active in bone mobilization [390]. OCT, like 1,25(OH)2D3, also suppresses PTH production and is a potent inhibitor of renal 1-hydroxylase activity. OCT is being studied for use in chronic renal failure patients to inhibit excessive PTH secretion [370]. A number of analogues synthesized by the Roche company have shown to have increased antiproliferative activity and decreased calcemic activity [368]. A number of these analogues have been studied in various in vitro and in vivo models of cancer with promising results (see Section X) [391 – 397].

B. Antagonists Among the various naturally occurring metabolites of 1,25(OH)2D is (23S, 25R)-1-hydroxyvitamin D3-26,23-lactone. This molecule has some 1,25(OH)2D-like activities in dogs and humans, but when given to vitamin D-deficient rats in pharmacological doses, it led to hypocalcemia [398]. This finding suggested possible antagonist activities, and a series of analogues were studied to determine whether a vitamin D antagonist could be developed. The most promising candidate thus far is the analogue (23S)-25-dehydro-1-hydroxyvitamin D3-26,23-lactone (TEI-9647), which has been shown to have antagonist activity in several systems, including HL-60, SaOS-2, and MG-63 cells [398, 399]. The analogue binds to the VDR and appears to prevent heterodimer formation with RXR and subsequent recruitment of the coactivator SRC-1 [399]. TEI-9647 has a small amount of agonist activity, suggesting that it is a partial agonist/antagonist. The major action is as an antagonist, which may be clinically useful in selected hypercalcemic states.

IX. ACTIONS OF VITAMIN D IN CLASSICAL TARGET ORGANS TO REGULATE MINERAL HOMEOSTASIS The classical actions of 1,25(OH)2D on intestine, bone and kidney (see Fig. 7) include improved efficiency of in-

testinal calcium absorption, increased calcium mobilization from bone, and maintenance of adequate concentrations of calcium and phosphate in the extracellular fluid to promote the normal mineralization of bone [14,33,127]. In recent years, additional mechanisms by which 1,25(OH)2D modulates calcium homeostasis have been demonstrated, including autoregulation of 1,25(OH)2D synthesis as well as regulation of the calciotropic peptides PTH and calcitonin. These classical actions of 1,25(OH)2D to regulate calcium homeostasis are discussed in this section. The nonclassical, newly recognized actions of 1,25(OH)2D on many additional target cells, apparently unrelated to maintenance of systemic mineral homeostasis, are discussed in Section X.

A. 1,25(OH)2D Actions in Intestine 1. OVERVIEW OF CALCIUM ABSORPTION Calcium and phosphate are absorbed along the length of the small intestine. Calcium is mostly absorbed proximally in the duodenum, whereas vitamin D-dependent phosphate absorption occurs more distally in the jejunum and ileum [29]. VDR are present along the entire course of the small intestine, with the highest concentration proximally and the levels decreasing distally [141]. The abundance of VDR in the duodenum is the highest of all organs reported, and at any cross-sectional level along the intestine, VDR content is highest in crypts and decreases as the cells progress up the villus [400]. VDR are also present throughout the colon [401] and are expressed in colon cancer cell lines as well as in cancer specimens removed at surgery [218, 402 – 404]. Three mechanisms for intestinal calcium absorption have been described [29]. (i) In the absence of vitamin D, calcium is absorbed by a paracellular passive route; the rate of absorption is driven by mass action and is a function of the calcium concentration. (ii) In the presence of 1,25(OH)2D, a transcellular, saturable process is stimulated. This 1,25(OH)2D-dependent process takes some hours to develop and enhances calcium absorption greatly. (iii) Transcaltachia is the process of very rapid change in calcium flux that occurs within minutes in isolated perfused duodenum and is believed to be stimulated by vitamin D but via a nongenomic pathway [246]. Transcaltachia is described further in Section IV,B on nongenomic actions of vitamin D. Details of the process of calcium absorption are still not completely delineated at the molecular level [29]. In pioneering studies, Wasserman et al. [405] found that vitamin D induced calcium-binding proteins, now known as calbindins, in chick intestinal mucosa. 1,25(OH)2D induces the 9000 Mr vitamin D-dependent calbindin-D9k in mammalian intestine and also calbindin-D28k and calbindin-D9k in mammalian kidney [406,407]. In intestine, these

282 high-affinity calcium-binding proteins are located in the cytosol of columnar epithelial cells of the intestine and are involved in the translocation of calcium. However, their role is not completely defined and their induction alone cannot explain the process of calcium absorption, as calbindin-D levels are not correlated directly with the time course of 1,25(OH)2D-mediated calcium transport. Nevertheless, the calbindins have been an important tool for studying 1,25(OH)2D action on the intestine [29,406,407]. Two other proteins that are also regulated by vitamin D and probably play a role in calcium absorption are the integral membrane calcium-binding protein (IMCAL) [408] and Ca,Mg-dependent ATPase (Ca2 pump) located in the basolateral membrane of intestinal and renal cells, which uses ATP as an energy source to pump Ca2 out of the cell across a concentration gradient [409]. In addition to regulating these gene products, 1,25(OH)2D also plays a role in intestinal cell differentiation, elongating the villi, and inducing the polyamine pathway [410]. A combination of these and other effects coordinately expressed in the intestine mediate 1,25(OH)2D action to enhance calcium absorption. A model of transcellular intestinal calcium transport includes three sequential steps with calcium moving from protein to protein along an uphill gradient of calcium-binding affinities from apical to basolateral membrane [29]: (i) rapid entry of calcium at the brush border with transient sequestration subjacent to the microvillar membrane; (ii) transfer of calcium from the brush border complex to calbindin-D, which has a higher affinity for Ca2; and (iii) eventual transfer to the Ca/Mg-dependent ATPase or Ca2 pump, which has the highest affinity for Ca2 and extrusion against a concentration gradient at the basolateral membrane. This process may be similar in kidney where equivalent molecular species exist [257]. The overall contribution to calcium transport of the transcaltachia process [246], an endocytic pathway proposed to rapidly transport calcium via exocytosis of Ca2 containing vesicles by a nongenomic pathway, remains to be determined [29]. 2. ACTION OF VITAMIN D METABOLITES ON CALCIUM ABSORPTION Heaney and colleagues [255, 411, 412] have investigated the calcium absorptive response to graded doses of vitamin D3, 25(OH)D, and 1,25(OH)2D in healthy adult men. While no relationship was found between baseline absorption and serum vitamin D metabolite levels, all three vitamin D compounds significantly elevated 45Ca absorption from a 300-mg calcium load given as part of a standard test meal. 1,25(OH)2D was active even at the lowest dose (0.5 g/day), and the slope was such that doubling of absorption would occur at an oral dose of approximately 3 g/day. 25(OH)D was also active in elevating absorption and did so without raising total 1,25(OH)2D levels. On the

FELDMAN, MALLOY, AND GROSS

basis of the dose – response curves for 1,25(OH)2D and 25(OH)D, the two compounds exhibited a molar ratio for physiological potency of approximately 100:1. The absorptive effect of vitamin D3 was seen only at the highest dose level (1250 g, or 50,000 IU/day) and was apparently mediated by conversion to 25(OH)D. Analysis of pooled 25(OH)D data from both 25(OH)D- and vitamin D3-treated groups suggests that approximately one-eighth of circulating vitamin D-like absorptive activity under untreated conditions in winter may reside in 25(OH)D. This is a substantially larger share than has been predicted from studies of in vitro receptor binding [255,411,412]. 3. CHANGES IN CALCIUM ABSORPTION WITH AGE Osteoporosis is often associated with decreased intestinal calcium absorption with increasing age, and this phenomena is speculated to contribute to its pathogenesis [413, 414]. In rats there is an age-related decrease in the induction of calbindin protein in response to 1,25(OH)2D in the duodenum, but not in the ileum or kidney [415]. This decline in protein expression may be due to decreased translation of calbindin D9k mRNA in the duodenum with age. Several studies have suggested that intestinal VDR declines with age in the rat [213]. Duodenal biopsies of human subjects showed a slight trend toward a decrease of VDR abundance in the intestine with age [416]. However, the change in VDR abundance did not correlate with calcium absorption efficiency [417]. Estradiol may be an additional regulator of calcium absorption, as a direct effect of estradiol on intestinal calcium absorption independent of 1,25(OH)2D has been demonstrated [418]. 4. HYPERCALCIURIA Idiopathic hypercalciuria, the commonest form of renal stone disease, is characterized by the hyperabsorption of calcium, hypercalciuria, and normal or elevated 1,25(OH)2D concentrations [419]. Hypercalciuria in genetic hypercalciuric stone-forming (GHS) rats has been studied as a model for human intestinal calcium hyperabsorptive conditions [419,420]. GHS rats with normal serum 1,25(OH)2D levels are hyperabsorptive and have a greater number of VDRs than normal in the intestine, kidney, and bone. Administration of 1,25(OH)2D3 increases VDR gene expression significantly in GHS but not normocalciuric animals. Results suggest that GHS rats hyperrespond to minimal doses of 1,25(OH)2D3 by upregulating VDR gene expression. This unique characteristic suggests that GHS rats may be susceptible to minimal fluctuations in serum 1,25(OH)2D3, which may pathologically amplify the actions of 1,25(OH)2D3 on Ca2 metabolism that thus contribute to the hypercalciuria and stone formation [420]. Whether this mechanism also causes some form of human hypercalciuria and renal stones remains to be proven.

283

CHAPTER 9 Vitamin D

B. 1,25(OH)2D Actions in Bone Bone undergoes constant remodeling involving osteoclast-mediated bone resorption and osteoblast-mediated bone formation (see Chapter 12). 1,25(OH)2D is a major regulator of both formation and resorption. The detailed actions of 1,25(OH)2D on bone are discussed more completely in Chapter 2 on osteoblasts, Chapter 3 on osteoclasts, and Chapter 4 on matrix proteins and mineralization. Vitamin D is necessary for normal mineralization of the skeleton, and when it is deficient, a mineralization defect develops, causing rickets in growing children and osteomalacia in adults [421]. 1,25(OH)2D actions on bone are complex, and both direct and indirect effects have been described. Direct actions on the bone are complicated further because 1,25(OH)2D actions appear to be induced in several bone cell types, including osteoblasts, bone stromal cells, and osteoclasts. In addition, the nature of the response to 1,25(OH)2D is dependent on the differentiation state of the bone cell. VDRs are expressed in osteoblasts [422,423] and direct actions of 1,25(OH)2D3 on these cells include the modulation of cell growth [210] and stimulation of differentiation [424]. 1,25(OH)2D3 induces osteoblasts to progress from immature, proliferating cells to differentiated, nondividing cells that synthesize matrix proteins and mineralize bone. Several 1,25(OH)2D3 upregulated gene products have been identified, including osteocalcin, Eta-1 (osteopontin), alkaline phosphatase, matrix gla protein (osteocalcin), and IMCAL (Chapter 2). Collagen synthesis, however, is downregulated by 1,25(OH)2D3 in osteoblasts [425]. Although 1,25(OH)2D has been well known to promote bone mineralization since its discovery as an antirachitic agent many years ago, there is no definitive evidence that direct actions of 1,25(OH)2D on bone are required for normal bone mineralization. Effects to promote mineralization appear to be due mainly to 1,25(OH)2D actions on the intestine to enhance calcium and phosphate absorption to ensure optimal delivery of these ions to bone-forming cells. This concept of permissive action is supported by studies showing a restoration of normal bone mineralization in the absence of vitamin D action when adequate calcium and phosphorous are provided by intravenous infusion to vitamin D-deficient rats, VDR knockout mice, and children with HVDRR [7]. In the latter situation, chronically administered i.v. calcium infusions, which bypass the intestinal site of 1,25(OH)2D action, can achieve normalization of serum calcium levels, reversal of secondary hyperparathyroidism, and promote healing of the mineralization defect of rickets, despite the fact that 1,25(OH)2D action at the bone is prevented because of defective VDR [7]. These studies highlight the essential role of 1,25(OH)2D action on the intestine and indicate that the actions of the hormone on bone are indirect in regard to the process of mineralization.

There are nontheless many consequential effects of 1,25(OH)2D on bone, often in conjunction with PTH [424]. It has been known for many years that 1,25(OH)2D stimulates bone resorption [426]. This effect appears to be due to 1,25(OH)2D actions to directly stimulate the differentiation of precursor cells, mononuclear phagocytes of the macrophage lineage, to fuse into mature multinucleated osteoclasts [427]. This process, osteoclastogenesis, involves a complex interaction of osteoclast precursor cells, osteoblasts, and bone stromal cells. Together with other factors, 1,25(OH)2D promotes the early stages of osteoclastogenesis by direct actions on the osteoclast precursor cells. During the later stages of this differentiation process, the developing osteoclasts seem to lose their VDR, and 1,25(OH)2D stimulation of differentiation becomes indirect by acting on cells in the osteoblast lineage, possibly osteoblast stromal cells, to induce osteoclast differentiating factor(s) (see Chapter 3). Studies reveal that osteoclastogenesis is regulated by osteoclast differentiation factor (ODF) or RANKL (RANK ligand), an osteoclastogenic factor of osteoblastic origin, and osteoprotegerin (OPG), a potent inhibitor of osteoclastogenic activity. These factors are discussed extensively in Chapters 3, 6, and 12–14. RANKL is a member of the tumor necrosis factor (TNF) ligand family and OPG is a member of the tumor necrosis factor receptor family. Overexpression of OPG in transgenic mice leads to osteopetrosis, whereas OPG knockout mice develop severe osteoporosis. The ratio of RANKL to OPG determines the level of osteoclastogenic activity, and 1,25(OH)2D3 has been shown to regulate both of these factors. However, like PTH and IL-2, 1,25(OH)2D3 increases the ratio of RANKL:OPG, leading to increased osteoclastogenesis and bone resorption [428].

C. 1,25(OH)2D Actions in Kidney VDR are present in kidney [429], and renal calbindinD28k [406] and calmodulin [430], among other proteins, are regulated by 1,25(OH)2D [431]. However, the most important renal actions of 1,25(OH)2D are probably the regulation of the 1- and 24-hydroxylases (see Fig. 2). 1,25(OH)2D has a short and a long loop feedback to regulate its own production (see Fig. 7). In the presence of adequate 1,25(OH)2D levels, the short loop feedback is a direct renal action of 1,25(OH)2D to inhibit 1-hydroxylase and to induce 24-hydroxylase gene expression. The two actions coordinately drive 25(OH)D into 24,25(OH)2D, an inactivation pathway, and inhibit further 1,25(OH)2D synthesis. The long loop feedback is via 1,25(OH)2D inhibition of PTH gene expression, as PTH is the major stimulator of 1-hydroxylase activity. The 1,25(OH)2D action on PTH is also mediated indirectly via 1,25(OH)2D regulation of

284

FELDMAN, MALLOY, AND GROSS

the serum Ca2 concentration, which will rise subsequent to the calcemic actions of 1,25(OH)2D on intestine and bone (see Fig. 7). 1,25(OH)2D has been implicated in regulating renal calcium and phosphate excretion; however, its role here is not as well defined. The components of the intestinal calcium transport system are also present in the kidney, including VDR and 1,25(OH)2D-dependent calbindin-D28k, as well as calbindin-D9k and a membrane calcium pump, all located within the same cells of the distal convoluted tubule. It is likely that 1,25(OH)2D stimulates calcium transport in the distal tubule via a calbindin D-dependent mechanism similar to the intestine [431]. In chronic renal failure, as the mass of functional renal tissue declines, the production of 1,25(OH)2D diminishes with resultant vitamin D insufficiency and secondary hyperparathyroidism. Coexisting hyperphosphatemia leads to the development of renal osteodystrophy [87,256]. In addition to preventing hyperphosphatemia, replacement of 1,25(OH)2D has become a cornerstone of managing patients with this syndrome. Initially, oral 1,25(OH)2D3 and then intravenous 1,25(OH)2D3 were used. Currently, newer less calcemic analogues are being studied for possible improved results, including 22-Oxa-1,25(OH)2D3 (22-oxacalcitriol or OCT), 19-nor-1,25(OH)2D2 (19-norD2), and 1(OH)D2 [363]. These analogues have been tested in animal models of uremia and in clinical trials [80]. Intravenous 19-norD2 and oral 1(OH)D2 have been approved for use in the United States; OCT is currently under review. The mechanisms by which these analogues exert their selective actions on the parathyroid glands, to suppress secondary hyperparathyroidism without causing hypercalcemia, are still under investigation.

D. 1,25(OH)2D Action on the Parathyroid Glands and Regulation of PTH The parathyroid glands possess VDR and are an important component of the systemic regulation of calcium homeostasis by 1,25(OH)2D [80,432]. The major effect of 1,25(OH)2D in this site is to suppress PTH secretion by inhibiting mRNA and protein synthesis. The other major regulator of PTH secretion is serum Ca2, which acts via the calcium sensing receptor in the parathyroid glands [94]. It has been suggested that the weight of parathyroid adenomas is related to vitamin D nutrition, indicating the importance of the feedback of vitamin D to inhibit parathyoid growth [245]. Patients with chronic renal failure develop secondary hyperparathyroidism, partly due to the decreased renal production of 1,25(OH)2D by the diseased kidneys. In addition, inappropriately elevated PTH secretion may result from decreased levels of VDR in the parathyroid gland of

uremic patients, resulting in the less efficient suppression of PTH synthesis by 1,25(OH)2D [242]. Studies indicate that the decrease in VDR is not distributed uniformly in parathyroid glands from chronic renal failure patients and that selected areas of low VDR content exhibit the most severe hyperplasia [433]. Suppression of elevated PTH in secondary hyperparathyroidism of chronic renal failure may be accomplished by the administration of 1,25(OH)2D3 or its analogues as described earlier. Better PTH suppression with less hypercalcemia is achieved with intermittent i.v. administration of 1,25(OH)2D3, which results from higher peak serum levels that are achieved with this regimen. The use of vitamin D analogues that elicit a reduced calcemic response, especially i.v. 19-norD2, oral 1(OH)D2 and oral OCT, may in the future yield improved PTH suppression without hypercalcemia and provide a more effective treatment for secondary or tertiary hyperparathyroidism in chronic renal failure [363].

E. Regulation of PTHrP and Calcitonin 1,25(OH)2D3 inhibits PTHrP expression in many normal tissues as well as malignant cells [434] but not all tissues (e.g., prostate), [435]. This may add to the beneficial effects of 1,25(OH)2D3 in the treatment of cancer with metastases to bone and/or in humeral hypercalcemia of malignancy. The less calcemic analogue of 1,25(OH)2D3, EB1089, was shown to adequately suppress PTHrP production by a squamous cell cancer xenografted into mice and reverse the hypercalcemic state caused by excess PTHrP [436]. Although use of a vitamin D preparation in a hypercalcemic state might at first appear counterintuitive, the less calcemic analogues may have a role in suppressing pathologic levels of PTHrP in humoral hypercalcemia of malignancy. Calcitonin is another calciotropic peptide hormone regulated by 1,25(OH)2D3 [437]. Inhibition of mRNA and protein expression has been demonstrated in vivo in rat and in vitro in medullary thyroid cancer cells. These issues are discussed in more detail in Chapters 7 and 8.

X. ACTIONS OF 1,25(OH)2D IN NONCLASSICAL TARGET ORGANS In recent years a number of additional actions of 1,25(OH)2D beyond merely regulating mineral homeostasis have been discovered in numerous nonclassical target organs. Many of these actions involve the promotion of cellular differentiation and inhibition of cell growth and appear to be unrelated to the regulation of total body calcium metabolism. VDR expression and 1,25(OH)2D effects have been demonstrated in a variety of cells and tissues, including hematopoietic, immunologic, epidermal, and cancer

285

CHAPTER 9 Vitamin D

cell systems. These diverse actions of 1,25(OH)2D and its analogues have been the subject of several reviews [152,365,434,438 – 440].

A. 1,25(OH)2D Effects on Cell Growth and Differentiation in Normal and Malignant Tissues VDR are expressed in many normal and malignant cell types, indicating a wide array of previously unrecognized potential targets for 1,25(OH)2D action [152]. In many of these nonclassical tissues, 1,25(OH)2D acts to inhibit cell proliferation [152,439], although in selected settings, 1,25(OH)2D may stimulate cell proliferation [441,442]. Also, in a number of systems, 1,25(OH)2D affects cells by promoting cellular differentiation [152,434,438,439]. The hormonal modulation of intracellular calcium and the regulation of expression of nuclear oncogenes [439,443] have been raised as possible mechanisms for these antiproliferative and differentiating effects of 1,25(OH)2D. Additional mechanisms are postulated to include inhibition of the passage of cells through the cell cycle, regulation of paracrine growth factors, stimulation of terminal differentiation, and induction of apoptosis [397,444 – 447]. The induction of p21 and p27, cell cycle-dependent kinase (Cdk) inhibitors, has emerged as a major mechanism for cell cycle arrest [386,448], whereas the downregulation of Bcl-2 and perhaps other antiapoptotic factors play a major role in the induction of apoptosis [444,447].

B. Possible Role of Vitamin D in Cancer Prevention or Therapy Because of its actions to inhibit cellular proliferation and stimulate cellular differentiation, 1,25(OH)2D has been considered for possible “chemoprevention” or “differentiation” therapy in a number of malignant cell types that possess VDR [434]. 1. COLON CANCER VDR are present in the colon [401] and in colon cancer cell lines, as well as in surgically removed colon cancers [449]. The possibility that calcium and/or vitamin D may be active in decreasing colon cancer has been examined by several groups [446,450]. Garland et al. [451] have suggested that higher levels of 25(OH)D protect against colon cancer. Eisman and coworkers showed that 1,25(OH)2D3 administration could inhibit the growth of colon cancer xenografts in nude mice [452]. A number of colon cancer models have been shown to be inhibited and/or differentiated by 1,25(OH)2D or its analogues, both in vitro and in vivo [218, 404, 446, 453].

2. BREAST CANCER VDR are present in normal breast and breast cancer cell lines and in many human cancer specimens [391,444]. 1,25(OH)2D3 was shown to suppress the growth of breast cancer cell lines, xenografts in nude mice, and carcinogen (nitrosomethyl urea, NMU)-induced breast cancer in rats [391]. Also, EB1089 (see Fig. 12), an analogue with reduced calcemic potency, has been shown to inhibit the proliferation of MCF-7 human breast cancer cells in vitro. In in vivo studies, EB1089 was found to be more potent than 1,25(OH)2D3 at inhibiting tumor growth induced by the carcinogen NMU and less likely to induce hypercalcemia. It therefore has been considered to have therapeutic potential as an antitumor agent. Clinical trials are currently underway to determine whether this antiproliferative action of vitamin D analogues on breast cancer will be effective in humans [387]. In a separate breast cancer prevention trial in rats, also using the NMU-induced breast cancer model, Anzano et al. [454] tested the effectiveness of the vitamin D analogue 1,25(OH)2D-16-ene-23-yne-26,27-hexafluoro. After 5 – 7 months of therapy they found that the compound extended tumor latency, lessened tumor incidence, enhanced the tamoxifen effect to reduce tumor burden, and increased the number of tumor-free rats. Also, the rats did not develop hypercalcemia as a result of treatment with this analogue. A number of investigators have shown that 1,25(OH)2D or its analogues are antiproliferative in breast cancer cell models and involve a number of different pathways [444,455 – 459]. In addition to its antiproliferative effects, evidence shows that 1,25(OH)2D stimulates apoptosis in some breast cancer cells [394]. 3. PROSTATE CANCER On the basis of geographic patterns of ultraviolet radiation throughout the contiguous United States and epidemiological data on prostate cancer incidence, a hypothesis was raised by Schwartz and colleagues that vitamin D deficiency may be a risk factor [460] and that increased sunlight exposure may protect against clinical prostate cancer [461]. In a prediagnostic study with stored sera, 1,25(OH)2D blood levels were found to be an important predictor for palpable and anaplastic tumors in men over 57 years of age but not for incidentally discovered or well-differentiated tumors [462]. VDR are present in prostate cancer cell lines [288,463] and in normal prostate [289], and 1,25(OH)2D3 has been shown to inhibit the growth of all these cell types in culture [288,289,392]. Furthermore, the prostate growth-inhibiting activity of selected vitamin D analogues with reduced calcemic potency (e.g., EB1089 and OCT) was even greater than 1,25(OH)2D3 [291], raising the possibility of the therapeutic potential of these drugs in the treatment of prostate cancer. A number of studies have demonstrated an antiproliferative activity of

286 1,25(OH)2D3 and vitamin D analogues in multiple prostate cancer models [292,393,395,397,447]. The induction of apoptosis may play some role in the antiproliferative activity in some prostate cancer cells [464]. Clinical trials have begun to address the utility of 1,25(OH)2D3 in treating prostate cancer patients [465,466]. 4. HEMATOPOIETIC CELLS: MYELOID CELLS AND LEUKEMIA In addition to promoting osteoclastogenesis from macrophage precursors described earlier in the section on bone (Section IXB), 1,25(OH)2D3 has been shown to stimulate a variety of immature hematopoietic myeloid cells to differentiate into mature cells, including M-1 mouse myeloid leukemic cells, HL-60 human promyelocytic leukemia cells, U-937 human monocytic cells, and peripheral human monocytes [445]. In these cells, VDR were present and the differentiation process was accompanied by the inhibition of cell proliferation. In HL-60 cells, the 1,25(OH)2D3-induced response appears to be due to the induction of terminal differentiation mediated by the inhibition of expression of the c-myc oncogene [467]. Liu et al. [448] have shown that 1,25(OH)2D stimulates myeloid leukemic cells lines to terminally differentiate into monocytes/macrophages. Using the myelomonocytic U937 cell line, they showed that 1,25(OH)2D induces the expression of the Cdk inhibitor p21 (WAF1/CIP1), which caused the cells to terminally differentiate [448]. These effects on leukemic cells in vitro as well as prolongation of survival time of mice inoculated with myeloid leukemia cells [468] have led to the consideration of using 1,25(OH)2D3 therapeutically in human leukemia as “differentiation” therapy [445]. Munker et al. [469] reviewed the potential use of 1,25(OH)2D and analogues to treat leukemia. Their work showed that 1,25(OH)2D differentiated both normal and leukemic cells. They suggested that 1,25(OH)2D alone, or in combination with retinoids or chemotherapeutic agents, would be useful in the treatment of patients and warrants clinical trials in patients with leukemia.

C. Immune System: 1,25(OH)2D Actions on Immunosuppression and Cytokines In addition to 1,25(OH)2D3 effects on myeloid cells described earlier and on monocytic/macrophage precursors that are differentiated into osteoclasts (described in Section IXB), 1,25(OH)2D3 has many important immunomodulatory effects [365,389,440,470]. Circulating resting T and B cells do not express VDR but when blast transformed or mitogen activated these cells do express VDR and respond to 1,25(OH)2D3 [439]. 1,25(OH)2D3 treatment of

FELDMAN, MALLOY, AND GROSS

VDR-positive cells results in growth inhibition and decreased IL-2 production. Monocytes and macrophages also express VDR. 1,25(OH)2D3 modulates the production of numerous interleukins, cytokines, and various oncogenes and transcription factors, including interleukins 1, 2, 3, and 6; interferon-; leukemia inhibitory factor (LIF); tumor necrosis factor (TNF); and granulocyte/macrophage colony-stimulating factor (GM-CSF), as well as several oncogenes, including c-myc, fos, myb, fms, and fgr [152, 438,439]. 1,25(OH)2D3 also suppresses the production of helper T cells. Although the actions of 1,25(OH)2D3 are predominantly inhibitory, some genes may be stimulated, depending on the activation state of the immune cells. In various animal models, 1,25(OH)2D3 reduces immune responses when administered prior to induction or early in the disease process [389,470]. 1,25(OH)2D3 inhibits the induction of several autoimmune diseases in animals, such as experimental autoimmune encephalomyelitis, lupus, thyroiditis, and type I diabetes. 1,25(OH)2-24-oxo-16ene-D3, a natural metabolite of the vitamin D analogue 1,25(OH)216ene-D3, can initiate immunosuppressive effects equal to the parent compound without causing hypercalcemia in vivo [471,472]. The significance of the immunomodulating properties of 1,25(OH)2D3 remains poorly understood; however, the possible applications to the clinical setting of leukemia, autoimmune disease, and transplantation are currently being explored. As described in Section VI,C, activated macrophages can synthesize 1,25(OH)2D from circulating 25(OH)D, as has been shown in sarcoid, tuberculosis, and other granulomatous diseases. Interferon- has been shown to stimulate the local production of 1,25(OH)2D in these cells [261]. 1,25(OH)2D inhibits lymphocyte proliferation, interleukin 2, and interferon- production, as well as other cytokines (IL-12) and immunoglobulins [438]. The local production of 1,25(OH)2D may have an autocrine/paracrine function, which may act locally to suppress the inflammatory response, particularly leading to the inhibition of helper T cell subset type 1 (Th1) [389]. The various findings suggest the possibility of therapeutic application of 1,25(OH)2D3 and analogues in autoimmunity and transplantation [473]. 1,25(OH)2D immunosuppressive activity has been well studied in the autoimmune model of diabetes that develops spontaneously in nonobese diabetic (NOD) mice [362]. Type I diabetes can be prevented without generalized immunosuppression by nonhypercalcemic analogues of 1,25(OH)2D when treatment is started early, i.e., before the autoimmune attack, reflected by insulitis. Even if the autoimmune disease is already active, treatment with 1,25(OH)2D analogues can prevent clinical diabetes when this therapy is combined with a short induction course of an immunosuppressant such as cyclosporin A.

287

CHAPTER 9 Vitamin D

D. 1,25(OH)2D Effects on Skin: Use in the Treatment of Psoriasis In addition to being the site of initiation of vitamin D synthesis, skin is also a 1,25(OH)2D target organ [35]. Human dermal fibroblasts and keratinocytes possess VDR and are 1,25(OH)2D3 responsive [474]. For this reason, cultured dermal fibroblasts are used frequently to study HVDRR [7,309, 310]. 1,25(OH)2D3 inhibits proliferation and stimulates terminal differentiation of keratinocytes, including stimulation of involucrin, cornified envelop development, and transglutaminase activity [35]. Other cells within the skin also contain VDR and appear to be 1,25(OH)2D3 targets as well. Melanoma cells express VDR, and 1,25(OH)2D3 induces differentiation and inhibits cell proliferation [475]. VDR are also present in the hair follicle within the skin [151]. It is of interest that children with HVDRR [476], as well as VDRablated mice [331,332], have alopecia; however, the mechanism is unknown. When HVDRR patients respond to treatment with calcium infusion and the bone mineralization defect of rickets is reversed, no improvement in alopecia is noted either in HVDRR patients [7] or in VDR-ablated mice [333]. Evidence from studies in VDR-ablated mice suggests that skin differentiation is not dependent on 1,25(OH)2D3 acting through its nuclear receptor, and the authors speculate that the associated alopecia in these animals may be due to nongenomically mediated vitamin D toxicity from elevated 1,25(OH)2D3 concentrations [477]. Psoriasis is a hyperproliferative disorder of the epidermis, which responds to treatment with vitamin D preparations applied topically or administered systemically [295,478]. The antipsoriatic effect may be due to the antiproliferative action of the hormone, but may also involve immunosuppressive properties induced by 1,25(OH)2D3 [296]. Newer vitamin D analogues with reduced calcemic activity are being developed to improve the therapeutic potential of treating this disease. Interestingly, in keratinocytes, the VDR levels are downregulated within a few hours after UV B irradiation [36]. These results strongly suggest the existence of a feedback mechanisms in that UV B initiates vitamin D synthesis in keratinocytes and at the same time limits VDR abundance. These findings provide an explanation for the reported lack of any additive effect between 1,25(OH)2D and UV B phototherapy in the treatment of psoriasis.

E. 1,25(OH)2D Action in the Nervous System: NGF, Alzheimer’s Disease, and Aging The first evidence for the presence of VDR in brain came from autoradiographic studies using [3H]1,25(OH)2D3

to localize the receptor [479]. In rodents, [3H]1,25(OH)2D3 binding sites were located throughout the brain from the basal forebrain to the midbrain and hindbrain [479, 480]. VDR were detected biochemically in the hippocampus by showing VDR mRNA expression using in situ hybridization [481]. Very little is known about 1,25(OH)2D actions in the central nervous system [482]. Calbindin-D28k in the brain is not vitamin D dependent; however, 1,25(OH)2D3 was found to stimulate choline acetyl transferase activity in the bed nucleus of the stria terminalis [483]. Furthermore, nerve growth factor (NGF) mRNA levels were stimulated by 1,25(OH)2D3 in mouse L929 fibroblasts, an in vitro model of nerve cell function [484,485], and studies demonstrated that 1,25(OH)2D3 induced NGF mRNA levels in the hippocampus and cortex [486]. In the intact organism, 1,25(OH)2D3 treatment results in improved memory performance of young adult rats in the Morris water maze test [487]. Interestingly, VDR mRNA expression was found to be decreased in the hippocampus of patients with Alzheimer’s disease [481]. A possible role of decreased 1,25(OH)2D or VDR with aging, leading to decreased NGF production in the brain, has raised conjecture about a possible role of decreased vitamin D action in the neurodegeneration found with aging or Alzheimer’s disease [482]. VDR levels have been shown to decrease with aging in the intestine [416], and although a connection to the brain is highly speculative at this time, some role for 1,25(OH)2D in the central nervous system seems clear. Alzheimer’s disease (AD) patients are susceptible to hypovitaminosis D due to their being elderly and confined to a hospital. A study of 46 ambulatory elderly women with AD showed that 26% had decreased 25(OH)D (5 – 10 ng/ml) and 54% had osteomalacic levels ( 5 ng/ml) [488]. Those with decreased vitamin D had increased PTH and decreased BMD. Many AD patients were sunlight deprived and consumed less than 100 IU of vitamin D per day. Vitamin D deficiency due to sunlight deprivation and malnutrition, together with compensatory hyperparathyroidism, contributes significantly to reduced BMD and increased risk of hip fractures in patients with AD [488].

F. 1,25(OH)2D Action on the Reproductive System The role of 1,25(OH)2D in reproduction has been examined in chickens and rats where 1,25(OH)2D appears to play a role in normal ovulation, fetal and neonatal bone development, milk production, and maintenance of normocalcemia and mineral homeostasis in the neonate [489]. Extrarenal synthesis of 1,25(OH)2D takes place in the placenta, which also expresses VDR. In addition, 1,25(OH)2D stimulates human placental lactogen (hPL) expression from

288 trophoblast cells, and a VDRE has been demonstrated in the 5 upstream region of the hPL gene, supporting a role for 1,25(OH)D in placental function [490]. VDR and vitamin D-dependent Ca2-binding protein are found in a number of additional tissues, including testis, uterus, pancreas, pituitary, thyroid, gonads, and muscle, including the heart [152,438,491], but the functional role of 1,25(OH)2D in these sites is unclear and will require further investigation. In VDR-ablated mice, uterine hypoplasia and ovarian abnormalities were detected in females and testicular defects and sperm abnormalities in males [331]. However, many of these defects were improved after calcium nutrition was normalized [334]. However, in females the estradiol levels were still somewhat reduced and gonadotropin levels somewhat elevated, suggesting a residual defect unrelated to calcium. These parameters normalized with administered estradiol. Because the aromatase gene is regulated by vitamin D [335], an effect on estradiol synthesis may affect fertility in VDR-ablated mice and HVDRR subjects [334].

XI. VITAMIN D AND OSTEOPOROSIS The importance of vitamin D in the etiology and treatment of osteoporosis will be discussed in detail in a number of subsequent chapters in this book, especially Chapters 40 and 68. The use of vitamin D and its analogues to prevent and treat ostoporosis has been reviewed [366,492,493]. In brief, several potential mechanisms have been put forward to implicate vitamin D in the development of osteoporosis. (i) The possibility that polymorphisms within the gene encoding the VDR contribute substantially to genetic differences in osteoporosis risk has been raised by Morrison et al. [342]. The basis for this genetic effect on osteoporosis risk is presumably as a hereditary factor affecting “peak bone mass,” but the mechanism is unknown. At this time the VDR genotype hypothesis remains controversial, as other groups, using subjects of different ethnic background, have not found a similar correlation between the different VDR alleles and bone density [344,345] (see Chapter 26). (ii) An age-related decline in renal 1,25(OH)2D production, due in part to a diminished renal response to PTH and reduced intestinal calcium absorption [414]. There appears to be a defect in the renal response to PTH so that older women with osteoporosis require greater amounts of PTH to stimulate 1,25(OH)2D production. (iii) A relative decrease in circulating 1,25(OH)2D has been considered a contributing factor in the development of senile osteoporosis [494]. A low vitamin D state from an inadequate diet and decreased exposure to sunlight as people age, especially in the house-bound elderly, contribute to malabsorption of calcium and vitamin D “insufficiency” in the elderly [17]. Other studies concur that there is a high prevalence of vitamin D insufficiency in the elderly, even in

FELDMAN, MALLOY, AND GROSS

an active community. These individuals may have established vertebral osteoporosis with increased bone turnover, decreased BMD at the hip, and thus an enhanced risk of further osteoporotic fractures in comparison with vitamin D-sufficient subjects [495]. (iv) An age-related decline in intestinal VDR creating a relative 1,25(OH)2D-resistant state and impairing intestinal calcium absorption [416]. All of these factors coordinately contribute to age-related bone loss, which, according to some studies, can be ameliorated by vitamin D and calcium supplements, [496 – 498]. It is interesting to note that there is a rapid recovery of BMD following the resolution of vitamin D insufficiency [499]. In a paper discussing vitamin D and osteoporosis, Lau and Baylink [500] argue that the type of vitamin D deficiency that exists determines the response to vitamin D or 1,25(OH)2D3 and its analogues. Vitamin D deficiency may be (1) primary vitamin D deficiency, which is due to a deficiency of vitamin D, the parent compound; (2) a deficiency of 1,25(OH)2D3 resulting from decreased renal production; and (3) resistance to 1,25(OH)2D3 action at the target tissues, which could be related to decreased VDR levels in the intestine with age. Each type of deficiency has been implicated as a potential cause of intestinal calcium malabsorption, secondary hyperparathyroidism, and senile osteoporosis. Primary vitamin D deficiency can be corrected by vitamin supplements, whereas 1,25(OH)2D3 deficiency or resistance may require 1,25(OH)2D3 or an analogue to correct the high serum PTH and the calcium malabsorption. In addition, some elderly patients have decreased intestinal Ca absorption, but with apparently normal vitamin D metabolism. Although the cause is unclear, these patients, as well as other patients with secondary hyperparathyroidism (not due to decreased renal function), show a decrease in serum PTH and an increase in calcium absorption in response to therapy with 1,25(OH)2D3 or an analogue. Some form of vitamin D therapy—vitamin D, 1,25(OH)2D3, or an analogue—can be used to correct all types of age-dependent impairments in intestinal calcium absorption and secondary hyperparathyroidism during aging. With respect to postmenopausal osteoporosis, there is strong evidence that 1,25(OH)2D3 or its analogues may have bone-sparing actions. However, these effects appear to be the result of pharmacologic actions on bone formation and resorption rather than through replenishing a deficiency [500]. Vitamin D has direct actions to affect estrogen synthesis by regulating the activity of aromatase in osteoblasts [335] and estrogen half-life by regulaing 17-hydroxysteroid dehydrogenase in keratinocytes [501]. The impact of these effects on multiple organs and their potential role in modulating vitamin D actions remain to be fully clarified. Vitamin D has effects on muscle, and findings suggest that vitamin D insufficiency may be associated with decreased muscle strength [502] and therefore increased rates of falling [503]. In ambulatory nursing home and hostel

289

CHAPTER 9 Vitamin D

residents, individuals who fall have lower serum 25hydroxyvitamin D and higher serum parathyroid hormone concentrations than other residents. The association between falling and serum PTH persists after adjustment for other variables [503]. An increased rate of falling can contribute to increased fractures in vitamin D-insufficient individuals. In a special case of glucocorticoid-induced osteoporosis, vitamin D plus calcium is superior to no therapy or calcium alone in the management of patients according to a recent meta-analysis. However, vitamin D is less effective than some osteoporosis therapies, such as bisphosphonates (see Chapter 44). Therefore, treatment with vitamin D plus calcium, as a minimum, should be recommended to patients receiving long-term corticosteroids [504]. Although most studies address the common problem of insufficient vitamin D, there are also difficulties with overdoses of vitamin D. Hypervitaminosis D may result from drinking milk that is incorrectly fortified with vitamin D [505] or from excessive dietary supplements [506]. After resolution of occult vitamin D intoxication in patients using dietary supplements that contained unadvertised high levels of vitamin D, resolution of vitamin D intoxication was associated with a rebound in bone mineral density [506]. The relationship of vitamin D to osteoporosis is discussed completely in Chapter 68.

References 1. S. Christakos, M. Raval-Pandya, R. P. Wernyj, and W. Yang, Genomic mechanisms involved in the pleiotropic actions of 1,25-dihydroxyvitamin D3. Biochem. J. 316, 361 – 371 (1996). 2. M. R. Haussler, G. K. Whitfield, C. A. Haussler, J. Hsieh, P. D. Thompson, S. H. Selznick, C. E. Dominguez, and P. W. Jurutka, The nuclear vitamin D receptor: Biological and molecular regulatory properties revealed. J. Bone Min. Res. 13, 325 – 349 (1998). 3. L. L. Issa, G. M. Leong, and J. A. Eisman, Molecular mechanism of vitamin D receptor action. Inflamm. Res. 47, 451 – 475. 4. G. Jones, S. A. Strugnell, and H. F. DeLuca, Current understanding of the molecular actions of vitamin D. Physiol. Rev. 78, 1193 – 1231 (1998). 5. A. W. Norman, Receptors for 1,25(OH)2D3: Past, present, and future. J. Bone Miner. Res. 13, 1360 – 1369 (1998). 6. A. J. Brown, A. Dusso, and E. Slatopolsky, Vitamin D. Am. J. Physiol. 277, F157 – F175 (1999). 7. P. J. Malloy, J. W. Pike, and D. Feldman, The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocr. Rev. 20, 156 – 188 (1999). 8. N. J. McKenna, R. B. Lanz, and B. W. O’Malley, Nuclear receptor coregulators: Cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 9. C. Rachez and L. P. Freedman, Mechanisms of gene regulation by vitamin D3 receptor: A network of coactivator interactions. Gene 246, 9 – 21 (2000). 10. D. Feldman, F. H. Glorieux, and J. W. Pike, (eds.) (1997). “Vitamin D.” Academic Press, San Diego, 1997. 11. R. L. Horst and T. A. Reinhardt, Vitamin D metabolism. In “Vitamin D” (D. Feldman, F. Glorieus, and J. W. Pike, eds.), pp. 13 – 31. Academic Press, San Diego 1997.

12. M. F. Holick, McCollum Award Lecture, 1994: Vitamin D — new horizons for the 21st century. Am. J. Clin. Nutr. 60, 619 – 630 (1994). 13. E. Kodicek, The story of vitamin D, from vitamin to hormone. Lancet 1, 325 – 329 (1974). 14. H. F. DeLuca, The vitamin D story: A collaborative effort of basic science and clinical medicine. FASEB J. 2, 224 – 236 (1988). 15. H. F. DeLuca, Historical overview. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 3 – 11. Academic Press, San Diego, 1997. 16. M. F. Holick, Q. Shao, W. W. Liu, and T. C. Chen, The vitamin D content of fortified milk and infant formula. N. Engl. J. Med. 326, 1178 – 1181 (1992). 17. M. C. Chapuy, P. Preziosi, M. Maamer, S. Arnaud, P. Galan, S. Hercberg, and P. J. Meunier, Prevalence of vitamin D insufficiency in an adult normal population. Osteoporos. Int. 7, 439 – 443 (1997). 18. M. F. Holick, Vitamin D requirements for humans of all ages: New increased requirements for women and men 50 years and older. Osteoporos. Int. 8 (Suppl. 2), S24 – S29 (1998). 19. H. Glerup, K. Mikkelsen, L. Poulsen, E. Hass, S. Overbeck, J. Thomsen, P. Charles, and E. F. Eriksen, Commonly recommended daily intake of vitamin D is not sufficient if sunlight exposure is limited. J. Intern. Med. 247, 260 – 268 (2000). 20. M. C. Chapuy, A. M. Schott, P. Garnero, D. Hans, P. D. Delmas, and P. J. Meunier, Healthy elderly French women living at home have secondary hyperparathyroidism and high bone turnover in winter: EPIDOS Study Group. J. Clin. Endocrinol. Metab. 81, 1129 – 1133 (1996). 21. F. M. Gloth, III, C. M. Gundberg, B. W. Hollis, J. G. Haddad, Jr., and J. D. Tobin, Vitamin D deficiency in homebound elderly persons. JAMA 274, 1683 – 1686 (1995). 22. M. C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994). 23. R. A. Bouillon, J. H. Auwerx, W. D. Lissens, and W. K. Pelemans, Vitamin D status in the elderly: Seasonal substrate deficiency causes 1,25-dihydroxycholecalciferol deficiency. Am. J. Clin. Nutr. 45, 755 – 763 (1987). 24. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, S. Harris, L. J. Sokoll, and G. Falconer, Effect of vitamin D supplementation on wintertime and overall bone loss in healthy postmenopausal women. Ann. Intern. Med. 115, 505 – 512 (1991). 25. M. J. McKenna, Differences in vitamin D status between countries in young adults and the elderly. Am. J. Med. 93, 69 – 77 (1992). 26. M. K. Thomas, D. M. Lloyd-Jones, R. I. Thadhani, A. C. Shaw, D. J. Deraska, B. T. Kitch, E. C. Vamvakas, I. M. Dick, R. L. Prince, and J. S. Finkelstein, Hypovitaminosis D in medical inpatients. N. Engl. J. Med. 338, 777 – 783 (1998). 27. M. S. LeBoff, L. Kohlmeier, S. Hurwitz, J. Franklin, J. Wright, and J. Glowacki, Occult vitamin D deficiency in postmenopausal US women with acute hip fracture. JAMA 281, 1505 – 1511 (1999). 28. M. C. Chapuy and P. J. Meunier, Vitamin D insufficiency in adults and the elderly. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 679 – 693. Academic Press, San Diego, 1997. 29. R. H. Wasserman, Vitamin D and the intestinal absorption of calcium and phosphorus. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 259 – 273. Academic Press, San Diego, 1997. 30. E. B. Mawer and M. Davies, Bone disorders associated with gastrointestinal and hepatobiliary disease. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 831 – 847. Academic Press, San Diego, 1997. 31. C. W. Lo, P. W. Paris, T. L. Clemens, J. Nolan, and M. F. Holick, Vitamin D absorption in healthy subjects and in patients with intestinal malabsorption syndromes. Am. J. Clin. Nutr. 42, 644 – 649 (1985). 32. T. D. Thacher, P. R. Fischer, J. M. Pettifor, J. O. Lawson, C. O. Isichei, J. C. Reading, and G. M. Chan, A comparison of calcium,

290

33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52. 53.

FELDMAN, MALLOY, AND GROSS vitamin D, or both for nutritional rickets in Nigerian children. N. Engl. J. Med. 341, 563 – 568 (1999). M. F. Holick, Vitamin D: Photobiology, metabolism, and clinical applications. In “Endocrinology” (L. J. DeGroot, M. Besser, H. G. Burger, D. L. Loriaux, J. C. Marshall, W. D. Odell, J. T. Potts, and A. H. Rubentstein, eds.), 3rd Ed., pp. 990 – 1014. Saunders, Philadelphia, 1995. M. F. Holick, Photobiology of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 33 – 39. Academic Press, San Diego, 1997. D. D. Bikle, Vitamin D and skin. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 379 – 394. Academic Press, San Diego, 1997. S. J. Courtois, S. Segaert, H. Degreef, R. Bouillon, and M. Garmyn, Ultraviolet B suppresses vitamin D receptor gene expression in keratinocytes. Biochem. Biophys. Res. Commun. 246, 64 – 69 (1998). L. Y. Matsuoka, J. Wortsman, T. C. Chen, and M. F. Holick, Compensation for the interracial variance in the cutaneous synthesis of vitamin D. J. Lab. Clin. Med. 126, 452 – 457 (1995). W. F. Loomis, Skin-pigment regulation of vitamin-D biosynthesis in man. Science 157, 501 – 506 (1967). L. Y. Matsuoka, J. Wortsman, J. G. Haddad, P. Kolm, and B. W. Hollis, Racial pigmentation and the cutaneous synthesis of vitamin D. Arch. Dermatol. 127, 536 – 538 (1991). N. H. Bell, A. Greene, S. Epstein, M. J. Oexmann, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in blacks. J. Clin. Invest. 76, 470 – 473 (1985). M. F. Holick, J. A. MacLaughlin, and S. H. Doppelt, Regulation of cutaneous previtamin D3 photosynthesis in man: Skin pigment is not an essential regulator. Science 211, 590 – 593 (1981). J. Hirschfeld, Immune-electrophoretic demonstration of qualitative differences in human sera and their relation to the haptoglobins. Acta Pathol. Microbiol. Scand. 47, 160 – 168 (1959). S. P. Daiger, M. S. Schanfield, and L. L. Cavalli-Sforza, Groupspecific component (Gc) proteins bind vitamin D and 25-hydroxyvitamin D. Proc. Natl. Acad. Sci. USA 72, 2076 – 2080 (1975). N. E. Cooke, J. F. McLeod, X. K. Wang, and K. Ray, Vitamin D binding protein: Genomic structure, functional domains, and mRNA expression in tissues. J. Steroid Biochem. Mol. Biol. 40, 787 – 793 (1991). N. E Cooke and J. G. Haddad, Vitamin D binding protein. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 87 – 101. Academic Press, San Diego 1997. A. Sabbatini, M. Petrini, L. Mattii, P. Arnaud, and R. M. Galbraith, Vitamin D binding protein is produced by human monocytes. FEBS Lett. 323, 89 – 92 (1993). H. Cleve and J. Constans, The mutants of the vitamin-D-binding protein: More than 120 variants of the GC/DBP system. Vox. Sang. 54, 215 – 225 (1988). L. L. Cavalli-Sforza, P. Menozzi, and A. Piazza, “The History and Geography of Human Genes.” Princeton Univ. Press, Princeton, NJ, 1994. J. G. Haddad, L. Y. Matsuoka, B. W. Hollis, Y. Z. Hu, and J. Wortsman, Human plasma transport of vitamin D after its endogenous synthesis. J. Clin. Invest. 91, 2552 – 2555, (1993). R. Bouillon, K. Allewaert, D. Z. Xiang, B. K. Tan, and H. van Baelen, Vitamin D analogs with low affinity for the vitamin D binding protein: Enhanced in vitro and decreased in vivo activity. J. Bone Miner. Res. 6, 1051 – 1057 (1991). J. P. Berg, Vitamin D-binding protein prevents vitamin D deficiency and presents vitamin D for its renal activation. Eur. J. Endocrinol. 141, 321 – 322 (1999). W. M. Lee and R. M. Galbraith, The extracellular actin-scavenger system and actin toxicity. N. Engl. J. Med. 326, 1335 – 1341 (1992). M. Guoth, A. Murgia, R. M. Smith, M. B. Prystowsky, N. E. Cooke, and J. G. Haddad, Cell surface vitamin D-binding protein

54.

55.

56.

57.

58.

59.

60.

61. 62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

(GC-globulin) is acquired from plasma. Endocrinology 127, 2313 – 2321 (1990). F. F. Safadi, P. Thornton, H. Magiera, B. W. Hollis, M. Gentile, J. G. Haddad, S. A. Liebhaber, and N. E. Cooke, Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J. Clin. Invest. 103, 239 – 251 (1999). A. Nykjaer, D. Dragun, D. Walther, H. Vorum, C. Jacobsen, J. Herz, F. Melsen, E. I. Christensen, and T. E. Willnow, An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3 Cell 96, 507 – 515 (1999). J. R. Leheste, B. Rolinski, H. Vorum, J. Hilpert, A. Nykjaer, C. Jacobsen, P. Aucouturier, J. O. Moskaug, A. Otto, E. I. Christensen, and T. E. Willnow, Megalin knockout mice as an animal model of low molecular weight proteinuria. Am. J. Pathol. 155, 1361 – 1370 (1999). M. A. Gacad, H. Chen, J. E. Arbelle, T. LeBon, and J. S. Adams, Functional characterization and purification of an intracellular vitamin D-binding protein in vitamin D-resistant new world primate cells: Amino acid sequence homology with proteins in the hsp-70 family. J. Biol. Chem. 272, 8433 – 8440 (1997). M. A. Gacad and J. S. Adams, Proteins in the heat shock-70 family specifically bind 25-hydroxyvitamin D3 and 17-estradiol. J. Clin. Endocrinol. Metab. 83, 1264 – 1267 (1998). B. W. Hollis, Detection of vitamin D and its major metabolites. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 587 – 606. Academic Press, San Diego 1997. M. Gascon-Barre, The vitamin D 25-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 41 – 55. Academic Press, San Diego 1997. K. I. Okuda, Liver mitochondrial P450 involved in cholesterol catabolism and vitamin D activation. J. Lipid Res. 35, 361 – 372 (1994). E. Usui, M. Noshiro, and K. Okuda, Molecular cloning of cDNA for vitamin D3 25-hydroxylase from rat liver mitochondria. FEBS Lett. 262, 135 – 138 (1990). J. J. Cali, C. L. Hsieh, U. Francke, and D. W. Russell, Mutations in the bile acid biosynthetic enzyme sterol 27-hydroxylase underlie cerebrotendinous xanthomatosis. J. Biol. Chem. 266, 7779 – 7783 (1991). Y. D. Guo, S. Strugnell, D. W. Back, and G. Jones, Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc. Natl. Acad. Sci. USA 90, 8668 – 8672 (1993). I. Bjorkhem and S. Skrede, Familial diseases with storage of sterols other than cholesterol: Cerebrotendinous xanthomatosis and phytosterolemia. In “The Metabolic Basis of Inherited Disease.” (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, eds.), pp. 1283 – 1302. McGraw Hill, New York, 1999. V. M. Berginer, S. Shany, D. Alkalay, J. Berginer, S. Dekel, G. Salen, G. S. Tint, and D. Gazit, Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis. Metabolism 42, 69 – 74 (1993). M. R. Hughes, D. J. Baylink, P. G. Jones, and M. R. Haussler, Radioligand receptor assay for 25-hydroxyvitamin D2/D3 and 1, 25dihydroxyvitamin D2/D3. J. Clin. Invest. 58, 61 – 70 (1976). N. H. Bell, S. Shaw, and R. T. Turner, Evidence that 1,25-dihydroxyvitamin D3 inhibits the hepatic production of 25-hydroxyvitamin D in man. J. Clin. Invest. 74, 1540 – 1544 (1984). B. P. Halloran, D. D. Bikle, M. J. Levens, M. E. Castro, R. K. Globus, and E. Holton, Chronic 1,25-dihydroxyvitamin D3 administration in the rat reduces the serum concentration of 25-hydroxyvitamin D by increasing metabolic clearance rate. J. Clin. Invest. 78, 622 – 628 (1986). M. R. Clements, L. Johnson, and D. R. Fraser, A new mechanism for induced vitamin D deficiency in calcium deprivation. Nature 325, 62 – 65 (1987). H. L. Henry, The 25-hydroxyvitamin D 1-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 57 – 68. Academic Press, San Diego 1997.

CHAPTER 9 Vitamin D 72. K. Takeyama, S. Kitanaka, T. Sato, M. Kobori, J. Yanagisawa, and S. Kato, 25-Hydroxyvitamin D3 1-hydroxylase and vitamin D synthesis. Science 277, 1827 – 1830 (1997). 73. R. St-Arnaud, S. Messerlian, J. M. Moir, J. L. Omdahl, and F. H. Glorieux, The 25-hydroxyvitamin D 1--hydroxylase gene maps to the pseudovitamin D-deficiency rickets (PDDR) disease locus. J. Bone Miner. Res. 12, 1552 – 1559 (1997). 74. T. Monkawa, T. Yoshida, S. Wakino, T. Shinki, H. Anazawa, H. F. Deluca, T. Suda, M. Hayashi, and T. Saruta, Molecular cloning of cDNA and genomic DNA for human 25-hydroxyvitamin D3 1hydroxylase. Biochem. Biophys. Res. Commun. 239, 527 – 533 (1997). 75. T. Shinki, H. Shimada, S. Wakino, H. Anazawa, M. Hayashi, T. Saruta, H. F. DeLuca, and T. Suda, Cloning and expression of rat 25-hydroxyvitamin D3-1-hydroxylase cDNA. Proc. Natl. Acad. Sci. USA 94, 12920 – 12925 (1997). 76. G. K. Fu, D. Lin, M. Y. Zhang, D. D. Bikle, C. H. Shackleton, W. L. Miller, and A. A. Portale, Cloning of human 25-hydroxyvitamin D-1 -hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol. Endocrinol. 11, 1961 – 1970 (1997). 77. S. Kitanaka, K. Takeyama, A. Murayama, T. Sato, K. Okumura, M. Nogami, Y. Hasegawa, H. Niimi, J. Yanagisawa, T. Tanaka, and S. Kato, Inactivating mutations in the 25-hydroxyvitamin D3 1hydroxylase gene in patients with pseudovitamin D-deficiency rickets. N. Engl. J. Med. 338, 653 – 661 (1998). 78. M. Labuda, K. Morgan, and F. H. Glorieux, Mapping autosomal recessive vitamin D dependency type I to chromosome 12q14 by linkage analysis. Am. J. Hum. Genet. 47, 28 – 36 (1990). 79. M. Labuda, T. M. Fujiwara, M. V. Ross, K. Morgan, J. Garcia-Heras, D. H. Ledbetter, M. R. Hughes, and F. H. Glorieux, Two hereditary defects related to vitamin D metabolism map to the same region of human chromosome 12q13-14. J. Bone Miner. Res. 7, 1447 – 1453 (1992). 80. E. S. Slatopolsky and A. J. Brown, Vitamin D and renal failure. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 849 – 865. Academic Press, San Diego 1997. 81. A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, and S. Kato, Positive and negative regulations of the renal 25-hydroxyvitamin D3 1-hydroxylase gene by parathyroid hormone, calcitonin, and 1,25(OH)2D3 in intact animals. Endocrinology 140, 2224 – 2231 (1999). 82. X. F. Kong, X. H. Zhu, Y. L. Pei, D. M. Jackson, and M. F. Holick, Molecular cloning, characterization, and promoter analysis of the human 25-hydroxyvitamin D3-1-hydroxylase gene. Proc. Natl. Acad. Sci. USA 96, 6988 – 6993 (1999). 83. T. Shinki, Y. Ueno, H. F. DeLuca, and T. Suda, Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1hydroxylase gene in normocalcemic rats. Proc. Natl. Acad. Sci. USA 96, 8253 – 8258 (1999). 84. S. J. Smith, A. K. Rucka, J. L. Berry, M. Davies, S. Mylchreest, C. R. Paterson, D. A. Heath, M. Tassabehji, A. P. Read, A. P. Mee, and E. B. Mawer, Novel mutations in the 1-hydroxylase (P450c1) gene in three families with pseudovitamin D-deficiency rickets resulting in loss of functional enzyme activity in blood-derived macrophages. J. Bone Miner. Res. 14, 730 – 739 (1999). 85. N. H. Bell, Vitamin D-endocrine system. J. Clin. Invest. 76, 1 – 6 (1985). 86. P. Beckerman and J. Silver, Vitamin D and the parathyroid. Am. J. Med. Sci. 317, 363 – 369 (1999). 87. J. Silver, C. Yalcindag, A. Sela-Brown, R. Kilav, and T. Naveh-Many, Regulation of the parathyroid hormone gene by vitamin D, calcium and phosphate. Kidney Int. Suppl. 73, S2 – S7 (1999). 88. N. Chattopadhyay, A. Mithal, and E. M. Brown, The calciumsensing receptor: A window into the physiology and pathophysiology of mineral ion metabolism. Endocr. Rev. 17, 289 (1996).

291 89. E. Slatopolsky, A. Dusso, and A. J. Brown, The role of phosphorus in the development of secondary hyperparathyroidism and parathyroid cell proliferation in chronic renal failure. Am. J. Med. Sci. 317, 370 – 376 (1999). 90. A. A. Portale, B. P. Halloran, M. M. Murphy, and R. Morris, Jr., Oral intake of phosphorus can determine the serum concentration of 1,25dihydroxyvitamin D by determining its production rate in humans. J. Clin. Invest. 77, 7 – 12 (1986). 91. B. P. Halloran and M. E. Castro, Vitamin D kinetics in vivo: Effect of 1,25-dihydroxyvitamin D administration. Am. J. Physiol. 256, E686 – E691 (1989). 92. N. A. Breslau, Normal and abnormal regulation of 1,25-(OH)2D synthesis. Am. J. Med. Sci. 296, 417 – 425 (1988). 93. J. Wortsman, J. G. Haddad, J. T. Posillico, and E. M. Brown, Primary hyperparathyroidism with low serum 1,25-dihydroxyvitamin D levels. J. Clin. Endocrinol. Metab. 62, 1305 – 1308 (1986). 94. E. M. Brown, G. Gamba, D. Riccardi, M. Lombardi, R. Butters, O. Kifor, A. Sun, M. Hediger, J. Lytton, and S. C. Hebert, Cloning, expression, and characterization of an extracellular Ca2 sensing receptor from bovine parathyroid. Nature 366, 575 – 580 (1993). 95. D. Riccardi, J. Park, W. Lee, G. Gamba, E. M. Brown, and S. Hebert, Cloning and functional expression of a rat kidney extracellular calcium/polyvalent cation-sensing receptor. Proc. Natl. Acad. Sci. USA 92, 131 – 135 (1995). 96. P. Jaeger, W. Jones, T. L. Clemens, and J. P. Hayslett, Evidence that calcitonin stimulates 1,25-dihydroxyvitamin D production and intestinal absorption of calcium in vivo. J. Clin. Invest. 78, 456 – 461 (1986). 97. M. J. Econs, B. Lobaugh, and M. K. Drezner, Normal calcitonin stimulation of serum calcitriol in patients with X-linked hypophosphatemic rickets. J. Clin. Endocrinol. Metab. 75, 408 – 411 (1992). 98. T. Nesbitt, B. Lobaugh, and M. K. Drezner, Calcitonin stimulation of renal 25-hydroxyvitamin D-1-hydroxylase activity in hypophosphatemic mice: Evidence that the regulation of calcitriol production is not universally abnormal in X-linked hypophosphatemia. J. Clin. Invest. 79, 15 – 19 (1987). 99. N. H. Bell, Renal and nonrenal 25-hydroxyvitamin D-1hydroxylases and their clinical significance. J. Bone Miner. Res. 13, 350 – 353 (1998). 100. G. A. Howard, R. T. Turner, D. J. Sherrard, and D. J. Baylink, Human bone cells in culture metabolize 25-hydroxyvitamin D3 to 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3. J. Biol. Chem. 256, 7738 – 7740 (1981). 101. J. S. Adams, O. P. Sharma, M. A. Gacad, and F. R. Singer, Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J. Clin. Invest. 72, 1856 – 1860 (1983). 102. H. S. Cross, M. Peterlik, G. S. Reddy, and I. Schuster, Vitamin D metabolism in human colon adenocarcinoma-derived Caco-2 cells: Expression of 25-hydroxyvitamin D3-1-hydroxylase activity and regulation of side-chain metabolism. J. Steroid Biochem. Mol. Biol. 62, 21 – 28 (1997). 103. Y. Weisman, A. Harell, S. Edelstein, M. David, Z. Spirer, and A. Golander, 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 in vitro synthesis by human decidua and placenta. Nature 281, 317 – 319 (1979). 104. G. G. Schwartz, L. W. Whitlatch, T. C. Chen, B. L. Lokeshwar, and M. F. Holick, Human prostate cells synthesize 1,25-dihydroxyvitamin D3 from 25-hydroxyvitamin D3. Cancer Epidemiol. Biomarkers Prev. 7, 391 – 395 (1998). 105. M. Davies, M. E. Hayes, J. A. Yin, J. L. Berry, and E. B. Mawer, Abnormal synthesis of 1,25-dihydroxyvitamin D in patients with malignant lymphoma. J. Clin. Endocrinol. Metab. 78, 1202 – 1207 (1994). 106. N. H. Bell, J. Shary, S. Shaw, and R. T. Turner, Hypercalcemia associated with increased circulating 1,25 dihydroxyvitamin D in a patient with pulmonary tuberculosis. Calcif. Tissue Int. 37, 588 – 591 (1985).

292 107. J. S. Adams, F. R. Singer, M. A. Gacad, O. P. Sharma, M. J. Hayes, P. Vouros, and M. F. Holick, Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J. Clin. Endocrinol. Metab. 60, 960 – 966 (1985). 108. J. S. Adams and S. Y. Ren, Autoregulation of 1,25-dihydroxyvitamin D synthesis in macrophage mitochondria by nitric oxide. Endocrinology 137, 4514 – 4517 (1996). 109. J. Omdahl and B. May, The 25-hydroxyvitamin D 24-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 69 – 85. Academic Press, San Diego, 1997. 110. R. St-Arnaud, Targeted inactivation of vitamin D hydroxylases in mice. Bone 25, 127 – 129 (1999). 111. J. P. Goff, T. A. Reinhardt, G. W. Engstrom, and R. L. Horst, Effect of dietary calcium or phosphorus restriction and 1,25-dihydroxyvitamin D administration on rat intestinal 24-hydroxylase. Endocrinology 131, 101 – 104 (1992). 112. R. Kumar, J. Schaefer, J. P. Grande, and P. C. Roche, Immunolocalization of calcitriol receptor, 24-hydroxylase cytochrome P-450, and calbindin D28k in human kidney. Am. J. Physiol. 266, F477 – F485 (1994). 113. Y. Ohyama, M. Noshiro, and K. Okuda, Cloning and expression of cDNA encoding 25-hydroxyvitamin D3 24-hydroxylase. FEBS Lett. 278, 195 – 198 (1991). 114. C. N. Hahn, E. Baker, P. Laslo, B. K. May, J. L. Omdahl, and G. R. Sutherland, Localization of the human vitamin D 24-hydroxylase gene (CYP24) to chromosome 20q13.2 : q13.3. Cytogenet. Cell. Genet. 62, 192 – 193 (1993). 115. K. S. Chen and H. F. DeLuca, Cloning of the human 1 alpha,25-dihydroxyvitamin D3 24-hydroxylase gene promoter and identification of two vitamin D-responsive elements. Biochem. Biophys. Acta. 1263, 1 – 9 (1995). 116. Y. Ohyama, K. Ozono, M. Uchida, T. Shinki, S. Kato, T. Suda, O. Yamamoto, M. Noshiro, and Y. Kato, Identification of a vitamin D-responsive element in the 5-flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J. Biol. Chem. 269, 10545 –10550 (1994). 117. C. Zierold, H. M. Darwish, and H. F. DeLuca, Identification of a vitamin D-response element in the rat calcidiol (25-hydroxyvitamin D3) 24-hydroxylase gene. Proc. Natl. Acad. Sci. USA 91, 900 – 902 (1994). 118. P. J. Malloy, J. W. Pike, and D. Feldman, Hereditary 1,25-dihydroxyvitamin D resistant rickets. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 756 – 788. Academic Press, San Diego 1997. 119. T. Shigematsu, N. Horiuchi, Y. Ogura, T. Miyahara, and T. Suda, Human parathyroid hormone inhibits renal 24-hydroxylase activity of 25-hydroxyvitamin D3 by a mechanism involving adenosine 3,5monophosphate in rats. Endocrinology 118, 1583 – 1589 (1986). 120. A. Nishimura, T. Shinki, C. H. Jin, Y. Ohyama, M. Noshiro, K. Okuda, and T. Suda, Regulation of messenger ribonucleic acid expression of 1,25-dihydroxyvitamin D3-24-hydroxylase in rat osteoblasts. Endocrinology 134, 1794 – 1799 (1994). 121. M. J. Beckman, J. P. Goff, T. A. Reinhardt, D. C. Beitz, and R. L. Horst, In vivo regulation of rat intestinal 24-hydroxylase: Potential new role of calcitonin. J. Clin. Endocrinol. Metab. 135, 1951 – 19551 (1994). 122. R. St-Arnaud and F. H. Glorieux, 24,25-Dihydroxyvitamin D: Active metabolite or inactive catabolite? Endocrinology 139, 3371 – 3374 (1998). 123. R. Brommage and H. F. DeLuca, Evidence that 1,25-dihydroxyvitamin D3 is the physiologically active metabolite of vitamin D3. Endocr. Rev. 6, 491 – 511 (1985). 124. M. Uchida, K. Ozono, and J. W. Pike, Activation of the human osteocalcin gene by 24R,25-dihydroxyvitamin D3 occurs through the vitamin D receptor and the vitamin D-responsive element. J. Bone Miner. Res. 9, 1981 – 1987 (1994).

FELDMAN, MALLOY, AND GROSS 125. D. Somjen, G. J. Somjen, Y. Weisman, and I. Binderman, Evidence for 24,25-dihydroxycholecalciferol receptors in long bones of newborn rats. Biochem. J. 204, 31 – 36 (1982). 126. H. A. Pedrozo, Z. Schwartz, S. Rimes, V. L. Sylvia, I. Nemere, G. H. Posner, D. D. Dean, and B. D. Boyan, Physiological importance of the 1,25(OH)2D3. J. Bone Miner. Res. 14, 856 – 867 (1999). 127. A. W. Norman, J. Roth, and L. Orci, The vitamin D endocrine system: Steroid metabolism, hormone receptors, and biological response (calcium binding proteins). Endocr. Rev. 3, 331 – 366 (1982). 128. T. Yamate, T. Hiroyuki, Y. Nagai, H. Yamato, N. Taniguchi, T. Nakamura, and Y. Seino, Bone-forming ability of 24R,25-dihydroxyvitamin D3 in the hypophosphatemic mouse. J. Bone Miner. Res. 9, 1967 – 1974 (1994). 129. R. St. Arnaud, A. Arabian, R. Travers, F. Barletta, M. Raval-Pandya, K. Chapin, J. Depovere, C. Mathieu, S. Christakos, M. B. Demay, and F. H. Glorieux, Deficient mineralization of intra membranous bone in vitamin D-24-hydroxylase ablated mice is due to elevated 1,25-dihydroxy vitamin D and not to the absence of 24,25-dihydroxy vitamin D. Endocrinology 141, 2658–2666 (2000). 130. B. Endres, S. Kato, and H. F. DeLuca, Metabolism of 1,25dihydroxyvitamin D3 in vitamin D receptor-ablated mice in vivo. Biochemistry 39, 2123 – 2129 (2000). 131. M. F. Holick, A. Kleiner-Bossaller, H. K. Schnoes, P. M. Kasten, I. T. Boyle, and H. F. DeLuca, 1,24,25-Trihydroxyvitamin D3: A metabolite of vitamin D3 effective on intestine. J. Biol. Chem. 248, 6691 – 6696 (1973). 132. L. Castillo, Y. Tanaka, H. F. DeLuca, and N. Ikekawa, On the physiological role of 1,24,25-trihydroxyvitamin D3. Miner. Electrolyte Metab. 1, 198 – 207 (1978). 133. G. Jones, Analog metabolism. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 973 – 994. Academic Press, San Diego 1997. 134. M. G. Bischof, M. L. Siu-Caldera, A. Weiskopf, P. Vouros, H. S. Cross, M. Peterlik, and G. S. Reddy, Differentiation-related pathways of 1,25-dihydroxycholecalciferol metabolism in human colon adenocarcinoma-derived Caco-2 cells: Production of 1,25dihydroxy-3epi-cholecalciferol. Exp. Cell. Res. 241, 194 – 201 (1998). 135. M. L. Siu-Caldera, H. Sekimoto, A. Weiskopf, P. Vouros, K. R. Muralidharan, W. H. Okamura, J. Bishop, A. W. Norman, M. R. Uskokovic, I. Schuster, and G. S. Reddy, Production of 1,25dihydroxy-3-epi-vitamin D3 in two rat osteosarcoma cell lines (UMR 106 and ROS 17/2.8): Existence of the C-3 epimerization pathway in ROS 17/2.8 cells in which the C-24 oxidation pathway is not expressed. Bone 24, 457 – 463 (1999). 136. H. Sekimoto, M. L. Siu-Caldera, A. Weiskopf, P. Vouros, K. R. Muralidharan, W. H. Okamura, M. R. Uskokovic, and G. S. Reddy, 1,25-dihydroxy-3-epi-vitamin D3: In vivo metabolite of 1,25dihydroxyvitamin D3 in rats. FEBS Lett. 448, 278 – 282 (1999). 137. A. J. Brown, C. Ritter, E. Slatopolsky, K. R. Muralidharan, W. H. Okamura, and G. S. Reddy, 1,25-dihydroxy-3-epi-vitamin D3, a natural metabolite of 1,25-dihydroxyvitamin D3, is a potent suppressor of parathyroid hormone secretion. J. Cell. Biochem. 73, 106 – 113 (1999). 138. M. R. Haussler, C. A. Haussler, P. W. Jurutka, P. D. Thompson, J. C. Hsieh, L. S. Remus, S. H. Selznick, and G. K. Whitfield, The vitamin D hormone and its nuclear receptor: Molecular actions and disease states. J. Endocrinol. 154(Suppl.), S57 – S73 (1997). 139. G. K. Whitfield, P. W. Jurutka, C. A. Haussler, and M. R. Haussler, Steroid hormone receptors: Evolution, ligands, and molecular basis of biologic function. J. Cell. Biochem. 32 – 33(Suppl.), 110 – 122 (1999). 140. M. R. Haussler and A. W. Norman, Chromosomal receptor for a vitamin D metabolite. Proc. Natl. Acad. Sci. USA 62, 155 – 162 (1969).

CHAPTER 9 Vitamin D 141. D. Feldman, T. A. McCain, M. A. Hirst, T. L. Chen, and K. W. Colston, Characterization of a cytoplasmic receptor-like binder for 1 alpha, 25-dihydroxycholecalciferol in rat intestinal mucosa. J. Biol. Chem. 254, 10378 – 10384 (1979). 142. E. A. Allegretto and J. W. Pike, Trypsin cleavage of chick 1,25-dihydroxyvitamin D3 receptors: Generation of discrete polypeptides which retain hormone but are unreactive to DNA and monoclonal antibody. J. Biol. Chem. 260, 10139 – 10145 (1985). 143. J. W. Pike, Evidence for a reactive sulfhydryl in the DNA binding domain of the 1,25-dihydroxyvitamin D3 receptor. Biochem. Biophys. Res. Commun. 100, 1713 – 1719 (1981). 144. J. W. Pike and M. R. Haussler, Purification of chicken intestinal receptor for 1,25-dihydroxyvitamin D. Proc. Natl. Acad. Sci. USA 76, 5785 – 5789 (1979). 145. R. U. Simpson, A. Hamstra, N. C. Kendrick, and H. F. DeLuca, Purification of the receptor for 1,25-dihydroxyvitamin D3 from chicken intestine. Biochemistry 22, 2586 – 2594 (1983). 146. J. W. Pike, Monoclonal antibodies to chick intestinal receptors for 1,25-dihydroxyvitamin D3: Interaction and effects of binding on receptor function. J. Biol. Chem. 259, 1167 – 1173 (1984). 147. M. C. Dame, E. A. Pierce, J. M. Prahl, C. E. Hayes, and H. F. DeLuca, Monoclonal antibodies to the porcine intestinal receptor for 1,25-dihydroxyvitamin D3: Interaction with distinct receptor domains. Biochemistry 25, 4523 – 4534 (1986). 148. W. E. Stumpf, M. Sar, F. A. Reid, Y. Tanaka, and H. F. DeLuca, Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid. Science 206, 1188 – 1190 (1979). 149. K. Colston, M. Hirst, and D. Feldman, Organ distribution of the cytoplasmic 1,25-dihydroxycholecalciferol receptor in various mouse tissues. Endocrinology 107, 1916 – 1922 (1980). 150. T. L. Clemens, K. P. Garrett, X. Y. Zhou, J. W. Pike, M. R. Haussler, and D. W. Dempster, Immunocytochemical localization of the 1,25dihydroxyvitamin D3 receptor in target cells. Endocrinology 122, 1224 – 1230 (1988). 151. U. Berger, P. Wilson, R. A. McClelland, K. Colston, M. R. Haussler, J. W. Pike, and R. C. Coombes, Immunocytochemical detection of 1,25-dihydroxyvitamin D receptors in normal human tissues. J. Clin. Endocrinol. Metab. 67, 607 – 613 (1988). 152. M. R. Walters, Newly identified actions of the vitamin D endocrine system. Endocr. Rev. 13, 719 – 764 (1992). 153. J. W. Pike and N. M. Sleator, Hormone-dependent phosphorylation of the 1,25-dihydroxyvitamin D3 receptor in mouse fibroblasts. Biochem. Biophys. Res. Commun. 131, 378 – 385 (1985). 154. T. A. Brown and H. F. DeLuca, Phosphorylation of the 1,25dihydroxyvitamin D3 receptor: A primary event in 1,25-dihydroxyvitamin D3 action. J. Biol. Chem. 265, 10025 – 10029 (1990). 155. J. C. Hsieh, P. W. Jurutka, M. A. Galligan, C. M. Terpening, C. A. Haussler, D. S. Samuels, Y. Shimizu, N. Shimizu, and M. R. Haussler, Human vitamin D receptor is selectively phosphorylated by protein kinase C on serine 51, a residue crucial to its trans-activation function. Proc. Natl. Acad. Sci. USA 88, 9315 – 9319 (1991). 156. P. W. Jurutka, J. C. Hsieh, P. N. MacDonald, C. M. Terpening, C. A. Haussler, M. R. Haussler, and G. K. Whitfield, Phosphorylation of serine 208 in the human vitamin D receptor: The predominant amino acid phosphorylated by casein kinase II, in vitro, and identification as a significant phosphorylation site in intact cells. J. Biol. Chem. 268, 6791 – 6799 (1993). 157. G. M. Hilliard, R. G. Cook, N. L. Weigel, and J. W. Pike, 1,25-dihydroxyvitamin D3 modulates phosphorylation of serine 205 in the human vitamin D receptor: Site-directed mutagenesis of this residue promotes alternative phosphorylation. Biochemistry 33, 4300 – 4311 (1994). 158. P. W. Jurutka, J. C. Hsieh, and M. R. Haussler, Phosphorylation of the human 1,25-dihydroxyvitamin D3 receptor by cAMP-dependent protein kinase, in vitro, and in transfected COS-7 cells. Biochem. Biophys. Res. Commun. 191, 1089 – 1096 (1993).

293 159. P. W. Jurutka, J. C. Hsieh, S. Nakajima, C. A. Haussler, G. K. Whitfield, and M. R. Haussler, Human vitamin D receptor phosphorylation by casein kinase II at Ser-208 potentiates transcriptional activation. Proc. Natl. Acad. Sci. USA 93, 3519 – 3524 (1990). 160. J. C. Hsieh, P. W. Jurutka, S. Nakajima, M. A. Galligan, C. A. Haussler, Y. Shimizu, N. Shimizu, G. K. Whitfield, and M. R. Haussler, Phosphorylation of the human vitamin D receptor by protein kinase C: Biochemical and functional evaluation of the serine 51 recognition site. J. Biol. Chem. 268, 15118 – 15126 (1993). 161. D. P. McDonnell, D. J. Mangelsdorf, J. W. Pike, M. R. Haussler, and B. W. OMalley, Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235, 1214 – 1217 (1987). 162. A. R. Baker, D. P. McDonnell, M. Hughes, T. M. Crisp, D. J. Mangelsdorf, M. R. Haussler, J. W. Pike, J. Shine, and B. W. OMalley, Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc. Natl. Acad. Sci. USA 85, 3294 – 3298 (1988). 163. D. J. Mangelsdorf, C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, et al. (1995). The nuclear receptor superfamily: The second decade. Cell 83, 835 – 839 (1995). 164. D. P. McDonnell, R. A. Scott, S. A. Kerner, B. W. O’Malley, and J. W. Pike, Functional domains of the human vitamin D3 receptor regulate osteocalcin gene expression. Mol. Endocrinol. 3, 635 – 644 (1989). 165. J. H. Faraco, N. A. Morrison, A. Baker, J. Shine, and P. M. Frossard, ApaI dimorphism at the human vitamin D receptor gene locus. Nucleic Acids Res. 17, 2150 (1989). 166. J. Szpirer, C. Szpirer, M. Riviere, G. Levan, P. Marynen, J. J. Cassiman, R. Wiese, and H. F. DeLuca, The Sp1 transcription factor gene (SP1) and the 1,25-dihydroxyvitamin D3 receptor gene (VDR) are colocalized on human chromosome arm 12q and rat chromosome 7. Genomics 11, 168 – 173 (1991). 167. K. Miyamoto, R. A. Kesterson, H. Yamamoto, Y. Taketani, E. Nishiwaki, S. Tatsumi, Y. Inoue, K. Morita, E. Takeda, and J. W. Pike, Structural organization of the human vitamin D receptor chromosomal gene and its promoter. Mol. Endocrinol. 11, 1165 – 1179 (1997). 168. L. A. Crofts, M. S. Hancock, N. A. Morrison, and J. A. Eisman, Multiple promoters direct the tissue-specific expression of novel N-terminal variant human vitamin D receptor gene transcripts. Proc. Natl. Acad. Sci. USA 95, 10529 – 10534 (1998). 169. T. Saijo, M. Ito, E. Takeda, A. H. Huq, E. Naito, I. Yokota, T. Sone, J. W. Pike, and Y. Kuroda, A unique mutation in the vitamin D receptor gene in three Japanese patients with vitamin D-dependent rickets type II: Utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am. J. Hum. Genet. 49, 668 – 673 (1991). 170. H. Arai, K. I. Miyamoto, Y. Taketani, H. Yamamoto, Y. Iemori, K. Morita, T. Tonai, T. Nishisho, S. Mori, and E. Takeda, A vitamin D receptor gene polymorphism in the translation initiation codon: Effect on protein activity and relation to bone mineral density in Japanese women. J. Bone Miner. Res. 12, 915 – 921 (1997). 171. C. Gross, A. V. Krishnan, P. J. Malloy, T. R. Eccleshall, X. Y. Zhao, and D. Feldman, The vitamin D receptor gene start codon polymorphism: A functional analysis of FokI variants. J. Bone Miner. Res. 13, 1691 – 1699 (1998). 172. P. W. Jurutka, L. S. Remus, G. K. Whitfield, P. D. Thompson, J. C. Hsieh, H. Zitzer, P. Tavakkoli, M. A. Galligan, H. T. Dang, C. A. Haussler, and M. R. Haussler, The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol. Endocrinol. 14, 401 – 420 (2000). 173. A. W. Norman, D. Adams, E. D. Collins, W. H. Okamura, and R. J. Fletterick, Three-dimensional model of the ligand binding domain of the nuclear receptor for 1,25-dihydroxyvitamin D3. J. Cell. Biochem. 74, 323 – 333 (1999).

294 174. K. Yamamoto, H. Masuno, M. Choi, K. Nakashima, T. Taga, H. Ooizumi, K. Umesono, W. Sicinska, J. VanHooke, H. F. DeLuca, and S. Yamada, Three-dimensional modeling of and ligand docking to vitamin D receptor ligand binding domain. Proc. Natl. Acad. Sci. USA 97, 1467 – 1472 (2000). 175. N. Rochel, J. M. Wurtz, A. Mitschler, B. Klaholz, and D. Moras, The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol. Cell. 5, 173 – 179 (2000). 176. K. Paech, P. Webb, G. G. Kuiper, S. Nilsson, J. Gustafsson, P. J. Kushner, and T. S. Scanlan, Differential ligand activation of receptors ER and ER at AP1 sites. Science 277, 1508 – 1510 (1997). 177. A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A. Agard, and G. L. Greene, The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927 – 937 (1998). 178. K. Kristjansson, A. R. Rut, M. Hewison, J. L. ORiordan, and M. R. Hughes, Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25 dihydroxyvitamin D3. J. Clin. Invest. 92, 12 – 16 (1993). 179. P. J. Malloy, T. R. Eccleshall, C. Gross, L. Van Maldergem, R. Bouillon, and D. Feldman, Hereditary vitamin D resistant rickets caused by a novel mutation in the vitamin D receptor that results in decreased affinity for hormone and cellular hyporesponsiveness. J. Clin. Invest. 99, 297 – 304 (1997). 180. S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc Natl. Acad. Sci. USA 86, 4455 – 4459 (1989). 181. J. Lian, C. Stewart, E. Puchacz, S. Mackowiak, V. Shalhoub, D. Collart, G. Zambetti, and G. Stein, Structure of the rat osteocalcin gene and regulation of vitamin D-dependent expression. Proc. Natl. Acad. Sci. USA 86, 1143 – 1147 (1989). 182. N. A. Morrison, J. Shine, J. C. Fragonas, V. Verkest, M. L. Mc Menemy, and J. A. Eisman, 1,25-dihydroxyvitamin D-responsive element and glucocorticoid repression in the osteocalcin gene. Science 246, 1158 – 1161 (1989). 183. J. Liao, K. Ozono, T. Sone, D. P. McDonnell, and J. W. Pike, Vitamin D receptor interaction with specific DNA requires a nuclear protein and 1,25-dihydroxyvitamin D3. Proc. Natl. Acad. Sci. USA 87, 9751 – 9755 (1990). 184. T. Sone, S. Kerner, and J. W. Pike, Vitamin D receptor interaction with specific DNA. Association as a 1,25-dihydroxyvitamin D3modulated heterodimer. J. Biol. Chem. 266, 23296 – 23305 (1991). 185. T. Sone, K. Ozono, and J. W. Pike, A 55-kilodalton accessory factor facilitates vitamin D receptor DNA binding. Mol. Endocrinol. 5, 1578 – 1586 (1991). 186. P. N. MacDonald, D. R. Dowd, S. Nakajima, M. A. Galligan, M. C. Reeder, C. A. Haussler, K. Ozato, and M. R. Haussler, Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol. Cell. Biol. 13, 5907 – 5917 (1993). 187. D. J. Mangelsdorf, E. S. Ong, J. A. Dyck, and R. M. Evans, Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345, 224 – 229 (1990). 188. A. A. Levin, L. J. Sturzenbecker, S. Kazmer, T. Bosakowski, C. Huselton, G. Allenby, J. Speck, C. Kratzeisen, M. Rosenberger, A. Lovey et al., 9-cis retinoic acid stereoisomer binds and activated the nuclear receptor RXR alpha. Nature 355, 359 – 361 (1992). 189. R. A. Heyman, D. J. Mangelsdorf, J. A. Dyck, R. B. Stein, G. Eichele, R. M. Evans, and C. Thaller, 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68, 397 – 406 (1992). 190. D. J. Mangelsdorf and R. M. Evans, The RXR heterodimers and orphan receptors. Cell 83, 841 – 850 (1995).

FELDMAN, MALLOY, AND GROSS 191. C. Solomon, J. H. White, and R. Kremer, Mitogen-activated protein kinase inhibits 1,25-dihydroxyvitamin D3-dependent signal transduction by phosphorylating human retinoid X receptor alpha. J. Clin. Invest. 103, 1729 – 1735 (1999). 192. D. Robyr, A. P. Wolffe, and W. Wahli, Nuclear hormone receptor coregulators in action: Diversity for shared tasks. Mol. Endocrinol. 14, 329 – 347 (2000). 193. T. E. Spencer, G. Jenster, M. M. Burcin, C. D. Allis, J. Zhou, C. A. Mizzen, N. J. McKenna, S. A. Onate, S. Y. Tsai, M. J. Tsai, and B. W. OMalley, Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389, 194 – 198 (1997). 194. R. K. Gill, L. M. Atkins, B. W. Hollis, and N. H. Bell, Mapping the domains of the interaction of the vitamin D receptor and steroid receptor coactivator-1. Mol. Endocrinol. 12, 57 – 65 (1998). 195. D. M. Kraichely, J. J. R. Collins, R. K. DeLisle, and P. N. Mac Donald, The autonomous transactivation domain in helix H3 of the vitamin D receptor is required for transactivation and coactivator interaction. J. Biol. Chem. 274, 14352 – 14358 (1999). 196. J. C. Blanco, I. M. Wang, S. Y. Tsai, M. J. Tsai, B. W. OMalley, P. W. Jurutka, M. R. Haussler, and K. Ozato, Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc. Natl. Acad. Sci. USA 92, 1535 – 1539 (1995). 197. P. N. MacDonald, D. R. Sherman, D. R. Dowd, S. C. Jefcoat, Jr., and R. K. DeLisle, The vitamin D receptor interacts with general transcription factor IIB. J. Biol. Chem. 270, 4748 – 4752 (1995). 198. P. W. Jurutka, J. C. Hsieh, L. S. Remus, G. K. Whitfield, P. D. Thompson, C. A. Haussler, J. C. Blanco, K. Ozato, and M. R. Haussler, Mutations in the 1,25-dihydroxyvitamin D3 receptor identifying C-terminal amino acids required for transcriptional activation that are functionally dissociated from hormone binding, heterodimeric DNA binding, and interaction with basal transcription factor IIB, in vitro. J. Biol. Chem. 272, 14592 – 14599 (1997). 199. H. Masuyama, S. C. Jefcoat, Jr., and P. N. MacDonald, The N-terminal domain of transcription factor IIB is required for direct interaction with the vitamin D receptor and participates in vitamin D-mediated transcription. Mol. Endocrinol. 11, 218 – 228 (1997). 200. T. A. Baudino, D. M. Kraichely, S. C. Jefcoat, Jr., S. K. Winchester, N. C. Partridge, and P. N. MacDonald, Isolation and characterization of a novel coactivator protein, NCoA-62, involved in vitamin D-mediated transcription. J. Biol. Chem. 273, 16434 – 16441 (1998). 201. C. Rachez, Z. Suldan, J. Ward, C. P. Chang, D. Burakov, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, A novel protein complex that interacts with the vitamin D3 receptor in a liganddependent manner and enhances VDR transactivation in a cell-free system. Genes Dev. 12, 1787 – 1800 (1998). 202. C. Rachez, B. D. Lemon, Z. Suldan, V. Bromleigh, M. Gamble, A. M. Naar, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398, 824 – 828 (1999). 203. Y. Yanagi, M. Suzawa, M. Kawabata, K. Miyazono, J. Yanagisawa, and S. Kato, Positive and negative modulation of vitamin D receptor function by transforming growth factor- signaling through smad proteins. J. Biol. Chem. 274, 12971 – 12974 (1999). 204. J. Yanagisawa, Y. Yanagi, Y. Masuhiro, M. Suzawa, M. Watanabe, K. Kashiwagi, T. Toriyabe, M. Kawabata, K. Miyazono, and S. Kato, Convergence of transforming growth factor- and vitamin D signaling pathways on SMAD transcriptional coactivators. Science 283, 1317 – 1321 (1999). 205. J. Barsony, J. W. Pike, H. F. DeLuca, and S. J. Marx, Immunocytology with microwave-fixed fibroblasts shows 1, 25-dihydroxyvitamin D3-dependent rapid and estrogen-dependent slow reorganization of vitamin D receptors. J. Cell Biol. 111, 2385 – 2395 (1990). 206. J. Barsony and W. McKoy, Molybdate increases intracellular 3,5guanosine cyclic monophosphate and stabilizes vitamin D receptor

CHAPTER 9 Vitamin D

207.

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

218.

219.

220.

221.

222.

223.

224.

association with tubulin-containing filaments. J. Biol. Chem. 267, 24457 – 24465 (1992). A. Racz and J. Barsony, Hormone-dependent translocation of vitamin D receptors is linked to transactivation. J. Biol. Chem. 274, 19352 – 19360 (1999). K. Prüfer, C. Schröder, A. Racz, and J. Barsony. Cell cycle dependence of vitamin D receptor expression. In “vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 379 – 382. University of California, Printing and Reprographics, Riverside, 2000. M. Hirst and D. Feldman, Regulation of 1,25(OH)2 vitamin D3 receptor content in cultured LLC-PK1 kidney cells limits hormonal responsiveness. Biochem. Biophys. Res. Commun. 116, 121 – 127 (1983). T. L. Chen, J. M. Li, T. Van Ye, C. M. Cone and D. Feldman, Hormonal responses to 1,25-dihydroxyvitamin D3 in cultured mouse osteoblast-like cells-modulation by changes in receptor level. J. Cell. Physiol. 126, 21 – 28 (1986). A. V. Krishnan and D. Feldman, Cyclic adenosine 3,5-monophosphate upregulates 1,25-dihydroxyvitamin D3 receptor gene expression and enhances hormone action. Mol. Endocrinol. 6, 198 – 206 (1992). A. V. Krishnan, S. D. Cramer, F. R. Bringhurst, and D. Feldman, Regulation of 1,25-dihydroxyvitamin D3 receptors by parathyroid hormone in osteoblastic cells: Role of second messinger pathways. Endocrinology 136, 705 – 712 (1995). A. V. Krishnan and D. Feldman, Regulation of vitamin D receptor abundance. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 179 – 200. Academic Press, San Diego 1997. E. M. Costa, M. A. Hirst, and D. Feldman, Regulation of 1,25dihydroxyvitamin D3 receptors by vitamin D analogs in cultured mammalian cells. Endocrinology 117, 2203 – 2210 (1985). E. M. Costa and D. Feldman, Homologous up-regulation of the 1,25 (OH)2 vitamin D3 receptor in rats. Biochem. Biophys. Res. Commun. 137, 742 – 747 (1986). M. J. Favus, D. J. Mangelsdorf, V. Tembe, B. J. Coe, and M. R. Haussler, Evidence for in vivo upregulation of the intestinal vitamin D receptor during dietary calcium restriction in the rat. J. Clin. Invest. 82, 218 – 224 (1988). M. Strom, M. E. Sandgren, T. A. Brown, and H. F. DeLuca, 1,25-Dihydroxyvitamin D3 up-regulates the 1,25-dihydroxyvitamin D3 receptor in vivo. Proc. Natl. Acad. Sci. USA 86, 9770 – 9773 (1993). X. Zhao and D. Feldman, Regulation of vitamin D receptor abundance and responsiveness during differentiation of HT-29 human colon cancer cells. Endocrinology 132, 1808 – 1814 (1993). R. J. Wiese, A. Uhland-Smith, T. K. Ross, J. M. Prahl, and H. F. DeLuca, Up-regulation of the vitamin D receptor in response to 1,25-dihydroxyvitamin D3 results from ligand-induced stabilization. J. Biol. Chem. 267, 20082 – 20086 (1992). D. Santiso-Mere, T. Sone, G. T. Hilliard, J. W. Pike, and D. P. McDonnell, Positive regulation of the vitamin D receptor by its cognate ligand in heterologous expression systems. Mol. Endocrinol. 7, 833 – 839 (1993). N. C. Arbour, J. M. Prahl, and H. F. DeLuca, Stabilization of the vitamin D receptor in rat osteosarcoma cells through the action of 1,25-dihydroxyvitamin D3. Mol. Endocrinol. 7, 1307 – 1312 (1993). M. L. Chen, A. Perez, D. K. Sanan, G. Heinrich, T. C. Chen, and M. F. Holick, Induction of vitamin D receptor mRNA expression in psoriatic plaques correlates with clinical response to 1,25-dihydroxyvitamin D3. J. Invest. Dermatol. 106, 637 – 641 (1996). T. L. Chen and D. Feldman, Regulation of 1,25-dihydroxyvitamin D3 receptors in cultured mouse bone cells: Correlation of receptor concentration with the rate of cell division. J. Biol. Chem. 256, 5561 – 5566 (1981). A. Krishnan and D. Feldman Stimulation of 1,25-dihydroxyvitamin D3 receptor gene expression in cultured cells by serum and growth factors. J. Bone Miner. Res. 6, 1099 – 1107 (1991).

295 225. S. Pillai D. D. Bikle, and P. M. Elias, 1,25-Dihydroxyvitamin D production and receptor binding in human keratinocytes varies with differentiation. J. Biol. Chem. 263, 5390 – 5395 (1988). 226. X. P. Yu, H. Mocharla, F. G. Hustmyer, and S. C. Manolagas, Vitamin D receptor expression in human lymphocytes: Signal requirements and characterization by western blots and DNA sequencing. J. Biol. Chem. 266, 7588 – 7595 (1991). 227. A. V. Krishnan and D. Feldman, Activation of protein kinase-C inhibits vitamin D receptor gene expression. Mol. Endocrinol. 5, 605 – 612 (1991). 228. J. P. van Leeuwen, J. C. Birkenhager, T. Vink-van Wijngaarden, G. J. van den Bemd, and H. A. Pols, Regulation of 1,25-dihydroxyvitamin D3 receptor gene expression by parathyroid hormone and cAMPagonists. Biochem. Biophys. Res. Commun. 185, 881 – 886 (1992). 229. F. Jehan and H. F. DeLuca, Cloning and characterization of the mouse vitamin D receptor promoter. Proc. Natl. Acad. Sci. USA 94, 10138 – 10143 (1997). 230. S. J. Smith, L. M. Green, M. E. Hayes, and E. B. Mawer, Prostaglandin E2 regulates vitamin D receptor expression, vitamin D-24-hydroxylase activity and cell proliferation in an adherent human myeloid leukemia cell line (Ad-HL60). Prostaglandins Other Lipid Mediat. 57, 73 – 85 (1999). 231. H. Yamamoto, K. Miyamoto, B. Li, Y. Taketani, M. Kitano, Y. Inoue, K. Morita, J. W. Pike, and E. Takeda, The caudal-related homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J. Bone Miner. Res. 14, 240 – 247 (1999). 232. T. L. Chen, C. M. Cone, E. Morey-Holton, and D. Feldman, Glucocorticoid regulation of 1,25(OH)2-vitamin D3 receptors in cultured mouse bone cells. J. Biol. Chem. 257, 13564 – 13569 (1982). 233. M. Hirst and D. Feldman, Glucocorticoid regulation of 1,25(OH)2 vitamin D3 receptors: Divergent effects on mouse and rat intestine. Endocrinology 111, 1400 – 1402 (1982). 234. M. Hirst and D. Feldman, Glucocorticoids down-regulate the number of 1,25-dihydroxyvitamin D3 receptors in mouse intestine. Biochem. Biophys. Res. Commun. 105, 1590 – 1596 (1982). 235. H. K. Nielsen, E. F. Eriksen, T. Storm, and L. Mosekiilde, The effects of short-term, high-dose treatment with prednisone on the nuclear uptake of 1,25-dihydroxyvitamin D3 in monocytes from normal human subjects. Metabolism 37, 109 – 114 (1988). 236. M. R. Walters, An estrogen-stimulated 1,25-dihydroxyvitamin D3 receptor in rat uterus. Biochem. Biophys. Res. Commun. 103, 721 – 726 (1981). 237. Y. Liel, S. Kraus, J. Levy, and S. Shany, Evidence that estrogens modulate activity and increase the number of 1,25-dihydroxyvitamin D receptors in osteoblast-like cells (ROS 17/2.8). Endocrinology 130, 2597 – 2601 (1992). 238. P. M. Petkovich, J. N. Heersche, D. O. Tinker, and G. Jones, Retinoic acid stimulates 1,25-dihydroxyvitamin D3 binding in rat osteosarcoma cells. J. Biol. Chem. 259, 8274 – 8280 (1984). 239. T. L. Chen and D. Feldman, Retinoic acid modulation of 1,25(OH)2 vitamin D3 receptors and bioresponse in bone cells: Species differences between rat and mouse. Biochem. Biophys. Res. Commun. 132, 74 – 80 (1985). 240. T. A. Reinhardt and R. L. Horst, Parathyroid hormone downregulates 1,25-dihydroxyvitamin D receptors (VDR) and VDR messenger ribonucleic acid in vitro and blocks homologous upregulation of VDR in vivo. Endocrinology 127, 942 – 948 (1990). 241. S. B. Borwn, T. T. Brierley, N. Palanisamy, I. B. Salusky, W. Goodman, M. Brandi, T. B. Drueke, E. Sarfati, P. Urena, R. S. Chaganti, J. W. Pike, and A. Arnold, Vitamin D receptor as a candidate tumor-suppressor gene in severe hyperparathyroidism of uremia. J. Clin. Endocrinol. Metab. 85, 868 – 872 (2000). 242. A. B. Korkor, Reduced binding of [3H]1,25-dihydroxyvitamin D3 in the parathyroid glands of patients with renal failure. N. Engl. J. Med. 316, 1573 – 1577 (1987).

296 243. S. J. Silverberg, E. Shane, D. W. Depster, and J. P. Bilezikian, The effects of vitamin D insufficiency in patients with primary hyperparathyroidism. Am. J. Med. 107, 561 – 567 (1999). 244. D. Feldman, Vitamin D, parathyroid hormone, and calcium: A complex regulatory network. Am. J. Med. 107, 637 – 639 (1999). 245. D. S. Rao, M. Honasoge, G. W. Divine, E. R. Phillips, M. W. Lee, M. R. Ansari, G. B. Talpos, and A. M. Parfitt, Effect of vitamin D nutrition on parathyroid adenoma weight: Pathogenetic and clinical implications. J. Clin. Endocrinol. Metab. 85, 1054 – 1058 (2000). 246. A. W. Norman, Rapid biological responses mediated by 1,25dihydroxyvitamin D3: A case study of transcaltachia (rapid hormonal stimulation of intestinal calcium transport). In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 233 – 256. Academic Press, San Diego, 1997. 247. A. W. Norman, X. Song, L. Zanello, C. Bula, and W. H. Okamura, Rapid and genomic biological responses are mediated by different shapes of the agonist steroid hormone, 1,25(OH)2 vitamin D3. Steroids 64, 120 – 128 (1999). 248. J. Barsony and S. J. Marx, Rapid accumulation of cyclic GMP near activated vitamin D receptors. Proc. Natl. Acad. Sci. USA 88, 1436 – 1440 (1991). 249. D. T. Baran, A. M. Sorensen, V. Shalhoub, T. Owen, A. Oberdorf, G. Stein and J. Lian, 1-,25-dihydroxyvitamin D3 rapidly increases cytosolic calcium in clonal rat osteosarcoma cells lacking the vitamin D receptor. J. Bone Miner. Res. 6, 1269–1275 (1991). 250. A. W. Norman, R. Bouillon, M. C. Farach-Carson, J. E. Bishop, L. X. Zhou, I. Nemere, J. Zhao, K. R. Muralidharan, and W. H. Okamura, Demonstration that 1,25-dihydroxyvitamin D3 is an antagonist of the nongenomic but not genomic biological responses and biological profile of the three A-ring diastereomers of 1,25-dihydroxyvitamin D3. J. Biol. Chem. 268, 20022 – 20030 (1993). 251. A. W. Norman, W. H. Okamura, M. W. Hammond, J. E. Bishop, M. C. Dormanen, R. Bouillon, H. van Baelen, A. L. Ridall, E. Daane, R. Khoury, and M. C. Farach-Carson, Comparison of 6-s-cis- and 6-s-trans-locked analogs of 1,25-dihydroxyvitamin D3 indicates that the 6-s-cis conformation is preferred for rapid nongenomic biological responses and that neither 6-s-cis- nor 6-s-trans-locked analogs are preferred for genomic biological responses. Mol. Endocrinol. 11, 1518 – 1531 (1997). 252. M. Bhatia, J. B. Kirkland, and K. A. Meckling-Gill, Monocytic differentiation of acute promyelocytic leukemia cells in response to 1,25-dihydroxyvitamin D3 is independent of nuclear receptor binding. J. Biol. Chem. 270, 15962 – 15965 (1995). 253. M. Kajikawa, H. Ishida, S. Fujimoto, E. Mukai, M. Nishimura, J. Fujita, Y. Tsuura, Y. Okamoto, A. W. Norman, and Y. Seino, An insulinotropic effect of vitamin D analog with increasing intracellular Ca2 concentration in pancreatic -cells through nongenomic signal transduction. Endocrinology 140, 4706 – 4712 (1999). 254. I. Nemere, Z. Schwartz, H. Pedrozo, V. L. Sylvia, D. D. Dean, and B. D. Boyan, Identification of a membrane receptor for 1,25-dihydroxyvitamin D3 which mediates rapid activation of protein kinase C. J. Bone Miner. Res. 13, 1353 – 1359 (1998). 255. R. P. Heaney, Vitamin D: Role in the calcium economy. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 485 – 497. Academic Press, San Diego, 1997. 256. E. Slatopolsky, A. Brown, and A. Dusso, Pathogenesis of secondary hyperparathyroidism. Kidney Int. Suppl. 73, S14 – 19 (1999). 257. J. A. Johnson and R. Kumar, Renal and intestinal calcium transport: Roles of vitamin D and vitamin D-dependent calcium binding proteins. Semin. Nephrol. 14, 119 – 128 (1994). 258. A. J. Brown, M. Zhong, J. Finch, C. Ritter, R. McCracken, J. Morrissey, and E. Slatopolsky, Rat calcium-sensing receptor is regulated by vitamin D but not by calcium. Am. J. Physiol. 270, F454 – F460 (1996).

FELDMAN, MALLOY, AND GROSS 259. J. L. Finch, A. J. Brown, and E. Slatopolsky, Differential effects of 1,25-dihydroxyvitamin D3 and 19-nor-1,25-dihydroxyvitamin D2 on calcium and phosphorus resorption in bone. J. Am. Soc. Nephrol. 10, 980 – 985 (1999). 260. J. S. Adams, Extrarenal production and action of active vitamin D metabolites in human lymphoproliferative diseases. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 903–921. Academic Press, San Diego, 1997. 261. A. Dusso, A. Brown, and E. Slatopolsky, Extrarenal production of calcitriol. Semin. Nephrol. 14, 144 – 155 (1994). 262. L. E. Siltzbach, D. G. James, E. Neville, J. Turiaf, J. P. Battesti, O. P. Sharma, Y. Hosoda, R. Mikami, and M. Odaka, Course and prognosis of sarcoidosis around the world. Am. J. Med. 57, 847 – 852 (1974). 263. G. L. Barbour, J. W. Coburn, E. Slatopolsky, A. W. Norman, and R. L. Horst, Hypercalcemia in an anephric patient with sarcoidosis: evidence for extrarenal generation of 1,25-dihydroxyvitamin D. N. Engl. J. Med. 305, 440 – 443 (1981). 264. R. S. Mason, T. Frankel, Y. L. Chan, D. Lissner, and S. Posen, Vitamin D conversion by sarcoid lymph node homogenate. Ann. Intern. Med. 100, 59 – 61 (1984). 265. J. F. Seymour and R. F. Gagel, Calcitriol: The major humoral mediator of hypercalcemia in Hodgkins disease and non-Hodgkin’s lymphomas. Blood 82, 1383 – 1394 (1993). 266. J. F. Seymour R. F. Gagel, F. B. Hagemeister, M. A. Dimopoulos, and F. Cabanillas, Calcitriol production in hypercalcemic and normocalcemic patients with non-Hodgkin lymphoma. Ann. Intern. Med. 121, 633 – 640 (1994). 267. R. Kumar, W. R. Cohen, P. Silva, and F. H. Epstein, Elevated 1,25dihydroxyvitamin D plasma levels in normal human pregnancy and lactation. J. Clin. Invest. 63, 342 – 344 (1979). 268. D. D. Bikle, E. Gee, B. Halloran, and J. G. Haddad, Free 1,25-dihydroxyvitamin D levels in serum from normal subjects, pregnant subjects, and subjects with liver disease. J. Clin. Invest. 74, 1966 – 1971 (1984). 269. C. Cheema, B. F. Grant, and R. Marcus, Effects of estrogen on circulating “free” and total 1,25-dihydroxyvitamin D and on the parathyroid-vitamin D axis in postmenopausal women. J. Clin. Invest. 83, 537 – 542 (1989). 270. S. Vargas, R. Bouillon, H. Van Baelen, and L. G. Raisz, Effects of vitamin D-binding protein on bone resorption stimulated by 1,25 dihydroxyvitamin D3. Calcif. Tissue Int. 47, 164 – 168 (1990). 271. T. K. Gray, G. E. Lester, and R. S. Lorenc, Evidence for extra-renal 1-hydroxylation of 25-hydroxyvitamin D3 in pregnancy. Science 204, 1311 – 1313 (1979). 272. J. A. Whitsett, M. Ho, R. C. Tsang, E. J. Norman, and K. G. Adams, Synthesis of 1,25-dihydroxyvitamin D3 by human placenta in vitro. J. Clin. Endocrinol. Metab. 53, 484 – 488 (1981). 273. N. A. Breslau and J. E. Zerwekh, Relationship of estrogen and pregnancy to calcium homeostasis in pseudohypoparathyroidism. J. Clin. Endocrinol. Metab. 62, 45 – 51 (1986). 274. A. D. Care, Vitamin D in pregnancy, the fetoplacental unit, and lactation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 437 – 446. Academic Press, San Diego, 1997. 275. A. R. Glass and C. Eil, Ketoconazole-induced reduction in serum 1,25-dihydroxyvitamin D and total serum calcium in hypercalcemic patients. J. Clin. Endocrinol. Metab. 66, 934 – 938 (1988). 276. J. S. Adams, O. P. Sharma, M. M. Diz, and D. B. Endres, Ketoconazole decreases the serum 1,25-dihydroxyvitamin D and calcium concentration in sarcoidosis-associated hypercalcemia. J. Clin. Endocrinol. Metab. 70, 1090 – 1095 (1990). 277. N. A. Breslau, G. M. Perminger, B. V. Adams, J. Otey, and C. Y. C. Pak, Use of ketoconazole to probe the pathogenetic importance of 1,25-dihydroxyvitamin D in absorptive hypercalciuria. J. Clin. Endocrinol. Metab. 75, 1446 – 1452 (1992).

CHAPTER 9 Vitamin D 278. O. P. Sharma, Vitamin D, calcium, and sarcoidosis. Chest 109, 535 – 539 (1996). 279. D. Feldman, Ketoconazole and other imidazole derivatives as inhibitors of steroidogenesis. Endocr. Rev. 7, 409 – 420 (1986). 280. D. S. Loose, P. B. Kan, M. A. Hirst, R. A. Marcus, and D. F. Feldman, Ketoconazole blocks adrenal steroidogenesis by inhibiting cytochrome P450-dependent enzymes. J. Clin. Invest. 71, 1495 – 1499 (1983). 281. H. L. Henry, Effect of ketoconazole and miconazole on 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells. J. Steroid Biochem. 23, 991 – 994 (1985). 282. A. S. Douglas, S. Ali, and S. S. Bakhshi, Does vitamin D deficiency account for ethnic differences in tuberculosis seasonality in the UK? Ethn. Health 3, 247 – 253 (1998). 283. R. J. Wilkinson, M. Llewelyn, Z. Toossi, P. Patel, G. Pasvol, A. Lalvani, D. Wright, M. Latif, and R. N. Davidson, Influence of vitamin D deficiency and vitamin D receptor polymorphisms on tuberculosis among Gujarati Asians in west London: A case-control study. Lancet 355, 618 – 621 (2000). 284. K. A. Rockett, R. Brookes, I. Udalova, V. Vidal, A. V. Hill, and D. Kwiatkowski, 1,25-Dihydroxyvitamin D3 induces nitric oxide synthase and suppresses growth of Mycobacterium tuberculosis in a human macrophage-like cell line. Infect. Immun. 66, 5314 – 5321 (1998). 285. S. Ashkar, G. F. Weber, V. Panoutsakopoulou, M. E. Sanchirico, M. Jansson, S. Zawaideh, S. R. Rittling, D. T. Denhardt, M. J. Glimcher, and H. Cantor, Eta-1 (osteopontin): An early component of type-1 (cell-mediated) immunity. Science 287, 860 – 864 (2000). 286. E. B. Mawer, M. E. Hayes, P. E. Still, M. Davies, G. A. Lumb, J. Palit, and P. J. Holt, Evidence for nonrenal synthesis of 1,25-dihydroxyvitamin D in patients with inflammatory arthritis. J. Bone Miner. Res. 6, 733 – 739 (1991). 287. S. J. Smith, M. E. Hayes, P. L. Selby, and E. B. Mawer, Autocrine control of vitamin D metabolism in synovial cells from arthritic patients. Ann. Rheum. Dis. 58, 372 – 378 (1999). 288. R. Skowronski, D. Peehl, and D. Feldman, Vitamin D and prostate cancer: 1,25-Dihydroxyvitamin D3 receptors and actions in prostate cancer cell lines. Endocrinology 132, 1952 – 1960 (1993). 289. D. M. Peehl, R. J. Skowronski, G. K. Leung, S. T. Wong, T. A. Stamey, and D. Feldman, Antiproliferative effects of 1,25-dihydroxy-vitamin D3 on primary cultures of human prostatic cells. Cancer Res. 54, 805 – 810 (1994). 290. G. G. Schwartz, T. A. Oeler, M. R. Uskokovic, and R. R. Bahnson, Human prostate cancer cells: Inhibition of proliferation by vitamin D analogs. Anticancer Res. 14, 1077 – 1081 (1994). 291. R. J. Skowronski, D. M. Peehl, and D. Feldman, Actions of vitamin D3 analogs on human prostate cancer cell lines: Comparison with 1,25-dihydroxyvitamin D3. Endocrinology 136, 20 – 26 (1995). 292. B. R. Konety, G. G. Schwartz, J. S. Acierno, Jr., M. J. Becich, and R. H. Getzenberg, The role of vitamin D in normal prostate growth and differentiation. Cell Growth Differ. 7, 1563 – 1570 (1996). 293. G. K. Fu, A. A. Portale, and W. L. Miller, Complete structure of the human gene for the vitamin D 1-hydroxylase, P450c1. DNA Cell. Biol. 16, 1499 – 1507 (1997). 294. D. D. Bikle, E. Gee, and S. Pillai, Regulation of keratinocyte growth, differentiation, and vitamin D metabolism by analogs of 1,25-dihydroxyvitamin D. J. Invest. Dermatol. 101, 713 – 718 (1993). 295. K. Kragballe, Psoriasis and other skin diseases. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1213 – 1225. Academic Press, San Diego, 1997. 296. J. Reichrath, A. Perez, S. M. Muller, T. C. Chen, A. Kerber, F. A. Bahmer, and M. F. Holick, Topical calcitriol (1,25-dihydroxyvitamin D3) treatment of psoriasis: An immunohistological evaluation. Acta Derm. Venereol. 77, 268 – 272 (1997).

297 297. J. Reichrath, S. M. Muller, A. Kerber, H. P. Baum, and F. A. Bahmer, Biologic effects of topical calcipotriol (MC 903) treatment in psoriatic skin. J. Am. Acad. Dermatol. 36, 19 – 28 (1997). 298. R. St-Arnaud and F. H. Glorieux, Vitamin D and bone development. In “Vitamin D” (D. Feldman, F. H. Gloriux, and J. W. Pike, eds.), pp. 293 – 304. Academic Press, San Diego, 1997. 299. M. K. Drezner, Clinical disorders of phosphate homeostasis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 733 – 753. Academic Press, San Diego, 1997. 300. T. O. Carpenter and K. L. Insogna, The hypocalcemic disorders: Differential diagnosis and therapeutic use of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 923 – 936. Academic Press, San Diego, 1997. 301. R. Bouillon, The many faces of rickets. N. Engl. J. Med. 338, 681 – 682 (1998). 302. W. L. Miller and A. A. Portale, Genetic disorders of vitamin D biosynthesis. Endocrinol. Metab. Clin. North Am. 28, 825 – 840 (1999). 303. M. K. Drezner, PHEX gene and hypophosphatemia. Kidney Int. 57, 9 – 18 (2000). 304. V. A. Prader, R. Illig, and E. Heierli, Eine besondere form der primaren vitamin-D-resistenten rachitis mit hypocalcamie und autosomal-dominantem erbgang: Die hereditare pseudo-mangelrachitis. Helv. Paediatr. Acta. 16, 452 – 468 (1961). 305. F. H. Glorieux and R. St-Arnaud, Molecular cloning of (25-OH D)1 -hydroxylase: An approach to the understanding of vitamin D pseudo-deficiency. Recent Prog. Horm. Res. 53, 341 – 349; discussion 350 (1998). 306. S. Kato, Genetic mutation in the human 25-hydroxyvitamin D3 1-hydroxylase gene causes vitamin D-dependent rickets type I. Mol. Cell. Endocrinol. 156, 7 – 12 (1999). 307. F. H. Glorieux and R. St-Arnaud, Vitamin D pseudodeficiency. In “Vitamin D” (D. Feldman, F. H. Glorieux, and J. W. Pike, eds.), pp. 755 – 764. Academic Press, San Diego, 1997. 308. S. Kato, J. Yanagisawa, A. Murayama, S. Kitanaka, and K. Takeyama, The importance of 25-hydroxyvitamin D3 1hydroxylase gene in vitamin D-dependent rickets. Curr. Opin. Nephrol. Hypertens. 7, 377 – 383 (1998). 309. D. Feldman and P. J. Malloy, Hereditary 1,25-dihydroxyvitamin D resistant rickets: Molecular basis and implications for the role of 1,25(OH)2D3 in normal physiology. Mol. Cell. Endocrinol. 72, C57 – C62 (1990). 310. M. R. Hughes, P. J. Malloy, B. W. OMalley, J. W. Pike, and D. Feldman, Genetic defects of the 1,25-dihydroxyvitamin D3 receptor. J. Recept. Res. 11, 699 – 716 (1991). 311. P. J. Malloy and D. Feldman, “Molecular Defects in the Vitamin D Receptor Associated with Hereditary 1,25-Dihydroxyvitamin D Resistant Rickets” (M. Holick, ed.). Humana Press, 1998. 312. P. J. Malloy and D. Feldman, Vitamin D resistance. Am. J. Med. 106, 355 – 370 (1999). 313. M. R. Hughes, P. J. Malloy, D. G. Kieback, R. A. Kesterson, J. W. Pike, D. Feldman, and B. W. OMalley, Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242, 1702 – 1705 (1988). 314. M. A. Hirst, H. I. Hochman, and D. Feldman, Vitamin D resistance and alopecia: A kindred with normal 1,25-dihydroxyvitamin D binding, but decreased receptor affinity for deoxyribonucleic acid. J. Clin. Endocrinol. Metab. 60, 490 – 495 (1985). 315. P. J. Malloy, Z. Hochberg, J. W. Pike, and D. Feldman, Abnormal binding of vitamin D receptors to deoxyribonucleic acid in a kindred with vitamin D-dependent rickets, type II. J. Clin. Endocrinol. Metab. 68, 263 – 269 (1989). 316. T. Sone, R. A. Scott, M. R. Hughes, P. J. Malloy, D. Feldman, B. W. O’Malley, and J. W. Pike, Mutant vitamin D receptors which confer hereditary resistance to 1,25-dihydroxyvitamin D3 in humans are transcriptionally inactive in vitro. J. Biol. Chem. 264, 20230 –20234 (1989).

298 317. T. Sone, S. J. Marx, U. A. Liberman, and J. W. Pike, A unique point mutation in the human vitamin D receptor chromosomal gene confers hereditary resistance to 1,25-dihydroxyvitamin D3. Mol. Endocrinol. 4, 623 – 631 (1990). 318. H. Yagi, K. Ozono, H. Miyake, K. Nagashima, T. Kuroume, and J. W. Pike, A new point mutation in the deoxyribonucleic acidbinding domain of the vitamin D receptor in a kindred with hereditary 1,25-dihydroxyvitamin D-resistant rickets. J. Clin. Endocrinol. Metab. 76, 509 – 512 (1993). 319. P. J. Malloy, Y. Weisman, and D. Feldman, Hereditary 1,25dihydroxyvitamin D-resistant rickets resulting from a mutation in the vitamin D receptor deoxyribonucleic acid-binding domain. J. Clin. Endocrinol. Metab. 78, 313 – 316 (1994). 320. H. H. Ritchie, M. R. Hughes, E. T. Thompson, P. J. Malloy, Z. Hochberg, D. Feldman, J. W. Pike, and B. W. OMalley, An ochre mutation in the vitamin D receptor gene causes hereditary 1,25-dihydroxyvitamin D3-resistant rickets in three families. Proc. Natl. Acad. Sci. USA 86, 9783 – 9787 (1989). 321. P. J. Malloy, Z. Hochberg, D. Tiosano, J. W. Pike, M. R. Hughes, and D. Feldman, The molecular basis of hereditary 1,25-dihydroxyvitamin D3 resistant rickets in seven related families. J. Clin. Invest. 86, 2071 – 2079 (1990). 322. R. J. Wiese, H. Goto, J. M. Prahl, S. J. Marx, M. Thomas, A. al-Aqeel, and H. F. DeLuca, Vitamin D-dependency rickets type II: Truncated vitamin D receptor in three kindreds. Mol. Cell. Endocrinol. 90, 197 – 201 (1993). 323. G. K. Whitfield, S. H. Selznick, C. A. Haussler, J. C. Hsieh, M. A. Galligan, P. W. Jurutka, P. D. Thompson, S. M. Lee, J. E. Zerwekh, and M. R. Haussler, Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D3: Point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol. Endocrinol. 10, 1617 – 1631 (1996). 324. M. Hewison, A. R. Rut, K. Kristjansson, R. E. Walker, M. J. Dillon, M. R. Hughes, and J. L. ORiordan, Tissue resistance to 1,25-dihydroxyvitamin D without a mutation of the vitamin D receptor gene. Clin. Endocrinol. 39, 663 – 670 (1993). 325. M. I. New, S. Nimkarn, D. D. Brandon, S. Cunningham-Rundles, R. C. Wilson, R. S. Newfield, J. Vandermeulen, N. Barron, C. Russo, D. L. Loriaux, and B. OMalley, Resistance to several steroids in two sisters. J. Clin. Endocrinol. Metab. 84, 4454 (1999). 326. Y. Weisman, N. Jaccard, C. Legum, Z. Spirer, G. Yedwab, L. Even, S. Edelstein, A. M. Kaye, and Z. Hochberg, Prenatal diagnosis of vitamin D-dependent rickets, type II: Response to 1,25-dihydroxyvitamin D in amniotic fluid cells and fetal tissues. J. Clin. Endocrinol. Metab 71, 937 – 943 (1990). 327. Y. Weisman, P. J. Malloy, A. V. Krishnan, N. Jaccard, D. Feldman, and Z. Hochberg, In “Ninth workshop on Vitamin D,” p. 106 (1994). 328. S. Balsan, M. Garabedian, M. Larchet, A. M. Gorski, G. Cournot, C. Tau, A. Bourdeau, C. Silve, and C. Ricour, Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J. Clin. Invest. 77, 1661 – 1667 (1986). 329. Z. Hochberg, D. Tiosano, and L. Even, Calcium therapy for calcitriol-resistant rickets. J. Pediatr. 121, 803 – 808 (1992). 330. A. al-Aqeel, P. Ozand, S. Sobki, W. Sewairi, and S. Marx, The combined use of intravenous and oral calcium for the treatment of vitamin D dependent rickets type II (VDDRII). Clin. Endocrinol. 39, 229 – 237 (1993). 331. T. Yoshizawa, Y. Handa, Y. Uematsu, S. Takeda, K. Sekine, Y. Yoshihara, T. Kawakami, K. Arioka, H. Sato, Y. Uchiyama, S. Masushige, A. Fukamizu, T. Matsumoto, and S. Kato, Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nature Genet. 16, 391 – 396 (1997).

FELDMAN, MALLOY, AND GROSS 332. Y. C. Li, A. E. Pirro, M. Amling, G. Delling, R. Baron, R. Bronson, and M. B. Demay, Targeted ablation of the vitamin D receptor: An animal model of vitamin D-dependent rickets type II with alopecia. Proc. Natl. Acad. Sci. USA 94, 9831 – 9835, (1997). 333. M. Amling, M. Priemel, T. Holzmann, K. Chapin, J. M. Rueger, R. Baron, and M. B. Demay, Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: Formal histomorphometric and biomechanical analyses. Endocrinology 140, 4982 – 4987 (1999). 334. K. Kinuta, H. Tanaka, T. Moriwake, K. Aya, S. Kato, and Y. Seino, Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology 141, 1317 – 1324 (2000). 335. S. Tanaka, M. Haji, R. Takayanagi, Y. Sugioka, and H. Nawata, 1,25Dihydroxyvitamin D3 enhances the enzymatic activity and expression of the messenger ribonucleic acid for aromatase cytochrome P450 synergistically with dexamethasone depending on the vitamin D receptor level in cultured human osteoblasts. Endocrinology 137, 1860 – 1869 (1996). 336. F. Francis, T. M. Strom, S. Hennig, A. Boddrich, B. Lorenz, O. Brandau, K. L. Mohnike, M. Cagnoli, C. Steffens, S. Klages, K. Borzym, T. Pohl, C. Oudet, M. J. Econs, P. S. Rowe, R. Reinhardt, T. Meitinger, and H. Lehrach, Genomic organization of the human PEX gene mutated in X-linked dominant hypophosphatemic rickets. Genome Res. 7, 573 – 585 (1997). 337. P. S. Rowe, C. L. Oudet, F. Francis, C. Sinding, S. Pannetier, M. J. Econs, T. M. Strom, T. Meitinger, M. Garabedian. A. David, M. A. Macher, E. Questiaux, E. Popowska, E. Pronicka, A. P. Read, A. Mokrzycki, F. H.Glorieux, M. K. Drezner, A. Hanauer, H. Lehrach, J. N. Goulding, and J. L. ORiordan, Distribution of mutations in the PEX gene in families with X-linked hypophosphataemic rickets (HYP). Hum. Mol. Genet. 6, 539 – 549 (1997). 338. Q. Cai, S. F. Hodgson, P. C. Kao, V. A. Lennon, G. G. Klee, A. R. Zinsmiester, and R. Kumar, Brief report: Inhibition of renal phosphate transport by a tumor product in a patient with oncogenic osteomalacia. N. Engl. J. Med. 330, 1645 – 1649 (1994). 339. J. A. Eisman, Genetics of osteoporosis. Endocr. Rev. 20, 788 – 804 (1999). 340. J. M. Zmuda, J. A. Cauley, and R. E. Ferrell, Recent progress in understanding the genetic susceptibility to osteoporosis. Genet. Epidemiol. 16, 356 – 367 (1999). 341. A. G. Uitterlinden, H. Burger, Q. Huang, E. Odding, C. M. van Duijn, A. Hofman, J. C. Birkenhager, J. P. T. M. van Leeuwen, and H. A. P. Pols, Vitamin D receptor genotype is associated with radiographic osteoarthritis at the knee. J. Clin. Invest. 100, 259 – 263 (1997). 342. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 343. C. Gross, T. R. Eccleshall, and D. Feldman, Vitamin D receptor gene alleles and osteoporosis. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 917 – 933. Academic Press, San Diego, 1996. 344. G. S. Cooper and D. M. Umbach, Are vitamin D receptor polymorphisms associated with bone mineral density? A meta-analysis. J. Bone Miner. Res. 11, 1841 – 1849 (1996). 345. G. Gong, H. S. Stern, S. C. Cheng, N. Fong, J. Mordeson, H. W. Deng, and R. R. Recker, The association of bone mineral density with vitamin D receptor gene polymorphisms. Osteoporos. Int. 9, 55 – 64 (1999). 346. G. Howard, T. Nguyen, N. Morrison, T. Watanabe, P. Sambrook, J. Eisman, and P. Kelly, Genetic influences on bone density: Physiological correlates of vitamin D receptor gene alleles in premenopausal women. Notification of genotype corrections. J. Clin. Endocrinol. Metab. 83, 1043 (1998).

CHAPTER 9 Vitamin D 347. C. Gross, T. R. Eccleshall, P. J. Malloy, M. L. Villa, R. Marcus, and D. Feldman, The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal Mexican-American women. J. Bone Miner. Res. 11, 1850 – 1855 (1996). 348. S. S. Harris, T. R. Eccleshall, C. Gross, B. Dawson-Hughes, and D. Feldman, The vitamin D receptor start codon polymorphism (FokI) and bone mineral density in premenopausal American black and white women. J. Bone Miner. Res. 12, 1043 – 1048 (1997). 349. L. Gennari, L. Becherini, R. Mansani, L. Masi, A. Falchetti, A. Morelli, E. Colli, S. Gonnelli, C. Cepollaro, and M. L. Brandi, FokI polymorphism at translation initiation site of the vitamin D receptor gene predicts bone mineral density and vertebral fractures in postmenopausal Italian women. J. Bone Miner. Res. 14, 1379 – 1386 (1999). 350. S. Ferrari, D. Manen, J. P. Bonjour, D. Slosman, and R. Rizzoli, Bone mineral mass and calcium and phosphate metabolism in young men: Relationships with vitamin D receptor allelic polymorphisms. J. Clin. Endocrinol. Metab. 84, 2043 – 2048 (1990). 351. T. R. Eccleshall, P. Garnero, C. Gross, P. D. Delmas, and D. Feldman, Lack of correlation between the start codon polymorphism of the vitamin D receptor and bone mineral density in premenopausal French women. J. Bone Miner. Res. 13, 31 – 35 (1998). 352. S. K. Ames, K. J. Ellis, S. K. Gunn, K. C. Copeland, and S. A. Abrams, Vitamin D receptor gene Fokl polymorphism predicts calcium absorption and bone mineral density in children. J. Bone Miner. Res. 14, 740 – 746 (1999). 353. D. Feldman, Androgen and vitamin D receptor gene polymorphisms: The long and short of prostate cancer risk. J. Natl. Cancer Inst. 89, 109 – 111 (1997). 354. S. A. Ingles, R. K. Ross, M. C. Yu, R. A. Irvine, G. La Pera, R. W. Haile, and G. A. Coetzee, Association of prostate cancer with genetic polymorphisms in vitamin D receptor and androgen receptor. Cancer Inst. 89, 166 – 170 (1997). 355. J. A. Taylor, A. Hirvonen, M. Watson, G. Pittman, J. L. Mohler, and D. A. Bell, Association of prostate cancer with vitamin D receptor gene polymorphism. Cancer Res. 56, 4108 – 4110 (1990). 356. A. S. Kibel, S. D. Isaacs, W. B. Isaacs, and G. S. Bova, Vitamin D receptor polymorphisms and lethal prostate cancer. J. Urol. 160, 1405 – 1409 (1998). 357. J. E. Curran, T. Vaughan, R. A. Lea, S. R. Weinstein, N. A. Morrison, and L. R. Griffiths, Association of A vitamin D receptor polymorphism with sporadic breast cancer development. Int. J. Cancer 83, 723 – 726 (1999). 358. A. C. Lundin, P. Soderkvist, B. Eriksson, M. Bergman-Jungestrom, and S. Wingren, Association of breast cancer progression with a vitamin D receptor gene polymorphism, South-East Sweden Breast Cancer Group. Cancer Res. 59, 2332 – 2334 (1999). 359. M. Ruggiero, S. Pacini, M. Amato, S. Aterini, and V. Chiarugi, Association between vitamin D receptor gene polymorphism and nephrolithiasis. Miner. Electrolyte Metab. 25, 185 – 190 (1999). 360. S. A. Ingles, D. G. Garcia, W. Wang, A. Nieters, B. E. Henderson, L. N. Kolonel, R. W. Haile, and G. A. Coetzee, Vitamin D receptor genotype and breast cancer in Latinas (United States). Cancer Causes Control 11, 25 – 30 (2000). 361. R. Bouillon, W. H. Okamura, and A. W. Norman, Structure-function relationships in the vitamin D endocrine system. Endocr. Rev. 16, 200 – 257 (1995). 362. K. M. Casteels, C. Mathieu, M. Waer, D. Valckx, L. Overbergh, J. M. Laureys, and R. Bouillon, Prevention of type I diabetes in nonobese diabetic mice by late intervention with nonhypercalcemic analogs of 1,25-dihydroxyvitamin D3 in combination with a short induction course of cyclosporin A. Endocrinology 139, 95 – 102 (1998).

299 363. A. J. Brown and E. Slatopolsky, Vitamin D analogs: Perspectives for treatment. Miner. Electrolyte Metab. 25, 337 – 341 (1999). 364. E. Slatopolsky, A. Dusso, and A. Brown, New analogs of vitamin D3. Kidney Int. Suppl. 73, S46 – 51 (1999). 365. A. Verstuyf, L. Verlinden, S. Segaert, E. Van Etten, C. Mathieu, and R. Bouillon, Nonclassical effects of l, 25-dihydroxyvitamin D3 and its analogs. Miner. Electrolyte Metab. 25, 345 – 348 (1999). 366. J. A. Kanis, Vitamin D analogs: From renal bone disease to osteoporosis. Kidney Int. Suppl. 73, S77 – 81 (1999). 367. M. Quack and C. Carlberg, Selective recognition of vitamin D receptor conformations mediates promoter selectivity of vitamin D analogs. Mol. Pharmacol. 55, 1077 – 1087 (1999). 368. M. R. Uskokovic, G. P. Studzinski, and S. G. Reddy, The 16-ene vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1045 – 1070. Academic Press, San Diego, 1997. 369. L. Binderup, E. Binderup, and W. O. Godtfredsen, Development of new vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1027 – 1043. Academic Press, San Diego 1997. 370. N. Kubodera, K. Sato, and Y. Nishii, Characteristics of 22-oxacalcitriol (OCT) and 2-(3-hydroxypropoxy)-calcitriol (ED-71). In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1071 – 1086. Academic Press, San Diego 1997. 371. W. H. Okamura and G. D. Zhu, Chemistry and design: Structural biology of vitamin D action. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 939 – 971. Academic Press, San Diego 1997. 372. S. Peleg, C. Nguyen, B. T. Woodard, J. K. Lee, and G. H. Posner, Differential use of transcription activation function 2 domain of the vitamin D receptor by 1,25-dihydroxyvitamin D3 and its A ring-modified analogs. Mol. Endocrinol. 12, 525 – 535 (1998). 373. R. Bouillon, A. Verstuyf, J. Zhao, B. K. Tan, and H. Van Baelen, Nonhypercalcemic vitamin D analogs: Interactions with the vitamin D-binding protein. Horm. Res. 45, 117 – 121 (1996). 374. A. J. Brown and E. Slatopolsky, Mechanisms for the selective actions of vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 995 – 1009. Academic Press, San Diego, 1997. 375. F. J. Dilworth, G. R. Williams, A. M. Kissmeyer, J. L. Nielsen, E. Binderup, M. J. Calverley, H. L. Makin, and G. Jones, The vitamin D analog, KH1060, is rapidly degraded both in vivo and in vitro via several pathways: Principal metabolites generated retain significant biological activity. Endocrinology 138, 5485 – 5496 (1997). 376. X. Y. Zhao, T. R. Eccleshall, A. V. Krishnan, C. Gross, and D. Feldman, Analysis of vitamin D analog-induced heterodimerization of vitamin D receptor with retinoid X receptor using the yeast twohybrid system. Mol. Endocrinol. 11, 366 – 378 (1997). 377. W. Yang and L. P. Freeman, 20-Epi analogues of 1,25-dihydroxyvitamin D3 are highly potent inducers of DRIP coactivator complex binding to the vitamin D3 receptor. J. Biol. Chem. 274, 16838 – 16845 (1999). 378. S. Peleg, M. Sastry, E. D. Collins, J. E. Bishop, and A. W. Norman, Distinct conformational changes induced by 20-epi analogues of 1,25-dihydroxyvitamin D3 are associated with enhanced activation of the vitamin D receptor. J. Biol. Chem. 270, 10551 – 10558 (1995). 379. S. Peleg, Molecular basis for differential action of vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1011 – 1025. Academic Press, San Diego 1997. 380. T. Jaaskelainen, S. Ryhanen, A. Mahonen, H. F. DeLuca, and P. H. Maenpaa, Mechanism of action of superactive vitamin D analogs through regulated receptor degradation. J. Cell. Biochem. 76, 548 – 558 (2000). 381. A. Verstuyf, L. Verlinden, H. Van Baelen, K. Sabbe, C. D’Hallewyn, P. De Clercq, M. Vandewalle, and R. Bouillon, The biological activity

300

382.

383.

384.

385.

386.

387.

388.

389.

390.

391.

392.

393. 394.

395.

396.

FELDMAN, MALLOY, AND GROSS of nonsteroidal vitamin D hormone analogs lacking both the C- and D-rings. J. Bone Miner. Res. 13, 549 – 558 (1998). T. C. Polek, S. Murthy, S. E. Blutt, E. A. Allegretto, and N. Wrigel, Vitamin D receptor (VDR) agonists inhibit prostate cancer cell growth in vitro and in a nude mouse tumor model. American Assoc. Cancer Research Annual Meeting, San Francisco, CA, April 1 – 5, 2000, Abstract 27/2, p. 426. K. W. Colston, A. G. Mackay, S. Y. James, L. Binderup, S. Chander, and R. C. Coombes, EB-1089: A new vitamin D analogue that inhibits the growth of breast cancer cells in vivo and in vitro. Biochem. Pharmacol. 44, 2273 – 2280 (1992). M. Haq, R. Kremer, D. Goltzman, and S. A. Rabbani, A vitamin D analogue (EB1089) inhibits parathyroid hormone-related peptide production and prevents the development of malignancy-associated hypercalcemia in vivo. J. Clin. Invest. 91, 2416 – 2422 (1993). B. L. Lokeshwar, G. G. Schwartz, M. G. Selzer, K. L. Burnstein, S. H. Zhuang, N. L. Block, and L. Binderup, Inhibition of prostate cancer metastasis in vivo: A comparison of 1,25-dihydroxyvitamin D (calcitriol) and EB1089. Cancer Epidemiol. Biomarkers Prev. 8, 241 – 248 (1999). W. H. Park, J. G. Seol, E. S. Kim, C. W. Jung, C. C. Lee, L. Binderup, H. P. Koeffler, B. K. Kim, and Y. Y. Lee, Cell cycle arrest induced by the vitamin D3 analog EB1089 in NCI-H929 myeloma cells is associated with induction of the cyclin-dependent kinase inhibitor p27. Exp. Cell Res. 254, 279 – 286 (2000). T. Gulliford, J. English, K. W. Colston, P. Menday, S. Moller, and R. C. Coombes, A phase I study of the vitamin D analogue EB 1089 in patients with advanced breast and colorectal cancer. Br. J. Cancer 78, 6 – 13 (1996). T. R. J. Evans, K. J. Hamberg, T. Skov, H. L. O. Haahr, P. Menday, C. Bay, and L. Binderup. Seocalcitol (EB 1089)-clinical experience to date. In “Vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 485 – 488, University of California, Printing and Reprographics, Riverside, 2000. J. Lemire, The role of vitamin D3 in immunosuppression: Lessons from autoimmunity and transplantation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1167 – 1181. Academic Press, San Diego, 1997. J. Abe, M. Morikawa, K. Miyamoto, S. Kaiho, M. Fukushima, C. Miyaura, E. Abe, T. Suda, and Y. Nishii, Synthetic analogues of vitamin D3 with an oxygen atom in the side chain skeleton: A trial of the development of vitamin D compounds which exhibit potent differentiation-inducing activity without inducing hypercalcemia. FEBS Lett. 226, 58 – 62 (1987). K. Colston, Vitamin D and brest cancer: Therapeutic potential of new vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1107 – 1123. Academic Press, San Diego, 1997. C. Gross, D. M. Peehl, and D. Feldman, Vitamin D and prostate cancer. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1125 – 1139. Academic Press, San Diego, 1997. G. J. Miller, Vitamin D and prostate cancer: Biologic interactions and clinical potentials. Cancer Metastasis Rev. 17, 353 – 360 (1998). J. Welsh, K. Van Weelden, L. Flanagan, I. Byrne, E. Nolan, and C. J. Narvaez (1998). The role of vitamin D3 and antiestrogens in modulating apoptosis of breast cancer cells and tumors. Subcell. Biochem. 30, 245 – 270. B. R. Konety, C. S. Johnson, D. L. Trump, and R. H. Getzenberg, Vitamin D in the prevention and treatment of prostate cancer. Semin. Urol. Oncol. 17, 77 – 84 (1999). H. Hofer, G. Ho, M. Peterlik, M. R. Uskokovic, J. K. Lee, M. C. White, G. H. Posner, and H. S. Cross, Biological effects of 1hydroxy- and 1-(hydroxymethyl)-vitamin D compounds relevant for potential colorectal cancer therapy. J. Pharmacol. Exp. Ther. 291, 450 – 455 (1999).

397. D. Feldman, X. Y. Zhao, and A. V. Krishnan, Vitamin D and prostate cancer. Endocrinology 141, 5 – 9 (2000). 398. D. Miura, K. Manabe, K. Ozono, M. Saito, Q. Gao, A. W. Norman, and S. Ishizuka, Antagonistic action of novel 1,25-dihydroxyvitamin D3-26, 23- lactone analogs on differentition of human leukemia cells (HL-60) induced by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 274, 16392 – 16399 (1999). 399. K. Ozono, M. Saito, D. Miura, T. Michigami, S. Nakajima, and S. Ishizuka, Analysis of the molecular mechanism for the antagonistic action of a novel 1,25-dihydroxyvitamin D3 analogue toward vitamin D receptor function. J. Biol. Chem. 274, 32376 – 32381 (1999). 400. S. D. Chan and D. Atkins, The temporal distribution of the 1,25dihydroxycholecalciferol receptor in the rat jejunal villus. Clin. Sci. 67, 285 – 290 (1984). 401. M. A. Hirst and D. Feldman, 1,25-Dihydroxyvitamin D3 receptors in mouse colon. J. Steroid Biochem. 14, 315 – 319 (1981). 402. P. Lointier, F. Meggouh, D. Pezet, M. Dapoigny, P. N. Dieng, S. Saez, and J. Chipponi, Specific receptors for 1,25-dihydroxyvitamin D3 (1,25-DR) and human colorectal carcinogenesis. Anticancer Res. 9, 1921 – 1924 (1989). 403. P. Lointier, F. Meggouh, P. Dechelotte, D. Pezet, C. Ferrier, J. Chipponi, and S. Saez, 1,25-Dihydroxyvitamin D3 receptors and human colon adenocarcinoma. Br. J. Surg. 78, 435 – 439 (1991). 404. W. M. Tong, G. Bises, Y. Sheinin, A. Ellinger, D. Genser, R. Potzi, F. Wrba, E. Wenzl, R. Roka, N. Neuhold, M. Peterlik, and H. S. Cross, Establishment of primary cultures from human colonic tissue during tumor progression: Vitamin D responses and vitamin-Dreceptor expression. Int. J. Cancer 75, 467 – 472 (1998). 405. R. H. Wasserman, R. A. Corradino, C. S. Fullmer, and A. N. Taylor, Some aspects of vitamin D action; calcium absorption and the vitamin D-dependent calcium-binding protein. Vitam. Horm. 32, 299 – 324 (1974). 406. S. Christakos, J. D. Beck, and S. J. Hyllner, Calbindin-D28K. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 209–221. Academic Press, San Diego 1997. 407. M. Thomasset, Calbindin-D9K. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 223 – 232. Academic Press, San Diego, 1997. 408. S. Kowarski, L. A. Cowen, M. T. Takahashi, and D. Schachter, Tissue distribution and vitamin D dependence of IMCAL in the rat. Am. J. Physiol. 253, G411 – 419 (1987). 409. R. H. Wasserman, J. S. Chandler, S. A. Meyer, C. A. Smith, M. E. Brinkak, C. S. Fullmer, J. T. Penniston, and R. Kumar, Intestinal calcium transport and calcium extrusion processes at the basolateral membrane. J. Nutr. 122, 662 – 671 (1992). 410. T. Suda, T. Shinki, and N. Takahashi, The role of vitamin D in bone and intestinal cell differentiation. Annu. Rev. Nutr. 10, 195 – 211 (1990). 411. R. P. Heaney, M. J. Barger-Lux, M. S. Dowell, T. C. Chen, and M. F. Holik, Calcium absorptive effects of vitamin D and its major metabolites. J. Clin. Endocrinol. Metab. 82, 4111 – 4116 (1997). 412. R. P. Heaney, Lessons for nutritional science from vitamin D. Am. J. Clin. Nutr. 69, 825 – 826 (1999). 413. B. L. Riggs, H. W. Wahner, L. D. Melton, L. S. Richelson, H. L. Judd, and W. M. O’Fallon, Dietary calcium intake and rates of bone loss in women. J. Clin. Invest. 80, 979 – 982 (1987). 414. B. P. Halloran and A. A. Portale, Vitamin D metabolism: The effects of aging. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 541 – 554. Academic Press, San Diego, 1997. 415. H. J. Armbrecht, M. A. Boltz, S. Christakos, and M. E. Bruns, Capacity of 1,25-dihydroxyvitamin D to stimulate expression of calbindin D changes with age in the rat. Arch. Biochem. Biophys. 352, 159 – 164 (1998). 416. P. R. Ebeling, M. E. Sandgren, E. P. DiMagno, A. W. Lane, H. F. DeLuca, and B. L. Riggs, Evidence of an age-related decrease in

CHAPTER 9 Vitamin D

417.

418.

419.

420.

421.

422.

423.

424.

425.

426.

427.

428.

429.

430.

431.

432.

433.

intestinal responsiveness to vitamin D: relationship between serum 1,25-dihydroxyvitamin D3 and intestinal vitamin D receptor concentrations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). M. J. Barger-Lux, R. P. Heaney, S. J. Lanspa, J. C. Healy, and H. F. DeLuca, An investigation of sources of variation in calcium absorption efficiency [published erratum appears in J. Clin. Endocrinol. Metab. 80(7), 2068 (1995)]. J. Clin. Endocrinol. Metab. 80, 406 – 411 (1995). E. M. Colin, G. J. Van Den Bemd, M. Van Aken, S. Christakos, H. R. De Jonge, H. F. Deluca, J. M. Prahl, J. C. Birkenhager, C. J. Buurman, H. A. Pols, and J. P. Van Leeuwen, Evidence for involvement of 17beta-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the Rat. J. Bone Miner. Res. 14, 57 – 64 (1999). M. J. Favus and F. L. Coe, Idiopathic hypercalciuria and nephrolithiasis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 867 – 881. Academic Press, San Diego 1997. J. Yao, P. Kathpalia, D. A. Bushinsky, and M. J. Favus, Hyperresponsiveness of vitamin D receptor gene expression to 1,25-dihydroxyvitamin D3: A new characteristic of genetic hypercalciuric stoneforming rats. J. Clin. Invest. 101, 2223 – 2232 (1995). A. M. Parfitt, Vitamin D and the pathogenesis of rickets and osteomalacia. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 645 – 662. Academic Press, San Diego, 1997. B. E. Kream, M. Jose, S. Yamada, and H. F. DeLuca, A specific high-affinity binding macromolecule for 1,25-dihydroxyvitamin D3 in fetal bone. Science 197, 1086 – 1088 (1977). T. L. Chen, M. A. Hirst, and D. Feldman, A receptor-like binding macromolecule for 1,25-dihydroxycholecalciferol in cultured mouse bone cells. J. Biol. Chem. 254, 7491 – 7494 (1979). J. E., Aubin and J. N. M. Heersche, Vitamin D and osteoblasts. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 313 – 328. Academic Press, San Diego, 1997. B. E. Kream and A. C. Lichtler, Vitamin D regulation of type I collagen synthesis and gene expression in bone. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 201 – 207. Academic Press, San Diego, 1997. L. G., Raisz, C. L. Trummel, M. F. Holick, and H. F. DeLuca, 1,25dihydroxycholecalciferol: A potent stimulator of bone resorption in tissue culture. Science 175, 768 – 769 (1972). T. Suda and N. Takahashi, Vitamin D and osteoclastogenesis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 329 – 340. Academic Press, San Diego, 1997. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). K. W. Colston and D. Feldman, Demonstration of 1, 25-dihydroxycholecalciterol cytoplasmic receptor-like binder in mouse kidney cytosol. J. Clin. Endocrinol. 49, 798 – 800 (1979). S. D. Antrobus and M. R. Walters, 1,25-Dihydroxyvitamin D effects in the kidney: Induction of calmodulin binding proteins. Biochem. Biophys. Res. Commun. 185, 636 – 640 (1992). R. Kumar, Vitamin D and the kidney. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 275 – 292. Academic Press, San Diego, 1997. J. Silver and T. Naveh-Many, Vitamin D and the parathyroid glands. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 353 – 367. Academic Press, San Diego, 1997. N. Fokuda, H. Tanaka, Y. Tominaga, M. Fukagawa, K. Kurokawa, and Y. Seino, Decreased 1,25-dihydroxyvitamin D3 receptor density is associated with a more severe form of parathyroid hyperplasia in chronic uremic patients. J. Clin. Invest. 92, 1436 – 1443 (1993).

301 434. J. P. T. M. Van Leeuwen and H. A. P. Pols, Vitamin D: Anticancer and differentiation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1089 – 1105. Academic Press, San Diego, 1997. 435. S. D. Cramer, D. M. Peehl, M. G. Edgar, S. T. Wong, L. J. Deftos, and D. Feldman, Parathyroid hormone-related protein (PTH-P) is an epidermal growth factor-regulated secretory product of human prostatic epithelial cell. Prostate 29, 20 – 29 (1996). 436. K. El Abdaimi, V. Papavasiliou, S. A. Rabbani, J. S. Rhim, D. Goltzman, and R. Kremer, Reversal of hypercalcemia with the vitamin D analogue EB1089 in a human model of squamous cancer. Cancer Res. 59, 3325 – 3328 (1999). 437. R. F. Gagel and S. Peleg, Thyroidal C cells and medullary thyroid carcinoma as target tissues for vitamin D action. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 369 – 377. Academic Press, San Diego, 1997. 438. D. D. Bikle, Clinical counterpoint: Vitamin D: new actions, new analogs, new therapeutic potential. Endocr. Rev. 13, 765 – 784 (1992). 439. S. C. Manolagas, X. P. Yu, G. Girasole, and T. Bellido, Vitamin D and the hematolymphopoietic tissue: A 1994 update. Semin. Nephrol. 14, 129 – 143 (1994). 440. M. Hewison and J. L. H. O’Riordan, Immunomodulatory and cell differentiation, effects of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 447 – 462. Academic Press, San Diego, 1997. 441. R. Baier, A. Grauer, M. Lazaretti-Castro, R. Ziegler, and F. Raue, Differential effects of 1,25-dihydroxyvitamin D3 on cell proliferation and calcitonin gene expression. Endocrinology 135, 2006 – 2011 (1994). 442. P. H. Itin, M. R. Pittelkow, and R. Kumar, Effects of vitamin D metabolites on proliferation and differentiation of cultured human epidermal keratinocytes grown in serum-free or defined culture medium. Endocrinology 135, 1793 – 1798 (1994). 443. G. P. Studzinski, J. A. McLane, and M. R. Uskokovic, Signaling pathways for vitamin D-induced differentiation: Implications for therapy of proliferative and neoplastic diseases. Crit. Rev. Eukaryot. Gene Expr. 3, 279 – 312 (1993). 444. J. Welsh, Vitamin D compounds as potential therapeutics for estrogen-independent breast cancer. Nutrition 13, 915 – 917 (1997). 445. R. Morosetti and H. P. Koeffler, Vitamin D compounds and analogs: Effects on normal and abnormal hematopoiesis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1155 – 1166. Academic Press, San Diego, 1997. 446. W. M. Tong, H. Hofer, A. Ellinger, M. Peterlik, and H. S. Cross, Mechanism of antimitogenic action of vitamin D in human colon carcinoma cells: Relevance for suppression of epidermal growth factor-stimulated cell growth. Oncol. Res. 11, 77 – 84 (1999). 447. S. E. Blutt and N. L. Weigel, Vitamin D and prostate cancer. Proc. Soc. Exp. Biol. Med. 221, 89 – 98 (1999). 448. M. Liu M. H. Lee, M. Cohen, M. Bommakanti, and L. P. Freedman, Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev. 10, 142 – 153 (1996). 449. T. A. Brasitus and M. D. Sitrin, Chemoprevention of colon cancer by vitamin D3 and its metabolites/analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1141 – 1154. Academic Press, San Diego 1997. 450. C. F. Garland, F. C. Garland, and E. D. Gorham, Can colon cancer incidence and death rates be reduced with calcium and vitamin D? Am. Clin. Nutr. 54 (Suppl. 1), 193S – 201S (1991). 451. C. F. Garland, G. W. Comstock, F. C. Garland, K. J. Helsing, E. K. Shaw, and E. D. Gorham, Serum 25-Hydroxyvitamin D and colon cancer: Eight-year prospective study. Lancet 2, 1176 – 1178 (1989). 452. J. A. Eisman, D. H. Barkla, and P. J. Tutton, Suppression of in vivo growth of human cancer solid tumor xenografts by 1,25-dihydroxyvitamin D3. Cancer Res. 47, 21 – 25 (1987).

302 453. S. R. Evans, A. M. Schwartz, E. I. Shchepotin, M. Uskokovic, and I. B. Shchepotin, Growth inhibitory effects of 1,25-dihydroxyvitamin D3 and its synthetic analogue, 1,25-dihydroxy-16-ene-23yne-26,27hexafluoro-19-nor-cholecalcifero1(Ro 25-6760), on a human colon cancer xenograft. Clin. Cancer Res. 4, 2869 – 2876 (1998). 454. M. A. Anzano, J. M. Smith, M. R. Uskokovic, C. W. Peer, L. T. Mullen, J. J. Letterio, M. C. Welsh, M. W. Shrader, D. L. Logsdon, C. L. Driver, C. C. Brown, A. B. Roberts, and M. B. Sporn, (1994). 1-,25-Dihydroxy-16-ene-23-yne-26,27-hexafluorocholecalciferol (Ro24-5531), a new deltanoid (vitamin D analogue) for prevention of breast cancer in the rat. Cancer Res. 54, 1653 – 1656 (1994). 455. E. Elstner, D. Heber, and H. P. Koeffler, 20-Epi-vitamin D3 analogs: Potent modulators of proliferation and differentiation of breast cancer cell lines in vitro. Adv. Exp. Med. Biol. 399, 53 – 70. 456. T. Vink-van Wijngaarden, H. A. Pols, C. J. Buurman, J. C. Birkenhager, and J. P. van Leeuwen, Inhibition of insulin- and insulin-like growth factor-I-stimulated growth of human breast cancer cells by 1,25-dihydroxyvitamin D3 and the vitamin D3 analogue EB1089. Eur. J. Cancer 32A, 842 – 848 (1996). 457. M. Koike, E. Elstner, M. J. Campbell, H. Asou, M. Uskokovic, N. Tsuruoka, and H. P. Koeffler, 19-nor-hexafluoride analogue of vitamin D3: A novel class of potent inhibitors of proliferation of human breast cell lines. Cancer Res. 57, 4545 – 4550 (1997). 458. G. Wu, R. S. Fan, W. Li, V. Srinivas, and M. G. Brattain, Regulation of transforming growth factor- type II receptor expression in human breast cancer MCF-7 cells by vitamin D3 and its analogues. J. Biol. Chem. 273, 7749 – 7756 (1998). 459. K. W. Colston, C. M. Perks, S. P. Xie, and J. M. Holly, Growth inhibition of both MCF-7 and Hs578T human breast cancer cell lines by vitamin D analogues is associated with increased expression of insulin-like growth factor binding protein-3. J. Mol. Endocrinol. 20, 157 – 162 (1998). 460. G. G. Schwartz and B. S. Hulka, Is vitamin D deficiency a risk factor for prostate cancer? Anticancer Res. 10, 1307 – 1311 (1990). 461. C. L. Hanchette and G. G. Schwartz, Geographic patterns of prostate cancer mortality. Evidence for a protective effect of ultraviolet radiation. Cancer 70, 2861 – 2869 (1992). 462. E. H. Corder, H. A. Guess, B. S. Hulka, G. D. Friedman, M. Sadler, R. T. Vollmer, B. Lobaugh, M. K. Drezner, J. H. Vogelman, and N. Orentreich, Vitamin D and prostate cancer: A prediagnostic study with stored sera. Cancer Epidemiol. Biomarkers Prev. 2, 467 – 447 (1993). 463. G. J. Miller, G. E. Stapleton, J. A. Ferrara, M. S. Lucia, S. Pfister, T. E. Hedlund, and P. Upadhya, The human prostatic carcinoma cell line LNCaP expresses biologically active, specific receptors for 1,25-dihydroxyvitamin D3. Cancer Res. 52, 515 – 520 (1992). 464. S. E. Blutt, T. J. McDonnell, T. C. Polek, and N. L. Weigel, Calcitriol-induced apoptosis in LNCaP cells is blocked by overexpression of Bcl-2. Endocrinology 141, 10 – 17 (2000). 465. C. Gross, T. Stamey, S. Hancock, and D. Feldman, Treatment of early recurrent prostate cancer with 1,25-dihydroxyvitamin D3 (calcitriol) [published erratum appears in J. Urol. 160(3 Pt 1), 840]. J. Urol. 159, 2035 – 2039; discussion 2039 – 2040 (1998). 466. D. Trump, P. A. Hershberger R. A. Modzelewski, R. J. Bernardi, J. Muindi, R. M. Rueger, W. Yu, M. J. Egorin, K. E. Blum, and C. S. Johnson. Clinical development of calcitriol and calcitriol analogues as anticancer agents. In “Vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 469 – 475. University of California, Printing and Reprographics, Riverside, 2000. 467. P. H. Reitsma, P. G. Rothberg, S. M. Astrin, J. Trial, Z. Bar-Shavit, A. Hall, S. L. Teitelbaum, and A. J. Kahn, Regulation of myc gene expression in HL-60 leukaemia cells by a vitamin D metabolite. Nature 306, 492 – 494 (1983). 468. Y. Honma, M. Hozumi, E. Abe, K. Konno, M. Fukushima, S. Hata, Y. Nishii, H. F. DeLuca, and T. Suda, 1,25-Dihydroxyvitamin D3

FELDMAN, MALLOY, AND GROSS

469.

470.

471.

472.

473.

474.

475.

476.

477.

478.

479.

480.

481.

482.

483.

484.

485.

and 1-hydroxyvitamin D3 prolong survival time of mice inoculated with myeloid leukemia cells. Proc. Natl. Acad. Sci. USA 80, 201 – 204 (1983). R. Munker, W. Zhang, E. Elstner, and H. P. Koeffler, Vitamin D analogs, leukemia and WAF1. Leuk. Lymphoma 31, 279 – 284 (1998). C. Mathieu, K. Casteels, and R. Bouillon, Vitamin D and diabetes. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1183 – 1196. Academic Press, San Diego, 1997. J. M. Lemire, D. C. Archer, and G. S. Reddy, (1994). 1,25Dihydroxy-24-OXO-16ene-vitamin D3, exerts immunosuppressive activity equal to its parent without causing hypercalcemia in vivo. Endocrinology 135, 2818 – 2821 (1994). C. Mathieu, K. Casteels, M. Waer, J. Laureys, D. Valckx, and R. Bouillon, Prevention of diabetes recurrence after syngeneic islet transplantation in NOD mice by analogues of 1,25(OH)2D3 in combination with cyclosporin A: mechanism of action involves an immune shift from Th1 to Th2. Transplant Proc. 30, 541 (1998). A. Verstuyf, S. Segaert, L. Verlinden, K. Casteels, R. Bouillon, and C. Mathieu, Recent developments in the use of vitamin D analogues. Curr. Opin. Nephrol. Hypertens. 7, 397 – 403 (1998). D. Feldman, T. Chen, M. Hirst, K. Colston, M. Karasek, and C. Cone, Demonstration of 1,25-dihydroxyvitamin D3 receptors in human skin biopsies. J. Clin. Endocrinol. Metab. 51, 1463 – 1465 (1980). K. Colston, J. M. Colston, and D. Feldman, 1,25-Dihydroxyvitamin D3 and malignant melanoma: The presence of receptors and inhibition of cell growth in culture. Endocrinology 108, 1083 – 1086 (1981). S. J. Marx, M. M. Bliziotes, and M. Nanes, Analysis of the relation between alopecia and resistance to 1,25-dihydroxyvitamin D. Clin. Endocrinol. 25, 373 – 381 (1986). Y. Sakai and M. B. Demay, Evaluation of keratinocyte proliferation in vitamin D receptor knockout mice. Endocrinology 141, 2043 – 2049 (2000). E. L. Smith, S. H. Pincus, L. Donovan, and M. F. Holick, A novel approach for the evaluation and treatment of psoriasis. Oral or topical use of 1,25-dihydroxyvitamin D3 can be a safe and effective therapy for psoriasis. J. Am. Acad. Dermatol. 19, 516 – 528 (1988). W. E. Stumpf, H. J. Bidmon, L. Li, C. Pilgrim, A. Bartke, A. Mayerhofer, and C. Heiss, Nuclear receptor sites for vitamin D-soltriol in midbrain and hindbrain of Siberian hamster (Phodopus sungorus) assessed by autoradiography. Histochemistry 98, 155 – 164 (1992). I. M. Musiol, W. E. Stumpf, H. J. Bidmon, C. Heiss, A. Mayerhofer, and A. Bartke, Vitamin D nuclear binding to neurons of the septal, substriatal and amygdaloid area in the Siberian hamster (Phodopus sungorus) brain. Neuroscience 48, 841 – 848 (1992). M. K. Sutherland, M. J. Somerville, L. K. Yoong, C. Bergeron, M. R. Haussler, and D. R. McLachlan, Reduction of vitamin D hormone receptor mRNA levels in Alzheimer as compared to Huntington hippocampus: Correlation with calbindin-28k mRNA levels. Brain Res. Mol. Brain Res. 13, 239 – 250 (1992). S. Carswell, Vitamin D in the nervous system: Actions and therapeutic potential. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1197 – 1211. Academic Press, San Diego 1997. J. Sonnenberg, V. N. Luine, L. C. Krey, and S. Christakos, 1,25Dihydroxyvitamin D3 treatment results in increased choline acetyltransferase activity in specific brain nuclei. Endocrinology 118, 1433 – 1439 (1986). D. Wion, D. MacGrogan, I. Neveu, F. Jehan, R. Houlgatte, and P. Brachet, 1,25-Dihydroxyvitamin D3 is a potent inducer of nerve growth factor synthesis. J. Neurosci. Res. 28, 110 – 114 (1991). I. M. Musiol and D. Feldman, 1,25-dihydroxyvitamin D3 induction of nerve growth factor in L929 mouse fibroblasts: Effect of vitamin D receptor regulation and potency of vitamin D3 analogs. Endocrinology 138, 12 – 18 (1997).

CHAPTER 9 Vitamin D 486. M. S. Saporito, E. R. Brown, K. C. Hartpence, H. M. Wilcox, J. L. Vaught, and S. Carswell, Chronic 1,25-dihydroxyvitamin D3-mediated induction of nerve growth factor mRNA and protein in L929 fibroblasts and in adult rat brain. Brain Res. 633, 189 – 196 (1994). 487. P. A. De Viragh, D. Wolfer, H. A. Lipp, and M. R. Celio, (eds.), “Behavioral Changes in Chronically D-Hyper-vitaminotic Animals.” Walter de Gruyter, New York, 1988. 488. Y. Sato, T. Asoh, and K. Oizumi, High prevalence of vitamin D deficiency and reduced bone mass in elderly women with Alzheimer’s disease. Bone 23, 555 – 557 (1998). 489. B. P. Halloran, Is 1,25-dihydroxyvitamin D required for reporduction? Proc. Soc. Exp. Biol. Med. 191, 227 – 232 (1989). 490. A. Stephanou, R. Ross, and S. Handwerger, Regulation of human placental lactogen expression by 1,25-dihydroxyvitamin D3. Endocrinology 135, 2651 – 2656 (1994). 491. H. Reichel, H. P. Koeffler, and A. W. Norman, The role of the vitamin D endocrine system in health and disease. N. Engl. J. Med. 320, 980 – 991 (1989). 492. R. Eastell and B. L. Riggs, Vitamin D and osteoporosis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 695–711. Academic Press, San Diego, 1997. 493. R. Nuti, E. Bonucci, D. Brancaccio, J. C. Gallagher, C. Gennari, G. Mazzuoli, M. Passeri, and P. Sambrook, The role of calcitriol in the treatment of osteoporosis. Calcif. Tissue Int. 66, 239 – 240 (2000). 494. B. L. Riggs and L. J. Melton, Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 495. O. Sahota, T. Masud, P. San, and D. J. Hosking, Vitamin D insufficiency increases bone turnover markers and enhances bone loss at the hip in patients with established vertebral osteoporosis. Clin. Endocrinol. (Oxf.) 51, 217 – 221 (1999). 496. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in the elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992).

303 497. M. W. Tilyard, G. F. Spears, J. Thomson, and S. Dovey, Treatment of postmenopausal osteoporosis with calcitriol or calcium. N. Engl. J. Med. 326, 357 – 362 (1992). 498. J. C. Gallagher, Prevention of bone loss in postmenopausal and senile osteoporosis with vitamin D analogues. Osteoporos. Int. 1, 172 – 175 (1993). 499. J. S. Adams, V. Kantorovich, C. Wu, M. Javanbakht, and B. W. Hollis, Resolution of vitamin D insufficiency in osteopenic patients results in rapid recovery of bone mineral density. J. Clin. Endocrinol. Metab. 84, 2729 – 2730 (1999). 500. K. H. Lau and D. J. Baylink, Vitamin D therapy of osteoporosis: Plain vitamin D therapy versus active vitamin D analog (D-hormone) therapy. Calcif. Tissue Int. 65, 295 – 306 (1999). 501. S. V. Hughes, E. Robinson, R. Bland, H. M. Lewis, P. M. Stewart, and M. Hewison, 1,25-dihydroxyvitamin D3 regulates estrogen metabolism in cultured keratinocytes. Endocrinology 138, 3711 – 3718 (1997). 502. H. A. Bischoff, H. B. Stahelin, N. Urscheler, R. Ehrsam, R. Vonthein, P. Perrig-Chiello, A. Tyndall, and R. Theiler, Muscle strength in the elderly: Its relation to vitamin D metabolites. Arch. Phys. Med. Rehabil. 80, 54 – 58 (1999). 503. M. S. Stein, J. D. Wark, S. C. Scherer, S. L. Walton, P. Chick, M. Di Carlantonio, J. D. Zajac, and L. Flicker, Falls relate to vitamin D and parathyroid hormone in an Australian nursing home and hostel. J. Am. Geriatr. Soc. 47, 1195 – 1201 (1999). 504. S. Amin, M. P. LaValley, R. W. Simms, and D. T. Felson, The role of vitamin D in corticosteroid-induced osteoporosis: A meta-analytic approach. Arthritis Rheum. 42, 1740 – 1751 (1999). 505. C. H. Jacobus, M. F. Holick, Q. Shao, T. C. Chen, I. A. Holm, J. M. Kolodny, G. E. Fuleihan, and E. W. Seely, Hypervitaminosis D associated with drinking milk. N. Engl. J. Med. 326, 1173 – 1177 (1992). 506. J. S. Adams and G. Lee, Gains in bone mineral density with resolution of vitamin D intoxication. Ann. Intern. Med. 127, 203 – 206 (1997).

CHAPTER 10

Regulation of Bone Cell Function by Estrogens BARRY S. KOMM AND PETER V. N. BODINE Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087

I. II. III. IV. V. VI.

VII. Estrogenic Responses in Bone Cells VIII. Estrogen-Related Receptor- and Osteopontin Gene Expression IX. Nongenomic Actions of Estrogens in Bone Cells X. Summary References

Introduction What Is an Estrogen? Estrogen Receptors ER and ER Knockout Mice Estrogens and Bone: Overview Estrogen Receptors in Bone Cells

I. INTRODUCTION

number of molecules, both steroidal and nonsteroidal in nature. Endogenous vertebrate estrogens are 18 carbon, four-ringed structures [7] (Fig. 1, see also color plate) derived from cholesterol. The most abundant estrogenic steroids in humans include estrone (E1), 17-estradiol (E2), and estriol (E3). In addition, there is an array of estrogenic metabolites that display variable estrogenic activity as well as several well-characterized B-ring-saturated estrogens [8]. Beyond these classic estrogens, several estrogenic substances obtained from plant sources (phytoestrogens), synthetic estrogens (i.e., diethylstibestrol), and a relatively large group of xenobiotics (e.g., DDT, biphenols) have been classified as estrogens. Finally, there is a growing number of molecules, originally classified as antiestrogens, but currently undergoing reclassification (based on their biological activity) that are represented by a diverse set of chemical structures (Fig. 1) and are collectively referred to as selective estrogen receptor modulators (SERMs) [9,10].

Estrogens and their diverse effects on bone remodeling are perhaps less well characterized than one would predict. The positive impact of estrogens on the skeleton has been well known and documented since the early 1940s, and estrogen remains the primary form of osteoporosis treatment in the world [1 – 6]. However, the mechanisms by which estrogens regulate bone remodeling and therby protect the skeleton continue to undergo intense evaluation.

II. WHAT IS AN ESTROGEN? Before discussing the role estrogens play in bone, it is important to define what an estrogen is and the abundance of basic science that describes the multiple facets of estrogenic activity. Estrogens are represented by a large

OSTEOPOROSIS, SECOND EDITION VOLUME 1

305

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

306

KOMM AND BODINE

FIGURE 1

Structures of a variety of compounds that can be classified as members of the family of estrogens. In red are classical steroidal estrogens represented by the three predominant circulating estrogens detected in mammals. In pink is the nonsteroidal and potent estrogen diethylstilbestrol. In green are two phytoestrogens, both nonsteroidal but functionally characterized as estrogens. In black is the potent steroidal antiestrogen ICI182780. This compound has been described as a pure estrogen receptor antagonist; however, its characterization is still under examination. At the bottom of the figure in blue are three generations of selective estrogen receptor modulators (SERMs). Originally referred to, like ICI182780, as antiestrogens, this group of compounds exhibits mixed functional activity all seemingly transduced by estrogen receptors. What all these compounds (and there are hundreds more) have in common is that they bind to the estrogen receptors and functionally affect estrogen receptor activity. In some cases the effects are only as agonists, or as relatively potent antagonists, but most commonly as mixed function ligands with their effects related to the cellular target and the specific genes that are being monitored. (See also color plate.)

III. ESTROGEN RECEPTORS A. Members of the Nuclear Receptor Superfamily What this assortment of compounds has in common is that they exert their function via a single class of nuclearlocalized proteins: estrogen receptors. The two currently recognized members of the estrogen receptor family are referred to as estrogen receptor  (ER) [11 – 13] and estrogen receptor  (ER) [14,15]. Estrogen receptors belong to a large superfamily (Table 1) of nuclear acting receptors represented by members that bind the classical group of steroid hormones, including glucocorticoids, progestins, androgens and mineralocorticoids. In addition to these, other members include the receptors for Vitamin D, retinoids, thyroid hormones, oxysterols, farnesol, prostanoids, and ecdysone. Well over 50 members of this superfamily remain for which a ligand has not been identified, and they are referred to as orphan nuclear receptors [16 – 19].

Steroid receptors share many common features. Structurally, this group of proteins can be dissected into discrete regions with different functions [11,20]. The regions are designated simply as A,B,C,D,E, and F (Fig. 2, see also color plate). The unifying feature characteristic of each nuclear receptor family member is a two zinc finger domain (region C) associated with DNA binding (DNA-binding TABLE 1

Members of the Steroid/Thyroid/Retinoid Nuclear Receptor Superfamily

Androgen

Estrogen (,)

Glucocorticoid

Mineralocorticoid

Progesterone (A,B)

Thyroid hormone (,)

Vitamin D

Retinoic acid (,,)

Retinoid X receptor (,,)

Peroxisome proliferator activating receptor (,,)

Pregnane receptor

Ecdysone

Orphan receptors (50)

307

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

B. Coactivators and Corepressors

FIGURE 2

Schematic structure of nuclear hormone receptors. This family of receptors, which includes the estrogen receptors, can be represented as cassettes with interchangeable units. The A/B domain at the N terminus contains at least one transactivation domain (AF-1) that is ligand independent. The A/B domain is adjacent to the C domain, which represents the DNA-binding domain containing two cysteine loops that each intercalate one zinc molecule to form DNA-binding fingers. This domain is highly conserved among the family members. The D domain is much less well defined but has been described as the hinge domain and contains a nuclear localization signal; however, other sites in the ER have been linked to nuclear localization outside of the D domain. The E domain represents the ligand-binding domain and is not as conseved as the C domain. Additionally, embedded within the ligand binding domain is another transactivation domain: AF-2, which is ligand dependent, unlike AF-1. The F domain of the receptor does not display any clear function; however, removal of small parts of this domain can affect receptor function (both ligand binding and transactivation). (See also color plate.)

domain=DBD). The receptors are DNA-binding proteins that interact with specific DNA sequences [21,22] [e.g., estrogen receptor element (ERE) and androgen receptor element (ARE)] via two cysteine-rich domains that intercalate zinc to form binding “fingers.” Sequence homology in this domain among members of this family is relatively high, and while there is amino acid disparity in the DBD, the cysteine residues can be aligned for all of the receptors, supporting their derivation from a common ancestral protein. The other domains are a ligand-binding domain (region D,E, and F), nuclear localization domain (D), and a hinge domain (D). In addition, two transactivation domains, AF-1, and AF-2, reside in the amino (A/B)and carboxy (e)- terminal portions of the protein, respectively [23]. The mechanism through which information is transduced from the ligand by the receptor has been the subject of intense research since the 1960s. It has become clear that ligand binding to the estrogen receptor initiates a number of processes. Ligand binding produces a change in conformtion, which for several members of the family, including the ER, appears to begin with the displacement of heat shock proteins [24,25]. Subsequently, two liganded estrogen receptors dimerize [26], are biochemically modified (e.g., acetylation and phosphorylation) [27], and then bind to specific DNA sequences. In this simple model, the “activated” ER complex can act as an enhancer or repressor of gene transcriptional activity [28,29].

The model for ER regulation of gene transcription has gained complexity with the discovery of several proteins that interact with the estrogen receptor, as well as other members of the steroid hormone receptor superfamily. These proteins, referred to as coregulators, or comodulators are represented by both coactivators [30,31] and corepressors [32,33]. Several coregulators have been identified and are represented by a diverse group of proteins and RNA [34]. Not unlike the nuclear receptors, several of these proteins contain specific regions associated with independent function [35], including histone acetylation, CREB binding protein interaction domains, and a nuclear receptor interaction domain (NRID) [36,37]. The corepressors contain histone deacetylase domains [33]. Within the NRID, one or more LXXLL motifs interact with the estrogen receptor and other members of the superfamily [38,39]. This binding has been verified by co-crystalization of the ER ligand-binding domain with a small peptide containing an LXXLL domain from the coactivator protein, GRIP 1 (SRC-2) [40], and has been shown to interact specifically with a region of the receptor represented by helices 3,4,5, and 12 [40,41]. The interaction of these coactivators via the NRID has also been demonstrated to be associated with increased transcriptional activity of the ER [42]. The transcriptional complex is composed of an array of proteins, which would include several coactivators whose roles may vary; however, some definitely serve to bridge the enhancer region of ER binding on DNA with the basal transcriptional machinery. The vitamin D receptor interacting protein/thyroid receptor activator protein (DRIP/TRAP) complex of proteins (10 proteins) plays the dual role of transcriptional activation and bridging the transcriptional enhancer complex with the basal transcriptional complex. Not all proteins in the DRIP complex have been shown to interact with the ER, and this complex does not play only a functional role in transcriptional enhancement with nuclear steroid hormone receptors [43 – 45]. Additionally, this is not to say that the estrogen receptor cannot interact directly with proteins associated with the basal transcriptional machinery, as has been suggested for the vitamin D receptor.

C. Alternate Pathways for Estrogenic Activity The model just described for ER activity is characterized as a multiple series of steps initiated by ligand binding (Fig. 3, see also color plate). However, it has become clear that ERs function through other mechanistic pathways to affect various physiologic functions in both a ligand dependent and independent fashion. Data demonstrate that the ER can be activated by growth factors working through protein kinase A and C or in concert with these kinases [46,47].

308

KOMM AND BODINE

FIGURE 3

Estrogens can affect cell function through several pathways. Classically, an estrogen diffuses through the plasma membrane to interact with a nuclear localized receptor (ER or ER or both). The binding of ligand results in a rapid conformational change in the receptor and other biochemical modifications, such as phosphorylation and acetylation. Associated with the changes in conformation are interactions and displacements of proteins from the class of coactivators and corepressors. These proteins form a transcriptional complex linking the receptor DNA complex to the basal transcriptional machinery, resulting in changes in gene transcriptional activity. The liganded receptor can also interact with the AP-1 and SF-1 complex to regulate gene transcription indirectly (i.e., not through estrogen receptors directly binding to DNA). The liganded receptor also interacts with the NF-B protein complex to inhibit gene activity regulated by NF-B. The unliganded ER can be activated by growth factors such as EGF and IGF-1. The stimulated phosphorylation of the ER is sufficient to generate genetic responses regulated by ER without any ligand-induced change in conformation. Other data also support the contention that a plasma membrane-associated ER exists and can transduce information via second messengers. (See also color plate.)

These kinases phosphorylate the receptor (also a result of ligand binding) predominantly at serine 118 [48]. This phosphorylation appears to be sufficient to activate the receptors to then recruit the appropriate coactivators, bind to DNA, and enhance transcriptional activity. This has been shown to occur in a human breast tumor cell model (MCF7 cells) [48,49], a rat osteosarcoma cell line overexpressing ER [50], and a human ovarian adenocarcinoma cell line, BG-1 [51]. In these cell models, treatment with IGF-1, EGF, and other activators of the A and C kinases activates the ER in these cells, as shown by transactivation of an ERE-driven promoter construct. These promoters are minimally active and require functional ER to detect transcriptional activity. It is important to note that in these models, ER antagonists such as ICI-164384 block the stimulation

whether activation is attributed to an estrogen such as 17 estradiol or to kinase activators, supporting the contention that the effects are ER dependent. Another classic model to evaluate estrogen action is the rodent uterine response to estrogens. Here, relatively low doses of an estrogen, typically 17- estradiol, stimulate an increase in uterine wet weight coupled with various uterine histological changes and the expression of several marker genes. If the animals are estrogen deficient (i.e., ovaries removed) and treated with epidermal growth factor, the uterine response in many aspects is quite similar to that seen with 17- estradiol treatment alone [52]. Data utilizing the ER knockout mouse has verified the estrogen receptor requirement for this response [46]. Additionally, it has also been shown that 17-estradiol can activate the MAPK

309

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

pathway, resulting in a rapid alteration in intracellular calcium [53]. Because the estrogen receptor is required for these effects to be manifested, these data strongly support the conclusion that the ER can function in ways other than its role as a transcription factor in the nucleus. While growth factors like IGF-1 and EGF have been demonstrated to activate the ER, it has also been shown that estrogens can inhibit the response associated with various interleukins, especially IL-1, TNF-, and IL-6. This has specific relevance to the role estrogens play in the skeleton. IL-6 and TNF- function via (NF-B) activation and the subsequent interaction of this complex with specific DNA sequences. The skeletal effects of these cytokines are negative (i.e., they increase resorption resulting in a loss of bone mass). ERs interact with NF-B in a ligand-dependent manner to inhibit its activity [54 – 56]. The exact mechanism for suppression is not known, but may be through interference of ER with NF-B binding to DNA, or ER inhibition of appropriate NF-B activation. ER:NF-B interactions and involvement in bone will be discussed in more detail later in this chapter. In addition to the inhibitory effect that the liganded ER has on NF-B, ERs also affect signaling through the AP-1 (JUN-FOS) [57], Sp1 [58], and SF-1 [59,60] transcriptional complexes. Like NF-B, this involves protein:protein interaction. For the AP-1 complex, the effect is compound specific. A compound such as tamoxifen, known as an ER antagonist, functions as a relatively potent agonist through AP-1 and the ER. Activation by tamoxifen is through the AF-1 domain in the N-terminal (A/B) region of the ER [57]. This type of response cannot be evoked with a more classical ERE-driven promoter by tamoxifen. These alternate pathways may partially explain the mixed functional activity that has been reported for SERMs (see Section IV). The final and, at this time, most controversial mechanism through which estrogens can elicit a response appears to involve a process that does not utilize nuclear-localized estrogen receptors or cytosolic ER. Instead, the response to compounds that would be classified as estrogenic appears to function via plasma membrane-associated estrogen receptors [61 – 64]. This would account for physiological responses to estrogens that occur rapidly and would be considered unlikely to be mediated through a nuclear receptor-mediated transcriptional process. Various examples representing different mechanistic explanations include plasma membrane residing ER that bind ligand and transmit information via G-coupled protein receptors [64]; for example, an associated increase in cAMP in response to 17-estradiol has been demonstrated. A twist on this paradigm involves sex hormone-binding globulin (SHBG), a serum protein that transports sex steroid hormones and is capable of binding to plasma membrane receptors. Cells with attached SHBG treated with dihydrotestosterone or 17-estradiol exhibit changes in intracellular cAMP levels,

suggesting an alternative mechanism for 17-estradiol signaling, but only through unliganded SHBG [65]. Additionally, calcium transport has been shown to be directly influenced by estrogens at the level of the calcium channel [66]. Quite recently, the MaxiK channel associated with ion flux via a voltage-gated channel in bladder smooth muscle has been shown to be directly regulated by estrogens [67].

D. ER Estrogens can elicit a variety of physiological responses and, until 1996, it was believed that transduction of information occurred through one nuclear receptor protein (ER). However, as mentioned earlier, a second protein has been identified that also exhibits high-affinity binding for estrogens and dubbed estrogen receptor  (ER) [14,15,68]. Its chromosome location differs from that of human ER (14 vs 6, respectively) [69]. The two transcripts differ in length with ER coding for a protein of 530 amino acids [70] and ER coding for a protein of 595 amino acids [71]. Additionally, their tissue distribution varies, especially in the brain, ovary, uterus, and prostate [72]. At this point in time, the functional role of ER remains to be proven. In vitro transcription assays have shown that ER, like ER, dimerizes and binds to DNA (specifically EREs). However, it has been shown that under appropriate conditions, ER heterodimerizes with ER, and the resulting complex binds to DNA more avidly than the ER homodimer [73]. However, the transcriptional activity of the heterodimer is similar to that of the ER homodimer, but differs from the ER homodimer. The affinity of 17-estradiol for the two receptors is essentially identical, but clearly under in vitro conditions, ER is a more effective activator of transcription than ER [70]. Another characteristic difference between these two receptors is their apparent variation in ligand affinity. Whereas 17-estradiol binding affinity is the same, another estrogen, the phytoestrogen genestein, shows a remarkable preference for ER (30-fold) [74]. The interaction of coactivators with these two proteins also differs. This information, coupled with the distinct tissue distribution and apparent differences in ligand preference, suggests that specific ligands may exist that activate one receptor preferentially over the other [75]. If this is the case, then it also seems quite possible that these compounds could be synthesized and specifically activate only one of the receptors. The pharmaceutical implications of this possibility are obvious.

E. Structure of ER and ER by X-Ray Crystallography Both ER and ER ligand-binding domains (LBDs) have been crystallized [40,76,77]. ER cocrystallized with DES,

310

KOMM AND BODINE

17-estradiol, and 4-OH tamoxifen demonstrates that these ligands generate two different conformations of the ER LBD. With a natural agonist, 17-estradiol, or a synthetic agonist, DES, the ligand fits snuggly into a pocket, and helix 12 (12 of 12 helices in LDB crystallized) appears to fold over the binding pocket [40]. With the SERM 4-OH tamoxifen, helix 12 no longer covers the binding pocket, and now shifts in position to a region that masks amino acids in helices 3,4, and 5. As it turns out, the hydrophobic surface created by those amino acids is critical for the interaction of members of the p160 coactivator family (SRC1,2, and 3) [40]. Indeed, transcriptional activation studies performed with these coactivators in the presence of various anti-estrogens reveal little to no activity, thereby supporting structural data and the importance of the AF-2 domain in estrogen receptor transactivation.

F. Tissue Selective Estrogens (SERMs) What has become clear recently is the fact that ERs are rather willing partners for a wide variety of ligands. This is unlike the other members of the steroid receptor superfamily, which at least, to date, demonstrate more stringent ligand-binding parameters. Compounds with rather diverse structures have been demonstrated to bind with high affinity to the estrogen receptor and exhibit various potencies depending on the end points evaluated. Classically, the targets of estrogen action were the uterus, breast, and liver. In the past two decades, it has also been shown that estrogens directly affect the skeleton, central nervous system (CNS), immune system, cardiovascular system, and the gastrointestinal tract. The discovery of ER has led to the inclusion of the prostate as an estrogen target tissue in males, along with some tissues common to both sexes (i.e., bone, cardiovascular, immune). Obviously, depending on the tissue, the genetic response to estrogens varies. There may be a group of genes that respond similarly in all tissues to a particular agonist, but the key end responses are most likely tissue selective as a result of the responsiveness of a specific set of genes. Thus, in the uterus, a collection of genetic end points can be quantitated that are distinct from the those of breast. This is a critical premise defining the role of tissue selective estrogens (or SERMs) and their clinical applications [8,10]. Perhaps all estrogens are selective and a change in nomenclature is in order. Nevertheless, one example of a tissue selective estrogen would be a compound that behaves as an ER agonist in the skeleton, but as an antagonist (no intrinsic activity but would antagonize estrogens) in the uterus. Tamoxifen, which was originally targeted for contraception, turned out to be a better antiestrogen on breast tissue and was developed as a treatment for hormone (estrogen) responsive breast cancer. As more data were generated using tamoxifen, it was seen to affect several other tissues besides the breast [78]. Some of the effects were positive or

desirable (estrogen agonist activity), such as on the skeleton and lipid profiles, whereas others were considered negative or undesirable such as the antagonist effect in the CNS and the agonist effect on the uterus [79 – 82]. How could this be? Clearly, all “tissue selective estrogens” do not behave identically. Because of structural diversity, their impact on ER function due to different receptor conformation varies [83] and, conceptually, this together with differences in the various target tissues must account for the differences in responses that are seen when comparing these compounds.

IV. ER AND ER KNOCKOUT MICE In an effort to define more clearly the physiologic role(s) of both ER and ER, knockout (KO) mice have been generated [84,85]. Neither KO is lethal and the phenotype exhibited by mice was not as predictable as anticipated. ERKO and ERKO (ER knockout) animals do not demonstrate a striking skeletal phenotype, suggesting that presence of either receptor suffices to maintain skeletal estrogen responsiveness. There is a small, but significant decrease in bone length in both sexes of ERKO animals. This is not seen in ERKOs. Bone mineral density is affected minimally in both KO strains [86]. Ovariectomy of either knockout results in osteopenia typical of wild-type mice and rats, supporting the fact that either receptor is capable of maintaining “normal” modeling in the mouse. One dramatic example of a human man who suffers from ER inactivation (a point mutation resulting in a premature stop condon) has been reported [87]. This man exhibits an overt phenotype where longitudinal bone growth has not terminated (no epiphyseal closure) and bone mineral density has been compromised. Although not published, it appears that this man expresses normal ER and normal androgen receptors. The patients skeletal phenotype is opposite that seen in mice lacking ER, which should warn us (once again) about extrapolating results from rodents to humans. Human data also, at least in this man, suggest that ER and androgen receptors are not sufficient to overcome the inactivation of ER in all aspects of skeletal function where estrogens are required. ERKO mice are characterized by atrophic uteri, ovarian malfunction, and tremendously increased circulating concentrations of estrogens. The testes are abnormal in appearance, wet weight, and function. Successful production of ERKO animals requires heterozygote crossing due to the reproductive impairment in both sexes when both ER alleles are inactivated. ERKO animals, like their ERKO counterparts, exhibit ovarian changes; however, unlike ERKOs, which have hemorrhagic ovaries, ERKOs demonstrate some mature follicles, but reduced numbers compared to normal wild-type mice, resulting in reduced fecundity. Uteri of these mice are normal, and circulating estrogens

311

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

are normal. Testicular histology and function are normal as is male reproductive behavior; however, with age, prostate and bladder hyperplasia has been reported. ER receptor distribution is clearly distinct from ER; there is some overlap, but there is absolutely no ER in specific CNS regions, in ovarian granulosa cells, or, in males, in prostate. Animal data indicate that ER plays a dominant role in the uterus and the ovary, which raises questions as to the absolute necessity of ER in the granulosa cell. One hopes that double knockout animals, which are becoming available, will aid in elucidating ER function more clearly than the individual gene knockout examples. Early data on males reveal that the bone phenotype resembles ERKO animals, again bringing into question the role of ER in the normal developing and remodeling skeleton [88].

V. ESTROGENS AND BONE: OVERVIEW Estrogens are important regulators of skeletal development and homeostasis [89]. This is demonstrated by the dramatic loss of bone that occurs after menopause [90,91] (see Chapter 41). Moreover, estrogens are considered to be firstline therapy for the treatment of postmenopausal osteoporosis [5,92] (see Chapter 69). The reason is that these steroid hormones not only suppress bone resorption and turnover, but also relieve additional menopausal symptoms, such as hot flashes [5,92]. However, the impact of estrogens on bone goes beyond the female skeleton. It is becoming increasingly recognized that these hormones not only play a major role in the cause and prevention of postmenopausal osteoporosis, but also contribute to the development of age-related bone loss and so-called type II osteoporosis, which affects both older women and men [93] (see Chapter 38). Estrogens have both direct and indirect effects on the skeleton [89,94,93]. The extraskeletal actions relevant to calcium homeostasis include the regulation of intestinal calcium absorption [95,96] or secretion [97] the modulation of serum 1,25-dihydroxyvitamin D (vitamin D) concentrations renal calcium excretion, and the secretion of parathyroid hormone (PTH) [93,94]. The direct action of estrogens on bone cells is the subject of the reminder of this chapter. Although some of this work has been reviewed previously [e.g., 89,98 – 101], our goal is to provide a comprehensive upto date review of the literature and some insights into the complexities and mechanisms of estrogen action in the skeleton.

VI. ESTROGEN RECEPTORS IN BONE CELLS Many cell types in the skeleton express ERs. These include cells of both osteoblast and osteoclast lineages, as

well as chondrocytes and endothelial cells. For historical reasons, our discussion of this work will begin with cells of the osteoblast lineage, as these were the first bone-derived cells reported to express the ER.

A. Estrogen Receptors in Osteoblasts Prior to 1987, bone cells were not generally considered to be direct targets for estrogens [102]. However, this view began to change in 1987 when Gray et al. [103] reported that 17-estradiol decreased proliferation and increased alkaline phosphatase activity in rat UMR-106 osteosarcoma cells, which are an in vitro model for the osteoblast or bone-forming cell [104] (see Chapter 2). This report was followed the subsequent year by four publications demonstrating that rat and human osteoblastic cells expressed ERs and/or exhibited estrogenic responses. Komm et al. [105] showed specific binding sites for 125I labeled 17-estradiol in nuclear extracts from rat ROS 17/2.8 and human HOSTE85 osteosarcoma cells, as well as ER mRNA expression by these cells. These authors also reported that 17-estradiol upregulated type I procollagen and transforming growth factor (TGF)-1 mRNA levels in HOS-TE85 cells. Eriksen et al. [106] described specific nuclear-binding sites for [3H]17-estradiol in explant cultures of normal human osteoblasts (hOBs), in addition to ER mRNA expression by these cells. This group also demonstrated that 17-estradiol upregulated the nuclear progesterone receptor (PR) content of hOB cells. Kaplan et al. [107] showed by both immunocytochemistry and ligand-binding assays that osteoblasts in cystic bone lesions from a female patient with McCune – Albright syndrome (fibrous dysplasia) expressed ER. Finally, Ernst et al. [108] reported that 17-estradiol increased the proliferation of primary rat osteoblasts (ROBs) and upregulated 1 type I procollagen mRNA levels in these cells. Since these initial observations over a decade ago, ER expression has been reported to occur in a dozen different in vitro osteoblast models as well as in osteoblasts from in situ studies of bone (Table 2). These models represent a variety of mammalian and avian species. Moreover, ER expression has been determined using either Northern blot or reverse transcriptase-polymerase chain reaction (RTPCR) analysis for mRNA and Western blot or immunocytochemistry for protein. In addition, ER function has been determined by ligand-binding, DNA-binding, and estrogen response element (ERE) reporter gene assays as well as endogenous responses (which will be discussed later). Analysis of ligand-binding data indicates that osteoblasts express relatively low numbers of ERs (60-4500/cell) of high-affinity (Kd  0.05 – 1.1 nM for 17-estradiol) [105 – 107, 109 – 115]. Although these levels are much lower than is observed in uterine and breast cells, which express high amounts of ER, they are consistent with the degree of

312

KOMM AND BODINE

expression seen in other “nonclassical” estrogen responsive tissues [116]. Altogether, these results provide unequivocal evidence that osteoblasts express functional ERs and are one of the direct targets of estrogen action in the skeleton. In 1996, the discovery of the second ER, ER, was reported [117]. This discovery resulted in renaming the original ER as ER. Because each of these had a distinct — albeit overlapping — tissue distribution, investigators began to reexamine ER expression in osteoblasts in the light of these new findings. As outlined in Table 2, in situ studies of rat and human bone have demonstrated that osteoblasts express both ER isoforms [118 – 124]. Moreover, several in vitro osteoblast models, including primary rat and human osteoblasts, have been shown to express both ER and ER [115,122,126 – 129]. However, after reexamining the earlier literature, it is unclear in some instances if a specific osteoblastic cell line expresses either one or the other — or both — ER isoforms. This is particularly true for the human osteosarcoma cell lines HOS-TE85, SaOS-2, and MG-63 (Table 2). Unpublished results from our laboratory using RTPCR analysis indicate that these human osteosarcoma cell lines express only ER mRNA. Although osteoblasts appear to express both ER and ER, it is not known if the isoforms heterodimerize in these cells and what impact this may have on estrogenic responses. Moreover, the ER isoforms appear to be independently regulated during osteoblast development, which may account for the differential effects of estrogens on these cells. In primary rat osteoblasts (ROBs) [125,128] and in SV-HFO transformed human fetal osteoblastic cells [126], ER mRNA expression increases with the increasing stage of differentiation. However, ER mRNA levels either remain constant [125] or also increase [126] with advancing cellular development. Thus, the ratio of ER to ER in osteoblasts may vary as the cells progress from the preosteoblast to the mature osteocyte. Such variation might contribute to the differential estrogenic responses that have been observed in these cells [128] (this will be discussed further in a later section). Support for this idea comes from work by Hall and McDonnell [130]. Using transient transfection assays, these authors showed the following: (1) ER functions as a transdominant inhibitor of ER transcriptional activity at subsaturating steroid levels, (2) ER and ER can heterodimerize in cells, and (3) ER can interact with target gene promoters in the absence of ligand. Thus, Hall and McDonnell concluded that the relative levels of expression of these two receptor isoforms would determine how a cell responds to either estrogens or antiestrogens.

B. Estrogen Receptors in Osteocytes and Lining Cells Osteoblasts, which arise from mesenchymal stem cells in the bone marrow, undergo further differentiation to either

lining cells or osteocytes [104] (see Chapter 2). Lining cells are thought to be quiescent osteoblasts that line the mineralized bone matrix and regulate access of the osteoclasts to this tissue [131]. In contrast, osteocytes are osteoblasts that become embedded within the mineralized matrix and assume a stellate or dendritic morphology [132,133]. The primary function of osteocytes, which are the most abundant cell type in mature bone, appears to be mechanosensory [132,133]. As such, they are involved in strain perception and the adaptive mediation of physical forces on bone modeling and remodeling [134,133]. Osteocytes and lining cells may also be targets for estrogens [133]. As outlined in Table 3, evidence has been obtained from in situ studies that mammalian and avian osteocytes express ERs. Receptor expression in these cells has been shown to occur using in situ hybridization for mRNA and immunocytochemistry for protein. Moreover, as with osteoblasts, human osteocytes have been reported to express both ER and ER [118,122,124,135,136]. Unpublished observations from our laboratory indicate that a conditionally immortalized human osteocyte cell line (HOB-05-T1) expresses both ER and ER mRNAs (as measured by RT-PCR), and that these receptors are functional based on the transactivation of an ERE reporter gene by 17-estradiol. Estrogenic responses in osteocytes will be discussed in a later section. At least two publications document ER expression in bone-lining cells. Ohashi et al. [137] reported that lining cells in Japanese quail bone contained ERs, where Kusec et al. [119] showed ER mRNA and protein expression in human-lining cells. Although these studies suggest that estrogens may play a role in the physiology of these cells, there are as yet no identified estrogenic responses in lining cells. One limitation to these types of investigations is the absence of an in vitro model to study lining cell biology.

C. Estrogen Receptors in Bone Marrow Stromal Cells Pluripotent mesenchymal stem cells of bone marrow have the capacity to become osteoblasts, as well as chondrocytes, adipocytes, myoblasts, and fibroblasts [138,139]. Like other cells of the osteoblast lineage, these bone marrow stromal cells (BMSCs) express ERs and are estrogen responsive. As summarized in Table 4, primary BMSCs from rodents and humans, as well as some immortalized bone marrow stromal cell lines, have been reported to express ER and ER. In these studies, ER expression was demonstrated using RT-PCR and Northern hybridization for mRNA, immunocytochemistry for protein, and cytosolic ligand-binding assays for receptor function. Oreffo et al. [140] reported that human BMSCs express ER mRNA based on Northern blot analysis and that its expression increases as the cells undergo differentiation to osteoblasts.

313

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

TABLE 2 Isoform

Estrogen Receptors in Osteoblasts System

Observation

Ref.

ER and ER

Rat ROS 17/2.8 osteosarcoma cells

mRNA Ligand binding Protein ERE-tk-CAT

[105] [113] [114] [125]

ER (?)

Human HOS-TE85 osteosarcoma cells

mRNA Ligand binding Protein

[105] [109] [294]

ER and ER

Primary human OB (hOB) cells

mRNA Ligand binding Protein ERE-tk-Luc/Cat

[106] [110] [295] [224] [127] [296] [122]

ER and ER

Human bone

Protein mRNA

[107] [118] [119] [122] [124]

ER (?) and ER

Human SaOS-2 osteosarcoma cells

Ligand binding mRNA Protein

[109] [122]

ER and ER

Rat bone

mRNA

[297] [121] [123]

ER and ER

Primary rat OB (ROB) cells

mRNA ERE-tk-CAT

[237] [125] [128] [137]

ER (?)

Japanese quail bone

Protein

ER (?)

Immortalized human HOBIT cells

mRNA

[111]

ER and ER

Immortalized mouse MC-3T3-E1 cells

mRNA Protein

[112] [294] [298]

ER (?)

Primary mouse OB cells

mRNA Protein

[294]

ER (?) and ER

Human MG-63 osteosarcoma cells

mRNA Protein

[299] [122]

ER and ER

Rat UMR-106 osteosarcoma cells

Ligand binding Protein mRNA ERE-tk-CAT

[114] [298]

ER and ER

Immortalized human HOB-03-CE6 cells

mRNA Ligand binding DNA binding ERE-tk-Luc

[115],[129]

ER

Rabbit bone

mRNA Protein

[119]

ER and ER

Transformed human SV-HFO cells

mRNA

[126]

ER

Mouse bone

mRNA Protein

[122]

314

KOMM AND BODINE

TABLE 3

Estrogen Receptors in Osteocytes

TABLE 5

Estrogen Receptors in Cells of the Osteoclast Lineage

Isoform

System

Observation

Ref.

ER (?)

Japanese quail bone

Protein

[137]

ER and ER

Human bone

Protein mRNA

[135] [118] [119] [122] [136] [124]

ER and ER (?)

Human bone

Protein mRNA

[144] [118] [122]

ER (?)

Chicken osteoclasts

Ligand binding mRNA Protein

[145]

ER

Human giant cell tumors

mRNA Protein

[146]

ER

Human FLG-29.1 preosteoclastic cells

Ligand binding mRNA Protein ERE-tk-Cat

[151]

Isoform

ER (?)

Pig bone

Protein

[135]

ER (?)

Guinea pig bone

Protein

[119]

ER

Rabbit bone

mRNA Protein

[119]

ER

Mouse bone

mRNA Protein

[122]

ER and ER

Immortalized human HOB-05-T1 cells

mRNA EREtk-Luc

Bodine and Komm, unpublished

Likewise, Dieudonne et al. [141] stated that immortalized human bone marrow stromal fibroblasts (BMSFs) isolated from a patient with a mutated ER gene, as well as nonimmortalized control BMSFs from normal patients, expressed ER mRNA as determined by RT-PCR. Moreover, the nonimmortalized control BMSFs were acknowledged to express the wild-type ER message. Estrogenic responses in BMSCs will be discussed in a later section.

D. Estrogen Receptors in Cells of the Osteoclast Lineage Osteoclasts are multinucleated giant cells responsible for bone resorption [142,143] (see Chapter 3). They arise from hemopoietic stem cells of the monocyte/macrophage linTABLE 4 Isoform

Estrogen Receptors in Bone Marrow Stromal Cells System

Observation

Ref.

ER

Mouse / LDA11 cells

Ligand binding mRNA

[300]

ER

Mouse MBA 13.2 cells

Ligand binding mRNA

[300]

ER

Mouse BMSCs

mRNA

[300] [301] [298]

ER and ER

Rat BMSCs

mRNA

[121]

ER and ER

Mouse ST2 cells

mRNA Protein

[298]

ER and ER

Human BMSCs

mRNA

[141] [140]

System

Observation

Ref.

ER (?)

Rabbit osteoclasts

mRNA

[148]

ER

Mouse hemopoietic blast cells

mRNA

[152]

ER

Rat preosteoclasts

mRNA

[302]

ER

Primary human osteoclasts

mRNA

[147]

ER

Human TCG 51 preosteoclastic cells

Protein

[303]

ER (?)

Mouse bone

Protein

[122]

eages, which, like BMSCs, are found in the bone marrow [143]. Because the primary therapeutic effect of estrogens on the postmenopausal skeleton is to suppress bone resorption [90,91], it seems logical that cells of the osteoclastic lineage would express ERs. However, the direct action of estrogens on osteoclasts is less generally accepted by workers in the field than is an indirect effect mediated by cells of the osteoblast lineage. Table 5 summarizes the evidence for ER expression by osteoclastic cells. In 1990, Pensler et al. [144] reported that human osteoclasts isolated from membranous bone (pediatric craniotomies) expressed ERs based on immunocytochemistry of fixed cells and radioimmunoassay (RIA) of cell lysates. Subsequently, Oursler and colleagues described the presence of ERs in osteoclasts purified from either chicken long bones [145] or human giant cell tumors (hGCTs) of bone (i.e., osteoclastomas) [146]. For these studies, the authors used a monoclonal antibody (121F) generated to chicken osteoclasts to purify mature osteoclasts ( 90% pure) from these tissues. ER expression was then demonstrated using either Northern blot analysis [145] or RT-PCR [146] for ER mRNA, Western blot analysis for receptor protein [145], and a nuclear ligand-binding assay, which indicated that the chicken osteoclasts contained 5000 – 6000 ERs/nucleus [145]. Two groups confirmed that human osteoclasts express ER mRNA. Hoyland et al. [118] used in situ RT-PCR to demonstrate the presence of ER message in normal human bone samples, whereas Sunyer et al. [147] used RT-PCR to reveal the expression of this mRNA in

315

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

purified normal human osteoclasts (hOCLs). ER mRNA has also been reported to be expressed by isolated mature rabbit osteoclasts [148]. Thus, at least five separate laboratories have found evidence for ER expression in osteoclasts. In contrast, Collier et al. [149] failed to detect either ER or ER mRNAs in pure preparations of microisolated osteoclasts from hGCTs. Moreover, the authors confirmed their results using fluorescence in situ hybridization (FISH), which showed that the tumor mononuclear cells expressed ER mRNA, whereas the multinuclear osteoclasts did not express this gene. The reason for this discrepancy is not clear. However, Oursler [142] has postulated that prior in vivo exposure to estrogens may have downregulated ER levels in the osteoclasts examined by Collier and co-workers [149]. This conclusion is based on the work of Pedersen et al. [150], who reported that in vivo treatment of 5-week-old chickens with 17-estradiol dramatically suppressed ER protein levels in the purified osteoclasts. Preosteoclasts also appear to express ERs (Table 5). For example, Fiorelli et al. [151] used RT-PCR (for ER), Western blot analysis, a nuclear extract ligand-binding assay, and an ERE reporter gene assay to demonstrate the presence of functional ERs in human leukemic FLG 29.1 cells. The ligand-binding assay showed that this cell line, which can be induced to express an osteoclast-like phenotype, contained approximately 400 ERs/nucleus. Moreover, Kanatani et al. [152] demonstrated that mouse hemopoietic blast cells, which contain osteoclast progenitors, express ER mRNA based on RT-PCR. Estrogenic responses in osteoclastic cells will be discussed in a later section.

E. Estrogen Receptors in Chondrocytes and Other Bone-Associated Cells Estrogens play an important role in the regulation of human longitudinal bone growth and skeletal maturation [89] (see Chapter 25). These steroid hormones accelerate endochondral bone formation in early adolescence, but also initiate epiphyseal growth plate fusion in late adolescence. Consistent with these observations, chondrocytes express both ER and ER. As outlined in Table 6, rabbit, mouse, rat, human, and pig chondrocytes are all reported to possess ERs. These observations are based on in situ hybridization for ER mRNA [119], immunocytochemistry for ER and ER proteins [119,153 – 156], and cytosolic ligand-binding assays [157 – 160]. Scatchard analysis of ligand-binding data indicates that chondrocytes express relatively low amounts of ER (3.9 – 11.2 fmol/mg protein) [160] of high affinity (Kd  0.12 – 0.87 nM for 17-estradiol) [157,160]. Thus, these ER parameters are comparable to those found in osteoblasts [115]. In human growth plate chondrocytes, ER was reported to be expressed by resting, proliferative, and hypertrophic cells [119] whereas ER expression was shown to be restricted to hypertrophic cells [156]. Thus, these ER iso-

TABLE 6 Isoform

Estrogen Receptors in Chondrocytes System

Observation

Ref.

ER

Rabbit chondrocytes

Ligand binding mRNA Protein

[157] [119] [155]

ER and ER

Human chondrocytes

Protein

[153] [154] [119] [156] [303]

mRNA

ER (?)

Rat chondrocytes

Protein

[160] [155]

ER (?)

Pig bone

Protein

[135]

ER (?)

Guinea pig

Protein

[135]

forms may have distinct roles in the regulation of endochondral bone growth and maturation. Estrogenic responses in chondrocytes will be discussed in a later section. At least one report describes the expression of ERs in bone-derived endothelial cells [161]. Using bovine bone endothelial (BBE) cells, the authors showed that these cells expressed ER mRNA by Northern hybridization and contained specific binding sites for [3H]17-estradiol (Kd  17.2 nM, Bmax  32,000 sites/cell). Treatment of the cells with 17-estradiol enhanced proliferation and suppressed PTH-stimulated cyclic adenosine monophosphate (cAMP) accumulation. As described in more detail later, both of these estrogenic responses have also been observed in osteoblasts. Thus, this study suggests that estrogens may regulate bone angiogenesis as well as bone formation and resorption.

F. Summary It is clear from the numerous studies reviewed in this section that many cell types in the skeleton express ERs. These estrogen responsive cell types include bone marrow progenitor cells as well as mature osteoblasts, osteoclasts, and chondrocytes. In the osteoblast lineage, each cell type — from the BMSC to the osteocyte or lining cell — has been shown to be a potential estrogen target. Thus, the totality of the effects of estrogen on the skeleton may, to a large extent, be equivalent to the sum of its action on all of these cell types. The following section reviews the estrogenic responses of skeletal cells and places them in the context of in vivo knowledge of estrogen action.

VII. ESTROGENIC RESPONSES IN BONE CELLS Consistent with the expression of ERs by many bone cell types, there are also many estrogenic responses in these

316

KOMM AND BODINE

cells. Our review of the literature on estrogenic responses indicates that findings are sometimes contradictory. We will attempt to place the estrogenic response in the context of estrogen’s known physiologic and therapeutic function in the skeleton.

A. Estrogenic Responses in Cells of the Osteoblast Lineage Due to the profusion of in vitro models, much of what we know about estrogen action on bone cells is in relationship to the osteoblast. As summarized in Table 7, 43 estrogenic responses have been identified in 15 different in vitro osteoblast models. To make sense of these observations we have separated them into six different categories: regulation of osteoblast number; regulation of matrix production and mineralization; regulation of growth factor expression and responsiveness; regulation of factors that modulate bone resorption; regulation of receptor expression and signal transduction; and miscellaneous responses. Moreover, we have indicated which in vitro model(s) was reported to exhibit each of the responses. The reason for doing this is to determine if a given response represents a generalized estrogenic effect in an osteoblast or whether it might be specific to a particular cell line (e.g., immortalized MC-3T3-E1 mouse cells) or cell type (e.g., osteosarcoma-derived cells). From our viewpoint, primary cultures are the most pertinent osteoblast models for attempting to extrapolate in vitro estrogen affects to in vivo relevance. On the other hand, caution should be applied to observations that are only made in osteosarcoma cells, as these are generally considered to be unreliable models of osteoblasts biology [162,163]. When available, we have also noted when an in vitro estrogenic response has been observed in vivo and therefore may be relevant physiologically or pharmacologically (see Chapter 37). 1. REGULATION OF OSTEOBLAST NUMBER Using UMR-106 rat osteosarcoma cells, Gray et al. [103] reported that 17-estradiol decreases osteoblastic cell proliferation. In the same study, 17-estradiol also increased alkaline phosphatase activity. Given the limitations of osteosarcoma cells as models of osteoblast biology [162,163] these results suggested that estrogens might potentiate cellular differentiation, as the mature rat osteoblast expresses high levels of alkaline phosphatase and no longer divides [162,104]. Subsequent to this publication, other researches described similar results using four additional in vitro osteoblast models (Table 7). These include primary osteoblasts isolated from the tibias of 17-estradiol-treated ovariectomized (OVX) rats [164]. Moreover, Westerlind et al. [165] confirmed these observations in vivo by showing that the potent nonsteroidal estrogen, diethylstilbestrol (DES), reduces the [3H]thymidine-labeling index of tibial

osteoblasts in OVX rats. Thus, a suppressive effect of estrogens on osteoblast proliferation is consistent with an inhibitory action of the steroid on bone turnover [89 – 91]. In contrast, other laboratories using additional in vitro models, as well as ROBs, have reported that estrogens increase osteoblast proliferation and DNA synthesis (Table 7). There are several possible explanations for these discrepancies. First, with the exception of studies using UMR106 and ROBs, other publications that showed that 17 estradiol suppresses proliferation used cell lines that overexpressed ER. Thus, as was concluded by Watts and King [166], overexpression of the ER may inhibit cell proliferation by interfering with transcription artifactually. If this is true, then a transfected ER may not necessarily function the same as the endogenous ER. However, studies reporting that 17-estradiol stimulated osteoblast proliferation all used in vitro models that naturally expressed ERs. At least two groups have reported that in vitro treatment of ROBs with 17-estradiol enhances cell proliferation or DNA synthesis [108,167], whereas Modrowski et al. [164] used isolated osteoblasts from in vivo-treated OVX rats to show that the steroid inhibits proliferation. Consequently, these two experimental paradigms may generate cells that are in different stages of differentiation (e.g., preosteoblastic versus mature osteoblasts), and these stages may respond differently to estrogens [168,128]. While a suppressive effect of estrogens on osteoblast proliferation is consistent with a potentiation of differentiation or a suppression of bone turnover, a stimulatory effect might reflect an expansion of the preosteoblast pool [169]. For example, Qu et al. [170] presented evidence that treatment of primary mouse BMSC cultures with 17-estradiol stimulates cellular proliferation and differentiation into osteoblastic cells. This in vitro observation is consistent with an in vivo study of Somjen et al. [171], which reported that 17-estradiol stimulates DNA synthesis in rat bone. Moreover, in OVX Swiss – Webster mice, high doses of 17 estradiol (50 – 100 g/mouse/week, s.c., for 4 weeks) increased both endosteal and cancellous bone formation, as well as inhibited bone resorption [172]. Thus, under some circumstances, estrogens may stimulate bone formation [173] as well as inhibit resorption and turnover. However, the stimulatory action of the steroid may represent a pharmacological or toxic effect rather than a physiologic or therapeutic response [174]. In addition to regulating cell division, estrogens have now been shown to control osteoblast apoptosis. Gohel et al. [175] reported that 17-estradiol blocks the induction of apoptosis by cortisol in primary rat and mouse osteoblasts. These in vitro observations were confirmed by an in vivo experiment showing that 17-estradiol decreased the number of apoptotic osteoblasts in the calvaria of dexamethasone-treated mice. Consequently, estrogens may modulate osteoblast number by regulating both proliferation and

317

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

TABLE 7

Estrogenic Responses in Cells of the Osteoblast Lineage

Response

System

Ref.

Regulation of cell number Decreases proliferation and decreases DNA synthesis

Rat UMR-106 osteosarcoma cells Human HTB-96 cells overexpressing ER Primary rat OB (ROB) cells Rat ROS.SMER-14 cells overexpressing ER Human hFOB/ER9 cells overexpressing ER

Rat ROS 17/2.8 osteosarcoma cells Rat bone Increases proliferation and increases DNA synthesis

Primary ROB cells

[103] [254] [304] [166] [164] [167] [113] [212] [168] [178] [203] [305] [165]

Primary human OB (hOB) cells Human HOS-TE85 osteosarcoma cells Primary mouse bone marrow stromal cells Rat bone Mouse bone

[108] [181] [167] [181] [112] [177] [176] [306] [170] [171] [169]

Primary ROB cells Primary mouse OB cells Mouse bone

[175] [175] [175]

Rat UMR-106 osteosarcoma cells

Immortalized human HOB-03-CE6 cells Primary ROB cells

[103] [254] [113] [176] [177] [168] [178] [115] [128]

Decreases alkaline phosphatase

Primary ROB cells Rat bone

[128] [180]

Increases osteocalcin

Primary ROB cells

[128]

Decreases osteocalcin

Rat ROS 17/2.8 osteosarcoma cells Human hFOB/ER9 cells overexpressing ER

[179] [168] [178] [128] [180] [179] [182]

Transformed rat RCT-1 and -3 cells Immortalized mouse MC-3T3-E1 cells

Inhibits glucocorticoid-induced apoptosis

Regulation of matrix production and mineralization Increases alkaline phosphatase

Rat ROS.SMER-14 cells overexpressing ER Primary hOB cells Immortalized mouse MC-3T3-E1 cells Human hFOB/ER9 cells overexpressing ER

Primary ROB cells Rat bone

Increases osteonectin

Primary ROB cells

[128]

Decreases osteonectin

Primary ROB cells Rat bone

[128] [180]

Increases type I collagen

Human HOS-TE85 osteosarcoma cells Primary ROB cells

[105] [108] [181] [128] [181]

Transformed rat RCT-1 and -3 cells

(continues)

318

KOMM AND BODINE

TABLE 7

(continued)

Response

System Primary hOB cells

Immortalized mouse MC-3T3-E1 cells

Ref. [110] [127] [296] [177]

Decreases type I collagen

Primary hOB Cells Rat bone

[296] [180] [182]

Increases mineralization

Human HOS-TE85 osteosarcoma cells Primary human OB (SaM-1) cells

[185] [185]

Regulation of growth factor expression and responsiveness Increases TGF-1

Human HOS-TE85 osteosarcoma cells Rat UMR-106 osteosarcoma cells Primary hOB Cells Primary mouse OB cells Primary ROB cells Rat ROS 17/2.8 osteosarcoma cells Rat bone

[105] [197] [198] [199] [128] [179] [199] [179]

Increases TGF-3

Human MG-63 osteosarcoma cells Rat bone

[202] [200]

Increases TIEG

Human hFOB/ER9 cells overexpressing ER

[203]

Increases BMP-6

Human hFOB/ER9 cells overexpressing ER

[205]

Increases IGF-I

Rat UMR-106 osteosarcoma cells Primary ROB cells

[207] [181] [307] [181] [308]

Transformed Rat RCT-1 and -3 cells Human hFOB/ER9 cells overexpressing ER Increases growth hormone Receptor

Rat UMR-106 osteosarcoma cells Primary hOB cells

[208] [208]

Increases IGF-BPs

Primary ROB cells Human hFOB/ER9 cells overexpressing ER Human SaOS-2 osteosarcoma cells

[211] [212]

Decreases IGF-BP3

Primary human bone marrow stromal cells

[215]

Blocks PGE2-induced IGF-1

ROB cells overexpressing ER

[216]

Mouse / LDA11 marrow stromal cells Primary hOB cells Primary ROB cells Primary mouse OB cells

[217] [217] [217] [217] [309] [217] [310] [225]

[213]

Regulation of factors that modulate bone resorption Decreases IL-6

Immortalized mouse MC-3T3-E1 cells Human SaOS-2 cells overexpressing ER Human hFOB/ER9 cells overexpressing ER Immortalized human HOB-03-CE6 cells Human MG-63 osteosarcoma cells Primary human bone marrow stromal cells In vivo (mice)

[115] [311] [219] [218] [312]

Decreases TNF-

Primary hOB cells

[223]

Decreases gp80 and gp130

Mouse / LDA11 marrow stromal cells Immortalized mouse MC-3T3-E1 cells

[227] [227]

319

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

TABLE 7

(continued)

Response

System

Ref.

Increases OPG

Human hFOB/ER9 cells overexpressing ER Primary hOB Cells

[229]

Suppresses PTH action

Human SaOS-2 osteosarcoma cells

[232] [233] [235] [213] [181] [237] [181] [234] [198] [177] [115] [236]

Transformed rat RCT-1 and -3 Cells Primary ROB cells Primary mouse OB cells Primary hOB cells Immortalized mouse MC-3T3-E1 cells Immortalized human HOB-03-CE6 cells In vivo (humans) Enhances PTH action

Increases IL-1

Human SaOS-2 osteosarcoma cells

[229]

Primary ROB cells Primary hOB cells

[238] [239] [239] [239]

Immortalized human HOBIT cells

[241]

Regulation of receptor expression and signal transduction Increases PR

Primary hOB cells Human hFOB/ER9 cells overexpressing ER

[106] [242] [225]

Antagonizes VD3 responsiveness

Rat UMR-106 osteosarcoma cells

[207]

Increases VDR and VD3 responsiveness

Rat ROS 17/2.8 osteosarcoma cells Human OGA osteosarcoma cells

[243] [243]

Increases ER

Primary hOB cells

[127] [296] [128] [136] [124]

Primary ROB cells In vivo (human bone) Decreases ER

Primary ROB cells In vivo (human bone)

[128] [136]

Decreases IP3 receptor

Rat UMR-106 osteosarcoma cells

[245] [247] [245] [247] [245] [247] [245] [247] [245] [247]

Human SaOS-2 osteosarcoma cells Primary ROB cells Immortalized mouse MC-3T3-E1 cells G-292 human osteosarcoma cells Increase basal NOS

Human HOS-TE85 osteosarcoma cells In vivo (rats)

[248] [250]

Decrease cytokine-induced NO

Immortalized mouse MC-3T3-E1 cells

[251]

Enhances bradykinin action

Primary hOB cells

[252]

Increases CK

Primary ROB cells Immortalized mouse MC-3T3-E1 cells Rat ROS 17/2.8 osteosarcoma cells Rat bone

[171] [171] [171] [171]

Increases HSP-27

Immortalized mouse MC-3T3-E1 cells

[253]

Increases AST, GGT, LDH, and transferrin

Rat UMR-106 osteosarcoma cells

[254]

Miscellaneous response

320 viability. As will be reviewed in later sections, estrogens may also suppress osteocyte apopotosis but induce the programmed cell death of osteoclasts. 2. REGULATION OF MATRIX PRODUCTION MINERALIZATION

AND

One of the most commonly observed estrogenic responses in osteoblasts is the upregulation of alkaline phosphatase expression, which is an important phenotypic marker of the osteoblast lineage [104] (see Chapter 2). Estrogens have been reported to increase either alkaline phosphatase mRNA levels and/or activity in seven different in vitro osteoblast models (Table 7). These models include rat osteosarcoma cell lines [103,113], primary cultures of ROB or hOB cells [128,176], immortalized mouse MC-3T3-E1 cells [177], and the conditionally immortalized human osteoblast cell lines hFOB/ER9 and HOB-03-CE6 [115,168,178]. However, in the case of ROB cells, 17estradiol has also been reported to downregulate alkaline phosphatase expression [128]. The explanation for this discrepancy is that 17-estradiol regulates the steady-state mRNA levels of this enzyme in a differentiation selective manner [128]. In post-proliferative/nodule-forming stage ROB cells (i.e., mature osteoblasts), 17-estradiol estradiol suppresses alkaline phosphatase expression, whereas in postmineralization stage cells (i.e., osteocytes), the steroid hormone increases enzyme message levels. This same pattern of regulation also holds true for the noncollagenous bone matrix proteins osteocalcin and osteonectin [128]. Estrogens also regulate the expression of osteocalcin (Table 7), which is the most selective phenotypic marker of the osteoblast lineage [104]. As noted previously, 17estradiol downregulates steady-state osteocalcin mRNA levels in postproliferative/nodule-forming stage ROB cells, but upregulates it in postmineralization stage cells [128]. Moreover, estrogens decrease osteocalcin expression in ROS 17/2.8 osteosarcoma cells [179] and in hFOB/ER9 cells, which over express human ER [168,178]. Confirmation that estrogens downregulates alkaline phosphatase, osteocalcin and osteonectin mRNA levels in vivo comes from Turner et al. [180], who reported that DES treatment of OVX rats decreased the expression of these messages in periosteal osteoblasts isolated from long bones. Again, suppression of osteoblastic activity as measured by the expression of bone matrix proteins would be consistent with a reduction in bone turnover. The most abundant bone matrix protein is of course type I collagen [104] (see Chapter 4), and it is perhaps not surprising that estrogens have been shown to regulate its expression (Table 7). Komm et al. [105] and Ernst et al. [108] were the first to report that 17-estradiol upregulated 1 type 1 procollagen mRNA levels in HOS-TE85 human osteosarcoma cells and ROB cells, respectively. Subsequent studies confirmed these observations in hOBs [110], MC-

KOMM AND BODINE

3T3-E1 [177], and transformed rat RCT-1 and RCT-3 cell lines [181]. In contrast to these in vitro studies, type I collagen expression does not appear to be upregulated by estrogens in vivo. In fact, mRNA levels for this bone matrix protein have been reported to increase in OVX rat bones [182,183], and estrogens have been observed to either suppress this increase [180,182], or have no effect [184]. Once again, the in vivo observations are consistent with the concept that estrogen deficiency increases bone resorption and bone turnover, and that estrogens reduce these effects [89 – 91]. Finally, at least one report describes the effects of estrogens on mineralization. Takeuchi et al. [185] showed that 17-estradiol at concentrations of 1 – 100 nM increased the calcium content of extracellular matrix laid down in vitro by either HOS-TE85 human osteosarcoma cells or primary human osteoblasts (referred to as SaM-1 cells). 3. REGULATION OF GROWTH FACTOR EXPRESSION RESPONSIVENESS

AND

Another aspect of osteoblast biology that estrogens have been shown to regulate is growth factor expression and responsiveness. Bone is an abundant reservoir for several growth factors, including isoforms of transforming growth factor (TGF)-, bone morphogenetic proteins (BMPs), and insulin-like growth factors (IGFs) [186 – 191] (see Chapters 13 and 14). These peptides are synthesized and secreted by cells of the osteoblast and/or osteoclast lineages, and they regulate the proliferation, differntiation, and activities of these cell types [186,187,189 – 193]. In fact, growth factors, together with other cytokines, provide the elaborate communication network that couples osteoclastic bone resorption to osteoblastic bone formation [89,99] (see Chapter 12). Moreover, it is the disruption of this network that, to a large extent, leads to accelerated bone resorption and increased bone turnover after menopause [99,194 – 196]. The first bone cell-derived growth factor whose expression was shown to be regulated by estrogens was TGF-1. Komm et al. [105] reported in 1988 that 17-estradiol treatment of HOS-TE85 human osteosarcoma cells upregulated the steady-state levels of TGF-1 mRNA. As outlined in Table 7 estrogens have also been shown to increase TGF1 mRNA expression and/or TGF- protein secretion in rodent osteosarcoma cell lines [179,197], as well as primary cultures of human, mouse, and rat osteoblasts [128,198,199]. Moreover, estrogens have been observed to increase TGF- expression in bone in vivo. Finkelman et al. [199] reported that treatment of OVX rat with 17 estradiol upregulated TGF- protein levels in long bones, and lkeda et al. [179] demonstrated that TGF-1 mRNA levels decreased in the tibia of OVX rats. However, neither Westerlind et al. [183] nor Yang et al. [200] were able to confirm these findings. Although TGF- regulates osteoblast proliferation, differentiation, and activity in vitro

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

and promotes bone formation in vivo [186,187,190], it has also been reported to inhibit osteoclast differentiation and activity in vitro [89,99]. Thus, an increase in osteoblastic TGF- production would be consistent with a therapeutic antiresorptive effect of estrogens [90,91]. Estrogens, as well as tissue-selective estrogens (TSEs) [92] or SERMs [201], have also been reported by at least one group to increase TGF-3 expression by osteoblastic cells (Table 7). Yang et al. [200] observed an increase in TGF-3 mRNA levels in the femurs of OVX rats that were treated with either 17-estradiol or the SERM raloxifene; in contrast, the message levels for either TGF-1 or TGF2 were unaffected by these treatments. Although in situ studies to identify the celltype(s) responsible for this expression were not reported, this same group subsequently demonstrated that 17-estradiol or raloxifene upregulated TGF-3 mRNA levels in MG-63 human osteosarcoma cells [202]. These observations were extended by contransfection studies in MG-63 cells using human TGF-3 promoter – reporter gene constructs and human ER expression vectors [200,202]. These experiments indicated that a variety of estrogens and TSEs/SERMs upregulated TGF-3 promoter activity in an ER-dependent manner. Although these results were intriguing, an apparent disconnection occurred between in vitro and in vivo pharmacology, as the potency and efficacy of compounds in this in vitro assay did not correlate with their bone-sparing activities in vivo. Moreover, 17-estradiol antagonized raloxifene in this in vitro system [202]. In any event, as with TGF-1, upregulation of TGF-3 expression in bone by either estrogens or a TSE/SERM would be consistent with an antiresorptive effect, as this isoform also inhibits in vitro osteoclastic differentiation and activity [200]. In addition to upregulating TGF- expression in osteoblasts, estrogens may act like these peptides in terms of their downstream effects. For instance, 17-estradiol has been reported by Tau et al. [203] to increase the expression of TIEG (TGF- inducible early gene) in conditionally immortalized hFOB/ER9 human fetal osteoblasts. The expression of this gene is also increased by TGF- in human osteoblastic cells [204]. Treatment of this cell line with 17-estradiol, or overexpression of TIEG, causes a reduction in DNA synthesis. These results suggest that at least part of the mechanism by which estrogens inhibit osteoblast proliferation may involve upregulation of TIEG. Estrogens appear to regulate the expression of additional members of the TGF- superfamily. In 1998, Rickard et al. [205] reported that treatment of hFOB/ER9 cells with 17estradiol increased both steady-state mRNA levels and protein levels of BMP-6 (Table 7). In contrast, the steroid hormone had no effect on TGF-1, TGF-2, BMP-2, BMP-4, or BMP-5 expression. Like TGF-s, BMPs also have autocrine and paracrine effects on a variety of skeletal cells [186,189]. More recently, van den Wijngaard et al. [206]

321 reported that antiestrogens or TSEs/SERMs such as tamoxifen, raloxifene, and ICI-164,384 upregulated human BMP-4 promoter-luciferase expression in U2-OS human osteosarcoma cells that were cotransfected with hER but not hER. However, this response required expression of relatively high receptor levels and was blocked by cotreatment with 17-estradiol. As there is no evidence to date that endogenous BMP-4 expression is increased in osteoblasts without ER overexpression, it is unclear whether this observation has any bearing on the pharmacological actions of TSEs/SERMs in the skeleton. In addition to members of the TGF-/BMP family, estrogens have been observed to regulate the expression of components of the osteoblastic IGF/growth hormone (GH) system as well. Gray et al. [207] were the first to report that 17-estradiol treatment upregulated the secretion of IGF-I and IGF-II from UMR-106 rat osteosarcoma cells. These results were confirmed, at least for IGF-I, in three additional osteoblast models, including ROBs (Table 7). Likewise, 17-estradiol increased GH receptor expression and GH action in UMR-106 cells and normal human osteoblast cultures [208]. In contrast, in vivo studies by Turner and coworkers [184,209] in OVX rats failed to verify these in vitro observations. In fact, these authors demonstrated that estrogen loss resulted in an increase in IGF-I mRNA expression in calvarial periosteum and that DES treatment suppressed this increase. Because IGFs increase bone formation, resorption, and turnover [187,188], upregulation of osteoblastic IGF expression following 17-estradiol treatment in vitro is inconsistent with a suppressive effect of the steroid hormone on resorption and turnover in vivo [89 – 91]. However, the in vitro studies were confirmed by Erdmann et al. [210], who showed that supraphysiological doses of 17-estradiol increased IGF-I protein content in femoral shaft bone matrix of OVX rats. However, these authors cautioned that this stimulatory effect of estrogens occurred only at relatively high concentrations of steroid and that this may not be relevant to the normal physiological actions of the hormone. Because high estrogen doses stimulate bone formation in OVX mice [169,172], upregulation of IGF-I levels in bone may be part of the mechanism by which this pharmacological effects occurs. Estrogens also increase IGF-binding protein (IGF-BP) secretion and expression by ROBs [211], hFOB/ER9 [212], and SaOS-2 human osteosarcoma cells [213] (Table 7). IGF-BPs are secreted proteins that bind IGF-I and IGF-II and regulate their bioavailability and activity [192,214]. Consequently, the IGF-BPs can either enhance or inhibit IGF action. Moreover, in some instances, these BPs may also act independently of IGFs. Of the six IGF-BPs, all of which are expressed by human osteoblasts [214], IGF-BP4 is considered to the most inhibitory to IGF activity [192]. In 1996, Kassem et al. [212] demonstrated that 17-estradiol increased IGF-BP4 mRNA expression and secretion in

322 hFOB/ER9 conditionally immortalized fetal human osteoblasts that overexpress hER. In contrast, the steroid had no effect on either IGF-II or IGF-BP3 expression. In addition, 17-estradiol decreased IGF-BP4 proteolysis. Because 17-estradiol also inhibited DNA synthesis by these cells, the authors proposed that upregulation of IGF-BP4 levels in the bone microenvironment might contribute to the suppressive action of estrogens on bone formation observed in vivo [89]. However, Rosen et al. [215] reported that 17-estradiol suppressed IGF-BP3 secretion from a primary culture of human BMSCs. Another potential mechanism by which estrogens may suppress IGF-dependent bone turnover is through antagonism of induced IGF-1 expression. Using ROBs that were contransfected with a human ER expression vector, McCarthy et al. [216] reported that 17-estradiol suppressed PGE2-induced rat IGF-I promoter-luciferase activity. However, basal promoter function was unaffected by the hormone. 4. REGULATION OF FACTORS THAT MODULATE BONE RESORPTION As noted in preceding sections, the therapeutic actions of estrogens in vivo primarily involve the suppression of bone resorption and turnover [5,89]. One of the chief estrogenic targets for these antiresorptive effects are cells of the osteoblast lineage [99,194 – 196]. As outline, in Table 7, at least five different effects of estrogens on osteoblasts and their progenitors involve the suppression of cytokine production, cytokine action, or bone-resorbing hormone activity (see Chapter 13). One of the most commonly reported estrogenic effects in cells of the osteoblast lineage is the downregulation of synthesis of interleukin (IL)-6, a cytokine that promotes differentiation of osteoclast progenitors to mature bone-resorbing cells [138,194 – 196] (see Chapter 41). In 1992, Girasole et al. [217] reported that 17-estradiol suppressed the induction of IL-6 secretion by tumor necrosis factor (TNF)- or IL-1 in mouse / LDA11 stromal cells, MC-3T3-E1 immortalized mouse osteoblastic cells, or primary cultures of rat and human osteoblasts. Moreover, in neonatal mouse calvarial-derived bone cell cultures that contain osteoblasts as well as osteoclast progenitors, 17estradiol inhibited both TNF--stimulated IL-6 production and osteoclast development. In addition, equivalent suppression was also observed with an anti-IL-6 antibody, indicating that IL-6 was involved in this process. These in vitro observations were confirmed by the same group later that year in an in vivo study in mice [218]. These findings were also corroborated by Cheleuitte et al. [219], who used cultured BMSCs isolated from postmenopausal women. These authors showed that basal and IL-1-stimulted IL-6 secretion from incubated BMSCs was reduced significantly relative to age-matched control when cells were isolated

KOMM AND BODINE

from women using estrogen replacement therapy (ERT). The mechanism for the inhibition of IL-6 expression by 17-estradiol was determined by Pottratz et al. [220], who showed that it was through an ER mediated indirect effect on IL-6 promoter activity. Subsequent studies have demonstrated that the ER interferes with NF-B activity, although the precise molecular events involved in this suppression await eluctation [reviewed in 221]. Although several other research groups using a variety of in vitro osteoblast models have corroborated these findings (Table 7), others have been unable to verify IL-6 as a target for estrogen action [222 – 225]. These reports used primary cultures of hOBs or human BMSCs, which are known to express relatively low and variable amounts of ER [106,224]. Our laboratory offered a possible explanation for this discrepancy. Using conditionally immortalized human HOB-03-CE6 cells that naturally express functional ERs [115], we showed that the bone-resorbing cytokines TNF- and IL-1/ are potent suppressors of ligand-dependent receptor activity [129]. In this cell line, 17-estradiol downregulates basal IL-6 mRNA levels [115], but does not block the induction of IL-6 secretion by either TNF- or IL-1 [129]. Thus, we postulated that in osteoblasts that normally express low ER levels, TNF- and IL-1/ may inactivate the receptor before it can blunt IL-6 production. Although Rickard et al. [223] were unable to demonstrate that 17-estradiol suppressed IL-1-induced IL-6 secretion from hOB cells, they did show that the steroid downregulated the release of TNF- from these cells in response to IL-1 stimulation. Estrogens have also been shown to blunt IL-6 responsiveness in osteoblastic and BMSCs cells. The IL-6 receptor is a bipartite complex composed of two transmembrane glycoproteins. One is an 80,000-Da protein (gp80) that binds the cytokine, whereas the other is a dimer of 130,000 Da proteins (gp130) that is involved in signal transduction to the JAK/STAT (Janus kinase/signal transducer and activator of transcription) pathway [226]. Lin et al. [227] reported that 17-estradiol downregulated gp80 and gp130 mRNA levels, as well as gp130 protein levels, in /LDA11 stromal cells. Likewise, the steroid hormone also suppressed the induction of gp130 mRNA by PTH, IL11, or leukemia inhibitory factor (LIF) in MC-3T3-E1 osteoblastic cells. Although cells of the osteoblast lineage produce many proteins that potentiate osteoclastogenesis and osteoclastic activity, one termed RANKL (receptor activator of NF-B ligand) appears to be critical for this process [reviewed in [143,228] (see also Chapters 3, 12, and 13) RANKL is a membrane protein found on the surface of osteoblasts and BMSCs. Moreover, it is the ligand for RANK (receptor activator of NF-B), a transmembrane protein that is expressed by osteoclast progenitors and mature bone-resorbing cells. The binding of RANKL to RANK stimulates the

323

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

differentiation of osteoclast progenitors to mature osteoclasts. Additionally, it activates mature cells. However, RANKL is also a ligand for a secreted decoy receptor called osteoprotegerin (OPG). Osteoblasts and BMSCs synthesize OPG as well as RANKL [99,228], and OPG suppresses bone resorption by sequestering RANKL [99,143,228]. Consequently, given the antiresorptive nature of estrogens, it is not surprising that these hormones have been observed to increase OPG expression by osteoblasts. Using both conditionally immortalized hFOB/ER9 fetal human osteoblastic cells as well as hOBs, Hofbauer et al. [229] demonstrated that 17-estradiol up regulated OPG mRNA levels and increased OPG secretion. One potential mechanism by which estrogens suppress cytokine expression in BMSCs has been elucidated by Srivastava et al. [230]. Using primary cultures of BMSCs isolated from mice, these authors showed that ovariectomy results in increased nuclear levels of phosphorylated Egr-1, a transcription factor that modulates expression of the cytokine macrophage colony-stimulating factor (M-CSF). M-CSF, in turn, is an important inducer (together with RANKL) of osteoclast differentiation [143]. When compared to non phosphorylated Egr-1, the phosphorylated protein binds less well to another transcription factor, Sp-1; this results in increased nuclear levels of free Sp-1, which leads to increased transctivation of the M-CSF gene in BMSCs. Conversely, treatment of wild-type OVX mice with 17-estradiol decreases the levels of phosphorylated Egr-1 in the nucleus of BMSCs and therefore down regulates M-CSF expression. Protein antagonists of IL-1 and TNF- mimic this down regulation. In contrast, 17-estradiol has no effect on M-CSF expression in OVX mice that lack Egr-1. Another commonly reported osteoblastic response to estrogens is the suppression of PTH action. Like estrogens, PTH is an important hormonal regulator of bone metabolism [231] (see Chapter 7). Osteoblasts are the primary targets for PTH action in bone and mediate both anabolic and catabolic activities of this hormone. In fact, one of the bone-resorbing effects of PTH on osteoblastic cells is the upregulation of RANKL expression [143]. As summarized in Table 7, treatment of seven different in vitro osteoblast models with 17estradiol has been shown to block the ability of those cells to respond to PTH. Typically, 17-estradiol has been observed to inhibit the PTH-stimulated increase in intracellular cAMP content [115,177, 181,232,233]. However, the steroid has also been reported to interfere with some of the downstream effects of the peptide as well [198,213,227,234]. In at least one instance, PTH has also been shown to block an estrogenic effect in an osteoblast [203]. Furthermore, the suppressive effect of estrogens on PTH activity has also been observed clinically. Using urinary biochemical markers of bone resorption (see Chapter 60), Cosman et al. [236] reported that postmenopausal women treated with estrogens

exhibited a markedly blunted response to a continuous intravenous infusion of h PTH (1 – 34). The mechanism by which estrogens interfere with PTH signaling is not clear. Using SaOS-2 human osteosarcoma cells, Monroe and Tashjian [233] proposed that suppression was due to a decrease in membrane-associated adenylyl cyclase activity. However, this mechanism does not appear to be applicable to HOB-03-CE6 conditionally immortalized human osteoblasts, as the inhibitory actions of 17-estradiol are selective for PTH over PGE2 and forskolin-stimulated cAMP production [115]. Ernst et al. [237] suggested that the ability of 17-estradiol to reduce PTH-stimulated cAMP production in RCT-3 transformed rat osteoblasts was due to a nongenomic action of the steroid, because it was observed within 4 h of treatment and was not enhanced by overexpression of ER. Although these data are suggestive of a nongenomic effect, they are by no means conclusive. While most studies have demonstrated an antagonistic effect of estrogens on PTH activity or cytokine expression, a few reports have shown the opposite to occur (Table 7). For example, 17-estradiol has been observed to enhance PTH responsiveness. In dexamethasone-conditioned SaOS2 cells, 17-estradiol and PTH potentiate each others stimulatory effect on alkaline phosphatase activity [238]. Whereas in SaOS-2 cells, as well as in primary rat and human osteoblasts, the steroid enhances the ability of PTH to stimulate fibronectin production [239]. Although these reports appear to contradict the antagonistic effects of estrogens on PTH activity in osteoblasts, PTH receptors are coupled to at least two signal transduction pathways [240], and estrogens may act differentially on these second messenger systems. Likewise, using a T-antigen-transformed human osteoblast cell line (HOBIT), Pivirotto et al. [241] presented evidence that 17-estradiol upregulates IL-1 mRNA levels. However, because this effect has only been reported to occur in HOBIT cells, its biological significance is questionable. 5. REGULATION OF RECEPTOR EXPRESSION SIGNAL TRANSDUCTION

AND

Estrogens modulate the expression of several receptors in osteoblasts. At least three members of the nuclear receptor superfamily are known to be regulated by these steroids. As occurs in uterine and breast cells, treatment of either hOBs or conditionally immortalized hFOB/ER9 cells with 17-estradiol upregulates progesterone receptor (PR) expression [106,225,242]. The steroid has also been observed to increase vitamin D receptor (VDR) levels and vitamin D responsiveness in two osteosarcoma cell lines [243,244]. In addition, it either increases [127,128] or decreases [128] ER mRNA levels in primary cultures of human and rat osteoblasts, respectively. In the case of ROB cells, our laboratory demonstrated that 17-estradiol downregulates ER expression in day 14 nodule-forming cultures (osteoblastic

324 cells), while it upregulates receptor expression in day 30 late mineralization stage cultures (osteocytic cells) [128]. Consistent with these observations, Hoyland et al. [136] have reported that ERT or hormone replacement therapy (HRT) decreases the number of ER mRNA-positive osteoblasts in human bone biopsies. However, ERT/HRT increases the number of ER protein-positive osteocytes in these biopsies. Thus, estrogens play a role in both directly regulating osteoblastic activity and modulating the hormonal responsiveness of the cells. Estrogens have also been reported to regulate additional signal transduction pathways in osteoblasts (Table 7). One interesting finding is that 17-estradiol downregulates mRNA expression of the type I inositol trisphosphate (IP3) receptor in several in vitro osteoblast models [245]. This receptor is a transmembrane calcium channel found on the “calciosome,” a specialized component of endoplasmic reticulum that is involved in the storage and release of IP3 sensitive intracellular calcium [246]. This receptor is therefore essential for the phosphoinositide-signaling pathway. Because bone-resorbing agents such as PTH, prostaglandins, and bradykinin utilize this pathway, suppression of type I IP3 receptor expression by estrogens in osteoblasts may lead to decreased bone resorption and turnover. Although the human type I IP3 receptor promoter does not contain a consensus ERE, 17-estradiol nevertheless downregulates promoter – reporter gene constructs when transfected transiently into G292 human osteosarcoma cells [247]. Another interesting observation is the upregulation of endothelial nitric oxide synthase (ecNOS or NOS-1) mRNA expression and enzyme activity in estrogen-treated HOS TE-85 human osteosarcoma cells [248]. Because a high nitric oxide (NO) content has been reported to inhibit in vitro osteoclastic bone resorption [249], this estrogenic effect is also consistent with an antiresorptive role for the steroid. An in vivo study with OVX rats confirms these results. Wimalawansa et al. [250] reported that treatment of OVX rats with either 17-estradiol or nitroglycerine (an NO donor) reversed lumbar spine bone loss as measured by dual-energy X-ray absorptiometry (DXA). In contrast, cotreatment with 17-estradiol and NG-nitro-L-arginine methyl ester (L-NAME, a NOS inhibitor) blocked the bone-sparing effects of the steroid hormone. In contrast to these observations regarding basal NO production, Van Bezooijen et al. [251] reported that 17-estradiol treatment of mouse-immortalized MC-3T3-E1 osteoblasts suppressed cytokine-induced (NOS-2 mediated) NO synthesis. This finding may reflect the generally antagonistic nature of estrogens toward cytokine action (i.e., IL-1 and TNF-) in the skeleton. Finally, pretreatment of hOBs with 17-estradiol has been reported to increase bradykinin responsiveness as measured by the release of arachindonic acid from the cells [252]. However, because bradykinin stimulates bone re-

KOMM AND BODINE

sorption, the physiological significance of this observation is unclear. 6. MISCELLANEOUS RESPONSES As outlined at the end of Table 7, treatment of several rodent osteoblastic cell models with 17-estradiol has been reported to increase creatine kinase (CK) [171], heat shock protein (HSP)-27 [253] and aspartate aminotransferase (AST), -glutamyl transferase (GGT), lactate dehydrogenase (LDH), and transferrin [254]. However, the physiological or therapeutic significance of these responses is unclear. The upregulation of CK activity by 17-estradiol was also observed in rat bone in vivo, and this may represent another anabolic effect of the steroid [171]. 7. SUMMARY As described in the preceding sections, about a third (15/43) of the estrogenic responses observed in a broad range of in vitro osteoblast and BMSC models are consistent with the suppressive effects of estrogens on bone resorption and bone turnover in vivo. In other instances, such as anabolic effects, a disconnection occurs between the in vitro responses and the in vivo physiology of these steroids. In vivo, increased bone turnover on estrogen depletion is driven primarily by increased osteoclastic bone resorption and the subsequent inadequate ability of osteoblastic bone formation to keep pace with this accelerated bone loss [89 – 91]. However, in vitro studies with osteoblasts are almost always performed with pure cultures of cells (i.e., cloned osteoblastic cell lines) and in the absence of osteoclasts. Consequently, the opportunity for coupling between the two cell types is lost [255]. Thus, in isolation, estrogens appear to exert both stimulatory and inhibitory effects on osteoblastic function. In some in vitro models, such as hFOB/ER9 cells [168] or ROBs [128], these differential effects seem to occur as a result of changes that arise during cellular differentiation. However, it is not known if estrogens have divergent actions on osteoblasts as they undergo maturation in vivo. Another possible explanation for the apparent anabolic effects of estrogens on osteoblasts in vitro is that these may represent a pharmacological response to the steroid and not a physiological one [174].

B. Estrogenic Responses in Osteocytes Only a few of estrogenic responses have been observed in osteocytes, and all of these reports come from in situ studies. In what may well be the first publication on this subject, Whitson [256] described the results of an electron microscopic analysis of metatarsal bones isolated from vehicle and 17-estradiol-n-valerate-treated female rabbits. Although the results were not quantitative, the author noted that the number of tight junctions (possibly gap junctions)

325

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

formed between osteocytes was greater in bones from estrogen-treated animals. Moreover, he suggested that this increased tight junction formation might be related to an accelerated osteogenesis. Twenty-five years later, Tomkinson et al. [257] reported the findings of a clinical study of premenopausal women who were treated with a gonadotropin-releasing hormone (GnRH) analogue for endometriosis. Transiliac biopsies were taken before and after GnRH analogue therapy, which resulted in a dramatic decrease in serum 17-estradiol concentrations. Although osteocyte lacuna density was not affected by the treatment, the percentage of lacunae containing viable osteocytes (as determined by cell-associated lactate dehydrogenase activity) was reduced in all but one of the six patients, suggesting that estrogen deficiency is associated with increased osteocyte apoptosis [133]. Because one of the functions of osteocytes is to serve as mechanosensors [132,134,258], these observations also implied that estrogen deficiency could lead to increased bone fragility (and therefore increased fracture) at weight-bearing skeletal sites with or without an accompanying net bone loss. The same group confirmed this clinical study the following year using OVX rats [259]. In this preclinical model of estrogen deficiency, OVX increased the number of apoptotic osteocytes (as determined by DNA strand fragmentation) in both trabecular and cortical bone of the tibia. In addition, repletion with 17-estradiol reversed this increase and returned the apoptotic index to sham values. In another in situ study of OVX rats, lkeda et al. [260] observed that osteopontin mRNA expression increased after OVX in osteocytes that were located in metaphyseal trabecular bone of the femur, but not in those found in the epiphysis. Because osteopontin is one of the bone matrix proteins to which osteoclasts are known to bind [104], these data suggested a possible role for osteocytes in regulating bone resorption. Our laboratory has also presented evidence that osteocytic cells may play a role in modulating osteoclastic activity [261]. Using a conditionally immortalized human preosteocytic (i.e., osteoid osteocyte) cell line (HOB-01C1), we showed that these cells secrete high amounts of IL-6 and monocyte chemoattractant protein (MCP)-1 in response to treatment with the bone-resorbing cytokines IL-1 and TNF-. Together, IL-6 and MCP-1, in addition to other factors, might stimulate osteoclast differentiation and recruitment to a specific bone-remodeling site. Another potential regulatory target for estrogens in osteocytes is ER. Using immunofluorescence to study ER protein expression in human bone biopsies, Braidman and colleagues [124,136] reported that ERT/HRT increases the number of ER protein-positive osteocytes and osteoblasts. Curiously, the number of ER mRNA-positive osteoblasts was observed to decline with ERT/HRT [136]. As noted previously, osteocytes are postulated to serve as mechansenors [132,134,258]. As such, they are thought

to translate the effects of weight bearing or weightlessness into either increases or decreases in bone mineral density, respectively. Several studies suggest that estrogens regulate the process of mechanosensory stimulation, and that mechanical strain and estrogen action may share common signaling pathways. Using organ cultures of rat ulnae isolated from female rats, Cheng et al. [262,263] reported that both 17-estradiol and mechanical loading stimulated [3H]thymidine and [3H]proline incorporation into the bones. Moreover, when the treatments were combined, a synergistic effect was observed. Thus, estrogens appeared to enhance the osteogenic response of the bones to mechanical strain. A subsequent study by the same group using primary cultures of rat long bone-derived osteoblasts demonstrated that both 17-estradiol and mechanical strain increase cellular DNA synthesis [167]. Furthermore, these increases were suppressed by cotreatment with the antiestrogen ICI-182,780. Although osteoblasts are probably not the targets for mechanical loading in vivo [258], these results nevertheless suggest that mechanical strain can activate the ER. The observation that mechanical strain and estrogens appear to share common signal transduction pathways is supported by an in vivo study of Westerlind et al. [264]. Using OVX rats, these authors showed that estrogen deficiency resulted in a preferential lose of cancellous bone from a site that experiences low mechanical strain (distal femur metaphysis), whereas one that experiences high-strain energies (distal femur epiphysis) does not lose bone (even though bone turnover was increased at both sites). In addition, increased mechanical loading (treadmill exercise) suppressed OVX-induced cancellous bone loss from the proximal tibial metaphysis. Conversely, treatment of OVX animals with 17-estradiol suppressed tibial cancellous bone loss that resulted from decreased mechanical loading (unilateral sciatic neurotomy). Finally, there is also evidence that these preclincal findings may translate to humans. For example, in a small clinical study of postmenopausal women, Kohrt et al. [265] reported that HRT and weight-bearing exercise had an additive effect on total body bone mineral accretion. Thus, the efficacies of HRT and weight-bearing exercise on the skeleton seem to be enhanced by concurrent use. Although the just-mentioned studies do not specifically address the role of estrogens in osteocyte biology per se, the implications of this work is that osteocytes — as the major mechanosensory cell in bone — are at least one of the targets for these effects.

C. Estrogenic Responses in Cells of the Osteoclast Lineage In addition to indirectly inhibiting bone resorption through cells of the osteoblast lineage, estrogens have also

326 been reported to have direct suppressive effects on cells of the osteoclast lineage [142]. The most extensive evidence for a direct inhibitory effect of estrogens on mature osteoclasts comes from the work of Oursler [142]. Using both avian and hGCT-derived osteoclasts that were highly purified (90% homogeneous) with an osteoclast-specific monoclonal antibody (121F), Oursler reported that 17-estradiol inhibits in vitro bone resorption by these preparations [145,146,150,266 – 268]. Estrogenic responses in these studies include the upregulation of c-fos, c-jun, TGF-2, TGF3, and TGF-4 mRNA levels; the downregulation of tartrate-resistant acid phosphatase (TRAP), cathepsin B, cathepsin D, LEP-100, and lysozyme message levels; the induction of total TGF- protein secretion (due mostly to an increase in TGF-3); and the suppression of TRAP, cathepsin B, cathepsin L, and -glucuronidase activity, as well as lysozyme protein production. The majority of these effects are consistent with an estrogen-mediated decrease in osteoclast activity and subsequent bone resorption. For example, TGF- is an inhibitor of bone resorption, whereas lysozomal proteases like the cathepsins are involved in digesting the bone matrix [142]. Confirmation that estrogens suppress osteoclastic gene expression in vivo comes from the studies of Zheng et al. [269], who demonstrated that treatment of OVX rats with 17-estradiol decreased the expression of TRAP mRNA in bone. Additional support for a direct effect of estrogens on osteoclasts comes from the work of Sunyer et al. [147]. Employing normal human osteoclasts (hOCLs) that were also purified to 90% homogeneity with the 121F monoclonal antibody, these authors reported that 17-estradiol decreased the mRNA levels of the signaling receptor for IL-1 (IL-1RI) and increased the message levels of the IL-1 decoy receptor (IL-1RII). This change in receptor expression correlated with a suppression of IL-1-mediated IL-8 expression by the steroid hormone. Moreover, 17estradiol pretreatment abrogated the reduction of hOCL apoptosis by IL-1. Finally, Mano et al. [148] have demonstrated that 17-estradiol also inhibits the in vitro bone resorption of purified rabbits osteoclasts and reduces the expression of cathepsin K mRNA by these cells. However, some studies have failed to detect a direct inhibitory effect of estrogens on mature osteoclasts. For example, Williams et al. [270] were unable to suppress bone resorption of purified avian osteoclasts with either 17estradiol or DES. However, high ( M) levels of the TSE/ SERM tamoxifen decreased osteoclast activity. Likewise, calmodulin antagonists had a similar effect. Additional experiments led the authors to conclude that tamoxifen acted through a membrane-associated target to suppress osteoclastic bone resorption independently of the ER. This target appeared to be similar or related to the target for the calmodulin inhibitors. As indicated earlier and in preceding sections, estrogens have been observed to increase the expression of TGF- by

KOMM AND BODINE

both osteoblasts and osteoclasts. In addition, these steroids suppress osteoblast apoptosis, but enhance programmed osteoclast cell death [271,272]. Hughes et al. [273] elegantly demonstrated a connection among estrogens, TGF-, and osteoclast apoptosis. These authors showed that treatment of marrow culture-derived murine osteoclasts with 17 estradiol increased the percentage of cells undergoing apoptosis. Likewise, treatment of the cultures with TGF-1 also increased osteoclast apoptosis. Moreover, the induction of osteoclast programmed cell death by 17-estradiol was blocked by coincubation with a pan-specific TGF- antibody. Consistent with its bone-sparing effects [201], treatment of the osteoclast-containing cultures with tamoxifen also increased apoptosis of these cells. These in vitro observations were confirmed by an in vivo study in which OVX mice were treated with 17-estradiol. Because the marrow culture system used by Hughes et al. [273] was a heterogeneous cell population, the promotion of osteoclast apoptosis by 17-estradiol could have resulted from either a direct action of the steroid on osteoclasts or an indirect effect on another cell type like osteoblasts or BMSCs. In addition to inducing apoptosis of mature osteoclasts, estrogens may also have similar effects on osteoclast progenitors. Zecchi-Orlandini et al. [274] reported that 17 estradiol induced apoptosis of the human monoblastic leukemia cell line FLG 29.1, which has characteristics resembling preosteoclasts. Moreover, treatment of this cell line with the TSE/SERM raloxifene [201] also induced apoptosis [275]. FLG 29.1 cells can be stimulated to form osteoclast-like cells in vitro by treatment with phorbol ester, 1,25(OH)2vitamin D, or osteoblast-derived factors [276]. These agents also induce the expression of a novel superoxide dismutase-related membrane glycoprotein, which is the osteoclast-specific antigen that is recognized by the 121F monoclonal antibody. Incubation of cells with 17-estradiol suppresses the induction of this antigen by phorbol ester [276]. Thus, these results suggest that estrogens may also suppress osteoclast differentiation by acting directly on their progenitors. Additional reports also indicate that estrogens can suppress osteoclast differentiation. Schiller et al. [277] have demonstrated that 17-estradiol antagonizes the induction of osteoclast-like cell formation by 1,25(OH)2 vitamin D in primary cultures of mouse bone marrow cells. In addition, these authors showed that the ability of 1,25(OH)2 vitamin D to stimulate osteoclast differentiation is at least partially mediated by an upregulation of IL-6 secretion and that 17-estradiol blocks this effect as well. Estrogens also suppresses PTH-stimulated osteoclast formation. Using primary mouse hemopoietic blast cell cultures, which were reportedly free of stromal cells and osteoblasts, Kanatani et al. [152] presented evidence that these osteoclast precursors contain PTH receptor mRNA based on RT-PCR. These cells also express ER message. Treatment of mouse

327

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

hemopoietic blast cell cultures with either 1,25(OH)2 vitamin D or PTH (1 – 34) induces the formation of osteoclastlike cells (i.e., TRAP-positive multinucleated cells). However, cotreatment of the cultures with 17-estradiol blunts the stimulation of osteoclast differentiation by PTH, but not by 1,25(OH)2 vitamin D. These authors also demonstrated that 17-estradiol blocks osteoclast-like cell formation induced by agents that activate adenylyl cyclase or mimic cAMP, but not ones that activate protein kinase C or increase intracellular calcium. Although an earlier report from the same group suggested that estrogens suppress PTH-induced osteoclast differentiation indirectly through an effect on osteoblasts [235], the study by Kanatani et al. [152] concluded that this inhibitory effect may also be due to a direct action on osteoclast progenitor cells. In summary, there is substantial evidence to conclude that estrogens inhibit osteoclast differentiation and activity in two ways: one that occurs indirectly via the osteoblast and stromal cell and a second that occurs directly through interaction with the ER in osteoclast progenitors and mature osteoclasts. However, as with other aspects of estrogen action on bone cells, this area of research is controversial.

D. Estrogenic Responses in Chondrocytes Another important target cell in the skeleton for estrogens is the chondrocyte. As noted in an earlier section, these cells have been shown to express both ER and ER. Moreover, chondrocytes have also been reported to exhibit estrogenic responses. In vivo, it is well known that estrogens accelerate endochondral growth during puberty and potentiate epiphyseal closure at the end of the growth spurt [89] (see Chapter 25). Consistent with these physiological responses, 17-estradiol has been observed to decrease the in vitro proliferation and/or DNA synthesis of embryonic duck [278] and rat chondrocytes [279]. In duck chondrocytes, 17-estradiol also suppressed sulfated proteoglycan synthesis [278], whereas in fetal rabbit [280] and human chondrocytes [281], the steroid had the opposite effect. Additional in vitro estrogenic effects in rat chondrocytes include the upregulation of alkaline phosphatase activity and collagen production, which are consistent with a potentiation of cellular differentiation by the steroid [279].

VIII. ESTROGEN-RELATED RECEPTOR- AND OSTEOPONTIN GENE EXPRESSION In addition to expressing ER and ER, osteoblasts also express a related member of the nuclear receptor superfamily known as estrogen-related receptor (ERR)-1 or  [282 – 284]. ERR- is an orphan receptor that shares

68% amino acid identity with ER and ER in the DNAbinding domain, but only 36% identity in the ligand-binding domain [284]. Consequently, it does not bind 17estradiol but instead is constitutively active in serum-containing medium [284]. However, this constitutive activity is diminished upon charcoal treatment of the serum [284]. ERR-, as well as the related ERR-, transactivates promoters containing either an ERE or a SF-1-response element (SFRE) [284]. ER also binds to both of these DNA response elements, whereas ER does not bind to the SFRE [284]. ERR- mRNA is highly expressed in the ossification zones of the developing mouse skeleton (long bones, vertebrae, ribs, and skull), as well as in some human osteosarcoma cell lines (HOS-TE85 and SaOS-2) and hOBs [282]. Given this expression pattern, as well as the knowledge that the osteopontin promoter contains an SFRE, it is perhaps not surprising that cotransfection of rat ROS 17/2.8 osteosarcoma cells with ERR and an osteopontin promoter – reporter gene construct resulted in the transactivation of this promoter [282 – 284]. Moreover, transient transfection of ROS 17/2.8 cells and immortalized mouse MC-3T3-E1 cells with ERR produced an upregulation of endogenous osteopontin mRNA levels [283]. Taken together, these data demonstrate that osteopontin gene expression in the osteoblast is regulated not only by ER in an estrogen-dependent manner, but also by ERR in an estrogen-independent manner [284]. In contrast, ER does not appear to regulate this gene [284]. Thus, these observations also point to a potential functional difference between the biological roles of ER and ER in the osteoblast. However, because osteopontin is an apparent binding site for osteoclasts to the bone matrix [104], the physiological significance of its upregulation by estrogens via either ER or ERR- in a ligand-independent manner is unclear.

IX. NONGENOMIC ACTIONS OF ESTROGENS IN BONE CELLS Although the majority of estrogenic effects are believed to be mediated by one of the nuclear ERs, some responses may also originate at the plasma membrane [285,286]. Estrogens have been reported to produce rapid effects (within seconds or minutes) on a variety of cell types, including bone cells [285,286]. These nongenomic actions are thought to be mediated via a membrane receptor. However, it is unclear whether this receptor is a membrane-localized form of a nuclear ER or if it is a distinct transmembrane protein, such as a G-protein-coupled receptor (GPCR) [285,286]. In a series of papers using primary female rat osteoblasts, Lieberherr and co-workers presented convincing evidence for rapid, membrane-derived effects of 17estradiol [287 – 289]. Treatment of ROB cells with low con-

328 centrations (1 pM – 1 nM) of 17-estradiol increased intracellular calcium concentrations within 10 to 30 [287]. Through the use of various inhibitors, the source of this calcium was shown to be both extracellular, via plasma membrane channels, and intracellular from the endoplasmic reticulum or calciosome. Cells within the same time frame also produced IP3 and diacylglycerol (DAG) after treatment with the steroid. Because inhibitors of both phospholipase C (PLC) and Gi proteins blocked the release of IP3 and DAG, the authors concluded that 17-estradiol acted through a GPCR [287]. Consistent with estrogens working through a distinct membrane receptor and not simply a membrane-localized ER, tamoxifen was neither an agonist nor an antagonist of 17-estradiol. Subsequent studies by this group refined the model to include activation of PLC2 by  subunits [288,289]. In contrast, 1,25(OH)2 vitamin D, which also has rapid effects on female ROB cells, was shown to act via modulation of PLC-1 by G (q/11) [288,289]. A potential downstream target for the rapid generation of a membrane-derived signal by 17-estradiol was reported by Endoh et al. [290]. These authors showed that treatment of ROS 17/2.8 cells with 17-estradiol activated the mitogen-activated protein kinase (MAPK) within 5 min. Estrogens may also produce rapid nongenomic effects in cells of the osteoclast lineage [41,291 – 293]. For example, using the human preosteoclastic cell line FLG 29.1, Fiorelli et al. [61] demonstrated that 17-estradiol stimulated an increase in intracellular pH within 50, as well as an increase in intracellular cAMP and cGMP after 30 min. In addition, Brubaker and Gay [293] reported that treatment of isolated avian osteoclasts with 17-estradiol caused a depolarization of the plasma membrane potential within seconds of adding the steroid to the cells. The mechanism for the depolarization appeared to be due to the regulation of potassium channel activity. The net effect of this rapid, nongenomic estrogenic response could be an inhibition of osteoclastic acidification.

X. SUMMARY Estrogens clearly play a critical role in bone biology. The burst in research aimed at elucidating the functional role of estrogens in bone remodeling that has occurred since the mid-1970s has led to the discovery of a multitude of potential pathways that are affected by estrogens in the skeleton. The sheer abundance of estrogenic-related regulated events in bone cells supports the contention that estrogens, working through their receptors, play key roles in the development and maintenance of a normal skeleton. Questions that remain to be answered relate to the differences in the skeletal response to the various types of estrogens (estradiol vs phytoestrogens vs SERMS). All estrogens do not evoke the same response in bone whether looking at a

KOMM AND BODINE

specific gene’s regulation in isolated osteoblasts or a global skeletal response in vivo. Why this occurs is not known. The complexity of the bone-remodeling process coupled with multiple sites where an estrogen could elicit an effect will make it difficult to fully answer the question, but as technology advances, so will the possibility of answering our tough questions.

References 1. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis: Its clinical features. JAMA 116, 2465 – 2474 (1941). 2. A. W. Diddle and I. Q. Smith, Postmenopausal osteoporsis: The role of estrogens. South. Med. J. 77, 868 – 874 (1984). 3. M. C. Chapuy and P. J. Meunier, Prevention and treatment of osteoporosis. Aging Clin. Exp. Res. 7, 164 – 173 (1995). 4. J. P. Bilezikian, Estrogens and postmenopausal osteoporosis: Was Albright right after all? J. Bone Miner. Res. 13, 774 – 776 (1998). 5. R. Eastell, Treatment of postmenopausal osteoporosis. Engl. J. Med. 338, 736 – 746 (1998). 6. A. Horsman, M. Jones, R. Francis, and C. Nordin, The effect of estrogen dose on postmenopausa bone loss. N. Engl. J. Med. 309, 1404 – 1407 (1983). 7. T. M. Willson, Chemistry: Structure and function relationships. Estrogen Antiestrogens Basic Clin. Aspects 21 – 22 (1997). 8. M. O Connell, Pharmacokinetic and pharmacologic variation between different estrogen products. J. Clin. Pharmacol. 35, 18s – 24s (1995). 9. H. U. Bryant and W. H. Dere, Selective estrogen receptor modulators: An alternative to hormone replacement therapy (44204). P.S.E.B.M. 217, 45 – 52 (1998). 10. T. A. Grese and J. A. Dodge, Selective estrogen receptor modulators (SERMs). Curr. Pharm. Design 4, 71 – 92 (1998). 11. V. Kumar, S. Green, G. Stack, M. Berry, J.-R. Jin, and P. Chambon, Functional domains for the human estrogen receptor. Cell. 51, 941 – 951 (1987). 12. S. Green, P. Walter, G. Greene, A. Krust, C. Goffin, E. Jensen, G. Scrace, M. Waterfield, and P. Chambon, Cloning of the human oestrogen receptor cDNA. J. Steroid Biochem. 24, 77 – 83 (1986). 13. P. Walter, S. Green, G. Greeene, A. Krust, J.-M. Bornert, J.-M. Jeltsch, A. Staub, E. Jensen, G. Scrace, M. Waterfield, and P. Chambon, Cloning of the human estrogen receptor cDNA. Proc. Natl. Acad. Sci. USA 82, 7889 – 7893 (1985). 14. G. J. M. Kuiper, E. Enmark, M. Pelto-Huikko, S. Nilsson, and J. Gustafsson, Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. USA 93, 5925 – 5930 (1996). 15. S. Mosselman, J. Polman, and R. Dijkema, ER-beta: Identification and characterization of a novel human estrogen receptor. FEBS Lett. 392, 49 – 53 (1996). 16. B. O Malley, The steroid receptor superfamily: More excitement predicted for the future. Mol. Endocrinol. 4, 363 – 369 (1990). 17. C. Weinberger, V. Giguere, S. Hollenberg, M. G. Rosenfeld, and R. M. Evans, Human steroid receptors and erbA proto-oncogene products: Members of a new superfamily of enhancer binding proteins. Cold Spring Harbor Symp. Quant. Biol. 51, 759 – 772 (1986). 18. Q. Zhao, S. Khorasanizadeh, Y. Miyoshi, M. A. Lazar, and F. Rastinejad, Structural elements of an orphan nuclear receptorDNA complex. Mol. Cell. 1, 849 – 861 (1998). 19. B. Blumberg and R. M. Evans, Orphan nuclear receptors: New ligands and new possibilities. Genes Dev. 12, 3149 – 3155 (1998). 20. R. Kumar and E. Thompson, The structure of the nuclear hormone receptors. Steroids 64, 310 – 319 (1999).

CHAPTER 10 Regulation of Bone Cell Function by Estrogens 21. L. Klein-Hitpass, S. Y. Tsai, G. L. Greene, J. H. Clark, M.-J. Tsai, and B. W. O’Malley, Specific binding of estrogen receptor to the estrogen response element. Mole. Cell. Biol. 9, 43 – 49 (1989). 22. G. Klock, U. Strahle, and G. Schutz, Oestrogen and glucocorticoid responsive elements are closely related by distinct. Nature 329, 734 – 736 (1987). 23. M. T. Tzukerman, A. Esty, D. Santiso-Mere, P. Danielian, M. G. Parker, R. B. Stein, J. W. Pike, and D. P. McDonnell, Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two fuctionally distinct intramolecular regions. Mol. Endocrinol. 8, 21 – 30 (1994). 24. K. Inno, M. Haino, M. Iwasaki, N. Ono, T. Horigome, and H. Sugano, Reconstruction of the 9 S estrogen receptor with heat shock protein 90. FEBS. Lett. 267, 157 – 159 (1990). 25. J. Devin-Leclerc, X. Meng, F. Delahaye, P. Leclerc, E.-E. Baulieu, and M.-G. Catelli, Interaction and dissociation by ligands of estrogen receptor and Hsp90: The antiestrogen RU 58668 induces a protein synthesis-dependent clustering of the receptor in the cytoplasm. Mol. Endocrinol. 12, 842 – 854 (1998). 26. V. Kumar and P. Chambon, The estrogen receptor binds tightly to its responsive element as a ligand-induced homodimer. Cell 55, 145 – 156 (1988). 27. F. Auricchio, Phosphorylation of steroid receptors. J. Steroid Biochem. 32, 613 – 622 (1989). 28. W. W. Grody, W. T. Schrader, and B. W. O Malley, Activation transformation, and subunit structure of steroid hormone receptors. Endocr. Rev. 3, 141 – 163 (1982). 29. B. Katzenellenbogen, B. Bhardwaj, H. Fang, B. Ince, F. Pakdel, J. Reese, D. Schodin, and C. Wrenn, Hormone binding and transcription activation by estrogen receptors: Analyses using mammalian and yeast systems. J. Steroid Biochem. Mol. Biol. 47, 39 – 48 (1993). 30. N. J. McKenna, R. B. Lanz, and B. W. O Malley, Nuclear receptor coregulators: Cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 31. J. D. Chen and H. Li, Coactivation and corepression in transcriptional regulation by steroid/nuclear hormone receptors. Crit. Rev. Eukary. Gene Express. 8, 169 – 190 (1998). 32. P. Clement-Lacroix, C. Ormandy, L. Lepescheux, P. Ammann, D. Damotte, V. Goffin, B. Bouchard, M. Amling, M. Gaillard-Kelly, N. Binart, R. Baron, and P. A. Kelly, Osteoblasts are a new target for prolactin: Analysis of bone formation in prolactin receptor knockout mice. Endocrinology 140, 96 – 105 (1999). 33. N. Huang, E. vom Baur, J.-M. Garnier, T. Lerouge, J.-L. Vonesch, Y. Lutz, P. Chambon, and R. Losson, Two distinct nuclear receptor interaction domains in NSD1, a novel SET protein that exhibits characteristics of both copressors and coactivators. EMBO J. 17, 3398 – 3412 (1998). 34. R. M. Lavinski, K. Jepsen, T. Heinzel, J. Torchia, T.-M. Mullen, R. Schiff, A. L. Del-Rio, M. Ricote, S. Ngo, J. Gemsch, S. G. Hilsenbeck, C. K. Osborne, C. K. Glass, M. G. Rosenfeld, and D. W. Rose, Diverse signaling pathways modulate nuclear receptor recruitment of N-CoR and SMRT complexes. Proc. Natl. Acad. Sci. USA 95, 2920 – 2925 (1998). 35. C. M. Klinge, Estrogen receptor interaction with co-activators and co-repressors. Steroids 65, 227 – 251 (2000). 36. J. T. Kadonaga, Eukaryotic transcription: An interlaced network of transcription factors and chromatin-modifying machines. Cell 92, 307 – 313 (1998). 37. E. Korzus, J. Torchia, D. W. Rose, L. Xu, R. Kurokawa, E. M. Mclnerney, T. Mullen, C. K. Glass, and M. G. Rosenfeld, Transcription factor-specific requirements for coactivators and their acetyltransferase functions. Science 279, 703 – 707 (1998). 38. H. Y. Mak, S. Hoare, P. M. A. Henttu, and M. G. Parker, Molecular determinants of the estrogen receptor-coactivator interface. Mol. Cell. Biol. 19, 3895 – 3903 (1999).

329 39. P. M. A. Henttu, E. Kalkhoven, and M. G. Parker, AF-2 activity and recruitment of steroid receptor coactivator 1 to the estrogen receptor depend on a lysine residue conserved in nuclear receptors. Mol. Cell. Biol. 17, 1832 – 1839 (1997). 40. A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A. Agard, and G. L. Greene, The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell. 95, 927 – 937 (1998). 41. W. Feng, R. C. J. Ribeiro, R. L. Wagner, H. Nguyen, J. W. Apriletti, R. J. Fletterick, J. D. Baxter, P. J. Kushner, and B. L. West, Hormone-dependent coactivator binding to a hydrophobic, cleft on nuclear receptors. Science 280, 1747 – 1749 (1998). 42. W. L. Kraus and J. T. Kadonaga, p300 and estrogen receptor cooperatively activate transcription via differential enhancement of initiation and reinitiation. Genes Dev. 12, 331 – 342 (1998). 43. C. Rachez, Z. Suldan, J. Ward, C.-P. B. Chang, D. Burakov, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, A novel protein complex that interacts with the vitamin D3 receptor in a liganddependent manner and enhances VDR transactivation in a cell-free system. Genes. Dev. 12, 1787 – 1800 (1998). 44. L. P. Freedman, Increasing the complexity of coactivation in nuclear receptor signaling. Cell 97, 5 – 8 (1999). 45. M. Ito, C.-X. Yaun, S. Malik, W. Gu, J. D. Fondell, S. Yamamura, Z.-Y. Fu, X. Zhang, J. Qin, and R. G. Roeder, Identity between TRAP and SMCC complexes indicates novel pathways for the function of nuclear receptors and diverse mammalian activators. Mol. Cell 3, 361 – 370 (1999). 46. S. W. Curtis, T. Washburn, C. Sewall, R. DiAugustine, J. F. Couse, and K. S. Korach, Cross-talk between growth factors and estrogen receptor. Estrogen Vessel Wall 31 – 41. 47. B. S. Katzenellenbogen and M. J. Norman, Multihormonal regulation of the progesterone receptor in MCF-7 human breast cancer cells: Interrelationships among insulin/insulin-like growth factor-I, serum, and estrogen. Endocrinology 126, 891 – 898 (1990). 48. P. Joel, A. Traish and D. Lannigan, Estradiol-induced phosphorylation of serine 118 in the estrogen receptor is independent of p42/p44 mitogen-activated protein kinase. J. Biol. Chem. 273, (1998). 49. H. Cho and B. Katzenellenbogen, Synergistic activation of estrogen receptor-mediated transcription by estradiol and protein kinase activators. Mol. Endocrinol. 7, 441 – 452 (1993). 50. S. Migliccio, S. Bernardini, W. Westsel, K. Korach, T. Faraggiana, and A. Teti, Protein kinase C modulates estrogen receptors in differentiated osteoblastic cells in vitro. Steroids 63, 352 – 354 (1998). 51. D. Ignar-Trowbridge, M. Pimentel, M. Parker, J. McLachlan, and K. Korach, Peptide growth factor cross-talk with the estrogen receptor requires the A/B domain and occurs independently of protein kinase C or estradiol. Endocrinology 137, 1735 – 1744 (1996). 52. D. Ignar-Trowbridge, M. Pimentel, C. Teng, K. Korach, and J. McLachlan, Cross talk between peptide growth factor and estrogen receptor signaling systems. Environ Health Perspect. 103, 35 – 38 (1995). 53. T. Improta-Brears, A. Whorton, F. Codazzi, J. York, T. Meyer, and D. McDonnell, Estrogen-induced activation of mitogen-activated protein kinase requires mobilization of intracellular calcium. Proc. Natl. Acad. Sci. USA 96, 4686 – 4691 (1999). 54. S. Manolagas, The role of IL-6 type cytokines and their receptors in bone. Ann. N. Y. Acad. Sci. 98, 194 – 204 (1998). 55. T. Bellido, C. O’Brien, P. Roberson, and S. Manolagas, Transcriptional activation of the p21WAF1, C1P1, SD11 gene by interluekin-6 type cytokines. J. Biol. Chem. 273, 21137 – 21144 (1998). 56. G. Cerillo, A. Rees, N. Manchanda, C. Reilly, I. Brogan, A. White, and M. Needham, The oestrogen receptor regulates NFB and AP-1 activity in a cell-specific manner. J. Steroid. Biochem. Mol. Bio. 67, 79 – 88 (1998).

330 57. P. Webb, G. Lopez, R. Uht, and P. Kushner, Tamoxifen activation of the estrogen receptor/AP-1 pathway: Potential origin for the cellspecific estrogen-like effects of antiestrogens. Mol. Endocrinol. 9, 443 – 456 (1995). 58. C. Qin, P. Singh, and S. Safe, Transcriptional activation of insulinlike growth factor-binding protein-4 by 17-estradiol in MCF-7 cells: Role of estrogen receptor-Sp1 complexes. Endocrinology 140, 2501 – 2508 (1999). 59. Y. Sadovsky and P. Crawford, Developmental and physiological roles of the nuclear receptor steroidogenic factor-1 in the reproductive system. J. Soc. Gynecol. Invest. 5, 6 – 12 (1998). 60. J. Bertherat, The nuclear receptor SF-1 (steroidogenic factor-1) is no longer an orphan. Eur. J. Endocrinol. 138, 32 – 33 (1998). 61. G. Fiorelli, F. Gori, U. Frediani, F. Franceschelli, A. Tanini, C. TostiGuerra, S. Benvenuti, L. Gennari, L. Becherini, and M. L. Brandi, Membrane binding sites and non-genomic effects of estrogen in cultured human pre-osteoclastic cells. J. Steroid Biochem. Mol. Biol. 59, 233 – 240 (1996). 62. P. S. Cooke, D. L. Buchanan, D. B. Lubahn, and G. R. Cunha, Mechanism of estrogen action: Lessons from the estrogen receptoralpha knockout mouse. Biol. Reprod. 59, 470 – 475 (1998). 63. P. Monje and R. Boland, Characterization of membrane estrogen binding proteins from rabbit uterus. Mol. Cell. Endocrinol. 147, 75 – 84 (1999). 64. M. Razandi, A. Pedram, G. Greeene, and E. Levin, Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: Studies of ER and ER expressed in Chinese hamster ovary cells. Mol. Endocrinol. 13, 307 – 319 (1999). 65. W. Rosner, D. Hryb, M. Khan, A. Nakhla, and N. Romas, Sex hormone-binding globulin binding to cell membranes and generation of a second messenger. J. Androl. 13, 101 – 106 (1992). 66. E. Colin, G. Van Den Bemd, M. Van Aken, S. Christakos, H. De Jonge, H. DeLeuca, J. Prahl, J. Birkenhager, C. Buurman, H. Pols, and J. Van Leeuwen, Evidence for involvement of 17-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the rat. J. Bone. Miner. Res. 14, 57 – 64 (1999). 67. M. Valverde, P. Rojas, J. Amigo, D. Cosmelli, P. Orio, M. Bahamonde, G. Mann, C. Vergara, and R. Latorre, Acute activation of Maxi-K channels (hSlo) by estradiol binding to the  subunit. Science 285, 1929 – 1931 (1999). 68. G. Tremblay, A. Tremblay, N. Copeland, D. Gilbert, N. Jenkins, F. Labrie, and V. Giguere, Cloning, chromosomal localization, and functional anaysis of the murine estrogen receptor . Mol. Endocrinol 11, 353 – 365 (1997). 69. E. Enmark, P. Pelto-Huikko, K. Grandien, S. Lagercrantz, J. Lagercrantz, G. Fried, M. Nordenskjold, and J.-A. Gustafsson, Human estrogen receptor -Gene structure, chromosomal localization, and expression pattern. J. Clin. Endocrinol. Metab. 82, 4258 – 4265 (1997). 70. R. Bhat, D. Harnish, P. Stevis, C. Lyttle, and B. Komm, A novel human estrogen receptor : Identification and functional analysis of aditional N-terminal amino acids. J. Steroid Biochem. Mol. Biol. 67, 233 – 240 (1998). 71. D. Chen, P. Pace, R. Coombes, and S. Ali, Phosphorylation of human estrogen receptor  by protein kinase A regulates dimerization. Mol. Cell. Bio. 19, 1002 – 1015 (1999). 72. P. Shughrue, M. Lane, P. Scrimo, and I. Mechenthaler, Comparative distribution of estrogen receptor- (ER-) and  (ER-) mRNA in the rat pituitary, gonad, and reproductive tract. Steroids 63, 498 – 504 (1998). 73. S. Cowley, S. Hoare, S. Mosselman, and M. Parker, Estrogen receptors  and  form heterodimers on DNA. J. Biol. Chem. 272, 19858 – 19862 (1997). 74. G. Kuiper, S. Nilsson, and J.-A. Gustafsson, Characteristics and function of the novel estrogen receptor . Horm. Signal. 1, 89 – 112 (1998).

KOMM AND BODINE 75. J. Sun, M. Meyers, B. Fink, R. Rajendran, J. Katzenbellenbogen, and B. Katzenbellenbogen, Novel ligands that function as selective estrogens or antiestrogens for estrogen receptor- or estrogen receptor-. Endocrinology 140, 800 – 804 (1999). 76. A. Brzozowski, A. Pike, Z. Dauter, R. Hubbard, T. Bonn, O. Engstrom, L. Ohman, G. Greene, J.-A. Gustafsson, and M. Carlquist, Molecular basis of agonism and antagonism in the oestrogen receptor. Nature. 389, 753 – 758 (1997). 77. A. Pike, A. Brzozowski, R. Hubbard, T. Bonn, A.-G. Thorsell, O. Engstrom, J. Ljunggren, J.-A. Gustafsson, and M. Carlquist, Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist. EMBO J. 18, 4608 – 4618 (1999). 78. W. Gradishar and V. Jordan, Clinical potential of new antiestrogens. J. Clin. Oncol. 15, 840 – 852 (1997). 79. R. Love, H. Barden, R. Mazess, S. Epstein, and R. Chappell, Effect of tamoxifen on lumbar spine bone mineral density in postmenopausal women after 5 years. Arch. Intern. Med. 154, 2585 – 2588 (1994). 80. C. A. Frolik, H. U. Bryant, E. C. Black, D. E. Magee, and S. Chandrasekhar, Time-dependent changes in biochemical bone markers and serum cholesterol in ovariectomized rats: Effects of raloxifene HC1, tamoxifen, estrogen, and alendronate. Bone 18, 621 – 627 (1996). 81. P. Carthew, R. Edwards, and B. Nolan, Uterotrophic effects of tamoxifen, toremifene, and raloxifene do not predict endometrial cell proliferation in the ovariectomized CD1 mouse. Toxicol. Appl. Pharmacol. 158, 24 – 32 (1999). 82. T. Nickelsen, E. G. Lufkin, B. L. Riggs, D. A. Cox, and T. H. Crook, Raloxifene hydrochloride, a selective estrogen receptor modulator: Safety assessment of effects on cognitive function and mood in postmenopausal women. Psychoneuroendocrinology 24, 115 – 128 (1999). 83. J. Norris, L. Paige, D. Christensen, C.-Y. Chang, M. Huacani, D. Fan, P. Hamilton, D. Fowlkes, and D. Mcdonnell, Peptide antagonists of the human estrogen receptor. Science 285, 744 – 746 (1999). 84. D. Lubahn, J. Moyer, T. Golding, J. Couse, K. Korach, and O. Smithies, Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. USA 90, 11162 – 11166 (1993). 85. J. Krege, J. Hodgin, J. Couse, E. Enmark, M. Warner, J. Mahler, M. Sar, K. Korach, J.-A. Gustatsson, and O. Smithies, Generation and reproductive phenotypes of mice lacking estrogen receptor Proc. Natl. Acad. Sci. USA 95, 15677 – 15682 (1998). 86. J. Couse and K. Korach, Estrogen receptor null mice: What have we learned and where will they lead us? Endocrinol. Rev. 20, 358 – 417 (1999). 87. E. Smith, J. Boyd, G. Frank, H. Takahashi, R. Cohen, B. Specker, T. Williams, D. Lubahn, and K. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. Engl. J. Med. 331, 1056 – 1061 (1994). 88. O. Vidal, M. K. Lindberg, K. Hollberg, D. J. Baylink, G. Andersson, D. B. Lubahn, S. Mohan, J.-A. Gustafsson, and C. Ohlsson, Estrogen receptor specificity in the regulation of skeletal growth and maturation in male mice. Proc. Natl. Acad. Sci. USA 97, 5474 – 5479 (2000). 89. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994). 90. B. L. Riggs and L. J. I. Melton, The prevention and treatment of osteoporosis. Engl. Med. 327, 620 – 627 (1992). 91. D. W. Dempster and R. Lindsay, Pathogenesis of osteoporosis. Lancet 341, 797 – 805 (1993). 92. B. S. Komm and P. V. Bodine, The ongoing saga of osteoporosis treatment. J. Cell. Biochem. Suppl. 30/31, 277 – 283 (1998).

CHAPTER 10 Regulation of Bone Cell Function by Estrogens 93. B. L. Riggs, S. Khosla, and L. J. Melton, A unitary model for involutional osteoporosis: Estrogen deficiency causes both type I and tyep II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13, 763 – 773 (1998). 94. R. L. Prince, Counterpoint: Estrogen effects on calciotropic hormones and calcium homeostasis. Endoc. Rev. 15, 301 – 309 (1994). 95. Y. Liel, S. Shany, P. Smirnoff, and B. Schwartz, Estrogen increases 1,25-dihydroxyvitamin D receptors expression and bioresponse in the rat duodenal mucosa. Endocrinology 140, 280 – 285 (1999). 96. M. ten Bolscher, J. C. Netelenbos, R. Barto, L. M. Van Buuren, and W. J. F. Van der Vijgh, Estrogen regulation of intestinal calcium absorption in the intact and ovariectomized adult rat. J. Bone Miner. Res. 14, 1197 – 1202 (1999). 97. C. R. Draper, I. M. Dick, and R. L. Prince, The effect of estrogen deficiency on calcium balance in mature rats. Calcif. Tissue Int. 64, 325 – 328 (1999). 98. M. J. Oursler, M. Kassem, R. Turner, B. L. Riggs, and T. C. Spelsberg, Regulation of bone cell function by gonadal steroids. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey), pp. 237 – 260. Academic Press, San Diego, 1996. 99. T. C. Spelsberg, M. Subramaniam, B. L. Riggs, and S. Khosla, The actions and interactions of sex steroids and growth factors cytokines on the skeleton. Mol. Endocrinol. 13, 819 – 828 (1999). 100. D. J. Rickard, M. Subramaniam, and T. C. Spelsberg, Molecular and cellular mechanisms of estrogen action on the skeleton. J. Cell. Biochem. Suppl. 32/33, 123 – 132 (1999). 101. R. Bland, Steroid hormone receptor expression and action in bone. Clin. Sci. 98, 217 – 240 (2000). 102. B. L. Riggs and L. J. I. Melton, Involutional osteoporosis. Engl. J. Med. 314, 1676 – 1686 (1986). 103. T. K. Gray, T. C. Flynn, K. M. Gray, and L. M. Nabell, 17beta-estradiol acts directly on the clonal osteoblastic cell line UMR106. Proc. Natl. Acad. Sci. USA 84, 6267 – 6271 (1987). 104. J. B. Lian, G. S. Stein, E. Canalis, P. Gehron Robey, and A. L. Boskey, Bone formation: Osteoblast lineage cells, growth factors, matrix proteins, and the mineralization process. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus), pp. 14 – 29. Lippincott Williams & Wilkins, Philadelphia, 1999. 105. B. S. Komm, C. M. Terpening, D. J. Benz, K. A. Graeme, A. Gallegos, M. Korc, G. L. Greene, B. W. OMalley, and M. R. Haussler, Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241, 81 – 84 (1988). 106. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 107. F. S. Kaplan, M. D. Fallon, S. D. Boden, R. Schmidt, M. Senior, and J. G. Haddad, Estrogen receptors in bone in a patient with polyostotic fibrous dysplasia (McCune-Albright Syndrome). N. Engl. J. Med. 319, 421 – 425 (1988). 108. M. Ernst, C. Schmid, and E. R. Froesch, Enhanced osteoblast proliferation and collagen gene expression by estradiol. Proc. Natl. Acad. Sci. USA 85, 2307 – 2310 (1988). 109. M. C. Etienne, J. L. Fischel, G. Milano, P. Formento, J. L. Formento, M. Francoual, M. Frenay, and M. Namer, Steroid receptors in human osteoblast-like cells. Eur. J. Cancer 26, 807 – 810 (1990). 110. D. J. Benz, M. R. Haussler, and B. S. Komm, Estrogen binding and estrogenic responses in normal human osteoblast-like cells. J. Bone Miner. Res. 6, 531 – 541 (1991). 111. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7, 127 – 136 (1992).

331 112. A. Masuyama, Y. Ouchi, T. Sato, T. Hosoi, T. Nakamura, and H. Orimo, Characteristics of steroid receptors in cultured MC3T3-E1 osteoblastic cells and effect of steroid hormones on cell proliferation. Calcif. Tissue Int. 51, 376 – 381 (1992). 113. S. Migliaccio, V. L. Davis, M. K. Gibson, T. K. Gray, and K. S. Korach, Estrogens modulate the responsiveness of osteoblast-like cells (ROS 17/2.8) stably transfected with estrogen receptor. Endocrinology 130, 2617 – 2624 (1992). 114. V. L. Davis, J. F. Couse, T. K. Gray, and K. S. Korach, Correlation between low levels of estrogen receptors and estrogen responsiveness in two rat osteoblast-like cell lines. J. Bone Miner. Res. 9, 983 – 991 (1994). 115. P. V. Bodine, J. Green, H. A. Harris, R. A. Bhat, G. S. Stein, J. B. Lian, and B. S. Komm, functional properties of a conditionally phenotypic, estrogen-responsive, human osteoblast cell line. J. Cell. Biochem. 65, 368 – 387 (1997). 116. D. R. Ciocca and L. M. Vargas Roig, Estrogen receptors in human nontarget tissues: Biological and clinical implications. Endocr. Rev. 16, 35 – 62 (1995). 117. S. Nilsson, G. Kuiper, and J.-A. Gustafsson, ER-beta: A novel estrogen receptor offers the potential for new drug development. Trends Endrocrinol. Metab. 9, 387 – 395 (1998). 118. J. A. Hoyland, A. P. Mee, P. Baird, I. P. Braidman, E. B. Mawer, and A. J. Freemont, Demonstration of estrogen receptor mRNA in bone using in situ reverse-transcriptase polymerase chain reaction. Bone 20, 87 – 92 (1997). 119. V. Kusec, A. S. Virdi, R. Prince, and J. T. Triffitt, Localization of estrogen receptor-alpha in human and rabbit skeletal tissues. J. Clin. Endocrinol. Metab. 83, 2421 – 2428 (1998). 120. D. N. Petersen, G. T. Tkalcevic, P. H. Koza-Taylor, T. G. Turi, and T. A. Brown, Identification of estrogen receptor beta2, a functional variant of estrogen receptor beta expressed in normal rat tissues. Endocrinology 139, 1082 – 1092 (1998). 121. S. K. Lim, Y. J. Won, H. C. Lee, K. B. Huh, and Y. S. Park, A PCR analysis of ER alpha and ER beta mRNA abundance in rats and the effect of ovariectomy. J. Bone Miner. Res. 14, 1189 – 1196 (1999). 122. O. Vidal, L. G. Kindblom, and C. Ohlsson, Expression and localization of estrogen receptor-beta in murine and human bone. J. Bone Miner. Res. 14, 923 – 929 (1999). 123. S. H. Windahl, M. Norgard, G. Kuiper, J. A. Gustafsson, and G. Andersson, Cellular distribution of estrogen receptor beta in neonatal rat bone. Bone 26, 117 – 121 (2000). 124. I. Braidman, C. Baris, L. Wood, P. Selby, J. Adams, A. Freemont, and J. Hoyland, Preliminary evidence for impaired estrogen receptor-alpha protein expression in osteoblasts and osteocytes from men with idiopathic osteoporosis. Bone 26, 423 – 427 (2000). 125. Y. Onoe, C. Miyaura, H. Ohta, S. Nozawa, and T. Suda, Expression of estrogen receptor beta in rat bone. Endocrinology 138, 4509 – 4512 (1997). 126. J. Arts, G. G. Kuiper, J. M. Janssen, J. A. Gustafsson, C. W. Lowik, H. A. Pols, and J. P. van Leeuwen, Differential expression of estrogen receptors alpha and beta mRNA during differentiation of human osteoblast SV-HFO cells. Endocrinology 138, 5067 – 5070 (1997). 127. M. A. Ankrom, J. A. Patterson, P. Y. dAvis, U. K. Vetter, M. R. Blackman, P. D. Sponseller, M. Tayback, P. G. Robey, J. R. Shapiro, and N. S. Fedarko, Age-related changes in human oestrogen receptor alpha function and levels in osteoblasts. Biochem. J. 333, 787 – 794 (1998). 128. P. V. Bodine, R. A. Henderson, J. Green, M. Aronow, T. Owen, G. S. Stein, J. B. Lian, and B. S. Komm, Estrogen receptor-alpha is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139, 2048 – 2057 (1998). 129. P. V. N. Bodine, H. A. Harris, and B. S. Komm, Suppression of ligand-dependent estrogen receptor activity by bone-resorbing

332

130.

131.

132. 133. 134. 135.

136.

137.

138.

139. 140.

141.

142.

143.

144.

145.

146.

147.

148.

KOMM AND BODINE cytokines in human osteoblasts. Endocrinology 140, 2439 – 2451 (1999). J. M. Hall and D. P. McDonnell, The estrogen receptor beta-isoform (ERbeta) of the human estrogen receptor modulates ERalpha transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology 140, 5566 – 5578 (1999). R. R. Recker, Emryology, anatomy, and microstructure of bone. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus eds.), pp. 219 – 240. Raven Press, New York, 1992. E. M. Aarden, E. H. Burger, and P. J. Nijweide, Functions of osteocytes in bone. J. Cell. Biochem. 55, 287 – 299 (1994). B. S. Noble and J. Reeve, Osteocyte function, osteocyte death and bone fracture resistance. Mol. Cell. Endocrinol. 159, 7 – 13 (2000). L. E. Lanyon, Osteocytes, strain detection, bone modeling and remodeling. Calcif. Tissue Int. 53 (Suppl. 1), S102 – S107 (1993). I. P. Braidman, L. K. Davenport, D. H. Carter, P. L. Selby, E. B. Mawer, and A. J. Freemont, Preliminary in situ identification of estrogen target cells in bone. J. Bone Miner. Res. 10, 74 – 80 (1995). J. A. Hoyland, C. Baris, L. Wood, P. Baird, P. L. Selby, A. J. Freemont, and I. P. Braidman, Effect of ovarian steroid deficiency on oestrogen receptor alpha expression in bone. J. Pathol. 188, 294 – 303 (1999). T. Ohashi, S. Kusuhara, and K. Ishida, Immunoelectron microscopic demonstration of estrogen receptors in osteogenic cells of Japanese quail. Histochemistry 96, 41 – 44 (1991). S. C. Manolagas and R. L. Jilka, Bone marrow, cytokines, and bone remodeling: Emerging insights into the pathophysiology of osteoporosis. N. Engl. J. Med. 332, 305 – 311 (1995). J. E. Aubin, Bone stem cells. J. Cell. Biochem. Suppl. 30/31, 73 – 82 (1998). R. O. C. Oreffo, V. Kusec, S. Romberg, and J. T. Triffitt, Human bone marrow osteoprogenitors express estrogen receptor-alpha and bone morphogenetic proteins 2 and 4 mRNA during osteoblastic differentiation. J. Cell. Biochem. 75, 382 – 392 (1999). S. C. Dieudonne, T. Xu, J. Y. Chou, S. A. Kuznetsov, K. Satomura, M. Mankani, N. S. Fedarko, E. P. Smith, P. G. Robey, and M. F. Young, Immortalization and characterization of bone marrow stromal fibroblasts from a patient with a loss of function mutation in the estrogen receptor-alpha gene. J. Bone Miner. Res. 13, 598 – 608 (1998). M. J. Oursler, Estrogen regulation of gene expression in osteoblasts and osteoclasts. Crit. Rev. Eukary. Gene Express. 8, 125 – 140 (1998). T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteclast differentiation and function by new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). J. M. Pensler, C. B. Langman, J. A. Radosevich, M. L. Maminta, M. Mangkornkanok, R. Higbee, and A. Molteni, Sex steroid hormone receptors in normal and dysplastic bone disorders in children. Bone Miner. Res. 5, 493 – 498 (1990). M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Nat. Acad. Sci. USA 88, 6613 – 6617 (1991). M. J. Oursler, L. Pederson, L. Fitzpatrick, B. L. Riggs, and T. Spelsberg, Human giant cell tumors of the bone (osteoclastomas) are estrogen target cells. Proc. Nal. Acad. Sci. USA 91, 5227 – 5231 (1994). T. Sunyer, J. Lewis, P. Collin-Osdoby, and P. Osdoby, Estrogen’s bone-protective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). H. Mano, T. Yuasa, T. Kameda, K. Miyazawa, Y. Nakamaru, M. Shiokawa, Y. Mori, T. Yamada, K. Miyata, H. Shindo, H. Azuma, Y. Hakeda, and M. Kumegawa, Mammalian mature osteoclasts as

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

estrogen target cells. Biochem. Biophys. Res. Commun. 223, 637 – 642 (1996). F. M. Collier, W. H. Huang, W. R. Holloway, J. M. Hodge, M. T. Gillespie, L. L. Daniels, M. H. Zheng, and G. C. Nicholson, Osteoclasts from human giant cell tumors of bone lack estrogen receptors. Endocrinology 139, 1258 – 1267 (1998). L. Pederson, M. Kremer, N. T. Foged, B. Winding, C. Ritchie, L. A. Fitzpatrick, and M. J. Oursler, Evidence of a correlation of estrogen receptor level and avian osteoclast estrogen responsiveness. J. Bone Miner. Res. 12, 742 – 752 (1997). G. Fiorelli, F. Gori, M. Petilli, A. Tanini, S. Benvenuti, M. Serio, P. Bernabel and M. L. Brandi, Functional estrogen receptors in a human preosteoclastic cell line. Proc. Natl. Acad. Sci. USA 92, 2672 – 2676 (1995). M. Kanatani, T. Sugimoto, Y. Takahashi, H. Kaji, R. Kitazawa, and K. Chihara, Estrogen via the estrogen receptor blocks cAMPmediated parathyroid hormone (PTH)-stimulated osteoclast formation. J. Bone Miner. Res. 13, 854 – 862 (1998). H. Ben-Hur, G. Mor, I. Blickstein, I. Likhman, F. Kohen, R. Dgani, V. Insler, P. Yaffe, and A. Ornoy, Localization of estrogen receptors in long bones and vertebrae of human fetuses. Calci. Tissue Int. 53, 91 – 96 (1993). H. Ben-Hur, H. H. Thole, A. Mashiah, V. Insler, V. Berman, E. Shezen, D. Elias, A. Zuckerman, and A. Ornoy, Estrogen, progesterone and testosterone receptors in human fetal cartilaginous tissue: Immunohistochemical studies. Calci. Tissue Int. 60, 520 – 526 (1997). J. Kennedy, C. Baris, J. A. Hoyland, P. L. Selby, A. J. Freemont, and I. P. Braidman, Immunofluorescent localization of estrogen receptoralpha in growth plates of rabbits, but not in rats, at sexual maturity. Bone 24, 9 – 16 (1999). L. O. Nilsson, A. Boman, L. Savendahl, G. Grigelioniene, C. Ohlsson, E. M. Ritzen, and J. Wroblewski, Demonstration of estrogen receptor-beta immunoreactivity in human growth plate cartilage. J. Clin. Endocrinol. Metab. 84, 370 – 373 (1999). N. Dayani, M. T. Corvol, P. Robel, B. Eychenne, B. Moncharmont, L. Tsagris, and R. Rappaprot, Estrogen receptors in cultured rabbit articular chondrocytes: Influence of age. J. Steroid Biochem. 31, 351 – 356 (1988). B. A. Monaghan, F. S. Kaplan, C. R. Lyttle, M. D. Fallon, S. D. Boden, and J. G. Haddad, Estrogen receptors in fracture healing. Clin. Orthop. Rel. Res. 280, 277 – 280 (1992). H. Pinus, A. Ornoy, N. Patlas, P. Yaffe, and Z. Schwartz, Specifi beta-estradiol binding in cartilage and serum from young mice and rats is age dependent. Connect. Tissue Res. 30, 85 – 98 (1993). E. Nasatzky, Z. Schwartz, W. A. Soskolne, B. P. Brooks, D. D. Dean, B. D. Boyan, and A. Ornoy, Evidence for receptors specific for 17beta-estradiol and testosterone in chondrocyte cultures. Connect. Tissue Res. 30, 277 – 294 (1994). M. L. Brandi, C. Crescioli, A. Tanini, U. Frediani, D. Agnusdei, and C. Gennari, Bone endothelial cells as estrogen targets. Calci. Tissue Int. 53, 312 – 317 (1993). G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation/differentiation interrelationships during progressive development of the osteoblastic phenotype. Endocr. Rev. 14, 424 – 442 (1993). J. Clover and M. Gowen, Are MG-63 and HOS TE85 human osteosarcoma cell lines representative models of the osteoblastic phenotype? Bone 15, 585 – 591 (1994). D. Modrowski, L. Miravet, M. Feuga, and P. J. Marie, Increased proliferation of osteoblast precursor cells in estrogen-deficient rats. Am. J. Physiol. 264, E190 – E196 (1993). K. C. Westerlind, G. K. Wakley, G. L. Evans, and R. T. Turner, Estrogen does not increase bone formation in growing rats. Endocrinology 133, 2924 – 2934 (1993).

333

CHAPTER 10 Regulation of Bone Cell Function by Estrogens 166. C. K. Watts and R. J. King, Overexpression of estrogen receptor in HTB 96 human osteosarcoma cells results in estrogen-induced growth inhibition and receptor cross talk. J. Bone Miner. Res. 9, 1251 – 1258 (1994). 167. E. Damien, J. S. Price, and L. E. Lanyon, The estrogen receptor’s involvement in osteoblasts adaptive response to mechanical strain. J. Bone Miner Res. 13, 1275 – 1282 (1998). 168. J. A. Robinson, S. A. Harris, B. L. Riggs, and T. C. Spelsberg, Estrogen regulation of human osteoblastic cell proliferation and differentiation. Endocrinology 138, 2919 – 2927 (1997). 169. A. Samuels, M. J. Perry, and J. H. Tobias, High-dose estrogeninduced osteogenesis in the mouse is partially suppressed by indomethacin. Bone 25, 675 – 680 (1999). 170. Q. Qu, P. L. Harkonen, and H. K. Vaananen, Comparative effects of estrogen and antiestrogens on differentiation of osteoblasts in mouse bone marrow culture. J. Cell. Biochem. 73, 500 – 507 (1999). 171. D. Somjen, Y. Weisman, A. Harell, E. Berger, and A. M. Kaye, Direct and sex-specific stimulation by sex steroids of creatine kinase activity and DNA synthesis in rat bone. Proc. Natl. Acad. Sci. USA 86, 3361 – 3365 (1989). 172. S. D. Bain, M. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High-dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 173. C. H. Turner, Editorial: Do estrogens increase bone formation? Bone 12, 305 – 306 (1991). 174. R. T. Turner, Mice, estrogen, and postmenopausal osteoporosis. Bone Miner. Res. 14, 187 – 191 (1999). 175. A. Gohel, M. B. McCarthy, and G. Gronowicz, Estrogen prevents glucocorticoid-induced apoptosis in osteoblasts in vivo and in vitro. Endocrinology 140, 5339 – 5347 (1999). 176. H. J. Verhaar, C. A. Damen, S. A. Duursma, and B. A. Scheven, A comparison of the action of progestins and estrogen on the growth and differentiation of normal adult human osteoblast-like cells in vitro. Bone 15, 307 – 311 (1994). 177. R. J. Majeska, J. T. Ryaby, and T. A. Einhorn, Direct modulation of osteoblastic activity with estrogen. J. Bone Joint Surg. 76, 713 – 721 (1994). 178. J. A. Robinson, K. M. Waters, R. T. Turner, and T. C. Spelsberg, Direct action of naturally occurring estrogen metabolites on human osteoblastic cells. J. Bone Miner. Res. 15, 499 – 506 (2000). 179. T. Ikeda, C. Shigeno, R. Kasai, H. Kohno, S. Ohta, H. Okumura, J. Konishi, and T. Yamamuro, Ovariectomy decreases the mRNA levels of transforming growth factor-beta1 and increases the mRNA levels of osteocalcin in rat bone in vivo. Biochem. Biophy. Commun. 194, 1228 – 1233 (1993). 180. R. T. Turner, D. S. Colvard, and T. C. Spelsberg, Estrogen inhibition of periosteal bone formation in rat long bones: Down-regulation of gene expression for bone matrix proteins. Endocrinology 127, 1346 – 1351 (1990). 181. M. Ernst, J. K. Heath, and G. A. Rodan, Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bones. Endocrinology 125, 825 – 833 (1989). 182. M. A. Salih, C. C. Liu, B. H. Arjmandi, and D. N. Kalu, Estrogen modulates the mRNA levels for cancellous bone protein of ovariectomized rats. Bone Miner. 23, 285 – 299 (1993). 183. K. C. Westerlind, T. J. Wronski, G. L. Evans, and R. T. Turner, The effect of long-term ovarian hormone deficiency on transforming growth factor-beta and bone matrix protein mRNA expression in rat femora. Biochem. Biophys. Res. Commun 200, 283 – 289 (1994). 184. R. T. Turner, P. Backup, P. J. Sherman, E. Hill, G. L. Evans, and T. C. Spelsberg, Mechanism of action of estrogen on intramembra-

185.

186.

187.

188.

189.

190.

191.

192.

193.

194. 195. 196. 197.

198.

199.

200.

201. 202.

203.

nous bone formation: Regulation of osteoblast differentiation and activity. Endocrinology 131, 883 – 889 (1992). M. Takeuchi, M. Tokin, and K. Nagata, Tamoxifen directly stimulates the mineralization of human osteoblast-like osteosarcoma cells through a pathway independent of estrogen response element. Biochem. Biophys. Res. Commun. 210, 295 – 301 (1995). M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-B gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). T. A. Linkhart, S. Mohan, and D. J. Baylink, Growth factors for bone growth and repair: IGF, TGF-beta and BMP. Bone 19, 1S – 12S (1996). R. N. Margolis, E. Canalis, and N. C. Partridge, Invited review of a workshop: anabolic hormones in bone: basic research and therapeutic potential. J. Clin. Endocrinol. Metab. 81, 872 – 877 (1996). J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogeneitic protein gene family in bone formation and repair. Clin. Orthop. Rel. Res. 346, 26 – 37 (1998). L. F. Bonewald, Regulation and regulatory activities of transforming growth factor beta. Crit. Rev. Eukary. Gene Express. 9, 33 – 44 (1999). C. J. Rosen, Serum insulin-like growth factors and insulin-like growth factor-binding proteins: Clinical implications. Clin. Chem. 45, 1384 – 1390 (1999). J. L. Jones and D. R. Clemmons, Insulin-like growth factors and their binding proteins: Biological actions. Endocr. Rev. 16, 3 – 34 (1995). P. P.-C. Hu, M. B. Datto, and X.-F. Wang, Molecular mechanisms of transforming growth factor-beta signaling. Endocr. Rev. 19, 349 – 363 (1998). M. C. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). R. Pacifici, Estrogen, cytokines, and pathogenesis of postmenopausal osteoporosis. J. Bone Miner. Res. 11, 1043 – 1051 (1996). R. L. Jilka, Cytokines, bone remodeling, and estrogen deficiency: A 1998 update. Bone 23, 75 – 81 (1998). T. K. Gray, B. Lipes, T. Linkhart, S. Mohan, and D. J. Baylink, Transforming growth factor beta mediates the estrogen induced inhibition of UMR106 cell growth. Connect. Tissue Res. 20, 23 – 32 (1989). M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). R. D. Finkelman, N. H. Bell, D. D. Strong, L. M. Demers, and D. J. Baylink, Ovariectomy selectively reduces the concentration of transfoming growth factor b in rat bone: Implications for estrogen deficiency-associated bone loss. Proc. Natl. Acad. Sci. USA 89, 12190 – 12193 (1992). N. N. Yang, H. U. Bryant, S. Hardikar, M. Sato, R. J. Galvin, A. L. Glasebrook, and J. D. Termine, Estrogen and raloxifene stimulate transforming growth factor-beta 3 gene expression in rat bone: A potential mechanism for estrogen- or raloxifene-mediated bone maintenance. Endocrinology 137, 2075 – 2084 (1996). D. P. McDonnell, The molecular pharmacology of SERMs. Trends Endocrinol. Metab. 10, 301 – 311 (1999). N. N. Yang, M. Venugopalan, S. Hardikar, and A. Glasebrook, Identification of an estrogen response element activated by metabolites of 17beta-estradiol and raloxifene. Science 273, 1222 – 1225 (1996). K. R. Tau, T. E. Hefferan, K. M. Waters, J. A. Robinson, M. Subramaniam, B. L. Riggs, and T. C. Spelsberg, Estrogen regulation of a transforming growth factor-beta inducible early gene that inhibits deoxyribonucleic acid synthesis in human osteoblasts. Endocrinology 139, 1346 – 1353 (1998).

334 204. M. Subramaniam, S. A. Harris, M. J. Oursler, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Identification of a novel TGF-betaregulated gene encoding a putative zinc finger protein in human osteoblasts. Nucleic Acids Res. 23, 4907 – 4912 (1995). 205. D. J. Rickard, L. C. Hofbauer, S. K. Bonde, F. Gori, T. C. Spelsberg, and B. L. Riggs, Bone morphogenetic protein-6 production in human osteoblastic cell lines. Selective regulation by estrogen. J. Clin. Invest. 101, 413 – 422 (1998). 206. A. van den Wijngaard, W. R. Mulder, R. Dijkema, C. J. C. Boersma, S. Mosselman, E. J. J. van Zoelen, and W. Olijve, Antiestrogens specifically up-regulate bone morphogenic protein-4 promoter activity in human osteoblastic cells. Mol. Endocrinol. 14, 623 – 633 (2000). 207. T. K. Gray, S. Mohan, T. A. Linkhart, and D. J. Baylink, Estradiol stimulates in vitro the secretion of insulin-like growth factors by the clonal osteoblastic cell line, UMR106. Biochem. Biophys. Res. Commun. 158, 407 – 412 (1989). 208. M. C. Slootweg, D. Swolin, J. C. Netelenbos, O. G. Isaksson, and C. Ohlsson, Estrogen enhances growth hormone receptor expression and growth hormone action in rat osteosarcoma cells and human osteoblast-like cells. J. Endocrinol. 155, 159 – 164 (1997). 209. R. T. Turner, L. S. Kidder, M. Zhang, S. A. Harris, K. C. Westerlind, A. Maran, and T. J. Wronski, Estrogen has rapid tissue-specific effects on rat bone. J. Appl. Physiol. 86, 1950 – 1958 (1999). 210. J. Erdmann, S. Storch, J. Pfeilschifter, P. Ochlich, R. Ziegler, and F. Bauss, Effects of estrogen on the concentration of insulin-like growth factor-1 in rat bone. Bone 22, 503 – 507 (1998). 211. C. Schmid, M. Ernst, J. Zapf, and E. R. Froesch, Release of insulinlike growth factor carrier proteins by osteoblasts: Stimulation by estradiol and growth hormone. Biochem. Biophys. Res. Commun. 160, 788 – 794 (1989). 212. M. Kassem, R. Okazaki, D. De Leon, S. A. Harris, J. A. Robinson, T. C. Spelsberg, C. A. Conover, and B. L. Riggs, Potential mechanism of estrogen-mediated decrease in bone formation: Estrogen increases production of inhibitory insulin-like growth factor-binding protein-4. Proc. Assoc. Am. Phys. 108, 155 – 164 (1996). 213. Y. Kudo, M. Iwashita, T. Iguchi, Y. Takeda, N. Hizuka, K. Takano, and T. Muraki, Estrogen and parathyroid hormone regulate insulinlike growth factor binding protein-4 in SaOS-2 cells. Life Sci. 61, 165 – 170 (1997). 214. V. Hwa, Y. Oh, and R. G. Rosenfeld, The insulin-like growth factor binding protein (IGFBP) superfamily. Endocr. Rev. 20, 761 – 787 (1999). 215. C. J. Rosen, D. Verault, C. Steffens, D. Cheleuitte, and J. Glowacki, Effects of age and estrogen status on the skeletal IGF regulatory system. Endocrin. 7, 77 – 80 (1997). 216. T. L. McCarthy, C. Ji, H. Shu, S. Casinghino, K. Crothers, P. Rotwein, and M. Centrella, 17beta-estradiol potently suppresses cAMP-induced insulin-like growth factor-1 gene activation in primary rat osteoblast cultures. J. Biol. Chem. 272, 18132 – 18139 (1997). 217. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17b-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblastic cells in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 218. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257, 88 – 91 (1992). 219. D. Cheleuitte, S. Mizuno, and J. Glowacki, In vitro secretion of cytokines by human bone marrow: Effects of age and estrogen status. J. Clin. Endocrinol. Metab. 83, 2043 – 2051 (1998). 220. S. T. Pottratz, T. Bellido, H. Mocharla, D. Crabb, and S. C. Manolagas, 17beta-estradiol inhibits expression of human interleukin-6 pro-

KOMM AND BODINE

221.

222.

223.

224.

225.

226. 227.

228.

229.

230.

231. 232.

233.

234.

235.

236.

237.

moter-reporter constructs by a receptor-dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). L. I. McKay and J. A. Cidlowski, Molecular control of immune/ inflammatory responses: Interactions between nuclear factor-kappaB and steroid receptor-signaling pathways. Endocr. Rev. 20, 435 – 459 (1999). L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 beta and tumor necrosis factor-alpha: Lack of regulation by 17 beta-estradiol. Endocrinology 130, 2528 – 2534 (1992). D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumor necrosis factor but not interleukin-6 from adult human osteoblasts in vitro. Osteoporosis Int. 2, 94 – 102 (1992). L. Rifas, J. S. Kenney, M. Marcelli, R. Pacifici, S.-L. Cheng, L. L. Dawson, and L. V. Avioli, Production of interleukin-6 in human osteoblasts and human bone marrow stromal cells: Evidence that induction by interleukin-1b and tumor necrosis factor-a is not regulated by ovarian steroids. Endocrinology 136, 4056 – 4067 (1995). M. Kassem, S. A. Harris, T. C. Spelsberg, and B. L. Riggs, Estrogen inhibits interleukin-6 production and gene expression in a human osteoblstic cell line with high levels of estrogen receptors. J. Bone Miner. Res. 11, 193 – 199 (1996). T. Taga and T. Kishimoto, GP130 and the interkeukin-6 family of cytokines. Annu. Rev. Immunol. 15, 797 – 819 (1997). S. C. Lin, T. Yamate, Y. Taguchi, V. Z. Borba, G. Girasole, C. A. OBrien, T. Bellido, E. Abe, and S. C. Manolagas, Regulation of the gp80 and gp130 subunits of the IL-6 receptor by sex steroids in the murine bone marrow. J. Clin. Invest. 100, 1980 – 1990 (1997). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370 (1999). S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). T. Chevalley and R. Rizzoli, Influence de I’hormone parathyroidienne sur Ios. Press Med. 28, 547 – 553 (1999). S. Fukayama and A. H. Tashjian, Direct modulation by estradiol of the response of human bone cells (SaOS-2) to human parathyroid hormone (PTH) and PTH-related protein. Endocrinology 124, 397 – 401 (1989). J. J. Monroe and A. H. J. Tashjian, Pretreatment with 17betaestradiol attenuates basal- and PTH-stimulated membrane adenylyl cyclase activity in human osteoblast-like SAOS-2 cells. Bichem. Biophys. Res. Commun. 225, 320 – 325 (1996). C. C. Pilbeam, J. Klein-Nulend, and L. G. Raisz, Inhibition by 17bestradiol of PTH-stimulated resorption and prostaglandin production in cultured neonatal mouse calvariae. Biochem. Biophys. Res. Commun. 163, 1319 – 1324 (1989). H. Kaji, T. Sugimoto, M. Kanatani, M. Nasu, and K. Chihara, Estrogen blocks parathyroid hormone (PTH)-stimulated osteoclast-like cell formation by selectively affecting PTH-responsive cyclic adenosine monophosphate pathway. Endocrinology 137, 2217 – 2224 (1996). F. Cosman, V. Shen, F. Xie, M. Seibel, A. Ratcliffe, and R. Lindsay, Estrogen protects against bone resorbing effects of parathyroid hormone infusion: Assesment by use of biochemical markers. Ann. Intern. Med. 118, 337 – 343 (1993). M. Ernst, M. G. Parker, and G. A. Rodan, Functional estrogen receptors in osteoblastic cells demonstrated by transfection with a reporter

CHAPTER 10 Regulation of Bone Cell Function by Estrogens

238.

239

240.

241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

251.

252.

253.

254.

gene containing an estrogen response element. Mol. Endocrinol. 5, 1597 – 1606 (1991). L. G. Rao, J. N. Wylie, M. S. K. Sutherland, and T. M. Murray, 17bestradiol and parathyroid hormone potentiate each others stimulatory effects on alkaline phosphatase activity in SaOS-2 cells in a differeniation- dependent manner. Endocrinology 134, 614 – 620 (1994). C. Eielson, D. Kaplan, M. A. Mitnick, I. Paliwal, and K. Insogna, Estrogen modulates parathyroid hormone-induced fibronectin production in human and rat osteoblast-like cells. Endocrinology 135, 1639 – 1644 (1994). V. Behar, C. Nakamoto, Z. Greenberg, A. Bisello, L. J. Suva, M. Rosenblatt, and M. Chorev, Histidine at position 5 is the specificity “switch” between two parathyroid hormone receptor subtypes. Endocrinology 137, 4217 – 4224 (1996). L. A. Pivirotto, D. S. Cissel, and P. E. Keeting, Sex hormones mediate interleukin-1b production by human osteoblastic HOBIT cells. Mol. Cell Endocrinol. 111, 67 – 74 (1995). S. A. Harris, K. R. Tau, R. J. Enger, D. O. Toft, B. L. Riggs, and T. C. Spelsberg, Estrogen response in the hFOB 1.19 human fetal osteoblastic cell line stably transfected with the human estrogen receptor gene. J. Cell. Biochem. 59, 193 – 201 (1995). Y. Liel, S. Kraus, J. Levy, and S. Shany, Evidence that estrogens modulate activity and increase the number of 1,25-dihydroxyvitamin D receptors in osteoblast-like cells (ROS 17/2.8). Endocrinology 130, 2597 – 2601 (1992). M. Ishibe, T. Nojima, T. Ishibashi, T. Koda, K. Kaneda, R. N. Rosier, and J. E. Puzas, 17b-Estradiol increases the receptor number and modulates the action of 1,25-dihydroxyvitamin D3 in human osteosarcoma-derived osteoblast-like cells. Calcif. Tissue Int. 57, 430 – 435 (1995). K. L. Kirkwood, R. Dziak, and P. G. Bradford, Inositol trisphosphate receptor gene expression and hormonal regulation in osteoblast-like cell lines and primary osteoblast cell cultures. J. Bone Miner. Res. 11, 1889 – 1896 (1996). J. W. J. Putney and G. S. J. Bird, The inositol phosphate-calcium signaling system in nonexcitable cells. Endocr. Rev. Rev. 14, 610 – 631 (1993). K. L. Kirkwood, K. Homick, M. B. Dragon, and P. G. Bradford, Cloning and characterization of the type I inositol 1,4,5-trisphosphate receptor gene promoter: Regulation by 17beta-estradiol in osteoblasts. J. Biol. Chem. 272, 22425 – 22431 (1997). K. E. Armour and S. H. Ralston, Estrogen upregulates endothelial constitutive nitric oxide synthase expression in human osteoblastlike cells. Endocrinology 139, 799 – 802 (1998). S. H. Ralston, L.-P. Ho, M. H. Helfrich, P. S. Grabowski, P. W. Johnston, and N. Benjamin, Nitric oxide: a cytokine-induced regulator of bone resorption. J. Bone Miner. Res. 10, 1040 – 1049 (1995). S. J. Wimalawansa, G. De Marco, P. Gangula, and C. Yallampalli, Nitric oxide donor alleviates ovariectomy-induced bone loss. Bone 18, 301 – 304 (1996). R. L. Van Bezooijen, C. Van der Bent, S. E. Papapoulos, and C. W. G. M. Lowik, Oestrogenic compounds modulate cytokine-induced nitric oxide production in mouse osteoblast-like cells. J. Pharm. Pharmacol. 51, 1409 – 1414 (1999). D. S. Cissel, M. Murty, D. L. Whipkey, J. D. Blaha, G. M. Graeber, and P. E. Keeting, Estrogen pretreatment increases arachidonic acid release by bradykinin stimulated normal human osteoblast-like cells. J. Cell. Biochem. 60, 260 – 270 (1996). L. F. Cooper and K. Uoshima, Differential estrogenic regulation of small Mr heat shock protein expression in osteoblasts. J. Biol. Chem. 269, 7869 – 7873 (1994). D. D. Bankson, N. Rifai, M. E. Williams, L. M. Silverman, and T. K. Gray, Biochemical effects of 17 beta-estradiol on UMR106 cells. Bone Miner. 6, 55 – 63 (1989).

335 255. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr. Rev. 7, 379 – 408 (1984). 256. S. W. Whitson, Tight junction formation in the osteon. Clin. Orthop. Relat. Res. 86, 206 – 213 (1972). 257. A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). 258. R. L. Duncan and C. H. Turner, Mechanotransduction and the functional response of bone to mechanical strain. Calcif. Tissue Int. 57, 344 – 358 (1995). 259. A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 13, 1243 – 1250 (1998). 260. T. Keda, A. Yamaguchi, S. Yokose, Y. Nagal, H. Yamato, T. Nakamura, H. Tsurukami, T. Tanizawa, and S. Yoshiki, Changes in biological activity of bone cells in ovariectomized rats revealed by in situ hybrdization. J. Bone Miner. Res. 11, 780 – 788 (1996). 261. P. V. Bodine, S. K. Vernon, and B. S. Komm, Establishment and hormonal regulation of a conditionally transformed preosteocytic cell line from adult human bone. Endocrinology 137, 4592 – 4604 (1996). 262. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, R. F. L. Suswillo, and L. E. Lanyon, Mechanical loading and sex hormone interactions in organ cultures of rat ulna. J. Bone Miner. Res. 11, 502 – 511 (1996). 263. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, A. A. Pitsillides, R. F. L. Suswillo, and L. E. Lanyon, Enhancement by sex hormones of the osteoregulatory effects of mechanical loading and prostaglandins in explants of rat ulna. J. Bone Miner. Res. 12, 1424 – 1430 (1997). 264. K. C. Westerlind, T. J. Wronski, E. L. Ritman, Z.-P. Luo, K.-N. An, N. H. Bell, and R. T. Turner, Estrogen regulates the rate of bone turnover but bone balance in ovariectomized rats is modulated by prevailing mechanical strain. Proc. Natl. Acad. Sci. USA 94, 4199 – 4204 (1997). 265. W. M. Kohrt, D. B. Snead, E. Slatopolsky, and S. J. J. Birge, Additive effects of weight-bearing exercise and estrogen on bone mineral density in older women. J. Bone Miner. Res. 10, 1303 – 1311 (1995). 266. M. J. Oursler, L. Pederson, J. Pyfferoen, P. Osdoby, L. Fitzpatrick, and T. C. Spelsberg, Estrogen modulation of avian osteoclast lysosomal gene expression. Endocrinology 132, 1373 – 1380 (1993). 267. M. Kremer, J. Judd, B. Ritkin, J. Auszmann, and M. J. Oursler, Estrogen modulation of osteoclast lysosomal enzyme secretion. J. Cell. Biochem. 57, 271 – 279 (1995). 268. J. A. Robinson, B. L. Riggs, T. C. Spelsberg, and M. J. Oursler, Osteoclasts and transforming growth factor-beta: Estrogen-mediated isoform-specific regulation of production. Endocrinology 137, 615 – 621 (1996). 269. M. H. Zheng, T.-T. A. Lau, R. Prince, A. Criddle, S. Wysocki, M. Beilharz, J. M. Papadimitriou, and D. J. Wood, 17beta-estradiol suppresses gene expression of tartrate-resistant acid phosphatase and carbonic anhydrase II in ovariectomized rats. Calcif. Tissue Int. 56, 166 – 169 (1995). 270. J. P. Williams, H. C. Blair, M. A. McKenna, S. E. Jordan, and J. M. McDonald, Regulation of avian osteoclastic H -ATPase and bone resorption by tamoxifen and calmodulin antagonists. Effects independent of steroid receptors. J. Biol. Chem. 271, 12488 – 12495 (1996). 271. B. F. Boyce, D. E. Hughes, K. R. Wright, L. Xing, and A. Dai, Recent advances in bone biology provide insight into the pathogenesis of bone diseases. Lab. Invest. 79, 83 – 94 (1999). 272. S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms for the pathogenesis and treatment of osteoporosis. Endocr. Rev. 21, 115 – 137 (2000). 273. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nature Med. 2, 1132 – 1136 (1996).

336 274. S. Zecchi-Orlandini, L. Formigli, A. Tani, S. Benvenuti, G. Fiorelli, L. Papucci, S. Capaccioli, G. E. Orlandini, and M. L. Brandi, 17 beta-estradiol induces apoptosis in the preosteoclastic FLG 29.1 cell line. Biochem. Biophys. Res. Commun. 255, 680 – 685 (1999). 275. G. Fiorelli, V. Martineti, F. Gori, S. Benvenuti, U. Frediani, L. Formigli, S. Zecchi, and M. L. Brandi, Heterogeneity of binding sites and bioeffects of raloxifene on the human leukemic cell line FLG 29.1. Biochem. Biophy. Res. Commun. 240, 573 – 579 (1997). 276. Z. Khalkhali-Ellis, P. Collin-Osdoby, L. Li, M. L. Brandi, and P. Osdoby, A human homolog of the 150 kD avian osteoclast membrane antigen related to superoxide dismutase and essential for bone resorption is induced by developmental agents and opposed by estrogen in FLG 29.1 cells. Calcif. Tissue Int. 60, 187 – 193 (1997). 277. C. Schiller, R. Gruber, K. Redlich, G. M. Ho, F. Katzgraber, M. Willheim, P. Pietschmann, and M. Peterlik, 17Beta-estradiol antagonizes effects of 1 alpha,25-dihydroxyvitamin D3 on interleukin-6 production and osteoclast-like cell formation in mouse bone marrow primary cultures. Endocrinology 138, 4567 – 4571 (1997). 278. M. M. Takahashi and T. Noumura, Sexually dimorphic and laterally asymmetric development of the embryonic duck syrinx: Effect of estrogen on in vitro cell proliferation and chondrogenesis. Dev. Biol. (Orlando) 121, 417 – 422 (1987). 279. E. Nasatzky, Z. Schwartz, B. D. Boyan, W. A. Soskolne, and A. Ornoy, Sex-dependent effects of 17beta-estradiol on chondrocyte differentiation in culture. J. Cell. Physiol. 154, 359 – 367 (1993). 280. M.-T. Corvol, A. Carrascosa, L. Tsagris, O. Blanchard, and R. Rappaport, Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: Influence of age and sex. Endocrinology 120, 1422 – 1429 (1987). 281. O. Blanchard, L. Tsagris, R. Rappaport, G. Duval-Beaupere, and M.-T. Corvol, Age-dependent responsivness of rabbit and human cartilage cells to sex steroids in vitro. J. Steroid Biochem. Mol. Biol. 40, 711 – 716 (1991). 282. E. Bonnelye, J. M. Vanacker, T. Dittmar, A. Begue, X. Desbiens, D. T. Denhardt, J. E. Aubin, V. Laudet, and B. Fournier, The ERR-1 orphan receptor is a transcriptional activator expressed during bone development. Mol. Endocrinol. 11, 905 – 916 (1997). 283. J. M. Vanacker, C. Delmarre, X. J. Guo, and V. Laudet, Activation of the osteopontin promoter by the orphan nuclear receptor estrogen receptor related alpha. Cell Growth Differ. 9, 1007 – 1014 (1998). 284. J. M. Vanacker, K. Pettersson, J. A. Gustafsson, and V. Laudet, Transcriptional targets shared by estrogen receptor- related receptors (ERRs) and estrogen receptor (ER) alpha, but not by ERbeta. EMBO J. 18, 4270 – 4279 (1999). 285. E. R. Levin, Cellular functions of the plasma membrane estrogen receptor. Trends Endocrinol. Metab. 10, 374 – 377 (1999). 286. M. J. Kelly and E. J. Wagner, Estrogen modulation of G-protein coupled receptors. Trends Endocrinol. Metab. 10, 369 – 374 (1999). 287. M. Lieberherr, B. Grosse, M. Kachkache, and S. Balsan, Cell signaling and estrogens in female rat osteoblasts: A possible involvement of unconventional nonnuclear receptors. J. Bone Miner. Res. 8, 1365 – 1376 (1993). 288. V. Le Mellay, B. Grosse, and M. Lieberherr, Phospholipase C beta and membrane action of calcitriol and estradiol. J. Biol. Chem. 272, 11902 – 11907 (1997). 289. V. Le Mellay, F. Lasmoles, and M. Lieberherr, Galpha q/11 and Gbeta gamma proteins and membrane signaling of calcitriol and estradiol. J. Cell. Biochem. 75, 138 – 146 (1999). 290. H. Endoh, H. Sasaki, K. Maruyama, K.-I. Takeyama, I. Waga, T. Shimizu, S. Kato, and H. Kawashima, Rapid activation of MAP kinase by estrogen in the bone cell line. Biochem. Biophys. Res. Commun. 235, 99 – 102 (1997). 291. C. V. Gay, N. L. Kief, and P. J. Bekker, Effect of estrogen on acidification in osteoclasts. Biochem. Biophys. Res. Commun. 192, 1251 – 1259 (1993).

KOMM AND BODINE 292. K. D. Brubaker and C. V. Gay, Specific binding of estrogen to osteoclast surfaces. Biochem. Biophys. Res. Commun. 200, 899 – 907 (1994). 293. K. D. Brubaker and C. V. Gay, Depolarization of osteoclast plasma membrane potential by 17 beta-estradiol. J. Bone Miner. Res. 14, 1861 – 1866 (1999). 294. A. Ikegami, S. Inoue, T. Hosoi, Y. Mizuno, T. Nakamura, Y. Ouchi, and H. Orimo, Immunohistochemical detection and northen blot analysis of estrogen receptors in osteoblastic cells. J. Bone Miner. Res. 8, 1103 – 1109 (1993). 295. K. Grandien, M. Backdahl, O. Ljunggren, J.-A. Gustafsson, and A. Berkenstam, Estrogen target tissue determines alternative promoter utilization of the human estrogen receptor gene in osteoblasts and tumor cell lines. Endocrinology 136, 2223 – 2229 (1995). 296. R. Delaveyne-Bitbol and M. Garabedian, In vitro responses to 17 beta-estradiol throughout pubertal maturation in female human bone cells. J. Bone Miner. Res. 14, 376 – 385 (1999). 297. M. E. Bolander, M. E. Joyce, S. E. Boden, B. Oliver, and A. Heydemann, Estrogen receptor mRNA expression during fracture healing in the rat detected by polymerase chain reaction amplification. In “Calcium Regulation and Bone Metabolism” (D. V. Cohn, F. H. Glorieux, and T. J. Martin eds.). Elsevier, New York, 1990. 298. R. Gruber, K. Czerwenka, F. Wolf, G. M. Ho, M. Willheim, and M. Peterlik, Expression of the vitamin D receptor, of estrogen and thyroid hormone receptor alpha- and beta-isoforms, and of the androgen receptor in cultures of native mouse bone marrow and of stromal/osteoblastic cells. Bone 24, 465 – 473 (1999). 299. A. Mahonen and P. H. Maenpaa, Steroid hormone modulation of vitamin D receptor levels in human MG-63 osteosarcoma cells. Biochem. Biophys. Res. Commun. 205, 1179 – 1186 (1994). 300. T. Bellido, G. Girasole, G. Passeri, X. P. Yu, H. Mocharla, R. L. Jilka, A. Notides, and S. C. Manolagas, Demonstration of estrogen and vitamin D receptors in bone marrow-derived stromal cells: Upregulation of the estrogen receptor by 1,25-dihydroxyvitamin-D3. Endocrinology 133, 553 – 562 (1993). 301. Q. Qu, M. Perala-Heape, A. Kapanen, J. Dahllund, J. Salo, H. K. Vaananen, and P. Harkonen, Estrogen enhances differentiation of osteoblasts in mouse bone marrow culture. Bone 22, 201 – 209 (1998). 302. W. H. Huang, A. T. Lau, L. L. Daniels, H. Fujii, U. Seydel, D. J. Wood, J. M. Papadimitriou, and M. H. Zheng, Detection of estrogen receptor alpha, carbonic anhydrase II and tartrate-resistant acid phosphatase mRNAs in putative mononuclear osteoclast precursor cells of neonatal rats by fluorescence in situ hybridization. J. Mol. Endocrinol. 20, 211 – 219 (1998). 303. R. O. C. Oreffo, V. Kusec, A. S. Virdi, A. M. Flanagan, M. Grano, A. Zambonin-Zallone, and J. T. Triffitt, Expression of estrogen receptor-alpha in cells of the osteoclastic lineage. Histochem. Cell. Biol. 111, 125 – 133 (1999). 304. C. K. Watts, M. G. Parker, and R. J. King, Stable transfection of the oestrogen receptor gene into a human osteosarcoma cell line. J. Steroid Biochem. 34, 483 – 490 (1989). 305. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, S. Mohan, D. J. Baylink, and L. E. Lanyon, Mechanical strain stimulates ROS cell proliferation through IGF-II and estrogen through IGF-I. J. Bone. Miner. Res. 14, 1742 – 1750 (1999). 306. A. Ikegami, S. Inoue, T. Hosoi, M. Kaneki, Y. Mizuno, Y. Akedo, Y. Ouchi, and H. Orimo, Cell cycle-dependent expression of estrogen receptor and effect of estrogen on proliferation of synchronized human osteoblast-like osteosarcoma cells. Endocrinology 782 – 789 (1994). 307. M. Ernst and G. A. Rodan, Estradiol regulation of insulin-like growth factor-I expression in osteoblastic cells: Evidence for transcriptional control. Mol. Endocrinol. 5, 1081 – 1089 (1991). 308. M. Kassem, R. Okazaki, S. A. Harris, T. C. Spelsberg, C. A. Conover, and B. L. Riggs, Estrogen effects on insulin-like

CHAPTER 10 Regulation of Bone Cell Function by Estrogens growth factor gene expression in a human osteoblastic cell line with high levels of estrogen receptor. Calcif. Tissue Int. 62, 60 – 66 (1998). 309. Q. Qu, P. L. Harkonen, J. Monkkonen, and H. K. Vaananen, Conditioned medium of estrogen-treated osteoblasts inhibits osteoclast maturation and function in vitro. Bone 25, 211 – 215 (1999). 310. B. Huo, D. A. Dossing, and M. T. Dimuzio, Generation and characterization of a human osteosarcoma cell line stably transfected with

337 the human estrogen receptor gene. J. Bone Miner. Res. 10, 769 – 781 (1995). 311. S. Koka, T. M. Petro, and R. A. Reinhardt, Estrogen inhibits interleukin-1beta-induced interleukin-6 production by human osteoblastlike cells. J. Interf. Cytokine Res. 18, 479 – 483 (1998). 312. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993).

CHAPTER 11

Skeletal Biology of Androgens KRISTINE M. WIREN AND ERIC S. ORWOLL Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201

VI. The Role of Aromatization: Effects of Replacement Sex Steroids after Castration VII. Gender Specificity in the Actions of Sex Steroids VIII. The Animal Model of Androgen Resistance IX. Summary References

I. Introduction II. Cellular Biology of the Skeletal Effects of Androgens III. Androgen Actions in Bone IV. The Role of Androgen Metabolism V. Animal Studies of the Skeletal Effects of Androgen

I. INTRODUCTION

estrogen replacement alone [8 – 11]. At the same time, treatment with nonaromatizable androgen alone and in combination with estrogen also result in distinct changes in bone mineral density in females [12]. These reports illustrate the independent actions of androgens and estrogens on the skeleton. Thus, in both men and women it is probable that androgens and estrogens each have important, yet distinct, functions during bone development and in the subsequent maintenance of skeletal homeostasis. With the awakening awareness of the importance of the effects of androgen on skeletal homeostasis, and the potential to make use of this information for the treatment of bone disorders, much is to be learned.

The obvious impact of menopause on skeletal health has focused much of the research on the general action of gonadal steroids on the specific effects of estrogen. However, androgens, independently, have important beneficial effects on skeletal development and on the maintenance of bone mass in both men and women. Thus, androgens (1) influence growth plate maturation and closure, helping to determine longitudinal bone growth during development, (2) mediate dichotomous regulation of cancellous and cortical bone mass, leading to a sexually dimorphic skeleton, (3) modulate peak bone mass acquisition, and (4) inhibit bone loss [1 – 4]. In castrate animals, replacement with nonaromatizable androgens (e.g., dihydrotestosterone) yields beneficial effects that are clearly distinct from those observed with estrogen replacement [5, 6]. In intact females, blockade of the androgen receptor with the specific androgen receptor antagonist hydroxyflutamide results in osteopenia [7]. Data suggest that combination therapy with both estrogen and androgenic steroids is more effective than

OSTEOPOROSIS, SECOND EDITION VOLUME 1

II. CELLULAR BIOLOGY OF THE SKELETAL EFFECTS OF ANDROGENS The mechanisms by which androgens affect skeletal homeostasis have been the focus of intensified research [13]. Androgen receptors are present in a variety of cells found in bone [14], clearly identifying bone as a target

339

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

340

WIREN AND ORWOLL

tissue for androgen. The direct actions of androgen that influence the complex processes of proliferation, differentiation, mineralization, and gene expression in the osteoblast have also been documented. Androgen effects on bone may also be indirectly modulated and/or mediated by other autocrine and paracrine factors in the bone microenvironment.

A. Molecular Mechanisms of Androgen Action in Bone Cells: The Androgen Receptor Direct characterization of AR expression in a variety of tissues, including bone, was made possible by the cloning of the AR cDNA [15,16]. The AR is a member of the class I (so-called classical or steroid) nuclear receptor superfamily, as are estrogen, progesterone, mineralocorticoid, and glucocorticoid receptors [17]. These steroid receptors are ligandinducible transcription factors with a highly conserved modular design. In the absence of ligand, AR is found in the nucleus [18] in a large complex of molecular chaperonins, consisting of loosely bound heat shock and other accessory proteins. As lipids, androgens can diffuse freely through the plasma membrane into the nucleus to bind the AR. Once bound by ligand, the AR is activated and released from this protein complex, allowing the formation of homodimers (or potentially heterodimers) that activate a cascade of nuclear events [19]. Bound to DNA, the AR influences transcription and translation of a specific network of genes, leading to the cellular response to the steroid. A steroid hormone target tissue is frequently defined as one that possesses both functional levels of the steroid receptor and a measurable response in the presence of hormone. Bone tissue clearly meets this standard with respect to androgen. Colvard et al. [20] first reported the presence of AR mRNA and specific androgen-binding sites in normal human osteoblastic cells. The abundance of both androgen and estrogen receptor (ER) proteins was similar, suggesting that androgens and estrogens each play important roles in skeletal physiology (Fig. 1). Subsequent reports have confirmed AR mRNA expression and/or the presence of androgen-binding sites in both normal and clonal, transformed osteoblastic cells derived from a variety of species [18, 21 – 25]. The size of the AR mRNA transcript in osteoblasts (about 10 kb) is similar to that described in prostate and other tissues [15], as is the size of the AR protein analyzed by Western blotting ( 110 kDa) [24]. There is a report of two isoforms of AR protein in human osteoblast-like cells ( 110 and 97 kDa) [26] similar to that observed in human fibroblasts [27]. Whether these isoforms possess similar functional activities in bone when expressed at similar levels as described in other tissue [28] is yet to be determined. The number of specific androgen-binding sites in osteoblasts varies, depending on methodology and the cell source, from 1000 to 14000 sites/cell [23,24,29,30], but is

FIGURE 1 Nuclear AR and ER binding in normal human osteoblast-like cells. Dots represent the mean calculated number of molecules per cell nucleus for each cell strain. (Left) Specific nuclear binding of [3H]R1881 (methyltrienolone, an androgen analogue) in 12 strains from normal men and 13 strains from normal women. (Right) Specific nuclear [3H]estradiol binding in 15 strains from men and 15 strains from women. Horizontal lines indicate mean receptor concentrations [20]. in a range seen in other androgen target tissues. Furthermore, the binding affinity of the AR found in osteoblastic cells (Kd  0.5 – 2  109) is typical of that found in other tissues. Androgen binding is specific, without significant competition by estrogen, progesterone, or dexamethasone [20,24,30]. Finally, testosterone and dihydrotestosterone (DHT) appear to have similar binding affinities [21,24]. All these data are consistent with the notion that the direct biologic effects of androgenic steroids in osteoblasts are mediated at least in part via classic mechanisms associated with the AR as a member of the steroid hormone receptor superfamily as described earlier. In addition to the classical AR present in bone cells, several other androgen-dependent signaling pathways may be present. Specific binding sites for weaker adrenal androgens [dehydroepiandrosterone (DHEA)] have been described [31], raising the possibility that DHEA or similar androgenic compounds can also have direct effects in bone. In fact, Bodine et al. [32] showed that DHEA caused a rapid inhibition of c-fos expression in human osteoblastic cells that was more robust than seen with the classical androgens (DHT, testosterone, androstenedione). Nevertheless, all androgenic compounds significantly increased transforming growth factor- (TGF-) activity in osteoblastic cells. Androgens may also be specifically bound in osteoblastic cells by a 63-kDa cytosolic protein [33]. There are reports of distinct AR polymorphisms identified in different races that may have a biological impact on androgen responses [34], but this has not been explored with respect to bone tissue. These different isoforms have the potential to interact in distinct fashions with other signaling molecules, such as cJun [35]. Finally, androgens may regulate osteoblast activity

CHAPTER 11 Skeletal Biology of Androgens

341

via rapid nongenomic mechanisms [36] through receptors at the bone cell surface [37], as has also been shown for estrogen [38]. The role and biologic significance of these nonclassical signaling pathways in androgen-mediated responses in bone are still relatively unexplored.

osteoclasts in human bone slices [14]. Because the major effects of androgens on skeletal remodeling and maintenance of bone mineral density seem to be mediated by cells of osteoblast lineage [43], the biologic relevance of potential AR expression osteoclasts is unclear.

B. Localization of Androgen Receptor Expression

C. Regulation of Androgen Receptor Expression

Clues about the potential sequella of AR signaling might be derived from a better understanding of the cell types in which expression is documented. In the bone microenvironment, the localization of AR expression has been described in intact human bone by Abu et al. [14] using immunocytochemical methods. In developing bone from young adults, ARs were predominantly expressed in active osteoblasts at sites of bone formation (Fig. 2). ARs were also observed in osteocytes embedded in the bone matrix. Importantly, the pattern of both AR distribution and the level of expression was similar in males and in females. Furthermore, AR was also observed within the bone marrow in mononuclear cells and endothelial cells of blood vessels. Expression of the AR has also been characterized in cultured osteoblastic cell populations isolated from bone biopsy specimens, determined at both the mRNA level and by binding analysis [30]. Expression varied according to the skeletal site of origin and age of the donor of the cultured osteoblastic cells: AR expression was higher at cortical and intramembranous sites and lower in cancellous bone. This distribution pattern correlates with androgen responsiveness. AR expression was highest in osteoblastic cultures generated from young adults and somewhat lower in samples from either prepubertal or senescent bone. Again, no differences were found between male and female samples, suggesting that differences in receptor number per se do not underlie development of a sexually dimorphic skeleton. AR and ER have also been shown in bone marrow-derived stromal cells [39], which are responsive to sex steroids during the regulation of osteoclastogenesis. Because androgens are so important in bone development at the time of puberty, it is not surprising that ARs are also present in epiphyseal chondrocytes [14,40]. The expression of ARs in such a wide variety of cell types known to be important for bone modeling during development, and remodeling in the adult, provides evidence for direct actions of androgens in bone and cartilage tissue, and these results illustrate the complexity of androgen effects on bone. Osteoclasts may be a target for sex steroid regulation, as they have been shown to possess ERs [41], but a direct effect of androgens on osteoclast function has not been demonstrated. Mizuno et al. [42] described the presence of AR immunoreactivity in mouse osteoclast-like multinuclear cells, but expression was not detected in bona fide

The regulation of AR expression in osteoblasts is incompletely characterized. Homologous regulation of AR by androgen has been described that is tissue specific. Upregulation by androgen exposure is seen in a variety of osteoblastic cells [18,25,44,45], whereas in prostatic tissue, downregulation of AR after androgen exposure is observed. The androgen-mediated upregulation observed in osteoblasts, at least in part, occurs through changes in AR gene transcription (Fig. 3). As in other tissues, increased AR protein stability may also play a part. No effect, or even inhibition, of AR mRNA by androgen exposure in other osteoblastic models has also been described [30,46]. The mechanism(s) that underlies tissue specificity in autologous AR regulation, and the possible biological significance of distinct autologous regulation of AR, is not yet understood. It is possible that AR upregulation by androgen in bone may result in an enhancement of androgen responsiveness at times when androgen concentrations are rising or elevated. In addition, AR expression in osteoblasts may be upregulated by exposure to glucocorticoids, estrogen, or 1, 25-dihydroxyvitamin D3 [26]. Except for the immunocytochemical detection of AR expression in bone slices described previously, regulation during osteoblast differentiation has not been described. Whether other hormones, growth factors, or agents influence AR expression in bone is also unknown. Finally, whether the AR in osteoblasts undergoes posttranslational processing that might influence receptor signaling (stabilization, phosphorylation, and so on) as described in other tissues [47,48], and the potential functional implications [49], is also unknown. Ligandindependent activation of AR has been described in other tissues [50], but has not yet been explored in bone. AR activity may also be influenced by receptor modulators, such as nuclear receptor coactivators or corepressors [51,52]. These coactivators/corepressors can influence the downstream signaling of nuclear receptors to reflect both the cellular context and the particular promoter. AR-specific coactivators have been identified [53], many of which interact with the ligand-binding domain of the receptor [54]. Expression and regulation of these modulators may thus influence the ability of steroid receptors to regulate gene expression in bone [55], but this has been underexplored with respect to androgen action. A preliminary report suggested the presence of androgen-specific coactivators in osteoblastic cells [56].

342

WIREN AND ORWOLL

FIGURE 2

Localization of AR in normal tibial growth plate and adult osteophytic human bone. (a) Morphologically, sections of the growth plate consist of areas of endochondral ossification with undifferentiated (small arrowhead), proliferating (large arrowheads), mature (small arrow), and hypertrophic (large arrow) chondrocytes. Bar: 80 m. An inset of an area of the primary spongiosa is shown in b. (b) Numerous osteoblasts (small arrowheads) and multinucleated osteoclasts (large arrowheads) on the bone surface. Mononuclear cells within the bone marrow are also present (arrows). Bar: 60 m. (c) In the growth plate, AR is predominantly expressed by hypertrophic chondrocytes (large arrowheads). Minimal expression is observed in mature chondrocytes (small arrowheads). Receptors are rarely observed in the proliferating chondrocytes (arrow). (d) In the primary spongiosa, the AR is predominantly and highly expressed by osteoblasts at modeling sites (arrowheads). Bar: 20 m. (e) In osteophytes, AR is also observed at sites of endochondral ossification in undifferentiated (small arrowheads), proliferating (large arrowheads), mature (small arrows), and hypertrophic-like (large arrow) chondrocytes. Bar: 80 m. (f) A higher magnification of e showing proliferating, mature, and hypertrophic-like chondrocytes (large arrows, small arrows, and very large arrows, respectively). Bar: 40 m. (g) At sites of bone remodeling, the receptors are highly expressed in osteoblasts (small arrowheads) and also in mononuclear cells in the bone marrow (large arrowheads). Bar: 40 m. (h) AR is not detected in osteoclasts (small arrowheads). Bar: 40 m. B, bone: C, cartilage; BM, bone marrow [14].

III. ANDROGEN ACTIONS IN BONE A. Effects of Androgens on Proliferation and Apoptosis in Osteoblastic Cells Androgens have direct effects on osteoblast proliferation and expression in vitro, but the nature of these effects remains controversial; both stimulation and inhibition of

osteoblast proliferation have been reported. Benz and coworkers. showed that prolonged androgen exposure in the presence of serum inhibited proliferation (cell counts) by 15 – 25% in a transformed human osteoblastic line (TE-85), with testosterone and DHT being nearly equally effective. Hofbauer et al. [57] examined the effect of DHT in hFOB/AR-6, an immortalized human osteoblastic cell line stably transfected with an AR expression construct (with

343

CHAPTER 11 Skeletal Biology of Androgens

FIGURE 3 Dichotomous regulation of AR mRNA levels in osteoblast-like and prostatic carcinoma cell lines after exposure to androgen. (A) Time course of changes in AR mRNA abundance after DHT exposure in human SaOS-2 osteoblastic cells and human LNCaP prostatic carcinoma cells. To determine the effect of androgen exposure on hAR mRNA abundance, confluent cultures of either osteoblast-like cells (SaOS-2) or prostatic carcinoma cells (LNCaP) were treated with 10 8 M DHT for 0, 24, 48, or 72 h. Total RNA was then isolated and subjected to RNase protection analysis with 50 g total cellular RNA from SaOS-2 osteoblastic cells and 10 g total RNA from LNCaP cultures. (B) Densitometric analysis of AR mRNA steady-state levels. The AR mRNA to -actin ratio is expressed as the mean  SEM compared to the control value from three to five independent assessments [44].

 4000 receptors/cell). In this line, DHT treatment inhibited cell proliferation by 20 – 35%. Finally, Kasperk et al. [26] reported that prolonged DHT treatment inhibited normal human osteoblastic cell proliferation (cell counts) in cultures pretreated with DHT. In contrast, Kasperk et al. [58,59] also demonstrated in serum free primary cultures of murine and passaged human osteoblast-like cells that a variety of androgens increase DNA synthesis (assessed by [3H]thymidine incorporation) up to nearly 300% (Table 1) and cell counts by 200%. Again, testosterone and nonaromatizable androgens (DHT and fluoxymesterone) were nearly equally effective regulators. Consistent with increased proliferation, testosterone and DHT have also been reported to cause an increase in creatine kinase activity and [3H]thymidine incorporation into DNA in rat diaphyseal bone [60]. The differences observed with androgen-mediated changes in osteoblastic cell proliferation may be due to the different model systems employed (transformed osteoblastic cells vs second to fourth

passage normal human cells) and/or may reflect differences in the culture conditions (e.g., state of differentiation, receptor number, times of treatment, phenol red-containing vs phenol red-free, or serum containing vs serumfree). These differences also suggest an underlying complexity and subtlety for androgen regulation of osteoblast proliferation. TABLE 1

Effect of Androgens (1 nM) on [3H]Thymidine Uptake into DNA of Mouse Bone Cellsa Counts per minute

% control

Control

235  14

100  6

DHT

479  23

204  9

 0.001

Testosterone

429  47

182  20

 0.01

Fluoxymesterone

423  44

180  18

 0.01

Methenolone

633  55

269  23

 0.001

a

Data from Kasperk et al. [58].

P

344

WIREN AND ORWOLL

As a component of the control of osteoblast survival, it is also important to consider programmed cell death, or apoptosis [61]. A variety of skeletal cell types have been shown to undergo apoptosis [62,63]. In particular, as the osteoblast population differentiates in vitro, the mature bone cell phenotype undergoes apoptosis [64]. Modulation of bone cell apoptosis by steroid hormones has been shown: glucocorticoids enhance and estrogen treatment prevents apoptosis of osteoclasts [65] and osteoblasts/osteocytes [66 – 68]. Furthermore, evidence shows that the osteocytic population is particularly sensitive to the effects of estrogen withdrawal, which induces apoptosis [69,70]. Androgen exposure has been shown to influence apoptosis in other tissues [71,72], but the effects of either androgen exposure or androgen withdrawal in bone have not been described.

variety of model systems that androgens either inhibit [57] or have no effect on alkaline phosphatase activity [25,74], which may reflect both the complexity and the dynamics of osteoblastic differentiation. There are also reports of androgen-mediated increases in type I -1 collagen protein and mRNA levels [21,73,74] and increased osteocalcin secretion [26]. Consistent with increased collagen production, androgen treatment has also been shown to stimulate mineral accumulation in a time-and dose-dependent manner [25,26,75]. These results suggest that under certain conditions androgens enhance osteoblast differentiation and may thus play an important role in the regulation of bone matrix production and/or organization. This effect is also consistent with an overall stimulation of bone formation, as is observed clinically after androgen treatment.

B. Effects of Androgens on Differentiation of Osteoblastic Cells

C. Interaction with Other Factors to Modulate Bone Formation and Resorption

Osteoblast differentiation can be characterized by changes in alkaline phosphatase activity and/or alterations in the expression of important extracellular matrix proteins, such as type I collagen, osteocalcin, and osteonectin. Enhanced osteoblast differentiation, as measured by increased matrix production, has been shown to result from androgen exposure. Androgen treatment in both normal osteoblasts and transformed clonal human osteoblastic cells (TE-89) appears to increase the proportion of cells expressing alkaline phosphatase activity, thus representing a shift toward a more differentiated phenotype (Fig. 4) [58]. Kasperk and colleogues. reported dose-dependent increases in alkaline phosphatase activity in both high and low alkaline phosphatase subclones of SaOS2 cells [73] and human osteoblastic cells [26]. However, there are also reports in a

The effects of androgens on osteoblast activity must certainly also be considered in the context of the very complex endocrine, paracrine, and autocrine milieu in the bone microenvironment. Systemic and/or local factors presumably act in concert to influence bone cell function. This has been well described with regard to modulation of the effects of estrogen on bone [see, for example, 76 – 78]. Androgens have also been shown to regulate well-known modulators of osteoblast proliferation or function. The most extensively characterized growth factor influenced by androgen exposure is TGF-. TGF- is stored in bone (the largest reservoir for TGF-) in a latent form and has been show to be a mitogen for osteoblasts [79,80]. Androgen treatment increases TGF- activity in human osteoblast primary cultures (Fig. 5). The expression of some TGF- mRNA transcripts (apparently TGF-2) was increased, but no effect on TGF-1 mRNA abundance was observed [32,59]. At the protein level, specific immunoprecipitation analysis reveals DHTmediated increases in TGF- activity to be predominantly in the form of TGF-2 [26,32]. DHT has also been shown to inhibit both TGF- and TGF--induced early gene (TIEG) expression that correlates with growth inhibition in this cell line [57]. TIEG has been shown to be a transcription factor that may mediate some TGF- effects [81]. These results are consistent with the notion that TGF- may mediate androgen effects on osteoblast proliferation. However, TGF1 mRNA levels are increased by androgen treatment in human clonal osteoblastic cells (TE-89), under conditions where osteoblast proliferation is slowed [21]. Thus, the specific TGF- isoform may determine osteoblast responses. It is interesting to note that at the level of bone, orchiectomy drastically reduces bone content of TGF- levels, and testosterone replacement prevents its occurrence

FIGURE 4

Effect of DHT on ALP-positive and ALP-negative cells in normal mouse, human osteosarcoma (TE89) monolayer cell culture, and normal human osteoblast line. (*** p0.001; ** p0.01; * p0.1). Control values in cells per mm2 for mouse bone cells were: 90  5; TE89 cells: 75  7; and human bone cells: 83  14 [58].

CHAPTER 11 Skeletal Biology of Androgens

Induction of total TGF- activity by gonadal and adrenal androgens in human osteoblast (hOB) cell-conditioned media. Cells were treated for 24 – 48 h with vehicle or steroids. After treatment, conditioned media were saved and processed for the TGF- bioassay. Results are presented as the mean  SEM of three to four experiments; *P  0.05; **P  0.02, ***P  0.0005 (Behren’s – Fisher t test) compared to the 48-h ethanol control. ETOH, ethanol; TEST, testosterone; DHT, DHT; ASD, androstenedione; DHEA, dehydroepiandrosterone; DHEA-S, DHEA-sulfate [32].

FIGURE 5

345 Other growth factor systems may also be influenced by androgens. Conditioned media from DHT-treated normal osteoblast cultures are mitogenic, and DHT pretreatment increases the mitogenic response to fibroblast growth factor and to insulin-like growth factor II (IGF-II) [59]. In part, this may be due to slight increases in IGF-II binding in DHT-treated cells [59], as the IGF-I and IGF-II content of osteoblast-conditioned media is not affected by androgen [59,83]. Although most studies have not found regulation of IGF-I or IGF-II abundance by androgen exposure [24,59,83], it has been shown that IGF-I mRNA levels are significantly upregulated by DHT [84]. Androgens also modulate the expression of components of the AP-1 transcription factor, as has been shown with the inhibition of cfos expression in proliferating normal osteoblast cultures [32]. Thus, androgens may accelerate osteoblast differentiation via a mechanism whereby growth factors or other mediators of differentiation are regulated by androgen exposure. Finally, androgens modulate responses to other important osteotropic hormones/regulators. Testosterone and DHT specifically inhibit the cAMP response elicited by parathyroid hormone or parathyroid hormone-related protein in the human clonal osteoblast-like cell line SaOS-2, whereas the inactive or weakly active androgen 17 epitestosterone had no effect (Fig. 7). This inhibition may

[82] (Fig. 6). These data support the findings that androgens influence cellular expression of TGF- and suggest that the bone loss associated with castration is related to a reduction in growth factor abundance induced by androgen deficiency.

FIGURE 6

Effects of orchiectomy and T replacement on isoforms of TGF- in long bones. Results are mean  SE of four to six animals. Rats underwent sham operation or orchiectomy (ORX), and 1 week later were given either placebo or 100 mg of testosterone in 60-day slow-release pellets. Specimens were obtained 6 weeks after surgery. All forms of TGF- were reduced by orchiectomy (at least p0.0002), whereas there was no change in those with testosterone replacement [82].

FIGURE 7 Actions of testosterone and 17 -epitestosterone on cAMP accumulation stimulated by hPTH1-34 (5.0 nM) or hPTHrP1-34 (5.0 nM) in human SaOS-2 cells. Cells were pretreated without or with the steroid hormones (10 9M) for 24 h. Each bar gives the mean value, and brackets give the SE for four to five dishes [85].

346 be mediated via an effect on the parathyroid hormone receptor-Gs-adenylyl cyclase [85 – 87]. The production of prostaglandin E2 (PGE2), another important regulator of bone metabolism, is also affected by androgens. Pilbeam and Raisz [88] showed that both DHT and testosterone potently inhibited both parathyroid hormone (Fig. 8) and interleukin-1-stimulated PGE2 production in cultured neonatal mouse calvaria. The effects of androgens on parathyroid hormone action and PGE2 production suggest that androgens could act to modulate (reduce) bone turnover in response to these agents. Finally, both androgen [13] and estrogen [77, 89] inhibit the production of interleukin-6 by osteoblastic cells. In stromal cells of the bone marrow, androgens have been shown to have potent inhibitory effects on the production of interleukin-6 (Table 2) and the subsequent stimulation of osteoclastogenesis by marrow osteoclast precursors [39]. Interestingly, adrenal androgens (androstenediol, androstenedione, dehydroepiandrosterone) have similar inhibitory activities on interleukin-6 gene expression and protein production by stroma [39]. The loss of inhibition of interleukin-6 production by androgen may contribute to the marked increase in bone remodeling and resorption that follows orchidectomy. Moreover, androgens inhibit the expression of the genes encoding the two subunits of the IL-6 receptor (gp80 and gp130) in murine bone marrow, another mechanism that may blunt the effects of this osteoclastogenic cytokine in intact animals [90]. In these aspects, the effects of androgens seem to be very similar to those of estrogen, which may also inhibit osteoclastogenesis via mechanisms that involve interleukin-6 inhibition.

WIREN AND ORWOLL

TABLE 2 Effect of Androgens on CytokineInduced IL-6 Production by Murine Bone Marrow Stromal Cells a Treatment

IL-6

IL-1 TNF

4.27  1.43

IL-1 TNF testosterone (10 12 M)

3.87  0.33

IL-1 TNF testosterone (10 11 M)

2.90  0.42*

IL-1 TNF testosterone (10 10 M)

2.09  0.33*

IL-1 TNF testosterone (10 9 M)

1.12  0.49*

IL-1 TNF testosterone (10 8 M)

1.03  0.04*

7

IL-1 TNF testosterone (10

M)

1.01  0.48*

IL-1 TNF DHT (10 12 M)

4.05  0.19

IL-1 TNF DHT (10 11 M)

2.97  0.48*

IL-1 TNF DHT (10 10 M)

2.31  0.86*

9

IL-1 TNF DHT (10

M)

1.72  0.43*

IL-1 TNF DHT (10 8 M)

0.65  0.21*

IL-1 TNF DHT (10 7 M)

1.41  0.82*

a

Murine stromal cells ( / LDA11 cells) were cultured for 20 h in the absence or the presence of different concentrations of either testosterone or DHT. Then IL-1 (500 U/ml) and TNF (500 U/ml) were added and cells were maintained for another 24 h in culture. Values indicate means (  SD) of triplicate cultures from one experiment. Data were analyzed by one-way ANOVA. IL-1, interleukin-1; TNF, tumor necrosis factor. *P  0.05, versus cells not treated with steroids as determined by Dunnet’s test. Neither testosterone nor DHT had an affect on cell number [39].

IV. THE ROLE OF ANDROGEN METABOLISM A. Metabolism of Androgens in Bone: Aromatase and 5 -Reductase Activities

FIGURE 8 Effect of testosterone on PTH-stimulated PGE2 production in cultured neonatal calvariae as a function of time. Each bone was precultured for 24 h in 1 ml medium with or without 10 9 M T and then transferred to similar medium with 2.5 nM PTH. Media were sampled (0.1 ml) at the indicated times. Data were corrected for the volume of medium removed. Each point represents the mean  SEM for six bones in one experiment. The effect of T on PTH-stimulated PGE2 production was significant (P  0.05) at 6, 12, and 24 h [88].

There is abundant evidence in a variety of tissues that the ultimate cellular effects of androgens may be the result not only of direct action of androgen, but also of the effects of sex steroid metabolites formed as the result of local enzyme activities. The most important of these androgen metabolites are estradiol (formed by the aromatization of testosterone) and 5 -DHT (the result of 5 reduction of testosterone). Evidence shows that both aromatase and 5 -reductase are present in bone tissue, at least to some measurable extent, but their biologic relevance is still controversial. 5 -reductase activity was first described in crushed rat mandibular bone by Vittek et al. [91]; and Schweikert et al. [92] reported similar findings in crushed human spongiosa. Two different 5 -reductase genes encode type 1 and type 2 isozymes in many mammalian species [93], but the

CHAPTER 11 Skeletal Biology of Androgens

isozyme present in human bone has not been characterized. In osteoblast-like cultures derived from orthopedic surgical waste, androstenedione (the major circulating androgen in women) can be reversibly converted to testosterone via 17-hydroxysteroid dehydrogenase activity and to 5 -androstanedione via 5 -reductase activity, whereas testosterone is converted to DHT via 5 -reductase activity [94]. The principal metabolites of androstenedione are 5 -androstanedione in the 5 -reductase pathway and testosterone in the 17-hydroxysteroid dehydrogenase pathway. Essentially the same results were reported in experiments with human epiphyseal cartilage and chondrocytes [95]. In general, the Km values for bone 5 -reductase activity are similar to those in other androgen responsive tissues [24,92]. The cellular populations in these studies were mixed and hence the specific cell type responsible for the activity is unknown. Turner and co-workers found that periosteal cells do not have detectable 5 -reductase activity [96], raising the possibility that the enzyme may be functional in only selected skeletal compartments, and that testosterone may be the active metabolite at this clinically important site. From a clinical perspective, the general importance of this enzymatic pathway is suggested by the presence of skeletal abnormalities in patients with 5 -reductase type 2 deficiency [97]. However, Bruch et al. [94] found no significant correlation between enzyme activities and bone volume. In mutant null mice lacking 5 -reductase type 1 (mice express very little type 2 isozyme), the effect on the skeleton cannot be analyzed due to midgestational fetal death [98]. Treatment of male animals with finasteride (an inhibitor of type 1 5 -reductase activity) does not recapitulate the effects of castration [99], indicating that reduction of testosterone to DHT by the type 1 isozyme is not a major determinant in the effects of gonadal hormones on bone. While available data point to a possible role for 5 reduction in the mechanism of action for androgen in bone, the clinical impact of this enzyme, which isozyme may be involved, whether it is uniformly present in all cells involved in bone modeling/remodeling, or whether local activity is important at all remains uncertain. The biosynthesis of estrogens from androgen precursors is catalyzed by the microsomal enzyme aromatase cytochrome P450 (P450arom, the product of the CYP19 gene). It is an enzyme well known to be both expressed and regulated in a very pronounced tissue-specific manner [100]. Aromatase activity has been reported in bone from mixed cell populations derived from both sexes [101 – 103] and from osteoblastic cell lines [24,104,105]. Aromatase expression in intact bone has also been documented by in situ hybridization and immunohistochemical analysis [103]. Aromatase mRNA is expressed predominantly in lining cells, chondrocytes, and some adipocytes, but there is no detectable expression in osteoclasts. At least in vertebral bone, the aromatase fibroblast (1b type) promoter is

347 predominantly utilized [103]. The enzyme kinetics in bone cells seem to be similar to those in other tissues, although the Vmax may be increased by glucocorticoids [105]. Aromatase can produce the potent estrogen estradiol, but can also form the weaker estrogen estrone from its adrenal precursors androstenedione and dehydroepiandrosterone [101]. In addition to aromatase itself, osteoblasts contain enzymes that are able to interconvert estradiol and estrone (estradiol-17B hydroxysteroid dehydrogenase) and to hydrolyze estrone sulfate to estrone (estrone sulfatase) [104]. Nawata et al. [101] have reported that dexamethasone and 1,25-dihydroxyvitamin D3 synergistically enhance aromatase activity and aromatase mRNA (P450arom) expression in human osteoblast-like cells. There is no other information concerning the regulation of aromatase in bone, although this is an area of obvious interest given the potential importance of the enzyme and its regulation by a variety of mechanisms (including androgens and estrogens) in other tissues [100,106]. The clinical impact of aromatase activity has been suggested by the reports of women [107] and men [108,109] with aromatase deficiencies who presented with a skeletal phenotype. The presentation of men with aromatase deficiency is very similar to that of a man with ER- deficiency, namely an obvious delay in bone age, lack of epiphyseal closure, and tall stature [110], suggesting that aromatase (and thus estrogen action) has a substantial role to play during skeletal development in the male (see Chapter 10 on estrogen action). In one case, estrogen therapy for a man with an aromatase deficiency was associated with an increase in bone mass [111]. Inhibition of aromatization in young growing orchidectomized male rats, with a nonsteroidal inhibitor, vorozole, results in decreases in bone mineral density and changes in skeletal modeling, as does castration with a resulting reduction in both androgens and estrogens. However, vorozole therapy induces less dramatic effects on bone turnover [112]. Inhibition of aromatization in older orchidectomized males resembles castration with similar increases in bone resorption and bone loss, suggesting that aromatase activity may also play a role in skeletal maintenance in males [113]. These studies herald the importance of aromatase activity (and estrogen) in the mediation of androgen action in bone. The finding of these enzymes in bone clearly raises the difficult issue of the origin of androgenic effects. Do they arise from direct androgen effects (as is suggested by the actions of nonaromatizable androgens) or to some extent from the local production of estrogenic intermediates? Nevertheless, there is substantial evidence that some, if in fact not most, of the biologic actions of androgens in the skeleton are direct. As noted previously, both in vivo and in vitro systems reveal the effects of the nonaromatizable androgen DHT to be essentially the same as those of testosterone (vida infra). In addition, blockade of the AR with the

348 receptor antagonist flutamide results in osteopenia as a result of reduced bone formation [7]. These reports clearly indicate that androgens, independent of estrogenic metabolites, have primary effects on osteoblast function. However, the clinical reports of subjects with aromatase deficiency also highlight the relevance of androgen metabolism to biopotent estrogens in bone. Elucidation of the regulation of steroid metabolism, and the potential mechanisms by which androgenic and estrogenic effects are coordinated, may have physiological, pathophysiological, and therapeutic implications.

B. Direct Effects of Androgens on Other Cell Types in Bone in Vitro Similar to the effects noted in osteoblastic populations, androgens regulate chondrocyte proliferation and expression. Androgen exposure promotes chondrogenesis as shown with increased creatine kinase and DNA synthesis after androgen exposure in cultured epiphyseal chondrocytes [40,114]. Increased [35S]sulfate incorporation into newly synthesized proteoglycan [115] and increased alkaline phosphatase activity [116] are androgen mediated. Regulation of these effects are obviously complex, as they were dependent on the age of the animals and the site from which chondrocytes were derived. Thus, in addition to effects on osteoblasts, multiple cell types in the skeletal milieu are regulated by androgen exposure.

V. ANIMAL STUDIES OF THE SKELETAL EFFECTS OF ANDROGEN The effects of androgens on bone remodeling have been examined fairly extensively in animal models. Much of this work has been in species not perfectly suited to reflect human bone metabolism (rodents), and certainly the field remains incompletely explored. Nevertheless, animal models do provide valuable insights into the effects of androgens at organ and cellular levels. Most studies of androgen action have been performed in male rats, in which rapid skeletal growth occurs until about 4 months of age, at which time epiphyseal growth slows markedly (although never completely ceases at some sites). Because the effects of androgen deficiency may be different in growing and more mature animals [1], it is appropriate to consider the two situations independently.

A. Effects on Epiphyseal Function and Bone Growth during Skeletal Development In most mammals there is a marked gender difference in bone morphology. The mechanisms responsible for these differences are complex and presumably involve both androgenic and estrogenic actions. Estrogens are particu-

WIREN AND ORWOLL

larly important for the regulation of epiphyseal function and act to reduce the rate of longitudinal growth via influences on chondrocyte proliferation and action, as well as on the timing of epiphyseal closure [117]. Androgens appear to have somewhat opposite effects and tend to promote long bone growth, chondrocyte maturation, and metaphyseal ossification. Androgen deficiency retards those processes [118]. Nevertheless, excess concentrations of androgen will accelerate aging of the growth plate and reduce growth potential [119], possibly via conversion to estrogens. Although the specific roles of sex steroids in the regulation of epiphyseal growth and maturation remain somewhat unresolved, evidence shows that androgens do have direct effects independent of those of estrogen. For instance, testosterone injected directly into the growth plates of rats increases plate width [120]. In a model of endochondral bone development based on the subcutaneous implantation of demineralized bone matrix in castrate rats, both testosterone and DHT increase the incorporation of calcium during osteoid formation [75]. Interestingly, in this model androgens reduced the incorporation of [35S]sulfate into glycosaminoglycans early in the developing cartilage. Certainly, androgens are known to interact considerably with the growth hormone – IGF system in the coordination of skeletal growth. Whereas androgens can induce a clear stimulation of growth in intact animals, in growth hormone deficiency that effect is essentially eliminated [121], underscoring the codependence of these two hormonal systems in the control of pubertal skeletal change. In sum, these data support the contention that androgens play a direct role in chondrocyte physiology, but how these actions are integrated with those of other regulators is unclear.

B. Effects on Bone Mass in Growing Male Animals The most dramatic effect of androgens during growth is on bone size. Male animals have larger bones, and particularly thicker cortices than females [117,122]. The effects of androgens on bone mass maturation can to some extent be assessed by observing the results of androgen withdrawal. In most studies, orchiectomy in young rats results in deficits in cortical bone mass within 2 – 4 weeks. The calcium content of the femur or tibia [123 – 127], whole femoral, tibial or body bone mineral density [99,127,128], and tibial diaphyseal cortical area [6] have been shown to be lower in castrated than in sham-operated controls. Similar trends have been reported in young, castrate male mice [129]. In animals followed for longer periods after castration (90 days), the density of cortical bone was reduced slightly (but not significantly), but bone area was clearly lessened in the diaphysis of the femur [126]. At least in part, the reduction in cortical bone mass appears to result

349

CHAPTER 11 Skeletal Biology of Androgens

from a reduction in the periosteal bone formation rate induced by gonadectomy in males [5,6]. This response differs from that induced by oophorectomy, which results in an increase in periosteal apposition in the period immediately after surgery (Fig. 9). This divergent trend in the periosteal response to castration in male and female animals abolishes the sexual dimorphism usually present in radial bone growth. Endosteal bone formation does not seem to be affected by orchiectomy [5]. As another indication that the cortical skeleton is affected by androgens, the characteristic acute increase in creatine kinase activity induced from diaphyseal bone by androgen treatment is abolished by orchiectomy [130]. For unclear reasons, it remains intact in epiphyseal specimens. Although castration in the male tends to slow growth and weight gain, the effects on cortical bone histomorphometry are present in pair-fed rats and in groups in which there was no difference in growth rates [5,6], indicating that the skeletal effects are not merely the indirect result of changes in body size or composition. The effects of androgens on cortical bone architecture have biomechanical implications. For instance, in studies of long-term androgen administration to female primates, Kasra and Grynpas [122] showed that cortical bone dimensions were increased and that tibias in treated animals were stronger, tougher, and stiffer. Thus, androgen deficiency during growth reduces cortical bone mass and strength primarily by blunting periosteal bone apposition. The lack of significant change in bone density following castration suggests that there is not a major impact of androgen deficiency on cortical porosity. Whereas estrogens appear to increase endosteal bone apposition, androgens probably have little effect at that site [117,122].

Cancellous bone mass is also reduced in castrate young male rats. Tibial metaphyseal bone volume and vertebral bone mineral density are clearly reduced [5,99,126], an effect that is seen rapidly following castration [5]. The reduction in bone volume is dramatic, with differences between control and castrate of 40 – 50% appearing in 4 – 10 weeks [99,131]. Rosen et al. [99] showed that measures of trabecular bone volume and mineral density diverged much more than areal measures of the proximal tibia or distal femur (by dual-energy X-ray absorptiometry) and speculated that this difference reflected a more intense bone deficit from trabecular than from cortical compartments. An important issue that remains unresolved is whether the bone deficit is a result of actual loss of bone mass following castration or whether the differences between castrate and control animals result from a failure of castrate animals to accrue bone normally. Nevertheless, bone changes following orchiectomy occur in the presence of an increase in skeletal blood flow [124,127], osteoclast numbers and surface [5], serum and urine calcium levels [5], and increased serum tartrateresistant acid phosphatase activity [128]. All these findings strongly suggest an increase in bone remodeling and bone resorption. However, in one report, distal femoral bone loss following castration was accompanied by a reduction in bone remodeling [132]. Parathyroid hormone concentrations have not been measured in these experiments, but vitamin D concentrations do not appear to be altered by orchiectomy [5]. In sum, trabecular bone mass, as well as cortical mass, appears to be dependent on adequate androgen action in the growing male animal, but the specific mechanisms that mediate that effect are not well delineated. For instance, the role of aromatization (see after) and the interaction of androgens and growth factors (e.g., IGF-1) are not clear.

C. Mature Male Animals

FIGURE 9

(A) The effect of ovariectomy (OVX) on periosteal bone formation rate. The mean  SE (vertical bar) and tetracycline-labeling period (horizontal line) for intact controls () and OVX () rats are shown as a function of time after OVX. p  0.01 for all OVX time points compared to intact controls. (B) The effect of orchiectomy (ORX) on periosteal bone formation rate. The mean  SE and tetracycline-labeling period for intact controls () and ORX () are shown as a function of time after ORX. p  0.01 for all ORX time points compared to same labeling period in intact controls [6].

In mature rats, castration also results in osteopenia. At a time when longitudinal growth has slowed markedly, pronounced differences between intact and castrate animals appear in cortical bone ash weight per unit length, crosssectional area, thickness, and bone mineral density [133 – 138] (Fig. 10 and 11). Periosteal bone accretion is reduced [138]. Because periosteal bone formation is known to be increased after oophorectomy, reduced by estrogens, and increased by androgens [6], the reduction in periosteal bone formation after orchidectomy is presumably mediated by a reduction in androgen action. Endocortical bone loss is accelerated in orchiectomized animals [135,139], and because endocortical remodeling is apparently not affected strongly by androgens, this may be a result of estrogen withdrawal by castration [5]. As might be expected in light of these changes, the maximum compressive load is

350

FIGURE 10

WIREN AND ORWOLL

Bone mass (unit ash: milligram of ash per millimeter of height), cross-sectional area, and wall thickness of the proximal femur cylinder from intact (I) and orchiectomized (O) rats (vertical bar  SE). Comparison among mean values at different ages for intact (I**) and castrated (O*) rats. *P 0.05; *** P 0.001 [135].

FIGURE 11 Microphotographs of 200-m-thick mid diaphyseal cross sections from 24-month-old intact (a) and orchidectomized (b) rats taken with a polarization microscope. Magnification  14 [135].

decreased in cortical bone, although when corrected for cortical mass, bone strength is normal [135]. In addition to changes in bone size, increased intracortical resorption cavities are reported to result from orchiectomy [133,140]. Cancellous bone volume is reduced rapidly after castration as well (Fig. 12) [133,138], and osteopenia becomes pronounced with time [137]. This bone loss appears to result in part from increased bone resorption, as it is associated with increased resorption cavities, osteoclasts, and blood flow [133,134,138]. Dynamic histomorphometric and biochemical measures of bone remodeling increase quickly [126,137], with evidence of increased osteoclast numbers only 1 week after castration [138] (Table 3). These changes include an increase in osteoblastic activity, as well as increased bone resorption [138,140]. In the SAMP6 mouse, a model of accelerated senescence in which osteoblastic function is impaired, the rise in remodeling following orchidec-

tomy is blunted, which has been interpreted as evidence that the early changes after gonadectomy are dependent on osteoblast-derived signals [43]. Although estrogen has been shown to increase osteoprotegerin (OPG) production (and thus to reduce osteoclastognesis), the effects of androgens on OPG are unknown [141]. This information will be key to the understanding of the effects of androgen on remodeling, as androgens reduce osteoclast formation [142]. This initial phase of increased bone remodeling activity appears to subside somewhat with time [126,134] and by 4 months there is evidence of a depression in bone turnover rates in some skeletal areas (Fig. 13) [134]. As in younger animals, indices of mineral metabolism are not altered by these changes in skeletal metabolism [137]. As a potential model for the effects of hypogonadism in humans, animal models therefore suggest an early phase of high bone turnover and bone loss after orchidectomy,

351

CHAPTER 11 Skeletal Biology of Androgens

FIGURE 12 (2) Microradiograph of a midsagittal longitudinal section from the proximal end of a femur in a control rat illustrating the normal appearance of distribution of compact and spongy bone ( 7). (3) Microradiograph of a midsagittal longitudinal section from the proximal end of a femur in a 4-month postcastrate rat. Much of the spongiosa has been lost. ( 7). (4) Microradiograph of a midsagittal longitudinal section from the distal end of a femur in a control rat ( 14). (5) Microradiograph of a midsagittal longitudinal section from the distal end of a femur in a 4-month postcastrate rat. Metaphaseal spongiosa has almost disappeared and compacta is thin ( 14) [133].

followed by a later reduction in remodeling rates. How long bone loss continues, and at what rate, is unclear. Both cortical and trabecular compartments are affected. The remodeling imbalance responsible for loss of bone mass appears complex, as there are changes in rates of both bone formation and resorption, and patterns that vary from one skeletal compartment to another. These changes are very similar to those noted in female animals after castration, in which a loss of estrogen has been associated with a stimulation of osteoblast progenitor differentiation, in association with an

increase in osteoclast numbers, bone resorption, and bone loss [143]. The mechanisms by which androgens mediate their skeletal effects in mature animals are not known. The direct effects of androgen on bone cells suggest that androgens act in part without intermediates. However estrogens derived from aromatization (either in bone or other tissues) play a role. Because sex steroids influence growth hormone and growth factor regulation, they may also be involved. In fact, growth hormone treatment of orchidectomized animals

352

WIREN AND ORWOLL

TABLE 3

Static Histomorphometry of Proximal Tibial Metaphysis Cancellous Bone in Sham-Operated or Orchiectomized (ORX) 4-Month-Old Male Rats after 1, 2, or 4 Weeksa

N

Bone volume a (%)

Trabecular number (n/mm)

Trabecular thickness (m)

Osteoblast surface b (%)

Osteoclast surface c (%)

Number of osteoclasts d (n/mm)

Sham

10

10  1

2.6  0.3

41.0  2.6

4.9  0.7

2.1  0.3

0.67  0.07

ORX

9

10  1

2.4  0.3

41.6  2.7

3.8  0.8

3.2  0.7

1.03  0.22

Sham

9

91

2.4  0.3

39.8  2.6

4.4  1.0

1.7  0.3

0.50  0.1

ORX

9

71

2.0  0.4

38.6  2.2

8.3  2.8

2.8  0.5

0.81  0.1

1 week

2 weeks

4 weeks Sham

8

91

2.2  0.4

41.9  4.0

1.9  0.g

2.0  0.5

0.57  0.2

ORX

12

61

1.4  0.2

41.8  2.5

12.8  2.2*

3.8  0.5

1.8  0.1

a Data expressed as mean  SEM. Significant treatment effect by two-way ANOVA: ap  0.02; cp  0.001, dp  0.002. Significant interaction effect by two-way ANOVA: bp  0.003. n, number. * Significantly different from sham , p  0.01. Data from Gunness and Orwoll [138].

recapitulates many (although not all) of the effects of testosterone replacement [140].

D. Androgens in the Female Animal

FIGURE 13 Evolution of the bone calcium turnover rate after castration (ratio of castrated/sham-operated animals). * P  0.05 [134].

Of course androgens are present in females as well as males and may affect bone metabolism. In castrate female rats, DHT administration suppresses elevated concentrations of bone resorption markers, as well as those of increased osteocalcin [144]. However, alkaline phosphatase activity increases further. Additional evidence to support the contention that androgens play a role in females includes the fact that antiandrogens are capable of evoking osteopenia in intact (i.e., fully estrogenized) female rats [7, 145] (Fig. 14). This obviously suggests that androgens provide crucial support to bone mass independent of estrogens. Of interest, the character of the bone loss induced by flutamide suggested that estrogen prevents bone resorption whereas androgens stimulate bone formation. In periosteal bone, DHT and testosterone appear to stimulate bone formation in orchidectomized young male rats, whereas in castrate females they suppress bone formation [6], perhaps reflecting an interaction or synergism between sex steroids and their effects on bone. There is also some information concerning androgens in other animal models, including primates. For instance, in adult female cynomolgus monkeys, testosterone treatment increased cortical and trabecular bone density as well as biomechanical strength [122].

353

CHAPTER 11 Skeletal Biology of Androgens

FIGURE 14

FIGURE 16

VI. THE ROLE OF AROMATIZATION: EFFECTS OF REPLACEMENT SEX STEROIDS AFTER CASTRATION

cancellous bone loss. In fact, estradiol resulted in an absolute increase in trabecular bone volume not achieved with androgen replacement. Similarly, estrogen was reported to antagonize the increase in blood flow resulting from castration and to increase bone ash weight more consistently than testosterone. Although data thus far available are incomplete, these studies raise obvious questions of the overlap between the actions of androgens and estrogens in bone. The gender reversal of estrogen replacement in male animals is also instructive. Nonaromatizable androgens are capable of preventing or reversing osteopenia and abnormalities in bone remodeling in oophorectomized females [6,146]. These actions apparently result from the suppression of cancellous bone resorption as well as stimulation of periosteal and endosteal bone formation [146]. Very similar results have been reported following the treatment of oophorectomized animals with dehydroepiandrosterone [6]. Moreover, blockage of androgen action with an AR antagonist in female rats already treated with an estrogen antagonist increases bone loss and indices of osteoclast activity more than treatment with an estrogen antagonist alone [147], indicating that ovarian androgens (apart from estrogens) exert a protective effect on bone. Analogously, androstenedione reduces (although does not abrogate) cancellous bone loss (and remodeling alterations) in oophorectomized animals treated with an aromatase inhibitor [148,149]. This protective effect was blocked by the addition of an AR antagonist [149]. Finally, whereas aromatase inhibition in male rats reduces bone mass, the large increase in remodeling induced by orchidectomy does not occur in these animals [112]. Also, orchidectomy in ERKO mice further reduces bone mass [150]. The latter observation implicates a role for androgens in the maintenance of bone mass in ERKO mice, but ER may also be playing a

Effects of buserelin alone, flutamide alone, and buserelin and flutamide in combination on total body calcium values after 4 weeks. Results are mean  SD, n  7 – 8 [7].

Essentially all of the alterations induced by orchiectomy (in both growing and mature animals) can be prevented at least in part by replacement with either testosterone or nonaromatizable androgens [6,124,127,130,131,139,140] (Figs. 15 and 16). These results strongly suggest that aromatization of androgens to estrogens cannot fully explain the actions of androgens on bone metabolism. However estrogens also seem to prevent bone loss following castration in male animals. Vanderschueren et al. [137] reported that estradiol (and nandrolone) was capable of not only preventing the increase in biochemical indices stimulated by orchiectomy, but also preventing cortical and

FIGURE 15

Cancellous bone volume (BV/TV) in control and androgen-treated orchiectomized rats (vertical) compared to a reference group consisting of age-matched untreated intact rats of group 1 (horizontal lines). *p  0.05 [131].

Osteoblast-lined cancellous bone surface (percentage of total surface) in control and androgen-treated orchiectomized rats (vertical bars) compared to a reference group consisting of age-matched untreated intact rats (horizontal lines). *p  0.05 [131].

354 role. Newly available ER and  KO mice have not yet been completely characterized. Aromatase-deficient mice have been developed [151], and the study of their skeletal phenotype should be very revealing. This should be a particularly useful model in view of human males with aromatase deficiency and impaired skeletal development [111]. In sum, these studies strongly support an independent role for androgens in skeletal homeostasis. Clearly estrogens play an important role in both genders. The interaction of the two is not yet defined but is of major interest.

VII. GENDER SPECIFICITY IN THE ACTIONS OF SEX STEROIDS Although still controversial, there may be gender specific responses in osteoblastic cells to sex steroids. Somjen and colleagues have shown that the increase in creatine kinase that occurs from bone cells in vivo and in vitro is gender specific (i.e., male animals or cells derived from male bones respond only to androgens, whereas females or female-derived cells respond only to estrogens) [130,152]. This gender specificity appears to depend on the previous history of exposure of animals to androgens (or estrogens). How much gender-specific effects might affect bone metabolism in the intact animal is completely unknown. In addition, in most mammals, there is a marked gender difference in morphology that results in a sexually dimorphic skeleton. The mechanisms responsible for these differences are obviously complex and presumably involve both androgenic and estrogenic actions on the skeleton. It is becoming increasingly clear that estrogens are particularly important for the regulation of epiphyseal function and act to reduce the rate of longitudinal growth via influences on chondrocyte proliferation and function, as well as on the timing of epiphyseal closure [117]. Androgens, however appear to have opposite effects on the skeleton, tending to promote long bone growth, chondrocyte maturation, and metaphyseal ossification. Furthermore, the most dramatic effect of androgens is on bone size, particularly cortical thickness [117,122], as androgens appear to have genderspecific effects on periosteal bone formation [6]. This difference has biomechanical implications, with thicker bones having greater resistance to fracture. At the same time, the response of the adult skeleton to the same intervention is distinct in males and females. For example, in a model of disuse osteopenia, antiorthostatic suspension significantly reduces the bone formation rate at the endocortical perimeter in males, but in females it decreases the bone formation rate along the periosteal perimeter [153]. Gender-specific responses in vivo and in vitro, and the mechanism(s) that underlies such responses in bone cells, may thus have significant implications in treatment options for metabolic bone disease.

WIREN AND ORWOLL

VIII. THE ANIMAL MODEL OF ANDROGEN RESISTANCE The testicular feminized (Tfm, AR deficient) male rat provides an interesting model for the study of the unique effects of androgens in bone. In these rats, androgens are presumed to be incapable of action, but estrogen and androstenedione concentrations are considerably higher than in normal males [154,155]. Clear increases also exist in Tfm male rats in serum concentrations of calcium, phosphorus, and osteocalcin, whereas IGF-1 concentrations are decreased. Estimates of bone mass suggest that Tfm rats have reduced longitudinal and radial growth rates, but that cancellous volume and density are similar to those of normal rats. In selected sites, measures of bone mass and remodeling were intermediate between normal male and female values. However, castration reduced bone volume markedly in Tfm male rats, suggesting a major role for estrogens in skeletal homeostasis (Fig. 17). This model indicates that androgens have an independent role to play in normal bone growth and metabolism, but the model is complex and not easily dissected. This is very similar to the study of humans with the androgen insensitivity syndrome. Marcus et al. [156] reported that there is a deficit in bone mineral density in women with androgen insensitivity even when compliance with estrogen replacement is excellent [156]. However, inadequate estrogen replacement appeared to worsen the deficit, and other environmental factors are difficult to quantitate.

IX. SUMMARY The influences of androgens on bone are obviously pervasive and complex. Androgens have a multiplicity of effects on skeletal cells in vitro. In vivo, they are particularly

FIGURE 17

Cancellous bone volume of the proximal metaphysis of the tibia in male, female, testicular feminized (Tfm), and orchiectomized (orch) male rats [155].

CHAPTER 11 Skeletal Biology of Androgens

dramatic during growth in males, but almost certainly play an important role during this period in females as well. Throughout the rest of life, androgens affect skeletal function in both sexes. Nevertheless, relatively little has been done to unravel the mechanisms by which androgens contribute to the physiology and pathophysiology of bone, and there is still much to be learned about the roles of androgens at all levels. The interaction of androgens and estrogens and how their respective actions can be utilized for specific diagnostic and therapeutic benefit are important but unanswered issues. With an increase in the understanding of the nature of androgen effects will come greater opportunities to use their positive actions in the prevention and treatment of a wide variety of bone disorders.

References 1. D. Vanderschueren and R. Bouillon, Androgens and bone. Calcif Tissue Int. 56, 341 – 346 (1995). 2. E. S. Orwoll, Androgens. In “Principles of Bone Biololy” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 563 – 580. Academic Press, San Diego, 1996. 3. D. Vanderschueren, Androgens and their role in skeletal homeostasis. Horm. Res. 46, 95 – 98 (1996). 4. E. S. Orwoll, (ed.), Androgen and bone: Clinical aspects. In “Osteoporosis in Men,” pp. 247 – 274. Academic Press, San Diego, 1999. 5. R. T. Turner, K. S. Hannon, L. N. Demers, J. Buchanan, and N. Bell, Differential effects of gonadal function in bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). 6. R. T. Turner, G. K. Wakley, and K. S. Hannon, Differential effects of androgens on cortical bone histomorphometry in gonadectomized male and female rats. J. Orthop. Res. 8, 612 – 617 (1990). 7. A. Goulding and E. Gold, Flutamide-mediated androgen blockade evokes osteopenia in the female rat. J. Bone Miner. Res. 8, 763 – 769 (1993). 8. N. B. Watts, M. Notelovitz, M. C. Timmons, W. A. Addison, B. Wiita, and L. J. Downey, Comparison of oral estrogens and estrogens plus androgen on bone mineral density, menopausal symptoms, and lipid-lipoprotein profiles in surgical menopause. Obstet. Gynecol. 85, 529 – 537 (1995). 9. L. G. Raisz, B. Wiita, A. Artis, A. Bowen, S. Schwartz, M. Trahiotis, K. Shoukril, and J. Smith, Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. J. Clin. Endocrinol. Metab. 81, 37 – 43 (1996). 10. M. J. Rosenberg, T. D. N. King, and M. C. Timmons, Estrogen-androgen for hormone replacement. J. Reprod. Med. 42, 394 – 404 (1997). 11. E. Barrett-Connor, Efficacy and safety of estrogen/androgen therapy. J. Reprod. Med. 43, 746 – 752 (1998). 12. V. Coxam, B. M. Bowman, M. Mecham, C. M. Roth, M. A. Miller, and S. A. Miller, Effects of dihydrotestosterone alone and combined with estrogen on bone mineral density, bone growth, and formation rates in ovariectomized rats. Bone 19, 107 – 114 (1996). 13. L. C. Hofbauer and S. Khosla, Androgen effects on bone metabolism: Recent progress and controversies. Eur. J. Endocrinol. 140, 271 – 286 (1999). 14. E. O. Abu, A. Horner, V. Kusec, J. T. Triffitt, and J. E. Compston, The localization of androgen receptors in human bone. J. Clin. Endocrinol. Metab. 82, 3493 – 3497 (1997).

355 15. C. Chang, J. Kokontis, and S. Liao, Structural analysis of complementary DNA and amino acid sequences of human and rat androgen receptors. Proc. Natl. Acad. Sci. USA 85, 7211 – 7215 (1988). 16. D. B. Lubahn, D. R. Joseph, M. Sar, J. Tan, H. N. Higgs, R. E. Larson, F. S. French, and E. M. Wilson, The human androgen receptor: Complementary deoxyribonucleic acid cloning, sequence analysis, and gene expression in prostate. Mol. Endocrinol. 2, 1265 – 1275 (1988). 17. D. J. Mangelsdorf, C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, and R. M. Evans, The nuclear receptor superfamily: The second decade. Cell 83, 835 – 839 (1995). 18. Y. H. Zhuang, M. Blauer, A. Pekki, and P. Tuohimaa, Subcellular location of androgen receptor in rat prostate, seminal vesicle and human osteosarcoma MG-63 cells. J. Steroid Biochem. Mol. Biol. 41, 693 – 696 (1992). 19. C. Chang, N. Saltzman, S. Yeh, W. Young, E. Keller, H.-J. Lee, C. Wang, and A. Mizokami, Androgen receptor: An overview. Crit. Rev. Eukaryot. Gene Expr. 5, 97 – 125 (1995). 20. D. Colvard, T. Spelsberg, E. Eriksen, P. Keeting, and L. B. Riggs, Evidence of steroid receptors in human osteoblast-like cells. Connect. Tissue Res. 20, 33 – 40 (1989). 21. D. J. Benz, M. R. Haussler, M. A. Thomas, B. Speelman, and B. S. Komm, High-affinity androgen binding and androgenic regulation of a1(I)-procollagen and transforming growth factor-B steady state messenger ribonucleic acid levels in human osteoblast-like osteosarcoma cells. Endocrinology 128, 2723 – 2730 (1991). 22. E. S. Orwoll, L. Stribrska, E. E. Ramsay, and E. Keenan, Androgen receptors in osteoblast-like cell lines. Calcif. Tissue Int. 49, 182 – 187 (1991). 23. P. Liesegang, G. Romalo, M. Sudmann, L. Wolf, and H.-U. Schweikert, Human osteoblast-like cells contain specific saturable, high-affinity glucocorticoid, androgen, estrogen and 1a,25-dihydroxycholecalciferol receptors. J. Andro. 15, 194 – 199 (1994). 24. Y. Nakano, I. Morimoto, O. Ishida, T. Fujihira, A. Mizokami, A. Tanimoto, N. Yanagihara, F. Izumi, and S. Eto, The receptor, metabolism and effects of androgen in osteoblastic MC3T3-E1 cells. Bone Miner. 26, 245 – 259 (1994). 25. M. Takeuchi, H. Kakushi, and M. Tohkin, Androgens directly stimulate mineralization and increase androgen receptors in human osteoblast-like osteosarcoma cells. Biochem. Biophys. Res. Commun. 204, 905 – 911 (1994). 26. C. H. Kasperk, G. K. Wakley, T. Hierl, and R. Ziegler, Gonadal and adrenal androgens are potent regulators of human bone cell metabolism in vitro. J. Bone Miner. Res. 12, 464 – 471 (1997). 27. C. M. Wilson and M. J. McPhaul, A and B forms of the androgen receptor are present in human genital skin fibroblasts. Proc. Natl. Acad. Sci. USA 91, 1234 – 1238 (1994). 28. T. Gao and M. J. McPhaul, Functional activities of the A and B forms of the human androgen receptor in response to androgen receptor agonists and antagonists. Mol. Endocrinol. 12, 654 – 663 (1998). 29. A. Masuyama, Y. Ouchi, F. Sato, T. Hosoi, T. Nakamura, and H. Orimo, Characteristics of steroid hormone receptors in cultured MC3T3-E1 osteoblastic cells and effect of steroid hormones on cell proliferation. Calcif. Tissue Int. 51, 376 – 381 (1992). 30. C. Kasperk, A. Helmboldt, I. Borcsok , S. Heuthe, O. Cloos, F. Niethard, and R. Ziegler, Skeletal site-dependent expression of the androgen receptor in human osteoblastic cell populations. Calcif. Tissue Int. 61, 464 – 473 (1997). 31. A. W. Meikle, R. W. Dorchuck, B. A. Araneo, J. D. Stringham, T. G. Evans, S. L. Spruance, and R. A. Daynes, The presence of a dehydroepiandrosterone-specific receptor binding complex in murine T cells. J. Steroid Biochem. Mol. Biol. 42, 293 – 304 (1992). 32. P. V. N. Bodine, B. L. Riggs, and T. C. Spelsberg, Regulation of c-fos expression and TGF-B production by gonadal and adrenal

356

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

WIREN AND ORWOLL androgens in normal human osteoblastic cells. J. Steroid Biochem. Mol. Biol. 52, 149 – 158 (1995). K. Wrogemann, G. Podolsky, J. Gu, and E. Rosenmann, A 63 -kDa protein with androgen-binding activity is not from the androgen receptor. Biochem. Cell. Biol. 69, 695 – 701 (1991). C. A. Pettaway, Racial differences in the androgen/androgen receptor pathway in prostate cancer. J. Natl. Med. Assoc. 91, 653 – 660 (1999). A. J. Grierson, R. C. Mootoosamy, and C. C. J. Miller, Polyglutamine repeat length influences human androgen receptor/c-Jun mediated transcription. Neurosci. Lett. 277, 9 – 12 (1999). H. Peterziel, S. Mink, A. Schonert, M. Becker, H. Klocker, and A. C. B. Cato, Rapid signalling by androgen receptor in prostate cancer cells. Oncogene 18, 6322 – 6329 (1999). M. Lieberherr and B. Grosse, Androgens increase intracellular calcium concentration and inositol 1,4,5-triphosphate and diacylglycerol formation via a pertussis toxin-sensitive G-protein. J. Biol. Chem. 269, 7217 – 7223 (1994). M. Lieberherr, B. Grosse, M. Kachkache, and S. Balsan, Cell signaling and estrogens in female rat osteoblasts: A possible involvement of unconventional nonnuclear receptors. J. Bone Miner. Res. 8, 1365 – 1376 (1993). T. Bellido, R. J. Jilka, B. F. Boyce, G. Girasole, H. Broxmeyer, A. Dalrymple, R. Murry, and S. C. Manolagas, Regulation of interleukin-6, osteoclastogenesis and bone mass by androgens: The role of the androgen receptor. J. Clin. Invest. 95, 2886 – 2895 (1995). A. Carracosa, L. Audi, M. A. Ferrandez, and A. Ballabriga, Biological effects of androgens and identification of specific dihydrotestosterone-binding sites in cultured human fetal epiphyseal chondrocytes. J. Clin. Endocrinol. Metab. 70, 134 – 140 (1990). M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Natl. Acad. Sci. USA 88, 6613 – 6617 (1991). Y. Mizuno, T. Hosoi, S. Inoue, A. Ikegami, M. Kaneki, Y. Akedo, T. Nakamura, Y. Ouchi, C. Chang, and H. Orimo, Immunocytochemical identification of androgen receptor in mouse osteoclast-like multinucleated cells. Calcif. Tissue Int. 54, 325 – 326 (1994). R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, The effects of androgen deficiency on murine bone remodeling and bone mineral density are mediated via cells of the osteoblastic lineage. Endocrinology 138, 4013 – 4021 (1997). K. M. Wiren, X. Zhang, C. Chang, E. Keenan, and E. Orwoll, Transcriptional up-regulation of the human androgen receptor by androgen in bone cells. Endocrinology 138, 2291 – 2300 (1997). K. Wiren, E. Keenan, X. Zhang, B. Ramsey, and E. Orwoll, Homologous androgen receptor up-regulation in osteoblastic cells may be associated with enhanced functional androgen responsiveness. Endocrinology 140, 3114 – 3124 (1999). L. C. Hofbauer, K. C. Hicok, M. J. Schroeder, S. A. Harris, J. A. Robinson, and S. Khosla, Development and characterization of a conditionally immortalized human osteoblastic cell line stably transfected with the human androgen receptor gene. J. Cell. Biochem. 66, 542 – 551 (1997). J. A. Kemppainen, M. V. Lane, M. Sar, and E. M. Wilson, Androgen receptor phosphorylation, turnover, nuclear transport, and transcriptional activation. J. Biol. Chem. 267, 968 – 974 (1992). T. Ikonen, J. J. Palvimo, P. J. Kallio, P. Reinikainen, and O. A. Janne, Stimulation of androgen-regulated transactivation by modulators of protein phosphorylation. Endocrinology 135, 1359 – 1366 (1994). L. J. Blok, P. E. de Ruiter, and A. O. Brinkmann, Androgen receptor phosphorylation. Endocrin. Res. 22, 197 – 219 (1996). A. Culig, A. Hobisch, M. V. Cronauer, C. Radmayr, J. Trapman, A. Hittmair, G. Bartsch, and H. Klocker, Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res. 54, 5474 – 5478 (1994).

51. K. B. Horwitz, T. A. Jackson, D. L. Bain, J. K. Richer, G. S. Takimoto, and L. Tung, Nuclear receptor coactivators and corepressors. Mol. Endocrinol. 10, 1167 – 1177 (1996). 52. N. J. McKenna, R. B. Lanz, and B. W. O’Malley, Nuclear receptor coregulators: cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 53. H. E. MacLean, G. L. Warne, and J. D. Zajac, Localization of functional domains in the androgen receptor. J. Steroid Biochem. Mol. Biol. 62, 233 – 242 (1997). 54. S. Yeh and C. Chang, Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells. Proc. Natl. Acad. Sci. USA 93, 5517 – 5521 (1996). 55. M. R. Haussler, C. A. Haussler, P. W. Jurutka, P. D. Thompson, J. C. Hsieh, L. S. Remus, S. H. Selznick, and G. K. Whitfield, The vitamin D hormone and its nuclear receptor: Molecular actions and disease states. J. Endocrinol. 154 (Suppl), S57 – S73 (1997). 56. K. M. Wiren, J. R. Lundblad, X.-W. Zhang, and E. S. Orwoll, Androgen receptor function in osteoblasts: Expression of coactivators and cointegrators. Endocr. Soc. [Abstract P2-466] 57. L. C. Hofbauer, K. C. Hicok, and S. Khosla, Effects of gonadal and adrenal androgens in a novel androgen-responsive human osteoblastic cell line. J. Cell Biochem. 71, 96 – 108 (1998). 58. C. H. Kasperk, J. E. Wergedal, J. R. Farley, T. A. Linkhart, R. T. Turner, and D. J. Baylink, Androgens directly stimulate proliferation of bone cells in vitro. Endocrinology 124(3), 1576 – 1578 (1989). 59. C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). 60. D. Somjen, Y. Weisman, A. Harell, E. Berger, and A. M. Kaye, Direct and sex-specific stimulation by sex steroids of creatine kinase activity and DNA synthesis in rat bone. Proc. Natl. Acad. Sci. USA 86, 3361 – 3365 (1989). 61. A. H. Wyllie, J. F. R. Kerr, and A. R. Currie, Cell death: The significance of apoptosis. Int. Rev. Cytol. 68, 251 – 307 (1980). 62. D. E. Hughes and B. F. Boyce, Apoptosis in bone physiology and disease. J. Clini. Pathol. Mol. Pathol. 50, 132 – 137 (1997). 63. S. C. Manolagas and R. S. Weinstein, New developments in the pathogenesis and treatment of steroid-induced osteoporosis. J. Bone. Miner. Res. 14, 1061 – 1066 (1999). 64. M. P. Lynch, C. Capparelli, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during bone-like tissue development in vitro. J. Cell. Biochem. 68, 31 – 49 (1998). 65. D. W. Dempster, B. S. Moonga, L. S. Stein, W. R. Horbert, and T. Antakly, Glucocorticoids inhibit bone resorption by isolated rat osteoclasts by enhancing apoptosis. J. Endocrinol. 154, 397 – 406 (1997). 66. S. C. Manolagas, Cellular and molecular mechanisms of osteoporosis. Aging (Milano) 10, 182 – 190 (1998). 67. R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, Inihibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoid: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 104, 1363 – 1374 (1999). 68. A. Gohel, M. B. McCarthy, and G. Gronowicz, Estrogen prevents glucocorticoid-induced apoptosis in osteoblasts in vivo and in vitro. Endocrinology 140, 5339 – 5347 (1999). 69. A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). 70. A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 18, 1243 – 1250 (1998). 71. M. T. Abreu-Martin, A. Chari, A. A. Palladino, N. A. Craft, and C. L. Sawyers, Mitogen-activated protein kinase kinase kinase 1 activates androgen receptor-dependent transcription and apoptosis in prostate cancer. Mol. Cell. Biol. 19, 5143 – 5154 (1999).

CHAPTER 11 Skeletal Biology of Androgens 72. I. K. Lim, H. J. Joo, K. S. Choi, E. Sueoka, M. S. Lee, M. S. Ryu, and H. Fuijiki, Protection of 5-dihydrotestosterone against TGF-Binduced apoptosis in FaO cells and induction of mitosis in HepG2 cells. Int. J. Cancer 72, 351 – 355 (1997). 73. C. H. Kasperk, K. Faehling, I. Borcsok, and R. Ziegler, Effects of androgens on subpopulations of the human osteosarcoma cell line SaOS2. Calcif. Tissue Int. 58, 376 – 382 (1996). 74. C. Gray, K. W. Colston, A. G. Mackay, M. L. Taylor, and T. R. Arnett, Interaction of androgen and 1,25-dihydroxyvitamin D3: Effects on normal rat bone cells. J. Bone Miner. Res. 7, 41 – 46 (1992). 75. S. P. Kapur and A. H. Reddi, Influence of testosterone and dihydrotestosterone on bond-matrix induced endochondral bone formation. Calcif. Tissue Int. 44, 108 – 113 (1989). 76. M. C. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). 77. M. Kassem, S. A. Harris, T. C. Spelsberg, and B. L. Riggs, Estrogen inhibits interleukin-6 production and gene expression in a human osteoblastic cell line with high levels of estrogen receptors. J. Bone. Miner. Res. 11, 193 – 199 (1996). 78. H. Kawaguchi, C. C. Pilbeam, S. J. Vargas, E. E. Morse, J. C. Lorenzo, and L. G. Raisz, Ovariectomy enhances and estrogen replacement inhibits the activity of bone marrow factors that stimulate prostaglandin production in cultured mouse calvariae. J. Clin. Invest. 96, 539 – 548 (1995). 79. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor- gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 80. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. Dallas, J. Q. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of transforming growth factor  on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, osteopontin, alkaline phosphatase, and type I collagen mRNa in long-term cultures of fetal rat calvarial osteoblasts. J. Bone. Miner. Res. 9, 855 – 863 (1994). 81. M. Subramaniam, S. A. Harris, M. J. Oursler, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Identification of a novel TGF-betaregulated gene encoding a putative zinc finger protein in human osteoblasts. Nucleic Acids Res. 23, 4907 – 4912 (1995). 82. R. K. Gill, R. T. Turner, T. J. Wronski, and N. H. Bell, Orchiectomy markedly reduces the concentration of the three isoforms of transforming growth factor  in rat bone, and reduction is prevented by testosterone. Endocrinology 139, 546 – 550 (1998). 83. E. Canalis, M. Centrella, and T. L. McCarthy, Regulation of insulinlike growth factor-II production in bone cultures. Endocrinology 129, 2457 – 2462. 84. F. Gori, L. C. Haufbauer, C. A. Conover, and S. Khosla, Effects of androgens on the insulin-like growth factor system in an androgenresponsive human osteoblastic cell line. Endocrinology 140, 5579 – 5586 (1999). 85. S. Fukayama and H. J. Tashjian, Direct modulation by androgens of the response of human bone cells (SaOS-2) to human parathyroid hormone (PTH) and PTH-related protein. Endocrinology 125, 1789 – 1794 (1989). 86. A. Gray, H. A. Feldman, J. B. McKinlay, and C. Longcope, Age, disease and changing sex hormone levels in middle-aged men: Results of the Massachusetts male aging study. J. Clin. Endocrinol. Metab. 73, 1016 – 1025 (1991). 87. A. Vermeulen, Clinical review 24: Androgens in the aging male. J. Clin. Endocrinol. Metab. 73, 221 – 224 (1991). 88. C. C. Pilbeam and L. G. Raisz, Effects of androgens on parathyroid hormone and interleukin-1-stimulated prostaglandin production in cultured neonatal mouse calvariae. J. Bone Miner. Res. 5, 1183 – 1188 (1990). 89. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993).

357 90. S.-C. Lin, T. Yamate, Y. Taguchi, V. Z. C. Borba, G. Girasole, C. A. O’Brian, T. Bellido, E. Abe, and S. C. Manolagos, Regulation of the gp80 and gp130 subunits of the IL-6 receptor by sex steroids in the murine bone marrow. J. Clin. Invest. 101, 1942 – 1950 (1997). 91. J. Vittek, K. Altman, G. G. Gordon, and A. L. Southren, The metabolism of 7a-3H-testosterone by rat mandibular bone. Endocrinology 94, 325 – 329 (1974). 92. H. U. Schweikert, W. Rulf, N. Niederle, H. E. Schafer, E. Kreck, and F. Kruck, Testosterone metabolism in human bone. Acta Endocrinol. 95, 258 – 264 (1980). 93. D. W. Russell and J. D. Wilson, Steroid 5-reductase: Two genes/two enzymes. Annu. Rev. Biochem. 63 (1994). 94. H.-R. Bruch, L. Wolf, R. Budde, G. Romalo, and H.-U. Schweikert, Androstenedione metabolism in cultured human osteoblast-like cells. J. Clin. Endocrinol. Metab. 75, 101 – 105 (1992). 95. L. Audi, A. Carrascosa, and A. Ballabriga, Androgen metabolism by human fetal epiphyseal cartilage and its condrocytes in primary culture. J. Clin. Endocrinol. Metab. 58, 819 – 825 (1984). 96. R. T. Turner, B. Bleiberg, D. S. Colvard, P. E. Keeting, G. Evans, and T. C. Spelsberg, Failure of isolated rat tibial periosteal cells to 5a reduce testosterone to 5a-dihydrotestosterone. J. Bone Miner. Res. 5, 775 – 779 (1990). 97. L. Fisher, M. D. Kogut, R. J. Moore, U. Goebelsmann, J. J. Weitzmann, H. Isaacs, J. E. Griffin, and J. D. Wilson, Clinical, endocrinological and enzymatic characterization of two patients with 5areductase deficiency: Evidence that a single enzyme is responsible for the 5a-reduction of cortisol and testosterone. J. Clin. Endocrinol. Metab. 47, 653 – 664 (1978). 98. M. S. Mahendroo, K. M. Cala, C. P. Landrum, and D. W. Russell, Fetal death in mice lacking 5-reductase type 1 caused by estrogen excess. Mol. Endocrinol. 11, 917 – 927 (1997). 99. H. N. Rosen, S. Tollin, R. Balena, V. L. Middlebrooks, W. G. Beamer, L. R. Donohue, C. Rosen, A. Turner, M. Holick, and S. L. Greenspan, Differentiating between orchiectomized rats and controls using measurements of trabecular bone density: A comparison among DXA, histomorphometry, and peripheral quantitative computerized tomography. Calcif. Tissue Int. 57, 35 – 39 (1995). 100. E. R. Simpson, M. S. Mahendroo, G. D. Means, M. W. Kilgore, M. M. Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C. R. Fisher, M. D. Michael, C. R. Mendelson, and S. E. Bulun, Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr. Rev 15, 342 – 355 (1994). 101. H. Nawata, S. Tanaka, R. Takayanagi, Y. Sakai, T. Yanase, S. Ikuyama, and M. Haji, Aromatase in bone cell: Association with osteoporosis in postmenopausal women. J. Steroid Biochem. Mol. Biol. 53, 165 – 174 (1995). 102. H.-U. Schweikert, L. Wolf, and G. Romalo, Estrogen formation from androstenedione in human bone. Clin. Endocrinol. 43, 37 – 42 (1995). 103. H. Sasano, M. Uzuki, T. Sawai, H. Nagura, G. Matsunaga, O. Kashimoto, and N. Harada, Aromatase in human bone tissue. J. Bone Miner. Res. 12, 1416 – 1423 (1997). 104. A. Purohit, A. M. Flanagan, and M. J. Reed, Estrogen synthesis by osteoblast cell lines. Endocrinology 131, 2027 – 2029 (1992). 105. S. Tanaka, M. Haji, Y. Nishi, T. Yanase, R. Takayanagi, and H. Nawata, Aromatase activity in human osteoblast-like osteosarcoma cell. Calcif. Tissue Int. 52, 107 – 109 (1993). 106. S. E. Abdelgadir, J. A. Resko, S. R. Ojeda, E. D. Lephart, M. J. McPhaul, and C. E. Roselli, Androgens regulate aromatase cytochrome P450 messenger ribonucleic acid in rat brain. Endocrinology 135, 395 – 401 (1994). 107. F. A. Conte, M. M. Grumbach, Y. Ito, C. R. Fisher, and E. R. Simpson, A syndrome of female pseudohermaphrodism, hypergonadotropic hypogonadism, and multicystic ovaries associated with missense mutations in the gene encoding aromatase (P450arom). J. Clin. Endocrinol. Metab. 78, 1287 – 1292 (1994).

358 108. C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effect of testosterone and estradiol in a man with aromatase deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). 109. A. Morishima, M. M. Grumbach, E. R. Simpson, C. Fisher, and K. Qin, Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J. Clin. Endocrinol. Metab. 80, 3689 – 3698 (1995). 110. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 111. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 112. D. Vanderschueren, E. Van Herck, J. Nijs, and A. G. H. Ederveen, Aromatase inhibition impairs skeletal modeling and decreases bone mineral density in growing male rats. Endocrinology 138, 2301 – 2307 (1997). 113. D. Vanderschueren, E. Van Herck, R. De Coster, and R. Bouillon, Aromatization of androgens is important for skeletal maintenance of aged male rats. Calcif. Tissue Int. 59, 179 – 183 (1996). 114. D. Somjen, Y. Weisman, Z. Mor, A. Harell, and A. M. Kaye, Regulation of proliferation of rat cartilage and bone by sex steroid hormones. J. Steroid Biochem. Mol. Biol. 40, 717 – 723 (1991). 115. M. T. Corvol, A. Carracosa, L. Tsagris, O. Blanchard, and R. Rappaport, Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: Influence of age and sex. Endocrinology 120, 1422 – 1429 (1987). 116. Z. Schwartz, E. Nasatzky, A. Ornoy, B. P. Brooks, W. A. Soskolne, and B. D. Boyan, Gender-specific, maturation-dependent effects of testosterone on chondrocytes in culture. Endocrinology 134, 1640 – 1647 (1994). 117. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994). 118. H. E. Lebovitz and G. S. Eisenbarth, Hormonal regulation of cartilage growth and metabolism. Vitam. Horm. (NY) 33, 575 – 648 (1975). 119. J. P. Iannotti, Growth plate physiology and pathology. Orthop. Clin. North Am. 21, 1 – 17 (1990). 120. S. G. Ren, S. Malozowski, P. Sanchez, D. E. Sweet, D. L. Loriaux, and F. Cassorla, Direct administration of testosterone increases rat tibial epiphyseal growth plate width. Acta Endocr. (Copenhagen) 21, 401 – 405 (1989). 121. L. E. Underwood and J. J. Van Wyk, Normal and aberrant growth. In “Williams Textbook of Endocrinology” (J. D. Wilson and D. W. Foster, eds.), 8th Ed., pp. 1079 – 1138. Saunders, Philadelphia, PA, 1992. 122. M. Kasra and M. D. Grynpas, The effects of androgens on the mechanical properties of primate bone. Bone 17, 265 – 270 (1995). 123. P. D. Saville and C. S. Lieber, Increases in skeletal calcium and femur cortex thickness in undernutrition. J. Nutr. 99, 141 – 144 (1969). 124. A. Schoutens, M. Verhas, M. L’Hermite-Baleriaux, M. L’Hermite, A. Verschaeren, N. Dourov, M. Mone, A. Heilporn, and A. Tricot, Growth and bone haemodynamic responses to castration in male rats: Reversibility by testosterone. Acta Endocrinol. 107, 428 – 432 (1984). 125. J. M. Hock, I. Gera, J. Fonseca, and L. G. Raisz, Human parathyroid hormone (1 – 34) increases bone mass in ovariectomized and orchidectomized rats. Endocrinology 122, 2899 – 2904 (1988). 126. D. Vanderschueren, I. Jans, E. Van Herck, K. Moermans, J. Verhaeghe, and R. Bouillon, Time-related increase of biochemical markers of bone turnover in androgen-deficient male rats. Bone Miner. 26, 123 – 131 (1994). 127. J. Kapitola, J. Kubickova, and J. Andrle, Blood flow and mineral content of the tibia of female and male rats: Changes following cas-

WIREN AND ORWOLL

128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

tration and/or administration of estradiol or testosterone. Bone 16, 69 – 72 (1995). B. Ongphiphadhanakul, S. Alex, L. F. Braverman, and D. T. Baran, Excessive L-thyroxine therapy decreases femoral bone densities in the male rat: Effect of hypogonadism and calcitonin. J. Bone Miner. Res. 7, 1227 – 1231 (1992). A. Ornoy, S. Giron, R. Aner, M. Goldstein, B. D. Boyan, and Z. Schwartz, Gender dependent effects of testosterone and 17bestradiol on bone growth and modelling in young mice. Bone Miner. 24, 43 – 58 (1994). D. Somjen, Z. Mor, and A. M. Kaye, Age dependence and modulation by gonadectomy of the sex-specific response of rat diaphyseal bone to gonadal steroids. Endocrinology 134, 809 – 814 (1994). G. K. Wakley, H. D. J. Schutte, K. S. Hannon, and R. T. Turner, Androgen treatment prevents loss of cancellous bone in the orchidectomized rat. J. Bone Miner. Res. 6, 325 – 330 (1991). J. Tuukkanen, Z. Peng, and H. K. Vaananen, Effect of running exercise on the bone loss induced by orchidectomy in the rat. Calcif. Tissue Int. 55, 33 – 37 (1994). C. S. Wink and W. J. L. Felts, Effects of castration on the bone structure of male rats: a model of osteoporosis. Calcif. Tissue Int. 32, 77 – 82 (1980). M. Verhas, A. Schoutens, M. L’hermite-Baleriaux, N. Dourov, A. Verschaeren, M. Mone, and A. Heilporn, The effect of orchidectomy on bone metabolism in aging rats. Calcif. Tissue Int. 39, 74 – 77 (1986). C. C. Danielsen, L. Mosekilde, and T. T. Andreassen, Long-term effect of orchidectomy on cortical bone from rat femur: Bone mass and mechanical properties. Calcif. Tissue Int. 50, 169 – 174 (1992). D. Vanderschueren and E. Van Herck, Androgen deficiency effects in the rat skeleton. Second Workshop on Using the Live Rate in Skeletal Studies, Chicago. D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Schot, and R. Bouillon, Bone and mineral metabolism in aged male rats: Short and long term effects of androgen deficiency. Endocrinology 130, 2906 – 2916 (1992). M. Gunness and E. S. Orwoll, Early induction of alterations in cancellous and cortical bone histology after orchiectomy in mature rats. J. Bone. Miner. Res. 10, 1735 – 1744 (1995). D. Vanderschueren, E. Van Herck, P. Schot, E. Rush, T. Einhorn, P. Geusens, and R. Bouillon, The aged male rat as a model for human osteoporosis: Evaluation by nondestructive measurements and biomechanical testing. Calcif. Tissue Int. 53, 342 – 347 (1993). G. Prakasam, J. K. Yeh, M.-M. Chen, M. Castro-Magana, C. T. Liang, and J. F. Aloia, Effects of growth hormone and testosterone on cortical bone formation and bone density in aged orchiectomized rats. Bone 24, 491 – 497 (1999). N. O. Vidal, H. Brandstrom, K. B. Jonsson, and C. Ohlsson, Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: Downregulation by gluccocorticoids. J. Endocrinol. 159, 191 – 195 (1998). S. Manolagas, T. Bellido, and R. Jilka, Sex steroids, cytokines and the bone marrow: New concepts on the pathogenesis of osteoporosis. In “Non-reproductive actions of Sex Steroids,” pp. 187 – 202. Wiley, Chichester, 1995. R. L. Jilka, K. Takahashi, M. Munshi, D. C. Williams, P. K. Roberson, and S. C. Manolagas, Loss of estrogen upregulates osteoblastogenesis in the murine bone marrow. J. Clin. Invest. 101, 1942 – 1950 (1998). R. A. Mason and H. A. Morris, Effects of dihydrotestosterone on bone biochemical markers in sham and oophorectomized rats. J. Bone Miner. Res. 12, 1431 – 1437 (1997). C. Lea, N. Kendall, and A. M. Flanagan, Casodex (a nonsteroidal antiandrogen) reduces cancellous, endosteal, and periosteal bone formation in estrogen-replete female rats. Calcif. Tissue Int. 58, 268 – 272 (1996).

CHAPTER 11 Skeletal Biology of Androgens 146. J. H. Tobias, A. Gallagher, and T. J. Chambers, 5a-dihydrotestosterone partially restores cancellous bone volume in osteopenic ovariectomized rats. Am J. Physiol. 267, E853 – E859 (1994). 147. C. K. Lea and A. M. Flanagan, Ovarian androgens protect against bone loss in rats made oestrogen deficient by treatment with ICI 182,780. J. Endocrinol. 160, 111 – 117 (1999). 148. C. K. Lea, V. Moxham, M. J. Reed, and A. M. Flanagan, Androstenedione treatment reduces loss of cancellous bone volume in ovariectomized rats in a dose-responsive manner and the effect is not mediated by oestrogen. J. Endocrinol. 156, 331 – 339 (1998). 149. C. K. Lea and A. M. Flanagan, Physiological plasma levels of androgens reduce bone loss in the ovariectomized rat. Am. J. Physiol. 274, E328 – 335 (1998). 150. M. Taki, K. S. Kimbro, T. F. Washburn, A. Kiyokawa, F. Okada, C. Fukuda, M. Miyamoto, and K. S. Korach, Effects of estrogen receptor destruction on femurs from male mice. J. Bone Miner. Res. 12 (Suppl. 1), S430 (1997). [Abstract] 151. C. R. Fisher, K. H. Graves, A. F. Parlow, and E. R. Simpson, Characterization of mice deficient in aromatase (arKO) because of targeted disruption of the cyp19 gene. Proc. Natl. Acad. Sci. USA 95, 6965 – 6970 (1998).

359 152. Y. Weisman, F. Cassoria, S. Malozowski, R. J. J. Krieg, D. Goldray, A. M. Kaye, and D. Somjen, Sex-specific response of bone cells to gonadal steroids: Modulation in perinatally androgenized females and in testicular feminized male rats. Steroids 58, 126 – 133 (1993). 153. T. A. Bateman, J. J. Broz, M. L. Fleet, and S. J. Simske, Differing effects of two-week suspension on male and female mouse bone metabolism. Biomed. Sci. Instrum. 34, 374 – 379 (1997). 154. D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Scot, K. Chung, R. S. Lucas, T. A. Einhorn, and R. Bouillon, Bone and mineral metabolism in the androgen-resistant (testicular feminized) male rat. J. Bone Miner. Res. 8, 801 – 809 (1993). 155. D. Vanderschueren, E. Van Herck, P. Geusens, A. Suiker, W. Visser, K. Chung, and R. Bouillon, Androgen resistance and deficiency have difference effects on the growing skeleton of the rat. Calcif. Tissue Int. 55, 198 – 203 (1994). 156. R. Marcus, D. Leary, D. L. Schneider, E. Shane, M. Favus, and C. A. Quigley, The contribution of testosterone to skeletal development and maintenance: Lessons from the androgen insensitivity syndrome. J. Clin. Endocrinol. Metab. 85, 1032 – 1037. 72 (2000).

CHAPTER 12

Coupling of Bone Resorption and Formation during Bone Remodeling THOMAS J. MARTIN GIDEON A. RODAN

I. II. III. IV. V.

St. Vincent’s Institute of Medical Research, Melbourne 3065, Australia Department of Bone Biology and Osteoporosis Research, Merck Sharp & Dohme Research Laboratories, West Point, Pennsylvania 19486

VI. Bone Mass Homeostasis VII. Role of Mechanical Function (Strain) in the Coupling of Bone Resorption to Bone Formation VIII. Integrated View of the Coupling of Bone Resorption and Bone Formation References

Introduction Sequence of Cellular Events in Bone Remodeling Interaction of Osteoblast Lineage Cells with Osteoclasts Similarities between Bone Remodeling and Inflammation Factors Proposed to Mediate the Coupling of Bone Formation to Bone Resorption

I. INTRODUCTION

In order to maintain skeletal balance, which one could name skeletal homeostasis, bone resorption intiates bone formation, which restores the amount of bone removed by resorption. Two main sets of observations pointed to the concept of coupling. Kinetic studies using radiotracers of calcium or strontium to estimate the rates of bone formation and resorption in animals or humans have shown that when bone resorption increases under physiological or pathological conditions, bone formation increases as well [1]. Hyperparathyroidism and estrogen deficiency are examples of conditions in which resorption and formation are increased. Similarly, when bone resorption decreases, e.g., during estrogen replacement therapy, bone formation does too [2]. The second type of evidence is histological. Examination of bone sections showed that osteoclastic bone resorption and osteoblastic bone formation are contiguous during bone remodeling and can be logically conceived to

Bone remodelling refers to the renewal process whereby small pockets of old bone, disposed throughout the skeleton and separated from others anatomically as well as chronologically, are replaced by new bone throughout adult life. Remodeling is essential for the maintenance of normal bone structure and for calcium homeostasis. The process is such that the entire adult human skeleton is replaced on the average in 10 years. The concept that bone formation and resorption are coupled during the bone remodeling process was developed in the 1970s. It is based on the principle that bone resorption occurs in order to release calcium for physiological needs and to reshape the bone structure to equip it better for its mechanical function, and formation restores the bone that is lost.

OSTEOPOROSIS, SECOND EDITION VOLUME 1

361

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

362 follow each other in the “basic multicellular unit” (BMU), which describes a packet of bone being resorbed and rebuilt [3]. The histomorphometric estimation of remodeling (or turnover) rates is based on the assumption that resorption precedes formation [3]. This “coupling” has been amply confirmed and, with some exceptions discussed later, is a general characteristic of bone remodeling. A number of discoveries in the late 1990s have revealed much about the molecular signaling pathways that influence local processes in bone, including osteoblast differentiation and function and the control of osteoclast formation and activity. As yet it is unknown what controls the extent of bone resorption and the extent of bone formation that replaces it and how in particular the two are contrived to be equal. This chapter outlines the cellular aspects of the bone remodeling process, how cells of the osteoblast lineage influence the resorption process, as well as bone formation, and considers current views of possible cellular and molecular mechanisms by which bone formation is coupled to resorption.

II. SEQUENCE OF CELLULAR EVENTS IN BONE REMODELING Cancellous bone remodeling starts on the bone surface, usually covered by lining cells, a single layer of flat cells derived from osteoblasts that have ceased to deposit matrix. These cells are proposed to respond to stimuli that initiate the remodeling cycle, of which there are many candidates both among circulating hormones and among cytokines generated locally, either by stromal/osteoblastic cells or cells of the immune system. The very nature of the remodeling process, occurring as it does in different parts of the skeleton at different times, highlights the importance of locally generated, regulatory factors in the process. When the cycle is initiated, say by parathyroid hormone (PTH) or by mechanical strain, which would generate cytokines or prostanoids [4], it has been proposed that the thin layer of nonmineralized matrix under these cells is initially digested by collagenase to expose the mineralized matrix that osteoclasts can resorb [5 – 7]. PTH is known to stimulate collagenase production and secretion in osteoblastic cells [8], and evidence in strong support of the role of collagenase is provided by a study in genetically mutated mice whose collagenase is unable to digest type I collagen [9]. In these mice the bone-resorbing action of PTH is severely attenuated. According to this model of initiation of resorption, osteoclasts (or their late precursors) near the stage of final maturation and activation must be available close to the required sites. Although this model is likely, there is no direct in vivo evidence for it. Certainly, once the process starts, then new osteoclasts must be generated, and all the bone-resorbing

MARTIN AND RODAN

hormones, cytokines, and prostanoids can promote this process through mechanisms discussed later in this chapter. Through the interaction of osteoblast lineage cells with osteoclasts, the latter start resorbing bone, a process lasting 2 to 4 weeks and carried out by groups of osteoclasts to a depth of about 30 m [10]. Toward the end of resorption, mononuclear cells, proposed to complete the resorption process, are seen at the bottom of resorption pits [11]. These are followed by cells, which may be related to macrophages, which appear along with preosteoblasts on the resorption surface. The transition from resorption to formation is called reversal, and the reversal plane can be identified microscopically with certain stains [12] or polarized light. The reversal line (cement line) contains a large abundance of osteopontin [13], which is produced both by osteoclasts and by osteoblasts. This is an arginine – glycine – aspartic acid (RGD) containing extracellular matrix protein [14], which interacts with vitronectin receptors v3 in osteoclasts and primarily v5 in osteoblasts. These integrin receptors were shown not only to mediate cell attachment to the extracellular matrix, but also to act as signal-transducing receptors [15]. Studies on mice that lack expression of the osteopontin gene indicate that osteopontin plays a role in bone resorption [16]. During maturation, osteoblasts become cuboidal, polarized cells, which are rich in endoplasmic reticulum and contain a large oval nucleus. Osteoblasts are connected to each other and form a contiguous layer. They seem to cooperate in the production of the extracellular bone matrix, as the dimensions of the fibrillar organization of collagen exceed the size of single cells. Moreover, because the organization of collagen is so well suited to withstand the tensile mechanical forces exerted on bone, osteoblasts probably sense and respond to mechanical strain. Osteocytes, which are embedded in bone and connected with each other and with surface cells by canaliculae, are particularly well situated to carry out this function. Another indication that this process may be important is the reduction in the rate of bone formation caused by immobilization [17] and weightlessness [18]. To replace the bone resorbed during a couple of weeks, bone formation continues for several months. Osteoblasts then become gradually flatter and the osteoid surface thins until a very flat layer of lining cells covers a very thin layer of nonmineralized matrix on quiescent bone surfaces. The remodeling of cortical bone follows similar stages, triggered by cues that may start in cells lining the Haversian canals or in osteocytes [19]. Osteoclasts excavate a “cutting cone,” which is refilled by osteoblast activity. Open questions in this sequence of events relate primarily to the signals that govern osteoclast and osteoblast recruitment and termination of osteoclast and osteoblast activity, the identity of cells at the reversal

363

CHAPTER 12 Coupling of Bone Resorption and Formation

phase, and the precise composition of the matrix on the reversal line.

III. INTERACTION OF OSTEOBLAST LINEAGE CELLS WITH OSTEOCLASTS Chambers [20] and Rodan and Martin [21] proposed that the stimulation of osteoclast activity by most agents is mediated by osteoblast lineage cells. This hypothesis was based on the fact that PTH receptors and PTH responses were much better documented for osteoblasts than osteoclasts, prostanoid stimulation of bone resorption correlated with effects on cyclic AMP accumulation in osteoblasts, and PTH and 1,25(OH)2D3 induced pronounced shape changes in osteoblastic cells. Similar changes in lining cells could expose the bone matrix, making it accessible to osteoclasts. Bone resorption studies with isolated osteoclasts supported the indirect modes of action of PTH, interleukin (IL)-1, tumor (TNF)-necrosis factor , 1,25-dihydroxyvitamin D, and the cytokines that use gp 130 as a signal transducer [22 – 25]. In vitro studies of osteoclast formation from bone marrow cells have convicingly demonstrated the requirement for osteoblasts or stromal cells [26]. This, together with the fact that actual contact between these cells and osteoclast precursors is necessary [26], strongly indicated that a molecule expressed on the cell membrane of osteoblast/stromal cells is important in promoting osteoclast formation. This prediction was fulfilled with the discovery of RANK ligand (RANKL), also known as “osteoclast differentiation factor” (ODF), a 316 amino acid, type II transmembrane protein, which is a member of the TNF ligand family [27,28] (see Chapter 3). Produced by osteoblastic stromal cells and activated T cells, RANKL, in the presence of colony stimulating factor1 (M-CSF), but without any accompanying stromal/osteoblastic cells, promotes the formation of osteoclasts from hemopoietic cells. When RANKL (-/-) mice were generated, they were found to be osteopetrotic because of failed osteoclast formation [29]. The action of RANKL is antagonized by osteoprotegerin (OPG), a soluble member of the TNF receptor family, produced by osteoblastic stromal cells as a decoy receptor, which inhibits RANKL action. Overexpression of OPG in transgenic mice results in osteopetrosis [30], and OPG (-/-) mice exhibit severe bone loss through excessive osteoclast formation and bone resorption [31]. The receptor for RANKL on hemopoietic cells is RANK (receptor activator of NF-KappaB), and RANK (-/-) mice are also osteopetrotic [32]. In addition to these effects on osteoclast formation, RANKL is able to activate mature osteoclasts [33] and OPG to inhibit their activity [34]. The formation of RANKL and of OPG in osteoblastic stromal cells is regulated by the hormones and cytokines that influence bone

resorption [35]. The identification in the promoter region of OPG of functional binding sites for the osteoblast master switch transcription factor, Cbfal [36] provides a molecular mechanism for the control of bone homeostasis by the osteoblast lineage. In addition to promoting and maintaining the osteoblast phenotype and thereby favoring bone formation, Cbfal can drive the stromal lineage toward the capacity to inhibit bone resorption by promoting OPG formation (see Chapter 6). These discoveries of the late 1990s have revealed much more of the cellular and molecular processes involved in the generation of resorption sites, and therefore in the bone remodeling process. There is little doubt of the importance in remodeling of the TNF and TNF receptor ligand family members. Their roles, to be considered in the context of hormone and drug actions upon bone, will undoubtedly have applications for new therapeutic approaches.

IV. SIMILARITIES BETWEEN BONE REMODELING AND INFLAMMATION In searching for molecules that link bone resorption to formation, it may be useful to point to striking similarities between bone remodeling and inflammation (Table 1). Inflammation starts with trauma produced by injury or by a foreign body. Bone remodeling starts with a stimulus, sometimes mechanical, which exposes the mineralized bone surface. In inflammation, the foreign body is recognized by white blood cells, e.g., macrophages, which start secreting cytokines and growth factors. The cytokines stimulate the production and migration of other white blood cells to the site of inflammation. Bone exposed to mechanical strain, which probably initiates remodeling, attracts mononuclear cells, which stain positively for nonspecific esterases [37]. Many cytokines involved in inflammation are potent stimulators of osteoclastic bone resorption and osteoclast differentiation in vitro [26] (see Chapter 13). IL-1 and TNF- are among the most powerful stimulators of bone resorption yet identified, inducing TABLE 1 Comparison of Sequence of Events and Cellular Interactions in Bone Remodeling and Inflammation Stage

Inflammation

Bone remodeling

Injury

Tissue damage foreign body

Mechanical?

Reaction

White blood cells, macrophage (local and hematogenous)

Osteoclasts

Repair

Mesenchymal cells (perivascular, fibroblasts)

Osteoblast lineage

Fibrosis, scar formation

Bone formation

364 osteoclast formation by effects on osteoblast/stromal cells to produce RANKL and M-CSF [38] and reduce OPG production. Most importantly a RANKL/RANK-independent pathway of the osteoclastogenic effect of TNF- has been demonstrated [39], through which TNF directly programs bone marrow macrophage precursors to osteoclasts, with their activation dependent on IL-1. The potential importance of this for inflammatory bone disease is evident. Indeed the impact of the immune system on bone cell function has become increasingly apparent. T cells produce many cytokines that have an impact on osteoblast or osteoclast differentiation. Although T cells represent about 2 to 3% of bone marrow cells, they become an abundant population in inflammatory states, e.g., periodontal disease and rheumatoid arthritis. IL-1 and TNF- are predominantly derived from monocytes. Among T-cell-derived cytokines, interferon (IFN)-, granulocyte macrophage colony stimulating factor (GM-CSF), IL-4, and IL-13 function as negative regulators of osteoclastogenesis [40]. IL-17 is a T-cell cytokine that promotes osteoclast formation and bone resorption through a prostaglandin dependent mechanism, similar to that of IL-1 [41]. IL-18, a stromal/osteoblastic product, inhibits osteoclast formation by acting on T cells to promote GM-CSF production [42]. It may be that local imbalances of pro and antiosteoclastogenic cytokines determine whether there is a net loss of bone in inflammatory conditions affecting bone directly. These discoveries of the late 1990s have revealed further cellular and molecular processes involved in the generation of resorption sites, and therefore in the bone remodeling process. One of the main antiosteoporotic effects of estrogen is to inhibit proliferation and differentiation of osteoclast precursors. The precise mechanism of these effects and the cellular targets of estrogen have yet to be fully elucidated. Estrogen receptors are expressed by monocytes, osteoblasts, and osteoclast precursors, as well as osteoclasts. Thus, estrogen could suppress osteoclastogenesis by regulating any one or more of these cell types. Current evidence indicates that production of at least five factors — IL-1, TNF, IL-6, and IL-6 receptor complex, M-CSF and GMCSF — is enhanced in conditions of estrogen deficiency [40]. In view of their pro-osteoclastogenic effects, all of these cytokines are considered potential mediators of the effects of estrogen on bone. A late phase in inflammation is the recruitment of fibroblasts, which produce matrix and encapsulate the foreign body. Fibroblast growth factor (FGF) and other growth factors are involved in this process [43]. The analogous phase in bone remodeling is the recruitment of osteoblasts, which cover the resorption surface with mineralized matrix. PGE, IL-1, transforming growth factor  (TGF-), and FGF were all shown to stimulate bone formation in vivo [44 – 47]. An important part of inflammation is neovascularization, probably stimulated by FGF, vascular endothelial growth factor

MARTIN AND RODAN

(VEGF), and other cytokines [48]. The importance of angiogenesis in osteogenesis has long been recognized, and bone-derived cells were shown to produce VEGF in response to PGE [49], and VEGF was shown to promote osteoclast formation in vitro [50]. This analogy suggests that in inflammation, T cells can substitute for osteoblastic stromal cells in promoting osteoclast formation, and evidence has been produced for that [51,52]. It suggests further that during the resorption process or at its termination, factors released by osteoclasts or cells present on the reversal surface, e.g., macrophages, attract the preosteoblasts to that surface. Interestingly, osteopontin, an extracellular molecule made by macrophages and osteoclasts, is found in inflammatory and atherosclerotic lesions [53 – 55], is present on the reversal surface [13], and is chemotactic [14]. Osteopontin is also produced abundantly by macrophages found in tumors [56] and tumors are often encapsulated by fibrous tissue. Osteopontin seems to be one of the molecules that plays a role in bone remodeling [16].

V. FACTORS PROPOSED TO MEDIATE THE COUPLING OF BONE FORMATION TO BONE RESORPTION Baylink and colleagues [57,58] suggested that “coupling” is due to bone formation factors released from the bone matrix during bone resorption. Indeed, a large number of substances that are mitogenic to osteoblasts or stimulate bone formation in vivo could be extracted from bone matrix [59]. These include insulin-like growth factor, (IGF) I and II [60], acidic and basic fibroblast growth factor (FGF) [61], transforming growth factor  1 and 2 [62] and TGF- heterodimers [63], bone morphogenetic proteins (BMPs) 2, 3, 4, 6, and 7 [64 – 66], platelet-derived growth factor (PDGF) [67], and probably others. Several questions should be considered regarding the role of these substances in the coupling of bone formation to bone resorption: (i) which cells produce them and under what circumstances, (ii) do they stimulate bone formation in vivo, (iii) can they be released from the matrix in active from and in controlled amounts during bone resorption, (iv) is there evidence for an increase in the abundance of these substances at sites of bone remodeling, and (v) are there regulatory mechanisms by which they are activated? IGF-I, IGF-II, bFGF, TGF-, and PDGF are produced by rat osteoblastic cells. IGF-I and IGF-II production is enhanced by stimulators of bone formation, such as PGE and PTH [68,69]. Elevated levels of IGF-I mRNA were found in bone from estrogen-deficient rats, where bone turnover is increased [70]. During bone growth in rats, there is a close association between osteogenesis and IGF-I expression [71]. However, following marrow ablation, which causes a

365

CHAPTER 12 Coupling of Bone Resorption and Formation

substantial increase in bone formation, the rise in IGF-I mRNA was seen after the histological appearance of differentiated osteoblasts, suggesting that it did not initiate bone formation in that system [72]. In human bone, the major form of IGF is IGF-II, which was also shown to be produced by human bone cells in culture [57,73]. Bone is one of the most abundant sources of TGF [74]. This growth factor is produced by all osteoblastic cells examined and its production is increased by estrogen and FGF (in osteosarcoma cells) [75,76]. BMPs are members of the TGF- superfamily. BMP-2 and BMP-4 are produced in adult bovine preodontoblasts [77], as well as in human fetal teeth [78]. BMP-7 (OP-1) was localized in human embryos in hypertrophic chondrocytes, osteoblasts, and periosteum, as well as other tissues [65], whereas BMP-3 was found in human embryonic lung and kidney, in addition to perichondrium, periosteum, and osteoblasts [79]. Both bFGF [80] and PDGF [67] were shown to be produced by bone cells or bone explants in culture. These factors could thus be involved in bone remodeling, but the time and site for their synthesis and secretion in vivo have not yet been determined. Prostaglandin E, primarily E2, is another bone cellproduced cytokine, which in vitro is upregulated by mechanical strain [81] and stimulates both bone resorption and formation [44,82]. Further light on mechanisms by which new bone is formed comes from the discovery of Cbfal, an essential transcription factor required for osteoblast differentiation [83,84]. Cbfal not only programs the primitive mesenchymal cell to express osteoblast-specific genes, but has also been shown to be important for maintaining the osteoblast phenotype in mature bone [85]. In remodeling, osteoblasts are recruited from a pool of committed cells and need to sustain the osteoblast phenotype. The regulated expression of Cbfal may be an important aspect of this. Study of its regulation is at an early stage, but it will be important to know whether growth factor effects proceed through the Cbfal pathway in the remodeling process. Several growth factors stimulate bone formation in vivo. IGF-I, injected into humans or rats, increases both bone resorption and bone formation [86], and reports on its effect on the bone balance are inconclusive [87]. When injected together with the IGF-binding protein IGF BP-3 into rats, it was reported to increase bone volume [88]. BMPs injected into bone stimulate bone formation locally and produce a positive bone balance. TGF-, from the same family of proteins, has a similar effect. When injected next to the periosteum or endosteum, there is substantial augmentation in local bone formation in rats and other species [46,89,90]. At the same time, there is an increase in endocortical bone resorption; thus, like IGF-I, TGF- seems to stimulate both resorption and formation; however, the local balance is clearly positive. bFGF, injected both locally and systemically, was also reported to increase bone formation [47,91]. Prostaglandin (PG)E1 and E2 have long been known to be

potent stimulators of bone formation when given either locally [92] or systemically, both to humans or to experimental animals [93,94]. These substances could thus contribute to the bone formation observed in remodeling if secreted or released in active form at the appropriate site and time. It was proposed that TGF-, which is produced as an inactive precursor in bone and bone cells [95,96], is present in the matrix and can be activated by acidification or proteolytic cleavage and is activated by resorbing osteoclasts [97]; TGF- activity was recovered from conditioned media of in vitro-resorbing osteoclasts [98]. It remains to be shown if the other growth factors also survive the proteolytic cleavage of the acidic hydrolases present in the resorption lacunae. Other questions raised by this model of coupling, via growth factor release from the matrix, relate to the time course and the distance between resorption and formation processes and whether activation can be controlled with sufficient precision in this way. Osteoclastic bone resorption proceeds for about 2 – 3 weeks before formation follows and continues for 3 – 4 months. The osteoblast precursors, which should respond to the “coupling factors”, could be many micrometers away from where active osteoclast resorption is in progress. Osteoblastic lineage cells produce TGF- in latent form, and IGFs are complexes bound to a family of specific, high-affinity binding proteins (IGFBPs), which regulate their bioavailability [99]. TGF- may be released from latent complexes at appropriate sites in bone by plasmin generated locally through the action of plasminogen activators in a manner that is controlled temporally and spatially by hormones and cytokines [100]. A similar local control could free IGF-I from association with its inhibitory binding protein [101]. Although there is no obvious skeletal phenotype in mice with inactivated genes for plasminogen activators, in vivo investigation of such possibilities would require treatment of such animals with anabolic agents such as PTH. Another way to explain coupling using an embryological paradigm is by implicating the surface left by osteoclasts, the so-called reversal surface, as an initiating influence. If active growth and/or differentiation factors are contained in this surface, they clearly could play a role by acting on osteoblasts or intermediary cells, which recruit the osteoblasts. Local matrix molecules, such as osteopontin, could also play such a role. Most of all, in vivo evidence is needed to show the presence by immunochemistry and the activity by bioassays, as illustrated for TGF- in vitro, of specific growth factors at bone remodeling sites. The technology for such investigations may become available soon.

VI. BONE MASS HOMEOSTASIS The putative biochemical mediators of bone resorption and bone formation described earlier do not explain a major aspect of “coupling,” namely, what determines the extent of

366 bone resorption and bone formation in each remodeling cycle. There clearly is bone mass homeostasis. All healthy individuals have a bone mineral density or bone mineral content that distributes normally around a mean with a standard deviation of about 10%. Bone mass is clearly genetically controlled [102,103] and there is much interest in possible genes that may be involved [103,104] (see chapter 26). Bone mass or bone mineral content, as measured, for example, in the lumbar spine noninvasively by dual energy X-ray absorptiometry (DXA), is determined by the amount of both cortical and cancellous bone. The amount of cortical bone is determined by periosteal bone formation, which continues throughout life, as well as endosteal and Haversian bone remodeling. Cancellous bone volume is determined by the relative extent of bone resorption and bone formation on the cancellous bone surface. The genetic determinants of bone mass thus should control these processes. Steroid hormones and sex hormones in particular are likely to participate in the genetic determination of bone mass. Men clearly have larger and thicker bones than women. The reduction in bone mass due to estrogen or androgen deficiency is well documented. Moreover, an estrogen receptor-deficient man [105], as well as mice in which the estrogen receptor was “knocked out” [106,107], was reported to have bone defects. In addition, the epiphyses closed very late in that man, suggesting an estrogen role in that function in males, as well as in females. It is not known exactly how steroids control bone formation or bone resorption. Receptors for sex steroids have been detected in osteoblastic cells from various species, including humans, and estrogens were shown to inhibit osteoclast activity in vitro [108]. The effect of estrogens on bone is discussed in detail in chapter 41. The sex steroids could have both direct and indirect effects, acting both on bone resorption and bone formation. Because these are systemic hormones and their concentration is most likly not determined by skeletal function, they do not generate the signals that terminate resorption or formation. They can provide a general background for the cellular responses to such signals. Frost [109] has indeed proposed that estrogen concentrations determine the “set point” for the response of the skeleton to mechanical signals. Mechanical stimuli may be the most direct input in the maintenance of bone mass and thus play a central role in bone mass homeostasis and by extension in the coupling of bone resorption and bone formation.

VII. ROLE OF MECHANICAL FUNCTION (STRAIN) IN THE COUPLING OF BONE RESORPTION TO BONE FORMATION The effect of mechanical forces on bone formation and resorption, mediated by the strain in the matrix, has long

MARTIN AND RODAN

been known and is very well documented. A decrease in mechanical load produced by immobilization or weightlessness causes a reduction in bone mass, which is due both to increased bone resorption, which occurs initially, and to decreased bone formation, which is sustained for a longer duration [17,110]. Eventually the system reaches a new steady state where the available bone mass is probably adequate for the prevailing mechanical load. The effects of weightlessness are a clear illustration of “uncoupling”, bone resorption being increased and bone formation decreased, implicating mechanical load in the “coupling” phenomenon. Examination of trabeculae from human vertebrae by scanning electron microscopy provided a visual illustration of this phenomenon [111]. Trabeculae, which were not mechanically loaded because one of their edges was loose and disconnected, showed very extensive resorption without evidence of bone formation. However, trabeculae, which were connected at both ends and thus mechanically loaded, had shallower resorption lacunae and evidence of bone formation. The increase in bone turnover produced by a reduction in mechanical load, the lower bone formation rate produced by immobilization, and the stabilization of bone mass at a new lower steady state pointed to mechanical strain as a factor that couples bone resorption to bone formation. In trabeculae mechanically weakened by resorption, and possibly cortical bone as well, bone formation would be stimulated until the strain is dissipated. The resulting structure would thus be ideally suited to sustain the prevailing strain. This would explain trabecular architecture, which matches the strain distribution in the bone and would explain the increase in the diameter of long bones to compensate for decreased bone mass observed in mice with osteogenesis imperfecta [112]. It could explain why the gain produced by an inhibitor of bone resorption, such as estrogen or bisphosphonates, eventually levels off, possibly when the existing bone mass has maximized its resistance to the prevailing loads. Consistent with this model is the fact that the potent bone resorption inhibitor alendronate did not cause any changes in bone mass in nonosteopenic minipigs [113]. Furthermore, in osteoporotic patients treated with inhibitors of bone resorption, bone mass continues to increase for some time after the filling of the remodeling spaces and the increment in mechanically loaded cortical bones at the hip, for example, is larger than in less loaded ones, such as the wrist. Experimental studies suggest that relatively limited mechanical input is probably sufficient to maintain the “genetically programmed” skeletal mass [114], that short-term bone loss can be caused by total rest or weightlessness (hypogravity) [115], and that very strenuous exercise, such as professional tennis playing, is necessary to produce exercise dependent significant increases in bone mass [116]. Thus, if we accept the fact that bone mass and bone structure are controlled by mechanical strain and that bone formation is

367

CHAPTER 12 Coupling of Bone Resorption and Formation

proportional to mechanical strain, we have to conclude that mechanical strain is at least one of the factors that couples bone resorption to formation. How this is brought about at biochemical and molecular levels has not been satisfactorily elucidated and remains one of the current challenges of skeletal reseach.

TABLE 2

Bone Remodeling/Skeletal Homeostasis

Skeletal functions Homeostasis of calcium and other ions Mechanical support and levers for muscle action Support of hemopoiesis Participating cells Osteoblasts (mesenchyme-derived cells)

VIII. INTEGRATED VIEW OF THE COUPLING OF BONE RESORPTION AND BONE FORMATION Bone has three major functions: mechanical support, homeostasis of calcium and other ions, and housing of hemopoiesis (see Table 2). Bone remodeling is initiated by stimuli generated to fulfill one of these functions. Mechanical stimuli are clearly local and were shown to be able to initiate remodeling [117]. The change in extracellular matrix strain, perceived by lining cells or osteocytes, likely primes a specific site in bone for remodeling. Local events could include the release of arachidonic acid metabolites, probably prostaglandin E, and other cytokines, plus possibly direct activation of osteoclasts or osteoclast precursors, leading to a local round of bone remodeling. It is not known what determines the extent of resorption and the depth or the size of the resorption lacunae. Bone formation is then initiated through (i) direct stimulation of osteoblast precursors by the initial factors, such as PGE; (ii) release of growth factors from the matrix or from other cells at the resorption site, such as vascular cells or macrophages; (iii) interaction with matrix molecules at the resorption surfaces, such as osteopontin; or (iv) all of the above. Once in progress, bone formation probably continues as long as the bone-forming cells perceive the osteogenic stimulus of mechanical strain. Likely transducers of that strain are integrins, through which cells are anchored in the matrix. Integrins were shown to act as signal transducing receptors [15] and to affect the phosphorylation of intracellular molecules in ways similar to those produced by growth factor receptors [118,119], possibly leading to similar outcomes of gene expression and protein synthesis. Both bone resorption and bone formation, presumbly controlled by the homeostatic inputs of mechanical forces, occur in an endocrine “field.” Thus, factors that suppress osteoclast activity, such as estrogens, would modulate the rate and possibly the extent of the resorptive phase, which would increase in the absence of estrogen or in the presence of stimulators of osteoclast activity, such as interleukins or parathyroid hormone. The same may hold true for bone formation where factors reported to enhance osteoblast activity, such as IGF, androgens, TGF- and BMPs, and others, may augment the rate and possibly the extent of the bone-forming phase. If the kinetic constraints, determined primarily by the rate of

Osteocytes (osteoblast lineage cells) Lining cells (osteoblast lineage cells) Marrow stromal cells (mesenchyme-derived cells) Osteoclasts (hemopoietically derived) B lymphocytes (hemopoietically derived) T lymphocytes (hemopoietically derived) Molecular mediators Major endocrine factors Parathyroid hormone Sex steroids (estrogens and androgens) Calcitonin Glucocorticoids Calcitriol [1,25(OH)2D] Thyroid hormones Paracrine/autocrine factors Insulin-like growth factors (IGFs) and IGF-binding proteins Transforming growth factor  family, including bone morphogenetic proteins (BMPs, 2, 4, 6, and others) Fibroblast growth factor family Prostanoids (PGE2 and others) Interleukins (IL-1, -6, -11, -17, and others) Colony-stimulating factors (M-CSF and GM-CSF) Tumor necrosis factors (RANK Ligand, TNF- and others) TNF receptors (osteoprotegerin) Parathyroid hormone-related peptide Matricrine factors Collagen (type I) Osteopontin Fibronectin Vitronectin Thrombospondin Mechanical stimuli

bone resorption, are not rate limiting, the steady-state bone density is most likely determined by the mechanical load. However, if bone resorption proceeds at an excessive pace that becomes rate limiting, such as in estrogen deficiency, bone formation, albeit increased, will not keep pace and bone loss will occur. Once bone resorption is slowed down by estrogen or other therapy, the bone mass can again reach its homeostatic level, determined by mechanical loads and possibly systemic regulators [120]. Thus, the relative

368 effects of various hormones and other factors would be to modulate the resorption or formation arm of the equation, permitting or preventing the maintenance of the homeostatic bone mass and the rate at which it is reached. This is another way of expressing the “set point hypothesis” for the mechanical control of bone mass [109]. For example, not all estrogen-deficient women or hyperthyroid patients with increased bone turnover lose bone to the same extent. This could also explain why exercise may be more effective in maintaining or gaining bone mass in estrogen-replete postmenopausal women. The second stimulus for bone turnover is calcium recruitment from the skeleton, initiated by PTH. Cortical bone seems to be a preferential target for PTH-stimulated bone resorption, possibly a reflection of the distribution of PTH receptors among osteoblast lineage cells [121,122]. However, elevated PTH concentrations may increase the general level of bone resorption wherever it occurs, augmenting bone loss produced by a lack of mechanical function. The beneficial effects of calcium supplements and vitamin D on hip fractures are consistent with this effect, as well as the bone gain observed after parathyroidectomy in vertebral BMD, which also contains a considerable amount of cancellous bone [123]. The third function of the skeleton, housing of the hemopoietic system, probably does not affect bone mass significantly under usual circumstances, but may lie at the basis of the response of the skeleton to lymphokines and other cytokines and explain the bone loss associated with inflammation in periarticular regions and the periodontium. It has been reported that increased red blood cell formation enlarges the marrow cavity [for review, see 124], and malignancies of the bone marrow, such as multiple myeloma, are clearly associated with extensive bone resorption. The feedback mechanisms, which come into play for enlarging the marrow cavity when increased hemopoiesis is needed, are probably mediated by the interleukins, which increase osteoclastogenesis, such as IL-1, IL-6, IL-11, and TNF-. Production of these interleukins during inflammation or in response to local tumors would lead to similar bone destruction. The similarity between the phases of inflammation and bone remodeling was pointed out earlier, but it is not yet known if factors involved in the later steps of inflammation, probably FGF and TGF-, which were shown to stimulate osteoblast proliferation, play a role in bone formation during normal bone remodeling. In addition to the local control described earlier, an intriguing possibility of central control of bone remodeling and homeostasis comes from the discovery that both ob/ob mice (leptin gene mutated to inactivity) and db/db mice (leptin receptor inactive) have greatly increased bone mass despite their hypogonadism and increased circulating glucocorticoid. Strikingly, this phenotype is corrected by intracerebroventricular injection of leptin [120], suggesting that the hypothalamus releases a bone mass regulatory substance.

MARTIN AND RODAN

In conclusion, an integrated view of the bone remodeling process should take into account that bone mass is controlled homeostatically by mechanical function in a hormonal environment (or by hormones in a mechanical field) and that there is a close relationship between bone and hemopoiesis and a similarity between bone remodeling and the cycle of inflammation and tissue repair. Rapidly accumulating new information should test and undoubtedly modify these hypotheses.

References 1. W. H. Harris and R. P. Heaney, Skeletal renewal and metabolic bone disease. N. Engl. J. Med. 280, 193 – 202 (1969). 2. K. M. Prestwood, C. C. Pilbeam, J. A. Burleson, F. N. Woodiel, P. D. Delmas, L. J. Deftos, and L. G. Raisz, The short term effects of conjugated estrogen on bone turnover in older women. J. Clin. Endocrinol. Metab. 79, 366 – 371 (1994). 3. H. M. Frost, Dynamics of bone remodeling. In “Bone Biodynamics,” pp. 315 – 333. Littel, Brown, Boston, 1964. 4. G. A. Rodan, Perspectives: Mechanical loading, estrogen deficiency, and the coupling of bone formation to bone resorption. J. Bone Miner. Res. 6, 527 – 530 (1991). 5. S. Sakamoto and M. Sakamoto, Bone collagenase, osteoblasts and cell-mediated bone resorption. In “Bone and Mineral Research,” pp. 49 – 102. Elsevier, Amsterdam, 1986. 6. T. J. Chambers, J. A. Darby, and K. Fuller, Mammalian collagenase predisposes bone surfaces to osteoclastic resorption. Cell Tissue Res. 241, 671 – 675 (1985). 7. T. J. Chambers and K. Fuller, Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic contact. J. Cell Sci. 6, 155 – 165 (1985). 8. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 – 1962 (1987). 9. W. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenaseresistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999). 10. N. J. Garrahan, P. I. Croucher, and J. E. Compston, A computerized technique for the quantitative assessment of resorption cavities in trabecular bone. Bone 11, 241 – 246 (1990). 11. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic disease. Endocr. Rev. 4, 379 – 408 (1986). 12. A. R. Villanueva, C. Sypitkowski, and A. M. Parfitt, A new method for identification of cement lines in undecalcified, plastic embedded sections of bone. Stain Technol. 61, 83 – 88 (1986). 13. J. Chen, M. D. McKee, A. Nanci, and J. Sodek, Bone sialoprotein mRNA expression and ultrastructural localization in fetal porcine calvarial bone: Comparisons with osteopontin. Histochem. J. 26, 67 – 78 (1994). 14. D. T. Denhardt and X. Guo, Osteopontin: A protein with diverse functions. FASEB J. 7, 1475 – 1482 (1993). 15. R. Hynes, Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69, 11 – 25 (1992). 16. H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontin-deficient mice are resistant to ovariectomy-induced bone resorption. Proc. Natl. Acad. Sci. USA 96, 10944 (1999).

CHAPTER 12 Coupling of Bone Resorption and Formation 17. M. Weinreb, G. A. Rodan, and D. D. Thompson, Osteopenia in the immobilized rat hind limb is associated with increased bone resorption and decreased bone formation. Bone 10, 187 – 194 (1989). 18. P. Backup, K. Westerlind, S. Harris, T. Spelsberg, B. Kline, and R. Turner, Spaceflight results in reduced mRNA levels for tissuespecific proteins in the musculoskeletal system. Am. J. Physiol. 266, E567 – E573 (1994). 19. T. M. Skerry, L. Bitensky, J. Chayen, and L. E. Lanyon, Early strainrelated changes in enzyme activity in osteocytes following bone loading in vivo. J. Bone Miner. Res. 4, 783 – 788 (1989). 20. T. J. Chambers, The cellular basis for bone resorption. Clin. Orthop. Relat. Res. 151, 283 – 293 (1980). 21. G. A. Rodan and T. J. Martin, Role of osteoblasts in hormonal control of bone resorption: A hypothesis. Calcif. Tissue Int. 33, 349 – 351 (1981). 22. P. M. J. McSheehy and T. J. Chambers, Osteoblastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology 118, 824 – 828 (1986). 23. B. M. Thomson, J. Saklatvala and T. J. Chambers, Osteoblasts mediate interleukin-1 stimulation of bone resorption by rat osteoclasts. J. Exp. Med. 164, 104 – 112 (1986). 24. H. M. Perry, W. Skogen, J. Chappel, A. J. Khan, G. Wilner and S. L. Teitelbaum, Partial characterization of a parathyroid hormone stimulated resorption factor(s) from osteoblast-like cells. Endocrinology 125, 2075 – 2082 (1989). 25. P. M. J. McSheehy and T. J. Chambers, 1,25-Dihydroxyvitamin D3 stimulates rat osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption. J. Clin. Invest. 80, 425 – 429 (1987). 26. T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endoc. Rev. 13, 66 – 80 (1992). 27. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, et al., Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 28. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, et al., Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 29. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, et al., OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 30. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, et al., Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 31. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, I. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. I. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 32. W. C. Dougall, M. Glaccum, K. Charrier, K. Rohirbach, K. Brasel, et al., RANK is essential for osteoclast and lymph node development. Genes Dev. 13, 2412 – 2424 (1999). 33. T. L. Burgess, Y. Qian, S. Kautman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. I. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin OPGL directly activates mature osteoclasts. J. Cell. Biol. 145, 527 – 538 (1999). 34. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentsz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998).

369 35. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). 36. K. Thirunavukkarasu, D. L. Halladay, R. R. Miles, Yang, R. I. S. Galvin, S. Chandrasekhar, T. S. Martin, and J. E. Onyia, The osteoblast-specific transcription factor Cbfal regulates the expression of osteoprotegerin (OPG), a potent inhibitor of osteoclast differentiation and function. J. Biol. Chem. in press. 37. R. L. Baron, L. Neff, P. Tran Van, J. R. Nefussi, and A. Vignery, Kinetic and cytochemical identification of osteoclast precursors and their differentiation into multinucleated osteoclasts. Am. J. Pathol. 122, 363 – 378 (1986). 38. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the human necrosis factor receptor and ligand families. Endocrin. Rev. 20, 345 – 357 (1999). 39. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of ODF/RANKL-RANK intereaction. J. Exp. Med. 191, 275 – 286. 40. T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentiation. Crit. Rev. Euk. Gene Expr. 8, 107 – 123 (1998). 41. S. Kotake, N. Udagawa, N. Takahashi, K. Matsuzaki, K. Itoh, S. Ishiyama, S. Saito, K. Inoue, N. Kamateni, M. T. Gillespie, T. J. Martin, and T. Suda, IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J. Clin. Invest. 103, 1345 – 1352 (1999). 42. N. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. Dunn, T. J. Martin, and M. T. Gillespie, IL-18 inhibits osteoclast formation via T-cell production of GM-CSF. J. Clin. Invest. 103, 1345 – 1352 (1999). 43. D. Gospodarowicz, Fibroblast growth factor. Clin. Orthop. Relat. Res. 257, 231 – 248 (1990). 44. W. S. S. Jee, H. Z. Ke, and X. J. Li, Long-term anabolic effects of prostaglandin-E2 on tibial diaphyseal bone in male rats. Bone Miner. 15, 33 – 55 (1991). 45. B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interluekin-1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989). 46. M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor. Endocrinology 124, 2991 – 2994 (1989). 47. T. Nakamura, K. Hanada, M. Tamura, T. Shibanushi, H. Nigi, M. Tagawa, S. Fukumoto, and T. Matsumoto, Stimulation of endosteal bone formation by systemic injections of recombinant basic fibroblast growth factor in rats. Endocrinology 136, 1276 – 1284 (1995). 48. J. Folkman, Angiogenesis in cancer, vascular rheumatoid and other disease. Nature Med. 1, 27 – 31 (1995). 49. S. Harada, J. A. Nagy, K. A. Sullivan, K. A. Thomas, N. Endo, G. A. Rodan, and S. B. Rodan, Induction of vascular endothelial growth factor expression by prostaglandin E2 and E1 in osteoblasts. J. Clin. Invest. 93, 2490 – 2496 (1994). 50. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa, and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 51. N. J. Horwood, V. Kartsogiannis, J. M. W. Quinn, E. Romas, T. J. Martin, and M. T. Gillespie, Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144 – 150 (1999).

370 52. Y. Y. Kong, U. Feige, U. Sarosi, B. Bolon, A. Tafuri, S. Morony, et al., Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 53. C. M. Giachelli, N. Bae, M. Almeida, D. T. Denhardt, C. E. Alpers, and S. M. Schwartz, Osteopontin is elevated during neointima formation in rat arteries and is a novel component of human atherosclerotic plaques. J. Clin. Invest. 92, 1686 – 1696 (1993). 54. E. R. OBrien, M. R. Garvin, D. K. Stewart, T. Hinohara, J. B. Simpson, S. M. Schwatz, and C. M. Giachelli, Osteopontin is synthesized by macrophage, smooth muscle, and endothelial cells in primary and restenotic human coronary atherosclerotic plaques. Arterioscler. Thromb. 14, 1648 – 1656 (1994). 55. T. Ikeda, T. Shirasawa, Y. Esaki, S. Yoshiki, and K. Hirokawa, Osteopontin mRNA is expressed by smooth muscle-derived foam cells in human atherosclerotic lesions of the aorta. J. Clin. Invest. 92, 2814 – 2820 (1993). 56. D. R. Senger, C. A. Peruzzi, C. F. Gracey, A. Papadopoulos, and D. G. Tenen, Secreted phosphoproteins associated with neoplastic transformation: Close homology with plasma proteins cleaved during blood coagulation. Cancer Res. 48, 5770 – 5774 (1988). 57. S. Mohan, J. C. Jennings, T. A. Linkhart, and D. J. Baylink, Primary structure of human skeletal growth factor: Homology with human insulin-like growth factor-II. Biochim. Biophys. Acta 966, 44 – 55 (1988). 58. S. Mohan and D. J. Baylink, The role of IGF-II in the coupling of bone formation to resorption. In “Modern Concepts of Insulin-like Growth Factors,” pp. 169 – 184. Elsevier, New York, 1991. 59. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsburn, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 60. C. Schmid and M. Ernst, Insulin-like growth factors. In “Cytokines and Bone Metabolism,” pp. 229 – 265. CRC Press, Boca Raton, FL, 1992. 61. S. B. Rodan and G. A. Rodan, Fibroblast growth factor and plateletdrived growth factor. In “Cytokines and Bone Metabolism,” pp. 115 – 145. CRC Press, Boca Raton, FL, 1992. 62. M. Centrella, T. L. McCarthy, and E. Canalis, Transforming growth factor-beta and remodeling in bone. J. Bone Joint Surg. 73-A, 1418 – 1428 (1991). 63. Y. Ogawa, D. K. Schmidt, J. R. Daschy, R. J. Chang, and C. B. Glaser, Purification and characterization of transforming growth factor beta2.3 and beta1.2 heterodimers from bovine bone. J. Biol. Chem. 267, 2325 – 2328 (1992). 64. S. E. Harris, M. Sabatini, M. A. Harris, J. Q. Feng, J. Wozney, and G. R. Mundy, Expression of bone morphogenetic protein messenger RNA in prolonged cultures of fetal rat calvarial cells. J. Bone Miner. Res. 9, 389 – 394 (1994). 65. S. Vukicevic, V. Latin, P. Chen, R. Batorsky, A. H. Reddi, and K. Sampath, Localization of osteogenic protein-1 (bone morphogenetic protein-7) during human embryonic development: High affinity to basement membranes. Biochem. Biophys. Res. Commun. 198, 693 – 700 (1994). 66. S. Vukicevic, V. M. Paralkar, N. S. Cunningham, J. S. Gutkind, and A. H. Reddi, Autoradiographic localization of osteogenin binding sites in cartilage and bone during rat embryonic development. Dev. Biol. 140, 209 – 214 (1990). 67. S. Rydziel, C. Ladd, T. L. McCarthy, M. Centrella, and E. Canalis, Determination and expression of platelet-derived growth factorAA in bone cell cultures. Endocrinology 130, 1916 – 1922 (1992). 68. T. L. McCarthy, M. Centrella, and E. Canalis, Parathyroid hormone enhances the transcript and polypeptide levels of insulin-like growth factor I in osteoblast-enriched cultures from fetal rat bone. Endocrinology 124, 1247 – 1253 (1989).

MARTIN AND RODAN 69. T. L. McCarthy, M. Centrella, L. G. Raisz, and E. Canalis, Prostaglandin E2 stimulates insulin-like growth factor I synthesis in osteoblast-enriched cultures from fetal rat bone. Endocrinology 128, 2895 – 2900 (1991). 70. R. T. Turner, P. H. Backup, P. H. Sherman, E. Hill, G. L. Evans, and T. C. Spelsberg, Mechanism of action of estrogen on intramembranous bone formation and regulation of osteoblast differentiation and activity. Endocrinology 131, 883 – 889 (1992). 71. D. M. Shinar, N. Endo, D. Halperin, G. A. Rodan, and M. Weinreb, Differential expression of insulin-like growth factor-I (IGF-I) and IGF-II messenger ribonucleic acid in growing rat bone. Endocrinology 132, 1158 – 1167 (1993). 72. L. J. Suva, J. G. Seedor, N. Endo, H. A. Quartuccio, D. D. Thompson, I. Bab, and G. A. Rodan, Pattern of gene expression following rat tibial marrow ablation. J. Bone Miner. Res. 8, 379 – 388 (1993). 73. S. Mohan and D. J. Baylink, Evidence that the inhibition of TE85 human bone cell proliferation by agents which stimulate cAMP production may in pat be mediated by changes in the IGF-II regulatory system. Growth Regul. 1, 110 – 118 (1991). 74. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 75. B. S. Komm, C. M. Terpening, D. J. Benz, K. A. Graeme, A. Gallegos, M. Korc, G. L. Greene, B. W. OMalley, and M. R. Haussler, Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241, 81 – 84 (1988). 76. M. Noda and R. Vogel, Fibroblast growth factor enhances type 1 transforming growth factor gene expression in osteoblast-like cells. J. Cell Biol. 109, 2529 – 2535 (1989). 77. M. Nakashima, H. Nagasawa, Y. Yamada, and A. H. Reddi, Regulatory role of transforming growth factor-beta, bone morphogenetic protein-2, and protein-4 on gene expression of extracellular matrix proteins and differentiation of dental pulp cells. Dev. Biol. 162, 18 – 28 (1994). 78. K. Heikinheimo, Stage-specific expression of decapentaplegic-Vgrelated genes 2,4 and 6 (bone morphogenetic proteins 2,4 and 6) during human tooth morphogenesis. J. Dent. Res. 73, 590 – 597 (1994). 79. S. Vukicevic, M. N. Helder, and F. P. Luyten, Developing human lung and kidney are major sites for synthesis of bone morphogenetic protein-3 (osteogenin). J. Histochem. Cytochem. 42, 869 – 875 (1994). 80. R. K. Globus, J. Plouet, and D. Gospodarowicz, Cultured bovine bone cells synthesize basic fibroblast growth factor and store it in their extracellular matrix. Endocrinology 124, 1539 – 1547 (1989). 81. D. Somjen, I. Binderman, E. Berger, and A. Harell, Bone remodelling induced by physical stress is prostaglandin E2 mediated. Biochim. Biophys. Acta 627, 91 – 100 (1980). 82. L. G. Raisz and T. J. Martin, Prostaglandins in bone and mineral metabolism. In “Bone and Mineral Research” (W. A. Pecked), pp. 286 – 310. Elsevier, Amsterdam, 1984. 83. P. Ducy, R. Zhan, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfal: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 84. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disrutpion of Cbfal results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 85. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfal-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1994). 86. E. M. Spencer, C. C. Liu, E. C. C. Si, and G. A. Howard, In vivo actions of insulin-like growth factor (IGF-I) on bone formation and resorption in rats. Bone 12, 21 – 26 (1991).

371

CHAPTER 12 Coupling of Bone Resorption and Formation 87. J. H. Tobias, J. W. M. Chow, and T. J. Chambers, Opposite effects of insulin-like growth factor-I on the formation of trabecular and cortical bone in adult female rats. Endocrinology 131, 2387–2392 (1992). 88. C. M. Bagi, R. Brommage, L. Deleon, S. Adams, D. Rosen, and A. Sommer, Benefit of systemically administered rhIGF-I/IGFBP-3 on cancellous bone in ovariectomized rats. J. Bone Miner. Res. 9, 1301 – 1312 (1994). 89. M. A. Critchlow, Y. S. Bland, and D. E. Ashhurst, The effects of age on the response of rabbit periosteal osteoprogenitor cells to exogeneous transforming growth factor-beta2. J. Cell Sci. 107, 499 – 516 (1994). 90. M. E. Joyce, A. B. Roberts, M. B. Sporn, and M. E. Bolander, Transforming growth factor and the initiation of chondrogenesis and osteogenesis in the rat femur. J. Cell Biol. 110, 2195 – 2207 (1990). 91. H. Kawaguchi, T. Kurokawa, K. Hanada, Y. Hiyama, M. Tamura, E. Ogata, and T. Matsumoto, Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 135, 774 – 781 (1994). 92. S. C. Marks, Jr. and S. C. Miller Local infusion of prostaglandin- E1 stimulates mandibular bone formation in vivo. J. Oral Pathol. 17, 500 – 505 (1988). 93. J. E. Lund, The toxicology of PGE1 and PGI2. In “Prostaglandins in Clinical Medicine: Cardiovascular and Thrombotic Disorders,” pp. 93 – 109. Yearbook Medical Publishers, New York, 1982. 94. W. S. S. Jee, S. Mori, X. J. Li, and S. Chan, Prostaglandin E2 enhances cortical bone mass and activates intracortical bone remodeling in intact and ovariectomized female rats. Bone 11, 253 – 266 (1990). 95. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardwik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of laten transforming growth factor beta (TGFbeta) complexes in osteoblast-like cell lines. J. Biol. Chem. 269, 6815 – 6822 (1994). 96. L. F. Bonewald, L. Wakefield, R. O. C. Oreffo, A. Escobedo, D. R. Twardzik, and G. R. Mundy, Latent forms of transforming growth factor beta (TGFbeta) derived from bone cultures: Identification of a naturally occurring 100-kDa complex with similarity to recombinant TGFbeta. Mol. Endocrinol. 5, 741 – 745 (1991). 97. R. O. C. Oreffo, G. R. Mundy, S. M. Seyedin, and L. F. Bonewald, Activation of the bone-derived latent TGF beta complex by isolated osteoclasts. Biochem. Biophys. Res. Commun. 158, 817 – 823 (1989). 98. J. Pfeilschifter and G. R. Mundy, Modulation of type beta transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 99. L. E. Underwood, J. P. Thissen, S. Lemozy, J. M. Ketelslegers, and D. R. Clemmons, Hormonal and nutritional regulation if IGF-I and its binding proteins. Horm. Res. 42, 145 – 151 (1994). 100. T. J. Martin, E. H. Allan, and S. Fukumoto, The plasminogen activator and inhibitor system in bone remodelling. Growth Regul. 3, 209 – 214 (1993). 101. P. G. Campbell, J. F. Novak, T. B. Yanosick, and J. H. McMaster, Involvement of the plasmin system in dissociation of the insulin-like growth factor-binding protein complex. Endocrinology 130, 1401 – 1412 (1992). 102. J. C. Christian, C. Slemenda, and C. C. Johnston, (1989). Heritability of bone mass: A longitudinal study in aging male twins. Am. J. Hum. Genet. 44, 429 – 433 (1989). 103. N. A. Morrison, J. Cheng Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 104. L. D. Spotila, C. D. Constantinou, L. Sereda, A. Ganguly, B. L. Riggs, and D. J. Prockop, Mutation in a gene for type I procollagen (COL1A2) in a woman with post-menopausal osteoporosis: Evidence for phenotypic and genotypic overlap with mild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423 – 5427 (1991). 105. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen

106.

107. 108.

109.

110.

111.

112.

113.

114. 115.

116.

117. 118.

119.

120.

121.

122.

123.

124.

resistance caused by a mutation in the estrogen-receptor gene in a man. N. Eng. J. Med. 331, 1056 – 1061 (1994). D. B. Lubahn, J. S. Moyer, T. S. Golding, J. F. Couse, K. S. Korach, and O. Smithies, Alteration of reproductive function but not prenatal sequal development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. USA 90, 11162 – 11166 (1993). K. S. Korach, Insights from the study of animals lacking functional estrogen receptor. Science 266, 1524 – 1527 (1994). M. J. Oursler, L. Pederson, L. Fitzpatrick, B. L. Riggs, and T. Spelsberg, Human giant cell tumors of the bone (osteoclastomas) are estrogen target cells. Proc. Natl. Acad. Sci. USA 91, 5227 – 5231 (1994). H. M. Frost, The mechanostat: A proposed pathogenic mechanism of osteoporosis and the bone mass effects of mechanical and nonmechanical agents. Bone 2, 73 – 85 (1987). D. T. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner Res. 3, 409 – 414 (1988). L. Mosekilde, Consequences of the remodeling process for vertebral trabecular bone structure: A scanning electron microscopy study (uncoupling of unloaded structures). Bone Miner. 10, 13 – 35 (1990). J. Bonadio, K. J. Jepsen, M. K. Mansoura, R. Jaenisch, J. L. Kuhn, and S. A. Goldstein, A murine skeletal adaptation that significantly increases cortical bone mechanical properties. J. Clin. Invest. 92, 1697 – 1705 (1993). M.-H. Lafage, R. Balena, M. A. Battle, M. Shea, J. G. Seedor, H. Klein, W. C. Hayes, and G. A. Rodan, Comparison of alendronate and sodium fluoride effects on cancellous and cortical bone in minipigs. J. Clin. Invest. 95, 2127 – 2133 (1995). C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). V. Schneider, V. Ogancy, A. LeBlanc, A. Rakhmanov, A. Bakulin, A. Grigoriev, and L. VaRonin, Spaceflight bone loss and changes in fat and lean body mass. J. Bone Miner. Res. 7, S122 (1992). A. L. Huddleston, D. Rockwell, D. N. Kulund, and R. B. Harrison, Bone mass in lifetime tennis athletes. J. Am. Med. Assoc. 244, 1107 – 1109 (1980). S. Morii and D. B. Burr, Increased intracortical remodeling following fatigue damage. Bone 14, 103 – 109 (1993). L. J. Kornberg, S. H. Earp, C. E. Turner, C. Prockop, and R. L. Juliano, Signal transduction by integrins: Increased protein tyrosine phosphorylation caused by clustering betal integrins. Proc. Natl. Acad. Sci. USA 88, 8392 – 8396 (1991). J. L. Guan, J. E. Trevithick, and R. O. Fibronectin/integrin interaction induces tyrosine phosphorylation of a 120-kDa protein. Cell Regul. 2, 951 – 954 (1991). P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. H. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). M. F. Rouleau, J. Mitchell, and D. Goltzman, In vivo distribution of parathyroid hormone receptors in bone: Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 123, 187 – 191 (1988). M. F. Rouleau, J. Mitchell, and D. Goltzman, Characterization of the major parathyroid hormone target cell in the endosteal metaphysis of rat long bones. J. Bone Miner. Res. 5, 1043 – 1053 (1990). S. J. Silverberg, F. Gartenberg, D. McMahon, and J. P. Bilezikian, Parathyroidectomy improves bone density at cancellous sites in asymptomatic primary hyperparathyroidism. J. Bone Miner. Res. 8, S169 (1993). D. Shinar and G. A. Rodan, Relationships and interactions between bone and bone marrow. In “The Hematopoietic Microenvironment: The Functional and Structural Basis of Blood Cell Development” (M. W. Long and M. S. Wicha, (eds.), pp. 79 – 109. Johns Hopkins Univ. Press, Baltimore, MD, 1993.

CHAPTER 13

Cytokines and Bone Remodeling GREGORY R. MUNDY, BABATUNDE OYAJOBI, KATHY TRAIANEDES, SARAH DALLAS, AND DI CHEN Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284

VIII. Interleukin (IL)-1 and IL-1 Signaling in Osteoclast-like Cells IX. Tumor Necrosis Factor X. Interleukin-6 XI. Vascular Endothelial Growth Factor XII. IL-15, IL-17, and IL-18 XIII. Archidonic Acid Metabolites: Prostaglandins and Leukotrienes XIV. Transforming Growth Factor- XV. Bone Morphogenetic Proteins XVI. Conclusion References

I. Introduction II. Evidence for a Role of Cytokines in Osteoclastic Bone Resorption III. The Osteoclast as a Cell Source of Cytokines Involved in Osteoclastic Resorption IV. The Osteoblast as a Cell Source of Cytokines Involved in Osteoclastic Resorption V. RANK Ligand and Its Signaling Receptor, RANK VI. Osteoprotegerin VII. Macrophage – Colony-Stimulating Factor and Its Signaling Receptor, c-fms

I. INTRODUCTION

recent advances. A list of the cytokines reviewed in this chapter is provided in Table 1. Recent advances in molecular biological techniques have meant that most of the biological activities ascribed to cytokines have now been associated with specific molecules, and their receptors identified and molecularly cloned. Several of these cytokines and their cognate receptors have been shown to be expressed by bone cells, marrow cells, or accessory cells in the bone microenvironment. Moreover, studies using knockout and transgenic mice have increased our understanding of the complex signal transduction mechanisms utilized by cytokines and are opening up new and exciting areas of study. Cytokines tend to be pleiotropic and multifactorial and may have overlapping and seemingly redundant biological effects. Some of this redundancy is

In normal individuals, bone is continuously being remodeled and this is achieved via a finely regulated balance between the processes of bone formation and resorption mediated by osteoblasts and osteoclasts, respectively. This bone remodeling is regulated, in part, by local factors including cytokines generated in the bone microenvironment. The purpose of this chapter is to summarize what is currently known about the role of cytokines and their receptors in bone remodeling. In the past few years, there has been an explosion of information on multiple aspects of the effects of cytokines on bone. This has become an enormous topic, and it will not be possible to cover all aspects in this chapter. Rather, it will focus on important

OSTEOPOROSIS, SECOND EDITION VOLUME 1

373

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

374 TABLE 1 Cytokines Produced in Bone Microenvironment with Major Effects on Osteoclasts and Osteoblasts Osteoclastogenic cytokines RANK ligand Osteoprotegrin Macrophage – colony stimulating factor Interleukin-1 Tumor necrosis factor Interleukin-6 Vascular endothelial growth factor Interleukin-15, -16, and -17 Prostaglandins and leukotrienes Osteoblastogenic cytokines Transforming growth factor- Bone morphogenetic proteins

apparent in the receptor mechanisms and signal transduction pathways used by groups of cytokines. Classic examples that illustrate this vividly are the various cytokines belonging to the interleukin (IL)-6 family, such as IL-6, leukemia inhibitory factor (LIF), oncostatin-M, and IL-11, which utilize a common signal transduction protein known as gp130. These cytokines bind to distinct membrane-associated receptors, which form hetero- or homodimers upon binding to the ligand. These dimers then complex with gp 130, leading to its activation by the phosphorylation of tyrosine residues. This subsequently activates several tyrosine kinase cascades within the cells by a common tyrosine kinase known as JAK2. One of these cascades involves phosphorylation of a transcription factor known as STAT-2. Another involves ras and MAP-2 kinase and leads to phosphorylation of the transcription factor, nuclear factor (NF) – IL-6 [1]. These signal transduction pathways and those used by other cytokines are only now being studied in bone cells, but observations already made in other cells and tissues are holding true for bone with just a few exceptions. The reasons that individual members of cytokine families have seemingly distinct effects on cells involved in bone remodeling remain unclear.

II. EVIDENCE FOR A ROLE OF CYTOKINES IN OSTEOCLASTIC BONE RESORPTION A considerable amount of data has been accumulate since the mild-1970s that indicate that cytokines play a role in both physiological bone remodeling. As mentioned previously, osteoclast formation and activity are regulated by factors that are generated in the bone microenvironment acting in an autocrine, paracrine, or juxtacrine fashion.

GREGORY R. MUNDY ET AL.

These include macrophage – colony-stimulating factor (M-CSF), also known as colony-stimulating factor-1 (CSF1), IL-6, IL-1, IL-11, tumor necrosis factor (TNF)-, TNF, granulocyte – macrophage colony-stimulating factor (GM-CSF), transforming growth factor (TGF)-, TGF-, leukemia inhibitory factor (LIF), and bone morphogenetic proteins (BMPs). For some of these cytokines, the precise cellular source within the bone microenvironment has not been defined, although possibilities include immune cells and bone cells of either osteoblastic or osteoclastic lineages. In addition, biologically active forms of some of these cytokines may be derived from sequestered stores within bone matrix. However, the relative importance of cytokines in the formation of new osteoclasts and the activation of mature osteoclasts in vivo is still unclear. The availability of transgenic animal technology has meant that the role of cytokines in bone metabolism is now increasingly being examined in an in vivo context. However, the functional redundancies among related cytokines or groups of cytokines that share common signaling pathways mean that the direct ablation of individual genes for these factors does not always impact bone remodeling adversely. With the exception of M-CSF, as exemplified by the op/op variant of osteopetrosis, naturally occurring models of total deficiency of any of the factors mentioned previously are rare. In the op/op mouse model of osteopetrosis, a frameshift mutation in the coding region of the csf-1 gene leads to the failure of secretion of biologically active M-CSF by stromal cells, osteoblasts, or other accessory cells. Consequently, mature macrophages do not survive for long and osteoclasts fail to form during the neonatal period, resulting in inadequately remodeled bone. However, osteoclasts do form beyond the neonatal period with sufficient function to reverse the osteopetrosis by 22 weeks, indicating that MCSF is not required for osteoclast formation beyond the first few weeks of life. These data show that in the mouse, secretion of biologically active M-CSF is an absolute requirement for normal osteoclast formation during this early period of life. The impairment in osteoclastic resorption and osteopetrosis can be rescued by the exogenous administration of M-CSF during the neonatal period [2 – 4]. Moreover, studies have also shown that the exogenous administration of vascular endothelial growth factor (VEGF) to neonatal op/op mutant also reverse osteopetrosis, suggesting that this factor may be responsible, in part, for the spontaneous improvement observed in affected op/op mice as they mature [5]. Interestingly, GM-CSF and IL-3, which are other major growth factors for cells of the monocyte/ macrophage lineage, can also partially reverse the osteopetrosis in these mutant mice [6], implying that other factors are essential for normal bone remodeling in this form of osteopetrosis. Consistent with this hypothesis, enforced expression of bcl-2 in cells of the monocyte/macrophage lineage resulting in their prolonged survival partially rescues

375

CHAPTER 13 Cytokines and Bone Remodeling

the osteoclast defect in csf-1 mutant op/op mice [7]. In summary, studies with the mutant M-CSF op/op mouse provide clear evidence that cytokines are involved in physiological bone resorption. There are other lines of evidence from studies using transgenic mice and inhibitors of cytokine activites that indicate that cytokines such as IL-1, IL-6, and TNF are important in disorders of bone remodeling in vivo. However, available data are complex to interpret and there are often conflicting reports. IL-1 and TNF have consistently been shown to play a major role in the rapid bone loss associated with estrogen-depleted states such as postmenopause and after ovariectomy [8], and there is also evidence from studies with mice lacking the IL-1 type 1 receptor that imply that IL-1 may be an important mediator of the effects of ovariectomy on bone mass. Mice deficient in the type I IL-1 receptor (IL-1R1), which is the signaling receptor for both IL-1 and IL-1, do not lose bone after ovariectomy [9]. The soluble p75 TNF receptor blocks the osteoclastogenic effect of TNF [10], and mice engineered to overexpress this soluble TNF receptor do not lose bone after ovariectomy [11]. Regarding IL-6, the increase in bone resorption observed following ovariectomy in mice can be corrected by the administration of neutralizing antibodies to IL-6, as shown by the experiments of Jilka and colleagues [12]. The same abnormality can be reversed by treatment of mice with estrogen. Also, IL-6 knockout mice are protected against bone loss induced by ovariectomy, further implicating IL-6 in bone remodeling in vivo [13]. Surprisingly, transgenic mice overexpressing IL-6 do not have osteopenia as would be predicted [14]. These seemingly discrepant observations are probably related to the fact that in the presence of estrogen deficiency, there is likely increased production of several cytokines by cells in the bone marrow microenvironment leading to increased osteoclastic activity. Whether production of each cytokine is largely by a particular cell type has yet to be defined. There is considerable evidence that IL-6 and/or TNF may synergize with IL-1 to enhance osteoclastic bone resorption. For example, Pacifici and colleagues have suggested that the simultaneous block of IL-1 and TNF may be necessary to completely abrogate the rapid bone loss seen in the early postovariectomy period [15,16]. However, neutralization of either of these factors may nevertheless lead to some decrease in bone resorption in certain situations. There is also evidence that other osteotropic cytokines may be involved in other disease states. Much of this evidence comes from “gain of function” rather than “loss of function” experiments, with evidence that there is increased production of certain cytokines associated with increased bone loss. This is certainly true in myeloma, in some solid tumors, and in chronic inflammatory diseases associated with a local increase in bone loss such as rheumatoid arthritis and periodontal diseases [17]. It is also true in Paget’s

disease where there is increased production of proresorptive cytokines by multinucleated osteoclasts, especially IL-6 [18 – 20]. Because overproduction of these cytokines in these conditions may enhance bone resorption through the stimulation of osteoclast formation and differentiation, pathologic bone lesions associated with a large increase in osteoclasts may be self-perpetuating.

III. THE OSTEOCLAST AS A CELL SOURCE OF CYTOKINES INVOLVED IN OSTEOCLASTIC RESORPTION Abundant evidence indicates that osteoclast formation and activity are regulated by factors generated in the bone cell microenvironment. As mentioned in the preceding section, these factors may be produced by immune cells or cells in the osteoblast lineage or be derived from the bone matrix itself. However, convincing data support the notion that the osteoclast itself may also be a source of autocrine or paracrine factors, which can modulate bone remodeling. The subject of osteoclast as a secretory cell has been reviewed comprehensively elsewhere [21 – 24] (see also Chapter 3). The osteoclast expresses IL-6 in prodigious amounts. Moreover, IL-6, at least in human systems, can stimulate the formation of cells with osteoclast characteristics [18]. Antibodies to IL-6 inhibit bone resorption by isolated human giant cells on calcified matrices, and, similarly, antisense oligonucleotides to IL-6 inhibit the capacity of human giant cells to form resorption pits on sperm whale dentine [25]. Furthermore, it appears that IL-6 may mediate some of the effects of IL-1 and TNF on bone resorption, as an anti-IL-6 neutralizing antibody and a potent IL-6 antagonist that binds to IL6 receptor but does no dimerize with gp130 both blocked IL-1 and TNF-induced osteoclast formation in human marrow cultures [26]. However, IL-6 is not the only cytokine that is produced by isolated osteoclasts. TGF-, interleukin1, annexin-II (lipocortin-II), and human stem cell antigen I [24,27,28] have all been shown to be expressed by osteoclasts, and each of these factors may regulate osteoclasts function. TGF- inhibits osteoclast formation [29,30] and is a powerful stimulator of osteoclast apoptosis [31]. These effects are probably mediated, in part, via paracrine mechanisms involving alterations in the stromal/osteoblastic cell expression of the receptor activator of NF-B ligand (RANKL) and osteoprotegerin (OPG) [32,33]. TGF- may also generate prostaglandins in the microenvironment of osteoclasts [30], which can exert independent effects on osteoclast formation and activity, most probably via modulating RANKL expression [34,35]. IL-1 and annexin-II both stimulate osteoclast formation, whereas human stem cell antigen I inhibits osteoclast formation. The relative importance of all these osteoclast products in the formation of new osteoclasts is not clear. However, one possibility is that

376 as osteoclasts undergo apoptosis within the bone remodeling unit, at the conclusion of the remodeling sequence, some of these cytokines may be released by the dying osteoclast to produce a new generation of osteoclasts derived from their marrow precursors.

IV. THE OSTEOBLAST AS A CELL SOURCE OF CYTOKINES INVOLVED IN OSTEOCLASTIC RESORPTION There is compelling evidence from ex vivo studies that the commitment of osteoclast progenitors (spleen, bone marrow, or peripheral blood derived) to differentiate to multinucleated cells with characteristics of mature osteoclasts requires direct cell – cell contact with osteoblastic or related marrow stromal cells [36]. Furthermore, it has been known for some time that almost all of the known bone-resorbing cytokines, such as IL-1, IL-6 and IL-11, as well as the systemic bone-resorbing hormones, such as parathyroid hormone (PTH), PTHrelated protein (PTHrP), 1,25-dihydroxyvitamin D3, and PGE2, appear to exert their effect only in the presence of stromal/osteoblastic cells [36]. Because these agents activate different signal transduction pathways on osteogenic cells, it was recognized that there is a convergence in their downstream response, and the existence of a membrane-associated factor on the surface of cells of the osteoblastic lineage essential for osteoclast progenitors to proliferate and differentiate was therefore proposed. This factor, which was variously termed “osteoclast differentiation factor” (ODF) and stromal cell-derived osteoclast formation activity (SOFA), was postulated to be inducible by cytokines and hormones known to regulate osteoclast differentiation. Although M-CSF was known to be membrane associated and to be important for osteoclastic bone resorption, recombinant M-CSF alone could not induce osteoclast formation in the absence of stromal/osteoblastic cells. Anderson and colleagues [37] reported the molecular cloning of a novel membrane-bound member of the TNF receptor (TNFR) family from a cDNA library established from human bone marrow-derived myeloid dendritic cells. Simultaneously, they reported the cloning of the mouse orthologue of the receptor from a fetal mouse liver cDNA library. This receptor, which activated (NF-B) activity, was designated receptor activator of NF-B (RANK) and a search for its cognate ligand led to the cloning of RANK ligand (RANKL). RANKL was shown to be identical to TNF--related activation-induced cytokine (TRANCE), a TNF ligand family member cloned from murine thymoma EL40.5 cells and shown to activate c jun-Nterminal kinase [38]. Subsequently, using a novel secreted TNFR homologue known as osteoprotegerin (OPG)/osteoclastogenesis-inhibitory factor (OCIF) as a probe, two groups independently reported the cloning of the same molecule, which they designated OPG ligand (OPGL) and osteoclast

GREGORY R. MUNDY ET AL.

differentiation factor (ODF), respectively [39,40]. As will be discussed later, RANKL/TRANCE/OPGL/ODF have now been shown to be the same molecule whose expression is obligatory for osteoclastic resorption and normal bone modeling and remodeling. As proposed by Suda and others, we will refer to this cytokine hereafter as RANKL [41]. This subject is also discussed in Chapters 3 and 12.

V. RANK LIGAND AND ITS SIGNALING RECEPTOR, RANK RANKL is synthesized as a type II integral membrane protein with its N terminus in the cytoplasm and a C terminus extending extracellularly. Expression of RANKL in a human embryonic kidney fibroblast (293) cell line was reported to generate a membrane-bound as well as a secreted form of RANKL representing the extracellular C-terminal domain [40]. It has also been reported that the extracellular domain of RANKL can be cleaved by a TNF-convertase (TACE)-like enzyme in vitro and that recombinant RANKL can be cleaved by purified TACE [42]. This cleaved form retains some biological activity in osteoclast assay systems [42]. Although it has been reported that T cells shed RANKL on activation in vitro [43], there is as yet no evidence that a soluble form of RANKL exists in vivo or is generated by proteolytic cleavage in the bone microenvironment. Nevertheless, it remains a possibility that the extracellular domain of RANKL is shed by tumor-associated metalloproteinases as has been described for other members of the TNF ligand family, such as TNF and Fas. In this regard, our group identified a factor from a human tumor associated with osteoclastosis and hypercalcemia that appears to be a novel cytokine that stimulates osteoclast formation in the presence of M-CSF [44], and another group also identified another factor from a mouse tumor with similar biological activity [45]. RANKL, in the presence of M-CSF, induces osteoclast formation in all model systems presently available to study osteoclast development. For example, RANKL stimulated the formation of osteoclasts from spleen-derived osteoclast progenitors in the absence of osteoblasts/stromal cells and this was abolished completely by simultaneously adding OPG [46] or a recombinant soluble form of the extracellular domain of RANK fused to the Fc region of human immunoglobulin (RANK.Fc) [47]. In the presence of M-CSF, RANKL also stimulated osteoclast formation in human and murine bone marrow cultures and also in human peripheral blood monocyte cultures [48 – 50], and it induced the formation of TRAP-positive colonies in an agar culture of bone marrow cells [40]. Treatment of stromal/osteoblastic cells of human and murine origins with known stimulators of osteoclast formation, 1,25(OH)2D3, PTH, PGE2, IL-11,

CHAPTER 13 Cytokines and Bone Remodeling

IL-6, IL-1, and TNF, induces or enhances RANKL messenger RNA levels [34,35,40,46,51 – 54]. Treatment of 45Ca-labeled fetal mouse or rat long bones with a recombinant soluble form of RANKL also stimulated the release of 45Ca from bone, which was completely inhibited by simultaneously adding OPG or RANK.Fc [46,47]. Like OPG, polyclonal antibodies against RANKL inhibited bone resorption in organ cultures induced not only by soluble RANKL but also by 1,25(OH)2D3, PTH, PGE2, and IL-1 [51,55] These results clearly indicate that bone resorption induced by these osteotropic factors is mediated by RANKL. In culture systems where there is essentially no continuing osteoclast formation, such as isolated rat osteoclasts, recombinant RANKL has been shown to induce actin rings in the cells rapidly and to increase their bone-resorbing activity markedly, independent of stromal/osteoblastic cells [55,56]. Administration of recombinant OPG to mice injected with saline or with RANKL daily for 7 days led to a rapid disappearance of osteoclasts (as early as 6 h) from bone surfaces with evidence of osteoclast apoptosis [57]. A single parenteral administration of recombinant RANKL increased blood ionized calcium within 1 h in mice [56]. Also, systemic injection of RANKL twice daily for 3 days led to a sustained hypercalcemia, althouth the number of osteoclasts was almost identical to those of untreated mice [40]. Taken together, these data suggest that RANKL stimulates not only osteoclast differentiation but also activates mature osteoclasts as well as prolongs their survival, therby having a direct impact on their function. RANKL knockout mice have been generated that exhibit typical osteopetrosis with total occlusion of bone marrow space within endosteal bone in the early neonatal period. The bones of these RANKL null mutant mice lack osteoclasts, although osteoclast progenitors were present that were shown to differentiate into functionally active osteoclasts when cocultured with normal osteoblasts/stromal cells from wild-type litter mates [58]. Although these results suggest that RANKL is an absolute requirement for osteoclast development, it remains unknown whether there is a spontaneous reversal of the osteopetrotic phenotype with age in RANKL (-/-) mice as seen in adult csf-1 null mutant mice as there are presently no data on older RANKL null mutant mice. It is likely that the increased expression of RANKL may play a role in pathological situations associated with bone destruction such as malignancies. We have reported that direct cell – cell contact between myeloma cells and marrow stroma-derived ST2 cells enhances RANKL expression in the myeloma and also stimulates production of a soluble factor(s) capable of enhancing RANKL expression in marrow stromal cells [59]. We postulated that increased RANKL expression within the bone microenvironment cells may explain the increased osteoclastic activity and destructive bone lesions that are characteristic of multiple myeloma. This is likely to be true for other tumors metastatic to bone such as

377 breast cancer. In this regard, it has also been demonstrated that although RANKL expression was undetectable in either mouse breast cancer cells or bone marrow stromal cells, its expression was elevated markedly in cocultures of both cell types [60]. It has also been reported that substantial numbers of multinucleated cells with osteoclastic characteristic form in coculture of activated human CD4 T helper cells and adherent murine splenic osteoclast precursors in the presence of M-CSF, independent of stromal/osteoblastic cells [61]. This suggests that in chronic inflammatory tissues characterized by CD4 T-cell infiltration such as rheumatoid arthrititic synovium, this mechanism may be responsible for the extensive localized bone destruction. As mentioned earlier, RANKL was originally cloned as a ligand for the receptor, RANK. RANK is a type I transmembrane protein with a C-terminal cytoplasmic tail much longer than that of all known members of the TNFR superfamily. Like other members of the family, RANK has four extracellular cysteine-rich domains. However, unlike most other TNFR family members, RANK messenger RNA is expressed ubiquitously with highest levels in skeletal muscle and thymus and in spleen- and bone-marrow-derived osteoclast precusors [37,63]. To date, RANK has been shown to bind only to RANKL; it does not bind other members of the TNF ligand family, such as lymphotoxin, TNF, Fas ligand, CD27 ligand, CD30 ligand, CD40 ligand, 4-1BB ligand, or TRAIL. It has also been demonstrated conclusively that the formation of mature osteoclasts from osteoclasts precursors as well as activation of mature osteoclasts can only be induced via RANK signaling [55,62 – 64]. As with RANKL knockout mice, RANK null mutant mice also exhibit severe osteopetrotic phenotype with a complete absence of osteoclasts [63]. RANK also activates c-jun N-terminal kinase (JNK) in immune cells such as T cells, dendritic cells, and spleen-derived hematopoietic osteoclast progenitors [55,62]. RANK also activates NF-B and JNK in a macrophage cell line RAW 264.7, which has been shown to differentiate into osteoclast-like cells when treated with RANKL and M-CSF [62]. However, it remains unclear whether RANK also activates the JNK pathway in marrow-derived osteoclast precursors. However, there is no unequivocal evidence to date that the JNK pathway is at all involved in osteoclast formation and activation. Interestingly, overexpression of RANK in human embryonic kidney fibroblast 293 cells induces ligand-independent NF-B and JNK activation, suggesting that pathological conditions associated with RANK expression may result in increased osteoclast formation independent of RANKL. As RANK has no intrinsic kinase activity, it activates NF-B via interactions with the TNF receptor-associated family (TRAFs) of adaptor molecules [65]. Several members of the TRAF family have been implicated in regulating signals from various TNF/TNFR family members. Evidence shows that TRAF2, TRAF5,

378 and TRAF6 interact with the C-terminal 85 amino acid cytoplasmic tail of RANK, and it is likely that the signals through RANK are mediated primarily through these TRAFs [63,66 – 70]. Of these three TRAFs, TRAF6 appears unique in several respects. First it interacts with a novel Cterminal domain of the cytoplasmic tail of RANK distinct from the known binding motifs for TRAF1, TRAF2, TRAF3, and TRAF5, although TRAF6 also associates with a short N-terminal sequence within the cytoplasmic domain [67,68]. Second, overexpression of an N-terminal truncated TRAF6, acting as a dominant negative, inhibited RANKLinduced NF-B activation in the human embryonic kidney 293 cell line. Third, unlike other TRAFs, TRAF6 has also been implicated in IL-1-induced NF-B activation [71]. Finally, whereas other TRAF null mutant mice currently available, such as TRAF2 null mutant mice, have a normal skeletal phenotype, TRAF6 knockout mice exhibit severe osteopetrosis with defective bone remodeling and delayed tooth eruptions [69]. However, unlike in RANKL (-/-) mice, the bones of TRAF6 (-/-) mice had a few osteoclasts, suggesting that there might be some redundancy in TRAF usage in osteoclast development. A few years ago, two groups independently generated mice that were lacking both p50 and p52 subunits of NFb [72,73] and reported that these double knockout (nf-b-1 and nf-B-2) mice developed severe osteopetrosis because of a defect in osteoclast differentiation. There was a complete absence of osteoclasts, although there were osteoclast progenitors and the number of macrophages was increased. Surprisingly, the osteopetrotic phenotype could be rescued by bone marrow transplant from wild-type littermates. Recent developments in our understanding of the molecular mechanisms of the intracellular signal transduction pathways induced when RANKL binds to its cognate receptor RANK have now clarified observations made with double NF-B null mutant mice. Although overexpression of RANK in human embryonic kidney 293 cells stimulated JNK and NF-B, when the C-terminal cytoplasmic tail necessary for TRAF binding was deleted, the truncated RANK receptor was still capable of stimulating JNK activity but not NF-B. This suggests that interaction with TRAFs is critical for NF-B activation but not for the activation of the JNK pathway. Taken together with the complete absence of osteoclasts in NF-B double knockout mice, the fact that the JNK pathway remains intact even in the absence of NF-B activation indicates that it is most unlikely that the JNK pathway is involved in osteoclast development. We have reported that a genetically engineered form of RANK generated by fusing the entire extracellular domain to the Fc region of human lgG1 (RANK.Fc) blocks hypercalcemia induced by PTHrP-secreting human tumor xenografts in nude mice. This further confirms the critical role of RANKL/RANK interaction in osteoclastic bone

GREGORY R. MUNDY ET AL.

resorption [47]. Transgenic mice overexpressing a soluble RANK.Fc fusion protein have severe osteopetrosis because of a marked reduction in osteoclast numbers and a decrease in bone resorption indices [62]. Figure 1 represents a diagram showing the interactions among RANK, RANK ligand, and OPG.

VI. OSTEOPROTEGERIN Our understanding of the biology of bone modeling and remodeling was given an impetus with the discovery of a novel secreted member of the TNFR superfamily. One group isolated a heparin-binding protein from conditioned media of human fibroblast cultures, which profoundly inhibited osteoclast formation. This protein was thus designated “osteoclastogenesis inhibitory factor” (OCIF) [74]. Independently, another group cloned a novel TNFR family member that constitutively lacked a transmembrane domain and was thus secreted. When expressed, the recombinant protein was shown to inhibit both physiological and pathological bone resorption, and hepatic overexpression of the gene in transgenic mice resulted in severe osteopetrosis. The receptor was therefore termed osteoprotegerin [75]. Subsequent molecular cloning of the cDNA coding for OCIF revealed that it was identical to OPG [39]. Other groups also independently cloned the same receptor molecule and the TNF receptorlike molecule 1 (TR1) and follicular dendritic cell-derived receptor I (FDCR-1) [76 – 78] have each been shown to have complete sequence identity to OPG/OCIF. As proposed by Suda et al. [41], we will hereafter refer to the protein (including OCIF, FDCR-1, and TR1) as OPG. Like other members of the TNF receptor family, OPG has four cysteine-rich domains (DI – D4). In addition, there are two homologous death domain regions (D5 and D6) in OPG. Both D5 and D6 share structural features with other death domains previously described in other members of the TNFR family, including the TNF receptor p55, Fas, DR3, and TRAIL receptor. These death domains have been shown to mediate apoptotic signals. Although the precise role of D5 and D6 of OPG is still not known, the death domain-homologous regions are active in mediating apoptotic signals [79]. OPG has only two known ligands, RANKL and TRAIL, both of which are type II membrane-bound TNF homologues [40,80]. In contrast, OPG circulates in vivo in measurable quantities in both human and rodent sera [81,82], and there is now incontrovertible evidence that it acts as a nonsignaling decoy receptor for RANKL and thereby as a regulator of bone turnover [41] (also see Chapter 12). Serum concentrations of OPG increase with age in both men and women and were significantly higher in postmenopausal osteoporotic women compared to age-matched

379

CHAPTER 13 Cytokines and Bone Remodeling

FIGURE 1

Interactions among RANK ligand, its receptor RANK, and the decoy receptor/in-

hibitor OPG.

controls. It was suggested that the increased levels of OPG in the former group reflect a compensatory response to the enhanced bone resorption in the postmenopausal period rather than a cause of the osteoporosis [82]. This is the only study to determine serum OPG levels in humans and it remains to be seen whether these early data are reproduced in different cohorts of patients. The role of OPG in normal bone remodeling has been highlighted in more detail in studies with OPG-deficient mice produced by targeted disruption of the gene [83,84]. OPG (-/-) mice are viable and fertile, but they exhibit profound osteoporosis from birth caused by enhanced osteoclast formation and formation and function as well as prolonged osteoclast survival. Histological analyses showed a destruction of growth plates and a lack of trabeculae, and histomorphometrical analyses revealed an increase in bone resorption indices in long bones of adult OPG null mutant mice. This is accompanied by a marked decrease in the strength and mineral density of their bones. Interestinly, the osteoblast surface area was also increased in OPG-deficient mice. OPG (-/-) mice also develop calcification of the aorta and renal arteries. These results indicate that OPG is a physiological regulator of osteoclast-mediated bone resorption during postnatal bone growth. It also suggests that OPG may play a role in preventing the calcification of larger arteries.

In the presence of M-CSF, RANKL induced osteoclast formation from spleen cells in the absence of osteoblasts/ stromal cells, and this was abolished completely by simultaneously adding OPG. OPG also strongly inhibits osteoclast formation induced by a range of osteotropic agents, including 1,25(OH)2D3, PTH, PGE2, IL-1, and IL-11, in cocultures of osteoblasts/stromal cells and hemapoietic osteoclast progenitors. Interestingly, in contrast to their stimulatory effects on RANKL mRNA expression, PGE2, 1,25(OH)2D3, and dexamethasone strongly inhibit OPG mRNA expression, suggesting that the regulation of OPG levels is also critical for osteoclastogenesis induced by known osteotropic factors [39,40,85–88]. This has led some workers to postulate that downregulation of OPG may be one of the mechanisms involved in glucocorticoid-induced osteoporosis [86]. OPG also directly inhibits the bone-resorbing activity of isolated mature osteoclasts [89]. As mentioned earlier, treatment of 45Ca-radiolabeled fetal mouse long bones with a soluble form of RANKL also stimulated the release of 45Ca from the bone tissues, which was completely inhibited by the simultaneous addition of OPG [46]. This effect of OPG to inhibit bone resorption is due, in part, to its ability to suppress osteoclast survival [90]. In contrast, OPG gene expression and production in marrow stromal/osteoblastic cells are markedly upregulated by TGF-, [32,33], which likely explains the powerful effect of TGF- to inhibit osteoclast formation [30] and enhance

380 osteoclast apoptosis [31]. In vivo, parenteral administration of OPG results in a marked increase in bone mineral density and bone volume associated with a decrease of active osteoclast number in normal and ovariectomized rats [91]. Serum calcium concentration was also decreased rapidly by the parenteral administration of OPG, independent of any changes in urinary calcium excretion, in thyroparathyroidectomized rats whose serum calcium levels were raised acutely by the administration of PTH [92]. This suggests that OPG, in addition to its effect on osteoclastogenesis, also affects the function and/or survival of mature osteoclasts. OPG also decreased serum calcium levels in tumor-bearing nude mice [91,93], suggesting that it has therapeutic potential for the treatment of hypercalcemic conditions, such as those associated with malignancy.

VII. MACROPHAGE – COLONYSTIMULATING FACTOR AND ITS SIGNALING RECEPTOR, c-fms One of the cytokines that has been clearly shown to play an important role in bone resoption is M-CSF. M-CSF on its own does not stimulate osteoclastic bone resorption in organ culture assays. However, it has been known since the mid-1980s that it is capable of stimulating the formation of cells with osteoclast characteristics in long-term human marrow cultures, as well as in murine marrow cultures [94 – 97]. Like IL-1, M-CSF induces the fusion of preosteoclasts [98,99] and prolongs survival of the multinucleated osteoclast-like cells [100]. However, unlike IL-1, M-CSF does not augment their pit-forming capacity when seeded on calcified matrices [55,99 – 101]. M-CSF has also been implicated in the bone disease osteopetrosis [2,3,102]. As discussed earlier, studies on the murine op/op model of osteopetrosis have clearly shown that M-CSF is required for normal osteoclastogenesis and bone remodeling in the mouse, at least up until the late neonatal period. Studies have also suggested a role for M-CSF in adult bone remodeling. Pacifici and colleagues have provided evidence that IL-1 and TNF concentrations are increased in vivo in estrogen-deficiency states. There is also a substantial body of evidence to indicate that the production of both secreted and cell surface forms of M-CSF by bone marrow stromal cells (BMSC) and osteoblastic cells is regulated by osteotropic cytokines, including IL-1 and TNF [103 –106]. Furthermore, the increased ability of BMSC to support osteoclast formation in the estrogen-deficient state is via IL-1 and TNF-mediated stimulation of M-CSF production [107]. Finally, it has been demonstrated that estrogen blocks M-CSF production by BMSC by directly inhibiting its gene expression [108]. M-CSF mediates its effects on osteoclastic bone resorption through a receptor tyrosine kinase, the protooncogene known as c-fms. Presumably, the presence of this receptor

GREGORY R. MUNDY ET AL.

tyrosine kinase on osteoclast precursors is responsible for M-CSF mediating its effects on osteoclast formation. There may be a hierarchy of receptor tyrosine kinases involved in normal and pathological bone resorption. Other stimulators of bone resorption that mediate their effects on osteoclasts and receptor tyrosine kinases include epidermal growth factor (EGF), TGF-, and platelet-derived growth factor (PDGF). The EGF receptor itself is a receptor tyrosine kinase. PDGF mediates its effects on osteoclast formation presumably through a receptor tyrosine kinase. Because activation of these different receptor tyrosine kinases leads to osteoclast formation and osteoclastic bone resorption, they likely play an important role in bone resorption. However, these receptor tyrosine kinases may not be the only tyrosine kinases involved in bone resorption. Mice deficient in expression of the nonreceptor tyrosine kinase c-src also develop osteopetrosis with failure of osteoclastic bone resorption [109]. However, in these mice, the defect differs from that which occurs in mice with op/op osteopetrosis. In c-src-deficient osteopetrotic mice, there is a failure of ruffled border formation and polarization of the osteoclasts [110]. Nevertheless, osteoclastic form normally. It appears that this receptor tyrosine kinase may, among other things, be involved in osteoclast polarization, which is required for normal osteoclastic bone resorption. Thus, there may be a hierarchy of tyrosine kinases involved in normal osteoclastic bone resorption. Interestingly, c-src has been implicated in signaling by M-CSF. Treatment of normal isolated osteoclasts with M-CSF results in increased osteoclast size and cytoplasmic spreading [111 – 113], which is associated with increased src kinase activity [111].

VIII. INTERLEUKIN (IL)-1 AND IL-1 SIGNALING IN OSTEOCLAST-LIKE CELLS IL-1 is the osteotropic cytokine about which the most is known, at least as far as its effects on bone are concerned. Both IL-1 and IL-1 are powerful stimulators of osteoclastic bone resorption in vitro [114 – 116] and in vivo [116 – 118]. The effects of both Il-1 and IL-1 on bone appear to be identical. IL-1 affects cells at all stages in the osteoclast lineage. This has been demonstrated both in vitro and in vivo [119 – 120]. In vitro studies using cultures of human and murine marrow mononuclear cells that contain committed osteoclasts show that IL-1 increases the formation of multinucleated osteoclasts. When IL-1 is infused or injected in vivo, there is increased appearance not only of mature multinucleated cells, but also of marrow mononuclear cells. This result is indicative of not only an increase in more mature cells in the osteoclast lineage, but also in

CHAPTER 13 Cytokines and Bone Remodeling

granulocyte – macrophage colony-forming units representative of early osteoclast precursors (CFU-GM) [121], an effect probably mediated by IL-1-induced IL-6 [122]. The major effects of IL-1 are probably on the earlier steps in the osteoclast lineage. IL-1 exerts, at least part of, its in vitro and in vivo effects by stimulating prostaglandin synthesis [117]. When IL-1 is injected into the subcutaneous tissue overlying the calvariae, there is accumulation of chronic inflammatory cells associated with an increase in osteoclast activity. This increase can be partially inhibited by the administration of indomethacin, a prostaglandin synthesis inhibitor [123]. IL-1 exerts systemic as well as local actions. Infusions of IL-1 by an osmotic minipump show evidence of increased osteoclastic bone resorption at distant sites [116]. This is associated with a progressive increase in blood-ionized calcium concentrations. IL-1 also has complex effects on osteoblasts [124,125]. These probably depend on whether the interleukin is administered intermittently or continuously. If administered continuously, IL-1 inhibits bone formation. In contrast, when administered intermittently, interleukin-1 induces the proliferation of bone cells, which stimulate osteoblast differentiation into mature bone-forming osteoblasts upon withdrawal [117]. This has been demonstrated clearly in vivo, where intermittent injection of IL-1 leads to a transient increase in bone resorption followed by prolonged osteoblastic bone formation and repair of the defect. Continuous or local administration of IL-1 in sufficient doses leads to a progressive increase in extracellular fluid calcium concentrations [116,117]. However, following injections of IL-1, there is an initial transient decrease in whole blood-ionized calcium levels [118]. This decrease is relatively small, but nevertheless extremely reproducible. It seems to be mediated via prostaglandin synthesis, as it can be abrogated by the concomitant administration of indomethacin. IL-1 has been implicated in a number of disease states associated with increased bone destruction, including all diseases where proinflammatory cells infiltrate and accumulate in tissues adjacent to bone. Thus, it has been implicated in the localized osteolysis that occurs in rheumatoid arthritis and in periodontal disease. It has also been shown that IL-1 is produced by solid tumors associated with the hypercalcemia of malignancy [126]. The form of IL-1 produced in these tumors is IL-1, and it is usually produced in conjunction with the parathyroid hormonerelated protein (PTHrP). The effects of IL-1 and PTHrP are synergistic on bone resorption [126]. This likely occurs because they act at different stages in the osteoclast lineage. IL-1 increases the pool of uncommitted osteoclast progenitors or CFU-GM. In contrast, PTHrP has no effect on CFUGM, but expands markedly the pools of cells at later stages in the osteoclast lineage. IL-1 has also been implicated in the localized bone destruction associated with myeloma [127]. Freshly isolated marrow cells derived from patients

381 with myeloma release IL-1 into the media, and the boneresorbing activity present in the conditioned media can be inhibited by neutralizing antibodies to IL-1 or by IL-1 receptor antagonists (IL-1RA). Possibly the most controversial suggestion for a role of IL-1 in disease states is in postmenopausal osteoporosis. Pacifici and co-workers [128] have suggested that mononuclear cells in the estrogen-deficient state release excessive amounts of IL-1. They postulate that this action is responsible for the increase in bone turnover seen following ovariectomy or at menopause [8]. As support for their concepts, they have shown that peripheral blood monocytes in patients with postmenopausal osteoporosis produce excessive amounts of IL-1 and that this increased production can be inhibited by treatment of the patients with estrogen. They have also shown that bone loss associated with ovariectomy in the rat can be reduced by treatment with IL1RA [15,129]. Other workers have also provided data that support the notion that elevated IL-1 levels may play a role in the bone loss associated with estrogen withdrawal. For example, it has been reported that IL-1R1-deficient mice do not lose bone after ovariectomy [9]. As with most of the other osteotropic cytokines, IL-1 probably mediate its effects on  4 differentiation of osteoclast progenitors indirectly, as it does not induce osteoclast formation in the absence of stromal/osteoblastic cells. Indeed, IL-1 has been reported to induce RANKL expression in human and murine osteoblastic cells [130,131]. However, IL-1 can also promote osteoclastic bone resorption by mechanisms independent of stromal/osteoblastic cells. Moreover, it appears that of the known osteotropic cytokines and hormones, only IL-1 and RANKL are capable of acting directly on mature osteoclasts, which is consistent with the demonstration that osteoclasts express the IL-1 receptor type 1 (IL1R1) [132]. In studies using near homogeneous populations of postmitotic osteoclast precursors (prefusion osteoclasts), it has been demonstrated that like RANKL, IL-1 can prolong the survival of osteoclasts in vitro, independent of stromal/osteoblastic cells. Furthermore, IL-1 can also enhance the resorptive capacity of these osteoclasts on calcified matrices in the absence of stromal/osteoblastic cells and independent of RANKL. It is now known that there are similarities in the signaling pathways utilized by RANKL and IL-1, and the specific receptors for both cytokines, RANK and IL1 receptor, share certain intracellular signal transducers. The effect of IL-1 to induce the fusion of preosteoclasts as well as promote the survival of the multinucleated osteoclasts formed is mediated via IL-1 binding to IL-1 type 1 receptors (IL-1R) on preosteoclasts and direct activation of NF-B, independent of RANKL [99,100,132,133]. As already mentioned, binding of either cytokine to its cognate receptor activates NF-B, and in either case, NF-B activation is preceded by the recruitment of TRAF6 [55,69,99]. Indeed the inflammatory response of TRAF6-deficient mice to IL-1

382 challenge is defective [69]. Interestingly, OPG inhibits the activation of NF-B and JNK induced by RANKL, but not by IL-1 [55]. Because no osteoclasts formed at all in bones of RANKL (-/-) mice, it has been proposed that RANKL is obligatory for osteoclast formation and that it is also responsible for maintaining osteoclast activity under physiological conditions. In contrast, although its presence is not obligatory for osteoclastogenesis, by prolonging the life span of osteoclasts and enhancing their resorptive activity, IL-1 may play a major contributory role in the osteoclastic bone destruction associated with chronic inflammatory diseases such as periodontitis.

IX. TUMOR NECROSIS FACTOR TNF and lymphotoxin are classic type II membrane cytokines that stimulate osteoclastic bone resorption both in vitro and in vivo [134,135]. Their effects in vivo have been demonstrated using Chinese hamster ovarian (CHO) cells transfected with the human TNF- gene. Nude mice bearing tumors that express TNF in large amounts develop hypercalcemia and demonstrate increased osteoclastic bone resorption [135]. TNF stimulates cells at all stages in the osteoclast lineage, in much the same way as does IL-1. The effects of TNF on bone-forming cells have been less well studied but are also probably similar to those of IL-1 [125]. TNF has been implicated in hypercalcemia in several human and animal tumors associated with the humoral hypercalcemia of malignancy [136 – 138]. Antibodies to TNF reduce the blood-ionized calcium in these models as well as some of the other paraneoplastic syndromes associated with maligancy, including leukocytosis and cachexia. In these models, TNF is not produced by the tumor cells but rather by the host immune cells, possibly as part of the immune defense mechanism generated by the presence of the tumor [136]. Lymphotoxin (tumor necrosis factor-) has been implicated in the same sorts of diseases of bone as for tumor necrosis factor. Lymphotoxin is produced by stable myeloma cell lines and may be responsible for the localized osteolysis associated with myeloma [139]. When given by injection or infusion, lymphotoxin causes hypercalcemia and increased bone resorption in rodents [139].

X. INTERLEUKIN-6 Interleukin-6 is a multifunctional cytokine that appears to have a number of unique effects in bone. These appear to differ from those of other osteotropic cytokines and include the following.

GREGORY R. MUNDY ET AL.

1. IL-6 is generated in bone in response to osteotropic hormones such as PTH and cytokines such as IL-1 and TNF- [140,141]. It appears that the production of IL-6 in bone is much greater in response to cytokines than to systemic hormones [142]. For example, IL-1 induces approximately 10 times more IL-6 in bone than PTH [142]. There is not a perfect relationship between IL-6 production in bone and bone resorption. Although IL-6 mediates the effects of PTHrP, IL-1, and TNF on osteoclastic bone resorption in murine organ cultures, it remains unclear whether IL-6 mediates the effects of PTH and PTHrP on osteoclast formation in human bone marrow cultures [26]. 2. IL-6 appears to enhance the effects of other cytokines and systemic hormones on bone resorption both in vitro and in vivo [143]. We have found that not only does IL-6 have synergistic effects with interleukin-1 and PTH in organ culture and cell culture systems for assessing bone resorption, it also has synergistic effects on the boneresorbing capacity of PTH in vivo. This has been shown using CHO cells transfected with PTH and with CHO cells transfected with IL-6 [143]. The effects of both agents together are much greater than either agent alone. 3. The effects of IL-6 on bone resorption in vivo alone are modest. We have found that when IL-6 is expressed by CHO cells, there are only modest effects on serum calcium, and bone resorption is not observed unless enormous amounts of circulating IL-6 are present [144]. This is in contrast to other cytokines, such as IL-1, TNF-, and lymphotoxin. 4. It has been suggested that the effects of IL-6 on bone resorption in vivo may be enhanced by the presence of the soluble IL-6 receptor (sIL-6R). Simultaneous treatment with IL-6 and sIL-6R induced the formation of multinucleated cells with features of authentic osteoclasts in cocultures of hematopoietic osteoclast precursors and osteoblastic cells [145]. This finding may be very important, as sIL-6R is present in the circulation of patients with multiple myeloma in increased amounts [146], and it may explain enhanced effects of IL-6 on bone in this condition. 5. The source of IL-6 in bone has not been definitively clarified. IL-6 may be present in bone matrix, but it is produced in prodigious amounts by osteoclasts as well as stromal cells and osteoblasts. Whether the most important source of IL-6 is osteoclasts or cells in the osteoblast lineage remains to be determined, as many more osteoblasts than osteoclasts are present in bone preparations. 6. IL-6 receptors are present on CFU-GM colonies, which are presumed to be the earliest stage in the macrophage/osteoclast lineage, and IL-6 stimulates the formation of early osteoclast precursors from these CFUGM colonies. However, it is now clear that IL-6 induction of osteoclast differentiation is dependent solely on IL-6 receptors expressed on cells of the osteogenic lineage and

383

CHAPTER 13 Cytokines and Bone Remodeling

not on osteoclast progenitors [147]. This effect of IL-6 is probably due to its ability to enhance RANKL expression by stromal cells [130]. IL-6 has been implicated in a number of disease states. Jilka and colleagues [12] have suggested that excess production of IL-6 may account in large part for the bone loss associated with ovariectomy and estrogen withdrawal. These workers have shown that neutralizing antibodies to IL-6 reduces osteoclastic resorption associated with ovariectomy in mice. They propose that IL-6 production by stromal cells and cells in the osteoblast lineage is enhanced in the presence of estrogen deficiency. There were no significant differences in osteoclast numbers between IL-6-deficient and wild-type mice. Furthermore, ovariectomy did not induce any change in osteoclast number in IL-6-deficient mice compared to wild-type mice [13]. These data suggest that IL-6 may play a more important role in osteoclast development in pathological conditions such as estrogen-depleted states rather than in normal development. This is further emphasized by observations of morphologically normal osteoclasts in bones of gp 130-deficient mice, at least in the early neonatal period [148]. IL-6 has been implicated in the bone loss associated with myeloma. It is clear that IL-6 is expressed by some malignant plasma cells in multiple myeloma, although the actual amounts of IL-6 expressed in myeloma have been conflicting and controversial [149]. Some groups believe that the major source of IL-6 expression is not myeloma cells but rather stromal cells in the bone marrow [150]. Whether IL-6 is an autocrine or paracrine growth factor in myeloma remains controversial. Because IL-6 has such powerful effects to potentiate the bone-resorbing actions of factors such as PTH and PTHrP, its excess production may be important in some disease states where there is overproduction of factors such as PTH or PTHrP, which could lead to the local production of IL-6 in bone. This may be true in some patients with severe primary hyperparathyroidism and secondary hyperparathyroidism and in some malignancies. In each of these conditions, there is excess production of peptides, which induce IL-6 production in osteoblasts, and experimentally there is good evidence to believe that IL-6 production in bone may enhance the bone-resorbing effects of other factors such as PTH or PTHrP on murine osteoclasts. In a search for naturally occurring inhibitors of IL-6, conditioned media harvested from human and murine immune cells were examined. It was found that the monocyte – macrophage cell lines U937 and P388DI produce a biological activity, which impaired the proliferative effects of IL-6 on bone [151]. These factors were purified to homogeneity and it was found that they could be ascribed to a factor in the TGF- superfamily, activin A. Activin A has

previously been shown to be present in considerable amounts in the bone matrix and therefore may act as a stored, endogenous inhibitor of IL-6.

XI. VASCULAR ENDOTHELIAL GROWTH FACTOR VEGF has been implicated in hypertrophic cartilage remodeling, endochondral ossification, and angiogenesis [152]. It appears that VEGF-mediated capillary invasion is an essential signal regulating growth plate morphogenesis and triggering cartilage remodeling. Interestingly, it has been shown that, as with M-CSF, a single injection of recombinant VEGF can induce osteoclast recruitment and survival in the neonatal period in osteopetrotic (op/op) mice [5]. Also, recombinant VEGF can substitute for M-CSF in the formation of osteoclast-like cells in vitro in the presence of RANKL [153]. Although both cytokines are not related, these data suggest that M-CSF and VEGF have overlapping functions with regards to osteoclastic bone resorption. However, a clear role for VEGF in adult bone remodeling remains to be demonstrated.

XII. IL-15, IL-17, AND IL-18 IL-15 is an IL-2-like cytokine produced almost exclusively by T cells and which binds the same receptor as IL-2. IL-15 has been reported to stimulate the formation of TRAP-positive, calcitonin receptor-positive multinucleated osteoclast-like cells in rat bone marrow cultures that resorb calcified matrices [154]. Although IL-15 is a potent inducer of TNF, this effect to stimulate the formation of osteoclastlike cells is not blocked by a specific anti-TNF-neutralizing antibody. Although IL-15 and IL-2 also share some receptor components, IL-2 does not stimulate the formation of osteoclast-like cells. IL-15 levels in synovial fluids of rheumatoid arthritis patients are markedly elevated [155], raising the possibility that this cytokine may play a role in the local destruction of bone associated with chronic inflammatory disease. IL-17 is also a product of activated T cells that induces the production of prostaglandin E2 and IL-6 by bone marrow stromal cells and has been demonstrated to stimulate osteoclast-like cells in vitro via a PGE2-dependent mechanism [156]. Furthermore, although it had no effect on either basal or IL-1-induced bone resorption in bone organ cultures, IL-17 markedly enhanced TNF--induced osteoclastic bone resorption in fetal mouse long bones [157]. It was proposed that this cytokine also plays a role in the bone destruction associated with rheumatoid arthritis. The levels of IL-17 are also markedly elevated in rheumatoid arthritis synovial fluids compared to osteoarthritis synovial fluids of

384 from normal controls. More recent studies have shown that the effect of IL-17 to stimulate osteoclast formation is via cyclooxygenase-2-dependent PGE2 production, which in turn stimulates RANKL expression in stromal/osteoblastic cells [158]. IL-18 is a proinflammatory cytokine originally described as a product of activated macrophages but since demonstrated to be produced by marrow stromal/osteoblastic cells [157,159]. Unlike IL-15 and IL-17, IL-18 inhibits osteoclast formation in cocultures of murine spleen cells and osteoblasts, an effect likely mediated via T-cell-produced GM-CSF, as neutralizing antibodies to GM-CSF abolished osteoclast formation [159]. IL-18, which is homologous to IL-1, signals by binding to IL-1R-related protein I (IL1RrP-1), which is in turn highly homologous to IL-1R. Both IL-1R and IL-1RrP-1 associate with IL-1R-associated kinase (IRAK) and both recruit TRAF6 and both activate NF-B [120]. It remains to be seen how the two seemingly divergent downstream responses can be generated by a near identical signaling cascade.

XIII. ARACHIDONIC ACID METABOLITES: PROSTAGLANDINS AND LEUKOTRIENES Cellular activation by a variety of different stimuli results in the remodeling of membrane phospholipids to generate biologically active lipid mediators that can function intracellularly and extracellularly. A number of distinct classes of lipid mediators, or eicosanoids, are derived from a common precursor, arachidonic acid (AA). Eicosanoids, which include the prostaglandins (PGs), leukotrienes (LTs), lipoxins, and epoxides [160,161] (Fig. 2), are implicated in a variety of pathological and physiological processes, and individual members can exert opposing responses either to stimulate or to inhibit inflammation [161]. For example, thromboxane and prostacyclin have opposing functions in hemodynamics; and leukotrienes are proinflammatory, whereas the lipoxins act as endogenous antiinflammatory eicosanoids. In general, activation of phospholipase A2 results in the hydrolysis of membrane phospholipids and the subsequent release of AA. AA metabolites are then synthesized by the action of specific enzymes on AA. PGs and LTs are synthesized locally and function as autocrine, paracrine, and perhaps intracrine mediators to elicit signals in response to ligand binding and initiate responses in the immediate environment. Both cyclooxygenases and lipoxygenases translocate from the cytosolic compartment to the nuclear membrane or to the nucleus [162,163]. Translocation to the nucleus or nuclear membrane suggests that these enzymes produce eicosanoids that are active within the nucleus, as well as producing secreted eicosanoids capable of paracrine

GREGORY R. MUNDY ET AL.

FIGURE 2 Arachidonic acid metabolic pathway: 5-LO, 5-lipoxygenase; FLAP, 5-lipoxygenase-activating protein; leukotriene (LT) A4, B4, C4, D4, E4; 5-HPETE, 5-hydroperoxytetraenoic acid; 5-HETE, 5-hydroxytetraenoic acid; and COX-1 and -2, cyclooxygenase 1 and 2.

stimulation. This suggestion has merit, as metabolites of both cyclooxygenase and 5-lipoxygenase pathways show binding and activation of nuclear receptors. PG-J2, a metabolite of PG-D2, activates the peroxisome proliferatoractivated receptor- (PPAR-), a transcription factor that regulates adipocyte formation [164]. PPAR- is also a target for thiazolidinediones [165]. Leukotriene B4 (LTB4) binds PPAR-, a transcription factor that regulates the expression of enzymes involved in the oxidation of fatty acids [164 – 166]. PPAR- has been identified in osteoblasts and is associated with a potential switch to the adipocyte phenotype in these cells [167,168]. A possible role for PPAR- in osteoblasts has not been clearly established. Both types of metabolites also bind to classic seven-transmembrane Gcoupled cell surface receptors. For example, PGE2 binds to four isoforms of the PGE receptor, EP1 – EP4 [169], and their presence in osteoblasts has been demonstrated previously [170]. In bone, EP4 and EP2 appear to be critical for the anabolic effects of PGE2 [171], as well as indirect stimulation of osteoclastogenesis [172] and direct inhibition of osteoclast function [173]. Cell surface receptors for leukotriene LTB4 [174] and cysteinyl leukotrienes LTC4, D4, and E4 have also been identified [175]. The pathway most studied with respect to bone is the cyclooxygenase (COX) pathway. Two isoforms of cyclooxygenase have been identified: prostaglandin synthase1 (PGS-1, COX-1) and prostaglandin synthase-2 (PGS-2, COX-2) [176]. COX-1 is expressed constitutively and is primarily responsible for maintaining prostaglandin-mediated physiological functions. COX-2 responds to various stimuli producing prostaglandins involved in inflammation and growth regulation [177,178]. Prostaglandins have many diverse functions in humans, including blood clotting,

CHAPTER 13 Cytokines and Bone Remodeling

nerve growth, would healing, kidney function, and blood vessel tone [179]. Also, the stimulation of COX-2 can be inhibited by glucocorticoids, whereas COX-1 cannot be inhibited by glucocorticoids [180,181]. This differential response to glucocorticoids has been the impetus to develop nonsteroidal anti-inflammatory drugs that selectively inhibit the proinflammatory COX-2 enzyme [182]. Severe bone loss is a major adverse effect of glucocorticoid therapy (see Chapter 44) [183] and may be due, in part, to the combined inhibition of PG synthesis as well as stimulation of LT synthesis (discussed later). The role of the PGs in bone metabolism have been studied extensively for two decades [179,184,185]. Prostaglandins of the E series stimulate bone resorption in vitro [186] but they also stimulate bone formation in vitro and in vivo [187 – 189]. in vivo studies using rodents [190 – 192], dogs [193], and humans [194] have demonstrated the anabolic effects of PGE2. PGE2 has been shown to stimulate bone formation, improve bone structure, and increase bone mass and strength [191,195]. The bone resorptive effects of PGE2 may be critical for the initial phase of bone remodeling. A significant increase in total and osteoid-covered eroded surfaces was observed in cancellous bone sites after 5 days of PGE2 treatment in rats [196]. Although the eroded surface was not elevated in the tibial shaft at the fifth day of treatment, the eroded surface covered with osteoid was increased on the endocortical surface, which the authors concluded was due to PGE2-stimulated bone resorption on this surface prior to day 5. The osteoid perimeter was increased with PGE2 treatment as early as 5 days, and bone formation indices were increased after 10 days of treatment. This study indicates that resorption and formation were rapid with a tapering off of the resorptive phase of the response. There appears to be a dose-dependent response to PGE2 by osteoclasts; at lower doses PGE2 stimulates bone formation, whereas at higher doses PGE2 leads to bone resorption, possibly without diminishing bone formation. Because prostaglandins and leukotrienes are known inflammatory mediators within the body, the bone remodeling unit could be considered an organ-specific inflammatory response to certain proinflammatory stimuli (cytokines, bone defects, micro cracks, stress, exposure of matrix components). Inflammation is a normal response during healing, but the inflammatory effect and responses are cell-type specific. The response is extremely localized, in which the osteoclast is responsible for removing defective bone and the osteoblast is necessary for the healing phase of the inflammatory response. During the resorptive phase, the production of PGE2 is elevated. This is followed by the bone formation, or healing phase, which most likely correlates with a reduction in PGE2 levels. This would suggest that bone formation proceeds without further effects on resorption. Under pathological conditions, constant stimulus and production

385 of inflammatory mediators may lead to progressive bone loss. To this end, it is not surprising to find that cytokines can stimulate bone cells to proliferate, differentiate, and undergo apoptosis in a similar manner to other cell types. Although the in vivo anabolic effects of PGE2 have been well established, the tissue nonspecificity of the metabolite makes it an unlikely therapeutic agent for bone loss. Side effects such as severe diarrhea, hair loss, and decreased physical activity accompanied systemic or oral treatment in animals. The leukotrienes and peptidoleukotrienes are 5-lipoxygenase (5-LO) metabolites of arachidonic acid that appear to have unique effects on bone. Previous work examining the role of leukotrienes in bone metabolism has mainly focused on their effects on osteoclasts. In vitro and in vivo evidence shows that leukotriene 5-LO metabolites (namely 5-HETE and the cysteinyl-leukotrienes LTC4, LTD4, and LTE4) stimulate the formation and activity of avian osteoclasts in vitro [197]. Studies show that leukotriene B4 (LTB4) stimulates bone resorption both in vitro and in vivo [198,199]. When LTB4 was injected over the calvaria of mice, there was a significant increase in osteoclast numbers per unit surface area of bone [198]. Meghji and co-workers [200] had also found LTB4 to be a more potent activator of bone resorption in the mouse calvarial assay. These studies indicated that 5-LO metabolites stimulate the recruitment, formation, and activation of osteoclasts. However, these effects may be indirect, as mature mouse osteoclasts and human giant cells do not express mRNA for either the 5-LO enzyme or the LTB4 receptor. The response of avian osteoclasts may be species specific or the effects of the leukotrienes occur in a precursor population. Leukotrienes and lipoxins may be important in hematopoiesis by regulating the production of committed progenitors: CFU-GM and BFU-E [201 – 204]. It may be that fully differentiated, mature osteoclasts lose the ability to synthesize leukotrienes or do not express cell surface receptors. Few studies have directly addressed the effects of 5-LO metabolites on osteoblast function. Previous reports have shown that LTB4 inhibited the proliferation of normal osteoblastic rat calvarial cells, as well as the osteoblastic cell lines SaOS-2 and G292, and increased intracellular calcium release in osteoblasts derived from neonatal mice calvaria [205]. Data suggest that mice lacking the functional gene for 5-LO have increased cortical bone thickness compared to wild-type mice [206] and have significantly different mechanical properties of bone compared to wild-type animals [207]. These observations suggest that increased bone formation may occur in the absence of leukotriene synthesis. More recently, the capacity of osteoblasts to differentiate and to form bone was inhibited in both the bone nodule formation assay and the calvarial organ culture assay in the presence of 5-LO metabolites [208]. The exogenous addition of 5-hydroxyeicosatetraenoic acid (5-HETE) and LTB4

386 showed a dose-dependent decrease in alkaline phosphatase activity consistent with the inhibition of osteoblast differentiation and inhibition of bone nodule formation in fetal rat calvarial cultures [208]. These studies all support the hypothesis that 5-LO metabolites are negative regulators of bone formation. Leukotrienes and cysteinyl-leukotrienes have been implicated in a number of chronic inflammatory conditions, such as rheumatoid arthritis (RA), asthma, psoriasis, periodontal disease, and inflammatory bowel disease [209]. 5-LO metabolites may be responsible for decreased osteoblast function or decreased bone formation in conditions of elevated 5-LO metabolite production, such as the acutephase inflammatory response and rheumatoid arthritis. Osteoblasts express mRNA for all enzymes in the leukotriene pathway necessary for leukotriene synthesis. LTB4 receptor mRNA was also expressed by these cells, potentially indicating an autocrine/paracrine role for these metabolites in osteoblasts. In contrast to this, because mammalian osteoclasts (mouse and human giant cells) [210] do not express LTB4 receptor mRNA, the regulation by leukotrienes may be either at the precursor level or indirectly through the osteoblast. Leukotriene synthesis inhibitors or receptor antagonists have been used successfully to treat asthma [211,212]. These compounds may have therapeutic potential with regards to bone formation. Inhibition of leukotriene synthesis with the use of 5-LO enzyme inhibitors or leukotriene receptor antagonists resulted in increased bone-like nodule formation, which was negated by the coaddition of indomethacin, indicating cross talk between the 5-LO and COX pathways. The possible interaction between cyclooxygenase and lipoxygenase pathways was demonstrated in the HT29 cl.19A enterocyte cell line, which has 5-LO metabolism without 5-LO activating protein (FLAP) [213]. FLAP cDNA-transfected clones resulted in increased COX-2 expression and PGE2 synthesis. If 5-LO metabolites regulate bone formation negatively, and in the complete absence of 5-LO expression and hence leukotriene synthesis, the basal prostaglandin levels (COX-1 and 2) may be sufficient to maintain bone formation, which eventually leads to the accumulation of bone in these animals (Fig. 3). Glucocorticoids are the mainstay treatment of many inflammatory conditions, including arthritis and asthma. The net in vivo effect of glucocorticoids on bone is to inhibit osteoblast differentiation and function [214]. They activate osteoclasts, inhibit intestinal calcium absorption, and suppress the gonadal axis. At supraphysiological concentrations in vivo, as observed in patients with hypercortisolism (Cushing’s syndrome) [215] and steroid-induced osteoporosis [216], there is an increased bone turnover rate, resulting from increased bone resorption and decreased bone formation. A fine balance exists between soluble mediators released by activated cells of the immune system (cytokines) and

GREGORY R. MUNDY ET AL.

FIGURE 3

Potential shunting between 5-lipoxygenase and cyclooxygenase pathways. Leukotrienes negatively regulate bone formation mediated by prostaglandins. Physiological levels of leukotrienes and prostaglandins work in concert to regulate bone formation. Inhibition of LT synthesis or blockade of their receptors may lead to increased bone formation or a reduction in bone loss associated with inflammatory conditions. Basal cyclooxygenase expression is controlled by COX1, which is not affected by glucocorticoids. COX2, the inducible enzyme, is inhibited by glucocorticoids. Low levels of PGE2 are necessary and stimulate bone formation in vivo. 5-HETE and the leukotrienes stimulate bone resorption and inhibit formation.

products released by the neuroendocrine system in response to these inflammatory mediators. For the most part, homeostasis is maintained under conditions of “stress”, however, in conditions of chronic inflammation, such as rheumatoid arthritis, there remains an imbalance [217]. In general, inflammatory cytokines, such as IL-1, IL-6, and TNF-, stimulate the production of corticotrophin-releasing hormone and arginine vasopressin from the hypothalamus. This in turn stimulates the release of ACTH from the pituitary, followed by glucocorticoid secretion by the adrenal cortex and indirect effects on gonadal function. The hormone products of the hypothalamic – pituitary – adrenal and the hypothalamic – pituitary – gonadal axes are capable of modulating cytokine production [217]. Glucocorticoids are the most potent endogenous inhibitors of immune and inflammatory processes, including the production of inflammatory cytokines [217]. In newly diagnosed, untreated RA patients, elevated levels of IL-6 stimulate the production of CRH, ACTH, and cortisol [218]. However, although overall hypothalamic – pituitary – adrenal axis activity appears to be normal, it may be insufficient to inhibit the chronic inflammatory condition. IL-6 levels were increased significantly in RA patients compared to controls in the early morning hours. In the face of elevated cytokines, perhaps due to glucocorticoid insufficiency or resistance of target tissues [219], these cytokines are able to stimulate osteoclast formation and activation. This effect appears to be mediated through the IL-6 receptor present in osteoblasts [147]. In coculture experiments, dexamethasone treatment increased IL-6 receptor mRNA in osteoblasts. Although the effects of cytokines on bone have been well documented [119 – 222], only a few groups have demonstrated their effects on the 5-LO pathway in osteoblasts. IL-1 has been shown to stimulate 5-hydroxyeicosatetraenoic

387

CHAPTER 13 Cytokines and Bone Remodeling

acid (5-HETE) and PGE2 production in G292 osteoblasticlike cells [223]. The same investigators later demonstrated dose-dependent effects on the cyclooxygenase and lipoxygenase pathways by TNF- [224]. High-dose TNF- (108M) stimulated PGE2 production, whereas low-dose TNF- (1010M) stimulated 15-HETE production. An intermediate dose resulted in 5, 12, and 15-HETE and leukotriene B4 (LTB4) production. In other cell types, IL-1 and IL-6 stimulate the production of 5-HETE [225]. Although glucocorticoids have been used effectively for the symptomatic treatment of RA, their long-term consequence is osteopenia. The anti-inflammatory use of glucocorticoid has been mainly as a replacement therapy in regimens that vary from low-dose oral administration to high-dose pulse therapy and intra-articular injection, which can modulate the activity of activated T lymphocytes and macrophages [226]. The most noted effect of glucocorticoid administration is related to the inhibitory effect on the inducible cyclooxygenase enzyme, COX2. O’Banion et al. [227] demonstrated inhibition of COX-2 by glucocorticoids. Messenger RNA for COX-2 was completely blocked by dexamethasone treatment in murine fibroblasts and IL-1-stimulated human monocytes. Similarly, dexamethasone inhibited COX-2 stimulation and PGE2 production by IL-1 and TNF- [228] and PTH [229] in human osteoblasts. However, differential regulation of COX-2 and 5-LO by dexamethasone has been reported. Goppelt-Struebe et al. [230] showed increased 5-LO activating protein (FLAP) and decreased COX-2 expression in the human monocytic leukemia cell line THP-1. In a longer, conditioning experiment, Riddick et al. [231] showed that 5-LO and FLAP mRNA were increased by dexamethasone in these cells. These experiments suggest a shunting of the arachidonic acid cascade from the cyclooxygenase pathway to the 5-LO pathway (Fig. 3). These data indicate that metabolites of the 5-LO pathway are negative regulators of bone formation. The coordinate regulation of COX and LOX pathways may be required to maintain bone metabolism, and it is the imbalance between the pathways that may account for bone loss. The continued presence of these metabolites in the bone environment might account, in part, for the bone loss associated with chronic inflammatory conditions.

XIV. TRANSFORMING GROWTH FACTOR- Transforming growth factor- has potent effects on the activity of both osteoblasts and osteoclasts and is therefore another important modulator of bone remodeling (see also Chapter 14). Large amounts of TGF- are stored in bone matrix in a latent form, making bone the most abundant source of TGF- in the body [232,233]. TGF-1 is the predominant isoform, making up 80 – 90% of matrix-stored

TGF-, with smaller amounts of TGF-2 and -3 [234 – 237]. Bone matrix-bound TGF- can be released and activated by resorbing osteoclasts [238 – 239]. Results from in vitro and in vivo studies on the effects of TGF- in bone have been somewhat controversial. However, a general consensus is that active TGF- has stimulatory effects on bone formation and inhibitory effects on bone resorption. Thus TGF- has been viewed as a coupling factor, which links bone resorption to subsequent bone formation.

A. Effects of TGF- in Bone In vivo studies with TGF- have shown that, unlike bone morphogenetic proteins, TGF- is not able to stimulate new bone formation in ectopic sites. However, TGF-1 does stimulate new bone formation when injected in close proximity to bone. TGF-1 injected over the calvaria of mice induces new bone formation [240 – 243], as does TGF-1 injected into skull defects [244] and TGF-1 and -2 injected systemically [245 – 247]. The overall effect of TGF- to enhance bone formation appears to be due to TGF- exerting effects at multiple stages in the osteoblast life cycle. Thus, TGF- is a potent chemotactant, which attracts osteoblast precursors to the resorption defect [248]. TGF- also appears to be mitogenic for these osteoblast precursors [249], stimulating them to proliferate at the site of new bone formation. The effect of TGF- on the mature osteoblast may then be to stimulate the production of matrix proteins, such as type I collagen, leading to the production of osteoid. However, in order for mineralization to proceed, it appears that TGF- must be withdrawn. Thus, when primary cultures of fetal rat calvarial osteoblasts or mineralizing bone organ cultures are treated continuously with TGF-, mineralization is inhibited [250 – 252]. Similarly, in animal models of TGF--induced bone formation, TGF- injections must be given and then bone formation allowed to continue in the absence of further treatment. If TGF- injections continue, bone formation will be inhibited. Interestingly, in transgenic mice, which chronically overexpress TGF-2 driven by the osteocalcin promoter, a similar inhibition of mineralization is seen, producing an osteopenic phenotype [253]. The effects of TGF- on bone resorption remain controversial. Following local injection over the calvaria in mice, TGF- does not stimulate bone resorption locally on the periosteal side of the bone, as seen with interleukin-1 or PTHrP. However, it does cause an increase in osteoclast number in the marrow spaces [241]. Its effects in in vitro models of bone resportion appear to be system specific. In cultured fetal rat long bone and in human and murine bone marrow cultures, TGF- inhibits osteoclast formation and osteoclastic bone resorption [29,30]. In contrast, in neonatal mouse calvariae, it stimulates osteoclastic bone resorption, probably through the stimulation of prostaglandin production [254].

388

GREGORY R. MUNDY ET AL.

An exciting recent finding is that in addition to inhibiting osteoclast formation and activity, TGF- may inhibit bone resorption by stimulating osteoclast apoptosis. Furthermore, TGF- appears to mediate the stimulation of osteoclast apoptosis by estrogen, suggesting that it may play a role in postmenopausal osteoporosis [31].

B. Targeted Disruption of TGF- Genes Reveals Isoform-Specific Roles in the Skeleton The generation of mice lacking the genes for TGF-1, 2, and 3 has provided valuable insights into the role of TGFs in bone [255]. Each of these gene knockouts has specific, but nonoverlapping, bone phenotypes, suggesting a specific role for each isoform. Mice lacking the TGF-1 gene, which are born to heterozygous mothers, die shortly after weaning due to inflammatory disease [256 – 257]. Although no gross abnormalities are observed in the bones of these animals, they do show decreased bone mass, reduced bone length, decreased bone elasticity, and are smaller than their normal siblings [258]. The potential contribution of TGF-1 to bone loss in postmenopausal or aging animals has not yet been assessed using this model due to the difficulty in maintaining mice to adulthood. Additional roles for TGF-1 may yet be revealed following various pathological challenges in these animal models. Mice that lack the gene for the TGF-2 isoform show a wide range of developmental defects, including cranial defects, cardiac, lung, and urogenital defects [259]. Mice lacking the gene for TGF-3 show failure of the palatial shelves to fuse, resulting in cleft palate. These mice also show abnormal lung development and die soon after birth [260]. The TGF-1 knockout model has been complicated by the observation that TGF- can pass from mother to fetus both across the placenta and in the mother’s milk [261]. Thus, in TGF-1 knockouts born to heterozygous mothers, this “maternal rescue” phenomenon may mean that the mice are not totally lacking TGF-1, but should be viewed more correctly as TGF-1 deficient. To eliminate the effects of multifocal inflammation, TGF- null deletions were created in severe combined immunodeficient (SCID) mice [262]. This enabled some of the knockout mice to survive to adulthood. These mice are 50 – 80% the size of their normal littermates, show a lack of vigor, and do not thrive, however, these investigators were able to breed a TGF-1 null female with a heterozygous male. The two TGF-1 null pups born to this TGF-1 null female were viable, suggesting that lack of TGF-1 is not lethal in the embryo.

C. TGF- Gene Polymorphisms Various polymorphisms of the TGF-1 gene have been associated with low bone mass and fracture risk in osteoporotic

women [263 – 265]. Some of these polymorphisms have been shown to result in reduced circulating TGF-1 concentrations. This suggests that TGF-1 allelic variants may be important determinants of bone mass and that analysis of the TGF-1 genotype may prove useful in the prevention and management of osteoporosis. In support of these genetic findings, animal studies have shown that osteopenia in old male mice was due to reduced TGF- content of the bone matrix and reduced TGF- responsiveness of marrow osteoprogenitors [266].

D. TGF- Receptors TGF-s signal through serine/threonine kinase receptors, which are composed of type I and type II subunits [267 – 269]. TGF- binds to the type II receptor, which then recruits the type I receptor to the complex. The type II receptor then phosphorylates the type I receptor, which initiates the signaling cascade through phosphorylation of a family of downstream signaling molecules known as Smads. TGF- signals through Smads 2 and 3, which then associate with Smad 4. Smad 4 translocates the complex to the nucleus, where the complex, in conjunction with other DNA-binding proteins, initiates the transcription of TGF-regulated genes. In addition, these activated Smads can be inhibited by the negative regulators, Smad 6 and Smad 7. Bone cells express both type I and type II receptors and are capable of transducing TGF- signals via the Smad pathway [235,270,271]. Experiments using transgenic mice have emphasized the importance of TGF- receptor signaling in skeletal tissue. Expression of a dominant-negative type II TGF- receptor in osteoblasts under control of the osteocalcin promoter resulted in increased trabecular bone in transgenic mice [272]. Expression of the same dominant-negative TGF- type II receptor in mouse perichondrium/periosteum, synovium, and articular cartilage, under control of a metallothionine-like promoter, resulted in progressive skeletal degeneration resembling osteoarthritis, suggesting that TGF- may be important for the normal maintenance of synovial joints [273]. Bone cells also express betaglycan, which is known as the type III TGF- receptor. This membrane-associated proteoglycan binds TGF- with high affinity but does not appear to signal directly. The type III receptor is thought to function as a cell surface reservoir for TGF-, which presents TGF- to the type II receptor [274].

E. TGF- Latency TGF- is produced by most cells, including bone cells, as one or more latent complexes that must be activated in order for TGF- to exert its biological effects (Fig. 4).

389

CHAPTER 13 Cytokines and Bone Remodeling

The forms of TGF- found in the bone microenvironment and the potential mechanisms for the regulation of TGF- activity by bone cells. TGF- is made by osteoblasts in two latent forms: a small latent complex (100 kDa) that lacks the 190-kDa latent TGF--binding protein (LTBP-1) and a large latent complex that contains LTBP-1. LTBP-1 appears to target the latent complex to the bone matrix for storage in a network of fibrillar structures. Distinct from this function associated with latent TGF-, LTBP-1 may also function as an extracellular matrix protein that influences the structure of the bone matrix. Latent TGF- can be activated by numerous mechanisms. One well-known cellular mechanism is the activation by resorbing osteoclasts. Another mechanism is through the action of proteases. Proteases, such as plasmin, can easily activate the small latent complex through proteolytic digestion of the precursor portion. However, in the large latent complex, the hinge region of LTBP-1 is highly susceptible to cleavage by proteases, thereby releasing the large latent complex containing a truncated form of LTBP-1. The target for this form of latent TGF- is being investigated.

FIGURE 4

Because virtually all cells express surface receptors for TGF-, activation of latent TGF- may be key point for the regulation of TGF- activity. Bone cells produce at least three forms of latent TGF-, including small (100 Da) and large (290 Da) latent TGF-1 and TGF-2 complexes [275,276]. The 100-Da small latent TGF- complex consists of the mature 25-kDa TGF- homodimer, which is cleaved from but remains noncovalently associated with a 75-kDa portion of the propeptide homodimer. The large (290-kDa) complex is identical, except that one of the propeptide chains is disulfide linked to a third protein called the latent transforming growth factor -binding protein-1 (LTBP1). This protein, originally identified as a component of the large latent TGF- complex [277,278], is clearly a matrix protein, which colocalizes with fibrillin-1 in microfibrillar structures in bone matrix [279,280]. LTBP1 is thought to play an important role in the regulation of TGF- at multiple levels. Thus, LTBP1 has been shown to facilitate the secretion of latent TGF- from the cell [281] and is also important in targeting TGF- for storage in the extracellular matrix [279,282,283]. By undergoing proteolytic cleavage, LTBP1 may also provide a mechanism for the release of latent TGF- from the extracellular matrix [279,282,284]. LTBP1 belongs to a family of recently identified matrix proteins, which share homology with fibrillins 1 and 2 and appear to be important components of connective tissue microfibrils [285]. Four LTBPs

have been identified, which range in size from approximately 180 to 310 kDa. LTBPs 1,3, and 4 all appear to bind small latent TGF-, but there are conflicting reports as to whether LTBP2 binds to TGF-. Although it is known that bone cells express LTBP2 and LTBP3 [286; Dallas, unpublished observations], future studies are clearly warranted to determine their specific roles in bone cell function and in regulation of TGF-. Activation of latent TGF- complexes involves the dissociation of mature TGF- from the propeptide. This can be achieved in vitro by extremes of pH, by chaotropic agents, by heat or via the action of proteases. Several studies have documented factors that stimulate or inhibit the activation of latent TGF- in various cell systems. However, very little is known about the actual mechanisms for activation of TGF- by cells [287]. A protease-mediated mechanism appears the most likely mechanism for the majority of cell types. In support of this, evidence for a plasmin-mediated activation mechanism has been reported in UMR-106 osteosarcoma cells [288]. Osteoclasts are interesting in that they are one of the few cell types for which an acid-mediated mechanism of activation for TGF- seems likely. The pH in the sealed zone, under the ruffled border of the osteoclast, has been reported to be as low as 5 [289]. Significant amounts of TGF- would be expected to be activated at this pH. Thus, osteoclasts may be unique in utilizing an acid-mediated mechanism of activation of latent TGF-.

390 In addition to its roles in matrix storage and release of latent TGF-, LTBP1 may play a role in the activation of latent TGF-. Flaumenhaft and co-workers [290] showed that antibodies to LTBP1 inhibited the activation of latent TGF- by cocultures of smooth muscle and endothelial cells. LTBP1 may also modulate activation by protecting latent TGF- from activation by proteases until the complex is bound to the surface of an appropriate cell that expresses cell surface-bound protease systems for activation. An important role for thrombospondin in the activation of TGF- has been suggested by studies showing that thrombospondin can activate latent TGF- [291] and, more recently, by the observation that thrombospondin null mice show a phenotype resembling the TGF- knockout mouse [292]. However, the role of thrombospondin in the activation of TGF- in bone cells remains undetermined. Future studies are clearly required to unravel the mechanisms of activation of TGF- in bone and will be the key to understanding the regulation of this important factor in bone. Figure 4 shows the relationship between TGF- and its associated binding proteins.

XV. BONE MORPHOGENETIC PROTEINS Studies of stromal cell transplantation in vivo [293] and analysis of different mouse or rat clonal cell populations in vitro [294 – 296] provide evidence that multipotential mesenchymal cells can differentiate into different cell types, including osteoblasts and chondroblasts. These mesenchymal cells have the capacity to undergo the commitment process to give rise to progeny with more limited or monopotential differentiation capacity. The mechanism of commitment and specification of uncommitted mesenchymal precursor cells to the osteoblast lineage is not fully understood. BMPs appear to play regulatory roles in the commitment of mesenchymal precursor cells to the osteoblast lineage [297]. BMPs are originally identified from bone matrix using an ectopic bone formation assay [298,299]. When BMPs are implanted subcutaneously or intramuscularly in mice or rats, they induce massive amounts of new cartilage and bone at implantation sites. BMP-2 has been shown to induce embryonic limb cells to differentiate into mature chondroblasts and osteoblasts [300]. BMP-2, BMP-4, and BMP-7 induce mesenchymal precursor C3H10T1/2 cells to differentiate into mature chondroblasts and osteoblasts [301,302]. These results indicate the regulatory roles of BMPs in the commitment of mesenchymal cells to the osteoblast and chondroblast lineages (also see Chapters 5 and 14). BMPs mediate their functions through type I and type II serine/threonine kinase receptors. To determine which type I BMP receptor is responsible for the commitment of mesenchymal precursor cells to the osteoblast lineage and for

GREGORY R. MUNDY ET AL.

BMP-induced osteoblast differentiation, we have stably transfected dominant-negative truncated and constitutively active type IA and IB BMP receptors (BMPR-IA and BMPR-IB) into a mesenchymal precursor cell line, 2T3. Overexpression of the truncated type IB BMP receptor, but not the truncated type IA BMP receptor, blocks the commitment of 2T3 cells to the osteoblast lineage and BMP-2induced osteoblast differentiation [167]. The differentiation pathway of 2T3 cells is respecified to the adipocyte lineage after type IB BMP receptor signaling is blocked [167]. These findings suggest that the temporal acquisition of the type IB BMP receptor in mesenchymal precursor cells may be a key step for the commitment of mesenchymal cells to the osteoblast lineage.

A. Role of BMPs and BMP Receptors in Osteoblast and Chondroblast Differentiation BMPs play important roles in osteoblast differentiation. BMP-2 stimulates osteoblast differentiation in primary osteoblastic cells [303,304] and in cell lines derived from osteogenic tissues [305 – 307]. BMPs also induce nonosteogenic precursor cells to differentiate into cells with osteoblast phenotypes. For example, BMP-2 induces myoblast C2C12 cells to differentiate into osteoblasts [308]. BMP-2, BMP-6, and BMP-7 have all been shown to induce both hypertrophy in cultured chondrocytes and osteogenesis from mesenchymal stem cells [309]. BMP-6 may play a central role in chondrocyte differentiation and endochondral bone formation, as glucocorticoid and estrogen selectively upregulate BMP-6 expression [310,311], and PTHrP, a key factor in chondrocyte differentiation, inhibits BMP-6 expression in chondrocytes [312]. BMP-2 upregulates BMP-2 and BMP-4 mRNA expression in primary osteoblastic cells [303,304], suggesting that BMP-2 acts as an autocrine and paracrine factor during osteoblast differentiation. The autocrine effect of BMP-2 may be mediated in part through transcription factors, Dlx2 and Dlx5, as BMP2 stimulates Dlx2 and Dlx5 mRNA expression and Dlx2 can bind and activate BMP-2 gene transcription [313,314]. BMPs may stimulate osteoblast differentiation in part through activating the osteoblast-specific transcription factor, Osf2/Cbfa1. In Osf2/Cbfa1 knockout mice, both membranous bones of the skull and endochondral bones in the rest of the skeleton are absent [315]. Heterozygous mutations of the Osf2/Cbfa1 gene are found in humans with cleidocranial dysplasia (CCD) [316,317]. BMPs may be important regulators of Osf2/Cbfa1, as BMP-7 and BMP-2 stimulate Osf2/Cbfa1 expression in pluripotent mouse fibroblast C3H10T1/2 cells and in mesenchymal precursor 2T3 cells [167,318]. In vitro studies using dominant-negative truncated and constitutively active type IA and IB BMP receptors reveal

391

CHAPTER 13 Cytokines and Bone Remodeling

the essential roles of the type IB BMP receptor in chondroblast and osteoblast differentiation. Expression of a dominant-negative type II and type IB BMP receptors in immature chondrocytes leads to a loss of differentiated function and inhibits the expression of chondrocyte marker genes such as type II collagen and aggrecan [319]. When the truncated type IB BMP receptor is overexpressed in mesenchymal precursor 2T3 cells, the osteoblast differentiation properties of 2T3 cells are blocked. Both BMP-2-induced mineralized bone matrix formation and osteoblast-specific gene expression such as Osf2/Cbfa1 and osteocalcin are inhibited [167].

in the shapes of the epiphyses of the long bones, suggesting that the type IB BMP receptor is essential for the formation of joints [329]. This phenotype shows close similarities to that of the bp/bp (brachypodism) mouse, which contains the mutated GDF-5 gene [326]. GDF-5 has been shown to bind specifically to the type IB BMP receptor but not to the other type I receptors [330]. In BMPR-IA homozygous null mutant mice, morphological defects are first detected at 7.0 days postcoitum (dpc) and no mesoderm is formed in the mutant embryos, suggesting that BMP signaling through the type IA BMP receptor is essential for mesoderm formation during gastrulation [331].

B. Role of BMPs and BMP Receptors in Bone Development and Bone Formation

C. Signaling Mechanism of BMP Receptors (Fig. 5)

The function of BMP-2, BMP-6, and BMP-7 in ectopic bone formation and in fracture repair have been well characterized [309,320,321]. These peptides all induce ectopic bone formation and accelerate the healing processes during fracture repair. Generation of null mutant mice lacking BMP-2 or BMP-4 genes results in early embryonic lethality, before any skeletal formation has been initiated [322,323]. Tissue-specific knockout of these genes will clarify the specific roles of these genes in bone development and bone formation. In BMP-7-deficient mice, skeletal abnormalities are identified in discrete areas: the rib cage, the skull, and the hindlimbs [324], suggesting that BMP-7 plays a role in bone development and patterning. Consistent with the roles for BMPs in bone development and bone formation, studies of naturally occurring mutations have shown that different members of the BMP family may control the formation of different morphological features in the mammalian skeleton. Mutation in the BMP5 gene is associated with a wide range of skeletal defects, including reductions in long bone width and the size of several vertebral processes and an overall lower body mass [325]. Mutations in the GDF5 (CDMP-1/BMP-11) gene result in brachypodism in mice [326] and Hunter – Thompson-type chondrodysplasia in humans [327]. These findings demonstrate the critical roles of BMPs in bone development and bone formation. In vivo studies using virus-mediated kinase defective and constitutively active type IA and IB BMP receptors demonstrate that the type IB BMP receptor is required for cartilage formation. In chicken embryos, chondrogenesis of limb mesenchymal cells and bone formation in the distal limb bud are markedly inhibited by dominant-negative type IB BMP receptors, but not by dominant-negative type IA BMP receptors [328]. Mice lacking the type IB BMP receptor gene are viable and display brachydactyly due to fusion and severe reduction in length of the first and second phalages. X-ray analysis of BMPR-IB-/- mice reveals defects

BMPs signal through serine/threonine kinase receptors, composed of type I and type II components. Three type I receptors have been shown to bind to BMP ligands: BMPRIA, BMPR-IB, and ActR-IA (also termed ALK-3, ALK-6, and ALK-2, respectively) [332,333]. Three type II receptors for BMPs are also identified. BMPR-II, ActR-IIA, and ActR-IIB [334 – 337]. Whereas BMPR-IA, BMPR-IB, and BMPR-II are specific to BMPs, ActR-IIA and ActR-IIB are also signaling receptors for activins. ActR-IA has been shown to bind both BMP-7 and activin when expressed in COS cells. In embryonic P19 cells and MC3T3-E1 cells, endogenous ActR-IA mediates only BMP signaling, but not activin signaling [338]. No cross interaction between BMP and TGF- receptors has yet been demonstrated. Signal transduction through serine/threonine kinase receptors has been best characterized in the TGF- receptor system. It is likely that BMPs transduce signals in a similar

FIGURE 5

The BMP ligand – receptor and signal transduction pathway.

392

GREGORY R. MUNDY ET AL.

fashion. Type I and type II BMP receptors are both indispensable for signal transduction; after ligand binding they form a heteromeric-activated receptor complex. Both type I and type II BMP receptors are composed of three parts: short extracellular domains with 10 – 12 cysteine residues; single membrane-spanning domains; and intracellular domains with a serine/threonine kinase region [332 – 336]. Preceding the serine/threonine kinase domain, type I BMP receptors, but not type II BMP receptors, have a domain that contains a characteristic SGSGS motif (GS domain) [332,333]. The GS domain, which plays an important role in signal transduction, distinguishes type I serine/threonine kinase receptors from type II receptors. In the TGF- and activin receptor systems, ligands bind to type II receptors in the absence of type I receptors. Type I receptors can bind ligands only in the presence of type II receptors. However, BMPs bind weakly to type I receptors in the absence of type II receptors. In the presence of type II receptors, the binding of BMPs to type I receptors is accelerated [332,333]. In the TGF- receptor system, the type II receptor kinase transphosphorylates the GS domain in the type I receptor, which leads to activation of the type I receptor kinase [337]. Phosphorylation of the type I receptor is required for signal transduction, as TGF--mediated responses are impaired after mutation of serine and threonine residues in the GS domain of type I receptor or after mutation in the type II receptor, rendering it incapable to phosphorylate the type I receptor [339,340]. These observations indicate that the type II receptor is a primary binding protein for ligand and that the type I receptor acts as an effector in the signal transduction. This notion is supported by the observation that the mutation of glutamine in the GS domain of the type I BMP receptor to aspartic acid results in a receptor with a constitutively activated kinase. In these mutants, signals are transduced from the type I BMP receptor in the absence of ligand and the type II BMP receptor [341].

D. Downstream Molecules of BMP Receptor Signaling Type I BMP receptor substrates include a recently identified protein family, the Smad proteins, that play a central role in the relay of BMP signals from the receptor to target genes in the nucleus. Smad 1 [341], Smad 5 [342], and Smad 8 [343] are phosphorylated by BMP receptors in a ligand-dependent manner. These receptor-regulated Smads physically associate with the ligand-activated receptor complex and undergo phosphorylation at the C terminus [342]. After release from the receptor, Smad proteins associate with the related protein Smad 4, which acts as a shared partner. This complex translocates into the nucleus and

participates in gene transcription. In vertebrates, Smad proteins consist of three regions: a conserved N-terminal domain (MH1 domain), a conserved C-terminal domain (MH2 domain), and a more divergent linker region. In the transcriptional complex, the Smads contact DNA via their N-terminal domains [344]. The C-terminal domain of Smads mediates Smad – receptor interaction [345]. Activation of specific genes by Smads is conducted by interaction with specific DNA-binding proteins. The Xenopus protein Fast1 is the prototypic Smad-recruiting, DNAbinding factor [346]. Fast1, which contains a “winged helix” DNA-binding domain, binds to the activin response element (ARE) and is absolutely required for activation of the Mix.2 gene in response to activin or TGF-. Fast1 bound to DNA alone does not activate transcription. However, recruitment of an activated Smad2 – Smad 4 complex to the ARE by Fast1 results in activation of Mix.2 expression [347]. One of the transcription factors, which interacts with Smad 1, has been identified as a homeodomain DNAbinding protein, Hoxc-8. Hoxc-8 can serve as a transcriptional repressor for osteopontin gene transcription. Interaction of Smad 1 with Hoxc-8 relieves the repressive activity of Hoxc-8 and activates target gene transcription. The interactive regions of Smad 1 with Hoxc-8 are in the MH1 domain and linker region [348]. Consistent with these findings, transgenic mice overexpressing Hoxc-8 show abnormal cartilage, which is characterized by an accumulation of proliferating chondrocytes and reduced maturation [349]. Interestingly, Smad 2, a downstream molecule for TGF- signaling, interacts with a homeodomain DNAbinding protein, TGIF, a repressor of transcription, when it moves into the nucleus with Smad 4. The Smad 2 – Smad 4 complex can recruit TGIF and histone deacetylases (HDACs) to a Smad target promoter, repressing transcription [350]. These findings provide molecular evidence for the functional connection between growth factors activin/BMP/TGF- and homeodomain DNA-binding transcription factors. The immediate downstream target genes for BMP signaling during osteoblast and chondroblast differentiation have yet to be identified. Dlx2 and Dlx5 are candidate target genes as immediate downstream molecules for BMP signaling, as BMP-2 induces Dlx2 and Dlx5 mRNA expression within 1 h and without the synthesis of new proteins [313,314].

XVI. CONCLUSION Rapidly accumulating evidence shows cytokines generated in the bone microenvironment control bone remodeling. There is now a body of data derived from in vivo studies in animals which show that over- or underproduction of certain cytokines cause profound effects on bone. These

393

CHAPTER 13 Cytokines and Bone Remodeling

fundamental observations have the potential of not only increasing our understanding of the pathophysiology of osteoporosis, but also leading to new and better forms of therapy using these molecules as targets for drug discovery programs.

13.

14.

Acknowledgments

15.

We are grateful to Nancy Garrett for her assistance in the preparation of the manuscript. This work was supported by Grants P01 CA40035, AR28149, AR07464, RR01346, DK45229, and AR42372 from the National Institutes of Health.

16.

17.

References 1. S. Akira and T. Kishimoto, NF-IL-6 and NF-B in cytokine gene regulation. Adv. Immunol. 65, 1 – 46 (1997). 2. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage-colony stimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 127, 2592 – 2594 (1990). 3. H. Kodama, A. Yamasaki, M. Nose, S. Niida, Y. Ohgame, U. Abe, M. Kumegawa, and T. Suda, Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony stimulating factor. J. Exp. Med. 173, 269 – 272 (1991). 4. H. Kodama, A. Yamasaki, M. Abe, S. Niida, Y. Yakeda, and H. Kawashima, Transient recruitment of osteoclasts and expression of their function in osteopetrotic (op/op) mice by a single injection of macrophage colony-stimulating factor. J. Bone Miner. Res. 8, 45 – 50 (1993). 5. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa, and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 6. Y. Y. Myint, K. Miyakawa, M. Naito, L. D. Shultz, Y. Oike, K. I. Yamamura, and K. Takahashi, Granulocyte/macrophage colonystimulating factor and interleukin-3 correct osteopetrosis in mice with osteopetrosis mutation. Am. J. Pathol. 154, 553 – 566 (1999). 7. E. Lagasse and I. L. Weissman, Enforced expression of Bcl-2 in monocytes rescues macrophages and partially reverses osteopetrosis in op/op mice. Cell 89, 102 – 131 (1997). 8. R. Pacifici, Cytokines, estrogen and postmenopausal osteoporosis. Endocrinology 139, 2659 – 2661 (1998). 9. J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, G. Gronowicz, J. Kalinowski, and C. C. Pilbeam, Mice lacking the type I interleukin-1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). 10. Y. Abu-Amer, F. P. Ross, J. Edwards, and S. L. Teitelbaum, Lipopolysaccharide-stimulated osteoclastogenesis is mediated by tumor necrosis factor via its P55 receptor. J. Clin. Invest. 100, 1557 – 1565 (1997). 11. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vasalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1702 (1997). 12. R. L. Jilka, G. Hagoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased

18.

19.

20.

21. 22.

23.

24. 25.

26.

27.

28.

29.

osteoclast development after estrogen loss: Mediation by interleukin6. Science 257, 88 – 91 (1992). V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Cilberto, G. A. Rodan, and F. Constantini, Interleukin6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994). H. Kitamura, H. Kawata, F. Takahashi, Y. Higuchi, T. Furuichi, H. Ohkawa, Bone marrow neutrophilia and suppressed bone turnover in human interleukin-6 transgenic mice. Am. J. Pathol. 147, 1682 – 1692 (1995). R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin-1 and tumor necrosis factor is required to completely prevent bone loss in the early post-ovariectomy period. Endocrinology 136, 3054 – 3061 (1995). G. R. Mundy and T. J. Martin, Pathophysiology of skeletal complications of cancer. In “Physiology and Pharmacology of Bone: Handbook of Experimental Pharmacology” (G. R. Mundy and T. J. Martin, eds.), Chap. 18, pp. 641 – 671. Springer, Berlin. G. D. Roodman, N. Kurihara, Y. Ohsaki, A. Kukita, D. Hosking, A. Demulder, J. F. Smith, and F. R. Singer, Interleukin 6: A potential autocrine/paracrine factor in Paget’s disease of bone. J. Clin. Invest. 89, 46 – 52 (1992). R. J. O’Keefe, L. A. Teot, D. Singh, J. E. Puzas, R. N. Rosier, and D. G. Hicks, Osteoclasts constitutively express regulators of bone resorption: An immunohistochemical and in situ hybridization study. Lab. Invest. 76, 457 – 465 (1997). B. G. Mills and A. Fausta, Cytokines expressed in multinucleated cells: Paget’s disease and giant cell tumors versus normal bone. Calcif. Tissue Int. 61, 16 – 21 (1997). S. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 1 – 17 (1998). T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 107 – 123 (1998). D. Heyman, J. Guicheux, F. Gouin, N. Passuti, and G. Dalculsi, Cytokines, growth factors and osteoclasts. Cytokine 10, 155 – 168 (1998). G. D. Roodman, Cell biology of the osteoclast. Exp. Hematol. 27, 1229 – 1241 (1999). S. V. Reddy, S. Takahashi, M. Dallas, R. E. Williams, L. Neckers, and G. D. Roodman, IL-6 antisense deoxyoligonucleotides inhibit bone resorption by giant cells from human giant cell tumors of bone. J. Bone Miner. Res. 9, 753 – 757 (1994). R. D. Devlin, S. V. Reddy, sR. Savino, G. Ciliberto, and G. D. Roodman, IL-6 mediates the effects of IL-1 or TNF, but not PTHrP or 1,25(OH)2D3, on osteoclast-like cell formation in normal human bone marrow cultures. J. Bone Miner. Res. 13, 393 – 399 (1998). S. J. Choi, R. D. Devlin, C. Menaa, H. Chung, G. D. Roodman, and S. V. Reddy, Cloning and identification of human Sca as a novel inhibitor of osteoclast formation and bone resorption. J. Clin. Invest. 102, 1360 – 1368 (1998). C. Menaa, R. D. Devlin, S. V. Reddy, Y. Gazitt, S. J. Choi, and G. D. Roodman, Annexin II increases osteoclast formation by stimulating the proliferation of osteoclast precursors in human marrow cultures. J. Clin. Invest. 103, 1605 – 1613 (1999). C. Chenu, J. Pfeilschifter, G. R. Mundy, and G. D. Roodman, Transforming growth factor  inhibits formation of osteoclast-like cells in long term human marrow cultures. Proc. Natl. Acad. Sci. USA 85, 5683 – 5687 (1988).

394 30. J. Pfeilschifter, S. Seyedin, and G. R. Mundy, Transforming growth factor  inhibits bone resorption in fetal rat long bone cultures. J. Clin. Invest. 82, 380 – 385 (1988). 31. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-. Nature Med. 2, 1132 – 1136 (1996). 32. H. Takai, M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, and Y. Yamada, Transforming growth factor- stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells. J. Biol. Chem. 278, 27091 – 27096 (1998). 33. T. Murakami, M. Yamamoto, M. Yamamoto, K. Ono, M. Nishikawa, N. Nagata, K. Motoyoshi, and T. Akatsu, Transforming growth factor-1 increases mRNA levels of osteoclastogenesis inhibitory factor in osteoblastic/stromal cells and inhibits the survival of murine osteoclast-like cells. Biochem. Biophys. Res. Commun. 252, 747 – 752 (1998). 34. K. Fuller, B. Wong, S. Fox, Y. Choi, and T. J. Chambers, TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188, 997 – 1001 (1998). 35. M. R. Wani, K. Fuller, N. S. Kim, Y. Choi, and T. J. Chambers, Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: Synergistic activation of differentiation, cell spreading and fusion. Endocrinology 140, 1927 – 1935 (1999). 36. T. Suda, I. Nakamura, E. Jimi, and N. Takahashi, Regulation of osteoclast function. J. Bone Miner. Res. 12, 869 – 879 (1997). 37. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tomesko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic cell function. Nature 390, 175 – 179 (1997). 38. B. R. Wong, J. Rho, J. Arron, E. Robinson, J. Orlinick, M. Chao, S. Kalachikov, E. Cayani, F. S. Bartlett, W. N. Frankel, S. Y. Lee, and Y. Choi, TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-jun N-terminal kinase in T cells. J. Biol. Chem. 272, 25190 – 25194 (1997). 39. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 40. W. S. Simonet, D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 41. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 42. L. Lum, B. R. Wong, R. Josien, J. D. Becherer, H. ErdjumentBromage, J. Schonlondorff, P. Tempsta, Y. Choi, and C. P. Blobel, Evidence for a role of a tumor necrosis factor  (TNF-) converting enzyme-like protease in shedding of TRANCE, a TNF family member involved in osteoclastogenesis and dendritic cell survival. J. Biol. Chem. 274, 13613 – 13618 (1999). 43. B. R. Wong, R. Josien, and Y. Choi, TRANCE is a TNF family member that regulates dendritic cell and osteoclast function. J. Leukocyte Biol. 65, 715 – 724 (1999). 44. T. Yoneda, I. Kato, L. F. Bonewald, W. H. Burgess, and G. R. Mundy, A novel cytokine with osteoclastopoietic activity. J. Periodontal Res. 28, 521 – 522 (1993).

GREGORY R. MUNDY ET AL. 45. M. Y. Lee, D. R. Eyre, and W. R. A. Osborne, Isolation of a murine osteoclast colony stimulating factor. Proc. Natl. Acad. Sci. USA 88, 8500 – 8504 (1991). 46. K. Tsukii, N. Shima, S. I. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Yano, O. Shibata, N. Udagawa, H. Yasuda, T. Suda, and K. Higashio, Osteoclast differentiation factor mediates an essential signal for bone resorption induced by 1,25-dihydroxyvitamin D3, prostaglandin E2, or parathyroid hormone in the microenvironment of bone. Biochem. Biophys. Res. Commun. 246, 337 – 341 (1998). 47. B. O. Oyajobi, P. J. Williams, K. Traianedes, T. Yoneda, D. M. Anderson, and G. R. Mundy, Efficacy of a genetically-engineered soluble receptor activator of NF-B (RANK) fusion protein on bone resorption in vitro and in vivo. J. Bone Miner. Res. 14, 1124 (1999). 48. V. Shalhoub, J. Faust, W. J. Boyle, C. R. Dunstan, M. Kelley, S. Kaufman, S. Scully, G. Van, and D. L. Lacey, Osteoprotegerin and osteoprotegerin ligand effects on osteoclast formation from human peripheral blood mononuclear cell precursors. J. Cell. Biochem. 72, 251 – 261 (1999). 49. J. M. W. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 50. K. Matsuzaki, N. Udagawa, N. Takahashi, K. Yamaguchi, H. Yasuda, N. Shima, T. Morinaga, Y. Toyama, Y. Yabe, K. Higashio, and T. Suda, Osteoclast differentiation factor (ODF) induces osteoclastlike cell formation in human peripheral blood mononuclear cell cultures. Biochem. Biophys. Res. Commun. 246, 199 – 204 (1998). 51. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. I. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998). 52. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology, 139, 4743 – 4746 (1998). 53. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 54. L. C. Hofbauer, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, B. L. Riggs, and S. Khosla, Interleukin-1 and tumor necrosis factor-, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells. Bone 25, 255 – 259 (1999). 55. E. Jimi, N. Akiyama, T. Tsurukai, N. Okahashi, K. Kobayashi, N. Udagawa, T. Nishihara, N. Takahashi, and T. Suda, Osteoclast differentiation factor (ODF) acts as a multifunctional regulator in murine osteoclast differentiation and function. J. Immunol. 162, 434 – 442 (1999). 56. T. L. Burgess, Y. X. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol., 145, 527 – 538 (1999). 57. D. L. Lacey, H. L. Tan, J. Lu, A. Eli, G. Van, F. Fletcher, W. J. Boyle, T. Juan, and A. Polverino, Osteoprotegerin ligand is essential for osteoclast survival in vitro and in vivo. J. Bone Miner. Res. 14(suppl), S179 (1999) [Abstract 1182]. 58. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Cappparelli, S. Morony, A. J. Olivera-dos Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penniger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph node organogenesis. Nature 397, 315 – 323 (1999).

CHAPTER 13 Cytokines and Bone Remodeling 59. B. O. Oyajobi, K. Traianedes, T. Yoneda, and G. R. Mundy, Expression of RANK Ligand (RANKL) by myeloma cells requires binding to bone marrow stromal cell via an 41-VCAM-1 interaction. Bone 23(Suppl.), S180 (1998). 60. N. Chikatsu, Y. Takeuchi, T. Kikuchi, E. Ogata, and T. Fujita, Adhesion of cancer cells to bone marrow stromal cells induces ODF/OPGL expression and osteoclast-like cell formation in vitro. 2nd International Conference on Cancer-Induced Bone Diseases, Davos, Switzerland. 61. N. J. Horwood, V. Kartsogiannis, M. H. C. Iam, E. Romas, J. M. W. Quinn, T. J. Martin, and M. T. Gillespie, Activated T cells are capable of inducing osteoclast formation: A mechanism for rheumatoid arthritis. Bone 23 (Suppl.), S214 (1998). 62. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). 63. B. Dougall, M. Glaccum, K. Charrier, K. Rohrbach, K. Brasel, T. De Smedt, E. Daro, J. Smith, M. E. Tomesko, C. R. Maliszewski, A. Armstrong, V. Shen, S. Bain, D. Cosman, D. Anderson, P. J. Morrisey, J. J. Peschon, and J. Schuh, RANK is essential for osteoclast and lymph node development. Genes Dev. 13, 2412 – 2424 (1999). 64. N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, K. Yano, T. Morinaga, and K. Higashio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400 (1998). 65. R. H. Arch, R. W. Gedrich, and C. B. Thompson, Tumor necrosis factor receptor-associated factors (TRAFs): A family of adapter proteins that regulate life and death. Genes Dev. 12, 2821 – 2830 (1998). 66. B. G. Darnay, J. Ni, P. A. Moore, and B. B. Aggarwal, Activation of NF-B by RANK requires tumor necrosis factor-receptor associated factor (TRAF) 6 and -inducing kinase: Identification of a novel TRAF6 interaction motif. J. Biol. Chem. 274, 7724 – 7731 (1999). 67. B. R. Wong, R. Josien, S. Y. Lee, M. Vologodskaia, R. M. Steinman, and Y. Choi, The TRAF family of signal transducers mediates the NF-B activation by the TRANCE receptor. J. Biol. Chem. 273, 28355 – 28359 (1998). 68. L. Galibert, M. E. Tomesko, D. M. Anderson, D. Cosman, and W. C. Dougall, The involvement of multiple tumor necrosis factor receptor (TNFR)-associated factors in signaling mechanisms of receptor activator of NF-B, a member of the TNFR superfamily. J. Biol. Chem. 273, 34120 – 34127 (1998). 69. M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, A. van der Heiden, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. Cao, J. M. Penninger, C. J. Page, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF-6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). 70. H. H. Kim, D. E. Lee, J. N. Shin, Y. S. Lee, Y. M. Jeon, C. H. Chung, J. Ni, B. S. Kwon, and Z. H. Lee, Receptor activator of NF-B recruits multiple TRAF family adaptors and activates c-jun N-terminal kinase. FEBS Lett. 443, 297 – 302 (1999). 71. Z. Cao, J. Xiong, M. Takeuchi, T. Kurma, and D. V. Goeddel, TRAF6 is a signal transducer for interleukin-1. Nature 383, 443 – 446 (1996). 72. V. Iotsova, J. Caamano, J. Loy, Y. Yang, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NF-B1 and NF-B2. Nature Med. 3, 1285 – 1289 (1997). 73. G. Franzoso, L. Carlson, L. Xing, L. Pljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-B in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997).

395 74. E. Tsuda, M. Goto, S. I. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 75. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, S. Sander, G. Van, J. Tarpley, P. Derby, R. Lee, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 76. K. B. Tan, J. Harrop, M. Reddy, P. Young, J. Terrett, J. Emery, G. Moore, and A. Truneh, Characterization of a novel TNF-like ligand and recently described TNF ligand and TNF receptor superfamily genes and their constitutive and inducible expression in hematopoietic and non-hematopoietic cells. Gene 204, 35 – 46 (1997). 77. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentsz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 78. T. J. Yun, P. M. Chaudhary, G. L. Shu, J. K. Frazer, M. K. Ewings, S. M. Schwartz, V. Pascual, L. E. Hood, and E. A. Clark, OPG/ FDCR-1, a TNF receptor family member, is expressed in lymphoid cells and is up-regulated by ligating CD40. J. Immunol. 161, 6113 – 6121 (1998). 79. T. Morinaga, N. Nakagawa, H. Yasuda, E. Tsuda, and K. Higashio, Cloning and characterization of the gene encoding human osteoprotegerin/osteoclastogenesis-inhibitory factor. Eur. J. Biochem. 254, 685 – 691 (1998). 80. J. G. Emery, P. McDonnell, M. B. Burke, K. C. Deen, S. Lyn, C. Silverman, E. Dul, E. R. Appelbaum, C. Eichman, R. DiPrinizio, I. E. Dodds, James, M. Rosenberg, J. C. Lee, and P. R. Young, Osteoprotegerin is a receptor for the cytotoxic ligand TRAIL. J. Biol. Chem. 273, 14363 – 14367 (1998). 81. H. Kawaguchi, C. Manabe, C. Miyaura, H. Chikuda, K. Nakamura, and M. Kuro-o, Independent impairment of osteoblast and osteoclast differentiation in klotho mouse exhibiting low-turnover osteopenia. J. Clin. Invest. 104, 229 – 237 (1999). 82. K. Yano, E. Tsuda, N. Washida, F. Kobayashi, M. Goto, A. Harada, K. Ikeda, K. Higashio, and Y. Yamada, Immunological characterization of circulating osteoprotegerin/osteoclastogenesis inhibitory factor: Increased serum concentrations in post-menopausal osteoporosis. J. Bone Miner. Res. 14, 518 – 527 (1999). 83. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H-L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 84. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. A. Mochizuki, A. Tomoyasu, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 85. H. Brandstrom, K. B. Jonsson, C. Ohlsson, O. Vidal, S. Ljunghall, and O. Ljunggren, Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 248, 338 – 341 (1998). 86. L. C. Hofbauer, F. Gori, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140, 4382 – 4389 (1999).

396 87. O. N. A. Vidal, H. Brandstrom, K. B. Jonsson, and C. Ohlsson, Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: Downregulation by glucocorticoids. J. Endocrinol. 159, 191 – 195 (1998). 88. O. N. A. Vidal, K. Sjogren, B. I. Eriksson, O. Ljunggren, and C. Ohlsson, Osteoprotegerin mRNA is increased by interleukin-1 in the human osteosarcoma cell line MG-63 and in human osteoblastlike cells. Biochem. Biophys. Res. Commun. 248, 696 – 700 (1998). 89. Y. Hakeda, Y. Kobayashi, K. Yamaguchi, H. Yasuda, E. Tsuda, K. Higashio, T. Miyata, and M. Kumegawa, Osteoclastogenesisinhibitory factor (OCIF) directly inhibits bone-resorbing activity of isolated mature osteoclasts. Biochem. Biophys. Res. Commun. 251, 796 – 801 (1998). 90. T. Akatsu, T. Murakami, M. Nishikawa, K. Ono, N. Shinomiya, E. Tsuda, S.-I. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Higashio, M. Yamamoto, and N. Nagata, Osteoclastogenesis-inhibitory factor suppresses osteoclast survival by interfering in the interaction of stromal cells with osteoclasts. Biochem. Biophys. Res. Commun. 250, 229 – 234 (1998). 91. T. Akatsu, T. Murakami, K. Ono, M. Nishikawa, E. Tsuda, S-I. Mochizuki, N. Fujise, K. Higashio, K. Motoyoshi, M. Yamamoto, and N. Nagata, Osteoclastogenesis-inhibitory factor exhibits hypocalcemic effects in normal mice and in hypercalcemic nude mice carrying tumors associated with humoral hypercalcemia of malignancy. Bone 23, 495 – 498 (1998). 92. M. Yamamoto, T. Murakami, M. Nishikawa, E. Tsuda, S. Mochizuki, K. Higashio, T. Akatsu, K. Motoyoshi, and N. Nagata, Hypocalcemic effect of osteoclastogenesis inhibitory factor/osteoprotegerin in the thyroparathyroidectomized rat. Endocrinology 139, 4012 – 4015 (1998). 93. S. Morony, C. Caparelli, R. Lee, G. Shimamoto, T. Boone, D. L. Lacey, and C. R. Dunstan, A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1, TNF-, PTH, PTHrP and 1,25(OH)2D3. J. Bone Miner. Res. 14, 1478 – 1485 (1999). 94. B. R. MacDonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 233 (1986). 95. S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J. Clin. Invest. 91, 257 – 263 (1993). 96. U. Sarma and A. M. Flanagan, Macrophage colony stimulating factor induces substantial osteoclast generation and bone resorption in human bone marrow cultures. Blood 88, 2531 – 2540 (1996). 97. C. S. Lader and A. M. Flanagan, Prostaglandin E2, interleukin 1 and tumor necrosis factor- increase human osteoclast formation and bone resorption in vitro. Endocrinology 139, 3157 – 3164 (1998). 98. H. Amano, S. Yamada, and R. Felix, Colony-stimulating factor-1 stimulates the fusion process in osteoclasts. J. Bone Miner. Res. 13, 846 – 843 (1998). 99. E. Jimi, I. Nakamura, L. T. Duong, T. Ikebe, N. Takahashi, G. A. Rodan, and T. Suda, Interleukin 1 induces multinucleation and boneresorbing activity of osteoclasts in the absence of osteoblast/stromal cells. Exp. Cell Res. 247, 84 – 93 (1999). 100. E. Jimi, T. Shuto, and T. Koga, Macrophage colony stimulating factor and interleukin 1  maintain the survival of osteoclast-like cells. Endocrinology 136, 808 – 811 (1995). 101. K. Fuller, J. M. Owens, C. J. Jagger, A. Wilson, and R. Moss, Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J. Exp. Med. 178, 1733 – 1744 (1993).

GREGORY R. MUNDY ET AL. 102. H. Yoshida, S. I. Hayashi, T. Kunisada, M. Ogawa, S. Nishikawa, H. Okamura, T. Sudo, L. D. Shultz, and S. I. Nishikawa, The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442 – 443 (1990). 103. M. L. Sherman, B. L. Weber, R. Datta, and D. W. Kufe, Transcriptional and posttranscriptional regulation of macrophage-specific colony stimulating factor gene expression by tumor necrosis factor. J. Clin. Invest. 85, 442 – 447 (1990). 104. J. H. Falkenburg, M. A. Harrington, R. A. de Paus, W. K. Walsh, R. Daub, J. E. Landegent, and H. E. Broxmeyer, Differential transcriptional and posttranscriptional regulation of gene expression of the colony-stimulating factors by interleukin-1 and fetal bovine serum in murine fibroblasts. Blood 78, 658 – 665 (1991). 105. S. D. Isaacs, X. Fan, D. Fan, H. Gewant, T. C. Murphy, P. Farmer, W. R. Taylor, M. S. Nanes, and J. Rubin, Role of NF-B in the regulation of macrophage colony stimulating factor by tumor necrosis factor- in 5T2 bone stromal cells. J. Cell. Physiol. 179, 193 – 200 (1999). 106. G. Q. Yao, B. H. Sun, E. E. Hammond, E. N. Spencer, M. C. Horowitz, K. L. Insogna, and E. C. Weir, The cell-surface form of colony-stimulating factor-1 is regulated by osteotropic agents and supports formation of multinucleated osteoclast-like cells. J. Biol. Chem. 273, 4119 – 4128 (1998). 107. R. B. Kimble, S. Srivastava, F. P. Ross, A. Matayoshi, and R. Pacifici, Estrogen deficiency increases the ability of stromal cells to support osteoclastogenesis via an IL-1 and TNF mediated stimulation of M-CSF production. J. Biol. Chem. 271, 28890 – 28897 (1996). 108. S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). 109. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 110. B. F. Boyce, T. Yoneda, C. Lowe, P. Soriano, and G. R. Mundy, Requirement of pp60c-src expression of osteoclasts to form ruffled borders and resorb bone. J. Clin. Invest. 90, 1622 – 1627 (1992). 111. K. L. Insogna, M. Sahni, A. B. Grey, S. Tanaka, W. C. Horne, L. Neff, M. Mitnick, J. B. Levy, and R. Baron, Colony stimulating factor-1 induces cytoskeletal re-organization and c-src dependent tyrosine phosphorylation of selected cellular proteins in rodent osteoclasts. J. Clin. Invest. 100, 2476 – 2485 (1997). 112. A. Teti, A. Taranta, S. Migliaccio, A. Degiorgi, E. Santandrea, I. Villanova, T. Faraggiana, M. Chellaiah, and K. A. Hruska, Colony stimulating factor-1-induced osteoclast spreading depends on substrate and requires vitronectin receptor and c-src proto-oncogene. J. Bone Miner. Res. 13, 50 – 58 (1998). 113. R. L. Lees, and J. H. M. Heersche, Macrophage colony stimulating factor increases bone resorption in dispersed osteoclast cultures by increasing osteoclast size. J. Bone Miner. Res. 14, 937 – 945 (1999). 114. M. Gowen, M. C. Meikle, and J. J. Reynolds, Stimulation of bone resorption in vitro by a non-prostanoid factor released by human monocytes in culture. Biochem. Biophys. Acta 762, 471 – 474 (1983). 115. M. Gowen, and G. R. Mundy, Actions of recombinant interleukin-1, interleukin-2 and interferon gamma on bone resorption in vitro. J. Immunol. 136, 2478 – 2482 (1986). 116. M. Gowen, G. Nedwin, and G. R. Mundy, Preferential inhibition of cytokine stimulated bone resorption by recombinant interferon gamma. J. Bone Miner. Res. 1, 469 – 474 (1986). 117. B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interleukin-1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989).

CHAPTER 13 Cytokines and Bone Remodeling 118. B. F. Boyce, A. J. P. Yates, and G. R. Mundy, Bolus injections of recombinant human interleukin-1 cause transient hypocalcemia in normal mice. Endocrinology 125, 2780 – 2783 (1989). 119. J. Pfeilschifter, C. Chenu, A. Bird, G. R. Mundy, and G. D. Roodman, Interleukin-1 and tumor necrosis factor stimulate the formation of human osteoclastlike cells in vitro. J. Bone Miner. Res. 4, 113 – 118 (1989). 120. E. Thomassen, T. A. Bird, B. R. Renshaw, M. K. Kennedy, and J. E. Sims, Binding of interleukin-18 to the interleukin-1 receptor homologous receptor IL-1RP1 leads to activation of signaling pathways similar to those used by interleukin-1. J. Interferon Cytokine Res. 18, 1077 – 1088 (1998). 121. H. L. Uy, T. A. Guise, J. DeLaMata, S. D. Taylor, B. M. Story, M. R. Dallas, B. F. Boyce, G. R. Mundy, and G. D. Roodman, Effects of PTH-rP and PTH on osteoclasts and osteoclast precursors in vivo. Endocrinology 136, 3207 – 3212 (1995). 122. C. A. Dinarello, Biologic basis for interleukin-1 in disease. Blood 77, 1627 – 1652 (1996). 123. R. M. Hinson, J. A. Williams, and E. Shacter, Elevated interleukin-6 is induced by prostaglandin E2 in a murine model of inflammation: Possible role of cyclooxygenase-2. Proc. Natl. Acad. Sci. USA 93, 4885 – 4890 (1996). 124. M. Gowen, D. D. Wood, G. R. Mundy, and R. G. G. Russell, Studies on the actions of interleukin-1 on bone metabolism: IL-1 stimulation of bone cell proliferation, and inhibition of IL-1-induced bone resorption by interferon gamma. Brit. J. Rheum. 24, 147 – 149 (1985). 125. D. Smith, M. Gowen, and G. R. Mundy, Effects of interferon gamma and other cytokines on collagen synthesis in fetal rat bone cultures. Endocrinology 120, 2494 – 2499 (1987). 126. K. Sato, Y. Fujii, K. Kasono, M. Ozawa, H. Imamura, Y. Kanaji, H. Kurosawa, T. Tsushima, and K. Shizume, Parathyroid hormonerelated protein and interleukin-1 synergistically stimulate bone resorption in vitro and increase the serum calcium concentration in mice in vivo. Endocrinology 124, 2172 – 2178 (1989). 127. M. Kawano, H. Tanaka, H. Ishikawa, M. Nobuyoshi, K. Iwato, H. Asaoku, O. Tanabe, and A. Kuramoto, Interleukin-1 accelerates autocrine growth of myeloma cells through interleukin-6 in human myeloma. Blood 73, 2145 – 2148 (1989). 128. R. Pacifici, J. L. Vannice, L. Rifas, and R. B. Kimble, Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrin. Metab. 77, 1135 – 1141 (1993). 129. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 130. C. A. O’Brien, I. Gubrij, S. C. Lin, R. L. Saylors, and S. C. Manolagas, STAT3 activation in stromal/osteoblastic cells is required for induction of the receptor activator of NF-B ligand and stimulation of osteoclastogenesis by gp 130-utilizing cytokines or interleukin-1 but not 1,25-dihydroxyvitamin D3 or parathyroid hormone. J. Biol. Chem. 274, 19310 – 19308 (1999). 131. L. C. Hofbauer, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Glucocorticoids stimulate osteoprotegerin ligand and inhibit osteoprotegerin production in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoidinduced osteoporosis. 18th Endocrine Society Meeting, San Diego, CA. 132. E. Jimi, I. Nakamura, T. Ikebe, N. Akiyama, N. Takahashi, and T. Suda, Activation of NF-B is involved in the survival of osteoclasts promoted by interleukin 1. J. Biol. Chem. 273, 8799 – 8805 (1998). 133. E. Jimi, T. Ikebe, N. Takahashi, N. Hirata, T. Suda, and T. Koga, Interleukin 1 activates an NF-B-like factor in osteoclast-like cells. J. Biol. Chem. 271, 4605 – 4608 (1996).

397 134. D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 319, 516 – 518 (1986). 135. R. A. Johnson, B. F. Boyce, G. R. Mundy, and G. D. Roodman, Tumors producing human TNF induce hypercalcemia and osteoclastic bone resorption in nude mice. Endocrinology 124, 1424 – 1427 (1989). 136. T. Yoneda, M. M. Alsina, J. B. Chavez, L. Bonewald, R. Nishimura, and G. R. Mundy, Evidence that tumor necrosis factor plays a pathogenetic role in the paraneoplastic syndromes of cachexia, hypercalcemia, and leukocytosis in a human tumor in nude mice. J. Clin. Invest. 87, 977 – 985 (1991). 137. M. Sabatini, A. J. Yates, R. Garrett, J. Chavez, J. Dunn, L. Bonewald, and G. R. Mundy, Increased production of tumor necrosis factor by normal immune cells in a model of the humoral hypercalcemia of malignancy. Lab. Invest. 63, 676 – 682 (1990). 138. M. Sabatini, J. Chavez, G. R. Mundy, and L. F. Bonewald, Stimulation of tumor necrosis factor release from monocytic cells by the A375 human melanoma via granulocyte-macrophage colony stimulating factor. Cancer Res. 50, 2673 – 2678 (1990). 139. I. R. Garrett, B. G. M. Durie, G. E. Nedwin, A. Gillespie, T. Bringman, M. Sabatini, D. R. Bertolini, and G. R. Mundy, Producion of lymphotoxin, a bone-resorbing cytokine, by cultured human myeloma cells. N. Engl. J. Med. 317, 526 – 532 (1987). 140. J. H. M. Feyen, P. Elford, F. E. Dipadova, and U. Trechsel, Interleukin-6 is produced by bone and modulated by parathyroid hormone. J. Bone Miner. Res. 4, 633 – 638 (1989). 141. Y. Ishimi, C. Miyaura, C. H. Jin, T. Akatsu, E. Abe, Y. Nakamura, A. Yamaguchi, S. Woshiki, T. Matsuda, and T. Hirano, IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297 – 3303 (1990). 142 J. Gilles, M. Qiao, G. R. Mundy, and L. F. Bonewald, Preferential stimulation of interleukin-6 production in bone by cytokines compared with systemic factors. J. Bone Miner. Res. 9, B433 (1994) [Abstract]. 143. J. De La Mata, H. Uy, T. A. Guise, B. Story, B. F. Boyce, G. R. Mundy, and G. D. Roodman, IL-6 enhances hypercalcemia and bone resorption mediated by PTH-rP in vivo. J. Clin. Invest. 95, 2846 – 2852 (1995). 144. K. Black, I. R. Garrett, and G. R. Mundy, Chinese hamster ovarian cells transfected with the murine interleukin-6 gene cause hypercalcemia as well as cachexia, leukocytosis and thrombocytosis in tumor-bearing nude mice. Endocrinology 128, 2657 – 2659 (1991). 145. T. Tamura, N. Udagawa, N. Takahashi C. Miyaura, S. Tanaka, Y. Yamada, Y. Koishihara, Y. Ohsugi, K. Kumaki, and T. Taga, Soluble interleukin-6 receptor triggers osteoclast formation by interleukin-6 Proc. Natl. Acad. Sci. USA 90, 11924 – 11928, (1993). 146. J. P. Gaillard, R. Bataille, H. Brailly, C. Zuber, K. Yasukawa, M. Attal, N. Maruo, T. Tage T. Kishimoto, and B. Klein, Increased and highly stable levels of functional soluble interleukin-6 receptor in sera of patients with monoclonal gammopathy. Eur. J. Immunol. 23, 820 – 824, (1993). 147. N. Udagawa, N. Takahashi, T. Katagiri, T. Tamura S. Wada, D. M. Findlay, T. J. Martin, H. Hirota, T. Taga, T. Kishimoto, and T. Suda, Interleukin (IL)-6 induction of osteoclast differentiation depends on IL-6 receptors expressed on osteoblastic cells but not on osteoclast progenitors. J. Exp. Med. 182, 1461 – 1468, (1995). 148. K. Kawasaki, Y. H. Gao, S. Yokose, Y. Kaji, T. Nakamura, T. Suda, K. Yoshida, T. Taga, T. Kishimoto, H. Kataoka, T. Yuasa, H. Norimatsu, and A. Yamaguchi, Osteoclasts are present in gp 130-deficient mice. Endocrinology 138, 4959 – 4965 (1997). 149. N. Nishimoto, Y. Shima, K. Yoshizaki, and T. Kishimoto, Myeloma biology and therapy: Present status and future developments. Hematol. Oncol. Clinics North Am. 11, 159 – 172 (1997).

398 150. B. Klein, X. G. Zhang, Z. Y. Lu, and R. Bataille, Interleukin-6 in human multiple myeloma. Blood 74, 863 – 872 (1995). 151. G. Gutierrez, G. Rossini, J. Chavez, T. Nishihara, T. Yoneda, and G. R. Mundy, Identification, purification and expression of a naturally-occurring inhibitor of interleukin-6 (IL-6). J. Bone Miner. Res. 9, 79 (1994) [Abstract]. 152. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartillage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999). 153. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 154. Y. Ogata, A. Kukita, T. Kukita M. Komine, A. Miyahara, S. Miyazaki, and O. Kohashi, A novel role of IL-15 in the development of osteoclasts: Inability to replace its activity with IL-2. J. Immunol. 162, 2754 – 2760 (1999). 155. I. B. Mclnnes, J. AI-Mughales, M. Field, B. P. Leung, F. P. Huang R. Dixon, R. D. Sturrock, P. C. Wilkinson, and F. Y. Lew, The role of interleukin-15 in T-cell migration and activation in rheumatoid arthritis. Nature Med. 2, 175 – 178 (1996). 156. N. Udagawa, N. J. Horwood, J. Elliot, A. Mackay, J. Owens, H. Okamura, M. Kurimoto, T. J. Chambers, T. J. Martin, and M. T. Gillespie, Interleukin 18 (interferon- inducing factor) is produced by osteoblasts and acts via granulocyte-macrophage colony stimulating factor and not via interferon- to inhibit osteoclast formation. J. Exp. Med. 185, 1005 – 1017 (1997). 157. R. L. Van Bezooijen, H. C. M. Farih-Sips, S. E. Papapoulos, and C. W. G. M. Lowik, Interleukin 17: A new bone acting cytokine in vitro. J. Bone Miner. Res. 14, 1513 – 1521 (1999). 158. S. Kotake, N. Udagawa, N. Takahashi, K. Matzuzaki, K. Itoh, S. Lyama, S. Saito K. Inoue, N. Kamatani, M. T. Gillespie, T. J. Martin, and T. Suda, IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J. Clin. Invest. 103, 1345 – 1352 (1999). 159. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. R. Dunn, T. J. Martin and M. T. Gillespie, Interleukin 18 inhibits osteoclast formation via T cell production of granulocyte macrophage colony stimulating factor. J. Clin. Invest. 101, 595 – 603 (1998). 160. E. Sigal, The molecular biology of mammalian arachidonic acid metabolism. Am. J. Physiol. 260, L13 – L28 (1991). 161. C. N. Serhan, J. Z. Haeggstrom, and C. C. Leslie, Lipid mediator networks in cell signaling: Update and impact of cytokines. FASEB J. 10, 1147 – 1158 (1996). 162. J. C. Otto and W. L. Smith, Prostaglandin endoperoxide synthases-1 and -2. J. Lipid Mediat. Cell Signal. 12, 139 – 156 (1995). 163. M. Peters-Golden, Cell biology of the 5-lipoxygenase pathway. Am. J. Respir. Crit. Care. Med. 157, S227 – S232 (1998). 164. K. Schoonjans, B. Staels, and J. Auwerx, The peroxisome proliferator activated receptors (PPARS) and their effects on lipid metabolism and adipocyte differentiation. Biochim. Biophys. Acta 1302, 93 – 109 (1996). 165. B. M. Forman, J. Chen, and R. M. Evans, Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors alpha and delta. Proc. Natl. Acad. Sci. USA 94, 4312 – 4317 (1997). 166. P. R. Devchand, H. Keller, J. M. Peters, M. Vazquez, F. J. Gonzalez, and W. Wahli, The PPARalpha-leukotriene B4 pathway to inflammation control. Nature 384, 39 – 43 (1996). 167. D. Chen, X. Ji, M. A. Harris, J. Q. Feng, G. Karsenty, A. J. Celeste, V. Rosen, G. R. Mundy, and S. E. Harris, Differential roles for bone morphogenetic protein (BMP) receptor type IB and IA in dif-

GREGORY R. MUNDY ET AL.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

ferentiation and specification of mesenchymal precursor cells to osteoblast and adipocyte lineages. J. Cell Biol. 142, 295 – 305 (1998). D. D. Diascro, Jr. R. L. Vogel, T. E. Johnson, K. M. Witherup, S. M. Pitsenberger, S. J. Rutledge D. J. Prescott, G. A. Rodan, and A. Schmidt, High fatty acid content in rabbit serum is responsible for the differentiation of osteoblasts into adipocyte-like cells. J. Bone Miner. Res. 13, 96 – 106 (1998). A. Lchikawa, Molecular biology of prostaglandin E receptors – Expression of multi-function by PGE receptor subtypes and isoforms Nippon Rinsho 56, 1813 – 1818 (1998). M. Suda, K. Tanaka, K. Natsui, T. Usui, I. Tanaka, M. Fukushima, C. Shigeno, J. Konishi, S. Narumiya, A. Ichikawa, and N. Nakao, Prostaglandin E receptor subtypes in mouse osteoblastic cell line. Endocrinology 137, 1698 – 1705 (1996). M. Weinreb, A. Grosskopf, and N. Shir, The anabolic effect of PGE2 in rat bone marrow cultures is mediated via the EP4 receptor subtype. Am. J. Physiol. 276, E376 – E383 (1999). K. Ono, T. Akatsu, T. Murakami, M. Nishikawa, M. Yamamoto, N. Kugai, K. Motoyoshi, and N. Nagata, Important role of EP4, a subtype of prostaglandin (PG) E receptor, in osteoclast-like cell formation form mouse bone marrow cells induced by PGE2. J. Endocr. 158, R1 – R5 (1998). M. Mano, Y. Hakeda, H. Mano, H. Kiyomura, and M. Kumegawa, Prostaglandin E2 directly inhibits osteoclastic bone resorption through EP2 receptor. Bone 23, (Suppl.) W066 (1998). T. Yokomizo, T. Izumi, K. Chang, Y. Takuwa, and A. Shimizu, G-protein-coupled receptor for leukotriene B4 that mediates chemotaxis. Nature 387, 620 – 624 (1997). K. R. Lynch, G. P. O’Neill, Q. Liu, D. S. Im, N. Sawyer, K. M. Metters, N. Coulombe M. Abramovitz, D. J. Figueroa, Z. Zeng, B. M. Connolly, C. Bai, C. P. Austin, A. Chateauneuf, R. Stocco, G. M. Greig, S. Kargman, S. B. Hooks, E. Hosfield, D. L. Williams, Jr. A. W. Ford-Hutchinson, C. T. Caskey, and J. F. Evans, Characterization of the human cysteinyl leukotriene CysL T1 receptor. Nature 399, 789 – 793 (1999). H. R. Herschman, Regulation of prostaglandin synthase-1 and prostaglandin synthase-2. Cancer Metastasis Rev. 13, 241 – 256 (1994). B. P. Kennedy, C. C. Chan, S. A. Culp, and W. A. Cromlish, Cloning and expression of rat prostaglandin endoperoxide synthase (cyclooxygenase)-2 cDNA. Biochem. Biophys. Res. Commun. 197, 494 – 500 (1993). Y. Harada, K. Hatanaka, M. Kawamura, M. Saito, M. Ogino, M. Majima, T. Ohno, K. Ogino, K. Yamamoto, Y. Taketani, S. Yamamoto, and M. Katori Role of prostaglandin H synthase-2 in prostaglandin E2 formation in rat carrageenin-induced pleurisy. Prostaglandins 51, 19 – 33 (1996). R. N. Dubois, S. B. Abramson, L. Crofford, R. A. Gupta, L. S. Simon, L. B. Van De Putte, and P. E. Lipsky, Cyclooxygenase in biology and disease. FASEB J. 12, 1063 – 1067 (1998). D. A. Kujubu and H. R. Herschman, Dexamethasone inhibits mitogen induction of the TIS 10 prostaglandin synthase/cyclooxygenase gene. J. Biol. Chem. 267, 7991 – 7994 (1992). J. L. Masferrer, K. Seibert, B. Zweifel, and P. Needleman, Endogenous glucocorticoids regulate an inducible cyclooxygenase enzyme. J. Biol. Chem. 89, 3917 – 3921 (1992). H. H. Versteeg, P. M. van Bergen en Henegouwen, S. J. van Deventer, and M. P. Peppelenbosch, Cyclooxygenase-dependent signalling: Molecular events and consequences. FEBS Lett. 445, 1 – 5 (1999). E. J. Canalis, Clinical review 83: Mechanisms of glucocorticoid action in bone: Implications to glucocorticoid-induced osteoporosis. Clin. Endocrinol. Metab. 81, 3441 – 3447 (1996). S. C. Miller and S. C. Marks, Effects of prostaglandins on the skeleton. Clin. Plast. Surg. 21, 393 – 400 (1994).

CHAPTER 13 Cytokines and Bone Remodeling 185. H. Kawaguchi, C. C. Pilbeam, J. R. Harrison, and L. G. Raisz, The role of prostaglandins in the regulation of bone metabolism. Clin. Orthop. 313, 36 – 46 (1995). 186. D. C. Klein and L. G. Raisz, Prostaglandins: Stimulation of bone resorption in tissue culture. Endocrinology 86, 1436 – 1440 (1970). 187. K. Ueda, A. Saito, H. Nakano, M. Aoshima, M. Yokata, R. Muraoka, and T. Iwaya, Cortical hypersteosis following long-term administration of prostaglandin E2 in infants with cyanotic congenital heart disease. J. Pediatr. 97, 834 – 836 (1980). 188. R. W. Norrdin, W. S. Jee, and W. B. High, The role of prostaglandins in bone in vivo. Prostaglandins Leukot. Essent. Fatty Acids 41, 139 – 149 (1990). 189. T. Nagata, K. Kaho, S. Nishikawa, H. Shinohara, Y. Wakano, and H. Ishida, Effect of prostaglandin E2 on mineralization of bone nodules formed by fetal rat calvarial cells. Calcif. Tissue Int. 55, 451 – 457 (1994). 190. W. S. Jee, B. Y. Lin, Y. F. Ma, and H. Z. Ke, Extra cancellous bone induced by combined prostaglandin E2 and risedronate administration is maintained after their withdrawal in older female rats. J. Bone Miner. Res. 10, 963 – 970 (1995). 191. W. S. Jee and Y. F. Ma, The in vivo anabolic actions of prostaglandins in bone. Bone 21, 297 – 304 (1997). 192. Y. F. Ma, Z. Pan, W. S. Jee, C. H. Lin, H. H. Liang, H. Chen, S. Pun, and X. J. Li, Intermittent on/off prostaglandin E2 and risedronate are equally anabolic as daily PGE2 alone treatment in cortical bone of ovariectomized rats. J. Bone Miner. Res. 12, 2108 – 2112 (1997). 193. X. J. Li, W. S. Jee, Y. L. Li, and P. Patterson-Buckendahl, Transient effects of subcutaneously administered prostaglandin E2 on cancellous and cortical bone in young adult dogs. Bone 11, 353 – 364 (1990). 194. H. R. Jorgensen, H. Svanholm, and A. Host, Bone formation induced in an infant by systemic prostaglandin-E2 administration. Acta. Orthop. Scand. 59, 464 – 466 (1988). 195. H. Z. Ke, V. W. Shen, H. Qi, D. T. Crawford, D. D. Wu, X. G. Liang, K. L. Chidsey-Frink, C. M. Pirie, H. A. Simmons, and D. D. Thompson, Prostaglandin E2 increases bone strength in intact rats and in ovariectomized rats with established osteopenia. Bone 23, 249 – 255 (1998). 196. C. H. Lin, W. S. Jee, Y. F. Ma, and R. B. Setterberg, Early effects of prostaglandin E2 on bone formation and resorption in different bone sites of rats. Bone 17, 255S – 259S (1995). 197. W. E. Gallwitz, G. R. Mundy, C. H. Lee, M. Qiao, G. D. Roodman, M. Raftery, S. J. Gaskell, and L. F. Bonewald, 5-lipoxygenase metabolites of arachidonic acid stimulate isolated osteoclasts to resorb calcified matrices. J. Biol. Chem. 268, 10087 – 10094 (1993). 198. C. Garcia, B. F. Boyce, J. Gilles, M. Dallas, M. Qiao, G. R. Mundy, and L. F. Bonewald, Leukotriene B4 stimulates osteoclastic bone resorption both in vitro and in vivo. J. Bone Miner. Res. 11, 1619 – 1627 (1996). 199. C. Garcia, M. Qiao, D. Chen, M. Kirchen, W. Gallwitz, G. R. Mundy, and L. F. Bonewald, Effects of synthetic peptidoleukotrienes on bone resorption in vitro. J. Bone Miner. Res. 11, 521 – 529 (1996). 200. S. Meghji, J. R. Sandy, A. M. Scutt, W. Harvey, and M. Harris, Stimulation of bone resorption by lipoxygenase metabolites of arachidonic acid. Prostaglandins 36, 139 – 149 (1988). 201. Z. Estrov, D. S. Halperin, F. Coceani, and M. H. Freedman, Modulation of human marrow haematopoiesis by leucotrienes in vitro. Br. J. Haematol. 69, 321 – 327 (1988). 202. D. Pasquale and G. Chikkappa, Lipoxygenase products regulate proliferation of granulocyte-macrophage progenitors. Exp. Hematol. 21, 1361 – 1365 (1993). 203. L. Stenke, P. Reizenstein, and J. A. Lindgren, Leukotrienes and lipoxins: New potential performers in the regulation of human myelopoiesis. Leukemia Res. 18, 727 – 732 (1994).

399 204. F. Dupuis, V. Desplat, V. Praloran, and Y. E. Denizot, Effects of lipidic mediators on the growth of human myeloid and erythroid marrow progenitors. J. Lipid Mediat. Cell Signal. 16, 117 – 125 (1997). 205. W. Ren and R. Dziak, Effects of leukotrienes on osteoblastic cell proliferation. Calcif. Tissue Int. 49, 197 – 201 (1993). 206. L. F. Bonewald, M. Flynn, M. Qiao, M. R. Dallas, G. R. Mundy, and B. F. Boyce, Mice lacking 5-lipoxygenase have increased cortical bone thickness. Adv. Exp. Med. Biol. 433, 299 – 302 (1997). 207. D. P. Nicolella, A. E. Nicholls, L. F. Bonewald, and J. Lankford, Mechanical properties of bone from mice lacking 5-lipoxygenase differ significantly from normal mice. J. Bone Miner. Res. 12 (Suppl) 223 (1997). 208. K. Traianedes, M. R. Dallas, I. R. Garrett, G. R. Mundy, and L. F. Bonewald, 5-Lipoxygenase metabolites inhibit bone formation in vitro. Endocrinology 139, 3178 – 3184 (1998). 209. W. R. Henderson, The role of leukotrienes in inflammation. Ann. Intern. Med. 121, 684 – 697 (1994). 210. F. M. Collier, W. H. Huang, W. R. Holloway, J. M. Hodge, M. T. Gillespie, L. L. Daniels, M. H. Zheng, and G. C. Nicholson, Osteoclasts from human giant cell tumors of bone lack estrogen receptors. Endocrinology 139, 1258 – 1267 (1998). 211. P. M. Renzi, Antileukotriene agents and asthma: The dart that kills the elephant? CMAJ 160, 217 – 223 (1999). 212. S. E. Wenzel, Leukotriene receptor antagonists and related compounds. Can. Respir. J. 6, 189 – 193 (1999). 213. S. Battu, J. L. Beneytout, M. Pairet, and M. Rigaud, Cyclooxygenase-2 up-regulation after FLAP transfection in human adenocarcinoma cell line HT29 cl. 19A. FEBS Lett. 437, 49 – 55 (1998). 214. R. S. Weinstein, J. A. Jilka, A. M. Parfitt, and S. C. Manolagos, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). 215. A. Sartorio, A. Conti, and B. Ambrosi, Bone and collagen turnover in patients with active and preclinical Cushings syndrome and in subjects with adrenal incidentaloma. J. Clin. Endocrinol. Metab. 83, 2605 – 2606 (1998). 216. D. J. Baylink, Glucocorticoid-induced osteoporosis. N. Engl. J. Med. 309, 306 – 308 (1983). 217. R. L. Wilder, Neuroendocrine-immune system interactions and autoimmunity. Annu. Rev. Immunol. 13, 307 – 338 (1995). 218. L. J. Crofford, T. K. Konstantine, G. Mastorakos, M. A. Magiakou, J. Wells, K. S. Kanik, P. W. Gold, G. P. Chrousos, and R. L. Wilder, Circadian relationships between interleukin (IL)-6 and hypothalamic-pituitary-adrenal axis hormones: Failure of IL-6 to cause sustained hypercortisolism in patients with early untreated rheumatoid arthritis. J. Clin. Endocrinol. Metab. 82, 1279 – 1283 (1997). 219. M. R. Podgorski, N. J. Goulding, N. K. Hall, R. J. Flower, and P. J. Maddison, Autoantibodies to lipocortin-1 are associated with impaired glucocorticoid responsiveness in rheumatoid arthritis. J. Rheumatol. 19, 1668 – 1671 (1992). 220. G. R. Mundy, G. D. Roodman, L. F. Bonewald, T. Yoneda, and M. Sabatini, Effects of TNF and lymphotoxin on bone cells. Immunol. Ser. 56, 483 – 498 (1992). 221. G. R. Mundy, Myeloma bone disease. Eur. J. Cancer 34, 246 – 51 (1998). 222. S. C. Manolagas, The role of IL-6 type cytokines and their receptors in bone. Ann. N.Y. Acad. Sci. 840, 194 – 204 (1998). 223. W. Zhang and R. Dziak, Effects of interleukin-1 alpha on arachidonic acid metabolism in human osteosarcoma osteoblastic cells. Prostagland. Leukot. Essent. Fatty Acids 49, 521 – 526 (1993). 224. W. Zhang and R. Dziak, Tumor necrosis factor alpha stimulates arachidonic acid metabolism in human osteoblastic osteosarcomal cells. Prostagland. Leukot. Essent. Fatty Acids 54, 427 – 431 (1996).

400 225. S. S. Edwin, S. L. LaMarche, D. Thai, D. W. Branch, and M. D. Mitchell, 5-Hydroxyeicosatetraenoic acid biosynthesis by gestational tissues: Effects of inflammatory cytokines. Am. J. Obstet. Gynecol. 169, 1467 – 1471 (1993). 226. J. W. Bijlsma, Prevention of glucocorticoid induced osteoporosis. Ann. Rheum. Dis. 56, 507 – 509 (1997). 227. M. K. OBanion, V. D. Winn, and D. A. Young, cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc. Natl. Acad. Sci. USA 89, 4888 – 4892 (1992). 228. A. J. de Brum-Fernandes, S. Laporte, M. Heroux, M. Lora, C. Patry, H. A. Menard, R. Dumais, and R. Leduc, Expression of prostaglandin endoperoxide synthase-1 and prostaglandin endoperoxide synthase-2 in human osteoblasts. Biochem. Biophys. Res. Commun. 198, 955 – 960 (1994). 229. F. M. Maciel, P. Sarrazin, S. Morisset, M. Lora, C. Patry, R. Dumais, and A. J. de Brum-Fernandes, Induction of cyclooxygenase-2 by parathyroid hormone in human osteoblasts in culture. J. Rheumatol. 24, 2429 – 2435 (1997). 230. M. Goppelt-Struebe, D. Schaefer, and A. J. Habenicht, Differential regulation of cyclo-oxygenase-2 and 5-lipoxygenase-activating protein (FLAP) expression by glucocorticoids in monocytic cells. Br. J. Pharmacol. 122, 619 – 624 (1997). 231. C. A. Riddick, W. L. Ring, J. R. Baker, C. R. Hodulik, and T. D. Bigby, Dexamethasone increases expression of 5-lipoxygenase and its activating protein in human monocytes and THP-1 cells. Eur. J. Biochem. 246, 112 – 118 (1997). 232. P. V. Hauschka, A. E. Mavrakos, S. E. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix. J. Biol. Chem. 261, 12665 – 12674 (1986). 233. S. M. Seyedin, T. C. Thomas, A. Y. Thompson, D. M. Rosen, and K. A. Piez, Purification and characterization of two cartilage-inducing factors from bovine demineralized bone. Proc. Natl. Acad. Sci. USA 82, 2267 – 2271 (1985). 234. L. F. Bonewald and S. L. Dallas, The role of active and latent TGF in bone formation. J. Cell Biochem. 55, 350 – 357 (1994). 235. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 236. A. B. Roberts, Molecular and cell biology of TGF-beta. Miner. Electrolyte Metab. 24, 111 – 119 (1998). 237. L. F. Bonewald, Regulation and regulatory activities of transforming growth factor . Crit. Rev. Eukary. Gene Express. 9, 33 – 44 (1999). 238. J. Pfeilschifter, and G. R. Mundy, Modulation of transforming growth factor beta activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 239. S. L. Dallas, G. R. Mundy, and L. F. Bonewald, Proteolytic cleavage of the latent TGF binding protein-1 by avian osteoclasts: A possible mechanism for release of latent TGF from bone matrix. J. Bone Miner. Res. 12 (Suppl. 1), S418 (1997). 240. E. J. Mackie and U. Trechsel, Stimulation of new bone formation in vivo by transforming growth factor : remodeling of woven bone and lack of inhibition by indomethacin. Bone 11, 295 – 300 (1990). 241. C. Marcelli, A. J. Yates, and G. R. Mundy, In vivo effects of human recombinant transforming growth factor beta on bone turnover in normal mice. J. Bone Miner. Res. 5, 1087 – 1096 (1990). 242. M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). 243. R. Fujimoto, T. Tanizawa, S. Nishida, N. Yamamoto, S. Soshi, N. Endo, and H. E. Takahashi, Local effects of transforming growth factor-beta 1 on rat calvaria: Changes depending on the dose and the injection site. J. Bone Miner. Metab. 17, 11 – 17 (1999). 244. L. S. Beck, L. DeGuzman, W. P. Lee, Y. Xu, L. A. McFatridge, N. A. Gillett, and E. P. Amento, TGF1 induces bone closure of skull defects. J. Bone Miner. Res. 6, 1257 – 1265 (1991).

GREGORY R. MUNDY ET AL. 245. D. Rosen, S. C. Miller, E. DeLeon, A. Y. Thompson, H. Bentz, M. Matthews, and S. Adams, Systemic administration of recombinant transforming growth factor beta 2 (rTGF-2) stimulates parameters of cancellous bone formation in juvenile and adult rats. Bone 15, 355 – 359 (1994). 246. M. Machwate, E. Zerath, X. Holy, M. Hott, D. Godet, A. Lomri, and P. J. Marie, Systemic administration of transforming growth factor2 prevents the impaired bone formation and osteopenia induced by unloading in rats. J. Clin. Invest. 96, 1245 – 1253 (1995). 247. D. Gazit, Y. Zilberman, L. Passi-Even, R. Ebner, and A. Kahn, In vivo administration of TGF beta 1 increases marrow osteoprogenitor number and selectively reverses the bone loss seen in old “osteoporotic” mice. J. Bone Miner. Res. 10 (Suppl. 1), 25 (1995). 248. J. Pfeilschifter, O. Wolf, A. Nautmann, H. W. Minne, G. R. Mundy, and R. Ziegler, Chemotactic response of osteoblast-like cells to TGF. J. Bone Miner. Res. 5, 825 – 830 (1990). 249. J. M. Hock, E. Canalis, and M. Centrella, Transforming growth factor  stimulates bone matrix apposition and bone cell replication in cultured fetal rat calvariae. Endocrinology 126, 421 – 426 (1990). 250. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Effects of transforming growth factor  and epidermal growth factor on cell proliferation and the formation of bone nodules in isolated fetal rat calvaria cells. J. Cell. Physiol. 140, 386 – 395 (1989). 251. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. L. Dallas, J. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of TGF on bone nodule formation and expression of bone morphogenetic protein-2, osteocalcin, osteopontin, alkaline phosphatase and type I collagen mRNA in prolonged cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 252. T. L. Chen and R. L. Bates, Recombinant human transforming growth factor beta 1 modulates bone remodeling in a mineralizing bone organ culture. J. Bone Miner. Res. 8, 423 – 434 (1993). 253. A. Erlebacher and R. Derynck, Increased expression of TGF2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). 254. A. H. Tashjian, E. F. Voelkel, M. Lazzaro, F. R. Singer, A. B. Roberts, R. Derynck, M. E. Winkler, and L. Levine, Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorption in cultured mouse calvaria. Proc. Natl. Acad. Sci. USA 82, 4535 – 4538 (1985). 255. J. J. Letterio and E. P. Bottinger, TGF-beta knockout and dominantnegative receptor transgenic mice. Miner. Electrolyte Metab. 24, 161 – 167 (1998). 256. M. M. Shull, I. Ormsby, A. B. Kier, S. Pawlowski, R. J. Diebold, M. Yin, R. Allen, C. Sidman, G. Proetzel, D. Calvin, N. Annunziata, and T. Doetschman, Targeted disruption of the mouse transforming growth factor 1 gene results in multifocal inflammatory disease. Nature 359, 693 – 699 (1992). 257. A. B. Kulkarni, C. G. Huh, D. Becker, A. Geiser, M. Lyght, K. C. Fkander, A. B. Roberts, M. B. Sporn, J. M. Ward, and S. Karlsson, Transforming growth factor 1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90, 770 – 774 (1993). 258. A. G. Geiser, Q. Q. Zeng, M. Sato, L. M. Helverding, T. Hirano, and C. H. Turner, Decreased bone mass and bone elasticity in mice lacking the transforming growth factor beta-1 gene. Bone 23, 87 – 93 (1998). 259. L. P. Sanford, I. Ormsby, A. C. Gittenberger-de Groot, H. Sariola, R. Friedman, G. P. Boivin, E. I. Cardell, and T. Doetschman, TGF beta 2 knockout mice have multiple developmental defects that are non-overlapping with other TGF beta knockout phenotypes. Development 124, 2659 – 2670 (1997). 260. G. Proetzel, S. A. Pawlowski, M. V. Wiles, M. Yin, G. P. Boivin, P. N. Howles, J. Ding, M. W. Ferguson, and T. Doetschman, Transforming growth factor beta 3 is required for secondary palate fusion. Nature Genet. 11, 409 – 414 (1995).

CHAPTER 13 Cytokines and Bone Remodeling 261. J. L. Letterio, A. G. Geiser, A. B. Kulkarni, N. S. Roche, M. B. Sporn, and A. B. Roberts, Maternal rescue of transforming growth factor-1 null mice. Science 264, 1936 – 1938 (1994). 262. R. J. Diebold, M. J. Eis, M. Yin, I. Ormsby, G. P. Boivin, B. J. Darrow, J. E. Saffitz, and T. Doetschman, Early onset multifocal inflammation in the TGF1-null mouse is lymphocyte-mediated. Proc. Natl. Acad. Sci. USA 92, 12215 – 12219 (1995). 263. B. L. Langdahl, J. Y. Knudsen, H. K. Jensen, N. Gregersen, and E. F. Eriksen, A sequence variation: 713-8delC in the transforming growth factor-beta 1 gene has higher prevalence in osteoporotic women than in normal women and is associated with very low bone mass in osteoporotic women and increased bone turnover in both osteoporotic and normal women. Bone 20, 289 – 294 (1997). 264. Y. Yamada, A. Miyauchi, J. Goto, Y. Takagi, H. Okuizumi, M. Kanematsu, H. Takai, A. Harada, and K. Ikeda, Association of a polymorphism of the transforming growth factor-beta 1 gene with genetic susceptibility to osteoporosis in postmenopausal japanese women. J. Bone Miner. Res. 13, 1569 – 1576 (1998). 265. Y. Yamada, T. Hosoi, F. Makimoto, H. Tanaka, Y. Seino, and K. Ikeda, Transforming growth factor beta-1 gene polymorphisms and bone mineral density in Japanese adolescents. Am. J. Med. 106, 477 – 479 (1999). 266. D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF. J. Cell. Biochem. 70, 478 – 488 (1998). 267. J. L. Wrana, TGF-beta receptors and signaling mechanisms. Miner. Electrolyte Metab. 24, 120 – 130 (1998). 268. J. Massague, TGF signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998). 269. R. Derynck and X. H. Feng, TGF-beta receptor signalling. Biochim. Biophys. Acta 1333, F105 – F150 (1997). 270. M. Centrella, C. Ji, and T. L. McCarthy, Control of TGF-beta receptor expression in bone. Front. Biosci. 3, D113 – 124 (1998). 271. A. Horner, P. Kemp, C. Summers, S. Bord, N. J. Bishop, A. W. Kelsall, N. Coleman, and J. E. Compston, Expression and distribution of transforming growth factor-beta isoforms and their signalling receptors in growing human bone. Bone 23, 95 – 102 (1998). 272. E. H. Filvaroff, A. Erlebacher, J. O. Ye, S. E. Gitelman, J. C. Lotz, M. R. Heilmann, and R. Derynck, Increased trabecular bone in transgenic mice expressing a dominant-negative, type II TGF receptor in osteoblasts. J. Bone Miner. Res. 12(Suppl. 1), S118 (1997). 273. R. Serra, M. Johnson, E. H. Filvaroff, J. LaBorde, D. M. Sheehan, R. Derynck, and H. L. Moses, Expression of a truncated, kinase-defective TGF type II receptor in mouse skeletal tissue promotes terminal chondrocyte differentiation and osteoarthritis. J. Cell Biol. 139, 541 – 552 (1997). 274. F. Lopez-Casillas, J. L. Wrana, and J. Massague, Betaglycan presents ligand to the TGF signaling receptor. Cell 73, 1435 – 1444 (1993). 275. L. F. Bonewald, L. Wakefield, R. O. C. Oreffo, A. Escobedo, D. R. Twardzik, and G. R. Mundy, Latent forms of transforming growth factor-beta (TGF) derived from bone cultures: Identification of a naturally occurring 100-kDa complex with similarity to recombinant latent TGF. Mol. Endocrinol. 5, 741 – 751 (1991). 276. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardzik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of latent transforming growth factor  (TGF) complexes in osteoblastlike cell lines. J. Biol. Chem. 269, 6815 – 6822 (1994). 277. T. Kanzaki, A. Olofsson, A. Moren, C. Wernstedt, U. Hellman, K. Miyazono, L. Claesson-Welsh, and C. H. Heldin, TGF1 binding protein: A component of the large latent complex of TGF1, with multiple repeat sequences. Cell 61, 1051 – 1061 (1990). 278. T. Tsuji, F. Okada, K. Yamaguchi, and T. Nakamura, Molecular cloning of the large subunit of transforming growth factor type  masking protein and expression of the mRNA in various rat tissues. Proc. Natl. Acad. Sci. USA 87, 8835 – 8839 (1990).

401 279. S. L. Dallas, K. Miyazono, T. M. Skerry, G. R. Mundy, and L. F. Bonewald, Dual role for the latent TGF binding protein (LTBP) in storage of latent TGF in the extracellular matrix and as a structural matrix protein. J. Cell Biol. 131, 539 – 549 (1995). 280. S. L. Dallas, D. R. Keene, S. P. Bruder, J. Saharinen, L. Y. Sakai, G. R. Mundy, and L. F. Bonewald, Role of the latent-TGF binding protein-1 in fibrillin-containing microfibrils in bone cells in vitro and in vivo. J. Bone Miner. Res. 15, 68 – 81 (2000). 281. K. Miyazono, A. Olofsson, P. Colosetti, and C. H. Heldin, A role of the latent TGF1-binding protein in the assembly and secretion of TGF1. EMBO J. 10, 1091 – 1101 (1991). 282. J. Taipale, K. Miyazono, C. H. Heldin, and J. Keski-Oja, Latent transforming growth factor-1 associates to fibroblast extracellular matrix via latent TGF- binding protein. J. Cell Biol. 124, 171 – 181 (1994). 283. J. Taipale, J. Saharinen, K. Hedman, and J. Keski-Oja, Latent transforming growth factor-1 and its binding protein are components of extracellular matrix microfibrils. J. Histochem. Cytochem. 44, 875 – 889 (1996). 284. H. A. Pedrozo, Z. Schwartz, R. Gomez, A. Ornoy, W. Xin-Sheng, S. L. Dallas, L. F. Bonewald, D. D. Dean, and B. D. Boyan, Growth plate chondrocytes store latent TGF in their matrix through latent TGF binding protein. J. Cell. Physiol. 177, 343 – 354 (1998). 285. S. Sinha, C. Nevett, C. A. Shuttleworth, and C. M. Kielty, Cellular and extracellular biology of the latent transforming growth factorbeta binding proteins. Matrix Biol. 17, 529 – 545 (1999). 286. W. Yin, E. Smiley, J. Germiller, R. P. Mecham, J. V. Florer, R. J. Wenstrup, and J. Bonadio, Isolation of a novel latent transforming growth factor- binding protein gene (LTBP-3). J. Biol. Chem. 270, 10147 – 10160 (1995). 287. R. Flaumenhaft, S. Kojima, M. Abe, and D. B. Rifkin, Activation of latent transforming growth factor beta. Adv. Pharmacol. 24, 51 – 76 (1993). 288. J. A. Yee, L. Yan, J. C. Dominguez, E. H. Allan, and T. J. Martin, Plasminogen-dependent activation of latent transforming growth factor beta (TGF) by growing cultures of osteoblast-like cells. J. Cell. Physiol. 157, 528 – 534 (1993). 289. I. A. Silver, R. J. Murrills, and O. J. Etherington, Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp. Cell Res. 175, 266 – 276 (1998). 290. R. Flaumenhaft, M. Abe, Y. Sato, K. Miyazono, C. H. Heldin, and D. B. Rifkin, Role of the latent TGF- binding protein in the activation of latent TGF by co-cultures of endothelial and smooth muscle cells. J. Cell Biol. 120, 995 – 1002 (1993). 291. S. Schultz-Cherry, H. Chen, D. F. Mosher, T. M. Misenheimer, H. C. Krutzsch, D. D. Roberts, and J. E. Murphy-Ullrich, Regulation of transforming growth factor- activation by discrete sequences of thrombospondin 1. J. Biol. Chem. 269, 26783 – 26788 (1995). 292. S. E. Crawford, V. Stellmach, J. E. Murphy-Ullrich, S. M. F. Ribeiro, J. Lawler, R. O. Hynes, G. P. Boivin, and N. Bouck, Thrombospondin1 is a major activator of TGF in vivo. Cell 93, 1159 – 1170 (1998). 293. A. J. Friedenstein, P. K. Chailakhyan, and U. V. Gerasimov, Bone marrow osteogenic stem cells: In vitro cultivation and transplantation in diffusion chambers. Cell Tissue Kinet. 20, 263 – 272 (1987). 294. S. M. Taylor and P. A. Jones, Multiple new phenotypes induced in 10T1/2 and 3T3 cells treated with 5-azacytidine. Cell 17, 771 – 779 (1979). 295. J. H. Bennett C. J. Joyner, J. T. Triffitt, and M. E. Owen, Adipocytic cells cultured from marrow have osteogenic potential. J. Cell Sci. 99, 131 – 139 (1991). 296. J. N. Beresford, J. H. Bennett, C. Delving, P. S. Leboy, and M. E. Owen, Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures. J. Cell Sci. 102, 341 – 351 (1992). 297. J. E. Aubin, F. Liu, L. Malaval, and A. K. Gupta, Osteoblast and chondroblast differentiation. Bone 17, 77S – 83S (1995).

402 298. M. R. Urist, Bone formation by autoinduction. Science 150, 893 – 899 (1965). 299. J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. Kriz, R. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528 – 1534 (1988). 300. V. Rosen, J. Nove, J. J. Song, R. S. Thies, K. Cox, and J. M. Wozney, Responsiveness of clonal limb bud cell lines to bone morphogenetic protein 2 reveals a sequential relationship between cartilage and bone cell phenotypes. J. Bone Miner. Res. 9, 1759 – 1768 (1994). 301. M. Ahrens, T. Ankenbauer, D. Schroder, A. Hollnagel, H. Mayer, and G. Gross, Expression of human bone morphogenetic protein-2 or -4 in murine mesenchymal progenitor C3H10T1/2 cells induces differentiation into distinct mesenchymal cell lineages. DNA Cell Biol. 12, 871 – 880 (1993). 302. I. Asahina, T. K. Sampath, and P. V. Hauschka, Human osteogenic protein-1 induces chondroblastics, osteoblastics and/or adipocyte differentiation of clonal murine target cells. Exp. Cell Res. 222, 38 – 47 (1996). 303. S. E. Harris, J. Q. Feng, M. A. Harris, N. Ghosh-Choudhury, M. R. Dallas, J. M. Wozney, and G. R. Mundy, Recombinant bone morphogenetic protein 2 accelerates bone cell differentiation and stimulates BMP-2 mRNA expression and BMP-2 promoter activity in primary fetal rat calvarial osteoblast cultures. Mol. Cell. Differ. 3, 137 – 155 (1995). 304. D. Chen, M. A. Harris, G. Rossini, C. R. Dunstan, S. L. Dallas, J. Q. Feng, G. R. Mundy, and S. E. Harris, Bone morphogenetic protein 2 (BMP-2) enhances BMP-3, BMP-4 and bone cell differentiation marker gene expression during the induction of mineralized bone matrix formation in cultures of fetal rat calvarial osteoblasts. Calcif. Tissue Int. 60, 283 – 290 (1997). 305. A. Yamaguchi, T. Katagiri, T. Ikeda, J. M. Wozney, V. Rosen, E. A. Wang, A. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell. Biol. 113, 681 – 687 (1991). 306. Y. Takuwa, C. Ohse, E. A. Wang, J. M. Wozney, and K. Yamashita, Bone morphogenetic protein-2 stimulates alkaline phosphatase activity and collagen synthesis in cultured osteoblastic cells. Biochem. Biophys. Res. Commun. 174, 96 – 101 (1991). 307. R. S. Thies, M. Bauduy, B. A. Ashton, L. Kurtzberg, J. M. Wozney, and V. Rosen, Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130, 1318 – 1324 (1992). 308. T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J. M. Wozney, A. Fujisawa-Sehara, and T. Suda, Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J. Cell Biol. 127, 1755 – 1766 (1994). 309. J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). 310. S. D. Boden, G. Hair, L. Titus, M. Racine, K. Mccuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 311. D. Rickard, L. C. Hofbauer, S. Bonde, F. Gori, T. C. Spelsberg, and B. L. Riggs, Bone morphogenetic protein-6 production in human osteoblastic cell lines: selective regulation by estrogen. J. Clin. Invest. 101, 413 – 422 (1998). 312. C. D. Grimsbud, P. R. Romano, M. DSouza, J. D. Puzas, P. R. Reynolds, R. N. Rosier, and R. J. OKeefe, BMP-6 is an autocrine stimulator of chondrocyte differentiation. J. Bone Miner. Res. 14, 475 – 482 (1999).

GREGORY R. MUNDY ET AL. 313. K. Miyama, G. Yamada, T. S. Yamamoto, C. Takagi, K. Miyado, M. Sakai, N. Ueno, and H. Shibuya, A BMP-inducible gene, Dlx5, regulates osteoblast differentiation and mesoderm induction. Dev. Biol. 208, 123 – 133 (1999). 314. S. E. Harris, M. A. Harris, N. Ghosh-Choudhury, C. Xu, D. Chen, L. Wang, J. Rubenstein, and G. R. Mundy, Dlx2 homeobox gene is rapidly induced by BMP-2 and directly regulates BMP-2 transcription through a cell-specific enhancer region. Bone 23, S197 (1998). 315. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. Gao, M. Inada, M. Sato, R. Okamoto, R. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 316. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor Cbfa1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 317. F. Otto, A. P. Thornell, T. Crompton, A. Denze, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. Selby, and M. Owen, Cbfa 1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). 318. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa 1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 319. M. Enomoto-Iwamoto, M. Iwamoto, Y. Mukudai, Y. Kawakami, T. Nohno, Y. Higuchi, S. Takemoto, H. Ohuchi, S. Noji, and K. Kurisu, Bone morphogenetic protein signaling is required for maintenance of differentiated phenotype, control of proliferation and hypertrophy in chondrocytes. J. Cell Biol. 140, 409 – 418 (1998). 320. S. E. Gitelman, M. S. Kobrin, J. Q. Ye, A. R. Lopez, A. Lee, and R. Derynck, Recombinant Vgr-1/BMP-6-expressing tumors induce fibrosis and endochondral bone formation in vitro. J. Cell Biol. 126, 1595 – 1609 (1994). 321. T. K. Sampath, J. C. Maliakal, P. V. Hauschka, W. K. Jones, H. Sasak, R. F. Tucker, K. H. White, J. E. Coughlin, M. M. Tucker, and R. H. Pang, Recombinant human osteogenic protein-1 (hOP-1) induces bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol. Chem. 267, 20352 – 20362 (1992). 322. H. Zhang and A. Bradley, Mice deficient for BMP-2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 323. G. Winnier, M. Blessing, P. A. Labosky, and B. L. M. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 324. G. Luo, C. Hofmann, A. L. J. J. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). 325. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 326. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF-superfamily. Nature 368, 639 – 643 (1994). 327. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGF superfamily member. Nature Gene. 12, 315 – 317 (1996). 328. Y. Kawakami, T. Ishikawa, M. Shimabara, N. Tanda, M. EnomotoIwamoto, M. Iwamoto, T. Kuwana, A. Ueki, S. Noji, and T. Nohno,

CHAPTER 13 Cytokines and Bone Remodeling

329.

330.

331.

332.

333.

334.

335.

336.

337.

BMP signaling during bone pattern determination in the developing limb. Development 122, 3557 – 3566 (1996). S. Yi, A. Daluiski, M. Zhao, V. Rosen, and K. Lyons, The bone morphogenetic protein receptor type IB is required for the development of digits and the maintenance of articular cartilage. 45th Annual Meeting of the Orthopaedic Research Society, Anaheim, CA (1999). [Abstract 61] H. Nishitoh, H. Ichijo, M. Kimura, T. Matsumoto, F. Makishima, A. Yamaguchi, H. Yamashita, S. Enomoto, and K. Miyazono, Identification of type I and type II serine/threonine kinase receptors for growth/differentiation factor-5. J. Biol. Chem. 271, 21345 – 21352 (1996). Y. Mishina, A. Suzuki, N. Ueno, and R. B. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). B. B. Koenig, J. S. Cook, D. H. Wolsing, J. Ting, J. P. Tiesman, P. E. Correa, C. A. Olson, A. L. Pecquet, F. Ventura, R. A. Grant, G. Chen, J. L. Wrana, J. Massague, and J. S. Rosenbaum, Characterization and cloning of a receptor for BMP-2 and BMP-4 from NIH 3T3 cells. Mol. Cell. Biol. 14, 5961 – 5974 (1994). P. ten Dijke, H. Yamashita, T. K. Sampath, A. H. Reddi, M. Estevez, D. L. Riddle, H. Ichijo, C. H. Heldin, and K. Miyazono, Identification of type I receptors for osteogenic protein-1 and bone morphogenetic protein-4. J. Biol. Chem. 269, 16985 – 16988 (1994). M. Kawabata, A. Chytil, and H. L. Moses, Cloning of a novel type II serine/threonine kinase receptor through interaction with the type I transforming growth factor- receptor. J. Biol. Chem. 270, 5625 – 5630 (1995). T. Nohno, T. Ishikawa, T. Saito, K. Hosokawa, S. Noji, D. H. Wolsing, and J. S. Rosenbaum, Identification of a human type II receptor for bone morphogenetic protein-4 that forms differential heteromeric complexes with bone morphogenetic protein type I receptors. J. Biol. Chem. 270, 22522 – 22526 (1995). B. L. Rosenzweig, T. Imamura, T. Okadome, G. N. Cox, H. Yamashita, P. ten Dijke, C. Heldin, and K. Miyazono, Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc. Natl. Acad. Sci. USA 92, 7632 – 7636 (1995). H. Yamashita, P. ten Dijke, D. Huylebroeck, T. K. Sampath, M. Andries, J. S. Smith, C. H. Heldin, and K. Miyazono, Osteogenic protein-1 binds to activin type II receptors and induces certain activinlike effects. J. Cell. Biol. 130, 217 – 226 (1995).

403 338. M. Macias-Silva, P. A. Hoodless, S. J. Tang, M. Buchwald, and J. L. Wrana, Specific activation of Smad 1 signaling pathways by the BMP7 type I receptor, ALK2. J. Biol. Chem. 273, 25628 – 25636 (1998). 339. J. Carcamo, A. Zentella, and J. Massague, Disruption of transforming growth factor b signaling by a mutation that prevents transphosphorylation within the receptor complex. Mol. Cell. Biol. 15, 1573 – 1581 (1995). 340. R. Wieser, J. L. Wrana, and J. Massague, GS domain mutations that constitutively activeate TbR-I, the downstream signaling component in the TGF- receptor complex. EMBO J. 14, 2199 – 2208 (1995). 341. P. A. Hoodless, T. Haerry, S. Abdollah, M. Stapleton, M. B. OConnor, L. Attisano, and J. L. Wrana, MADR1, a MAD-related protein that functions in BMP2 signaling pathways. Cell 85, 489 – 500 (1996). 342. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP2-induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 343. Y. Chen, A. Bhushan, and W. Vale, Smad 8 mediates the signaling of the receptor serine kinase. Proc. Natl. Acad. Sci. USA 94, 12938 – 12943 (1997). 344. Y. Shi, Y. F. Wang, L. Jayaraman, H. Yang, J. Massague, and N. P. Pavletich, Crystal structure of Smad MH1 domain bound to DNA: insights on DNA binding in TGF signaling. Cell 94, 585 – 594 (1998). 345. R. S. Lo, Y. G. Chen, Y. Shi, N. Pavletich, and J. Massague, The L3 loop: A structural motif determining specific interactions between SMAD proteins and TGF receptors. EMBO J. 17, 996 – 1005 (1998). 346. X. Chen, M. J. Rubock, and M. Whitman, A transcriptional partner of MAD proteins in TGF- signaling. Nature 383, 691 – 696 (1996). 347. X. Chen, E. Weisberg, V. Fridmacher, M. Watanabe, G. Naco, and M. Whitman, Smad 4 and FAST-1 in the assembly of activin-responsive factor. Nature 389, 85 – 89 (1997). 348. X. Shi, X. Yang, D. Chen, Z. Chang, and X. Cao, Smad 1 interacts with homeobox DNA-binding proteins in bone morphogenetic protein signaling. J. Biol. Chem. 274, 13711 – 13717 (1999). 349. Y. G. Tueh, D. P. Gardner, and C. Kappen, Evidence for regulation of cartilage differentiation homeobox gene Hoxc-8. Proc. Natl. Acad. Sci. USA 95, 9956 – 9961 (1998). 350. D. Wotlon, R. S. Lo, S. Lee, and J. Massague, A Smad transcriptional corepressor. Cell 97, 29 – 39 (1999).

CHAPTER 14

Bone Growth Factors XUEZHONG QIN, REINHARD GYSIN, SUBBURAMAN MOHAN, AND DAVID J. BAYLINK Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357

I. Introduction II. Bone Growth Factors

III. Conclusions References

factor availability to cells of osteoblastic lineage is through the release of growth factors from bone during bone resorption [11]. Although this latter growth factor-releasing process is part of the mechanism of the coupling of bone formation to resorption, which is itself a very important topic, we will not deal with this process, as it is covered in detail in Chapter 12. Therefore, this chapter focuses on the molecular mechanisms of action, the in vitro and in vivo functions, and the potential therapeutic relevance of each growth factor. We will also provide the latest information available from knockout, transgenic, and natural mutant studies on these growth factors, as this information is important in bringing to our attention potential adverse effects of a growth factor on other tissues, in addition to its beneficial effect on bone formation. Finally, we will briefly comment on the potential role of a given growth factor in the pathogenesis of osteoporosis.

I. INTRODUCTION Since the early discovery by Marshal Urist [1] that demineralized bone extract could induce bone formation ectopically, scores of papers have been published demonstrating that bone is a storehouse for growth factors that are capable of stimulating both osteoblast cell proliferation and osteoblast differentiation [2 – 10]. Moreover, the finding that osteoblast line cells secrete factors that are mitogenic to bone cells in vitro led to studies on the identification of the active principle from culture media conditioned by a variety of untransformed and transformed bone cell lines in vitro [2,3,5 – 7,9]. These findings showing mitogenic activity in bone or in conditioned medium (CM) from osteoblastic cultures have been extended to demonstrate specific classes of growth factors produced by osteoblasts, including transforming growth factor- (TGF-), bone morphogenetic proteins (BMPs), insulin-like growth factors (IGFs), and fibroblast growth factors (FGFs). A comprehensive treatise on bone growth factors would include regulation of growth factor secretion, together with molecular mechanisms of action and in vitro and in vivo functions. Regarding growth factor secretion, osteoblasts along the trabecular bone and the endosteal surfaces are exposed to growth factors secreted by cells of osteoblast lineage and osteoclast lineage, as well as bone marrow cells. In addition, a unique aspect of growth

OSTEOPOROSIS, SECOND EDITION VOLUME 1

II. BONE GROWTH FACTORS A. The TGF- System To date, five TGF- isoforms have been identified [12] (also see Chapter 13). TGF-1, 2, and 3 are present in mammalian species and TGF-4 and TGF-5 are present only in chicken and Xenopus, respectively. Two heterodimers

405

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

406 termed TGF-1,2 and TGF-2,3 have also been identified in bovine bone extracts [4]. Of the three members of the TGF- family known to be produced by various mammalian tissues, TGF-1 appears to be the major member produced by human osteoblasts and stored in bone. TGF-2 is also known to be produced by human bone cells and stored in bone but to a lesser extent [8,13,14]. Active TGF- exists as disulfidelinked dimers comprised of identical 12,500-Dalton subunits. Studies from TGF- transgenic and knockout experiments indicate that TGF-s are multifunctional genes, although their effects on skeletal tissues appear to be similar. This section focuses on studies related to the in vitro and in vivo effects of TGF- on bone metabolism, as well as the mechanism of TGF- action. 1. IN VITRO STUDIES TGF- produces diverse effects on skeletal tissues and cells in vitro. Studies on the effects of TGF- on cultured osteoblasts and bones in organ cultures have led to a large number of conflicting reports. TGF- stimulates prostaglandin-mediated bone resorption in mouse bone organ culture [15] and inhibits proliferation of isolated fetal rat calvaria cells at low cell density [16], as well as clonal mouse calvaria-derived cells (MC3T3E1) [17]. TGF- also inhibits osteoblast differentiation in MC3T3E1 cells [17] and normal rat osteoblasts [18]. In contrast, TGF- stimulates cell growth in fetal rat calvaria [19]. TGF- increases protein synthesis [19] and collagen expression in fetal rat calvaria cells [16] and stimulates alkaline phosphatase (ALP) in rat osteosarcoma cells [20]. Treatment of rat osteoblastic ROS 17/2.8 cells with a TGF- increases ALP activity in a dose-dependent fashion with an ED50 of approximately 0.2 ng/ml. This effect is more profound during the logarithmic growth than at confluence [17]. Chondrocytes also respond to TGF- in vitro by increasing ALP activity in a biphasic manner [21 – 23] with a maximal response observed at a dose of 0.22 ng/ml [22]. However, TGF- inhibits glycosaminoglycan and collagen production in fully committed mature articular chondrocytes [24]. The discrepancy among studies regarding the effects of TGF- on bone cells appears to be related to the stages of cell differentiation, cell culture conditions, and TGF- doses, as well as the variables that are measured. Differences in TGF- receptor signaling could also contribute to the observed differences in target cell response to TGF-. In general, both chondrocytes and osteoblasts appear to be more sensitive to TGF- at their early developmental stages. Although the effects of TGF- in vitro on bone cells are controversial, TGF- has been shown to be a strong and consistent bone-inductive molecule in vivo (see Section II,A,4). The effect of TGF- on bone cells can be modulated by other factors. For example, TGF- and 1,25(OH)2D3 synergistically increase ALP activity in normal human

XUEZHONG QIN ET AL.

osteoblasts [8,14] and ALP activity and collagen synthesis in rat chondrocytes [22]. Yaeger et al. [25] found that IGFI, insulin, or TGF-s alone did not stimulate induction of aggrecan and type II collagen expression by adult human articular chondrocytes. However, TGF-1 or TGF-2, together with IGF-I or insulin, strongly induces the expression of these proteins. Similarly, acidic FGF (aFGF) or basic FGF (bFGF) alone is unable to induce odontoblast differentiation, whereas both aFGF and bFGF act synergistically with TGF-1 or IGF-I to induce odotoblast differentiation [26]. These data suggest that responses of cells to exogenously added TGF- may vary depending on the type and amount of growth factors produced endogenously by cells in vitro. 2. MECHANISM OF ACTION TGF- has been known to regulate the expression of a number of genes associated with osteoblast activity [8,27 – 29]. TGF- action is mediated through its binding to a distinct heteromeric receptor complex, including type I and type II serine/threonine kinase type receptors. Activation of the receptor complex is initiated upon the binding of the ligand to type II receptor, which then recruits, phosphorylates, and activates the type I receptor [12,30,31]. The activated type I receptor propagates the signal to downstream targets, including the Smad proteins [12,30,31]. To date, at least seven Smads have been identified; their potential role in mediating TGF- action is illustrated in Fig. 1. Understanding the mechanism of TGF- action has been enhanced by studies on TGF- regulation of transcription factors and the interaction of these transcription factors with Smads. Over the last decade, several TGF- and activin-responsive elements, including AP-1, Sp-1, and CTF/NF-1-binding sites, have been identified in the promoters of various genes [30,31]. Among the various TGF-regulated transcription factors, AP-1, a heterodimer of Fos and Jun proteins, has been the most extensively studied, as it clearly plays an important role in the regulation of proliferation and differentiation of osteoblast cell lines [28,32,33]. AP-1-regulated promoters are transcriptionally activated by TGF- [34 – 36]. This activation appears to be mediated through the TGF--stimulated c-Jun/c-Fos expression [37]. Studies have provided evidence for an interaction between AP-1 and Smads in mediating TGF- action [36,38]. Moreover, the osteoblast-specific transcription factor, core-binding factor (Cbfa1), has been suggested to be involved in Smad-mediated TGF- signaling in bone cells [39]. Thus, TGF- effects on target cells may, in part, be mediated via the activation of several known transcription factors. In addition to the AP-1-mediated PKC pathway, previous studies also suggest that TGF--induced gene expression may involve the MAP kinase pathway [30]. It is anticipated that identification of additional TGF--regulated

407

CHAPTER 14 Bone Growth Factors

Proposed model for the role of Smads in TGF- signaling pathway. Binding of a TGF family member to its type II receptor in concert with a type I receptor leads to formation of a receptor complex and phosphorylation of the type I receptor. The activated type I receptor subsequently phosphorylates Smad-2 or Smad-3 (this step can be inhibited by Smad 6/7). Upon phosphorylation, the Smad-2 or Smad-3 homodimer associates with the Smad-4 homodimer to form a heterohexamer that translocates into the nucleus. The Smad complex then activates transcription of target genes through an intermediary transcription factor or by binding to DNA directly [30] (reproduced with permission).

FIGURE 1

transcription factors in these signaling pathways will further advance our understanding of the molecular mechanism of TGF- action in bone. 3. TRANSGENIC/KNOCKOUT STUDIES To date, null mice have been generated for each of the three TGF- isoforms [40 – 43]. TGF-1 knockout mice develop a rapid wasting syndrome with excessive inflammatory response and die by 3 – 4 weeks of age [40,41]. The longitudinal growth and total mineral content are decreased in these knockout mice [44]. In TGF-1 transgenic mice, overexpression of TGF-1 increases plasma TGF-1 concentration and causes glomerulosclerosis [45]. Effects on skeletal tissues in these mice have not been examined. TGF-2 knockout mice exhibit perinatal mortality and a wide range of developmental defects, including some skeletal tissues, such as the spinal column [43]. Because there is little phenotypic overlap between TGF-1 and TGF-2 null mice, these two isoforms appear to have different functions. In contrast to the positive effect of TGF-2 on bone formation via local administration (see Section II,A,4), overexpression of TGF-2 in osteoblasts using the osteocalcin

promoter in mice resulted in progressive bone loss associated with increases in osteoblastic matrix deposition and osteoclastic bone resorption [46]. TGF-3 knockout mice exhibit an incomplete penetrant failure of the palatal shelves to fuse, leading to cleft palate [42,47]. Results from these studies demonstrate that TGF-s are multifunctional genes and that different forms of TGF- may have distinct biological functions. In TGF- knockout mice, one would expect to see a reduction in bone formation based on the in vivo studies described later. However, such effects have not been obvious. Furthermore, instead of producing an increase in bone formation, the overexpression of TGF-2 in osteoblasts led to an increase in bone resorption (see earlier discussion). This latter observation may not be relevant to the effects of TGF- in humans for the following reason. TGF- in mice is known to stimulate prostaglandin synthesis, which in turn is a strong osteolytic agent in mice. In contrast, TGF- does not stimulate prostaglandin synthesis in rats, and the effect of TGF- in rats is to decrease rather than increase bone resorption. Therefore, whereas the genetic studies using these mice provide important information on the in vivo function of TGF-s, these data need to

408 be interpreted with caution when they are applied to other species. 4. IN VIVO TREATMENT STUDIES The findings that TGF- family members are expressed during normal fracture healing and are actively produced by bone cells suggest that TGF- plays an important role in fracture healing [48,49]. Therefore, a number of studies using various animal model systems have been conducted to evaluate the effect of TGF- on bone formation. The results from representative studies are briefly discussed. a. Local Effects The calvaria model has been used frequently to evaluate the effect of TGF- on bone formation in either intact animals [50 – 52] or animals with critical sized defects [53 – 56]. Direct injection of TGF- onto the periostea of parietal bones of neonatal rats [50] or to the subcutaneous tissue overlying the calvaria of mice [51] increases the thickness of parietal bone. Because the stimulatory effect is partially inhibited by concomitant treatment with indomethacin, an inhibitor of prostaglandin synthesis, it is speculated that TGF-1 effects on bone formation may, in part, be mediated by prostaglandin synthesis in mice [51]. However, this contention is not supported by studies using rats [57]. More recently, Fujimoto, et al. [52] reported that subcutaneous administration of TGF-1 onto the parietal surface in rats had a negative effect on bone formation at the injection site, whereas direct injection of TGF-1 in the subperiosteal layer induced new bone formation. These data suggest that the TGF- effect can vary depending on the site. Studies using calvarial critical sized defect models have generally shown that TGF-, via appropriate carriers, can heal the defects in various animal models [53,56], although a lack of response has also been described [54]. The effect of local TGF- administration on bone regeneration in long bones has also been evaluated in intact animals [48,49] or animals with critical sized bone defects [58 – 62]. Subperiosteal injection of TGF- into the nonfractured rat femur results in mesenchymal cell proliferation and the subsequent initiation of chondrogenesis and intramembranous bone formation [48,49]. Short-term administration of TGF-1 directly into the bone marrow space attenuates the stimulation of osteoclastic resorption induced by ovariectomy in rats [60]. Repeated applications of TGF- to rabbit and rat tibial fractures for 6 weeks resulted in calluses that are larger and mechanically stronger than those in controls [58,59]. In contrast, a single injection of TGF-2 [60 or 600 ng) into the developing callus of rabbit tibial fractures 4 days after fracture had no significant effect on fracture healing within 5 to 14 days [61]. These studies suggest that multiple injections of TGF into the callus of the fractured bone are needed to enhance the healing process. Beck et al. [62] reported that

XUEZHONG QIN ET AL.

implantation of TGF-1 (1 g), together with autologous bone marrow, is as effective as the implantation of autogenous cortical bone graft in healing segmental defects in the radius of rabbits, suggesting that bone marrow may serve as an ideal carrier for TGF-. The effect of TGF- on bone formation in other skeletal sites has also been evaluated [56,63,64]. For example, implantation of TGF-1 in a 5-mm critical size defect in the rat mandible results in a dose-dependent (0.1 to 20 g) bone bridging at both 12 and 24 days [56]. In contrast, TGF-1 has a negative effect on bone regeneration when the access of cells from the periosteum to the defect is prevented by microporous membrane. Wikesjo et al [63] reported that implantation of 20 g rhTGF-1 in CaCO3/hydroxyethyl starch carrier into periodontal defects in dogs is not effective in regenerating bone. The authors concluded that rhTGF-1 has a limited potential in periodontal repair. In contrast, local TGF-1 administration has been shown to be effective in the stimulation of articular cartilage repair in mice with arthritis induced by zymosan injection [64]. Overall, the studies just described demonstrate that TGF-, via local administration, is effective in bone regeneration at various skeletal sites in small experimental animals. Whether TGF- is effective in fracture healing in large animals, such as nonhuman primates, needs to be evaluated in the future before its application to patients with large bone fractures. b. Systemic Effects Previous studies have shown that systemic TGF- administration also enhances bone formation [65 – 68]. In ovariectomized rats, daily administration of TGF-2 for 5 weeks by subcutaneous injections stimulates bone formation and prevents ovariectomy-induced bone loss [65]. Short-term systemic administration of rhTGF-2 (5 or 14 days) also stimulates cancellous bone formation in both juvenile and adult rats [66]. Histomorphometric analysis revealed that TGF-2 treatment did not affect the number of osteoclasts or the number of osteoclast nuclei per cell. Consistent with these anabolic effects, Zerath et al. [67] found that treatment of rats with TGF-2 at a dose of 2 g/kg/day prevented the decrease in chondrocyte number and proliferative zone growth in the proliferative zone induced by unloading. Gazit et al. [68] reported that the effect of systemic injection of TGF-1 on bone formation in mice is affected by the age of the animals. Daily administration of TGF-1 at 0.5 or 5 g/day for 20 days into 24-month-old mice caused a significant increase in trabecular bone volume, bone formation, and mineral apposition rate and enhanced fracture healing. However, these stimulatory effects are either not seen or are less significant in younger mice (4 weeks old). The studies just summarized suggest that TGF- is unique in that it increases bone formation and, simultaneously, decreases bone resorption. However, the potential

409

CHAPTER 14 Bone Growth Factors

adverse effects associated with the systemic administration of TGF- have not been fully examined in these studies. It has been shown previously that systemic TGF-1 administration in mice causes severe anemia [69] and that overexpression of TGF-1 in transgenic mice leads to progressive glomerulosclerosis [45,70]. Thus, it will be essential to determine if the systemic administration of TGF- at an effective dose for promoting fracture healing has any significant systemic deleterious effects before the therapeutic potential of TGF- can be fully evaluated. 5. IMPAIRMENT IN TGF- SYSTEM AND PATHOGENESIS OF BONE LOSS Regarding the functional significance of growth factors, such as TGF-, stored in bone, it has been proposed that growth factors are released in a bioactive form during osteoclastic bone resorption to act in a paracrine manner on osteoblast precursors and mature osteoblasts to ensure sitespecific bone replacement [5,7]. Thus, if the extent of the refill of the resorption cavity depends partly on the amount of TGF- stored in bone, then the level of the TGF- sequestered in bone should vary in response to physiological and pathological conditions. Consistent with this argument, we have found an age-related decline in the amount of TGF- stored in the cortical bone of both men and women [71]. More recently, Gazit et al. [72] reported that the matrix of long bones of old mice contains significantly less TGF- than that of young mice. In addition, bone marrow osteoblast progenitor cells isolated from old mice produce less TGF- in vitro than those isolated from young mice. The molecular mechanisms responsible for this age-related decrease in TGF- synthesis can only be speculated upon at this time. In this regard, studies have shown that estrogen deficiency and vitamin D deficiency both resulted in reduced cortical TGF- content in rats [73,74]. In addition, treatment of bone cells with estradiol, testosterone, or 1,25(OH)2D3 has also been reported to increase TGF- production [73 – 76]. Taken together these findings suggest that the decreased serum concentrations of sex steroid hormones with advancing age could lead to a decrease in the expression of TGF- by osteoblasts and thus could contribute to the age-related decline in the skeletal concentration of TGF-. Because physical activity also seems to regulate TGF- production, the decrease in physical activity with aging could also contribute to the observed decrease in skeletal TGF- content with age [77]. In view of the findings that the amount of TGF- in human bone decreases with age and that the osteoblast production of TGF- is regulated by sex steroid hormones and physical activity, it is tempting to speculate that the uncoupling between bone formation and resorption in the elderly could, in part, be contributed by a decreased amount of growth factors, such as TGF-, secreted by osteoblasts for contemporary use and storage in bone for later use (Fig. 2).

B. The BMP System The discovery of BMPs stems from the findings that implantation of demineralized bone matrix at ectopic sites causes an induction of heterotopic (extraskeletal) bone development [1]. Subsequently, investigation on the identification of osteoinductive molecules [78] has led to the discovery of several BMPs, including BMP-2, BMP-3, and BMP-4 [79,80], as well as BMP-7 and BMP-8 [81,82]. To date, more than 15 BMPs have been identified [10,83 – 86]. More recently, Paralkar et al [87] cloned a novel member of the BMP family from prostate, designated prostate-derived factor (PDF), which shows a structural and functional relationship to the BMPs. All these BMPs, except BMP-1, belong to the TGF- superfamily and share significant sequence homology in the carboxy-terminal region with a conserved pattern of seven cysteine residues. BMPs are synthesized as precursor forms and are cleaved at the C-terminal region to yield mature proteins. Active BMPs exist as dimers formed by disulfide bond bridging [88]. Genetic studies suggest that BMPs are produced by multifunctional genes that control not only the growth and development of skeletal tissue, but also the morphogenesis and function of other tissues and organs [10,83 – 86]. This section focuses on current understanding of the role of BMPs in regulating bone cell functions and in vivo studies on BMPs in bone regeneration (also see Chapter 13). 1. IN VITRO STUDIES Recruitment of active osteoblasts is essential for the regeneration of bone at the site of osteoclastic bone resorption. This process is dependent on the continuous supply of mature osteoblasts differentiated from hematopoietic stem cells. In this regard, treatment with BMP-2 increases ALP activity and osteocalcin production in human bone marrow stromal cells [89,90] and stimulates ALP activity and collagen synthesis in primary cultures of fetal calvarial osteoblasts [91]. The stimulatory effect of BMP-2 on osteoblast differentiation can be enhanced by ascorbic acid [91,92]. Treatment of MC3T3-E1 cells with BMP-4 also increases ALP activity [93]. This effect is enhanced by IL-1 (IL-1 alone has no effect on ALP activity) but is suppressed by TNF-. Treatment of secondary rat calvarial cultures with antisense oligodeoxynucleotides to BMP-6 suppresses glucocorticoidinduced osteoblast differentiation [94]. This inhibition can be rescued by treatment with rhBMP-6. In fetal rat calvarial cell cultures, treatment with BMP-7 dose dependently increases bone nodule formation and osteoblast marker gene expression [95]. Erlacher et al. [96] compared the effect of BMP-7 with other BMP subfamily members, i.e., the cartilagederived morphogenetic protein (CDMP)-1 (BMP-15) and CDMP-2 (BMP-13) in stimulating osteogenic differentiation and chondrogenic differentiation. It was found that these three BMP members are equally potent in stimulating

410

XUEZHONG QIN ET AL.

Model of potential role of a deficiency of TGF- in the pathogenesis of bone loss. Based on previous findings that the production of TGF- by osteoblasts is increased by treatment with sex steroid hormones [74 – 76] and that the serum levels of sex steroid hormones decrease with age, it is postulated that the osteoblast cell production of TGF- decreases with advances in age. The decrease in osteoblast cell production of TGF- is a likely explanation for why the skeletal content of TGF- is decreases with age [71]. Because TGF- treatment increases bone formation parameters in vitro and in vivo and decreases osteoclast cell survival in vitro, the underproduction of TGF- could, in part, lead to the age-related uncoupling of bone formation to resorption.

FIGURE 2

chondrogenic differentiation, whereas BMP-7 is more potent than CDMPs in promoting osteogenic differentiation in both primary fetal chondrocytes and established osteoblast cell lines. Thus, while many of the BMPs examined thus far induce differentiated functions of osteoblasts, BMP members belonging to the CDMP subfamily are more potent in stimulating chondrogenic differentiation compared with other BMP subfamily members. Studies suggest that heterodimeric BMPs are more potent than homodimers in stimulating osteoblast differentiation [97 – 99]. Compared with the BMP-4 or BMP-7 homodimer, the BMP-4/7 heterodimer more potently stimulates osteocalcin production in MC3T3-E1 cells [97] and ALP activity in rat bone marrow stromal cells [98]. Coexpression of BMP-2 with BMP-7 in Chinese hamster ovary cells has yielded a BMP-2/7 heterodimer with a specific activity of 20-fold higher than BMP homodimers in stimulating ALP production [99]. Similar results have also been obtained with the BMP-2/6 heterodimer. In addition to these in vitro data, it has also been demonstrated that these BMP heterodimers are more potent in stimulating bone formation in vivo (see

Section II,B,5). Although these research findings are intriguing, the mechanism by which these BMP heterodimers are more potent than homodimers has not yet been determined. Taken together, these findings demonstrate that BMPs consistently stimulate osteoblast differentiation, as well as chondrocyte differentiation, from stem cells and early differentiated bone cells, and that these effects can be modulated by other factors, as well as BMP heterodimer formation. 2. MECHANISM OF ACTION The actions of BMPs are mediated through the BMP receptors (BMPR), which are structurally and functionally similar to the receptors for activin and TGF-. Two types of BMP serine/threonine kinase receptors, BMPR-I and BMPR-II, have been identified. Type I BMP receptors (BMPR-I) exist as two forms: BMPR-IA and BMPR-IB (BMPR-Is). In vitro studies suggest that optimal binding of BMPs requires the coordination of both BMPR-Is and BMPR-II [85,100]. It is generally accepted that BMPR-II activates BMPR-Is and that signals are mediated through the BMPR-Is [10,85].

411

CHAPTER 14 Bone Growth Factors

Understanding of BMP signaling pathways has been enhanced greatly by the discovery of a group of molecules named “Smads,” which mediate BMP/BMPR signaling in either a positive or a negative manner. Models of BMP intracellular signal transduction have been proposed by various investigators [10,84 – 86]. The Smad-mediated signaling pathway of BMP and TGF- (Fig. 1) is very similar, except that the activated BMPR-Is specifically phosphorylate Smad 1 or Smad 5 and, possibly, Smad 8 (refer to Section II,A). Similarly, phosphorylation of these Smads is inhibited and modulated by Smads 6 and 7. Phosphorylated Smad 1 or 5 interacts with Smad 4 and enters the nucleus to activate transcription of the early BMP response genes. Overexpression of mutant Smad 5 or DPC-4/Smad 4 has been shown to block BMP-2 induction of ALP [101]. The role of Smad 5 in BMP signaling is supported by the findings that Smad 5 knockout mice exhibit phenotypes similar to those of BMP-2 and BMP-4 knockout mice [102]. Although the nuclear targets of BMP signaling are largely unknown, studies suggest that homeodomain transcription factors such as T-cell leukemia homeobox genes (Tlx), muscle segment homeobox homologue genes (Msx), and distal-less homeobox genes (Dlx) are induced by BMPs. BMP-2 has been shown to rapidly activate Tlx-2 expression in mouse embryos [103]. BMP-4 can increase the expression of Dlx-1 and -2 in dental mesenchyme [104]. In fetal mouse calvaria organ culture, BMP-4 increases the expression of both Msx-1 and Msx-2, whereas FGF-4 increases only the expression of Msx-1 [105]. In addition to these homeodomain transcription factors, the osteoblastspecific transcription factor Cbfa1 is also suggested to be an important downstream mediator of BMPs, based on the finding that BMP-7 increases Cbfa1 expression in bone cells [106]. Further studies on the relationship between these BMP-regulated transcription factors and BMP-specific Smads will be essential to enhance our understanding of the mechanisms involved in BMP signaling. 3. BMP ANTAGONISTS Studies demonstrate that the biological action of BMPs can be modulated by BMP antagonists, such as noggin and chordin [107 – 109]. Noggin, originally identified in Xenopus embryos [110], blocks BMP-4-induced ALP activity in murine bone marrow stromal cells [107] and decreases the stimulatory effects of BMP-2, -4, and -6 on DNA synthesis, collagen synthesis, and ALP activity in rat osteoblasts [111]. Studies on the mechanism by which noggin inhibits BMP activity revealed that noggin binds to BMP-2 and BMP-4 with high affinity, but not to TGF- [107]. In vitro, noggin prevents BMP-4 from binding to its receptor [107]. Noggin knockout mice are lethal and reveal severe developmental skeletal abnormalities [112]. The excess BMP activity in noggin knockout mice increases the recruitment of cells to cartilage, expanding the cartilage at the expense

of other tissues and causing larger growth plates [112]. Consistent with the idea that noggin may act to prevent excess BMP activity, Gazzerro et al. [111] found that treatment of rat osteoblasts with BMP-2 dramatically increased noggin expression as early as 2 h. Chordin is another BMP antagonist [109,113]. Similar to noggin, chordin binds to BMP-2 and -4, but not to activin or TGF- [109]. Chordin prevents the binding of BMP-4 to its receptors and, consequently, inhibits BMP-4-stimulated ALP activity in the multipotential mesodermal cell line 10T1/2 . The first mammalian form of chordin cDNA has been cloned from mice [114]. It has been speculated that chordin may play a major role during gastrulation of the mammalian embryo. The role of chordin in bone metabolism has not yet been established. Based on in vitro and in vivo findings that noggin and chordin are potent inhibitors of BMP actions, it can be speculated that the balance between the production of BMP and the production of these BMP antagonists may be crucial for the regulation of new bone formation. Thus, a better understanding of the mechanisms controlling the synthesis of these antagonists will be important in terms of improving BMP therapy or new drug discovery using inhibitors of BMP antagonists in the treatment of bone loss. 4. TRANSGENIC/KNOCKOUT/MUTANT STUDIES Characterization of the abnormalities associated with BMP gene mutations has provided important information on the in vivo role of BMPs (Table 1). BMP-2 and BMP-4 knockout mice have defects in mesoderm formation and die at an early stage of embryonic development [115,116]. BMP-4 heterozygous mice show abnormalities in various organs/tissues, including hindlimb [117]. Natural mutations in the BMP-5 gene in mice have short ears and exhibit abnormalities in the skull and axial parts of the skeleton [118]. BMP-6 knockout mice are viable and fertile with normal skeletal development at birth [119]. However, in late gestation embryos, a delay in ossification of the developing sternum is revealed, and these defects are slightly exacerbated by the disruption of both BMP-5 and BMP-6 genes. BMP-7 knockout mice die shortly after birth with developmental defects in the kidney, eyes, and skeleton [120]. BMP-2/7 and BMP-5/7 double heterozygous mice do not reveal any abnormalities, whereas BMP-4/7 double heterozygotes develop minor skeletal defects in the rib cage and the distal part of the limbs [121]. BMP-8B (not present in humans) knockout mice exhibit germ cell deficiency and infertility [122]. Natural mutations in growth and differentiation factor (GDF)-5/CDMP-1 gene cause brachypodism in mice [123] and chondrodysplasia in humans [124]. GDF-8 knockout mice exhibit increased skeletal muscle mass [125]. GDF-9 knockout mice have defects in ovarian folliculogenesis [126]. GDF-10 (BMP-3B) knockout mice do not exhibit apparent abnormalities [127]. It is suggested that GDF-10 may be functionally redundant with other

412

XUEZHONG QIN ET AL.

TABLE 1

Abnormality Associated with BMP Gene Mutations

Genetic pertubation

Phenotype

Ref.

Single BMP knockout BMP-2

Nonviable; defects in mesoderm formation and cardiac development

[115]

BMP-4

Death at early embryo, defect in mesodermal development

[116]

BMP-6

Viable, fertile, but delay in sternum ossification

[119]

BMP-7

Die shortly after birth, defect in development of multiple tissues, including bone

[120]

BMP-8B

Germ cell deficiency and infertility

[122]

GDF-8

Excessive muscle development

[125]

GDF-10

No apparent abnormalities

[127]

Similar to BMP-6 knockout

[119]

BMP-2 and BMP-7

No apparent abnormalities

[121]

BMP-5 and BMP-7

No apparent abnormalities

[121]

BMP-4 and BMP-7

Minor defect in rib cage and limbs

[121]

Double BMP knockout Homozygous BMP-5 and BMP-6 Heterozygotes

Natural mutants BMP-5

Short ear mice, abnormalities in skull and axial parts of skeleton

[118]

GDF-5 (CDMP-1)

Brachypodism in mice and chondrodysplasia in human

[123,124]

growth factor-like molecules. These genetic studies suggest that some BMPs may have distinct functions, whereas others may functionally overlap each other or have redundant functions with other growth factors. Finally, these studies emphasize that BMPs control the development of not only skeleton but also other tissues. 5. IN VIVO TREATMENT STUDIES a. Local Effects Studies of BMP actions in vivo have concentrated on the development of local delivery systems for the BMPs. Materials that have frequently served as BMP carriers include hydroxyapatite, biodegradable polymers, titanium, and fibrous glass membrane [128 – 133]. By using appropriate delivery systems, BMPs have been shown to be effective in bone regeneration in segmental fracture [134 – 136] and spinal fusion models in experimental animals [137,138]. The following briefly summarizes the findings of several representative studies. Cook et al. [135] evaluated the effect of rhBMP-7 (OP-1) using bovine collagen as a carrier, on bone formation in segmental defects in monkeys. The majority of the ulnae treated with 0.25 to 2 mg BMP-7/400 mg bovine collagen exhibited complete healing at 6 to 8 weeks. Boden et al. [138] evaluated the efficacy of using a ceramic material (hydroxyapatite-tri-calcium phosphate) loaded with rhBMP-2 as a bone graft substitute in a primate model of intertransverse spinal fusion. All monkeys treated with ceramic

blocks loaded with BMP-2 (6 to 12 mg) achieved solid fusion, whereas no fusion occurred in monkeys treated with an autogenous iliac crest bone graft. In a rabbit critical-sized defect model, insertion of a porous poly DL-lactic acid implant soaked with rhBMP-2 dose and time dependently induced new bone formation in the 20-mm-long defect [136]. By 8 weeks postimplantation, 35 or 70 g BMP-2 restored the cortices and marrow elements. When adsorbed onto porous hydroxyapatite, BMP-3 has been shown to induce rapid bone formation in calvarial defects in baboons [128]. BMP-2 and BMP-4 have also been shown to induce dentin formation by stimulating the differentiation of dental pulp cells into odontoblasts [130,139]. These studies demonstrate that BMP, when administered with a suitable carrier, effectively regenerates bone at various skeletal sites. It has been reported that a combination of BMP-2 and TGF- increases bone formation more potently than BMP-2 alone in ceramic bovine bone implanted into the thigh muscle of mice [140]. This synergistic effect is suggested to be due to their different preferential effects on cells at different stages. Ono et al. [141] reported that PGE1 promoted the osteogenic effect of BMP-2 in a rabbit model using porous hydroxyapatite ceramic pellets as the carrier. This effect was suggested to be mediated indirectly through PGE1-induced IL-6 and TGF- production. Thus, it appears that the in vivo bone-forming activity of BMPs can be mediated by other locally produced osteoregulatory factors.

413

CHAPTER 14 Bone Growth Factors

Interestingly, it has been shown that the osteoinductive effect of BMPs in vivo can be enhanced by heterodimer formation. Implantation of Xenopus BMP-4 or -7, together with a pure collagen carrier in rats, dose dependently induces new bone formation, but administration of the BMP4/7 heterodimer is more effective [98]. In addition, Israel et al. [99] reported that coexpression of various BMPs in mammalian cells led to BMP heterodimer formation. The purified BMP-2/7 or BMP-2/6 heterodimer is 5- to 10-fold more potent than BMP homodimers in inducing cartilage and bone formation [99]. These in vivo data are, thus, consistent with in vitro data showing that BMP heterodimers are more potent in stimulating osteoblast differentiation (see Section II,B,1). Future studies should emphasize the mechanism by which these BMP heterodimers exhibit more potent osteoinductive activity. Efforts to identify the molecules that stimulate BMP-2 promoter activity have led to the discovery that “statins” (which are the commonly prescribed HMG-lA reductase inhibition drugs for reducing serum cholesterol concentrations and lowering the risk of heart attack) are stimulators of BMP-2 production [142]. Statins, such as lovostatin, increased bone formation when injected subcutaneously over the calvaria of mice and increased cancellous bone volume when administered orally to rats. Whether the stimulation of bone formation by statins is solely due to their stimulatory effect on BMP-2 production remains to be studied. b. Gene Therapy Studies described earlier have demonstrated the clinical utility of BMPs in bone regeneration. However, these studies do not completely mimic the clinical situation in which defects are often much larger. Therefore, a single application of BMP may not be very effective in these situations. Moreover, this approach has been limited by the lack of availability of an ideal carrier that can provide a sustained release of BMP. Thus, BMP gene therapy offers an alternative strategy. In this regard, Fang et al. [143] reported that the implantation of degradable matrices (mainly bovine collagen) soaked with BMP4 expression plasmid led to functional union of the segmental defect in rats 9 weeks postimplantation. The fracture healing is facilitated by codelivery of the PTH(1 – 84) gene. However, further studies are warranted to evaluate the reproducibility of this strategy as the use of plasmid, in general, cannot achieve high transfection efficiency and only provides a transient expression of the therapeutic genes. Subsequently, it has been reported that the delivery of the BMP-2 gene using either the ex vivo [144 – 146] or the in vivo [147,148] approach is able to induce bone formation in muscle or heal a bone fracture. However, a significant osteoinductive effect is only observed in immunodeficient animals. The lack of significant effect in immunocompetent animals is due to the production of the adenoviral proteins encoded by the vectors,

which causes a severe immune response that ultimately inactivates the expression system. Therefore, the use of first-generation adenoviral vector (E1 and E3 deleted) to deliver BMP genes in clinical bone repair may have very limited potential. Despite the various problems associated with the current strategies, it is conceivable that BMP gene therapy may become a powerful therapeutic tool in bone repair with the development of new vector systems, such as the “gutless” adenoviral vector that does not cause a severe immune response. In addition, application of BMP gene therapy can be improved further if long-term, regulatable gene expression and systemic delivery with tissue-specific expression can be achieved in the future.

C. The IGF System IGFs represent the most abundant growth factors produced by osteoblasts and stored in bone. Previous studies have provided strong support for the concept that IGFs function in an autocrine/paracrine manner to regulate bone metabolism [149 – 152]. It has been demonstrated that the actions of IGFs on bone metabolism are modulated by multiple regulatory components of the IGF system, which includes IGF-I and IGF-II, type I and type II IGF receptors, at least six high-affinity IGF-binding proteins (IGFBP-1 to -6), and IGFBP proteases. In this regard, understanding the role of each component of the IGF system in bone metabolism is critical to understanding the pathogenic role of the IGF system in the development of osteoporosis and, also, in improving the utility of IGFs as a future therapy in the treatment of osteoporosis. Therefore, our primary focus in this section is on the scientific advancement related to studying the action of each IGF system component in bone. 1. IGFS a. In Vitro Studies IGFs have been shown to stimulate the proliferation of serum-free cultures of osteoblasts derived from the bone of various species [149 – 152]. Approximately 40 – 50% of total basal cell proliferation in serumfree bone cell culture has been shown to be attributable to IGFs [153]. IGFs also promote osteoblast differentiation based on the findings that treatment with IGFs in bone cell cultures increases the activity of ALP [153 – 156]. Hill et al. [157] demonstrated that IGF-I and IGF-II are among a few growth factors that strongly inhibit cell apoptosis in primary mouse osteoblast cultures. The antiapoptotic effect of IGFs is enhanced by PDGF, which by itself has no effect on osteoblast survival. However, Kawakami et al. [158] reported that IGF-I potentiated the Fas-mediated apoptosis of human primary osteoblasts and MG63 osteosarcoma cells. These data contrast to findings in our laboratory that IGF-II

414 dose and time dependently inhibited both basal and okadaic acid- or TNF-induced apoptosis in primary human osteoblasts and human osteosarcoma cells [159]. IGFs also stimulate collagen synthesis and decrease collagen degradation [153,155,160,161], which is consistent with the inhibitory effect of IGF-I on interstitial collagenase expression in rat osteoblast cultures [161]. These data suggest that IGFs play an important role in osteoblast proliferation, differentiation, and apoptosis. New evidence suggests that IGFs may increase osteoclastogenesis [162]. PTH-induced osteoclastogenesis and GH-induced osteoclastic resorption in vitro can be blocked significantly by IGF-I neutralization antibody [163,164]. In addition, Hou et al. [165] reported that the induction of nuclear fragmentation in serum-depleted cultures of purified mature osteoclasts was inhibited dose dependently by IGF-I in the picomolar range but not by 1 nM insulin. The stimulatory effect of IGFs on the formation and activity of both osteoblasts and osteoclasts may, in part, contribute to IGFinduced bone formation, as well as bone resorption in vivo under certain conditions. b. Mechanism of Action Although IGFs clearly play an important role in bone metabolism, the molecular mechanism by which they exert their biological actions in bone is just beginning to be understood. Accordingly, studies using cell model systems have begun to elucidate the IGF signaling pathways [166,167], and may provide important information for future studies in bone. Because it is well accepted that the major biological actions of both IGF-I and IGF-II are mediated through the type I IGF receptor, studies on the IGF signaling mechanism have predominantly been focused to that receptor and its downstream mediators. The type I IGF receptor controls cell proliferation by at least three different means: it is mitogenic, it causes transformation, and it protects cells from apoptosis [168,169]. Studies suggest that the functional domain in the type I receptor that controls each of these cellular responses is different. The domain conferring the transforming signal is located in the C terminus of the receptor, which is separated from the mitogenic signal located within the tyrosine kinase domain [168,170,171]. Type I IGF receptor mutants with disrupted kinase domain are capable of suppressing apoptosis, suggesting that the antiapoptotic domain is different from the mitogenic domain [171]. Although the postreceptor signaling mechanism for these responses is not well defined, previous studies have clearly demonstrated that type I receptor, upon binding to IGFs, and after autophosphorylation directly phosphorylates at least two classes of substrate proteins: insulin receptor substrates (IRSs) and SH-2-containing proteins (Shc). Phosphorylation of these molecules leads to activation of the MAP kinase cascade and PI-3 kinase pathways [166,167]. The role of IRS-1 in IGF receptor signaling is strongly supported by

XUEZHONG QIN ET AL.

the findings that targeted IRS-I gene disruption in mice led to a reduction in muscle PI-3 and MAP kinase activity, severe growth retardation, and resistance to both IGFs and insulin [167]. However, further downstream mediators that control various IGF biological actions remain to be identified. Because both insulin and type I IGF receptors are able to phosphorylate IRS and Shc, future identification of the downstream mediators diverging the actions of insulin and IGFs could provide important information with regard to reducing the hypoglycemic effect of IGFs. c. Transgenic/Knockout Studies IGF transgenic and knockout studies in mice have provided animal models to study the functions of IGFs in vivo. IGF-II knockout mice are apparently normal and fertile, but their body weights are only 60% of the wild type [172,173]. While survival is normal in mice lacking IGF-II, the majority of IGF-I knockout mice die shortly after birth, and surviving mice are infertile and exhibit delayed bone development [174]. In IGF-I and IGF-II double knockout mice, birth weight is further reduced, and all the mice die shortly after birth [174]. In contrast, transgenic mice overexpressing hIGF-I exhibit 30% higher body weight 4 weeks after birth than the wild type [175]. Moreover, transgenic mice overexpressing IGF-I in osteoblasts using an osteocalcin promoter to target expression exhibit increased cancellous bone volume [176]. These studies provide strong evidence that IGFs are crucial for both somatic growth and skeletal development. d. In Vivo Animal Treatment Studies To evaluate the effect of systemic IGF administration on bone formation, studies have been conducted using various animal model systems. Spencer et al. [177] showed that continuous infusion of IGF-I for 14 days in normal adult female rats increases the formation of both cortical and trabecular bone. Administration of IGF-I to 10 – 12-week-old growing rats for 4 weeks increases bone size and mineral content, although no effect on bone density is observed [178]. We have found that the systemic administration of IGF-I to 6month-old mice increases serum concentrations of bone formation markers [179]. In hypophysectomized rats, continuous IGF-I infusion for 18 days (0.3 mg/day) increases tibia epiphyseal width, longitudinal bone growth, and trabecular bone formation [180]. More relevant to osteoporosis, a number of studies have been performed to determine the effect of IGF-I treatment on osteopenia induced by ovariectomy [152]. Delivery of IGF-I via an osmotic minipump implanted subcutaneously for 6 weeks in ovariectomized rats results in a dose-dependent increase in bone mineral density in the lumbar spine and proximal and midshaft tibia [181]. Administration of human GH and IGF-I to aged ovariectomized rats prevents further loss of bone strength at sites containing trabecular bone [182]. In

415

CHAPTER 14 Bone Growth Factors

addition, IGF-I administration also increased bone formation in rats with osteopenia induced by liver cirrhosis [183]. Because hypoglycemia is a major adverse effect associated with systemic IGF administration, several studies have been performed to evaluate the efficacy of local IGF delivery on bone formation. Isgaard et al. [184] reported that local injection of IGF-I into the epiphyseal growth plate increased unilateral bone growth significantly. Local infusion of IGF-I (50 ng/day for 2 weeks) into the epiphyseal and metaphyseal junction at the distal femur of old rats increased trabecular bone volume and bone formation rate significantly without increasing bone resorption [185]. In another study [186], 12 daily injections of IGF-II (10 to 500 ng/day) into the outer periostea of parietal bones of neonatal rats dose dependently increased parietal bone mineral density and thickness. Consistent with these data, we found that a single local injection of IGF-I (0.2 or 1 g) over the parietal bone of old mice is able to increase ALP activity significantly in parietal bone extracts [179]. Taken together, these studies demonstrate that systemic or local IGF-I delivery to various skeletal sites is effective in promoting bone formation in experimental animals under both normal and pathological conditions, although the degree of effect varies among studies. e. In Vivo Human Treatment Studies A number of human clinical trials using IGF-I aimed at osteoporosis treatment have been conducted. Johansson et al. [187] first reported that bone formation and resorption markers in serum and urine are increased substantially during the treatment of osteoporotic men with rhIGF-I at a dose of 160 g/kg/day divided into two subcutaneous injections for 7 days. Subsequently, Ebeling et al. [188] showed that daily administration of IGF-I by a single subcutaneous injection for 6 days dose dependently increases the concentrations of markers for collagen synthesis and breakdown in healthy postmenopausal women. These authors demonstrated that the low dose of IGF-I (30 g/day) increases the level of bone formation but not the bone resorption markers. Johansson et al. [189,190] studied the effect of three doses of rhIGF-I (20, 40, or 80 g/kg/day) on bone turnover markers in 24 men with idiopathic osteoporosis. At a 20-g/kg dose, the increase in the serum PCP concentration remains elevated during the entire 6 weeks of injections. In contrast, urine deoxypyridinoline excretion is not altered by the same dose. Ghiron et al. [191] also found that 24 daily injections of IGF-I at 60 g/kg increases the levels of both resorption and formation markers, whereas treatment with IGF-I at a lower dose (15 g/day) only increased the levels of bone formation markers. Based on these preliminary data, a major problem with IGF treatment is that it usually increases not only bone formation but also bone resorption. Because IGFs are known to be the major mediator of GH actions, this may explain

why GH therapy also leads to an increase in both bone resorption and formation. Therefore, minimization of the stimulatory effect of IGFs on bone resorption represents an important issue in terms of improving IGF therapy. This goal may be achieved by the administration of IGF in combination with antiresorptive agents. 2. IGFBPS a. In Vitro Studies In addition to functioning as classic binding proteins (i.e., acting as transport proteins and prolonging the half-life of IGFs), recent studies demonstrate that different IGFBPs can exhibit independent actions which may either enhance or inhibit IGF actions. The following briefly describes our understanding of the biological effects of various IGFBPs in osteoblasts. Campbell and Novak [192] reported that purified IGFBP-1 inhibits IGF-I-induced cell proliferation in human MG63 osteosarcoma cells. In contrast to these results, we found that IGFBP-1, up to 30 ng/ml, had no effect on chick osteoblast cell proliferation [193]. rhIGFBP-2 inhibits IGF-I-induced bone cell proliferation at an apparent dose ratio of 10:1 [194]. The effect of IGFBP-3 seems to vary depending on the dose and culture conditions used. Ernst and Rodan [195] showed that IGFBP-3 augments the effects of IGF-I on rat osteoblast cell proliferation. We found that IGFBP-3 potentiates the IGF-I effect in osteoblasts at low concentrations, whereas it inhibits the IGF-I effect at high concentrations [196]. Among the six IGFBPs, IGFBP-4 is the only abundant osteoblast-produced IGFBP that consistently inhibits the mitogenic activity of IGFs in a variety of cell types, including osteoblasts [193,197 – 199]. In contrast to IGFBP-4, IGFBP-5 enhances IGF-induced cell proliferation in osteoblasts [199 – 201]. IGFBP-6 inhibits the mitogenic action of IGFII more potently than that of IGF-I in osteoblasts [202]. This differential effect is apparently due to the lower affinity of IGFBP-6 for IGF-I than for IGF-II. The distinct effect of each individual IGFBP, as well as the differential regulation of IGFBP production [152], may be critical in determining the degree of IGF-induced cellular responses in target tissues such as bone. b. Mechanism of Action The mechanisms by which IGFBPs modulate IGF actions in bone have not been clearly defined and may vary among IGFBPs. In this regard, evidence suggests that some of the IGFBPs may exert IGF-independent actions [199,203 – 205], in addition to modulating IGF actions [192,199,206 – 209]. Previous studies demonstrate that IGFBP – 4 acts primarily to inhibit human osteoblast proliferation by an IGF-dependent mechanism (Fig. 3), based on the following observations: (1) IGFBP-4 competes with IGF receptors for IGF binding in human osteoblasts in the monolayer culture and in purified type I IGF receptor preparations [199]; (2) IGFBP-4 has no

416

XUEZHONG QIN ET AL.

FIGURE 3

Proposed models for the actions of IGFBPs in bone cells. IGFBP-4 inhibits IGF binding to IGF receptors by binding IGFs near or at the receptor-binding site. This mechanism may also be applicable to the actions of IGFBP-1 and IGFBP-6 in osteoblasts, although further confirming studies need to be performed. With regard to IGFBP-5, three alternate models are proposed. In model 1, the complex of IGFBP5IGF binds to IGF receptors. In model 2, IGFBP-5 may bind to its own cell surface receptors and stimulate cell proliferation via IGF-independent mechanisms. In model 3, IGFBP-5-binding sites in the bone cell surface may recruit IGFs at the surface of cells, enabling the ligand to be captured easily by IGF receptors (reproduced with permission).

effect on human osteoblast proliferation induced by IGF-I or -II analogues with reduced affinity for IGFBP-4 [199]; and (3) IGFBP-4 proteolytic fragments or rhIGFBP-4 analogues with little IGF-binding activity are much less effective at inhibiting IGF-induced human osteoblast proliferation [208,209]. In addition to IGFBP-4, IGFBP-1 and IGFBP-6 may also act to inhibit osteoblast proliferation by an IGF-dependent mechanism based on the observations that: (1) IGFBP1 is less effective in inhibiting cell proliferation of MG63 cells induced by IGF analogues with low affinity for IGFBP-1 and (2) IGFBP-6 is less potent in inhibiting the mitogenic action of IGF-I than that of IGF-II (IGFBP-6 exhibits  100-fold higher affinity for IGF-II than for IGF-I) in oseoblasts [202]. IGFBP-5, which by itself associates with human osteoblast surfaces, can enhance the binding of IGFs to human osteoblasts in serum-free culture [199]. In addition, IGFBP-5 treatment further increases human osteoblast proliferation in the presence of IGF-I or IGF-II analogues with little IGFBP-5 binding affinity [199]. These findings suggest that IGFBP-5 may, in part, stimulate human osteoblast proliferation by an IGF-independent mechanism involving IGFBP-5-specific cell surface-binding sites (Fig. 3). In

addition, Jones and Clemmons [166] have proposed that association of IGFBP-5 with proteins on the cell surface or in the extracellular matrix (ECM) results in an increase in the local concentration of IGFs in the vicinity of the IGF receptor. Similarly, IGF-independent effects of IGFBP-3 in fibroblast cell cultures have also been described [203 – 205]. It has been reported that IGFBP-3 may be localized to the nuclei, although the physiological significance of this finding has not been determined [210]. It is anticipated that future studies on the role of IGFBP nuclear localization and cell surface IGFBP receptor/acceptor molecules will provide experimental data needed to determine the mechanism whereby some IGFBPs exert IGF-independent effects. c. Transgenic Studies To date, transgenic mice have been generated for IGFBP-1 [211], IGFBP-2 [212], IGFBP3 [213], and IGFBP-4 [214]. Transgenic mice overexpressing high levels of rat IGFBP-1 exhibit reduced birth weight (80 to 90% of the wild type), hyperglycemia, and impaired brain development [211,215]. IGFBP-2 transgenic mice also show reduced postnatal growth [212]. In IGFBP-3 transgenic mice, the rate of growth is not affected, although the weights of several tissues are higher than these of wild

CHAPTER 14 Bone Growth Factors

type mice [213]. Overexpression of rat IGFBP-4 in smooth muscle cells of transgenic mice using a myoblast-specific promoter causes muscle hypoplasia, a reciprocal phenotype to that of IGF-I transgenic mice [214]. Although the effects of overexpression of these IGFBPs on skeleton have not been examined, these transgenic studies suggest that IGFBP-1, -2, and -4 may play a negative role in growth regulation in vivo. d. In Vivo Treatment Studies Studies on the effect of IGFBPs on bone formation in vivo have been limited. Bagi et al. [216] first evaluated the effect of systemic administration of IGF-I alone or in combination with IGFBP-3 on bone formation in ovariectomized rats. At the highest dose used in this study, IGF-I (7.5 mg/kg), with an equimolar amount of IGFBP-3, significantly promotes bone formation. However, it is not known whether this effect is indeed due to the presence of IGFBP-3, as a treatment group receiving IGF-I only at this dose (7.5 mg/kg) was not included in this study. Subsequently, Narusawa et al. [217] compared the effect of administration of rhIGF-I alone or rhIGF-I/IGFBP-3 complex on bone formation in rats with combined ovariectomy and bilateral sciatic neurectomy. Injection of IGF-I alone three times a week for 4 weeks (0.3 or 3 mg/kg) did not significantly increase the bone formation rate. However, injection of the same dose of IGF-I along with an equimolar amount of IGFBP-3 significantly increased bone formation rate and cancellous bone volume of the lumbar vertebrae. These observations suggest that IGFBP-3 can potentiate the effect of IGF-I on bone formation (although the issue of whether IGFBP-3 alone affects bone formation was not addressed in this study). Regarding IGFBP-5, it has been demonstrated that nine daily subcutaneous injections of rhIGFBP-5 (50 g/day) in mice increase serum osteocalcin values as effectively as administration of IGF-I alone (13 g/day) or IGFIrhIGFBP-5 complex [218]. In addition, a single systemic administration of IGFBP-5 alone increased bone formation markers. Future studies using various doses of IGFBP-5 with or without IGF are needed to determine if the systemic administration of IGFBP-5 and IGF-I can cause an additive effect on bone formation. It is anticipated that further detailed studies on the effect of IGFBP, alone or in combination with IGFs, on bone formation could lead to increased understanding of the physiological role of each IGFBP in vivo and could provide important information for future optimization of IGF therapy in the treatment of osteoporosis and other diseases. 3. IGFBP PROTEASES In essentially all body fluids, IGFs form complexes with IGFBPs. The release of IGFs from these inactive IGF/IGFBP complexes is a prerequisite for IGFs to bind to their receptors and elicit a biological response. One mech-

417 anism would be through the degradation of IGFBPs by IGFBP proteases (Fig. 4). Although proteolysis of various IGFBPs by human osteoblast conditioned medium has been reported [152], the IGF-dependent IGFBP-4 protease produced by human osteoblasts is studied most extensively. This protease is unique in that the proteolysis of IGFBP-4 by cell-free human osteoblast CM can be enhanced by the addition of exogenous IGFs [207,219 – 221]. Based on the following, our characterization of human osteoblast-produced IGFBP-4 protease suggests that binding of IGF-II to IGFBP-4 may alter the IGFBP-4 conformation such that the cleavage site is more accessible to the protease (1) des [1– 6] IGF-II, with at least 200-fold reduced affinity for IGFBP-4, is much less effective than intact IGF-II in promoting IGFBP-4 proteolysis by human osteoblast CM [220] and (2) the IGFBP-4 analogue with deletion of Leu73, Met73, and His74 from the IGF- binding domain exhibits undetectable IGF-binding affinity and is not cleaved by the partially purified IGFBP-4 protease from human osteoblast CM [209]. Because IGFBP-4 is one of the most abundant IGFBPs produced by hOBs and its molar concentration in human osteoblast CM is at least 10-fold higher than that of IGF-II, we postulate that the IGFBP-4 protease may play an important role in the regulation of human osteoblast proliferation. This notion is supported by findings that protease-resistant IGFBP-4 analogues, which exhibit normal IGF-II and IGF-I binding affinity, are much more potent than the-wild-type IGFBP-4 in blocking IGF-II mitogenic actions in normal human osteoblast cultures, but not in the culture of MG63 cells, which do not produce IGFBP-4 protease [221]. In addition, IGFBP-4 protease activity has been known to be regulated by other osteoregulatory agents in addition to IGFs, such as TGF- and estrogen [219]. Lawrence et al. [222] reported that the IGF-II-dependent IGFBP-4 protease produced by human fibroblasts is identical to the previously identified pregnancy-associated plasma protein (PAPP)-A. Consistent with this new finding, we found that proteolysis of IGFBP-4 by human pregnancy serum is IGF-II dependent [223]. Moreover, pregnancy serum and human osteoblast-produced IGFBP-4 protease recognized the same cleavage site (Met135/Lys136). These data suggest that PAPP-A represents the major IGF-II dependent IGFBP-4 protease produced by human osteoblasts. Since the PAPPA cDNA has been cloned [224], the role of PAPP-A in osteoblast functions can be further confirmed by blocking its production using the antisense approach. In addition to the IGFBP-4 protease, human osteoblasts also produce a metal-dependent serine protease capable of degrading IGFBP-5 but not other IGFBPs [220,225]. However, the identity of the human osteoblast-produced IGFBP5 protease has not been determined. Based on findings that the activity of the IGFBP-5 protease changes in response to treatment with growth factors and osteoregulatory agents in

418

XUEZHONG QIN ET AL.

FIGURE 4 Potential role of IGFBP proteases in modulating IGF bioavailability to local tissues. Proteolysis of IGFBP-3 by serum protease leads to the disruption of IGF/IGFBP-3/ALS (available substance) complex and the release of IGFs. The released IGFs are either directly transported to the target tissues or bind to other IGFBPs to form smaller IGF/IGFBP complexes, which can cross the endothelium into the extracellular space. Upon transport into the extracellular space, IGFs are released from inactive IGFBP/IGF complexes (e.g., IGF/IGFBP-4] through action of the locally produced IGFBP proteases (e.g., the IGF-dependent IGFBP-4 protease/PAPP-A). However degradation of stimulatory IGFBPs, such as IGFBP-5, may attenuate the action of IGFs. Thus, the regulation of IGFBP protease production/activity may play a key role in controlling the availability of IGFs to the target tissues.

osteoblasts [220,225 – 228] and that IGFBP-5 proteolytic fragments are much less active compared to the intact IGFBP-5 in stimulating osteoblast proliferation [199], it can be proposed that the IGFBP-5-specific protease is also an important regulatory component of the IGF system in human osteoblasts. The regulation of IGFBP degradation and synthesis may be equally important in controlling the bioavailability of IGFBPs and, consequently, IGF actions in the local bone environment. Studies on the physiological role of IGFBP proteases have just begun and new understanding will emerge as the identity of various IGFBP proteases become established. 4. ROLE OF IGF SYSTEM IN THE PATHOGENESIS OF OSTEOPOROSIS The development of osteoporosis is determined by at least two major parameters: (1) peak bone density achieved during or shortly after puberty and (2) rate of bone loss with aging, particularly after menopause. These two parameters appear

to be affected by the IGF system. Studies demonstrate that the serum concentrations of IGF-I, IGFBP-3, and IGFBP-5, but not IGF-II, showed a significant positive correlation to bone mineral content, bone density, and metacarpal indexes during puberty in girls [229]. In addition, serum concentrations of IGFBP-3 and IGFBP-5 showed a significant positive correlation to the level of bone formation marker, osteocalcin. These data suggest that the IGF system may play an important role in skeletal development during sexual development, a stage of life that determines peak bone density. Although it has been well established that bone formation decreases with age, the mechanism which underlies this process awaits full elucdation. In this regard, serum concentrations of IGF-I are lower in elderly individuals [230,231] and postmenopausal women [232]. Serum IGF-I values are reduced in men with idopathic osteoporosis and correlate with lumbar spine bone mineral density [231]. Serum levels of IGF-I and other stimulatory IGF system components are significantly lower in hip fracture patients than in agematched controls [233]. The IGF-I content of cortical bone

419

CHAPTER 14 Bone Growth Factors

also decreases with age [71]. In addition, concentration of IGFBP-5 (a stimulator of IGF actions) in serum and bone decrease with aging [234]. In contrast, serum levels of IGFBP4 (an inhibitor of IGF actions) are elevated in elderly subjects with low bone density and correlate with serum concentrations of PTH [235,236]. Consistent with the hypothesis that the increase in serum PTH in osteoporotic patients may lead to overproduction of IGFBP-4, we found that treatment of human osteoblasts with PTH significantly increases IGFBP-4 concentration in the conditioned medium [236]. In addition to these age-related changes in the production of IGF system components, bone cells derived from aged rats [237] or humans [238,239] are much less responsive to IGFs than bone cells derived from their younger counterparts. Based on these findings, it is conceivable that the age-related decline in bone formation may be due, at least in part, to the overproduction of an inhibitory IGF component (i.e., IGFBP-4], underproduction of stimulatory IGF system components (IGF-I and IGFBP-5], and the decreased osteoblast responsiveness to IGFs in aged individuals. In addition to age-related osteoporosis, glucocorticoid induced osteoporosis represents a significant medical problem (see also Chapter 44). Although the mechanism for the glucocorticoid-induced bone loss is not clear, the impairment of the GH – IGF axis appears to play a role. Although results are not entirely consistent among studies, treatment of bone cells with glucocorticoid generally decreases the expression/production of stimulatory components of the IGF system, including IGF receptors, IGF-I, IGF-II, IGFBP-3, and IGFBP-5, but increases the inhibitory components of this system, such as IGFBP-6 [233,240 – 245]. Dexamethasone dose dependently suppressed basal, as well as GH or IGF-I stimulated rat chondrocyte proliferation and the GH stimulated IGF-I production in these cells [245]. In vivo, methylprednisolone treatment dose dependently decreased the concentrations of serum-free IGF-I in rats [246]. In a human longitudinal prospective study, glucocorticoid therapy decreased serum bone formation parameters and serum concentrations of IGF-I, IGF-II, and IGFBP-3 markedly [247]. In aggregate, these in vitro and in vivo studies provide evidence that the IGF system may be involved in the pathogenesis of osteoporosis induced by glucocorticoid treatment.

D. The FGF System Fibroblast growth factors belong to a large family of heparin-binding proteins that regulate cell proliferation, differentiation, and migration in many different tissues and play key roles during physiological and pathological conditions, such as wound healing, skeletal repair, neovascularization, and tumor growth [248]. FGFs were first isolated in the 1970s from bovine brain extracts based on their mitogenic and angiogenic activities [249], and currently 19

unique family members have been identified on the basis of their strong sequence homology in a core region of 120 amino acids. FGF1 (acidic FGF) and FGF2 (basic FGF) were the first two members of this family to be characterized and are the most widely studied in many tissues, including bone. In contrast to the almost ubiquitous tissue distribution and action of FGF2 and the wide distribution of FGF1, many other FGF family members have more restricted spatial and temporal expression patterns. Originally, FGF3 – 6 were isolated as protooncogenes from tumor cells [250 – 253]. FGF7 was originally known as keratinocyte growth factor, based on its identification as a selective mitogen for epithelial cells but not fibroblasts [254]. FGF8, or androgen-induced growth factor (AIGF), was isolated from a murine androgen-dependent carcinoma [255]. FGF9, originally termed GAF or glial-activating factor, was isolated from a human glioma cell line [256]. A group of 4 fibroblast homologous growth factors (FHFs) implicated in nervous system development were designated FGF11 – 14 [257]. The number of newly discovered FGFs has been growing rapidly, with FGF19 being the most recently added member [258]. Undoubtedly, more members of the FGF family will be discovered in the future and, with these discoveries, new functions are also likely to emerge. Work performed in many laboratories has established that FGFs play many important physiological roles in bone growth, remodeling, and repair. Both FGF1 and FGF2 stimulate osteoblast proliferation and promote bone growth. They are also synthesized by osteoblasts and are stored in bone. Because FGF2 is more potent than FGF1, most of the in vitro and in vivo studies have been performed with FGF2. In vitro, FGF2 has been shown to stimulate the proliferation of osteoblasts, chondrocytes, and periosteal cells, and to stimulate the formation of mineralized bone-like nodules in cultures of bone marrow stromal cells [259 – 262]. Based on these data and the following studies, it is suggested that FGF2 may play a role in normal bone remodeling and, in addition, may have therapeutic use in fracture repair and in the treatment of other local skeletal defects. 1. IN VITRO STUDIES FGF1 and FGF2 strongly stimulate the proliferation of bone cells under most culture conditions, but reduce markers of the differentiated phenotype, such as alkaline phosphatase and PTH-responsive adenylate cyclase activity [263,264]. In bovine, rodent, and human calvaria-derived cells, the effects on bone cell proliferation and differentiation appear to be uncoupled, indicating that FGFs have independent effects on cell replication and differentiation. The effects of FGF on osteoblastic cell differentiation and bone matrix formation are, however, conflicting. FGF2 has been reported to either stimulate or inhibit the production of type I collagen, the major differentiated product of

420 the osteoblast [265,266]. Two factors that influence the type of response are duration of FGF exposure and maturational stage of the cells being treated. Thus, short-term (24 h) treatment with FGF2 results in the stimulation of collagen I, whereas long-term exposure of osteoblast cultures to FGF2 inhibits type I collagen synthesis. Such inhibition by FGF appears to be transcriptional in nature and is mediated by DNA sequences in the promoter of the alpha1(I) collagen gene [267]. Further, the effect of FGF2 on human calvarial osteoblast cells is influenced by the level of cell maturation. FGF2 reduces the expression of osteoblast markers in less mature cells, while increasing osteocalcin production and matrix mineralization in more mature cells [268]. These observations indicate that FGF does not have a consistent effect on osteoblastic cell differentiation in vitro. Studies of the mechanism of action of FGF on osteoblasts should help clarify the function of FGF on osteoblastic differentiation. The observation of a dose-dependent biphasic effect of FGF2 on long bone growth has prompted several in vitro studies with isolated chondrocytes in order to elucidate this mechanism. In vitro, FGF2 inhibits terminal differentiation and hypertrophy of chondrocytes. Cultured chondrocytes that undergo terminal differentiation have fewer high-affinity binding sites for FGF [269]. This suggests that the action of FGF may be partially regulated at the level of receptor number. Similar to the situation with osteoblasts, there are discrepancies between different studies regarding the effect of FGF on thymidine incorporation and collagen synthesis in chondrocytes [270]. A possible explanation for the failure of these in vitro experiments to consistently represent the in vivo situation is that a cell culture system does not maintain the three-dimensional cellular interaction of the highly organized growth plate. The use of an organ system circumvents many of the limitations of cell culture. The addition of FGF2 to organ cultures from rat metatarsals inhibits longitudinal growth, which is consistent with the effect in vivo [271]. 2. MECHANISM OF ACTION The biological effects of FGFs are mediated through four high affinity transmembrane kinase receptors, known as FGFR1 – FGFR4. FGFRs structurally resemble other transmembrane protein kinase receptors and have been found in all contemporary vertebrates [272]. Four human genes encoding FGFR1 – FGFR4 have amino acid identities of 55 – 72% [273]. The receptors are monomeric in their native state and dimerize after binding with an FGF ligand. Dimerization activates tyrosine kinase and triggers downstream effects through multiple signaling pathways. The details of these pathways have not been fully established for all the FGFRs. Studies with FGFR1 and FGFR3 suggest that activation of Ras-dependent pathways results in the stimulation of mitogenesis, whereas STAT signaling

XUEZHONG QIN ET AL.

pathways are inhibitory [274]. Another signaling pathway may involve internalization and nuclear translocation of the ligand – FGFR complex [275]. An interesting aspect of the FGFR gene family is the existence of multiple variants generated by alternative splicing of mRNA, which can result in over 100 different protein sequences. Some of the variants have distinct affinities for individual FGF ligands and use different signaling pathways [276]. Taken together, the complex genetic organization of both FGF ligands and receptors presents a great number of regulatory mechanisms to mediate the biological effects of this extensive family of growth factors. This modulation can occur at transcriptional, translational, posttranslational, or signal pathway levels, thereby providing the means for fine-tuning the many biological processes that utilize FGF. In addition to high-affinity FGFRs, FGF also binds to heparin sulfate proteoglycans (HSPGs) [277] and to a lowaffinity, cysteine-rich transmembrane FGF-binding protein [278]. HSPGs are sulfated glycosaminoglycans bound to a core protein localized in the extracellular matrix. At this location, HSPGs may provide an extracellular storage compartment protecting FGFs from degradation. In response to external or internal stimuli, such as phosphorylation or proteolytic processing of the extracellular matrix, HSPGs may regulate the bioavailability of FGFs. Many studies have indicated an enhancement by HSPGs for FGF-induced biological activities. The most likely mechanism for this effect is by increasing the affinity of FGFs for FGFRs [279,280]. Heparin contains the same sugar residues as HSPGs and is often used in experiments to enhance FGF stability and potency. The role of the low-affinity, cysteine-rich binding protein has not been well studied. It binds FGF at a site that overlaps the high-affinity receptor binding site, as binding to both receptor types is mutually exclusive. A possible role for the low-affinity receptor may be in regulating intracellular FGF levels [281]. 3. TRANSGENIC/KNOCKOUT STUDIES Much can be learned about the role of FGF and FGFRs from transgenic studies or from the analysis of natural mutations [282]. Transgenic mice overexpressing FGF2 exhibit chondrodysplasia, which is characterized by long bones that are considerably shorter than those of wild-type animals [283]. Chondrodysplasia in animals overexpressing FGF2 is similar to skeletal malformations observed in human syndromes associated with mutations in FGFRs. Therefore, a possible explanation for this phenomenon is that high concentrations of FGF2 selectively activate certain subsets of FGFRs that exert a negative effect on bone growth. Surprisingly, FGF2 knockout mice were indistinguishable in appearance from their normal littermates [284], indicating that other members of the FGF family may be able

CHAPTER 14 Bone Growth Factors

to substitute for some FGF2 functions during early bone formation. However, significant neuronal defects in FGF2 knockout mice were observed in the neocortex, most notably a reduction in the neuronal density and disruption of the cytoarchitecture. In addition, the healing of skin wounds was delayed significantly in mice lacking FGF2. These results indicate that FGF2, while not essential for embryonic development, plays specific roles in cortical neurogenesis and wound healing that cannot be carried out by other FGF family members. Because of the redundancy of FGF signaling, it is difficult to assess the role of any particular member of the FGF family by elimination or overexpression of a single gene. Only specific functions that cannot be substituted by other FGFs will be manifest in obvious defects. FGF4, FGF8, and FGF10 appear to be the most crucial regulating factors during embryonic development. Mice lacking FGF4 or FGF8 are both embryonic lethal, whereas a targeted disruption in FGF10 resulted in mutant embryos without limbs [285 – 287]. The action of FGF8 and FGF10, both of which are expressed during limb development, is most likely mediated through FGFR2, as shown by expression of a soluble dominant-negative FGFR2 gene [288]. Knockout studies of all four high-affinity FGFRs have shown individually that FGFR1 and FGFR2 are the only receptors that are involved in limb development and are embryonic lethal [282,289]. FGFR1 mutants died early in gastrulation, displaying growth defects and axial mesodermal disorganization. In contrast to knockouts for FGFR1 and FGFR2, null mutations for FGFR3 and FGFR4, or both, have normal limbs. Of great interest regarding the regulation of bone growth was the finding that FGFR3 knockouts displayed excessive growth of long bones and vertebrae, indicating enhanced endochondral bone formation [290,291]. These findings have led to the suggestion that the FGFR3 kinase is a negative regulator of bone growth that acts by limiting endochondral ossification. This explanation is supported by observations of certain naturally occurring mutations of FGFR3 that cause short stature (achondroplasia and thanatophoric displasia). Because these syndromes exhibit increased ligand-independent activity of FGFR3, it is likely that the mutations act by enhancing the negative control normally exerted by FGFR3. Therefore, the identification of distinct FGF receptor mutations in individuals with human skeletal disorders provides supporting evidence that FGFs play an important role in growth and development of the skeleton. 4. IN VIVO TREATMENT STUDIES a. Local Applications During fracture repair, FGF1 and FGF2 are expressed at the earliest stage in macrophages and periosteal cells, and in osteoblasts and chondrocytes in later stages [49]. These observations provided the rationale for applications of FGF directly to the

421 fracture site in order to improve the healing process. In a diabetic rat model, where fracture healing is impaired due to reduced expression of FGF2 during early stages, a single application of recombinant FGF2 increased the volume and mineral content of the callus in both normal and diabetic rats [292]. In addition, FGF2 also provided a marked improvement in the mechanical stability of the healing fibula in all animals. While one study found no effect on fracture healing in the rabbit tibia with a single injection of FGF1 or FGF2 [293], more recent investigations have reported improved fracture healing by FGF in several species. In rabbits, single injections of FGF2 at 100 g or above in a 3-mm tibial bone defect resulted in increased volume and mineral content of newly made bone after 5 weeks [294]. A similar model in dogs used 200g of FGF2 and showed increased membraneous ossification after 2 weeks. After 4 weeks, the callus in the FGF-treated group was larger than in the vehicle-treated control group. Additionally, the fracture strength by week 16 was greater in FGF-treated animals than in controls [295]. Interestingly, the FGF group also had a markedly increased osteoclast number in the periosteal callus. This suggests that FGF2 stimulates both callus formation and osteoclastic resorption, thereby promoting fracture healing by stimulating bone remodeling. A single local application of FGF was also found to promote healing in a primate fracture model. As measured by radiography and mechanical and histological critera, a single injection of FGF2 in a stabilizing gel of hyaluronic acid was found to be more effective at promoting fracture healing in baboon fibulae than FGF2 alone or injected in fibronectin gel [296]. Based on these findings, it can be postulated that local delivery of FGF2 or FGF1 directly into the fracture site immediately after the injury may substantially enhance fracture healing in patients with both normal or impaired ability for fracture repair. In addition to promoting fracture healing, local application of FGFs has also been shown to promote regeneration at other sites. Intraosseous injection into the femur of ovariectomized rabbits to a depth of 3 – 5 mm from the cortex stimulated intraosseous bone formation after 4 weeks in a dose-dependent manner [297]. Slow release delivery methods for FGF are currently being developed aimed at increasing the efficacy of this growth factor. Specifically, FGF2 embedded in a gelatin-based hydrogel promoted complete closure within 21 days of a skull defect in primates [298]. In contrast, no closure occurred when the same dose of FGF2 was applied in solution. This type of experiment suggests that local delivery of FGF to sites of degeneration, such as the hip in generalized osteoporosis, may be used to promote new bone formation at this site and reduce the risk for hip fractures. b. Systemic Applications Similar to local application of FGFs, systemic application also promotes bone growth

422 and restoration in many systems. FGF1 prevented bone loss and increased new bone formation when injected at 0.2 mg/kg daily for 28 days into the tail vein of adult ovariectomized female rats [299]. FGF1 had substantial anabolic effects in sham-operated animals, where bone density increased two-fold. Without treatment, severe bone loss and trabecular disruption occurred in ovariectomized animals 6 months after the operation, similar to that seen in patients with osteoporosis. In these rats, FGF1 treatment induced extensive new woven bone formation with new trabecular-like structures, and bone density in the tibial metaphysis increased three-fold. More complex effects of systemically applied FGF on bone growth are suggested by the observation that systemic administration of FGF2 to growing rats events a dose-dependent biphasic effect on longitudinal bone growth [300]. In this study, daily intravenous injections of FGF2 at 0.1mg/kg in growing rats increased longitudinal bone growth and cartilage cell proliferation. In contrast, a higher dose (0.3 mg/kg) reduced the rate of bone growth and cartilage proliferation. As the rate of longitudinal bone growth depends on the rate of growth plate chondrogenesis, FGF2, at higher doses, may inhibit one of the cellular processes underlying chondrogenesis [271]. Pleiotropic effects of FGF on its target tissues throughout the body may preclude the clinical application of this very exciting therapeutic agent to systemically regenerate the skeleton. Such effects on nonskeletal tissues include hypotension [301] and hypertrophy of epithelial tissues, kidney, and lung, some of which would be considered adverse effects in patients who require only skeletal regeneration. Systemic application of FGF with gene therapy methods designed specificially to increase bone formation may circumvent such undesirable side effects of FGF action by targeting the vector delivery, or FGF expression, exclusively to bone tissue. New therapies using the growth factor potential of FGFs are currently being actively pursued, not only to stimulate bone formation, but also to stimulate new blood vessel formation in peripheral vascular disease and myocardial ischemia [302,303]. It seems likely that, in the future, FGF gene targeting to skeletal tissue will be accomplished so that we can realize the remarkable skeletal regenerative capacity of this growth factor.

III. CONCLUSIONS Growth factors, such as TGF-, BMP, IGF, and FGF, play a unique role in bone in that they are not only actively produced by bone cells for contemporary use, but are also stored in bone as an immediately available source of growth factors for future use when the old bone is destroyed by resorption. One or both of these mechanisms for growth factor elaboration in bone may be involved in the

XUEZHONG QIN ET AL.

several key physiological processes in bone metabolism: namely, development of peak bone density, bone remodeling (in which the coupling of bone formation to bone resorption plays a pivotal role), fracture repair, and the adaptive response in terms of density and architecture that bone exhibits in response to mechanical strains. Based on these above observations, it is not surprising that bone growth factor deficiencies have also been implicated in the pathogenesis of disease processes, such as osteoporosis. In this regard, both deficiencies of TGF- and the IGFs have been associated with osteoporosis. If a deficiency of growth factor can lead to bone loss, it follows that treatment with a growth factor might be an effective therapy to promote bone gain in osteoporosis. With regard to growth factor therapy, it is clear from animal studies that a number of growth factors, including TGF-, BMP, IGF, and FGF, can locally stimulate bone formation in vivo in a number of animal models, and it seems likely that these agents could be used in the near future in human conditions requiring stimulation of local bone growth, such as fracture healing and the treatment of local lesions in bone. However, the potential usefulness of skeletal growth factors to treat bone loss in generalized skeletal disorders, such as osteoporosis, requires more knowledge, as these multifunctional factors affect diverse tissues besides bone. Therefore, from the standpoint of safety, TGF-, BMP, and FGF can now be administered only locally. In the future it may be possible to develop the necessary methodologies to target these growth factors for the skeleton or specific sites in the skeleton. In this regard, the development of suitable delivery systems to target growth factors to areas of interest would be very critical. In contrast to TGF-s, BMPs, and FGFs, IGFs are produced locally as growth factors and also circulate in body fluids as hormones. Therefore, the IGFs may represent a growth factor group that can possibly be used to treat bone loss via systemic administration. Because IGF action can be enhanced by certain IGFBPs, the combination IGF/IGFBPs may prove more useful than IGF alone in the treatment of bone loss. In addition, simultaneous administration of antiresorptive agents may help reduce the bone-resorbing activity of IGF. This important area should demand attention in the future.

References 1. M. R. Urist, Bone: Formation by autoinduction. Science 150, 893 – 899 (1965). 2. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin-sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 3. S. Mohan and D. J. Baylink, Bone growth factors. Clin. Orthop. 263, 30 – 48 (1991).

CHAPTER 14 Bone Growth Factors 4. Y. Ogawa, D. K. Schmidt, J. R. Dasch, R. J. Chang, and C. B. Glaser, Purification and characterization of transforming growth factor-beta 2.3 and -beta 1.2 heterodimers from bovine bone. J. Biol. Chem. 267, 2325 – 2328 (1992). 5. D. J. Baylink, R. D. Finkelman, and S. Mohan, Growth factors to stimulate bone formation. J. Bone Miner. Res. 8 (Suppl. 2), S565 – S572 (1993). 6. E. Canalis, J. Pash, and S. Varghese, Skeletal growth factors. Crit. Rev. Eukaryot. Gene Expr. 3, 155 – 166 (1993). 7. G. R. Mundy, Factors which stimulate bone growth in vivo. Growth Regul. 3, 124 – 128 (1993). 8. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 9. A. H. Reddi, Bone morphogenetic proteins, bone marrow stromal cells, and mesenchymal stem cells: Maureen Owen revisited. Clin. Orthop. 313, 115 – 119 (1995). 10. T. Sakou, Bone morphogenetic proteins: From basic studies to clinical approaches. Bone 22, 591 – 603 (1998). 11. D. J. Baylink, J. C. Jennings, and S. Mohan, Calcium and bone homeostasis and changes with aging In “Principles of Geriatric Medicine” (W. R. Hazzard, J. P. Blass, W. H. Ettinger, J. B. Halter, and J. G. Ouslander, eds.), pp. 1041 – 1056 (1999). 12. P. P. Hu, M. B. Datto, and X. F. Wang, Molecular mechanisms of transforming growth factor-beta signaling. Endocr. Rev. 19, 349 – 363 (1998). 13. J. C. Jennings and S. Mohan, Heterogeneity of latent transforming growth factor-beta isolated from bone matrix proteins. Endocrinology 126, 1014 – 1021 (1990). 14. L. F. Bonewald and S. L. Dallas, Role of active and latent transforming growth factor beta in bone formation. J. Cell. Biochem. 55, 350 – 357 (1994). 15. A. H. Tashjian, Jr., E. F. Voelkel, M. Lazzaro, F. R. Singer, A. B. Roberts, R. Derynck, M. E. Winkler, and L. Levine, Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorption in cultured mouse calvaria. Proc. Natl. Acad. Sci. USA 82, 4535 – 4538 (1985). 16. M. Centrella, T. L. McCarthy, and E. Canalis, Transforming growth factor beta is a bifunctional regulator of replication and collagen synthesis in osteoblast-enriched cell cultures from fetal rat bone. J. Biol. Chem. 262, 2869 – 2874 (1987). 17. M. Noda and G. A. Rodan, Type-beta transforming growth factor inhibits proliferation and expression of alkaline phosphatase in murine osteoblastlike cells. Biochem. Biophys. Res. Commun. 140, 56–65 (1986). 18. E. C. Breen, R. A. Ignotz, L. McCabe, J. L. Stein, G. S. Stein, and J. B. Lian, TGF beta alters growth and differentiation related gene expression in proliferating osteoblasts in vitro, preventing development of the mature bone phenotype. J. Cell. Physiol. 160, 323 – 335 (1994). 19. M. Centrella, J. Massague, and E. Canalis, Human platelet-derived transforming growth factor-beta stimulates parameters of bone growth in fetal rat calvariae. Endocrinology 119, 2306 – 2312 (1986). 20. J. Pfeilschifter and G. R. Mundy, Modulation of type beta transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 21. Z. Schwartz, L. F. Bonewald, K. Caulfield, B. Brooks, and B. D. Boyan, Direct effects of transforming growth factor-beta on chondrocytes are modulated by vitamin D metabolites in a cell maturation-specific manner. Endocrinology 132, 1544 – 1552 (1993). 22. Z. Schwartz, V. L. Sylvia, Y. Liu, D. D. Dean, and B. D. Boyan, Treatment of resting zone chondrocytes with transforming growth factor- beta 1 induces differentiation into a phenotype characteristic of growth zone chondrocytes by downregulating responsiveness to 24,25-(OH)2D3 and upregulating responsiveness to 1,25-(OH)2D3. Bone 23, 465 – 470 (1998).

423 23. V. L. Sylvia, S. Mackey, Z. Schwartz, L. Schuman, R. Gomez, and B. D. Boyan, Regulation of protein kinase C by transforming growth factor beta 1 in rat costochondral chondrocyte cultures. J. Bone Miner. Res. 9, 1477 – 1487 (1994). 24. K. A. Skantze, C. E. Brinckerhoff, and J. P. Collier, Use of agarose culture to measure the effect of transforming growth factor beta and epidermal growth factor on rabbit articular chondrocytes. Cancer Res. 45, 4416 – 4421 (1985). 25. P. C. Yaeger, T. L. Masi, J. L. de Ortiz, F. Binette, R. Tubo, and J. M. McPherson, Synergistic action of transforming growth factor-beta and insulin-like growth factor-I induces expression of type II collagen and aggrecan genes in adult human articular chondrocytes. Exp. Cell Res. 237, 318 – 325 (1997). 26. A. Martin, F. J. Unda, C. Begue-Kirn, J. V. Ruch, and J. Arechaga, Effects of aFGF, bFGF, TGFbeta1 and IGF-I on odontoblast differentiation in vitro. Eur. J. Oral Sci. 106 (Suppl. 1), 117 – 121 (1998). 27. J. E. Wergedal, T. Matsuyama, and D. D. Strong, Differentiation of normal human bone cells by transforming growth factor-beta and 1,25(OH)2 vitamin D3. Metabolism 41, 42 – 48 (1992). 28. Y. Chen, A. Takeshita, K. Ozaki, S. Kitano, and S. Hanazawa, Transcriptional regulation by transforming growth factor beta of the expression of retinoic acid and retinoid X receptor genes in osteoblastic cells is mediated through AP – 1. J. Biol. Chem. 271, 31602 – 31606 (1996). 29. C. Pilbeam, Y. Rao, O. Voznesensky, H. Kawaguchi, C. Alander, L. Raisz, and H. Herschman, Transforming growth factor-beta1 regulation of prostaglandin G/H synthase-2 expression in osteoblastic MC3T3-E1 cells. Endocrinology 138, 4672 – 4682 (1997). 30. J. Massague, TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998). 31. R. Derynck, Y. Zhang, and X. H. Feng, Smads: Transcriptional activators of TGF-beta responses. Cell 95, 737 – 740 (1998). 32. Z. Q. Wang, C. Ovitt, A. E. Grigoriadis, U. Mohle-Steinlein, U. Ruther, and E. F. Wagner, Bone and haematopoietic defects in mice lacking c-fos. Nature 360, 741 – 745 (1992). 33. N. Udagawa, J. Chan, S. Wada, D. M. Findlay, J. A. Hamilton, and T. J. Martin, c-Fos antisense DNA inhibits proliferation of osteoclast progenitors in osteoclast development but not macrophage differentiation in vitro. Bone 18, 511 – 516 (1996). 34. S. J. Felts, E. S. Stoflet, C. T. Eggers, and M. J. Getz, Tissue factor gene transcription in serum-stimulated fibroblasts is mediated by recruitment of c-Fos into specific AP-1 DNA-binding complexes. Biochemistry 34, 12355 – 12362 (1995). 35. G. Jin and P. H. Howe, Regulation of clusterin gene expression by transforming growth factor beta. J. Biol. Chem. 272, 26620 – 26626 (1997). 36. Y. Zhang, X. H. Feng, and R. Derynck, Smad3 and Smad4 cooperate with c-Jun/c-Fos to mediate TGF-beta-induced transcription. Nature 394, 909 – 913 (1998). 37. A. Takeshita, K. Imai, S. Kato, S. Kitano, and S. Hanazawa, 1alpha,25-dehydroxyvitamin D3 synergism toward transforming growth factor-beta1-induced AP – 1 transcriptional activity in mouse osteoblastic cells via its nuclear receptor. J. Biol. Chem. 273, 14738 – 14744 (1998). 38. L. J. Jonk, S. Itoh, C. H. Heldin, P. ten Dijke, and W. Kruijer, Identification and functional characterization of a Smad binding element (SBE) in the JunB promoter that acts as a transforming growth factor-beta, activin, and bone morphogenetic protein-inducible enhancer. J. Biol. Chem. 273, 21145 – 21152 (1998). 39. J. Li, K. Tsuji, T. Komori, K. Miyazono, J. L. Wrana, Y. Ito, A. Nifuji, and M. Noda, Smad2 overexpression enhances Smad4 gene expression and suppresses CBFA1 gene expression in osteoblastic osteosarcoma ROS17/2.8 cells and primary rat calvaria cells. J. Biol. Chem. 273, 31009 – 31015 (1998). 40. M. M. Shull, I. Ormsby, A. B. Kier, S. Pawlowski, R. J. Diebold, M. Yin, R. Allen, C. Sidman, G. Proetzel, D. Calvin, et al., Targeted

424

41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

54.

55.

56.

57.

XUEZHONG QIN ET AL. disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 359, 693 – 699 (1992). A. B. Kulkarni, C. G. Huh, D. Becker, A. Geiser, M. Lyght, K. C. Flanders, A. B. Roberts, M. B. Sporn, J. M. Ward, and S. Karlsson, Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90, 770 – 774 (1993). G. Proetzel, S. A. Pawlowski, M. V. Wiles, M. Yin, G. P. Boivin, P. N. Howles, J. Ding, M. W. Ferguson, and T. Doetschman, Transforming growth factor-beta 3 is required for secondary palate fusion. Nature Genet. 11, 409 – 414 (1995). L. P. Sanford, I. Ormsby, A. C. Gittenberger-de Groot, H. Sariola, R. Friedman, G. P. Boivin, E. L. Cardell, and T. Doetschman, TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes. Development 124, 2659 – 2670 (1997). A. G. Geiser, Q. Q. Zeng, M. Sato, L. M. Helvering, T. Hirano, and C. H. Turner, Decreased bone mass and bone elasticity in mice lacking the transforming growth factor-beta1 gene. Bone 23, 87 – 93 (1998). M. M. Mozes, E. P. Bottinger, T. A. Jacot, and J. B. Kopp, Renal expression of fibrotic matrix proteins and of transforming growth factor-beta (TGF-beta) isoforms in TGF-beta transgenic mice. J. Am. Soc. Nephrol. 10, 271 – 280 (1999). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). Y. Chai, Y. Sasano, P. Bringas, Jr., M. Mayo, V. Kaartinen, N. Heisterkamp, J. Groffen, H. Slavkin, and C. Shuler, Characterization of the fate of midline epithelial cells during the fusion of mandibular prominences in vivo. Dev. Dyn. 208, 526 – 535 (1997). M. E. Joyce, A. B. Roberts, M. B. Sporn, and M. E. Bolander, Transforming growth factor-beta and the initiation of chondrogenesis and osteogenesis in the rat femur. J. Cell Biol. 110, 2195 – 2207 (1990). M. E. Bolander, Regulation of fracture repair by growth factors. Proc. Soc. Exp. Biol. Med. 200, 165 – 170 (1992). M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). C. Marcelli, A. J. Yates, and G. R. Mundy, In vivo effects of human recombinant transforming growth factor beta on bone turnover in normal mice. J. Bone Miner. Res. 5, 1087 – 1096 (1990). R. Fujimoto, T. Tanizawa, S. Nishida, N. Yamamoto, S. Soshi, N. Endo, and H. E. Takahashi, Local effects of transforming growth factor-beta1 on rat calvaria: changes depending on the dose and the injection site. J. Bone Miner. Metab. 17, 11 – 17 (1999). L. S. Beck, E. P. Amento, Y. Xu, L. Deguzman, W. P. Lee, T. Nguyen, and N. A. Gillett, TGF-beta 1 induces bone closure of skull defects: temporal dynamics of bone formation in defects exposed to rhTGF-beta 1. J. Bone Miner. Res. 8, 753 – 761 (1993). C. Bosch, B. Melsen, R. Gibbons, and K. Vargervik, Human recombinant transforming growth factor-beta 1 in healing of calvarial bone defects. J. Craniofac. Surg. 7, 300 – 310 (1996). L. McKinney and J. O. Hollinger, A bone regeneration study: Transforming growth factor-beta 1 and its delivery. J. Craniofac. Surg. 7, 36 – 45 (1996). G. Zellin, S. Beck, R. Hardwick, and A. Linde, Opposite effects of recombinant human transforming growth factor-beta 1 on bone regeneration in vivo: Effects of exclusion of periosteal cells by microporous membrane. Bone 22, 613 – 620 (1998). E. J. Mackie and U. Trechsel, Stimulation of bone formation in vivo by transforming growth factor-beta: Remodeling of woven bone and lack of inhibition by indomethacin. Bone 11, 295 – 300 (1990).

58. M. Lind, B. Schumacker, K. Soballe, J. Keller, F. Melsen, and C. Bunger, Transforming growth factor-beta enhances fracture healing in rabbit tibiae. Acta Orthop. Scand. 64, 553 – 556 (1993). 59. H. M. Nielsen, T. T. Andreassen, T. Ledet, and H. Oxlund, Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop. Scand. 65, 37 – 41 (1994). 60. J. Beaudreuil, G. Mbalaviele, M. Cohen-Solal, C. Morieux, M. C. De Vernejoul, and P. Orcel, Short-term local injections of transforming growth factor-beta 1 decrease ovariectomy-stimulated osteoclastic resorption in vivo in rats. J. Bone Miner. Res. 10, 971 – 977 (1995). 61. M. A. Critchlow, Y. S. Bland, and D. E. Ashhurst, The effect of exogenous transforming growth factor-beta 2 on healing fractures in the rabbit. Bone 16, 521 – 527 (1995). 62. L. S. Beck, R. L. Wong, L. DeGuzman, W. P. Lee, B. Ongpipattanakul, and T. H. Nguyen, Combination of bone marrow and TGFbeta1 augment the healing of critical-sized bone defects. J. Pharm. Sci. 87, 1379 – 1386 (1998). 63. U. M. Wikesjo, S. S. Razi, T. J. Sigurdsson, D. N. Tatakis, M. B. Lee, B. Ongpipattanakul, T. Nguyen, and R. Hardwick, Periodontal repair in dogs: Effect of recombinant human transforming growth factor-beta1 on guided tissue regeneration. J. Clin. Periodontol. 25, 475 – 481 (1998). 64. H. L. Glansbeek, H. M. van Beuningen, E. L. Vitters, P. M. van der Kraan, and W. B. van den Berg, Stimulation of articular cartilage repair in established arthritis by local administration of transforming growth factor-beta into murine knee joints. Lab. Invest. 78, 133 – 142 (1998). 65. D. N. Kalu, E. Salerno, Y. Higami, C. C. Liu, F. Ferraro, M. A. Salih, and B. H. Arjmandi, In vivo effects of transforming growth factorbeta 2 in ovariectomized rats. Bone Miner. 22, 209 – 220 (1993). 66. D. Rosen, S. C. Miller, E. DeLeon, A. Y. Thompson, H. Bentz, M. Mathews, and S. Adams, Systemic administration of recombinant transforming growth factor beta 2 (rTGF-beta 2] stimulates parameters of cancellous bone formation in juvenile and adult rats. Bone 15, 355 – 359 (1994). 67. E. Zerath, X. Holy, J. M. Mouillon, B. Farbos, M. Machwate, C. Andre, S. Renault, and P. J. Marie, TGF-beta2 prevents the impaired chondrocyte proliferation induced by unloading in growth plates of young rats. Life Sci. 61, 2397 – 2406 (1997). 68. D. Gazit, Y. Zilberman, G. Turgeman, S. Zhou, and A. Kahn, Recombinant TGF-beta1 stimulates bone marrow osteoprogenitor cell activity and bone matrix synthesis in osteopenic, old male mice. J. Cell. Biochem. 73, 379 – 389 (1999). 69. J. A. Carlino, H. R. Higley, J. R. Creson, P. D. Avis, Y. Ogawa, and L. R. Ellingsworth, Transforming growth factor beta 1 systemically modulates granuloid, erythroid, lymphoid, and thrombocytic cells in mice. Exp. Hematol. 20, 943 – 950 (1992). 70. J. B. Kopp, V. M. Factor, M. Mozes, P. Nagy, N. Sanderson, E. P. Bottinger, P. E. Klotman, and S. S. Thorgeirsson, Transgenic mice with increased plasma levels of TGF-beta 1 develop progressive renal disease. Lab. Invest. 74, 991 – 1003 (1996). 71. V. Nicolas, A. Prewett, P. Bettica, S. Mohan, R. D. Finkelman, D. J. Baylink, and J. R. Farley, Age-related decreases in insulin-like growth factor-I and transforming growth factor-beta in femoral cortical bone from both men and women: Implications for bone loss with aging. J. Clin. Endocrinol. Metab. 78, 1011 – 1016 (1994). 72. D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF-beta. J. Cell. Biochem. 70, 478 – 488 (1998). 73. R. D. Finkelman, T. A. Linkhart, S. Mohan, K. H. Lau, D. J. Baylink, and N. H. Bell, Vitamin D deficiency causes a selective reduction in deposition of transforming growth factor beta in rat bone: Possible mechanism for impaired osteoinduction. Proc. Natl. Acad. Sci. USA 88, 3657 – 3660 (1991). 74. R. D. Finkelman, N. H. Bell, D. D. Strong, L. M. Demers, and D. J. Baylink, Ovariectomy selectively reduces the concentration of

CHAPTER 14 Bone Growth Factors

75.

76.

77.

78.

79.

80.

81.

82.

83.

84. 85.

86.

87.

88.

89.

90.

91.

transforming growth factor beta in rat bone: Implications for estrogen deficiency-associated bone loss. Proc. Natl. Acad. Sci. USA 89, 12190 – 12193 (1992). C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). K. C. Westerlind and R. T. Turner, The skeletal effects of spaceflight in growing rats: Tissue-specific alterations in mRNA levels for TGFbeta. J. Bone Miner. Res. 10, 843 – 848 (1995). T. K. Sampath and A. H. Reddi, Dissociative extraction and reconstitution of extracellular matrix components involved in local bone differentiation. Proc. Natl. Acad. Sci. USA 78, 7599 – 7603 (1981). J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. W. Kriz, R. M. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528–1534 (1988). V. Rosen, J. M. Wozney, E. A. Wang, P. Cordes, A. Celeste, D. McQuaid, and L. Kurtzberg, Purification and molecular cloning of a novel group of BMPs and localization of BMP mRNA in developing bone. Connect. Tissue Res. 20, 313 – 319 (1989). E. Ozkaynak, D. C. Rueger, E. A. Drier, C. Corbett, R. J. Ridge, T. K. Sampath, and H. Oppermann, OP-1 cDNA encodes an osteogenic protein in the TGF-beta family. EMBO J. 9, 2085 – 2093 (1990). E. Ozkaynak, P. N. Schnegelsberg, D. F. Jin, G. M. Clifford, F. D. Warren, E. A. Drier, and H. Oppermann, Osteogenic protein-2: A new member of the transforming growth factor-beta superfamily expressed early in embryogenesis. J. Biol. Chem. 267, 25220 – 25227 (1992). A. H. Reddi, Cartilage morphogenesis: Role of bone and cartilage morphogenetic proteins, homeobox genes and extracellular matrix. Matrix Biol. 14, 599 – 606 (1995). A. H. Reddi, Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nature Biotechnol. 16, 247 – 252 (1998). M. Kawabata, T. Imamura, and K. Miyazono, Signal transduction by bone morphogenetic proteins. Cytokine Growth Factor Rev. 9, 49 – 61 (1998). J. M. Schmitt, K. Hwang, S. R. Winn, and J. O. Hollinger, Bone morphogenetic proteins: An update on basic biology and clinical relevance. J. Orthop. Res. 17, 269 – 278 (1999). V. M. Paralkar, A. L. Vail, W. A. Grasser, T. A. Brown, H. Xu, S. Vukicevic, H. Z. Ke, H. Qi, T. A. Owen, and D. D. Thompson, Cloning and characterization of a novel member of the transforming growth factor-beta/bone morphogenetic protein family. J. Biol. Chem. 273, 13760 – 13767 (1998). T. K. Sampath, J. E. Coughlin, R. M. Whetstone, D. Banach, C. Corbett, R. J. Ridge, E. Ozkaynak, H. Oppermann, and D. C. Rueger, Bovine osteogenic protein is composed of dimers of OP-1 and BMP-2A, two members of the transforming growth factor-beta superfamily. J. Biol. Chem. 265, 13198 – 13205 (1990). F. Lecanda, L. V. Avioli, and S. L. Cheng, Regulation of bone matrix protein expression and induction of differentiation of human osteoblasts and human bone marrow stromal cells by bone morphogenetic protein-2. J. Cell. Biochem. 67, 386 – 396 (1997). O. Fromigue, P. J. Marie, and A. Lomri, Bone morphogenetic protein – 2 and transforming growth factor-beta2 interact to modulate human bone marrow stromal cell proliferation and differentiation. J. Cell. Biochem. 68, 411 – 426 (1998). A. Chaudhari, E. Ron, and M. P. Rethman, Recombinant human bone morphogenetic protein-2 stimulates differentiation in primary cultures of fetal rat calvarial osteoblasts. Mol. Cell. Biochem. 167, 31 – 39 (1997).

425 92. Y. Torii, K. Hitomi, and N. Tsukagoshi, Synergistic effect of BMP-2 and ascorbate on the phenotypic expression of osteoblastic MC3T3E1 cells. Mol. Cell. Biochem. 165, 25 – 29 (1996). 93. T. Nakase, K. Takaoka, K. Masuhara, K. Shimizu, H. Yoshikawa, and T. Ochi, Interleukin-1 beta enhances and tumor necrosis factor-alpha inhibits bone morphogenetic protein-2-induced alkaline phosphatase activity in MC3T3-E1 osteoblastic cells. Bone 21, 17 – 21 (1997). 94. S. D. Boden, G. Hair, L. Titus, M. Racine, K. McCuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 95. I. W. Li, S. Cheifetz, C. A. McCulloch, K. T. Sampath, and J. Sodek, Effects of osteogenic protein-1 (OP-1, BMP-7) on bone matrix protein expression by fetal rat calvarial cells are differentiation stage specific. J. Cell. Physiol. 169, 115 – 125 (1996). 96. L. Erlacher, J. McCartney, E. Piek, P. ten Dijke, M. Yanagishita, H. Oppermann, and F. P. Luyten, Cartilage-derived morphogenetic proteins and osteogenic protein-1 differentially regulate osteogenesis. J. Bone Miner. Res. 13, 383 – 392 (1998). 97. M. Hazama, A. Aono, N. Ueno, and Y. Fujisawa, Efficient expression of a heterodimer of bone morphogenetic protein subunits using a baculovirus expression system. Biochem. Biophys. Res. Commun. 209, 859 – 866 (1995). 98. A. Aono, M. Hazama, K. Notoya, S. Taketomi, H. Yamasaki, R. Tsukuda, S. Sasaki, and Y. Fujisawa, Potent ectopic boneinducing activity of bone morphogenetic protein-4/7 heterodimer. Biochem. Biophys. Res. Commun. 210, 670 – 677 (1995). 99. D. I. Israel, J. Nove, K. M. Kerns, R. J. Kaufman, V. Rosen, K. A. Cox, and J. M. Wozney, Heterodimeric bone morphogenetic proteins show enhanced activity in vitro and in vivo. Growth Factors 13, 291 – 300 (1996). 100. H. Yamashita, P. Ten Dijke, C. H. Heldin, and K. Miyazono, Bone morphogenetic protein receptors. Bone 19, 569 – 574 (1996). 101. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy, and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP-2induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 102. H. Chang, D. Huylebroeck, K. Verschueren, Q. Guo, M. M. Matzuk, and A. Zwijsen, Smad5 knockout mice die at mid-gestation due to multiple embryonic and extraembryonic defects. Development 126, 1631 – 1642 (1999). 103. S. J. Tang, P. A. Hoodless, Z. Lu, M. L. Breitman, R. R. McInnes, J. L. Wrana, and M. Buchwald, The Tlx-2 homeobox gene is a downstream target of BMP signalling and is required for mouse mesoderm development. Development 125, 1877 – 1887 (1998). 104. M. Bei and R. Maas, FGFs and BMP4 induce both Msx1-independent and Msx1-dependent signaling pathways in early tooth development. Development 125, 4325 – 4333 (1998). 105. H. J. Kim, D. P. Rice, P. J. Kettunen, and I. Thesleff, FGF-, BMPand Shh-mediated signalling pathways in the regulation of cranial suture morphogenesis and calvarial bone development. Development 125, 1241 – 1251 (1998). 106. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 107. L. B. Zimmerman, J. M. De Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). 108. S. A. Holley, J. L. Neul, L. Attisano, J. L. Wrana, Y. Sasai, M. B. O’Connor, E. M. De Robertis, and E. L. Ferguson, The Xenopus dorsalizing factor noggin ventralizes Drosophila embryos by preventing DPP from activating its receptor. Cell 86, 607 – 617 (1996). 109. S. Piccolo, Y. Sasai, B. Lu, and E. M. De Robertis, Dorsoventral patterning in Xenopus: Inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86, 589 – 598 (1996).

426 110. W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992). 111. E. Gazzerro, V. Gangji, and E. Canalis, Bone morphogenetic proteins induce the expression of noggin, which limits their activity in cultured rat osteoblasts. J. Clin. Invest. 102, 2106 – 2114 (1998). 112. L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). 113. Y. Sasai, B. Lu, H. Steinbeisser, D. Geissert, L. K. Gont, and E. M. De Robertis, Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 79, 779 – 790 (1994). 114. W. N. Pappano, I. C. Scott, T. G. Clark, R. L. Eddy, T. B. Shows, and D. S. Greenspan, Coding sequence and expression patterns of mouse chordin and mapping of the cognate mouse chrd and human CHRD genes. Genomics 52, 236 – 239 (1998). 115. H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 116. G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 117. N. R. Dunn, G. E. Winnier, L. K. Hargett, J. J. Schrick, A. B. Fogo, and B. L. Hogan, Haploinsufficient phenotypes in Bmp4 heterozygous null mice and modification by mutations in Gli3 and Alx4. Dev. Biol. 188, 235 – 247 (1997). 118. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF beta superfamily. Cell 71, 399 – 410 (1992). 119. M. J. Solloway, A. T. Dudley, E. K. Bikoff, K. M. Lyons, B. L. Hogan, and E. J. Robertson, Mice lacking Bmp6 function. Dev. Genet. 22, 321 – 339 (1998). 120. N. Gena, C. Martin-Seisdedos, P. McCue, and C. M. Groce, BMP-7 null mutation in mice: Developmental defects in skeleton, kidney and eye. Exp. Cell Res. 230, 28 – 37 (1997). 121. T. Katagiri, S. Boorla, J. L. Frendo, B. L. Hogan, and G. Karsenty, Skeletal abnormalities in doubly heterozygous Bmp4 and Bmp7 mice. Dev. Genet. 22, 340 – 348 (1998). 122. G. Q. Zhao, K. Deng, P. A. Labosky, L. Liaw, and B. L. Hogan, The gene encoding bone morphogenetic protein 8B is required for the initiation and maintenance of spermatogenesis in the mouse. Genes Dev. 10, 1657 – 1669 (1996). 123. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF beta-superfamily. Nature 368, 639 – 643 (1994). 124. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGFbeta superfamily member. Nature Genet. 12, 315 – 317 (1996). 125. A. C. McPherron, A. M. Lawler, and S. J. Lee, Regulation of skeletal muscle mass in mice by a new TGF-beta superfamily member. Nature 387, 83 – 90 (1997). 126. J. Dong, D. F. Albertini, K. Nishimori, T. R. Kumar, N. Lu, and M. M. Matzuk, Growth differentiation factor – 9 is required during early ovarian folliculogenesis. Nature 383, 531 – 535 (1996). 127. R. Zhao, A. M. Lawler, and S. J. Lee, Characterization of GDF-10 expression patterns and null mice. Dev. Biol. 212, 68 – 79 (1999). 128. U. Ripamonti, S. S. Ma, B. van den Heever, and A. H. Reddi, Osteogenin, a bone morphogenetic protein, adsorbed on porous hydroxyapatite substrata, induces rapid bone differentiation in calvarial defects of adult primates. Plast. Reconstr. Surg. 90, 382 – 393 (1992). 129. R. Langer and J. P. Vacanti, Tissue engineering. Science 260, 920 – 926 (1993).

XUEZHONG QIN ET AL. 130. E. Hedner and A. Linde, Efficacy of bone morphogenetic protein (BMP) with osteopromotive membranes: An experimental study in rat mandibular defects. Eur. J. Oral Sci. 103, 236 – 241 (1995). 131. M. Niederwanger and M. R. Urist, Demineralized bone matrix supplied by bone banks for a carrier of recombinant human bone morphogenetic protein (rhBMP-2): A substitute for autogeneic bone grafts. J. Oral Implantol. 22, 210 – 215 (1996). 132. Y. Yamazaki, S. Oida, K. Ishihara, and N. Nakabayashi, Ectopic induction of cartilage and bone by bovine bone morphogenetic protein using a biodegradable polymeric reservoir. J. Biomed. Mater. Res. 30, 1 – 4 (1996). 133. H. Takita, E. Tsuruga, I. Ono, and Y. Kuboki, Enhancement by bFGF of osteogenesis induced by rhBMP-2 in rats. Eur. J. Oral Sci. 105, 588 – 592 (1997). 134. T. N. Gerhart, C. A. Kirker-Head, M. J. Kriz, M. E. Holtrop, G. E. Hennig, J. Hipp, S. H. Schelling, and E. Wang, Healing segmental femoral defects in sheep using recombinant human bone morphogenetic protein. Clin. Orthop. 293, 317 – 326 (1993). 135. S. D. Cook, M. W. Wolfe, S. L. Salkeld, and D. C. Rueger, Effect of recombinant human osteogenic protein-1 on healing of segmental defects in non-human primates. J. Bone Joint. Surg. Am. 77, 734 – 750 (1995). 136. H. D. Zegzula, D. C. Buck, J. Brekke, J. M. Wozney, and J. O. Hollinger, Bone formation with use of rhBMP-2 (recombinant human bone morphogenetic protein-2). J. Bone Joint Surg. Am. 79, 1778 – 1790 (1997). 137. J. P. Sheehan, D. F. Kallmes, J. M. Sheehan, J. A. Jane, Jr., A. H. Fergus, C. G. diPierro, N. E. Simmons, D. D. Makel, and G. A. Helm, Molecular methods of enhancing lumbar spine fusion. Neurosurgery 39, 548 – 554 (1996). 138. S. D. Boden, G. J. Martin, Jr., M. A. Morone, J. L. Ugbo, and P. A. Moskovitz, Posterolateral lumbar intertransverse process spine arthrodesis with recombinant human bone morphogenetic protein 2/hydroxyapatite-tricalcium phosphate after laminectomy in the nonhuman primate. Spine 24, 1179 – 1185 (1999). 139. M. Nakashima, Induction of dentin formation on canine amputated pulp by recombinant human bone morphogenetic proteins (BMP)-2 and -4. J. Dent. Res. 73, 1515 – 1522 (1994). 140. X. Si, Y. Jin and L. Yang, Induction of new bone by ceramic bovine bone with recombinant human bone morphogenetic protein 2 and transforming growth factor beta. Int. J. Oral Maxillofac. Surg. 27, 310 – 314 (1998). 141. I. Ono, M. Inoue, and Y. Kuboki, Promotion of the osteogenetic activity of recombinant human bone morphogenetic protein by prostaglandin E1. Bone 19, 581 – 588 (1996). 142. G. Mundy, R. Garrett, S. Harris, J. Chan, D. Chen, G. Rossini, B. Boyce, M. Zhao, and G. Gutierrez, Stimulation of bone formation in vitro and in rodents by statins. Science 286, 1946 – 1949 (1999). 143. J. Fang, Y. Y. Zhu, E. Smiley, J. Bonadio, J. P. Rouleau, S. A. Goldstein, L. K. McCauley, B. L. Davidson, and B. J. Roessler, Stimulation of new bone formation by direct transfer of osteogenic plasmid genes. Proc. Natl. Acad. Sci. USA 93, 5753 – 5758 (1996). 144. J. R. Lieberman, L. Q. Le, L. Wu, G. A. Finerman, A. Berk, O. N. Witte, and S. Stevenson, Regional gene therapy with a BMP-2-producing murine stromal cell line induces heterotopic and orthotopic bone formation in rodents. J. Orthop. Res. 16, 330 – 339 (1998). 145. K. D. Riew, N. M. Wright, S. Cheng, L. V. Avioli, and J. Lou, Induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene in a rabbit spinal fusion model. Calcif. Tissue Int. 63, 357 – 360 (1998). 146. J. Lou, F. Xu, K. Merkel, and P. Manske, Gene therapy: Adenovirus-mediated human bone morphogenetic protein-2 gene transfer induces mesenchymal progenitor cell proliferation and differentiation in vitro and bone formation in vivo. J. Orthop. Res. 17, 43 – 50 (1999).

CHAPTER 14 Bone Growth Factors 147. D. S. Musgrave, P. Bosch, S. Ghivizzani, P. D. Robbins, C. H. Evans, and J. Huard, Adenovirus-mediated direct gene therapy with bone morphogenetic protein-2 produces bone. Bone 24, 541 – 547 (1999). 148. T. D. Alden, D. D. Pittman, G. R. Hankins, E. J. Beres, J. A. Engh, S. Das, S. B. Hudson, K. M. Kerns, D. F. Kallmes, and G. A. Helm, In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum. Gene Ther. 10, 2245 – 2253 (1999). 149. E. Canalis, Insulin like growth factors and the local regulation of bone formation. Bone 14, 273 – 276 (1993). 150. C. J. Rosen, L. R. Donahue, and S. J. Hunter, Insulin-like growth factors and bone: the osteoporosis connection. Proc. Soc. Exp. Biol. Med. 206, 83 – 102 (1994). 151. C. J. Rosen and L. R. Donahue, Insulin-like growth factors and bone: the osteoporosis connection revisited. Proc. Soc. Exp. Biol. Med. 219, 1 – 7 (1998). 152. S. Mohan and D. J. Baylink, IGF system components and their role in bone metabolism In “Contemporary Endocrinology: The IGF System” (R. Rosenfeld and C. Roberts, eds.), pp. 457 – 496, Humana Press, Inc., Totowa, NJ, (1999). 153. S. Mohan and D. J. Baylink, The role of insulin-like growth factor-II in the coupling of bone formation and resorption In “Modern Concepts of Insulin-like Growth Factors” (E. M. Spencer, ed.), pp. 169 – 184, Elsevier Science, New York, 1991. 154. C. Schmid, T. Steiner and E. R. Froesch, (1984). Insulin-like growth factor I supports differentiation of cultured osteoblast-like cells. FEBS Lett 173, 48 – 52 (1984). 155. D. D. Strong, A. L. Beachler, J. E. Wergedal, and T. A. Linkhart, Insulin-like growth factor II and transforming growth factor beta regulate collagen expression in human osteoblast-like cells in vitro. J. Bone Miner. Res. 6, 15 – 23 (1991). 156. T. Onishi, S. Kinoshita, S. Shintani, S. Sobue, and T. Ooshima, Stimulation of proliferation and differentiation of dog dental pulp cells in serum-free culture medium by insulin-like growth factor. Arch. Oral Biol. 44, 361 – 371 (1999). 157. P. A. Hill, A. Tumber, and M. C. Meikle, Multiple extracellular signals promote osteoblast survival and apoptosis. Endocrinology 138, 3849 – 3858 (1997). 158. A. Kawakami, T. Nakashima, M. Tsuboi, S. Urayama, N. Matsuoka, H. Ida, Y. Kawabe, H. Sakai, K. Migita, T. Aoyagi, M. Nakashima, K. Maeda, and K. Eguchi, Insulin-like growth factor I stimulates proliferation and Fas-mediated apoptosis of human osteoblasts. Biochem. Biophys. Res. Commun. 247, 46 – 51 (1998). 159. J. Farley and B. Coffing, Apoptosis may determine the rate of appearance of skeletal alkaline phosphatase in serum. J. Bone Miner. Res. 14 (Suppl 1), (1999), [Abstract SA148]. 160. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122, 254 – 260 (1988). 161. E. Canalis, S. Rydziel, A. M. Delany, S. Varghese, and J. J. Jeffrey, Insulin-like growth factors inhibit interstitial collagenase synthesis in bone cell cultures. Endocrinology 136, 1348 – 1354 (1995). 162. P. A. Hill, J. J. Reynolds, and M. C. Meikle, Osteoblasts mediate insulin-like growth factor-I and -II stimulation of osteoclast formation and function. Endocrinology 136, 124 – 131 (1995). 163. H. Kaji, T. Sugimoto, M. Kanatani, K. Nishiyama, M. Nasu, and K. Chihara, Insulin-like growth factor-I mediates osteoclast-like cell formation stimulated by parathyroid hormone. J. Cell. Physiol. 172, 55 – 62 (1997). 164. J. Guicheux, D. Heymann, A. V. Rousselle, F. Gouin, P. Pilet, S. Yamada, and G. Daculsi, Growth hormone stimulatory effects on osteoclastic resorption are partly mediated by insulin-like growth factor I: An in vitro study. Bone 22, 25 – 31 (1998). 165. P. Hou, T. Sato, W. Hofstetter, and N. T. Foged, Identification and characterization of the insulin-like growth factor I receptor in mature rabbit osteoclasts. J. Bone Miner. Res. 12, 534 – 540 (1997).

427 166. J. I. Jones and D. R. Clemmons, Insulin-like growth factors and their binding proteins: Biological actions. Endocr. Rev. 16, 3 – 34 (1995). 167. T. Kadowaki, K. Tobe, R. Honda-Yamamoto, H. Tamemoto, Y. Kaburagi, K. Momomura, K. Ueki, Y. Takahashi, T. Yamauchi, Y. Akanuma, and Y. Yazaki, Signal transduction mechanism of insulin and insulin-like growth factor-1. Endocr. J. 43 (Suppl), S33 – 41 (1996). 168. S. Li, M. Resnicoff, and R. Baserga, Effect of mutations at serines 1280 – 1283 on the mitogenic and transforming activities of the insulin-like growth factor I receptor. J. Biol. Chem. 271, 12254 – 12260 (1996). 169. R. Baserga, M. Resnicoff, C. D’Ambrosio, and B. Valentinis, The role of the IGF-I receptor in apoptosis. Vitam. Horm. 53, 65 – 98 (1997). 170. E. Surmacz, C. Sell, J. Swantek, H. Kato, C. T. Roberts, Jr., D. LeRoith, and R. Baserga, Dissociation of mitogenesis and transforming activity by C-terminal truncation of the insulin-like growth factor-I receptor. Exp. Cell Res. 218, 370 – 380 (1995). 171. R. O’Connor, A. Kauffmann-Zeh, Y. Liu, S. Lehar, G. I. Evan, R. Baserga, and W. A. Blattler, Identification of domains of the insulin-like growth factor I receptor that are required for protection from apoptosis. Mol. Cell. Biol. 17, 427 – 435 (1997). 172. T. M. DeChiara, E. J. Robertson, and A. Efstratiadis, Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64, 849 – 859 (1991). 173. J. Wang, J. Zhou, L. Powell-Braxton, C. Bondy, Effects of IGF – 1 gene deletion on postnatal growth patterns. Endocrinology 140, 3391 – 3394 (1999). 174. J. Baker, J. P. Liu, E. J. Robertson, and A. Efstratiadis, Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75, 73 – 82 (1993). 175. L. S. Mathews, R. E. Hammer, R. R. Behringer, A. J. D’Ercole, G. I. Bell, R. L. Brinster, and R. D. Palmiter, Growth enhancement of transgenic mice expressing human insulin-like growth factor I. Endocrinology 123, 2827 – 2833 (1988). 176. G. Zhao, M. C. Faugere, H. Malluche, J. Wang, T. Nakayama, J. Pike, J. Fagin, and T. Clemens, Osteoblast-specific overexpression of IGF-I in transgenic mice increases cancellous bone volume. J. Bone Miner. Res. 23 (Suppl 1), (1988) [Abstract 1099]. 177. E. M. Spencer, C. C. Liu, E. C. Si, and G. A. Howard, In vivo actions of insulin-like growth factor-I (IGF-I) on bone formation and resorption in rats. Bone 12, 21 – 26 (1991). 178. H. N. Rosen, V. Chen, A. Cittadini, S. L. Greenspan, P. S. Douglas, A. C. Moses, and W. G. Beamer, Treatment with growth hormone and IGF-I in growing rats increases bone mineral content but not bone mineral density. J. Bone Miner. Res. 10, 1352 – 1358 (1995). 179. N. Miyakoshi, C. Richman, X. Qin, D. J. Baylink, and S. Mohan, Effects of recombinant insulin-like growth factor-binding protein-4 on bone formation parameters in mice. Endocrinology 140, 5719 – 5728 (1999). 180. H. P. Guler, J. Zapf, E. Scheiwiller, and E. R. Froesch, Recombinant human insulin-like growth factor I stimulates growth and has distinct effects on organ size in hypophysectomized rats. Proc. Natl. Acad. Sci. USA 85, 4889 – 4893 (1988). 181. P. Ammann, R. Rizzoli, K. Muller, D. Slosman, and J. P. Bonjour, IGF-I and pamidronate increase bone mineral density in ovariectomized adult rats. Am. J. Physiol. 265, E770 – E776 (1993). 182. J. Verhaeghe, R. van Bree, E. Van Herck, H. Thomas, A. Skottner, J. Dequeker, L. Mosekilde, T. A. Einhorn, and R. Bouillon, Effects of recombinant human growth hormone and insulin-like growth factor-I, with or without 17 beta-estradiol, on bone and mineral homeostasis of aged ovariectomized rats. J. Bone Miner. Res. 11, 1723 – 1735 (1996). 183. A. Cemborain, I. Castilla-Cortazar, M. Garcia, J. Quiroga, B. Muguerza, A. Picardi, S. Santidrian, and J. Prieto, Osteopenia in

428

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

196. 197.

198.

199.

200.

XUEZHONG QIN ET AL. rats with liver cirrhosis: beneficial effects of IGF-I treatment. J. Hepatol. 28, 122 – 131 (1998). J. Isgaard, A. Nilsson, A. Lindahl, J. O. Jansson, and O. G. Isaksson, Effects of local administration of GH and IGF-1 on longitudinal bone growth in rats. Am. J. Physiol. 250, E367 – E372 (1986). A. Wakisaka, H. Tanaka, J. Barnes, and C. T. Liang, Effect of locally infused IGF-I on femoral gene expression and bone turnover activity in old rats. J. Bone Miner. Res. 13, 13 – 19 (1998). M. Ishibe, T. Ishibashi, K. Kaneda, T. Koda, R. N. Rosier, and J. E. Puzas, Stimulation of bone formation in vivo by insulin-like growth factor-II in rats. Calcif. Tissue Int. 63, 36 – 38 (1998). A. G. Johansson, E. Lindh, and S. Ljunghall, Insulin-like growth factor I stimulates bone turnover in osteoporosis. Lancet 339, 1619 (1992). P. R. Ebeling, J. D. Jones, W. M. O’Fallon, C. H. Janes, and B. L. Riggs, Short-term effects of recombinant human insulin-like growth factor I on bone turnover in normal women. J. Clin. Endocrinol. Metab. 77, 1384 – 1387 (1993). A. G. Johansson, E. Lindh, G. Kollerup, and S. Ljunghall, Dosedependent effects of IGF-I on the metabolism of collagen type I in male osteoporosis. J. Bone Miner. Res. 10 (Suppl 1), S249 (1995) [Abstract S393]. A. G. Johansson, E. Lindh, W. F. Blum, G. Kollerup, O. H. Sorensen, and S. Ljunghall, Effects of growth hormone and insulin-like growth factor I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 81, 44 – 48 (1996). L. J. Ghiron, J. L. Thompson, L. Holloway, R. L. Hintz, G. E. Butterfield, A. R. Hoffman, and R. Marcus, Effects of recombinant insulinlike growth factor-I and growth hormone on bone turnover in elderly women. J. Bone Miner. Res. 10, 1844 – 1852 (1995). P. G. Campbell and J. F. Novak, Insulin-like growth factor binding protein (IGFBP) inhibits IGF action on human osteosarcoma cells. J. Cell. Physiol. 149, 293 – 300 (1991). S. Mohan, C. M. Bautista, J. Wergedal, and D. J. Baylink, Isolation of an inhibitory insulin-like growth factor (IGF) binding protein from bone cell-conditioned medium: A potential local regulator of IGF action. Proc. Natl. Acad. Sci. USA 86, 8338 – 8342 (1989). J. H. Feyen, D. B. Evans, C. Binkert, G. F. Heinrich, S. Geisse, and H. P. Kocher, Recombinant human [Cys281]insulin-like growth factor-binding protein 2 inhibits both basal and insulin-like growth factor I-stimulated proliferation and collagen synthesis in fetal rat calvariae. J. Biol. Chem. 266, 19469 – 19474 (1991). M. Ernst and G. A. Rodan, Increased activity of insulin-like growth factor (IGF) in osteoblastic cells in the presence of growth hormone (GH): Positive correlation with the presence of the GH-induced IGFbinding protein BP – 3. Endocrinology 127, 807 – 814 (1990). S. Mohan, Insulin-like growth factor binding proteins in bone cell regulation. Growth Regul. 3 67 – 70 (1993). S. Amarnani, H. L. Merriman, T. A. Linkhart, D. J. Baylink, and S. Mohan, Autocrine regulators of MC3T3-E1 cell proliferation. J. Bone Miner. Res. 8, 157 – 165 (1993). S. K. Durham, M. C. Kiefer, B. L. Riggs, and C. A. Conover, Regulation of insulin-like growth factor binding protein 4 by a specific insulin-like growth factor binding protein 4 proteinase in normal human osteoblast-like cells: Implications in bone cell physiology. J. Bone Miner. Res. 9, 111 – 117 (1994). S. Mohan, Y. Nakao, Y. Honda, E. Landale, U. Leser, C. Dony, K. Lang, and D. J. Baylink, Studies on the mechanisms by which insulin-like growth factor (IGF) binding protein-4 (IGFBP-4) and IGFBP-5 modulate IGF actions in bone cells. J. Biol. Chem. 270, 20424 – 20431 (1995). C. M. Bautista, D. J. Baylink, and S. Mohan, Isolation of a novel insulin-like growth factor (IGF) binding protein from human bone: A potential candidate for fixing IGF-II in human bone. Biochem. Biophys. Res. Commun. 176, 756 – 763 (1991).

201. D. L. Andress and R. S. Birnbaum, Human osteoblast-derived insulin-like growth factor (IGF) binding protein-5 stimulates osteoblast mitogenesis and potentiates IGF action. J. Biol. Chem. 267, 22467 – 22472 (1992). 202. N. Srinivasan, D. Edwall, T. A. Linkhart, D. J. Baylink, and S. Mohan, Insulin-like growth factor-binding protein-6 produced by human PC-3 prostate cancer cells: isolation, characterization and its biological action. J. Endocrinol. 149, 297 – 303 (1996). 203. B. Valentinis, A. Bhala, T. DeAngelis, R. Baserga, and P. Cohen, The human insulin-like growth factor (IGF) binding protein-3 inhibits the growth of fibroblasts with a targeted disruption of the IGF-I receptor gene. Mol. Endocrinol. 9, 361 – 367 (1995). 204. C. Lalou, C. Lassarre, and M. Binoux, A proteolytic fragment of insulin-like growth factor (IGF) binding protein-3 that fails to bind IGFs inhibits the mitogenic effects of IGF-I and insulin. Endocrinology 137, 3206 – 3212 (1996). 205. S. Mohseni-Zadeh and M. Binoux, Insulin-like growth factor (IGF) binding protein – 3 interacts with the type 1 IGF receptor, reducing the affinity of the receptor for its ligand: An alternative mechanism in the regulation of IGF action. Endocrinology 138, 5645 – 5648 (1997). 206. S. D. Chernausek, C. E. Smith, K. L. Duffin, W. H. Busby, G. Wright, and D. R. Clemmons, Proteolytic cleavage of insulin-like growth factor binding protein 4 (IGFBP-4). Localization of cleavage site to non-homologous region of native IGFBP-4. J. Biol. Chem. 270, 11377 – 11382 (1995). 207. C. A. Conover, S. K. Durham, J. Zapf, F. R. Masiarz, and M. C. Kiefer, Cleavage analysis of insulin-like growth factor (IGF)-dependent IGF-binding protein-4 proteolysis and expression of proteaseresistant IGF-binding protein-4 mutants. J. Biol. Chem. 270, 4395 – 4400 (1995). 208. X. Qin, D. D. Strong, D. J. Baylink, and S. Mohan, Structurefunction analysis of the human insulin-like growth factor binding protein-4. J. Biol. Chem. 273, 23509 – 23516 (1998). 209. X. Qin, D. Byun, D. D. Strong, D. J. Baylink, and S. Mohan, Studies on the role of human insulin-like growth factor-II (IGF-II)-dependent IGF binding protein (hIGFBP)-4 protease in human osteoblasts using protease-resistant IGFBP-4 analogs. J. Bone Miner. Res. 14, 2079 – 2088 (1999). 210. W. Li, J. Fawcett, H. R. Widmer, P. J. Fielder, R. Rabkin, and G. A. Keller, Nuclear transport of insulin-like growth factor-I and insulinlike growth factor binding protein-3 in opossum kidney cells. Endocrinology 138, 1763 – 1766 (1997). 211. K. Rajkumar, D. Barron, M. S. Lewitt, and L. J. Murphy, Growth retardation and hyperglycemia in insulin-like growth factor binding protein-1 transgenic mice. Endocrinology 136, 4029 – 4034 (1995). 212. A. Hoeflich, M. Wu, S. Mohan, J. Foll, R. Wanke, T. Froehlich, G. J. Arnold, H. Lahm, H. J. Kolb, and E. Wolf, Overexpression of insulin-like growth factor-binding protein-2 in transgenic mice reduces postnatal body weight gain. Endocrinology 140, 5488 – 5496 (1999). 213. L. J. Murphy, P. Molnar, X. Lu, and H. Huang, Expression of human insulin-like growth factor-binding protein-3 in transgenic mice. J. Mol. Endocrinol. 15, 293 – 303 (1995). 214. J. Wang, W. Niu, D. P. Witte, S. D. Chernausek, Y. E. Nikiforov, T. L. Clemens, B. Sharifi, A. R. Strauch, and J. A. Fagin, Overexpression of insulin-like growth factor-binding protein-4 (IGFBP-4) in smooth muscle cells of transgenic mice through a smooth muscle alpha-actin-IGFBP-4 fusion gene induces smooth muscle hypoplasia. Endocrinology 139, 2605 – 2614 (1998). 215. W. Ni, K. Rajkumar, J. I. Nagy, and L. J. Murphy, Impaired brain development and reduced astrocyte response to injury in transgenic mice expressing IGF binding protein-1. Brain Res. 769, 97 – 107 (1997). 216. C. M. Bagi, R. Brommage, L. Deleon, S. Adams, D. Rosen, and A. Sommer, Benefit of systemically administered rhIGF-I and

CHAPTER 14 Bone Growth Factors

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230. 231.

rhIGF-I/IGFBP-3 on cancellous bone in ovariectomized rats. J. Bone Miner. Res. 9, 1301 – 1312 (1994). K. Narusawa, T. Nakamura, K. Suzuki, Y. Matsuoka, L. J. Lee, H. Tanaka, Y. Seino, The effects of recombinant human insulin-like growth factor (rhIGF-1 and rhIGF-1/IGF binding protein-3 administration on rat osteopenia induced by ovariectomy with concomitant bilateral sciatic neurectomy. J. Bone Miner. Res. 10, 1853 – 1864 (1995). C. Richman, D. J. Baylink, K. Lang, C. Dony, and S. Mohan, Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140, 4699 – 4705 (1999). S. K. Durham, B. L. Riggs, and C. A. Conover, The insulin-like growth factor-binding protein-4 (IGFBP-4)-IGFBP-4 protease system in normal human osteoblast-like cells: regulation by transforming growth factor-beta. J. Clin. Endocrinol. Metab. 79, 1752 – 1758 (1994). S. Kanzaki, S. Hilliker, D. J. Baylink, and S. Mohan, Evidence that human bone cells in culture produce insulin-like growth factor-binding protein-4 and -5 proteases. Endocrinology 134, 383 – 392 (1994). X. Qin, D. Byun, D. J. Baylink, and S. Mohan, Evidence that IGF binding domain in IGF binding protein (IGFBP)-4 is essential for the IGF-II enhancement of IGFBP-4 proteolysis by osteoblast-produced IGFBP-4 protease. J. Bone Miner. Res. 14 (Suppl 1), S462 (1999) [Abstract SU123]. J. B. Lawrence, C. Oxvig, M. T. Overgaard, L. Sottrup-Jensen, G. J. Gleich, L. G. Hays, J. R. Yates, III, and C. A. Conover, The insulinlike growth factor (IGF)-dependent IGF binding protein-4 protease secreted by human fibroblasts is pregnancy-associated plasma protein-A. Proc. Natl. Acad. Sci. USA 96, 3149 – 3153 (1999). D. Byun, S. Mohan, C. Kim, K. Suh, M. Yoo, H. Lee, D. J. Baylink, and X. Qin, Studies on the human pregnancy-induced insulin-like growth factor binding protein (IGFBP)-4 poetesses in serum: Determination of IGF-II dependency and localization of cleavage site. J. Clin. Endocrinol. Metab. 85, 373 – 381 (2000). T. Kristensen, C. Oxvig, O. Sand, N. P. Moller, and L. SottrupJensen, Amino acid sequence of human pregnancy-associated plasma protein-A derived from cloned cDNA. Biochemistry 33, 1592 – 1598 (1994). C. A. Conover and M. C. Kiefer, Regulation and biological effect of endogenous insulin-like growth factor binding protein-5 in human osteoblastic cells. J. Clin. Endocrinol. Metab. 76, 1153 – 1159 (1993). C. A. Conover, L. K. Bale, J. T. Clarkson, and O. Torring, Regulation of insulin-like growth factor binding protein-5 messenger ribonucleic acid expression and protein availability in rat osteoblast-like cells. Endocrinology 132, 2525 – 2530 (1993). R. Knutsen, Y. Honda, D. D. Strong, T. K. Sampath, D. J. Baylink, and S. Mohan, Regulation of insulin-like growth factor system components by osteogenic protein-1 in human bone cells. Endocrinology 136, 857 – 865 (1995). K. M. Thrailkill, L. D. Quarles, H. Nagase, K. Suzuki, D. M. Serra, and J. L. Fowlkes, Characterization of insulin-like growth factorbinding protein 5-degrading proteases produced throughout murine osteoblast differentiation. Endocrinology 136, 3527 – 3533 (1995). C. Libanati, D. J. Baylink, E. Lois-Wenzel, N. Srinvasan, and S. Mohan, Studies on the potential mediators of skeletal changes occurring during puberty in girls. J. Clin. Endocrinol. Metab. 84, 2807 – 2814 (1999). M. R. Hammerman, Insulin-like growth factors and aging. Endocrinol. Metab. Clin. North Am. 16, 995 – 1011 (1987). E. S. Kurland, C. J. Rosen, F. Cosman, D. McMahon, F. Chan, E. Shane, R. Lindsay, D. Dempster, and J. P. Bilezikian, Insulin-like growth factor-I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 82, 2799 – 2805 (1997).

429 232. E. Romagnoli, S. Minisola, V. Carnevale, R. Rosso, M. T. Pacitti, A. Scarda, L. Scarnecchia, and G. Mazzuoli, Circulating levels of insulin-like growth factor binding protein 3 (IGFBP-3) and insulinlike growth factor I (IGF-I) in perimenopausal women. Osteoporos. Int. 4, 305 – 308 (1994). 233. S. Boonen, S. Mohan, J. Dequeker, J. Aerssens, D. Vanderschueren, G. Verbeke, P. Broos, R. Bouillon, and D. J. Baylink, Down-regulation of the serum stimulatory components of the insulin-like growth factor (IGF) system (IGF-I, IGF-II, IGF binding protein [BP]-3, and IGFBP-5) in age-related (Type II) femoral neck osteoporosis. J. Bone Miner. Res. 14, 2150 – 2158 (1999). 234. S. Mohan, J. R. Farley, and D. J. Baylink, Age-related changes in IGFBP-4 and IGFBP-5 levels in human serum and bone: Implications for bone loss with aging. Prog. Growth Factor Res. 6, 465 – 473 (1995). 235. C. Rosen, L. R. Donahue, S. Hunter, M. Holick, H. Kavookjian, A. Kirschenbaum, S. Mohan, and D. J. Baylink, The 24/25-kDa serum insulin-like growth factor-binding protein is increased in elderly women with hip and spine fractures. J. Clin. Endocrinol. Metab. 74, 24 – 27 (1992). 236. Y. Honda, E. C. Landale, D. D. Strong, D. J. Baylink, and S. Mohan, S. Recombinant synthesis of insulin-like growth factor-binding protein-4 (IGFBP-4): Development, validation, and application of a radioimmunoassay for IGFBP-4 in human serum and other biological fluids. J. Clin. Endocrinol. Metab. 81, 1389 – 1396 (1996). 237. H. Tanaka, Liang, C. T. Mitogenic activity but not phenotype expression of rat osteoprogenitor cells in response to IGF-I is impaired in aged rats. Mech. Aging Dev. 92, 1 – 10 (1996). 238. J. Pfeilschifter, I. Diel, U. Pilz, K. Brunotte, A. Naumann, and R. Ziegler, Mitogenic responsiveness of human bone cells in vitro to hormones and growth factors decreases with age. J. Bone Miner. Res. 8, 707 – 717 (1993). 239. Y. D’ Avis P, C. R. Frazier, J. R. Shapiro, and N. S. Fedarko, Agerelated changes in effects of insulin-like growth factor I on human osteoblast-like cells. Biochem. J. 324, 753 – 760 (1997). 240. A. M. Delany and E. Canalis, Transcriptional repression of insulinlike growth factor I by glucocorticoids in rat bone cells. Endocrinology 136, 4776 – 4781 (1995). 241. S. Rydziel and E. Canalis, Cortisol represses insulin-like growth factor II receptor transcription in skeletal cell cultures. Endocrinology 136, 4254 – 4260 (1995). 242. B. Gabbitas and E. Canalis, Cortisol enhances the transcription of insulin-like growth factor-binding protein-6 in cultured osteoblasts. Endocrinology 137, 1687 – 1692 (1996). 243. B. Gabbitas J. M. Pash, A. M. Delany, and E. Canalis, Cortisol inhibits the synthesis of insulin-like growth factor-binding protein-5 in bone cell cultures by transcriptional mechanisms. J. Biol. Chem. 271, 9033 – 9038 (1996). 244. T. Chevalley, D. D. Strong, S. Mohan, D. Baylink, and T. A. Linkhart, Evidence for a role for insulin-like growth factor binding proteins in glucocorticoid inhibition of normal human osteoblast-like cell proliferation. Eur. J. Endocrinol. 134, 591 – 601 (1996). 245. C. Jux, K. Leiber, U. Hugel, W. Blum, C. Ohlsson, G. Klaus, and O. Mehls, Dexamethasone impairs growth hormone (GH)-stimulated growth by suppression of local insulin-like growth factor (IGF)-I production and expression of GH- and IGF-I-receptor in cultured rat chondrocytes. Endocrinology 139, 3296 – 3305 (1998). 246. C. Skjaerbaek, J. Frystyk, T. Grofte, A. Flyvbjerg, M. S. Lewitt, R. C. Baxter, and H. Orskov, Serum free insulin-like growth factor-I is dose-dependently decreased by methylprednisolone and related to body weight changes in rats. Growth Horm. IGF Res. 9, 74 – 80 (1999). 247. C. Libanati, S. Rajaram, D. J. Baylink, and S. Mohan, In a longitudinal prospective study, glucocorticoid therapy causes a marked reduction in serum bone formation parameters and stimulatory insulin-like

430

248. 249.

250. 251.

252.

253.

254.

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

XUEZHONG QIN ET AL. growth factor (IGF) system components. J. Bone Miner. Res. 11 (Suppl 1), S150 (1966) [Abstract P239]. G. Szebenyi and J. F. Fallon, Fibroblast growth factors as multifunctional signaling factors. Int. Rev. Cytol. 185, 45 – 106 (1999). H. A. Armelin, Pituitary extracts and steroid hormones in the control of 3T3 cell growth. Proc. Natl. Acad. Sci. USA 70, 2702 – 2706 (1973). C. Dickson and G. Peters, Potential oncogene product related to growth factors (letter). Nature 326, 833 (1987). M. Taira, T. Yoshida, K. Miyagawa, H. Sakamoto, M. Terada, and T. Sugimura, cDNA sequence of human transforming gene hst and identification of the coding sequence required for transforming activity. Proc. Natl. Acad. Sci. USA 84, 2980 – 2984 (1987). X. Zhan, B. Bates, X. G. Hu, and M. Goldfarb, The human FGF – 5 oncogene encodes a novel protein related to fibroblast growth factors. Molec. Cell. Biol. 8, 3487 – 3495 (1988). I. Marics, J. Adelaide, F. Raybaud, M. G. Mattei, F. Coulier, J. Planche, O. de Lapeyriere, and D. Birnbaum, Characterization of the HST-related FGF6 gene, a new member of the fibroblast growth factor gene family. Oncogene 4, 335 – 340 (1989). P. W. Finch, J. S. Rubin, T. Miki, D. Ron, and S. A. Aaronson, Human KGF is FGF-related with properties of a paracrine effector of epithelial cell growth. Science 245, 752 – 755 (1989). A. Tanaka, K. Miyamoto, N. Minamino, M. Takeda, B. Sato, H. Matsuo, and K. Matsumoto, Cloning and characterization of an androgen-induced growth factor essential for the androgen-dependent growth of mouse mammary carcinoma cells. Proc. Natl. Acad. Sci. USA 89, 8928 – 8932 (1992). M. Miyamoto, K. Naruo, C. Seko, S. Matsumoto, T. Kondo, and T. Kurokawa, Molecular cloning of a novel cytokine cDNA encoding the ninth member of the fibroblast growth factor family, which has a unique secretion property. Mol. Cell. Biol. 13, 4251 – 4259 (1993). P. M. Smallwood, I. Munoz-Sanjuan, P. Tong, J. P. Macke, S. H. Hendry, D. J. Gilbert, N. G. Copeland, N. A. Jenkins, and J. Nathans, Fibroblast growth factor (FGF) homologous factors: New members of the FGF family implicated in nervous system development. Proc. Natl. Acad. Sci. USA 93, 9850 – 9857 (1996). T. Nishimura, Y. Utsunomiya, M. Hoshikawa, H. Ohuchi, and N. Itoh, Structure and expression of a novel human FGF, FGF – 19, expressed in the fetal brain. Biochim. Biophys. Acta 1444, 148 – 151 (1999). E. Canalis, M. Centrella, and T. McCarthy, Effects of basic fibroblast growth factor on bone formation in vitro. J. Clin. Invest. 81, 1572 – 1577 (1988). R. K. Globus, P. Patterson-Buckendahl, and D. Gospodarowicz, Regulation of bovine bone cell proliferation by fibroblast growth factor and transforming growth factor beta. Endocrinology 123, 98 – 105 (1988). C. H. Kasperk, J. E. Wergedal, S. Mohan, D. L. Long, K. H. Lau, and D. J. Baylink, Interactions of growth factors present in bone matrix with bone cells: effects on DNA synthesis and alkaline phosphatase. Growth Factors 3, 147 – 158 (1990). D. Noff, S. Pitaru, and N. Savion, Basic fibroblast growth factor enhances the capacity of bone marrow cells to form bone-like nodules in vitro. FEBS Lett. 250, 619 – 621 (1989). S. B. Rodan, G. Wesolowski, K. Yoon, and G. A. Rodan, Opposing effects of fibroblast growth factor and pertussis toxin on alkaline phosphatase, osteopontin, osteocalcin, and type I collagen mRNA levels in ROS 17/2.8 cells. J. Biol. Chem. 264, 19934 – 19941 (1989). V. Shen, G. Kohler, J. Huang, S. S. Huang, and W. A. Peck, An acidic fibroblast growth factor stimulates DNA synthesis, inhibits collagen and alkaline phosphatase synthesis and induces resorption in bone. Bone Miner. 7, 205 – 219 (1989).

265. E. Canalis and J. B. Lian, Effects of bone associated growth factors on DNA, collagen and osteocalcin synthesis in cultured fetal rat calvariae. Bone 9, 243 – 246 (1988). 266. T. L. McCarthy, M. Centrella, and E. Canalis, Effects of fibroblast growth factors on deoxyribonucleic acid and collagen synthesis in rat parietal bone cells. Endocrinology 125, 2118 – 2126 (1989). 267. M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type I collagen gene expression in osteoblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). 268. F. Debiais, M. Hott, A. M. Graulet, and P. J. Marie, The effects of fibroblast growth factor – 2 on human neonatal calvaria osteoblastic cells are differentiation stage specific. J. Bone Miner. Res. 13, 645 – 654 (1998). 269. M. Iwamoto, A. Shimazu, K. Nakashima, F. Suzuki, and Y. Kato, Reduction of basic fibroblasts growth factor receptor is coupled with terminal differentiation of chondrocytes. J. Biol. Chem. 266, 461–467 (1991). 270. F. De Luca and J. Baron, Control of bone growth by fibroblast growth factors. Trends in Endocrinol. Metab. 10, 61 – 65 (1999). 271. E. E. Mancilla, F. De Luca, J. A. Uyeda, F. S. Czerwiec, and J. Baron, Effects of fibroblast growth factor-2 on longitudinal bone growth. Endocrinology 139, 2900 – 2904 (1998). 272. F. Coulier, P. Pontarotti, R. Roubin, H. Hartung, M. Goldfarb, and D. Birnbaum, Of worms and men: An evolutionary perspective on the fibroblast growth factor (FGF) and FGF receptor families. J. Mol. Evol. 44, 43 – 56 (1997). 273. K. Keegan, D. E. Johnson, L. T. Williams, and M. J. Hayman, Isolation of an additional member of the fibroblast growth factor receptor family, FGFR-3. Proc. Natl. Acad. Sci. USA 88, 1095 – 1099 (1991). 274. W. C. Su, M. Kitagawa, N. Xue, B. Xie, S. Garofalo, J. Cho, C. Deng, W. A. Horton, and X. Y. Fu, Activation of Stat1 by mutant fibroblast growth-factor receptor in thanatophoric dysplasia type II dwarfism. Nature 386, 288 – 292 (1997). 275. D. M. Kilkenny and D. J. Hill, Perinuclear localization of an intracellular binding protein related to the fibroblast growth factor (FGF) receptor 1 is temporally associated with the nuclear trafficking of FGF-2 in proliferating epiphyseal growth plate chondrocytes. Endocrinology 137, 5078 – 5089 (1996). 276. R. E. Friesel and T. Maciag, Molecular mechanisms of angiogenesis: fibroblast growth factor signal transduction. FASEB J. 9, 919 – 925 (1995). 277. C. Basilico and D. Moscatelli, The FGF family of growth factors and oncogenes In “Advances in Cancer Research.” pp. 115 – 165. Department of Microbiology, New York University School of Medicine, New York, (1992). 278. L. W. Burrus, M. E. Zuber, B. A. Lueddecke, and B. B. Olwin, Identification of a cysteine-rich receptor for fibroblast growth factors. Mol. Cell Biol. 12, 5600 – 5609 (1992). 279. M. Fannon and M. A. Nugent, Basic fibroblast growth factor binds its receptors, is internalized, and stimulates DNA synthesis in Balb/c3T3 cells in the absence of heparan sulfate. J. Biol. Chem. 271, 17949 – 17956 (1996). 280. M. Roghani, A. Mansukhani, P. Dell’Era, P. Bellosta, C. Basilico, D. B. Rifkin, and D. Moscatelli, Heparin increases the affinity of basic fibroblast growth factor for its receptor but is not required for binding. J. Biol. Chem. 269, 3976 – 3984 (1994). 281. Z. Zhou, M. E. Zuber, L. W. Burrus, and B. B. Olwin, Identification and characterization of a fibroblast growth factor (FGF) binding domain in the cysteine-rich FGF receptor. J. Biol. Chem. 272, 5167 – 5174 (1997). 282. X. Xu, M. Weinstein, C. Li, and C. Deng, Fibroblast growth factor receptors (FGFRs) and their roles in limb development. Cell Tissue Res. 296, 33 – 43 (1999). 283. J. D. Coffin, R. Z. Florkiewicz, J. Neumann, T. Mort-Hopkins, G. W. Dorn, P. Lightfoot, R. German, P. N. Howles, A. Kier, and B. A.

CHAPTER 14 Bone Growth Factors

284.

285.

286.

287.

288.

289.

290.

291.

292.

293.

O’-Toole, Abnormal bone growth and selective translational regulation in basic fibroblast growth factor (FGF-2) transgenic mice. Mol. Biol. Cell 6, 1861 – 1873 (1995). S. Ortega, M. Ittmann, S. H. Tsang, M. Ehrlich, and C. Basilico, Neuronal defects and delayed wound healing in mice lacking fibroblast growth factor 2. Proc. Natl. Acad. Sci. USA 95, 5672 – 5677 (1998). B. Feldman, W. Poueymirou, V. E. Papaioannou, T. M. DeChiara, and M. Goldfarb, Requirement of FGF-4 for postimplantation mouse development. Science 267, 246 – 249 (1995). E. N. Meyers, M. Lewandoski, and G. R. Martin, An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nature. Genet. 18, 136 – 141 (1998). H. Min, D. M. Danilenko, S. A. Scully, B. Bolon, B. D. Ring, J. E. Tarpley, M. DeRose, and W. S. Simonet, Fgf-10 is required for both limb and lung development and exhibits striking functional similarity to Drosophila branchless. Genes Dev. 12, 3156 – 3161 (1998). G. Celli, W. J. LaRochelle, S. Mackem, R. Sharp, and G. Merlino, Soluble dominant-negative receptor uncovers essential roles for fibroblast growth factors in multi-organ induction and patterning. EMBO J. 17, 1642 – 1655 (1998). E. Arman, R. Haffner-Krausz, Y. Chen, J. K. Heath, and P. Lonai, Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development. Proc. Natl. Acad. Sci. USA 95, 5082 – 5087 (1998). J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). H. Kawaguchi, T. Kurokawa, K. Hanada, Y. Hiyama, M. Tamura, E. Ogata, T. Matsumoto, Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 135, 774 – 781 (1994). Y. S. Bland, M. A. Critchlow, and D. E. Ashhurst, D. E. Exogenous fibroblast growth factors-1 and -2 do not accelerate fracture healing in the rabbit. Acta Orthop. Scand. 66, 543 – 548 (1995).

431 294. T. Kato, H. Kawaguchi, K. Hanada, I. Aoyama, Y. Hiyama, T. Nakamura, K. Kuzutani, M. Tamura, T. Kurokawa, and K. Nakamura, Single local injection of recombinant fibroblast growth factor-2 stimulates healing of segmental bone defects in rabbits. J. Orthop. Res. 16, 654 – 659 (1998). 295. T. Nakamura, Y. Hara, M. Tagawa, M. Tamura, T. Yuge, H. Fukuda, and H. Nigi, Recombinant human basic fibroblast growth factor accelerates fracture healing by enhancing callus remodeling in experimental dog tibial fracture. J. Bone Miner. Res. 13, 942 – 949 (1998). 296. M. L. Radomsky, T. B. Aufdemorte, L. D. Swain, W. C. Fox, R. C. Spiro, and J. W. Poser, Novel formulation of fibroblast growth factor – 2 in a hyaluronan gel accelerates fracture healing in nonhuman primates. J. Orthop. Res. 17, 607 – 614 (1999). 297. K. Nakamura, H. Kawaguchi, I. Aoyama, K. Hanada, Y. Hiyama, T. Awa, M. Tamura, and T. Kurokawa, Stimulation of bone formation by intraosseous application of recombinant basic fibroblast growth factor in normal and ovariectomized rabbits. J. Orthop. Res. 15, 307 – 313 (1997). 298. Y. Tabata, K. Yamada, L. Hong, S. Miyamoto, N. Hashimoto, and Y. Ikada, Skull bone regeneration in primates in response to basic fibroblast growth factor. J. Neurosurg. 91, 851 – 856 (1999). 299. C. R. Dunstan, R. Boyce, B. F. Boyce, I. R. Garrett, E. Izbicka, W. H. Burgess, and G. R. Mundy, Systemic administration of acidic fibroblast growth factor (FGF-1) prevents bone loss and increases new bone formation in ovariectomized rats. J. Bone Miner. Res. 14, 953 – 959 (1999). 300. H. Nagai, R. Tsukuda, and H. Mayahara, Effects of basic fibroblast growth factor (bFGF) on bone formation in growing rats. Bone 16, 367 – 373 (1995). 301. P. Cuevas, F. Carceller, S. Ortega, M. Zazo, I. Nieto, G. GimenezGallego, Hypotensive activity of fibroblast growth factor. Science 254, 1208 – 1210 (1991). 302. T. K. Rosengart, S. R. Patel, and R. G. Crystal, Therapeutic angiogenesis: protein and gene therapy delivery strategies. J. Cardiovasc. Risk 6, 29 – 40 (1999). 303. J. M. Isner, The role of angiogenic cytokines in cardiovascular disease. Clin. Immunol. Immunopathol. 80, S82 – S91 (1996).

CHAPTER 15

Skeletal Heterogeneity and the Purposes of Bone Remodeling Implications for the Understanding of Osteoporosis A. M. PARFITT

Division of Endocrinology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205

I. Introduction II. Skeletal Heterogeneity III. Purposes of Bone Remodeling

IV. Implications for Understanding Osteoporosis References

I. INTRODUCTION

sense, implies a target. The target value of any regulatory process in biology has been optimized by natural selection. Mechanisms have evolved that ensure that deviations from the target are detected and that corrective measures to restore the target value are carried out. In this sense, body temperature, extracellular fluid osmolality, tissue oxygen tension, and countless other physiologic quantities are regulated, but the mechanisms of regulation could not be determined until the existence of the target had been recognized and its precise nature defined. Is there a target for bone remodeling or for some characteristic of bone that is influenced by remodeling? The piecemeal, quantal nature of bone remodeling is well known. The process is carried out by temporary anatomic structures known as basic multicellular units (BMUs) [2 – 4],

The cells of bone influence its structure by means of four processes: growth, repair, modeling, and remodeling, the last being the basis of bone tissue turnover in the adult skeleton. The purposes of growth and repair are obvious. Modeling serves to adapt bones to changes in mechanical loading, a process that is most effective during growth [1]. But why does a tissue that can survive for thousands of years after death need to be maintained by periodic replacement during life? Most of those interested in bone, whether as physicians, as clinical investigators, or as basic scientists, show remarkably little interest in this fundamental question. Many articles and book chapters discuss the regulation of bone remodeling, but regulation, at least in the physiologic

OSTEOPOROSIS, SECOND EDITION VOLUME 1

433

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

434

A. M. PARFITT

which excavate and replace tunnels through cortical bone (osteonal remodeling) or trenches across the surface of cancellous bone (hemiosteonal remodeling). Each BMU includes two teams of executive cells (osteoclasts and osteoblasts), supported by blood vessels, nerves, and loose connective tissue. The life span of the BMU, which comprises separate stages of origination, progression, and termination, is measured in months, but the life span of osteoblasts while they are making bone is measured in weeks, and the life span of osteoclast nuclei is measured in days. During progression of the BMU through or across the surface of bone the spatial and temporal relationships between its components are maintained by the continued growth of the central capillary [3] and recruitment of new cells [3 – 5]. These cells, like the formed elements of the blood, originate from stem cells in the bone marrow [6], except that in the peripheral skeleton, osteoblasts are derived from local precursors [5]. For blood cells, as for other short-lived cells, control of cell production and survival is more important than control of differentiated cell function; although the details are less clear, the same also applies to bone cells [6]. Each type of blood cell is normally produced at a basal rate that is sufficient for ordinary purposes but which can be increased when needed [7]. For each cell type, the circumstances under which demand is increased are well known and are related to the function of the particular cell, although the cell types differ with respect to the time scale of this response, its specificity, the relative importance of reactive and anticipatory homeostasis [8], and the extent to which the control mechanisms have been elucidated. The importance of these relationships between supply and demand, and between demand and function, also applies to bone cells. For osteoblasts in the adult nongrowing skeleton, the demand is created by bone resorption, as the function of osteoblasts is to replace the bone removed by osteoclasts. However, the circumstances that create a demand for osteoclasts are much less well defined, as these circumstances are dictated by the purposes of bone remodeling. Indeed, the questions “What are the purposes of bone remodeling and how are they achieved?” are essentially equivalent to the questions “Where and when are osteoclasts needed, and how is this need recognized and satisfied?” The answers to these questions are different in different types of bone and in different regions of the skeleton.

II. SKELETAL HETEROGENEITY A. Structure and Function The structural differences between cortical bone, in which porosity and surface-to-volume ratio are low, and cancellous bone, in which these geometric quantities are high [9], are now widely recognized. All intermediate values for these quantities can occur, but they are infrequent,

TABLE 1

Subdivisions of the Skeleton

Feature

Central

Peripheral

Main bone tissue

Cancellous

Cortical

Main soft tissue

Viscera

Muscle

Main joint type

Various

Synovial

Cortices

Thin

Thick

Marrow

Hematopoietic

Fatty

Turnover

High

Low

implying that transitional structures tend to be temporary and short-lived [10]. Less often noted are the differences between axial and appendicular subdivisions of the skeleton (Table 1); the pelvis, defined anatomically as appendicular, behaves functionally as part of the axial skeleton, so that it is more accurate to contrast central with peripheral regions. This distinction is important because the different functions of the skeleton are divided differently between central and peripheral components. The primary function is load bearing: to support posture, permit movement (including locomotion), and provide protection for the soft tissues. Subsidiary functions participate in mineral homeostasis and provide a favorable microenvironment for hematopoiesis. For convenience the former functions will be referred to as “mechanical” and the latter as “metabolic” [7]. It is commonly believed that the mechanical functions are carried out mainly by cortical bone and the metabolic functions mainly by cancellous bone, regardless of their central or peripheral locations. In fact, the functions of the peripheral skeleton, cancellous as well as cortical, are mainly mechanical, whereas the central skeleton, cortical as well as cancellous, in addition to its mechanical function, participates to a much greater extent in the metabolic functions of bone. This revision in functional attribution is most striking for peripheral cancellous bone, such as in the metaphyses of the long bones [11]. As is evident from the orientation of the trabeculae (Fig. 1a), metaphyseal cancellous bone transmits loads from the joint surfaces to diaphyseal cortical bone. Indeed, the metaphyses are flared in shape precisely to make such load transmission possible. Similar functional and architectural considerations apply to the cancellous bone in the small bones of the hands and feet (Fig. 1b). As will subsequently be discussed in detail, there is no evidence that such peripheral cancellous bone participates to a significant extent in the metabolic functions of the skeleton, whether related to mineral homeostasis or to hematopoiesis.

B. Remodeling and Turnover The frequent assertion that cancellous bone has higher turnover than cortical bone is usually supported by comparing

435

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling

the vertebral bodies in particular [15], is quite unrepresentative of the peripheral skeleton [16,17]. Based on a variety of indirect methods, turnover in peripheral cortical bone is lower by about half than in the ribs (around 2%/year vs 4%/year) [18]. For peripheral cancellous bone, estimates of turnover are based on fewer data, but they suggest that the central – peripheral difference is greater than for cortical bone. During the treatment of osteomalacia, the increase in cancellous bone mineral was about 35% in the ilium, measured histologically, but only 1 – 2% in the distal radius, measured by single photon absorptiometry [19]. On the reasonable assumption that unmineralized osteoid tissue accumulates during the evolution of osteomalacia in proportion to the initial rate of turnover, this rate in the cancellous bone of the distal radius is normally only about 2%/year. This is similar to the estimate for peripheral cortical bone; because the surfaceto-volume ratio would be higher in cancellous bone, the activation frequency would be even lower than on the intracortical surfaces. Direct measurements of turnover in peripheral and central cancellous bone in the beagle confirm a much lower value for the former, even though the absolute values for both were higher than in human subjects [20]. FIGURE 1

(a) Examples of trabecular orientation in metaphyseal cancellous bone in the appendicular skeleton. Alignment with stress trajectories facilitates the transmission of loads from the joints to diaphyseal cortical bone. (b) Examples of trabecular orientation in the small bones of the feet. Alignment with stress trajectories facilitates the transmission of loads during locomotion to the ankle joint and hence to diaphyseal cortical bone in the tibia. Modified from Meyer [11].

central cancellous with peripheral cortical bone, but this is to confuse the geometrical and biological factors that influence turnover. The remodeling process occurs only on bone surfaces, and the intensity of remodeling is expressed by the activation frequency, which is the reciprocal of the average time interval between the initiation of consecutive cycles of remodeling at the same surface location, referred to as the regeneration period [12]. Turnover refers to volume replacement, which depends not only on the surface-defined activation frequency but on the surface-to-volume ratio. This geometrical property is about four to five times higher in typical cancellous bone than in typical cortical bone [9,13]. Consequently, the former could have a higher turnover despite a lower intensity of remodeling. Systematic site-specific measurements of turnover in the human skeleton are available only for the rib [2] and for the ilium. In the latter, because activation frequency is similar on cancellous, endocortical, and intracortical subdivisions of the endosteal envelope [14,15], the difference in turnover between cortical and cancellous bone at this site depends entirely on the difference in the surface-to-volume ratio. Unfortunately, the ilium, although probably representative of the central skeleton in general and of

C. Relationship to Marrow Composition In the embryo, hematopoietic marrow appears first in the yolk sac and subsequently migrates to the liver and spleen and then to the marrow cavities. At birth, hematopoiesis is active in cancellous bone throughout the skeleton but has virtually ceased at extramedullary sites [21,22]. During growth, there is gradual conversion of red to yellow marrow, a process that begins in the distal extremities and proceeds centripetally. By age 25, hematopoiesis has disappeared from the peripheral skeleton, except to a limited extent in the upper femora [22]. Macroscopically visible hematopoiesis continues in the central skeleton throughout life, although there is a gradual increase in the number of fat cells at the expense of hematopoietic cells. At any age, a sustained increase in demand can lead to the reappearance of hematopoiesis in the extremities [22,23]. Whether this results from reactivation of dormant local stem cells or from recolonization of fatty marrow by circulating stem cells is not known, although there is strong evidence that hematopoietic stem cells do circulate [24]. Data presented, although incomplete, indicate that in the adult human skeleton central cancellous bone has persistent hematopoiesis and high bone turnover, whereas peripheral cancellous bone has absent hematopoiesis and low bone turnover (Table 2). Furthermore, based on external radionuclide counting, there is a close correlation between the extent of hematopoiesis and bone blood flow [25]. When different bones sampled at autopsy were compared, there

436

A. M. PARFITT

TABLE 2

Cancellous Bone and Its Marrow

Feature

Red marrow

Yellow marrow

Bone type

Metabolic

Mechanical

Location

Central

Peripheral

Main function

Calcium homeostasis

Transmit loads

Support hematopoiesis

Absorb energy

Cellularity

High

Low

Blood flow

High

Low

Turnover

High

Low

III. PURPOSES OF BONE REMODELING

was a good relationship between the proportion of cancellous bone surface in contact with hematopoietic cells and the proportion engaged in bone remodeling [26]. In adult beagles, there is an even more striking correspondence between marrow composition and bone remodeling (Table 3). Adjacent to red marrow there is a 15% higher mineral apposition rate and an almost 10-fold higher bone formation rate than adjacent to yellow marrow, with corresponding differences in the uptake of plutonium [27,28]. If, as seems likely, there are no hematopoietic stem cells in yellow marrow, all osteoclasts in the peripheral skeleton, cancellous as well as cortical, must be derived from circulating mononuclear precursor cells [5,7]. In the axial skeleton, osteoclast precursors might be able to migrate directly to the bone surface, but participation of the local microcirculation has now been established [5,29]. The relationship between marrow composition and remodeling can be disturbed in pathologic conditions. For example, after ovariectomy, both bone turnover and amount of fat in the marrow increase [30]. No relationship between marrow composition and bone remodeling was found in a single patient with osteoporosis who died from an unrelated cause after the administration of tetracycline labels in preparation for bone biopsy [31]. The relationship is also disturbed by proximity to synovial joints; turnover is higher within 1 mm of the articular surface than at more distant locations [32]. Nevertheless, the spatial association between TABLE 3

Cancellous Bone Turnover in Normal Beaglesa

Site

Marrow

MARb(m/day)

BFRc(%/year)

Lumbar vertebra

Red

1.29  0.10

Proximal humerus

Red

1.23  0.10

89  18

Pelvis

Red

1.26  0.10

83  25

Proximal ulna

Yellow

0.90  0.06

13  6

Distal ulna

Yellow

0.97  0.07

73

106  9

Data expressed as mean  SE. From Wronski et al. [27,28]. Mineral apposition rate; n  8. c Bone formation rate; n  4. a b

hematopoiesis and active remodeling appears to be characteristic of the healthy skeleton. To most observers, this is simply the expected consequence of the presence or absence of precursor cells in close proximity to the bone surface, but this is a superficial view. Why does cancellous bone need to turn over so much faster in some locations than in others?

There is probably no physiologic function other than bone remodeling that has attracted so much study in the face of so much uncertainty about why it occurs. Many in the field act as if they believed that the only purpose of remodeling was to cause osteoporosis and thus provide employment for scientists and business opportunities for the pharmaceutical industry! In the analysis of this problem, it seems reasonable to make two assumptions. First, periodic replacement of bone serves to maintain its ability to carry out its functions, as summarized previously. Second, because the most obvious difference between the old bone removed and the new bone put in its place is in their ages, excessive age of bone in some way compromises its functional capacity; in general, there is a reciprocal relationship between the rate of turnover and the mean age of the bone. Because the primary function of bone is mechanical, the primary purpose of remodeling of bone is to maintain its load-bearing capacity. This is accomplished partly by preventing structural damage at the microscopic and submicroscopic levels and partly by repairing such damage after it occurs.

A. Fatigue Damage and Mechanical Competence All structural materials that undergo repetitive cyclical loading are subject to fatigue, a phenomenon that has been studied most extensively in fabricated materials such as steel [33]. After a certain number of load cycles, tiny cracks appear, which are detectable at first at the ultramicroscopic level, but were probably preceded by damage at submicroscopic and molecular levels. If cyclical loading continues, the cracks extend and accumulate into macroscopic damage and eventually into overt structural failure. In bone, subtle molecular changes in matrix constituents may also appear with increasing material age [34]. The essence of fatigue is that in each cycle, the load is far below the instantaneous breaking strength of the intact material. Biological materials such as bone also undergo fatigue damage, but differ from man-made materials in their capacity for self-repair. The occurrence of fatigue damage has been demonstrated unequivocally in cortical bone [35,36], and there is compelling evidence that experimentally induced fatigue damage in

437

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling

cortical bone induces repair by remodeling so that the damaged bone is removed and replaced by new undamaged bone [36,37]. It is reasonable to assume that the same applies to load-bearing cancellous bone, although this has not yet been demonstrated. Microcracks occur in human cancellous bone [38,39], and various degrees of microdamage can be identified [40]; because they do not accumulate with age, unlike those in femoral cortical bone [41], they must be repaired by remodeling. However, there is no evidence that these lesions are due to fatigue; indeed, identical lesions can be produced experimentally by compression [42]. Microfractures in cancellous bone heal by callus formation rather than by remodeling [43], and most of them (at least in the vertebrae) can also be explained, not by fatigue, but by instantaneous overload, leading to failure by buckling [44]. Evidently, one function of remodeling in load-bearing bone is to provide a means for replacing bone that has undergone fatigue microdamage. However, this cannot be the only way in which remodeling maintains the mechanical competence of bone. The similarity between different members of the same species in the spatial distribution of remodeling activity at different skeletal sites [45] cannot be explained by a mechanism that is purely reparative. One of the most striking aspects of such remodeling maps is their bilateral symmetry, such that cross sections at the same level of bones on opposite sides of the body are virtually mirror images of one another [33,46]. It is exceedingly unlikely that such consistent symmetry could be the expression of fatigue damage repair, but it could well be an expression of fatigue damage prevention. Increased mechanical usage might stimulate remodeling before the occurrence of damage [47], but there are several difficulties with this proposal [2,33,46,48]. For material of the same mechanical properties, the major determinant of fatigue damage is the number of load cycles, and for the same level of physical activity, the major determinant of the number of load cycles is the age of the structure. The customary pattern and intensity of physical activity are species specific and so are genetically programmed [1]. Consequently, it seems possible that the remodeling map is the expression of a genetic program to prevent bone age from exceeding some critical level, a level that is different in different regions of the skeleton [45]. The contrast between the prevention of fatigue and other forms of damage by keeping bone age below some critical value and the repair of such damage by removal of the bone involved is analogous to the contrast between anticipatory and reactive homeostasis [8], except that the basis of the anticipation is genetic rather than physiologic. More specifically, it exemplifies the distinction between stochastic1 remodeling and targeted remodeling, a distinction that establishes an order of priority for different remodeling projects. There is a wide range of turnover rates consistent with skeletal health [2,18], and the low rates that occur in hypothyroidism [50] and hypoparathyroidism [51] do not

appear to increase fracture risk. Presumably this is because stochastic remodeling to prevent excessive bone age provides a substantial margin of safety. Consequently, curtailing the progression of a particular BMU, which is a likely basis of stochastic remodeling [52], is unlikely to have any harmful effects. However, targeted remodeling, to remove fatigued bone before the damage escalates from microscopic to macroscopic, must be carried out promptly or else it will fail in its purpose. The existence of such a temporal hierarchy has an important impact on the therapeutic reduction of bone turnover, a point that will subsequently be discussed in more detail. The mechanism of targeted remodeling is becoming a little clearer. The only cell that is in the right location to detect microscopic damage is the osteocyte. This cell can be activated by mechanically induced strain to increase protein synthesis [53], but the relationship of this phenomenon to damage detection is unknown. The osteocytes must transmit a signal to the cells lining the nearest free bone surface, instructing them to originate a new BMU, but whether the signal is biochemical, electrical, hydraulic, or neural is unknown [7]. In low turnover load-bearing bone, whether cortical or cancellous, there will be no osteoclast precursor cells in the vicinity of the surface so that the lining cells must transmit a second signal to the nearest capillary, which induces circulating mononuclear osteoclast precursors to leave the circulation by an area code mechanism, analogous to that which attracts circulating neutrophils to sites of inflammation [4,5,7]. Once the new BMU is in place, it must find its way to the site of damage. Matrix microdamage induces local osteocyte apoptosis in the bone that will be removed, and molecules released by dying cells could serve as a homing signal for the approaching BMU [54]. Many other factors can influence one or more steps in this complex process, but their role is more likely to be permissive than regulatory [3].

B. Metabolic Functions of Remodeling The foregoing argument has established three interconnected facts. First, the primary function of metaphyseal cancellous bone in the extremities is mechanical load bering. Second, the reason why load-bearing bone must be remodeled is to maintain its mechanical competence. Third, the rate of turnover of load-bearing bone adjacent to fatty marrow, whether cortical or cancellous, is low. Clearly, a low rate of turnover, of the order of 2 – 5%/year, is sufficient to maintain the mechanical competence of bone, regardless of its location in the skeleton or its geometric features. Consequently,

1 By stochastic is meant a succession of events that are individually random but collectively constitute a process that is amenable to study [49].

438 the rate of turnover of axial cancellous bone adjacent to hematopoietic marrow (15 – 35%/year) is much higher (by a factor of at least five) than is necessary to maintain mechanical competence. Unless this mechanically surplus or spare remodeling is simply a form of occupational therapy for cells with nothing better to do, it must subserve an entirely different purpose. This conclusion will not surprise the many endocrinologists who have always believed that the main purpose of bone remodeling was to support calcium homeostasis, but the restriction of this function mainly to cancellous bone adjacent to red marrow has not previously been emphasized. The relative importance of the mechanical and metabolic aspects of remodeling, debated inconclusively for many years [46], is evidently different in different regions of the skeleton, although both are essential to the organism as a whole. The most important nonmechanical function of bone remodeling concerns the regulation of calcium homeostasis. Bone is involved in both determining the steady-state target value for plasma-free calcium and correcting deviations from the target value [55]. Both of these processes depend on a relatively high rate of bone remodeling, but in quite different ways. Bone mineral also functions as a reservoir for sodium and as a buffer for hydrogen ion regulation. Bone remodeling may also provide biochemical support for hematopoiesis as well as the mechanical support provided by the bone itself. Both stem cell number and proliferative activity are greatest adjacent to the endosteal surface [56], and for this reason, bone-lining cells may need timely replacement. Bone matrix contains growth factors and other regulatory molecules, some of which may act on blood-forming cells rather than on bone cells. For several reasons, it could be advantageous for such molecules to be released into the bone marrow during bone resorption rather than directly from the cells involved in their biosynthesis. Possible reasons include cell polarization, with osteoblasts transporting substances away from, and osteoclasts toward, the marrow, the high proton concentration within the ruffled border of osteoclasts, and a need for intermittent rapid release rather than more continuous slow release. However, this is speculative, and the remainder of the discussion focuses on the relationship between bone remodeling and calcium homeostasis. Except under conditions of extreme calcium deprivation, the calcium homeostatic function of remodeling is not antagonistic to the mechanical function, as normally calcium homeostasis does not depend on continued net loss of calcium from bone. Steady-state levels of plasma-free calcium can be high, normal, or low, regardless of the directional changes in osteoclastic bone resorption or in calcium balance [57]. Plasma-free calcium is regulated by the joint effects of parathyroid hormone (PTH) on the renal tubular reabsorption of calcium and on the blood – bone equilibrium. This equilibrium is achieved when the inward and outward fluxes of calcium at quiescent bone surfaces are equal, and the calcium level at which this occurs is determined by

A. M. PARFITT

some effect of PTH on bone-lining cells [55]. For this mechanism to be effective, several conditions must be met. First, there must be a high blood flow, which is ensured by the proximity of hematopoietic marrow. Second, the bone at the surface must retain enough water to permit rapid diffusion of minerals, which is ensured by a high rate of remodeling. As bone ages, secondary mineralization proceeds slowly to completion by crystal enlargement and displacement of water, with a progressive decline in its ability to support the rapid mineral exchanges on which plasmacalcium homeostasis depends [55]. Stochastic remodeling could prevent excessive aging of surface bone, but as for fatigue damage, from time to time targeted remodeling will be needed to remove bone that has become hypermineralized. The mechanism of targeting is even less understood than for fatigue damage but should be simpler, as the bone to be removed is on rather than beneath the surface. As well as determining the steady-state target level of plasma-free calcium, the bone also participates in the correction of deviations from the target value. A fall in plasmafree calcium stimulates PTH secretion, which increases the outflow of calcium from bone, not only by shifting the balance of exchange at quiescent bone surfaces but also by increasing the resorptive activity of existing osteoclasts. This acute effect is quite separate from the long-term effect of PTH to increase activation frequency, osteoclast recruitment, and bone turnover in primary and secondary hyperparathyroidism. Obviously, the rapidity of the correction depends on the number of osteoclasts available, which is determined by the number of BMUs present and by the efficiency of the local circulation. The most important use for this mechanism is to accommodate the circadian changes in the supply of calcium from intestinal absorption, with an approximately 12- to 16-h period of eating, followed by an 8- to 12-h period of fasting, during which both PTH secretion and bone resorption increase [58]. In each BMU, the cutting cone (in osteonal remodeling) or hemicone (in hemiosteonal remodeling) advances more rapidly at night and slows down to allow the closing cone (or hemicone) to catch up during the day. This concertina-like action (Fig. 2) allows the skeleton to supply calcium at night when it is needed, without affecting the terminal balance of the BMUs and so without causing an irreversible loss of bone. A final aspect of remodeling and homeostasis is that the remodeling apparatus can supply a temporary but sustained demand for calcium lasting for many months by a temporary increase in BMU origination and a corresponding increase in the remodeling-dependent reversible mineral deficit [59]. The best known example is cyclic physiologic osteoporosis in deer, in which a seasonal increase in cortical porosity is entrained to the antler growth cycle [60]. The phenomenon has been demonstrated only in ribs; whether it is confined to the central skeleton or affects the peripheral skeleton as well is not known. The

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling

439

1. MECHANISMS OF BONE LOSS

FIGURE 2

Contribution of BMU-based remodeling to short-term demands for calcium. During the night, osteoclasts of the cutting cone (Rs) advance more quickly than osteoblasts of the closing cone (F), increasing the reversal zone (Rv). During the day, the cutting cone slows down and the closing cone catches up. The same concertina-like action can occur with cancellous BMUs (hemiosteonal remodeling). Reprinted with permission from Parfitt [55].

same phenomenon can partly satisfy the increased demand for calcium that occurs during growth, pregnancy, and lactation; based on densitometric data, in these circumstances the peripheral skeleton is also involved. During the adolescent growth spurt, some of the calcium needed for endochondral ossification and subperiosteal apposition is provided by a further increase in the already high cortical porosity, which subsides after cessation of longitudinal growth [1].

IV. IMPLICATIONS FOR UNDERSTANDING OSTEOPOROSIS “Osteoporosis” is a convenient term with which to cover the health implications of two related phenomena. First, bone mass in individuals falls with age. Second, partly as a result, the incidence of fractures in the population rises with age. Regrettably, for a variety of nonmedical and nonscientific reasons, it has become fashionable to define “osteoporosis” as a disease that is either present or absent, but in this text the term is used only in the former sense.

A. Pathogenesis of Fractures The relationship of bone remodeling to bone loss and to bone fragility is considered separately, as bone loss is not the only cause of increased bone fragility.

The most remarkable feature of age-related bone loss is its universality. There are useful analogies between osteoporosis and hypertension [61,62], but there are also differences. In some communities remote from western civilization, mean blood pressure does not rise with age. However, there is no subset of the human species in which mean bone mass does not fall with age, although the rate and magnitude of loss may differ between individuals and between groups [63]. Bone loss not only affects almost all persons but almost every bone, and it is of interest to compare the observed rates of loss at different skeletal sites with those predicted from remodeling theory. There are many problems in comparing rates of loss between different sites [64], including differences in methodology, instrumentation, and units. Rates of bone loss are usually expressed as a percentage of the initial value per year. This is not the best way of comparing measurements at the same site between individuals or groups [65], but in the absence of a better mathematical model, it is the most practical way of comparing different sites. For a few years after menopause, the rate of loss is substantially faster for vertebral cancellous bone than for cancellous bone at other sites and for cortical bone [66,67], but the wider the age range over which data are collected, the more similar the rates become. For example, 25 years after menopause the average amount of bone that has been lost in healthy women is about 35% of the initial value (or about 1.4%/year) in both the vertebral bodies and the distal forearm [66]. In the ilium, loss of cancellous bone, measured histologically in autopsy specimens, is about 1%/year in women between the ages of 25 and 75 [68,69]. About the same rate of loss is found in biopsy specimens, and the proportional loss of cancellous and cortical bone is very similar [9,13]. Likewise, in healthy women studied between the ages of 55 and 75 years, the average rates of loss (%/year) were 1.0 in the distal radius, 1.2 in the calcaneum, and 1.4 in the proximal radius [70]. Thus, at both central and peripheral sites, comprising various proportions of cortical and cancellous bone, the long-term rates of bone loss measured cross-sectionally are in the range of 1 – 1.5%/year. In cross-sectional studies the subjects differ not only in age but in year of birth and so may have been subject to different environmental influences [71]. This generational or cohort effect could increase the apparent rate of loss compared to longitudinal studies, but the latter can be continued only for relatively short periods. Furthermore, because the cohort effect would apply to every site, the real differences between sites are even smaller than they appear. All bone loss occurs from one of the internal surfaces of bone, and the rate of loss from any surface location depends on the average bone deficit at the end of each cycle of remodeling and the frequency with which cycles occur on that surface [12,72]. Thus for the same focal imbalance,

440 the rate of bone loss from a surface is proportional to the rate of remodeling on that surface [15,72]. It is impossible to measure remodeling rates at individual surface locations noninvasively, but biochemical indices of bone turnover reflect the aggregate of the separate contributions of each BMU currently present in the skeleton, although each index is also influenced by several other factors [73]. In accordance with remodeling theory, differences in these indices between persons are significantly correlated with differences in the rate of bone loss [74]. However, when different sites are compared, a serious paradox emerges. Remodeling theory predicts that for the same focal imbalance, the average rate of loss will be about five times higher from cancellous bone adjacent to red marrow than from cancellous bone adjacent to yellow marrow because of their difference in turnover, but sustained differences of even half this magnitude have never been demonstrated. The inescapable conclusion is that the degree of focal remodeling imbalance in, for example, the calcaneum is much greater than in the ilium, the only site where such imbalance has so far been measured [14,15]. For the same absolute rates of bone loss from a surface, the fractional loss depends on the thickness of bone beneath the surface, and hence is proportional to the surfaceto-volume ratio [65]. Accordingly, it would be expected that for the same degree of remodeling imbalance and the same frequency of remodeling activation, the average fractional rate of bone loss would be about five times higher in cancellous than in cortical bone because of their difference in surface-to-volume ratio. However, again, sustained differences in rates of bone loss of even half this magnitude have never been demonstrated. Only in the ilium have rates of both remodeling and bone loss been measured at both cortical and cancellous sites in the same bone. As mentioned previously, the results indicated similar rates of surface remodeling, similar fractional rates of bone loss, much larger absolute rates of loss from the endocortical surface, and, by inference, much greater remodeling imbalance on this surface [9,14,15]. In primary hyperparathyroidism, in normal age and menopause-related bone loss, and in patients with vertebral fracture, cortical thinning is mainly the result of increased resorption depth [14,15], which is the two-dimensional reflection of deeper penetration by endocortical BMUs. The same phenomenon has been demonstrated in the rib [75] and inferred for the metacarpal [72] and is presumably a universal feature of cortical bone loss throughout the skeleton. Furthermore, the similarity in fractional rates of bone loss indicates that the increase in resorption depth at different sites is inversely related to the customary rate of turnover and is positively related to the usual thickness of cortical bone, at each site. This is a remarkable and unexpected conclusion. When bone loss is both generalized and sustained, as in normal

A. M. PARFITT

aging, it appears that resorption depth at different sites increases to the extent necessary to bring about roughly the same rates of fractional bone loss and, as it were, “compensates” for differences in bone turnover contingent on differences in marrow composition and for differences in local bone structure and geometry. The only conceivable kind of explanation for such a phenomenon is biomechanical [2]. All mechanical influences on bone remodeling are mediated by strain, the technical term for relative deformation of a structural material as the result of load bearing. Similar fractional rates of bone loss throughout the skeleton will produce similar proportional changes in the strains that occur in different bones as a result of the same pattern and intensity of physical activity. Frost [76], building on earlier work by others [33,77], has proposed the existence of the “mechanostat,” which orchestrates the recruitment and activity of osteoclasts and osteoblasts in such a way that strain is maintained within an acceptable range [78]. The primary function of the mechanostat is to ensure that each bone during growth acquires the strength it needs to support the species-specific pattern and intensity of physical activity customary during adult life [1]. After growth has ceased, the mechanostat is much less effective in adapting the bones to an increase in mechanical demand, but is highly effective in adapting them to a decrease, accounting for the rapidity, severity, and usual irreversibility of bone loss consequent on disuse [79]. As a result of the sedentary lifestyle made possible by economic development, aging in most persons is accompanied by a progressive reduction in physical activity and muscle strength, of earlier onset and greater severity than is biologically mandated [80]. According to biomechanical theory, this should not increase the risk of fracture, as the reduced bone mass would remain appropriate to the reduced level of activity, but this does not take into account the age-related increase in the liability to fall, to which the mechanostat is blind. Frost has postulated that as a result of estrogen deficiency, the mechanostat is reset so that the skeleton responds not so much to actual but to erroneously perceived disuse [76]. A universal resetting of the mechanostat would not account for the disproportionately rapid loss of central cancellous bone in the first few years after menopause, so it is more likely not the result of estrogen deficiency, but of the aging process. 2. MECHANISMS OF BONE FRAGILITY Bone mass is related inversely to fracture risk, both current and future, but there are also qualitative abnormalities in bone that contribute to its fragility [81]. The best known and most well established of these relates to cancellous bone architecture. When cancellous bone is lost as a result of estrogen deficiency, whole structural elements are removed, leaving those that remain more widely separated and less well connected [68]. As a result, vertebral fracture

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling

risk is increased to a greater extent than would be expected for the reduction in bone mass [82]. This is probably why the presence of at least one vertebral fracture is an independent risk factor for further fractures [83]. The structural changes are the result of perforation of trabecular plates because the cutting hemicones of individual BMUs penetrate more deeply into the bone away from the surface [12,84]. This qualitative abnormality in osteoclast function appears now to be more likely due to delayed osteoclast apoptosis [85] than to excessively rapid resorption by individual osteoclasts [72]. However, a more fundamental problem may be loss of BMU directional control [86]. These various changes were attributed by Frost to resetting of the mechanostat [76], but estrogen deficiency leads for a few years to a disproportionately rapid loss of central cancellous bone adjacent to hematopoietic marrow. Furthermore, the occurrence of severe vertebral osteopenia in elite athletes with exercise-associated amenorrhea [87] indicates that the effects of estrogen deficiency are not prevented by increased physical activity. The second, more controversial, qualitative factor in bone fragility is the accumulation of fatigue microdamage. Frost [88] has proposed that normally there is such a wide margin of safety that the adverse effect of bone loss on bone fragility is mediated, not by a reduction in instantaneous breaking strength, but by fatigue damage accumulation due to increased strain in the bone that remains. However, most investigators believe that the margin of safety is not as great as Frost has claimed [89]. Frost has further proposed that a defective damage repair mechanism could be overwhelmed by even normal damage production. Increased bone age would increase susceptibility to fatigue damage, both directly (by exceeding the fatigue life) and indirectly (Fig. 3). Osteocyte death, which can occur spontaneously when bone age exceeds about 20 years [90], would impair the detection of fatigue damage, and the consequent perilacunar hypermineralization (or micropetrosis) would make the bone more brittle and more susceptible to fatigue damage [91]. The repair of microdamage by a new BMU could be delayed by an age-related decline in any of the intervening steps outlined previously or by a loss of the directional control needed for the new BMU to find its target [86], another possible consequence of osteocyte death [54]. Excessive bone age and its adverse consequences are most likely to occur in peripheral cortical bone because of its low rate of turnover, but even in central cancellous bone, turnover may be sufficiently low in some subjects that a significant fraction of interstitial bone may be older than 20 years, and this fraction is much higher in patients with vertebral fracture [92]. On the basis of this reasoning, it has been proposed that the subnormal turnover of central cancellous bone, a common finding in patients with vertebral fracture [72], was of pathogenetic significance, via increases in bone age and

441

FIGURE 3

Mechanisms whereby increased bone age could lead to accumulation of fatigue damage. Some effects would increase fatigue damage production, and some effects would decrease its detection and repair. Reproduced with permission from Parfitt [100].

susceptibility to fatigue damage [92]. At the time the previous version of this chapter was written [93], no evidence in support of this proposal had appeared and at an international meeting on osteoporosis held in 1993, it was withdrawn [44]. However, recent findings, although by no means establishing the hypothesis, suggest that this act of apostasy was premature. Osteocyte death, which increases in prevalence with subject age in the upper femur [94], appears not to change with age in the vertebral body when autopsy specimens are studied by respiratory enzyme histochemistry [95], but does increase with age in biopsy specimens of iliac cancellous bone studied by fluorescence and confocal microscopy [96]. The microcracks that occur in peripheral cortical bone have now also been observed in central cancellous bone, as mentioned previously [38], although it is not known whether they are more common in patients with vertebral fracture. It remains true that vertebral cancellous microfractures have not been shown to result from fatigue, even though they are frequently referred to as “fatigue fractures.” In the vertebral body, the perforations and loss of structural elements mentioned previously occur preferentially in horizontal rather than in vertical trabeculae. The compressive strength of a vertical trabecula will decline in proportion to the square of the unsupported length so that a 50% reduction in the number of horizontal trabeculae will lead to a fourfold increase in the susceptibility to buckling [44]. Based on estimates of in vivo stresses during normal activity [97] and on the production of microcracks by experimental compression [42], vertebral microfractures can be explained by instantaneous overload as a result of the architectural changes mentioned previously without the need to invoke a fatigue-based mechanism [44, 97,98]. Lower than normal turnover in central cancellous bone may not have the adverse effects on bone fragility that were

442 predicted because the turnover of cancellous bone adjacent to hematopoietic marrow is much higher than is needed to maintain its mechanical competence, which had not yet been deduced when the hypothesis was first proposed. In patients with vertebral fracture, central cancellous bone turnover may be much lower than in healthy subjects, but still high enough to prevent fatigue damage accumulation. In patients with osteoporotic vertebral fracture, the mineral density of iliac bone is decreased rather than increased [99,100]. This suggests that in some patients with osteoporotic vertebral fracture there is a substantial delay in secondary mineralization, the process whereby mineral crystals enlarge at the expense of water [55], which would remove much of the need for bone remodeling to prevent hypermineralization. Hypomineralization would be expected to reduce the strength of bone as a material [33] and to be an independent risk factor for bone fragility so that it is important to discover its pathogenesis. However, even if lower turnover of central cancellous bone is a consequence of hypomineralization rather than a cause of hypermineralization, there would still be accumulation of very old bone [92], which would be expected to exaggerate the age-related increase in osteocyte death and to compromise microdamage repair. Furthermore, the ability to direct a new BMU to regions of microdamage in interstitial bone could be impaired by poorly understood physical and cellular changes. If so, then the relationship between bone fragility and bone turnover is U shaped, as high bone turnover is a risk factor, because each remodeling site constitutes both a region of focal weakness and a stress concentrator [44,101]. Hip fractures share with vertebral fractures the inverse relationship of risk to bone mass, but differ from vertebral fractures with respect to the qualitative contribution to bone fragility. Loss of cancellous bone connectivity due to estrogen deficiency is less important, whereas fatigue damage accumulation is more important. Although small islands of hematopoietic tissue can persist in the upper femur much longer than at more distal sites, particularly in the femoral head, the proportion of red marrow is much lower than in the ilium [17]. There are no tetracycline-based measurements of bone remodeling in the upper femur, but other indices of bone remodeling are lower than in the ilium or vertebral body [26], and this difference is exaggerated in patients with hip fracture [17]. The proportion of osteocytes that are viable declines progressively with increasing subject age in the femoral neck [94]. True bone mineral density increases with age in the femoral shaft cortex [102], but not in the spine. Fatigue microdamage occurs in the cortical bone of both the femoral neck [103] and the femoral shaft, and in the latter, crack density increases exponentially with age, more so in women than in men [104]. Cancellous microfractures in the femoral head increase in number with age and with reduction in mineral density [105,106] and are significantly more frequent in hip fracture patients than in

A. M. PARFITT

TABLE 4

Fracture Pathogenesis at Different Sites Vertebra

Femoral neck

Function of cancellous bone

Metabolic

Mechanical

Marrow/turnover

Red/high

Yellow/low

Osteocyte death

Yes

Yes

Increase with age

Small

Large

Fatigue damage

?a

Yesb

Hypermineralization

No

Yesb

Main qualitative factor

Architecture

Bone age

a

Microdamage, not shown to be due to fatigue. In femoral cortical bone, not necessarily at fracture site.

b

controls, despite a statement by the authors to the contrary [107]; because of the lower bone turnover and differences in architecture, there is greater reason to invoke a fatiguebased mechanism than in the spine [100,105]. All these data indicate that increased bone age and its adverse effects on bone fragility (Fig. 3) are likely to be of major importance in the pathogenesis of hip fracture [100] (Table 4).

B. Prevention of Fractures It is customary to discuss the “prevention” and “treatment” of osteoporosis separately, but this is a misleading distinction, as the only therapeutic goal is to prevent fractures; whether one’s aim is to prevent the first fracture or a subsequent fracture does not alter this principle. Of the several aspects of fracture prevention, the theme of this chapter relates most clearly to the prevention and restoration of bone loss. Agents that accomplish these aims are usually referred to respectively as “inhibitors of bone resorption” and “stimulators of bone formation,” but these vague terms betray a serious lack of comprehension of bone remodeling.They ignore the indivisible unity of the BMU as a structural and functional entity, obscure the crucial distinction between effects on cell recruitment and effects on differentiated cell function, and engender the absurd notions that all bone resorption is bad and all bone formation is good. The former error is potentially more dangerous than the latter, so this aspect of therapy will be the focus of subsequent discussion. A long-term reduction in the rate of bone loss can be accomplished by a long-term reduction in activation frequency and a consequent reduction in bone turnover. How is this possible without frustrating the purposes of bone remodeling? Activation frequency is the best histologic index of the intensity of bone remodeling on a surface and is the main determinant of the rate of bone turnover, but it is not a measure of the frequency of BMU origination because it

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling

FIGURE 4 Relationship between BMU origination and remodeling activation. Activation frequency represents the product of frequency of BMU origination and the average distance of BMU progression. In this example, activation frequency would be the same with one BMU that progresses for 9 units distance, two BMUs that each progress for 4.5 units of distance, or three BMUs that each progress for 3 units of distance. However, the biological significance would be different because each BMU represents a separate remodeling project. Copyright 1995, A. M. Parfitt, used with permission.

also depends on the mean distance of BMU progression [4,108 – 110]. The effects on all histologic, biochemical, and radiokinetic indices of bone turnover would be the same whether, for example, one BMU traveled for nine units of distance through or across the surface of bone or each of three BMUs traveled for three units of distance (Fig. 4). The biological significance would be different, however, because each new BMU represents a separate remodeling project. Approximately 90% of new mononuclear osteoclast precursor cells are used to sustain the progression of existing BMUs, and only 10% are used to originate a new BMU [4]. Consequently, substantial changes in activation frequency and bone turnover can be brought about by manipulating the distance and duration of BMU progression without changing the frequency of BMU origination. Each episode of targeted remodeling requires a new BMU, but stochastic remodeling could be accomplished if each BMU progressed for a variable distance beyond its target [4]. This arrangement makes it possible for therapeutic agents to reduce activation frequency and bone turnover by curtailing BMU progression, without inhibiting BMU origination, and so to reduce stochastic remodeling without interfering with targeted remodeling. Obviously, the ability to prioritize different remodeling tasks is a feature of the remodeling system itself, not of the individual therapeutic agents. It must be assumed that the signals for osteoclast precursors to arrive at a particular location are more compelling for BMU origination than for BMU progression, more compelling for BMU progression toward its target than beyond its target, and more compelling for

443

the peripheral than for the central skeleton because of the difference in margin of safety. These hierarchies could reflect differences in the types as well as the amounts of signal molecules. However, therapeutic agents may differ in their ability to exploit these differences in signal strength. Agents that act directly on osteoclasts to reduce their resorptive activity are more likely to act indiscriminately on all osteoclasts throughout the skeleton, and in some locations this is likely to negate their purpose; consequently, the net outcome of the intervention could be harmful rather than beneficial. However, agents that reduce the supply of osteoclast precursor cells leave the remodeling system able to deploy its more limited resources to the best advantage. Not surprisingly, hormone replacement therapy (HRT) is the most effective means of preventing the adverse effects on bone of the hormone deficiency that results from menopausal ovarian failure. Estrogen deficiency increases the availability of osteoclast precursor cells [6] and so increases the stochastic component of bone remodeling by removing a constraint on BMU progression, particularly in the central skeleton. However, the most destructive consequence of estrogen deficiency is delayed osteoclast apoptosis [4,85], leading to deeper BMU penetration (reflected in two-dimensional histologic sections as increased resorption depth), trabecular plate perforation, and loss of connectivity. Both of these effects — increased osteoclast recruitment and delayed osteoclast apoptosis — are prevented by HRT, and ideally both of them should be prevented by any agent that is used as a substitute for HRT. Until recently, the most widely used substitute was calcitonin, but it has not been shown to promote earlier osteoclast apoptosis [111], and its effects on resorption depth are uncertain. Furthermore, its long-term use can be complicated by secondary hyperparathyroidism of sufficient severity that bone turnover may be increased rather than decreased [112]. The newer bisphosphonates appear to be more complete substitutes for HRT. Although their best known effect is to acutely inhibit the function of existing osteoclasts, in the long term, they reduce osteoclast recruitment by mechanisms that remain uncertain [4,113], promote earlier osteoclast apoptosis [114], and reduce resorption depth [115]. The safety of reducing bone turnover depends on the ability to limit stochastic remodeling preferentially in the central skeleton without interfering with targeted remodeling at any skeletal site. Obviously, there is a lower limit to osteoclast precursor cell recruitment below which the purposes of remodeling will be frustrated. As would be predicted from the earlier discussion, complete suppression of remodeling in beagles leads to the occurrence of spontaneous fractures after a few months [116]; this occurred with etidronate, which causes osteomalacia, but also with clodronate, which does not. A dangerous reduction in bone turnover could never occur with physiological agents such

444 as estrogen or calcitonin, but can be produced readily by bisphosphonates if given in excessive dose. Regrettably, there is very little information on what lower limit is safe. The safe level will be different in different regions of the skeleton, which reduces the value of biochemical indices of turnover to determine safety, as these are necessarily blind to regional differences. Quite low levels of whole body bone turnover are consistent with skeletal health when they occur naturally, but might conceal regional ill health when induced by therapeutic intervention. Reducing osteoclast recruitment to a level just sufficient to allow the completion of targeted remodeling but leaving no room for stochastic remodeling would also lead eventually to spontaneous fractures, but the time required would probably be measured in years rather than in months, which limits the use of animal models to determine long-term safety. As explained earlier, when vertical trabeculae have lost their horizontal supports, even normal remodeling may constitute a mechanical threat. In this situation, reducing turnover even within the normal range (defined by biochemical indices) may be useful in the prevention of vertebral fractures [44]. In a study of transdermal estrogen therapy, reduced turnover contributed independently to reduced vertebral fracture occurrence, in addition to the effect of increased bone mass [117], but the long-term effects of reducing turnover on hip fracture risk are less easily predictable. For reasons given previously, the adverse effects of prolonged bone age on bone fragility (Fig. 3) are likely to be more serious in the upper femur than in the spine (Table 4). The large increase in the use of bisphosphonates that followed the approval of alendronate by the FDA should reduce the incidence of vertebral fractures [118] and, for a few years, the incidence of other fractures, but data available today do not exclude the possibility that 10 – 20 years later there will be an epidemic of hip fractures. By then, a large proportion of the elderly population will have levels of bisphosphonate of one kind or another in the femoral heads and necks that are possibly dangerous, and there will be nothing that can be done about it. Two years of treatment with risedronate did not increase microdamage in the canine femoral neck [119], but a higher effective dose of alendronate did increase microdamage in the canine rib [120]. It is not too late to find out what is really going on in the bones of hip fracture patients, but only if we abandon the exclusive reliance on biochemical and densitometric methods and on histologic examination at a site chosen for its convenience rather than its relevance to the problem of greatest importance. An important recent discovery from direct examination is that clustered remodeling and giant resorption cavities due to confluence of clusters are more common in hip fracture patients than in age-matched control subjects [121]. A reduction in the frequency of such clusters could account for the early reduction in fracture risk by correction of vitamin D deficiency

A. M. PARFITT

[122] and also for the beneficial effect of bisphosphonate administration. However, if the clusters are an effect of defective directional control of BMUs, then reducing their number could compromise microdamage repair. Obviously a great deal more research will be needed to resolve these uncertainties.

References 1. A. M. Parfitt, The two faces of growth-Benefits and risks to bone integrity. Osteopor. Int. 4, 382 – 398 (1994). 2. H. M. Frost, “Intermediary Organization of the Skeleton.” CRC Press, Boca Raton, FL, 1986. 3. A. M. Parfitt, Osteonal and hemi-osteonal remodeling: The spatial and temporal framework for signal traffic in adult human bone. J. Cell. Biochem. 55, 273 – 286 (1994). 4. A. M. Parfitt, G. R. Mundy, G. D. Roodman, D. E. Hughes, and B. Boyce, A new model for the regulation of bone resorption. J. Bone Miner. Res. 11, 150 – 159 (1996). 5. A. M. Parfitt. Min: Review — Osteoclast precursors as leucocytes: Importance of the area code. Bone 23, 491 – 494 (1998). 6. S. C. Manolagas, Cell number versus cell vigor: What really matters to a regenerating skeleton? Endocrinology 140, 4377 – 4381. (1999). 7. A. M. Parfitt, Problems in the application of in vitro systems to the study of human bone remodeling. Proceedings from the Workshop on Human Models of Skeletal Aging, NIH. Calcif. Tissue Int. 56 (Suppl. 1), S5 – S7 (1995). 8. M. C. Moore-Ede, Physiology of the circadian timing system: Predictive versus reactive homeostasis. Am. J. Physiol. 250, R735 – R752 (1986). 9. J. Foldes, A. M. Parfitt, M.-S. Shih, D. S. Rao, and M. Kleerekoper, Structural and geometric changes in iliac bone: Relationship to normal aging and osteoporosis. J. Bone Miner. Res. 6, 759 – 766 (1991). 10. D. P. Fyhrie, S. M. Lang, and A. M. Parfitt, “Cortical and Cancellous Bone Structure in the Ilia of Normals and Osteoporotics,” p. 443. Orthopaedic Research Society, 40th Annual Meeting, 1994. [Abstract] 11. G. H. Meyer, Die Architectur der spongiosa. Arch. Anat. Physiol. Wissensch. 34, 615 – 628 (1867). 12. A. M. Parfitt, The physiologic and pathogenetic significance of bone histomorphometric data. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), 2nd Ed. Lippincott, Williams, and Wilkins, Philadelphia, in press. 13. Z.-H. Han, S. Palnitkar, D. S. Rao, D. Nelson, and A. M. Parfitt, Effect of ethnicity and age or menopause on the structure and geometry of iliac bone. J. Bone Miner. Res. 11, 1967 – 1975 (1996). 14. A. M. Parfitt, Surface specific bone remodeling in health and disease. In “Clinical Disorders of Bone and Mineral Metabolism” (M. Kleerekoper and S. Krane, eds.), pp. 7 – 14. Mary Ann Liebert, New York, 1989. 15. Z.-H. Han, S. Palnitkar, D. Rao, Sudhaker, D. Nelson, and A. M. Parfitt, Effects of ethnicity and age or menopause on the remodeling and turnover of iliac bone: Implications for mechanisms of bone loss. J. Bone Miner. Res. 12, 498 – 508 (1997). 16. D. W. Dempster, Relationship between the iliac crest bone biopsy and other skeletal sites. In “Clinical Disorders of Bone and Mineral Metabolism” (M. N. Kleerekoper and S. M. Krane, eds.), pp. 247 – 252. Mary Ann Liebert, New York, 1989. 17. I. Eventov, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodelling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 12, 1 – 6 (1991).

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling 18. A. M. Parfitt, The physiologic and clinical significance of bone histomorphometric data. In “Bone Histomorphometry: Techniques and Interpretations” (R. Recker, ed.), pp. 142 – 223. CRC Press, Boca Raton, FL, 1983. 19. A. M. Parfitt, D. S. Rao, J. Stanciu, A. R. Villenueva, M. Kleerekoper, and B. Frame, Irreversible bone loss in osteomalacia: Comparison of radial photon absorptiometry with iliac bone histomorphometry during treatment. J. Clin. Invest. 76, 2403 – 2412 (1985). 20. D. B. Kimmel and W. S. S. Jee, A quantitative histologic study of bone turnover in young adult beagles. Anat. Rec. 203, 31 – 45 (1982). 21. A. J. Erslev, Medullary and extramedullary blood formation. Clin. Orthop. 52, 25 – 36 (1967). 22. S. N. Wickramasinghe, “Human Bone Marrow.” Blackwell Scientific, Oxford, 1975. 23. R. P. Custer, “An Atlas of the Blood and Bone Marrow.” Saunders, Philadelphia, 1949. 24. A. Kessinger, Circulating stem cells: Waxing hematopoietic. N. Engl. J. Med. 333(5), 315 – 316 (1995). 25. D. Van Dyke, Similarity in distribution of skeletal blood flow and erythropoietic marrow. Clin. Orthop. 52, 37 – 51 (1967). 26. B. Krempien, F. M. Lemminger, E. Ritz, and E. Weber, The reaction of different skeletal sites to metabolic bone disease: A micromorphometric study. Klin. Wochenschr. 56, 755 – 759 (1978). 27. T. J. Wronski, J. M. Smith, and W. S. S. Jee, The microdistribution and retention of injected 239Pu on trabecular bone surfaces of the beagle: Implications for the induction of osteosarcoma. Radiat. Res. 83, 74 – 89 (1980). 28. T. J. Wronski, J. M. Smith, and W. S. S. Jee, Variations in mineral apposition rate of trabecular bone within the beagle skeleton. Calcif. Tissue Int. 33, 583 – 586 (1981). 29. F. Melsen, L. Mosekilde, and E. F. Eriksen, Spatial distribution of sinusoids in relation to remodeling sites: Evidence for specialized sinusoidal structures associated with formative sites. J. Bone Miner. Res. 10(Suppl.), S209 (1995). 30. R. B. Martin, B. D. Chow, and P. A. Lucas, Bone marrow fat content in relation to bone remodeling and serum chemistry in intact and ovariectomized dogs. Calcif. Tissue Int. 46, 189 – 194 (1990). 31. J. Podenphant and U. Engel, Regional variations in histomorphometric bone dynamics from the skeleton of an osteoporotic woman. Calcif. Tissue Int. 40, 184 – 188 (1987). 32. H. Duncan, A. R. Villanueva, C. H. Mathews, J. C. Leisen, and A. M. Parfitt, Dynamics of bone destruction and repair in rheumatoid joints. Arthritis Rheum. 24, S97 (1981). 33. J. Currey, “The Mechanical Adaptations of Bones.” Princeton Univ. Press, Princeton, NJ, 1984. 34. J. J. Tomasek, S. W. Meyers, J. B. Basinger, D. T. Green, and R. L. Shew, Diabetic and age-related enhancement of collagen-linked fluorescence in cortical bones of rats. Life Sci. 55(11), 855 – 861 (1994). 35. D. B. Burr, M. R. Forwood, D. P. Fyhrie, R. B. Martin, M. B. Schaffler, and C. H. Turner, Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J. Bone Miner. Res. 12, 6 – 15 (1997). 36. V. Bentolila, T. M. Boyce, D. P. Fyhrie, R. Drumb, T. M. Skerry, and M. B. Schaffler, Intracortical remodeling in adult rat long bones after fatigue loading. Bone 23, 275 – 281 (1998). 37. S. Mori and D. B. Burr, Increased intracortical remodeling following fatigue damage. Bone 14, 103 – 109 (1993). 38. T. E. Wenzel, M. B. Schaffler, and D. P. Fyhrie, In vivo trabecular microcracks in human vertebral bone. Bone 19(2), 89 – 95 (1996). 39. S. Mori, R. Harruff, W. Ambrosius, and D. B. Burr, Trabecular bone volume and microdamage accumulation in the femoral heads of women with and without femoral neck fractures. Bone 2 1, 521 – 526 (1997).

445

40. N. H. Fazzalari, M. R. Forwood, B. A. Manthey, K. Smith, and P. Kolesik, Three-dimensional confocal images of microdamage in cancellous bone. Bone 23, 373 – 378 (1998). 41. M. B. Schaffler, K. Choi, and C. Milgrom, Aging and matrix microdamage accumulation in human compact bone. Bone 17, 521 – 525 (1995). 42. D. P. Fyhrie and M. B. Schaffler, Failure mechanisms in human vertebral cancellous bone. Bone 15, 105 – 109 (1994). 43. E. P. M. Urovitz, V. L. Fornasier, M. I. Risen, and I. MacNab, Etiological factors in the pathogenesis of femoral trabecular fatigue fractures. Clin. Orthop. 127, 275 – 280 (1977). 44. A. M. Parfitt, Pathophysiology of bone fragility. In “Proceedings of the Fourth International Symposium on Osteoporosis” (C. Christiansen and B. Riis, eds.), pp. 164 – 166. Hong Kong, Handelstrykkeriet Aalborg ApS, Aalborg, Denmark, 1993. 45. G. Marotti, Map of bone formation rate values recorded throughout the skeleton of the dog. In “Bone Morphometry: Proceedings of the First Workshop” (Z. F. G. Jaworski, ed.), pp. 202 – 207. Univ. of Ottawa Press, 1976. 46. D. Dempster, Bone remodeling. In “Disorders of Bone and Mineral Metabolism” (F. Coe and M. Z. Favus, eds.), pp. 355 – 380. Raven Press, New York, 1992. 47. Z. F. G. Jaworski, Haversian systems and Haversian bone. In “Bone — A Treatise: Bone Metabolism and Mineralization” (B. K. Hall, ed.), Vol. 4, pp. 21 – 45. CRC Press, Boca Raton, FL, 1992. 48. R. B. Martin and D. B. Burr, “Structure, Function, and Adaptation of Compact Bone.” Raven Press, New York, 1989. 49. C. Morris, “Academic Press Dictionary of Science and Terminology.” Academic Press, San Diego, 1992. 50. E. F. Eriksen, T. Steiniche, L. Mosekilde, and F. Melsen, Histomorphometric analysis of bone in metabolic bone disease. Endocrinol. Metabol. Clin. North Am. 18, 919 – 954 (1989). 51. A. M. Parfitt, Idiopathic, surgical and other varieties of parathyroid hormone deficient hypoparathyroidism. In “Endocrinology” (L. DeGroot, ed.), 2nd Ed., pp. 1049 – 1064. Saunders, Philadelphia, 1989. 52. A. M. Parfitt, BMU Origination and Progression: Relationship to Targeted and Nontargeted Remodeling. Abstract for VIIIth Congress of the International Society of Bone Morphometry, 1999. 53. C. H. Turner and M. R. Forwood, What role does the osteocyte network play in bone adaptation? Bone 16, 283 – 285 (1995). 54. O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Min. Res. 15, 60 – 67 (2000). 55. A. M. Parfitt, Calcium homeostasis. In “Handbook of Experimental Pharmacology: Physiology and Pharmacology of Bone” (G. R. Mundy and T. J. Martin, eds.), Vol. 107, pp. 1 – 65. Springer-Verlag, Heidelberg, 1993. 56. A. Deldar, H. Lewis, and L. Weiss, Bone lining cells and hematopoiesis: An electron microscopic study of canine bone marrow. Anat. Rec. 213, 187 – 201 (1985). 57. A. M. Parfitt, Equilibrium and disequilibrium hypercalcemia: New light on an old concept. Metab. Bone Dis. Rel. Res. 1, 279 – 293 (1979). 58. A. Blumsohn, K. Herrington, R. A. Hannon, P. Shao, D. R. Eyre, and R. Eastell, The effect of calcium supplementation on the circadian rhythm of bone resorption. J. Clin. Endocrinol. Metab. 79, 730 – 735 (1994). 59. A. M. Parfitt, Morphologic basis of bone mineral measurements: Transient and steady state effects of treatment in osteoporosis. Miner. Electrolyte Metab. 4, 273 – 287 (1980). 60. A. M. Parfitt, Integration of skeletal and mineral homeostasis. In “Osteoporosis: Recent Advances in Pathogenesis and Treatment” (H. F. DeLuca, H. Frost, W. Jee, C. Johnston, and A. M. Parfitt, eds.), pp. 115 – 126. University Park Press, Baltimore, MD, 1981.

446 61. A. M. Parfitt, Definition of osteoporosis: Age-related loss of bone and its relationship to increased fracture risk. In “Osteoporosis.” NIH Consensus Development Conference, 1984. 62. M. Kleerekoper and D. Nelson, Osteoporosis as a community health problem: Lessons learned from studying hypertension. HFH Med. J. 36, 113 – 116 (1988). 63. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone.” Thomas, Springfield, IL, 1970. 64. A. M. Parfitt, Interpretation of bone densitometry measurements: The disadvantages of a percentage scale and a discussion of available alternatives. J. Bone Miner. Res. 5, 537 – 540 (1990). 65. A. M. Parfitt, Is the rate of bone loss influenced by the initial value? Biological and statistical issues. Abstract, Perth International Bone Meeting, Fremantle, Australia. Osteopor. Int. 5, 309 – 310 (1995). 66. B. E. C. Nordin, A. G. Need, R. L. Prince, M. Horowitz, D. H. Gutteridge, and S. E. Papapoulos, Osteoporosis. In “Metabolic Bone and Stone Disease” (B. E. C. Nordin, A. G. Need, and H. A. Morris, eds.), Churchill Livingstone, London, 1993. 67. C. W. Slemenda, Adult bone loss. In “Osteoporosis” (R. Marcus, ed.), Blackwell Scientific, London, 1994. 68. A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationship between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis: Implications for the microanatomic and cellular mechanism of bone loss. J. Clin. Invest. 72, 1396 – 1409 (1983). 69. P. J. Meunier, Assessment of bone turnover by histomorphometry in osteoporosis. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), Raven Press, New York, 1988. 70. J. W. Davis, J. S. Grove, P. D. Ross, J. M. Vogel, and R. D. Wasnich, Relationship between bone mass and rates of bone change at appendicular measurement sites. J. Bone Miner. Res. 7, 719 – 725 (1992). 71. K. M. Davies, R. R. Recker, M. R. Stegman, and R. P. Heaney, Tallness versus shrinkage: Do women shrink with age or grow taller with recent birth date? J. Bone Miner. Res. 6, 1115, 1120 (1991). 72. A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), Raven Press, London, 1988. 73. B. J. Pedersen, A. Schlemmer, C. Hassager, and C. Christiansen, Changes in the carboxyl-terminal propeptide of type I procollagen and other markers of bone formation upon five days of bed rest. Bone 17, 91 – 95 (1995). 74. C. Christiansen, Postmenopausal bone loss and the risk of osteoporosis. Osteopor. Int.(Suppl. 1), S47 – S51 (1994). 75. K. Wu, S. Jett, and H. M. Frost, Bone resorption rates in rib in physiological, senile, and postmenopausal osteoporoses. J. Lab. Clin. Med. 69, 810 – 818 (1967). 76. H. M. Frost, The mechanostat: A proposed pathogenetic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone Miner. 2, 73 – 85 (1987). 77. L. E. Lanyon, A. E. Goodship, C. J. Pye, and J. H. MacFie, Mechanically adaptive bone remodeling. J. Biomech. 15, 141, 154 (1982). 78. H. M. Frost, Structural adaptations to mechanical usage (SATMU). Anat. Rec. 226, 414 – 422 (1990). 79. Z. F. G. Jaworski and H. K. Uhthoff, Disuse osteoporosis: Current status and problems. In “Current Concepts of Bone Fragility” (H. K. Uhthoff, ed.), p. 181. Springer-Verlag, Berlin, 1986. 80. H. M. Frost, On our age-related bone loss: Insights from a new paradigm. J. Bone Miner. Res. 12, 1539 – 1546 (1997). 81. R. P. Heaney, Is there a role for bone quality in fragility fractures? Calcif. Tissue Int. 53 (Suppl. 1), S3 – S6 (1993). 82. M. Kleerekoper, A. R. Villanueva, J. Stanciu, D. S. Rao, and A. M. Parfitt, The role of three dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif. Tissue Int. 37, 594 – 597 (1985).

A. M. PARFITT 83. P. D. Ross, J. W. Davis, R. S. Epstein, and R. D. Wasnich, Preexisting fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 84. EF Eriksen, B. Langdahl, A. Vesterby, J. Rungby, and M. Kassem, Hormone replacement therapy prevents osteoclastic hyperactivity: A histomorphometric study in early postmenopausal women. J. Bone Miner. Res. 14, 1217 – 1221 (1999). 85. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G.R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF. Nature Med. 2, 1132 – 1135 (1996). 86. A. M. Parfitt, Abnormal structure and fragility of bone as expressions of disordered remodeling. Abstract XXIV, European Symposium on Calcified Tissues, Aarhus, Calcif. Tissue Int. 56, 423 (1995). 87. B. L. Drinkwater, K. Nilson, C. H. Chesnut, W. J. Bremner, S. Shainholtz, and M. B. Southworth, Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311, 277 – 281 (1984). 88. H. M. Frost, The pathomechanics of osteoporosis. Clin. Orthop. 200, 198 – 225 (1985). 89. R. McN. Alexander, Optimum strengths for bones liable to fatigue and accidental fracture. J. Theor. Biol. 109, 621 – 636 (1984). 90. H. M. Frost, In vivo osteocyte death. J. Bone Jt. Surg. 42A, 138 – 143 (1960). 91. H. M. Frost, Micropetrosis. J. Bone Jt. Surg. 42A, 144 – 150 (1960). 92. A. M. Parfitt, M. Kleerekoper, and A. R. Villanueva, Increased bone age: Mechanisms and consequences. In “Osteoporosis” (C. Christiansen, C. Johansen, and B. J. Riis, eds.), pp. 301 – 308. Osteopress ApS, Copenhagen, 1987. 93. A. M. Parfitt, Skeletal heterogeneity and the purposes of bone remodeling: Implications for the understanding of osteoporosis. In “Osteoporosis” (R. Marcus, D. Feldman and J. Kelsey, eds.), pp. 315 – 329. Academic Press, San Diego, 1996. 94. S. Y. P. Wong, J. Kariks, R. A. Evans, C. R. Dunstan, and E. Hills, The effect of age on bone composition and viability in the femoral head. J. Bone Jt. Surg. 67A, 274 – 283 (1985). 95. C. R. Dunstan, N. M. Somers, and R. A. Evans, Osteocyte death and hip fracture. Calcif. Tissue Int. 53(Suppl. 1), S113 – S117 (1993). 96. S. J. Qui, S. Palnitkar, and D. S. Rao, Age-related changes in osteocyte density and distribution in human cancellous bone. J. Bone Miner. Res. 14(Suppl. 1), S308 (1999). 97. B. D. Snyder, S. Piazza, W. T. Edwards, and W. C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue Int. 53 (Suppl. 1), S14 – S22 (1993). 98. B. Vernon-Roberts and C. J. Pirie, Healing trabecular microfractures in the bodies of lumbar vertebrae. Ann. Rheum. Dis. 32, 406 – 412 (1973). 99. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, M. W. Mathews, and E. Teubner, Bone matrix and mineral abnormalities in postmenopausal osteoporosis. Metabolism 31, 1113 – 1120 (1982). 100. A. M. Parfitt, Bone age, mineral density and fatigue damage. Calcif. Tissue Int. 53 (Suppl. 1), S82 – S86 (1993). 101. T. A. Einhorn, Bone strength: The bottom line. Calcif. Tissue Int. 51, 333 – 339 (1992). 102. M. Grynpas, Age and disease-related changes in the mineral of bone. Calcif. Tissue Int. 53 (Suppl. 1), S57 – S64 (1993). 103. M. B. Schaffler, T. M. Boyce, K. D. Lundin-Cannon, C. Milgrom, and D. P. Fyhrie, “Age-Related Architectural Changes and Microdamage Accumulation in the Human Femoral Neck Cortex,” p. 549. 41st Annual Meeting, Orthopaedic Research Society, 1995. [Abstract] 104. M. B. Schaffler, K. Choi, and C. Milgrom, Microcracks and aging in human femoral compact bone. Bone 17 (1995). 105. M. A. R. Freeman, R. C. Todd, and C. J. Pirie, The role of fatigue in the pathogenesis of senile femoral neck fractures. J. Bone Jt. Surg. 56B, 698 – 702 (1974).

CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling 106. N. L. Fazzalari, Trabecular microfracture. Calcif. Tissue Int. 53(Suppl. 1), S143 – S147 (1993). 107. M. Wicks, R. Garrett, B. Vernon-Roberts, and N. Fazzalari, Absence of metabolic bone disease in the proximal femur in patients with fracture of the femoral neck. J. Bone Jt. Surg. 64B, 319 – 322 (1982). 108. A. M. Parfitt, The actions of parathyroid hormone on bone: Relation to bone remodeling and turnover, calcium homeostasis and metabolic bone disease. I. Mechanisms of calcium transfer between blood and bone and their cellular basis. Morphologic and kinetic approaches to bone turnover. Metabolism 25, 809 – 844 (1976). 109. R. B. Martin, On the histologic measurement of osteonal BMU activation frequency. Bone 15, 547 – 549 (1994). 110. C. J. Hernandez, S. J. Hazelwood, and R. B. Martin, The relationship between basic multicellular unit activation and origination in cancellous bone. Bone 25, 583 – 587 (1999). 111. D. E. Hughes, K. R. Wright, A. Sasaki, T. Yoneda, H. Uy, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphonates induce osteoclast apoptosis in vivo and in vitro, but calcitonin does not. J. Bone Miner. Res. 9 (Suppl. 1), S347 (1994). 112. R. Boyce, personal communication. 113. H. Fleisch, “Bisphosphonates in Bone Disease: From the Laboratory to the Patient,” 3rd Ed. Patterson Publishing Group/New York, 1997. 114. D. E. Hughes, K. R. Wright, H. L. Uy, S. Saski, T. Yoned, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphates promote apoptosis in murine osteoclasts in vitro and in vivo. J. Bone Miner. Res. 10, 1478 – 1487 (1995). 115. R. W. Boyce, C. L. Paddock, J. R. Gleason, W. K. Sletsema, and E. F. Eriksen, The effects of risedronate on canine cancellous bone remodeling: Three-dimensional kinetic reconstruction of the remodeling site. J. Bone Miner. Res. 10, 211 – 221 (1995).

447

116. L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. In “Bone Histomorphometry: Third International Workshop” (W. S. S. Jee and A. M. Parfitt, eds.), pp. 389 – 407. ArmourMontagu, Paris, 1981. 117. B. L. Riggs, L. J. Melton III, and W. M. O’Fallon, Postmenopausal osteoporosis: Evidence that antiresorptive and formation-stimulating regimens decrease vertebral fracture rate by independent mechanisms. In “Proceedings of the Fourth International Symposium on Osteoporosis” (C. Christiansen and B. Riis, eds.), Vol. 102, pp. 13 – 15. Hong Kong, Handelstrykkeriet Aalborg ApS, Aalborg, Denmark, 1993. 118. U. A. Liberman, S. R. Weiss, J. Broll et al., Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N. Engl. J. Med. 333, 1437 – 1443 (1995). 119. M. R. Forwood, D. B. Burr, Y. Takano, D. F. Eastman, P. N. Smith, and J. D. Schwardt, Risedronate treatment does not increase microdamage in the canine femoral neck. Bone 16, 643 – 650 (1995). 120. T. Mashiba, T. Hirano, C. H. Turner, M. R. Forwood, C. C. Johnston, and D. B. Burr, Suppressed bone turnover by bisphophonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J. Bone Miner. Res. 15, 613 – 620 (2000). 121. K. L. Bell, N. Loveridge, G. Jorda, J. Power, and J. Reeve, Merging of haversian canals within remodeling clusters; a cause of focal weakness in the femoral neck cortex? J. Bone Miner. Res. 14 (Suppl. 1), S265 (1999). 122. M. C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994).

CHAPTER 16

Basic Biology of Bisphosphonates H. FLEISCH

University of Berne, CH-3008 Berne, Switzerland

I. Introduction II. Chemistry III. Actions

IV. Pharmacokinetics V. Animal Toxicology References

I. INTRODUCTION

P – C – P structure, the compounds are called geminal bisphosphonates. They are therefore analogues of pyrophosphate that contain a carbon instead of an oxygen atom. For the sake of simplicity, and because only P – C – P bisphosphonates have been found to exert strong activity on the skeleton, the geminal bisphosphonates will simply be called bisphosphonates in this review. This simplification is usually also made in the literature (Fig. 1). The geminal bisphosphonates have been known for a long time, the first synthesis by German chemists dating back to 1865 [7]. Etidronate was synthesized as early as 1897 [8]. They were used for a variety of industrial applications, among them as antiscaling agents [8]. The P – C – P structure allows a great number of possible variations, either by changing the two lateral chains on the carbon atom or by esterifying the phosphate groups. The first report on the biological action of bisphosphonates dates back to 1968/1999 [9– 11] Since then, many bisphosphonates have been investigated in animals and humans with respect to their effect on bone. Today, seven — alendronate, clodronate, etidronate, ibandronate, pamidronate, risedronate, and tiludronate — are available commercially in some countries for use in human bone disease (Fig. 2).

Bisphosphonates are a class of drugs developed in the past three decades for use in various diseases of bone and calcium metabolism. This chapter deals with the preclinical aspects of these compounds, with emphasis on those related to osteoporosis. The topics discussed are divided into the following sections: chemistry, effects on bone resorption and their mechanisms, effects on mineralization and their mechanisms, other effects, pharmacokinetics, and animal toxicology. Because the literature in this field is very large, references are generally restricted in the specific topic to the original ones and to those bringing new knowledge, as well as to newer reviews. For a more complete general update of the preclinical aspects of bisphosphonates, several newer reviews are available [1 – 6]. The clinical aspects of bisphosphonates will be covered in Chapter 72.

II. CHEMISTRY Bisphosphonates, formerly called diphosphonates, are compounds characterized by two C – P bonds. If the two bonds are located on the same carbon atom, resulting in a

OSTEOPOROSIS, SECOND EDITION VOLUME 1

449

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

450

H. FLEISCH

Each bisphosphonate has its own physicochemical and biological characteristics. This variability in effect makes it impossible to extrapolate with certainty from data for one compound to others so that each compound has to be considered on its own, with respect to both its use and its toxicology. The P – C – P bonds of bisphosphonates are stable to heat and most chemical reagents, and completely resistant to enzymatic hydrolysis, but can be hydrolyzed in solution by FIGURE 1

Chemical structure of pyrophosphate and bisphosphonates.

FIGURE 2 Chemical structure of bisphosphonates investigated for their effect on bone in humans. *Available commercially. From Fleisch [5].

451

CHAPTER 16 Basic Biology of Bisphosphonates

FIGURE 2

ultraviolet light. These compounds have a strong affinity for metal ions, among them calcium, with which they can form both soluble and insoluble complexes and aggregates, depending on the pH of the solution and the specific metal [12]. This can occur in vivo when large amounts are infused rapidly so great care must be taken when these compounds are given intravenously. Some uncertainty still exists as to the state of bisphosphonates when in solution.

III. ACTIONS A. Inhibition of Bone Resorption The main effect of pharmacologically active bisphosphonates is to inhibit bone resorption. Indeed, these compounds proved to be extremely powerful inhibitors of resorption when tested in a variety of conditions, both in vitro and in vivo.

(continued)

1. EFFECTS IN VITRO Bisphosphonates block bone resorption induced by various means in cell and organ culture. In the former, they inhibit the formation of pits by isolated osteoclasts cultured on mineralized substrata [13,14]. In organ culture they decrease the destruction of bone in embryonic long bones and in neonatal calvaria [15,16]. This inhibition is present whether resorption is stimulated or not. Up until now, the effect of all the stimulators of bone resorption, such as parathyroid hormone, 1,25(OH)2D, prostaglandins, and products of tumor cells, as well as others, has been inhibited by biphosphonates. In the past, the correlation between results obtained in calvaria in vitro and those obtained in vivo was rather poor. However, a more recent study performed with nine compounds varying in their activity by five to six orders of magnitude showed a satisfactory correlation using the 4- to 7-day-old mouse calvaria assay [17].

452

H. FLEISCH

2. EFFECTS IN VIVO a. Intact Animals In growing rats, bisphosphonates can block the degradation of both primary and secondary trabeculae, thus arresting the modeling and remodeling of metaphyses [18]. The latter therefore become club-shaped and radiologically more dense than normal, leading to a picture similar to that seen in congenital osteopetrotic animals. This effect, called the “Schenk assay,” is often used as an experimental assay to estimate the potency of new compounds [19] (Fig. 3). The inibition of bone resorption by bisphosphonates has also been documented using 45Ca kinetic studies and hydroxyproline excretion [20], as well as by other means. The effect occurs within 24 – 48 h [21] and is therefore slower than that of calcitonin. The decrease in resorption is accompanied, at least in the growing animal, by a positive calcium balance and an increase in the mineral content of bone and in bone mass

FIGURE 3

[20]. This is possible because of an increase in intestinal calcium absorption, consequent to an elevation of 1,25(OH)2D. The increase is, however, smaller than expected considering the dramatic decrease in bone resorption. This is due to the fact that, after a certain time, bone formation also decreases [20] because of the so-called coupling between formation and resorption characteristic of a decrease in remodeling. The main effect of bisphosphonates is therefore a reduction in bone turnover. It is not known how long the increase in balance lasts in the rat after discontinuation of the bisphosphonate. This increase is the basis for the administration of these compounds to prevent and treat osteoporosis in humans. Less is known about the effect in the normal adult animal. In dogs and minipigs, the long-term administration of alendronate did not lead to an increase in bone mass [22]. This might be explained by the physiological biomechanical homeostasis of bone structure, which would eliminate a

Inhibition of metaphyseal modeling and remodeling by a bisphosphonate in the growing rat. (Top) Locations of bone resorption in the rat tibia during longitudinal growth (left): osteoclasts resorb calcified cartilage (1), subperiosteal bone (2), and primary spongiosa (3), therefore enlarging the marrow cavity. Effect of clodronate (right). (Bottom) Microradiograph of a normal tibia (left) and of a bone from an animal treated with clodronate (right). Adapted from Schenk et al. (1973). Reproduced from Calcif. Tissue Res. 11, 196 – 214, with copyright permission from the author and Springer-Verlag, Heidelberg.

CHAPTER 16 Basic Biology of Bisphosphonates

453

biomechanically unnecessary excess of bone. This fact suggests that the fear of the dangers of long-term use of therapeutic doses may not be warranted. b. Animals with Experimentally Increased Bone Resorption Bisphosphonates can also prevent an experimentally induced increase in bone resorption. Thus, they impair resorption induced by many bone-resorbing agents such as, among others, parathyroid hormone [10,15], 1,25(OH)2D, and retinoids [23], the latter effect having been used to develop a powerful and rapid screening assay for new compounds. c. Animals with Experimentally Induced Osteoporosis Many osteoporosis models have been investigated for their response to the administration of bisphosphonates. All showed that bone loss could be prevented by bisphosphonates (Fig. 4). The first used was immobilization by a sciatic nerve section in the rat. The administration of etidronate and especially clodronate not only prevented bone loss, but actually led to an increase of radiological bone density and calcium content of the immobilized bone [24,25]. Similar results were obtained with spinal cord section [26] and other means of immobilization (Fig. 5). Because ovariectomy produces significant bone loss in animals it has been used frequently as a model for human postmenopausal osteoporosis. Bisphosphonates have been found to be very active in preventing the induced bone loss in rats [24 – 33], monkeys [34 – 36], dogs [37,38], and other animals. Also the loss induced by castration in males is attenuated by bisphosphonates [39,40]. Another model of clinical interest is corticosteroid induced bone loss. This loss is inhibited by bisphosphonates in the rabbit [41] and other animals. The same is true for the loss induced by other hormones, such as thyroid hormones [42,43]. Effects on the loss induced by low calcium diet are ambiguous. An interesting new model has been developed to mimic osteolysis and aseptic loosening around total hip arthroplasty. In this model, alendronate also inhibited bone destruction [44]. Of clinical importance is the finding that treatments that increase bone formation, such as prostaglandins, IGF-I and

FIGURE 4 Fleisch [5].

Osteoporosis models improved by bisphosphonates. From

FIGURE 5 Effect of 10 mg phosphorus/kg of clodronate subcutaneously on the bone loss induced by a sciatic nerve section in the rat. From Fleisch [5]. PTH, are still effective or maintained after their discontinuation in bisphosphonates-treated animals [32,45], resulting sometimes even in an additive effect of the two treatments. This additive effect of stimulators of bone formation and inhibitors of bone destruction opens an interesting possibility for future therapy. Another question was whether bisphosphonates could display an additive effect together with another inhibitor of bone resorption. This has been shown to be the case with estrogen in humans [46,47]. Practically all of the bisphosphonates tested have been effective. Listed in order of increasing potency in animals, these include etidronate [24,25,27,29,33], tiludronate [31,37,40], clodronate [24,25,39,41], pamidronate [42], olpadronate [26], incadronate [45], alendronate [30,34,35,43], risedronate [28,29,33], ibandronate [38], minodronate, and zoledronate [36]. In the case of etidronate, the effect was blurred at higher doses because they also inhibited mineralization. In the ovariectomized rat the preventive effect was maintained for a long period after discontinuation of the drug. However, many of these studies were performed in growing animals, in which the bisphosphonates increased bone mass conspicuously by inhibiting the resorption of the metaphyseal bone. In such animals, it is most often not possible to know whether an inhibition of the bone loss induced by the various procedures is due only to the effect of the compound on the induced bone loss or to a general effect on endogenous bone resorption. This is especially the case because the proper controls, namely nonosteoporotic animals, are often not available. d. Mechanisms of Action in Bone Loss The mechanism of action of the bisphosphonates in osteoporosis is still not completely understood. The prevention of bone loss is probably explained to a large extent by the decrease

454 in bone turnover, which by itself slows bone loss. This decrease also diminishes the fracture rate, as fewer trabeculae are destroyed. One explanation of the initially occurring increase in bone is that the decrease in bone resorption is followed only later by the “coupling-induced” diminution in formation, which brings an initial gain in calcium and bone balance through the reduction of the so-called remodeling space. In addition, bisphosphonates may also act at the individual bone mineral density (BMU). Indeed, it has been shown that they decrease the depth of the resorption site [35,48]. Because the amount of bone formed at each individual BMU is not decreased, but may possibly even be somewhat increased in animals [35,48] and humans [49], bone balance may be increased at this site. Whether this occurs is under discussion. Actually no increase of the trabecular bone has been reported [49] (Fig. 6). It must be remembered that in nearly all studies BMD and not real bone mass has been measured. A lower turnover will lengthen the life span of the BMU, thus permitting it to mineralize more completely, which will increase BMD without increasing bone mass. This has been described in alendronate-treated baboons [50]. Finally, it has been suggested that bisphosphonates may, to some extent, increase bone formation. Thus very low concentrations of bisphosphonates were found to increase cell multiplication, colony formation, nodule formation, mineralization, and osteocalcin synthesis in bone cell cultures (51– 53). As mentioned previously, some results suggest that the bone formed in individual BMU is possibly somewhat increased under bisphosphonate treatment [48]. Finally, since statins, which are inhibitors of

H. FLEISCH

the mevalonate pathway, increase bone formation at least in vitro, probably through an elevation of BMP-2, it could be conceivable that the bisphosphonates, which also inhibit the mevalonate pathway (see later), have a similar action. Thus, it could be that bisphosphonates might, under certain conditions, increase bone formation in vivo. However, this still needs to be verified.

e. Effect on Mechanical Properties of Bone The effects of bisphosphonates on the mechanical properties of the skeleton have been addressed only recently. This issue is of importance, as it is known that a long-lasting, extreme inhibition of bone resorption can lead to increased bone fragility, with the human osteopetrosis described by Albers – Schönberg being a good illustration. It appears that when not given in excess, bisphosphonates have a positive effect on mechanical characteristics, such as torsional torque, ultimate bending strength, stiffness, maximum elastic strength, Young’s modulus of elasticity, and others, both in normal animals and in various experimental osteoporosis models. The bisphosphonates that proved to be active include alendronate [22,54], clodronate, etidronate [27, 55,57], ibandronate, incadronate [56], minodronate, neridronate, olpadronate [57], pamidronate [55,58], risedronate, and tiludronate [59]. In contrast, when given in excess, bisphosphonates can induce bone to become more prone to fractures both because of an inhibition of mineralization, mainly with etidronate [60], or because of a excessive inhibition of bone resorption [60]. For review, see Ferretti [61] (Fig. 7).

FIGURE 7

FIGURE 6 Possible effect of bisphosphonates at the level of the individual BMU. From Fleisch [5].

Effect of alendronate given intravenously every 2 weeks for a period of 2 years to ovariectomized baboons on bone mineral density and mechanical strength. Squares, not ovariectomized; triangles, ovariectomized; circles, ovariectomized  0.05 mg/kg; diamonds, ovariectomized  0.25 mg/kg. MPa, megapascal. Adapted from Balena et al. [35]. Reproduced from J. Clin. Invest. 92, 2577 – 2586, with copyright permission from the author and the American Society of Clinical Investigation.

455

CHAPTER 16 Basic Biology of Bisphosphonates

The mechanisms leading to the improved mechanical strength are still poorly understood. They may not be caused uniquely, as previously thought, by a higher bone mass, but also by an improvement in architecture and probably a reduction in bone remodeling. Indeed, a higher number of remodeling sites, in which there is excessive osteoclastic destruction of bone, leads to the development of areas of stress concentration, and hence to increased fracture risk. Bisphosphonates may serve as a means of reducing these effects, hence reducing the incidence of new fractures. Another mechanism may be related to the increase in the mineral density following the lower bone turnover [50]. f. Animals with Experimentally Induced Osteolytic Tumors Bisphosphonates also inhibit bone resorption induced experimentally in vitro [62] or in vivo [63,64] by implantation of various tumor cells. They reduce the local destruction of bone near the invading tumor cells, as well as the resorption induced by systemically circulating factors. These effects lead to a partial or total prevention of hypercalcemia and hypercalciuria [63] and to a decrease of metastases and tumor burden [64,65]. This effect is the basis for their use in tumor-related bone disease. For review, see Fleisch [66] and Russell and Croucher [67]. g. Others Of interest in the dental field is the fact that they also slow down periodontal bone destruction in rats susceptible to periodontal disease and in experimental periodontitis in monkeys. Furthermore, they inhibit tooth movement induced by orthodontic procedures [68], and these effects can be achieved even when the compounds are administered topically [69]. Finally, several bisphosphonates inhibit local bone and cartilage resorption, preserve the joint architecture, and decrease the inflammatory reaction in various types of experimental arthritis [70]. 3. RELATIVE ACTIVITY OF BISPHOSPHONATES ON BONE RESORPTION The potency of bisphosphonates on bone resorption varies greatly from compound to compound. For etidronate, the dose required to inhibit resorption is relatively high: in the rat above 1 mg/kg parenterally per day. This dose is

FIGURE 8

very near that which impairs normal mineralization. One of the aims of bisphosphonate research has therefore been to develop compounds with a more powerful antiresorptive activity, without a stronger inhibition of mineralization. This has proven to be possible. Clodronate was already more potent than etidronate [10,15], and pamidronate was found to be even more active [71]. In more recent years, compounds have been developed that are up to 10,000 times more powerful than etidronate in the inhibition of bone resorption in experimental animals without being more active in inhibiting mineralization (Fig. 8). For the development of future compounds, it is of relevance that, so far, the potency evaluated in the rat corresponds quite well to that found in humans, at least with respect to their relative place in the scale of potency. However, the difference of activity between the least and the most potent compound is less in humans and depends on the disease for which the compounds are used. It is much smaller for osteoporosis, less so for Paget’s disease and tumor-induced hypercalcemia. At present, the structural requirements for activity are only partially defined. The length of the aliphatic carbon chain is important, with the effect on bone resorption increasing and then decreasing again with increasing chain length [72]. Adding a hydroxyl group to the carbon atom at position 1 increases activity, and compounds with a nitrogen atom in the side chains are more active. The first compound of the latter kind to be described, pamidronate, has an amino group at the end of the alkyl chain [71]. When the chain is altered in its length, the highest activity is present with a backbone of four carbons, as seen in alendronate [19]. A primary amine is not necessary for this activity, as dimethylation of the amino nitrogen of pamidronate, as seen in olpadronate, increases potency [73]. The latter can still be increased further when other groups are added to the nitrogen, as is the case for ibandronate, [1-hydroxy3[methylpentylamino)propylidene]bisphosphnate, which is extremely potent [74]. Geminal bisphosphonates containing cyclic substituents are also very potent, especially those containing a nitrogen atom in the ring, such as risedronate [75]. The most active compounds described so far, zoledronate [17] and minodronate, belong to this class and contain an imidazole ring.

Potency of some bisphosphonates to inhibit bone resorption in the rat. Compounds in each column are listed in alphabetical order. From Fleisch [5].

456 It must be noted that, at present, all effective compounds have a P – C – P structure, which appears to be a prerequisite for activity. The intensity of the effect is, however, also dependent on the side chain. A three-dimensional structural requirement appears to be involved. Indeed, stereoisomers of the same chemical structure have shown 10-fold differences in activity [76]. This opens the possibility of binding onto some kind of receptor. 4. MECHANISMS OF ACTION ON BONE RESORPTION a. General Concepts Our understanding of the mode of action of the bisphosphonates has made great progress. There is no doubt that the action in vivo is mediated mostly, if not completely, through mechanisms other than the physicochemical inhibition of crystal dissolution, as was initially postulated. However, the exact nature of these mechanisms is still not entirely unraveled. It may well be that several mechanisms operate simultaneously. b. Levels of Action The mechanism of action of the bisphosphonates can be considered at three levels, which are, however, tightly linked one to another. At the tissue level their main effect is a decrease in bone turnover, which is secondary to the inhibition of bone resorption. This effect is due to a decrease in the number and the activity of osteoclasts destroying bone [49,77], which leads to a decrease in the number of new BMUs. Because bone loss is intimately linked to turnover in diseases such as tumor-induced bone disease and osteoporosis, this loss will be reduced by the bisphosphonates. Furthermore, the bisphosphonates act to a certain extent at the individual BMU level by decreasing the depth of the resorption site [35,48,77]. Because the amount of new bone formed in the BMU is not decreased, but possibly even increased [35,48,50,77], local and consequently the whole body bone balance will be less negative or possibly sometimes even positive. Four mechanisms possibly can be involved at the cellular level. 1. Inhibition of osteoclast recruitment. Several bisphosphonates inhibit osteoclast differentiation in various culture systems of both cells [78] and bones [79,80]. Some experiments suggest that the effect occurs at the terminal step of the differentiation process [81]. 2. Inhibition of osteoclastic adhesion to the mineralized matrix. One study in vitro reports such an effect [82]. Whether this takes place in vivo is not yet established. However, there is now excellent evidence that bisphosphonates can inhibit the adhesion of tumor cells in vitro [83,84]. 3. Shortening of the life span of osteoclasts due to earlier apoptosis. It has been reported that bisphosphonates induce

H. FLEISCH

osteoclast programmed cell death (apoptosis], both in vitro and in vivo, and both in normal mice and in mice with increased bone resorption [85]. The ranking of effectiveness of clodronate, pamidronate, and risedronate was the same as seen in vivo. The effect was not due to toxic cell death. 4. Inhibition of osteoclast activity. While the former three mechanisms will lead to a decrease in the number of osteoclasts, which is usually seen after longer treatment, the fourth will lead to inactive osteoclasts, often seen at the beginning of treatment. Several facts suggest that this effect must be present. Thus, following bisphosphonate administration, the number of multinucleated osteoclasts on the bone surface often increases initially, despite a reduced bone resorption [18], however, the cells show changes in morphology and look inactive [18]. The changes are numerous and include alterations in the cytoskeleton, among others, in actin [14,86,87] and vinculin [87], as well as disruption of the ruffled border [18,88 – 90]. It is only later, after chronic administration, that the osteoclast number decreases. The cause for the initial increase is unknown. One possibility is that it could reflect a stimulation of osteoclast formation to compensate for the decrease in osteoclast activity. At the molecular level, the low concentration necessary for activity suggests some sort of “pocket” that induces a cellular transduction mechanism. This site could be either on the cell membrane or within the cell and might be an enzyme, a pump, or some other intracellular protein involved in the signaling cascade. One of these could be farnesyl pyrophosphate synthase, which has been found to be inhibited by nitrogen-containing bisphosphonates and responsible for at least some of their effects (see later) [91 – 93]. A great number of different biochemical effects on various cell types have been described in vitro, but fewer data exist on the osteoclasts themselves. Some of the changes that possibly relate to bone resorption, include reduction in lactic acid production by calvaria [94], various cells [95], and osteoclasts [96,97]. In the latter, an inhibition of the acid extrusion performed by a sodium-independent mechanism and of the vacuolar-type proton ATPase present in the ruffled border has been shown [96,97]. Bisphosphonates also inhibit lysosomal enzyme activity in vitro [98], and decrease prostaglandin synthesis in bone when added both in vitro and in vivo [99,100]. Furthermore, they inhibit matrix metalloproteinases [101], which might be also of interest in view of a possible use in arthritis [102]. An inhibition of certain protein tyrosine phosphatases (PTPase), namely PTPase  and , has also been described [103]. Unfortunately, in many cases there is no structure – effect correlation between these effects in vitro and those on bone resorption in vivo when bisphosphonates of various antiresorbing activity are compared. It has been found that nitrogen-containing bisphosphonates can inhibit the mevalonate pathway [104,105] by in-

457

CHAPTER 16 Basic Biology of Bisphosphonates

It was also shown that some nonnitrogen-containing bisphosphonates that closely resemble pyrophosphate, such as etidronate, tiludronate, and clodronate, can be incorporated into the phosphate chain of ATP-containing compounds so that they become nonhydrolyzable. The new P – C – P containing ATP analogues inhibit cell function and may lead to apoptosis and cell death [107 – 109]. Thus, the bisphosphonates can be classified into two major groups with different modes of action. The latter may explain the various cellular changes described earlier (Fig. 10).

FIGURE 9

Effect of bisphosphonates on the mevalonate pathway. From

Fleisch [5].

hibiting farnesyl pyrophosphate synthase [91 – 93]. This leads to a decrease of the formation of isoprenoid lipids such as farnesyl- and geranylgeranylpyrophosphates. These are required for the posttranslational prenylation (transfer of fatty acid chains) of proteins, including the GTP-binding proteins Ras, Rho, Rac, and Rab. These proteins are important for many cell functions, including cytoskeletal assembly and intracellular signaling. Therefore, disruption of their activity will induce a series of changes leading to decreased activity, probably the main effect, and to earlier apoptosis in several cell types, including osteoclasts [104,105]. In osteoclasts the lack of geranylgeranylpyrophosphate is probably responsible for the effects [106] (Fig. 9).

FIGURE 10

c. Direct vs Indirect Effect through Other Cells It is probable that bisphosphonates influence osteoclasts either directly as a result of their cellular binding or intracellular uptake or indirectly via other cells. The direct effect is supported by many facts. As mentioned earlier, bisphosphonates inhibit the formation of resorption cavities by isolated osteoclasts deposited on calcified matrices in vitro [13,14]. One study [87] showed that morphological changes occurred only when the cells were actively resorbing the calcified matrix or if the bisphosphonate was injected into the cells. No changes occurred when the osteoclasts were not active, showing that the bisphosphonates has to be taken up with the resorbed mineral. A direct action on osteoclasts is also supported by the fact that, under certain conditions, bisphosphonates can enter cells [90,95,110], particularly those of the macrophage lineage [111]. They are taken up by the osteoclasts during the resorption process, a process favored by the fact that bisphosphonates also deposit preferentially under osteoclasts where they can attain very high concentrations, in the range

The two classes of bisphosphonates. Courtesy of Dr. M. Rogers.

458

H. FLEISCH

of 104 M or higher [90], and are then released from the mineral at the acid pH prevailing at this location. It is likely that bisphosphonates also act, at least in part, through other cells. One candidate is the osteoblast. It is now generally accepted that cells of osteoblastic lineage control the recruitment and activity of osteoclasts. One of the modulators involved in this mechanism appears to be bisphosphonates. Indeed, these compounds induce the osteoblasts to synthesize an inhibitor(s) for osteoclast recruitment and therefore of bone resorption [112 – 114]. In this regard it might be relevant that bisphosphonates prevent the apoptotic effect of various bone resorbers on osteoblasts and osteocytes [115]. Another candidate target cell population are the macrophages [111], which release many cytokines, which are able to modulate osteoclasts and are influenced by bisphosphonates. Thus, under certain conditions, bisphosphonates inhibit macrophage release of interleukin (IL)-1, IL-6, and TNF in vitro. Alternatively, at high concentrations, such as after intravenous administration, the release of these cytokines can be stimulated, producing an acute phase reaction in humans [116]. It is not known at present to which extent these mechanisms — the direct effect on the osteoclast or indirect action through osteoblastic or other cells — are operating in vivo and, if both do, which of the two is more important (Fig. 11).

B. Inhibition of Mineralization 1. EFFECTS IN VITRO The physicochemical effects of many of the bisphosphonates are very similar to those of pyrophosphate. Thus, they inhibit the formation [11,117,118], the aggregation, and

FIGURE 11 Summary of the effects of bisphosphonates on the osteoclast. From Fleisch [5].

also slow down the dissolution of calcium phosphate crystals [10]. All these effects are related to the marked affinity of these compounds for solid-phase calcium phosphate, on the surface of which they bind strongly [119]. This property is of great importance because it is the basis for the use of these compounds as skeletal markers in nuclear medicine and the basis for their selective localization in bone when used as drugs (Fig. 12). The inhibition of calcium phosphate formation is closely related to the affinity of the bisphosphonate to the solidphase calcium phosphate. The binding can be bidentate through the two phosphates, as is the case for clodronate, or it can be tridentate [120] through a third moiety, such as a hydroxyl or a nitrogen attached to the carbon atom. This is the case for most bisphosphonates used clinically today. The third binding site increases the affinity and hence the inhibitory effect on calcification. Bisphosphonates also inhibit the formation and the aggregation of calcium oxalate crystals [121]. 2. EFFECTS IN VIVO a. Ectopic Mineralization Like pyrophosphate, bisphosphonates inhibit calcification in vivo very efficiently. Thus, they prevent experimentally induced calcification of many soft tissues such as arteries and kidneys [11,117] and skin. They are active when administered either orally or parenterally. In arteries, they decrease not only mineral deposition, but also the accumulation of cholesterol, elastin, and collagen. Bisphosphonates, such as etidronate, can also inhibit the calcification of bioprosthetic heart valves, either when administered subcutaneously or when released locally from various matrices. Etidronate also inhibits ectopic ossification when given either systemically [122] or locally. This effect has led to the clinical use of etidronate in ectopic ossification, but normal mineralization is unfortunately inhibited as well. Similarly, certain bisphosphonates such as etidronate decrease the formation of experimental urinary stones [121]. However, the active dose also leads to the inhibition of normal mineralization of bone so that these compounds cannot be administered in this condition. Finally, topical administration of etidronate induces a decrease in the formation of dental calculus [123], a property that is exploited in some toothpastes.

FIGURE 12 Physicochemical effects of bisphosphonates on calcium phosphate. From Fleisch [5].

459

CHAPTER 16 Basic Biology of Bisphosphonates

b. Normal Mineralization The dose, at least of etidronate, that inhibits experimental ectopic mineralization also impairs the mineralization of normal calcified tissues such as bone and cartilage (Fig. 3) [18], dentine [124,125], and enamel [125]. The amount required to have this effect varies according to animal species and length of treatment. In contrast to bone resorption, where the different compounds vary greatly in their activity, this does not seem to be so much the case for the inhibition of mineralization. For most species and compounds, the effective daily parenteral dose is in the order of 1 – 10 mg of compound phosphorus per kilogram. Interestingly, clodronate inhibits normal mineralization somewhat less than etidronate, despite the fact that it is more active on bone resorption. This may be due to the fact that clodronate has no hydroxyl side chain and is therefore less bound to the mineral. The inhibition of calcification with high doses can lead to fractures [60] and to an impairment of fracture healing. The mineralization defect is eventually reversed after discontinuation of the drug. Bisphosphonates also inhibit calcification of bone in humans when given in larger amounts. This propensity to inhibit the mineralization of normal bone has hampered the therapeutic use of bisphosphonates in ectopic calcification. This is not the case for their use in bone resorption, as compounds have been developed that inhibit this process at doses at least 1000 times lower than those that inhibit mineralization. OF

3. MECHANISMS OF ACTION IN THE INHIBITION CALCIFICATION

There is a close relationship between the ability of an individual bisphosphonate to inhibit the formation of calcium phosphate in vitro and its effectiveness on calcification in vivo [117,126], strongly suggesting that the latter can be explained in terms of a physicochemical mechanism. However, additional effects on matrix formation, involving changes in glycosaminoglycan and collagen synthesis, may occur. These may be direct via cellular effects or mediated indirectly by effects on crystals. In contrast, there is no relation between crystal binding and bone resorption [127].

C. Other Actions In view of the large array of their effects on cells, it is surprising that bisphosphonates act almost exclusively on calcified tissues. This selectivity is explained by the strong affinity of these compounds for calcium phosphate, which allows them to be cleared very rapidly from blood and to be incorporated into calcified tissues, especially bone (see pharmacokinetics). However, some effects exist that are not, or not entirely, explained by the effects on bone. Thus several bisphosphonates inhibit local bone and cartilage resorption, preserve

the joint architecture, and decrease the inflammatory reaction in various types of experimental arthritis, such as that induced by Freund’s adjuvant, carrageenin, or collagen [128,129]. This effect is especially pronounced when bisphosphonates are encapsulated in liposomes [130]. The fact that not only is bone resorption decreased, but also the inflammatory reaction in the joint and in the paw itself is diminished [131] suggest that mechanisms other than those in bone, possibly involving the mononuclear phagocyte system, are operating. These results open the exciting possibility of using bisphosphonates in inflammatory arthritis, given either systemically or locally, possibly encapsulated in liposomes. Of interest, bisphosphonates or phosphonosulfonates linked to an isoprene chain are potent inhibitors of squalene synthase and hence are cholesterol-lowering agents in animals [132], which may open some interesting new therapeutic possibilities for these drugs.

IV. PHARMACOKINETICS Bisphosphonates are synthetic compounds that have not yet been found to occur naturally in animals or humans. No enzymes able to cleave the P – C – P bonds have been described. The bisphosphonates on which data have been published so far appear to be absorbed, stored, and excreted unaltered from the body. Therefore, these bisphosphonates seem to be nonbiodegradable in solution and in animals. However, it cannot be excluded that some bisphosphonates may be metabolized, especially in their side chains. Data from relatively few pharmacokinetic studies are available. Most of the published data have been obtained with alendronate, clodronate, etidronate, pamidronate, and tiludronate [For reviews, see 133 – 135].

A. Intestinal Absorption The bioavailability of an oral dose of a bisphosphonate in animals as well as in humans is low, lying between  1 and 10% [136], probably because of their low lipophilicity, which prevents transcellular transport, and their high negative charge, which prevents paracellular transport. Absorption is proportionally greater when large doses are given, such as with etidronate and clodronate. This is possibly the reason why it is generally lower for the more potent bisphosphonates, which are administered in lower amounts. Absorption is in general higher in the young and shows great inter- and intraspecies variation. This variability can present a problem in humans, especially for compounds such as etidronate, where the dose that shows an adverse event, such as an inhibition of mineralization, is close to that which inhibits bone resorption. The location of the

460

H. FLEISCH

absorption in the gastrointestinal tract is not yet elucidated, although it can occur in the stomach as well as in the small intestine. It appears to occur by passive diffusion, probably through a paracellular pathway [137]. Absorption is diminished substantially when the drug is given with meals, especially in the presence of calcium and iron. The mechanism of this reduction may be due to the conversion of the bisphosphonate into a nonabsorbable form or to a decrease of the absorption process itself. Therefore, bisphosphonates should never be given at mealtimes and never together with milk or dairy products or with iron supplements. For some unknown reasons, orange juice and coffee also decrease absorption [138].

B. Distribution In the blood, only part of the bisphosphonates are ultrafilterable [139,140]. The values vary between about twothirds to only a few percent and are strongly species dependent, being low for rats and higher for larger animals and humans. The nonfilterable fraction is either bound to proteins, especially albumin [140], or present in very small calcium-containing aggregates. Some 20 – 80% of the absorbed bisphosphonate is then taken up by bone, with the remainder being rapidly excreted in the urine. Skeletal uptake varies with species, sex, and age and with the dose and nature of the compound. In humans receiving clinical doses, values are about 20% for clodronate, 50% for etidronate, and more for alendronate and pamidronate. Sometimes bisphosphonates, especially pamidronate, deposit in other organs, mostly the liver and the spleen [141]. Such deposition is proportionally greater when large amounts of the compounds are given. At least part of this extraosseous deposition appears to be due to the formation of complexes with metals or to aggregates after too high or too rapid intravenous injection. These complexes are then phagocytosed by the macrophages of the reticuloendothelial system. Therefore, data obtained from studies using large amounts of labeled bisphosphonate given rapidly intravenously should be interpreted with caution. The formation of aggregates in the blood is thought to occur in humans following rapid intravenous injections of large quantities, possibly explaining the renal failure that can ensue. The circulating half-life of bisphosphonates is short, in the order of minutes in the rat [142]. In humans it is somewhat longer, 0.5 – 2 h. The rate of entry into bone is very fast, similar to that of calcium and phosphate. Bone clearance is compatible with a complete extraction by the skeleton after the first passage [142] so that skeletal uptake might be determined to a large extent by skeletal vascularization and blood flow. The areas of deposition were generally thought to be mostly those of bone formation. This property is used to measure areas of high bone turnover in nuclear medicine by

means of 99mTc-linked bisphosphonates. However, alendronate, when given in therapeutic doses, has been found to accumulate preferentially under osteoclasts [143]. This is also the case, although to a lesser extent, for etidronate when given in the same amount. When given at a therapeutic dose, the latter, however, accumulates equally under both cells. This suggests that when a bisphosphonate is given in small doses, which is the case for all newer compounds, it is likely to deposit preferentially in locations of bone resorption. The rapid uptake by bone means that the soft tissues are exposed to bisphosphonates for only short periods, explaining why practically only bone is affected in vivo. When bisphosphonates are given to humans in clinically effective doses, there seems to be no saturation in their total skeletal uptake, at least within periods as long as years or decades. In contrast, with continuous administration, the antiresorbing effect reaches a maximum relatively rapidly [71], both in animals and in human. The level of this maximal effect depends on the dose administered, as does the duration of the effect after discontinuation of the drug. The fact that a plateau of activity is reached, despite the fact that the bisphosphonate continues to be incorporated, suggests that the compounds are buried in the bone and become inactive (Fig. 13).

FIGURE 13 Effect of various doses of pamidronate administered daily sc on urinary hydroxyproline excretion in the rat. The maximal effect is obtained rapidly and depends on the dose given. Adapted and reproduced from Reitsma et al. [143]. Calcif. Tissue Int. 32, 145 – 157, with copyright permission from the author and Springer-Verlag, Heidelberg.

461

CHAPTER 16 Basic Biology of Bisphosphonates

Once deposited in the skeleton, part of the bisphosphonate is liberated again by physicochemical mechanisms. Once buried under new layers of bone, they will be released to a large extent only when the bone in which they were deposited is resorbed. Thus the half-life in the body depends on the rate of bone turnover itself. As the bisphosphonates slow down the resorption of the bone in which they are deposited, their half-life may be even longer than the normal half-life of the skeleton. The half-life of various bisphosphonates is between 3 months and up to a year in mice or rats [144], with clodronate being cleared somewhat faster than etidronate and pamidronate. For humans it is much longer, for some bisphosphonates over 10 years [145], and it is possible that a portion of the administered compounds remains in the body for life. However, this is also true for other bone seekers such as tetracyclines, heavy metals, and fluoride. There is no indication that the bisphosphonate buried in the skeleton has any pharmacological activity. On the contrary, in the rat, bone formed under administration of even high doses of alendronate can be resorbed normally. However, at sites where the bisphosphonate is deposited in large amounts, such as in high turnover locations of patients with bone metastases or with Paget’s disease, the long skeletal retention may explain why one single administration of a bisphosphonate can be active for long periods of time, both in animals and in humans.

FIGURE 14

Pharmacokinetics of bisphosphonates. From Fleisch [5].

Acute, subacute, and chronic administration in several animal species have in general revealed little toxicity. Teratogenicity, mitogenicity, and carcinogenicity tests have been negative. When bisphosphonates are administered subcutaneously, local toxicity can occur, with local necrosis. This is especially the case for the nitrogen-containing derivatives such as pamidronate.

A. Acute Toxicity

The renal clearance of bisphosphonates is high. When taking into account their only partial ultrafilterability, it can, at least in animals, exceed that of inulin, indicating active secretion [146,147]. The secretory pathway involved is not yet characterized. Urinary excretion is decreased in renal failure and the removal by peritoneal dialysis is poor, which has to be accounted for when the compounds are administered to patients with kidney disease (Fig. 14).

Acute toxicity can include induction of hypocalcemia and appears to be due mainly to the formation of complexes or aggregates with calcium, which lead to a decrease in ionized calcium. Toxicity therefore varies with the speed of infusion when the compounds are administered intravenously, so that the rate of infusion in humans must be controlled carefully when given in high doses. In contrast, it appears that very potent bisphosphonates, such as ibandronate, may be given safely given in doses up to 2 – 3 mg as an intravenous bolus injection. In the event of hypocalcemia, calcium infusion can correct the signs and symptoms rapidly, Some acute toxicity can also be due to renal tubular effects.

D. Other Modes of Application

B. Nonacute Toxicity

Bisphosphonates are also bioavailable to some extent when given intranasally and through the skin. This may open new modes of administration in clinical practice.

In view of the large array of cellular effects obtained in vitro with the bisphosphonates, one would have expected a large number of toxic effects. This is not the case, and when administered in pharmacological doses, bisphosphonates seem to act almost exclusively on calcified tissues and secondarily on plasma calcium, but are otherwise very well tolerated. This selectivity is explained by the strong affinity of these compounds for calcium phosphate, which allows them to be incorporated rapidly into calcified tissues, especially bone, and therefore to be cleared quickly from the blood.

C. Renal Clearance

V. ANIMAL TOXICOLOGY Published animal toxicological data are scanty and deal mostly with alendronate, clodronate, etidronate, incadronate, pamidronate, and tiludronate. Unfortunately, little is published about other compounds.

462 The nonskeletal toxicity associated with compounds used clinically occurs only when doses substantially larger than those which inhibit bone resorption are used. In general, the first organ to show cellular alterations with all bisphosphonates, as well as with polyphosphates and phosphate itself, is the kidney [148,149]. The liver, as well as the testis, the epididymis, the prostate, and possibly the lung, can in some cases also show alterations. Some inflammatory gastrointestinal changes have been described, with parenteral administration at high doses. After the appearance in humans of gastrointestinal adverse events after oral administration of nitrogen-containing bisphosphonates [150], special attention has been given to the effects of oral bisphosphonates in animals. Thus alendronate, when given orally to rats at suprapharmacological doses, has been reported to occasionally induce gastric and esophageal erosions and ulcerations and delay healing of indomethacin-induced gastric erosions. These effects are not attributable to changes in gastric acid secretion or prostaglandin synthesis, but are thought to be due to a topical irritant effect. No esophageal irritation occurred when the pH was above 3.5 in a dog model. Similar effects were reported with etidronate, risedronate, and tiludronate when given at pharmacologically equivalent doses. All these effects were obtained at doses much larger than those given to humans [151,152]. The most relevant toxicity associated with bisphosphonates is the inhibition of bone and cartilage calcification [17,22,153]. This starts to occur at parenteral doses of approximately 5 – 10 mg/kg daily. The radiological appearance resembles rickets or osteomalacia, although there are some histological differences. Fractures can occur after the long-term administration of high doses and are probably the result of defective mineralization. Developmental disturbances of enamel can also appear at high systemic doses. Fractures can also be induced occasionally by very large amounts inducing a pathological decrease in bone resorption without signs of osteomalacia or rickets [22]. The fractures may be caused by the extreme long-term decrease in bone turnover, which can itself lead to an increased fragility, as is well known in human congenital osteopetrosis, or by bone cell toxicity. Finally, certain bisphosphonates, such as etidronate and pamidronate, cross the placenta and can affect the fetus. Very large doses can lead to a decrease in the number of live pups, to fetal abnormalities of the skeleton and the skin, and to malformations and hemorrhages [154 – 158). In view of these results, bisphosphonates should, as a general rule, not be administered to pregnant women. It must be stressed that the results with one bisphosphonate cannot necessarily be extrapolated to other bisphosphonates. Indeed, toxicity, both in cell and organ culture and in vivo, varies greatly from one compound to another.

H. FLEISCH

References 1. G. A. Rodan and H. A. Fleisch, Bisphosphonates: Mechanisms of action. J. Clin. Invest. 97, 2692 – 2699 (1996). 2. F. H. Ebetino, M. D. Francis, M. J. Rogers, and R. G. G. Russell, Mechanisms of action of etidronate and other bisphosphonates. Rev. Contemp. Pharmacother. 9, 233 – 243 (1998). 3. H. Fleisch, Bisphosphonates: Mechanisms of action. Endocr. Rev. 19, 80 – 100 (1998). 4. G. A. Rodan, Mechanisms of action of bisphosphonates. Annu. Rev. Pharmacol. Toxicol. 38, 375 – 388 (1998). 5. H. Fleisch, “Bisphosphonates in Bone Disease: From the Laboratory to the Patient.” Academic Press San Diego, 2000. 6. M. J. Rogers, J. C. Frith, S. P. Luckman, F. P. Coxon, H. L. Benford, J. Monkkonen, S. Auriola, K. M. Chilton, and R. G. G. Russell, Molecular mechanisms of action of bisphosphonates. Bone 24, 73S – 79S (1999). 7. N. Menschutkin, Ueber die Einwirkung des Chloracetyls auf phosphorige Säure. Ann. Chem. Pharm. 133, 317 – 320 (1865). 8. L. J. M. J. Blomen, History of bisphosphonates: Discovery and history of the non-medical uses of bisphosphonates. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A., Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 111 – 124. Elsevier, Amsterdam, 1995. 9. H. Fleisch, R. G. G. Russell, S. Bisaz, P. A. Casey, and R. C. Mühlbauer, The influence of pyrophosphate analogues (diphosphonates) on the precipitation and dissolution of calcium phosphate in vitro and in vivo. Calcif. Tissue Res. 2(Suppl.), 10 – 10A (1968). 10. H. Fleisch, R. G. G. Russell, and Francis, M. D. Diphosphonates inhibit hydroxyapatite dissolution in vitro and bone resorption in tissue culture and in vivo. Science 165, 1262 – 1264 (1969). 11. M. D. Francis, R. G. G. Russell, and H. Fleisch, Diphosphonates inhibit formation of calcium phosphate crystals in vitro and pathological calcification in vivo. Science, 165, 1264 – 1266 (1969). 12. M. L. Lamson, J. L. Fox, and W. I. Huguchi, Calcium and 1-hydroxyethylidene-1,1-bisphosphonic acid: Polynuclear complex formation in the physiological range of pH. Int. J. Pharm. 21, 143 – 154 (1984). 13. A. M. Flanagan and T. J. Chambers, Dichloromethylene-bisphosphonate (Cl2MBP) inhibits bone resorption through injury to osteoclasts that resorb Cl2MBP-coated bone. Bone Miner. 6, 33 – 43 (1989). 14. M. Sato and W. Grasser, Effects of bisphosphonates on isolated rat osteoclasts as examined by reflected light microscopy. J. Bone Miner. Res. 5, 31 – 40 (1990). 15. R. G. G. Russell, R. C. Mühlbauer, S. Bisaz, D. A. Williams, and H. Fleisch, The influence of pyrophosphate, condensed phosphates, phosphonates and other phosphate compounds on the dissolution of hydroxyapatite in vitro and on bone resorption induced by parathyroid hormone in tissue culture and in thyroparathyroidectomised rats. Calcif. Tissue Res. 6, 183 – 196 (1970). 16. J. J. Reynolds, C. Minkin, D. B. Morgan, D. Spycher, and H. Fleisch, The effect of two diphosphonates on the resorption of mouse calvaria in vitro. Calcif. Tissue Res. 10, 302 – 313 (1972). 17. J. R. Green, K. Müller, and K. A. Jaeggi, Preclinical pharmacology of CGP 42’446, a new, potent, heterocyclic bisphosphonate compound. J. Bone Miner. Res. 9, 745 – 751 (1994). 18. R. Schenk, W. A. Merz, R. Mühlbauer, R. G. G. Russell, and H. Fleisch, Effect of ethane-1-hydroxy-1,1-diphosphonate (EHDP) and dichloromethylene diphosphonate (Cl2MDP) on the calcification and resorption of cartilage and bone in the tibial epiphysis and metaphysis of rats. Calcif. Tissue Res. 11, 196 – 214 (1973). 19. R. Schenk, P. Eggli, H. Fleisch, and S. Rosini, Quantitative morphometric evaluation of the inhibitory activity of new aminobisphosphonates on bone resorption in the rat. Calcif. Tissue Int. 38, 342 – 349 (1986).

CHAPTER 16 Basic Biology of Bisphosphonates 20. A. B. Gasser, D. B. Morgan, H. A. Fleisch, and L. J. Richelle, The influence of two diphosphonates on calcium metabolism in the rat. Clin. Sci. 43, 31 – 45 (1972). 21. R. C. Mühlbauer and H. Fleisch, A method for continual monitoring of bone resorption in rats: Evidence for a diurnal rhythm. Am. J. Physiol. 259, R679 – R689 (1990). 22. C. P. Peter, J. Guy, M. Shea, W. Bagdon, W. F. Kline, and W. C. Hayes, Long-term safety of the aminobisphosphonate alendronate in adult dogs. I. General safety and biomechanical properties of bone. J. Pharmacol. Exp. Ther. 276, 271 – 276 (1996). 23. U. Trechsel, A. Stutzer, and H. Fleisch, Hypercalcemia induced with an arotinoid in thyroparathyroidectomized rats: New model to study bone resorption in vivo. J. Clin. Invest. 80, 1679 – 1686 (1987). 24. W. R. Michael, W. R. King, and M. D. Francis, Effectiveness of diphosphonates in preventing “osteoporosis” of disuse in the rat. Clin. Orthopaed. 78, 271 – 276 (1971). 25. R. C. Mühlbauer, R. G. G. Russell, D. A. Williams, and H. Fleisch, The effects of diphosphonates, polyphosphates, and calcitonin on “immobilisation” osteoporosis in rats. Eur. J. Clin. Invest. 1, 336 – 344 (1971). 26. A. Schoutens, M. Verhas, N. Dourov, P. Bergmann, F. Caulin, A. Verschaeren, M. Mone, and A. Heilporn, Bone loss and bone blood flow in paraplegic rats treated with calcitonin, diphosphonate, and indomethacin. Calcif. Tissue Int. 42, 136 – 142 (1988). 27. E. Shiota, Effects of diphosphonate on osteoporosis induced in rats: Roentgenological, histological and biomechanical studies. Fukuoka Acta Med. 76(6), 317 – 342 (1985). 28. J. E. McOsker and W. K. Sietsema, Preclinical pharmacology of risedronate, a novel bisphosphonate. In “Osteoporosis 1990, Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), Osteopress ApS, Copenhagen, 1990. 29. T. J. Wronski, C-F. Yen, and K. S. Scott, Estrogen and diphosphonate treatment provide long-term protection against osteopenia in ovariectomized rats. J. Bone Miner. Res. 6, 387 – 394 (1991). 30. J. G. Seedor, H. A. Quartuccio, and D. D. Thompson, The bisphosphonate alendronate (MK-217) inhibits bone loss due to ovariectomy in rats. J. Bone Miner. Res. 6, 339 – 346 (1991). 31. P. Ammann, R. Rizzoli, J. Caverzasio, T. Shigematsu, D. Slosman, and J. P. Bonjour, Effects of the bisphosphonate tiludronate on bone resorption, calcium balance, and bone mineral density. J. Bone Miner. Res. 8, 1491 – 1498 (1993). 32. W. S. S. Jee, L. Tang, H. Z. Ke, R. B. Setterberg, and D. B. Kimmel, Maintaining restored bone with bisphosphonate in the ovariectomized rat skeleton: Dynamic histomorphometry of changes in bone mass. Bone 14, 493 – 498 (1993). 33. T. J. Wronski, L. M. Dann, H. Qi, and C. F. Yen, Skeletal effects of withdrawal of estrogen and diphosphonate treatment in ovariectomized rats. Calcif. Tissue Int. 53, 210 – 216 (1993). 34. D. D. Thompson, J. G. Seedor, H. Quartuccio, H. Solomon, C. Fioravanti, J. Davidson, H. Klein, R. Jackson, J. Clair, D. Frankenfield, E. Brown, H. A. Simmons, and G. A. Rodan, The bisphosphonate, alendronate, prevents bone loss in ovariectomized baboons. J. Bone Miner. Res. 7, 951 – 960 (1992). 35. R. Balena, B. C. Toolan, M. Shea, A. Markatos, E. R. Myers, S. C. Lee, E. E. Opas, J. G. Seedor, H. Klein, D. Frankenfield, H. Quartuccio, C. Fioravanti, J. Clair, E. Brown, W. C. Hayes, and G. A. Rodan, The effects of 2-year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry, and bone strength in ovariectomized non human primates. J. Clin. Invest. 92, 2577 – 2586 (1993). 36. N. Binkley, D. Kimmel, J. Bruner, A. Haffa, B. Davidowitz, C. Meng, V. Schaffer, and J. Green, Zoledronate prevents the development of absolute osteopenia following ovariectomy in adult Rhesus monkeys. J. Bone Miner. Res. 13, 1775 – 1782 (1998).

463 37. M. C. Vernejoul, de, Y. Jiang, F. Lacheretz, A. Barbier, P. Geusens, C. Morieux, and C. Pfersdorff, Prevention of bone loss following tiludronate administration to ovariectomized beagle dogs. In “Osteoporosis 1990, Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), pp. 1119 – 1122. Osteopress ApS, Copenhagen, 1990. 38. M.-C. Monier-Faugère, R. M. Friedler, F. Bauss, and H. H. Malluche, A new bisphosphonate, BM 21.0955, prevents bone loss associated with cessation of ovarian function in experimental dogs. J. Bone Miner. Res. 8, 1345 – 1355 (1993). 39. C. S. Wink, M. S. Onge, and B. Parker, The effects of dichloromethylene bisphosphonate on osteoporotic femora of adult castrated male rats. Acta Anat. 124, 117 – 121 (1985). 40. A. Barbier, X. Edmonds-Alt, J. C. Brelière, and D. Ethgen, In vitro and in vivo osseous pharmacological profile of tiludronate: Implication for osteoporosis treatment. In “Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), pp.1127 – 1130 Osteopress ApS, Copenhagen, 1990. 41. W. S. S. Jee, H. E. Black, and J. E. Gotcher, Effect of dichloromethane diphosphonate on cortisol-induced bone loss in young adult rabbits. Clin. Orthop. Res. 158, 39 – 51 (1981). 42. H. N. Rosen, E. K. Sullivan, V. L. Middlebrooks, A. J. Zeind, C. Gundberg, R. Dresner-Pollak, L. A. Maitland, J. M. Hock, A. C. Moses, and S. L. Greenspan, Parenteral pamidronate prevents thyroid hormone-induced bone loss in rats. J. Bone Miner. Res. 8(10), 1255 – 1261 (1993). 43. M. Yamamoto, A. Markatos, J. G. Seedor, P. Masarachia, M. Gentile, G. A. Rodan, and R. Balena, The effects of the aminobisphosphonate alendronate on thyroid hormone-induced osteopenia in rats. Calcif. Tissue Int. 53, 278 – 282 (1993). 44. A. S. Shanbhag, C. T. Hasselman, and H. E. Rubash, Inhibition of wear debris mediated osteolysis in a canine total hip arthroplasty model. Clin. Orthop. 344, 33 – 43 (1997). 45. Y. Takano, T. Tanizawa, T. Mashiba, N. Endo, S. Nishida, and H. E. Takahashi, Maintaining bone mass by bisphosphonate incadronate disodium (YM175) sequential treatment after discontinuation of intermittent human parathyroid hormone (1 – 34) administration in ovariectomized rats. J. Bone Miner. Res. 11, 169 – 177 (1996). 46. S. J. Wimalawansa, A four-year randomized controlled trial of hormone replacement and bisphosphonate, alone or in combination, in women with postmenopausal osteoporosis. Am. J. Med. 104, 219 – 226 (1998). 47. R. Lindsay, F. Cosman, R. O. Lobo, B. W. Walsh, S. T. Harris, J. T. Reagan, C. L. Liss, M. E. Melton, and C. A. Byrnes, Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84, 3076 – 3081 (1999). 48. R. W. Boyce, C. L. Paddock, J. R. Gleason, W. K. Sietsema, and E. F. Eriksen, The effects of risedronate on canine cancellous bone remodeling: Three-dimensional kinetic reconstruction of the remodeling site. J. Bone Miner. Res. 10, 211 – 215 (1995). 49. P. M. Chavassieux, M. E. Arlot, C. Reda, L. Wei, A. J. Yates, and P. M. Meunier, Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. J. Clin. Invest. 100, 1475 – 1480 (1997). 50. P. J. Meunier and G. Boivin, Bone mineral density reflects bone mass but also the degree of mineralization of bone: Therapeutic implications. Bone 5, 373 – 377 (1997). 51. H. L. Guenther, H. E. Guenther, and H. Fleisch, The effects of 1-hydroxyethane-1, 1-diphosphonate and dichloromethanedisphosphonate on collagen synthesis by rabbit articular chondrocytes and rat bone cells. Biochem. J. 196, 293 – 301 (1981). 52. M. Tsuchimoto, Y. Azuma, O. Higuchi, I. Sugimoto, N. Hirata, M. Kiyoki, and I. Yamamoto, Alendronate modulates osteogenesis

464

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

H. FLEISCH of human osteoblastic cells in vitro. Jpn. J. Pharmacol. 66, 25 – 33 (1994). N. Giuliani, M. Pedrazzoni, G. Negri, G. Passeri, M. Impicciatore, and G. Girasole, Bisphosphonates stimulate formation of osteoblast precursors and mineralized nodules in murine and human bone marrow cultures in vitro and promote early osteoblastogenesis in young and aged mice in vivo. Bone 22, 455 – 461 (1998). J. A. Guy, M. Shea, C. P. Peter, R. Morrissey, and W. C. Hayes, Continuous alendronate treatment throughout growth, maturation, and aging in the rat results in increases in bone mass and mechanical properties. Calcif. Tissue Int. 53, 283 – 288 (1993). J. L. Ferretti, C. J. Delgado, R. F. Capozza, G. Cointry, E. Montuori, E. Roldán, A. Pérez-Lloret, and J. R. Zanchetta, Protective effects of disodium etidronate and pamidronate against the biomechanical repercussion of betamethasone-induced osteopenia in growing rat femurs. Bone Miner. 20, 265 – 276 (1993). H. Motoie, T. Nakamura, N. O’Uchi, H. Nishikawa, H. Kanoh, T. Abe, and H. Kawashima, Effects of the bisphosphonate YM175 on bone mineral density, strength, structure, and turnover in ovariectomized beagles on concomitant dietary calcium restriction. J. Bone Miner. Res. 10, 910 – 920 (1995). J. L. Ferretti, N. Mondelo, R. F. Capozza, G. R. Cointry, J. R. Zanchetta, and E. Montuori, Effects of large doses of olpadronate (dimethyl-pamidronate) on mineral density, cross-sectional architecture, and mechanical properties of rat femurs. Bone 16 (Suppl. 4), 285S – 293S 57 (1995). J. L. Ferretti, G. Cointry, R. Capozza, E. Montuori, E. Roldan, and A. Perez Lloret, Biomechanical effects of the full range of useful doses of (3-amino-1-hydroxypropylidene)-1,1-bisphosphonate (APD) on femur diaphyses and cortical bone tissue in rats. Bone Miner. 11, 111 – 122 (1990). P. Geusens, G. Nijs, G. Van der Perre, R. Van Audekercke, G. Lowet, S. Goovaerts, A. Barbier, F. Lacheretz, B. Remandet, Y. Jiang, and J. Dequeker, Longitudinal effect of tiludronate on bone mineral density, resonant frequency, and strength in monkeys. J. Bone Miner. Res. 7(6), 599 – 608 (1992). L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. Metab. Bone Dis. Relat. Res. 2, 389 – 407 (1980). J. L. Ferretti, Effects of bisphosphonates on bone biomechanics. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A. Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 211 – 229. Elsevier, Amsterdam, 1995. A. Jung, B. Mermillod, C. Barras, M. Baud, and B. Courvoisier, Inhibition by two diphosphonates of bone lysis in tumor-conditioned media. Cancer Res. 41, 3233 – 3237 (1981). R. R. Martodam, K. S. Thornton, D. A. Sica, S. M. D’Souza, L. Flora, and G. R. Mundy, The effects of dichloromethylene diphosphonate on hypercalcemia and other parameters of the humoral hypercalcemia of malignancy in the rat Leydig cell tumor. Calcif. Tissue Int. 35, 512 – 519 (1983). A. Jung, J. Bornand, B. Mermillod, C. Edouard, and P. J. Meunier, Inhibition by diphosphonates of bone resorption induced by the Walker tumor of the rat. Cancer Res. 44, 3007 – 3011 (1984). A. Sasaki, B. F. Boyce, B. Story, K. R. Wright, M. Chapman, R. Boyce, G. R. Mundy, and T. Yoneda, Bisphosphonate risedronate reduces metastatic human breast cancer burden in bone in nude mice. Cancer Res. 55, 3551 – 3557 (1995). H. Fleisch, Bisphosphonates: Introduction and mechanisms of action in tumor prevention. In “Tumor Bone Diseases and Osteoporosis in Cancer Patients. Pathophysiology, Diagnosis, and Therapy” (J.-J. Body, ed.), pp. 357 – 376. Dekker, New York, 2000. R. G. Russell and P. J. Croucher, Clinical pharmacology of bisphosphonates. In “Tumor Bone Diseases and Osteoporosis in Cancer

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

Patients: Pathophysiology, Diagnosis, and Therapy” (J.-J. Body, ed.), pp. 377 – 392. Dekker, New York, 2000. K. Shoji, H. Horiuchi, and H. Shinoda, Inhibitory effects of a bisphosphonate (risedronate) on experimental periodontitis in rats. J. Periodont. Res. 30, 27 – 84 (1995). K. Igarashi, H. Adachi, H. Mitani, and H. Shinoda, Inhibitory effect of the topical administration of a bisphosphonate (risedronate) on root resorption incident to orthodontic tooth movements in rats. J. Dent. Res. 75, 1644 – 1659 (1996). M. D. Francis, K. Hovancik, and R. W. Boyce, NE-58095: A diphosphonate which prevents bone erosion and preserves joint architecture in experimental arthritis. Int. J. Tissue React. 11, 239 – 252 (1989). P. H. Reitsma, O. L. M. Bijvoet, H. Verlinden-Ooms, and L. J. A. van der Wee-Pals, Kinetic studies of bone and mineral metabolism during treatment with (3-amino-1-hydroxypropylidene)-1,1-bisphosphonate (ADP) in rats. Calcif. Tissue Int. 32, 145 – 157 (1980). H. Shinoda, G. Adamek, R. Felix, H. Fleisch, R. Schenk, and P. Hagan, Structure-activity relationships of various bisphosphonates. Calcif. Tissue Int. 35, 87 – 99 (1983). P. M. Boonekamp, C. W. G. M. Löwik, L. J. A. van der Wee-Pals, M. L. L.van Wijk-van Lennep, and O. L. M. Bijvoet, Enhancement of the inhibitory action of APD on the transformation of osteoclast precursors into resorbing cells after dimethylation of the amino group. Bone Miner. 2, 29 – 42 (1987). R. C. Mühlbauer, F. Bauss, R. Schenk, M. Janner, E. Bosies, K. Strein, and H. Fleisch, BM 21.0955, a potent new bisphosphonate to inhibit bone resorption. J. Bone Miner. Res. 6, 1003 – 1011 (1991). W. K. Sietsema, F. H. Ebetino, A. M. Salvagno, and J. A. Bevan, Antiresorptive dose-response relationships across three generations of bisphosphonates. Drugs Exp. Clin. Res. XV, 389 – 396 (1989). M. Takeuchi, S. Sakamoto, M. Yoshida, T. Abe, and Y. Isomura, Studies on novel bone resorption inhibitors. I. Synthesis and pharmacological activities of aminomethylenebisphosphonate derivatives. Chem. Pharm. Bull. 41, 688 – 693 (1993). T. Storm, T. Steiniche, G. Thamsborg, and F. Melsen, Changes in bone histomorphometry after long-term treatment with intermittent, cyclic etidronate for postmenopausal osteoporosis. J. Bone Miner. Res. 8, 199 – 208 (1993). D. E. Hughes, B. R. MacDonald, R. G. G. Russell, and M. Gowen, Inhibition of osteoclast-like cell formation by bisphosphonates in long-term cultures of human bone marrow. J. Clin. Invest. 83, 1930 – 1935 (1989). P. M. Boonekamp, L. J. A. van der Wee-Pals, M. M. L. van Wijk-van Lennep, C. W. Thesing, and O. L. M. Bijvoet, Two modes of action of bisphosphonates on osteoclastic resorption of mineralized matrix. Bone Miner. 1, 27 – 39 (1986). P. M. Boonekamp, C. W. G. M. Löwik, L. J. A. van der Wee-Pals, M. L. L. van Wijk-van Lennep, and O. L. M. Bijvoet, Enhancement of the inhibitory action of APD on the transformation of osteoclast precursors into resorbing cells after dimethylation of the amino group. Bone Miner. 2, 29 – 42 (1987). C. W. G. M. Löwik, G. van der Pluijm, L. J. A. van der Wee-Pals, H. Bloys van Treslong-de Groot, and O. L. M. Bijvoet, Migration and phenotypic transformation of osteoclast precursors into mature osteoclasts: The effect of a bisphosphonate. J. Bone Miner. Res. 3, 185–192 (1988). S. Colucci, V. Minielli, G. Zambonin, N. Cirulli, G. Mori, M. Serra, V. Patella, A. Zambonin Zallone, and M. Grano, Alendronate reduces adhesion of human osteoclast-like cells to bone and bone proteincoated surfaces. Calcif. Tissue Int. 63, 230 – 235 (1998). G. van der Pluijm, H. Vloedraven, E. van Beek, L. van der Wee-Pals, C. Löwik, and S. Papapoulos, Bisphosphonates inhibit the adhesion of breast cancer cells to bone matrices in vitro. J. Clin. Invest. 98, 698 – 705 (1996).

465

CHAPTER 16 Basic Biology of Bisphosphonates 84. S. Boissier, S. Magnetto, L. Frappart, B. Cuzin, F. H. Ebetino, P. D. Delmas, and P. Clezardin, Bisphosphonates inhibit prostate and breast carcinoma cell adhesion to unmineralized and mineralized bone extracellular matrices. Cancer Res. 57, 3890 – 3894 (1997). 85. D. E. Hughes, K. R. Wright, H. L. Uy, A. Sasaki, T. Yoneda, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphonates promote apoptosis in murine osteoclasts in vitro and in vivo. J. Bone Miner. Res. 10, 1478 – 1487 (1995). 86. K. Selander, P. Lehenkari, and H. K. Väänänen, The effects of bisphosphonates on the resorption cycle of isolated osteoclasts. Calcif. Tissue Int. 55, 368 – 375 (1994). 87. H. Murakami, N. Takahashi, T. Sasaki, N. Udagawa, S. Tanaka, I. Nakamura, D. Zhang, A. Barbier, and T. Suda, A possible mechanism of the specific action of bisphosphonates on osteoclasts: Tiludronate preferentially affects polarized osteoclasts having ruffled borders. Bone 17, 137 – 144 (1995). 88. S. C. Miller and W. S. S. Jee, The effect of dichloromethylenediphosphonate, a pyrophosphate analog, on bone and bone cell structure in the growing rat. Anat. Rec. 193, 439 – 462 (1979). 89. C. M. T. Plasmans, P. H. K. Jap, W. Kuijpers, and T. J. J. H. Slooff, Influence of a diphosphonate on the cellular aspect of young bone tissue. Calcif. Tissue Int. 32, 247 – 256 (1980). 90. M. Sato, W. Grasser, N. Endo, R. Akins, H. Simmons, D. D. Thompson, E. Golub, and G. A. Rodan, Bisphosponate action. Alendronate localization in rat bone and effects on osteoclast ultrastructure. J. Clin. Invest. 88, 2095 – 2105 (1991). 91. E. van Beek, E. Pieterman, L. Cohen, C. Löwik, and S. Papapoulos, Nitrogen-containing bisphosphonates inhibit isopentenyl pyrophosphate isomerase/farnesyl pyrophosphate synthase activity with relative potencies corresponding to their antiresorptive potencies in vitro and in vivo. Biochem. Biophys. Res. Commun. 255, 491 – 494 (1999). 92. E. van Beek, E. Pieterman, L. Cohen, C. Löwik, and S. Papapoulos, Farnesyl pyrophosphate synthase is the molecular target of nitrogencontaining bisphosphonates. Biochem. Biophys. Res. Commun. 264, 108 – 111 (1999). 93. J. D. Bergstrom, R. G. Bostedor, P. J. Masarachia, A. A. Reszka, and G. Rodan, Alendronate is a specific, nanomolar inhibitor of farnesyl diphosphate synthase. Arch. Biochem. Biophys. 373, 231 – 241 (2000). 94. D. B. Morgan, A. Monod, R. G. G. Russell, and H. Fleisch, Influence of dichloromethylene diphosphonate (Cl2MDP) and calcitonin on bone resorption, lactate production and phosphatase and pyrophosphatase content of mouse calvaria treated with parathyroid hormone in vitro. Calcif. Tissue Res. 13, 287 – 294 (1973). 95. D. K. Fast, R. Felix, C. Dowse, W. F. Neuman, and H. Fleisch, The effects of diphosphonates on the growth and glycolysis of connective-tissue cells in culture. Biochem. J. 172, 97 – 107 (1978). 96. Z. Zimolo, G. Wesolowski, and G. A. Rodan, Acid extrusion is induced by osteoclast attachment to bone: Inhibition by alendronate and calcitonin. J. Clin. Invest. 96, 2277 – 2283 (1995). 97. P. David, H. Nguyen, A. Barbier, and R. Baron, The bisphosphonate tiludronate is a potent inhibitor of the osteoclast vacuolar H-ATPase. J. Bone Miner. Res. 11, 1498 – 1507 (1996). 98. R. Felix, R. G. G. Russell, and H. Fleisch, The effect of several diphosphonates on acid phosphohydrolases and other lysosomal enzymes. Biochim. Biophys. Acta 429, 429 – 438 (1976). 99. R. Felix, J. D. Bettex, and H. Fleisch, Effect of diphosphonates on the synthesis of prostaglandins in cultured calvaria cells. Calcif. Tissue Int. 33, 549 – 552 (1981). 100. K. Ohya, S. Yamada, R. Felix, and H. Fleisch, Effect of bisphosphonates on prostaglandin synthesis by rat bone cells and mouse calvaria in culture. Clin. Sci. 69, 403 – 411 (1985). 101. O. Teronen, Y. T. Konttinen, C. Lindqvist, T. Salo, T. Ingman,

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

H. Lauhio, S. Santavirta, H. Valleala, and T. Sorsa, Inhibition of matrix matalloproteinase-1 by dichloromethylene bisphosphonate (clodronate). Calcif. Tissue Int. 61, 59 – 61 (1997). O. Teronen, P. Heikkila, Y. T. Konttinen, M. Laitinen, T. Salo, R. Haanemaaijer, A. Teronen, P. Maisi, and T. Sorsa, MMP inhibition and downregulation by bisphosphonates. Ann. N. Y. Acad. Sci. 878, 453 – 465 (1999). A. Schmidt, S. J. Rutledge, N. Endo, E. E. Opas, H. Tanaka, G. Wesolowski, C. T. Leu, Z. Huang, C. Ramachandaran, S. B. Rodan, and G. A. Rodan, Protein-tyrosine phosphatase activity regulates osteoclast formation and function: Inhibition by alendronate. Proc. Natl. Acad. Sci. USA 93, 3068 – 3073 (1996). S. P. Luckman, F. P. Coxon, F. H. Ebetino, R. G. G. Russell, and M. J. Rogers, Heterocycle-containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: Evidence from structure-activity relationships in J774 macrophages. J. Bone Miner. Res. 13, 1668 – 1678 (1998). S. P. Luckman, D. E. Hughes, F. P. Coxon, R. G. G. Russell, and M. J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including Ras. J. Bone Miner. Res. 13, 581 – 589 (1998). J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J. Masarachia, G. Wesolowski, R. G. G. Russell, G. A. Rodan, and A. A. Reszka, Alendronate mechanism of action: Gernylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc. Natl. Acad. Sci. USA 96, 133 – 138 (1999). M. J. Rogers, R. G. Brown, V. Hodkin, R. G. G. Russell, and D. J. Watts, Bisphosphonates are incorporated into adenine nucleotides by human aminoacyl-tRNA synthetase enzymes. Biochem. Biophys. Res. Commun. 224, 863 – 869 (1996). S. Auriola, J. Frith, M. J. Rogers, A. Koivuniemi, and J. Mönkkönen, Identification of adenine nucleotide-containing metabolites of bisphosphonate drugs using ion-pair liquid chromatographyelectrospray mass spectrometry. J. Chromatogr. B 704, 187 – 195 (1997). J. C. Frith, J. Mönkkönen, G. M. Blackburn, R. G. G. Russell, and M. J. Rogers, Clodronate and liposome-encapsulated clodronate are metabolised to a toxic ATP analog, adenosine5 ,-dichloromethylene) triphosphate, by mammalian cells in vitro. J. Bone Miner. Res. 12, 1358 – 1367 (1997). R. Felix, H. L. Guenther, and H. Fleisch, The subcellular distribution of [14C] dichloromethylenebisphosphonate and [14C]1-hydroxyethylidene-1,1-bisphosphonate in cultured calvaria cells. Calcif. Tissue Int. 36, 108 – 113 (1984). M. G. Cecchini, R. Felix, H. Fleisch, and P. H. Cooper, Effects of bisphosphonates on proliferation and viability of mouse bone marrow-derived macrophages. J. Bone Miner. Res. 2, 135 – 142 (1987). M. Sahni, H. L. Guenther, H. Fleisch, P. Collin, and T. J. Martin, Bisphosphonates act on rat bone resorption through the mediation of osteoblasts. J. Clin. Invest. 91, 2004 – 2011 (1993). C. Vitté, H. Fleisch, and H. L. Guenther, Bisphosphonates induce osteoblasts to secrete an inhibitor of osteoclast-mediated resorption. Endocrinology 137, 2324 – 2333 (1996). M. Nishikawa, T. Akatsu, Y. Katayama, Y. Yasutomo, S. Kado, N. Kugai, M. Yamamoto, and N. Nagata, Bisphosphonates act on osteoblastic cells and inhibit osteoclast formation in mouse marrow cultures. Bone 18, 9 – 14 (1996). L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Roberson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999).

466 116. S. Adami, A. K. Bhalla, R. Dorizzi, F. Montesanti, S. Rosini, G. Salvagno, and V. Lo Cascio, The acute-phase response after bisphosphonate administration. Calcif. Tissue Int. 41, 326 – 331 (1987). 117. H. Fleisch, R. G. G. Russell, S. Bisaz, R. C. Mühlbauer, and D. A. Williams, The inhibitory effect of phosphonates on the formation of calcium phosphate crystals in vitro and on aortic and kidney calcification in vivo. Eur. J. Clin. Invest. 1, 12 – 18 (1970). 118. M. D. Francis, The inhibition of calcium hydroxyapatite crystal growth by polyphosphonates and polyphosphates. Calcif. Tissue Res. 3, 151 – 162 (1969). 119. A. Jung, S. Bisaz, and H. Fleisch, The binding of pyrophosphate and two diphosphonates by hydroxyapatite crystals. Calcif. Tissue Res. 11, 269 – 280 (1973). 120. B. L. Barnett and L. C. Strickland, Structure of disodium dihydrogen 1-hydroxyethylidene-diphosphonate tetrahydrate: A bone growth regulator. Acta Cryst. B35, 1212 – 1214 (1979). 121. D. Fraser, R. G. G. Russell, O. Pohler, W. G. Robertson, and H. Fleisch, The influence of disodium ethane-1-hydroxy-1,1-diphosphonate (EHDP) on the development of experimentally induced urinary stones in rats. Clin. Sci. 42, 197 – 207 (1972). 122. C. M. T. Plasmans, W. Kuypers, and T. J. J. H. Slooff, The effect of ethane-1-hydroxy-1,1-diphosphonic acid (EHDP) on matrix induced ectopic bone formation. Clin. Orthop. Rel. Res. 132, 233 – 243 (1978). 123. H. R. Mühlemann, D. Bowles, A. Schatt, and J. P. Bernimoulin, Effect of diphosphonate on human supragingival calculus. Helv. Odont. Acta 14, 31 – 33 (1970). 124. A. Larsson, The short-term effects of high doses of ethylene-1hydroxy-1,1-diphosphonates upon early dentin formation. Calcif. Tissue Res. 16, 109 – 127 (1974). 125. Y. Ogawa, Disturbances in enamel and dentin formation of rat incisor following injection with sodium fluoride, strontium(II) chloride and EHDP. Jpn. J. Oral Biol. 22, 199 – 226 (1980). 126. E. van Beek, M. Hoekstra, M. van de Ruit, C. Löwik, and S. Papapoulos, Structural requirements for bisphosphonate actions in vitro. J. Bone Miner. Res. 9, 1875 – 1882 (1994). 127. E. van Beek, C. Löwik, I. Oue, and S. Papapoulos, Dissociation of binding and antiresorptive properties of hydroxybisphosphonates by substitution of the hydroxyl with an amino group. J. Bone Miner. Res. 11, 1492 – 1499 (1996). 128. M. D. Francis, L. Flora, and W. R. King, The effects of disodium ethane-1-hydroxy-1,1-diphosphonate on adjuvant induced arthritis in rats. Calcif. Tissue Res. 9, 109 – 121 (1972). 129. C. J. Dunn, L. A. Galinet, H. Wu, R. A. Nugent, S. T. Schlachter, N. D. Staite, D. G. Aspar, G. A. Elliott, N. A. Essani, N. A. Rohloff, and R. J. Smith, Demonstration of novel anti-arthritic and antiinflammatory effects of diphosphonates. J. Pharmacol. Exp. Ther. 266, 1691 – 1698 (1993). 130. P. L. E. M. van Lent, L. A. M. van den Bersselaar, A. E. M. Holthuyzen, N. van, Rooijen, L. B. A. van de Putte, and W. B. van den Berg, Phagocytic synovial lining cells in experimentally induced chronic arthritis: Down-regulation of synovitis by CL2MDPliposomes. Rheumatol. Int. 13, 221 – 228 (1994). 131. T. Österman, K. Kippo, L. Laurén, R. Hannuniemi, and R. Sellman, Effect of clodronate on established adjuvant arthritis. Rheumatol. Int. 14, 139 – 147 (1994). 132. C. P. Ciosek, D. R. Magnin, T. W. Harrity, J. V. H. Logan, J. K. Dickson, E. M. Gordon, K. A. Hamilton, K. G. Jolibois, L. K. Kunselman, R. M. Lawrence, K. A. Mookhtiar, L. C. Rich, D. A. Slusarchyk, R. B. Sulsky, and S. A. Biller, Lipophilic 1,1-bisphosphonates are potent squalene synthase inhibitors and orally active cholesterol lowering agents in vivo. J. Biol. Chem. 268, 24832 – 24837 (1993). 133. J. H. Lin, Bisphosphonates: A review of their pharmacokinetic properties. Bone 18, 75 – 85 (1996).

H. FLEISCH 134. S. E. Papapoulos, Pharmacodynamics of bisphosphonates in man; implications for treatment. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A. Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 231 – 263. Elsevier, Amsterdam, 1995. 135. A. G. Porras, S. D. Holland, and B. J. Gertz, Pharmacokinetics of alendronate. Clin. Pharmacokinet. 36, 315 – 328 (1999). 136. R. R. Recker and P. D. Saville, Intestinal absorption of disodium ethane-1-hydroxy-1,1-diphosphonate (disodium etidronate) using a deconvolution technique. Toxicol. Appl. Pharmacol. 24, 580 – 589 (1973). 137. X. Boulenc, E. Marti, H. Joyeux, C. Roques, Y. Berger, and G. Fabre, Importance of the paracellular pathway for the transport of new bisphosphonate using the human CACO-2 monolayers model. Biochem. Pharmacol. 46(9), 1591 – 1600 (1993). 138. B. J. Gertz, S. D. Holland, W. F. Kline, B. K. Matuszewski, A. Freeman, H. Quan, K. C. Lasseter, J. C. Mucklow, and A. G. Porras, Studies of the oral bioavailability of alendronate. Clin. Pharmacol. Ther. 58, 288 – 298 (1995). 139. W. H. Wiedmer, A. M. Zbinden, U. Trechsel, and H. Fleisch, Ultrafiltrability and chromatographic properties of pyrophosphate, 1-hydroxyethylidene-1,1 bisphosphonate, and dichloromethylenebisphosphonate in aqueous buffers and in human plasma. Calcif. Tissue Int. 35, 397 – 400 (1983). 140. J. H. Lin, I.-W. Chen, and F. A. De Luna, Nonlinear kinetics of alendronate: Plasma protein binding and bone uptake. Drug Metab. Dispos. 22, 400 – 405 (1994). 141. F. Wingen, and D. Schmähl, Distribution of 3-amino-1-hydroxypropane-1,1-diphosphonic acid in rats and effects on rat osteosarcoma. Arzneimittelforschung 35, 1565 – 1571 (1985). 142. S. Bisaz, A. Jung, and H. Fleisch, Uptake by bone of pyrophosphate, diphosphonates and their technetium derivatives. Clin. Sci. Mol. Med. 54, 265 – 272 (1978). 143. P. Masarachia, M. Weinreb, and G. A. Rodan, Comparison of the distribution of 3H-alendronate and 3H-etidronate in rat and mouse bones. Bone 19, 281 – 290 (1996). 144. J. Mönkkönen, H. M. Koponen, and P. Ylitalo, Comparison of the distribution of three bisphosphonates in mice. Pharmacol. Toxicol. 66, 294 – 298 (1990). 145. G. B. Kasting and M. D. Francis, Retention of etidronate in human, dog and rat. J. Bone Miner. Res. 7, 513 – 522 (1992). 146. U. Troehler, J. P. Bonjour, and H. Fleisch, Renal secretion of diphosphonates in rats. Kidney Int. 8, 6 – 13 (1975). 147. J. H. Lin, I.-W. Chen, A. Florencia, A. Deluna, and M. Hichens, Renal handling of alendronate in rats: An uncharacterized renal transport system. Drug Metab. Dispos. 20, 608 – 613 (1992). 148. C. L. Alden, R. D. Parker, and D. F. Eastman, Development of an acute model for the study of chloromethanediphosphonate nephrotoxicity. Toxicol. Pathol. 17, 27 – 32 (1989). 149. J. C. Cal and P. T. Daley-Yates, Disposition and nephrotoxicity of 3-amino-1-hydroxypropylidene-1,1-bisphosphonate (APD) in rats and mice. Toxicology 65, 179 – 197 (1990). 150. P. C. De Groen, D. F. Lubbe, L. J. Hirsch, A. Daifotis, W. Stephenson, D. Freedholm, S. Pryor-Tillotson, M. J. Seleznick, H. Pinkas, and K. K. Wang, Esophagitis associated with the use of alendronate. N. Engl. J. Med. 355, 1016 – 1021 (1996). 151. C. P. Peter, K. K. Handt, and S. M. Smith, Esophageal irritation due to alendronate sodium tablets. Digest. Dis. Sci. 43, 1998 – 2002 (1998). 152. C. P. Peter, M. V. Kindt, and J. A. Majka, Comparative study of potential for bisphosphonates to damage gastric mucosa of rats. Digest. Dis. Sci. 43, 1009 – 1015 (1998). 153. W. R. King, M. D. Francis, and W. R. Michael, Effect of disodium ethane-1-hydroxy-1,1-diphosphonate on bone formation. Clin. Orthop. 78, 251 – 270 (1971).

CHAPTER 16 Basic Biology of Bisphosphonates 154. M. Eguchi, T. Yamaguchi, E. Shiota, and S. Handa, Fault of ossification and calcification and angular deformities of long bones in the mouse fetuses caused by high doses of ethane-1-hydroxy-1, 1-diphosphonate (EHDP) during pregnancy. Cong. Anom. 22, 47 – 52 (1982). 155. P. Graepel, P. Bentley, H. Fritz, M. Miyamoto, and S. R. Slater, Reproduction toxicity studies with pamidronate. Arzneim.-Forsch. Drug Res. 42, 654 – 667 (1992). 156. A. Okazaki, T. Matsuzawa, M. Takeda, R. G. York, P. C. Barrow, V. C. King, and G. P. Bailey, Intravenous reproductive and develop-

467 mental toxicity studies of cimadronate (YM175), a novel bisphosphonate, in rats and rabbits. J. Toxicol. Sci. 20, 1 – 13 (1995). 157. A. Okazaki, H. Sakai, T. Matsuzawa, C. J. Perkin, and P. W. East, Intravenous single and repeated dose toxicity studies of cimadronate (YM175), a novel bisphosphonate, in rats. J. Toxicol. Sci. 20, 15 – 26 (1995). 158. A. Okazaki, T. Matsuzawa, C. J. Perkin, and M. H. Barker, Intravenous single and repeated dose toxicity studies of cimadronate (YM175), a novel bisphosphonate, in beagle dogs. J. Toxicol. Sci. 20, 27 – 36 (1995).

CHAPTER 17

Skeletal Development Mechanical Consequences of Growth, Aging, and Disease MARJOLEIN C. H. VAN DER MEULEN

DENNIS R. CARTER AND GARY S. BEAUPRÉ

Sibley School of Mechanical and Aerospace Engineering, Cornell University, Ithaca, New York 14853; and Biomechanics and Biomaterials, Hospital for Special Surgery, New York, New York 10021 Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304

I. Developmental Mechanics in Skeletogenesis II. Mechanical Regulation of Bone Biology III. Mechanobiologic Self-Design of Bones

IV. Adaptational Mechanics in Aging and Disease References

I. DEVELOPMENTAL MECHANICS IN SKELETOGENESIS

the normal adult skeleton is “inherently” well designed from a mechanical point of view. Changes in bone tissue quality and/or quantity as a result of aging or disease then diminish the normal mechanical integrity of the skeleton, thereby increasing the risk of fracture. Mechanical deficits in the osteoporotic skeleton are thus often assessed with respect to a “normal” young or age-matched control group. The approach taken in this chapter is a bit different. We believe that the mechanics of the osteoporotic skeleton can best be understood when one appreciates the role of mechanics in skeletal development. We consider the mechanical integrity of the skeleton at any age to be a reflection of intrinsic genetic factors and the entire prior life history of mechanical and chemical epigenetic events. These events include numerous factors that are related to hormones, diet, and physical activity. Biomechanical

The bones of the adult skeleton are well designed for supporting the forces that are created during normal physical activities. The tubular shape of the diaphyses of long bones is ideal for withstanding the bending and torsional loads imposed on the bone shaft. Bone tissue at the ends of long bones and in short bones serves to support and distribute joint contact forces. The intricate architecture of the cancellous bone in these regions is well suited for this task. Indeed, the form and internal architecture of the entire skeleton are exquisitely matched to its mechanical function. Biomechanical considerations of the osteoporotic skeleton usually concentrate on the material and structural changes that compromise its mechanical integrity and thus can lead to bone fracture. Such an approach assumes that

OSTEOPOROSIS, SECOND EDITION VOLUME 1

471

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

472 changes in osteoporosis can then be directly linked to the physicochemical events that precede the disease and continue to affect the skeleton during the progression and treatment of the disease. Mechanical regulation of bone biology begins very early. At approximately 5 to 7 weeks of prenatal life, most of the skeletal elements, muscles, tendons, and ligaments characteristic of the adult have formed. Involuntary contractions of the newly formed muscle fibers commence and ossification is initiated in the cartilaginous endoskeleton. The intermittently imposed skeletal tissue stresses, deformations, and motions caused by muscular contractions then play an increasingly important role in modulating cartilage growth, ossification, and bone modeling and remodeling throughout the skeleton except for the cranium. By 15 weeks, all of the basic movements characteristic of full-term newborn infants can be observed [1]. The inhibition of muscular contractions and movements in utero results in abnormally low skeletal mass and strength (Fig. 1) [2 – 4]. After birth, further growth and ossification of the skeleton continues to be strongly influenced by physical activity and externally applied forces [5]. Long bones such as the femur begin to ossify when the primary bone collar appears at the midshaft of the cartilage anlage where the chondrocytes have reached a hypertrophic

VAN DER

state. Some have speculated that the hypertrophy of the chondrocytes at the midshaft may result in the release of chemical factors that induce osteogenesis in the perichondrium. Vascular invasion of the hypertrophic cartilage inside the primary bone collar results in a transient stage of endochondral bone formation followed by osteoclastic resorption. The medullary canal and endosteal surface are established. The entire anlage continues to grow in length by the proliferation and ossification of cartilage. Further growth and development of the diaphyseal cross section are achieved by direct bone apposition and resorption on the periosteal and endosteal surfaces, respectively. When one examines the structure of a typical long bone, the bone in different regions is associated with different ontogenetic processes (Fig. 2). A significant portion of the compact bone in the diaphysis has a developmental history that includes initial appositional bone formation. Extending from the center of the bone toward both of the bone ends, however, are conical regions that include the cancellous bone of the metaphyses and epiphyses. The bone in these areas was initially formed by endochondral ossification. With increasing age, secondary bone remodeling throughout the skeleton will progressively diminish the distinctions associated with the primary bone formed in the different regions. In considering the influence of mechanical factors on bone development, appositional bone

FIGURE 2 FIGURE 1 Radiographs of tibiae from (a) normal infant and (b) newborn with spinal muscular atrophy. Used with permission from Rodriguez et al., Calcif. Tissue Int. 43, 335 – 339 (1988).

MEULEN, CARTER, AND BEAUPRÉ

Schematic representation of long bone growth. The shaded regions on the right-hand figure are areas of appositional growth. Metaphyseal and epiphyseal cancellous bone is formed by endochondral ossification, shown in the white conical regions and bone ends. Used with permission from Carter et al., Bone 18 (Suppl. 1), 5S – 10S (1996).

473

CHAPTER 17 Skeletal Development

formation must be considered as a different process from endochondral bone formation. The subsequent bone modeling and remodeling in these two regions, however, are similar.

II. MECHANICAL REGULATION OF BONE BIOLOGY Wolff [7] wrote extensively on the relationship between physical forces and bone structure. He was influenced by the work of Wilhelm Roux [8], who was interested not only in the morphology of tissues and organ systems but also the mechanisms responsible for the development of specific morphological features. Roux was convinced that physical forces play a major role in development. He referred to the processes by which physicochemical factors regulate development as “Entwicklungsmechanik” or “developmental mechanics” [8]. Wolff stated, “Roux, as I do myself, distinguishes two periods in the life of every organism. One is embryonic. During this period the ‘organs expand, differentiate and grow.’ The other period is adulthood. During this period, growth and replacement of what is worn out takes place ‘only when stimulated.’ . . . ” (translation by Maquet and Furlong, 1986). In the writings of Roux and Wolff we see the seeds of a fundamental question regarding the relative importance of biological and mechanical regulation of skeletogenesis. Building on these concepts first introduced a century ago, we have developed a theory for bone adaptation in which biologic factors play an important role only in the initial phases of skeletal development and their influence diminishes over time [9,10]. Mechanobiologic influences, however, remain a fundamental influence on bone apposition and resorption throughout life. In this theory we describe the intensity of bone tissue mechanical loading in terms of a daily stress (or strain)

FIGURE 3

stimulus, b, that takes into account both the magnitude and the number of cycles of loading applied during daily activities [11]. For example, we might consider the daily stress stimulus for a nonathletic individual to consist of contributions from walking, stair climbing, and rising from a chair. For an individual who is athletic, we might include additional contributions from jogging, bicycling, running, and so on. In mathematical terms we define the daily stress stimulus as

b 

n   i b

m

1/m

,

(1)

day

where ni is the number of cycles of each load type i, b is a measure of stress intensity within the bone tissue, and the stress exponent, m, is an empirical constant. The stress exponent can be thought of as a weighting factor for the relative importance of the stress magnitude and the number of load cycles. For m  1, the stress magnitude and the number of load cycles are equally important. For m  1, those activities having high stress magnitudes would contribute more to the total stimulus. Alternatively, for m  1, those activities that are repeated many times each day would be relatively more important. Experimental data suggest that m lies in the range of 3 to 8, indicating that the magnitude of the cyclic stress is more important than the number of loading cycles [12]. We assume that if the imposed daily stress stimulus is greater than some target level of stimulus, the attractor stress stimulus (AS), then bone apposition will occur. If the imposed daily stress stimulus is less than the attractor stimulus, bone will be resorbed. The stress stimulus error, e, is the difference between the daily stress stimulus and the attractor stress stimulus, and is the driving force for bone adaptation. The block diagram shown in Fig. 3 is a schematic representation of our bone development and adaptation theory with the linear apposition/resorption rate, r. used as the feedback

Block diagram representation of bone remodeling having multiple feedback loops. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).

474

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

FIGURE 4

Simplified block diagram representation assuming that local nonstress effects (see Fig. 3) do not occur and that the bone attractor stress stimulus does not change appreciably due to formation or resorption. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).

parameter. In this particular block diagram two feedback loops are shown. The upper feedback loop implies that bone adaptation and factors related to metabolic status, genotype, and local tissue interactions may influence the attractor stress stimulus, as well as the local tissue response. The lower feedback loop illustrates the interaction between bone adaptation and purely mechanical factors. By assuming that neither the attractor state stimulus nor the local tissue response changes during the course of bone adaptation, the upper loop in Fig. 3 can be eliminated (Fig. 4). In this representation we have divided the lower, mechanical feedback loop into two parallel paths: one path corresponding to changes in geometry (typically cortical changes) and the other path corresponding to changes in apparent bone density (typically cancellous changes). To represent the time-dependent nature of bone adaptation, we must establish a quantitative relationship between the stress stimulus error and the rates of bone apposition and resorption. We believe, as others do, that the rate relationship that describes the bone response to a given remodeling error is nonlinear [13 – 15]. Specifically, we think that when the stress stimulus error is within a range near zero (i.e., within the normal activity range) the rate of net bone apposition or resorption will be small. When the remodeling error is outside this range, however, the rates of bone apposition and resorption can increase dramatically. Three hypothetical rate relationships are shown in Fig. 5. The curve labeled “1” might represent the skull, which is shown having a lower attractor state stress and a relative insensitivity to unloading and bone resorption. The curve

labeled “2” might represent the periosteum of the femur, with a low sensitivity for bone resorption with unloading and a high sensitivity for bone apposition with increased loading. Finally, the curve labeled “3” might represent the endosteum of the femur, with a higher sensitivity for resorption than for apposition. These differences among the three curves could be related to local tissue interactions and cell populations associated with the different bone surfaces in question. In the following sections, this approach for describing the regulation of bone biology is presented in more detail and implemented for particular skeletal elements.

FIGURE 5 Hypothetical curves for three bone regions showing the rate of surface response as a function of the tissue level stress stimulus. See text for further discussion. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).

475

CHAPTER 17 Skeletal Development

III. MECHANOBIOLOGIC SELF-DESIGN OF BONES Development and subsequent growth involve the coordinated change of shape, size, and material. The conceptual approach presented in Section II directly relates the in vivo mechanical loading environment to these changes during the development, growth, and adaptation of the skeleton. Using mathematical implementations on computers, we can model the various skeletal elements that are influenced by mechanics to simulate bone appositional and endochondral growth. The following sections first show how mechanics guides modeling changes in the long bone diaphysis during growth and development. Thereafter, a parallel development is demonstrated for the density and morphology changes in trabecular bone sites. The same fundamental relationships are used in both cases. Osteoporotic changes caused by alterations in bone loading are also simulated with the same models used for development.

A. Diaphyseal Compact Bone 1. DEVELOPMENT AND ADAPTATION In mammals the ossification of the long bones initiates with endochondral ossification and the formation of the primary bone collar at the middiaphysis of the cartilage anlage. At this time during development, fetal muscle contractions also commence [1]. Thereafter, the primary ossification center forms and the processes of endochondral ossification and direct bone apposition begin. Whereas the initial bone collar appears without the stimulation of fetal muscle contractions, further normal development and growth of the diaphysis are dependent on the mechanical loading environment created by these forces [2 – 4,16 – 18]. Therefore, the radial growth of the diaphysis may be considered as a combination of intrinsic biologic growth and mechanically regulated biologic (“mechanobiologic”) processes. Using this fundamental concept, we have developed an analytical model to simulate the roles of biological and mechanobiological factors in the development of the crosssectional geometry of the human femur [10]. We modeled the long bone diaphysis as a circular cross section defined by a periosteal (outer) and an endosteal (inner) radius. These sections were “grown” using bone surface apposition rates determined from underlying biologic growth and mechanically regulated biologic stimuli. Both growth processes were assumed to be functions of time and the particular bone surface. This model was used to simulate development under normal and altered loading conditions as well as adaptation to increased and decreased loading in the adult. To model intrinsic biological processes, we assumed that purely biological factors play a significant role in early

cross-sectional development and that their contribution gradually diminishes with time, becoming negligible in the later half of maturation. Biological factors were modeled as a periosteal surface apposition rate, which was a decaying exponential function of age. In our model of human femoral development, the biological growth rate decayed to zero by 6 years of age (Fig. 6). Mechanically regulated surface bone growth rates were calculated for the periosteum and endosteum based on the daily stress stimulus,  [Section II, Eq. (1)], and added directly to the biological rate. Whereas the contribution of purely biological growth processes diminishes with time, extrinsic mechanical influences on long bone cross-sectional growth are fundamental processes that remain active throughout an individual’s lifetime. In this model, the mechanobiologic stimulus is the sole regulator of long bone cross-sectional geometry once the biologic contribution has disappeared. Using the daily stress stimulus to describe the cyclic in vivo load history, the stress stimulus attractor was chosen consistent with experimental data from the literature [19]. In vivo experimental studies using strain gages bonded to adult bone surfaces have shown that the magnitude of bone strains created during physical activity is similar across different bones in a variety of animals over several orders of body mass [20]. Similar constant peak strain levels have also been measured in growing animals [21 – 23]. In our model, therefore, the stress stimulus attractor was assumed to be identical at all diaphyseal locations and constant throughout life. The surface modeling rate – stress stimulus relationship (see Fig. 5) used to model the mechanobiological responses included a “lazy zone” in the region near the stress stimulus attractor. The relationship between the bone stress stimulus and the surface apposition rates was modeled differently on the periosteum and endosteum, reflecting different tissue

FIGURE 6

Periosteal apposition rate used to simulate intrinsic biologic growth. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).

476 properties and interactions on the two surfaces [24 – 27]. The bone apposition rates were implemented identically for the periosteum and endosteum; there was no resorptive response allowed on the periosteum. A symmetric stimulus – rate relationship was modeled on the endosteum. The width of the lazy zone was chosen to be 20% of the attractor stimulus value. This value was based on the bone’s mature thicknessto-radius ratio. Values of the apposition rates were based on experimental measurements of diaphyseal changes with aging [28]. To implement this model, one must apply an assumed mechanical loading that changes as a function of age. In vivo strain gage studies have shown that the surface stresses in the weight-bearing long bones are primarily longitudinal normal and shear stresses from combined bending and torsional moments 1 [22,29,30]. These in vivo moments are produced by actions of the muscles on the skeleton and were assumed to scale in proportion to muscle mass [10]. Muscle mass is approximately proportional to body mass in adult mammals [31]. Assuming that torsional moments are directly proportional to body mass, we obtained human body mass data during growth and aging and used this to simulate the age-dependent in vivo loading history (Fig. 7). For modeling simplicity, we applied an axisymmetric torsional moment as a representative loading history. The stress distribution determines the cross-sectional morphology; torsional and bending moments both produce stresses that increase linearly from the inner surface to the outer surface in a cylindrical structure with constant material properties. The magnitude of the moments at maturity (age 20 years) was calculated as that which produced a bone stress stimulus at the corner of the lazy zone. A strength of materials analysis for a hollow circular cylinder was used to determine the stress stimulus on the periosteum and endosteum once the load was applied to the bone cross section. A normal developmental loading history was applied to the model, and cross-sectional morphologies were developed with time. For each simulation we calculated the following parameters: the periosteal and endosteal radii, surface apposition/resorption rates and stress stimuli, cortical area, polar moment of inertia, section modulus, and ratio of bone thickness to periosteal radius. The simulation values were compared with experimental measurements of human femoral cross-sectional morphology. Starting from an initial bone collar, the biological growth rate alone (without any mechanical sensitivity) produced a cross section completely dependent on the magnitude of the intrinsic growth rate. These results are shown in Fig. 8a. As implemented, only periosteal apposition was present and periosteal expansion ceased after 6 years. When both biologic

VAN DER

FIGURE 7

Torsional moment applied to simulate mechanobiological growth. The corresponding body mass at each time point is indicated on the right-hand scale. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).

and mechanobiologic responses were implemented (Fig. 8b), a rapid expansion of both the periosteum and the endosteum occurred during development and subsequent growth. The dimensional increases stabilized at maturity, and thereafter a gradual age-related expansion and thinning of the cortex occurred throughout the remainder of life. Comparison of the simulation radius values with those measured by other researchers shows a very good correspondence (Fig. 9) [32 – 34]. Although the loading is constant throughout maturity, gradual, age-related periosteal expansion occurs and results in a diminution of the cortical area and an increase in the polar moment of inertia. Numerous studies have measured this continuing subperiosteal expansion with age [28,34 – 37] and showed similar area and moment of inertia results. When a moment is applied to a beam, the resulting surface stresses are proportional to the applied moment divided by the section modulus, a cross-sectional shape parameter. For a given cross-sectional geometry, the section modulus is defined as the polar moment of inertia divided

FIGURE 8 1 A “moment” results from a force applied at a distance from the point of interest.

MEULEN, CARTER, AND BEAUPRÉ

Simulation results for normal cross-sectional growth of the human femur. (a) Biologic growth only and (b) biologic and mechanobiologic growth. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).

477

CHAPTER 17 Skeletal Development

FIGURE 9 Simulation results for periosteal and endosteal radii over time. Experimental data shown from McCammon ([33] mean data), Smith and Walker ([34] mean data), and Martin and Atkinson ([32] individual data points). Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993). by the periosteal diameter. If the moments increase with increasing body mass, as we have assumed, then the section modulus must also increase in the same fashion if the stresses are to remain constant. Our mechanobiological model reflects this adaptation, and the section modulus increases nearly linearly with body mass during development while the relationship with age is more complex (Fig. 10). Clinical measurements made in our laboratory [38 – 41] reveal a poor linear relationship between section modulus and age in adolescents and young adults (Fig. 11). When the same data are plotted against body mass, however, a very strong linear relationship is evident (Fig. 12). Further implications of the section modulus are discussed in Section III,A,2. Once the validity of this model was established for examining normal growth and development, we used the

FIGURE 10

model to examine skeletal ontogeny with mechanical loading reduced to 40% of normal [42]. Functional adaptation in the normal adult was simulated by altering the loading at maturity (20 years of age) in two ways: (i) a 60% decrease in load magnitudes and (ii) a 25% increase over normal load levels. In these analyses, only the load levels were altered; all other models parameters were maintained at their normal values. For all cases, the periosteal and endosteal radius values were calculated and compared to the results for normal development (Fig. 13). The trends represented in the results were also compared to experimental results by others under qualitatively similar conditions. Reduction of the normal loading history by 60% throughout the lifetime of the individual produces an overall diminished cross section. The adult periosteal radius was reduced 25% compared to normal, and the endosteal radius was approximately 80% of the normal value. The thickness-to-radius ratio of the sections was only slightly reduced; however, the section modulus was reduced approximately 40% as a result of the periosteal radius reduction (Fig. 14). These results cannot be fully experimentally validated, but are similar to those observed in growing animals with reduced skeletal loading [43 – 48] and are consistent with clinical observations of children born with neuromuscular defects [3,16,49]. The rapid reduction of loading after reaching maturity results in extreme cortical thinning through arrested periosteal growth and increased endosteal expansion. Although the overall cross-sectional dimensions were much greater than when the loading was reduced throughout development, the resulting cross-sectional strength changes were similar for the two cases; the section moduli were nearly equally reduced (Fig. 14). These simulations may be compared to clinical results of cortical bone adaptation following spinal cord injury. Experimental measurements of

Normal increase in femoral diaphysis section modulus with age and body mass predicted by our theoretical model [10]. Used with permission from Carter et al., Bone 18 (Suppl. 1), 5S – 10S (1996).

478

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

FIGURE 11

Section modulus plotted against age for Caucasian adolescents (r2  0.51). Male and female regressions are significantly different. Used with permission from van der Meulen et al., J. Orthop. Res. 14, 22 – 29 (1996).

diaphyseal bone mineral content (BMC) have been somewhat mixed, but several studies have shown decreased BMC [50 – 52] and increased fracture rates at femoral and tibial cortical sites [53,54]. An abrupt increase of applied loading at maturity produced approximately equivalent increases (6 – 7%) in endosteal and periosteal diameters and a 20% increase in the section modulus. These changes are consistent with increased bone strength observed in animal studies [55,56]. Direct comparison to clinical data is difficult for various reasons. Human exercise studies have been unable to produce bone hypertrophy with consistency and have reported increased, decreased, and unaltered bone mass. When increased bone density is

FIGURE 13

Effects of changes in bone loading during life on cross-sectional dimensions of the femoral diaphysis predicted by our theoretical models. Adapted from van der Meulen, Ph.D. thesis, Stanford University.

present, the changes are very modest, generally ranging from 0.5 to 3% [57]. Because the loading magnitudes are difficult to quantify, the true experimental levels are unknown and may be less than those that have been theoretically assumed. Finally, our simulations represent prolonged, sustained increased loading, and most in vivo studies follow the subjects for shorter periods. 2. MATERIAL AND STRUCTURAL STRENGTH

FIGURE 12 Regression of section modulus on body mass for Caucasian adolescents (r2  0.86). No significant effect of gender. Used with permission from van der Meulen et al., J. Orthop. Res. 14, 22 – 29 (1996).

In addition to mineral metabolism functions, the long bones of the skeleton primarily perform a structural role supporting our body mass and enabling locomotion. The structural response of a long bone to an applied force is a function of its material (or tissue) properties and geometry, and thus we need to examine these properties individually. The cortices of the long bones consist of dense cortical bone. Cortical bone is a transversely isotropic material with

479

CHAPTER 17 Skeletal Development

diaphyseal structural behavior. For example, the governing equation for the torsion of a hollow circular cylinder results in the following expression for the applied torsional moment, T, T

FIGURE 14 Effects of changes in bone loading on the section modulus of the femoral diaphysis predicted by our theoretical models.

a longitudinal elastic modulus that is nearly 50% greater than the transverse modulus [58]. Cortical bone is stronger in compression than in tension for both longitudinally and transversely applied loads and is weakest in shear loading (induced by torsion about the longitudinal axis) [58]. The material properties of bone tissue are affected by a variety of intrinsic and extrinsic factors. Contributing intrinsic properties include the bone microstructure, porosity, degree of mineralization, and density. In addition, age, race, hormones, and diet are determining factors. In the normal adult, the degree of mineralization, porosity, and density of cortical bone change relatively little. From childhood to maturity (age 8 to 26 years), the ash content of human femoral cortical bone increases only 6%, while its material strength increases 12.5% [59]. Femoral midshaft density increases 4.5% between adolescence and young adulthood [60]. In addition, no difference in bone material properties has been observed between males and females during the same period [61]. Because the material properties of cortical bone change little during postnatal growth, changes in femoral diaphyseal structural behavior are dominated by geometric changes. In the adult, when both material properties and geometry are relatively stable, differences in cortical structural strength and stiffness are also generally attributable to subtle geometric variations. Long bone cross-sectional geometry is fairly complex and varies along the bone length. Closed-form solutions to mechanical analyses, in contrast, are limited to known geometries and require many simplifying assumptions when applied to skeletal structures. In general, long bones are modeled as hollow, prismatic tubes of either a circular or a similar elliptical cross section [62 – 64]. Simple geometric relationships indicative of bone structural behavior can be derived from the governing equations for different loading conditions and used to understand

J , r

(2)

where J is the polar moment of inertia of the circular cross section, r is the outer (periosteal) radius of the tube, and  is the maximum shear stress of bone tissue. Assuming that bone material properties are relatively constant, we expect that the torsional moment that can be withstood by a bone with cross-sectional properties r and J is proportional to the geometric term J/r, which is directly proportional to the section modulus, Z. The section modulus is defined as Z

J , D

(3)

where D is the periosteal diameter of the circular cross section (twice the radius). An analogous analysis for bending demonstrates that the applied force during bending is also directly proportional to the section modulus. Therefore, the cross-sectional morphology of a long bone is critical in determining its structural behavior. This result is confirmed by experimental results, which show a very strong correlation between bending strength and section modulus (Fig. 15). Taking these concepts one step further, Selker and Carter [65] defined the “whole bone strength index,” SB, for torsion as SB 

J , DL

(4)

where L is the bone length. This index is proportional to the ultimate force required to fracture a long bone when it is

FIGURE 15

Linear regression of whole bone bending strength on section modulus for the human radius. Adapted from Martin and Burr, J. Biomech. 17, 195 – 201 (1984). The sample indicated by the square was omitted from linear regression because it was from the youngest individual and failed in a different mode than the others.

480 held by its ends and a torsional or transverse force is applied to the midshaft. This expression highlights the two important geometric aspects of whole bone strength: the cross-sectional resistance (indicated by J/D) and the bone length. To increase bone strength, there is a direct correspondence with the section modulus and an inverse relationship with bone length. Clinically, bone cross-sectional morphology is difficult to measure noninvasively and can be obtained only by tomographic techniques, which have several drawbacks and are not commonly used. Absorptiometric methods for measuring bone mass [e.g., single or dual photon absorptiometry, dual-energy X-ray absorptiometry (DXA)], however, can measure bone width and total bone mineral content only in the scan plane. The cross-sectional bending moment of inertia of the bone mineral can be directly determined for a plane perpendicular to the scan direction by integrating the absorption curve. With some assumptions similar to those presented earlier for determining whole bone structural behavior, it is possible to estimate the cross-sectional area, area moments of inertia, and section modulus of a cortical cross section [66,67]. The long bone diaphysis is modeled as a hollow circular tube with an outer diameter equal to the bone width measured in the scan plane, and the mineral density, porosity, and degree of mineralization are assumed [68]. Studies examining the relationships between cortical bone strength and the linear bone mineral density (BMC in g/cm) measured by projected radiography demonstrate a good correlation between whole bone strength and in vivo BMC at cortical sites (Fig. 16) [66,69]. These noninvasive imaging techniques have primarily been applied to cancellous bone sites in an attempt to predict fracture risk, which is discussed in the following sections (see Section II,B,2) and in other chapters of this book.

FIGURE 16

Ultimate bending moment regressed on bone mineral content at the fracture site for canine radii, ulnae, and tibiae. Adapted from Borders et al., Biochem. Eng. 99, 40 – 44 (1977).

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

B. Cancellous Bone 1. DEVELOPMENT AND ADAPTATION As noted earlier, the development of cancellous bone in the metaphysis and epiphysis is inextricably tied to the process of endochondral ossification. It is useful, therefore, to understand how endochondral ossification proceeds in the developing skeleton. In mammals, endochondral ossification commences in the central area of the cartilage anlage shortly after the primary bone collar forms. This region of primary endochondral ossification then expands and progresses toward the bones ends, establishing primary ossification fronts in both directions. The cartilage directly ahead of each ossification front exhibits the characteristic feature of interstitial cartilage growth in which the chondrocytes undergo proliferation, maturation, and hypertrophy prior to ossification. Endochondral growth and ossification can proceed without local tissue mechanical loading, provided that the biological environment is appropriate to support bone formation. However, cyclic mechanical stresses (strains) caused by physical activity provide a complex history of physical stimuli throughout the cartilage tissue of the anlage. As a result of the stress distributions created, the cartilage growth and ossification process will be accelerated in some areas and retarded in others [70 – 72]. Mechanical influences on the rate of endochondral ossification are responsible for establishing the geometry of the ossification fronts, the appearance of secondary ossification centers, and the geometry of the growth plates. The cartilage loading histories at the bone ends are also responsible for the stabilization of the subchondral growth front at skeletal maturity and, therefore, are a key factor in establishing the thickness of the articular cartilage covering the joint surfaces [73]. Later in life, articular cartilage stresses and strains play a critical role in the pathogenesis of osteoarthrosis, which can be viewed as the final stage of endochondral ossification in the cartilage anlage [70 – 72]. Cyclic tissue stresses not only regulate the process of endochondral ossification, but also directly influence the organization and remodeling of the cancellous bone initially formed at ossification fronts. The bone formed is immediately exposed to cyclic stresses that are a result of the physical activity of the fetus and, later, of the child. Cancellous bone tissue remodels toward the attractor stress stimulus by increasing or decreasing the local bone apparent density (which is inversely related to porosity) while at the same time adjusting the local trabecular orientation. In effect, bone apposition and resorption take place on the surfaces of trabeculae rather than on the periosteal and endosteal surfaces as in the appositional bone modeling of the diaphysis. The bony architecture of the proximal femur and the changes that occur to that architecture due to altered loads

481

CHAPTER 17 Skeletal Development

received considerable attention in the mid- and late 19th century [7,8,74]. The proximal femur continues to attract the interest of both clinicians and researchers for a number of reasons, including osteoporosis and the risk of hip fracture, as well as the challenges for improving the longevity of hip arthroplasty. Figure 17a shows an anteroposterior radiograph of the proximal femur. Key architectural features include the primary load-bearing trabecular system and the secondary or arcuate trabecular system within the femoral head and the region of low density bone near the center of the femoral neck referred to as Ward’s triangle. The bone slice in Fig. 17b shows additional features, including the detailed arrangement of trabecular struts in the metaphyseal and more proximal regions and the dense cortices and hollow medullary canal at the more distal regions. Our research group has conducted computer simulations of cancellous bone remodeling in the proximal femur using finite element models that represent the geometry and typical daily loading conditions. In these models, the bone apparent density is incrementally adjusted based on the error between the attractor state stimulus and the imposed stress stimulus value at each location throughout the bony model. When this process is implemented on the

computer, the entire distribution of normal bone density and architecture can be developed. Similar simulations have been conducted by others [15,75]. These results strongly suggest that the development of normal cancellous bone architecture in endochondrally derived bone is achieved primarily by epigenetic mechanobiologic processes. In addition, when these methods are used to simulate bone remodeling in response to altered stress states caused by prosthesis implantations or changes in physical activity, these computer models predict changes in bone density distributions that are consistent with the changes observed clinically and experimentally [15,19,76]. The process of cancellous bone adaptation in response to a daily loading history was illustrated by Beaupré et al. [19]. The geometry of the proximal femur was represented using a two-dimensional finite element model (Fig. 18a). To provide a rough approximation of the daily loading history of the bone, three separated loading conditions were considered that represented different activities and joint orientations encountered in a typical day. Initially it was assumed that the cancellous bone density was constant throughout the entire bone. The cumulative stress stimulus in each element was then calculated, assuming that each

FIGURE 17 Morphology of the proximal human femur by (a) radiograph illustrating the bone tissue density distribution and (b) histological section showing cancellous bone architecture.

482 loading case was imposed for many loading cycles over the course of the day. Based on the magnitude of the stress stimulus calculated relative to the attractor state stimulus, the apparent density of each element was adjusted incrementally according to a time-dependent, bone remodeling rate law similar to that used in our appositional modeling simulations. The distribution of bone apparent density after 1 and 30 remodeling increments is shown in Figs. 18b and 18c. Note the development distally of dense cortices, and proximally of a compressive trabecular column through the femoral head, the trabecular band corresponding to the arcuate system in the lateral superior neck and the lowdensity region corresponding to Ward’s triangle. In addition to the distribution of bone density, our bone remodeling simulations can also provide an indication of the directionality of the trabecular bone [77,78]. The polar plots show the equivalent normal stresses at selected locations within the remodeling femur (Fig. 19). The major and minor axes of these polar plots are an indication of trabecular orientation. One interesting note is that the trabecular orientations at a given location need not be perpendicular. Compressive equivalent normal stresses predominate in the femoral head (Fig. 19b), whereas tensile equivalent normal stresses are seen near the greater trochanter and, to a lesser extent, in the region of the superior neck and within the arcuate system (Fig. 19c).

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

Once the normal architecture of the proximal femur was created, three additional simulations were performed to predict bone changes at subsequent times. In the first simulation, the load magnitudes and the number of loading cycles were kept the same as during normal development. In the second simulation, the magnitudes of the loads and the number of cycles were reduced by 20% and in the third simulation the load magnitudes and number of cycles were increased by 20%. Continued normal loading caused little change in the bone density distributions (Fig. 20a). Reduced loading led to a general decrease in the bone density throughout the entire proximal femur (Fig. 20b). The general pattern of density distribution, however, remained similar to that of the normal femur. With increased bone loading, there was a general increase in the bone density everywhere but, again, the general pattern of density distribution remained unchanged (Fig. 20c). These simulation results are consistent with experimental and clinical studies of cancellous bone remodeling in response to changes in physical activity [79 – 81] and can be used to simulate architecture changes in the osteoporotic femur. 2. MATERIAL AND STRUCTURAL STRENGTH The local stress/strain history of the tissue strongly influences the cancellous bone microstructural characteristics that are established during morphogenesis and altered during

FIGURE 18 (a) Finite element mesh and loading conditions. Distribution of bone apparent density is shown after 1 (b) and 30 (c) remodeling increments. Adapted from Beaupré et al., J. Orthop. Res. 8, 662 – 670 (1990).

483

CHAPTER 17 Skeletal Development

FIGURE 19

Polar plots of bone equivalent normal stresses at selected locations in the proximal femur. Used with permission from Carter et al., J. Biomech. 22, 231 – 244 (1989).

functional adaptation. The tissue that is formed under this stress history exhibits material properties that can, in turn, be directly related to its chemical and microstructural character. Under normal circumstances, the two parameters that have been used most successfully to characterize cancellous bone tissue mechanical behavior are apparent density and trabecular orientation. In viewing sections of whole bones, one is often impressed by the strong variations of apparent density and trabecular orientation in the section. These distributions are directly associated with rather dramatic variations in tissue mechanical behavior throughout the bone. Furthermore, the

structural mechanical characteristics of whole bone strength and stiffness are determined by these variations. It is therefore important to appreciate the effect of apparent density and trabecular orientation on tissue mechanics. Clinical measures that reflect these characteristics in whole bones can then be better related to whole bone fracture risk. The apparent density of cancellous bone can be determined by cutting out a small (e.g., 5  5  5 mm) region of bone and washing the fat, marrow, and blood from the trabecular pores. The remaining tissue is then weighed to determine the mass of the mineralized tissue. The apparent density is the mass divided by the bulk volume (including

FIGURE 20 Bone density distributions for three different loading histories, starting with the distribution shown in Fig. 19c as the initial conditions: (a) normal loading, (b) load magnitude and number of cycles reduced by 20%, and (c) load magnitude and number of cycles increased by 20%. Adapted from Beaupré et al., J. Orthop. Res. 8, 662 – 670 (1990).

484

FIGURE 21 Relationship between trabecular bone ultimate compressive strength and apparent density. Adapted from Carter and Hayes, J. Bone Jt. Surg. 59A, 954 – 962 (1977).

pores). Although the extent of mineralization in cancellous bone is, on average, slightly less than that of cortical bone, the true density of bone tissue in cancellous bone is very close to that of cortical bone. The apparent density is, therefore, approximately proportional to the bone volume fraction, which is inversely proportional to the porosity. Bone strength is approximately proportional to the square of the apparent density (Fig. 21) [82,83]. This relationship can be used to describe the general strength characteristics of bone tissue from the most porous trabecular bone to fully compact bone. In compression, the range of strength that this represents is from less than 1 to more than 200 MPa. However, two bone specimens of the same apparent density may differ substantially in strength, depending on the trabecular microstructural characteristics. The most noticeable microstructural characteristic of cancellous bone of a specific apparent density is the organization of the trabeculae. As discussed in the previous section, trabeculae in a particular region tend to be preferentially oriented in the direction of the principal stresses imposed during daily activities. This organization imposes anisotropic characteristics on the tissue so that it is both stronger and stiffer in directions of most pronounced orientation. The anisotropic nature of cancellous bone can be documented using stereological methods that generally document the intersection of trabecular struts with a theoretical grid of parallel lines oriented in different directions [84]. Other stereological measures of secondary importance are mean trabecular width and the extent to which trabeculae are interconnected (trabecular connectivity).

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

The importance of trabecular orientation on bone strength was shown in the early study of Galante et al. [85], who examined the influence of apparent density and trabecular orientation on the compressive strength of vertebral bone specimens. In these specimens there is a pronounced orientation bias in the superior – inferior direction. They found, as have others, a positive relationship between strength and apparent density. Specimens tested in a superior – inferior direction were found to be more than twice as strong as specimens of comparable apparent density that were tested in the medial – lateral direction (Fig. 22). In considering the risk of fracture in whole bones, one should be aware that developmental and adaptational mechanics act to design whole bones for the loads experienced during normal activities and not necessarily for the loads imposed during a traumatic episode. The density distributions and trabecular orientations established in the skeleton may therefore not be well suited for specific traumatic loads that are likely to cause fracture. Nevertheless, some general statements can be made concerning the fracture resistance of cancellous bone regions that are of concern clinically. If we ignore microstructural characteristics and load direction (as a first approximation), the most important parameters to consider are bone size and bone density. Big, dense bones are stronger than small, osteoporotic bones. The DXA techniques that are widely used to measure both density and size can serve to generate predictors of whole bone strength. It is important to recognize that the projected areal bone density measures of BMD (g/cm2) do not provide a true volumetric measure of bone density (g/cm3) [86]. Instead the areal bone density BMD is positively biased by bone size. Large bones with the same apparent density as small bones will have greater DXA estimates of BMD. This

FIGURE 22

Regression of trabecular bone compressive strength on apparent density of human vertebral bone demonstrating dependence of strength on trabecular orientation. The primary trabecular orientation in these specimens was in the superior–inferior direction, and tests were performed in the superior–inferior (SI), anterior–posterior (AP), and lateral (Lat) directions. Data from Galante et al., Calcif. Tissue Res. 5, 236–246 (1970).

CHAPTER 17 Skeletal Development

inherent bias may be fortuitous, as by inherently containing a component of bone size as well as density, BMD values turn out to be good predictors of whole bone strength. To understand how bone density and size contribute to the strength of a whole bone, consider hypothetical cubes of bone tissue with apparent density  and cross-sectional area A. The force, F, required to fracture these cubes of bone would be proportional to 2A because the tissue fracture stress would be proportional to 2 and the tissue stress is equal to the applied force divided by A. The parameter 2A can be considered a strength index. If we take a DXA scan of these same cubes we would find that the areal density BMD (g/cm2) would be equal to t, where t is the thickness of the cube. The thickness of the cube is equal to the square root of A. Therefore, the areal density BMD is directly proportional to the square root of the strength index. One might, therefore, expect a correlation between areal density BMD and whole bone fracture strength. Although there is a good deal of scatter in whole bone testing, some investigators are finding reasonable correlations between areal density BMD values and strengths of cadaveric specimens tested in the laboratory. A relationship was found between failure load and femoral neck BMD by Courtney et al. [87], who tested young adult and elderly human femurs to failure in a fall-loading configuration. The ultimate load for both groups combined was found to be strongly positively correlated with the BMD measured at the femoral neck by DXA (Fig. 23).

IV. ADAPTATIONAL MECHANICS IN AGING AND DISEASE This chapter concentrated on the role of mechanical factors in the development and adaptation of the skeleton. This was done with the belief that the skeleton is a self-designing

FIGURE 23

Regression of failure load on BMD for the femoral neck of human femora. Adapted from Courtney et al., Calcif. Tissue Res. 55, 53 – 58 (1994).

485 structure and that the tissue mechanical characteristics and whole bone strength are primarily a consequence of the loading histories that are imposed during ontogeny. With this simple view one can demonstrate that increases or decreases in bone mass may appear as a direct result of changes in the intensity of daily physical activities. Whereas osteoporosis in many individuals may be partly due to decreases in skeletal loading, a purely mechanical view of the pathogenesis of osteoporosis is clearly an oversimplification. Skeletal developmental and adaptational mechanics must be evaluated in light of the many genetic, metabolic, and dietary factors that have been shown to influence bone density and strength in important ways. In general, nonmechanical factors can influence bone by either influencing the basic quality of bone (e.g., mineralization, chemical composition, ultrastructure) or simply increasing or decreasing the amount of bone that is present. In some instances, such as fluoride treatment, both bone quality and bone quantity are affected. Changes in bone quality are the result of changes in the basic biophysics of bone formation and mineralization. Changes in bone quantity alone can be realized by simply changing the balance of osteoblastic and osteoclastic activity as reflected by the number and activities of various populations of bone cells. The number of identifiable agents that influence bone cell biology are staggering and include estrogen, vitamin D, calcium, parathyroid hormone, calcitonin, bisphosphonates, and fluoride. The specific influences of these and other nonmechanical factors are described in other chapters of this book. The mechanisms of action of these factors may all differ but can be of two basic types. These chemical agents could either influence the mechanical regulation of bone cells or have a direct influence on bone cells and their precursors that is independent of the mechanical stimuli. In either case there would be apparent interactions between mechanical and nonmechanical factors. Such synergistic influences have been shown in both preclinical and clinical studies, indicating that mechanical stimuli for bone hypertrophy or atrophy can be altered by endocrine status, diet, or drugs [e.g., 88 – 90]. If one wishes to maintain a view of bone regulation that is dominated by mechanical factors, one can conceptually model the nonmechanical factors as agents that effectively alter the level of mechanical stimulus that is required to maintain bone. In the model of Fig. 3, we find that this viewpoint will cause variations in the attractor stress stimulus, AS, which regulates the mechanically related bone remodeling stimulus. This approach was used by Carter and Beaupré [91] to represent the influence of fluoride treatment on changes in bone volume fraction and was also used by Carter et al. [71] to represent genetic differences in bone mass among different individuals. Frost [92] has employed a similar perspective on factors that may alter the mechanical “set point” of bone. He proposed that such a perspective may be useful for viewing the many bone changes that are observed in osteoporosis.

486

VAN DER

We are only beginning to be able to understand and effectively model the many intrinsic and extrinsic (both mechanical and chemical) factors that control the development and maintenance of bone. It is clear, however, that the biomechanical characteristics of bone and its fracture risk are tied to its ontogenetic history. The structure and mechanical properties of the bones are, in fact, a direct reflection of prior mechanical loading, metabolic status, and diet. Prior mechanical function is perhaps the most dominant factor in determining the form and strength of bones.

References 1. J. I. P. de Vries, G. H. A. Visser, and H. F. R. Prechtl, The emergence of fetal behavior. I. Qualitative aspects. Early Hum. Dev. 7, 301 – 322 (1982). 2. B. K. Hall and S. W. Herring, Paralysis and growth of the musculoskeletal system in the embryonic chick. J. Morphol. 206, 45 – 56 (1990). 3. J. I. Rodriguez, A. Garcia-Alix, J. Palacios, and R. Paniagua, Changes in long bones due to fetal immobility caused by neuromuscular disease. J. Bone Jt. Surg. 70A, 1052 – 1060 (1988). 4. J. I. Rodriguez, J. Palacios, A. Garcia-Alix, I. Pastor, and R. Paniagua, Effects of immobilization on fetal bone development: A morphometric study in newborns with congenital neuromuscular diseases with intrauterine onset. Calcif. Tissue Int. 43, 335 – 339 (1988). 5. J. A. Ogden, Chondro-osseous development and growth. In “Fundamental and Clinical Bone Physiology” (M. R. Urist, ed.), pp. 108 – 171. Lippincott, Philadelphia, 1980. 6. D. R. Carter, M. C. H. van der Meulen, and G. S. Beaupré, Mechanical factors in bone growth and development. Bone 18(Suppl. 1), 5S – 10S (1996). 7. J. Wolff, “The Law of Bone Remodeling.” Springer-Verlag, Berlin, 1892. 8. W. Roux, “Gesammelte Abhandulgen über Entwicklungsmechanic der Organismen,” Vols. I and II. Wilhelm Engelmann, Leipzig, 1895. 9. D. R. Carter, Mechanical loading history and skeletal biology. J. Biomech. 20, 1095 – 1109 (1987). 10. M. C. H. van der Meulen, G. S. Beaupré, and D. R. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 11. G. S. Beaupré, T. E. Orr, and D. R. Carter, An approach for timedependent bone modeling and remodeling: Theoretical development. J. Orthop. Res. 8, 651 – 661 (1990). 12. R. T. Whalen, D. R. Carter, and C. R. Steele, Influence of physical activity on the regulation of bone density. J. Biomech. 21, 825 – 837 (1988). 13. D. R. Carter, The relationship between in vivo strains and cortical bone remodeling. In “CRC Critical Reviews in Biomedical Engineering” (J. R. Bourne, ed.), pp. 1 – 28. CRC Press, Boca Raton, FL, 1982. 14. H. Frost, “Intermediary Organization of the Skeleton,” Vol. 1. CRC Press, Boca Raton, FL, 1986. 15. R. Huiskes, H. Weinans, H. J. Grootenboer, M. Dalstra, B. Fudala, and T. J. Sloof, Adaptive bone-remodeling theory applied to prosthetic-design analysis. J. Biomech. 20, 1135 – 1150 (1987). 16. Z. A. Ralis, H. M. Ralis, M. Randall, G. Watkins, and P. D. Blake, Changes in shape, ossification and quality of bone in children with spina bifida. Dev. Med. Child. Neurol. 18 (Suppl 37), 29 – 41 (1976). 17. M. Wong, J. Germiller, J. Bonadio, and S. A. Goldstein, Neuromuscular atrophy alters collagen gene expression, pattern formation, and mechanical integrity of the chick embryo long bone. Prog. Clin. Biol. Res. 383B, 587 – 597 (1993).

MEULEN, CARTER, AND BEAUPRÉ

18. J. E. Bertram, L. S. Greenberg, T. Miyake, and B. K. Hall, Paralysis and long bone growth in the chick: Growth shape trajectories of the pelvic limb. Growth Dev. Aging 61, 51 – 60 (1997). 19. G. S. Beaupré, T. E. Orr, and D. R. Carter, An approach for timedependent bone modeling and remodeling — Application: A preliminary remodeling situation. J. Orthop. Res. 8, 662 – 670 (1990). 20. C. T. Rubin and L. E. Lanyon, Dynamic strain similarity in vertebrates; an alternative to allometric limb bone scaling. J. Theor. Biol. 107, 321 – 327 (1984). 21. A. A. Biewener, S. M. Swartz, and J. E. A. Bertam, Bone modeling during growth: Dynamic strain equilibrium in the chick tibiotarsus. Calcif. Tissue Int. 39, 390 – 395 (1986). 22. K. Indrekvam, O. Husby, N. Gjerdet, L. Engester, and N. Langeland, Age-dependent mechanical properties of rat femur: Measured in vivo and in vitro. Acta Orthop. Scand. 62, 248 – 252 (1991). 23. T. S. Keller and D. M. Spengler, Regulation of bone stress and strain in the immature and mature rat femur. J. Biomech. 22, 1115 – 1127 (1989). 24. J. L. Vaughan, “The Physiology of Bone,” 3rd Ed. Clarendon, Oxford, England, 1981. 25. C. B. Ruff, A. Walker, and E. Trinkaus, Postcranial robusticity in Homo. III. Ontogeny. Am. J. Phys. Anthrop. 93, 35 – 54 (1994). 26. J. D. Currey, “The Mechanical Adaptations of Bones.” Princeton Univ. Press, Princeton, 1984. 27. A. M. Parfitt, Bone-forming cells in clinical conditions. In “Bone” (B. K. Hall, ed.), pp. 351 – 429. Telford Press, Caldwell, NH, 1992. 28. C. B. Ruff and W. C. Hayes, Subperiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 947 (1982). 29. A. A. Biewener, Musculoskeletal design in relation to body size. J. Biomech. 24 (Suppl. 1), 19 – 29 (1991). 30. T. S. Gross, K. J. McLeod, and C. T. Rubin, Characterizing bone strain distributions in vivo using three triple rosette strain gages. J. Biomech. 25, 1081 – 1087 (1992). 31. H. N. Munro, Evolution of protein metabolism in mammals. In “Mammalian Protein Metabolism” (H. N. Munro, ed.), pp. 133 – 182. Academic Press, New York, 1969. 32. R. B. Martin and P. J. Atkinson, Age and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 33. R. W. McCammon, “Human Growth and Development.” Thomas, Springfield, IL, 1970. 34. R. W. Smith and R. R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 35. S. M. Garn, C. G. Rohmann, B. Wagner, and W. Ascoli, Continuing bone growth throughout life: A general phenomenon. Am. J. Phys. Anthrop. 26, 313 – 318 (1967). 36. C. B. Ruff and W. C. Hayes, Cross-sectional geometry of Pecos Pueblo femora and tibiae: A biomechanical investigation. II. Sex, age, and side differences. Am. J. Phys. Anthropol. 60, 383 – 400 (1983). 37. C. B. Ruff, Allometry between length and cross-sectional dimensions of the femur and tibia in Homo sapiens sapiens. Am. J. Phys. Anthrop. 65, 347 – 358 (1984). 38. M. C. H. van der Meulen, M. W. Ashford, Jr., B. J. Kiratli, L. K. Bachrach, and D. R. Carter, Determinants of femoral geometry and structure during adolescent growth. J. Orthop. Res. 14, 22 – 29 (1996). 39. M. Moro, M. C. H. van der Meulen, B. J. Kiratli, R. Marcus, L. K. Bachrach, and D. R. Carter, Body mass is the primary determinant of mid-femoral bone acquisition during adolescence. Bone 19, 519 – 526 (1996). 40. M. C. H. van der Meulen, R. Marcus, L. K. Bachrach, and D. R. Carter, Correspondence between theoretical models and DXA measurements of cross-sectional growth in adolescence. J. Orthop Res. 15, 473 – 476 (1997).

CHAPTER 17 Skeletal Development 41. M. C. H. van der Meulen, M. Moro, B. J. Kiratli, R. Marcus, and L. K. Bachrach, Mechanobiology of femoral neck structure during adolescence. J. Rehab. Res. Dev. 37, 201 – 209 (2000). 42. M. C. H. van der Meulen, “The Influence of Mechanics on Long Bone Development and Adaptation,” Ph.D. thesis, Stanford University, 1993. 43. M. Geiser and J. Trueta, Muscle action, bone rarefaction and bone formation. J. Bone Jt. Surg. 40B, 282 – 311 (1958). 44. L. E. Lanyon, The influence of function on the development of bone curvature. J. Zool. Lond. 192, 457 – 466 (1980). 45. S. R. Shaw, A. C. Vailas, R. E. Grindeland, and R. F. Zernicke, Effects of a one-week spaceflight on the morphological and mechanical properties of growing bone. Am. J. Physiol. Reg. Int. Comp. Physiol. 254, R78 — R83 (1988). 46. D. M. Spengler, E. R. Morey, D. R. Carter, R. T. Turner, and D. J. Baylink, Effects of spaceflight on structural and material strength of growing bone. Proc. Soc. Exp. Biol. Med. 174, 224 – 228 (1983). 47. H. K. Uhthoff and Z. F. G. Jaworski, Bone loss in response to longterm immobilisation. J. Bone Jt. Surg. 60B, 420 – 429 (1978). 48. M. C. H. van der Meulen, E. R. Morey-Holton, and D. R. Carter, Hindlimb suspension diminishes femoral cross-sectional growth in the rat. J. Orthop. Res. 13, 700 – 707 (1995). 49. S. W. Burke, V. P. Jameson, J. M. Roberts, C. E. Johnston, and J. Willis, Birth fractures in spinal muscular atrophy. J. Pediatr. Orthop. 6, 34 – 36 (1986). 50. F. Biering-Sørenson, H. H. Bohr, and O. P. Schaadt, Longitudinal study of bone mineral content in the lumbar spine, the forearm and the lower extremities after spinal cord injury. Eur. J. Clin. Invest. 20, 330 – 335 (1990). 51. A. Chantraine, B. Nugens, and C. M. Lapiere, Bone remodeling during the development of osteoporosis in paraplegia. Calcif. Tissue Int. 38, 323 – 327 (1986). 52. R. L. Prince, R. I. Price, and S. Ho, Forearm bone loss in hemiplegia: A model for the study of immobilization osteoporosis. J. Bone Miner. Res. 3, 305 – 310 (1988). 53. A. E. Comarr, R. H. Hutchinson, and E. Bors, Extremity fractures of patients with spinal cord injuries. Am. J. Surg. 103, 732 – 739 (1962). 54. K. T. Ragnarsson and G. H. Sell, Lower extremity fractures after spinal cord injury: A retrospective study. Arch. Phys. Med. Rehabil. 62, 418 – 423 (1981). 55. R. K. Martin, J. P. Albright, W. R. Clarke, and J. A. Niffenegger, Load-carrying effects on the adult beagle tibia. Med. Sci. Sports Exercise 13, 343 – 349 (1981). 56. S. L.-Y. Woo, S. C. Kuei, D. Amiel, M. A. Gomez, W. C. Hayes, F. C. White, and W. H. Akeson, The effect of prolonged physical training on the properties of long bone: A study of Wolff’s law. J. Bone Jt. Surg. 63A, 780 – 787 (1981). 57. R. Marcus and D. R. Carter, The role of physical activity in bone mass regulation. Adv. Sports Med. Fitness 1, 63 – 82 (1988). 58. D. T. Reilly and A. H. Burstein, The elastic and ultimate properties of compact bone tissue. J. Biomech. 8, 393 – 405 (1975). 59. J. D. Currey and G. Butler, The mechanical properties of bone tissue in children. J. Bone Jt. Surg. 57A, 810 – 814 (1975). 60. P. Atkinson and J. A. Weatherell, Variation in the density of the femoral diaphysis with age. J. Bone Jt. Surg. 49B, 781 – 788 (1967). 61. A. H. Burstein, D. T. Reilly, and M. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt. Surg. 58A, 82 – 86 (1976). 62. J. G. Kennedy and D. R. Carter, Long bone torsion. I. Effects of heterogeneity, anisotropy and geometric irregularity. J. Biomech. Eng. 107, 183 – 188 (1985). 63. M. E. Levenston, G. S. Beaupré, and M. C. H. van der Meulen, Improved method for analysis of whole bone torsion tests. J. Bone Miner. Res. 9, 1459 – 1465 (1994).

487 64. R. B. Martin, Determinants of the mechanical properties of bones. J. Biomech. 24 (Suppl. 1), 79 – 88 (1991). 65. F. Selker and D. R. Carter, Scaling of long bone fracture strength with animal mass. J. Biomech. 22, 1175 – 1183 (1989). 66. R. B. Martin and D. B. Burr, Non-invasive measurement of long bone cross-sectional moment of inertia by photon absorptiometry. J. Biomech. 17, 195 – 201 (1984). 67. T. M. Cleek and R. T. Whalen, Bone structural properties in the tibia and fibula using DXA. Int Workshop Bone Densitometry (2000). 68. V. K. Sarin, E. G. Laboa, G. S. Beaupré, B. J. Kiratli, D. R. Carter, and M. C. H. van der Meulen, DXA-derived section modulus and bone mineral, content predict long bone torsional strength. Acta Orthop. Scand. 70, 71 – 76 (1999). 69. S. Borders, K. R. Petersen, and D. Orne, Prediction of bending strength of long bones from measurements of bending stiffness and bone mineral content. J. Biomech. Eng. 99, 40 – 44 (1977). 70. D. R. Carter, T. E. Orr, D. P. Fyhrie, and D. J. Schurman, Influences of mechanical stress on prenatal and postnatal skeletal development. Clin. Orthop. 219, 237 – 250 (1987). 71. D. R. Carter, M. Wong, and T. E. Orr, Musculoskeletal ontogeny, phylogeny, and functional adaptation. J. Biomech. 24 (Suppl. 1), 3 – 16 (1991). 72. M. Wong and D. R. Carter, A theoretical model of endochondral ossification and bone architectural construction in long bone ontogeny. Anat. Embryol. 181, 523 – 532 (1990). 73. G. S. Beaupré, S. S. Stevens, and D. R. Carter, Mechanobiology in the development, maintenance, and degeneration of articular cartilage. J. Rehab. Res. Dev. 37, 145 – 151 (2000). 74. G. H. von Meyer, Die Architektur der Spongiosa. Arch. Anat. Physiol. Wiss. Med. 34, 615 – 628 (1867). 75. H. Weinans, R. Huiskes, and H. J. Grootenboer, The behavior of adaptive bone-remodeling simulation models. J. Biomech. 25, 1425 – 1441 (1992). 76. T. E. Orr, G. S. Beaupré, D. R. Carter, and D. J. Schurman, Computer predictions of bone remodeling around porous-coated implants. J. Arthrop. 5, 191 – 200 (1990). 77. D. R. Carter, T. E. Orr, and D. P. Fyhrie, Relationships between loading history and femoral cancellous bone architecture. J. Biomech. 22, 231 – 244 (1989). 78. C. R. Jacobs, J. C. Simo, G. S. Beaupré, and D. R. Carter, Adaptive bone remodeling incorporating simultaneous density and anisotropy considerations. J. Biomech. 30, 603 – 613 (1997). 79. G. P. Dalsky, K. S. Stocke, A. A. Ehsain, E. Slatopolsky, W. C. Lee, and S. J. Birge, Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann. Int. Med. 108, 824 – 828 (1988). 80. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers, and D. E. McMillan, Effect of prolonged bed rest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 81. W. S. S. Jee and X. J. Li, Adaptation of cancellous bone to overloading in the adult rat: A single photon absorptiometry and histomorphometry study. Anat. Rec. 227, 418 – 426 (1990). 82. D. R. Carter, W. C. Hayes, and D. J. Schurman, Fatigue life of compact bone. II. Effects of microstructure and density. J. Biomech. 9, 211 – 218 (1976). 83. D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two-phase porous structure. J. Bone Jt. Surg. 59A, 954 – 962 (1977). 84. A. D. Kuo and D. R. Carter, Computational methods for analyzing the structure of cancellous bone in planar sections. J. Orthop. Res. 9, 918 – 931 (1991). 85. J. Galante, W. Rostoker, and R. D. Ray, Physical properties of trabecular bone. Calcif. Tissue Res. 5, 236 – 246 (1970). 86. D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone Miner. Res. 7, 137 – 145 (1992).

488 87. A. C. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Effects of loading rate on strength of the proximal femur. Calcif. Tissue Int. 55, 53 – 58 (1994). 88. Y. Kodama, Y. Umemura, S. Nagasawa, W. G. Beamer, L. R. Donahue, C. R. Rosen, D. J. Baylink, and J. R. Farley, Exercise and mechanical loading increase periosteal bone formation and whole bone strength in C57BL/6J mice but not in C3H/Hej mice. Calcif. Tissue Int. 66, 298 – 306 (2000). 89. B. F. Halloran, D. D. Bikle, J. Harris, S. Tanner, T. Curren, and E. Morey-Holton, Regional responsiveness of the tibia to intermittent

VAN DER

MEULEN, CARTER, AND BEAUPRÉ

administration of parathyroid hormone as affected by skeletal unloading. J. Bone Miner. Res. 12, 1068 – 1074 (1997). 90. T. L. Robinson, C. Snow-Harter, D. R. Taaffe, D. Gillis, J. Shaw, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10, 26 – 35 (1995). 91. D. R. Carter and G. S. Beaupré, Effects of fluoride treatment on bone strength. J. Bone Miner. Res. 5 (Suppl. 1), S177 — S184 (1990). 92. H. M. Frost, The pathomechanics of osteoporoses. Clin. Orthop. 200, 198 – 225 (1985).

489

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

CHAPTER 18

Inhibition of Osteopenia by Biophysical Intervention CLINTON T. RUBIN,*,† STEFAN JUDEX,* KENNETH J. MCLEOD,* AND YI-XIAN QIN* *

Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, and the Center for Biotechnology, State University of New York, Stony Brook, New York 11794

†

I. II. III. IV.

V. Attenuation of Wolff’s Law by Systemic Disorders VI. Clinical Application of Biophysical Stimuli VII. Summary References

Introduction Wolff’s Law of Bone Adaptation Structural Demands on the Vertebrate Skeleton Modulation of Bone Tissue by Biophysical Stimuli

I. INTRODUCTION

mechanical environment of bone and demonstrating that physically based signals within these physiologic constraints can be osteogenic. Further, evidence will be provided that demonstrates that these regulatory signals diminish as we age because of sarcopenia (a state of diminished muscle mass and function), suggesting that a key etiologic factor in osteoporosis results from degeneration of the muscular system, rather than a primary dysfunction of the skeletal system per se. Finally, preliminary evidence is provided that extremely low-level mechanical signals are capable of inhibiting the rapid bone loss that typically follows menopause. While a biophysical approach contrasts sharply with pharmaceutical strategies for the treatment of osteoporosis, in essence, the structural success of the skeleton is a product of the ability of bone tissue to adapt to this constant barrage of mechanically based signals, and herein lies the basis for a unique treatment regimen for this debilitating skeletal disease. The elaborate cortical and trabecular morphology of the skeleton is a dynamic product of three competing and disparate goals: pressure to establish (and maintain) mechanical strength, the metabolic advantages inherent in a

Osteopenia, a condition of diminished bone mass, will become symptomatic only when mechanical demands exceed the structural viability of the skeleton. In other words, when function is no longer accommodated by form. While treatment of osteoporosis is principally oriented toward pharmaceutical prophylaxis, this chapter proposes a case for considering biophysical intervention in general (e.g., electrical, acoustic, thermal), and biomechanical treatment in particular. Much of the frustration and ambivalence toward biophysical treatment for osteoporosis may, in reality, be a result of a poorly defined osteogenic signal, exacerbated by decreasing sensitivity/responsiveness of the aging skeleton. However, there is accumulating basic science and clinical evidence that biophysical intervention may prove a safe and efficacious means of inhibiting and reversing osteoporosis, a goal that can be achieved by augmentation, rather than disruption, of bone remodeling processes. This chapter will try to build support for the clinical potential of biophysical stimuli by defining the functional

OSTEOPOROSIS, SECOND EDITION VOLUME 1

489

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

490 minimal mass, and tissue serving as the body’s principal mineral reservoir. This balance is achieved via site-specific formation and resorption of bone, reflecting a combative struggle between systemically based catabolic factors (e.g., parathyroid hormone, vitamin D) aimed at releasing calcium from the skeleton, and site-specific biophysical signals, which arise from function (e.g., mechanical strain, pressure generated fluid flow, stress-generated electrical potentials), which serve as local anabolic factors. When metabolic demands swamp the structural role, the skeleton can deteriorate to the point that it can no longer withstand normal functional load bearing. While osteopenia may result from an increased level of resorptive factors, it is as likely due to an age- or menopause-related attenuation of the physically based signals that serve to maintain bone structure. The premise of the chapter is that osteoporosis is an end product of a dysfunctional form/function relationship. This perspective is supported by the fact that this disease is most symptomatic at specific load bearing sites of the skeleton (e.g., femoral neck, lumbar spine). Paradoxically, even though the disease is focal in nature, the most accepted treatment protocols are administered systemically. Intervention that inhibits, disrupts, or even promotes the bone cell kinetics of the entire skeleton belies the need for a focal treatment strategy for a focal etiology. Further, an effective treatment for osteoporosis cannot realistically arise simply by statically retaining the bone mass or density that is present at any given point in time (e.g., antiresorptives). Bone quality is as important as bone quantity, emphasizing that the ideal therapy will be one that incorporates all aspects of normal bone turnover, not one that annihilates any given part of it. An optimal treatment would target a site-specific regimen for the inhibition and/or reversal of bone loss and achieve this without interrupting the delicate interplay between the cells responsible for bone remodeling. Ideally, this therapy would reestablish the osteogenic components of this balance to a level that would revitalize a normal remodeling equilibrium. The anabolic potential of load bearing, as well as the site-specific loss of bone density in osteoporosis, implies that an improved understanding of how biophysical signals mediate remodeling will facilitate their effective application to the treatment of these diseases.

II. WOLFF’S LAW OF BONE ADAPTATION The principal responsibility of the skeleton is to support the loads and bending 1moments that arise during activity, a responsibility that results in mechanical strain in the bone tissue. The ability of the skeleton to adapt to these functional demands was recognized over a century ago and is now referred to as Wolff’s law [1]. The basic premise of

CLINTON T. RUBIN

ET AL.

this “law” is that bone strives toward an optimized structure, which caters to an individual’s level of activity. Thus, each individual would tune the mass and morphology of the skeleton such that it was sufficient to withstand the extremes of functional loading, but not so massive as to make transportation a metabolic liability. While this Goldilocks paradigm of “just right” emphasizes the role of anabolic functional stimuli in defining skeletal morphology, identifying those components within the milieu that achieve this balance has proven difficult. Nonetheless, it is a venture worth pursuing. If the osteogenic constituents of the functional regime could be identified, focal modalities could be developed to treat skeletal disorders that exploit the extreme sensitivity of the musculoskeletal system to the mechanical domain. The search for the osteogenic components that control Wolff’s law has benefited from both qualitative and quantitative observations of the skeleton’s response to changes in its functional environment. The sensitivity of bone to physical and environmental stimuli is readily evident in a large body of clinically based studies that show the skeleton’s graded response to levels of exercise [2,3], as well as the bone lost due to reductions in gravitational force [4,5] and bed rest [6]. Importantly, evidence of local hypertrophy (e.g., humerus in tennis players [7,8]) or resorption (e.g., femoral neck after total arthroplasty [9,10]) following site-specific activities emphasizes that focally mediated adaptation is caused by changes in the local functional environment. While these studies portray the skeleton’s sensitivity to function, the difficulty in defining the complex loading history of the bone under study has precluded identification of the osteoregulatory component(s) embedded within the physical milieu. To address these limitations, analytic and empiric models have been developed to study physical influences on bone formation. Through the past two decades, specific components of the mechanical milieu have been proposed as the dominant stimulus for bone adaptation. These include strain magnitude [11], strain rate [12], electrokinetic currents [13], piezoelectric currents [14], fluid shear flow [15], and strain energy density [16]. While these parameters demonstrate strong correlations to specific skeletal morphologies, few have validated their prescience by predicting morphologic changes that would be stimulated by distinct loading condition [17]. The inability to identify a unifying principle for the mechanical control of bone adaptation may be aggravated by the underlying assumption that a structural efficiency paradigm (minimal skeletal strain/minimal skeletal mass) is itself the driving stimulus that regulates the remodeling process. Alternatively, bone cells may know little of the principles of structural mechanics and 1 Moment is an engineering term that describes the turning, twisting, or rotational effect of a force; M  Nm.

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

491

instead are responding to “biologically relevant” parameters of functional milieu that are not necessarily linked to peak structural challenges. This “other than peak” perspective is employed in several biologic systems, which perceive and respond to exogenous stimuli, such as vision, hearing, and touch. In considering the mechanically mediated control of bone remodeling, there is little argument that biophysical stimuli are potent determinants of skeletal morphology; but too much may only damage the system. Indeed, we shut our eyes when it is too bright, cover our ears when it is too loud, and shed tears when the pressure is too great. To identify criteria by which these processes are controlled, it is necessary to look beyond the material consequences of a structure subject to load and consider the biologic benefit of a viable tissue subject to functional levels of strain.

significant safety factor reigns and that the skeleton can survive an errant step into a plot hole or the odd trip over a curb. However, imagining a mechanism whereby the skeleton “knows” it is loaded to half its yield strength seems unlikely, somehow altering its morphology based on a paranoia of strains that are much higher. Instead, in vivo and in vitro data suggest that the functional criteria that regulate adaptation, and the means by which bone cells perceive and respond to their functional milieu, are far more sophisticated than a magnitude “brute force” or “accumulation of microdamage” perspective. Indeed, these data suggest that adaptation to biophysical stimuli occurs to encourage specific components of the strain milieu, rather than to necessarily annihilate them. Two thousand microstrain may sound like an ominously large number, but in reality it represents an exceedingly small change in length from a material’s original length. Cartilage is subject to 25% compressive deformations, tendons experience functional tensile strain upward of 20%, and ligaments may well stretch 4 – 5% during the extremes of functional loading. The 20% strain of these connective tissues is two orders of magnitude greater than the 0.2% (2000 ) peak strains experienced by bone. Admittedly, even a 0.2% strain seems unwieldy for a bridge support or skyscraper, yet by the time a 10 , bone-lining cell is subject to 2000 , such deformation is on the order of angstroms. Clearly, if deformations of this order are to affect cell metabolism, the bone cell mechanosensory system must be exceedingly sensitive (Fig. 1).

III. STRUCTURAL DEMANDS ON THE VERTEBRATE SKELETON A. Cross-Species Similarity of Peak Bone Strain Magnitudes Regardless of the design or function of a vertebrate, strain is a ubiquitous product of a functionally loaded skeleton. Mechanical strain, therefore, is commonly considered a reasonable and efficient means of translating the intensity, duration, and manner of functional loading into a site-specific, generic signal relevant to the cells responsible for osteoregulation. One obvious goal of this strain-mediated form/function formula is to avoid fracture. Thus, bone loading and architecture must be coordinated to avoid the tissue’s yield strain of 0.7% (7000 microstrain () [18]). To establish the role of functional strain in defining skeletal morphology, as well as to determine how closely bone approaches its point of failure, the mechanical signals to which the skeleton is exposed must be determined. This goal is achieved by attaching strain gauges directly to a bone in vivo [19,20], which permits a number of critical observations to be made regarding the structural demands made on the skeleton. Peak strain magnitudes measured in diverse vertebrates are remarkably similar, ranging in amplitude from 2000 to 3500  [21,22]. Whether measured in metacarpal bone of a galloping horse, the tibia of a running human, the humerus of a flying goose, the femur of a trotting sheep, or the mandible of a chewing macaque, this “dynamic strain similarity” suggests that skeletal morphology is adjusted in such a way that functional activity elicits a very specific (and perhaps beneficial) level of strain to the bone tissue [23]. That strains of this magnitude are a factor of two below the yield point of bone material emphasizes that a

B. Absence of a Uniform Peak Strain Stimulus Models aimed at defining the osteogenic components of the overall strain history of alone (accumulated strain information over time) have focused on correlating bone morphology to the predominant characteristics of the mechanical environment of the bone, including peak strain magnitude, peak strain rate, peak strain energy density, or number of peak loading cycles. Certainly, that peak strain magnitudes among vertebrates are all very similar would support a hypothesis that achieving a specific level of peak strain is the Holy Grail toward which the bone tissue strives. However, models based on this hypothesis also commonly assume that a homogeneous state of strain persists across the cortex [24 – 27]. In other words, at either that point in the stride where the strains are greatest, or when strains are integrated over some time-averaging period, it is assumed that each area of the cortex is subject to the identical strain information and therefore the same stimulus for remodeling. In vivo strain gauge data, however, demonstrate the spatial distribution of peak normal and shear strains, as well as strain energy density (an aggregate of the stress/strain state), to be extremely non uniform [20,28]. For example, for a horse

492

FIGURE 1

Strain is defined as a (load-induced) change in length relative to the structure’s original length. One thousand microstrain, or 0.1% strain, reflects the amount of strain experienced by bone tissue during an activity such as walking. For a structure such as the 170-m Washington monument, 1000  would represent a 17-cm change in length over the entire structure. In a giraffe tibia, 1000  would reflect a 1-mm change over the bone’s original 1000-mm length. At the level of a 10-m bone lining cell sitting on the periosteum of that giraffe tibia, its dimensional change when subject to 1000  would be 100 A. The mechanisms responsible for perceiving and responding to such small biophysical signals, whatever they may be, must be extremely sensitive. Reproduced with permission from Rubin et al. [90].

walking at 2 ms1, at the point in the stride in which peak strain is achieved, the spatial distribution of normal strain in the metacarpal ranges from plus 13  in tension to minus 1048  in compression. Further, shear strain ranges from 54 to 360  (indicating opposite directions of shear), and peak strain energy density, which accounts for all components of the strain tensor, spans two orders of magnitude, from 117 to 10,602 Pa. A uniform strain stimulus is certainly not readily apparent in the functionally loaded skeleton. Bone structure would realize an important benefit if it were to adapt to elicit a homogeneous strain distribution; peak strain could be minimized while simultaneously minimizing total bone mass. Imagine trying to break a pencil by loading it only in an axial fashion, along its longitudinal shaft. Short of enlisting a materials test machine, such an endeavor would prove difficult. As any frustrated writer knows, the effortless way to snap a pencil is to bend it. Taking this design hazard into account, structural models of bone adaptation insist that bone strives to minimize risk of failure by ensuring that the bone is loaded axially, hence the uniform distribution of strain. It certainly seems aesthetically reasonable that architectural embellishments of the skeleton such as cross-sectional morphology and longitudinal curvature, together with antagonistic and synergistic

CLINTON T. RUBIN

ET AL.

muscle activity, all conspire to minimize any bending in the bone. Empirical evidence, however, measured from the appendicular skeleton during functional activity, demonstrates that the predominant (85%) component of strain is generated by bending, even though far less bone mass would be required to support the same loads if the bone were loaded axially [21,29,30]. The bending is sufficient to subject a significant portion of the cortex to longitudinal tension, thus mimicking the perilous state of the doomed pencil. Further, committing one surface to tension and another to compression means that the transition between these two areas creates a region of the cortex which experiences very low peak strain magnitudes. Even though this “neutral axis” is far removed from the area of the cortex subject to the peak strains, somehow tissue is retained in this strain demilitarized zone (DMZ) [31–33]. Clearly, bone cannot be presumed to be solely a compressive element, and strain cannot be presumed to be uniform across the cortex. It could be argued that a more uniform strain stimulus could be achieved via integration of strain information over time [26,27]. Such a time-dependent “strain memory” in bone cell networks has already been demonstrated, with noncollagenous matrix proteins “responding” to load by changing their morphology [34]. Over the course of time, a homogeneous strain signal could be achieved by ensuring all areas of the cortex are subject — at one point or another within the tissue’s memory span — to the peak strain milieu. However, this “equilibration presumption” is not consistent with in vivo data that show that the inhomogeneity of the instantaneous strain distribution becomes even more disparate as the strain energy is summed over the course of a stride (Fig. 2). Summing the functional strain milieu over an entire 24-h period demonstrates the range of total strain experienced between areas of the cortex to be huge [35], approaching three orders of magnitude. While high degrees of bending may provoke distinctly non uniform strain distribution in cortical bone, it does not necessarily preclude the possibility of an adaptive mechanism mediated by some aspect of strain. One solution to this dilemma is to suggest that bone cells in different regions of the cortex may be differentially sensitive to strain (some cells strive to 3000  in compression, some to 1500  in tension; others — near the neutral axis — are content with strains of 50 or 100 ). While this is appealing in its simplicity, the genetic logistics of a spatially specific strain sensitivity would be astronomical. Alternatively, it is possible that strain information is spatially integrated in three dimensions via a cell network facilitated by gap junction intercellular communication [36], such that the are of the cortex subject only to 100  resists resorption due to sufficient homeostatic signals received from adjacent areas subject to much higher strain. This “information integration” perspective is supported by the observation that the bone loss that parallels disuse occurs

493

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

FIGURE 2 The distribution of strain energy about the midshaft of the horse cannon bone (MCIII) during a gallop. Shown is strain energy density (SED) during that point in the stride in which peak strain is achieved (left), as well as the SED averaged over the entire stance phase of the stride (right). In the first case, SED ranges from a minimum of 600 Pa to a maximum of 56,000 Pa and, when averaged over the stride, from 461 to 51,375 Pa. Importantly, while the distribution of the peak and time averaged SED is very nonuniform, the manner in which the bone is loaded remains constant (i.e., the site of peak and minimal SED varies very little). Adapted from Gross et al. [20] uniformly about the cortex and through the diaphysis, even though the net change in bone strain caused by the absence of function varies widely [37]. The degeneration of cell:cell communication, which occurs with age, may contribute to the etiology of osteopenia [38]; even though the physical signals persist, the information super highway needs repaving. Evidence suggests that individual osteocytes also contain the ability to modulate the architecture of their individual lacunae, thereby “autoregulating” the mechanical environment to which they are exposed [71]. The bone cell would modulate its perception of strain by regulating the level of its attachment to the matrix, as well as the size of the periosteocytic space, thereby controlling the amount of direct deformation and/or shear to which it would be subjected. It is entirely feasible, and also biologically straightforward, to deem the level of strain in the matrix as irrelevant and instead focus solely on the amount and type of mechanical stimulus imposed on the bone cells. Accept for a moment that the osteocyte has a generic strain goal, similar at all sites through the skeleton. It does not really matter what that amount is, assume it is 100 . The osteocyte, regardless of whether the matrix is exposed to strains of 3000  in the tibia cortex or 300  in the mandibular trabeculae, can achieve the goal of 100  by increasing or decreasing its lacunar attachments rather than altering the mass and morphology of the bone. Near the neutral axis, the cell is highly coupled to the matrix such that very little strain influences cell metabolism in the same manner as a poorly coupled cell in a highly strained region of the cortex. With the osteocyte also capable of detaching from the matrix, through expression of collagenase, it is able to modulate its own mechanical microenvironment through the active coupling and uncoupling of its

membrane to the matrix [39]. In this manner, the osteocyte, identical in the mandible or tibia, would achieve its goal of 100  simply by modulating its own private space. This may also, in part, explain the “lazy zone” in bone (that strain region where an alteration in a mechanical signals fail to stimulate changes in bone architecture); the cell, not the tissue, is accommodating the new mechanical stimulus. In reality, this indicates that the bone can adapt without necessarily requiring the formation or resorption of tissue. Indeed, some exercise regimes that do not increase bone density per se may nevertheless improve the viability of the tissue, and thus improve the quality of bone, without influencing quantity.

C. Contribution of Muscle Dynamics to the Strain Environment That bone morphology ignores so many structural paradigms suggests that models of bone adaptation may rely too heavily on teleologic-based engineering concepts of structural optimization and safety, and should consider principles of evolution and physiology, which emphasize undirected development. Recognizing that bone is first a tissue and second (and quite conveniently) a structure, it is important to consider the biologic implications associated with biophysical stimuli. Indeed, tissue viability may depend on aspects of the mechanical environment that may not be at all rooted in minimal strain/minimal mass criteria, such as strain-dependent perfusion or strain-induced electrical currents. Alternatively, bone adaptation may depend on some camouflaged subset of the mechanical milieu, e.g., the mechanical strains induced by muscle.

494 While the symbiotic relationship between muscle and bone is obvious, only seldom is it explicitly considered in the context of one defining the other [40]. As muscle contraction imposes far smaller strains on the skeleton than those caused by ground reaction loads (e.g., impact), their role in defining bone morphology has not received much consideration. Although muscle-induced strains may indeed be relatively small, they are sustained for extended periods of time (e.g., in postural muscle activity), and thus , over time , may dominate a bone’s characteristic “strain history.” Examining this hypothesis, strain data from a variety of animals reveal the existence of a broad frequency range of strains in the appendicular skeleton, even during activity such as quiet standing [35]. While reaction forces due to locomotion give rise to large distinct strain components, spectral analysis of standing strain recordings shows significant strain information extending out to the frequency range of 50 cycles per second (Hz). While the magnitudes and frequency content of gait-related strains change transiently as a function of speed and gait, time-averaged strains (strain history) are dominated by standing strain spectra and are therefore quite stable over time and more uniform (Fig. 3). The spectral content of time averaged strain history, therefore, may better portray the wide range of strain information present in the functional milieu of the skeleton. From a stimulus standpoint, these persistent, low-amplitude signals may, when summed, be at least as important as the seldom occurring, and somewhat unpredictable, peak strain events [41]. Whether the skeleton is preferentially sensitive to a few, large strain events or a continual barrage of low magnitude events, must be evaluated at the tissue level, where specific mechanical signals can be introduced, and the resultant remodeling evaluated.

IV. MODULATION OF BONE TISSUE BY BIOPHYSICAL STIMULI It is clear that the skeletal organ is subject to a wide range of mechanical signals, including low to highfrequency strains, normal and shear strains, and compressive and tensile strains. It is also clear that the cells on and within the mineralized matrix are subject not only to mechanical parameters such as strain, but derivatives of tissue deformation such as fluid flow and electrokinetic currents, parameters that may represent an important physiologic pathway in mediating an adaptive response. Unfortunately, the study of bone at this organ level makes it difficult to categorically identify the osteogenic parameters of the biophysical milieu. Addressing the “mechanism” of signal transduction necessitates a move to the level of bone tissue.

CLINTON T. RUBIN

ET AL.

FIGURE 3

(a) A 2-min strain recording from the caudal longitudinal gage of the sheep tibia while the animal took a few steps with peak strains on the order of 200 . (b) A 20-s portion of that strain record shows peak strain events as large as 40 . (c) Further scaling down to a 3-s stretch of the strain recording illustrates events on the order of 5 . Adapted from Fritton et al. [35].

A. Identifying Osteogenic Parameters of the Strain Milieu The most efficient means of studying adaptive tissue responses to biophysical stimuli is through animal models in which the mechanical environment can be controlled accurately. Investigations designed to identify those components of the mechanical milieu that regulate skeletal adaptation have used a number of experimental approaches, including stress protection adjacent to implants of varied stiffness [42,43], overload caused by osteomy [18,44], and externally applied loading [45 – 47]. Although these applied loading experiments have contributed to our understanding of adaptation, they too have distinct limitations, the greatest of which is that the loads are applied for a limited, arbitrarily chosen period of time, yet for the remainder of the day the animals are able to apply uncontrolled, unmonitored loading to the bone under investigation.

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

495

To ensure that the adaptive modeling and remodeling that one observes is uniquely a product of those mechanical parameters that are applied, it becomes essential that the bone be exposed to minimal spurious loading events. These requirements have been met through several different animal models, ranging from the loading of cancellous bone in the distal femur of dogs [48] to the loading of tail vertebrae in rats [49]. In our laboratory, a primary means of studying the adaptation of cortical bone to biophysical stimuli has been the functionally isolated turkey ulna [31]. The advantage of this model is that the bone tissue is subject only to the mechanical [32] or electrical [50] regimen prescribed by the investigators, with no aberrant biophysical signals entering the preparation. In short, the ulna of adult male turkeys is functionally isolated by proximal and distal epiphyseal osteotomies, leaving the entire diaphyseal shaft undisturbed. Caps filled with methylmethacrylate are placed over the ends of the bones and percutaneous transfixing pins are placed through the caps. This model has demonstrated that 8 weeks of functional isolation alone will consistently result in a 10 – 15% loss of bone in 8 weeks, where an externally applied mechanical strain regimen, physiological in strain magnitude, can lead to significant increases in cross-sectional bone area depending on the signal parameters. Specifically, alterations in bone mass, turnover, and internal replacement are sensitive to changes in the magnitude [11] and distribution [51] generated within the bone tissue. Further, a loading regimen must be dynamic (time-varying) in nature; static loads do not influence bone morphology [52]. Moreover, the full osteogenic potential of a large amplitude (2000 ), low frequency (1 Hz) regimen is realized following only an extremely short ( 1 min) exposure to this stimulus [31].

response to be an appropriate interim strategy in the adaptive response to intense new mechanical challenges, we undertook a series of longer duration adaptation studies [54]. Following 16 weeks of a load regimen of only 100 cycles per day, inducing a peak compressive strain of 2000 , new bone stimulated in the turkey ulna model was lamellar, composed of primary and secondary osteons toward the original cortex and circumferential lamellae at the periphery, (Fig. 4, see also color plate). Remnants of the initial woven bone response seen at 4 weeks remained clearly visible at both 8 and 16 weeks as diffusely labeled interstitial elements within the newly formed lamellar construct. The presence of secondary osteons, circumferential lamellae, and an osteocyte density and organization similar to that seen in controls suggests that the presence of woven bone in the initial stages of the adaptive process is not

B. Long-Term Modeling Response to Mechanical Stimuli In the tissue level experiments discussed earlier the efficacy of any given mechanical stimulus to affect bone remodeling has typically been evaluated following 8 weeks of either disuse or externally applied loading. When a strongly osteogenic load regimen is applied, the magnitude and location of the adaptive response the consistent, yet the character of the periosteal response is often woven in nature and is not considered an ideal adaptive response under any conditions. If mechanically mediated modeling stimulated only woven bone, its potential use to combat osteoporosis would be diminished. To determine if this non-optimal tissue type persists over time (or disappears altogether), suggesting its appearance to be an aberrant reaction to surgery [53], or if lamellar bone replaces the woven response, emphasizing the woven

FIGURE 4

A fluorescent photomicrograph of the periosteal surface of a turkey ulna diaphysis following 8 (top) and 16 (bottom) weeks of a mechanical regimen sufficient to cause a peak of 2000 . The 8-week response shows consolidating primary bone. By 16 weeks, remnants of the original woven response can be seen serving as interstitial elements of primary and secondarily remodeled bone. In essence, the woven bone response has served as a strategic stage in the achievement of a structurally appropriate increase in bone mass. Reproduced with permission from Rubin et al. [54]. (See also color plate.)

496 necessarily a pathologic or transient reaction to injury, but instead may represent a normal – and strategic – stage in response to a potent mechanical stimulus. It also demonstrates that the response of bone to mechanical loading is not only anabolic, but that it produces lamellar bone, the “gold standard” of bone formation.

C. Differential Modeling/Remodeling to Distinct Components of the Strain Tensor Mechanical factors such as magnitude and duration are essentially “organ” level stimuli. Out of the widely diverse range of mechanical signals to which the tissue is exposed, it is essential to determine which components of this strain tensor (i.e., the complete strain state of the bone tissue) actually influence the metabolism of the osteocyte, osteoblast, or osteoclast to retain the status quo, initiate modeling, or turn on remodeling. While the strain tensor of the functional regimen is very complex, it can be described in general terms by two predominant components: dilatation (i.e., dilate; volume changes caused by hydrostatic stress) and deviatoric (i.e., deviate; shape changes caused by shear stress) parameters. If the control of bone adaptation demonstrates a differential response to discrete parameters of the mechanical milieu, the mechanisms that control bone morphology can be elucidated. This goal has been approached using the turkey ulna model of disuse osteopenia, in which the modeling and remodeling response was quantified following 4 weeks of either axial or torsional loading or disuse [55]. Each of the two load groups were subject to peak principal strains of 1000  (predominately normal strain in the axial case, and shear strain when subject to torsion). Of the three distinct groups, only disuse caused a significant change in gross areal properties as compared to controls (13% loss of bone). This suggests that both axial and torsional loading conditions are reasonable substitutes for the functional signals normally responsible for retention of bone mass, leaving the periosteal and endosteal envelopes unphased by disparate components of the strain tensor. The intracortical response, however, was found to depend strongly on the manner in which the bone was loaded. Disuse failed to increase the number of sites within the cortex actively involved in bone turnover (intracortical events), yet significant area was lost within the cortex due to a threefold increase in the mean size of each porotic site. Axial loading increased the degree of intracortical turnover as compared to intact controls, yet the average size of each porotic event remained identical to that of the control. Conversely, compared to the control, torsion elevated neither the number of porotic events, the area of bone lost from within the cortex, nor the size of the porotic event. It appears that bone tissue can readily differentiate between

CLINTON T. RUBIN

ET AL.

distinct components of the strain tensor, with strain per se necessary to retain coupled formation and resorption, deviatoric strain achieving this goal by maintaining the status quo, whereas dilatation elevates intracortical turnover, but retains coupling. These experiments suggest two critically important characteristics of bone cellular activity: modeling (surface) and remodeling (intracortical) activity are not necessarily coupled and the osteocyte population can differentiate between dilatational and deviatoric strains in the tissue. This suggests that the ability of the cell to distinguish between volumetric and shape changes is achieved through several distinct mechanisms; perhaps dilatation is sensed directly by the degree of coupling between cell and matrix, whereas deviatoric stresses, which cause fluid flow, influence cell activity via second messenger gradients. This hypothesis is supported by in vitro findings that confirm that bone cells perceive and respond, albeit differently, to both hydrostatic and shear stresses [56,57]. Most importantly, these data suggest that different components of the strain tensor have distinct regulatory roles. . . it is not the aggregate of strain per se that defines remodeling, but that independent components of the strain tensor have differential responsibilities in achieving and maintaining bone mass. Other biophysical factors derived from the strain environment may also be responsible for cellular signaling. Dynamic loading of bone tissue not only results in a dynamic stress – strain environment, but is also associated with other matrix-related events related to the fluid content within the porous space of bone tissue [58 – 60]. These interstitial flow phenomena are driven by gradients in tissue dilatation [17,61], which will be significantly more pronounced under conditions of axial loading than with torsional loads. Thus, observed differences between axial and torsional loading may be due to differences in fluid pressurization [62], flow-induced shear stresses [15], or strain-generated electric potentials [13,14] in the cellular microenvironment.

D. Osteogenic Potential of Low-Level Electric Fields Mechanical data described earlier demonstrate a distinct relationship of function to form. They even suggest a means by which the mechanical stimulus can be perceived and regulated, via the cell – matrix tethers facilitated by integrins and matrix proteins. Just the same, for an osteocyte entombed within the tissue matrix, these deformations are on the order of a few angstroms, suggesting that alternatives to the strain of the cell per se should be considered. One such mechanism for the coupling of mechanical deformation to cellular activity is the strain-induced movement of interstitial fluid in the bone, similar to the water flow through a sponge caused either by stretching or squeezing.

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

497

Whereas strain-induced fluid flow will certainly contribute to increased nutrient and metabolite transport, this fluid movement will also cause an electrokinetic interaction with the bone tissue. The potential role of electricity in the regulation of bone tissue was first considered over 40 years ago with the report of electrical currents being generated with the loading of dry bone [63]. The discovery of load-induced piezoelectric potentials in bone provided a means by which stress or strain could intrinsically alter the biophysical environment of the bone cell, and thus influence proliferation and differentiation [14]. This hypothesis became even more attractive when it was demonstrated that, in wet bone, two sources of electrical current existed in parallel; piezoelectric currents induced by the deformation of collagen and the relatively large electrokinetic currents (streaming potentials) produced by the strain-induced flow of charged constituents of extracellular fluids flowing through the charged matrix [58,60]. Measurement of the electric potentials generated by functional levels of strain shows that the average field intensities in bone are quite small, on the order of 1 V/ (one microvolt per microstrain [59]). Considering that the adult skeleton is seldom subject to strains exceeding 2000  [22], if endogenous fields are to influence bone morphology they must do so at field intensities below 2 mV/cm (2000 V/cm). This field intensity, in and of itself, is certainly sufficient to perturb the membrane potential of an osteoblast to the point of having some biologic effect [64]. However, as described previously, a large percentage of bone tissue is rarely subject to strains greater than 500  [22], yet bone mass is retained in these areas [17]. Therefore, if each cell at each site of the bone is responsible for its own assessment of the biophysical milieu in its adjacent matrix, field intensities well below the “threshold” of 1 mV/cm would somehow have to regulate bone cell metabolism. Because of their suspected role in regulating bone morphology, electric fields have been used in the clinic for the treatment of delayed fracture unions [65]. The majority of the signals used for pseudoarthroses, delayed union, and even avascular necrosis are complex pulsed electromagnetic fields (PEMFs), which utilize time-varying magnetic fields exogenously to induce electrical currents into the local tissue. These devices induce relatively high electric field intensities (10 mV/cm), with the energy distributed over a broad frequency range, from one to more than 1,000,000 Hz. To identify the osteogenic components of the PEMF, we investigated how changes in PEMF characteristics would affect the remodeling response. In the first series of experiments, electromagnetic fields were induced in the turkey ulna model of disuse osteopenia using Helmholtz coil pairs strapped to the wing of the animal [66]. Keeping the peak

of the magnetic field levels constant yet varying their rise time, both the PEMF waveshape and, correspondingly, the power of the induced electric field were changed. Exposing the turkey ulna preparation to 1 h per day, the maximum osteogenic effect was seen with rise times of 3.8 to 5.5 ms. In contrast to the 11% loss of bone caused by 8 weeks of disuse, these waveforms stimulated a 12% increase in bone area, a net benefit over disuse exceeding 20%. Rise times above or below this range were less effective in generating bone formation and, in some cases, were incapable of inhibiting bone loss. Because changes in magnetic field rise time alter both pulse duration and pulse width in a PEMF, we considered the possibility that the osteoregulatory efficacy of the PEMF (all for a given magnetic field amplitude) was related to the spectral distribution of energy in specific PEMF signals [67]. This correlation analysis indicated that even though the frequency content of the PEMFs spanned the 1 to 250-Hz range,the component of the field energy that correlated most strongly with the ability of the PEMF’s to stimulate new bone formation was that induced at frequencies below 75 Hz. The osteogenic capability of this low frequency bandwidth was evident even though less than 0.1% of the total PEMF energy was contained in this range. Subsequent studies have validated this suspected lowfrequency affinity, with sinusoidal fields induced in the range of 15 – 35 Hz appearing the most potently osteogenic [68]. For example, a 15-Hz magnetic field signal inducing electric field intensities on the order of only 1 to 10 V/cm is more effective in stimulating new bone formation than a PEMF signal inducing fields on the order of 1 – 10 mV/cm (Fig. 5). However, at frequencies below 15 Hz (and thus more aptly associated with locomotion) the osteogenic potential decreased dramatically, such that at 5 Hz and below, induced electric fields were incapable even of preventing disuse bone loss [28]. That electric fields in the 10- 100-Hz range can affect bone remodeling activity at intensities below 10 V/cm suggests that strains below 10  — if induced within this frequency range — can play an important role in mediating bone remodeling even though these strains would be 300-fold less than the peak strains induced during maximal activity (based on 1 V/ [59]). In essence, very low-magnitude, high-frequency physical signals persist in functionally loaded bone, and very low-magnitude, high-frequency electric fields are strongly osteogenic.

E. Osteogenic Potential of Frequency The bone tissue modeling/remodeling response is sensitive only to dynamic (time-varying) strains; static strains are ignored as a source of osteogenic stimuli [52]. This observation indicates that — rather than strain magnitude per

498

CLINTON T. RUBIN

ET AL.

FIGURE 5

Microradiographs of transverse sections of the ulna midshaft following 8 weeks of 1 h per day of various electric field regimens. An ulna subject only to a “dummy” coil, resulting in a 12% bone loss via intracortical and endosteal resorption (top left). An ulna isolated from function but subject to a 75-Hz simusoidal electric field inducing 10 V/cm (bottom left) showed little modeling or remodeling activity, with a net increase in bone mass of 3%. A signal of the same magnitude but induced at 15 Hz resulted in substantial new bone formation on both endosteal and periosteal surfaces, with little evidence of intracortical porosis, resulting in a 14% increase in bone area (top right). Reproduced with permission from McLeod and Rubin [68].

se — the total number of strain events, the number of strain events per unit time, or the strain rates involved in the loading regimen may be critical to bone mass and morphology. In cortical bone, 2000  induced at 0.5 Hz (cycles/second) maintains bone mass and achieves this with just four cycles of loading encompassing 8 per day [31]. Reducing this strain to 1000  at 1 Hz requires 100 cycles, and 100, to maintain bone mass [11]. Raising the loading frequency to 3 Hz, bone mass can be retained with 1800 cycles (600 of load) with peak-induced strains of only 800  [61]. With the same 600 per day loading regimen, only 200  is necessary to maintain cortical bone mass if the strain is applied at 30 Hz, a protocol employing 18,000 cycles of loading. When these 30-Hz mechanical signals are induced for 1 h per day (108,000 cycles), only 70  is necessary to inhibit bone loss. Plotted together, these data demonstrate that the sensitivity of cortical bone to mechanical loading goes up quickly with frequency (Fig. 6), and thus much lower strains are necessary to of maintain bone mass.

These data indicate that bone is preferentially sensitive to high-frequency mechanical stimuli. However, all the experiments discussed to this point were invasive in nature, using the functionally isolated turkey ulna preparation. Unless we are able to induce these osteogenic stimuli noninvasively, application to humans in the clinic will be greatly limited. Using whole body vibration as a means of inducing mechanical signals into the skeleton, our first studies again used turkeys, but rather than invasive studies on the ulna, these animals simply stood on a small oscillating plate [69]. Over a 2-month period, each animal was subjected to a 30-Hz sinusoidal vibration for 10 min each day, five days per week. Five animals were in each of four groups to test acceleration intensity of 0.1, 0.2, 0.3, and 0.9g (where 1.0g  earth’s gravitational field, or 9.8 ms2). Eight control animals remained unstimulated. When not being loaded, animals roamed their pens freely. At 0.3g, this stimulation induced approximately 5 on the cortical surface of the tibia. Dynamic indices of bone formation (labeled surface, LS; mineral apposition rate; MAR) were determined for all animals via pulsed, double labels of tetracycline.

499

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

FIGURE 6

The sensitivity of cortical bone tissue to mechanical strain increases with loading frequency. The plot indicates the area increase (mm2) in cortical bone measured for each additional 1  imposed on the turkey ulna, at each of five loading frequencies spanning the 1- to 60-Hz range. Another way of interpreting these data is to consider that 1/10th of the strain is necessary to maintain cortical bone mass if the strain is induced at 60 Hz, rather than 1 Hz. Trabecular bone is even more responsive to frequency. Adapted from Qin et al. [61]

FIGURE 7

Trabeculae within the proximal tibial metaphysis of the eight control animals showed 1.9% labeled surface, with no measurable mineral apposition rate, whereas trabeculae in the femoral head of controls showed 1.8% LS, with no detectable MAR (Fig. 7, see also color plate). These data demonstrate that the skeletons of these adult animals were in a state of low bone turnover. The LS in trabeculae following 8 weeks of stimulation demonstrated a linear dose response with an increase in vibration intensity, extending to 50.7% in the tibia and 44.2% in the femur at 0.9g. In contrast, the MAR, once turned on at 0.1g (1.47 m/day in the tibia, 1.36 in the femur), failed to increase further by increasing intensity. These results suggest that brief exposure to extremely small strains, two orders of magnitude below the peak strains experienced by the skeleton, if induced at sufficiently high frequency, can be as important to skeletal design as large magnitude strains seen only occasionally. Further, that the strains induced noninvasively were far less than those in the invasive protocols (e.g., the isolated turkey ulna) suggests that trabeculae may be even more sensitive to strain than cortical bone. Regardless, the influence of the mechanical milieu on skeletal morphology as it

Fluorescent photomicrographs of trabeculae within the proximal tibial metaphysis of two 1.5-year-old turkeys. (Left) Trabeculae were harvested from a control (nonoscillated) animal. (Right) Trabeculae harvested from a turkey subject to 30 days of 5 min per day of plate vibrations at 30 Hz, inducing accelerations on the order of 0.3g. The fluorescent labeling schedule is identical in each animal. While mineral apposition rate and labeled surface were essentially zero in the skeletally mature controls, these dynamic indices of formation rose to 41  3.4% of labeled surface in the loaded group, with MAR rising to 1.9  0.22 m per day (N  4). By small fluctuations in gravity, induced at the proper frequency, trabecular bone formation is stimulated noninvasively in the weight bearing skeleton. Magnification 40. Adapted from Rubin et al. [93]. (See also color plate.)

500 relates to posture, speaking, and chewing, while generating relatively small signals, may be quite dramatic, depending on how long you stand, how loudly you speak, or how much you eat. If the small strains induced by the musculature during activities such as posture, speaking, or chewing are critically important to the establishment and maintenance of the skeleton, if those signals change with age, a key regulatory signal may be lost. The short-term (8 weeks) studies on turkeys show that noninvasive, very low intensity mechanical stimuli are osteogenic if applied above 10 Hz. It is essential, however, to demonstrate that this strongly anabolic stimulus can influence not only the quantity, but the quality of the bone. To determine if long-term (12 month) stimuli will ultimately improve the structural status of the bone, we have used DXA, peripheral quantitative computed tomography (pQCT), and histology (static and dynamic histomorphometry, CT) to evaluate the skeletal effects of a low-magnitude mechanical stimulus [70]. Eighteen adult female sheep, 5–7 years of age, were randomized into two groups; experimental and untreated controls. For 20 min/day 5 day/week, experimental sheep stood constrained in a chute such that only the hind limbs were subject to a vertical ground-based vibration, oscillating at 30 Hz, to create peak–peak accelerations of 0.3g. When the animals were not being treated, they joined the controls in a pasture area. Following 1 year of stimulation, the animals were euthanized and the femora and ulnae removed. Compared to untreated controls, the bone mineral density (BMD) of the proximal femur in stimulated animals was 5.4% greater, but this difference was not significant (p  0.1). Although pQCT also failed to demonstrate a significant difference in the total density of the proximal femur ( 6.5%; p  0.1) when this assay was used to selectively evaluate cortical and cancellous bone at the lesser trochanter, a 34.2% increase in trabecular density was observed in mechanically stimulated sheep (p , 0.01). This effect was substantiated by undecalcified bone histology, which demonstrated substantial increases in trabecular bone volume and trabecular number, and sharp decreases in trabecular spacing (Fig. 8, see also color plate). This effect appeared highly selective for cancellous bone, as there were no significant changes in any cortical bone parameters. To determine if the stimulus had any benefit to bone “quality,” high-resolution three-dimensional models were made of 1-cm cubes of trabecular bone harvested from the medical condyle of the femur, composed of 300 microtomographic slices for each cube [71]. The trabecular bone pattern factor, an index of connectivity, was decreased by 24.2% in animals subject to the noninvasive stimulus (p  0.03), reflecting a significant increase in connectivity and thus an improvement in the quality of bone. True physical property measurements of the bone samples, as performed using ultrasound and material testing, substantiated these

CLINTON T. RUBIN

ET AL.

findings, showing that the elastic modulus and stiffness of the bone subject to the low-level mechanical stimulus increased by 12% ( p 0.05), and strength by 27% ( p 0.05) [72]. Indeed, both the quantity and the quality of bone were enhanced by an extremely low-level mechanical stimulus.

V. ATTENUATION OF WOLFF’S LAW BY SYSTEMIC DISORDERS As sensitive as the skeletal system appears to be to mechanical stimuli, it becomes somewhat distressing that osteopenia should become symptomatic at skeletal sites subject to the greatest mechanical demand; as bone mass diminishes, a given load will engender a greater strain. However, if the strain increases, why then does the bone disappear? Something must happen to the level of the signal (e.g., suppressed streaming potentials caused by ageinduced change in interstitial fluid viscosity), the perception of that signal by the cells within bone (compromised matrix/cell interaction), or the means of responding to those signals (attenuated cell metabolism). Evidence is building that systemic distress such as age [73], calcium deficiency [74], or endocrine imbalance [75] will dramatically affect the interaction of biophysical stimuli with the modeling/remodeling response. For example, mechanical signals that are osteogenic in the young skeleton fail to stimulate bone formation in the old skeleton. Further, the osteopenia caused by disuse is markedly pronounced when combined with calcium deficiency (Fig. 9). Finally, even the modus operandi of disuse in normal bone is affected in conditions such as endocrinopathy: instead of a few, large porotic events occurring in the cortex, a huge number of smaller events arise. Perhaps it is this attenuation of the response of bone to mechanical signals that could explain why exercise is not considered ideal osteoporosis therapy for the postmenopausal or aging population [76]. Indeed, the risk of increasing signal strength to induce a response in the frail skeleton may induce the very failure that one is trying to prevent. To examine the inability of older bone to perceive mechanical stimuli, a series of experiments were devised to quantify the osteocyte population’s perception of exogenous signals. Obviously, the amplitude of the strain itself does not diminish in the older skeleton, if anything, it gets larger as bone mass disappears. Perhaps instead it is the means by which the cells perceive this deformation as a regulatory signal that has been affected. This hypothesis has been addressed in the turkey ulna model using the in situ reverse transcript polymerase chain reaction (RTPCR). These experiments show that the mRNA expression of many matrix proteins (e.g., type I collagen, osteopontin) and cell adhesion molecules (e.g., integrins) is reduced in the osteocytes of aging cortical bone [77]. For example,

501

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

FIGURE 8 Following 1 year of extremely low-level mechanical stimulation, parameters of both static and dynamic histomorphometry demonstrated a significant benefit to both the quantity and the quality of bone from exposure to the biophysical stimulus, as compared to control animals (top). Fluorescent photomicrographs of a transverse section at the lesser trochanter of the femur showing more trabeculae, which are thicker, than control. (See also color plate.)

approximately 3.0% of the osteocytes were positively labeled for collagen mRNA in the 1-year-old bone, with only 1.2% labeled in the old bone. Such a small, but significant, change will hardly convince anyone of a new etiologic factor in osteoporosis. However, while 12.8% of the osteocytes in the 1-year-old bone showed positive labeling for osteopontin, this labeling dropped 84% to 2.1% of osteocytes in the old bone. Finally, positive labeling for integrin 3 dropped from 5.3% in young bone to 0.8% in old bone, an 85% decrease in message for this protein [78]. These data reflect a diminished cell – matrix interaction; while the matrix still experiences strain, the regulatory information never reaches the osteocyte. Measuring the mRNA abundance in osteocytes following the load of old bone supports this mechanism of signal perception. Subjecting the older bone to a load regimen of 2000  at 1 Hz,

which is strongly osteogenic in young adult bone, was insufficient to stimulate any new bone formation, and protein expression remained essentially unchanged from the suppressed levels observed in the intact old bone. However, changing the mechanical signal to a low-magnitude, highfrequency signal increased the levels of expression of these critical proteins and stimulated substantial new bone formation (14% over the animals’ intact control). A 500- regimen at 30 Hz increased the number of osteocytes labeled for mRNA message of integrins to 7.4%, 2% above the basal level of young bone. Further, the osteopontin message increased to 13.9%, a level equivalent to that of young bone. Finally, the collagen type I message increased to 92.4% of osteocytes with label, a 300  increase in active osteocytes. This upregulation of mRNA activity is observed as early as 3 days into the loading regimen, long before the surfaces

502

CLINTON T. RUBIN

ET AL.

modulate its coupling to the matrix and morphology of its lacunae [54], it is further evidence of the ability of the cell to autoregulate its own mechanical environment (i.e., striving for the “optimal” mechanical milieu of 100). Age or menopause-induced reduction in the presence of these proteins would diminish the ability of the cell to sense some aspects of the mechanical signals within the bone tissue. However, aged bone tissue clearly retains the ability to respond to its mechanical environment, as reflected by the robust response to a high-frequency, low-magnitude mechanical stimulus.

VI. CLINICAL APPLICATION OF BIOPHYSICAL STIMULI The observations that relatively high-frequency, lowmagnitude mechanical stimuli can influence bone formation and resorption suggest that this signal can be used clinically to inhibit or reverse osteopenia. Certainly, lowlevel biophysical signals are simpler and safer to impose into the skeletal system than large stimuli. Recalling Goldilocks, it is important to remember that, before implementing these stimuli for a disease that may require three decades of treatment, the signal should not be “too big or too small, but just right.”

A. Deterioration of Muscle Dynamics as an Etiologic Factor in Osteopenia

Microradiographs of a 100-m midshaft section of an intact turkey hen ulna fed a normal diet (top). Following a 6-week period of a low calcium diet, the intact ulna shows a substantial amount of cortical thinning, endosteal resorption, subendosteal cavitations, and intracortical resorption (middle). From the functionally isolated ulna of the same bird, a more pronounced degree of tissue loss occurs in a bone subjected to the same calcium insufficiency, but not protected by functional loading (bottom). While the biophysical regulation of bone mass and morphology occurs locally, the systemic state of the animal can have an overwhelming effect. Reproduced with permission from Lanyon et al. [74].

FIGURE 9

have begun to produce osteoid [78]. These data suggest that extracellular matrix proteins provide a means of translating the mechanical loading dynamics of the skeleton to the osteocyte population. Recalling the ability of the osteocyte to

The suppression of the cell – matrix interaction in aged bone has already suggested that strain signals will not be transduced efficiently in older bone tissue [77]. Before completely blaming the bone cells’ inability to perceive or respond to the strain environment, it is worth considering whether the aged skeleton is instead lacking a critical component of the functionally induced endogenous signal. As suggested from strain gage recordings from the appendicular skeleton, low-level, high-frequency strains arise directly from muscle dynamics [35]. Further, these persistent, low-magnitude strains have been shown to be strongly osteogenic and may represent a strong stimulus in defining the morphology of the skeleton [69 – 72]. Therefore, if there is an age (or menopause)-induced change in the dynamics of these muscle oscillations, it could be argued that bone mass may deteriorate because these muscle-based signals also attenuate. To determine the role of muscle dynamics in the etiology of osteopenia, the spectral characteristics of muscle activity as a function of age were obtained through measurements of muscle surface vibration [79]. During the contraction of a muscle, radial expansion of the individual fibers results in fiber collisions and the

503

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

production of muscle sound or acoustic vibrations of the muscle body. The frequency of these vibrations reflects the firing rate of the motor units and, correspondingly, the force output of the motor unit, at least up to 80% maximum voluntary contraction during isometric contraction. The acoustic vibrations normal to the surface of the soleus muscle were recorded in 40 volunteers (20 – 83 years) using a low mass accelerometer. Recordings were made from the left and right soleus, a principal leg muscle associated with posture, while the volunteer sat (passive) and stood (active). Recordings from each leg were averaged and integrated to obtain spectral power in the boundaries of 1 – 50, 1 – 25 and 25 – 50 Hz. Spectra obtained from these recordings show muscle activity in the frequency range above 20 Hz decreases by a factor of three in the elderly as compared to that seen in young adults (Fig. 10), a sarcopenia consistent with loss of fast oxidative type fibers. As the high-frequency components seen in bone strain almost certainly arise through muscle activity, the deterioration of the postural muscle contraction spectra with age would contribute to a decrease in the spectral content of strain above 20 Hz. From this perspective, it can be argued that the sarcopenia that parallels aging, may well prove to be a principal etiologic factor in osteoporosis, as this portion of the strain spectra is demonstrably osteogenic. While the association between loss of muscle dynamics and bone mass does not demonstrate a causal relationship, it provides support for the concept that the chronic activity of postural muscles may be a dominant force in controlling bone mass. Further, if aging leads to the loss of specific muscle fibers critical to the maintenance of bone mass, osteoporosis could presumably be inhibited by providing a “surrogate” for the lost spectral strain history.

B. Transmissibility of Ground Reaction Forces to the Appendicular and Axial Skeleton The nature of the weight-supporting skeleton facilitates the transmission of mechanical energy into bone tissue in a relatively direct manner. That the weight-bearing skeleton certainly subjects the skeleton to strain, a dynamic strain on the skeleton can presumably be induced by perturbing its effective gravity. In other words, the static strain in your weight-bearing skeleton would rise if you suddenly found yourself on Jupiter, with the percentage increase in strain rising proportionally to the present increase in g force. If this change in g force, and therefore the change in strain that it caused, was achieved at a frequency that was demonstrated to be osteogenic, a unique means of influencing bone mass becomes possible. Fortunately, moving 20 million people to Jupiter may not be necessary, as the strains necessary to influence bone mass appear to be so small.

FIGURE 10

Age-related changes in soleus muscle dynamics during postural activity. Whereas low-frequency (1 – 25 Hz) spectra are only slightly affected by age, high-frequency muscle dynamics (25 – 50 Hz) are reduced markedly in the elderly. If these higher frequency vibrations are the dominant source of the high-frequency, low-magnitude strains in bone, it could be argued that the pathogenesis of osteopenia is rooted in degenerative changes in the neuromuscular system rather than bone tissue per se. Adapted from Huang et al. [80].

The modulation of g force can be accomplished by placing the standing human on a platform which is made to oscillate at a specific frequency and acceleration [80]. The strains arising from dynamic alterations in g force would be transfered into the skeleton along a normal trajectory, ensuring that the stimulus is concentrated at those sites with greatest weight-bearing responsibility (e.g., femoral neck), yet weak at sites not subject to resisting gravity (e.g., cranium). While conceptually simple, it must be demonstrated that groundbased accelerations are indeed transmitted through the bones and joints of the lower appendicular skeleton; little is known of the transmissibility of ground-based vibration at frequencies above 12 Hz [81]. To rigorously establish the relationship between acceleration at the plate surface and transmission of acceleration through the appendicular and axial skeleton, accelerations were measured from the femur and spine of the human standing on a vibrating platform [82]. Force transmission to these bones was determined using accelerometers mounted on Steinman pins imbedded transcutaneously in the spinous process of L3 and the greater trochanter of the right femur of six volunteers. To determine damping as a function of posture, data were also collected from subjects while standing with bent knees. For a constant force input (18N), plate accelerations increased with frequency at both the femur and spinous process of L4 (Fig. 11). When the subject stood erect, a negligible loss of acceleration was observed in the femur and spine in the lower frequency bands, yet transmissibility fell off by as much as 40% when the frequency approached 40 Hz. When the subject was asked to stand with bent knees, transmissibility fell to below 20% at the femur and spine. Presumably, this is due at least in part to the uncoupling of the body segments, such that they are no longer working efficiently as a fixed,

504

CLINTON T. RUBIN

ET AL.

FIGURE 11

Accelerometer readings as a function of time as measured from the center of the resonant plate, as well as the vertical component of L4 and femur of a 52-kg female oscillating at 29 Hz (left). Plotting site-specific accelerations as a function of frequency (right), it is clear that as frequency increases, the transmissibility of the ground-based vibrations fall in both the hip and the spine. Around 30 Hz, about 85% reaches the femur. In both figures, the solid line is the plate, the dashed line is the spine, and the dotted line is the hip. Adapted from Rubin et al. [82].

stiff system. More importantly, these measurements confirm the ability of the standing adult skeleton to transmit a substantial fraction of ground accelerations to regions of the weightbearing skeleton most susceptible to osteopenic bone loss.

C. Inhibition of Post Menopausal Bone Loss by Extremely Low-Level Mechanical Stimlui In vivo animal work, as well as the design and development of a prototype device suitable for humans, led to a 1-year feasibility trial on a small cohort of women [84]. The ability of a low-magnitude (0.2g), high-frequency (30 Hz) mechanical stimulation to inhibit postmenopausal osteopenia was evaluated in a prospective, randomized, double blind, placebo-controlled clinical trial. Sixty-two women, 3 – 8 years postmenopausal, enrolled in the pilot study. Thirty-one women underwent mechanical loading of the lower appendicular and axial skeleton for two 10 min periods per day, induced via floor-mounted devices that produced the mechanical stimulus. Accelerations of 0.2g are just perceptible and no adverse reactions were reported. Thirty-one women received placebo devices and underwent daily treatment for the same period of time. DXA was performed on the spine (L1-4), right and left proximal femurs, and nondominant radius at baseline and at approximately 3,6, and 12 months. A full complement of DXA data was obtained for 56 of the patients (28 treatment, 28 placebo; six subjects dropped from the study for reasons not related to the device). In a post hoc analysis, a linear re-

gression of the means was used to show that lumbar spine bone mineral density (BMD) declined by 3.3% ( 0.83, n  28) in the placebo group compared to only 0.8% (

0.82, n  28) in the treated group (p  0.03), reflecting a 2.5% benefit of the biomechanical intervention. A 3.3% treatment benefit was observed in the trochanter region of the hip, with a 2.9% ( 1.2) loss observed in the placebo group, yet with a 0.4% ( 1.2) gain in the treated group (p  0.03). At the distal radius, no significant differences were observed as a function of time or between groups, emphasizing the influence to be local. Stratifying the results based on patient body mass index (BMI) (kg/m 2), end point analysis confirmed the relationship between svelte stature and a greater degree of osteoporosis [84,85]; subjects with BMI  24 lost 2.5% ( 0.6) BMD over the course of the year (Fig. 12), where those with a BMI 24 did not show any change over the 12 month period. This stratification also demonstrates the ability of mechanical stimulation to inhibit this bone loss in the group at greatest risk; in subjects with BMI  24 who were exposed to the mechanical stimulus, bone loss in the spine was not significantly different than zero ( 0.2 0.7). The 2.7% difference between placebo and treatment groups was significant at p 0.01. Treated subjects with BMI 25 showed no apparent affect of treatment, perhaps because there was no bone loss to inhibit. Overall, these results indicate the potential of a unique, non-invasive biomechanical therapy for osteoporosis, representing a nonpharmacologic means of inhibiting the rapid decline of bone mineral density that follows menopause.

505

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

absence of a key regulatory stimulus to the bone tissue. To a certain extent, osteopenia may arise not through the inability of bone cells to respond to mechanical stimuli, but rather through a deficiency of a regulatory signal caused by muscle wasting. Regardless, the osteogenic potential of biophysical stimuli clearly points to their potential as a unique intervention for disorders and injuries of the musculoskeletal system. Whether such biophysical intervention will supersede pharmaceutical prophylaxes is doubtful, but not impossible.

Acknowledgments FIGURE 12

Stratification based on body mass index shows that the lighter women (BMI  24) lost on the order of 2.5% bone from the spine over the course of the year. Those thinner women, when exposed to lowlevel mechanical stimulation, inhibited this loss (p  0.005). As importantly, women with a BMI greater than 24 lost no bone over the course of the year, and thus it was not possible to demonstrate the efficacy of treatment to inhibit a loss that was not occurring (p  0.36). Adapted from Rubin et al. [83].

This work has been kindly supported by grants from the National Institutes of Health (AR39278, AR41011, AR41040, and AR43498), The Whitaker Foundation, Exogen, Inc., and The National Science Foundation (PYI 865105). The authors are grateful for the contributions made by our colleagues, in particular Jack Ryaby, Susannah and Chris Fritton, Yan-Qun Sun, Yi-Xian Qin, Steven Bain, Ted Gross, Simon Turner, and Robert Recker.

VII. SUMMARY

References

The role of biophysical stimuli in the achievement and maintenance of a structurally appropriate bone mass is clear. Indeed, these factors critical to retaining an effective skeletal structure have great potential for direct clinical applications, such as in fracture healing [86,87] or osseointegration [88]. In contrast to systemic, pharmaceutical intervention such as estrogens, bisphosphonates, or calcitonin, the attributes of such biophysical prophylaxes are that they are native to the bone tissue, safe at low intensities, incorporate all aspects of the remodeling cycle, will ultimately induce lamellar bone, and the relative amplitude of the signal will subside as formation persists (self-regulating and self-targeting). However, the widespread use of biophysical stimuli in the treatment of skeletal disorders will undoubtedly be delayed until we achieve a better understanding of the mechanisms by which they act [89]. At the organ level, biophysical signals exist as a normal, physiologic component of the functional milieu; strain energy appears on the occipital ridge of the macaque, the femur of the lizard, and the metacarpal of the horse. In addition to the large amplitude strains typically associated with functional activity, a strain signal, far less than 10  in amplitude, arises through muscular activity in the frequency band of 10 to 50 Hz. This signal is present in the cranial, axial, and appendicular skeleton and persists essentially at all times including passive actions such as standing and speaking. Indeed, the sarcopenia that parallels the aging process, and more specifically, the attenuation of the 20- to 50-Hz spectral content of muscle contraction, suggests that the absence of these signals may also indicate the

1. J. Wolff, “The Law of Bone Remodeling” (Das Gesetz der Transformation der Knochen). Originally published by Verlag von August Hirshwald, Berlin, 1892. (English translation by P. Maquet and R. Furlong, published by Springer Verlag, Berlin, 1986]. 2. T. Robinson, C. Snow-Harter, D. Taaffe, D. Gillis, J. Shaw, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10(1), 26 – 35 (1995). 3. C. Snow-Harter, M. Bouxsein, B. Lewis, D. Carter, and R. Marcus, Effects of resistance and endurance exercise on bone mineral status of young women: A randomized exercise intervention trail. J. Bone Miner. Res. 7(7), 761 – 769 (1992). 4. Hargens, Recent bed rest results and countermeasure development at NASA. Acta Physiol. Scand. 150(S616), 103 – 114 (1994). 5. D. Simmons, J. Russel, F. Winter, R. Baron, A. Vignery, T. Van Thuc, G. Rosenberg, and W. Walke, Space flight and the non-weight bearing bones of the rat skeleton (Cosmos 1129). Trans. 27th Orthop. Res. Soc. 6, 65 (1981). 6. Y. Nishimura, H. Fukuoka, M. Kiriyama, Y. Suzuki, S. Oyama, S. Ikawa, M. Higurashi, and A. Gunji, Bone turnover and calcium metabolism during 20 days bed rest in young healthy males and females. Acta Physiol. Scand. 150(S616), 27 – 35 (1994). 7. H. Jones, J. Priest, W. Hayes, C. Ticheno, and D. Nagel, Humeral hypertrophy in response to exercise. J. Bone J. Surg. 59A, 204 – 208 (1977). 8. Huddleston, D. Rockwel, D. Kurlund, and J. Harrison, Bone mass in lifetime tennis athletes. J. Am. Med. Assoc. 244, 1107 (1980). 9. P. Engh, Massin, and K. Suthers, Roentgenographic assessment of the biologic fixation of poroussurfaced femoral components. Clin. Orthop. Rel. Res. 257, 107 – 128 (1990). 10. J. Callaghan, The clinical results and basic science of total hip arthroplasty with porous coated prostheses. J. Bone J. Surg. 75A(2), 299 – 310 (1993). 11. L. Rubin, Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 12. J. O’Conner, L. Lanyon, and H. MacFie, The influence of strain rate on adaptive bone remodelling. J. Biochem. 15, 767 – 781 (1982).

506 13. S. Pollack, D. Berretta, H. Steinberg, D. Steinberg, and C. Brighton, Osteoporosis and the electrical double layer. 6th Trans. Biol. Reprod. Growth Soc. (1986). 14. Bassett, Biologic significance of piezoelectricity. Calcif. Tissue Int. 1, 252 – 272 (1971). 15. S. Weinbaum, S. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. J. Biomech. 27(3), 339-360 (1994). 16. Fyhrie and D. Carter, A unifying principle relating stress to trabecular bone morphology. J. Orthop. Res. 4, 304 – 331 (1986). 17. T. S. Gross, J. L. Edwards, K. J. McLeod, and C. T. Rubin, Strain gradients correlate with sites of periosteal bone formation. J. Bone Miner. Res. 12(6), 982 – 988 (1997). 18. D. Carter, W. Harris, R. Vasu, and W. Caler, The mechanical and biological response of cortical bone to in vivo strain histories, In Mechanical Properties of Bone (S. Cowin, ed.) Vol. 45, p. 81. Am. Soc. Mech. Eng. 1981. 19. L. Lanyon and R. Smith, Bone strain in the tibia during normal quadrupedal locomotion. Acta Orthop. Scand. 41, 238 – 248 (1970). 20. T. Gross, K. McLeod, and C. Rubin, Characterizing the bone strain distributions in vivo using three triple rosette strain gages. J. Biomech. 25, 1081 – 1087 (1992). 21. C. Rubin, Skeletal strain and the functional significance of bone architecture. Calcif. Tissue Int. 36, 11 – 18 (1984). 22. C. Rubin and L. Lanyon, Dynamic strain similarity in vertebrates: An alternative to allometric limb bone scaling. J. Theor. Biol. 107, 321 – 327 (1984). 23. C. Rubin, K. McLeod, and S. Bain, Functional strains and cortical bone adaptation: Epigenetic assurance of skeletal integrity. J. Biomech. 23, 43 – 54 (1990). 24. S. Cowin, Bone remodeling of diaphyseal surfaces by torsional loads: Theoretical predictions. J. Biomech. 20, 1111 – 1120 (1987). 25. R. Huiskes, H. Weinans, H. Grootenboer, M. Dalstra, B. Fudala, and T. Slooff, Adaptive bone remodeling theory applied toprosthetic:Design analysis. J. Biomech. 20, 1135 – 1150 (1987). 26. Beaupre, T. Orr, and D. Carter, An approach for timedependent bone modeling and remodeling: Theoretical development. J. Orthop. Res. 8, 651 – 661 (1990). 27. M. Van Der Meulen, G. Beaupre, and D. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 28. C. Rubin and K. McLeod, Endogenous control of bone morphology via frequency specific, low magnitude functional strain, In “Human Biology” (H. Weinans, ed.), pp. 79 – 90. World Scientific, Singapore, 1995. 29. C. Rubin and L. Lanyon, Limb mechanics as a function of speed and gait: Functional strains in the radius and tibia of horse and dog. J. Exp. Biol. 101, 187 – 211 (1982). 30. Biewener, Biomechanics of mammalian terrestrial locomotion. Science 250, 1097 – 1103 (1990). 31. C. Rubin and L. Lanyon, Regulation of bone formation by applied dynamic loads. J. Bone J. Surg. 66A, 397 – 402 (1984). 32. C. Rubin and L. Lanyon, Osteoregulatory nature of mechanical stimuli: Function as a determinant for adaptive remodeling in bone. J. Orthop. Res. 5, 300 – 310 (1987). 33. C. Rubin and K. McLeod, Biologic modulation of mechanical influences in bone remodeling, In “Biomechanics of Diarthroidal Joints” (V. Mow, A. Ratcliffe, and S. Woo, eds.), Chap. 21, pp. 97 – 118. Springer-Verlag, New York, 1990. 34. T. Skerry, R. Suswillo, A. El Haj, N. Ali, R. Dodds, and L. Lanyon, Load-induced proteoglycan orientation in bone tissue in vivo and in vitro. Calcif. Tissue Int. 46, 318 – 326 (1990). 35. S. P. Fritton, K. J. McLeod, and C. T. Rubin, Self-similarity in longterm strain recordings from the skeleton of disparate species: Evidence for the self-organization in bone. J. Biomech. 33, 317 – 326 (2000). 36. H. Donahue, K. McLeod, C. Rubin, J. Andersen, E. Grine, E. Hertzberg, and P. Brink, Cell to cell communication in osteoblastic

CLINTON T. RUBIN

37. 38.

39.

40. 41.

42.

43.

44. 45. 46.

47.

48.

49.

50.

51.

52. 53. 54.

55.

56.

57.

ET AL

networks: Cell line dependent hormonal regulation of gap junctions composed of connexin43. J. Bone Miner. Res. 10, 881 – 889 (1995). T. Gross, and C. Rubin, Uniformity of resorptive bone loss induced by disuse. J. Orthop. Res. 13(5), 708 – 714 (1995) H. Donahue, K. McLeod, C. Rubin, J. Andersen, E. Grine, E. Hertzberg, and P. Brink, Cell to cell communication in osteoblast networks: Cell line dependent hormonal regulation of gap junctions composed of connexin 43. J. Bone Mineral Res. 10(6), 881 – 889 (1995). C. T. Rubin, Y-Q. Sun, M. Hadjiargyrou, and K. McLeod, Increased expression of matrix metalloproteinase-1 mRNA in osteocytes precedes bone resorption as stimulated by disuse: Evidence for autoregulation of the cell’s mechanical environment? J. Orth op. Res. 17, 354 – 361 (1999). D. Carter and M. Wong, Mechanical stresses and endochondral ossification in the chondroepiphysis. J. Orthop. Res. 6, 148 – 154 (1988). D. Adams, A. Spirt, T. Brown, S. Fritton, C. Rubin, and R. Brand, Testing the “daily stress stimulus” hypothesis of bone remodeling with natural and experimentally controlled strain histories. J. Biomech. 30, 671 – 678 (1997). S. Woo, W. Akeson, R. Coutts, L. Rutherford, D. Doty, G. Jemmott, and D. Amiel, A comparison of cortical bone atrophy secondary to fixation with plates with large differences in bending stiffness. J. Bone J. Surg. 58A, 190 – 195 (1976). D. Carter, W. Harris, R. Vasu, and W. Caler, Stress fields in the implated and plated canine femur calculated from in vivo strain measurements. J. Biomech. 14, 63 – 70 (1981). L. Lanyon, A. Goodship, C. Pye, and J. MacFie, Mechanically adaptive bone remodeling. J. Biomech. 15(3), 141 – 154 (1982). J. Hert, M. Liskova, and J. Landa, Continuous and intermittent loading of tibia in rabbit, Folia Morphol. (Prague). 19, 290 – 317 (1971). A. Churches, C. Howlett, K. Waldron, and G. Ward, The response of living bone to controlled time varying loading. J. Biomech. 12, 35 – 45 (1979). C. Turner, M. Akhter, D. Raab, D. Kimmel, and R. Recker, A noninvasive, in vivo model for studying strain adaptive bone remodeling. Bone 12, 73 – 79 (1991). S. Goldstein, L. Matthews, J. Kuhn, and S. Hollister, Trabecular bone remodeling: An experimental model. J. Biomech. 24, 135 – 150 (1991). T. Chambers, M. Evans., T. Gardner, A. Turner-Smith, and J. Cho, Induction of bone formation in rat tail vertebrae by mechanical loading. Bone Miner. 20, 167 – 178 (1993). C. Rubin, H. Donahue, J. Rubin, and K. McLeod, Optimization of electric field parameters for the control of bone remodeling: Exploitation of an indigenous mechanism for the prevention of osteopenia. J. Bone Miner. Res. 8, S573 – S581 (1993). T. Brown, D. Pedersen, M. Gray, R. Brand, and C. Rubin, Identification of mechanical parameters initiating periosteal remodeling. J. Biochem. 23, 893 – 905 (1990). L. Lanyon and C. Rubin, Static versus dynamic loads as an influence on bone remodeling. J. Biomech. 17, 897 – 906 (1985). H. Frost, Perspectives: Bone’s mechanical usage windows. Bone Miner. 19, 257 – 271 (1992). C. Rubin, T. Gross, K. McLeod, and S. Bain, Stimulation of lamellar bone formation by potent mechanical stimuli: Morphologic stages in the achievement of lamellar bone. J. Bone Miner. Res. 10(3), 488 – 495 (1995). C. Rubin, T. Gross, Y. Qin, S. Fritton, F. Guilak, and K.J. McLeod, Differentiation of bone-tissue remodeling response to axial and torsional loading in the turkey ulna. J. Bone Jt. Surg. 78, 1523 – 1533 (1996). E. Burger, J. Klein-Nulend, and J. Veldhuuzen, Modulation of osteogenesis in fetal bone rudiments by mechanical stress in vitro. J. Biomech. 24 (Suppl. 1), 101 – 109 (1991). P. Davies and S. Trepathi, Mechanical stress mechanisms and cell: An endothelial paradigm. Circ. Res. 27, 239 – 245 (1993).

CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention

507

58. Y. Mak, Streaming potential in bone. Exercise Sports Sci. 17, 175 – 194 (1989). 59. M. Otter, V. Palmieri, D. Wu, K. K. Siez, L. MacGinitie, and G. Cochran, A comparative analysis of streaming potentials in vivo and in vitro. J. Orthop. Res. 10, 710 – 719 (1992). 60. R. Salzstein, S. Pollack, A. Mak, and N. Petrov, Electromechanical potentials in cortical bone. 1. A continuum approach. J. Biochem. 20(3), 261 – 270 (1987). 61. Y.-X. Qin, C. Rubin, and K. McLeod, Nonlinear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. J. Orthop Res. 16, 482 – 489 (1998). 62. Y. Zeng, S. Cowin, and S. Weinbaum, A fiber matrix model for fluid flow and streaming potentials in the canaliculi of an osteon. Ann. Biomed. Eng. 22, 280 – 292 (1994). 63. A. Yasuda, Piezoelectric activity of bone. J. Jpn. Orthop. Surg. Soc. 28, 267 (1954). 64. K. McLeod, H. Donahue, P. Levin, M. Fontaine, and C. Rubin, Electric fields modulate bone cell function in a density dependent manner. J. Bone Miner. Res. 8(8), 977 – 984 (1993). 65. A. Bassett, Beneficial effects of electromagnetic fields. J. Cell. Biochem. 51, 387 – 393 (1993). 66. C. Rubin, K. McLeod, and L. Lanyon, Prevention of osteoporosis by pulsed electromagnetic fields. J. Bone J. Surg. 71A(3), 411 – 418 (1989). 67. K. McLeod and C. Rubin, Frequency specific modulation of bone adaptation by induced electric fields. J. Theor. Biol. 145, 385 – 396 (1990) 68. K. McLeod and C. Rubin, The effect of low-frequency electrical fields on osteogenesis. J. Bone J. Surg. 74A, 920 – 929 (1992). 69. C. Rubin, C. Li, Y. Sun, C. Fritton, and K. McLeod, Non-invasive stimulation of trabecular bone formation via low magnitude, high frequency strain. 41st Orthop. Res. Soc. 20(2), 548 (1995). 70. C. Rubin, S. Turner, S. Bain, C. Mallinekrodt, and K. McLeod, Extremely low-level mechanical signals are anabolic to trabecular bone. Nature (in press) (2001). 71. C. Rubin, A. S. Turner, R. Muller, Y-X. Qin, and K. McLeod, Femoral bone density and trabecular number are increased by non-invasive low level mechanical stimuli, as quantified by micro-computed tomography. Trans. Orthop. Res. Soc. 25, 715 (2000). 72. Y. X. Qin, R. Mauser, C. Berman, W. Lin, and C. Rubin. The relationship between bone mineral density and ultrasonic velocity. Trans. Orthop. Res. Soc. 25, 568 (2000). 73. C. Rubin, S. Bain, and K. McLeod, Suppression of the osteogenic response in the aging skeleton. Calcif. Tissue Int. 50, 306 – 313 (1992). 74. L. Lanyon, C. Rubin, and G. Baust, Modulation of bone loss during calcium insufficiently by controlled dynamic loading. Calcif. Tissue Int. 38, 209 – 216 (1986). 75. S. Bain and C. Rubin, Metabolic modulation of disuse osteopenia: Endocrine-dependent site specificity of bone remodeling. J. Bone Miner. Res. 5(10), 1069 – 1075 (1990).

76. G. Rodan, Mechanical loading, estrogen deficiency, and the coupling of bone formation to bone resorption. J. Bone Miner. Res. 6(6), 527 – 530 (1991). 77. Y-Q. Sun, K. McLeod, and C. Rubin, Mechanically induced periosteal bone formation is paralleled by the upregulation of collagen type one mRNA in osteocytes as measured by in situ reverse transcript-polymerase chain reaction. Calcif. Tissue. Int. 57, 456 – 462 (1995). 78. C.T. Rubin, J. Zhi, and M. Hadjiargyrou, Expression of a novel and a known gene, upregulated by disuse, is downregulated by anabolic mechanical stimulation. J. Bone Miner. Res. 14(S1), S522 (1999). 79. Y-Q. Sun, K. J. McLeod, and C. Rubin, Rapid enhancement of coupling of osteocyte to matrix. Trans. 42nd. Ann. Mtg. Ortho. Res. Soc. 21(2), 267 (1996). 80. R. Huang, K. McLeod, and C. Rubin, Changes in the dynamics of muscle contraction as a function of age. A contributing factor to the etiology of osteoporosis? J. Geronto 54A(8), B352 – B357 (1999). 81. J. Fritton, C. Rubin, Y-X. Qin, and K. McLeod, Whole body vibration in the skeleton: Development of a resonance-based testing device. Ann. Biomed Eng. 25(5), 831 – 839 (1997). 82. M. Griffin, “Handbook of Human Vibration.” Academic Press, San Diego, (1990). 83. C. Rubin, K. McLeod, M. Pope, M. Magnysson, M. Rostedt, C. Fritton, and T. Hansson, Transmissibility of ground vibration to the axial and appendicular skeleton: An alternative strategy for the treatment of osteoporosis. 18th Am. Soc. Biomech. 79 – 80 (1994). 84. C. Rubin, R. Recker, D. Cullen, J. Ryaby, and K. McLeod, Prevention of bone loss in a postmenopausal population by low-level biomechanical intervention. Am. Soc. Bone Miner. Res. 23, 1106 (1998). 85. J. Aloia, and E. Flaster, Estimating the risk of fracture in osteopenic patients. Endocrinologist 5, 297 – 402 (1995). 86. P. Martin, C. Verhas, C. Als, I. Geerts, J. Paternot, and P. Bergmann, Influence of patient’s weight on measurements of bone mineral density. Osteopor. Int. 3, 199 – 203 (1993). 87. J. Heckman, J. Ryaby, J. McCabe, J. Frey, and R. Kilcoyne, Acceleration of tibial fracture healing by noninvasive low intensity ultrasound. J. Bone J. Surg. 76A(1), 26 – 34 (1994). 88. A.E. Goodship, T. Lawes, and C.T. Rubin, Low magnitude high frequency mechanical stimulation of endochondral bone repair. Trans. 43rd Orthop. Res. Soc. 22(1), 234 (1997). 89. C. Rubin and K. McLeod, Promotion of bony ingrowth by frequency specific, low amplitude mechanical strain. Clin. Orthop. Rel. Res. 298, 165 – 174 (1993). 90. J. Rubin, C. Rubin, and K. McLeod, Biophysical modulation of cell and tissue structure and function. Crit. Rev. Eukar. Gene Expr. 5, 177 – 191 (1995). 91. C. Rubin, T. Gross, H. Donahue, H. Guilak, and K. McLeod, Physical and environmental influences on bone formation. In “Bone Formation and Repair”, pp. 61 – 78. Amer. Acad. Orth. Surg. (1994).

CHAPTER 19

Biomechanics of Age-Related Fractures MARY L. BOUXSEIN

Orthopedic Biomechanics Laboratory, Department of Orthopedic Surgery, Beth Israel Deaconess Medical Center, and Harvard Medical School, Boston, Massachusetts 02215

I. Introduction II. Biomechanics of Bone: Basic Concepts and Age-Related Changes III. Biomechanics of Hip Fractures

IV. Biomechanics of Vertebral Fractures V. Summary and Clinical Implications References

I. INTRODUCTION

to the forces applied to the bone, as well as to its loadbearing capacity, are important determinants of fracture risk. In support of this concept, clinical studies have repeatedly shown that factors related to skeletal fragility, as well as to the loads applied to the skeleton, are important determinants of fracture risk [4 – 9]. Insight into the relative contributions of skeletal fragility versus skeletal loading may be gained by using a standard engineering approach for evaluating the risk of structural failure. To design a structure, engineers must consider the size and geometry of the structure, the materials from which it is to be made, and the types of loads to which it will be subjected. Using this information, the loads applied to the structure during its normal usage can be compared to the loads known to cause failure. This comparison of applied load versus failure load gives an estimate of how “safely” the structure is designed. If a structure’s design appears “unsafe,” it may be necessary to change the geometry of the structure (e.g., increase its size), use stronger materials, or reduce the applied loads. In practice, it is often difficult to estimate precisely the strength of a structure and the loads applied to it. Therefore, to reduce the likelihood of

Age-related fractures represent an immense and increasing public health issue. In the United States alone, there are an estimated 1.5 million fractures annually, with associated medical expenditures of nearly $14 billion [1,2]. Based on current demographic trends, the number of fractures and their associated costs are projected to double or triple in the near future [3]. Most importantly, the consequences of these fractures are enormous, as those who suffer fractures experience increased mortality rates, chronic pain and disability, and a decreased quality of life. Strategies designed to prevent fractures must be based on a sound understanding of their etiology. From an engineering viewpoint, fractures of any type are due to a structural failure of the bone. This failure occurs when the forces applied to the bone exceed its load-bearing capacity. The load-bearing capacity of a bone depends primarily on the material that comprise the bone (and its corresponding mechanical behavior), the geometry of the bone (its size, shape, and distribution of bone mass), and the specific loading conditions (Fig. 1). Thus it is clear that factors related

OSTEOPOROSIS, SECOND EDITION VOLUME 1

509

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

510

MARY L. BOUXSEIN

FIGURE 1 Characteristics of the spine that determine the capacity to carry load: The trabecular bone (left); the design and organization of the vertebral body (middle); and the loading conditions (right), which are illustrated as lifting in this figure, but could be any loading action. From Myers and Wilson [159].

unexpected failure, structures are often designed with very high safety factors. To apply these concepts in the study of the etiology of fractures, Hayes and colleagues introduced a parameter called the “factor of risk” [10]. The factor of risk, , is defined as the ratio of the load delivered to a bone (applied load) to the load-bearing capacity of that bone (failure load):   applied load / failure load. Thus, when the factor of risk is low (1), the forces applied to the bone are much lower than those required to fracture it, and the bone is at low risk for fracture. However, when the factor of risk is high (1), fracture of the bone is predicted to occur. A high factor of risk can occur either when the bone is very weak and its load-bearing capacity is compromised or when very high loads, such as those resulting from trauma, are applied to the bone. In elderly individuals, it is likely that the coupling of a weak bone with an increased incidence of traumatic loading leads to the dramatic rise in fracture incidence with age [11,12]. To apply the factor of risk concept in studies of hip and vertebral fracture, the loads applied to the bone of interest and the corresponding load required to fracture the bone must be identified. For example, the majority of hip fractures are associated with a fall. Therefore, to compute the factor of risk for hip fracture due to a fall, information about the loads applied to the femur during a fall and about the load-bearing capacity of the femur in a fall configuration is required. While this approach is relatively easy to

conceive, in practice it is difficult to apply. There are surprisingly few data describing the magnitude and direction of loads applied to the skeleton during the activities of daily living and even fewer data describing the loads engendered during traumatic events, such as a trip, slip, or fall. Moreover, due to the complex morphology of the skeleton and associated muscle and tendon attachments, it is difficult to design a laboratory study that mimics the loading environment encountered by the bone in vivo. Therefore, it is challenging to determine the load-bearing capacity of skeletal elements under realistic loading conditions. Moreover, because these are “biologic structures,” both applied loads and structural capacity can change with aging, pharmacologic intervention, and disease. Nevertheless, despite these uncertainties and limitations, rough estimates of the factor of risk for hip and vertebral fracture can be derived to provide insights into the complex roles of loading severity and skeletal fragility in the etiology of age-related fractures. This chapter reviews clinical and laboratory studies related to the biomechanics of age-related fractures. It first presents basic concepts related to the biomechanics of bone, including a summary of the factors that determine the material and structural behavior of bone. It then evaluates the roles of skeletal loading and bone fragility as they relate to hip and spine fractures. These sections discuss the factors that are related to the loads applied to the skeleton, either through traumatic events or everyday activities; the factors that are related to the structural capacity of skeletal elements; and how these factors interact to influence fracture risk.

CHAPTER 19 Biomechanics of Age-Related Fractures

511

II. BIOMECHANICS OF BONE: BASIC CONCEPTS AND AGE-RELATED CHANGES

return to its original shape with no residual deformation. The slope of the load – deformation curve in this elastic region defines the structural stiffness (or rigidity) of the bone. In contrast to a bone’s behavior in the elastic region, beyond the yield point, the bone undergoes permanent deformation and will not return to its original shape even when the load is removed completely. At this point, the bone is said to be in the plastic region. If the load continues to increase, the ultimate or failure load is reached, after which the structure often fails catastrophically. The energy required to produce structural failure (“yield”) is computed as the area under the load – displacement curve and is sometimes referred to as the work to fracture. To determine the mechanical behavior of bone material itself, the geometry of the specimen must be accounted for. Thus, mechanical tests are conducted on specimens of a standardized geometry under controlled conditions. As the load is applied, the specimen deforms and internal forces are generated within the specimen. The resulting relative deformation at any point is called the strain at that point. The “intensity” of the internal forces is referred to as the stress at that point. The material properties are analogous to the structural properties discussed earlier except that the properties are determined from a plot of stress versus strain instead of load vs deformation. In practice, the load – deformation curve can be converted to a stress – strain curve by correcting for specimen geometry by applying appropriate formulas for converting load to stress and deformation to strain. For example, for a specimen loaded in compression, stress is equal to the applied load divided by the crosssectional area of the specimen and strain is equal to the deformation divided by the original length of the specimen (Fig. 3). The resistance of the material to deformation is described by the elastic (or Young’s) modulus, defined as the slope of the stress – strain curve in the elastic region. As the load is increased, the specimen undergoes permanent deformation and begins to yield. If the load is increased beyond the yield point, the specimen will eventually fail, at which point the strength or ultimate stress and ultimate (or failure) strain can be determined. The biomechanical property termed toughness reflects the amount of work per unit volume of material required to yield or fracture the specimen and can be computed as the area under the stress – strain curve. Tough bone will be more resistant to fracture, although it may yield at a lower stress and, according to that measure, be considered weaker. In addition, examining the pre- and postyield regions of the stress – strain curve may provide information regarding the tendency of the bone material to accumulate damage and the mechanisms underlying its failure. A material that fractures soon after yielding, and therefore sustains little postyield strain before fracture, is termed brittle. In contrast, a material that sustains relatively large postyield strains before fracturing is considered ductile.

A. Structural vs Material Behavior To understand the nature of skeletal fragility, it is important to distinguish between factors that affect the mechanical behavior of a whole bone as a structure (structural behavior) and those that affect the mechanical behavior of the bony tissue itself (material behavior) [13]. In general, structural properties are determined by both the size and the shape of the bone, along with the mechanical properties of the tissue that comprise the bone. As such, a bone’s structural properties are in large part determined by the amount of bone present. In comparison, the material properties of bone tissue are independent of specimen size and shape, thereby reflecting the intrinsic characteristics of the bony tissue itself. During any physical activity, a complex distribution of forces, or loads, is applied to the skeleton, and with the imposition of these forces, the skeleton undergoes deformations. It is this relationship between the forces applied to a bone and the resulting deformations — characterized by a load – deformation curve — that define the structural behavior of the whole bone (Fig. 2). The load – deformation curve reflects the amount of load needed to produce a unit deformation. As mentioned earlier, the shape of this curve depends on the size/shape of the bone, as well as the properties of the tissue that comprise it. Generally, load and deformation are linearly related until the yield point is reached, at which time the slope of the load – deformation curve is reduced. Before the yield point, the bone is considered to be in the elastic region and, if unloaded, would

FIGURE 2

The load vs deformation plot is used to describe the structural behavior of a specimen. The elastic region is distinguished from the plastic region by the yield region. In the elastic region, when the load is removed there will be no residual deformation and the bone will return to its original shape. In contrast, in the plastic region, the bone will undergo permanent deformations that will remain even if the load is removed.

512

MARY L. BOUXSEIN

FIGURE 3 Typical stress versus strain diagram for longitudinally (L) and transversely (T) oriented specimens of cortical bone from the diaphysis. For specimens tested in compression, load and displacement are converted to stress and strain by dividing by the cross-sectional area and original length of the specimen, respectively. The figure shows the inherent anisotropy in bone, as specimens testing in the longitudinal direction are significantly stronger than those tested in the transverse direction.

The elastic properties of isotropic materials, such as steel or rubber, are the same in all directions. The elastic properties of bone, however, depend on the orientation of the material with respect to the loading direction. Materials whose elastic properties are sensitive to loading direction are referred to as anisotropic materials. For example, cortical bone from the femoral diaphysis has a higher modulus and is stronger when loaded in the longitudinal direction than when loaded in the transverse direction [14 – 16] (Fig. 3). The anisotropic nature of bone reflects its function as a load-bearing structure, as it is generally strongest in the primary loading direction. Hence, the degree of anisotropy in bone varies with anatomical site and functional loading [17 – 19]. For instance, human trabecular bone from the vertebral body is much stronger in the vertical direction than in the transverse direction [20 – 22], yet trabecular bone from the iliac crest and central femoral head are nearly isotropic [18,23]. In a heterogeneous material such as bone, the definition of material properties is not straightforward. In describing the properties of bone as a tissue, one could consider the mechanical properties of single trabeculae, the calcified bone matrix, or small specimens of cortical or trabecular bone. For purposes of this review, we consider bone “material” to include the calcified bone matrix, the marrow spaces in trabecular bone, and Haversian and Volkmann’s canals in cortical bone [24]. With this approach, we take a continuum mechanics view of bone in that the specimen is small enough to be homogeneous (uniform), but large enough to include a sufficient number of trabeculae (for trabecular bone) or osteons (for cortical bone) to characterize the overall material behavior.

B. Age-Related Changes in the Material Properties of Bone The elastic modulus and ultimate strength of cortical [25 – 32] and cancellous [21,22,33 – 37] bone decrease with increasing age in both men and women. In human cortical bone from the femoral middiaphysis, the tensile and compressive strengths (Fig. 4) and elastic moduli decrease approximately 2% per decade after age 20 [25]. In addition, the incurred deformation and energy absorbed before

FIGURE 4

Age-related changes in the ultimate stress of human femoral cortical bone in tension and compression (error bars represent 1 SD). The mean change per decade is -2.1% for tension and -2.5% for compression. These data indicate that femoral cortical bone becomes weaker with age. Data from Burstein et al. [25].

513

CHAPTER 19 Biomechanics of Age-Related Fractures

is characterized by a decline in the apparent density of cancellous bone. It appears that the amount of bone is reduced and therefore the integrity of the trabecular network is compromised; however, the remaining bone is histologically normal. Relevant to this is the observation, first reported by Carter and Hayes [43,44] and later confirmed by others [36,45 – 47], of a nonlinear relationship between bone density and strength (Fig. 5), whereby a given change in bone density leads to a relatively greater change in bone strength. In support of this, Mosekilde and colleagues conducted a series of experiments demonstrating an age-related decline in the mechanical properties of vertebral trabecular bone [21,22,34,36,40,41]. In one study [36], they found that the ash density declines approximately 50% from ages 20 to 80, whereas the material properties (compressive elastic modulus, ultimate stress, and energy to failure) decrease approximately 75 – 90% (Table 1). In trabecular bone of the proximal tibia, an age-related decline in apparent density of 25% is accompanied by a 30 – 40% reduction in compressive strength and energy absorption properties [37]. In addition, the strength anisotropy of trabecular bone from human lumbar vertebrae increases with age, as the ratio of compressive strengths of vertically and horizontally loaded specimens increases from about 2 at age 20 to 3.5 at age 80 [20,21]. This observation may reflect age-related changes in the trabecular architecture of vertebral bodies, whereby horizontally oriented trabeculae thin and disappear to a greater extent than vertically oriented trabeculae [36,40,48,49]. Finally, it is important to note that in trabecular bone specimens from the iliac crest that were matched pairwise for density, yield stress was approximately 40% lower in specimens from older donors (60 years) compared with younger donors (40 years) [39]. These data suggest that factors other than decreased bone density may

FIGURE 5 Compressive modulus as a function of apparent density for trabecular bone specimens from a wide variety of species and anatomic sites. In general, the modulus varies as a power-law function of density, with an exponent of approximately two. From Keaveny and Hayes [203], with permission.

fracture decrease approximately 5 – 12% per decade, suggesting the bone becomes more brittle and less tough with increasing age [25,31,32]. Moreover, the energy required to fracture a cortical bone specimen under impact loading decreases three-fold between the ages of 3 and 90 [38]. These changes in the elastic and ultimate properties of cortical bone are likely the result of porosity increases with age. McCalden and colleagues [31] found that age was strongly correlated with porosity (r  0.73) and that porosity explained over 75% of the variability in cortical bone strength. In summary, age-related changes in cortical bone lead to a weaker, more brittle material. Human cancellous bone exhibits a similar age-related decline in material properties [21,22,34 – 37,39 – 42]. Aging TABLE 1

Age-Related Changes in Vertically Oriented Trabecular Bone Specimens Compressed in Either the Vertical or Horizontal Directiona Specimens compressed in vertical direction % per decade

Specimens compressed in horizontal direction

Correlation with age (r)

% per decade

Correlation with age (r)

8.7

0.85b

8.7

Not reported

Ultimate stress

12.8

0.79

15.5

0.87c

Elastic modulus

13.5

0.83

15.9

0.83c

Energy to failure

14

0.75

c

15.2

0.88c

0.45b

3.1

0.30d

Ash density

Ultimate strain

4

c c

a The mean percentage change per decade and the linear correlation with age are presented. Specimens were taken from 42 persons, aged 15 to 87. Data from Mosekilde et al. [36]. b p0.01. c p0.001. d 0.05p0.06.

514

MARY L. BOUXSEIN

contribute to the age-related decline in material properties of trabecular bone [46,47,50,51].

C. Factors That Influence the Mechanical Behavior of Bone as a Material Whereas the age-related changes in the material properties of cortical and trabecular bone are influenced by many factors, the major determinant of the mechanical properties of bone is its porosity or apparent density. The mechanical properties of cortical bone are strongly related to porosity and the degree of matrix mineralization [52 – 56]. Over 80% of the variation in the elastic modulus of cortical bone can be explained by a power – law relationship with mineralization (defined as calcium content) and porosity as explanatory variables [52,53]. In general, with increasing age, the degree of mineralization of the matrix increases, leading to stiffer, but more brittle material behavior [26,57]. The material properties of cancellous bone are also determined to a great extent by bone density. As mentioned previously, power – law relationships with bone density as the explanatory variable explain 60 – 90% of the variation the modulus and strength of cancellous bone [36,43 – 46]. These power – law relationships indicate that small changes in apparent density can lead to dramatic changes in mechanical behavior. For instance, a 25% decrease in apparent density, approximately equivalent to 15 – 20 years of agerelated bone loss [58,59], would lead to a 44% decrease in the strength of cancellous bone. Given the anisotropic nature of trabecular bone and the variation in predicted modulus for a given density [39,60], it is clear that density alone cannot explain all of the variability in the mechanical behavior of trabecular bone. Empirical observations and theoretical analyses indicate that trabecular architecture plays an important role in determining the mechanical properties of trabecular bone [46,47,50]. Trabecular architecture can be characterized by the thickness, number, and separation of the individual trabecular elements, as well as the extent to which these elements are interconnected. Advances in nondestructive, high-resolution imaging techniques have provided new insights into the relative influence of architecture and density on age-related changes in the mechanical behavior of cancellous bone [19,61 – 64]. However, defining the precise role of microarchitecture in prediction of the mechanical behavior of bone and its influence on fracture risk is complicated by the fact that microarchitecture characteristics are strongly correlated to each other and to bone density. As such, changes in trabecular architecture accompany the age-related declines in bone density. Trabecular number, trabecular thickness, and connectivity all decline with decreasing density, whereas trabecular separation and anisotropy increase [40,41,51,60,65 – 69]. Previous studies using architectural

features derived from a model that assumes that cancellous bone architecture is “plate like” suggested that architectural features provided only modest improvements in the prediction of mechanical properties over those provided by bone density alone [67,68]. However, these previous findings should be interpreted with caution, as recent data indicate significant differences in structural indices derived from the traditional plate model compared to those computed directly from high-resolution three-dimensional images [70]. For instance, Ulrich et al. [19] reported that indices of trabecular structure determined directly from three-dimensional micro-computed tomography data significantly improve the prediction of the mechanical behavior of cancellous bone specimens from several skeletal sites. Several studies have indicated that trabecular architecture differs in fracture subjects compared to those who have not suffered a fracture [71 – 73]. However, few of these studies have controlled for the confounding influence of differences in bone density between the two groups and few have investigated microarchitecture at the sites of fracture. Ciarelli and colleagues [74] measured microarchitecture of cancellous bone specimens from the femoral neck in subjects with hip fracture compared to unfractured autopsy patients. Whereas there were no differences in trabecular thickness, number, separation, or connectivity among samples matched for equal bone density, the degree of anisotropy differed between the two groups even after controlling for density differences. These data suggest a role for trabecular architecture in the etiology of fractures that may be independent of changes in bone density. Clearly, this is an area of great interest, and additional studies are required to define the role of in vivo assessments of trabecular architecture in the prediction of fracture risk [71,75 – 78]. As mentioned previously, because changes in trabecular architecture are strongly intercorrelated, it is difficult to discern the relative effect on bone strength of reductions in trabecular number versus trabecular thickness for both vertically and horizontally oriented trabecular struts. To address this issue, Silva and Gibson [79] developed a two-dimensional model of vertebral trabecular bone to simulate the effects of age-related changes in trabecular microstructure. They found that reductions in the number of trabeculae decreased bone strength two to five times more than reductions in trabecular thickness, which resulted in an identical decrease in bone density (Fig. 6). For instance, removing longitudinally oriented trabecular elements to create a 10% reduction in density resulted in a 70% reduction in bone strength. In contrast, reducing trabecular thickness to achieve a 10% reduction in density resulted in only a 20% reduction in strength. This study implies that it is important to maintain trabecular number in order to preserve bone strength with aging. Consequently, therapies designed to counter age-related declines in bone strength should strive to maintain or restore the number of trabeculae rather

515

CHAPTER 19 Biomechanics of Age-Related Fractures

FIGURE 6

A plot of the predicted effect of bone density reductions, either by a reduction in trabecular thickness or loss of trabecular elements, on the strength of vertebral cancellous bone. Strength reductions were at least twice as sensitive to changes in the number of trabeculae as to changes in the thickness of trabeculae. Findings were similar for loading in the transverse direction. From Silva and Gibson [79].

than just increasing the thickness of existing trabecular struts. A final aspect of trabecular architecture that may have been underappreciated until recently is the potential detrimental effect of increased variability in trabecular thickness and number within a given cancellous bone specimen [80,81]. Other factors may influence age-related changes in the mechanical behavior of bone, including the histologic structure (primary vs osteonal bone); the collagen content and orientation of collagen fibers; the number and composition of cement lines; and the presence of fatigue microdamage and microfractures [55,82 – 89]. For example, an increase in osteonal remodeling (and the subsequent increase in the number of cement lines) reduces the strength of the bone for single load applications. However, the cement lines act as deterrents to crack proliferation, possibly improving the mechanical behavior of bone under repetitive loading conditions [24]. Burr and colleagues [89] reviewed the potential role of skeletal microdamage in age-related fractures. They suggest that microdamage due to repetitive loading of bone likely initiates at the level of the collagen fiber or below and may include collagen fiber – matrix debonding, disruption of the mineral – collagen aggregate, and failure of the collagen fiber itself. They hypothesize that the accumulation and coalescence of these small defects eventually lead to microcracks that are visible under light microscopy. Although the relationship between existing microcracks and bone mechanical properties has not been established in vivo, investigators have shown that damage accumulation in devitalized bone leads to a decrease in bone strength [85,90]. Thus, the accumulation of microdamage in vivo may contribute to the increased fragility of the aging skeleton.

Microcracks occur naturally in human specimens from several anatomic locations, including trabecular bone from the femoral head and vertebral body, as well as cortical bone from the femoral and tibial diaphyses [88,91 – 95]. It appears that the incidence of microcracks increases with age, probably in an exponential fashion, and that after age 40, microdamage accumulates faster in women than in men [92,94] (Fig. 7). For instance, Mori and colleagues [91] reported that the density of microcracks in the femoral head of older women is more than double the density seen in younger women. In addition, they observed an inverse, nonlinear relationship between microcrack density and trabecular bone area, indicating that microcracks accumulate more rapidly as bone mass decreases. Similar evidence for a nonlinear relationship between microcrack density and trabecular bone area has been reported for vertebral trabecular bone specimens [93]. Thus, the accumulation of microdamage in vivo may contribute to the increased fragility of the aging skeleton.

D. Age-Related Changes in Bone Geometry Age-related changes in the material properties of bone tissue are frequently accompanied by a redistribution of the cortical and trabecular bone material. It is likely that the structural rearrangement of bone tissue is driven by “preprogrammed” behavior of the endosteal and periosteal bone

FIGURE 7

Bone microcrack density vs age. There is an exponential increase in microdamage accumulation in the femoral cortex in both men and women after the age of 40 years. Damage accumulation occurs about twice as rapidly in women as in men. From Burr et al. [89].

516

MARY L. BOUXSEIN

FIGURE 8 Age-related changes in the femoral midshaft demonstrating periosteal expansion and endosteal resorption. Data represent the mean  2 standard errors. From Ruff and Hayes [99], with permission. cells, as well as the local mechanical loading environment and biochemical signals [24,96]. Hence, the adaptation pattern depends on age, gender, skeletal site, physical activity patterns, and expression (local and systemic) of cytokines and growth factors. The general pattern of adaptation in the appendicular skeleton includes endosteal resorption and periosteal apposition of bone tissue (Fig. 8). Thus, the diameter of the bone increases, but the thickness of the cortex decreases. This redistribution of bone tissue away from the center of the bone allows the bone to better resist bending and torsional loads. Resistance to bending and torsional loading is particularly important, as the highest stresses in the appendicular skeleton are due to these loading modes [24]. The most efficient design for resisting bending and torsional loads involves distributing the material far from the neutral axis of bending

or torsion (generally the center of the bone). The distribution of mass about the center of a structural element is quantitatively described by the area moment of inertia. For example, consider three circular bars, each composed of the same material (Fig. 9). The resistance of each bar to tensile and compressive loads is directly proportional to the cross-sectional area. However, the resistance to bending and torsional loads is influenced not only by how much bone (i.e., the crosssectional area), but also by how it is distributed. Therefore, the structural capacity of bar C in bending or torsion is twice that of bar A due to its greater moment of inertia. Some studies indicate that both men and women exhibit endosteal resorption accompanied by periosteal expansion [97 – 101], whereas others report that women undergo geometric changes that lead to decreased bone strength [102 – 106]. Smith and Walker [101] studied femoral

FIGURE 9 Illustration of the influence of cross-sectional geometry on the structural strength of circular structures.

517

CHAPTER 19 Biomechanics of Age-Related Fractures

radiographs of 2030 women aged 45 to 90 and reported that periosteal diameter and cortical cross-sectional area (assuming a circular cross-section) both increased approximately 11% in 35 years. Furthermore, the section modulus (an indicator of the resistance to bending loads) increased 32% in the same time period. In contrast, Ruff and Hayes [106], using direct assessment of cadaveric femurs and tibiae from 75 Caucasian adults, reported that although both men and women undergo endosteal resorption and medullary expansion with age, only men show subperiosteal expansion and bone apposition at the femoral diaphysis. They reported that, in men, cortical area is nearly constant and moments of inertia increase slightly with age. In women, however, both cortical area and moments of inertia decrease with age. The authors concluded therefore, that in this sample from modern humans, only men exhibit bone remodeling patterns that would compensate for the agerelated decline in bone material properties. Age-related changes in bone shape and size have been observed not only at diaphyseal sites, but also at the vertebral body and femoral neck. However, age-related changes at these latter sites are much smaller than those seen at diaphyseal sites. In a study of 338 skeletons from the Smithsonian Institution, Ericksen [107] found that the transverse breadths of the L3 and L4 vertebrae increase slightly with increasing age in both men and women. Using indirect measurements, Mosekilde and Mosekilde [34] also found that, in men, the cross-sectional area of lumbar vertebrae is only very weakly associated with age (r 2  0.11), increasing approximately 25 – 30% from age 20 to 90. In women, however, there was no age-related change in vertebral cross-sectional area. Based on methodology originally presented by Martin and Burr [108], techniques have been developed to assess femoral geometry from X-ray absorptiometry exams, thereby allowing in vivo assessment of bone structure [102,109 – 112]. Investigators have used these methods to investigate race [111]- and sex [102,110,112]-based differences in femoral geometry. For instance, Beck and colleagues [102] studied the cross-sectional relationship between femoral neck geometry and age in 1044 women aged 18 – 89. In women under age 50, femoral neck bone mineral density (BMD) decreased on average 4% per decade with no observable changes in femoral neck geometry. In women over age 50, femoral neck BMD declined 2.5 times faster than in the younger group (on average 7% per decade). However, in contrast to the younger group, the older women also showed changes in femoral geometry wherein the cross-sectional area and cross-sectional moment of inertia of the femoral neck declined approximately 7 and 5% per decade, respectively. These data suggest that after menopause women lose bone mass at an accelerated rate, but that unlike men [110], they do not exhibit changes in femoral geometry that would

compensate for this loss. As a result, in women, the stress in the femoral neck during walking is predicted to increase from 25 to 40% from age 50 to 80. In another study, however, there were no age-related changes in the moment of inertia of the femoral neck in either men or women [112]. As can be seen from the results of the previous studies, the sex-specific nature of age-related changes in skeletal structure remains controversial. The discrepancies in findings related to sex-specific bone adaptation patterns may be attributed to several factors. Most importantly, most of these studies use a cross-sectional design, thereby possibly introducing secular changes that confound data and eliminating the possibility of a causal relationship with age. In addition, differences in methodology (direct vs in vivo measurements), subject populations (archaeological vs modern human specimens), and measurement site (femoral shaft vs femoral neck) likely contribute to the conflicting findings. Thus, it appears that women may have a reduced capacity to alter their bone geometry in order to preserve bone strength with aging compared to men. However, the extent to which this reduced capacity contributes to the increased fracture risk observed in women is unknown.

III. BIOMECHANICS OF HIP FRACTURES Recall that the “biomechanics” view of fractures states that a fracture occurs when the loads applied to the bone exceed its load-bearing capacity. Therefore, to study the etiology of hip fractures it is important first to identify what event(s) is associated with hip fractures, and determine the loads that are applied to the bone during that event, and what the load-bearing capacity of the femur is during that loading situation. It is estimated that over 90% of hip fractures in the elderly are associated with a fall [8,113]. Thus, studies of the etiology of hip fractures are complicated by the need to examine risk factors for falls, as well as risk factors for fracture. In addition, given that fewer than 2% of falls in the elderly result in a hip fracture [114 – 116], investigations of hip fracture etiology must also distinguish factors related to “high-risk” falls that result in fracture. Therefore, this section reviews clinical and laboratory studies related to factors influencing the loads applied to the femur during a fall and the load-bearing capacity of the femur in a fall configuration.

A. Factors That Influence Fall Severity: Loads Applied to the Femur during a Fall A fall can be defined as a sudden, unexpected event that results in a person coming to rest on a horizontal surface [4,117]. A fall can be further characterized by several

518 phases: (1) an instability phase resulting in fall initiation, (2) a descent phase, (3) an impact phase, and (4) a postimpact phase during which the faller comes to rest [4]. The definition of “fall severity” is more difficult. From a biomechanical perspective, fall severity can be described by the magnitude and direction of the load applied to the hip and the impact site. From a clinical perspective, Cummings and Nevitt [118] suggest that a high-risk fall includes (1) impact on or near the hip, (2) lack of active protective mechanisms such as an outstretched arm to break the fall, and (3) insufficient energy absorption by local soft tissues. Thus, by these criteria a high-risk fall could transmit a force to the proximal femur that exceeds the force required to fracture the hip. A few surveillance studies have been conducted to more fully characterize falls as they relate to hip fracture [4 – 7,9,115]. Among nursing home residents, falling to the side and impacting the hip or side of the leg increased the risk of hip fracture approximately 20-fold relative to falling in any other direction [4]. An increase in the potential energy content of the fall, computed from fall height and body mass, was also associated with an increased risk for fracture. Similar results were reported in a nested case-control analysis of the Study of Osteoporotic Fractures cohort, a large, prospective study in community-dwelling women [7]. Women who suffered a hip fracture were more likely to have fallen sideways or straight down and to have landed on or near the hip than women who fell and did not suffer a fracture [7]. Thus, these surveillance studies have identified several factors that are related to the “severity” of a fall in terms of hip fracture risk. From these data it is clear that a fall to the side represents a particularly risky event. Several laboratory investigations have been conducted to further study the characteristics of sideways falls. In a study of the descent phase of sideways falls, van den Kroonenberg and colleagues [119] estimated the impact velocities and energies that may occur during falls from standing height, the effect of muscle activity on these impact velocities, and insights into the high-risk nature of sideways falls. Six young, healthy adults (age 19 – 30) were asked to fall sideways, as naturally as possible, onto a thick gymnastics mattress. To investigate the effect of muscle activity on fall dynamics, subjects were instructed to fall either as relaxed as they could or to fall naturally, using the musculature of the trunk and upper extremity as they would in a reflex-mediated fall. To investigate potential protective mechanisms, during some falls subjects were instructed to try to break the fall with their arm. The vertical velocity at impact with the floor ranged between 2.1 and 4.8 m/sec. The impact velocity was 7% lower in relaxed than in muscle-active falls, a finding attributed to the observation that hip impact occurs closer to the feet in the muscle-relaxed case. Despite instructions to break the fall with an outstretched arm, only two of six subjects were able to do so (Fig. 10). In the

MARY L. BOUXSEIN

FIGURE 10 Example of a sideways fall onto a thick gymnastics mattress. Despite instructions to break the fall with the hand, only two of six subjects were able to do so. In the other subjects, hip impact occurred first, thus providing insight into the high-risk nature of sideways falls From van den Kroonenberg et al. [119], with permission.

remaining subjects, hip impact occurred first, followed by impact of the arm or hand. Finally, the authors found that, in these young adults, approximately 70% of the total energy available is dissipated during the descent phase of a sideways fall from standing height. This energy dissipation is likely due to muscle activity and the stiffness and damping characteristics of the hip and knee joints. Sabick and colleagues [120] reported that “active responses,” such as using the arm to break a fall, reduce the impact forces experienced at the hip during falls to the side. However, despite the potential for reducing fall severity via active responses, it is likely that with age, the ability to dissipate energy during a fall or to activate protective responses will decrease, and therefore it is quite likely that elderly individuals “fall harder” than young adults. The forces applied to the proximal femur during a sideways fall depend not only on the dynamics of the descent phase of the fall, but also on characteristics of the impact phase of the fall. Robinovitch and colleagues have conducted a series of experiments to study the potential roles of trochanteric soft tissues, muscle contraction, and body configuration in determining the load applied to the femur during a sideways fall with impact to the greater trochanter [121 – 125]. In these experiments, they used a “pelvis release” system (Fig. 11), in which a small force is applied to the lateral aspect of the hip and the dynamic response of the body is measured [121]. This system allows impact forces from falls to be predicted with reasonable accuracy from the body’s response to safe, simulated collisions [122]. They found that during a sideways fall with impact to the greater trochanter, only about 15% of the total impact force is distributed to structures peripheral to the hip, whereas the remainder of the force is delivered along a load path directly in line with the hip [123]. In addition, for the same body mass and height, sideways falls with the trunk in a more upright position are predicted to result in greater

CHAPTER 19 Biomechanics of Age-Related Fractures

519

FIGURE 11 Schematic diagram of the setup used for “pelvis release” experiments. The subject’s pelvis was supported by the sling, raised a small amount, and then released onto the force platform, which recorded the body’s dynamic response. Experiments were conducted in two body configurations — in the trunk-straight and trunk-flexed positions — to determine the effect of trunk position on fall impact dynamics. From Robinovitch et al. [123], with permission.

impact forces on the proximal femur than falls where the trunk is more horizontal at impact. To study the force attenuation and energy absorption properties of the soft tissues overlying the greater trochanter, tissue samples were obtained from nine cadavers, positioned over a surrogate proximal femur and pelvis, and subjected to a typical impact load associated with a sideways fall [125]. For a constant impact energy, trochanteric soft tissue thickness was strongly negatively correlated with the peak femoral impact force (r 2  0.91), such that the force applied to the femur decreased approximately 70 N per 1-mm increase in tissue thickness (Fig. 12). However, the force attenuation due to trochanteric soft tissues alone is likely insufficient to prevent hip fracture in falls where an elderly person lands directly on the hip [125]. These findings suggest that trochanteric padding systems may be effective means of reducing the load applied to the femur during a fall [126]. Finally, van den Kroonenberg et al. [127] developed a series of biomechanical models to estimate peak impact forces delivered to the proximal femur during a sideways fall from standing height. The models incorporated stiffness and damping parameters from the “pelvis-release” experiments [121 – 123], and the models’ behavior was compared with previous observations of the dynamics of voluntary sideways falls [119]. Using the most accurate model, peak impact forces applied to the greater trochanter ranged from 2900 to 4260 N (  650 – 960 lbs) for 5th to 95th percentile woman based on weight and height. Thus, these findings support the idea that “the bigger they are, the harder they fall” [128]. Given an individual’s height and weight, these models can be used to estimate femoral impact forces associated with a sideways fall.

FIGURE 12 Effect of trochanteric soft tissue thickness on (top) the force delivered to the femur and (bottom) the energy absorbed by soft tissue for a constant energy impact directed laterally on the hip. From Robinovitch et al. [125].

B. Factors That Influence the Strength of the Proximal Femur As mentioned previously, several factors contribute to the load-bearing capacity of the proximal femur, including its intrinsic material properties, as well as the total amount (size) and spatial distribution (shape) of the bone tissue. Because the mechanical properties of both cortical and trabecular bone are strongly related to bone density, many have hypothesized that age-related bone loss is a primary contributor to the steep increase in hip fracture incidence with age. In support of this hypothesis, there is strong evidence from prospective clinical studies that low BMD, measured both at the hip and at other sites, is a risk factor for hip fracture [129 – 132]. Furthermore, case-control studies of elderly fallers have reported that low BMD of the hip is a risk factor for hip fracture that is independent of fall characteristics [5 – 7]. Several laboratory studies have evaluated the loadbearing capacity of the proximal femur using a configuration designed to simulate the single-leg stance phase of gait [109,133 – 139]. The loads required to fracture the femur in the stance phase of gait ranged from approximately 1000 to 13,000 N (225 – 3000 lbs). These studies demonstrated a

520 strong relationship between the load required to fracture the femur in this stance configuration and noninvasive measurements of bone geometry and bone mineral density or content. Other studies have evaluated the load-bearing capacity of the proximal femur in a configuration designed to simulate a sideways fall with impact to the greater trochanter [139 – 147]. Courtney and colleagues [141,142] studied the effect of age and loading rate on the failure load of the proximal femur in the fall configuration (Fig. 13). They found that at a slow loading rate (2 mm/sec), femurs from young individuals (age 17 – 51) were more than twice as strong as femurs from older individuals (age 59 – 83). At high loading rates (100 mm/sec), such as might be expected during a fall, femurs from both young and older individuals were approximately 20% stronger than at the slower loading rate [142]. However, femurs from the younger group were still approximately 80% stronger than those from the older group. Loading direction may also dramatically influence femoral failure loads. Greater loads are required to fracture femurs testing in a single-leg stance configuration than in a sideways fall configuration, further supporting the high risk of a sideways fall in terms of hip fracture risk [139]. Moreover, subtle differences in the direction of a sideways fall can influence femoral strength as much as 25 years of age-related bone loss [148,149]. In addition to age, loading rate, and loading direction, femoral geometry also influence the load-bearing capacity of the proximal femur. The relationship between femoral

FIGURE 13 Mean failure loads for cadaveric proximal femurs from young and elderly donors tested in a sideways fall configuration at slow and fast loading rates. For each loading rate, femurs from the younger individuals were 80 – 100% stronger than femurs from the older individuals. Data from Courtney et al. [141,142].

MARY L. BOUXSEIN

geometry and load-bearing capacity is not unexpected. Because the load-bearing capacity is a structural property, it is influenced by the size of the specimen. Therefore, larger femurs have a greater load-bearing capacity. Therefore, as expected, femoral neck area, neck width, and neck axis length are all positively correlated (r 2  0.21 – 0.79) with femoral failure loads [141,143,145]. It is interesting to note that the positive correlation between femoral neck length and femoral strength appears to contradict findings from clinical studies, where a longer hip axis length (HAL) is associated with a greater risk of hip fracture [150]. This discrepancy may be attributed to the differences in the portion of hip anatomy that is included in the in vitro measurements (neck axis length only) versus in vivo measurements (neck axis length plus acetabular thickness). Some evidence suggests that it is the “acetabular thickness” portion of the measurement that is associated with fracture risk, and not the “femoral neck length” portion [151]. Additional laboratory studies are required to understand the complex relationship between hip geometry and fracture risk. While it is important to understand what factors influence the load-bearing capacity of the femur in the laboratory environment, it is also critical to develop techniques that can be used clinically to predict femoral strength. Several studies have confirmed that noninvasive assessments of bone mineral density and geometry using dual-energy Xray absorptiometry (DXA) or quantitative computed tomography (QCT) are strongly correlated to the load-bearing capacity of human cadaveric femurs. Femoral bone mineral content and density explain over 80% of the variation in the load-bearing capacity of the proximal femur [138,141,143,145 – 147] (Fig. 14). In summary, the load-bearing capacity of cadaveric proximal femurs ranges from approximately 800 to 10,000 N (180 – 2250 lbs) and is influenced, at least in part, by femoral bone mineral density, femoral geometry, loading direction, and loading rate. At a given moment, an individual’s

FIGURE 14

Bone mineral density of the femoral neck versus femoral failure load of cadaveric proximal femurs. The femurs were tested to failure in a configuration designed to simulate a sideways fall with impact to the greater trochanter. From Bouxsein et al. [147], with permission.

521

CHAPTER 19 Biomechanics of Age-Related Fractures

bone density and geometry are constant, although they can readily change with age or therapeutic intervention. However, other factors, such as loading direction and loading rate that are influenced by the characteristics of the fall, may significantly influence fracture risk.

C. Interactions between Fall Severity and Femoral Strength: The Factor of Risk for Hip Fracture The concept of a factor of risk for fracture (introduced in Section I) suggests that low BMD is not the only indicator of risk, but rather that the loads applied to the bone must also be considered. Case-control studies have demonstrated the importance of both fall severity and bone mineral density as risk factors for hip fracture [5 – 7]. In a nested casecontrol analysis of the Study of Osteoporotic Fractures cohort, Nevitt and Cummings [7], studied 130 women who fell and suffered a hip fracture and a consecutive sample of 467 women who fell and did not fracture. They reported that among those who fell on or near their hip, those who fell sideways or straight down were at increased risk for hip fracture (odds ratio  4.3), whereas those who fell backward were less likely to suffer a hip fracture (odds ratio  0.2). Furthermore, low BMD at the femoral neck (odds ratio  2.6 for a 1 SD decrease) or calcaneus (odds ratio  2.4 for a 1 SD decrease) strongly increased the risk of fracture among those who fell on or near the hip. Greenspan and co-workers [5] reported similar findings in a study of 149 community-dwelling men and women, including 72 cases who fell and suffered a hip fracture and 77 control subjects who fell and did not fracture. They showed that in these elderly fallers, independent risk factors for hip fracture included characteristics related to fall severity, low bone mineral density at the hip, and body habitus (Table 2). The success of hip protectors in preventing hip fracture provides additional evidence of the strong relationship between falls and hip fracture risk [126,152,153].

Clinical studies provide valuable information about the independent contributions of fall severity and skeletal fragility to hip fracture risk. However, further insight may be achieved by considering a “factor of risk” for hip fracture. The previous two sections have described how laboratory techniques can be used to develop and validate methods for estimating the loads applied to the femur and the load-bearing capacity of the femur from data that can be acquired in a clinical setting. Thus, these findings can be used to estimate the factor of risk for hip fracture due to a sideways fall from standing height. Myers and co-workers [154] applied the factor of risk concept in a case-control study of elderly fallers. The numerator of the factor of risk, the applied load, was estimated from previous studies of the descent and impact phases of a sideways fall with impact to the lateral aspect of the hip [119,121 – 123,127]. Each individual’s body height and weight were used as input parameters for the model to estimate the impact force delivered to the proximal femur during a sideways fall from standing height. The denominator of the factor of risk, or load-bearing capacity of the proximal femur, was determined from linear regressions between noninvasive bone densitometry and femoral failure loads in a fall configuration [143]. For each subject, femoral bone mineral density was assessed by DXA and then used to estimate the femoral failure load. There was a strong association between the factor of risk and hip fracture in these elderly fallers, with the odds of hip fracture increasing by 5.1 for a 1 SD increase in the factor of risk (95% confidence interval: 2.9, 9.2) (Fig. 15). In comparison, the odds ratio for a 1 SD decrease in femoral BMD was 2.0 (95% confidence interval: 1.4, 2.6).

IV. BIOMECHANICS OF VERTEBRAL FRACTURES Investigations of the etiology and biomechanics of vertebral fractures are particularly difficult, as the precise definition of a vertebral fracture remains controversial [155,156].

TABLE 2 Multiple Logistic Regression Analysis of Factors Associated with Hip Fracture in Community-Dwelling Men and Women Who Fella Factor

Adjusted odds

95% Confidence interval

P

Fall to the side

5.7

2.3 – 14

0.001

Femoral neck BMD (g/cm2)b

2.7

1.6 – 4.6

0.001

Potential energy of fall (Joules)c

2.8

1.5 – 5.2

0.001

Body mass index (kg/m2)b

2.2

1.2 – 3.8

0.003

a

Data from Greenspan et al. [5]. Calculated for a decrease of 1 SD. c Calculated for an increase of 1 SD. b

522

MARY L. BOUXSEIN

FIGURE 15

Proportion of subjects with hip fracture in each quartile of the factor of risk for hip fracture (top) and femoral neck BMD (bottom). Data from Myers et al. [154].

In addition, many fractures that are identified by radiographic review are asymptomatic [157], further complicating the interpretation of many studies. In contrast to the growing recognition of the importance of bone fragility and fall severity in the etiology of hip fractures, the role of spinal loading in the etiology of age-related vertebral fractures has received relatively little attention [158,159]. As loads are applied to the spine during nearly every activity of daily living, it is crucial to distinguish which of these activities (and the resulting loads on the spine) are associated with vertebral fractures to try to understand the loading environment that leads to vertebral fractures.

A. Factors That Influence Loads Applied to the Spine Although no clinical study has yet examined the relative roles of bone fragility and load severity as risk factors for

TABLE 3

vertebral fracture, several investigators have reviewed medical records or interviewed patients to assess the “degree of trauma” associated with vertebral fractures [157,160 – 162]. A few observational studies have attempted to identify the activities surrounding the onset of a vertebral fracture. These investigators have reviewed medical records or interviewed patients to assess the “degree of trauma” associated with vertebral fractures [157,160 – 162]. Cooper et al. [157] reviewed medical records from a 5-year period to determine the circumstances associated with “clinically diagnosed” vertebral fractures in a populationbased sample of 341 Rochester, Minnesota, residents. In their study, a specific loading event was reported for approximately 50% of the total fractures (Table 3). In contrast to the commonly held belief that lifting plays a major role in the development of vertebral fractures, relatively few of the fractures were associated with lifting. Excluding fractures that were diagnosed incidentally, only 10% of fractures were associated with “lifting a heavy object,” whereas nearly 40% were associated with falling. In a hospitalbased study, Myers et al. [163] interviewed patients after diagnosis of vertebral fracture with respect to their activity at the time of fracture. Their results indicate that nearly 50% of acute, symptomatic vertebral fractures in individuals over age 60 are associated with a fall, whereas 20% are associated with “controlled” activities, such as bending, lifting, and reaching. Most of the remainder of the patients could not identify a specific activity at the time of fracture. Therefore determining the forces on the spine during controlled activities and falls may improve our understanding of the biomechanics of vertebral fractures. Although it is impossible to measure the loads on the vertebral bodies in vivo, investigators have used kinematic analysis, electromyographic measurements, and biomechanical modeling to estimate the loads on the lumbar spine during various activities [164 – 167]. The models use optimization techniques to estimate the trunk muscle forces and compressive forces on the spine during various tasks and

Circumstances Associated with Clinically Diagnosed Vertebral Fractures a

Reported activity/circumstance

Number of persons

% of symptomatic fractures

% of total fractures 3.5

Pathologic fracture

12

4

Traffic accident

20

7

6

Fall from greater than standing height

27

9

8

Fall from standing height or less

86

30

25

Lifting a heavy object

29

10

8.5

“Spontaneous”

113

39

33

Diagnosed incidentally (asymptomatic)

54

NA

16

a

Data from Cooper et al. [157].

523

CHAPTER 19 Biomechanics of Age-Related Fractures

have been verified by comparing predicted compressive spine loads and muscle activity with direct measurements of intradiscal pressure [167 – 170] and myoelectric trunk muscle activity [164,165,167,171 – 173]. These models were originally developed to study the potential origins and mechanisms of low back pain and injury in working adults. Therefore, they are generally based on anthropometric data from young, healthy adults and are limited to estimating the vertebral forces in the lumbar region. However, Wilson [174] extended these models to include the mid- and lower thoracic spine and incorporated geometric properties of the trunk using CT scans of older individuals. Using this model, the compressive forces applied to T8, T11, and L2 vertebrae during various activities for a woman who weighed 58 kg and was 1.6 m tall (mean values from a cohort of 120 women aged 65 yrs or older [175] were computed (Table 4). The estimated forces applied to the spine ranged from approximately 400 to 2100 N for typical activities. For example, rising from a chair without the use of one’s hands results in compressive forces equal to 60 and 173% of body weight on T11 and L2 vertebrae, respectively. These estimates reinforce the concept that subtle changes in body position can dramatically alter spinal loading. Standing straight and holding an 8-kg weight with the arms slightly extended creates a compressive load on L2 equal to 230% of body weight, whereas flexing the trunk forward 30° and holding the same weight generates a compressive force on L2 of over 320% of body weight. From these estimates, it is clear that everyday activities, such as rising from a chair or bending over and picking up a full grocery bag, can generate high forces on the spine. There are currently few biomechanicals models designed to estimate the load on the spine during falls. However, TABLE 4

based on data from a study of the dynamics of backwards falls [176] and from the previously described “pelvis-release” experiments [121], Myers and Wilson [159] estimated that the impact force on the pelvis due to a backward fall would be approximately 2000 – 2500 N. This impact force would be expected to be dissipated somewhat before reaching the thoracolumbar spine. Research is currently underway to further characterize the loads applied to the spine during falls in order to better our understanding of the circumstances surrounding acute vertebral fractures.

B. Factors That Influence Vertebral Strength The use of noninvasive assessments of skeletal status to predict vertebral strength in vivo is based on the assumption that much of the variability in the strength of whole vertebrae can be explained by variations in bone mineral density and/or geometry. As in other skeletal structures, the loadbearing capacity of a whole vertebra is determined by its intrinsic material properties, as well as its overall geometry and shape. The vertebral body is characterized by a central core of cancellous bone surrounded by a thin covering of condensed trabecular bone (often referred to as a “cortical shell“) [177,178]. In the spine, compressive loads are transferred from the intervertebral discs to adjacent vertebral bodies. Therefore, age-related changes in the properties of the intervertebral disc, the vertebral centrum, and the vertebral shell can each influence the load-bearing capacity of the vertebrae. For instance, the thickness of the outer shell decreases from approximately 400 – 500 m at age 20 – 40 to 200 – 300 m at age 70 – 80 and to 120 – 150 m in osteoporotic individuals [179]. This change in vertebral morphology likely influences the way that loads are transmitted

Predicted Compressive Loads on L2 and T11 Vertebrae during Various Activitiesa Predicted load on T11

Activity Relaxed standing

N 240

% of body weight 41

Predicted load on L2 N

% of body weight

290

51 173

Rising from a chair, without hands

340

60

980

Standing, holding 8 kg weight close to body

320

57

420

74

Standing, holding 8-kg weight arms extended

660

117

1302

230

Standing, trunk flexed 30°, extended

370

65

830

146

Standing, trunk flexed 30°, 18 kg with arms extended

760

135

1830

323

Lift 15 kg from floor, knees arms straight down

593

104

1810

319

a The loads were computed from the model developed by Wilson [174] for a woman who weighs 58 kg and is 162 cm tall, and are expressed as the absolute value (in Newtons, N) and as a percent of total body weight.

524 throughout the spine. For instance, whereas the relative contributions of the vertebral centrum and shell to overall vertebral strength remain controversial, it is suggested that the vertebral shell may account for 10 – 30% of vertebral strength in healthy individuals and, due to decreased bone mass in the trabecular centrum, from 50 to 90% in osteoporotic persons [178 – 183]. Understanding the relative contributions of the cortical shell and trabecular centrum throughout aging and disease may afford the development of therapeutic agents specifically designed to strengthen one of these bone compartments. A number of laboratory studies have investigated the relationships among the strength of human lumbar and thoracic vertebrae and age, bone density, and vertebral geometry [21,33,182,184 – 196]. These studies indicate that the strength of thoracolumbar vertebrae is reduced from a value of 8000 – 10,000 N at age 20 – 30 to 1000 – 2000 N by age 70 – 80 [179,197]. In severely osteoporotic individuals, the load-bearing capacity may be even less [191]. The strength of human vertebrae is strongly correlated with noninvasive estimates of vertebral bone density and geometry, with approximately 50 – 80% of the variance in load-bearing capacity explained by parameters measured noninvasively [21,182,185 – 194,196]. For example, strong correlations have been reported between bone density and vertebral cross-sectional area assessed by quantitative computed tomography (QCT) and vertebral failure loads [187,188]. In addition, several investigators have reported strong correlations between bone mineral density, assessed by dual-energy X-ray absorptiometry, and vertebral strength [191,192,194 – 196]. For example, Moro et al. [191] found that lumbar BMD assessed by DXA correlates strongly with the compressive failure load and energy to failure of both L2 and T11 vertebrae (Fig. 16). The standard error of the estimate for predicting vertebral failure load from lumbar BMD was 527 N (25% of the mean failure load) for T11 and 733 N (28% of the mean failure load) for L2.

FIGURE 16 Linear relationship between lumbar BMD and compressive failure loads of T11 and L2 vertebrae as reported by Moro et al. [191]. Correlation coefficients between lumbar BMD and failure loads of T11 and L2 were r  0.94 (p0.001) and r  0.89 (p0.001), respectively.

MARY L. BOUXSEIN

Thus, it appears that noninvasive assessments of bone mass and bone mineral density provide reasonable estimates of the failure loads of cadaveric vertebrae subjected to controlled compression tests in the laboratory. It remains to be seen whether BMD or other bone density parameters can predict the strength of vertebrae subjected to loading conditions that may more closely resemble the mechanical environment in vivo, such as falling or compression combined with forward flexion or compression combined with lateral bending.

C. Interactions between Spinal Loads and Vertebral Strength: The Factor of Risk for Vertebral Fracture Although it has not been clearly demonstrated by clinical surveillance studies, it seems reasonable to suggest that, similar to hip fractures, both bone fragility and skeletal loading are important factors in the etiology of vertebral fractures [198]. To investigate the potential roles of bone fragility and spinal loading, Myers and Wilson [159] estimated a factor of risk for vertebral fractures for various activities of daily living (Fig. 17). As before, the factor of risk was defined as the ratio between load applied to the bone and its load-bearing capacity for a given loading event. They estimated the numerator of the factor of risk (i.e., the applied load) using predictions of compressive loading in the spine from the model developed by Wilson [174]. The denominator of the factor of risk (i.e., the failure load) was estimated from linear regressions between lumbar BMD and the compressive failure load of cadaveric vertebrae [191]. Their predictions of the factor of risk indicate that osteopenic individuals may perform many activities wherein their factor of risk for vertebral fracture is close to or greater than one, suggesting that they are at high risk for fracture. These estimates show that a woman who bends over to pick up a 15-kg object is predicted to be at great risk for vertebral fracture (i.e., 1) when her lateral L2 BMD is less than 0.55 g/cm2. To put this in context, the mean lateral L2 BMD for a 65-year-old woman is 0.58  0.10 g/cm2 (199). Hence, for the same lifting activity (ie., picking up a 15-kg object), a 65-year-old woman whose spine BMD is 1 SD below the mean for her age would have a factor of risk equal to 1.4 and would be at high risk for fracture. To reduce her factor of risk below one without altering the applied load due to lifting, the osteopenic woman would have to increase her spine BMD by 20%, an increase much greater than is currently achieved through the use of pharmocologic agents [200]. Thus, individuals with extremely low bone mineral density may be at risk for vertebral fracture during simple activities such as tying one’s shoes or opening a window. Individuals with low bone mineral density (still in the osteopenic range) may be at risk for vertebral fracture when lifting groceries out of the car or

525

CHAPTER 19 Biomechanics of Age-Related Fractures

FIGURE 17 Factor of risk for vertebral fracture for eight common activities as a function of lumbar bone mineral density. The numerator of the factor of risk was determined from models of spine loading at L2 for an elderly woman of average height and weight. The demonimator was determined on the basis of regression analysis between lateral lumbar BMD and the load-bearing capacity of the L2 vertebrae. Values for lateral BMD cover a wide range. The t score (number of standard deviations from the mean value for BMD in young women) is approximately  1 for a BMD  0.9 g/cm2 and is 5 for BMD  0.4 g/cm2. The factor of risk is predicted to be greater than or close to 1 for low BMD values (shaded area). From Myers and Wilson [159], with permission.

picking up a toddler. These examples illustrate the need for strategies to prevent vertebral fractures, such as reducing spinal loading by avoiding certain “high risk” activities.

V. SUMMARY AND CLINICAL IMPLICATIONS This chapter emphasized the concept that age-related fractures represent a structural failure whereby the forces applied to the bone exceed its load-bearing capacity. Viewing fractures in this manner, it is clear that studies of their etiology must include factors that influence skeletal fragility or its load-bearing capacity, as well as those that influence the forces that are applied to the skeleton. The load-bearing capacity of a skeletal structure is determined by its intrinsic material properties as well as the total amount (size) and spatial distribution (shape) of the bone tissue. Considerable evidence indicates that the material properties, in particular the elastic modulus and ultimate strength, of both cortical and trabecular bone decrease with increasing age in both men and women. This decrease in material properties is likely due, in part, to age-related

reductions in bone mass, as the elastic modulus and strength of trabecular bone are related to density by a nonlinear relationship. Therefore, small changes in bone density can dramatically influence bone material properties. These decrements in bone density and material properties may be partially offset by geometric rearrangement of the bone tissue, particularly in the long bones, that helps preserve the bone’s ability to resist bending and torsional loads. Clinical investigations have confirmed that skeletal status and fall severity are both significant and independent risk factors for hip fracture [5 – 9]. Estimates of the forces applied to the proximal femur during a sideways fall range from 2900 to 4260 N for the 5th to 95th percentile woman based on height and weight. Factors that influence the load applied to the femur include, but are not limited to, fall height, fall direction, body habitus, muscle activity, trochanteric soft tissue thickness, and the intrinsic stiffness of the hip and knee joints. In comparison, estimates of the load required to fracture the elderly cadaveric femur in a configuration simulating a sideways fall range from 800 to 10,000 N. This femoral failure load is influenced by femoral bone mineral content and density; femoral geometry; and the direction and rate of

526 the applied load. In particular, it appears that subtle changes in the direction of the load applied to the femur during a fall can influence femoral failure loads as much as nearly 25 year’s worth of age-related bone loss. Many of these factors that influence fall severity and femoral strength are independent of femoral bone mineral density and thus may prove useful in improving current estimates of fracture risk that are based on bone densitometry alone. In contrast to hip fractures, relatively little is known about the combined roles of spinal loading and skeletal fragility in the etiology of vertebral fractures. Contrary to previously held beliefs that vertebral fractures are caused primarily by bending and lifting activities, evidence shows that falls may play a significant role in the etiology of vertebral fractures. In one study, nearly 40% of clinically diagnosed, symptomatic fractures were associated with falls, whereas 10% were attributed to lifting a heavy object [157]. Moreover, preliminary findings from a surveillance study of acute vertebral fractures indicate that approximately one-half of these fractures are associated with falls. Thus, future studies should incorporate assessments of fall severity in order to determine the characteristics of falls associated with vertebral fracture. In addition, models to estimate the loads applied to the spine during a fall should be developed. Mathematical models used to estimate the forces generated in the spine during bending and lifting activities indicate that compressive forces generated in the lower thoracic and upper lumbar spine range from approximately 400 to 2100 N. A comparison of these loads with predicted vertebral strengths suggests that activities of daily living may place the elderly, osteopenic person at high risk for vertebral fracture. To date, investigators have focused primarily on methods to prevent bone loss and to restore bone to the osteopenic skeleton. However, alternative approaches for fracture prevention that are directed at reducing the loads applied to the skeleton may prove to be both effective and cost-efficient. For example, trochanteric padding systems designed to reduce the load applied to the hip during a fall have shown great potential for reducing fracture risk [126,152,153]. In one study, analyzed on an intention to treat basis, hip fracture incidence was reduced 53% in the group assigned to wear the hip pads [152]. In addition, energy-absorbing floors, particularly in institutional environments, may help lower the risk of fractures due to falls [201,202]. Vertebral fracture incidence may be reduced by teaching high-risk patients to avoid activities that generate high loads on the spine and thereby put them at increased risk for fracture. Clearly, identification of these high-risk activities is critical to the success of this approach for preventing fractures. Ultimately, fracture prevention may be best achieved by an educational program designed to limit high-risk activities in conjunction with interventions targeted at increasing bone mass and reducing loads applied to the skeleton during traumatic events.

MARY L. BOUXSEIN

References 1. N. F. Ray, J. Chan, M. Thamer, and L. Melton, Medical expenditures for the treatment of osteoporotic fracture in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12(1), 24 – 35 (1997). 2. A. Praemer, S. Furner, and D. Rice, “Musculoskeletal Conditions in the United States.” American Academy of Orthopaedic Surgeons, Rosemont, IL, (1999). 3. C. Cooper, G. Campion, and L. J. Melton, Hip fractures in the elderly: A world-wide projection. Osteopor Int. 2, 285 – 289 (1992). 4. W. Hayes, E. Myers, J. N. Morris, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 5. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 217(2), 128 – 133 (1994). 6. S. L. Greenspan, E. R. Myers, D. P. Kiel, R. A. Parker, W. C. Hayes, and N. M. Resnick, Fall direction, bone mineral density, and function: Risk factors for hip fracture in frail nursing home elderly. Am. J. Med. 104(6), 539 – 545 (1998). 7. M. C. Nevitt and S. R. Cummings, Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). 8. J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, and F. Kaplan, Risk factors for falls as a cause of hip fracture in women. N. Engl. J. Med. 324, 1326 – 1331 (1991). 9. R. G. Cumming, and R. J. Klineberg, Fall frequency and characteristics and the risk of hip fractures. J. Am. Geriatr. Soc. 42(7), 774 – 778 (1994). 10. W. Hayes, S. Piazza, and P. Zysset, Biomechanics of fracture risk prediction of the hip and spine by quantitative computed tomography. In: “Radiologic Clinics of North America D. I. Rosenthal (ed.), Vol. 29, pp. 1 – 18 Saunders, Philadelphia. 11. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81(6), 1804 – 1809 (1988). 12. L. J. Melton III, Epidemiology of hip fractures: Implications of the exponential increase with age. Bone 18, 121S – 126S (1996). 13. C. H. Turner and D. B. Burr, Basic biomechanical measurements of bone: A tutorial. Bone 14, 595 – 608 (1993). 14. D. Reilly, A. Burstein, and V. Frankel, The elastic modulus for bone. J. Biomech. 7, 271 – 275 (1974). 15. D. Reilly and A. Burstein, The elastic and ultimate properties of compact bone tissue. J. Biomech. 8, 393 – 405 (1975). 16. R. B. Ashman, S. C. Cowin, W. C. Van Buskirk, and J. C. Rice, A continuous wave technique for the measurement of the elastic properties of cortical bone. J. Biomech. 17, 349 – 361 (1984). 17. T. Brown and A. Ferguson, Mechanical property distributions in the cancellous bone of the human proximal femur. Acta. Orthop. Scand. 51, 429 – 437 (1980). 18. M. J. Ciarelli, S. Goldstein, J. Kuhn, D. Cody, and M. Brown, Evaluation of orthogonal mechanical properties and density of human trabecular bone from the major metaphyseal regions with materials testing and computed tomography. J. Orthop. Res. 9, 674 – 682 (1991). 19. D. Ulrich, B. van Rietbergen, A. Laib, and P. Ruegsegger, The ability of three-dimensional structural indices to reflect mechanical aspects of trabecular bone. Bone 25(1), 55 – 60 (1999). 20. J. S. Arnold, M. H. Bartley, D. S. Bartley, S. A. Tont, and D. P. Henkins, Skeletal changes in aging and disease. Clin. Orthop. Rel. Res. 49, 37 (1966). 21. L. Mosekilde, A. Viidik, and L. Mosekilde, Correlation between the compressive strength of iliac and vertebral trabecular bone in normal individuals. Bone 6, 291 (1985).

CHAPTER 19 Biomechanics of Age-Related Fractures 22. L. Mosekilde and L. Mosekilde, Normal vertebral body size and compressive strength: Relations to age and to vertebral and iliac trabecular bone compressive strength. Bone 7, 207 – 212 (1986). 23. L. Mosekilde and L. Mosekilde, Iliac crest trabecular bone volume as predictor for vertebral compressive strength, ash density and trabecular bone volume in normal individuals. Bone 9, 195 – 199 (1988). 24. R. Martin, Aging and strength of bone as a structural material. Calcif. Tissue. Int. 53 (Suppl 1), S34 – S40 (1993). 25. A. Burstein, D. Reilly, and M. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt Surg. 58-A, 82 – 86 (1976). 26. J. Currey, The mechanical consequences of variation in the mineral content of bone. J. Biomech. 2, 1 – 11 (1969). 27. O. Lindahl and A. Lindgren, Cortical bone in man. II. Variation in tensile strength with age and sex. Acta. Orthop. Scand. 38, 141 – 147 (1967). 28. R. Melick and D. Miller, Variations of tensile strength of human cortical bone with age. Clin. Sci. 30, 243 – 248 (1966). 29. C. Smith and D. Smith, Relations between age, mineral density and mechanical properties of human femoral compacta. Acta. Orthop. Scand. 47, 496 – 502 (1976). 30. J. C. Wall, S. K. Chatterji, and J. W. Jeffrey, Age-related changes in the density and tensile strength of human femoral cortical bone. Calcif. Tissue. Int. 27, 10 – 108 (1979). 31. R. McCalden, J. McGeough, M. Barker, and C. Court-Brown, Agerelated changes in the tensile properties of cortical bone. J. Bone. Jt. Surg. 75-A, 1193 – 1205 (1993). 32. P. Zioupos and J. Currey, Changes in the stiffness, strength, and toughness of human cortical bone with age. Bone 22, 57 – 66 (1998). 33. G. H. Bell, O. Dunbar, and J. S. Beck, Variations in strength of vertebrae with age and their relation to osteoporosis. Calcif. Tissue. Res. 1, 75 – 86 (1967). 34. L. Mosekilde and L. Mosekilde, Sex differences in age-related changes in vertebral body size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 35. J. Weaver and J. Chalmers, Cancellous bone: Its strength and changes with ageing and an evaluation of some methods for measuring its mineral content. J. Bone. Jt Surg. 48-A, 289 – 299 (1966). 36. L. Mosekilde, L. Mosekilde, and C. C. Danielson, Biomechanical competance of vertebral trabecular bone in relation of ash density and age in normal individuals. Bone 8, 79 – 85 (1987). 37. M. Ding, M. Dalstra, C. Danielsen, J. Kabel, I. Hvid, and F. Linde, Age variations in the properties of human tibial trabecular bone. J. Bone. Jt Surg. 79-B, 995 – 1002 (1997). 38. J. D. Currey, Changes in the impact energy absorption of bone with age. J. Biomech. 12, 459 – 469 (1979). 39. J. M. Britton and M. W. J. Davie, Mechanical properties of bone from iliac crest and relationship to L5 vertebral bone. Bone 11, 21 – 28 (1990). 40. L. Mosekilde, Age-related changes in vertebral trabecular bone architecture: Assessed by a new method. Bone 9, 247 – 250 (1988). 41. L. Mosekilde, Sex differences in age-related loss of vertebral trabecular bone mass and structure: Biomechanical consequences. Bone 10, 425 – 432 (1989). 42. O. Lindahl, Mechanical properties of dried defatted spongy bone. Acta. Orthop. Scand. 47, 11 – 19 (1976). 43. D. R. Carter and W. C. Hayes, Bone compressive strength: The influence of density and strain rate. Science 194, 1174 – 1176 (1976). 44. D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two-phase porous structure. J. Bone. Jt. Surg. 59-A(7), 954 –962 (1977). 45. L. Gibson and M. Ashby, “Cellular Solids: Structure and Properties.” Pergamon Press, New York, (1988). 46. J. C. Rice, S. C. Cowin, and J. A. Bowman, On the dependence of the elasticity and strength of cancellous bone on apparent density. J. Biomech. 21(2), 155 – 168 (1988).

527 47. T. M. Keaveny and W. C. Hayes, A 20-year perspective on the mechanical properties of trabecular bone. J. Biomech. Eng. 115, 534–542 (1993). 48. P. Atkinson, Variation in trabecular structure of vertebrae with age. Calcif. Tissue. Res. 1, 24 – 32 (1967). 49. L. Twomey, J. Taylor, and B. Furniss, Age changes in the bone density and structure of the lumbar vertebral column. J. Anat. 136(1), 15 – 25 (1983). 50. L. Gibson, The mechanical behaviour of cancellous bone. J. Biomech. 18, 317 – 328 (1985). 51. J. Compston, Connectivity of cancellous bone: Assessment and mechanical implications. Bone 15, 463 – 466 (1994). 52. J. D. Currey, The effect of porosity and mineral content on the Young’s modulus of elasticity of compact bone. J. Biomech. 21(2), 131 – 139 (1988). 53. J. Currey, Physical characteristics affecting the tensile failure properties of compact bone. J. Biomech. 23, 837 – 844 (1990). 54. R. Martin and D. Burr, “The Structure, Function, and Adaptation of Compact Bone.” Raven Press, New York, (1989). 55. R. Martin and J. Ishida, The relative effects of collagen fiber orientation, porosity, density, and mineralization on bone strength. J. Biomech 22, 419 – 426 (1989). 56. M. Schaffler and D. Burr, Stiffness of compact bone: Effects of porosity and density. J. Biomech. 21, 13 – 16 (1988). 57. M. Grynpas, Age and disease-related changes in the mineral of bone. Calcif. Tissue. Int. 53 (Suppl. 1), S57 – S64 (1993). 58. P. Steiger, S. Cummings, D. Black, N. Spencer, and H. Genant, Agerelated decrements in bone mineral density in women over 65. J. Bone. Miner. Res. 7, 625 – 632 (1992). 59. S. Greenspan, L. Maitland, E. Myers, M. Krasnow, and T. Kido, Femoral bone loss progresses with age: A longitudinal study in women over age 65. J. Bone. Miner. Res. 9, 1959 – 1965 (1994). 60. D. Ulrich, B. Rietbergen, A. Laib, and P. Ruegsegger, Mechanical analysis of bone and its microarchitecture based on in vivo voxel images. Technol. Health. Care. 6(5-6), 421 – 427 (1998). 61. R. Muller and P. Ruegsegger, Micro-tomographic imaging for the nondestructive evaluation of trabecular bone architecture. Stud. Health. Technol. Inform. 40, 61 – 79 (1997). 62. P. Ruegsegger, B. Koller, and R. Muller, A microtomographic system for the nondestructive evaluation of bone architecture. Calcif. Tissue. Int. 58(1), 24 – 29 (1996). 63. S. Majumdar, D. Newitt, A. Mathur, D. Osman, A. Gies, E. Chiu, J. Lotz, J. Kinney, and H. Genant, Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with Xray tomographic microscopy and biomechanics. Osteopor. Int. 6(5), 376 – 385 (1996). 64. S. Majumdar, M. Kothari, P. Augat, D. C. Newitt, T. M. Link, J. C. Lin, T. Lang, Y. Lu, and H. K. Genant, High-resolution magnetic resonance imaging: Three-dimensional trabecular bone architecture and biomechanical properties. Bone 22(5), 445 – 454 (1998). 65. B. D. Snyder, S. Piazza, W. T. Edwards, and W. C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue. Int. 53, S14 – S22 (1993). 66. C. Bergot, A. Laval-Jeantet, F Prêteux, and A. Meunier, Measurement of anisotropic vertebral trabecular bone loss during aging by quantitative image analysis. Calcif. Tissue. Int. 43, 143 – 149 (1988). 67. R. Goulet, S. Goldstein, M. Ciarelli, J. Kuhn, M. Brown, and L. Feldkamp, The relationship between the structural and orthogonal compressive properties of trabecular bone. J. Biomech 27, 375 – 389 (1994). 68. B. D. Snyder and W. C. Hayes, Multiaxial structure-property relations in trabecular bone. In “Biomechanics of Diarthrodial Joints.” pp. 31 – 59. V. C. Mow, A. Ratcliffe, and S. L.-Y. Woo, (eds.), Springer-Verlag, New York, (1990). 69. C. Turner, S. Cowin, J. Rho, R. Ashman, and J. Rice, The fabric dependence of the orthotropic elastic constants of cancellous bone. J. Biomech. 23, 549 – 561 (1990).

528 70. T. Hildebrand, A. Laib, R. Muller, J. Dequeker, and P. Ruegsegger, Direct three-dimensional morphometric analysis of human cancellous bone: Microstructural data from spine, femur, iliac crest, and calcaneus. J. Bone. Miner. Res. 14(7), 1167 – 1174 (1999). 71. T. M. Link, S. Majumdar, S. Grampp, G. Guglielmi, C. van Kuijk, H. Imhof, C. Glueer, and J. E. Adams, Imaging of trabecular bone structure in osteoporosis. Eur. Radiol. 9(9), 1781 – 1788 (1999). 72. E. Legrand, D. Chappard, C. Pascaretti, M. Duquenne, S. Krebs, V. Rohmer, M. F. Basle, and M. Audran, Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in male osteoporosis. J. Bone. Miner. Res. 15(1), 13 – 19 (2000). 73. J. E. Aaron, P. A. Shore, R. C. Shore, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. II. Three-dimensional histology. Bone 27(2), 277 – 282 (2000). 74. T. E. Ciarelli, D. P. Fyhrie, M. B. Schaffler, and S. A. Goldstein, Variations in three-dimensional cancellous bone architecture of the proximal femur in female hip fractures and in controls. J. Bone. Miner. Res. 15(1), 32 – 40 (2000). 75. A. Laib, T. Hildebrand, H. J. Hauselmann, and P. Ruegsegger, Ridge number density: A new parameter for in vivo bone structure analysis. Bone 21(6), 541 – 546 (1997). 76. A. Laib, H. J. Hauselmann, and P. Ruegsegger, In vivo high resolution 3D-QCT of the human forearm. Technol. Health. Care 6(5-6), 329 – 337 (1998). 77. S. Majumdar, T. M. Link, P. Augat, J. C. Lin, D. Newitt, N. E. Lane, and H. K. Genant, Trabecular bone architecture in the distal radius using magnetic resonance imaging in subjects with fractures of the proximal femur. Osteoporos. Int. 10(3), 231 – 239 (1999). 78. T. Lang, P. Augat, S. Majumdar, X. Ouyang, and H. K. Genant, Noninvasive assessment of bone density and structure using computed tomography and magnetic resonance. Bone 22(5 Suppl.), 149S – 153S (1998). 79. M. Silva and L. Gibson, Modeling the mechanical behavior of vertebral bone: Effects of age-related changes in microstructure. Bone 21, 191 – 199 (1997). 80. O. C. Yeh and T. M. Keaveny, Biomechanical effects of intraspecimen variations in trabecular architecture: A three-dimensional finite element study. Bone 25(2), 223 – 228 (1999). 81. M. Kothari, T. M. Keaveny, J. C. Lin, D. C. Newitt, and S. Majumdar, Measurement of intraspecimen variations in vertebral cancellous bone architecture. Bone 25(2), 245 – 250 (1999). 82. D. B. Burr, M. B. Schaffler, and R. G. Frederickson, Composition of the cement line and its possible mechanical role as a local interface in human compact bone. J. Biomech. 21(11), 939 – 945 (1988). 83. A. Burstein, J. Zika, K. Heiple, and L. Klein, Contribution of collagen and mineral to the elastic-plastic properties of bone. J. Bone Jt. Surg. 57, 956 – 961 (1975). 84. S. Carando, M. Portigliatti Barbos, A. Ascenzi, and A. Boyde, Orientation of collagen in human tibial and fibular shafts and possible correlation with mechanical properties. Bone 10, 139 – 142 (1989). 85. D. R. Carter and W. C. Hayes, Compact bone fatigue damage. I. Residual strength and stiffness. J. Biomech. 10, 325 – 337 (1977). 86. F. Evans and R. Vincintelli, Relations of the compressive properties of human cortical bone to histological structure and calcification. J. Biomech. 7, 1 – 10 (1974). 87. J. Katz and A. Meunier, The elastic anisotropy of bone. J. Biomech. 20, 1063 – 1070 (1987). 88. M. B. Schaffler, K. Choi, and C. Milgrom, Microcracks and aging in human femoral compact bone. Trans. Orthop. Res. Soc. 19, 190 (1994). 89. D. Burr, M. Forwood, D. Fyhrie, R. Martin, M. Schaffler, and C. Turner, Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J. Bone Miner. Res. 12, 6 – 15 (1997).

MARY L. BOUXSEIN 90. S. Hoshaw, D. Cody, A. Saad, and D. Fhyrie, Decrease in canine proximal femoral ultimate strength and stiffnss due to fatigue damaage. J. Biomech. 30, 323 – 329 (1997). 91. S. Mori, R. Harruf, W. Ambrosius, and D. Burr, Trabecular bone volume and microdamage accumulation in the femoral heads of women with and without femoral neck fractures. Bone 21, 521 – 526 (1997). 92. M. Schaffler, K. Choi, and C. Milgrom, Aging and matrix microdamage accumulation in human compact bone. Bone 17, 521 – 525 (1995). 93. T. Wenzel, M. Schaffler, and D. Fyhrie, In vivo trabecular microcracks in human vertebral bone. Bone 19, 89 – 95 (1996). 94. T. Norman and Z. Wang, Microdamage of human cortical bone: Incidence and morphology in long bones. Bone 20, 375 – 379 (1997). 95. A. Courtney, W. Hayes, and L. Gibson, Age-related differences in post-yield damage in human cortical bone: Experiment and model. J. Biomech 29, 1463 – 1471 (1996). 96. M. C. H. van der Meulen, G. S. Beaupré, and D. R. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 97. M. Bouxsein, K. Myburgh, M. van der Meulen, E. Lindenberger, and R. Marcus, Age-related differences in cross-sectional geometry of the forearm bones in healthy women. Calcif. Tissue Int. 54, 113 – 118 (1994). 98. S. Garn, C. Rohmann, B. Wagner, and W. Ascoli, Continuing bone growth throughout life: A general phenomenon. Am. J. Phys. Anthrop. 26, 313 – 318 (1967). 99. C. Ruff and W. Hayes, Subperiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 947 (1982). 100. C. Ruff and W. Hayes, Cross-sectional geometry of Pecos Pueblo femora and tibiae: A biomechanical investigation. I. Method and general patterns of variation. Am. J. Phys. Anthropol. 60, 359 – 381 (1983). 101. R. Smith and R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 102. T. Beck, C. Ruff, and K. Bissessur, Age-related changes in female femoral neck geometry: Implications for bone strength. Calcif. Tissue Int. 53 (Suppl. 1), S41 – S46 (1993). 103. D. Burr and R. Martin, The effects of composition, structure and age on the torsional properties of the human radius. J. Biomech. 16(8), 603 – 608 (1983). 104. R. Martin and P. Atkinson, Age and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 105. H. Meema, Cortical bone atrophy and osteoporosis as a manifestation of aging. Am. J. Roentgenol. Radium Ther, Nucl. Med. 89, 1287 – 1295 (1963). 106. C. Ruff and W. Hayes, Sex differences in age-related remodeling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). 107. M. F. Ericksen, Some aspects of aging in the lumbar spine. Am. J. Phys. Anthrop. 45, 575 – 580 (1976). 108. R. B. Martin and D. B. Burr, Non-invasive measurement of long bone cross-sectional moment of inertia by photon absorptiometry. J. Biomech. 17(3), 195 – 201 (1984). 109. T. J. Beck, C. B. Ruff, K. E. Warden, W. W. Scott, and G. U. Rao, Predicting femoral neck strength from bone mineral data. Invest. Radiol. 25, 6 – 18 (1990). 110. T. Beck, C. Ruff, W. Scott. Jr. C. Plato, J. Tobin, and C. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 111. T. Nakamura, C. Turner, T. Yoshikawa, C. Slemenda, M. Peacock, D. Burr, Y. Mizuno, H. Orimo, Y. Ouchi, and C. Johnston, Do variations in hip geometry explain differences in hip fracture risk between Japanese and white Americans? J. Bone Minre. Res. 9, 1071 – 1076 (1994). 112. T. Yoshikawa, C. Turner, M. Peacock, C. Slemenda, C. Weaver, D. Teegarden, P. Markwardt, and D. Burr, Geometric structure of the

CHAPTER 19 Biomechanics of Age-Related Fractures

113.

114.

115.

116.

117. 118. 119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

femoral neck measured using dual-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 1053 – 1064 (1994). S. Cummings, D. Black, M. Nevitt, W. Browner, J. Cauley, H. Genant, S. Mascioli, and J. Scott, Appendicular bone density and age predict hip fractures in women. JAMA 263, 665 – 668 (1990). J. Michelson, A. Myers, R. Jinnah, Q. Cox, and M. Van Natta, Epidemiology of hip fractures among the elderly: Risk factors for fracture type. Clin. Orthop. Rel. Res. 311, 129 – 135 (1995). M. C. Nevitt, S. R. Cummings, and E. S. Hudes, Risk factors for injurious falls: A prospective study. J. Gerontol. 46, M164 – M170 (1991). M. Tinetti, Factors associated with serious injury during falls among elderly persons living in the community. J. Am. Geriatr. Soc. 35, 644 – 648 (1987). R. Sattin, Falls among older persons. Annu. Rev. Publ. Health. 13, 489 – 508 (1992). S. Cummings and M. Nevitt, A hypothesis: The causes of hip fracture. J. Gerontol. 44, M107-M111 (1989). A. van den Kroonenberg, W. Hayes, and T. McMahon, Hip impact velocities and body configurations for experimental falls from standing height. J. Biomech. 29, 807 – 811 (1996). M. Sabick, J. Hay, V. Goel, and S. Banks, Active responses decrease impact forces at the hip and shoulder in falls to the side J. Biomech. 32, 993 – 998 (1999). S. Robinovitch, W. Hayes, and T. McMahon, Prediction of femoral impact forces in falls on the hip. J. Biomech. Eng. 113, 366 – 374 (1991). S. Robinovitch, W. Hayes, and T. McMahon, Predicting the impact reponse of a nonlinear, single-degree-of-freedom shock-absorbing system from the measured step response. J. Biomech. Eng. 119, 221 – 227 (1997). S. Robinovitch, W. Hayes, and T. McMahon, Distribution of contact force during impact to the hip. Ann. Biomed. Eng. 25(3), 499 – 508 (1997). S. Robinovitch, T. McMahon, and W. Hayes, Energy-shunting hip padding system attenuates femoral impact force in a simulated fall. J. Biomech. Eng. 177, 409 – 413(1995). S. Robinovitch, T. McMahon, and W. Hayes, Force attenuation in trochanteric soft tissues during impact from a fall. J. Orthop. Res. 13, 956 – 962 (1995). J. Lauritzen, Hip fractures: Epidemiology, risk factors, falls, energy absorption, hip protectors, and prevention. Dan. Med. Bull. 44(2), 155 – 168 (1997). A. van den Kroonenberg, W. Hayes, and T. McMahon, Dynamic models for sideways falls from standing height. J. Biomech. Eng. 117, 309 – 318 (1995). D. Hemenway, D. Azrael, E. Rimm, D. Feskanich, and W. Willett, Risk factors for hip fracture in US men aged 40 through 75 years. Am. J. Public Health 84, 1843 – 1845 (1994). S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, C. Ensrud, H. K. Genant, L. Palermo, J. Scott, and T. M. Vogt, Bone density at various sites for prediction of hip fractures. Lancet 341, 72 – 75 (1993). M. Nevitt, O. Johnell, D. Black, K. Ensrud, H. Genant, and S. Cummings, Bone mineral density predicts non-spine fractures in very elderly women. Osteopor. Int. 4, 325 – 331 (1994). P. D. Ross, J. W. Davis, J. M. Vogel, and R. D. Wasnich, A critical review of bone mass and the risk of fractures in osteoporosis. Calcif. Tissue. Int. 46, 149 – 161 (1990). D. Hans, P. Dargent-Molina, A. Schott, J. Sebert, C. Cormier, P. Kotski, P. Delmas, J. Pouilles, G. Breart, and P. Meunier, Ultrasonographic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 348, 511 – 514 (1996).

529 133. N. Dalén, L. Hellström, and B. Jacobson, Bone mineral content and mechanical strength of the femoral neck. Acta Orthop. Scand. 47, 503 – 508 (1976). 134. I. Leichter, J. Y. Margulies, A. Weinreb, J. Mizrahi, G. C. Robin, B. Conforty, M. Makin, and B. Bloch, The relation between bone density, mineral content, and mechanical strength in the femoral neck. Clin. Orthop. Rel. Res. 163, 272 – 281 (1982). 135. J. Mizrahi, J. Y. Margulies, I. Leichter, and D. Deutsch, Fracture of the human femoral neck: Effect of density of the cancellous core. J. Biomed. Eng. 6, 56 – 62 (1984). 136. D. J. Sartoris, F. G. Sonner, J. Kosek, A. Gies, and D. R. Carter, Dualenergy projection radiography in the evaluation of femoral neck strength, density, and mineralization. Invest. Radiol. 20, 476 – 485 (1985). 137. A. Alho, T. Husby, and A. H iseth, Bone mineral content and mechanical strength: An ex vivo study on human femora at autopsy. Clin. Orthop. Rel. Res. 227, 292 – 297 (1988). 138. D. D. Cody, G. J. Gross, F. J. Hou, H. J. Spencer, S. A. Goldstein, and D. P. Fyhrie, Femoral strength is better predicted by finite element models than QCT and DXA. J. Biomech. 32(10), 1013 – 1020 (1999). 139. J. H. Keyak, S. A. Rossi, K. A. Jones, and H. B. Skinner, Prediction of femoral fracture load using automated finite element modeling. J. Biomech. 31(2), 125 – 133 (1998). 140. J. Lotz and W. Hayes, The use of quantitative computed tomography to estimate risk of fracture of the hip from falls. J. Bone Jt. Surg. 72-A(5), 689 – 700 (1990). 141. A. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Age-related reductions in the strength of the femur tested in a fall loading configuration. J. Bone Jt Surg. 77, 387 – 395 (1995). 142. A. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Effects of loading rate on the strength of the proximal femur. Calcif Tissue Int. 55, 53 – 58 (1994). 143. M. Bouxsein, A. Courtney, and W. Hayes, Ultrasound and densitometry of the calcaneus correlate with the failure loads of cadaveric femurs. Calcif. Tissue Int. 56, 99 – 103 (1995). 144. Deleted at proof. 145. X. Cheng, G. Lowet, S. Boonen, P. Nicholson, P. Brys, P. Nijs, and J. Dequeker, Assessment of the strength of proximal femur in vitro: Relationship to femoral bone mineral density and femoral geometry. Bone 20(3), 213 – 218 (1997). 146. E. M. Lochmuller, J. B. Zeller, D. Kaiser, F. Eckstein, J. Landgraf, R. Putz, and R. Steldinger, Correlation of femoral and lumbar DXA and calcaneal ultrasound, measured in situ with intact soft tissues, with the in vitro failure loads of the proximal femur. Osteopor. Int. 8(6), 591 – 598 (1998). 147. M. L. Bouxsein, B. S. Coan, and S. C. Lee, Prediction of the strength of the elderly proximal femur by bone mineral density and quantitative ultrasound measurements of the heel and tibia. Bone 25(1), 49 – 54 (1999). 148. T. P. Pinilla, K. C. Boardman, M. L. Bouxsein, E. R. Myers, and W. C. Hayes, Impact direction from a fall influences the failure load of the proximal femur as much as age-related bone loss. Calcif. Tissue Int. 58(4), 231 – 235 (1996). 149. C. Ford, T. Keaveny, and W. Hayes, The effect of impact direction on the structural capacity of the proximal femur during falls. J. Bone Miner. Res. 11, 377 – 383 (1996). 150. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fracture. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 151. C. C. Glüer, S. R. Cummings, A. Pressman, J. Li, K. Glüer, K. G. Faulkner, S. Grampp, and H. K. Genant, Prediction of hip fractures from pelvic radiographs: The study of osteoporotic fractures. J. Bone Miner. Res. 9, 671 – 677 (1994).

530 152. J. Lauritzen, M. Petersen, and B. Lund, Effect of external hip protectors on hip fractures. Lancet 341, 11 – 13 (1993). 153. M. Parker, L. Gillespie, and W. Gillespie, Hip protectors for preventing hip fractures in the elderly. Cochrane Database Syst. Rev. 2, CD001255 (2000). 154. E. R. Myers, S. N. Robinovitch, S. L. Greenspan, and W. C. Hayes, Factor of risk is associated with frequency of hip fracture in a casecontrol study. Trans. Orthop. Res. Soc. 19, 526 (1994) 155. T. O’Neill, J. Varlow, D. Felsenberg, O. Johnell, K. Weber, F. Marchant, P. Delmas, C. Cooper, J. Kanis, and A. Silman, Variation in vertebral height ratios in population studies. J. Bone Miner. Res. 9, 1895 – 1907 (1994). 156. M. Kleerekoper and D. Nelson, Vertebral fracture or vertebral deformity? Calcif. Tissue Int. 50, 5 – 6 (1992). 157. C. Cooper, E. Atkinson, W. Ó. Fallon, and L. Melton, Incidence of clinically diagnosed vertebral fractures: A population-based study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7(2), 221 – 227 (1992). 158. M. C. Nevitt, Epidemiology of osteoporosis. In “Rheumatic Disease Clinics of North America: Osteoporosis” N. E. Lane, (ed.), Vol. 20, pp. 535 – 560. Saunders, Philadelphia, (1994). 159. E. Myers and S. Wilson, Biomechanics of osteoporosis and vertebral fractures. Spine 22, 25S – 31S (1997). 160. U. Bengnér, O. Johnell, and I. Redlund-Johnell, Changes in incidence and prevalence of vertebral fractures during 30 years. Calcif. Tissue Int. 42, 293 – 296 (1988). 161. U. Patel, S. Skingle, G. Campbell, A. Crisp, and I. Boyle, Clinical profile of acute vertebral compression fractures in osteoporosis. Br. J. Rheumato. 30, 418 – 421 (1991). 162. S. Santavirta, Y. Konttinen, M. Heliövaara, P. Knekt, P. Lüthje, and A. Aromaa, Determinants of osteoporotic thoracic vertebral fracture. Acta Orthop. Scand. 63(2), 198 – 202 (1992). 163. E. Myers, S. Wilson, and S. Greenspan, Vertebral fractures in the elderly occur with falling and bending. J. Bone Miner. Res. 11, S355 (1996). 164. G. B. J. Andersson, R. Örtengren, and A. Schultz, Analysis and measurement of the loads on the lumbar spine during work at a table. J. Biomech. 13, 513 – 520 (1980). 165. A. B. Schultz, G. B. J. Andersson, K. Haderspeck, R. Örtengren, M. Nordin, and R. Björk, Analysis and measurement of lumbar trunk loads in tasks involving bends and twists. J. Biomech. 15(9), 669 – 675 (1982). 166. A. B. Schultz and G. B. J. Andersson, Analysis of loads on the lumbar spine. Spine 6(1), 76 – 82 (1981). 167. A. Schultz, G. Andersson, R. Ortengren, K. Haderspeck, and A. Nachemson, Loads on the lumbar spine: Validation of a biomechanical analysis by measurements of intradiscal pressures and myoelectric signals. J. Bone Jt. Surg. 64A, 713 – 720 (1982). 168. M. Jäger and A. Luttmann, Biomechanical analysis and assessment of lumbar stress during load lifting using a dynamic 19-segment human model. Ergonomics 32, 93 – 112 (1989). 169. A. Nachemson and G. Elfström, Intravital dynamic pressure measurements in lumbar discs. Scand. J. Rehab. Med. Suppl. 1, 3 – 40 (1970). 170. A. L. Nachemson, The lumbar spine: An orthopaedic challenge. Spine 1(1), 59 – 69 (1976). 171. J. Cholewicki, S. McGill, and R. Norman, Comparison of muscle forces and joint load from an optimization and EMG assisted lumbar spine model: Towards development of a hybrid approach. J. Biomech. 28, 321 – 331 (1995). 172. A. Schultz, G. B. J. Andersson, R. Örtengren, R. Björk, and M. Nordin, Analysis and quantitative myoelectric measurements of loads on the lumbar spine when holding weights in standing postures. Spine 7, 390 – 397 (1982).

MARY L. BOUXSEIN 173. A. Yettram and M. Jackman, Equilibrium analysis for the forces in the human spinal column and its musculature. Spine 5(5), 402 – 411 (1980). 174. S. Wilson, Development of a model to predict the compressive forces on the spine associated with age-related vertebral fractures. Massachusetts Institute of Technology (1994). 175. S. Greenspan, L. Maitland-Ramsey, and E. Myers, Classification of osteoporosis in the elderly is dependent on site-specific analysis. Calcif. Tissue Int. 58, 409 – 414 (1996). 176. A. van den Kroonenberg, S. Wilson, E. Myers, W. Hayes, and T. McMahon Impact velocities and body configurations for backward falls from standing height. Submitted for publication. 177. L. Mosekilde and L. Mosekilde, Vertebral structure and strength in vivo and in vitro. Calcif. Tissue Int. 53, S121 – S126 (1993). 178. M. Silva, T. Keaveny, and W. Hayes, Load sharing between the shell and centrum in the lumbar vertebral body. Spine 22(2), 140–150 (1997). 179. L. Mosekilde, Osteoporosis: Mechanisms and models. In “Anabolic Treatments for Osteoporosis.” J. Whitfield, and P. Morley, (eds.), pp. 31 – 58. CRC Press, Boca Raton, FL, (1998). 180. K. Faulkner, C. Cann, and B. Hasedawa, Effect of bone distribution on vertebral strength: Assessment with a patient-specific nonlinear finite element analysis. Radiology 179, 669 – 674 (1991). 181. S. D. Rockoff, E. Sweet, and J. Bleustein, The relative contribution of trabecular and cortical bone to the strength of human lumbar vertebrae. Calcif. Tissue. Res. 3, 163 – 175 (1969). 182. R. J. McBroom, W. C. Hayes, W. T. Edwards, R. P. Goldberg, and A. A. White, Prediction of vertebral body compressive fracture using quantitative computed tomography. J. Bone Jt. Surg. 67-A, 1206 – 1214 (1985). 183. N. Yoganandan, J. Myklebust, J. Cusick, C. Wilson, and A. Sances, Functional biomechanics of the thoracolumbar vertebral cortex. Clin. Biomech. 3, 11 – 18 (1988). 184. M. H. Bartley, J. S. Arnold, R. K. Haslam, and W. S. S. Jee, The relationship of bone strength and bone quantity in health, disease, and aging. J. Gerontol. 21, 517 – 521 (1966). 185. M. Biggemann, D. Hilweg, S. Seidel, M. Horst, and P. Brinckmann, Risk of vertebral insufficiency fractures in relation to compressive strength predicted by quantitative computed tomography. Eur. J. Radiol. 13, 6 – 10 (1991). 186. M. Biggemann, D. Hilweg, and P. Brinckmann, Prediction of the compressive strength of vertebral bodies of the lumbar spine by quantitative computed tomography. Skel. Radiol. 17, 264 – 269 (1988). 187. P. Brinckmann, M. Biggeman, and D. Hilweg, Prediction of the compressive strength of human lumbar vertebrae. Clin. Biomech. 4, S1 – S27. 188. D. Cody, S. Goldstein, M. Flynn, and E. Brown, Correlations between vertebral regional bone mineral density (rBMD) and whole bone fracture load. Spine 16, 146 – 154 (1991). 189. S. A. Eriksson, B. O. Isberg, and J. U. Lindgren, Prediction of vertebral strength by dual photon absorptiometry and quantitative computed tomography. Calcif. Tissue Int. 44, 243 – 250 (1989). 190. T. Hansson, B. Roos, and A. Nachemson, The bone mineral content and ultimate compressive strength of lumbar vertebrae. Spine 5(1), 46 – 55 (1980). 191. M. Moro, A. T. Hecker, M. L. Bouxsein, and E. R. Myers, Failure load of thoracic vertebrae correlates with lumbar bone mineral density measured by DXA. Calcif. Tissue Int. 56, 206 – 209 (1995). 192. B. Myers, K. Arbogast, B. Lobaugh, K. Harper, W. Richardson, and M. Drezner, Improved assessment of lumbar vertebral body strength using supine lateral dual-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 687 – 693 (1994). 193. A. Vesterby, L. Mosekilde, H. Gundersen, F. Melsen, L. Mosekilde, K. Holme, and S. S rensen, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991).

531

CHAPTER 19 Biomechanics of Age-Related Fractures 194. A. Tabensky, J. Williams, V. DeLuca, E. Briganti, and E. Seeman, Bone mass, areal, and volumetric bone density are equally accurate, sensitive, and specific surrogates of the breaking strength of the vertebral body: An in vitro study. J. Bone Miner. Res. 11, 1981 – 1988 (1996). 195. X. G. Cheng, P. H. Nicholson, S. Boonen, G. Lowet, P. Brys, J. Aerssens, G. Van der Perre, and J. Dequeker, Prediction of vertebral strength in vitro by spinal bone densitometry and calcaneal ultrasound. J. Bone Miner. Res. 12(10), 1721 – 1728 (1997). 196. E. M. Lochmuller, F. Eckstein, D. Kaiser, J. B. Zeller, J. Landgraf, R. Putz, and R. Steldinger, Prediction of vertebral failure loads from spinal and femoral dual- energy X-ray absorptiometry, and calcaneal ultrasound: An in situ analysis with intact soft tissues. Bone 23(5), 417 – 424 (1998). 197. M. Biggeman and P. Brinckman, Biomechanics of osteoporotic vertebral fractures. In “Vertebral Fracture in Osteoporosis.” H. Genant,

198.

199. 200. 201.

202. 203.

M. Jergas, C. van Kuijk, (eds.), pp. 21 – 34. Osteoporosis Research Group, University of California, San Francisco, (1995). W. C. Hayes, Biomechanics of cortical and trabecular bone: Implications for assessment of fracture risk. In “Basic Orthopaedic Biomechanics.” V. C. Mow, and W. C. Hayes (eds.), pp. 93 – 142. Raven Press, New York, (1991). Hologic, Hologic Reference Database, (1994). R. Eastell, Treatment of postmenopausal osteoporosis. N. Engl. J. Med. 338(11), 736 – 746 (1998). J. Casalena, D. Streit, P. Cavanaugh, Preliminary design of a dual stiffness floor for injury prevention. In “Injury Prevention through Biomechanics” K. Yang (ed.), p. 155. B. Maki, G. Fernie, Impact attenuation of floor coverings in simulated falling accidents. Appl. Ergon. 21, 107 – 114 (1990). T. M. Keaveny and W. Hayes, Mechanical properties of cortical and trabecular bone. In “Bone” (B. Hall, ed.), Vol. 7. CRC Press, Boca Raton, FL, (1993).

CHAPTER 20

Introduction to Epidemiologic Methods

I. II. III. IV. V. VI.

JENNIFER L. KELSEY

Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305

MARYFRAN SOWERS

Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan 48109

Introduction Descriptive and Analytic Studies Study Designs Some Useful Epidemiologic Concepts Some Frequently Used Statistics Criteria for Deciding Whether an Association Is Causal

VII. Sample Size Considerations VIII. Measurement Error IX. Measuring Diet and Bone Turnover Status as Examples of Measurement Issues X. Conclusions References

I. INTRODUCTION

magnitude and impact of diseases or other conditions in populations or in selected subgroups of populations. This information can be used in setting priorities for investigation and control, in deciding where preventive efforts should be focused, in evaluating the efficacy of therapeutic procedures, and in determining what type of treatment facilities are needed. For instance, knowledge of current hip fracture incidence rates in various parts of the world and projected large increases in the numbers of elderly in developing countries indicate that hip fractures will become major problems in all parts of the world in the future [3]. Accordingly, identifying risk factors in these regions and planning for more services become high priority. Epidemiologic studies may also be used to learn about the natural history, clinical course, and pathogenesis of diseases. Currently, several studies are being undertaken of the clinical course of vertebral osteoporosis. One study [4], for instance, reported that

Many of the chapters in this section focus on epidemiologic aspects of osteoporosis. Since many readers will have only a passing acquaintance with the terms, methods, and concepts used in epidemiology, we start this section with an introduction to epidemiologic methods. It is hoped that this introduction will be of help in reading subsequent chapters in this section, in reading and evaluating studies directly from the published literature, and in understanding some of the reasons that different results sometimes are reported from various studies of the same issue. Much of this material is adapted from a textbook of epidemiology [1] and from a paper on epidemiologic methods [2]. Epidemiology is the study of the occurrence and distribution of diseases and other health-related conditions in populations. It is used for many purposes. One is to determine the

OSTEOPOROSIS, SECOND EDITION VOLUME 1

535

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

536

KELSEY AND SOWERS

minor fractures, unless multiple, are not associated with back pain or loss of height. Another [5] found that most back pain cannot be attributed to vertebral deformities and that vertebral deformities cause significant pain, disability, or height loss only if vertebral height ratios fall four standard deviations below the normal mean. Most commonly, epidemiologic studies are undertaken to identify causes of disease, and it is this application of epidemiology that is the focus of this chapter.

II. DESCRIPTIVE AND ANALYTIC STUDIES A. Descriptive Epidemiology Descriptive studies provide information on patterns of disease occurrence in populations according to such attributes as age, gender, race, ethnicity, marital status, social class, occupation, geographic area, and time of occurrence. Usually, routinely collected data such as from hospital discharge records, death certificates, and general health surveys are used for descriptive studies. This information can be used to indicate the magnitude of a problem or to provide preliminary ideas about etiology. For instance, several decades ago, the marked loss of bone mass and increasing hip fracture incidence rates in women after age 50 or so suggested that menopause and its accompanying decreased estrogen levels might be involved in the etiology of osteoporosis [6]. This hypothesis has been borne out by many subsequent studies [7,8]. The variation in hip fracture incidence rates around the world has led to several hypotheses about reasons for the differences, including differences in diet, physical activity, frequency of falling [9], and, most recently, hip axis length [10]. The increasing hip fracture incidence rates in northern European countries have led to a great deal of speculation about reasons for the increases [9]. Correlations between a putative risk factor and a disease according to geographic region or over time, however, at most provide weak evidence that the factor causes the disease. There are so many differences between lifestyles and other characteristics of people living in different geographic areas and at different periods of time that singling out one factor as being the reason for the difference in incidence rates is usually impossible. Countries with high incidence rates of hip fracture compared to those with low incidence rates have many differences in their diets, as well as different levels of physical activity, neuromuscular functioning, medication use, and perhaps hip axis length and propensity to fall. Accordingly, analytic epidemiologic studies designed specifically to test hypotheses are used to provide more definitive information.

B. Analytic Epidemiology Analytic studies are designed to test causal hypotheses that have been generated from descriptive epidemiology, clinical observations, laboratory studies, and other sources, including analytic studies undertaken for other purposes. Whereas descriptive epidemiology describes how a disease is distributed in a population, analytic epidemiology tries to explain why. Because analytic studies often require the collection of new data, they tend to be more expensive than descriptive studies, but, if designed and executed properly, generally allow more definitive conclusions to be reached about causation. Most epidemiologic studies are observational; i.e., the investigator observes what is occurring in the study populations of interest and does not interfere with what he or she observes. For instance, an investigator could observe existing physical activity levels among individuals and relate those activity levels to bone mass or fracture occurrence. In contrast, in an experimental study, the investigator intervenes and assigns members of the study population to one exposure or treatment category or another, as in a randomized clinical trial. In such a trial, an investigator would randomly assign individuals (or communities) to programs with varying levels of physical activity and note changes in bone mass or the occurrence of fractures following implementation of the programs. Observational epidemiologic studies, to be described first, include case-control, cohort, and cross-sectional studies, as well as some hybrid designs. Then, experimental studies will be discussed briefly.

III. STUDY DESIGNS A. Case-Control Studies Case-control studies are those in which the investigator selects persons with a given disease (the cases) and persons without the given disease (the controls) for study. Usually the cases enter the study as they are diagnosed over time, and controls enter the study as they are identified over the same time period. The proportion of cases and controls with certain characteristics or past exposure to possible risk factors (e.g., ever used estrogen) are then determined and compared. For instance, Table 1 shows that in a case-control study of hip fractures in women [11], 16.6% of hip fracture cases and 23.5% of controls had ever used estrogen, suggesting some protection from use of estrogen. For a numerical measurement such as weight, the mean level of the characteristic of interest in the cases is compared to the mean level of the characteristic in the controls. In that same case-control study [11], the mean Quetelet index [weight (kg)/height2 (m2)] was 22.3 in cases and 25.6 in controls, suggesting an increased risk among thin women.

537

CHAPTER 20 Introduction to Epidemiologic Methods

TABLE 1

Case-Control Study Relating Hip Fracture in Women to Previous Estrogen Use Hip fracture Cases

Ever used estrogen Yes No

Controls

Total

26

39

65

131

127

258 323

Total

157

166

Percent having used estrogen

16.6

23.5

Note. Source: Nieves et al. [11].

Cases are generally persons seeking medical care for the disease. Only newly diagnosed cases are usually included in order to be more certain that the risk factor preceded the disease rather than being a consequence of the disease and so that rapidly fatal cases or cases of short duration are appropriately represented. For instance, fracture cases tend to change their physical activity patterns after the fracture, and if cases who had incurred their fracture in the past are included, it would be more difficult to differentiate the physical activity pattern that preceded the fracture from the physical activity pattern that resulted from the fracture. Many case-control studies have been undertaken to identify risk factors for hip fracture. Because hip fracture is almost always seen in a hospital, virtually all cases can be identified from hospital sources. For fractures of many other common sites, such as distal forearm fractures, only a select portion of all cases would be seen in hospital, so the representativeness of cases in hospital settings would be highly questionable. Efforts would have to be made to identify cases seen on an outpatient basis as well, a considerably more extensive undertaking in most settings. Choice of an appropriate control group is often one of the most difficult and controversial aspects of designing a case-control study. A useful working concept of what a control group should be has been provided by Miettinen [12]: the controls should be selected in an unbiased manner from those individuals who would have been included in the case series if they had developed the disease under study. The choice of which control group to use generally depends on the source of the cases, the relative costs of obtaining the various types of controls, and the facilities available to the investigator. If cases consist of all individuals developing the disease of interest in a defined population, then the single best control group would generally be a random sample of individuals (in the same age range and of the same sex, for instance) from the same source population who have not developed the disease. If cases are identified at certain hospitals that do not cover a defined geographic area, it is usually impossible to specify the source population from

which the cases arose. In this situation, controls are often chosen from among other patients admitted to similar services of the same hospitals as the cases, as one wants to obtain a source of controls subject to the same selective factors as the cases. It is usually desirable to include as controls people with a variety of other conditions so that no one disease is unduly represented among the control group. Generally, it is important to exclude potential controls who have had their disease for a long time because, like the cases, the presence of their disease may have influenced their exposure to possible risk factors. Such characteristics as physical activity, diet, weight, and medication use may change as a result of many diseases. Controls in case-control studies of hip fracture have included people from the same retirement community as the cases [13], people from the same prepaid health care plan as the cases [14], people selected at random from lists of Medicare recipients of the Health Care Financing Administration (HCFA) files [15], people sampled from the general population of the same geographic areas as the cases [16], and patients seen at the same hospitals as the cases for other conditions [17,18]. Obtaining controls through random digit dialing, a procedure frequently used in case-control studies in younger populations, often is not practical for case-control studies that include large numbers of elderly individuals, as an enormous number of telephone calls must be made to find an appropriate elderly person. In countries or other geographic units with population registries, controls might consist of a random sample of persons from the same geographic area as the cases in the appropriate age groups, as listed in the register. The various types of control groups have their own strengths and weaknesses. If hospital controls are used, the controls by definition are different from the cases in that they generally have another disease for which they have sought medical care. If smoking is the putative risk factor, for instance, there may be concern that hospital controls include more than their fair share of smokers, as smoking is associated with many diseases that require hospitalization. A major concern in using controls from HCFA files or obtained though random digit dialing in younger age groups is that a substantial proportion of potential controls (typically 30 – 40% in otherwise well-executed studies) may decline to participate, and it is possible that participants and nonparticipants differ in ways that affect study results. Cases and controls from prepaid health care plans or from retirement communities are generally more likely to participate in studies, thus giving higher response rates. In some situations when no single control group is obviously best, it may be helpful to have more than one control group with which to compare the cases. Information on exposure to putative risk factors may be obtained in several ways, depending on the nature of the exposure. Risk factor data are obtained most commonly by

538 means of questionnaires administered by trained interviewers to cases and controls. For instance, the only practical way to find out about a person’s smoking habits is to ask the person. Existing records may sometimes be used to find out about exposures such as medication use. Physical measurements or laboratory tests on sera or other tissue drawn from cases and controls may also be used, but it must be kept in mind that measurements of such attributes as bone density or markers of bone turnover made after the fracture has occurred may differ from the values of these attributes before the fracture occurred. Whichever methods are used, ensuring that ascertainment of exposure status is comparable in cases and controls is of the utmost importance. Certain cases and controls may be excluded from a study, such as those with other disorders that affect calcium metabolism and that are not of interest to the study being conducted. Although excluding cases and controls limits generalizability, the validity of the comparison between cases and controls must take highest priority. The general principle that the same exclusion criteria should be applied to cases and controls should be maintained whenever possible. If cases are restricted to a certain sex or age range, controls should be similarly restricted. If cases with certain medical conditions are excluded, then controls with those conditions should also be excluded. While equal application of exclusion criteria may sound reasonable and easy, in practice this may be more difficult. Undiagnosed disease such as Paget’s disease may exist among controls in a casecontrol study of hip fracture, as the controls may not have had as thorough a diagnostic workup as the cases. Inequitable access to health care between cases and controls can exacerbate this problem. In summary, case-control studies can provide much useful information about risk factors for diseases, including hip fracture and other fractures, in settings where fractures can be ascertained readily. Case-control studies are by far the most frequently undertaken type of analytic epidemiologic study. They can generally be carried out in a much shorter period of time than cohort studies (to be discussed later), do not require nearly so large a sample size, and consequently are less expensive. For a rare disease, case-control studies are usually the only practical approach to identifying risk factors. Certain potential problems and limitations that may arise in case-control studies need to be carefully considered before deciding whether a case-control study is appropriate in a given situation. Sackett [19] and Austin et al. [20] have listed and discussed a large number of possible sources of bias and error in case-control studies. Among the most common concerns are that (i) information on potential risk factors may not be available either from records or the participants’ memories, (ii) information on other relevant variables may not be available either from records or from the participants’ memories, (iii) cases may search for a cause for their

KELSEY AND SOWERS

disease and thereby be more likely to report an exposure than controls, (iv) the investigator may be unable to determine with certainty whether the agent was likely to have caused the disease or whether the occurrence of the disease was likely to have caused the person to be exposed to the agent, (v) identifying and assembling a case group representative of all cases may be unduly difficult, (vi) identifying and assembling an appropriate control group may be unduly difficult, and (vii) participation rates may be low. Because of these potential weaknesses, the case-control study is considered by some to be a type of study that merely provides leads to be followed up by more definitive cohort studies. However, decisions as to whether preventive actions should be taken often must be reached on the basis of information obtained from case-control studies. Each casecontrol study should be evaluated individually, as some studies are affected by error and bias very little and others a great deal.

B. Cohort Studies In a typical prospective cohort study, persons free of the disease of interest at the time of entry into the study are classified according to whether they are exposed to the risk factors of interest. The cohort is then followed for a period of time (which may be many years), and the incidence rates (number of new cases of disease per population at risk per unit time) or mortality rates (number of deaths per population at risk per unit time) in those exposed or not exposed are compared. A prospective cohort study may also involve measuring exposure status at the beginning of a study and determining how this relates to changes in an attribute (such as bone mass) over time. Cohort studies have a major advantage over case-control studies in that exposures or characteristics of interest are measured before the disease has developed (or before changes in an attribute take place). However, prospective cohort studies generally require large sample size, long-term follow-up of study subjects, large monetary expense, and complex administrative and organizational arrangements. The outcome of interest must be relatively common, or prohibitively large numbers of cohort members will be required in order to ensure adequate numbers experiencing that outcome. Therefore, prospective cohort studies are usually initiated under two circumstances: (1) when sufficient (but not definitive) evidence has been obtained from less expensive studies to warrant a more expensive cohort study and (2) when a new agent (e.g., a new widely used medication) is introduced that may alter the risk for several diseases. Results from a prospective cohort study of the association between bone mass measured at the mid-radius and fractures at all sites among a cohort of 521 Caucasian women of ages 22 – 95 years followed for an average of 6.5

539

CHAPTER 20 Introduction to Epidemiologic Methods

years are shown in Table 2 [21]. It can be seen that as bone mass increases, incidence rates for fracture decrease. Before this study was undertaken, it was considered possible on the basis of results from case-control studies that low bone mass could be a consequence rather than a cause of fractures. This prospective cohort study clearly showed that low bone mass preceded the fractures. In Table 2 the column headed “Person-years of followup” should be noted. The figures in this column indicate the total number of years that women in this cohort were actually under observation for the occurrence of a fracture. Because women entered and left the cohort at different times, the total number of years that each woman was under surveillance by the investigators for fracture occurrence had to be taken into account. The sum of the number of years each woman in a given bone mass category was at risk and under observation is the number of person-years for that category. In most studies, cohort members are removed from follow-up when they experience a first event because they are no longer at risk for being an incident case. In this study, fractures of different sites were each counted so that women were still considered at risk for a fracture of other sites even though they experienced a fracture of one site. Studies of bone mass are increasingly being designed as longitudinal studies in which individuals are measured repeatedly over time. This enables one to track an individual’s changes in bone mass or in a marker of bone turnover over time, or to track an entire cohort’s average changes in bone mass or in a marker of bone turnover over time. In contrast, in a cross-sectional study (see later), one could examine the relationship between age and bone mass in a group of people at one point in time, but could not determine how bone mass changes in individuals as they become older. When the same individuals are measured repeatedly over time, one can also more readily determine

TABLE 2 Prospective Cohort Study Relating Bone Mass (g/cm) at Midradius to Number of Fractures of All Sites in Caucasian Women

Bone mass

No. of fractures

Personyears of follow-up

Incidence rate per person-year

0.60

46

415.5

0.111

0.60 – 0.69

25

554.2

0.045

0.70 – 0.79

46

861.1

0.053

0.80 – 0.89

15

776.8

0.019

0.90 – 0.99

5

521.1

0.010

1.00

0

260.2

0

Note. Source: Hui et al. [21].

whether a change is attributable to age or to a specific event, such as menopause, that is correlated with age. Sowers et al. [22] provided an example of a prospective cohort study to identify attributes associated with changes in radial bone mineral density. They found that lower weight, smaller triceps, and less arm muscle mass were predictive of increased 5-year loss of bone mineral density in postmenopausal women, whereas use of estrogen for 5 years or longer and current use of thiazide diuretics were predictive of less loss of bone mineral density. Another example of repeated measurement of one variable over time would be a study to assess drift in a densitometer over a period of several months by measuring a phantom weekly over this period. Statistical analysis of repeated measurements requires specific methods that take into account that the measurements over time on each individual tend to be correlated with each other. (See Diggle et al. [23] for a description of methods of analysis of longitudinal data.) Cohorts are sometimes chosen because they are representative of the general population, such as in the Framingham Heart Study [24]. Studies to identify risk factors for hip fracture and loss of bone mass have been able to take advantage of data collected on these cohorts, even though they were not originally designed for this purpose. Although the ability to generalize from such studies makes them highly desirable, they are usually very expensive and tend to be associated with relatively large numbers of people lost to follow-up. Also, an exposure of interest may be uncommon in the general population so it sometimes may be more efficient to select a cohort with a higher proportion exposed or to select a cohort at higher risk of the disease, so that the sample size does not have to be so large. Other cohorts used in studies of fractures include the residents of retirement communities [25], members of prepaid health care plans [26], the Nurses’ Health Study cohort [27], and women recruited from available listings in four areas of the United States [28]. A related type of study is the retrospective cohort study (also called a historical cohort study). In this design, investigators assemble a cohort by reviewing records to identify exposures in the past (often decades ago). Based on recorded exposure histories, cohort members are divided into exposed and nonexposed groups or according to level of exposure. The investigator then reconstructs their subsequent disease experience up to some defined point in the more recent past or up to the present time. For instance, 35,767 people in a county in Norway participated in a health screening program in 1984 – 1986. Each participant had filled out a questionnaire on some background factors and personal health characteristics, including smoking habits, health status, weight, height, and physical activity. By linking this information to hospital admissions for hip fracture over the 3-year period 1986 – 1989, Forsen et al. [29] were able to determine that cigarette smoking was

540

KELSEY AND SOWERS

a risk factor for hip fracture independent of thinness, physical inactivity, and self-reported ill health. In the United States, the first National Health and National Examination Survey (NHANES I) has also provided information on potential risk factors for retrospective cohort studies. People examined in NHANES I during 1971 – 1975 were followed through 1987 in the NHANES I Epidemiologic Follow-Up Study. Looker et al. [30] related data obtained during 1971 – 1975 on calcium consumption of participants in NHANES I to subsequent occurrence of hip fracture. These data suggested that calcium may lower the risk for hip fracture in women who were at least 6 years past menopause and who did not use hormone replacement therapy. Retrospective cohort studies have many of the advantages of prospective cohort studies, but can be completed in a much more timely fashion and are therefore much less expensive. However, only when the necessary information on past exposure and other characteristics of interest has been reliably recorded can a retrospective cohort study be reasonably undertaken. In addition, it must be possible to trace a large proportion of the cohort members in order to determine whether they in fact developed the disease of interest. As with prospective cohort studies, it is usually feasible to carry out a retrospective cohort study only when the outcome of interest is relatively common. It is also frequently important to obtain information on characteristics of the cohort members other than the exposure history and the outcome of primary interest so as to make sure that those with and those without the exposure of interest are comparable in other relevant respects. If such information is not available, then the interpretation of the study results may be ambiguous. Thus, retrospective cohort studies are economical and useful when the necessary information on the past history of study subjects is available and when the outcome of interest is relatively common. In many situations, however, these conditions are not met.

C. Cross-Sectional Studies In a cross-sectional, or prevalence, study, exposure to a hypothesized risk factor or other characteristic of interest and the occurrence of a disease are measured at one time (or over a relatively short period of time) in a study population. Prevalence rates of disease (number of cases of existing disease per population at risk at a given point in time or time period) among those with and without the exposure or characteristic of interest are then compared. For a quantitative variable such as bone mineral density, the mean value of the exposed and nonexposed groups may be compared. For instance, in a cross-sectional study of the association between back muscle strength and spinal osteoporosis, it was found that osteoporotic women had lower muscle

strength than women without osteoporosis [31]. However, in cross-sectional studies, it is often difficult to differentiate cause and effect. As the authors themselves pointed out, a longitudinal study would be needed to determine whether weak back muscles contribute to the development of osteoporosis or are a consequence of osteoporosis. Cross-sectional studies of the association between calcium supplementation and bone mineral density would be difficult to interpret because people with low bone mineral density might take calcium once they were told about their low bone mineral density. Interpretation of findings from crosssectional studies is generally clear only for potential risk factors that will not change as a result of the disease, such as genotype. Prevalence studies include all cases of disease, new and old. Therefore, a second limitation of cross-sectional studies is that the case group tends to be weighted toward individuals with disease of long duration, as the chances for cases of long duration to be included are greater than those for cases who recover or die quickly. Thus, any associations found between an exposure and a disease may be more applicable to survivorship with disease rather than development of disease. Another use of prevalence studies is simply to describe the prevalence of a disease or condition in the population. For such studies to be useful, it is important that the individuals studied be representative of the population to which the results are to be generalized. Patients seen in tertiary care centers or in the practice of any one physician are seldom representative of all persons in the community with a disease, many of whom may not have even sought medical care. Accordingly, generalizations from such select groups of patients should be avoided.

D. Hybrid Study Designs It is sometimes possible to design a case-control study within either a retrospective or prospective cohort study. Consider a traditional cohort study in which an investigator wishes to find out whether a positive test result from a certain expensive serologic test is associated with an increased risk of hip fracture. In such a traditional cohort study, the investigator might start with blood samples drawn from 10,000 people free of hip fracture. The cohort might then be followed for 10 years to determine the incidence rate of hip fracture in those positive and in those negative on the serologic test. A modification of this traditional cohort design, called a nested case-control study, is illustrated in Fig. 1. The blood samples from the 10,000 people could be frozen and stored. Suppose that after 10 years had elasped, 200 people had incurred a hip fracture and 9800 had not. The stored serum samples from the 200 cases and a sample of, say, 400 of the

541

CHAPTER 20 Introduction to Epidemiologic Methods

FIGURE 1

Hypothetical nested case-control study of association between a serologic marker and hip

fracture.

9800 without the disease could then be tested. This sampling of nondiseased people would greatly reduce the cost from what it would be if the sera from all 10,000 cohort members had to be tested, yet the serologic status before disease occurrence would be measured. The proportion testing positive among the cases could then be compared to the proportion testing positive among the controls, as in a usual case-control study. Controls are selected from unaffected cohort members who are still alive and under surveillance at the time the cases developed the disease. Typically, the controls are matched to cases according to age, sex, and time of entry into the cohort. The availability of a variety of banks of stored serum around the world and the current interest in serologic predictors of disease make nested case-control studies an attractive and economical approach, as long as the serologic marker of interest does not undergo degradation over time. A case-cohort study is another method of increasing efficiency compared to a traditional retrospective or prospective cohort study. Like the nested case-control study, all cases and a sample of controls are selected. However, controls are sampled from the entire cohort, not just those free of disease, and are not matched to the cases. Rather, other relevant variables are taken into account in the statistical analysis. A case-cohort design is particularly useful when the associations between a serologic marker or other variable and several diseases are of interest.

E. Experimental Studies In general, the strongest evidence that a given exposure is a cause of a disease is produced from experimental

studies. In experimental studies, the investigator randomly assigns study subjects either to be exposed or not exposed to an agent, and then follows them through time to see what proportion of the exposed and unexposed develop certain diseases. Thus, the possibility that people choosing to be exposed to a certain factor are systematically different from those who do not is eliminated. Randomized clinical trials, which are one type of experimental study, have provided convincing evidence that estrogen replacement therapy protects against loss of bone mass, at least as long as it is used [7,8]. However, the protective effect of estrogen replacement therapy against coronary heart disease that has been reported from many observational studies was not uniformly accepted as causal because the association had not been tested in randomized trials, even though randomized trials have shown favorable effects on high-density lipoprotein cholesterol and low-density lipoprotein cholesterol levels [32]. Some people believed that women who use estrogen replacement therapy would be at lower risk for coronary heart disease even if they did not use estrogen. Users of replacement estrogen tend to be more physically active, healthier, and younger than women who do not use estrogen. Such characteristics would be associated with less coronary heart disease regardless of estrogen use [32]. In fact, the Heart Estrogen/Progestin Study (HERS), a randomized trial in women with previously diagnosed coronary heart disease (32a), found an initial increase in risk after initiation of use of an estrogen/progestin compound, followed by an apparent protective effect with increasing length of use. A more detailed description of issues that arise in randomized trials is included in Chapter 64.

542

KELSEY AND SOWERS

IV. SOME USEFUL EPIDEMIOLOGIC CONCEPTS A. Confounding Confounding is an issue that is especially likely to arise in observational studies. A statistical association between an exposure or other characteristic and a disease does not necessarily mean that one is a cause of the other. In ruling out other explanations for an association, confounding variables need to be considered. For instance, an investigator wishing to determine whether coffee drinking increases the risk for osteoporosis would have to be concerned with whether any observed statistical association between coffee consumption and osteoporosis was actually attributable to the tendency of coffee drinkers to smoke more cigarettes, not to use estrogen, to be thinner, or to have some other characteristic that puts them at an elevated risk for osteoporosis. Such variables are considered potential confounding variables and need to be considered in virtually all observational epidemiologic studies. A confounding variable is defined as a variable (e.g., cigarette smoking) that (i) is causally related to the disease under study independently of the exposure or characteristic of primary interest (e.g., coffee drinking) and (ii) is associated with the exposure or characteristic of primary interest in the study population, but (iii) is not a consequence of this exposure. Confounding variables may be taken into account in the study design by matching on the confounding variables, such as when cases are matched to controls on age and sex in a case-control study. For instance, Sowers et al. [33] described different hormone levels in premenopausal women with low bone mass compared to controls without low bone mass matched to cases by age, weight, and parity. Alternatively, confounding variables may be taken into account in the statistical analysis by using multivariate statistical methods, such as the Mantel – Haenzel procedure, logistic regression, Cox proportional hazard model, Poisson regression, or multiple linear regression. In the study in Table 2 relating bone mass to fracture risk, age was subsequently controlled in the analysis by a modification of Poisson regression [21]. Both matching in the study design and controlling in the analysis are valid ways of adjusting for confounding variables, and in fact it is possible to match roughly on certain variables in the study design and then control more finely in the analysis. Matching is used most frequently in case-control studies. The main considerations in deciding whether to match in the study design or control in the analysis in a case-control study are whether a given variable really is likely to be a confounder, the cost of obtaining information on the confounding variable so that it can be matched on in the study design, and whether the confounder is strongly related to the

disease and the exposure. The reader is referred elsewhere [1] for a discussion of which method of controlling confounding is more efficient under which circumstances. The same procedures that are used to determine exposure to putative risk factors are used to ascertain exposure to confounding variables, including questionnaires, medical records, laboratory tests, physical assessment, and special procedures. Measurement of potential confounding variables is highly important, as otherwise they cannot be controlled adequately in the analysis.

B. Effect Modification Effect modification, sometimes referred to as statistical interaction, also needs to be considered when studies are designed, analyzed, and interpreted. It occurs when the magnitude of the association between one variable and another differs according to the level of a third variable. For instance, it has been hypothesized [34], with some supporting empirical evidence [35], that any beneficial effect of supplemental calcium will be seen mainly in persons with low dietary calcium intake. Thus, the effect of supplemental calcium is modified by dietary calcium intake. It has been noted in at least one study [24] that use of estrogen replacement therapy in the immediate postmenopausal period protects against low bone mineral density only in women younger than about age 75 years. In other words, the effect of estrogen is modified by a person’s age (or years since menopause). Forsen et al. [29] found that the effect of cigarette smoking on hip fracture risk was greater in thin women than in heavy women. Detecting effect modification is an important component of the analysis of epidemiologic data.

V. SOME FREQUENTLY USED STATISTICS A. Relative Risk In cohort studies, the strength of the association between a putative risk factor and a disease is often measured by what is called a relative risk (or, more technically, a rate ratio or risk ratio; a discussion of the difference between a rate ratio and a risk ratio is beyond the scope of this chapter). A relative risk is simply the risk (or incidence rate) of disease in one group divided by the risk (or incidence rate) of disease in another group. For instance, a relative risk of 0.111/0.045  2.47 for fracture among women with bone mass of less than 0.60 g/cm compared to women with bone mass of 0.60 – 0.69 g/cm in Table 2 indicates that the former group has about two and one-half times the risk of

543

CHAPTER 20 Introduction to Epidemiologic Methods

fractures as women in the latter group. A relative risk of 0.68 for hip fracture among persons using thiazide diuretics compared to those not using them [36] indicates that the risk of hip fracture is reduced by almost one-third among users of these drugs. Relative risks give a meaningful idea of the extent to which an exposure elevates or decreases risk for disease and are important in assessing whether a causal relationship between an exposure and a disease exists.

B. Odds Ratio In case-control studies, risks and incidence rates are generally not available, so relative risks cannot be calculated. Instead, the odds ratio (ratio of exposed to nonexposed among cases divided by the ratio of exposed to nonexposed among controls) is calculated. It can be shown [37] that for all but the most common diseases ( 10% of the population affected, for instance), the odds ratio is a good approximation to the relative risk and can be interpreted similarly. In Table 1, an odds ratio of (26  127)/(39  131)  0.65 can be calculated, indicating that women who have ever used estrogen have about a 35% reduction in their risk of experiencing a hip fracture compared to women who have never used estrogen. In this same case-control study [18], an odds ratio of 1.9 was calculated for the association between lower extremity dysfunction and hip fracture, meaning that the risk of hip fracture is almost twice as great among those with lower extremity dysfunction as among those without.

C. Confidence Interval A confidence interval is often presented along with the estimate of the relative risk or odds ratio (or other parameter) in order to give a range of plausible values for the parameter being estimated. Confidence intervals provide more information than can be obtained simply by testing for statistical significance. A 95% confidence interval of 1.46 – 2.75 around a point estimate of relative risk of 2.00, for instance, gives the likely range of values for the true relative risk and indicates that a relative risk of less than 1.46 or greater than 2.75 can be ruled out with 95% confidence. The 95% confidence interval around the odds ratio of 1.9 for the association between lower extremity dysfunction and hip fracture mentioned earlier was 0.9 – 3.8 [18]. That 1.0 is included in the interval indicates that this association is not statistically significant at the P  0.05 level, although an odds ratio of greater than 1.0 is still quite likely and is certainly biologically plausible.

D. Statistically Adjusted Relative Risk or Odds Ratio When interpreting relative risks or odds ratios, the effects of confounding variables need to be taken into account. Statistical procedures for making adjustments for confounding variables are available and are described in textbooks of biostatistics [38] and epidemiology [1]. Briefly, a commonly used procedure for making statistical adjustments for confounding variables in case-control studies and cross-sectional studies when a disease is either present or absent (e.g., fracture vs no fracture) and when there are small numbers of confounding variables is the Mantel – Haenszel procedure. For instance, data in Table 1 might be subdivided by age group (which may be considered a potential confounding variable) and then summarized by the Mantel – Haenszel procedure to obtain an odds ratio adjusted for age. Logistic regression is used frequently when a disease is either present or absent and when there are several potential confounding variables or when a potential confounding variable is continuously distributed (e.g., weight). For instance, if an investigator wanted to examine the relationship between estrogen and hip fracture adjusting for age (measured as actual years of age or in broad age groups), weight, and several other variables simultaneously, he/she would use logistic regression to obtain an estimate of the odds ratio for the association between estrogen use and hip fracture, adjusted for any differences in the distributions of age, weight, and other variables between users and nonusers of estrogen. If cases and controls have been matched in the study design, then statistical methods that take the matching into account need to be employed. Procedures for making adjustments in cohort studies are based on similar principles, but must take into account the varying periods of time that different cohort members usually are followed and under observation. A form of the Mantel – Haenszel procedure that takes into account the length of follow-up of each cohort member may be used for small numbers of categorized potential confounding variables. Methods of analysis called Poisson regression and the Cox proportional hazard model can be used to calculate rate ratios adjusted for multiple potential confounding variables, taking into account the length of time that each cohort member has been followed. When the outcome of interest is a continuously distributed variable, such as bone mineral density or changes in bone mineral density over time, multiple regression can be used to estimate the relationship between an independent variable (e.g., age) and a dependent variable (e.g., bone mineral density), adjusting for potential confounding variables. If effect modification is present, statistics should be presented separately for groups in which the effects differ from each other.

544

VI. CRITERIA FOR DECIDING WHETHER AN ASSOCIATION IS CAUSAL Because most epidemiologic studies are observational rather than experimental, participants will usually differ in respect to other characteristics than just the exposure and disease of primary interest to the investigator. Sometimes confounding variables can be recognized, measured, and accounted for, but often they are unknown or only vaguely hypothesized. Also, any one study may have certain methodological deficiencies or may produce certain results by chance. Therefore, seldom will a single epidemiologic study provide definitive evidence for or against a hypothesis. Even results from several studies may not be convincing. With these considerations in mind, epidemiologists have developed criteria to be used as tests of whether a causal association exists. Not all criteria need to be fulfilled in all instances, nor are all equally important, but taken together they provide useful guidelines for determining whether an association between a given exposure and disease is causal. 1. Strength of association. For a positive association, the measure of association (relative risk or odds ratio) should be elevated, indicating that the exposed are at increased risk of disease over the unexposed or that those with disease are more likely to have histories of exposure than those without the disease. The greater the magnitude of these measures, the more likely the association is to be causal. As a rough rule of thumb, a relative risk or odds ratio of 2 indicates a moderate elevation in risk and a relative risk or odds ratio of 3 or more is considered strong. If no association between exposure and disease exists, the issue of causality does not arise, so establishing an association is an essential first step; the stronger the association, the more convincing is this aspect of the argument. Women with stroke, for instance, are 4 – 5 times more likely to fracture their hip than women not having had a stroke, making it more likely that this association is causal than if the relative risk were 1.5. Women in the highest quintile of body mass index (weight/height2) have only one-fifth the risk of experiencing a hip fracture compared to those in the lowest quintile [18]. 2. Statistical significance. A finding of statistical significance means that the result is unlikely to be a consequence of chance. Statistical significance depends on both the strength of the association and the number of people included in a study. If the sample size is inadequate, even relatively strong associations may not demonstrate statistical significance. Conversely, a tiny, biologically meaningless elevated risk can become “significant” with a very large sample size. For instance, among the 9704 women included in the baseline survey of the Study of Osteoporotic Fractures, lifetime caffeine consumption was inversely associated with bone mass, such that the

KELSEY AND SOWERS

equivalent of 10 cups of coffee per day over a period of 30 years was associated with a 1.1% decrease in radial bone mineral density [28]. Although statistically significant, the clinical significance of this finding is probably slight. 3. Ruling out alternative explanations. Once a significant association has been established (i.e., the exposure and disease are related, and the relationship is unlikely to be attributable to chance), other explanations for the observed association, such as methodologic deficiencies and confounding, should be carefully considered and ruled out. As mentioned earlier, if an association is found between coffee drinking and osteoporosis, it must be determined whether the association exists only because people who drink coffee also tend to smoke, be thin, not use estrogen, or have some other characteristic that influences their risk for osteoporosis. 4. Dose – response relationship. If increasing dose or length of exposure is associated with increasing risk, then the case for causality is considerably enhanced, as the likelihood is reduced that such a pattern could arise by chance or be attributable to confounding. Increasing length of use of hormone replacement therapy, for instance, is associated with a decreasing risk of hip fracture. Increasing weight is associated with a decreased risk [18]. The absence of a dose – response relationship does not disprove causality, however, as other patterns of association, such as a threshold effect, could also occur. 5. Removal of exposure. If the presence of an exposure increases risk of disease and removing the exposure reduces risk, belief in a causal association is strengthened. When estrogen replacement therapy is stopped, loss of bone mass resumes [8], thus strengthening the belief that the association is indeed causal. 6. Time order. As mentioned earlier in the discussion of the association between low back muscle strength and spinal osteoporosis, sometimes it is not clear whether an exposure caused a disease or the disease caused the exposure. This problem is particularly notable in crosssectional studies, where prevalent disease and exposure are determined simultaneously. Time order is unique among the causal criteria in that if disease can be shown to precede exposure, causality is definitely ruled out. 7. Predictive power. Hypotheses regarding presumed causal associations that can in turn be shown to predict future occurrences lend strong support to the belief in the causality. 8. Consistency. If associations of similar magnitude are found in different populations by different study methods, the likelihood of causality is increased substantially, as all studies are unlikely to have the same methodologic limitations or study population idiosyncrasies. Virtually all studies, for instance, show that estrogen replacement therapy protects against loss of bone mass, at least in the early postmenopausal period.

545

CHAPTER 20 Introduction to Epidemiologic Methods

9. Coherence with experimental data. When available, the results of well-designed experiments in which exposures are assigned at random are very convincing because the only factor on which groups differ, except by chance, is the exposure of interest. Randomized trials [7,8], for instance, have clearly shown that women in their early postmenopausal years who are randomly assigned to use estrogen replacement therapy have significantly less loss of bone mass than women randomized to placebo. However, many exposures cannot be ethically or practically assigned at random. In addition, some well-controlled experiments on a few carefully selected people may have little relevance to the general, free-living population. 10. Biologic plausibility. When a new finding fits well with the currently known biology of a disease, it is more plausible than if a whole new theory must be developed to explain the new finding. Protection against hip fracture from obesity, for instance, is biologically plausible because of the increased estrogen production in obese women [39] and the protection against hip fracture afforded during a fall from fatty tissue around the hips [40]. Another possible means of enhancing biologic plausibility is through laboratory experiments. However, what occurs in a laboratory setting or in experimental animals may have limited relevance to humans. It should be obvious from the discussion just given that decisions on the likelihood of causality are of necessity partly judgmental. What one person may believe is a causal association, another person may not. Lilienfeld [41] divided the degree of evidence for causation into three levels. At the first level, the evidence is considered sufficient for further study. For instance, studies suggesting a protective effect of vitamin C against osteoporosis [11] should be followed up with further study. At the second level, the evidence is considered sufficient to warrant public health action, even if the causal association has not been definitively established. Many people believe that the possible protection against osteoporosis and associated fractures from calcium supplementation fits into this category [35]. At the third level, the evidence is so strong that the causal association is considered part of the body of scientific knowledge. There is general agreement that the protective effect of estrogen replacement therapy against loss of bone mass in early postmenopausal women is established with this degree of certainty [42].

VII. SAMPLE SIZE CONSIDERATIONS There is little point in undertaking a study to determine whether an exposure is associated with a disease unless the number of study subjects is large enough that the association is in fact likely to be detected. Similarly, under most circumstances it would be wasteful to go to the expense of

including far more study subjects than are actually needed. Thus, determining the optimal sample size is an important component of planning a study. Many statistical and epidemiologic textbooks provide formulae for sample size estimation [1,37,43]. The sample size required depends on several conditions, all of which enter into the equations given later. First, what risk is one willing to take that the null hypothesis (of no difference) is rejected when it is in fact true? This is the  value, usually taken to be 0.05, meaning that the investigator is willing to reject a null hypothesis incorrectly 1 out of 20 times. Smaller values of  will require larger sample sizes. Second, what risk is the investigator willing to take that the null hypothesis is not rejected when it should be? This is the  value, usually taken to be 0.10 or 0.20, meaning that the power to reject the null hypothesis of no association is 0.90 or 0.80, respectively. The greater the power (and thus the smaller the value of ), the larger the sample size that is needed. Third, how large a difference does one want to be able to detect? The smaller the difference, the larger the required sample size. Fourth, if the outcome of interest is a yes/no variable, what proportion of the population develops the disease in a cohort study or is exposed in a case-control study? Outcomes that affect roughly half the population will require smaller sample sizes than outcomes that are either very rare or very common. Thus, studying rare diseases with a cohort design or rare exposures with a case-control design requires enormous sample sizes. Fifth, what is the variance of what is being measured in the population? The greater the variance, the larger the sample size that will be needed. Two formulae that may be used in estimating required sample size are given here when the objective is to detect difference between two groups. To detect differences between means, the appropriate formula is n

(Z  /2 Z )2  2 (d*) 2 r

To detect differences between proportions, the appropriate formula is n

(Z /2 Z )2 p(1 p) (r 1) (d*) 2 r

where d* is the value of the difference in means or proportions that one wishes to be able to detect; n is the number of exposed individuals in a cohort (or cross-sectional) study or the number of cases in a casecontrol study; r is the ratio of the number of unexposed individuals to the number of exposed individuals in a cohort (or crosssectional) study or the ratio of the number of controls to the number of cases in a case-control study; r  1 if the numbers in the two groups being compared are equal;

546

KELSEY AND SOWERS

 is the standard deviation in the population for a continuously distributed variable; p1 is the proportion of exposed individuals who develop (or have) the disease in a cohort (or cross-sectional) study or the proportion of cases who are exposed in a case-control study; p2 is the proportion of unexposed individuals who develop (or have) the disease in a cohort (or cross-sectional) study or the proportion of controls who are exposed in a case-control study; p rp2 p is the weighted average of p1 and p2 , or p  1 1 r Z is called a standard normal deviate because it rescales distributions of measurements to have a mean of 0 and a standard deviation of 1, thus enabling one to use standard tables of the normal distribution. Most textbooks of basic statistics have tables that enable one to determine the level of significance (i.e., p value) that corresponds to a given value of Z. The symbol Z/2 refers to the standard normal deviate for a two-tailed test of statistical significance. A two-tailed test provides for the possibility that a difference between two groups might be either positive or negative. For   0.05, Z/2  1.96. Z is related to the probability that one fails to reject a null hypothesis that should in fact be rejected. For   0.20, Z  0.84; for   0.10, Z  1.28; for   0.05, Z  1.64. As an example, suppose that an investigator wants to undertake a case-control study to determine whether there is an association between alcohol use (categorized as yes versus no) and distal forearm fracture. Suppose the investigator knows that about 20% (i.e., p2  0.20) of the population drinks alcoholic beverages and that it is desired to detect a 10% difference between cases and controls (i.e., d*  0.10, p1  0.20 0.10  0.30). Further, suppose that the investigator wants to be 90% certain of detecting a difference of this magnitude (i.e.,   1 0.90  0.10) and wants to find a difference when there really is none only 5% of the time (i.e.,   0.05). Suppose that the desired ratio of controls to cases is 2:1 (i.e., r  2). Note that 0.30 2(0.20) p  0.23 1 2 Then, substituting into the second formula just given, n

(1.96 1.28)2 0.23 (1 0.23) (3)  279 cases (0.10)2 2

Number of controls  2  279  558. In trying to keep the sample size as small as possible, it is important that measurements be as precise as possible. With poor measurement, not only will a larger sample size be needed, but the estimated magnitude of an association will be a poor approximation to the true association. These issues are discussed next.

VIII. MEASUREMENT ERROR A. Nature of the Problem and Definitions A certain amount of measurement error is almost inevitable, whether in measurement of potential risk factors, disease status, or potential confounding variables. This discussion will be referring both to the validity or accuracy of a measurement, or the closeness with which the measurement approaches the true value, and to the reliability or reproducibility of a measurement, or the extent to which the same measurement is obtained on the same occasion by the same observer, on multiple occasions by the same observer, or by different observers on the same occasion. It is well known by investigators in the field of osteoporosis that it is difficult to measure some of the major putative environmental risk factors, such as diet, physical activity, coffee consumption, and alcohol consumption. Because much of this information is obtained from questionnaires, the quality of data obtained is often no better than the imperfect memory of individuals about such factors as their dietary habits or physical activity. Also, questionnaires would become much too tedious if too much detail were required of study subjects. Thus, in obtaining summary indicators of such variables as calcium consumption or physical activity, accuracy of measurement will, to some extent, be compromised. Data obtained from questionnaires present particular problems in case-control studies in which information is sought about exposures that took place decades before the disease manifests itself. If physical activity or diet during adolescence affects risk for fractures in older individuals, obtaining accurate information in a case-control study would in all likelihood be impossible. Even if information on an exposure is obtained from biologic assays, there is often no assurance that a single measurement is indicative of the cumulative exposure or the exposure at the time the disease was developing. Errors in classification of disease status are probably as small for fractures (except vertebral fractures) as for any other disease. However, if measures of bone mass are of interest, several measurement issues must be considered. First, can bone mass be measured accurately considering that marrow fat may distort the true measure of mineral? Second, the measure of bone density is areal, not volumetric. Use of conventional units (g/cm2) may cause selective misrepresentation of density. Third, risk factors such as body size influence the precision of the bone mass measure. Finally, while measures of bone density have been used as an indicator of fracture risk, this measure does not include all dimensions of bone structure that affect fracture risk, such as bone structure and microarchitecture. Errors in the measurement of confounding variables are also of concern. Often the critical confounding variables are just as difficult to measure as the exposure of primary interest.

547

CHAPTER 20 Introduction to Epidemiologic Methods

TABLE 3

Definitions of Sensitivity and Specificity True classification Present

Absent

TABLE 4 Hypothetical Example of Effect of Differential Misclassification of Exposure in Case-Control Study Relating Hip Fracture to Medication Use in Past 24 h Hip fracture cases

Imperfect classification Present

a

b

Absent

c

d

a c

b d

Total

Note. Sensitivity  a/(a c); false negative rate  c/(a c); specificity  d/(b d); false positive rate  b/(b d).

Controls

True status a Medication use

20



80

80

Total

100

100

20

Observed status with recall bias b Medication use

Some potential confounding variables, such as socioeconomic status, are difficult to conceptualize let alone measure. As will be discussed later, inadequate measurement of important confounding variables can lead to biases just as serious as those arising from errors in the measurement of exposure or disease. For ease of presentation, this discussion focuses mainly on measurement of binary variables (i.e., variables that take on only one of two values, such as disease present or absent or exposure present or absent). Toward the end of this section, we will briefly extend the discussion to quantitative variables such as bone mineral density. Sensitivity is defined as the proportion of those who truly have the characteristic that are correctly classified as having it by the measurement technique. Specificity is the proportion of those who truly do not have the characteristic that are correctly classified as not having it by the measurement technique. Table 3 shows how sensitivity and specificity may be calculated from a 2  2 table. The proportion false positive is 1 specificity and the proportion false negative is 1 sensitivity. Measurement of a binary characteristic is perfect only when sensitivity and specificity are both 100%. Unfortunately, sensitivity and specificity close to perfect are seldom achieved in practice. When sensitivity equals 1.00 minus specificity, the measurement method is no better than entirely random classification of study subjects. Measurement error is said to be differential if the magnitude of the error for one variable differs according to the actual value of another variable. Table 4 shows a hypothetical example of differential misclassification. Suppose a case-control study of hip fracture is being undertaken, and the exposure of interest is whether a certain medication was taken in the 24 h before the fracture occurred. Cases might want to blame the fracture on some external agent such as a medication and therefore might report use in the last 24 h when such use did not occur. Controls, however, might forget that they even had taken the medication in the past 24 h and therefore might underreport its use. Thus, an association between the medication and hip fracture such as that



25



75

85

Total

100

100

a b

15

True odds ratio  (20  80)/(20  80)  1.00. Observed odds ratio  (25  85)/(15  75)  1.89.

shown in the lower table might be observed when in fact no such association exists. Measurement error is said to be nondifferential when the magnitude of error for one variable does not vary according to the actual value of the other variable of interest. In other words, both sensitivity and specificity remain constant irrespective of the value of the other variables.

B. Effects of Nondifferential Misclassification of Discrete Variables Nondifferential misclassification in a 2  2 table always causes the measure of association (i.e., relative risk or odds ratio) to become closer to the null value. Table 5 shows what happens in a hypothetical case-control study in which an exposure is measured with sensitivity of 0.60 and specificity of 0.50, and misclassification is nondifferential. The true odds ratio of 2.22 would be observed to be only 1.06. With some variables, such as diet and physical activity, for which measurement error is undoubtedly substantial, it is difficult to know whether results showing no association between these variables and fractures occur because there really is no association or because measurement is poor. The extent of attenuation of measures of association depends in part on how common the exposure is. Table 6 shows a hypothetical case-control study in which the true odds ratio is 2.11, the exposure is fairly rare, the disease is measured without error, and the sensitivity of exposure measurement is 100% but the specificity is only 60%. (That is, 40% of those without the exposure are classified incorrectly as having the exposure.) It may be seen that the

548

KELSEY AND SOWERS

TABLE 5 Hypothetical Example of Nondifferential Misclassification of Exposure in Case-Control Study Cases

Controls

Total

True status in case-control study a Exposure

85

50

135



115

150

265

Total

200

200

400

Observed status in case-control study b Exposure

51 57  108

30 75  105



34 58  92

20 75  95

187

200

200

400

Total

213

True odds ratio  (85  150)/(50  115)  2.22. Assume misclassification in measurement of exposure, but not disease, and sensitivity  0.60 and specificity  0.50. Observed odds ratio  (108  95)/(105  92)  1.06. a b

observed odds ratio is 1.13 instead of 2.11. However, if the sensitivity is 60% (i.e., 40% of those with the exposure are classified incorrectly as not having the exposure) and the specificity 100%, then the odds ratio is reduced only to 2.06. Thus, for an uncommon exposure, it is important to have a

TABLE 6 Hypothetical Example of Effects of Different Values of Specificity and Sensitivity of Exposure Measurement on Attenuation of Odds Ratio in Case-Control Study with Uncommon Exposurea True status b Cases

Controls

Exposure

10

Total

5

90

95

100

100

Observed status c

Observed status d

Cases

Controls

Cases

46

43

6

Controls

Exposure

Total a

3

54

57

94

97

100

100

100

100

Assume misclassification in measurement of exposure, but not disease. True odds ratio  (10  95)/(5  90)  2.11. c If specificity  0.60 and sensitivity  1.00, observed odds ratio  1.13. d If specificity  1.00 and sensitivity  0.60, observed odds ratio  2.06. b

highly specific measure in order to obtain a good estimate of the odds ratio. With a highly prevalent exposure, the situation is different. Here, with high sensitivity but only fair specificity, the odds ratio is attenuated only slightly, whereas with high specificity but low sensitivity, the odds ratio is reduced almost to 1.00. In this situation, a highly sensitive measure is desirable. When measurement error occurs when controlling for potential confounding variables, additional problems occur. If a confounding variable is measured imperfectly, then controlling for the confounding variable in the analysis will not entirely remove its effect because it was not measured with sufficient accuracy. Accordingly, if an investigator controls for the effect of physical activity when considering the possible effect of alcohol consumption on risk for hip fracture, the relative risk could change from 2.0 without adjustment for physical activity to 1.5 with adjustment. The investigator would not know whether there still is an independent effect of alcohol consumption or whether physical activity had been measured accurately, there would be no residual association between alcohol consumption and hip fracture. Furthermore, if the confounding variable is measured perfectly but the exposure variable is not (or vice versa), then the effect of the measurement error is typically to induce apparent effect modification when none exists [44]. Table 7, for instance, presents a hypothetical case-control study of the association between previous use of replacement estrogen and hip fracture. Suppose that the true situation (data on left) is that in both younger and older women, use of estrogen is associated with an odds ratio of 0.47. However, note that fewer women in the older strata have used estrogen. Suppose that sensitivity is 80% but that specificity is only 50%. (In other words, 50% of women who have not used estrogen say that they have used it.) Because the effect of misclassification on the odds ratio will be greater when prevalence of exposure is lower, the odds ratio becomes much closer to 1.0 in the stratum of older women than in the stratum of younger women (data on right). Thus, because of the poor specificity, it appears that there is effect modification by age when none exists. If measurement of exposure were perfect but measurement of the confounder less than ideal, then similar apparent effect modification could be induced. When both the exposure and the confounder are subject to nondifferential measurement error, the effects are less predictable. The adjusted estimate of the odds ratio may be even more biased than the unadjusted estimate that ignores confounding entirely [44]. Considering that epidemiologic studies are frequently trying to measure and control for several variables with substantial measurement error (e.g., food consumption, caffeine consumption, physical activity, alcohol consumption, cigarette smoking), it is no wonder that results from different studies are inconsistent, as measurement techniques and

549

CHAPTER 20 Introduction to Epidemiologic Methods

TABLE 7 Hypothetical Example in Which Confounding Variable (Age) and Disease (Hip Fracture) Are Measured without Error and Exposure (Estrogen Use) Is Imperfectly Measured (but Nondifferentially) Observed statusc

True status Hip fracture cases

Controls

Ages 50 – 74a

Hip fracture cases

Controls

Ages 50 – 74 d

Estrogen

320

500

Estrogen

596

Estrogen

680

500

Estrogen

404

350

1000

1000

1000

1000 515

Total Age  75b

Total

650

Age  75 e

Estrogen

24

50

Estrogen

507

Estrogen

976

950

Estrogen

493

485

1000

1000

1000

1000

Total

Total

True odds ratio  0.47. True odds ratio  0.47. c Sensitivity for measurement of estrogen use  0.80. Specificity for measurement of estrogen use  0.50. d Observed odds ratio  0.79. e Observed odds ratio  0.97. a b

the prevalence of exposures and confounders differ from study to study. It should also be mentioned that in analyzing data from tables other than simple 2  2 tables, there are circumstances under which nondifferential measurement error can make an association appear larger than it really is. The reader is referred to an article by Weinberg et al. [45] for a discussion of such situations.

C. Error Correction Methods for Discrete Variables Sometimes the accuracy of a measurement is known from previous studies or can be determined in a small substudy undertaken as part of an ongoing study, but it is impractical or too costly to use the accurate measurement on all study subjects. In such situations, error correction methods may be employed to correct for the effects of measurement error on the magnitude of the observed association. That is, the known values of specificity and sensitivity can be used to estimate the true proportion exposed from the observed proportion. If p denotes the observed proportion exposed, then the following formula may be used to estimate P, the true proportion exposed: p specificity 1 . P sensitivity specificity 1 Other methods of improving estimates even if sensitivity and specificity are not known are described elsewhere [1]. Also discussed [1] is the use of multiple imperfect measurements

of a given variable to improve accuracy rather than relying on a single imperfect measurement.

D. Quantification of the Reproducibility of Discrete Variables For quantifying the reproducibility of a discrete variable, the kappa statistic is used most frequently. Consider data in Table 8 [46]. Suppose it is known from medical records that 39 of 217 people were prescribed a certain medication. When administered a questionnaire, 14 of the 39 people who were prescribed a medication say that they were prescribed it, while 171 of 178 people who were not prescribed a medication say that they were not. The agreement between the two methods of ascertaining information (85%) is actually quite good. However, relatively few people in fact were prescribed the medication, so that even if all the study subjects said they had not had the medication prescribed, regardless of whether or not they had, agreement would still be high. Thus, a statistic is needed that takes into account the agreement that would be expected by chance. The kappa statistic, which takes into account chance agreement, is defined as observed agreement expected agreement . 1 expected agreement When two measurements agree only at the chance level, the value of kappa is zero. When the two measurements agree perfectly, the value of kappa is 1.0. In Table 8, the value of kappa is 0.39, indicating that the observed agreement is

550

KELSEY AND SOWERS

TABLE 8 Example of Calculation of kappa: Agreement between Personal Interview and Medical Chart for Use of Reserpine among Controls in a Case-Control Study in Two Retirement Communities History of use of reserpine according to medical chart Yes

No

Total

History of use of reserpine according to patient’s report Yes

14

7

21

No

25

171

196

Total

39

178

217

Note. Source: Paganini-Hill and Ross [46]. Chance-expected agreement  [(21)(39) (196)(178)]/(217)2  0.7583. Observed agreement  (14 171)/217  0.8525. kappa  (0.8525 0.7583)/(1 0.7583)  0.39.

only 39% of the way between chance agreement and perfect agreement. O’Neill et al. [47] assessed the reproducibility of answers obtained by the questionnaire used in the European Vertebral Osteoporosis Study of persons ages 50 – 85 years by having a different interviewer readminister the same questionnaire within a 28-day period at four of the study sites. The kappa coefficient was 1.00 at all four study sites for the variable of ever having been pregnant, but ranged between only 0.25 and 0.63 at the four sites for activity level when ages 15 – 25 and from 0.17 to 0.62 for milk consumed at ages 15 – 25. It should be noted that for measurements of conditions that are uncommon, the value of kappa will be lower than for common conditions, even though the values of specificity and sensitivity remain the same. This property needs to be taken into account when interpreting values of kappa.

E. Continuously Distributed Variables For continuously distributed variables such as bone mineral density, measurement error is again a concern. A number of indices can be calculated that reflect the accuracy of the measure of interest [1]. One frequently used measure to reflect the accuracy or lack of accuracy is the standardized bias, which is defined as mean of measurements true mean . standard deviation of the true values Thus, for example, if a given technique for measuring bone mineral density systematically overestimates the true value by 0.02 g/cm2 and the standard deviation for true bone mineral density in the population is 0.10 g/cm2, then the technique has a standardized bias of 0.02/0.10. That is, the

imperfect technique tends to give values that are 0.20 of a standard deviation higher than the true values. A measure of the extent to which imperfectly measured values tend to fall in the same position relative to their mean as do the corresponding true values relative to the true mean is the correlation coefficient, which can range from 1.0 to 1.0. The correlation coefficient of reproducibility (which also can range from 1.0 to 1.0) is often used to assess the extent to which two imperfect sets of measurements agree. This coefficient indexes the extent to which the measurement tends to fall in the same position relative to the mean for the first set as it does relative to the mean for the second set. The square of this correlation coefficient indicates that proportion of the variance in one set of measurements that is captured by the other set of measurements. A correlation coefficient of 0.60 between two types of measures of bone mass, for instance, would indicate that (0.60)2  0.36 of the variance in one type of measure was captured by the other type of measure. It is important to note that both the correlation coefficient of reproducibility and kappa may give misleading indications of the extent of reproducibility if the errors in measurement are not independent of each other. For instance, information recorded in medical records may have been obtained from the patient herself so that data subsequently elicited from the patient by an interviewer may not be independent of what is found in the medical record. The coefficient of variation, or the standard deviation divided by the mean, is sometimes used as an indicator of the precision of a measure and is best interpreted taking into account the numeric values of the mean and standard deviation. It is particularly useful for assessing the relative amount of variation in situations in which as the mean increases, so does the standard deviation or when investigators want to compare their precision to that reported by others. For instance, Schott et al. [48] reported a coefficient of variation of 0.84% for the broadband ultrasonic attenuation of the calcaneous and 0.15% for speed of sound by measuring a phantom daily for 45 days using a Lunar Achilles ultrasonic instrument. When 20 volunteers were measured three times each, these coefficients of variation were 0.93 and 0.15%, respectively. The authors reported that these results were comparable to other reports in the literature. As is the case for binary variables, nondifferential error in measurement generally results in attenuation of associations between continuously distributed variables, such as the association between dietary calcium intake and bone mineral density. If the accuracy or reproducibility of a continuously distributed variable is available from previous studies, then methods are available to correct for this attenuation. Approaches to correction in correlation analysis with one or two variables measured with known error are described by Liu et al. [49] and Rosner and Willett [50]. Liu et al. [49] also presented an approach for correction when regression analysis is used.

CHAPTER 20 Introduction to Epidemiologic Methods

Multiple measurements can be used to increase the accuracy of certain continuously distributed variables, such as levels of hormones or bone turnover markers. For example, in a longitudinal study, Sowers et al. [51] observed that the between-person variability for bone-specific alkaline phosphatase was two times greater than the within-person variation. In contrast, when measuring bone mineral density with dual energy X-ray absorptiometry (DXA), the between-person variability was 11 to 29 times greater than the within-person variability. To reduce the relatively high within-person variability for the bone-specific alkaline phosphatase, the investigators sampled individuals at five different times and used the mean of the five samples as their measure of bone-specific alkaline phosphatase.

IX. MEASURING DIET AND BONE TURNOVER STATUS AS EXAMPLES OF MEASUREMENT ISSUES This section addresses in greater detail issues related to the measurement of dietary status and of bone turnover status. The measurement of these characteristics illustrates the pitfalls and problems in epidemiologic studies in which the data sources are interview and biologic specimens, respectively.

A. Dietary Intake The study of calcium intake and other dietary characteristics has long been a focus of those interested in risk factors for osteoporosis. However, the results of studies concerned with dietary intake of calcium have been inconsistent. These discrepant results may have their origin in either (i) the nature of and relatively weak role of dietary intake in the development of osteoporosis or (ii) the manner in which data were collected. Four general categories of assessment methods have been used to obtain information about diet in epidemiologic studies [52]. In a dietary record the individual records the amounts of each food and beverage consumed over a specified number of days (usually 4 days). The respondent must be able to record food weights, volumetric measures, and the content of mixed dishes. Accordingly, the participants in a study using dietary records tend to be highly select individuals who have the literacy, endurance, and motivation to participate in a demanding assessment. This selectivity ultimately limits the generalizability of findings. In addition, during the period that food records are being used, respondents may alter their behavior to ease the burden of completing the assessment or to avoid the perceived positive or negative social value placed on certain eating patterns.

551 In 24-h recalls, the respondent is asked to describe all food and beverages consumed in the previous 24 h, typically with a standard set of question probes and food models for assistance in identifying serving sizes. Because most individuals’ diets vary greatly from day to day, it is not appropriate to use data from a single 24-h recall to characterize an individual’s usual intake. The principal use of 24-h recalls should be to describe the average dietary intake of a group. Measurements on at least 50 people should be made to categorize a group. The food frequency method asks the respondents to report their usual frequency of foods consumed from a list of foods (ranging in length from about 10 to more than 100) for a specific period of time, most often in the past year. Responses to food frequency questionnaires can be used to rank individuals within a population according to their consumption of foods specified on the frequency list. These responses can also be linked by computer to food composition data bases; the one generally used in the United States is that of the U.S. Department of Agriculture [53]. A food composition data base provides information on the nutrient content of food groupings. Because food groupings used in food frequency questionnaires are generally quite broad (e.g., fruit juices), an average nutrient content is estimated for each grouping. The averages are obtained by weighting the contributions of the different constituents based on consumption patterns in the population. Thus, in the United States, the nutrient contribution of orange juice might be, say, 50% of the nutrient estimate for fruit juices, whereas apple and grape juice would contribute smaller proportions. The exact proportions to be attributed to different foods should be determined by the investigator based on the food groupings used in his or her particular questionnaire and the consumption patterns in the population being studied. Whether it is appropriate to use the food frequency methodology to quantify nutrient intake is controversial. Although many investigators use quantitative estimates derived from food frequency questionnaires as approximations of the usual intake, food frequency questionnaires have been observed to perform inconsistently in different populations, and the same questionnaire performs differently when administered under different conditions to the same population. Table 9 shows considerable differences in estimated calcium intake when the same questionnaire was administered by an interviewer (row 1) or filled out by the respondent (rows 2 and 3). When a food frequency questionnaire is administered by an interviewer, typically more than 90% of the frequencies are considered usable; however, when the food frequency questionnaire is self-administered, between 20 and 40% of frequencies may not be usable. In a food frequency questionnaire, the lack of details about diet makes it likely that quantification of nutrient intake will not be as accurate as dietary records or 24-h recalls. For instance, portion size may not be estimated

552

KELSEY AND SOWERS

TABLE 9 Mean (and Standard Deviation) of Dietary Calcium Intake (mg/day) Estimated from the National Cancer Institute Food Frequency Questionnaire According to Whether the Questionnaire Was Interviewer-Administered or Self-Administered, Women Categorized by Reproductive Status Mode of administration Interviewer-administered

Menstruating women

Women with hysterectomy

Women with oophorectomy 1161 (702)

1012 (577)

995 (559)

Self-administered (baseline)

714 (457)

697 (434)

853 (602)

Self-administered (year 1)

754 (440)

749 (441)

800 (434)

Note. Source: Sowers et al., unpublished data.

adequately and certain unusual foods eaten in large quantities by only a small proportion of the population may be difficult to take into account. Diet histories are typically a combination of a food frequency questionnaire, an interview about usual patterns of eating, and a 3-day diet record. Each method uses a data base with the nutrient content of food and either a reported or an imputed measure of serving size to estimate nutrient intake. While the diet history is noted for the volume of data gathered, it is used infrequently because of the cost of administration and the burden to the respondent. Because the diet history includes a 3-day food record, it is likely that only select groups of people will participate in assessment. Also, people may modify their behavior during the assessment period. Because each of the parts of the diet history provides a unique piece of information, data reduction and reconciliation of the findings of the 3-day food record and the food frequency present major problems. Establishing the role of dietary calcium intake in osteoporosis and fractures also requires various other pieces of information, including (i) knowledge of the intake of other nutrients, such as vitamin D as well as sunlight exposure; (ii) knowledge of intake of other nutrients that might be deleterious in high concentrations, such as fluoride or aluminum; (iii) information about nutrient intake and its adequacy in critical periods during growth, development, reproduction, and the menopausal transition; (d) an assessment of factors such as dietary sodium that may influence the excretion of calcium in the urine; and (e) information about factors that may facilitate or impede the absorption of dietary calcium, including intestinal disorders such as sprue. Additional problems are that the nutrient data banks do not have consistent documentation about important nutrients, such as vitamin D, fluoride, or aluminum content of foods. For example, the U.S. Department of Agriculture has only provisional tables for vitamin D. Other nutrients, such as fluoride, may be in part determined by the fluoride content of the water in which the food is prepared. The responsibility then rests with the individual investigator to

ascertain how much total fluoride is ingested. There are no validated instruments to assess retrospectively dietary calcium intake during critical periods such as adolescence and young adulthood. Furthermore, those assessments must take into account the change in the food supply over time. For instance, were school lunch programs available to provide a daily source of calcium during the Depression years?

B. Bone Turnover Markers Bone turnover markers may help define which component of bone remodeling is responsible for reduction in bone density or alteration in bone microarchitecture. Additionally, bone turnover markers can help define the responsiveness of bone tissue to preventive or therapeutic interventions. While these markers hold greater potential for understanding bone biology at a cellular level, they have limitations in epidemiologic studies. For instance, bone turnover markers have a greater variability than the measurement of bone mineral density [51]. Also, these turnover markers reflect the short-term events of bone changes required for homeostasis. As such, their measurement may reflect immediate events, but their extrapolation to events over a broader time frame is questionable. Multiple measurements of bone turnover markers are required to characterize the individual’s current status in regard to osteocalcin, bone-specific alkaline phosphatase, or N-telopeptides. The use of urine-based turnover markers is particularly useful in epidemiologic studies because it is easier to collect multiple urine samples than multiple blood samples. However, these urinary markers reflect not only the bone turnover processes, but also the integrity of the individual’s renal function and any changes in fluid status. Thus, in epidemiologic studies, problems and pitfalls occur with biologic data as well as with interview data. As with interview data, multiple sampling of biological material can improve the precision of measurement of an individual’s status. There is also a need to collect specimens in

553

CHAPTER 20 Introduction to Epidemiologic Methods

ways that are acceptable to a broad range of individuals so that studies are not based on a highly select group whose findings cannot be extrapolated to a larger population.

X. CONCLUSIONS Epidemiologic studies have added considerably to our understanding of the etiology, course, and consequences of osteoporosis. As will be described in subsequent chapters in this section, a variety of risk factors have been established and intriguing leads for further study have been suggested. However, for some potential risk factors, results of different studies have been rather inconsistent, such as those pertaining to some aspects of diet, physical activity, alcohol, certain medications, and reproductive variables. When trying to interpret these discrepant results, it is important to keep in mind common reasons for inconsistent results. Sometimes discrepant results may be explained by major flaws in study designs, such as failure to realize that the time sequence is unclear in cross-sectional studies, poor choice of controls in case-control studies, large numbers of people lost to follow-up in cohort studies, and, in any type of study, sample sizes that are too small and the use of specialized populations that are not representative of any larger population. However, for the most part, the variables for which results have tended to be inconsistent are those that are difficult to measure. It is important to keep in mind that if a characteristic cannot be measured well or for some reason is not measured well in an otherwise methodologically sound study, associations between that risk factor and a disease will be difficult to detect. If measurements of varying quality are used in different studies, then discrepant results may be expected. Different results may also be obtained when the prevalence of a potential risk factor varies from one study to another if there is some error in the measurement of that risk factor. Another reason for discrepant results is the extent to which confounding variables have been accurately measured and taken into account. Because many established and potential risk factors for osteoporosis, such as body build, diet, physical activity, alcohol consumption, caffeine consumption, and cigarette smoking, are correlated with each other, separating out the effect of any one of these variables is difficult. Errors in measuring these characteristics exacerbate this problem. Another reason for inconsistent results is that a risk factor may have a different effect in one subgroup of the population than in another. If one study includes mostly members of one subgroup and another study has mostly people from another subgroup, then results could be different in the two studies. Sometimes results from different studies are said to be discrepant when they really are not. One common reason

for this is varying sample size. If sample sizes differ from one study to another, then statistical significance is also likely to vary. If a result is statistically significant in one study and not in another, but the magnitudes of the association between the risk factor and the disease are similar in the studies, then this is evidence for consistency between the studies, not inconsistency. For instance, Cumming [35] found that although an apparently protective effect of supplemental calcium on loss of bone mass in adult women was statistically significant in some studies and not others, the magnitude of the slight protective effect in almost all studies was actually quite consistent. It is thus very important to consider magnitudes of associations and their confidence limits, not just statistical significance. Finally, if an effect is relatively small and there is even a modest amount of measurement error, it will be difficult for epidemiologic studies to detect the effect. The association between caffeine intake and osteoporosis may fall into this category. As knowledge of good epidemiologic methods and awareness of potential pitfalls in epidemiologic studies have become more widespread, the quality of studies has improved considerably. Also, results have tended to be interpreted more cautiously when potential problems in studies have been recognized. One key to continued improvement will be better methods of measurement of exposures, confounding variables, and, to some extent, outcome variables, whether by questionnaire, laboratory assay, densitometry, or other approach. Such improvements should help advance knowledge of the epidemiology of osteoporosis and associated fractures and should also enable epidemiologists to continue to provide further ideas for investigators in other disciplines, including endocrinology, biomechanics, and other areas.

References 1. J. L. Kelsey, A. S. Whittemore, A. S. Evans, and W. D. Thompson, “Observational Epidemiology,” 2nd Ed. Oxford Univ. Press, New York, 1996. 2. J. L. Kelsey and S. Parker, Epidemiology as an alternative to animal research. In “Non-Animal Techniques in Biomedical and Behavioral Research and Testing” (M. B. Kapis and S. C. Gad, eds.). Lewis Publishers, Boca Raton, FL, 1993. 3. C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992). 4. T. D. Spector, E. V. McCloskey, D. V. Doyle, and J. A. Kanis, Prevalence of vertebral fracture in women and the relationship with bone density and symptoms: The Chingford Study. J. Bone Miner. Res. 8, 817 – 822 (1993). 5. B. Ettinger, D. M. Black, M. E. Nevitt, A. C. Rundle, J. C. Cauley, S. R. Cummings, H. K. Genant, and the Study of Osteoporotic Fractures Research Group, Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 455 (1992). 6. F. Albright, E. Bloomberg, and P. H. Smith, Postmenopausal osteoporosis. Trans. Assoc. Am. Phys. 55, 298 – 305 (1940).

554 7. R. Lindsay, D. M. Hart, A. MacLean, A. C. Clark, A. Kraszewski, and J. Garwood, Bone response to termination of estrogen treatment. Lancet 1, 1325 – 1327 (1978). 8. C. Christiansen, M. S. Christiansen, and I. Transbol, Bone mass in postmenopausal women after withdrawal of estrogen/gestagen therapy. Lancet 1, 459 – 461 (1981). 9. S. Maggi, J. L. Kelsey, J. Litvak, and S. P. Heyse, Incidence of hip fractures in the elderly: A cross-national analysis. Osteopor. Int. 1, 232 – 241 (1991). 10. S. R. Cummings, J. A. Cauley, L. Palermo, P. D. Ross, R. P. Wasnich, D. Black, and K. G. Faulkner for the Study of Osteoporotic Fractures Research Group, Racial differences in hip axis lengths might explain racial differences in rates of hip fracture. Osteopor. Int. 4, 226 – 229 (1994). 11. J. W. Nieves, J. A. Grisso, and J. L. Kelsey, A case-control study of hip fracture: Evaluation of selected dietary variables and teenage physical activity. Osteopor. Int. 2, 122 – 127 (1992). 12. O. S. Miettinen, The “case-control” study: Valid selection of subjects. J. Chron. Dis. 38, 543 – 548 (1985). 13. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 14. J. A. Grisso, J. L. Kelsey, L. A. O’Brien, C. G. Miles, S. Sidney, G. Maislin, K. LaPann, D. Moritz, B. Peters, and the Hip Fracture Study Group, Risk factors for hip fracture in men. Am. J. Epidemiology. 145, 786 – 793 (1997). 15. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, S. Hoffman, and the Northeast Hip Fracture Study Group, Risk factors for hip fracture in black women. N. Engl. J. Med. 330, 1555 – 1559 (1994). 16. N. S. Weiss, C. L. Ure, J. H. Ballard, A. R. Williams, and J. R. Daling, Decreased risk of fractures of the hip and lower forearm with postmenopausal use of estrogen. N. Engl. J. Med. 303, 1195 – 1198 (1980). 17. N. Kreiger, A. Gross, and G. Hunter, Dietary factors and fracture in postmenopausal women: A case-control study. Int. J. Epidemiol. 21, 953 – 958 (1992). 18. J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, F. Kaplan, and the Northeast Hip Fracture Study Group, Risk factors for falls as a cause of hip fracture in women. N. Engl. J. Med. 324, 1326 – 1331 (1991). 19. D. L. Sackett, Bias in analytic research. J. Chron. Dis. 32, 51– 68 (1979). 20. H. Austin, H. A. Hill, W. D. Flanders, and R. S. Greenberg, Limitations in the application of case-control methodology. Epidemiol. Rev. 16, 65 – 76 (1994). 21. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). 22. M. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, Body size, estrogen use and thiazide diuretic use affect 5-year radial bone loss in postmenopausal women. Osteopor. Int. 3, 314 – 321 (1993). 23. P. J. Diggle, K. Liang, and S. L. Zeger (eds.), “Analysis of Longitudinal Data.” Oxford Univ. Press, New York, 1998. 24. D. T. Felson, Y. Zhang, M. T. Hannon, D. P. Kiel, P. W. Wilson, and J. J. Anderson, The effect of postmenopausal estrogen therapy on bone density in elderly women. N. Engl. J. Med. 329, 1141 – 1146 (1993). 25. E. Barrett-Connor, J. C. Chang, and S. L. Edelstein, Coffee-associated osteoporosis offset by daily milk consumption, The Rancho Bernardo Study. JAMA 271, 280 – 283 (1994). 26. D. B. Petitti and S. Sidney, Hip fracture in women. Clin. Orthop. 246, 150 – 155 (1989). 27. M. Hernandez-Avila, G. A. Colditz, M. J. Stampfer, B. Rosner, F. E. Speizer, and W. C. Willett, Caffeine, moderate alcohol intake, and risk of fractures of the hip and forearm in middle-aged women. Am. J. Clin. Nutr. 54, 157 – 163 (1991).

KELSEY AND SOWERS 28. D. C. Bauer, W. S. Browner, J. A. Cauley, E. S. Orwell, J. C. Scott, D. M. Black, J. L. Tao, and S. R. Cummings for the Study of Osteoporotic Fractures Group, Factors associated with appendicular bone mass in older women. Ann. Intern. Med. 118, 657 – 665 (1993). 29. L. Forsen, A. Bjorndal, K. Bjartveit, T. H. Edna, J. Holman, V. Jessen, and G. Westberg, Interaction between current smoking, leanness, and physical inactivity in the prediction of hip fracture. J. Bone Miner. Res. 9, 1671 – 1678 (1994). 30. A. C. Looker, T. B. Harris, J. H. Madans, and C. T. Sempos, Dietary calcium and hip fracture risk: The NHANES I Epidemiologic Follow-Up Study. Osteopor. Int. 3, 177 – 184 (1993). 31. M. Sinaki, S. Khosla, P. J. Limburg, J. W. Rogers, and P. A. Murtaugh, Muscle strength in osteoporotic versus normal women. Osteopor. Int. 3, 8 – 12 (1993). 32. Institute of Medicine Committee to Review the NIH Women’s Health Initiative, “An Assessment of the NIH Women’s Health Initiative” (S. Thaul and D. Hotra, eds.). National Academy Press, Washington, DC, 1993. 32a. S. Hulley, D. Grady, T. Bush, C. Furberg, D. Herrington, B. Riggs, and E. Vittinghoff, Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women: Heart and Estrogen/Progestin Study (HERS) Research Group. JAMA 280, 605 – 613 (1998). 33. M. Sowers, B. Shapiro, M. A. Gilbraith, and M. Jannausch, Health and hormonal characteristics of premenopausal women with lower bone mass. Calcif. Tissue Int. 47, 130 – 135 (1990). 34. R. P. Heaney, Calcium, bone health, and osteoporosis. Bone Miner. Res. 4, 255 – 301 (1986). 35. R. G. Cumming, Calcium intake and bone mass: A quantitative review of the evidence. Calcif. Tissue Int. 47, 194 – 201 (1990). 36. A. Z. LaCroix, J. Wienpahl, L. R. White, R. B. Wallace, P. A. Scherr, L. K. George, J. Coroni-Huntley, and A. M. Ostfeld, Thiazide diuretic agents and the incidence of hip fracture. N. Engl. J. Med. 322, 286 – 290 (1990). 37. J. J. Schlesselman, “Case-Control Studies.” Oxford Univ. Press, New York, 1982. 38. S. Selvin, “Statistical Analysis of Epidemiologic Data.” Oxford Univ. Press, New York, 1991. 39. A. E. Schindler, A. Ebert, and E. Friedrich, Conversion of androstenedione to estrogen by human fat tissue. J. Endocrinol. Metab. 35, 627 – 630 (1972). 40. S. R. Cummings and M. C. Nevitt, A hypothesis: The causes of hip fractures. J. Gerontol. 44, M107 – M111 (1989). 41. A. M. Lilienfeld, Epidemiologic methods and inferences instudies of noninfectious diseases. Public Health Rep. 72, 51 – 60(1957). 42. R. Lindsay, Hormone replacement for prevention and treatment of osteoporosis. Am. J. Med. 95, 37S – 39S (1993). 43. J. F. Fleiss, “The Design and Analysis of Clinical Experiments.” Wiley, New York, 1986. 44. S. Greenland, The effect of misclassification in the presence of covariates. Am. J. Epidemiol. 112, 564 – 569 (1980). 45. C. R. Weinberg, D. M. Umbach, and S. Greenland, When will nondifferential misclassification of an exposure preserve the direction of a trend? Am. J. Epidemiol. 140, 565 – 571 (1994). 46. A. Paganini-Hill and R. K. Ross, Reliability of recall of drug usage and other health-related information. Am. J. Epidemiol. 116, 114 – 122 (1982). 47. T. W. O’Neill, C. Cooper, J. B. Cannata, J. B. Diaz Lopez, K. Hoszowski, O. Johnell, R. S. Lorene, B. Nilsson, H. Raspe, O. Stewart, and A. J. Silman on Behalf of the European Vertebral Osteoporosis (EVOS) Group, Reproducibility of a questionnaire on risk factors for osteoporosis in a multicentre prevalence survey: The European Vertebral Osteoporosis Study. Int. J. Epidemiol. 23, 559 – 565 (1994). 48. A. M. Schott, D. Hans, E. Somay-Rendu, P. D. Delmas, and P. J. Meunier, Ultrasound measurements on os calcis: Precision and

CHAPTER 20 Introduction to Epidemiologic Methods age-related changes in a normal female population. Osteopor. Int. 3, 249 – 254 (1993). 49. K. Liu, J. Stamler, A. Dyer, J. McKeever, and P. McKeever, Statistical methods to assess and minimize the role of intraindividual variability in obscuring the relationship between dietary lipids and serum cholesterol. J. Chron. Dis. 31, 399 – 418 (1978). 50. B. A. Rosner and W. C. Willett, Interval estimates for correlation coefficients corrected for within-person variation: Implications for

555 study design and hypothesis testing. Am. J. Epidemiol. 127, 337 – 386 (1988). 51. M. Sowers, Clinical epidemiology and ostoporosis: Measures and their interpretation. Clin. North Am. Endocrinol. 26, 219 – 231 (1997). 52. F. E. Thompson and T. Byers, Dietary Assessment Resource Manual. J. Nutr. 124, 11S, 2245S – 2317S (1994). 53. U. S. Department of Agriculture, “The Nutrient Composition of Foods,” Handbook 8, Vols. 1 – 23, Hyattsville, 1989.

CHAPTER 21

Magnitude and Impact of Osteoporosis and Fractures L. JOSEPH MELTON III CYRUS COOPER

I. II. III. IV.

Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905 MRC Environmental Epidemiology Unit, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, England

V. Future Projections VI. Conclusions References

Introduction Magnitude of the Problem Fracture Epidemiology Impact of Osteoporotic Fractures

I. INTRODUCTION

osteoporosis and osteoporosis-related fractures and to determine the impact that the condition has on society.

When the term “osteoporosis” entered medical parlance in France and Germany during the past century [1], it implied a histological diagnosis (“porous bone”) that was subsequently refined to mean that bone tissue, while normally mineralized, was present in reduced quantity. This definitional approach culminates today in attempts to define osteoporosis on the basis of low bone mass [2], which can be assessed in vivo by a variety of noninvasive densitometric techniques (see Chapter 59). Low bone mineral density (BMD), in combination with impaired bone “quality” [3], leads to skeletal fragility and an increased risk of fracture, the clinical manifestation of osteoporosis [4]. Indeed, the realization that these fractures might result from an age-related reduction in bone strength antedated the histological observations [5]. This line of thought suggests that any assessment of osteoporosis must also include the associated fractures. The purposes of this review are to summarize epidemiologic data concerning the frequency of

OSTEOPOROSIS, SECOND EDITION VOLUME 1

II. MAGNITUDE OF THE PROBLEM How many people suffer from osteoporosis? It is clear from the preceding section that the answer to this question will depend on whether osteoporosis is defined on the basis of low bone mass or whether the emphasis, instead, is on osteoporosis-related fractures. Implicit in the definition of osteoporosis by bone mass alone is the relationship between this parameter and fracture risk. Low bone density is therefore analogous to high blood pressure, and the risk of fracture increases when bone density declines, just as the risk of stroke increases when blood pressure rises. A working group of the World Health Organization operationalized this concept by considering osteoporosis to be present when BMD levels in

557

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

558

MELTON AND COOPER

white women are more than 2.5 SD below the young normal mean [6]. To provide some comparability with earlier definitions that incorporated fracture, the subset of women with presumptive osteoporosis who also have a history of one or more fragility fractures are deemed to have severe (“established”) osteoporosis. Low bone mass (“osteopenia”) is defined by bone density levels more than 1 SD below the young normal mean, but less than 2.5 SD below. Some of the best data on the prevalence of osteoporosis come from the Third National Health and Nutrition Examination Survey (NHANES III), a large probability sample of the United States population. As assessed at the femoral neck, 20% of postmenopausal white women have osteoporosis [7]. No criterion for osteoporosis has been established for nonwhite women, but 10% of Hispanic women and 5% of African-American women have femoral neck BMD values more than 2.5 SD below the young normal mean for white women. Using the same cut-off value, osteoporosis prevalence rates for white, Hispanic, and African-American men age 50 years and over are 4, 2, and 3% respectively [7]. However, these race and genderspecific differences partly relate to overestimation of areal BMD (g/cm2) in individuals with larger skeletons [8]. When bone size is taken into account with bone mineral apparent density (BMAD, g/cm3), differences in bone density between men and women [9,10] and between women of different races [11 – 13] are reduced. In addition, osteoporosis is a systemic disease, and a greater proportion of the population is seen to be affected when more skeletal sites are assessed (Table 1). Most white women under age 50 have normal bone density, but with advancing age, the proportion with osteoporosis increases dramatically. Among women age 80 years and over, for example, only 3% have normal bone density at the hip, spine,

and forearm, while 27% have osteopenia at one skeletal site or another and 70% have osteoporosis. As judged from these data, an estimated 16.8 million (54%) postmenopausal white women in the United States have osteopenia and another 9.4 million (30%) have osteoporosis [14]. However, results vary from one geographic region to another [15]. The alternative approach is to assess fracture frequency. There are no data regarding the prevalence of distal forearm fracture, while the prevalence of hip fracture has been estimated at 51 per 1000 women age 65 and over [16] or 6 per 1000 men and women of all ages [17]. Most available information relates to vertebral fractures. In the European Vertebral Osteoporosis Study, for example, 15,570 men and women aged 50 – 79 years were selected from population registers in 36 European centers and evaluated according to a standardized protocol [18]. The overall prevalence of morphometrically defined vertebral deformity was 12% in both men and women, but the increase in prevalence with age was steeper among the women (Fig.1). Thus, the frequency of deformities in men aged 50 – 54 years was around 10%, rising to 18% at age 75 – 79 years, while the prevalence rose from 5% among women aged 50 – 54 years to over 24% at age 75 – 79 years. Using a different approach to morphometry, other studies have estimated the overall prevalence of vertebral deformities among postmenopausal women at 20 to 24% [19 – 22]. A frequency measure of interest to patients and clinicians is the probability of fracture over an average lifetime. Using fracture incidence rates from the United States, the estimated lifetime risk of a hip fracture is 17% in white women and 6% in white men [23]. This compares with risks of 16 and 5% for clinically diagnosed vertebral fractures and 16 and 2% for distal forearm fractures in white women and men, respectively. The lifetime risk of any of

TABLE 1 Proportion (%) of Rochester, Minnesota, Women with Bone Density Measurements More Than 2.5 SD below the Mean for Young Normal Womena Lumbar spine (%)

Either hip site (%)

50 – 59

7.6

3.9

3.7

14.8

60 – 69

11.8

8.0

11.8

21.6

70 – 79

25.0

24.5

23.1

38.5

 80

32.0

47.5

50.0

70.0

Totalb

16.5

16.2

17.4

30.3

Age group

Midradius (%)

Spine, hip, or midradius(%)

a Mean is from 48 subjects under age 40 who were randomly sampled from the Rochester, Minnesota, population. None of them was known to have any disorder that might influence bone metabolism. From L. J. Melton III, How many women have osteoporosis now? J. Bone Miner. Res. 10, 175 – 177 (1995). b Age-adjusted to the population structure of 1990 United States white women 50 years of age and older.

CHAPTER

559

21 Magnitude and Impact of Osteoporosis and Fractures

FIGURE 1

Prevalence of vertebral deformities among European men and women by age. Data derived from the European Vertebral Osteoporosis Study. Reproduced from O’Neill et al. [18], with permission.

the three fractures is 40% for women and 13% for men from age 50 years onward [23]. In Great Britain, the lifetime risk of a hip fracture among 50-year-old women is 14%, compared to 3% for British men of comparable age [24]. This contrasts with lifetime risks of 11 and 2% for clinical diagnosed vertebral deformities and 13 and 2% for wrist fractures in white women and men, respectively. Because lifetime risk depends on life expectancy as well as fracture incidence, the lifetime risk of hip fracture in British women could rise to 24% by 2050 if life expectancy continues to increase [24].

III. FRACTURE EPIDEMIOLOGY Fracture incidence in the community is bimodal, with peaks in youth and advanced age [25,26]. In young people, fractures of the long bones predominate, often following substantial trauma, and the incidence is greater in young men than in young women. Above the age of 35 years, overall fracture incidence in women climbs steeply so that female rates become twice those in men [27]. At least 1.3 million fractures in the United States each year have been attributed to osteoporosis, presuming that 70% of all fractures in persons aged 45 years or over are due to the condition [28]. The three sites most closely associated with osteoporosis are fractures of the hip, spine, and distal forearm. However, the epidemiologic characteristics of these three fractures differ, suggesting the influence of different etiologic factors.

A. Hip Fracture In most populations, hip fracture incidence rates increase exponentially with age (Fig. 2). In Rochester, Minnesota, rates reach about 3.0% per year among women age 85 years and over and 1.9% among men in this age group

FIGURE 2

Age-specific incidence rates for hip, vertebral, and distal forearm fractures in men and women. Data derived from the population of Rochester, Minnesota. From C. Cooper and L. J. Melton III, Epidemiology of osteoporosis. Trends Endocrinol. Metab. 3, 224 – 229 (1992).

[29]. At all ages beyond 50 years, the incidence in women is about twice that in men. Because there are more elderly women than men, however, about 80% of all hip fractures occur in women. Worldwide, there were an estimated 1.7 million hip fractures in 1990, about 1,197,000 in women, and another 463,000 or so in men [30]. A minority of such fractures are due to overwhelming trauma or to specific lesions in the proximal femur, although severe trauma accounts for a greater proportion of the total in countries where hip fractures are uncommon [29]. The vast majority of hip fractures follow a fall from standing height or less in individuals with reduced bone strength [31]. Over a lifetime, bone density of the femoral neck declines an estimated 58 and 39% in women and men, respectively, while bone density of the intertrochanteric region of the proximal femur falls about 53 and 35% [32]. Each 1 SD decline in BMD is associated with a 2.0- to 3.6-fold increase in the age-adjusted risk of hip fracture, depending on the exact site in the proximal femur that is measured [33]. Simultaneously, there is a dramatic increase in the likelihood of falling each year, from about one in five women age 40 – 49 years to nearly half of women 85 years old and over, along with a third of men in the oldest age group [34]. The pathophysiology of falling is complex (see Chapter 32), but only about 1% of falls lead to a hip fracture [35]. This is because the amount of force delivered to the proximal femur depends on various protective responses and on the orientation of the faller (see Chapter 19). The seemingly inexorable bone loss and increased risk of falling that accompany the aging process suggest that hip fractures are inevitable. In fact, incidence rates vary substantially from one population to another. Thus, hip fractures are much less frequent among nonwhites than whites (see Chapter 22), but there is substantial variation even within populations of a given race and gender (Fig. 3). Thus, incidence rates are higher among white residents

560

MELTON AND COOPER

FIGURE 3 Hip fracture incidence around the world as a ratio of the rates observed in various populations to those expected for U.S. white women. From L. J. Melton III, Differing patterns of osteoporosis across the world, In “New Dimensions in Osteoporosis in the 1990s” (C. H. Chesnut, ed.), pp. 13 – 18. Excerpta Medica Asian, Hong Kong, 1991. of Scandinavia than comparable people in North America or Oceania [36]; even within Europe, hip fracture rates vary more than sevenfold from one country to another [37]. The higher incidence in urban than rural districts has been explained on the basis of lower bone mass among urban residents [38], but data from the United States show that the pattern is much more complex (Fig. 4). In over 2000 counties nationwide, the incidence of hip fractures in elderly white women was negatively associated with latitude (higher in the South), water hardness, and mean hours of January sunlight and positively associated with poverty levels, proportion of the land in farms, and proportion of the population with fluoridated water supplies [39]. Regional differences did not seem to be accounted for by variation in activity levels, obesity, cigarette smoking, alcohol consumption, or Scandinavian heritage. Additional studies are needed to identify the environmental factors associated with such regional differences.

B. Vertebral Fracture Vertebral fractures have been synonymous with osteoporosis since its earliest description as a metabolic bone disorder [40]. However, epidemiologic data remain scant because there is no universally accepted definition of vertebral fracture and because a substantial proportion are

asymptomatic (see Chapter 34). Although it has long been clear that some vertebral fractures do not reach clinical attention, the size of this fraction was unknown. The age-adjusted incidence of clinically diagnosed vertebral fractures has been estimated at 5.3 per 1000 person-years among white women aged 50 years and over [41]. This represents about 30% of the total incidence of vertebral fractures among American postmenopausal white women of 18 per 1000 person-years [21]. The incidence of clinically diagnosed vertebral fractures was 4.3 times greater than rates derived from United States hospital discharge data for spine fractures [42], suggesting that a third of all vertebral deformities come to medical attention with about 8% necessitating admission to hospital. Data for England and Wales imply that as few as 2% might be hospitalized [43], but this is almost certainly an underestimate due to incomplete diagnostic coding. Figure 2 illustrates the incidence of clinically diagnosed vertebral fractures. In men, incidence rates climb exponentially with age, adopting a pattern similar to that observed for hip fractures in the same population [29]. In women, there is a more linear increase such that vertebral fracture rates are higher than those for hip fracture before the age of 70 years, but not thereafter. Figure 2 also illustrates that vertebral fractures are a greater problem in men than previously recognized, with an overall age-adjusted incidence in women only 1.9 times greater than that in men [41], although sex ratios vary by geographic region [18]. Other

CHAPTER

561

21 Magnitude and Impact of Osteoporosis and Fractures

Age-adjusted incidence of hip fractures among white women  65 years of age by county of residence in the United States,(1984 – 1987). From S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Regional variation in the incidence of hip fracture. JAMA 264, 500 – 502 (1990).

FIGURE 4

than the expected relationship with secondary osteoporosis (see Chapter 51), strong underlying risk factors have been difficult to define [44 – 50]. Falls accounted for only a third of new vertebral fractures among Rochester men and women; the majority were due instead to the compressive loading associated with lifting, changing positions, and so on or were discovered only incidentally [41]. The most frequent vertebral levels involved were T8 and T12/L1. These correspond with biomechanically compromised regions of the spine: the midthoracic region, where the dorsal kyphosis is most pronounced, and the thoracolumbar junction, where the relatively rigid thoracic spine meets the freely moving lumbar segment [31].

C. Distal Forearm Fracture Distal forearm fractures almost always follow a fall on the outstretched arm [51]. They display a different pattern of incidence compared to hip or vertebral fractures (Fig. 2). In white women, incidence rates increase linearly from age 40 to age 65 years and then stabilize [29] for reasons that remain obscure but may relate to a change in the pattern of falling with advancing age so that elderly women with slower gait and impaired neuromuscular coordination are more likely to fall on their hip than on their wrist [51,52]. In addition, compared to hip fractures, a greater proportion

of distal forearm fractures occur outdoors, and a winter peak in incidence has been associated with periods of icy weather [53,54]. In men, the incidence of distal forearm fractures remains relatively constant and low between ages 20 and 80 years. Consequently, the majority of such fractures occur in women, and the age-adjusted female to male ratio of 4:1 is more marked than for hip or vertebral fractures. Nonetheless, the incidence of distal forearm fracture varies from one geographic area to another generally in parallel with hip fracture incidence rates [29].

D. Other Fractures Incidence rates for fractures of the proximal humerus, pelvis, proximal tibia and distal femur also increase with aging in elderly women but to a lesser extent in men. Along with most other fractures among elderly women, these have been directly associated with low bone density [55]. About 80% of proximal humerus fractures occur in individuals 35 years old and over, and three-fourths are in women [56]. The same general picture is seen in other populations [29]. Three-fourths or more of all proximal humerus fractures are due to moderate trauma, typically a fall from standing height or less [56,57]. Like the falls related to hip fracture, these seem to be more frequent in frail women with poor neuromuscular function [58].

562

MELTON AND COOPER

Dramatic age-related increases are also seen for pelvic fractures [29]. Overall, about two-thirds of all pelvic fractures occur among persons age 35 years or older, and nearly 70% are in women [59]. While multiple pelvic fractures and acetabular fractures are associated with severe trauma, moderate trauma accounts for nearly two-thirds of fractures of isolated pelvic bones and single breaks in the pelvic ring, which constitute about 80% of all pelvic fractures in the community [59]. Proximal tibia fractures have been classified as “composite” fractures, with peaks in incidence among the young and the old [60]. Incidence rates are highest in adolescent boys, but age-adjusted incidence rates are 30% greater among women [29]. Proximal tibia fractures in women outnumber those in men by about 2:1 after age 50 years [27,61]. Over 40% of proximal tibia fractures are due to severe trauma, while only a fourth are due to falls. Nonetheless, low bone mass increases the risk of leg fractures in elderly women [55], and the majority of proximal tibia fractures among older individuals are related to osteoporosis [62]. Distal femur fractures also have some of the characteristics of an age-related fracture [29]. The incidence of fractures of the femur shaft in Stockholm is greater in women than in men after age 65 years; the majority of the fractures in this age group are due to falls, and the incidence of femur shaft fractures due to moderate trauma increases exponentially with age in both sexes [63]. However, most fractures of the femur shaft are due to severe trauma. Moderate trauma, however, accounts for half of all distal femur fractures in Rochester, and this subgroup of cases exhibits age-related increases in incidence among both men and women like those seen for hip fractures [64].

IV. IMPACT OF OSTEOPOROTIC FRACTURES The adverse outcomes of osteoporotic fractures fall into three broad categories: mortality, morbidity, and cost.

A. Mortality The influence that hip, spine, and distal forearm fractures have on survival appears to differ with the type of fracture. Hip fractures are the most serious, leading to an overall reduction in survival of 10 – 20% [29]. Excess deaths occur mainly within the first 6 months and diminish with time, although the death rate may remain elevated for a number of years [65]. Mortality differs, however, by age and sex. In one population-based study, a relative survival of 92% was found for white hip fracture victims under 75 years of age, compared to only 83% for those aged 75 years

and over [29]. Despite their greater average age at the time of fracture, survival is better among women. This sex difference appears to arise from the greater frequency of other chronic diseases among men who sustain hip fractures [66]. Thus, while some deaths are attributable to acute complications of the fracture or of its surgical management [67], the majority appear to be due to serious coexisting illnesses [68]. There does not appear to be any excess mortality among patients who sustain distal forearm fractures [69] and, until recently, it was assumed that osteoporotic vertebral fractures were not attended by significant mortality either. However, 5-year survival among patients with clinically diagnosed vertebral fractures is only 61% compared to 76% expected for those of like age and sex (relative survival, 0.82) [69]. The proportionate excess of deaths is thus comparable to that following hip fracture (Table 2). Except for an excess of pulmonary deaths in women with severe vertebral deformities and kyphosis [70], strong associations with specific causes of death have not been identified [69,70], suggesting that impaired survival may be due to an indirect association with comorbid conditions that lead to an increased risk of osteoporosis. This explanation would accord with the observation that low bone density is itself associated with excess mortality from various causes [71 – 73].

B. Morbidity After allowing for the functional impairment expected in aged people, fractures of the hip, spine, and distal forearm result in an estimated 7% of women becoming dependent in the basic activities of daily living and cause nursing home care in a further 8% [74]. As reviewed elsewhere (see Chapter 34), hip fractures contribute most to this burden. They almost invariably necessitate hospitalization, and in 1985 the average length of hospital stay in England and Wales was 30 days so that 3500 National Health Service hospital beds were occupied daily. While these patients are at high risk of acute complications such as pressure sores, pneumonia, and urinary tract infections, the most important long-term outcome is impairment in the ability to walk. Around 20% of patients are nonambulatory even before fracture, but of those able to walk, half cannot walk independently afterward [75]. Ultimately, up to a third of hip fracture victims may become totally dependent [76], and the risk of institutionalization is great [77]. Nearly 140,000 nursing home admissions are attributable to hip fractures each year in the United States [78], and as many as 8% of all nursing home residents have had a hip fracture [79]. The health impact of vertebral fractures has proved considerably more difficult to quantify. As noted earlier, only a minority of incident vertebral deformities come to clinical attention. Nonetheless, vertebral fractures in patients aged

CHAPTER

563

21 Magnitude and Impact of Osteoporosis and Fractures

TABLE 2 Relative Survival Following Vertebral, Hip and Distal Forearm Fractures among Residents of Rochester, Minnesota, According to Duration of Follow-Up from Diagnosis Relative survival (95% CI) Time from diagnosis (year)

Vertebral

Hip

Forearm

1

0.96 (0.92 – 0.99)

0.88 (0.85 – 0.91)

1.00 (0.98 – 1.02)

2

0.93 (0.87 – 0.99)

0.87 (0.83 – 0.90)

1.00 (0.97 – 1.03)

3

0.92 (0.86 – 0.98)

0.86 (0.82 – 0.90)

1.01 (0.98 – 1.04)

4

0.84 (0.75 – 0.92)

0.83 (0.78 – 0.88)

0.99 (0.95 – 1.04)

5

0.82 (0.71 – 0.93)

0.83 (0.77 – 0.89)

1.00 (0.95 – 1.05)

45 years and older account for about 52,000 hospital admissions in the United States [79] and 2188 in England and Wales each year. The major consequences of vertebral fracture are back pain, kyphosis, and height loss. New compression fractures may give rise to severe back pain, which typically resolves over weeks or months [80]. A more protracted clinical course affects a proportion of patients who experience chronic pain while standing and during physical stress, particularly bending [81]. Not only physical function but self-esteem, body image, and mood also appear to be adversely affected in patients with vertebral fractures (see Chapter 61). Despite the fact that only about one-fifth of all patients with distal forearm fractures are hospitalized [82], they account for some 50,000 hospital admissions and over 400,000 physician visits in the United States each year [79]. Admission rates vary markedly with age, such that only 16% of forearm fractures occurring in women ages 45 – 54 years required inpatient care compared to 76% of those in women age 85 years and over [83]. There is a 30% risk of algodystrophy after these fractures [84], as well as an increased likelihood of neuropathies and posttraumatic arthritis [85]. Nearly half of all patients report only fair or poor functional outcomes at 6 months following a distal forearm fracture [86].

C. Economic Costs Fractures in the United States may cost as much as $20 billion per year, with hip fractures accounting for over a third of the total [87]. Because most hip fractures are in elderly individuals, wages foregone or years of life lost are not the primary determinants of cost. Eiskjaer et al. [88] found that the 9.2 years of potential life per 1000 women lost due to hip fracture was much lower than comparable figures for heart disease, stroke, and breast cancer (73, 29, and 20 per 1000 women, respectively). Instead, the greatest

expense is for inpatient medical services and nursing home care (Table 3). The direct costs in Table 3 include an estimated 547,000 hospitalizations and 4.6 million hospital bed days for the care of osteoporotic fractures in the United States in 1995 [78]. In Switzerland, osteoporotic fractures account for more hospital bed days than myocardial infarction and stroke at a cost of CHF 600 million [89]. In England, hip fractures alone consume one-fifth of all orthopedic beds, at a direct cost of £850 million per year in 1999 [90]. In France, an estimated 56,000 hip fractures annually cost about FF 3.5 billion [91]. Such figures are a source of concern to governmental leaders in almost every country.

V. FUTURE PROJECTIONS The cost of osteoporosis-related fractures can only rise in the future. Life expectancy is increasing around the globe and the number of elderly individuals is rising in every region. The estimated 323 million individuals in the world age 65 years or over at present is expected to rise to 1555 million by the year 2050 [30]. These demographic changes alone can be expected to cause the number of hip fractures throughout the world to increase from about 1.7 million in 1990 to 6.3 million in 2050 (Fig. 5). Around half of all hip fractures occurred in Europe and North America in 1990, but rapid aging of the Asian and Latin American populations will reduce the European and North American contribution to only 25% by 2050, with over half of all hip fractures occurring in Asia [30]. It is clear, therefore, that osteoporosis will truly become a global problem over the next half century and that measures are urgently required to avert this trend. Such projections would be worsened by any increase in hip fracture incidence. For example, an increase of only 1% in the age-adjusted incidence rate each year might cause the estimated number of hip fractures in 2050 to reach 8.1 million [92]. Fortunately, hip fracture rates appear to have

564

MELTON AND COOPER

TABLE 3

Health Care Expenditures Attributable to Osteoporotic Fractures in the United States by Type of Service and Type of Fracture (1995)a Type of service (millions of dollars)

Type of fracture Hip

Inpatient hospital

Emergency room

Outpatient physician

Outpatient hospital

Other outpatientb

Nursing home

Total

5576

130

67

9

90

2811

8682

Forearm

183

55

93

8

4

41

385

Spine

575

20

13

3

10

126

746

All other sites

2259

3632

297

45

91

899

3953

Total

8594

567

470

65

194

3875

13,764

a From N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton III, Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). b Includes home health care, ambulance services, and medical equipment.

leveled off in the northern United States [93], in parts of Scandinavia [94 – 97] and in Great Britain [98], although rates in Asia have risen substantially in recent years [99]. On the basis of current trends, hip fractures might increase in the United Kingdom from 46,000 in 1985 to 117,000 in 2016 [100]. In Australia, they could rise from 10,150 in 1986 to 18,550 in 2011, with a doubling in the cost of care in constant dollars from $38 to $69 million annually [101]. Health authorities in Finland expect to see a three-fold increase in the number of hip fractures between 1997 and 2030 [102], while in Canada, there could be a four-fold increase by 2041 [103]. Incidence rates for fractures at other skeletal sites have also risen during the last half century [104]. There are three broad explanations for this trend. First, it might reflect the influence of some increasingly prevalent risk factor for

osteoporosis or for falling. Time trends for a number of osteoporosis risk factors, including oophorectomy, estrogen replacement therapy, cigarette smoking, alcohol consumption, and dietary calcium intake, do not match those observed for hip fractures [105], but physical activity could be a candidate. Ample evidence links inactivity to hip fracture risk [106 – 109], and the steepest increases in incidence have been observed in Asian countries, which have witnessed dramatic reductions in customary activity levels [99]. A second possible explanation is increasing frailty among the elderly population, especially since many of the disorders contributing to frailty are also associated with osteoporosis and the risk of falling [110]. Finally, the trends could arise from a cohort phenomenon, i.e., some adverse influence that acted at an earlier time is now manifesting as a rising fracture incidence in successive generations. For example, it has been speculated that the increase in adult height during this century led to a secular trend toward longer hip axis length, which may increase the risk of hip fracture [111]. However, analysis of data from the Oxford Record Linkage Study revealed a declining incidence of hip fracture in more recent birth cohorts [112]. This is in accord with a recent survey in southern England, where fracture rates increased in the elderly between 1978 and 1995 but were tending to decrease among people under 70 years of age [113].

VI. CONCLUSIONS

FIGURE 5

Estimated number of fractures (in thousands) for men and women in different regions of the world in 1990, 2025, and 2050. Modified from C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992).

Osteoporosis is a complex, multifactorial chronic disorder in which a variety of pathophysiologic mechanisms lead to a progressive reduction in bone strength and an increased risk of fracture. Although viewed for many years as a major public health problem, the exact burden posed by

21 Magnitude and Impact of Osteoporosis and Fractures

565

osteoporosis is only now being rigorously assessed. Whether the disorder is defined by low bone mass or by the occurrence of specific fractures, osteoporosis is clearly a common condition. Thus, a third of postmenopausal white women in the United States can be expected to have osteoporosis in the lumbar spine, proximal femur, or midradius at any point in time, while the lifetime risk of a hip, spine, or distal forearm fracture from age 50 years onward in this group approaches 40%. However, the relative absence of symptoms until fractures occur makes effective therapeutic intervention difficult to implement. The public health burden will worsen dramatically in future decades, and the evaluation of strategies to prevent these fractures, both in individuals and in populations, has become an urgent priority.

density in the postmenopausal estrogen/progestin interventions trial. J. Bone Miner. Res. 9, 1467–1476 (1994). T. Cundy, J. Cornish, M. C. Evans, G. Gamble, J. Stapleton, and I. R. Reid, Sources of interracial variation in bone mineral density. J. Bone Miner. Res. 10, 368–373 (1995). G. S. Bhudhikanok, M.-C. Wang, K. Eckert, C. Matkin, R. Marcus, and L. K. Bachrach, Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J. Bone Miner. Res. 11, 1545–1556 (1996). L. J. Melton III, How many women have osteoporosis now? J. Bone Miner. Res. 10, 175–177 (1995). M. Lunt, D. Felsenberg, J. Adams, L. Benevolenskaya, J. Cannata, J. Dequeker, C. Dodenhof, J. A. Falch, O. Johnell, K.-T., Khaw, P. Masaryk, H. Pols, G. Poór, D. Reid, C. Scheidt-Nave, K. Weber, A. J. Silman, and J. Reeve, Population-based geographic variations in DXA bone density in Europe: The EVOS Study. Osteoporos. Int. 7, 175–189 (1997). M. C. Hochberg, J. Williamson, E. A. Skinner, J. Guralnik, J. D. Kasper, and L. P. Fried, The prevalence and impact of self-reported hip fracture in elderly community-dwelling women: The Women’s Health and Aging Study. Osteoporos. Int. 8, 385–389 (1998). L. J. Melton III, T. M. Therneau, and D. R. Larson, Long-term trends in hip fracture prevalence: The influence of hip fracture incidence and survival. Osteoporos. Int. 8, 68–74 (1998). T. W. O’Neill, D. Felsenberg, J. Varlow, C. Cooper, J. A. Kanis, and A. J. Silman, The prevalence of vertebral deformity in European men and women: The European Vertebral Osteoporosis Study. J. Bone Miner. Res. 11, 1010 – 1018 (1996). G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedüs, and I. Transbøl, Epidemiology of postmenopausal spinal and long bone fractures: A unifying approach to postmenopausal osteoporosis. Clin. Orthop. 166, 75 – 81 (1982). B. Ettinger, D. M. Black, M. C. Nevitt, A. C. Rundle, J. A. Cauley, S. R. Cummings, and H. K. Genant, The Study of Osteoporotic Fractures Research Group, Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). L. J. Melton III, A. W. Lane, C. Cooper, R. Eastell, W. M. O’Fallon, and B. L. Riggs, Prevalence and incidence of vertebral deformities. Osteoporos. Int. 3, 113 – 119 (1993). G. Jones, C. White, T. Nguyen, P. N. Sambrook, P. J. Kelly, and J. A. Eisman, Prevalent vertebral deformities: Relationship to bone mineral density and spinal osteophytosis in elderly men and women. Osteoporos. Int. 6, 233 – 239 (1996). L. J. Melton III, E. A. Chrischilles, C. Cooper, A. W. Lane, and B. L. Riggs, How many women have osteoporosis? J. Bone Miner. Res. 7, 1005 – 1010 (1992). C. Cooper, Epidemiology of osteoporosis. Osteoporos. Int. 9 (Suppl. 2), S2 – S8 (1999). W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. I. Frequency and distribution. Mayo Clin. Proc. 54, 701 – 707 (1979). Editorial, Fracture patterns revisited. Lancet 336, 1290 – 1291 (1990). L. J. Donaldson, A. Cook, and R. G. Thomson, Incidence of fractures in a geographically defined population. J. Epidemiol. Community Health 44, 241 – 245 (1990). J. L. Kelsey, Osteoporosis: Prevalence and incidence. In “Osteoporosis,” pp. 25 – 28. Proceedings of the NIH Consensus Development Conference, April 2 – 4, 1984. L. J. Melton III, Epidemiology of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 133 – 154. Raven Press, New York, 1988. C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A worldwide projection. Osteoporos. Int. 2, 285 – 289 (1992).

CHAPTER

12.

13.

14. 15.

16.

Acknowledgments

17.

The authors thank Mrs. Gill Strange for her assistance in preparing the manuscript. This work was supported in part by Research Grants AG04875 and AR27065 from the National Institutes of Health, U.S. Public Health Service, and by the Medical Research Council of Great Britain.

18.

19.

References 1. D. Schapira and C. Schapira, Osteoporosis: The evolution of a scientific term. Osteoporos. Int. 2, 164–167 (1992). 2. J. A. Kanis and the WHO Study Group, Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: Synopsis of a WHO report. Osteoporos. Int. 4, 368–381 (1994). 3. C. Cooper, The epidemiology of fragility fractures: Is there a role for bone quality? Calcif. Tissue Int. 53 (Suppl. 1), S23–S26 (1993). 4. Anonymous, Consensus Development Statement: Who are candidates for prevention and treatment for osteoporosis? Osteoporos. Int. 7, 1–6 (1997). 5. A. Cooper, “A Treatise on Dislocations and Fractures of the Joints” (B. B. Cooper, ed.). John Churchill, London, 1842 6. World Health Organization, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis,” WHO Technical Report Series. WHO, Geneva, 1994. 7. A. C. Looker, E. S. Orwoll, C. C. Johnston, Jr., R. L. Lindsay, H. W. Wahner, W. C. Dunn, M. S. Calvo, T. B. Harris, and S. P. Heyse, Prevalence of low femoral bone density in older US adults from NHANES III. J. Bone Miner. Res. 12, 1761–1768 (1997). 8. E. Seeman, Growth in bone mass and size: Are racial and gender differences in bone mineral density more apparent than real? J. Clin. Endocrinol. Metab. 83, 1414–1419 (1998). 9. R. A. Faulkner, R. G. McCulloch, S. L. Fyke, W. E. De Couteau, H. A. McKay, D. A. Bailey, C. S. Houston, and A. A. Wilkinson, Comparison of areal and estimated volumetric bone mineral density values between older men and women. Osteoporos. Int. 5, 271–275 (1995). 10. L. J. Melton III, E. J. Atkinson, M. K. O’Connor, W. M. O’Fallon, and B. L. Riggs, Bone density and fracture risk in men. J. Bone Miner. Res. 13, 1915–1923 (1998). 11. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral

20.

21.

22.

23.

24. 25.

26. 27.

28.

29.

30.

566 31. L. J. Melton III, E. Y. S. Chao, and J. Lane, Biomechanical aspects of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 111 – 131. Raven Press, New York, 1988. 32. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton III, Changes in bone mineral density of the proximal femur and spine with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982). 33. D. Marshall, O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312, 1254 – 1259 (1996). 34. S. J. Winner, C. A. Morgan, and J. G. Evans, Perimenopausal risk of falling and incidence of distal forearm fracture. Br. Med. J. 298, 1486 – 1488 (1989). 35. M. J. Gibson, The prevention of falls in later life. Dan. Med. Bull. 34 (Suppl. 4), 1 – 24 (1987). 36. L. J. Melton III, Differing patterns of osteoporosis across the world. In “New Dimensions in Osteoporosis in the 1990s” (C. H. Chesnut III, ed.), pp. 13 – 18. Asia Pacific Conference Series 125, Excerpta Medica, Hong Kong, 1991. 37. O. Johnell, B. Gullberg, E. Allander, J. A. Kanis, and the MEDOS Study Group, The apparent incidence of hip fracture in Europe: A study of national register sources. Osteoporos. Int. 2, 298 – 302 (1992). 38. P. Gärdsell, O. Johnell, B. E. Nilsson, and I. Sernbo, Bone mass in an urban and a rural population: A comparative, population-based study in southern Sweden. J. Bone Miner. Res. 6, 67 – 75 (1991). 39. S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Regional variation in the incidence of hip fracture: US white women aged 65 years and older. JAMA 264, 500 – 502 (1990). 40. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis: Its clinical features. JAMA 116, 2465 – 2474 (1941). 41. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, The incidence of clinically diagnosed vertebral fractures: A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 42. S. J. Jacobsen, C. Cooper, M. S. Gottlieb, J. Goldberg, D. P. Yahnke, and L. J. Melton III, Hospitalization with vertebral fracture among the aged: A national population-based study, 1986 – 1989. Epidemiology 3, 515 – 518 (1992). 43. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 44. R. D. Wasnich, P. D. Ross, C. J. MacLean, J. W. Davis, and J. M. Vogel, The relative strengths of osteoporotic risk factors in a prospective study of postmenopausal osteoporosis. In “Osteoporosis 1987” (C. Christiansen, J. S. Johansen, and B. J. Riis, eds.), Vol. 1, pp. 394 – 395. Osteopress ApS, Copenhagen, 1987. 45. C. Cooper, S. Shah, D. J. Hand, J. Adams, J. Compston, M. Davie, A. Woolf, and the Multicentre Vertebral Fracture Study Group, Screening for vertebral osteoporosis using individual risk factors. Osteoporos. Int. 2, 48 – 53 (1991). 46. C. Huang, P. D. Ross, S. Fujiwara, J. W. Davis, R. S. Epstein, K. Kodama, and R. D. Wasnich, Determinants of vertebral fracture prevalence among native Japanese women and women of Japanese descent living in Hawaii. Bone 18, 437 – 442 (1996). 47. M. Naves Diaz, T. W O’Neill, and A. J. Silman, The influence of alcohol consumption on the risk of vertebral deformity: European Vertebral Osteoporosis Study Group. Osteoporos. Int. 7, 65 – 71 (1997). 48. A. J. Silman, T. W. O’Neill, C. Cooper, J. Kanis, and D. Felsenberg, Influence of physical activity on vertebral deformity in males and women: Results from the European Vertebral Osteoporosis Study. J. Bone Miner. Res. 12, 813 – 819 (1997). 49. T. W. O’Neill, A. J. Silman, M. Naves Diaz, C. Cooper, J. Kanis, and D. Felsenberg, Influence of hormonal and reproductive factors on the

MELTON AND COOPER

50.

51.

52.

53. 54.

55.

56.

57.

58.

59.

60. 61.

62. 63. 64.

65.

66.

67.

68. 69.

70.

risk of vertebral deformity in European women: European Vertebral Osteoporosis Study Group. Osteoporos. Int. 7, 72 – 78 (1997). L. J. Melton III, E. J. Atkinson, S. Khosla, W. M. O’Fallon, and B. L. Riggs, Secondary osteoporosis and the risk of vertebral deformities in women. Bone 24, 49 – 55 (1999). M. C. Nevitt and S. R. Cummings, Type of fall and risk of hip and wrist fractures: The Study of Osteoporotic Fractures Research Group. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). W. C. Graafmans, M. E. Ooms, P. D. Bezemer, L. M. Bouter, and P. Lips, Different risk profiles for hip fractures and distal forearm fractures: A prospective study. Osteoporos. Int. 6, 427 – 431 (1996). C. Cooper, Osteoporosis: An epidemiological perspective: A review. J. R. Soc. Med. 82, 753 – 757 (1989). S. J. Jacobsen, D. J. Sargent, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Contribution of weather to the seasonality of distal forearm fractures: A population-based study in Rochester, Minnesota. Osteoporos. Int. 9, 254 – 259 (1999). D. G. Seeley, W. S. Browner, M. C. Nevitt, H. K. Genant, J. C. Scott, and S. R. Cummings, Which fractures are associated with low appendicular bone mass in elderly women? The Study of Osteoporotic Fractures Research Group. Ann. Intern. Med. 115, 837 – 842 (1991). S. H. Rose, L. J. Melton III, B. F. Morrey, D. M. Ilstrup, and B. L. Riggs, Epidemiologic features of humeral fractures. Clin. Orthop. 168, 24 – 30 (1982). B. Kristiansen, G. Barfod, J. Bredesen, J. Erin-Madsen, B. Grum, M. W. Horsnaes, and J. R. Aalberg, Epidemiology of proximal humeral fractures. Acta Orthop. Scand. 58, 75 – 77 (1987). J. L. Kelsey, W. S. Browner, D. G. Seeley, M. C. Nevitt, and S. R. Cummings, Risk factors for fractures of the distal forearm and proximal humerus: The Study of Osteoporotic Fractures Research Group. Am J. Epidemiol. 135, 477 – 489 (1992). L. J. Melton III, J. M. Sampson, B. F. Morrey, and D. M. Ilstrup, Epidemiologic features of pelvic fractures. Clin. Orthop. 155, 43 – 47 (1981). A. J. Buhr and A. M. Cooke, Fracture patterns. Lancet 1, 531 – 536 (1959). U. Bengnér, O. Johnell, and I. Redlund-Johnell, Increasing incidence of tibia condyle and patella fractures. Acta Orthop. Scand. 57, 334 – 336 (1986). E. Foltin, Bone loss and forms of tibial condylar fracture. Arch. Orthop. Trauma Surg. 106, 341 – 348 (1987). R. Hedlund and U. Lindgren, Epidemiology of diaphyseal femoral fracture. Acta Orthop. Scand. 57, 423 – 427 (1986). T. J. Arneson, L. J. Melton III, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin Orthop. 234, 188 – 194 (1988). L. Forsén, A. J. Søgaard, H. E. Meyer, T.-H. Edna, and B. Kopjar, Survival after hip fracture: Short- and long-term excess mortality according to age and gender. Osteoporos. Int. 10, 73 – 78 (1999). G. Poór, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Determinants of reduced survival following hip fractures in men. Clin. Orthop. 319, 260 – 265 (1995). M. T. Clayer and R. J. Bauze, Morbidity and mortality following fractures of the femoral neck and trochanteric region: Analysis of risk factors. J. Trauma 29, 1673 – 1678 (1989). G. Poór, S. J. Jacobsen, and L. J. Melton III, Mortality following hip fracture. Facts Res. Gerontol. 7, 91 – 109 (1994). C. Cooper, E. J. Atkinson, S. J. Jacobsen, W. M. O’Fallon, and L. J. Melton III, Population-based study of survival following osteoporotic fractures. Am. J. Epidemiol. 137, 1001 – 1005 (1993). D. M. Kado, W. S. Browner, L. Palermo, M. C. Nevitt, H. K. Genant, and S. R. Cummings, Vertebral fractures and mortality in older women: A prospective study. Arch. Intern. Med. 159, 1215 – 1220 (1999).

21 Magnitude and Impact of Osteoporosis and Fractures

567

71. W. S. Browner, D. G. Seeley, T. M. Vogt, and S. R. Cummings, Nontrauma mortality in elderly women with low bone mineral density: Study of Osteoporotic Fractures Research Group. Lancet 338, 355 – 358 (1991). 72. C. Johansson, D. Black, O. Johnell, A. Odén, and D. Mellström, Bone mineral density is a predictor of survival. Calcif. Tissue Int. 63, 190 – 196 (1998). 73. P. von der Recke, M. A. Hansen, and C. Hassager, The association between low bone mass at the menopause and cardiovascular mortality. Am. J. Med. 106, 273 – 278 (1999). 74. E. A. Chrischilles, C. D. Butler, C. S. Davis, and R. B. Wallace, A model of lifetime osteoporosis impact. Arch. Intern. Med. 151, 2026 – 2032 (1991). 75. W. Miller, Survival and ambulation following hip fracture. J. Bone Joint. Surg. 60A, 930 – 934 (1978). 76. J. S. Jensen and J. Bagger, Long-term social prognosis after hip fractures. Acta Orthop. Scand. 53, 97 – 101 (1982). 77. S. K. Bonar, M. E. Tinetti, M. Speechley, and L. M. Cooney, Factors associated with short- versus long-term skilled nursing facility placement among community-living hip fracture patients. J. Am. Geriatr. Soc. 38, 1139 – 1144 (1990). 78. N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton III, Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). 79. T. L. Holbrook, K. Grazier, J. L. Kelsey, and R. N. Stauffer, “The Frequency of Occurrence, Impact and Cost of Selected Musculoskeletal Conditions in the United States.” American Academy of Orthopedic Surgeons, Chicago, 1984. 80. P. D. Ross, J. W. Davis, R. S. Epstein, and R. D. Wasnich, Pain and disability associated with new vertebral fractures and other spinal conditions. J. Clin. Epidemiol. 47, 231 – 239 (1994). 81. G. Leidig, H. W. Minne, P. Sauer, C. Wuster, M. Logen, F. Raue, and R. Ziegler, A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. 8, 217 – 229 (1990). 82. W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. II. Orthopedic treatment and utilization of health care. Mayo Clin. Proc. 54, 708 – 713 (1979). 83. J. A. Kanis and F. A. Pitt, Epidemiology of osteoporosis. Bone 13 (Suppl. 1), S7 – S15 (1992). 84. R. M. Atkins, T. Duckworth, and J. A. Kanis, Features of algodystrophy following Colles fracture. J. Bone Joint Surg. 72B, 105 – 110 (1990). 85. H. P. de Bruijn, The Colles fracture, review of literature. Acta Orthop. Scand. 58 (Suppl. 223), 7 – 25 (1987). 86. J.-P. Kaukonen, E. O. Karaharju, M. Porras, P. Lüthje, and A. Jakobsson, Functional recovery after fractures of the distal forearm. Ann. Chir. Gynaecol. 77, 27 – 31 (1988). 87. A. Praemer, S. Furner, and D. P. Rice, “Musculoskeletal Conditions in the United States.” American Academy of Orthopaedic Surgeons, Park Ridge, IL, 1992. 88. S. Eiskjaer, S. E. Østgård, B. W. Jakobsen, J. Jensen, and U. Lucht, Years of potential life lost after hip fracture among postmenopausal women. Acta Orthop. Scand. 63, 293 – 296 (1992). 89. K. Lippuner, J. von Overbeck, R. Perrelet, H. Bosshard, and Ph. Jaeger, Incidence and direct medical costs of hospitalizations due to osteoporotic fractures in Switzerland. Osteoporos. Int. 7, 414 – 425 (1997). 90. Royal College of Physicians of UK. “Guidelines for the Prevention and Treatment of Osteoporosis.” Royal College of Physicians of UK, London, 1999. 91. E. Levy, Cost analysis of osteoporosis related to untreated menopause. Clin. Rheumatol. 8 (Suppl. 2), 76 – 82 (1989). 92. B. Gullberg, O. Johnell, and J. A. Kanis, World-wide projections for hip fracture. Osteoporos. Int. 7, 407 – 413 (1997). 93. L. J. Melton III, E. J. Atkinson, and R. Madhok, Downturn in hip fracture incidence. Public Health Rep. 111, 146 – 150 (1996).

94. T. Naessen, R. Parker, I. Persson, M. Zack, and H. O. Adami, Time trends in incidence rates of first hip fracture in the Uppsala Health Care Region, Sweden, 1965 – 1983. Am. J. Epidemiol. 130, 289 – 299 (1989). 95. L. Rehnberg, S. Nungu, and C. Olerud, The incidence of femoral neck fractures in women is decreasing. Acta Orthop. Scand. 63 (Suppl. 248), 92 – 93 (1992). 96. T. M. Huusko, P, Karppi, V. Avikainen, H. Kautiainen, and R. Sulkava, The changing picture of hip fractures: Dramatic change in age distribution and no change in age-adjusted incidence within 10 years in central Finland. Bone 24, 257 – 259 (1999). 97. C. Rogmark, I. Sernbo, O. Johnell, and J.-Å. Nilsson, Incidence of hip fractures in Malmö, Sweden, 1992 – 1995: A trend-break. Acta Orthop. Scand. 70, 19 – 22 (1999). 98. T. D. Spector, C. Cooper, and A. F. Lewis, Trends in admissions for hip fracture in England and Wales, 1968 – 85. Br. Med. J. 300, 1173 – 1174 (1990). 99. E. M. Lau and C. Cooper, The epidemiology of osteoporosis: The Oriental perspective in a world context. Clin. Orthop. 323, 65 – 74 (1996). 100. R. Hoffenberg, O. F. W. James, J. C. Brocklehurst, I. D. Green, P. Horracks, J. A. Kanis, N. J. Wald, G. E. MacLellan, and R. H. Vickers, Fractured neck of femur: Prevention and management. Summary and recommendations of a report of the Royal College of Physicians. J. R. Coll. Phys. London 23, 8 – 12 (1989). 101. S. R. Lord and P. F. Sinnett, Femoral neck fractures: Admissions, bed use, outcome and projections. Med. J. Aust. 145, 493 – 496, (1986). 102. P. Kannus, S. Niemi, J. Parkkari, M. Palvanen, I. Vuori, and M. Järvinen, Hip fractures in Finland between 1970 and 1997 and predictions for the future. Lancet 353, 802 – 805 (1999). 103. E. A. Papadimitropoulos, P. C. Coyte, R. G. Josse, and C. E. Greenwood, Current and projected rates of hip fracture in Canada. Can. Med. Assoc. J. 157, 1357 – 1363 (1997). 104. K. J. Obrant, U. Bengnér, O. Johnell, B. E. Nilsson, and I. Sernbo, Increasing age-adjusted risk of fragility fractures: A sign of increasing osteoporosis in successive generations? Calcif. Tissue Int. 44, 157 – 167 (1989). 105. L. J. Melton III, W. M. O’Fallon, and B. L. Riggs, Secular trends in the incidence of hip fractures. Calcif. Tissue Int. 41, 57 – 64 (1987). 106. C. Cooper, D. J. P. Barker, and C. Wickham, Physical activity, muscle strength and calcium intake in fracture of the proximal femur in Britain. Br. Med. J. 297, 1443 – 1446 (1988). 107. E. Lau, S. Donnan, D. J. P. Barker, and C. Cooper, Physical activity and calcium intake in fracture of the proximal femur in Hong Kong. Br. Med. J. 297, 1441 – 1443 (1988). 108. C. Wickham, K. Walsh, C. Cooper, D. J. Barker, B. M. Margetts, J. Morris, and S. A. Bruce, Dietary calcium, physical activity, and risk of hip fracture: A prospective study. Br. Med. J. 299, 889 – 892 (1989). 109. E. W. Gregg, J. A. Cauley, D. G. Seeley, K. E. Ensrud, and D. C. Bauer, Physical activity and osteoporotic fracture risk in older women: Study of Osteoporotic Fractures Research Group. Ann. Intern. Med. 129, 81 – 88 (1998). 110. C. Wickham, C. Cooper, B. M. Margetts, and D. J. P. Barker, Muscle strength, activity, housing and the risk of falls in elderly people. Age Ageing 18, 47 – 51 (1989). 111. I. R. Reid, K. Chin, M. C. Evans, and J. G. Jones, Relation between increase in length of hip axis in older women between 1950s and 1990s and increase in age specific rates of hip fracture. Br. Med. J. 309, 508 – 509 (1994). 112. J. G. Evans, V. Seagroatt, and M. J. Goldacre, Secular trends in proximal femoral fracture, Oxford record linkage study area and England, 1968 – 86. J. Epidemiol. Community Health 51, 424 – 429 (1997). 113. A. McColl, P. Roderick, and C. Cooper, Hip fracture incidence and mortality in an English region: A study using routine National Health Service data. J. Public Health Med. 20, 196 – 205 (1998).

CHAPTER

569

CHAPTER 22 Race, Ethnicity, and Osteoporosis

CHAPTER 22

CHAPTER 22

Race, Ethnicity, and Osteoporosis MARIE LUZ VILLA LORENE NELSON DOROTHY NELSON

Department of Medicine, Division of Gerontology and Geriatrics, University of Washington School of Medicine, Seattle, Washington 98104 Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201

IV. Racial and Ethnic Influences on Risk for Osteoporosis V. Summary References

I. We Are All Individuals II. Defining Terms III. Ethnoepidemiology of Osteoporosis

I. WE ARE ALL INDIVIDUALS Marked differences exist between groups that can be sorted according to descriptors such as race or ethnicity. Within racial and ethnic groups, large degrees of individual variation exist; this chapter attempts to address observed trends in bone dynamics that can be sorted by race or ethnic grouping. Sorting of humans, however, suffers from the inconstancy of racial and ethnic definitions. This chapter therefore defines terms used to describe human groups and then explores their application to the epidemiology of hip fracture and parameters that affect skeletal health. The diagnostic term “osteoporosis” refers to a skeletal condition that predisposes to multisite fragility fractures. Because only hip fractures reliably lead to hospitalization

OSTEOPOROSIS, SECOND EDITION VOLUME 1

and therefore documentation, we present hip fracture data in various populations as a reflection of morbidity resulting from osteoporosis.

II. DEFINING TERMS A. Race and Ethnicity “Ethnicity” and “race” appear interchangeably in many publications. “Race” in the United States reflects the belief that a limited number of genetically characterized human groups exist, exemplified by the list used by the U.S. Census: White/Caucasian, Black/African American, Native American/American Indian, Alaskan native/Eskimo/Aleut,

569

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

570

VILLA, NELSON, AND NELSON

Asian/Pacific Islander, and other (Spanish or Hispanic origin is asked separately). Most investigators recognize, however, that distinct racial lines may not be drawn due to significant genetic admixture that has occurred over time. Also, because environments change, and because populations move and interbreed, it is difficult if not impossible to identify discrete, biologically meaningful subgroups of humans. A factor that reflects cultural, religious, dietary, geographic, and other differences among races, known as ethnicity, then becomes important [1]. In fact, ethnicity plays an important role in disease prevalence even within races. For example, Hispanic Caucasians show different trends in disease incidence when compared to non-Hispanic Caucasians (an ethnic dichotomization of the Caucasian race specific to the Americas): Mexican Americans have a twoto fivefold greater risk of developing noninsulin dependent diabetes mellitus (NIDDM) than does the majority of the U.S. population [2]. Lack of ethnic definition of study groups affects the general applicability of data. A study reporting hip fracture rates of “Asians” does not help determine a Korean woman’s risk of suffering a hip fracture. Ethnic-specific data would be more valuable than a broad, racial summarization because bone mineral densities (BMD) and fracture rates vary among countries, as well as among ethnic subgroups. In a study comparing the average BMD of Japanese, Korean, and Taiwanese women, the Taiwanese had consistently greater BMD at the lumbar spine at almost every age [3]. In a study of ethnic/racial BMD differences among children, a large subgroup of “Whites” considered themselves Chaldean, an Iraqi ethnic group [4]. The Chaldean children’s whole body bone mass was significantly higher than– non-Chaldean White children and was not different from other study subjects who considered themselves Black. Because Middle Easterners are included in the U.S. Census category “White/Caucasian,” such a difference would not be expected a priori and would affect the results of the study if the Chaldeans were analyzed together with other White children.

B. Acculturation In addition to race and ethnicity, another factor known as acculturation contributes to nuances in measured variables. Acculturation scales measure how much an ethnic group assimilates the language, habits, and cultural values of the country or area to which it migrates [5 – 7]. Thus, a higher acculturation score means greater adoption of the dominant culture of a region. Degree of acculturation

may also affect dietary and lifestyle habits of people inhabiting a given region for many years: African Americans preserve many distinct customs and dietary habits when moving to regions outside the southern United States [8]. Returning to the example of ethnic differences in the incidence of diabetes, the prevalence of NIDDM varies within Hispanic ethnic groups, depending on the degree of acculturation: for Mexican Americans, the rate of NIDDM decreases with increasing acculturation [9]. Sometimes acculturation may serve as a proxy for factors that affect risk for osteoporosis but are themselves not easily measured, as the environment in which a group is born or raised can affect variance in observed fracture rates [10]. Hip fracture incidence for elderly Japanese women ranges from 450 to 1011 per 100,000 person-years/year [11], depending on their region of origin. Tools for retrospectively assessing fracture risk factors in elders suffer from substantial inaccuracy. If, in addition, the study group originated in another country or was reared under very different cultural conditions than the group for which the tool was validated, even greater measurement error may be introduced. Retrospectively recalled teenage physical activity and milk intake have no relationship to lumbar bone mass of postmenopausal Mexican American women, yet lumbar BMD is independently and positively related to acculturation, even when controlling for age and obesity [12]. Assessment of ethnicity and acculturation helps describe disease occurrence among human groups, contributing greatly to the description of factors that may affect observed differences in fracture rates. For this reason, as well as the pitfalls associated with “race” described earlier, we attempt to use “ethnic” instead of “racial” in most of the contexts that follow. However, data frequently are reported using “racial” categorizations without indication of ethnic grouping. We present those data as reported.

III. ETHNOEPIDEMIOLOGY OF OSTEOPOROSIS Great variation in the occurrence of osteoporotic fractures exists within and among different racial and ethnic groups. Some studies report wide ranges in hip fracture incidence rates within a given racial group, probably due to regional and/or cultural (ethnic and lifestyle) factors. (Geographic variation in fracture rates is discussed elsewhere in the text.) Fracture incidence rates in Caucasians vary among ethnic groups, as demonstrated in Table 1. The magnitude of these within-Caucasian differences approaches that noted between Caucasians and other racial groups [13 – 15].

571

CHAPTER 22 Race, Ethnicity, and Osteoporosis

A. Methodologic Issues We reviewed information on hip fracture incidence rates obtained from studies conducted among different racial and ethnic groups. Prior to summarizing these data, several methodologic issues are discussed that have important bearing on the ability to compare incidence rates among studies, including the need for age and sex standardization

to adjust for differences in the age and gender composition of the populations under study, differences between studies in definitions of hip fracture; and differences between studies in the methods that were used to identify individuals with hip fracture. Several recent articles provide insights into other methodologic issues that may also affect the ability make cross-national comparisons of hip fracture incidence rates [13,24,25,49].

TABLE 1 Age-Adjusted Ratesa of Hip Fracture per 100,000 Population for Females, Males, and Totalb and Year of Study Ethnic group Blacks

Hispanicsc Asians

Caucasiansd

Site (reference)

Years of study

Female

Male

Total

Female:male

USA [16]

1986 – 1989

214

179

200

1.2

Maryland [17]

1979 – 1988

345

191

283

1.8

USA [18]

1984 – 1985

344

235

300

1.5

California [19]

1983 – 1984

241

153

202

1.6

Texas [20]

1980

243

13

141

18.7

USA [14]

1974 – 1979

174

108

137

1.6

Johannesburg, South Africa [21]

1950 – 1964

26

20

23

1.3

California [19]

1983 – 1984

219

97

165

2.3

Texas [20]

1980

305

128

227

2.4

Tottori, Japan [22]

1994

342

136

249

2.5

Tottori, Japan [23]

1986 – 1987

227

79

163

2.9

Okinawa, Japan [11]

1984 – 1985

325

86

219

3.8

Beijing, China [24]

1990 – 1992

97

101

99

1.0

Beijing, China

1990 – 1992

96

107

**

0.9

Hong Kong [25]

1990 – 1992

428

270

**

1.4

Hong Kong [26]

1985

389

196

304

2.0

Hong Kong [27]

1965 – 1967

179

113

150

1.6

Singapore [28]

1955 – 1962

83

111

95

0.7

Kuwait [29]

1992 – 1995

378

279

333

1.4

New Zealand [30]

1973 – 1976

212

121

172

1.8

California [19]

1983 – 1984

383

116

265

3.3

Hawaii [11]

1979 – 1981

224

66

153

3.4

Sweden [31]

1985

714

268

517

2.7

Sweden [32]

1980

432

199

517

2.7

Sweden [33]

1980

984

338

705

2.9

Sweden [34]

1972 – 1981

714

319

540

2.2

Sweden [35]

1972 – 1981

730

581

664

1.3

Malmo, Sweden [36]

1950 – 1960

468

153

329

3.1

Norway [37]

1983 – 1984

737

298

543

2.5

Oslo, Norway [15]

1978 – 1979

850

329

620

2.6

1970

377

142

273

2.7

1990 – 1992

697

349

**

2.0 2.6

Finland [32] Reykjavik [25] Kuopio, Finland [38] Edin, Scotland [39] Oxford, England [40] Yorkshire, UK [41]

1968

280

107

204

1978 – 1979

529

174

376

3.0

1983

603

114

392

5.3

1973 – 1977

310

102

218

3.0 (continues)

572

VILLA, NELSON, AND NELSON TABLE 1

Ethnic group

(continued)

Site (reference)

Years of study

Alicante, Spain [42]

1974 – 1984

90

57

75

1.6

Italy [43]

1988 – 1989

287

110

207

2.6

Jerusalem, Israel [44]

1957 – 1966

355

168

272

2.1

Canada [45]

1976 – 1985

788

307

595

2.6

USA [16]

1986 – 1989

968

396

738

2.4

USA [18]

1984 – 1985

845

350

645

2.4

California [19]

1983 – 1984

617

215

439

2.9

Texas [20]

Female

Male

Total

Female:male

1980

593

223

430

2.7

1979 – 1988

950

358

712

2.7

Hawaii [11]

1979 – 1981

645

205

451

3.1

Minnesota [11]

1978 – 1982

613

285

468

2.2

USA [14]

1974 – 1979

422

151

285

2.8

USA [46]

1970 – 1983

705

244

506

2.9

Rochester [47]

1965 – 1974

559

191

396

2.9

New Zealand [30]

1973 – 1976

466

139

321

3.4

Australia [48]

1994 – 1996

575

244

425

2.4

Maryland [17]

a

Rates were age and gender adjusted to the 1990 U.S. non-Hispanic Caucasian population. Both age and gender adjusted. c Hispanic Caucasians. d Non-Hispanic Caucasians. b

1. NEED FOR AGE AND SEX STANDARDIZATION Because the number of elderly in the world’s population is increasing rapidly with time, and because studies differ with respect to the age and gender composition of the populations under study, hip fracture incidence rates obtained from different time periods and from different populations are not strictly comparable unless the age and gender differences between study populations have been taken into account. A method called standardization is used as a means to provide an estimate of the incidence rate in a given population if that population had the same gender and age composition as an arbitrarily selected standard population. The 1990 U.S. non-Hispanic Caucasian population served as the standard population for the age- and sex-adjusted incidence rates for hip fracture that are presented in Table 1. The studies summarized in Table 1 all contained information regarding age- and sex-specific incidence rates of hip fracture for individuals above 50 years of age so that age- and sex-adjusted incidence rates could be calculated. The differences among study populations in adjusted hip fracture incidence rates that exist after standardization are unlikely to be attributed to differences in age or gender composition that exist between the study populations. Of note, however, many studies treat individuals aged 80 and older as one group. Because fracture incidence rises steeply with age, the standardization process cannot adequately adjust for age if studies do not provide enough detailed data for the older age groups.

2. DIFFERENCES IN THE DEFINITION OF HIP FRACTURE Studies differ with respect to the amount of detail provided regarding the exact anatomic locations of the fractures. Each study summarized in Table 1 used one of the following definitions of hip fracture: (1) fracture of the femoral neck or proximal femur; (2) cervical, trochanteric, intracapsular, extracapsular, or intertrochanteric fracture; (3) hip fracture defined on the basis of International Classification of Disease (ICD) codes; or (4) hip fracture with no specification of fracture location. A small percentage of hip fractures result from severe trauma or from underlying pathology. In countries where hip fractures are uncommon, fractures due to severe trauma account for a larger proportion of the total number of hip fractures [50]. While some of the studies in Table 1 excluded fractures due to severe trauma, tumors, or metabolic bone diseases, others did not specify that these exclusions had occurred. 3. DIFFERENCES IN CASE ASCERTAINMENT METHODS Most studies of hip fracture suffer some degree of underascertainment due to the difficulties of identifying every person with hip fracture. All studies included in Table 1 ascertained cases of hip fracture through hospital records, usually through hospital discharge diagnoses. Reasons for underascertainment are that some fractures are misclassified as femoral shaft fractures, and some individuals with hip fracture are not hospitalized either because health services are not available or because they are treated in

CHAPTER 22 Race, Ethnicity, and Osteoporosis

another setting (i.e., at home, in a chronic care facility, or by a native healer). Although the latter sources of bias differ substantially from one country or study to another, they may not contribute substantially to fracture statistics in those countries listed in Table 1.

B. Racial and Ethnic Differences in Rates of Hip Fracture Despite methodologic difficulties that affect the comparison of hip fracture rates among studies, broad conclusions can be drawn regarding differences in hip fracture incidence rates for members of different races and ethnic groups. Age- and sex-adjusted incidence rates of hip fracture in Blacks, Hispanic Caucasians, Asian or Pacific Islanders, and non-Hispanic Caucasians are presented in Table 1. Within each race or ethnic group and each country or geographic region, studies are arranged in order of most recent to least recent so that cross-ethnic comparisons can be made between studies that have been conducted in similar time periods. In addition, many authors do not give detailed information about the racial and ethnic backgrounds of groups studied so data presented reflect use of ethnic and racial categories as published. Since few studies prior to 1980 were conducted in groups other than non-Hispanic Caucasians, our discussion will be largely focused on studies conducted in the 1980s and 1990s. Caucasians have the highest hip fracture incidence rates of any race or ethnic group, and this is particularly striking in northern Europe and North America. Studies in the United States demonstrate that hip fracture incidence in Asian Americans is intermediate to those of non-Hispanic Caucasians and Blacks. Although the number of studies in Hispanics is small, estimates of hip fracture incidence in this group are close to (and in some cases lower than) rates among Black Americans [19 – 21]. Hip fracture incidence among Black South Africans was reported to be very low in a study conducted between 1950 and 1964 [21], but there are no recent studies in African populations. Studies conducted in racially diverse populations using the same methodology for ascertaining hip fractures in all groups are particularly valuable for making inferences about racial differences in hip fracture incidence [16 – 20,30]. A cross-national study carried out in five geographic areas in the years 1990 – 1992 used similar methodologies in all areas and corrected for methodologic differences between studies [25]. The highest hip fracture incidence rates occurred in Iceland, intermediate rates in Hungary and Hong Kong, and lowest rates in Beijing. Other studies found that hip fracture incidence in Beijing and Singapore appears to be much lower than rates in Hong Kong or other Asian countries; reasons for these differences are not understood [24,28]. Another cross-national study re-

573 lied solely on hospital discharge diagnoses to identify hip fracture cases [49]. The highest rates were reported for the European and North American countries, and the lowest fractures rates were observed in Venezuela and Chile; however, sole reliance on hospital discharge data likely resulted in the underascertainment of hip fracture cases [25]. Many studies of racially diverse populations have been conducted in the United States and they have consistently reported higher rates among Caucasians than among other racial and ethnic groups. In a study conducted in Bexar County, Texas, age- and sex-adjusted hip fracture incidence was lowest in Blacks and highest in non-Hispanic Caucasians, with intermediate rates for Hispanics [20]. Studies in the United States have utilized Medicare data from the Health Care Financing Administration to estimate hip fracture incidence rates among elderly individuals (ages 65 and older). These studies have confirmed that U.S. Hispanic rates are intermediate between the lower Black rates and the higher Caucasian rates [51]. In contrast, Silverman et al. [19] found that age- and sex-adjusted incidence of hip fracture in California was lower among Hispanic Caucasians than among all other groups, including non-Hispanic Caucasians, Blacks, and Asians. A study comparing hip fracture incidence among native Japanese, Japanese Americans, and non-Hispanic Caucasian Americans [11] reported the lowest rates among Japanese Americans and the highest rates among non-Hispanic Caucasians. The analysis of Medicare data for individuals ages 65 and older found that age-adjusted hip fracture incidence rates were lower for all Asian groups (Chinese, Japanese, and Korean Americans) than for Caucasians [52]. Also of interest are racial and ethnic differences with respect to the magnitude of the female:male ratio of hip fracture incidence rates, which usually exceeds two in non-Hispanic Caucasian populations. Also of interest are racial and ethnic differences with respect to the magnitude of the female:male ratio of hip fracture incidence rates. Among Black Americans, hip fracture is more common among women than men; however, the female:male ratio is usually below two in contrast to non-Hispanic Caucasian populations where the gender ratio usually exceeds two. In the studies conducted in Hispanic populations, the gender ratio is close to two [19,20]. All studies but two [24,28] in Asian populations demonstrate a higher incidence of hip fracture in Asian women than men, with two studies reporting female:male ratios that exceed 3.0 [11,19]. Increasing age is an established risk factor for hip fracture in all racial and ethnic groups. An excellent review of differences in age-specific incidence rates of hip fracture between racial and ethnic groups is provided by Maggi et al. [13]. Although hip fracture incidence increases with age in all ethnic groups, the increase occurs earlier in non-Hispanic Caucasian populations than in

574

VILLA, NELSON, AND NELSON

Black, Asian, and Hispanic populations [13]. Studies conducted in non-Hispanic Caucasian populations report higher rates of hip fracture among men than women before 50 years of age, whereas after age 50, women have higher rates than men. Environmental factors such as diet, level of physical activity, frequency of cigarette smoking, and use of hormonal medications may explain some of the differences in hip fracture incidence observed between racial and ethnic groups. Factors that may contribute to racial and ethnic differences in skeletal health and risk for hip fracture are discussed in the remainder of this chapter.

IV. RACIAL AND ETHNIC INFLUENCES ON RISK FOR OSTEOPOROSIS Many factors affect the risk of developing osteoporosis or suffering nontraumatic hip fracture, as described elsewhere in this volume. It is not clear that all populations are similarly characterized with respect to osteoporotic risk factors, as most were established from studies of non-Hispanic Caucasians. It does seem intuitive that most people should respond similarly to factors such as reproductive hormone status, medication use, and physical activity. Other factors, such as calcium metabolism, bone mass, and body composition, may have different effects from one ethnic group to the next.

A. Bone Mass The term bone mass can refer to a variety of measurements, including bone mineral content (BMC in g), areal bone mineral density (BMD in g/cm2), and volumetric bone density (BMD in g/cm3). The degree of ethnic differences in bone mass reported by various investigators varies with the measurement used as well as other factors. American Blacks have significantly greater areal bone mass than Caucasians [53 – 59], which is thought to contribute to their lower rate of hip fracture. Kleerekoper and colleagues have shown that volumetric BMD measured by quantitative computed tomography (QCT) is 40% higher in African American compared with White women [60], considerably greater than the 5 – 15% difference in areal BMD generally reported for African American versus Caucasian adults. It is not well understood whether racial differences in bone mass exist at birth or develop at some point thereafter. Most studies based on absorptiometry (SPA, DXA) find higher bone mass (bone mineral content or areal bone density) in African Americans throughout childhood [61 – 66]. Some studies of volumetric bone density (g/cm3),

based on QCT, find no distinction in skeletal status between young Black and Caucasian children [67,68]. Gilsanz and colleagues [69] examined bone mineral density in Black and non-Hispanic Caucasian children at different stages of sexual development and found that significant racial differences did not occur until late puberty. In contrast, recent investigations describe significant differences in volumetric (BMAD) femoral neck bone density, based on DXA, at all stages of puberty [70,71]. It has been hypothesized that the higher bone mass seen in North American Blacks stems in part from genetic factors. However, the U.S. Blacks’ gene pool is very heterogeneous and is the result of much admixture over several centuries. It might be assumed that any population of African origin would have a high bone mass similar to African Americans, but this has not been borne out. Investigations of Blacks in South Africa [53,72 – 74] and the Gambia [67,75] showed that their bone mass does not exceed and, in some cases, is lower than that of age-matched African Caucasians. These data illustrate the difficulty in generalizing about a “racial” group, when obviously ethnic gradations in bone mass exist within people of African descent, with further differences introduced by acculturation in areas to which Black Africans migrated. The positive relationship between high bone mass and low incidence of hip fracture is seen in New Zealand Polynesians [30,76]. However, Hispanics have hip fracture incidence rates comparable to those of African Americans, yet have bone mass values closer to those of non-Hispanic Caucasians [12,58,77,78]; Asian Americans demonstrate a similar relationship [52,79 – 81]. One South American ethnic group in Vilcabamba, Ecuador, enjoys an extremely low rate of hip fracture despite bone mineral density values much lower than that of non-Hispanic Caucasians [82]. Few studies exist investigating bone health in American Indian tribes [83 – 86]. Data presented in abstract form by Chen and colleagues [87] showed BMD of Native Americans residing in the Southwest to be significantly lower than that of Whites, although the study groups were not well matched for age. Unfortunately, hip fracture studies entirely exclude Native Americans so that relationships between bone mass and fracture risk cannot be drawn. Bone mass therefore may not be the factor that best predicts fracture risk in many racial and ethnic groups, rendering World Health Organization guidelines for diagnosis of osteoporosis [88] only narrowly applicable.

B. Bone Turnover Risk for osteoporotic fracture depends not only on the mass of bone, but its quality as well. The rate and efficiency of bone turnover (affected by reproductive hormone status,

575

CHAPTER 22 Race, Ethnicity, and Osteoporosis

body composition, vitamin D/calcium nutriture, and physical activity) affects bone architecture, which is an essential component of skeletal strength. Low bone turnover in Black adults may partially explain their greater lifelong bone mass and lower fracture risk than non-Hispanic Caucasians. Some studies found biochemical evidence that African Americans have lower rates of bone turnover than non-Hispanic Caucasians [60,89,90]. In addition, histomorphometric studies showed mean rates of bone formation in American Blacks to be significantly lower than those of non-Hispanic Caucasians [90,91]. These authors concluded that most, if not all, ethnic differences observed in bone cell function could be the result of differences in bone accumulation during growth: higher bone mass would result in less fatigue damage and less need for repair by directed bone remodeling. However, studies comparing South African Blacks and Caucasians suggested higher bone turnover in Blacks, which was hypothesized to lead to fewer fractures because of better trabecular bone quality and less skeletal fragility [92,93]. Evidence of no black/white differences in bone turnover or mass has been presented as well [94]. These contrasting studies again highlight the pitfalls associated with assuming that subgroups (such as geographically different populations) of a “racial” group will be biologically similar. No studies have directly compared bone turnover of Asians, Blacks, and Caucasians, but it appears that the reported normal values for circulating osteocalcin in Japanese women are lower than those of non-Hispanic Caucasian women [95,96]. However, Polynesians and Caucasians do not manifest different serum concentrations of osteocalcin or parathyroid hormone (PTH), or urinary excretion of hydroxyproline, despite significant differences in BMD [97]. In one study of young Mexican American and non-Hispanic Caucasians, osteocalcin concentrations did not differ significantly between the two groups, despite differences in 25hydroxyvitamin D and PTH values [98]. Differences in reproductive hormone status may contribute to ethnic and racial variation in bone turnover, skeletal quality, and subsequent fracture risk. Androgens and estrogens contribute positively and independently to attainment of peak bone mass [99,100], and adult bone loss often stems from the increased bone turnover associated with decreased levels of reproductive hormones [95]. Furthermore, serum unbound sex steroid concentrations are lower in women with hip fracture than in controls [101]. Gilsanz and colleagues found that racial differences in bone mass develop during late stages of puberty (69), perhaps related to differences in serum sex hormone status. Bone loss in Japanese women appears to be greatest in the early postmenopausal period, but subsequently declines at rates similar to those for non-Hispanic Caucasians [95,102]. In a large study of postmenopausal women con-

ducted in the northeastern United States, race and serum estrone concentrations contributed independently to observed racial differences in bone mass [103]. Serum estrone values were significantly higher in African Americans, but this difference disappeared when analyses were adjusted for obesity (as determined by body mass index  27.3 kg/m2).

C. Body Size and Composition Body size appears to be an independent contributor to variance in BMD. Therefore, use of a mathematical correction for differences in densitometric bone size [104] from one population to another might correct for differences in body habitus and shed some light on the seeming discrepancies between bone mass and fracture risk across ethnic groups. In a multisite study of hormone replacement and its effects on bone mass in postmenopausal women, it was noted that although African Americans had the highest measured bone mass, when adjustments were made for bone size, the ethnic differences in bone density were significantly attenuated [55]. Analysis of Eskimo bone mineral content showed it to be lower than that of non-Hispanic Caucasians, but their low rate of fracture is thought to be due to relatively larger bone size [84]. Despite the widely accepted axiom that Asians have lower bone mass than non-Hispanic Whites, a recent comparison of closely matched non-Hispanic White and Chinese women found slightly higher bone mass in the Chinese when height and weight (and theoretically differences in bone size) were controlled [105]. Data suggest that differences in bone accumulation in multiethnic teenagers may predominantly reflect bone size [70,106]. Body weight factors importantly in the maintenance of bone density, and thinness is an important risk factor for hip fracture in Black, Caucasian, and Asian women [107 – 109]. However, it appears that both fat and lean body mass contribute to preservation of the skeleton [110,111], perhaps due in part to peripheral aromatization of androgen to estrogen that occurs in adipose tissue and skeletal muscle [101,112]. Serum estrone concentrations relate positively to degree of obesity, and bone mass correlates positively with body weight in both Black and Caucasian women [113]. However, differences in body weight do not explain the differences in bone mass between Blacks and Caucasians [114]. Epidemiologic studies indicate that African American women are classified as overweight twice as frequently as non-Hispanic Caucasian women [115], but that differences in body mass index (BMI) do not develop until after adolescence (interestingly, that is about the time differences in bone mass become apparent as well). Obesity classification typically depends on self-report of weight and height, with

576 subsequent computation of BMI. Obesity is then defined as BMI  27.3 kg/m2 [116]. This definition is based on statistics from the second National Health and Nutrition Examination Survey (NHANES II), using the 85th percentile of BMI for 20 to 29-year-old non-Hispanic Caucasians as an obesity cutoff point. Application of this definition for obesity has not been race or ethnicity adjusted in the majority of studies. Lopez and Masse [117] compared anthropometric data from NHANES II to that acquired during the 1982 – 1984 Hispanic Health and Nutrition Examination Survey (HHANES). They found that use of the NHANES II obesity cutoff consistently labeled 12 – 14% more Puerto Rican and Mexican American women as obese than if ethnicity-specific cutoff data from HHANES were used [117]. Cuban American women’s BMIs reflected those of NHANES II data, rendering the use of a “Hispanic”-specific cutoff meaningless because there are differences in height and weight distribution among the Hispanic ethnic groups [118]. The optimal formula for a body mass index, which is conventionally expressed as weight/height2, varies among populations. If the purpose of using a weight for height index is to minimize the effect of height on body mass, then one approach to identifying an appropriate index is to find the exponent for the denominator that minimizes the correlation between weight and height (such that correlation r  0) in a given population [119]. Kleerekoper et al. [120] applied this approach to 201 White and 77 African American postmenopausal women participating in a longitudinal study of bone mass and biochemical markers of bone remodeling. The African American women had a significantly greater BMI based on the conventional formula. However, the formula that best provided a height-free measure of weight was different for the White [weight (kg)/height(m)1.17] and the African American women [weight(kg)/height(m)1.30]. When BMI was calculated using these population-specific formulae, there was no significant difference between the two groups, suggesting that these African American women were not more “obese.” This underscores the need to consider population-specific approaches to studying human biologic phenomena. Although BMI is often used to estimate degree of obesity, body composition is much more complex. The simplest model uses bone, fat, and lean body mass as the three major components. Because bone and lean body mass are closely related [121], it stands to reason that groups with increased muscle mass, such as African Americans and Polynesians [97,122], have higher bone mass. Therefore, total body weight will be greater in those with higher bone mass, not necessarily because of obesity but also due to the contribution of bone and muscle weight. Body composition comparisons made between African American and Caucasian adults demonstrate consistently greater muscle and

VILLA, NELSON, AND NELSON

bone mass in the former [123 – 125], underscoring the importance of using race- and ethnicity-specific reference populations when interpreting body habitus data.

D. Calcium Nutriture Calcium intake affects attainment of peak bone mass as well as the ability to preserve skeletal calcium throughout life [126]. In an early study, Matkovic and colleagues demonstrated that hip fracture rates differed significantly within an ethnic group living in two regions of Croatia with divergent levels of dietary calcium intake [127] and concluded that these differences were due to differences in attainment of peak bone mass. Gradations in bone mass related to calcium intake are also observed in other racial and ethnic groups. In an excellent dietary study of Chinese women with similar ethnic backgrounds, Hu and collaborators demonstrated a wide range in BMD depending on dietary calcium intake [128]. In this group, although the women with higher calcium intake had higher BMD, the rate of bone loss with age was not affected by dietary calcium, supporting the hypothesis that the differences in bone mass observed in older women were realized earlier in life. Calcium nutriture may contribute to differences in the bone mass of Japanese and Japanese American groups as well [129,130], and calcium supplementation has been shown to reduce bone loss in elderly Chinese women [131]. 1. LACTOSE INTOLERANCE Dairy products serve as a major source of dietary calcium in many places. Because the majority of the world’s non-Caucasian population develops lactase deficiency relatively early in life, it would seem that impaired dairy tolerance should lead to low calcium intake and therefore suboptimal development and preservation of bone. Many investigators cite this rationale when attempting to explain ethnic differences in calcium and vitamin D metabolism. However, the presence of lactose malabsorption does not predict milk consumption in Mexican Americans, Blacks, or the elderly [132,133], and lactase deficiency does not cause adult bone loss [134]. In fact, milk is the primary source of calcium in the diets of three Hispanic ethnic groups living in North America [135]. When analyzing NHANES and NHANES II data, Looker and co-workers [135] found that calcium intakes in Hispanic diets paralleled those of non-Hispanic Caucasians and were somewhat higher than those of African Americans. Striking differences existed in the dietary sources of calcium for the three Hispanic groups, although total calcium intakes did not vary significantly. Milk was the single greatest contributor for all three, but corn tortillas were

CHAPTER 22 Race, Ethnicity, and Osteoporosis

second in importance for Mexican Americans alone. A listing of the top 10 contributors to dietary calcium for Mexican Americans also included flour tortillas and pinto beans, whereas for Cuban Americans and Puerto Ricans, pizza and rice were major sources. The bioavailability of calcium may vary widely among these diverse foods. For some, such as pinto beans, calcium absorbability is less than that of milk [136], but the total intake of calcium from all sources may be sufficient to exceed gastrointestinal and renal losses. Questionnaire assessment of dietary calcium intake should therefore be tailored to include ethnic foods in order to collect representative data. 2. CALCIUM METABOLISM Racial or ethnic differences in the absorption and excretion of calcium may affect overall calcium balance. One study conducted under conditions of severe calcium restriction found African Americans to have low vitamin D levels and compensatory hypersecretion of parathyroid hormone, theoretically maximizing urinary retention of calcium and thereby contributing to greater bone mass [96]. Another study that provided adequate dietary calcium found no evidence of a racial alteration in the vitamin D endocrine system [137]. However, despite lack of racial differences in dietary calcium and vitamin D intake, it was found that Blacks had significantly lower urinary calcium excretion and that calcium excretion was related inversely to radial BMD. Studies comparing postmenopausal African American and White women reported statistically significant differences in calciotropic hormones and biochemical markers of bone remodeling. PTH and 1,25 vitamin D concentrations tend to be higher in African Americans, and 25 hydroxyvitamin D(25OHD), osteocalcin, hydroxyproline, and bonespecific alkaline phosphatase values lower in AfricanAmerican compared to White women (p0.05), suggesting possible resistance to the skeletal actions of PTH in African-Americans [60,138]. Skeletal attributes develop early in life, as mentioned previously, so Abrams and colleagues investigated aspects of calcium metabolism in Mexican American and White children. They found higher PTH concentrations in Mexican American girls despite lack of vitamin D deficiency, although ethnic differences in 25OHD and PTH concentrations did not significantly affect calcium absorption, excretion, or bone calcium kinetics [139].

E. Vitamin D Exposure Hypovitaminosis D, when present in non-Hispanic Caucasian populations, predicts low bone mass and increased risk for hip fractures [140,141]. A similar relationship is

577 seen among inhabitants of Hong Kong, where hypovitaminosis D appears as a common problem in elders with hip fracture [142]. In this group, subclinical vitamin D deficiency is also associated with muscle weakness and increased risk of falling. Studies conducted in Japan note a marked beneficial effect on BMD and spinal fracture rate in patients treated with vitamin D [143]. It appears, therefore, that individuals from divergent racial and ethnic backgrounds respond similarly to the influence of circulating 25OHD. Regional differences in vitamin D status exist, perhaps due to differences in levels of solar radiation, individuals’ exposure to sunlight, and skin color. Differences in skin pigmentation reflect an evolutionary adaptation to solar radiation, according to one theory, so that those living closer to the equator would have greater amounts of skin pigment in order to reduce the risk of hypervitaminosis D [144]. Conversely, those living in higher latitudes would have fairer skin in order to make sufficient amounts of vitamin D. Data suggest that there is a difference in the gradient of skin color south of the equator compared with north of the equator [145]; this would suggest that factors other than solar radiation affect skin pigmentation. Some investigators find that individuals with high and low skin pigmentation possess similar 25OHD synthetic abilities in response to ultraviolet exposure [146]. Others imply that increased pigment reduces the capacity of skin to synthesize vitamin D [96,98,147]. Data presented in abstract form found a high prevalence of inadequate vitamin D nutriture in a group of elderly Mexican American women, but this was related to vitamin D exposure rather than skin pigmentation [148]. In a multiracial/ethnic study of the relationship between skin pigment and cutaneous synthesis of vitamin D, Matsuoka and colleagues [49] found that increased skin pigmentation had a photoprotective effect but did not impair adequate formation of vitamin D. Comparison of Blacks and Caucasians living in Zaire or Belgium found no racial differences when the study was conducted in Zaire. Evaluation of respective groups in Belgium, however, demonstrated lower vitamin D values in Blacks, with an inverse relationship noted between serum 25OHD and length of stay in Belgium [150]. Blacks living in the United States showed lower concentrations of 25OHD and higher levels of 1,25-dihydroxyvitamin D when compared with Whites in the same geographic area, but without discernible effects on levels of calciotropic hormones or renal calcium excretion [151]. It is possible that at northern latitudes, avoidance of lifestyle habits such as sun-seeking behavior and intake of vitamin D-fortified foods will result in lower levels of circulating 25OHD. This could manifest itself in the protective response of mild secondary hyperparathyroidism seen by some investigators [152].

578

VILLA, NELSON, AND NELSON

F. Vitamin D Receptor Gene In 1992, Morrison et al. [153] reported that allelic variation in several polymorphisms at the vitamin D receptor (VDR) gene locus could be used to predict bone turnover and later reported an association with bone mass in a large group of white Australian women [154]. Attempts to corroborate these findings in other populations have yielded variable results, as well as investigations of other restriction fragment length polymorphisms (RFLPs). Investigations of the VDR gene in U.S. populations mainly focus on white women, but occasionally include other ethnic groups, such as African Americans [155 – 158] and Mexican Americans [159]. Results of these are somewhat contradictory, as discussed later, which may reflect ethnic and/or environmental differences in the genotype frequencies as well as on the expression of the VDR. Outside the United States, data from a study of Japanese women suggest an association between VDR gene polymorphisms and both BMD [160] and the rate of postmenopausal bone loss [161]. Studies of VDR in Chinese [162,163] and Korean [164] women do not find an association with bone mass, although the study groups may have had low dietary calcium intake, which could have independently affected attainment of peak bone mass. A significant correlation of the BsmI VDR and BMD [165] was observed among premenopausal Brazilian women living in Sao Paulo, a population characterized by a high degree of miscegenation. Fleet et al. [156] reported no significant ethnic difference in genotype distribution in adult White and African American women. They noted no significant interaction of ethnicity and genotype on BMD of the femoral neck and lumbar spine, although a significant relationship between the genotypes and bone density existed in the group as a whole. Using a start codon polymorphism detected with the endonuclease FokI, Harris et al., found Black/White differences in its distribution among premenopausal women [157]. They suggested that this polymorphism may influence peak bone density and that ethnic differences in genotype frequencies may explain some ethnic differences in femoral neck BMD. Nelson et al. [158] drew a similar conclusion using the BsmI polymorphism, reporting a significant difference in genotype distribution between premenopausal African American and White women, as well as a significantly higher mean whole body bone mass in the high bone mass (bb) genotype in the groups combined. It is notable that the low bone mass genotype (BB) was absent among the African American women. Their data suggested that ethnic differences in the distribution of the BsmI genotypes may help explain observed ethnic difference in whole body bone mass in younger adult women. Zmuda et al. [155,166] investigated four VDR gene polymorphisms (BsmI, ApaI, FokI, and TaqI), bone

turnover, and rates of bone loss in older African American women. They did not find an association between VDR gene polymorphisms and BMD or indices of bone turnover in this group. McClure et al. [159] studied three RFLPs (BsmI, ApaI, and TaqI) for VDR in postmenopausal Mexican American women and did not find significant associations with BMD, but there were trends suggesting that a larger sample size may reveal such associations. It may also be that lifelong habits and exposures (such as vitamin D and calcium intake) muddy the relationship between VDR and BMD when studied in older adults. Interestingly, Sainz et al. conducted a study looking at BsmI, ApaI, and TaqI VDR polymorphisms in prepubertal Mexican American girls, finding a strong relationship between VDR and both femoral and vertebral bone density.

G. Physical Activity and Other Lifestyle Factors Physical activity benefits the skeleton throughout all stages of life (physical activity and risk for osteoporosis are discussed elsewhere in this publication) and is a lifestyle habit subject to great variation among different groups. Despite varying cultural values and attitudes toward physical activity throughout life, its beneficial effect on bone mass appears ubiquitous. Ethnic differences in bone mass and fracture risk related to physical activity habits are seen in studies from Sweden [10], Japan [168], Mexico [169], and Hawaii [170]. Although multicultural studies concerning type of physical activity and accretion of bone density are lacking, it would be interesting to see whether culturally determined activities during adolescence affect lifetime risk for fracture. It could be that habitual activities or work during youth affect adult bone mass and possibly skeletal structure. Physical activity and muscle strength also affect the risk of falling, another contributor to overall fracture risk. Tobacco use contributes to risk of osteoporotic fracture (discussed in greater detail elsewhere in this volume). Smoking is thought to reduce calcium absorption [171], and heavy smokers appear to have lower BMD than light smokers or nonsmokers [172]. Elderly Japanese American men who smoke have significantly faster bone loss rates than those who do not [173,174], but the relationship among smoking, bone mass, race, and ethnicity has not been widely studied. Alcohol intake contributes to risk for low bone density [175], but its relationship to fracture is not well established in multiethnic populations. Some data suggest that noninsulin dependent diabetes mellitus is positively associated with bone mineral density in women [176], and a positive association between serum glucose concentrations and bone mass in Mexican American women has been described [12]. Mexican Americans and Blacks experience a proportionately greater amount of diabetes than non-Hispanic Caucasians [2,177]. This may

579

CHAPTER 22 Race, Ethnicity, and Osteoporosis

contribute to differences in bone mineral density or rates of bone loss.

H. Falls Falling down contributes greatly to risk of suffering a hip fracture. Few studies of racial or ethnic differences in falls incidence exist. In a study of risk factors for hip fracture in African American women, Grisso and colleagues found that variables associated with risk for falling (previous stroke, use of ambulatory aids, lower-limb dysfunction, history of seizures) also predicted hip fracture risk [107]. A study of postmenopausal Mexican American Caucasians [178] found the falls rate and risk factors contributing to falls to be comparable to that reported for non-Hispanic Caucasians [179]. Rate of falls in Japanese Americans living in Hawaii appear to be lower than that published for Caucasians [180], but that risk for injury following a fall to be no different.

I. Bone Geometry Bone mineral density predicts fracture risk, but there is considerable overlap of BMD in non Hispanic Caucasian controls and hip fracture cases. Also, as noted previously, there is considerable variation in fracture risk among ethnic and racial groups with similar BMD values. Thus, factors other than bone mass also affect risk for osteoporotic fracture. Analysis of bone densitometry data collected in a large study of osteoporotic fracture suggests that a simple geometric measurement of femoral size, hip axis length (HAL), is related to hip fracture risk [181]. In this study, shorter HAL was associated with a decreased risk of hip fracture. Of the three racial groups represented in this study, African Americans and Asian Americans had a significantly lower HAL than the fracture group (which was predominantly non-Hispanic Caucasian) [182]. Additionally, the HAL of Mexican American Caucasian women, another ethnic group with a relatively low risk for hip fracture (see Table 1), averages about the same as for Asian and African Americans [12]. Radiographic studies of hip geometry have shown crossnational as well as ethnic differences that may relate to hip fracture risk. One study comparing Japanese and Caucasian differences in geometric properties of the femoral neck demonstrated an association between low fracture risk and short femoral neck [183], whereas a study of African Americans and U.S. Whites found significant ethnic differences in various measurements of hip geometry [184]. Other potentially important geometric variables have been assessed in the proximal femur using DXA data to describe cross-sectional geometry, using the method of “hip structure analysis” developed by Beck et al. [185].

Measurements include bone width (subperiosteal diameter), cortical thickness, cross-sectional bone area, cross-sectional moment of inertia, and section modulus, which contribute to the biomechanical strength of the hip. An investigation of data from NHANES III showed significant sex and ethnic differences in many of these variables [186]. Nelson et al. [187] used this method when analyzing data from a group of postmenopausal African American and white women, showing that the spatial distribution of bone in the femoral neck is arranged to resist greater loading in the African American women. A comparison of these results with hip structure analysis of data from Black and White postmenopausal women in Johannesburg showed that both U.S. ethnic groups have significantly greater indices than their South. African counterparts, although the Black women in both countries have a higher section modulus — an index of bending strength — in the femoral neck compared with White women (unpublished data). Thus it is clear that factors other than bone mass may be important in determining the biomechanical strength of the hip and that these may differ across ethnic groups. The same cautionary rule applies to assessment of HAL in determining risk for fracture: its application to risk may vary with ethnicity. In a comparison of bone geometric properties as risk factors for hip fracture in European, Chinese, Indian, and Polynesian premenopausal women, Chin and colleagues [188] found shorter HAL in the Chinese and Indian groups, but longer HAL in European and Polynesian women. Because Polynesian women enjoy a very low rate of hip fracture, osteoporotic risk factors other than HAL must be considered.

V. SUMMARY The international range in hip fracture incidence confirms the widely held notion that many factors enter into determination of skeletal health. Much information may be gleaned from interracial and interethnic studies that may help elucidate possible etiologies in the pathogenesis of osteoporosis. In order to best delineate these factors, investigators must explore contributions from the environmental and cultural milieu in which different groups of people reside. Description of study groups would ideally describe the criteria on which categories such as race and ethnicity were determined and would include a discussion of the degree of acculturation when appropriate. Description of study groups would ideally include race, ethnicity, and acculturation. Bone mass, in and of itself, is not the best predictor of fracture risk in all groups. Variables such as bone geometry, reproductive history, physical activity, dietary exposures, body composition, and others all contribute to fracture risk, reflecting the rich diversity peculiar to the human race.

580

VILLA, NELSON, AND NELSON

References 1. M. L. Villa, Cultural determinants of skeletal health: The need to consider both race and ethnicity in bone research. J. Bone Miner. Res. 9(9), 1329 – 1332 (1994). 2. C. L. Hanis, R. E. Ferrell, S. A. Barton, L. Aguilar, A. Garza-Ibarra, B. R. Tulloch, C. A. Garcia, and W. J. Schull, Diabetes among Mexican Americans in Starr County, Texas. Am. J. Epidemiol. 118(5), 659 – 672 (1983). 3. T. Sugimoto, M. Tsutsumi, Y. Fujii, M. Kawakatsu, H. Negishi, M. C. Lee, K. S. Tsai, M. Fukase, and T. Fujita, Comparison of bone mineral content among Japanese, Koreans, and Taiwanese assessed by dual-photon absorptiometry. J. Bone Miner. Res. 7(2), 153 – 159 (1992). 4. D. A. Nelson and D. A. Barondess, Whole body bone, fat and lean mass in children: Comparison of three ethnic groups. Am. J. Phys. Anthropol. 103(2), 157 – 62 (1997). 5. J. Anderson, M. Moeschberger, M. J. Chen, P. Kunn, M. E. Wewers, and R. Guthrie, An acculturation scale for Southeast Asians. Soc. Psychiatry Psychiatr. Epidemiol. 28(3), 134 – 141 (1993). 6. I. Cuellar, L. C. Harris, and R. Jasso, An acculturation scale for Mexican-American normal and clinical populations. Hisp. J. Behav. Sci. 2, 199 – 217 (1980). 7. H. P. Hazuda, M. P. Stern, and S. M. Haffner, Acculturation and assimilation among Mexican Americans: scales and population-based data. Soc. Sci. Quart. 69(3), 687 – 706 (1988). 8. N. W. Jerome, Diet and acculturation. In “Nutritional Anthropology: Contemporary Approaches to Diet and Culture” (N. W. Jerome, R. F. Kandel, and G. H. Pelto, eds.), pp 275 – 325. Redgrave Publishing Company, (1980). 9. H. P. Hazuda, S. M. Haffner, M. P. Stern, and C. W. Eifler, Effects of acculturation and socioeconomic status on obesity and diabetes in Mexican Americans: The San Antonio Heart Study. Am. J. Epidemiol. 128(6), 1289 – 1301 (1988). 10. B. Jónsson, P. Gärdsell, O. Johnell, I. Sernbo, and B. Gullberg, Lifestyle and different fracture prevalence: a cross-sectional populationbased study. Calcif. Tissue Int. 52, 425 – 433 (1993). 11. P. D. Ross, H. Norimatsu, J. W. Davis, K. Yano, R. D. Wasnich, S. Fujiwara, Y. Hosoda, and L. Melton, A comparison of hip fracture incidence among native Japanese, Japanese Americans, and American Caucasians. Am. J. Epidemiol. 133(8), 801 – 809 (1991). 12. M. L. Villa, R. Marcus, R. R. Delay, and J. L. Kelsey, Factors contributing to skeletal health of postmenopausal Mexican-American women. J. Bone Miner. Res. 10(8), 1233 – 1242 (1995). 13. S. Maggi, J. L. Kelsey, J. Litvak, and S. P. Heyse, Incidence of hip fractures in the elderly: A cross-national analysis. Osteopor. Int. 1(4), 232 – 241 (1991). 14. M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public Health 74 (12), 1374 – 1380 (1984). 15. J. A. Falch, A. Ilebekk, and U. Slungaard, Epidemiology of hip fractures in Norway. Acta Orthop. Scand. 56, 12 – 16 (1985). 16. J. A. Baron, J. Barrett, D. Malenka, E. Fisher, W. Kniffin, T. Bubolz, and T. Tosteson, Racial differences in fracture risk. Epidemiology 5, 42 – 47 (1994). 17. E. S. Fisher, J. A. Baron, D. J. Malenka, J. A. Barrett, W. D. Kniffin, F. S. Whaley, and T. A. Bubolz, Hip fracture incidence and mortality in New England. Epidemiology 2, 116 – 122 (1991). 18. R. Y. Hinton and G. S. Smith, The association of age, race, and sex with the location of proximal femoral fractures in the elderly. J. Bone Joint Surg. Am. 75(5), 752 – 759 (1993). 19. S. L. Silverman and R. E. Madison, Decreased incidence of hip fracture in Hispanics, Asians, and blacks: California Hospital Discharge Data. Am. J. Public Health 78(11), 1482 – 1483 (1988).

20. R. L. Bauer, Ethnic differences in hip fracture: A reduced incidence in Mexican Americans. Am. J. Epidemiol. 127(1), 145 – 149 (1988). 21. L. Solomon, Osteoporosis and fracture of the femoral neck in the South African Bantu. J. Bone Jt. Surg. 50B, 2 – 11 (1968). 22. H. Hagino, K. Yamamoto, H. Ohshiro, T. Nakamura, H. Kishimoto, and T. Nose, Changing incidence of hip, distal radius, and proximal humerus fractures in Tottori Prefecture, Japan. Bone 24, 265 – 270 (1999). 23. H. Hagino, K. Yamamoto, R. Teshima, H. Kishimoto, K. Kuranobu, and T. Nakamura, The incidence of fractures of the proximal femur and the distal radius in Tottori prefecture, Japan. Arch. Orthop. Trauma Surg. 109(1), 43 – 44 (1990). 24. X. Ling, L. Aimin, C. Xiaoshu, and S. R. Cummings, Very low rates of hip fracture in Beijing, People’s Republic of China: The Beijing Osteoporosis Project. Am. J. Epidemiol. 144, 901 – 907 (1996). 25. A. V. Schwartz, J. L. Kelsey, S. Maggi, M. Tuttleman, S. C. Ho, P. V. Jonsson, G. Poor, J. A. Sisson de Castro, L. Xu, C. C. Matkin, L. M. Nelson, and S. P. Heyse, International variation in the incidence of hip fractures: Cross-national project on osteoporosis for the World Health Organization Program for Research on Aging. Osteopor. Int. 9, 242 – 253 (1999). 26. E. M. C. Lau, C. Cooper, C. Wickham, S. Donnan, and D. J. P. Barker, Hip fracture in Hong Kong and Britain. Int. J. Epidemiol 19, 1119 – 1121 (1990). 27. J. Chalmers and K. C. Ho, Geographical variations in senile osteoporosis: The association with physical activity. J. Bone Jt. Surg. 52B, 667 – 675 (1970). 28. P. C.N. Wong, Fracture epidemiology in a mixed southeastern Asian community (Singapore). Clin. Orthop. 45, 55 – 61 (1966). 29. A. Memon, W. M. Pospula, A. Y. Tantawy, S. Abdul-Ghafar, A. Suresh, and A. Al Rowaih, Incidence of hip fracture in Kuwait. Int. J. Epidemiol. 27, 860 – 865 (1998). 30. S. Stott, D. H. Gray, and W. Stevenson, The incidence of femoral neck fractures in New Zealand. New Zeal. Med. J. 91, 6 – 9 (1980). 31. R. Nilsson, O. Lofman, K. Berglund, L. Larson, and G. Toss, Increased hip-fracture incidence in the county of Ostergotland, Sweden, 1940-1986, with forecasts up to the year 2000: An epidemiological study. Int. J. Epidemiol. 20, 1018 – 1024 (1991). 32. P. Luthje, Incidence of hip fracture in Finland: A forecast for (1990). Acta Orthop. Scand. 56, 223 – 225 (1985). 33. B. S. Z. Elabdien, S. Olerud, G. Karlstrom, and B. Smedby, Rising incidence of hip fracture in Uppsala, 1965 – 1980. Acta Orthop. Scand. 55, 284 – 289 (1984). 34. R. Hedlund, U. Lindgren, and A. Ahlbom, Age- and sex-specific incidence of femoral neck and trochanteric fractures. Clin. Orthop. Rel. Res. 222, 132 – 139 (1987). 35. R. Hedlund, A. Ahlbom, and U. Lindgren, Hip fracture incidence in Stockholm 1972 – 1981. Acta Orthop. Scand. 57, 30 – 34 (1985). 36. P. A. Alffram, An epidemiologic study of cervical and trochanteric fractures of the femur in an urban population. Acta Orthop. Scand. Suppl. 65, 101 – 109 (1964). 37. V. Finsen and P. Benum, Changing incidence of hip fractures in rural and urban areas of central Norway. Clin. Orthop. Rel. Res. 218, 104 – 110 (1987). 38. E. M. Alhava and J. Puittinen, Fractures of the upper end of the femur as an index of senile osteoporosis in Finland. Ann. Clin. Res. 5, 398 – 403 (1973). 39. A. L. Currie, D. M. Reid, and N. Brown, An epidemiological study of fracture of the neck of femur. Health Bull. 44, 143 – 148 (1986). 40. W. J. Boyce, and P. M. Vessey, Rising incidence of fracture of the proximal femur. Lancet 1, 150 – 151 (1985). 41. M. R. Baker, An investigation into secular trends in the incidence of femoral neck fracture using hospital activity analysis. Public Health 94, 368 – 374 (1980).

CHAPTER 22 Race, Ethnicity, and Osteoporosis 42. A. Lizaur-Utrilla, P. Orts, F. Sanchez Del Campo, A. A. Barrio, and P. Gutierrez-Carbonell, Epidemiology of trochanteric fractures of the femur in Alicante, Spain, 1974 – 1982. Clinic. Orthop. Rel. Res. 218, 24 – 31 (1987). 43. G. F. Mazzuoli, C. Gennari, M. Passeri, F. S. Celi, M. Acca, A. Camporeale, G. Pioli, and M. Pedrazzoni, Incidence of hip fracture: An Italian survey. Osteopor. Int. 3 (Suppl 1), 8 – 9 (1993). 44. S. Levine, M. Makin, J. Menczel, G. Robin, E. Naor, and R. Steinbert, Incidence of fractures of the proximal end of the femur in Jerusalem: A study of ethnic factors. J. Bone Jt. Surg. 52, 1193 – 1202 (1970). 45. W. A. Ray, M. R. Griffin, R. West, L. Strand, and L. J. Melton, Incidence of hip fracture in Saskatchewan, Canada, 1976 – 1985. Am. J. Epidemiol. 131, 502 – 509 (1990). 46. J. G. RodrÌguez, R. W. Sattin, and R. J. Waxweiler, Incidence of hip fractures, United States, 1970 – 83. Am. J. Prev. Med. 5(3), 175 – 181 (1989). 47. J. C. Gallagher, L. J. Melton, B. L. Riggs, and E. Bergstrath, Epidemiology of fractures of the proximal femur in Rochester, Minnesota. Clin. Orthop. Rel. Res. 150, 163 – 171 (1980). 48. K. M. Sanders, E. Seeman, A. M. Ugoni, J. A. Pasco, T. J. Martin, B. Skoric, G. C. Nicholson, and M. A. Kotowicz, Age- and genderspecific rate of fractures in Australia: A population-based study. Osteopor. Int. 10, 240 – 247 (1999). 49. W. E. Bacon, S. Maggi, A. Looker, T. Harris, C. R. Nair, J. Giaconi, R. Honkanen, S. C. Ho, K. A. Peffers, O. Torring, R. Gass, and N. Gonzalez, International comparison of hip fracture rates in 1988 – 89. Osteopor. Int. 6, 69 – 75 (1996). 50. L. J. Melton, Epidemiology of fractures. In “Osteoporosis: Etiology, Diagnosis and Management” (B. L. Riggs, L. J. D. Melton, eds.), pp. 111 – 154. Raven Press, New York, (1988). 51. D. S. Lauderdale, S. J. Jacobsen, S. E. Furner, P. S. Levy, J. A. Brody, and J. Goldberg, Hip fracture incidence among elderly Hispanics. Am. J. Public Health 88(8), 1245 – 1247 (1998). 52. D. S. Lauderdale, S. J. Jacobsen, S. E. Furner, P. S. Levy, J. A. Brody, and J. Goldberg, Hip fracture incidence among elderly Asian-American populations. Am. J. Epidemiol. 146(6), 502 – 509 (1997). 53. L. Solomon, Bone density in ageing Caucasian and African populations. Lancet 2, 1326 – 1330 (1979). 54. M. M. Luckey, D. E. Meier, J. P. Mandeli, M. C. DaCosta, M. L. Hubbard, and S. J. Goldsmith, Radial and vertebral bone density in white and black women: Evidence for racial differences in premenopausal bone homeostasis. J. Clin. Endocrinol. Metab. 69(4), 762 – 770 (1989). 55. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral density in the Postmenopausal Estrogen/Progestin Interventions Trial. J. Bone Miner. Res. 9(9), 1467 – 1476 (1994). 56. Y. Liel, J. Edwards, J. Shary, K. M. Spicer, L. Gordon, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip, and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66(6), 1247 – 1250 (1988). 57. J. F. Aloia, A. Vaswani, M. Mikhail, M. Badshah, and E. Flaster, Cancellous bone of the spine is greater in black women. Calcif. Tissue Int. 65(1), 29 – 33 (1999). 58. A. C. Looker, H. W. Wahner, W. L. Dunn, M. S. Calvo, T. B. Harris, S. P. Heyse, C. C. Johnston, Jr., and R. L. Lindsay, Proximal femur bone mineral levels of US adults. Osteopor. Int. 5(5), 389 – 409 (1995). 59. B. Ettinger, S. Sidney, S. R. Cummings, C. Libanati, D. D. Bikle, I. S. Tekawa, K. Tolan, and P. Steiger, Racial differences in bone density between young adult black and white subjects persist after adjustment for anthropometric, lifestyle, and biochemical differences. J. Clin. Endocrinol. Metab. 82(2), 429 – 434 (1997).

581 60. M. Kleerekoper, D. A. Nelson, E. L. Peterson, M. J. Flynn, A. S. Pawluszka, G. Jacobsen, and P. Wilson, Reference data for bone mass, calciotropic hormones, and biochemical markers of bone remodeling in older (55-75) postmenopausal white and black women. J. Bone Miner. Res. 9(8), 1267 – 1276 (1994). 61. J. Y. Li, B. L. Specker, M. L. Ho, and R. C. Tsang, Bone mineral content in black and white children 1 to 6 years of age: Early appearance of race and sex differences. Am. J. Dis. Child. 143(11), 1346 – 1349 (1989). 62. N. H. Bell, J. Shary, J. Stevens, M. Garza, L. Gordon, and J. Edwards, Demonstration that bone mass is greater in black than in white children. J. Bone Miner. Res. 6(7), 719 – 723 (1991). 63. D. P. McCormick, S. W. Ponder, H. D. Fawcett, and J. L. Palmer, Spinal bone mineral density in 335 normal and obese children and adolescents: Evidence for ethnic and sex differences. J. Bone Miner. Res. 6(5), 507 – 513 (1991). 64. D. Laraque, L. Arena, J. Karp, and D. Gruskay, Bone mineral content in black pre-schoolers: Normative data using single photon absorptiometry. Pediatr. Radiol. 20(6), 461 – 463 (1990). 65. S. M. Garn, S. T. Sandusky, J. M. Nagy, and M. B. McCann, Advanced skeleton development in low-income and Negro children. J. Pediatr. 80, 965 – 969 (1972). 66. D. A. Nelson, P. M. Simpson, C. C. Johnson, D. A. Barondess, and M. Kleerekoper, The accumulation of whole body skeletal mass in third- and fourth-grade children: Effects of age, gender, ethnicity, and body composition. Bone 20(1), 73 – 78 (1997). 67. A. Prentice, M. A. Laskey, J. Shaw, T. J. Cole, and D. R. Fraser, Bone mineral content of Gambian and British children aged 0 – 36 months. Bone Miner. 10, 211 – 224 (1990). 68. R. N. Southard, J. D. Morris, J. D. Mahan, J. R. Hayes, M. A. Torch, A. Sommer, and W. B. Zipf, Bone mass in healthy children: Measurement with quantitative DXA. Radiology 179, 735 – 738 (1991). 69. V. Gilsanz, T. F. Roe, S. Mora, G. Costin, and W. G. Goodman, Changes in vertebral bone density in black girls and white girls during childhood and puberty [see comments]. N. Engl. J. Med. 325(23), 1597 – 600 (1991). 70. M. C. Wang, M. Aguirre, G. S. Bhudhikanok, C. G. Kendall, S. Kirsch, R. Marcus, and L. K. Bachrach, Bone mass and hip axis length in healthy Asian, black, Hispanic, and white American youths. J. Bone Miner. Res. 12(11), 1922 – 1935 (1997). 71. L. K. Bachrach, T. Hastie, M. C. Wang, B. Narasimhan, and R. Marcus, Bone mineral acquisition in healthy Asian, Hispanic, black, and Caucasian youth: A longitudinal study. J. Clin. Endocrinol. Metab. 84(12), 4702 – 4712 (1999). 72. D. N. Patel, J. M. Pettifor, P. J. Becker, C. Grieve, and K. Leschner, The effect of ethnic group on appendicular bone mass in children. J. Bone Miner. Res. 7(3), 263 – 272 (1992). 73. E. D. Daniels, J. M. Pettifor, C. M. Schnitzler, S. W. Russell, and D. N. Patel, Ethnic differences in bone density in female South African nurses. J. Bone Miner. Res. 10(3), 359 – 367 (1995). 74. E. D. Daniels, J. M. Pettifor, C. M. Schnitzler, G. P. Moodley, and D. Zachen, Differences in mineral homeostasis, volumetric bone mass and femoral neck axis length in black and white South African women. Osteopor. Int. 7(2), 105 – 112 (1997). 75. T. J. Aspray, A. Prentice, T. J. Cole, Y. Sawo, J. Reeve, and R. M. Francis, Low bone mineral content is common but osteoporotic fractures are rare in elderly rural Gambian women. J. Bone Miner. Res. 11(7), 1019 – 1025 (1996). 76. I. R. Reid, M. Mackie, and H. K. Ibbertson, Bone mineral content in Polynesian and white New Zealand women. Br. Med. J. 292, 1547 – 1548 (1986). 77. B. W. Benson, T. J. Prihoda, and B. J. Glass, Variations in adult cortical bone mass as measured by a panoramic mandibular index. Oral Surg. Oral Med. Oral Pathol. 71(3), 349 – 356 (1991).

582 78. L. Haddock, V. Ortiz, M. D. Vazquez, F. Aguilo, E. Bernard, A. Ayala, and N. Mejias, The lumbar and femoral bone mineral densities in a normal female Puerto Rican population. P. R. Health Sci. J. 15(1), 5 – 11 (1996). 79. M. Russell-Aulet, J. Wang, J. Thornton, E. W. Colt, and R. J. Pierson, Bone mineral density and mass by total-body dual-photon absorptiometry in normal white and Asian men. J. Bone Miner. Res. 6(10), 1109 – 1113 (1991). 80. K. Yano, R. D. Wasnich, J. M. Vogel, and L. K. Heilbrun, Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am. J. Epidemiol. 119(5), 751 – 764 (1984). 81. S. Hagiwara, T. Miki, Y. Nishizawa, H. Ochi, Y. Onoyama, and H. Morii, Quantification of bone mineral content using dual-photon absorptiometry in a normal Japanese population. J. Bone Miner. Res. 4(2), 217 – 222 (1989). 82. R. B. Mazess, Bone mineral in Vilcabamba, Ecuador. Am. J. Roentgenol. 130, 671 – 674 (1978). 83. H. M. Perry III, M. Bernard, M. Horowitz, D. K. Miller, S. Fleming, M. Z. Baker, J. Flaherty, R. Purushothaman, R. Hajjar, F. E. Kaiser, P. Patrick, and J. E. Morley, The effect of aging on bone mineral metabolism and bone mass in Native American women. J. Am. Geriatr. Soc. 46(11), 1418 – 1422 (1998). 84. A. B. Harper, W. S. Laughlin, and R. B. Mazess, Bone mineral content in St Lawrence Island Eskimos. Hum. Biol. 56(1), 63 – 78 (1984). 85. R. B. Mazess and W. Mather, Bone mineral content of North Alaskan Eskimos. Am. J. Clin. Nutr. 27, 916 – 925 (1974). 86. D. McHugh, R. N. Baumgartner, P. M. Stauber, S. Wayne, V. L. Hicks, and V. H. Heyward, Bone mineral in southwest Native American women. Basic Life Sci. 60(87), 87 – 88 (1993). 87. Z. Chen, M. J. Maricic, S. B. Going, T. G. Lohman, B. R. Altimari, and T. L. Bassford, Comparative findings in bone mineral density among postmenopausal Native American women and postmenopausal White women residing in Arizona. Bone 23(5), S592 (1998). 88. J. A. Kanis, L. J. d. Melton, C. Christiansen, C. C. Johnston, and N. Khaltaev, The diagnosis of osteoporosis. J. Bone Miner. Res. 9(8), 1137 – 1141 (1994). 89. D. E. Meier, M. M. Luckey, S. Wallenstein, R. H. Lapinski, and B. Catherwood, Racial differences in pre- and postmenopausal bone homeostasis: Association with bone density. J. Bone Miner. Res. 7(10), 1181 – 1189 (1992). 90. Z. H. Han, S. Palnitkar, D. S. Rao, D. Nelson, and A. M. Parfitt, Effects of ethnicity and age or menopause on the remodeling and turnover of iliac bone: Implications for mechanisms of bone loss [published erratum appears in J. Bone Miner. Res. 14(4), 660 (1999)]. J. Bone Miner. Res. 12(4), 498 – 508 (1997). 91. R. S. Weinstein and N. H. Bell, Diminished rates of bone formation in normal Black adults. N. Engl. J. Med. 319(26), 1698 – 1701 (1988). 92. C. M. Schnitzler, Bone quality: A determinant for certain risk factors for bone fragility. Calcif. Tissue Int. 53 (Suppl 1), S27 – 531 (1993). 93. C. M. Schnitzler, J. M. Pettifor, J. M. Mesquita, M. D. Bird, E. Schnaid, and A. E. Smyth, Histomorphometry of iliac crest bone in 346 normal black and white South African adults. Bone Miner. 10(3), 183 – 199 (1990). 94. B. Dibba, A. Prentice, M. A. Laskey, D. M. Stirling, and T. J. Cole, An investigation of ethnic differences in bone mineral, hip axis length, calcium metabolism and bone turnover between West African and Caucasian adults living in the United Kingdom. Ann. Hum. Biol. 26(3), 229 – 242 (1999). 95. H. Hagino, K. Yamamoto, R. Teshima, H. Kishimoto, and T. Kagawa, Radial bone mineral changes in pre- and postmenopausal healthy Japanese women: Cross-sectional and longitudinal studies. J. Bone Miner. Res. 7(2), 147 – 152 (1992).

VILLA, NELSON, AND NELSON 96. N. H. Bell, A. Greene, S. Epstein, M. J. Oexman, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in Blacks. J. Clin. Invest. 76, 470 – 473 (1985). 97. I. R. Reid, S. Cullen, B. A. Schooler, N. E. Livingston, and M. C. Evans, Calcitropic hormone levels in polynesians: Evidence against their role in interracial differences in bone mass. J. Clin. Endocrinol. Metab. 70(5), 1452 – 1456 (1990). 98. C. Reasner, J. F. Dunn, D. A. Fetchick, Y. Liel, B. W. Hollis, S.Epstein, J. Shary, G. R. Mundy, and N. H. Bell, Alteration of vitamin D metabolism in Mexican-Americans. J. Bone Miner. Res. 5(1), 13 – 17 (1990). 99. J. R. Buchanan, C. A. Myers, T. Lloyd, P. Leuenberger, and L. M. Demers, Determinants of peak trabecular bone density in women: The role of androgens, estrogen, and exercise. J. Bone Miner. Res. 3(6), 673 – 680 (1988). 100. J. S. Finkelstein, R. M. Neer, B. M.K. Biller, J. D. Crawford, and A. Klibanski, Osteopenia in men with a history of delayed puberty. N. Engl. J. Med. 326(9), 600 – 604 (1992). 101. B. J. Davidson, R. K. Ross, A. Paganini-Hill, G. D. Hammond, P. K. Siiteri, and H. L. Judd, Total and free estrogens and androgens in postmenopausal women with hip fractures. J. Clin. Endocrinol. Metab. 54(1), 115 – 120 (1982). 102. K. Kin, K. Kushida, K. Yamazaki, S. Okamoto, and T. Inoue, Bone mineral density of the spine in normal Japanese subjects using dualenergy X-ray absorptiometry: Effect of obesity and menopausal status. Calcif. Tissue Int. 49(2), 101 – 106 (1991). 103. J. A. Cauley, J. P. Gutai, L. H. Kuller, J. Scott, and M. C. Nevitt, Black-white differences in serum sex hormones and bone mineral density. Am. J. Epidemiol. 139(10), 1035 – 1046 (1994). 104. D. K. Katzman, L. K. Bachrach, D. R. Carter, and R. Marcus, Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J. Clin. Endocrinol. Metab. 73(6), 1332 – 1339 (1991). 105. M. Russell-Aulet, J. Wang, J. C. Thornton, E. W. Colt, and R. J. Pierson, Bone mineral density and mass in a cross-sectional study of white and Asian women. J. Bone Miner. Res. 8(5), 575 – 582 (1993). 106. G. S. Bhudhikanok, M. C. Wang, K. Eckert, C. Matkin, R. Marcus, and L. K. Bachrach, Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J. Bone Miner. Res. 11(10), 1545 – 1556 (1996). 107. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, and S. Hoffman, Risk factors for hip fracture in black women. N. Engl. J. Med. 330(22), 1555 – 1559 (1994). 108. E. M. Lau, J. Woo, P. C. Leung, and R. Swaminthan, Low bone mineral density, grip strength and skinfold thickness are important risk factors for hip fracture in Hong Kong Chinese. Osteopor. Int. 3(2), 66–70 (1993). 109. M. E. Pruzansky, M. Turano, M. Luckey, and R. Senie, Low body weight as a risk factor for hip fracture in both black and white women. J. Orthop. Res. 7(2), 192 – 197 (1989). 110. I. R. Reid, L. D. Plank, and M. C. Evans, Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J. Clin. Endocrinol. Metab. 75(3), 779 – 782 (1992). 111. J. F. Aloia, D. M. McGowan, A. N. Vaswani, P. Ross, and S. H. Cohn, Relationship of menopause to skeletal and muscle mass. Am J. Clin. Nutr. 53, 1378 – 1383 (1991). 112. H. Matsumine, K. Hirato, T. Yanaihara, T. Tamada, and M. Yoshida, Aromatization by skeletal muscle. J. Clin. Endocrinol. Metab. 63(3), 717 – 720 (1986). 113. D. P. DeSimone, J. Stevens, J. Edwards, J. Shary, L. Gordon, and N. H. Bell, Influence of body habitus and race on bone mineral density of the midradius, hip, and spine in aging women. J. Bone Miner. Res. 4(6), 827 – 830 (1989). 114. D. A. Nelson, M. Kleerekoper, and A. M. Parfitt, Bone mass, skin color and body size among black and white women. Bone Miner. 4, 257 – 264 (1988).

583

CHAPTER 22 Race, Ethnicity, and Osteoporosis 115. S. Kumanyika, Obesity in Black women. Epidemiol. Rev. 9, 31 – 51 (1987). 116. Anonymous, Anthropometric reference data and prevalence of overweight, United States, 1976-1980. Dept of Health and Human Services, Washington, DC. 117. L. M. Lopez and B. Masse, Comparison of body mass indexes and cutoff points for estimating the prevalence of overweight in Hispanic women. J. Am. Diet. Assoc. 92(11), 1343 – 1347 (1992). 118. L. M. López and B. Masse, Comparison of body mass indexes and cutoff points for estimating the prevalence of overweight in Hispanic women. J. Am. Diet. Assoc. 92(11), 1343 – 1347 (1992). 119. J. Lee, L. Kolonel, and M. Hinds, Relative merits of the weight-corrected-for height indices. Am. J. Clin. Nutr. 34, 2521 – 2529 (1981). 120. M. Kleerekoper, D. Nelson, E. Peterson, P. Wilson, G. Jacobsen, and C. Longcope, Body composition and gonadal steroids in older white and black women. J. Endocrinol. Metab. 79, 775 – 779 (1994). 121. K. J. Ellis and S. H. Cohn, Correlation between skeletal calcium mass and muscle mass in man. J. Appl. Physiol. 38(3), 455 – 460 (1975). 122. S. H. Cohn, C. Abesamis, I. Zanzi, J. F. Aloia, S. Yasumura, and K. J. Ellis, Body elemental composition: Comparison between black and white adults. Am. J. Physio. 232, E419 – 422 (1977). 123. K. D. Côté and W. C. Adams, Effect of bone density on body composition estimates in young adult black and white women. Med. Sci. Sports Exerc. 25(2), 290 – 296 (1993). 124. O. Ortíz, M. Russell, T. L. Daley, R. N. Baumgartner, M. Waki, S. Lichtman, J. Wang, R. J. Pierson, and S. B. Heymsfield, Differences in skeletal muscle and bone mineral mass between black and white females and their relevance to estimates of body composition. Am. J. Clin. Nutr. 55(1), 8 – 13 (1992). 125. K. J. Ellis, S. A. Abrams, and W. W. Wong, Body composition of a young, multiethnic female population. Am. J. Clin. Nutr. 65(3), 724 – 731 (1997). 126. R. P. Heaney, Calcium, bone health and osteoporosis. In “Bone and Mineral Research” (W. A. Peck, ed.) Vol. 4, pp. 255 – 301. Elsevier, Amsterdam. 127. V. Matkovic, K. Kostial, I. Simonovic, R. Buzina, A. Brodarec, and B. E. C. Nordin, Bone status and fracture rates in two regions of Yugoslavia. Am. J. Clin. Nutr. 32, 540 – 549 (1979). 128. J. F. Hu, X. H. Zhao, J. B. Jia, B. Parpia, and T. C. Campbell, Dietary calcium and bone density among middle-aged and elderly women in China. Am. J. Clin. Nutr. 58(2), 219 – 227 (1993). 129. T. Fujita, Y. Okamoto, T. Tomita, Y. Sakagami, K. Ota, and M. Ohata, Calcium metabolism in aging inhabitants of mountain versus seacoast communities in the Kii peninsula. J. Am. Geriatr. Soc. 25, 254 – 258 (1977). 130. K. Yano, L. K. Heilbrun, R. D. Wasnich, J. H. Hankin, and J. M. Vogel, The relationship between diet and bone mineral content of multiple skeletal sites in elderly Japanese-American men and women living in Hawaii. Am. J. Clin. Nutr. 42, 877 – 888 (1985). 131. E. M.C. Lau, J. Woo, P. C. Leung, R. Swaminathan, and D. Leung, The effects of calcium supplementation and exercise on bone density in elderly Chinese women. Osteopor. Int. 2, 168 – 173 (1992) 132. C. E. Woteki, E. Weser, and E. A. Young, Lactose malabsorption in Mexican-American adults. Am. J. Clin Nutr. 30, 470 – 475 (1977). 133. M. H. Rorick and N. S. Scrimshaw, Comparative tolerance of elderly from differing ethnic backgrounds to lactose-containing and lactosefree dairy drinks: A double-blind study. J. Gerontol. 34(2), 191 – 196 (1979). 134. C. W. Slemenda, J. C. Christian, S. Hui, J. Fitzgerald, and C. C. Johnston, No evidence for an effect of lactase deficiency on bone mass in pre- or postmenopausal women. J. Bone Miner. Res. 6(12), 1367 – 1371 (1991). 135. A. C. Looker, C. M. Loria, M. D. Carroll, M. A. McDowell, and C. L. Johnson, Calcium intakes of Mexican Americans, Cubans,

136. 137.

138.

139.

140.

141.

142.

143. 144. 145. 146.

147.

148.

149.

150.

151.

152.

153.

154.

Puerto Ricans, non-Hispanic whites, and non-Hispanic blacks in the United States. J. Am. Diet. Assoc. 93(11), 1274 – 1279 (1993). R. P. Heaney and C. M. Weaver, Plant constituents and food calcium absorbability. J. Bone Miner. Res. 7 (Suppl), S136 (1992). [Abstract] D. E. Meier, M. M. Luckey, S. Wallenstein, T. L. Clemens, E. S. Orwoll, and C. I. Waslien, Calcium, vitamin D, and parathyroid hormone status in young white and black women: Association with racial differences in bone mass. J. Clin. Endocrinol. Metab. 72(3), 703 – 710 (1991). H. M. Perry III, M. Horowitz, J. E. Morley, S. Fleming, J. Jensen, P. Caccione, D. K. Miller, F. E. Kaiser, and M. Sundarum, Aging and bone metabolism in African American and Caucasian women. J. Clin. Endocrinol. Metab. 81(3), 1108 – 1117 (1996). S. A. Abrams, K. C. Copeland, S. K. Gunn, J. E. Stuff, L. L. Clarke, and K. J. Ellis, Calcium absorption and kinetics are similar in 7- and 8-year-old Mexican-American and Caucasian girls despite hormonal differences. J. Nutr. 129(3), 666 – 671 (1999). D. T. Villareal, R. Civitelli, A. Chines, and L. V. Avioli, Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J. Clin. Endocrinol. Metab. 72(3), 628 – 634 (1991). M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327(23), 1637 – 1642 (1992). K. K. Pun, F. H. W. Wong, C. Wang, P. Lau, P. W. M. Ho, W. K. Pun, S. P. Chow, C. L. Cheng, J. C. Y. Leong, and R. T. T. Young, Vitamin D status among patients with fractured neck of femur in Hong Kong. Bone 11, 365 – 368 (1990). T. Fujita, Studies of osteoporosis in Japan. Metabolism 39(4), 39 – 42 (1990). W. F. Loomis, Skin-pigment regulation of vitamin-D biosynthesis in man. Science 157, 501 – 506 (1967). J. Relethford, Hemispheric differences in human skin color. Am. J. Phys. Anthropol. 104, 449 – 457 (1998). W. F. Brazerol, A. J. McPhee, F. Mimouni, B. L. Specker, and R. C. Tsang, Serial ultraviolet B exposure and serum 25-hydroxyvitamin D response in young adult American Blacks and Whites: No racial differences. J. Am. Coll. Nutr. 7(2), 111 – 118 (1988). T. L. Clemens, S. L. Henderson, J. S. Adams, and M. F. Holick, Increased skin pigment reduces the capacity of skin to synthesise vitamin D3. Lancet 1, 74 – 76 (1982). M. Villa, J. Kelsey, J. Chen, and R. Marcus, Skin pigmentation does not affect vitamin D status in community-dwelling Mexican-American women J. Bone Miner. Res. 9 (Suppl 1), S418 (1994). [Abstract] L. Y. Matsuoka, J. Wortsman, J. G. Haddad, P. Kolm, and B. W. Hollis, Racial pigmentation and the cutaneous synthesis of vitamin D. Arch. Dermatol. 127(4), 536 – 538 (1991). J. R. M’Buyamba-Kabangu, R. Fagard, R. Lijnen, R. Bouillon, W. Lissens, and A. Amery, Calcium, vitamin D-endocrine system, and parathryroid hormone in black and white males. Calcif. Tissue Int. 41, 70 – 74 (1987). D. D. Bikle, B. Ettinger, S. Sidney, I. S. Tekawa, and K. Tolan, Differences in calcium metabolism between black and white men and women. Miner. Electrolyte Metab. 25(3), 178 – 184 (1999). H. M. I. Perry, D. K. Miller, J. E. Morley, M. Horowitz, F. E. Kaiser, H. M. J. Perry, J. Jensen, J. Bentley, S. Boyd, and D. Kraenzle, A preliminary report of vitamin D and calcium metabolism in older African Americans. J. Am. Geriatr. Soc. 41(6), 612 – 616 (1993). N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphism and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89(15), 6665 – 6669 (1992). N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles [published erratum

584

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

VILLA, NELSON, AND NELSON appears in Nature 387 (6628),106 (1997)] Nature 367(6460), 284 – 287 (1994). J. M. Zmuda, J. A. Cauley, M. E. Danielson, R. L. Wolf, and R. E. Ferrell, Vitamin D receptor gene polymorphisms, bone turnover, and rates of bone loss in older African-American women. J. Bone Miner. Res. 12(9), 1446 – 1452 (1997). J. C. Fleet, S. S. Harris, R. J. Wood, and B. Dawson-Hughes, The BsmI vitamin D receptor restriction fragment length polymorphism (BB) predicts low bone density in premenopausal black and white women. J. Bone Miner. Res. 10(6), 985 – 990 (1995). S. S. Harris, T. R. Eccleshall, C. Gross, B. Dawson-Hughes, and D. Feldman, The vitamin D receptor start codon polymorphism (FokI) and bone mineral density in premenopausal American black and white women. J. Bone Miner. Res. 12(7), 1043 – 1048 (1997). D. Nelson, P. VandeVord, and P. Wooley, Polymorphism in the vitamin D receptor gene and bone mass in African-American and white mothers and children: A preliminary report. Ann. Rheum Dis. 59, 626 – 630 (2000). L. McClure, T. R. Eccleshall, C. Gross, M. L. Villa, N. Lin, V. Ramaswamy, L. Kohlmeier, J. L. Kelsey, R. Marcus, and D. Feldman, Vitamin D receptor polymorphisms, bone mineral density, and bone metabolism in postmenopausal Mexican-American women. J. Bone Miner. Res. 12(2), 234 – 240 (1997). M. Tamai, M. Yokouchi, S. Komiya, K. Mochizuki, S. Hidaka, S. Narita, A. Inoue, and K. Itoh, Correlation between vitamin D receptor genotypes and bone mineral density in Japanese patients with osteoporosis. Calcif. Tissue Int. 60(3), 229 – 232 (1997). Z. Yamagata, T. Miyamura, S. Iijima, A. Asaka, M. Sasaki, J. Kato, and K. Koizumi, Vitamin D receptor gene polymorphism and bone mineral density in healthy Japanese women. Lancet 344(8928), 1027 (1994). A. W. Kung, S. S. Yeung, and K. S. Lau, Vitamin D receptor gene polymorphisms and peak bone mass in southern Chinese women. Bone 22(4), 389 – 393 (1998). K. S. Tsai, S. H. Hsu, W. C. Cheng, C. K. Chen, P. U. Chieng, and W. H. Pan, Bone mineral density and bone markers in relation to vitamin D receptor gene polymorphisms in Chinese men and women. Bone 19(5), 513 – 518 (1996). S. K. Lim, Y. S. Park, J. M. Park, Y. D. Song, E. J. Lee, K. R. Kim, H. C. Lee, and K. B. Huh, Lack of association between vitamin D receptor genotypes and osteoporosis in Koreans. J. Clin. Endocrinol. Metab. 80(12), 3677 – 3681 (1995). M. Lazaretti-Castro, M. A. Duarte-de-Oliveira, E. M. Russo, and J. G. Vieira, Vitamin D receptor alleles and bone mineral density in a normal premenopausal Brazilian female population. Braz. J. Med. Biol. Res. 30(8), 929 – 932 (1997). J. M. Zmuda, J. A. Cauley, M. E. Danielson, T. M. Theobald, and R. E. Ferrell, Vitamin D receptor translation initiation codon polymorphism and markers of osteoporotic risk in older African-American women. Osteopor. Int. 9(3), 214 – 219 (1999). J. Sainz, J. M. Van Tornout, M. L. Loro, J. Sayre, T. F. Roe, and V. Gilsanz, Vitamin D-receptor gene polymorphisms and bone density in prepubertal American girls of Mexican descent. N. Engl. J. Med. 337(2), 77 – 82 (1997). T. Hirota, M. Nara, M. Ohguri, E. Manago, and K. Hirota, Effect of diet and lifestyle on bone mass in Asian young women. Am. J. Clin. Nutr. 55(6), 1168 – 1173 (1992). S. Parra-Cabrera, M. Hernandez-Avila, J. Tamayo-y-Orozco, L. Lopez-Carrillo, and F. Meneses-Gonzalez, Exercise and reproductive factors as predictors of bone density among osteoporotic women in Mexico City. Calcif. Tissue Int. 59(2), 89 – 94 (1996). P. D. Ross, H. Orimo, R. D. Wasnich, J. M. Vogel, C. J. MacLean, J. W. Davis, and A. Nomura, Methodological issues in comparing genetic and environmental influences on bone mass. Bone Miner. 7, 67 – 77 (1989).

171. E. A. Krall and B. Dawson-Hughes, Smoking and bone loss among postmenopausal women. J. Bone Miner. Res. 6(4), 331 – 338 (1991). 172. C. W. Slemenda, S. L. Hui, C. Longcope, and C. J. Johnston, Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4(5), 737 – 741 (1989). 173. J. W. Davis, P. D. Ross, J. M. Vogel, and R. D. Wasnich, Effects of smoking and alcohol on the rates of bone loss among elderly JapaneseAmerican men. J. Bone Miner. Res. 7(Suppl), S138 (1992). [Abstract] 174. J. M. Vogel, J. W. Davis, A. Nomura, R. D. Wasnich, and P. D. Ross, The effects of smoking on bone mass and the rates of bone loss among elderly Japanese-American men. J. Bone Miner. Res. 12(9), 1495 – 501 (1997). 175. P. Clark, F.de la Pena, F. Gomez Garcia, J. A. Orozco, and P. Tugwell, Risk factors for osteoporotic hip fractures in Mexicans. Arch. Med. Res. 29(3), 253 – 257 (1998). 176. E. Barrett-Connor and T. L. Holbrook, Sex differences in osteoporosis in older adults with noninsulin-dependent diabetes mellitus. JAMA 268(23), 3333 – 3337 (1992). 177. L. S. Lieberman, Diabetes and obesity in elderly black Americans. In “The Black American Elderly” (J. S. Jackson, ed.). Springer, New York, (1988). 178. A. V. Schwartz, M. L. Villa, M. Prill, J. A. Kelsey, J. A. Galinus, R. R. Delay, M. C. Nevitt, D. A. Bloch, R. Marcus, and J. L. Kelsey, Falls in older Mexican-American women. J. Am. Geriatr. Soc. 47(11), 1371 – 1378 (1999). 179. J. L. O’Loughlin, Y. Robitaille, J-F. Boivin, and S. Suissa, Incidence of and risk factors for falls and injurious falls among the communitydwelling elderly. Am. J. Epidemiol. 137(3), 342 – 354 (1993). 180. J. W. Davis, P. D. Ross, M. C. Nevitt, and R. D. Wasnich, Incidence rates of falls among Japanese men and women living in Hawaii. J. Clin. Epidemiol. 50(5), 589 – 594 (1997). 181. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C-C. Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8(10), 1211 – 1217 (1993). 182. S. R. Cummings, J. A. Cauley, L. Palermo, P. D. Ross, R. D. Wasnich, D. Black, and K. G. Faulkner, Racial differences in hip axis length might explain racial differences in rates of hip fracture. Osteopor. Int. 4, 226 – 229 (1994). 183. T. Nakamura, C. H. Turner, T. Yoshikawa, C. W. Slemenda, M. Peacock, D. B. Burr, Y. Mizuno, H. Orimo, Y. Ouchi, and C. C. I. Johnston, Do variations in hip geometry explain differences in hip fracture risk between Japanese and White Americans? J. Bone Miner. Res. 9(7), 1071 – 1076 (1994). 184. T. M. Theobald, J. A. Cauley, C. C. Gluer, C. H. Bunker, F. A. Ukoli, and H. K. Genant, Black – white differences in hip geometry: Study of osteoporotic fractures Research group. Osteopor. Int. 8(1), 61 – 67 (1998). 185. T. Beck, C. B. Ruff, W. W. Scott, C. C. Pluto, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 186. T. J. Beck, A. C. Looker, C. B. Ruff, H. Sievanen, and H. W. Wahner, Structural trends in the aging femoral neck and proximal shaft: Analysis of the Third National Health and Nutrition Examination Survey dual-energy x-ray absorptiomerty data. J. Bone Miner. Res. 15, 2297 – 2304 (2000). 187. D. Nelson, D. Barondess, S. Hendrix, and T. Beck, Cross-sectional geometry and bone mass in the proximal femur in African-American and white postmenopausal women. J. Bone Miner. Res. 188. K. Chin, M. C. Evans, J. Cornish, T. Cundy, and I. R. Reid, Differences in hip axis and femoral neck length in premenopausal women of Polynesian, Asian and European origin. Osteopor. Int. 7(4), 344 – 347 (1997).

CHAPTER 23

Epidemiology of Osteoporotic Fractures in Europe CHRIS DE LAET JONATHAN REEVE

Institute for Medical Technology Assessment, Erasmus University, Rotterdam 3000 DR, The Netherlands Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Cambridge CB2 2QQ, England

I. Introduction II. Hip Fractures III. Vertebral Fractures

IV. General Conclusions References

I. INTRODUCTION

obtained from including subjects and/or populations at extremes of the range of experienced risk, as this tends to improve the power of any such study provided it is properly designed and executed. Partly with this in mind, the Mediterranean Osteoporosis Study (MEDOS) included a section in which 2816 cases of hip fracture were compared with 5369 controls. It was noteworthy that rates of hip fracture in the highest decile of age varied according to investigational center by an order of magnitude, while at the same time rates for men and women were highly correlated within the participating centers. All participating countries in MEDOS, apart from Portugal, have a Mediterranean seaboard [6]. At about the same time, methods were developed for the quantitative assessment of vertebral deformity, which made it possible to envisage large multicenter studies of vertebral fracture. This allowed a substantial group of European investigators from 36 centers in 18 countries to launch the European Vertebral Osteoporosis Study (EVOS) as a prevalence study [7] in population-based, age-stratified

The epidemiology of osteoporosis in Europe has been the subject of increasingly intensive investigational effort since mid-1970s [1]. Earlier work was largely based in single countries, particularly in northern Europe. Work such as that by Fenton-Lewis and colleagues [2] in Britain established that age-specific rates of hip fracture increased substantially between the 1950s and the 1980s and this has been confirmed for other European countries within a sometimes different time frame. It seems likely that a similar increase in age-specific rates of vertebral fracture has occurred in parallel [3]. This contrasts remarkably with the downward secular trend in vertebral fracture rates observed in the studies in Hiroshima, Japan, of Fujiwara [4]. Other work carried out in the 1970s and 1980s showed that there were considerable between-country differences in rates of osteoporotic fracture, which extended to differences between individual European countries [5]. Scientific advantage in cohort and case-control studies is to be

OSTEOPOROSIS, SECOND EDITION VOLUME 1

585

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

586

DE

cohorts of both sexes in the age range of 50 – 80. Most of the original EVOS centers, together with several new centers, have proceeded to follow their cohorts prospectively in the European Prospective Osteoporosis Study (EPOS), which has now been collecting fracture data for over 6 years. This chapter summarizes results of the considerable volume of recent epidemiological work done on hip and vertebral fractures in Europe. Europe is heterogeneous not only genetically but also because of the large variation in lifestyles, often different from those seen in the United States and Canada. However, it has been unclear whether there are true differences in fracture rates between North America and Europe or whether risk factors for fracture differ between continents. This is important, not so much because it may help us understand the etiology of fracture (comparative ecological studies of geographical differences in incidence often do not provide a very efficient platform for identifying true causes of disease). The importance of studying European data for Europeans is because the management of osteoporosis requires knowledge of the burden of disease as it affects the individuals and populations whose health is being cared for. Knowledge of the impact on quality of life of fractures is particularly important with the increasing competition for scarce health care resources. From the public health perspective, identifying risk factors for fracture is important in the near term for devising strategies for identifying individuals at high risk and in the longer term as part of a research strategy aimed at identifying and counteracting the biological pathways leading to osteoporotic fractures with populationbased approaches to prevention. For North Americans and others, there is much to be learned from published European data. This is for several reasons. Not only is the epidemiology of osteoporosis in men quite well advanced in Europe, but also some parts of Europe are given habitually to high rates of compliance with epidemiological studies, increasing confidence that such studies are reasonably representative of the populations from which the study subjects were drawn. Also, by comparing European with non-European studies it should be possible to derive new perspectives on the general validity of the conclusions and recommendations for action derived from both European and non-European studies.

II. HIP FRACTURES A. Current Incidence Currently, most hip fractures occur in Western industrialized countries, and Europe accounts for a large proportion of these. While the total number of hip fractures worldwide in 1990 was estimated at 1.7 million [8], 560,000 of those

LAET AND REEVE

occurred in Europe (including the former Soviet Union) and 360,000 in North America. Lifestyle and living conditions are more diverse in different countries around the European continent than they are in the United States. However, little information is available about the influence of racial differences on the incidence of hip fractures in Europe whether they might act through genetic or environmental mechanisms. Information about hip fractures is available for most countries, as virtually all hip fractures are treated during a hospital admission, making these fractures relatively easy to detect using discharge registries. The age-specific incidence of hip fractures in Europe shows a similar exponential increase with age as observed in other parts of the world, and in general, age-specific incidence is higher in women than in men after the age of 50. This higher incidence in women is reinforced by the fact that there are more elderly women than men who currently only account for approximately a quarter of the total. In 1998 the European Commission reported on the incidence of hip fractures in the 15 member states of the European Union (EU) [9]. Data were compiled from the most recent studies, including the previously mentioned MEDOS study, focusing on countries around the Mediterranean Sea and also from national studies[6,10 – 12]. Those data were then further extrapolated to the European population. Whenever country-specific data were lacking, information from neighboring countries was used [9]. For the year 2000, the total number of hip fractures in those 15 countries was estimated at over 400,000 [9]. Incidence showed a north – south gradient, with the highest incidence in Scandinavia and much lower incidences in the countries around the Mediterranean Sea. The incidence in Sweden was the highest; compared to the incidence of hip fractures in the United States [5], the relative incidence was 1.3 in women and 1.7 in men (own calculations based on the European Commission report). In Finland, by slight contrast, rates are comparable to U.S. rates in women but slightly higher in men. In the United Kingdom, the Netherlands, and Germany, the calculated incidence is very similar to the incidence in the United States. In southern European countries, the incidence was much lower: In France, Greece, and Spain, the incidence was about 70 % of the U.S. incidence, while in Italy and Portugal incidence was as low as 50% of comparable U.S. rates. Figures 1 and 2 (see also color plates) give yearly incidence rates/100,000 for a few selected countries. The MEDOS study also contained data on Turkey, a non-EU country with a highly diverse gene-pool, judged by its history of repeated population mixture in the last two millennia. Turkish data are quite different from data from the remainder of Europe. The low absolute and age-standardized rates approach the levels of some less developed countries and the increase with age was less pronounced than in European countries [13]. Especially in the rural

587

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

FIGURE 1

Age-specific incidence rates of hip fracture per 100,000 women in a few selected countries. (See also color plate.)

(Asia Minor) areas, but even in the principal cities of Istanbul and Ankara, incidence was very low. Incidence rates were about 10 – 20 % of comparable rates in U.S. in women and about 20 – 30% in men.

respectively). Again, Turkish data were strikingly different with a female to male ratio of about 1 in the large cities of Istanbul and Ankara, and a reversed ratio was found in rural areas (around 0.4), especially in older participants.

B. Gender Ratio

C. Future Trends

The global European female to male ratio for the total number of fractures was 3.7 for the year 2000. This high female to male ratio was largely due to demography, as women live longer than men. The female to male ratio of incidence rates was between 1.6 and 2 in most countries. Only in Sweden and Finland was a lower ratio found (1.3 and 1.4,

Future trends are influenced by both demography and trends in age-adjusted incidence. Demography is relatively easy to predict for the near future, all affected persons have already been born; but trends in hip fracture incidence are more difficult to predict and there are conflicting findings from different countries. An increase in the age-adjusted

FIGURE 2 Age-specific incidence rates of hip fracture per 100,000 men in a few selected countries. (See also color plate.)

588

DE

incidence between 1930 and 1980 was described in several countries and although the increase of those rates appeared to have leveled off in the 1980s in parts of the United States [14,15] and also in Sweden and the United Kingdom [2,5, 16], in other European countries, such as the Netherlands and Italy, the age adjusted incidence continued to rise during the beginning of the 1990s [5,17,18]. Because of this uncertainty, most predictions for the future are based on current incidence rates taking only demography into account. If the incidence continues to rise, the picture can only become bleaker. It is thought that the global number of hip fractures worldwide will increase to over 6 million by the year 2050. Although the high proportion of the world’s total of hip fractures that occur in Europe will decline in the next decades due to demographic growth in other parts of the world, the absolute number of hip fractures will also continue to rise in Europe to over 1 million in 2050. As in other areas of the world, the population in Europe is aging, mainly due to the post-World War II baby boom, which was followed by record low fertility rates, most extreme in Spain, Italy, and Germany. Moreover, with the continuing increase in life expectancy in both men and women there is a remorseless increase in the proportion of very elderly people. Over the next decades these trends will continue. Estimates of lifetime risk for hip fracture range from 13 to 18% in northern Europe and the United States, but taking into account expanding life expectancy, these rates are very likely underestimated [19]. The report from the European Commission [9] estimated that while the yearly number of hip fractures would increase to almost 1 million by the year 2050, the global female to male ratio within the European Union for the total number of fractures would decline from 3.7 for the year 2000 to about 3.2 in the year 2050.

D. Risk Indicators Although there is a multitude of known risk indicators for hip fractures, the real occurrence of a fracture remains a chance event and, by definition, chance events cannot be predicted accurately. Only the risk for the occurrence of fractures can be estimated. The list of known risk factors or indicators for hip, as it is for other fractures, is very long [20 – 24]. Broadly speaking, these may be grouped into indicators related to bone fragility and indicators related to the risk of a fall sideways onto the greater trochanter. If these are to be applied to selecting patients at high risk in populations a pragmatic approach should be followed, limiting this list of risk indicators to only those that are most important, both in terms of relative risk and in terms of their prevalence. In the setting of case finding, up until now the normal modus operandi of practicing physicians in this

LAET AND REEVE

field, success depends on an appreciation of the level of absolute risk faced by the individual patient in the consulting room. Thus a rare risk indicator that has a large impact on individual risk may be extremely important to the individual patient who is positive for that indicator. The successful clinician will identify such rare individuals, who may be missed by so-called “screening” of populations and therefore be at risk of not receiving counseling or treatment. If consensus is reached in the future that the screening of older populations is justifiable, the case-finding approach will remain important for this reason. Just as bone density measurements at different skeletal sites are correlated, many risk indicators of frailty are correlated, so that even in case finding one risk indicator can sometimes serve for several. The need in case finding or screening is to summate the overall risk faced by the individual and develop a risk reduction strategy if the absolute risk level is unacceptable. In its 1998 report [25], the American National Osteoporosis Foundation (NOF) proposed the use of a limited set of five risk indicators for the estimation of fracture risk in the setting of a population-based approach to screening women from the seventh decade of age and older. Those indicators were bone mineral density (BMD), a history of a prior fracture after the age of 40, a family history of fracture, “thinness” (which was categorized as the lowest quartile of body weight in U.S. white women), and current cigarette smoking. This selection was based on the large North American Study of Osteoporotic Fractures (SOF) [24] and so far has not been validated elsewhere. It also failed to include any validated indicator of the risk of falling. There is generally an inherent weakness in any such approach, which is that statistical analyses based on individual cohort studies tend to overestimate collectively the predictive power of the risk indicators derived from them. Because risk indicators are chosen purely because of their statistical associations with risk in the given study, each individual association has a certain level of imprecision or uncertainty. Also, in extrapolating the use of risk indicators to populations other than those from which the study population was drawn, differences between populations are likely to affect the relative strengths of different risk indicators. It is of considerable importance therefore to validate in other studies the strengths of the various associations selected, e.g., from the SOF study, for use in screening instruments. This has been addressed in the large Dutch ERGO (Rotterdam) study, in which family history and smoking were found not to be significant risk indicators [26], whereas other indicators, such as the use of a walking aid, were of increased apparent importance. Other groups have proposed alternative hip fracture risk assessment schemes, also based on specific cohorts [27–29]. While age and gender should be a necessary ingredient of any hip fracture prediction score, most such scores also include a measure of BMD and either low body weight or low body mass index (BMI:

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

weight/height2). Some studies, such as the European EPIDOS and Rotterdam Studies, captured the risk of falling by gait speed or by the use of walking aids, respectively, and these could be incorporated directly into a practical scoring instrument. Some scores, such as the NOF instrument, tried to incorporate the component of skeletal fragility, which is independent of bone mineral density: the indicator of this, which is usually preferred, is a prior fragility fracture, but ultrasound BUA has also been suggested for this purpose [27,30] and even biochemical markers of bone turnover [31–33]. The development of genetic markers is promising, and several candidate genes have been identified [34–37], but at the moment, apart from parental history of hip fracture, there is no practical routine way of capturing the genetic component of hip fracture risk. In any scoring system, a combination of these indicators will have to be used. A truly working instrument will probably have to include individual indicators of the principal etiological axes. Further, not only must risk indicators be strong, but in applying them to unselected populations they must be both universally applicable and of reasonably high prevalence. Moreover, it is essential for simplicity that these risk indicators each individually continue to have a significant contribution to fracture risk in a multivariate model, even when they are used in conjunction. These models will then have to be tested and validated in separate cohorts before any universal application can be recommended. Finally, in devising future efficient combinations of risk indicators for predicting hip fracture, more attention may have to be paid to the differences in etiology between intracapsular (femoral neck) and extracapsular (intertrochan teric) hip fracture. It seemed possible that current antiresorptive treatments, such as estrogens and bisphosphonates, may be more efficient at preventing extra- than intracapsular hip fractures. In the case of bisphosphonates this is because extracapsular hip fractures are associated more closely with low bone density and bisphosphonates appear decreasingly effective as bone density increases; and in the case of estrogens because their cessation leads to a more rapid increase in risk or intra- than extracapsular fractures [38]. Intriguingly, when Cummings and Palermo [38a] addressed this hypothesis they found the reverse to be the case. Also, comorbidities may affect relative risks, particularly coxarthrosis (hip osteoarthritis), which reduces the risk of intra- but not extracapsular fracture [39].

III. VERTEBRAL FRACTURES A. Pre-morphometry Studies of the Epidemiology of Vertebral Fracture Studies undertaken before the development of modern approaches to vertebral morphometry included some very

589 large studies, such as that of Santavirta et al. [40], who identified cases of thoracic vertebral fracture on miniature radiographs taken during screening for tuberculosis. The problems with these early studies were summarized succinctly by Kanis and McCloskey [41]. They centered on the variety of mutually incompatible methods used to identify cases. Sometimes clinical readings were used; otherwise one of a range of morphometric approaches was adopted. Concerning the morphometric methods, it was not always appreciated that when up to 39 vertebral height measurements are made on thoraciclumbar vertebral images from a single spine there is a 50%, not the anticipated 2.5%, chance that one of the heights that is in fact normal will be 2 SDs below the expected value. Therefore, many early studies using morphometric methods generated falsely high prevalence rates. The problems with using clinical readings were starkly revealed by the substantial disagreements between observers in a study comparing diagnoses in a representative sample of test X-rays sent blind to a number of expert radiologists and clinicians [42]. So it was generally agreed some 10 years ago that new studies would have to be undertaken using novel methodologies for identification of vertebral fracture cases. Against this background the European Vertebral Osteoporosis Study (EVOS) was conceived and implemented by a large group of European investigators.

B. Methodological Issues: The EVOS Study Because there is no central European source for comprehensive funding of multinational epidemiological studies for most categories of disease, central funding was available only for the design and analysis of results. The individual centers had to raise their own funds for participation, notably for recruitment, interview, and the X-ray of their subjects. A significant proportion of centers (13 of 36 in 19 countries) also performed dual X-ray absorptiometry (DXA) on the spine and/or hip of between 20 and 100% of their subjects. This was one practical example of the autonomy within the study of the individual centers, who all agreed to operate within a framework set by the study design partnership. At the study’s initiation the previously mentioned investigation of between-observer variation in the clinical identification of deformities on X-rays [42] made it mandatory that all X-rays should be assessed in a single center and that vertebral morphometry as well as clinical identification[43] should be carried out. In evaluating the vertebral morphometry results, the Eastell – Melton algorithm [44] was used alongside the McCloskey – Kanis algorithm [45]. In a pilot study, O’Neill et al. [46] showed that Eastell grade 1 (3 SD) deformities were more prevalent but less specific than McCloskey – Kanis 3 SD deformities and that Eastell grade

590

DE

2 deformities were of the same specificity but less prevalent. It was demonstrated by a three-center analysis that the use of a 3 SD cutoff to define a biologically significant deformity required the use of population- as well as vertebra level-specific normal ranges to define the appropriate cutoff [46]. At the same time as methods for defining vertebral deformities were being refined, considerable effort was devoted to validating the half-hour questionnaire for epidemiological risk factors [47], to establishing whether significant bias was likely to be introduced by varying nonresponse levels in different centers [48,49], and in demonstrating that the answers to the questions asked were reproducible on different occasions to a level that was scientifically acceptable [50]. In general, the questionnaire, which was constructed in close consultation with colleagues undertaking similar work in North America and elsewhere, performed well against these tests. There was no evidence of nonresponse bias substantially affecting the main findings of the study [48].

C. Vertebral Deformity Prevalence Results The main results of the study showed that there was significant variation in prevalence between centers, with rates for both men and women extending over a threefold range after age standardization [51]. Surprisingly, male and female rates were rather similar at over 10% using the McCloskey – Kanis 3 SD definition, with male rates being higher at the age of 50 and female rates being substantially higher than the male rates after the age of 75. Rates in Scandinavia were significantly higher than in other parts of Europe [51].

D. Vertebral Fracture Incidence Rates At the time of writing, the incidence data for vertebral fracture in the EPOS study have only been published in abstract form. An extremely rigorous quality assurance procedure was adopted to ensure that the number of false-positive fractures reported on the second (and the first) films was minimized. The approach adopted was essentially that described by Weber et al. [52] in the pilot study, with careful attention paid to ensuring that pairs of films were adjusted to the same magnification to avoid the artifactual appearance of vertebral height changes [53] and careful visual scrutiny by a senior radiologist of all candidate fractures for (i.e., point placement) diagnostic errors [54]. At age 65, women had a 1.0 – 1.1% risk of suffering an incident vertebral deformity in a 12 month period and this risk increased two-N-fold with each decade increase in age. Based on available North American data, these results do not so far support the hypothesis that American and

LAET AND REEVE

European incidence rates of vertebral fracture in women after the age of menopause are substantially different. Their rates of increase with age also appear similar. The principal difference between the results of the European incidence and prevalence studies concerned men. Before age 60, in the European prevalence study [51], male deformity rates were actually higher than those of women. This difference contributed to the overall similarity of male and female prevalence rates from age 50 to 80, since as expected female rates were substantially higher than male rates over age 70. When risk factors for prevalent deformity were incorporated into models of deformity risk, men appeared to be at twice the risk compared to women of a prevalent vertebral fracture for any given BMD value. In contrast, after adjusting for age and bone density, the risk of an incident fracture was similar in the two genders. The most plausible explanation for these findings is that the gender difference in incident fracture rates in the over 50s depends principally on the postmenopausal bone loss seen in women [55]. The higher prevalence of deformities experienced by men compared to women at younger ages, which is supported by previous single country studies [4, 40], suggests that men are at increased risk of vertebral fractures under the age of 50.

E. Epidemiological Risk Factors So far, the principal sources of data have been prevalence studies, in particular the EVOS study. In EVOS, analyses have been undertaken of both risk of vertebral deformities and risk of low bone density. Concerning risk of vertebral deformity, an important determinant of risk was found to be body mass index (BMI  weight /height2), with leanness being a significant positive risk factor [56]. A long fertile period (interval between menarche and menopause) was found to be significantly protective [57]. Also protective was past use of the oral contraceptive pill [57]. The protective effect of these estrogen-related factors on bone density was also demonstrated [58]. In women, measures of physical activity demonstrated a protective effect of moderate exercise against vertebral deformity, but in men, data suggested that this effect was biphasic. Intensive occupational physical activity was associated with increased risk of a vertebral deformity, whereas physical inactivity was associated with a (nonsignificant) increase in risk by comparison with men who walked or cycled for half an hour or more a day [59]. This suggested the possibility that trauma may play a role in the development of vertebral deformity in men younger than 50 years of age if they had previously experienced very high levels of load-bearing physical activity, perhaps applied to their vertebral column. Moderate alcohol

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

consumption, as in previous studies, was found to be somewhat protective against spinal deformity [60]. However, the presence of menopausal symptoms was unrelated to risk of vertebral deformity [61]. Lifestyle, gynecological, and environmental risk factors for incident vertebral fractures at the time of writing were still being evaluated in the EPOS incidence study. In confirmation of previous studies [62,63], a prevalent fracture was a strong determinant of incident vertebral fracture [54], even after adjusting for BMD. Large fractures (e.g., crush fractures as contrasted with smaller wedge deformities) conveyed a greater increase in risk of future fractures. Beyond that, a new finding, which might be important in relation to the clinical impact on future quality of life, was that the size of a new fracture in terms of vertebral body height lost (calculated as the sum of losses of all three measured heights) could be predicted from the clinical classification of a previous fracture as a wedge, a single end plate concavity, a double end plate biconcavity, or a complete crush fracture [64].

F. Genetic Determinants of Risk Twin studies and studies of mother daughter pairs have made it clear that there is a substantial heritable component to both fracture risk and adult bone density[65]. It is not clear, however, to what extent these associations depend on gene – environment interactions, for which there is some preliminary evidence [66]. The recent doubling of age-specific hip fracture rates in one generation in Europe clearly has environmental or lifestyle rather than directly genetic causes. However, this must not be allowed to detract from our understanding of the potential of heritable factors to explain variations in fracture risk within the present generation. Efforts are currently underway to use candidate gene and genome searching techniques to identify genes of potential importance. We know that concerning risk of vertebral fractures, fractures in a close relative are of predictive value, as a parental history of hip fracture is a positive risk factor for vertebral deformity [67]. So far there have been no large European multinational studies of the genetic determinants of vertebral fractures, but in two smaller single country studies [68,69], the SP1 polymorphism (genotypes sS and SS combined) in the regulatory region of the COL1A1 locus has been associated with a markedly increased risk of vertebral deformity of the order of two-to three-fold. This is particularly intriguing because this polymorphism is only weakly associated, if at all, with low bone density, and most of the known environmental and endocrine risk factors for fracture are associated with equivalent effects on bone density.

591

G. Low Bone Density in the Assessment of Fracture Risk Prevalence studies such as EVOS can provide important clues as to the true determinants of osteoporosis but cannot prove a causal link between risk factors and outcomes. For this reason, 13 out of 36 centers in EVOS explored the role of low bone density in the determination of vertebral deformity risk by measuring spine and hip bone density using DXA in as many as possible of their subjects. The 13 centers were equipped with bone densitometers from four different manufacturers: Hologic, Lunar, Norland, and Sopha. Because these machines all reported bone density in units that are normally not cross-calibrated between machine brands, the bone density studies in EVOS were undertaken in close collaboration with a contemporary concerted action: “The Quantitative Assessment of Osteoporosis” (QAO) [70]. The QAO study developed a semianthropomorphic phantom, the European Spine Phantom Prototype (ESPp), for the cross-calibration of bone densitometers [71 – 73]. With this phantom, most of the systematic differences between the different brands of densitometer were eliminated and, at the same time, differences in initial setting up between different densitometers of the same brand were removed [72]. The EVOS bone density study was one of the first to apply DXA bone densitometry to population samples of men and women aged 50 – 80. We found first that there were substantial differences between centers in mean bone density in the spine, the femoral trochanter, and the femoral neck even after adjusting for age, sex, height, and weight [74]. Apparent rates of change of bone density with age in this crosssectional study were also very different between centers. The size of these differences was surprisingly large; in relation to the overall standard deviation of the combined European population of subjects studied, the range of mean values at these three sites in women typically extended over half a standard deviation (Fig. 3) The size of these betweencenter differences had major implications for the diagnosis of osteoporosis using densitometry cutoffs such as the one proposed by the WHO study group [75]. Another important finding was that the population SD of bone density was dependent on both age and weight. Not only did the spread of values seen in our populations increase with age (with the exception of the femoral neck in women), but weight variation had a surprising and potentially important effect. A low body weight in women was associated with an increase in the standard deviation of bone density as well as being associated with lower mean bone density. This effect is illustrated in Fig. 4. The increasing prevalence of osteoporosis (as defined using the WHO cutoff of  2.5 SDs in relation to a young normal population) increases progressively with declining body weight as the result of a combination of the

592

DE

LAET AND REEVE

Mean density at the spine (  ), femoral neck (), and trochanter () in women in each center, after adjusting to age 65, height 1.65 m and weight 70 kg. Bone densitometers were cross-calibrated with the European Spine Phantom prototype (ESPp). Centers are shown in order of decreasing femoral neck bone density. Ab, Aberdeen (GB); Be, Berlin (DE); Bu, Budapest (HU); Ca, Cambridge (GB); Er, Erfurt (DE); Gr, Graz (AT); Ha, Harrow (GB); He, Heidelberg (DE); Le, Leuven (BE); Mal, Malmo (SE); Man, Manchester (GB); Mo, Moscow (RU); Os, Oslo (NO); Ov, Oviedo (ES); Pi, Piestany (SK); Ro, Rotterdam (NE). From Lunt et al. [74], with permission.

FIGURE 3

effects of weight on mean bone density and on the SD of bone density. These bone density studies also allowed us to explore the association between low bone density and vertebral deformity risk [55]. First we classified the subjects in the study according to whether they had a deformity or whether they were free of vertebral deformities according to the results of using one of the two algorithms [44,45] adopted for use in the study. In a preliminary evaluation we found the risk of being a ‘case’ defined by the McCloskey – Kanis algorithm was more strongly predicted by low bone density [76]; therefore, in the discussion that follows, we shall refer to the use of the McCloskey – Kanis algorithm with a 3 SD population-specific cutoff. In the morphometric assessment of vertebral deformity, three heights are measured per vertebra. These are placed anteriorly, posteriorly, and approximately midway between vertebral endplates [43]. This allows the classification of vertebral deformities into six groups: vertebrae with anterior height loss, those with loss of anterior  midbody height, those with loss of all three heights, those with loss of midbody height only, and, less frequently, those with posterior and midbody height loss or (unusually) posterior  anterior height loss without apparent loss of midbody height. It was found that the principal difference between genders was that bone density was related more strongly to loss of all three vertebral heights (crush fractures) in

FIGURE 4 Predicted prevalence of female femoral neck bone density values below 0.580 g/cm2 after adjustment to age 65 according to the statistical model. This is the value presented by Pearson et al. [71] for a mixed population of normal young European women, which represents a T score of <2.5 as defined by Kanis et al. [75]. For illustrative purposes, women are considered from the two EVOS centers with the highest (Harrow) and lowest (Oviedo) mean age and weight-adjusted BMD values in the femoral neck; both used Hologic machines. The predicted prevalence among women of 50, 60, and 70 kg body weight are shown to demonstrate the association of low bone density with low body weight. From Lunt et al. [74], with permission.

593

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

women than in men [55]. Bone density was strongly related to loss of anterior together with loss of midbody height and also with loss of midbody height alone (Fig. 5). However, the 20 – 27% of subjects with vertebrae that were deformed through loss of anterior height alone showed no significant association between vertebral deformity and low bone density. These subjects are of considerable interest because it seems possible that their deformities might be due to causes other than osteoporosis. Currently, interest centers on the degenerative disease of spinal intervertebral disks as a possible etiological cause [77,78]. We went on to model further the effect of age and sex as well as BMI on risk of vertebral deformity. After adjusting for bone density, high BMI was no longer significantly protective against vertebral deformity; indeed, BMI became weakly positively predictive of vertebral deformity after this adjustment was made [55]. This provides indirect evidence that BMI might protect against vertebral deformity through its effect on bone mineral density. The other principal finding of this analysis was that after adjusting for the difference in rates of bone loss between the sexes after the age of 50, rates of increase of vertebral deformity, so much more rapid in women than

FIGURE 5

in men before adjustment, became more nearly equal [55]. Indeed, the sole effect of sex that remained significant in the model was an approximately twofold increase in risk for men compared to women after adjusting to the same BMD, age, and BMI. Finally, none of the other adjustments was able to substantially reduce the effect of age per se, which remained the most important predictor of vertebral deformity risk in both sexes (Fig. 4).

H. Do Bone Density Variations at the Population Level Account for Geographical Variations in Deformity Risk Seen in Europe? Because there were only 13 centers participating in the bone density studies, there were rather few degrees of statistical freedom to address this question. There was an apparent sex difference in that the between-center differences in age and BMI adjusted BMD were more significant in women. In women, data were consistent with the hypothesis that variations in deformity risk between centers could directly or otherwise be attributed to geographical variations in bone density [55]. Factors relating to

Odds ratios for the increased risk of occurrence of a vertebral deformity of each of six types defined by reductions of one or more of the three measured vertebral heights in any combination (according to the McCloskey – Kanis algorithm) when trochanteric bone density is reduced by 1 population SD. Note similarity between men (left bars) and women (right bars) except when all three vertebral heights are reduced (a classical crush fracture) and the inability of bone density to predict loss of anterior height when unaccompanied by loss of height elsewhere in the vertebra. Means shown  95% CIs. Data from Lunt et al. [74].

594

DE

estrogen exposure had a measurable effect on BMD [58], but exposure to drugs (other than estrogens) had no more than a small effect at the population level (Fig. 5).

I. Effects of Vertebral Deformity on Other Outcome Variables and Quality of Life The European Prospective Osteoporosis Study has been used as a vehicle for studying personal impact-related variables. Vertebral deformities were confirmed to be associated with increased mortality after adjusting for other variables [79]. They predicted hip, but not wrist fractures in the study’s incidence phase, with a more substantial effect in women than in men [RR 4.5 (95% CI 2.1, 9.4) vs 1.7 (0.6, 5.1)] [80]. All types of vertebral deformity were associated with back pain [81], and vertebral deformity had a measurable adverse impact on several quality of life indices [82], an effect that was stronger in men than in women. Multiple severe deformities were associated with disabling back pain [82].Vertebral osteophytosis, sometimes a secondary outcome of vertebral fracture and otherwise usually secondary to spinal osteoarthrosis, was also associated with back pain and was more prevalent in subjects with a history of heavy physical activity [83] (Fig. 6).

LAET AND REEVE

Other multicenter European studies have addressed quality of life issues in osteoporosis. After developing a quality of life questionnaire instrument, the Qualeffo 41, it was validated by an international group in cases of spinal osteoporosis in secondary care [84]. In some respects (the pain and physical function domains), the Qualeffo 41 was more discriminatory than the generic SF-36 instrument. Subsequent work with the Qualeffo 41 has shown that it can be used in some European settings as a postal questionnaire, considerably reducing research costs. The Qualeffo 41 has been used in the European centers participating in the MORE randomized trial of raloxifene [85]. In these studies it was shown clearly that lumbar vertebral fractures had a greater impact on quality of life that thoracic fractures. Also, the increasing impact of osteoporotic spine fractures with numbers of affected vertebrae was clearly apparent.

J. Hospital Burden of Vertebral Fractures The hospital burden varies considerably between different European countries [86], but this is as much the consequence of different criteria for hospital admission, imposed by funding restraints, as the consequence of differing incidence rates.

FIGURE 6 Statistically modeled prevalence of low BMD deformities in men and women by age group adjusted to BMI  25 kg/m2, spinal BMD  1.0 g/cm2, and trochanteric BMD  0.7g/cm2 From Lunt et al. [74], with permission.

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

IV. GENERAL CONCLUSIONS Europe faces a substantial burden of osteoporotic fractures for the forseeable future. This is due in part to growth in the population of older men and women, but also to a secular increase in age-specific risk, at least in women, which may have occurred later in almost all parts of Europe than in Rochester, Minnesota [15]. Some parts of Europe, particularly in the south [6] and east [87], may be currently at lower risk than the north and west. There is no way of knowing what, if any, relationship exists between current rates of osteoporosis and the previous privations, extreme in some parts of Europe, resulting from global conflict. There are differences as well as similarities in risk factor patterns seen in Europe for hip and spine fractures when comparisons are made with U.S. data. However, it is unclear whether these differences are the result of different study designs, stochastic variations between the derived results from studies that are analyzed to the limits of their statistical power, or genuine differences between populations. Studies of male osteoporotic fractures are relatively advanced in Europe. Generally speaking, the European experience of researching the epidemiology and causation of osteoporotic fractures has added considerably to what has been learned from North American studies and has added the depth of insight that comes from different perspectives on the same problem. However, to understand the reasons why some human populations appear genuinely more resistant to osteoporotic fractures, despite in some cases showing no higher bone density that European and US whites, it might be necessary to examine in greater detail the causes of the low fracture rates seen in, for example, west Africa [88,89].

References 1. C. Cooper, Epidemiology and public health impact of osteoporosis. In, “Baillière’s Clinical Rheumatology” D. M. Reid, (ed.), Vol. 7:3, pp. 459 – 77. Baillière Tindall, London, (1993). 2. T. D. Spector, C. Cooper, and A. F. Lewis, Trends in admission for hip fracture in England and Wales, 1968 – 1985. B. Med J. 300, 173 – 174 (1990). 3. U. Bengner, O. Johnell, and I. Redlund-Johnell, Changes in incidence and prevalence of vertebral fractures during 30 years. Calcif. Tissue Int. 42, 293 – 296 (1988). 4. S. Fujiwara, S. Mizuno, Y. Ochi, et al., The incidence of thoracic vertebral fractures in a Japanese population. J. Clin. Epidemiol. 44, 1007 – 1014 (1991). 5. L. Melton III, Hip fractures: A worldwide problem today and tomorrow. Bone, 14, S1 – S8 (1993). 6. I. Elffors, E. Allander, J. A. Kanis, et al., The variable incidence of hip fracture in southern Europe: The MEDOS Study. Osteopor. Int. 4, 253 – 263 (1994). 7. A. J. Silman, on behalf of the European Vertebral Osteoporosis (EVOS) Group. The European Vertebral Osteoporosis Study (EVOS). In “Osteoporosis: Proceedings of the IVth International Symposium on Osteoporosis and Consensus Development” C. Christiansen, and B. Riis, (eds.), pp. 65 – 66. Hong Kong, (1993).

595 8. C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992). 9. European Commission, “Report on Osteoporosis in the European Community: Action for Prevention.” Brussels, (1998). 10. C. E. D. H. De Laet, B. A. Van Hout, and H. A. P. Pols, “Osteoporosis in the Netherlands; a Burden of Illness Study.” Institute for Medical Technology Assessment, Rotterdam, (1996). 11. W. E. Bacon, S. Maggi, A. Looker, et al., International comparison of hip fracture rates in 1988 – 89. Osteopor. Int. 6, 69 – 75 (1996). 12. R. Hedlund, “Incidence and Cause of Femur Fractures.” Karolinska Institutet, Stockholm, (1985). 13. L. Elffors, “Hip Fractures: A European Perspective.” Stockholm, (1999). 14. L. J. Melton III, T. M. Therneau, and D. R. Larson, Long-term trends in hip fracture prevalence: The influence of hip fracture incidence and survival. Osteopor. Int. 8, 68 – 74 (1998). 15. L. J. Melton III, W. M. O’Fallon, and B. L. Riggs, Secular trends in the incidence of hip fractures. Calcif. Tissue Int. 41, 57 – 64 (1987). 16. S. Nungu, C. Olerud, and L. Rehnberg, The incidence of hip fracture in Uppsala County: Change of time trend in women. Acta Orthoped. Scand. 64, 75 – 78 (1993). 17. D. Agnusdei, A. Camporeale, D. Gerardi, S. Rossi, L. Bocchi, and C. Gennari, Trends in the incidence of hip fracture in Siena, Italy, from to 1991. Bone 14, s31 – s34 (1993). 18. C. De Laet, B. van Hout, H. Burger, A. Hofman, and H. Pols, Bone density and the risk of hip fracture in men and women: Cross sectional analysis. B. Med. J. 315, 221 – 225 (1997). 19. A. Oden, A. Dawson, W. Dere, O. Johnell, B. Jonsson, and J. A. Kanis, Lifetime risk of hip fractures is underestimated. Osteopor. Int. 8, 599 – 603 (1998). 20. J. A. Kanis, O. Johnell, B. Gullberg, et al., Risk factors for hip fracture in men from southern Europe: The MEDOS study. Osteopor. Int. 9, 45 – 54 (1999). 21. S. L. Greenspan, E. R. Myers, D. P. Kiel, R. A. Parker, W. C. Hayes, and N. M. Resnick, Fall direction, bone mineral density, and function: Risk factors for hip fracture in frail nursing home elderly. Am. J. Med. 104, 539 – 545 (1998). 22. T. Suzuki, H. Yoshida, T. Hashimoto, et al., Case-control study of risk factors for hip fractures in the Japanese elderly by a Mediterranean Osteoporosis Study (MEDOS) questionnaire. Bone 21, 461 – 467 (1997). 23. S. Cummings, Treatable and untreatable risk factors for hip fracture. Bone 18 (No.3 Suppl.), 165s – 167s (1996). 24. S. R. Cummings, M. C. Nevitt, W. S. Browner, et al., Risk factors for hip fracture in white women. N. Engl. J. Med. 332, 767 – 773 (1995). 25. National Osteoporosis Foundation, Osteoporosis: Review of the evidence for prevention, diagnosis and treatment and cost-effectiveness analysis. Osteopor. Int. 8 (Suppl.) 4:s7 – s80 (1998). 26. C. E. D. H. De Laet, A. E. A. M. Weel, A. Hofman, and H. A. P. Pols, Value of the NOF guidelines for fracture risk assessment in other populations: The Rotterdam Study. Calcif. Tissue Int. 64:s38 (1999). 27. P. Dargent-Molina, A. M. Schott, D. Hans, et al., Separate and combined value of bone mass and gait speed measurements in screening for hip fracture risk: Results from the EPIDOS study. Osteopor. Int. 9, 188 – 192 (1999). 28. H. Burger, C. E. D. H De Laet, A. E. A. M. Weel, A. Hofman, and H. A. P. Pols, Added value of bone mineral density in hip fracture risk scores. Bone 25, 369 – 374 (1999). 29. C. E. De Laet, B. A. H. B. Van Hout, A. E. Weel, A. Hofman, and H. A. P. Pols, Hip fracture prediction in elderly men and women: Validation in the Rotterdam study. J. Bone Miner. Res. 13, 1587 – 1593 (1998). 30. P. L. van Daele, H. Burger, C. E. De Laet, et al., Longitudinal changes in ultrasound parameters of the calcaneus. Osteopor. Int. 7, 207 – 212 (1997).

596 31. P. Garnero, E. Sornay-Rendu, F. Duboeuf, and P. D. Delmas, Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: the OFELY study. J. Bone Miner. Res. 14, 1614 – 1621 (1999). 32. P. Garnero, P. Dargent-Molina, D. Hans, et al., Do markers of bone resorption add to bone mineral density and ultrasonographic heel measurement for the prediction of hip fracture in elderly women? The EPIDOS prospective study. Osteopor. Int. 8, 563 – 569 (1998). 33. P. Garnero, E. Hausherr, M. C. Chapuy, et al., Markers of bone resorption predict hip fracture in elderly women: The EPIDOS Prospective Study. J. Bone Miner. Res. 11, 1531 – 1538 (1996). 34. A. G. Uitterlinden, H. Burger, Q. Huang, et al., Relation of alleles of the collagen type Ialpha1 gene to bone density and the risk of osteoporotic fractures in postmenopausal women. N. Engl. J. Med. 338, 1016 – 1021 (1998). 35. A. Weel, A. G. Uitterlinden, I. C. D. Westendorp, et al., Estrogen receptor polymorphism predicts the onset of natural and surgical menopause. J. Clin. Endocrinol. Metab. 84, 3146 – 3150 (1999). 36. H. A. Pols, A. G. Uitterlinden, and J. P. van Leeuwen, How about vitamin D receptor polymorphisms? Osteopor. Int. 8 (Suppl. 2), s20 – s33 (1998). 37. P. Garnero, O. Borel, S. F. Grant, S. H. Ralston, and P. D. Delmas, Collagen Ialpha1 Sp1 polymorphism, bone mass, and bone turnover in healthy French premenopausal women: The OFELY study. J Bone Miner. Res. 13, 813 – 817 (1998). 38. K. Michaëlsson, E. Weiderpass, B. Y. Farahmand, et al., Differences in risk factor patterns between cervical and trochanteric hip fractures. Osteopor. Int. 10, 487 – 494 (1999). 38a.S. R. Cummings and L. Palermo, J. Bone Miner. Res. 15 (Suppl. 1), s150 (2000). 39. J. S. Wand, I. D. Hill, and J. Reeve, Coxarthrosis and femoral neck fracture. Clin. Orthopaed. Relat. Res. 278, 88 – 94 (1990). 40. S. Santavirta, Y. T. Konttinen, M. Heliövaara, P. Knekt, P. Lüthje, and A. Aromaa, Determinants of osteoporotic thoracic vertebral fracture: Screening of 57,000 Finnish women and men. Acta Orthop Scand. 63, 198 – 202 (1992). 41. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13 (Suppl. 2), s1 – s10 (1992). 42. H. Raspe, A. Raspe, M. Holzmann, et al., Die Reliabilitat radiologischer Befunde zur Differentialdiagnose der vertebralen Osteoporose. Med. Klinik. 93 (Suppl. 2), 34 – 40 (1998). 43. D. Felsenberg, E. Wieland, W. Gowin, et al., Morphometrische Analyse von Rontgenbildern der Wirbelsaule zur Diagnose einer osteoporotischen Fraktur. Med. Klinik 93 (Suppl. 2), 26 – 30 (1998). 44. R. Eastell, S. L. Cedel, H. W. Wahner, B. L. Riggs, and L. J. Melton III, Classification of vertebral fractures. J. Bone Miner. Res. 6, 207 – 215 (1991). 45. E. V. McCloskey, T. D. Spector, K. S. Eyres, et al., The assessment of vertebral deformity: A method for use in population studies and clinical trials. Osteopor. Int. 3, 138 – 147 (1993). 46. T. W. O’Neill, J. Varlow, D. Felsenberg, et al., Variation in vertebral height ratios in population studies. J. Bone Miner. Res. 9, 1895 – 1907 (1994). 47. T. W. O’Neill, C. Cooper, D. Algra, et al., Design and development of a questionnaire for use in a multicenter study of vertebral osteoporosis in Europe: The European Vertebral Osteoporosis Study (EVOS). Rheumatol. Eur. 24, 75 – 81 (1995). 48. T. W. O’Neill, D. Marsden, C. Matthis, H. Raspe, and A. J. Silman, On behalf of the European Vertebral Osteoporosis Study (EVOS) group, Survey response rates: National and regional differences in a European multicentre study of vertebral osteoporosis. J. Epidemiol. Community Health 49, 87 – 93 (1995). 49. T. W. O’Neill, D. Marsden, and A. J. Silman, On behalf of the European Vertebral Osteoporosis Study Group, Differences in the characteristics of responders and nonresponders in a prevalence survey of vertebral osteoporosis. Osteopor. Int. 5, 327 – 334 (1995).

DE

LAET AND REEVE

50. T. W. O’Neill, C. Cooper, J. B. Cannata, et al., Reproducibility of a questionnaire on risk factors for osteoporosis in a multicentre prevalence survey: The European Vertebral Osteoporosis Study. Int. J. Epidemiol. 23, 559 – 565 (1994). 51. T. W. O’Neill, D. Felsenberg, J. Varlow, et al., The prevalence of vertebral deformity in European men and women: The European Vertebral Osteoporosis Study. J. Bone Miner. Res. 11, 1010 – 1017 (1996). 52. K. Weber, M. Lunt, D. Felsenberg, et al., Measurement imprecision in digital vertebral morphometry of spinal radiographs obtained in the EPOS study: Consequences for the identification of prevalent and incident deformities. Br. J. Radiol. 72, 957 – 966 (1999). 53. M. Lunt, W. Gowin, O. Johnell, G. Armbrecht, D. Felsenberg, and J. Reeve, A statistical method to minimise magnification errors in serial X-rays. Bone 23 (Suppl.), s592 (1998) [Abstract] 54. D. Felsenberg, M. Lunt, G. Armbrecht, et al., Rates and determinants of vertebral fracture incidence in European men and women. J. Bone Miner Res. 14 (Suppl. 1), s159 (1999) [Abstract] 55. M. Lunt, D. Felsenberg, J. Reeve, et al., Bone density variation and its effects on risk of vertebral deformity in Men and Women studied in 13 European centres: The EVOS Study. J. Bone Miner Res. 12, 1883 – 1894 (1997). 56. O. Johnell, T. O’Neill, D. Felsenberg, et al., Anthropometric measurements and vertebral deformities. Am. J. Epidemiol. 146, 287 – 293 (1997). 57. T. W. O’Neill, A. J. Silman, M. Naves Diaz, et al., Influence of hormonal and reproductive factors on the risk of vertebral deformity in European women. Osteopor. Int. 7, 72 – 78 (1997). 58. P. Masaryk, M. Lunt, L. Benevolenskaya, et al., Effects of menstrual history and use of medications on bone mineral density: The EVOS study. Calcif. Tissue Int. 63, 271 – 276 (1998). 59. A. Silman, T. O’Neill, C. Cooper, J. Kanis, and D. Felsenberg, The EVOS Study Group, Influence of physical activity on vertebral deformity in men and women: Results from the European Vertebral Osteoporosis Study. J. Bone Miner Res. 12, 813 – 819 (1997). 60. M. Naves Diaz, T. O’Neill, and A. Silman, The EVOS Study Group, The influence of alcohol consumption on the risk of vertebral deformity. Osteopor. Int. 7, 65 – 71 (1997). 61. V. Scoutellas, Does the presence of post-menopausal symptoms influence susceptibility to vertebral deformity? Maturitas 32, 179 – 187 (1998). 62. P. D. Ross, J. N. Davis, R. S. Epstein, and R. D. Wasnich, Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 63. P. Ross, H. Genant, J. Davis, P. Miller, and R. Wasnich, Predicting vertebral fracture incidence from prevalent fractures and bone density among non-black, osteoporotic women. Osteopor. Int. 3, 120 – 126 (1993). 64. M. Lunt, D. Felsenberg, G. Armbrecht, et al., Incident vertebral fractures: Determinants of size and location. J. Bone Miner. Res. 14 (Suppl. 1), s259 (1999) [Abstract] 65. S. Ralston, The genetics of Osteoporosis. Bone 25, 85 – 86 (1999). 66. S. Ferrari, R. Rizzoli, T. Chevalley, D. Slosman, J. A. Eisman, J.-P. Bonjour, Vitamin-D-receptor-gene polymorphisms and change in lumbar-spine bone mineral density. Lancet 345, 423 – 424 (1995). 67. M. Naves Diaz, T. O’Neill, and A. Silman, The EVOS Study Group, The influence of family history of hip fracture on the risk of vertebral deformity in men and women: The European Vertebral Osteoporosis Study. Bone 20, 145 – 9 (1997). 68. S. F. A. Grant, D. M. Reid, G. Blake, R. Herd, I. Fogelman, and S. H. Ralston, Reduced bone density and osteoporosis associated with a polymorphic Sp1 binding site in the collagen type 1a1 gene. Nature Genet. 14, 203 – 205 (1996). 69. B. L. Langdahl, S. H. Ralston, S. F. Grant, and E. F. Eriksen, An Sp1 binding site polymorphism in the COL1A1 gene predicts osteoporotic

597

CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

fractures in both men and women. J. Bone Miner. Res. 13, 1384–1389 (1998). J. Dequeker, J. Reeve, J. Pearson, et al., Multicentre European COMAC-BME study on the standardisation of bone densitometry procedures. Technol. Health Care 1, 127–131 (1993). J. Pearson, J. Dequeker, J. Reeve, et al., Dual X-ray absorptiometry of the proximal femur: Normal European values standardised with the European spine phantom. J. Bone Miner. Res. 10, 315–324 (1995). J. Pearson, J. Dequeker, M. Henley, et al., European semi-anthropomorphic spine phantom for the calibration of bone densitometers: Assessment of precision, stability and accuracy. Osteopor. Int. 5, 174–184 (1995). J. Dequeker, J. Pearson, J. Reeve, et al., Dual X-ray absorptiometry: Cross-calibration and normative reference ranges for the spine: results of a European Community Concerted Action. Bone 17, 247–254 (1995). M. Lunt, D. Felsenberg, J. Adams, et al., Population-based geographic variations in DXA bone density in Europe: The EVOS study. Osteopor. Int. 7, 175–189 (1997). WHO Study Group, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis WHO Technical Report,” Vol. 843. World Health Organisation, Geneva, 1994. T. O’Neill, M. Lunt, D. Felsenberg, et al., The choice of algorithm for evaluating spine deformities from 6 point placements: Use of DXA densitometry as a referent. Calcif. Tissue Int. 56, 496 (1995). A. Abdel-Hamid Osman, H. Bassiouni, R. Koutri, J. Nijs, P. Geusens, and J. Dequeker, Aging of the thoracic spine: Distinction between wedging in osteoarthritis and fracture in osteoporosis—A cross-sectional and longitudinal study. Bone 15, 437–442 (1994). T. Masud, S. Langley, P. Wiltshire, D. V. Doyle, and T. D. Spector, Effect of spinal osteophytosis on bone mineral density measurements in vertebral osteoporosis. Br. Med. J. 307, 172–173 (1993). A. A. Ismail, T. W. O’Neill, C. Cooper, et al., Mortality associated with vertebral deformity in men and women: Results from the Euro-

80.

81.

82.

83.

84.

85.

86.

87. 88.

89.

pean Prospective Osteoporosis Study (EPOS). Osteopor. Int. 8, 291–297 (1998). A. A. Ismail, W. Cockerill, C. Cooper, et al., Prevalent vertebral deformity predicts future hip though not distal forearm fracture: Results from the European Prospective Osteoporosis Study. Osteopor. Int. 12, 85 –90 (2001). A. A. Ismail, C. Cooper, D. Felsenberg, et al., Number and type of vertebral deformities: Epidemiological characteristics and relation to back pain and height loss. Osteopor. Int. 9, 206–213 (1999). C. Matthis, U. Weber, T. W. O’Neill, and H. Raspe, The European Vertebral Osteoporosis Study Group. Health impact associated with vertebral deformities: Results from the European Vertebral Osteoporosis Study (EVOS). Osteopor. Int. 8, 364 – 372 (1998). T. O’Neill, E. McCloskey, J. Kanis, et al., The distribution, determinants and clinical impact associated with vertebral osteophytosis: A population-based survey. Rheumatol. 26, 842 – 848 (1999). P. Lips, C. Cooper, D. Agnusdei, et al., Quality of life in patients with vertebral osteoporosis: Validation of the quality of life questionnaire of the European Foundation for Osteoporosis (QUALEFFO). Osteopor. Int. 10, 150 – 160 (1999). A. Oleksik, P. Lips, A. Dawson, et al., Health-related quality of life in postmenopausal women with low BMD with or without prevalent vertebral fractures. J. Bone Miner Res., 15, 1384 – 1392 (2000). O. Johnell, B. Gullberg, and J. A. Kanis, The hospital burden of vertebral fracture in Europe: A study of national register sources. Osteopor. Int. 7, 138 – 144 (1997). L. I. Benevolenskaya, Osteoporosis: A genuine medical problem. Osteopor. Osteopath. 1, 4 – 7 (1998). [In Russian] T. J. Aspray, A. Prentice, T. Cole, Y. Sawo, J. Reeve, and R. M. Francis, Low bone mineral content is common but osteoporotic fractures are rare in elderly rural Gambian women. J. Bone Miner Res. 11, 1019 – 1025 (1996). A. O. Adebajo, C. Cooper, and J. G. Evans, Fractures of the hip and distal forearm in west Africa and the United Kingdom. Age Ageing 20, 435 – 438 (1991).

599

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

CHAPTER 24

Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood BONNY L. SPECKER RAN NAMGUNG REGINALD C. TSANG

I. II. III. IV.

Ethel Austin Martin Program in Human Nutrition, South Dakota State University, Brookings, South Dakota 57007 Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea Department of Pediatrics, University of Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio 45229

Introduction Measurement of Bone Mass in Infants and Children Bone Acquisition in Utero Bone Acquisition in Preterm Infants

V. Bone Mineral Acquisition in Term Infants and Children VI. Summary References

II. MEASUREMENT OF BONE MASS IN INFANTS AND CHILDREN

I. INTRODUCTION The two major factors determining whether an individual is at increased fracture risk due to low bone mass are the peak bone mass achieved early in life and the rate of bone loss thereafter. For many years research has focused on strategies to ameliorate bone loss; only recently has the importance of bone gain early in life been recognized and it has been suggested that the roots of osteoporosis may be based in childhood [1]. This chapter reviews the factors contributing to early bone gain, both genetic and environmental. For consideration of bone acquisition during adolescence, the reader is referred to Chapter 25.

OSTEOPOROSIS, SECOND EDITION VOLUME 1

Before commencing a description of the normal and abnormal progression of bone acquisition, it may be useful to review the technical aspects of skeletal assessment in children. In the 1970s, single photon absorptiometry (SPA) techniques were first used to study bone mass in infants. Prior to this time diagnosis relied on conventional radiological methods, which could detect osteopenia only if significant bone loss had already occurred [2]. SPA measurements typically involved the forearm, and normative pediatric data were reported for infants of various gestational ages [3 – 5] and for healthy children [2,6 – 11]. Numerous studies

599

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

600

SPECKER, NAMGUNG, AND TSANG

also were reported showing decreased bone mineralization in children with various clinical conditions ranging from renal disease to diabetes mellitus [12,13]. SPA also was used to study the effect of different infant feeding regimens on bone mineral content (BMC) in both term and preterm infants [14 – 19]. The major disadvantages of SPA were the need for a radioactive source that deteriorated over time and the inability to measure axial bone sites. In the early 1990s the measurement of BMC in children at axial and peripheral bone sites, as well as the total body, became possible with the development of pediatric software using dual- energy X-ray absorptiometry (DXA). Several reports have since validated the use of pencil beam DXA in children [20 – 25], and reference data are available for lumbar spine in preterm infants [26] and healthy children [27]. However, reference data differ among machine manufacturers, and differences in results obtained using pediatric versus adult software can be substantial. Figure 1 (see also color plate) illustrates the magnitude of difference in total body BMC comparing infant and adult software on the same machine and between different models from the same manufacturer. Data in Fig. 1 come from numerous completed or ongoing cross-sectional and longitudinal studies in healthy infants and children [28 – 30], as well as unpublished data (Specker). Currently, no consistent recommendation addresses the weight or age for which infant, pediatric or adult software should be applied to children. An additional problem with the clinical use of pediatric DXA is the lack of an adequate reference database. Several candidate databases have been evaluated to determine their relative performance in rendering a diagnosis of osteopenia (age-related Z score  2.0) [31]. The results showed sub-

stantial inconsistency, even using similar DXA machines and software, with 11 – 30% of children with chronic diseases being classified as osteopenic depending on the referent database. One of the difficulties in interpreting DXA results is that bone mineral density (BMD) is expressed as an areal measurement, i.e., BMC divided by the projected bone area. Calculation of BMD (BMC/area) assumes that BMC and bone area are directly proportional to each other, which is often not the case. Several mathematical methods have been proposed to adjust areal BMD to more closely reflect true volumetric BMD [32 – 35]. Measurement of the latter is possible using quantitative computed tomography (QCT), although the cost and accessibility of CT scanners, and the relatively high radiation exposures, limit the use of CT in pediatrics [36]. Several investigators are currently applying peripheral QCT (pQCT) to children. These machines are relatively inexpensive and radiation exposure is low. The technique allows for the measurement of volumetric density of both cortical and trabecular bone at peripheral sites, and estimates of bone strength can be calculated based on the geometric properties of the bone [37]. Accurate measurements of cortical thickness in the range commonly observed in young children have been reported, although cortical BMD values may be underestimated in young children whose cortical thickness is less than 2 mm [38]. Ultrasound has been used in the measurement of pediatric bone and is thought to be a measure of not only BMD but also bone quality (see also Chapter 59). Measurements in healthy populations tend to vary, but differences between osteopenic and normal children have been reported [39 – 42].

FIGURE 1 Measurement of total body BMC by body weight using the Hologic 1000W infant whole body software (squares, solid line), the Hologic 1000W adult whole body software (solid diamonds, lower dashed line), or the Hologic QDR4500A pediatric whole body software (triangles, upper dashed line). (See also color plate.)

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

The use of these different bone measurement techniques in healthy pediatric populations has contributed greatly to understanding the factors that influence normal bone accretion. Without an understanding of what occurs in the healthy child, it is difficult to determine what factors may optimize peak bone mass, or the pathogenesis of bone diseases that may occur in children.

III. BONE ACQUISITION IN UTERO During the period of a normal human pregnancy the fetus accumulates approximately 30 g of calcium; the majority of this is accrued during the last trimester [43]. The concentration of calcium in fetal blood exceeds that in maternal blood, and this maternofetal gradient emerges as early as 20 weeks of gestation [44]. Transplacental transfer of calcium, as well as phosphorus and magnesium, against concentration gradients, along with evidence for interference with them by metabolic inhibitors, suggests that placental mineral transport requires active processes [45,46]. The rate of maternofetal calcium transfer increases dramatically during the last trimester (from 24 weeks to term gestation), with peak accretion rates at 36 to 38 weeks gestation. The reported range of peak accretion for calcium at 34 to 36 weeks gestation is between 102 and 151 mg/ kg/day; for phosphorus is 65 to 85 mg/kg/day; and for magnesium is 2.5 to 3.7 mg/kg/day. At term, the body content of calcium, phosphorus, and magnesium, as determined by chemical analysis [47 – 49], neutron activation analysis [50], or DXA [21], are quite comparable and approximate 28 to 30 g, 17 to 19 g, and 0.6 to 0.8 g, respectively. Thus, about two-thirds of total body calcium, phosphorus, and magnesium accumulated in a healthy term human fetus is transported during the last trimester of pregnancy [51]. Factors that increase placental calcium transport capacity as gestation proceeds are genetically preprogrammed or achieved through hormones produced by mother, fetus, or placenta. Hormonal signals controlling placental calcium transfer might act on maternal or fetal sides of the placenta. Possible regulatory hormones are 1,25-dihydroxyvitamin D [1,25(OH)2D], parathyroid hormone (PTH), PTH-related peptide (PTHrP), calcitonin, or other peptides secreted by the parathyroid glands. PTHrP midmolecule fragments produced in fetal parathyroid glands also appear to stimulate maternofetal calcium transport [52]. The increased calcium demand due to fetal bone accretion is likely met by an increase in maternal intestinal calcium absorption. A recent study of pregnant women showed a dissociation between bone formation and resorption during pregnancy. An increase in maternal bone resorption was observed during the first trimester, yet an increase in markers of bone formation did not occur until the

601

third trimester [53] (see also Chapters 29 and 53). The physiological significance of the early increase in bone resorption is not clear as the majority of fetal mineral accretion occurs during the last trimester. Although PTHrP may play a role in the increase in bone resorption during pregnancy [54], not all studies support this hypothesis. Black and co-workers [53] observed an increase in circulating insulin-like growth factor-I (IGF-I) prior to an increase in biomarkers of bone formation and suggested that this hormone may play a role in stimulating bone formation in the last trimester. Cross-sectional studies of preterm infants have provided estimates of in utero BMC accretion, although it is not clear whether these infants are representative of healthy infants born at term. Figure 2 illustrates total body BMC by body weight for infants measured within several days of birth. Factors that influence the normal accretion of bone in utero or alter the calcium metabolism of mother or fetus may affect the amount of bone mineral present at birth. Because the majority of bone is gained during the last trimester, one of the major variables affecting bone mass at birth is gestational age. Other factors that may influence neonatal BMC include environmental variables such as season of birth and maternal or fetal disorders.

A. Gestational Age and Size As intrauterine bone mineral accretion is substantial during the last trimester, infants born prematurely have lower BMC than term infants of similar postconceptional age. Several studies have reported normative data for preterm

FIGURE 2 Scatter plot for total body BMC as a function of body weight within several days of birth. Reproduced from Koo et al. [21], with permission.

602 and term infants at birth [21,24,25,55,56]. Preterm infants are at increased risk for osteopenia and rickets due to postnatal nutritional deficits resulting from prolonged exclusive human milk feeding, prolonged total parenteral nutrition with low calcium and phosphorus content, and chronic diuretic therapy [57 – 60]. The development of high mineral containing preterm infant formulas and human milk fortifiers have reduced, but not eliminated, the frequency of osteopenia in this population [61,62]. Results of prospective studies of preterm infants indicate that, given formula with sufficient mineral content following discharge, a catch-up in bone mineralization can occur in these infants, and their BMC is appropriate for size [63 – 66]. However, former preterm infants tend to be shorter and lighter than their peers at 8 to 12 years of age [65,66]. Small-for-gestational age (SGA) infants have decreased BMC compared to appropriate-for-gestational age (AGA) infants even after controlling for body size [67]. Cord serum osteocalcin and 1,25(OH)2D concentrations were also reported to be lower in SGA infants, but serum PTH concentrations were similar in the two groups. The authors suggested that low uteroplacental blood flow may cause a reduction in fetal-placental 1,25(OH)2D production, leading to low bone turnover in SGA infants. It was noted that reduced blood flow leading to reduced mineral transfer across the placenta also might directly affect bone formation. Using an experimental model of fetal growth retardation in the rat (uterine artery ligation), Mughal et al. [68] showed that maternofetal transfer of calcium across the placenta is reduced proportional to the reduction in body size. This finding may simply reflect the smaller requirement of calcium in a small fetus, but it may also be due to depletion of the necessary energy supply for active placental Ca transport secondary to a chronic reduction in uteroplacental blood flow. In theory, conditions that affect fetal growth can lead to changes in the metabolism of type I collagen, the dominant collagen isoform in bone. Namgung et al. [69] examined whether alterations in type I collagen metabolism could be a cause for low BMC in SGA infants. They found that cord serum markers of type I collagen synthesis (PICP, carboxyterminal propeptide of type I collagen) and degradation (ICTP) did not differ between SGA and AGA infants, leading to the conclusion that low BMC in SGA vs AGA infants predominantly reflects low mineral supply rather than defective collagen metabolism. However, concentrations of bone turnover markers vary highly in the neonatal period and in infancy, which may have made it difficult to detect significant differences in this relatively small series. A proportion of SGA infants have dcreased growth [70] even into childhood, and these authors have suggested that an alteration in levels of IGF or other growth factors may be responsible for this long-term growth retardation.

SPECKER, NAMGUNG, AND TSANG

B. Environmental Variables Season of birth may influence bone mass due to alterations in vitamin D status. Adult and pediatric studies both report seasonal fluctuations in vitamin D concentrations, and it is likely that low maternal vitamin D status during pregnancy subsequently affects fetal bone mass. Studies at birth have found that the bone mass of infants born in the summer is lower than that of those born in the winter [67,71,72]. In Korea, where vitamin D supplementation is low, infants born in winter months have markedly (8%) lower total body bone mineral content (BMC) at birth, lower cord serum 25-hydroxyvitamin D (25-OHD), and higher bone resorption marker ICTP (cross-linked carboxyterminal telopeptide of type I collagen) than infants born in summer: infant total body BMC correlated positively with cord serum 25-OHD, and inversely with the resorption marker; the bone resorption marker also correlated negatively with vitamin D status (Figs. 3 and 4) [73]. These findings may imply that low vitamin D status in winter-born Korean infants may result in increased bone resorption and lower bone mineral status. In three separate North American studies, markedly lower (8 to 12%) BMC was observed in summer than in winter newborns, opposite to the findings of Korean neonates (Fig. 5) [67,71,72]. The major reason for these discrepant BMC results might be a markedly different maternal vitamin D status in Korea than in North

FIGURE 3 Newborn total body bone mineral content (TBBMC, g) in winter and summer in a population without routine vitamin D supplementation (Korea). Winter-born newborn infants had significantly lower TBBMC compared with summer-born newborn infants (p  0.0002). Reproduced from Namgung et al. [73], with permission.

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

603

FIGURE 4

Newborn mineral metabolism in relation to vitamin D status. (A) Relationship between total body bone mineral content (TBBMC) and serum 25-hydroxyvitamin D(25-OHD, ng/ml) (TBBMC  87.256  0.425 [25-OHD]; r  0.243, p  0.047). (B) Relationship between TBBMC and serum ICTP (ng/ml, a bone resorption marker)(TBBMC  103.74  0.149 [ICTP]; r  0.333, p  0.008). (C) Relationship between serum ICTP and 25-OHD (ICTP  98.3  1.5 [25-OHD]; r   0.391, p  0.001). Reproduced from Namgung et al. [73], with permission.

America: in Korea, cord serum 25-OHD concentrations in winter are extremely low (mean  SD: 4.3  3 ng/ ml), and 97% of winter-born neonates have suboptimal vitamin D concentrations related to lack of vitamin supplement throughout pregnancy. In contrast, in the United States, cord serum 25-OHD concentration did not differ between winter and summer newborns, probably related to vitamin supplement use in most pregnant women (especially beyond early pregnancy). During the first trimester of pregnancy, however, most women in the United States have not received much prenatal care and are less likely to be taking multivitamin pills that contain

vitamin D. Thus, the mother is relatively “unprotected” (i.e., not given vitamin D supplements), and seasonal changes of vitamin D status can occur at that time. The authors proposed that seasonal vitamin D status changes result in first-trimester bone alterations in winter that are evident 6 months later as reduced bone mass at birth in summer compared with winter. The period of greatest vulnerability of the fetus to altered maternal vitamin D status, if any, is unclear. However, from these studies, apparently it can happen either in the early and/or in the late trimester of pregnancy, depending on the maternal vitamin D status.

604

SPECKER, NAMGUNG, AND TSANG

FIGURE 5

Newborn bone mineral content (BMC, mg/cm) in winter and summer by race and gender in a population where vitamin D fortification is practiced (Cincinnati, OH). Values are least- square means adjusted for birth weight  SE with numbers of newborns in parentheses. Summer-born infants had significantly lower BMC than winter-born infants (p  0.009). Race and gender differences in BMC corresponded to birth weight differences; after adjusting for birth weight, race and gender differences were p  0.09 and p  0.17, respectively. Seasonal differences in BMC remained significant after adjusting for race and gender effects (p  0.002 and p  0.001, respectively). Reproduced from Namgung et al. [72], with permission.

C. Maternal Factors Decreased skeletal mineralization in utero may be manifested as rickets or osteopenia in the newborn infant. However, fetal or congenital rickets of the newborn is rare. Congenital rickets in newborn infants of mothers with severe nutritional osteomalacia has been reported [74 – 76]. Maternal vitamin D deficiency may also lead to craniotabes, or delayed ossification of the cranial vertex [77]. A study conducted in China also found evidence for a relationship between maternal vitamin D deficiency and impaired fetal bone ossification [78]. Wrist ossification centers were less likely to be present in infants born in northern China than in southern China, and the presence of ossification centers was associated with cord serum concentrations of 25-OHD. A higher rate of ossification centers in newborn infants of mothers with adequate vitamin D status was apparent when compared to infants of mothers with low vitamin status. Maternal hypo- or hyperparathyroid disease also is associated with neonatal osteopenia or rickets [79]. 1. INTRAVENOUS MAGNESIUM SULFATE THERAPY PREGNANCY

DURING

Long-term maternal intravenous magnesium sulfate therapy has been associated with newborn rickets [80,81]. It is presumed that magnesium may replace calcium in bone, thereby altering bone metabolism. However, a controlled study of mothers requiring prolonged bed rest for preterm labor showed that the issue is not simple. Preterm infants born to these mothers receiving long-term antenatal

therapy with magnesium sulfate were compared with those of mothers who were also at bed rest but not receiving magnesium sulfate. Infants born to the magnesium-treated mothers had delayed clearance of magnesium and phosphorus, and normalized serum calcium by 72 h after delivery. However, both groups had similar radius BMC measurements and anthropometric indices after delivery [82]. 2. MATERNAL DIABETES MELLITUS Maternal diabetes may affect BMC at birth through alterations in mineral metabolism [83]. Neonatal hypocalcemia and hypomagnesemia are frequently observed in infants of insulin-dependent diabetic mothers, and the severity of the hypocalcemia corresponds directly to the severity of maternal diabetes. Infants of insulin-dependent diabetic mothers (IDM) may have low BMC at birth (about 10% reduction compared with control infants), and infant BMC decreases progressively with poor control of maternal diabetes, specifically with first (but not later) trimester maternal mean capillary blood glucose concentration, implying that factors early in pregnancy might have an effect on fetal BMC (Fig. 6) [83]. The low BMC in IDMs may be related to the decreased transplacental mineral transfer. The bone resorption marker ICTP in cord serum was higher in IDM than in control subjects, implying increased intrauterine bone resorption [84]. Increased cord concentrations of 1,25(OH)2D reported in infants of diabetic mothers also may contribute to decreased BMC by promoting bone resorption [85]. The increased renal magnesium losses of insulin-dependent diabetic mothers may

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

605

FIGURE 6

Infants of diabetic mothers: neonatal bone mineral content values plotted against mean maternal first trimester postprandial blood glucose concentration. There is a significant inverse correlation; low neonatal bone mineral content relates to high maternal blood glucose (r  0.6636, p 0.01). Reproduced from Mimouni et al. [83], with permission.

lead to chronic magnesium deficiency, reducing the ability of the parathyroid glands to respond to low serum calcium concentrations. These alterations may also contribute to decreased BMC. In addition, the secondary hypoparathyroidism resulting from hypomagnesemia may be related to the high incidence of neonatal hypocalcemia in infants of diabetic mothers. A randomized trial of magnesium administration to newborn infants of diabetic mothers did not find a statistically significant decrease in the incidence of clinical hypocalcemia among 26 infants randomized to receive magnesium compared to 23 infants randomized to placebo. However, the decrease in serum calcium concentrations from birth to 72 h was smaller for the active treatment group [86]. Fetal histomorphometric analysis and ash studies [87 – 90] have been performed in the offspring of both genetically diabetic rats and those rendered diabetic by streptozotocin. These data indicate that net fetal accretion of bone minerals and fetal bone mineralization in the experimental diabetic pregnancy is reduced. The reduced fetal bone mineral accretion is likely to be associated with lower net placental mineral transport. In in situ-perfused rat placenta of the streptozotocin-induced diabetic rat, maternofetal fluxes of both calcium and magnesium are reduced significantly by between 20 and 40% in untreated maternal diabetes compared with a control group of animals. Net fetal accretion of calcium and magnesium was reduced in the untreated diabetic group of animals and largely normalized in the insulin-treated group. After statistical adjustment for low weight of fetuses of untreated diabetics, the calcium content, but not that of magnesium,

remained significantly lower in untreated animals, implying that the net flux of calcium across the diabetic rat placenta is reduced to a greater degree than that expected from growth retardation alone [91]. 3. MATERNAL ETHANOL CONSUMPTION, SMOKING, AND CAFFEINE INTAKE Ethanol consumption and smoking during pregnancy affect fetal skeletal development. Pregnant rats fed a liquid diet with ethanol showed low Ca content of maternal bone, low maternal blood-ionized Ca activity, and elevated serum PTH concentrations compared with rats fed diets without ethanol (pair-fed controls); mean fetal body weight and fetal skeletal ossification were reduced in the ethanol-fed rats compared with controls but there were no differences in fetal Ca content [92]. Maternal smoking is detrimental to the developing fetus: in rats there is dose-related retardation in embryonic growth, and also a reduction in the number of skeletal ossification centers (developmental delay) [93]. Maternal smoking during pregnancy is associated with deficits in growth (lower weight and height) in 8-year-old children born at term, and also with disproportionate deficits in bone mass. Children whose mothers smoked during pregnancy had lower size adjusted bone mass at the lumbar spine and femoral neck but not total body [94]. These associations may be mediated through placental size and function, as adjustment for placental weight led to a loss of significance for both growth and bone variables.

606

SPECKER, NAMGUNG, AND TSANG

Maternal caffeine intake during rat pregnancy resulted in decreased fetal weight and in fetal bone calcium and magnesium content. When data are expressed per gram of bone tissue, most of these decreases disappeared, suggesting that maternal caffeine intake influences overall growth and development [95]. 4. MATERNAL CALCIUM SUPPLEMENTATION Koo and associates [96] showed that maternal calcium supplementation of up to 2 g/day during the second and third trimesters can increase fetal BMC (about 15% increase in total body BMC vs placebo group) in adequately nourished women with low dietary calcium intake ( 600 mg of Ca intake/day). The effect of calcium supplementation remained significant after adjustment for maternal age and body mass index, and after normalization for skeletal area and body length of the infant. However, supplementation in pregnant women with adequate dietary calcium intake (600 – 1800 mg/day) was not associated with an increase in fetal bone mineral content. The authors suggested that a maternal homeostatic response to high calcium intakes might be a decrease in percentage absorption of calcium, and thus no further enhancement of calcium retention may occur with supplementation. Animal studies have shown that fetal skeleton may be affected by low [97] and high [98,99] maternal calcium intake. In ewes fed a low calcium diet (about one-third of normal) for 2 months beginning in the second trimester, the fetus showed delayed humoral ossification, a 20% decrease in the proportion of bone to cartilage, and an 11% decrease in fetal bone ash content [97]. Very high maternal calcium intake during pregnancy theoretically may be detrimental to the fetus. For example, intakes at 2.5 times the control diet in ewes can result in fetal osteochondrosis [98]. Fetuses of rats given calcium at 2.5 times the control diet show a decrease in whole body content of phosphorus by 3%, and magnesium by 2%, possibly related to mineral interaction [99].

D. Fetal Factors Any disorder that affects fetal growth is likely to affect bone accretion. Infants with osteogenesis imperfecta have been reported to have decreased BMD for their age and weight [100,101]. A possible role of activity and muscular contraction in fetal bone development can be speculated based on findings of osteopenia in newborn infants with intrinsic fetal akinesia [102]. Newborn infants with intrinsic akinesia resulting from neuromuscular diseases typically have a lack of muscle contraction. However, those with extrinsic akinesia caused by oligohydramnios, who still have muscle contractions, do not have osteopenia. Based on these findings, it has been proposed that the lack of muscle

contraction underlies the osteopenia in this group of infants [103,104]. In rat fetuses curarized from day 17 of gestation until term, long bone changes resembled those observed in human fetuses and newborns with congenital neuromuscular disease [104]. However, as in human fetuses, described earlier, fetal rats subjected to oligohydramnios had no alterations in femoral shape and transverse growth of the metaphysis and diaphysis, reinforcing the suggestion that the main mechanical factor related to fetal bone modeling is muscular strength, while motion would mainly affect fetal joint development (lack of motion in oligohydramniosis is associated with joint contractures).

IV. BONE ACQUISITION IN PRETERM INFANTS Osteopenia and rickets during the first year of life occur most often in very low birth-weight preterm infants. The cause of osteopenia and rickets in these infants is likely multifactorial. Because the fetus acquires more than twothirds of its body mineral content during the last trimester of pregnancy, preterm infants are theoretically at a disadvantage for mineral homeostasis because they have missed the peak mineral accretion period and have not acquired their full antenatal complement of minerals. Thus, inadequate mineral intake during their early postnatal life may result in biochemical abnormalities and failure of appropriate bone mineralization. Bone mineralization after birth in a premature infant does not occur in the same way as in utero. The difference in bone mineralization could be explained in part by the difference in the amount of minerals transferred to the fetus across the placenta in the face of limited enteral mineral intake in the preterm postnatal infant. Intrauterine calcium retention in the last trimester approximates 140 mg/kg/day [48], whereas retention from standard enteral feeding in preterm infant is limited to approximately 70 – 80 mg/ kg/day [105]. In preterm infants, daily growth may be similar to or greater than that seen in utero, but daily Ca retention is only about 50% of fetal accretion, which is thought to lead to an increased incidence of osteopenia [105]. Significantly lower (around 50%) BMC in premature infants (gestation  32 weeks) has been reported at term compared with full-term infants [106,107]. Normative data for BMC in preterm and term infants are very limited. Results are often difficult to compare because of different equipment and software, or differences in the population studied, especially related to vitamin D status and calcium supply of the pregnant mother. Using DXA (QDR 1000) in preterm infants, Koo et al. [21], reported normative data for total body BMC in preterm and term infants. They found a positive relationship between body

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

607

weight and bone mineral content [total body BMC (g)  24.2 study bare weight (kg)  11.1; r2  0.95, p0.0001] (Fig. 2). From that study, BMC is 25.2 g at 1500 g body weight and 73.0 g at 3500 g. In Rigo’s study [108] also using DXA QDR 2000 Infant Whole Body Software 5.65 (Hologic Inc., Waltham, Massachusetts), BMC is 20 g at 1500 g body weight and 63 g at 3500 g [total body BMC(g)  21.5 DXA body weight(kg)  12.2; r2  0.94, p0.0001].

A. Factors Affecting Bone Mineral Acquisition in Preterm Infants 1. DIETARY FACTORS a. During Hospitalization Premature infants have higher calcium requirements than full-term infants in early postnatal life. These may be met by using human milk fortified with additional minerals (cow milk-based fortifier) or with specially designed formulas for premature infants [62]. The exact amount and duration of calcium and phosphorus supplement required for preterm infants are not known. Human milk (which contains only approximately 25% of the amount of Ca and P needed for normal bone mineralization at intakes of 200 ml/kg/day) and standard humanized milk formulas do not provide enough calcium, phosphorus, or magnesium to meet the bone mineralization rate of these rapidly growing preterm infants [109,110]. Specially designed preterm formulas that are fortified with Ca and P can approach the fetal mineral accretion rate by providing a high mineral intake necessary for normal bone mineralization. An earlier study done by Steichen et al. [111] showed that preterm infants (gestation 33 – 35 weeks) fed a high mineral concentration formula (120 mg Ca and 63 mg P per 100 ml) achieve a BMC measured by single-photon absorptiometry equal to intrauterine levels and increase their BMC by 58% during the first 10 weeks of life (Fig. 7). BMC of preterm infants fed mineralenriched formula is not different from intrauterine bone mineral curve and significantly exceeds that of routinely fed infants at 4 – 6 weeks’ postnatal age. However, in extremely low birth weight infants with birth weight  800 g, 6 to 8 weeks of increased mineral supplementation may still be insufficient to prevent rickets and fractures [112 – 114]. Since the early 1990s, the effect of mineral supplementation of either human milk or formulas has been investigated in several randomized trials with conflicting results, some finding a positive effect of mineral supplementation on bone mineralization [115 – 118] and others, no effect [119 – 122]. In most studies the infants have only been followed up for a period of 4 – 5 weeks or until weighing 1800 g, and sample size may not be sufficient for significance. A randomized blinded dietary intervention study done in Danish preterm infants (gestation  32 weeks) [123] showed

FIGURE 7 Postnatal bone mineral content (BMC) of infants fed an experimental formula fortified with calcium and phosphorus. Infants were of gestational ages 33 – 35 weeks. The experimental group was compared with infants fed routine cow milk formula and with an intrauterine bone mineralization curve. BMC in the experimental group was not different from intrauterine bone mineral curve and was significantly higher than BMC in routinely fed infants at 4 – 6 weeks’ postnatal age. Values are mean  SEM. Reproduced from Steichen et al. [111], with permission. that mineral supplementation during hospitalization does not significantly improve bone mineralization outcome at term if infants are fed 200 ml/kg per day. Phosphorus supplementation of human milk (10 mg P/100 ml), fortified supplementation of human milk with calcium and phosphorus (35 mg Ca/100 ml and 17 mg P/100 ml), or preterm formula (70 mg Ca/100 ml and 35 mg P/100 ml) plus vitamin D supplementation of 800 IU/day resulted in BMC measured at term by DXA that did not differ by group or compared with infants fed only their own mother’s milk. This study is complicated by many dietary subgroups, each with very few subjects, and by only moderate levels of fortification of the preterm formulas. The 18% higher BMC in the preterm formula group (n  16) vs own mother’s milk group (n  21) may not have reached significance because of the small sample size. b. After Hospital Discharge Although postnatal intakes of Ca and P eventually may meet estimates of fetal mineral accretion, a cumulative deficit in mineral accumulation usually occurs by the time of hospital discharge. The consequences of inadequate mineral substrate during the

608

SPECKER, NAMGUNG, AND TSANG

period of rapid skeletal growth in early postnatal life may be poor bone mineralization, reduced linear growth, rickets, or fractures [124,125]. After hospitalization, there may be benefits to providing formula-fed premature infants formulas with higher Ca concentrations than those of routine cow milk-based formulas [126]. The optimal concentrations and length of time needed for such formulas are unknown. In the posthospital follow-up of preterm infants by 52 weeks, infants who receive human milk after discharge have a significantly lower BMC than that of infants receiving term formula, and also a lower serum phosphosrus concentration and higher serum alkaline phosphatase activity throughout the 52 weeks [127]. Infants who switch to term formula have a rapid increment in BMC from 10 to 25 weeks compared with human milk-fed infants. In both groups, however, the median weight and height are below the National Center for Health Statistics fiftieth percentile [128], and even at termequivalent age, both groups of preterm infants have radius BMC well below that of infants born at term (approximately 90 mg/cm). After hospital discharge the catch-up in bone mineralization is achieved between 6 and 12 months of corrected age for infants who switch to term formula. In contrast, infants receiving human milk after discharge fail to achieve catch-up bone mineralization until their second birthday [64]. Because large deficits in bone mass arise before preterm infants are discharged from hospital, an attempt should be made to use posthospital nutritional regimens that favor the accumulation of bone mineral mass, especially during the first months after hospital discharge [129,130].

Because total correction of mineral deficits before discharge from hospital is often not practical, especially in very low birth weight infants, there have been several studies in preterm infants evaluating the effects of a mineralenriched diet on growth, serum biochemical indices, and BMC after discharge home. Preterm infants receiving the enriched formula (70 mg of Ca/100 ml and 35 mg of P/ 100 ml) after hospital discharge had greater gains in weight and increments in length throughout the study period than infants fed standard “term” formula after discharge (Ca and P content not reported) [131]. BMC at the distal third of radius by 3 and 9 months corrected age was significantly greater in infants receiving the enriched diet than those fed standard formula after discharge (Fig. 8) [126]. Raupp et al. [132] showed that 36 preterm infants (birthweight  1800 g) who were fed calcium-and phosphorus-enriched formula (80 mg of Ca/100 ml and 45 mg of P/100 ml) exclusively for 3 months after hospital discharge had a significantly higher radial BMC at 3 months of age, but at 6 months, these differences were less pronounced. Thus, follow-up studies of preterm infants after discharge from hospital indicate that bone mineralization is affected negatively by feeding unfortified human milk and positively by feeding enriched formula diets. The long-term consequences of these studies indicate that catch-up mineralization is possible. Because not all preterm infants may be at nutritional risk at hospital discharge, specific criteria may be needed to indicate which infants need an enriched diet after discharge.

FIGURE 8 Radial bone mineral content (BMC) for preterm infants fed after discharge on standard or enriched formula compared with reference data from infants born at term: (A) at 3 and 9 months corrected postnatal age and (B) data plotted at ages appropriate to body size. Reproduced from Bishop et al. [126], with permission.

609

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

Infants at risk, i.e., those who are unable to consume sufficient volumes of milk ad libitum to satisfy growth needs and those who are noted to have biochemical indicators of nutritional risk (low serum albumin and P concentrations, and increased serum alkaline phosphatase activity), theoretically may benefit from an enriched diet after discharge. Early nutrition may affect later growth, specifically linear growth, in preterm infants. Low radius BMC at 1 year of age was related to the development of rickets and fractures in early infancy [124]. Abnormal biochemical indices of mineral status obtained during the neonatal period in preterm infants (e. g., increased activity of serum alkaline phosphatase) were associated with reduced linear growth in follow-up examination at 9 and 18 months. Long-term consequences for bone growth and development have been examined by Bishop et al. [125] in 54 children who were born preterm and fed either banked donor breast milk or preterm formula as a supplement to mother’s own milk. The BMC at average age 5 years was significantly related to the diet they received in the early weeks postpartum. There was a strong positive relationship between childhood BMC and the percentage intake of mother’s own breast milk for those children who received preterm formula as their dietary supplement, leading to the conclusion that diet in early life has implications for long-term bone growth and mineralization. The authors speculated that early bone mineral depletion in preterm infants fed unsupplemented human milk “programs” them to be conservative with bone mineral and also to reduce the overall rate of growth so that “overmineralization” occurs at a later stage when the intake of bone mineral substrates and other nutrients is normal.

V. BONE MINERAL ACQUISITION IN TERM INFANTS AND CHILDREN The rate of mineral accretion relative to body weight remains rapid in infants when compared with older children, as estimated biochemically [133,134] or by single photon absorptiometry [3,6,7,135]. The estimated content of calcium, phosphorus, and magnesium between birth and 2 years, adapted from Ziegler et al. [48] and Fomon and Nelson [134], is shown in Table 1. These data are based on bone calcium content, with proportional changes in phosphorus and magnesium (higher soft tissue P and Mg contents compared with those in bone); these minerals are adjusted to the fiftieth percentile of the National Center for Health Statistics growth chart for male infants [128]. The major determinants of mineral retention for enterally fed infants, particularly for calcium and magnesium, are the amounts of mineral intake and the rate of mineral absorption. However, differences in the distribution and concentration of minerals (and possibly vitamin D) between human milk and infant formulas may affect mineral absorption and retention.

TABLE 1 Estimated Average Body Content of Calcium Magnesium and Phosphorus from Birth to 2 Yearsa Birth

6 Months

12 Months

24 Months

Body weight (g)

3270

7850

10,150

12,590

Length (cm)

50.50

67.80

76.10

87.60

Calcium (g)

54.45

31.18

37.47

47.36

Magnesium (g)

0.82

1.78

2.97

6.54

Phosphorus (g)

17.13

21.19

26.89

31.37

a

From Koo and Tsang, Building better bones: Calcium, phosphorus and vitamin D. In “Nutrition during Infancy: Principles and Practice” (R. C. Tsang et al., eds.). Digital Educational Publishing, Inc., Cincinnati, 1997, with permission.

Postnatal development of total body and regional bone mineralization during infancy has been investigated by Koo et al. [136] and related to anthropometric measurements and other physiologic variables during the first year. During infancy, average total body BMC increases by 389% and total body bone mineral density increases by 157%. The best determinant of bone mineral status is body weight, which accounts for 97% of total body BMC. Postnatal age and body length jointly add only  1% to the explained variation of these DXA measurements; race, gender, and season do not reach statistical significance. Thus, in healthy infants, body weight is the major predictor of bone mineral status, consistent with previous findings, accounting for over 90% of the variance of total body BMC. These data in clinically healthy subjects reflect the postnatal development in bone mineral and can be used as reference data for this age range (Fig. 9) (Table 2) [136]. Whereas rickets in the preterm infant appears to result from low mineral intake, particularly calcium and phosphorus, rickets in the term infant usually results from vitamin

FIGURE 9

Scatter plot including mean regression line and 95% confidence interval for the mean for infant total body bone mineral content (TB BMC) as a function of different body weights (2 – 14 kg) during infancy. TB BMC (g)  77.24 (25.85)  24.94 (1.96) bare weight (kg)  0.21 (0.03) postnatal age (days)  1.889 (0.62) Length (cm).(r2  0.976). Reproduced from Koo et al. [136], with permission.

610

SPECKER, NAMGUNG, AND TSANG

TABLE 2 Dual-Energy X-Ray Absorptiometric Values for Infant Total Body Bone Mineral Content (TB BMC), Area (TB Area), Bone Mineral Density (TB BMD), and Calcium (TB Ca) at Various Postnatal Age (Days 1 – 390) Intervalsa Postnatal age

TB BMC (g)

Days

n

Mean

1–8

65

9 – 90

16

91 – 150

17

137.1

TB area (cm2)

TB BMD (cm2)

TB Cab (g)

SD

Mean

SD

Mean

SD

Mean

SD

68.2

10.2

308

26.4

0.221

0.017

23.2

3.5

103.4

21.4

431

58.1

0.238

0.022

35.2

7.3

20.0

527

45.4

0.259

0.024

46.6

6.8

151 – 270

12

196.4

26.6

650

64.3

0.302

0.018

66.8

9.0

271 – 390

20

253.2

41.3

754

87.8

0.335

0.029

86.1

14.0

a b

From Koo et al., J. Am. Coll. Nutr. 17, 65 – 70 (1998), with permission. Calculated from TB BMC based on 34% osseous mineral as calcium.

D deficiency and is seen primarily in dark-skinned infants exclusively fed human milk. Increased skin pigmentation is thought to reduce the ability to produce endogenous vitamin D [137], and human milk has low concentrations of the vitamin [138]. The human milk-fed infant therefore depends on sunlight exposure to maintain vitamin D levels within the normal range [139]. Cases of nutritional vitamin D deficiency rickets are still reported [140 – 142], despite national fortification of dairy products with vitamin D.

A. Factors Affecting Mineral Acquisition in Term Infants and in Children 1. DEMOGRAPHIC AND ANTHROPOMETRIC VARIABLES a. Infants Weight and length are strong predictors of areal BMD and BMC during infancy [7,28,30,143]. However, because no studies of volumetric BMD have been done during the first year of life, it is not clear whether true volumetric density changes during this period. Greater areal total body BMD and BMC within the first 18 months of age have been reported in male compared to female infants and black compared to white infants [143]. It is not clear whether these higher areal BMD values reflect differences in true volumetric BMD or in bone size, as discussed earlier. Garn [144] reported a greater bone width for black children than for white children 1–2 years of age, and Gilsanz et al. [145] found no significant racial differences in true volumetric density of the lumbar spine in older children until puberty. Older boys also have been found to have larger bone size than girls, which could provide higher areal BMD values [146]. The recent development of pQCT should allow determination of the age at which gender and ethnic differences in volumetric BMD emerge. b. Children Cross-sectional and longitudinal studies of bone accretion during childhood have been reported [23,147–151]. As observed for infants, weight and height

emerge as strong predictors of areal BMD and BMC. Gender and ethnic differences have also been reported. Boys have higher BMC/BMD than girls, and blacks have higher areal BMD than whites. Significant changes in vertebral volumetric BMD occur during puberty [145,152]. Factors affecting bone acquisition during this period are described in Chapter 25. Molgaard et al. [148] found that peak increases in bone area occur earlier than that for BMC, indicating a delay in the mineralization of newly formed bone. These findings were also observed by Bailey et al. [153], supporting the hypothesis of Blimkie et al. [154] that a lag in cortical bone thickness and mineralization relative to linear skeletal growth may be a predominant risk factor for fractures occurring during adolescence. Rates of increase in bone size and mass differ depending on the skeletal region and pubertal status [151], as well as the gender and race of the child [149]. Many of the studies reporting effects of anthropometric or demographic factors on bone have investigated the specific role of genetics, diet, or physical activity on bone accretion. Several of these are described. 2. GENETICS Results of family studies and the identification of several candidate genes that influence BMD or bone accretion have been described [155,156]. However, few studies have addressed whether genetic factors influence the attainment of high peak bone density by young adulthood. Hopper et al. [157] reported that genetic factors play a major role in explaining the variation in BMD at different bone sites during adolescence and young adulthood. Ferrari et al. [158] also found that age-adjusted BMD Z scores in young girls (average 8 years) correlated with their mothers’ BMD Z scores. Although these studies indicate that genetic influences may act prior to puberty, it is often difficult to distinguish the genetic component from an effect due to a shared or common environment (see Chapter 26). A relationship between growth and the vitamin D receptor (VDR) gene has been reported. One study described a

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

retrospective evaluation of 66 women for whom detailed birth and infant growth records were available [159]. Women with the tt genotype were heavier and tended to have greater spine BMD at 1 year of age than those with TT and Tt genotypes. These authors concluded that early infant environment may interact with an individual’s underlying genotype to program early skeletal growth and that this effect may track through later life and influence adult characteristics. Other investigators have also found the VDR gene to be associated with growth parameters, such as height, during childhood [160]. Differences in BMD by VDR genotype have also been reported in children. Sainz et al. [161] found differences in femoral and vertebral BMD by VDR genotype in prepubertal girls of Mexican descent. Girls with bb genotype, using the Bsm1 polymorphism, were found to have greater femoral cortical BMD and vertebral BMD measured by QCT than girls with the BB genotype. Genetic endowment also influences how a child responds to envirnonmental factors. Ferrari et al. [162] found that prepubertal bb genotype girls had the highest bone accrual but did not respond significantly to supplemental calcium. Girls with the BB and Bb genotypes had less accrual but showed larger increases in BMD in response to supplemental calcium (Fig. 10). This interaction between the VDR gene and diet has also been reported in postmenopausal women [163]. The underlying mechanisms behind these genetic influences are not known. If the genes that influence bone also influence early growth, this may lead to differences in bone size due to increased skeletal loading. However, the studies described previously have not found consistent effects. Using stable isotopes in girls 8 – 15 years of age, O’Brien et al. [164], found that girls from osteoporotic families had altered bone turnover responses to acute changes in calcium intake compared to those in girls from nonosteoporotic families. These findings, along with results indicating an interaction between VDR and dietary intake, indicate that

FIGURE 10 The effect of calcium supplementation (850 mg/day for 12 months) on increases in BMD in prepubertal girls varies depending on VDR genotype. Data represent mean changes in BMD per year averaged for six skeletal sites. Data from Ferrai et al. [162].

611

there likely are genes that influence the skeletal response to multiple environmental factors during childhood. 3. DIET a. Infants Vitamin D and Ca intakes influence bone accretion during infancy and early childhood. Whether changes of vitamin D concentrations in infancy, especially within the normal range, are associated with alterations of bone mineralization is not clear. In 2- to 5-month-old Korean infants, bone mineral content is not reduced, despite relatively low vitamin D status in human milk-fed infants compared with cow milk-based, formula-fed infants [165]. The effects of different diets on infant bone mineralization have been evaluated in several studies using human milk fortified with vitamin D versus human milk alone or human milk versus formula (either soy based or cow milk based). In some studies, a significantly lower BMC has been found in infants fed human milk alone than in infants fed human milk plus vitamin D [166], but others have found no effect of vitamin D supplements on BMC [167,168]. Numerous studies have compared regional bone mineral accretion in infants on different types of feeding regimens [5,16,18,28,169,170], but few have measured total body bone mineral accretion. Infants fed human milk consistently show lower total body BMC or regional BMD than those fed infant formulas. The low phosphorus concentration combined with lower calcium concentration of human milk may be the main factor explaining this difference. Infants fed soy formulas have lower BMC than those fed human milk or cow milk-based formula in some studies [16,169], but not all [15,170]; lack of difference may relate to insufficient power of the studies. Newer formulations of soy formulas have improved calcium and phosphorus contents and may correct for any mineral malabsorptive effect from soy [17]. Specker et al. [28] assessed the effect of varying mineral intakes on total body bone accretion during the first year of life in a partially randomized controlled trial in healthy infants. The study was designed as two phases: the first phase included a group of infants whose mothers elected to breast feed for approximately 6 months and another group included infants whose mothers had chosen to formula feed. Within the formula group, infants were randomized to receive either a low mineral-containing formula or a moderate mineral formula. The second phase of the study involved rerandomizing all infants at 6 months of age, including those fed human milk, to either a moderate mineral or a high mineral formula. Infants who were fed moderate mineral formula during the first 6 months of age had significantly greater bone mass at 3 and 6 months of age than those fed human milk or low mineral formula (Fig. 11). However, during the second 6 months of life, infants who received human milk during the first 6 months and then received moderate or high mineral formula had a greater gain in BMC than those who received either the

612

SPECKER, NAMGUNG, AND TSANG

FIGURE 11

Mean total body BMC at 1 and 6 months of age in infants fed human milk, low mineral formula, or high mineral formula and mean total body BMC at 6 and 12 months of age in the same infants after randomization at 6 months to either moderate or high mineral formula. Infants fed moderate mineral formula had significantly greater bone mass at 6 months of age than infants fed human milk or low mineral formula. There were no significant differences in 12-month total body BMC by type of formula consumed. Data from Specker et al. [28].

low or the moderate mineral formula during the first 6 months. By 12 months of age there were no differences in total body BMC by feeding regime during either the first or the second 6-month interval. These findings indicate that the effect of mineral intake on bone accretion appears to be transient rather than long term. The significance of high bone accretion early in life is not clear, and there is currently no evidence that a high BMD or BMC during this period is either beneficial or clinically important. No studies indicate that a high bone accretion rate early in life leads to greater mineralization in later childhood or adolescence. In contrast, Bishop et al. [171] reported that preterm infants fed human milk and high mineral formula during neonatal life had lower BMC at 5 years than those fed only human milk, with its relatively low mineral content. Those authors suggested that “programming” of infants may occur early and those who have lower intakes develop improved calcium retention later in life. This hypothesis has not been confirmed [172, 173]. In addition, these same authors reported the findings of a larger study that showed that at 8 to 12 years of age, preterm children are shorter, lighter, and have lower bone mass than their peers, but that the lower BMC values are appropriate for their body size. Moreover, they found that, despite large differences in mineral intakes in the neonatal period, early diet did not subsequently affect bone mass [65]. b. Children Various studies have explored the relationship among BMC, BMD, and calcium intake. These

include studies relating adult BMD to recall values for childhood diet, cross-sectional assessments of BMD/BMC and calcium intake during childhood, and randomized trials of calcium during childhood. Several retrospective studies among adults have found a relationship between BMD and recall of childhood calcium or dairy intake [174,175]. Others have reported no significant effect of childhood calcium intake on adult BMD [176,177]. A study by Nieves et al. [178] found that teenage calcium intake was related to hip, but not lumbar spine or forearm BMD. These studies indicate that effects on bone of calcium intake in early life are not consistent among skeletal sites. However, retrospective studies are always difficult to interpret because of problems with long-term recall of dietary intake. It is also possible that individuals who consumed high intakes of calcium or dairy products early in life also consume higher amounts as adults, making it difficult to distinguish early from current effects. Studies that have investigated the association between childhood bone status and calcium intake have also been in conflict. Some find an association at some bone sites [179 – 183], whereas others find none [176,177]. Interpretive difficulties arise from problems with methodology in these studies. The effect of calcium on bone accretion may also have a threshold, and in some of the studies there may be insufficient intake variation to observe a relationship. Several clinical trials of calcium supplementation in young children have been completed [184 – 190]. One of the first, reported by Johnston et al. [184], found a significant difference in the percentage BMD change at the

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

radius following 3 years of calcium administration. Mean calcium intakes in the placebo and calcium groups, respectively, were  900 and 1600 mg/day. This study was conducted in 45 twin pairs, with one twin in each pair randomized to receive placebo and the other to receive supplemental calcium. Although the difference between groups in the percentage BMD increase was statistically significant, mean BMD values after 3 years of supplementation did not differ. No significant differences were observed in BMD changes at lumbar spine or proximal femur. The benefits of calcium supplementation were found to depend on the maturational status of the child. The supplement resulted in significantly increased BMD for twins who were prepubertal throughout the study, but there was no benefit shown for pubertal children. BMD was measured in these children 1 to 2 years after cessation of the trial. At that time, no group differences were observed, indicating that the originally observed effects did not persist beyond the period of supplementation [191,192]. Other investigators have also reported results from intervention trials in which a greater BMD increase was associated with supplemental calcium [185,186,188 – 190]. The apparent benefits associated with supplementation are modest relative to annual gains typically observed at these ages. In 8 to 12-year-old Gambian children, calcium supplementation for 12 months led to a higher BMC compared to that in a placebo-treated group [189]. In 5-year-old Chinese children, BMC was significantly lower in children with a longterm habitual low Ca intake [280 g/day] [183], but after 18 months of calcium supplementation, there was a significant percentage gain in BMC compared with the placebo group [193]. To assess the durability of such an effect, an 18month follow-up study after the withdrawal of calcium supplementation was conducted in 84 children at age 8.5 and 10 years [193]. Dietary calcium intakes during follow-up were similar for the two groups. At the end of the follow-up

FIGURE 12

Total body BMD in prepubertal boys before and after an 8-month exercise trial. Boys in the exercise group had greater increase, in BMD than those in the control group. Data from Bradney et al. [205].

613

period, the achieved differences in lumbar spine BMC had disappeared. Thus, higher percentage gains in bone mineral in childhood by calcium supplementation are reversible, and the accrued benefits may disappear after treatment is withdrawn. However, Bonjour et al. [185] reported persistent significant increases in height and femoral neck BMD in prepubertal girls who received a supplemental food product enriched with milk calcium (850 mg/day) compared to girls receiving the same product without the added calcium. This persistent increase was obtained only in girls who had low initial dietary calcium intakes ( 880 mg/day). In summary, the calcium supplementation trials that have been reported in young children have found modest benefits of calcium that are more pronounced before puberty than after pubertal onset, more pronounced in children with low baseline calcium intakes, and do not persist for very long after the period of supplementation comes to an end. In addition to vitamin D and calcium, high sodium intake may reduce bone accretion, especially if the diet is low in calcium. In rats maintained on a low calcium (0.02%) diet, a high sodium diet (1.8% sodium chloride) caused a significant loss of calcium in bone [194]. In young girls (aged 5 to 17 years), urinary calcium is most strongly associated with urinary sodium excretion and is not affected by calcium intake. Increased sodium intake significantly increases urinary calcium excretion and may lead to reduced calcium retention and bone mineral accretion [195]. No data are available in infants on the role of sodium intake on bone accretion. 4. PHYSICAL ACTIVITY Several study designs have been used to assess the role of physical activity on bone development. The number of cross-sectional studies exceeds that of the more expensive and time-consuming longitudinal observational studies and clinical trials. Both designs have tested hypotheses concerning the role of physical activity during childhood on subsequent adult bone health, as well as the role of activity on bone during childhood (see Chapter 28). One of the first retrospective studies looking at the relationship between adult cortical BMD and early physical activity level was that of Kriska et al. [174]. Distal radius BMD was measured by CT in 223 postmenopausal women who completed a questionnaire on past activity during specific age groupings. Although the earliest age group to be studied was 14 – 21 years, there was a stronger correlation between postmenopausal BMD and activity in earliest age groups than with current activity. Other retrospective studies consistently report associations between adult BMD and early physical activity [175,176,196,197]. Fewer studies have also been done in adults who participated as children in competitive sport [198,199]. These studies found that women who competed in gymnastics around the time of puberty have higher BMD later in life than those who did not participate in that sport.

614 In a case-control study of women with hip fractures and controls, recreational activities during adolescence and early adult life appeared to afford protection against hip fracture [200]. The highest quartile of activity (  4 times/week) was associated with a lower odds ratio relative to the lowest quartile ( 1 time/week). Teenage calcium intake or milk drinking was not related to hip fracture risk. Joakimsen et al. [201] summarized 17 studies on physical activity and hip fracture risk. Although only a small number of these looked at childhood activity levels, the results support that increased activity protects against hip fracture and that this association is present from childhood throughout adult life. Because activity levels appear to track throughout life, it is difficult to determine the age at which increased physical activity may be beneficial to bone. The major limitation to these studies is their reliance on recall data for past activity levels. Two studies employed activity data collected longitudinally during childhood to determine whether adult BMD reflected previous activity levels [202,203]. Information on activity and dietary intake levels was obtained in the Cardiovascular Risk in Young Finns Study Group three times over 11 years in subjects aged from 9 to 18 years [203]. BMD of the lumbar spine and femoral neck was measured when these participants reached 20 – 29 years of age. Physical activity during childhood was a significant predictor of adult BMD. Similar results were obtained in the Amsterdam Growth and Health Study, consisting of 182 males and females who were followed from age 13 to 28 years [202]. Weight-bearing activity and calcium intakes were assessed during three periods: 13 – 17 years, 13 – 21 years, and 13 – 27 years. Only weight-bearing activity was significantly predictive of lumbar spine BMD at age 27. This relationship was found for all three age groupings. Two nonrandomized intervention studies have been conducted with prepubertal children [204,205]. Both found significantly greater increases in total body and regional BMD among children participating in bone-loading activities than in control children. One study was conducted among 8 to 12-year-old boys and involved three 20-min activity sessions each week for 8 months (Fig. 12) [205]. The other was conducted among 9 to 10-year-old girls and involved similar activity for 10 months [204]. These studies confirm a beneficial effect of physical activity on bone accrual in children prior to puberty. 5. DISEASES AFFECTING BONE ACQUISITION Genetic factors, diet, and activity are major predictors of bone acquisition and mineralization. Therefore, pediatric diseases that interfere with normal growth, nutritional adequacy, or spontaneous activity will result in decreased bone growth or mineralization (see Chapter 43). Decreased BMD or altered bone metabolism is observed in many diseases,

SPECKER, NAMGUNG, AND TSANG

including anorexia nervosa [206,207], cystic fibrosis [208], liver disease [209], inflammatory bowel disease [210], cerebral palsy [211], and juvenile rheumatoid arthritis [212,213]. Decreased BMD is also observed among asthmatic children treated with glucocorticosteroids, although the effect depends on the type and dose of steroid that is used [214 – 216] (see Chapter 44). Although many of these diseases exert direct effects on calcium or bone metabolism, it is possible that the decreased physical activity associated with chronic illness may be a contributory factor. Trials with therapeutic agents to prevent bone resorption in many of these diseases are being conducted as a means to improve BMD in these children.

VI. SUMMARY Gains in bone mass during childhood have been suggested to be a primary factor determining whether an individual will develop osteoporosis later in life. Knowledge concerning what factors influence bone acquisition has increased substantially with the advent of densitometric methods that can estimate bone mineral in infants and children with minimal exposure to ionizing radiation. The three major factors currently understood to affect bone acquisition are genetic endowment, diet, and physical activity. Their respective roles in maximizing bone acquisition are likely to be complex. Although not all studies report consistent findings, there is general agreement that adquate calcium intake and moderate physical activity are necessary to optimize bone acquisition during childhood. Interrelationships among these factors with genetics are only beginning to be investigated.

References 1. R. E. Kreipe, Bones of today, bones of tomorrow. AJDC 146, 22 – 25 (1992). 2. R. B. Mazess and J. R. Cameron, Growth of bone in school children: Comparison of radiographic morphometry and photon absorptiometry. Growth 36, 77 – 92 (1972). 3. F. R. Greer, J. Lane, S. Weiner, and R. B. Mazess, An accurate and reproducible absorptiometric techniques for determining bone mineral content in newborn infants. Pediatr. Res. 17, 259 – 262 (1983). 4. J. J. Steichen, B. Kaplan, N. Edwards, and R. C. Tsang, Bone mineral content in full-term infants measured by direct photon absorptiometry. Am. J. Roentgenol. 126, 1283 – 1285 (1976). 5. W. B. Pittard III, K. M. Geddes, S. E. Sutherland, M. C. Miller, and B. W. Hollis, Longitudinal changes in the bone mineral content of term and premature infants. Am. J. Dis. Child. 144, 36 – 40 (1990). 6. B. L. Specker, W. Brazerol, R. C. Tsang, R. Levin, J. Searcy, and J. Steichen, Bone mineral content in children 1 to 6 years of age: Detectable sex differences after 4 years of age. Am. J. Dis. Child. 141, 343 – 344 (1987). 7. J. Y. Li, B. L. Specker, M. L. Ho, and R. C. Tsang, Bone mineral content in black and white children 1 to 6 years of age. AJDC 143, 1346 – 1349 (1989).

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood 8. R. B. Mazess and J. R. Cameron, Growth of bone in school children: Maturation and bone mass. Am J. Phys. Anthropol. 35, 399 – 408 (1971). 9. L. Landin and B. E. Nilsson, Forearm bone mineral content in children. Acta Paediatr. Scand. 70, 919 – 923 (1981). 10. S. L. Hui, C. C. Johnston Jr., and R. B. Mazess, Bone mass in normal children and young adults. Growth 49, 34 – 43 (1985). 11. N. J. Bishop, J. A. dePreister, T. J. Cole, and A. Lucas, Reference values for radial bone width and mineral content using single photon absortiometry in healthy children aged 4 to 10 years. Acta Paediatr. 81, 463 – 468 (1992). 12. R. M. Shore, R. W. Chesney, R. W. Mazess, P. G. Rose, and G. J. Bargman, Osteopenia in juvenile diabetes. Calcif. Tissue Int. 33, 455 – 457 (1981). 13. R. W. Chesney, R. B. Mazess, P. Rose, and D. K. Jax, Effect of prednisone on growth and bone mineral content in childhood glomerular disease. Pediatrics 73, 459 – 466 (1984). 14. G. Chan, L. Mileur, and J. Hansen, Effects of increased calcium and phosphorus formulas and human milk on bone mineralization in preterm infants. J. Pediatr. Gastro. Nutr. 5, 444 – 449 (1986). 15. G. M. Chan, C. C. Robert, D. Folland, and R. Jackson, Growth and bone mineralization of normal breast-fed infants and the effects of lactation on maternal bone mineral status. Am. J. Clin. Nutr. 36, 438 – 443 (1982). 16. J. J. Steichen and R. C. Tsang, Bone mineralization and growth in term infants fed soy-based or cow milk-based formula. J. Pediatr. 110, 687 – 692 (1987). 17. R. R. Bainbridge, F. Mimouni, and R. C. Tsang, Bone mineral content of infants fed soy based formula. J. Pediatr. 113, 205 – 207 (1988). 18. F. R. Greer and A. McCormick, Improved bone mineralization and growth in premature infants fed fortified own mother’s milk. J. Pediatr. 112, 961 – 969 (1988). 19. F. R. Greer, J. E. Searcy, R. S. Levin, J. J. Steichen, P. S. SteichenAsche, and R. C. Tsang, Bone mineral content and serum 25hydroxyvitamin D concentrations in breast-fed infants with and without supplemental vitamin D: One-year follow-up. J. Pediatr. 100, 919 – 922 (1982). 20. H. Sievanen, M. C. Backstrom, A. L. Kuusela, R. S. Ikonen, and M. Maki, Dual energy x-ray absorptiometry of the forearm in preterm and term infants: Evaluation of the methodology. Pediatr. Res. 45, 100 – 105 (1999). 21. W. W. K. Koo, J. Walters, A. J. Bush, R. W. Chesney, S. E. Carlson, Dual energy x-ray absorptiometry studies of bone mineral status in newborn infants. J. Bone. Miner. Res. 11, 997 – 1002 (1996). 22. W. W. K. Koo, L. R. Massom, and J. Walters, Validation of accuracy and precision of dual energy x-ray absorptiometry for infants. J. Bone. Miner. Res. 10, 1111 – 1115 (1995). 23. K. J. Ellis, R. J. Shypailo, A. Hergenroeder, M. Perez, and S. Abrams, Total body calcium and bone mineral content: Comparison of dual-energy x-ray absorptiometry with neutron activation analysis. J. Bone. Miner. Res. 11, 843 – 848 (1996). 24. J. A. Brunton, H. S. Bayler, and S. A. Atkinson, Validation and application of dual-energy x-ray absorptiometry to measure bone mass and body composition in small infants. Am. J. Clin. Nutr. 58, 839 – 845 (1993). 25. P. M. Braillon, B. L. Salle, J. Brunet, F. H. Glorieux, P. D. Delmas, and P. J. Meunier, Dual energy x-ray absorptiometry measurement of bone mineral content in newborns: Validation of a technique. Pediatr. Res. 32, 77 – 80 (1992). 26. H. Tsukahara, M. Sudo, M. Umezaki, Y. Fujii, M. Kuriyama, K. Yamamoto, and Y. Ishii, Measurement of lumbar spine bone mineral density in preterm infants by dual-energy x-ray absorptiometry. Biol. Neonate. 64, 96 – 103 (1993). 27. R. N. Southard, J. D. Morris, M. D. Mahan, J. R. Hayes, M. A. Torch, A. Sommer, and W. B. Zipf, Bone mass in healthy children:

28.

29.

30.

31.

32.

33.

34.

35.

36. 37.

38. 39.

40.

41.

42.

43.

44.

45.

615

Measurement with quantitative DXA. Radiology 179, 735 – 738 (1991). B. L. Specker, A. Beck, H. Kalkwarf, and M. Ho, Randomized trial of varying mineral intake on total body bone mineral accretion during the first year of life. Pediatrics 99, e12 (1997). K. C. Mehta, B. L. Specker, S. Bartholmey, J Giddens, and M. L. Ho, Trial on timing of introduction to solids and food type on infant growth. Pediatrics 102, 569 – 573 (1998). B. L. Specker, L. Mulligan, and M. L. Ho, Longitudinal study of calcium intake, physical activity, and bone mineral content in infants 6 – 18 months of age. J. Bone Miner. Res. 14, 569 – 576 (1999). M. B. Leonard, K. J. Propert, B. S. Zemel, V. A. Stallings, and H. I. Feldman, Discrepancies in pedatric bone mineral density reference data: Potential for misdiagnosis of osteopenia. J. Pediatr. 135, 182 – 188 (1999). C. T. Cowell, P. W. Lu, S. A. Lloyd-Jones, J. N. Briody, J. R. Allen, I. R. J. Humphries, E. Reed, J. Knight, R. Howman-Giles, and K. Gaskin, Volumetric bone mineral density: A potential role in paediatrics. Acta Paediatr. Suppl. 411, 12 – 16 (1995). H. P. J. Kroger, Measurement of bone mass and density in children. In “Paediatric Osteology: New Developments in Diagnostics and Therapy” (E. Schonau, ed.), pp. 103 – 108. Elsevier, Amsterdam (1996). D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches to interpreting projected bone densitometry data. J. Bone Miner. Res 7, 137 – 145 (1992). A. Prentice, T. Parsons, and T. Cole, Uncritical use of bone mineral density in absorptiometry may lead to size related artifacts in the identification of bone mineral determinants. Am. J. Clin. Nutr. 60, 837 – 842 (1994). V. Gilsanz, Bone density in children: A review of the available techniques and indications. Europ. J. Radiol. 26, 177 – 182 (1998). H. Schiessl, J. L. Ferretti, G. Tysarczyk-Niemeyer, and J. Willnecker, Noninvasive bone strength index as analyzed by peripheral quantitative computed tomography (pQCT). In “Paediatric Osteology: New Developments in Diagnostics and Therapy (E. Schonau, ed.), pp 141 – 146 Elsevier, Amsterdam (1996). T. L. Binkley and B. L. Specker, pQCT measurement of bone parameters in young children. J. Clin. Densit. 3, 9 – 14 (2000). M. Jaworski, M. Lebiedowski, R. S. Lorenc, and J. Trempe, Ultrasound bone measurement in pediatric subjects. Calif. Tissue Int. 56, 368 – 371 (1995). M. Jaworski, M. Lebiedowksi, and R. S. Lorenc, Evaluation of ultrasound measurement in physiology and pathology condition in children. In “Paediatric Osteology: New Developments in Diagnostics and Therapy” (E. Schonau, ed.), pp. 169 – 174. Elsevier, Amsterdam (1996). M. Z. Mughal, S. Wilmshurst, K. Ward, J. E. Adams, B. L. Specker, Broadband ultrasound attenuation at the calcaneum in children with cerebral palsy: Relationship to mobility status and dietary calcium intake. In “Paediatric Osteology: Prevention of Osteoporosis — A Paediatric Task? (E. Schonau and V. Matkovic, eds.), pp. 203 – 207. Elsevier, Amsterdam (1998). S. Wilmhurst, K. Ward, J. E. Adams, C. M. Langton, and M. Z. Mughal, Mobility status and bone density in cerebral palsy. Arch. Dis. Child. 75, 164 – 165 (1996). E. M. Widdowson, D. A. T. Southgate, and E. Hey, Fetal growth and body composition. In “Perinatal Nutrition” (B. S. Landblad, ed.), Vol. 4 – 14. Academic Press, New York, (1988). F. Forester, F. Daffos, M. Rainaut, M. Bruneau, and F. Trivin, Blood chemistry of normal human fetuses at midtrimester of pregnancy. Pediatr. Res. 21, 579 (1987). A. J. Shaw, M. Z. Mughal, T. Mohammed, M. J. A. Maresh, and C. P. Sibley, Evidence for active maternofetal transfer of magnesium across the in situ perfused rat placenta. Pedatr Res. 27, 622 (1990).

616 46. J. Stulc and B. Stulcova, Transport of calcium by the placenta of the rat. J. Physiol. (Lond) 371, 1 (1986). 47. E. M. Widdowson and R. A. McCance, The metabolism of calcium, phosphorus, magnesium and strontium. Pediatr. Clin. North Am. 12, 595 – 614 (1945). 48. E. E. Ziegler, A. M. O’Donnell, S. E. Nelson, and S. J. Fomon, Bone composition of the reference fetus. Growth 40, 320 – 341 (1976). 49. J. W. Sparks, Human intrauterine growth and nutrient Semin. Perinatol. 8, 74 – 93 (1984). 50. K. J. Ellis, R. J. Shypailo, and R. J. Schanler, Body composition of infants: Human cadaver studies. In “Human Body Composition: In Vivo Methods, Models and Assessment” (K. J. Ellis and J. D. Eastman, eds.), pp. 147 – 160. Plenum Press, New York (1993). 51. E. M. Widdowson and C. M. Spray, Chemical development in utero. Arch. Dis. Child. 26, 205 – 214 (1951). 52. C. P. Rodda, M. Kubota, J. A. Heath, P. R. Ebeling, J. M. Moseley, A. D. Care, I. W. Caple, and T. J. Martin, Evidence for a novel parathyroid hormone-related protein in fetal lamb parathyroid glands and sheep placenta: Comparisons with a similar protein implicated in humoral hypercalcemia of malignancy. J. Endocrinol. 117, 261 – 271 (1988). 53. A. J. Black, J. Topping, B. Durham, R. G. Farquharson, and W. D. Fraser, A detailed assessment of alteration in bone turnover, calcium homeostasis and bone density in normal pregnancy. J. Bone Miner. Res. 15, 557 – 563 (2000). 54. N. A. Cross, L. S. Hillman, S. H. Allen, G. F. Krause, and N. E. Vieira, Calcium homeostasis and bone metabolism during pregnancy, lactation, and postweaning: A longitudinal study. Am. J. Clin. Nutr. 61, 514 – 523 (1995). 55. J. A. Brunton, H. A. Wiler, and S. A. Atkinson, Improvement in the accuracy of dual energy x-ray absorptiometry for whole body and regional analysis of body composition: Validation using piglets and methodologic considerations in infants. Pediatr. Res. 41, 590 – 596 (1997). 56. J. Rigo, M. DeCurtis, K. Nyamugabo, C. Pieltain, P. Gerard, and J. Senterre, Premature bone. In “Nutrition and Bone Development” (J. P. Bonjour and R. C. Tsang, eds.), Vol. 41, pp. 83 – 97. Lippincott-Raven, Vevey, (1999). 57. W. W. K. Koo and F. M. Mimouni, Calcium and magnesium metabolism in preterm infants. In “Calcium and Magnesium Metabolism in Early Life” (R. C. Tsang, ed.), pp. 55 – 70. CRC Press, Boca Raton, FL (1995). 58. L. L. Prestridge, R. J. Schanler, R. J. Shulman, P. A. Burns, and L. L. Laine, Effect of parenteral calcium and phosphorus therapy on mineral retention and bone mineral content in very low birth weight infants. J. Pediatr. 122, 761 – 768 (1993). 59. J. C. Rowe, D. E. Carey, C. A. Goetz, N. D. Adams, and E. Horak, Effect of high calcium and phosphorus intake on mineral retention in very low birth weight infants chronically treated with furosemide. J. Pediatr. Gastroenterol. Nutr. 9, 206 – 211 (1989). 60. F. Alpay, B. Unay, Y. Narin, R. Akin, N. Inanc, O. Ozcan, and E. Gokcay, Measurement of bone mineral desnity by dual energy x-ray absorptiometry in preterm infants fed human milk or formula. Eur. J. Pediatr. 157, 505 – 507 (1998). 61. I. P. Wauben, S. A. Atkinson, T. L. Grad, J. K. Shah, and B. Paes, Moderate nutrient supplementation of mother’s milk for preterm infants supports adequate bone mass and short-term growth: A randomized, controlled trial. Am. J. Clin. Nutr. 67, 465 – 472 (1998). 62. R. J. Schanler and S. A. Abrams, Postnatal attainment of intrauterine macromineral accretion rates in low birth weight infants fed fortified human milk. J. Pediatr. 126, 441 – 447 (1995). 63. M. C. Backstrom, R. Maki, A. L. Kuusela, H. Sievanen, A. M. Koivisto, M. Koskinen, R. S. Ikonen, and M. Maki, The long-term

SPECKER, NAMGUNG, AND TSANG

64.

65.

66. 67.

68.

69.

70.

71.

72.

73.

74. 75. 76.

77.

78.

79.

80. 81.

82.

effect of early mineral, vitamin D, and breast milk intake on bone mineral status in 9- to 11- year old children born prematurely. J. Pedatr. Gastroenterol. Nutr. 29, 575 – 582 (1999). R. J. Schanler, P. A. Burns, S. A. Abrams, and C. Garza, Bone mineralization outcomes in human milk-fed preterm infants. Pediatr. Res. 31, 583 – 586 (1992). M. S. Fewtrell, A. Prentice, S. C. Jones, N. J. Bishop, D. Stirling, R. Buffenstein, M. Lunt, and T. J. Cole, Bone mineralization and turnover in pretrm infants at 8 - 12 years of age: The effect of early diet. J. Bone Miner. Res. 14, 810 – 820 (1999). L. S. Bowden, C. J. Jones, and S. W. Ryan, Bone mineralisation in expreterm infants aged 8 years. Eur. J. Pediatr. 158, 658 – 661 (1999). R. Namgung, R. C. Tsand, B. L. Specker, R. I. Sierra, and M. L. Ho, Reduced serum osteocalcin and 1,25-dihydroxyvitamin D concentrations and low bone mineral content in small for gestational age infants: Evidence of decreased bone formation rates. J. Pediatr. 122, 269 – 275 (1993). M. Z. Mughal, R. Ross, and R. C. Tsang, Clearance of calcium across in situ perfused placentas in intrauterine growth-retarded rat fetuses. Pediatr. Res. 24, 420 – 422 (1989). R. Namgung, R. C. Tsang, R. I. Sierra, and M. L. Ho, Normal serum indices of bone collagen biosynthesis and degradation in small for gestational age infants. J. Pediatr. Gastroenterol. Nutr. 23, 224 – 228 (1996). G. R. Ambler and P. D. Gluckman, “Postnatal growth: An overview. In “Perinatal and Pediatric Pathophysology: A. Clinical Perspective” (P. D. Gluckman and M. A. Heymann, eds.), pp. 170 – 174. Hodder and Stoughton Limited, London (1992). R. Namgung, F. Mimouni, B. N. Campaigne, M. L. Ho, and R. C. Tsang, Low bone mineral content in summer compared with winter-born infants. J. Pediatr. Gastroenterol. Nutr. 15, 285 – 288 (1992). R. Namgung, R. C. Tsang, B. L. Specker, R. I. Sierra, and M. L. Ho, Low bone mineral content and high serum ostocalcin and 1,25-dihydroxyvitamin D in summer- versus winter-born newborn infants: An early fetal effect? J. Pediatr. Gastroenterol. Nutr. 19, 220 – 227 (1994). R. Namgung, R. C. Tsang, C. Lee, D. G. Han, M. L. Ho, and R. I. Sierra, Low total body bone mineral content and high bone resorption in Korean winter-born versus summer-born newborn infants. J. Pediatr 132 (1998). H. Zhou, Rickets in China. In “Rickets” (F. H. Glorieux, ed.), p. 253. Raven Press, New York (1991). J. G. B. Russell and L. F. Hill, True fetal rickets. Br. Radiol. 47, 732 (1974). M. Moncrief and T. O. Fadahunsi, Congenital rickets due to maternal vitamin D deficiency. Arch. Dis. Child. 49, 810 – 811 (1974). S. Reif, Y. Katzir, Z. Eisenberg, and Y. Weisman, Serum 25-hydroxyvitamin D levels in congenital craniotabes. Acta. Paediatr. Scand. 77, 167 – 168 (1988). B. Specker, M. Ho, T. Yin, Q. Shui, X. Chen, and R. Tsang. Prospective study of vitamin D supplementation and rickets in China. J. Pediatr. 120, 733 – 739 (1992). A. Hanukoglu, S. Chlen, and A. Kowarski, Late onset hypocalcemia, rickets and hypoparathyroidism in an infant of a mother with hyperparathyroidism. J. Pediatr. 112, 751 (1988). C. Lamberg-Allardt, Is there a role for vitamin D in osteoporosis? Calcif. Tissue Int. 49, S46 – S49 (1991). C. I. Lamm, K. I. Norton, R. J. C. Murphy, I. A. Wilkins, and J. G. Rabinowitz, Congential rickets associated with magnesium sulfate infusion for tocolysis. J. Pediatr. 113, 1078 – 1082 (1988). R. J. Schanler, Jr. L. G. Smith, and P. A. Burns, Effects of long-term maternal intravenous magnesium sulfate therapy on neonatal calcium metabolism and bone mineral content. Gynecol. Obstet. Invest. 43, 236 – 241 (1997).

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood 83. F. Mimouni, J. J. Steichen, R. C. Tsang, V. Hertzberg, and M. Miodovnik, Decreased bone mineral content in infnats of diabetic mothers. Am. J. Perinatol. 5, 339 – 343 (1988). 84. S. Demarini, B. L. Specker, R. I. Sierra, M. Miodovnik, and R. C. Tsang, Evidence of increased intrauterine bone resorption in term infants of mothers with insulin-dependent diabetes. J. Pediatr. 126, 796 – 798 (1995). 85. J. J. Steichen, R. C. Tsang, M. Ho, et al., Perinatal magnesium, calcium and 1,25-dihydroxyvitamin D in relation to propsective randomized management of maternal diabetes. Calcif. Tissue Int. 33, 317 (1981). 86. K. C. Mehta, H. J. Kalkwarf, F. Mimouni, J. Khoury, and R. C. Tsang, Randomized trial of magnesium administration to prevent hypocalcemia in infants of diabetic mothers. J. Perinatol. 18, 352 – 356 (1998). 87. J. Y. Uriu-Hare, J. S. Stern, G. M. Reaven, and C. L. Keen, The effect of maternal diabetes on trace element status and fetal development in the rat. Diabetes 34, 1031 – 1040 (1985). 88. J. Verhaeghe, R. Bouillon, B. L. Nyomba, W. Lissens, and F. A. Van Assche, Vitamin D and bone mineral homeostasis during pregnancy in the diabetic BB rat. Endocrinology 118, 1019 – 1025 (1986). 89. J. Demignon and C. Rebut- Bonneton, Effects of experimental diabetes of the vitamin D metabolism of pregnant rats and their fetuses. Clacif. Tissue Int. 42, 127 – 135 (1988). 90. J. Verhaeghe, R. Bouillon, W. Lissens, W. J. Visser, and F. A. Van Assche, Diabetes and low Ca-P diet have opposite effects on adult and fetal bone mineral metabolism. Am. J. Physiol. 254, E496 – 504 (1988). 91. S. M. Husain, T. J. Birdsey, J. D. Glazier, M. Z. Mughal, H. O. Garland, and C. P. Sibley, Effect of diabetes mellitus on maternofetal flux of calcium and magnesium and calbindin9K mRNA expression in rat placenta. Pediatr. Res. 35, 376 – 381 (1994). 92. K. Keiver, L. Ellis, A. Anzarut, and J. Weinberg, Effect on prenatal ethanol exposure on fetal calcium metabolism. Alcoh. Clin. Exp. Res. 21, 1612 – 1618 (1997). 93. M. J. Seller and K. S. Bnait, Effects of tobacco smoke inhalation on the developing mouse embryo and fetus. Repro. Toxicol. 9, 449 – 459 (1995). 94. G. Jones, M. Riley, and T. Dwyer, Maternal smoking during pregnancy, growth, and bone mass in prepubertal children. J. Bone Miner. Res. 14, 146 – 151 (1999). 95. T. Nakamoto, S. Grant, and M. Yazdani, The effects of maternal caffeine intake during pregnancy on mineral contents of fetal rat bone. Res. Exp. Med. (Berl). 189, 275 – 280 (1989). 96. W. W. K. Koo, J. C. Walters, J. Esterlitz, R. J. Levine, A. J. Bush, and B. Sibai, Maternal calcium supplementation and fetal bone mineralization. Obstet. Gynecol. 94, 577 – 582 (1999). 97. M. S. Lima, F. Kallfelz, L. Krook, and P. W. Nathanielsz, Humeral skeletal development and plasma constituent changes in fetuses of ewes maintained on a low calcium diet from 60 days of gestation. Calcif. Tissue Int. 52, 283 – 290 (1993). 98. C. N. Corbellini, L. Krook, P. W. Nathanielsz, F. A. Kallfelz, Osteochondrosis in fetuses of ewes overfed calcium. Calcif. Tissue Int. 48, 37 – 45 (1991). 99. M. E. Shackelford, T. F. Collins, T. N. Black, M. J. Ames, S. Dolan, N. S. Sheikh, R. K. Chi, and M. W. O’Donnell, Mineral interaction in rats fed AIN-76A diets with excess calcium. Food Chem. Toxicol. 32, 255 – 263 (1994). 100. M. W. J. Davies and M. J. Haddaway, Bone mineral content and density in healthy subjects and in osteogenesis imperfecta. Arch. Dis. Child. 70, 331 – 334 (1994). 101. L. E. Zionts, J. P. Nash, R. Rude, T. Ross, and N. S. Stott, Bone mineral density in children with mild osteogenesis imperfecta. J. Bone Joint Surg. Br. 77, 143 – 147 (1995). 102. J. I. Rodriquez, A. Garcia-Alix, J. Palacios, and R. Paniagua, Changes in the long bones due to fetal immobility caused by neuromuscular disease. J. Bone Joint Surg. 70A, 1052 – 1060 (1988).

617

103. J. Palacios and J. I. Rodriguez, Extrinsic fetal akinesia and skeletal development: A study in oligohydramnios sequence. Teratology 42, 1 – 5 (1990). 104. J. Palacios, J. I. Rodriguez, A. Ruiz, M. Sanchez, I. Alvarez, and E. DeMiguel, Long bone development in extrinsic fetal akinesia: An experimental study in rat fetuses subjected to oligohydramnios. Teratology 46, 79 – 84 (1992). 104a. J. I. Rodriquez, J. Palacios, M. Ruiz, I. Sanchez, and A. E. DeMiguel, Morphological changes in long bone development in fetal akinesia deformation sequence: An experimental study in curarized rat fetuses. Teratology 45, 213 – 221 (1992). 105. F. Bronner, B. L. Salle, G. Putut, J. Rigo, and J. Senterre, Net calcium absorption in premature infants: Results of 103 metabolic balance studies. Am. J. Clin. Nutr. 56, 1037 – 1044 (1992). 106. J. R. James, P. J. Congdon, J. Truscott, A. Horsman, and R. Arthur, Osteopenia of prematurity. Arch. Dis. Child. 61, 871 – 876 (1986). 107. A. Horsman, S. W. Ryan, P. J. Congdon, J. G. Truscott, and J. R. James, Osteopenia in extremely low birth weight infants. Arch. Dis. Child. 64, 485 – 488 (1989). 108. J. Rigo, M. De Curtis, C. Pieltain, K. Nyamugabo, and J. Senterre, Bone mineral density index (BMDI) determined by whole body dual energy x-ray absorptiometry (DEXA) in IDM, IUGR and preterm infants: Comparison to intrauterine reference values. Pediatr. Res. 40, A548 (1996). 109. P. Lemons, M. Stuart, and J. A. Lemons, Breast-feeding the premature infant. Clin. Perinatol. 13, 111 – 122 (1986). 110. G. H. Anderson, Human milk feeding. Pediatr. Clin. North Am. 32, 335 – 353 (1985). 111. J. J. Steichen, T. L. Cratton, and R. C. Tsang, Osteopenia of prematurity: The cause and possible treatment. J. Pediatr. 96, 528 – 533 (1980). 112. W. W. K. Koo, A. E. Oestreich, R. Sherman, D. Buckley, R. C. Tsang, and J. J. Steichen, Failure of high calcium and phosphorus fortification in the prevention of rickets of prematurity. Am. J. Dis. Child. 140, 857 – 858 (1986). 113. W. W. K. Koo, R. Sherman, P. Succop, S. Krug-Wispe, R. C. Tsang, J. J. Steichen, A. H. Crawford, and A. E. Oestreich, Fractures and rickets in very low birth weight infants: Conservative management and outcome. J. Pediatr. Orthop. 9, 326 – 330 (1989). 114. R. J. Cooke, Rickets in a very low birth weight infant. J. Pediatr. Gastroenterol. Nutr. 9, 397 – 399 (1989). 115. G. M. Chan, L. Mileur, and J. W. Hansen, Calcium and phosphorus requirements in bone mineralization of preterm infants. J. Pediatr. 113, 225 – 229 (1988). 116. A. Horsman, S. W. Ryan, P. J. Congdon, J. G. Truscott, and M. Simpson, Bone mineral accretion rate and calcium intake in preterm infants. Arch. Dis Child. 64, 910 – 918 (1989). 117. P. C. Holland, A. R. Wilkinsnson, J. Diez, and D. R. Lindsell, Prenatal deficiency of phosphate, phosphate supplementation and rickets in very-low-birth weight infants. Lancet 335, 697 – 701 (1990). 118. A. A. Lapillonne, F. H. Glorieux, B. L. Salle, P. M. Braillon, M. Chambon, J. Rigo, G. Putet, and J. Senterre, Mineral balance and whole body bone mineral content in very low-birth-weight infants. Acta Paediatr. Suppl. 405, 117 – 122 (1994). 119. H. D. Modanlou, M. O. Lim, J. W. Hansen, and V. Sickles, Growth, biochemical status and mineral metabolism in very-low-birth weight infants receiving fortified preterm human milk. J. Pediatr. Gastroenterol. Nutr. 5, 762 – 767 (1986). 120. S. J. Gross, Bone mineralization in preterm infants fed human milk with and without mineral supplementation. J. Pediatr. 111, 450 – 458 (1987). 121. F. R. Greer, Determination of radial bone mineral content in low birth weight infants by photon absorptiometry. J. Pediatr. 113, 213 – 219 (1988).

618 122. P. S. Venkataraman and K. E. Blick, Effect of mineral supplementation of human milk on bone mineral content and trace element metabolism. J. Pediatr. 113, 220 – 226 (1988). 123. J. Faerk, S. Petersen, B. Peitersen, and K. F. Michaelsen, Diet and bone mineral content at term in premature infants. Pediatr. Res. 47, 148 – 156 (2000). 124. W. W. K. Koo, R. Sherman, P. Succop, A. E. Oestreich, R. C. Tsang, S. K. Krug-Wispe, and J. J. Steichen, Sequential bone mineral content in small preterm infants with and without fractures and rickets. J. Bone Miner. Res. 3, 193 – 197 (1988). 125. N. J. Bishop, S. L. Dahlenburg, M. S. Fewtrell, R. Morley, and A. Lucas, Early diet of preterm infants and bone mineralization at age 5 years. Acta. Paediatr. 85, 230 – 236 (1996). 126. N. J. Bishop, F. J. King, and A. Lucas, (1993) Increased bone mineral content of preterm infants fed with a nutrient enriched formula after discharge from hospital. Arch. Dis. Child. 68, 573 – 578 (1993). 127. S. A. Abrams, R. J. Schanler, R. C. Tsang, and C. Graza, Bone mineralization in former very low birth weight infants fed either human milk or commercial formula: One year follow-up observation J. Pediatr. 114, 1041 – 1044 (1989). 128. P. V. V. Hamill, T. A. Drizd, S. L. Johnson, R. B. Reed, A. F. Roche, and W. M. Moore, Physical growth: National Center for Health Statistics percentiles. Am. J. Clin. Nutr. 32, 607 – 629 (1997). 129. P. J. Congdon, A. Horsman, S. W. Ryan, J. G. Truscott, and H. Durward, Spontaneous resolution of bone mineral depletion in preterm infants. Arch. Dis. Child. 65, 1038 – 1042 (1990). 130. A. Horsman, S. W. Ryan, P. J. Congdon, J. G. Truscott, and M. Simpson, Bone mineral content and body size 65 to 100 weeks postconception in preterm and full term infants. Arch. Dis. Child. 64, 1579 – 1586 (1989). 131. A. Lucas, N. J. Bishop, F. J. King, and T. J. Cole, Randomized trial of nutrition for preterm infants after discharge. Arch. Dis. Child. 67, 324 – 327 (1992). 132. P. Raupp, G. Poss, R. von Kries, and E. Schmidt, Effect of a calcium and phosphorus-enriched formula on bone mineralization and bone growth in preterm infants after discharge from hospital. Ann. Nutr. Metabol. 41, 358 – 364 (1997). 133. S. J. Fomon, F. Haschke, E. E. Ziegler, and S. E. Nelson, Body composition of reference children from birth to age 10 years. Am. J. Clin. Nutr. 35, 1169 – 1175 (1982). 134. S. J. Fomon and S. E. Nelson, Calcium, phosphorus, magnesium, and sulfate. In “Nutrition of Normal Infant” (S. J. Fomon, ed.), Mosby, St. Louis (1993). 135. S. D. Minton, J. J. Steichen, and R. C. Tsang, Bone mineral content in term and preterm appropriate-for-gestational-age infants. J. Pediatr. 95, 1037 – 1042 (1979). 136. W. W. K. Koo, A. J. Bush, J. Walters, and S. E. Carlson, Postnatal development of bone mineral status during infancy. J. Am. Coll. Nutr. 17, 65 – 70 (1998). 137. M. F. Holick, J. A. MacLaughlin, and S. H. Doppert, Regulation of cutaneous previtamin D3 photosynthesis in man: Skin pigment is not an essential regulator. Science 211, 590 – 593 (1981). 138. B. L. Specker, R. C. Tsang, and B. W. Hollis, Effect of race and diet on human milk vitamin D and 25-hydroxyvitamin D. Am. J. Dis. Child. 139, 1134 – 1137 (1985). 139. B. Specker, B. Valanis, V. Hertzberg, N. Edwards, and R. T.sang, Sunshine exposure and serum 25-hydroxyvitamin D concentrations in exclusively breast-fed infants. and J. Pediatr. 107, 372 (1985). 140. M. T. Pugliese, D. L. Blujberg, J. Hludzinski, and S. Kay, Nutritional rickets in suburbia. J. Am. Coll. Nutr. 17, 637 – 641 (1998). 141. B. D. Gessner, E. DeSchweintiz, K. M. Peterson, and C. Lewandowski, Nutritional rickets among breast-fed black and Alaska Native children. Am. J. Dis. Child. 145, 127 – 130 (1997).

SPECKER, NAMGUNG, AND TSANG 142. M. Z. Mughal, H. Salama, T. Greenaway, I. Laing, and E. B. Mawer, Lesson of the week: Florid rickets associated with prolonged breast-feeding without vitamin D supplementation. Br. Med. J. 318, 39 – 40 (1999). 143. R. C. Rupich, B. L. Specker, N. Lieuw-A-Fam, and M. Ho, Gender and race differences in bone mass during infancy. Calcif. Tissue Int. 58, 395 – 397 (1996). 144. S. M. Garn, Lifelong black-white differences in bone size and cortical area (letter). AJDC 144, 750 – 751 (1990). 145. V. Gilsanz, T. F. Roe, S. Mora, G. Costin, and W. G. Goodman, Changes in vertebral bone density in black girls and white girls during childhood and puberty. N. Engl. J. Med. 325, 1597 – 1600 (1991). 146. V. Gilsanz, M. I. Boechat, T. F. Roe, M. L. Loro, J. W. Sayre, and W. G. Goodman, Gender differences in vertebral body sizes in children and adolescents. Radiology 190, 673 – 677 (1994). 147. M. Moro, M. C. H. VanDerMeulen, B. J. Kiratli, R. Marcus, L. K. Bachrach, and D. R. Carter, Body mass is the primary determinant of midfemoral bone acquisition during adolescent growth. Bone 19, 519 – 526 (1996). 148. C. Molgaard, B. L. Thomsen, and K. F. Michaelsen, Whole body bone mineral accretion in healthy children and adolescents. Arch. Dis. Child. 81, 10 – 15 (1999). 149. L. K. Bachrach, T. Hastie, M. C. Wang, B. Narasimhan, and R. Marcus, Bone mineral acquisition in healthy Asian, Hispanic, black, and Caucasian youth: A longitudinal study. J. Clin. Endocrinol. Metab. 84, 4702 – 4712 (1999). 150. D. A. Nelson, P. M. Simpson, C. C. Johnson, D. A. Barondess, and M. Kleerekoper, The accumulation of whole body skeletal mass in third- and fourth-grade children: Effects of age, gender, ethnicity, and body composition. Bone 20, 73 – 78 (1997). 151. S. Bass, P. D. Delmas, G. Pearce, E. Hendrich, A. Tabensky, and E. Seeman, The differing tempo of growth in bone size, mass, and density in girls is region-specific. J. Clin. Invest. 104, 795 –804 (1999). 152. V. Gilsanz, D. Gibbens, T. Roe, M. Carlson, M. Senac, M. Ines Boechat, H. Huang, E. Schulz, C. Libanati, and C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). 153. D. A. Bailey, H. A. McKay, R. L. Mirwald, P. R. E. Crocker, and R. A. Faulkner, A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: The University of Saskatchewan bone mineral accrual study. J. Bone Miner. Res. 14, 1672 – (1679) (1999). 154. C. J. R. Blimkie, J. Lefevre, G. P. Beunen, R. Renson, J. Dequeker, and P. Van Damme, Fractures, physical activity, and growth velocity in adeolescent Belgian boys. Med. Sci. Sports Exerc. 25, 801 – 808 (1993). 155. W. S. Pollitzer and J. J. B. Anderson, Ethnic and genetic differences in bone mass: A review with a hereditary vs environmental perspective. Am. J. Clin. Nutr. 50, 1244 – 1259 (1989). 156. T. V. Nguyen, G. M. Howard, P. J. Kelly, and J. A. Eisman, Bone mass, lean mass and fat mass: Same genes or same environments? Am. J. Epidemiol. 147, 3 – 16 (1998). 157. J. L. Hopper, R. M. Green, C. A. Nowson, D. Young, A. J. Sherwin, B. Kaymakci, R. G. Larkins, and J. D. Wark, Genetic, common environment, and individual specific components of variance for bone mineral density in 10- to 26-year old females: A twin study. Am. J. Epidemiol. 147, 17 – 29 (1998). 158. S. Ferrari, R. Rizzoli, D. Slosman, and J. P. Bonjour, Familial resemblance for bone mineral mass is expressed before puberty. J. Clin. Endocrinol. Metab. 83, 358 – 361 (1998). 159. R. W. Keen, P. Egger, P. J. Major, T. D. Spector, and C. Cooper, Polymorphisms of the vitamin D receptor, inant growth and adult bone mass. J. Bone Miner. Res. 10, S189 (1995). 160. K. Minamitani, Y. Takahashi, M. Minagawa, T. Yasuda, and H. Niimi, Difference in height associated with a translation start site

CHAPTER 24 Bone Mineral Acquisition in Utero, during Infancy, and throughout Childhood

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

polymorphism in the vitamin D receptor gene. Pediatr. Res. 44, 628 – 632 (1998). J. Sainz, J. M. Van Tornout, M. L. Loro, J. Sayre, T. F. Bore, and V. Gilsanz, Vitamin D receptor gene polymorphisms and bone density in prepubertal American girls of Mexican descent. N. Engl. J. Med. 337, 77 – 82 (1997). S. L. Ferrari, R. Rizzoli, D. O. Slosman, and J. P. Bonjour, Do dietary calcium and age explain the controversy surrounding the relationship between bone mineral density and vitamin D receptor gene polymorphisms? J. Bone Miner. Res. 13, 363 – 370 (1998). E. A. Krall, P. Parry, J. B. Lichter, and B. Dawson-Hughes, Vitamin D receptor alleles and rates of bone loss: Influences of years since menopause and calcium intake. JBMR 10, 978 – 984 (1995). K. O. O’Brien, S. A. Abrams, L. K. Liang, K. J. Ellis, and R. F. Gagel, Bone turnover response to changes in calcium intakes is altered in girls and adult women in families with histories of osteoporosis. J. Bone Miner. Res. 13, 491 – 499 (1998). M. J. Park, R. Namgung, D. H. Kim, and R. C. Tsang, (1998) Bone mineral content is not reduced despite low vitamin D status in breast milk-fed infants versus cow’s milk-fed infants. J. Pediatr. 132, 641 – 645 (1998). F. R. Greer, J. F. Searcy, R. S. Levin, J. J. Steichen, P. SteichenAsch, and R. C. Tsang, Bone mineral content and serum 25-hydroxyvitamin D concenraion in breast-fed infants with and without supplemental vitamin D. J. Pediatr. 98, 696 – 701 (1981). C. C. Roberts, G. M. Chan, D. Follad, C. Rayburn, and R. Jackson, Adequate bone mineralization in breast fed infants. J. Pediatr. 99, 192 – 196 (1981). F. Greer, and S. Marshall, Bone mineral content, serum vitamin D metabolite concentrations and ultraviolet B light exposure in infants fed human milk with and without vitamin D2 supplements. J. Pediatr. 114, 204 – 211 (1989). G. M. Chan, L. Leeper, and L. S. Book, Effects of soy formulas on mineral metabolism in term infants. AJDC 141, 527 – 530 (1987). L. S. Hillman, Bone mineral content in term infants fed human milk, cow milk-based formula, or soy-based formula. J. Pediatr. 113, 208 – 212 (1988). N. J. Bishop, S. L. Dahlenburg, M. S. Fewtrell, R. Morley, and A. Lucas, Early diet of preterm infants and bone mineralization at age five years. Acta Paediatr. 85, 230 – 236 (1996). A. Rubinacci, P. Sirtori, G. Moror, L. Galli, I. Minoli, and L. Tessari, Is there an impact of birth weight and early life nutrition on bone mineral content in preterm infants and children? Acta Paediatr. 82, 711 – 713 (1993). I. Helin, L. A. Landin, and B. E. Nilsson, Bone mineral content in preterm infants at age 4 to 16. Acta Paediatr. Scand. 74, 264 – 267 (1985). A. M. Kriska, R. B. Sandler, J. A. Dauley, R. E. LaPorte, D. L. Hom, and G. Pambianco, The assessment of historical physical activity and its relation to adult bone parameters. Am. J. Epidemiol. 127, 1053 – 1063 (1988). C. Cooper, M. Cawley, A. Bhalla, P. Egger, F. Ring, L. Morton, and D. Barker, Childhood growth, physical activity, and peak bone mass in women. JBMR 10, 940 – 947 (1995). R. G. McCulloch, D. A. Bailey, C. S. Houston, and B. L. Dodd, Effects of physical activity, dietary calcium intake and selected lifestyle factors on bone density in young women. Can. Med. Assoc. J. 142, 221 – 227 (1990). M. F. Q. VandenBergh, S. A. DeMan, C. M. Witteman, A. Hofman, W. T. Trouerbach, and D. E. Grobbee, Physical activity, calcium intake, and bone mineral content in children in The Netherlands. J. Epidemiol. Community Health 49, 299 – 304 (1995). J. W. Nieves, A. L. Golden, E. Siris, J. L. Kelsey, and R. Lindsay, Teenage and current calcium intake are related to bone mineral den-

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

193a.

619

sity of the hip and forearm in women aged 39 – 39 years. Am. J. Epidemiol. 141, 342 – 351 (1995). K. Rubin, V. Schirduan, P. Gendreau, M. Sarfarazi, R. Mendola, and G. Dalsky, Predictors of axial and peripheral bone mineral density in healthy children and adolescents, with special attention to the role of puberty. J. Pediatr. 123, 863 – 870 (1993). J. C. Ruiz, C. Mandel, and M. Garabedian, Influence of spontaneous calcium intake and physical exercise on the vertebral and femoral bone mineral density of children and adolescents. J. Bone Miner. Res. 10, 675 – 682 (1995). A. M. Boot, M. A. J. de Ridder, A. P. Pols, E. P. Krenning, and S. M. P. F. de Munick Keizer-Schrama, Bone mineral density in children and adolescents: Relation to puberty, calcium intake, and physical activity. J. Clin. Endocrinol. Metab. 82,, 57 – 62 (1997). J. M. Sentipal, G. M. Wardlaw, J. Mahan, and V. Matkovic, Influence of calcium intake and growth indexes on vertebral bone mineral density in young females. Am. J. Clin. Nutr. 54, 425 – 428 (1991). W. T. K. Lee, S. S. F. Leung, and S. S. H. Lui, Relationship between long-term calcium intake and bone mineral content of children from birth to 5 years. Br. J. Nutr. 70, 235 – 248 (1993). C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). J. P. Bonjour, A. L. Carrie, S. Ferrari, H. Clavien, D. Slosman, G. Theintz, and R. Rizzoli, Calcium-enriched foods and bone mass growth in prepubertal girls: A randomized, double-blind, placebocontrolled trial. J. Clin. Invest. 99, 1287 – 1294 (1997). W. T. K. Lee, S. S. F. Leung, D. M. Y. Leung, H. S. Y. Tsang, J. Lau, and J. C. Y. Cheng, A randomized double-blind controlled calcium supplementation trial, and bone and height acquisition in children. Br. J. Nutr. 74, 125 – 139 (1995). W. T. K. Lee, S. S. F. Leung, S. H. Wang, Y. C. Xu, W. P. Zeng, J. Lau, S. J. Oppenheimer, and J. C. Y. Cheng, Double-blind, controlled calcium supplementation and bone mineral accretion in children accustomed to a low-calcium diet. Am. J. Clin. Nutr. 60, 744 – 750 (1994). J. Cadogan, R. Eastell, N. Jones, and M. E. Barker, Milk intake and bone mineral acquisition in adolescent girls: Randomised, controlled intervention trial. Br. Med. J. 815, 1255 – 1260 (1997). B. Dibba, A. Prentice, M. Ceesay, D. M. Stirling, T. J. Cole, and E. M. E. Poskitt, Effect of calcium supplementation on bone mineral accretion in Gambian children accustomed to a low-calcium diet. Am. J. Clin. Nutr. 71, 544 – 549 (2000). T. Lloyd, M. B. Andon, N. Rollings, J. K. Martel, J. R. Landis, L. M. Demers, D. F. Eggli, K. Kieselhorst, and H. E. Kulin, Calcium supplementation and bone mineral density in adolescent girls. JAMA 270, 841 – 844 (1993). C. W. Slemenda, M. Pacock, S. Hui, L. Zhou, and C. C. Johnston, Reduced rates of skeletal remodeling are associated with increased bone mineral density during the development of peak skeletal mass. J. Bone Miner. Res. 12, 676 – 682 (1997). C. W. Slemenda, T. K. Reister, M. Peacock, and C. C. Johnston, Bone growth in children following cessation of calcium supplementation. J. Bone Miner. Res. 8, S154 (1993). W. T. Lee, S. S. Leung, M. Y. Ng, S. F. Wang, Y. C. Xu, W. P. Zeng, J. Lau, S. J. Oppenheimer, and J. C. Cheng, Double-blind, controlled calcium supplementation and bone mineral accretion in children accustomed to a low-calcium diet. Am. J. Clin. Nutr. 60, 744 – 750. W. T. K. Lee, S. S. F. Leung, D. M. Y. Leung, and J. C. Y. Cheng, A follow-up study on the effects of calcium-supplement withdrawal and puberty on bone acquisition of children. Am. J. Clin. Nutr. 64, 71 – 77 (1996).

620 194. A. Y. Chan, P. Poon, E. L. Chan, S. L. Fung, R. Swaminathan, The effect of high sodium intake on bone mineral content in rats fed a normal calcium or a low calcium diet. Osteopor. Int. 3, 341 – 344 (1993). 195. K. O. O’Brien, S. A. Abrams, J. E. Stuff, L. K. Liang, and T. R. Welch, Variables related to urinary calcium excretion in young girls. J. Pediatr. Gastroenterol. Nutr. 23, 8 – 12 (1996). 196. D. Teegarden, W. R. Proulx, M. Kern, D. Sedlock, C. M. Weaver, C. C. Johnston, and R. M. Lyle, Previous physical activity relates to bone mineral measures in young women. Med. Sci. Sports Exerc. 28, 105 – 113 (1996). 197. E. Puntila, H. Kroger, T. Lakka, R. Honkanen, and M. Tuppurainen, Physical activity in adolescence and bone density in peri and postmenopausal women: A population based study. Bone 21, 363 – 367 (1997). 198. E. M. Kirchner, R. D. Lewis, and P. J. O’Connor, Effect of past gymnastics participation on adult bone mass. J. Appl. Physiol. 80, 226 – 232 (1996). 199. S. Bass, G. Pearce, M. Bradney, E. Hendrich, P. D. Delmas, A. Harding, and E. Seeman, Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepubertal and retired female gymnasts. J. Bone Miner. Res. 13, 500 – 507 (1998). 200. J. W. Nieves, J. A. Grisso, and J. L. Kelsey, A case-control study of hip fracture: Evaluation of selected dietary variables and teenage physical activity. Osteopor. Int. 2, 122 – 127 (1992). 201. R. M. Joakimsen, J. H. Magnus, and V. Foonebo, Physical activity and predisposition for hip fractures: A review. Osteopor. Int. 7, 503 – 513 (1997). 202. D. C. Welten, H. C. G. Kemper, G. B. Post, W. Van Mechelen, J. Twisk, P. Lips, and G. J. Teule, Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 9, 1089 – 1096 (1994). 203. M. J. Valimaki, M. Karkkainen, C. Lamberg-Allardt, K. Laitinen, E. Alhava, J. Heikkinen, O. Impivaara, P. Makela, J. Palmgren, R. Seppanen, and I. Vuori, Exercise, smoking, and calcium intake during adolescence and early adulthood as determinants of peak bone mass. Br. Med. J. 309, 230 – 231 (1994). 204. F. L. Morris, G. A. Naughton, J. L. Gibbs, J. S. Carlson, and J. D. Wark, Prospective ten-month exercise intervention in premenarcheal girls: Positive effects on bone and lean mass. J. Bone Miner. Res. 12, 1453 – 1462 (1997). 205. M. Bradney, G. Pearce, G. Naughton, C. Sullivan, S. Bass, T. Beck, J. Carlson, and E. Seeman, Moderate exercise during growth in pre-

SPECKER, NAMGUNG, AND TSANG

206.

207.

208.

209.

210.

211. 212.

213.

214.

215.

216.

pubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J. Bone Miner. Res. 13, 1814 – 1821 (1998). M. Hotta, I. Fukuda, K. Sato, N. Huzuka, T. Shibasaki, and K. Takano, The relationship between bone turnover and body weight, serum insulin-like growth factor (IGF) I, and serum IGFbinding protein levels in patients with anorexa nervousa. J. Clin. Endocrinol. Metab. 85, 200 – 206 (2000). L. A. Soyka, S. Grinspoon, L. L. Levitsky, D. B. Herzog, and A. Klibanski, The effects of anorexia nervosa on bone metabolism in female adolescents. J. Clin. Endocrinol. Metab. 84, 4489 – 4496 (1999). D. T. Gibbens, V. Gilsanz, M. I. Boechat, D. Dufer, M. E. Carlson, and C. I. Wang, Osteoporosis in cystic fibrosis. J. Pediatr. 113, 295 – 300 (1988). E. Argao, B. Specker, and J. Heubi, Bone mineral content in infants and children with chronic cholestatic liver disease. Pediatrics 91, 1151 – 1154 (1993). R. M. Issenman, S. A. Atkinson, C. Radoja, and L. Fraher, Longitudinal assessment of growth, mineral metabolism, and bone mass in pediatric Crohn’s disease. J. Pediatr. Gastro. Nutr. 17, 401 – 406 (1993). N. J. Shaw, C. P. White, W. D. Fraser, and L. Rosenbloom, Osteopenia in cerebral palsy. Arch. Dis. Child. 71, 235 – 2381 (9994). C. J. Henderson, G. D. Cawkwell, B. L. Specker, R. I. Sierra, R. W. Wilmott, B. N. Campaigne, and D. J. Lovell, Predictors of total body bone mineral density in non-corticosteroid-treated prepubertal children with juvenile rheumatoid arthritis. Arthritis Rheum. 40, 1967 – 1975 (1997). C. J. Henderson, B. L. Specker, R. I. Sierra, B. N. Campaigne, and D. J. Lovell, Total body bone mineral content in non-corticosteroidtreated postpubertal females with junveile rheumatoid arthritis. Arthritis Rheum. 43, 531 – 540 (2000). G. Hedlin, J. Svedmyr, and A. C. Ryden, Systemic effects of a short course of betamethasone compared with high-dose inhaled budesonide in early childhood asthma. Acta Paediatr. 88, 48 – 51 (1999). O. D. Wolthers, and C. Heuck, Differential effects of inhaled budesonide on serum osteocalcin in children and adolescents with asthma. Pediatr. Allergy Immunol. 9, 150 – 155 (1998). H. D. Allen, I. G. Thong, P. Clifton-Gligh, S. Holmes, L. Nery, and K. B. Wilson, Effects of high-dose inhaled corticosteroids on bone metabolism in prepubertal children with asthma. Pediatr. Pulmonol. 29, 188 – 193 (2000).

CHAPTER 25

Bone Acquisition in Adolescence JEAN-PHILIPPE BONJOUR AND RENE RIZZOLI Department of Internal Medicine, Division of Bone Diseases (World Health Organization Collaborating Center for Osteoporosis and Bone Diseases), University Hospital, Geneva CH-1211, Switzerland

I. Introduction II. Measurement of Bone Mass Development III. Bone Mass Gain during Puberty

IV. Calcium – Phosphate Metabolism during Puberty V. Determinants of Bone Mass Gain References

I. INTRODUCTION

mineralized tissue contained within its periosteal envelope. The mean volumetric mineral density of bony tissue (vBMD in g hydroxyapatite per cm3) can be determined noninvasively by quantitative computed tomography (QCT). The techniques of single and dual X-ray (SXA, DXA) absorptiometry provide measurements of the so-called “areal” or “surface” bone mineral density (BMD in g hydroxyapatite per cm2). The values generated by these techniques are directly dependent on both the size, or more precisely the thickness, and the integrated mineral density of the scanned skeletal tissue. The term bone mineral density without the additional areal qualification has been widely used with the understanding by specialists in the field that neither SPX nor DPX techniques provide a measurement of volumetric density. It will be useful to define here the meaning of areal BMD [1], as changes in this variable have been interpreted erroneously as changes in vBMD. Areal BMD is of clinical relevance in the context of osteoporosis. Indeed, it has been shown to be directly related to bone strength, i.e., to the resistance of the skeleton to mechanical stress both in vivo and in vitro (for review, see [2,3]). Thus, there is an inverse relationship between areal BMD values and the prevalence of osteoporotic fractures [4].

In adults the mass of bony tissue present at any time in life is a function of the amount achieved at maturity and that lost with aging. Hence, the bone mass acquired at the end of the growth period is a major determinant of the risk of fractures, such as those observed at the radial, vertebral, or femoral sites in osteoporotic patients. Puberty is a crucial period in the acquisition of bone mass. This chapter attempts to summarize some of the knowledge that has accrued recently on the characteristics of normal bone mass development during pre-, peri-, and postpubertal period.

II. MEASUREMENT OF BONE MASS DEVELOPMENT Most information on the characteristics of skeletal growth during childhood and adolescence has been obtained from noninvasive techniques that allow quantification with great precision and accuracy of bone mass at various sites of the skeleton. The bone mass of a part of the skeleton depends directly on both its volume or size and the density of the

OSTEOPOROSIS, SECOND EDITION VOLUME 1

621

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

622

BONJOUR AND RIZZOLI

In the spine, the total mineral content (BMC in g hydroxyapatite) of the vertebra, including the posterior arch, can be measured by dual-energy X-ray absorptiometry (DXA) using the classical anteroposterior (frontal) projection. The BMC and the derived areal BMD of the vertebral body “isolated” from the vertebral arch can also be obtained by using DXA in the lateral (sagittal) projection [5,6]. The so-called bone mineral “apparent” density (BMAD in g/cm3) represents an indirect estimate of the volumetric skeletal density [7]. This extrapolated variable can be expected to be less related to bone strength than areal BMD, as it does not take into account the important size component that influences the mechanical resistance [2,8]. Note that the values of vBMD as determined by QCT [9] correlate better with bone strength when they are multiplied by the area of the vertebral body [10]. This calculation gives a value equivalent to the bone mass that can be measured directly by dual-photon absorptiometry (DPA) or DXA [2]. It has been reported in two groups of postmenopausal women with equally low volumetric cancellous density, as determined by spinal QCT, that the group with vertebral fractures had a significantly smaller cross-sectional area of the vertebral bodies [11]. Therefore, in terms of overall bone strength prediction, the areal BMD value is more informative than isolated measurements of volumetric trabecular density, as the former variable includes both the bone thickness and its integrated volumetric density. This statement does not mean that other variables that are more difficult to assess precisely, such as the microarchitecture of the trabecular network and/or the intrinsic “quality” of the mineralized tissue, do not play contributory roles in the resistance to mechanical force. Furthermore, it is obvious that a full understanding of the fundamental mechanisms that underlie the marked interindividual variability observed in bone mass gain will require separate analysis of how bone size, cortical thickness, and volumetric trabecular density evolve during growth. The main genetic and environmental factors that determine the development of each of these three important contributors to bone strength remain to be identified.

III. BONE MASS GAIN DURING PUBERTY A. Gender Difference Puberty is the period during which the characteristic sex difference in bone mass observed in adults becomes fully expressed. There is no evidence for a gender difference in bone mass of either the axial or the appendicular skeleton at birth. Likewise, the volumetric bone mineral density also appears to be similar between female and male newborns [12]. This absence of sex difference in bone mass appears to be maintained until the onset of puberty [12 – 18]. Nevertheless, this general statement needs to be qualified some-

what. In one study, size and density components of lumbar vertebral bodies were determined by QCT [19]. In prepubertal girls and boys, the height of the vertebral bodies L1 – L3 did not differ, and no gender difference in cortical or cancellous bone mineral densities was found. However, a greater vertebral cross-sectional area was found in boys. This gender difference increased progressively during pubertal maturation, while the progression of the other densitometric or anthropometric values of L1 – L3 was quite similar in both females and males. It is well known that, in adults, most bones are on average larger in males than in females. Furthermore, morphometric studies document a significant gender difference in the cortical thickness of most axial and appendicular bones [20 – 22]. The important gender difference in bone mass that develops during pubertal maturation appears to result from a greater increase in bone size, a characteristic that is associated with a larger increment in the cortical shell in males than in females [23]. It is important to emphasize that gender differences in mean areal BMD/BMC observed at several sites of the skeleton after pubertal maturation do not appear to reflect differences in volumetric bone mineral density. Indeed, very consistent data obtained by histomorphometry [24], gravimetry [25,26], and QCT [19,27 – 31] indicate no sex difference in the volumetric trabecular density at the end of the period of maturation, i.e., in young healthy adults in their third decade. However, puberty appears to be the critical period during which racial differences in both volumetric trabecular density and bone mass become expressed. Blacks have greater volumetric density than whites [30,34,35]. Trabecular number appears to be similar, but the trabeculae are thicker in black than in white subjects [35,36]. In cortical peripheral bone, there is no gender difference in the cross-sectional area of the midfemoral shaft after adjustment for height and weight [29]. However, these areas are greater in blacks than in whites for an identical cortical thickness [30]. Thus, despite the same cortical thickness, larger bones result in greater bone mineral content. Bony tissue situated more distantly from the bone central axis confers a greater resistance to bending [8,37]. As discussed later, the significantly greater mean lumbar spine areal BMD of young healthy adult males compared to females (at least when assessed using lateral projection) [38] and similarly increased values at the midfemoral or midradial diaphysis appear to reflect a more prolonged duration of pubertal maturation rather than a greater maximal rate of bone accretion.

B. Variable Velocity in Bone Mass Gain According to Skeletal Sites A longitudinal study [39] using DXA indicated that during pubertal maturation the accumulation rate of areal bone mineral density (aBMD in g/cm2/year) or content (BMC in

623

CHAPTER 25 Bone Acquisition in Adolescence

g/year) at both lumbar spine (LS) and femoral neck (FN) sites increases by four to sixfold over a 3-year period in females (from 11 to 14 years) and a 4-year period in males (from 13 to 17 years) (Figs. 1 and 2). The increase in BMD/BMC gain during the corresponding pubertal period appears to be less marked in long bone diaphyses as observed in cross-sectional studies in the forearm [13 – 16] or more recently in a longitudinal study at the level of the midfemoral shaft (FS), a site where a twofold increase has been observed [17,39] (Fig. 2). There is heterogenity in skeletal growth. Growth of distal limb segments precedes that of proximal segments. Appendicular growth is more rapid than axial growth before puberty, but decelerates at puberty when axial growth accelerates. [40]. It is suggested that this differing tempo in bone growth could have important pathophysiological consequences [40]. Regions growing rapidly or distant from their peak may be more affected by illness than those growing slowly or nearer completion of growth. Depending on the age and pubertal maturation, deficits may occur in limb dimension (prepuberty), spine dimension (early puberty), or vBMD through interference with the phenomenon of “consolidation” that follows the marked decline in longitudinal growth [40].

C. Changes in Bone Mass Components There is a marked difference in the magnitude of gains in bone size compared to those in volumetric trabecular density. From Tanner pubertal stage P1 to P5 the bony mass of lumbar vertebrae increases about 10-fold more than the volumetric trabecular density. We observed that from stage P1 to P5, L2 – L4 BMC increased by about 175 and 160% in males and females, respectively [14], whereas corresponding mean increases in the volumetric trabecular density of the lumbar vertebrae, assessed by QCT, were 20 and

FIGURE 1

15% in males and females, respectively [28]. This means that a large portion of the increment in bone mass during puberty is due to an increment in size associated with an increase in thickness of the cortical shell.

D. Gain in Statural Height and Bone Mass Growth In prepubertal children, there is a tight relationship between bone mass at the spine and femur and statural height [16,17]. This close correlation vanishes during pubertal maturation, with the appearance of the pattern observed in adults in whom BMD/BMC values are poorly correlated to height (Fig. 3). This means that among the determinants of bone mass growth during puberty, some act in complete independence from those responsible for the gain in height. This concept is reinforced by the dissociation in time between the gain in height and the growth in bone mass at some important skeletal sites [39] (Fig. 4).

E. Transient Fragility There is an asynchrony between the gain in height and the bone mass growth at the levels of both lumbar spine and femoral neck [17,39,41]. This temporal dissociation appears to be maximal when height velocity reaches its greatest rate, i.e., at the ages of 11 – 12 and 13 – 14 years, in girls and boys, respectively [17,39,41]. Interestingly, in both genders, these are the ages at which the incidence of traumatic fractures is the highest [42,43]. This suggests that the peak of fracture incidence noticed during the second decade of life could be due, at least in part, to a transient fragility rather than merely resulting from particular risky behavior, as originally thought [44]. It has been postulated that an increase in the porosity of

Bone mass gain at lumbar spine level during adolescence. Yearly increase in lumbar BMD (A; L2 – L4 BMD) and BMC (B; L2 – L4 BMC). Results are the mean  SEM. From Theintz et al., J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992).

624

BONJOUR AND RIZZOLI

FIGURE 2

Bone mass gain at the femoral level during adolescence. Yearly increase in femoral neck (A) and mid-femoral shaft (B) BMD. Results are the mean  SEM. From Theintz et al., J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992).

the cortical shell could be responsible for this transient fragility that coincides with the adolescent growth spurt [45]. Both a delay in the thickening of the cortical shell [46] and an increase in its porosity could contribute to this phenomenon, as mechanical resistance depends on both the size and the volumetric density of the bones [8,47].

F. Bone Mass Variability during Puberty At the beginning of the third decade in both sexes, there is a large variability in the normal values of BMD/BMC. This large biological variability can be simply quantified by

FIGURE 3

calculating the coefficient of variation [(standard deviation/mean)  100]. It can be observed in both the axial and the appendicular skeleton, particularly at sites susceptible to osteoporotic fractures, such as the lumbar spine and femoral neck. Most important and as mentioned earlier, the variability in BMD/BMC is barely reduced after correction for stature. Note that in young healthy adults the biological variability of L2 – L4 BMC is four to five times larger than that of height [38]. This height-independent large variability in BMD/BMC appears to expand during pubertal development at sites such as lumbar spine and femoral neck [39], where the increase in accretion rate is particularly marked during this period. In contrast, at sites

Relationship between lumbar BMC and statural height in female and male adolescents. Each dot corresponds to one individual. Mean heights achieved at the different pubertal stages (P1 – P5) are indicated by the vertical bars drawn on the horizontal axis. From Bonjour et al., J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991).

625

CHAPTER 25 Bone Acquisition in Adolescence

FIGURE 4 Relationship between changes () in BMC and statural height in adolescents grouped according to pubertal stages. Each symbol corresponds to the mean  SEM of the different L2 – L4 BMC determined at 1-year intervals and plotted against the mean height gain  SEM. From Theintz et al., J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992). where pubertal acceleration is less pronounced, such as the femoral midshaft, the variability remains constant from Tanner stage P1 to P5. It is also important to add that variability in height does not increase during puberty. The mechanisms underlying this height-independent increment in bone mass variability at some sites during puberty are still poorly understood. Twin studies showed that the increase in the statistical variance of aBMD during puberty was much more pronounced in either lumbar spine or femoral neck than in the total forearm [48]. Mathematical modeling suggested that the increase in variance was due to both genetic and environmental factors and this in various proportions according to the stage of pubertal maturation and skeletal sites. The authors concluded that the genetic and environmental etiology of bone mineral density is more complex than previously thought [48]. In other words, the degree to which genetic or environmental factors underlie the difference in the size of the variance remains unknown.

and 20 years some increase in mean bone mass could still be seen at both L2 – L4 BMD/BMC and midfemoral shaft. In subjects reaching pubertal stage P5 and growing less than 1 cm/year, data suggest that a significant bone mass

G. Stabilization of Bone Mass Gain after Pubertal Maturation In adolescent females, the rate of increase in bone mass declines rapidly after menarche (Fig. 5). In healthy Caucasian females with apparently adequate intakes of energy, protein, and calcium, bone mass accumulation can virtually be completed before the end of the second decade at both the lumbar spine and the femoral neck [7,17,39,41,48,49]. The mean gains in lumbar spine, femoral neck, and midfemoral shaft BMD were found to be slight or nonexistent between 17 and 20 years [39]. In adolescent males, the gain in BMD/BMC, which was particularly high from 13 to 17 years, declined markedly thereafter, although between 17

FIGURE 5

Relationship between spinal and femoral BMD gains () and years after menarche. Results are the mean  SEM. Time after menarche: P5a, less than 1 year (n  11; age, 13.8  0.4 years); P5b, 1 – 2 years (n  7; age, 14.5  0.2 years); P5c, 2 – 4 years (n  24; age, 16.6  0.3 years); and P5d, more than 4 years (n  16; age, 17.4  0.3 years). Statistical significance of BMD increase: *P  0.05; **P  0.01; ***P  0.001 (compared to zero). The corresponding height gains were P5a, 2.9  0.7; P5b, 1.4  0.4; P5c, 0.5  0.2; and P5d, 0.1  0.2 cm/year. From Theintz et al., J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992).

626 gain is still present in males but not in females [39]. Thus, an important sex difference in the magnitude and/or duration of the so-called “consolidation” phenomenon that contributes to peak bone mass value may occur. That BMD of the proximal femur and vertebral body as determined by DXA will reach a maximal value by late adolescence has also been observed in cross-sectional surveys carried out in healthy Caucasian females living in Ohio [7,49], in agreement with previous [17,39] and more recent results obtained in similar cohorts living in Europe [50 – 52] and Australia [48,53]. Cross-sectional observations made with QCT technology also indicate that the maximal volumetric bone mineral density of the lumbar vertebral body will be achieved soon after menarche [54]. This is in keeping with numerous observations indicating that bone mass does not increase much from the third to the fifth decade [13,15,24,26, 27,55 – 66]. Nevertheless, a few studies suggest that bone mass acquisition could still be substantial during the third and fourth decades [67 – 69]. In any case, the weight of evidence does not sustain the concept that bone mass at any skeletal site, in either gender, in any race, or in any geographical area around the world, continues to accumulate substantially until the fourth decade. Note that the external frame of the bones can become larger during adult life. This phenomenon has been documented by measuring the external diameter of several bones by radiogrammetry [34,70 – 72] and may be the consequence of increased endosteal bone resorption with enlargement in the internal diameter. This would result in a transient reduction of the cortical thickness, leading to an augmentation in mechanical load on the remaining bony tissue. As a response to this increased load, the number and/or activity of osteogenic cells on the periosteal side of the cortex would be enhanced, giving rise to compensatory growth with enlargement of the external diameter. Thus, such a modeling phenomenon would be a response to bone loss, tending to compensate for the reduction in mechanical resistance [8]. Note that the modeling process does not seem to occur in parts of the skeleton deprived of periosteum, such as a portion of the proximal femur [47]. This increase in the external diameter of the bone should not be interpreted as a net increase in skeletal mass nor should it be used to support the concept that bone mass, as opposed to external bone size, can continue to grow in all parts of the skeleton many years after the arrest of statural growth.

IV. CALCIUM – PHOSPHATE METABOLISM DURING PUBERTY Several physiological functions influence bone accumulation during growth. Animal studies have identified physiological mechanisms that sustain increased bone mineral de-

BONJOUR AND RIZZOLI

mand in relation to variations in growth velocity. In this context, two adaptive mechanisms affecting calcium – phosphate metabolism appear to be particularly important: the increase in the plasma concentration of 1,25-dihydroxyvitamin D3 (calcitriol) and the stimulation of the renal tubular reabsorption of inorganic phosphate (Pi). The elevation in the production and plasma level of calcitriol enhances the capacity of the intestinal epithelium to absorb both calcium and Pi. The increase in the tubular reabsorption of Pi results in a rise in its extracellular concentration. Without these two concerted adaptive responses, growth and mineralization cannot be optimal. Note that the increase in the tubular Pi reabsorption is not mediated by a rise in the renal production or in the plasma level of calcitriol [73]. Analysis of cross-sectional studies suggests that these two adaptive mechanisms could be essential to cope with the increased bone mineral demand during the pubertal growth spurt. An increase in plasma calcitriol concentrations has been reported during pubertal maturation [74]. Both the pattern of this response and its consequence for intestinal calcium absorptive capacity in relation to pubertal bone mass acquisition remain difficult to document, as it would require a time-integrated estimate of the controlling (calcitriol and intestinal calcium absorption) elements. A tight relationship exists among the tubular reabsorption of Pi, the plasma Pi level, and growth velocity in children [75]. A rise in plasma Pi during puberty has been reported [76,77]. Precise quantitation of the relationship between changes in the regulatory component of the tubular Pi reabsorption and plasma concentration of Pi and bone mass gain during puberty remains to be done. However, similar to the calcitriol – intestinal calcium absorption regulatory pathway, a correct evaluation would require a time-integrated assessment of the changes in the tubular reabsorption and plasma concentration of Pi during the period of accelerated bone mass gain. The mechanism underlying the parallel rise in calcitriol and the tubular reabsorption of Pi has been clarified. In fact, experimental studies indicate that one single factor, namely insulin-like growth factor-I (IGF-I), could be responsible for the stimulation of both calcitriol production and tubular Pi reabsorption (TmPi/GFR) in relation to the increased calcium and Pi demand associated with bone growth [17,78]. In humans, the plasma level of IGF-I rises transiently during pubertal maturation, to reach a peak during midpuberty, the maximal level thus occurring at an earlier chronological age in females than in males [79]. The role of IGF-I in calcium phosphate metabolism during pubertal maturation in relation with essential nutrients for bone growth is illustrated in Fig. 6 (see also color plate). The rise in the plasma levels of IGF-I, calcitriol, and Pi are correlated with the elevation in indices of the bone appositional rate such as alkaline phosphatase [80,81,82,83] and osteocalcin [82 – 85]. Note that the plasma concentrations of gonadal sex hormones, as well as those of adrenal androgens (dehydroepiandrosterone and

627

CHAPTER 25 Bone Acquisition in Adolescence

FIGURE 6 Role of insulin-like growth factor-I (IGF-I) in calcium – phosphate metabolism during pubertal maturation in relation to essential nutrients for bone growth. During the pubertal bone growth spurt there is a rise in circulating IGF-I. The hepatic production of IGF-I is under the positive influence of growth hormone (GH) and essential amino acids (a.a.). IGF-I stimulates bone growth. At the kidney level, IGF-I increases both the 1,25-dihydroxyvitamin D (1,25 D) conversion from 25-hydroxyvitamin D (25D) and the maximal tubular reabsorption of Pi (TmPi). By this dual renal action IGF-I favors a positive calcium and phosphate balance as required by the increased bone mineral accrual. See Section IV for references supporting this integrated physiological concept. (See also color plate.) androstendione), which increase before and during pubertal maturation, do not seem to accord with the accelerated bone mass gain [81a,86,87]. Whether differences in the adaptive responses that control calcium and phosphate homeostasis could play a role in the increased variance in lumbar spine or femoral neck BMD/BMC remains to be explored. As reviewed recently, the interaction between the growth hormone – IGF-I axis and sex steroids is quite complex [83]. The effect of these interactions on the gains in bone size and mass during pubertal maturation, independent of their influence on the rate and duration of longitudinal growth, remains largely unknown.

A. Bone Biochemical Markers during Puberty The interpretation of the changes in bone biochemical markers during growth is more complex than in adulthood, particularly for the markers of bone resorption [see for review 83]. The plasma concentrations of the bone formation markers are highest when the velocity of bone mineral accrual is maximal. This suggests that the two phenomena are

related. The high urinary excretion of bone resorption markers, such as collagen pyridinium cross-links, observed during childhood, decreases after the growth spurt and reaches adult values at the end of pubertal maturation, i.e., at 15 – 16 and 17 – 18 years of age in girls and boys, respectively [see for review 83]. This probably reflects the decrease in the resorption rate associated with the reduction and arrest in longitudinal bone growth. In a longitudinal study in pubertal girls, bone turnover markers (osteolcalcin, bone-specific alkaline phosphatase, and collagen pyridium cross-links) were modestly related to statural height gain, but they were not predictive of gains in either total bone mineral content or density as assessed by DXA [88].

V. DETERMINANTS OF BONE MASS GAIN Many factors, more or less independently, are supposed to influence bone mass accumulation during growth. The list of these determinants classically includes heredity, sex, dietary components (calcium, proteins), endocrine factors

628

BONJOUR AND RIZZOLI

(sex steroids, calcitriol, IGF-I), mechanical forces (physical activity, body weight), and exposure to other risk factors [27,89 – 93]. Quantitatively, the most prominent determinant appears to be genetically related.

A. Genetic Determinant As mentioned earlier the variability in BMD/BMC at the level of the lumbar spine and of the proximal femur, unrelated to changes in the statural height, increases during pubertal maturation. The contribution of heredity, compared to that of the environment, to this increased bone mass variability is not clearly elucidated. As discussed in Chapter 26 genetic factors account for a large percentage of the population variability in BMD among age- and sex-matched normal individuals [89]. Daughters of osteoporotic women have a low BMD [94]. BMD is decreased among the relatives of 38 middle-aged men with severe idiopathic osteoporosis [95]. To investigate the proportion of the BMD variance across the population explained by genetic factors, known as its heritability [96], two human models mainly have been used. In the twin model, within-pairs correlations for BMD are compared between monozygotic (MZ) twins, who by essence share 100% of their genes, and dizygotic (DZ) twins, who have 50% of their genes in common. Stronger correlation coefficients among adult MZ as compared to DZ twins are indicative of the genetic influence on peak bone mass, accounting for as much as 80% of lumbar spine and proximal femur BMD variance [89]. Lean and fat mass are also genetically determined [97]. Indeed, it appears that 80 and 65% of variance of lean and fat mass, respectively, are attributable to genetic factors. However, genetic factors affecting lean and fat mass have only little influence on lumbar spine or femoral neck BMD. These results differ from previous evidence of indirect genetic effects on bone mass occurring through the determination of lean body mass [98]. Parents – offspring comparisons have also shown significant relationships for BMD, albeit heritability estimates have been somewhat lower (in the range of 60%) than in the twin model [99]. Actually, the magnitude of direct genetic effects on peak bone mass as evaluated in both human models may be overestimated by similarities in environmental covariates [91,100]. We investigated correlations for bone mineral content, areal and volumetric bone mineral density, and bone area in the lumbar spine and femur (neck, trochanter, and diaphysis) in premenopausal women and their prepubertal daughters [101]. Regressions were adjusted for height, weight, and calcium intake to minimize the impact of indirect genetic effects as well as of dietary influences on bone mineral mass resemblance among relatives. Results indicate that despite great disparities in the maturity of the various constituents of bone mass before puberty with respect to peak adult values, heredity by

maternal descent is detectable at all skeletal sites and affected virtually all bone mass constituents, including bone size and volumetric mineral density. Moreover, when daughters’ bone values were reevaluated 2 years later, while puberty had begun and bone mineral mass had increased considerably, measurements were highly correlated with prepubertal values and mother – daughter correlations had remained unchanged. Thus, a major proportion of this variance is due to genetic factors that are already expressed before puberty with subsequent tracking of the bone mass constituent through the phase of rapid pubertal growth until peak bone mass is achieved. Interestingly, it appears that male-to-male and male-to-female inheritance of bone mass may differ substantially [100]. It might be hazardous therefore to extrapolate genetic influences on bone mineral mass as identified in women to the male population, in which this question has virtually not yet been investigated. In contrast to the clear heritability of peak bone mass, the proportion of the variance in bone turnover that depends on genetic factors, as assessed in this model by various markers of bone formation and resorption, appears to be small [102]. Hence, peak bone mass is very likely determined by numerous gene products implicated in both bone modeling and remodeling. Among the multiple candidate genes harboring polymorphic loci investigated so far in relation to BMD and/or BMD changes, the vitamin D receptor (VDR ) – 3 end alleles are controversial [103 – 111]. A meta-analysis combining 16 separate studies examined the relationship between VDR genotypes and BMD [111] and found that subjects with the BB genotype had a 2.4, 2.5 and 1.7% nonsignificantly lower BMD as compared to bb at the level of the femoral neck, lumbar spine, and distal radius, respectively. The more recently described association between VDR-5 start codon polymorphism (Fokl) and BMD, at first observed in small cohorts of postmenopausal Mexican-American women [112], white premenopausal American women [113], and Japanese women [114], has not been confirmed in two larger European studies in healthy premenopausal women or prepubertal girls [115,116]. Several independent investigators have shown the importance of age, gene – environment, and gene – gene interactions to explain the inconsistent relationship between bone mineral mass and VDR-3 and 5genotypes. Thus, significant BMD differences between VDR-3BsmI genotypes were detected in children [117,118], but were absent in premenopausal women from the same genetic background [118]. Moreover, the latter study found that BMD gain in prepubertal girls was increased at several skeletal sites in Bb and BB subjects in response to calcium supplements, whereas it remained apparently unaffected in bb girls, who had a trend for spontaneously higher BMD accumulation on their usual calcium diet [118] (Fig. 7). Accordingly, a model taking into account the early influence of

629

CHAPTER 25 Bone Acquisition in Adolescence

FIGURE 7 Mean BMD changes of five appendicular skeleton sites (radial metaphysis and diaphysis, femoral neck, trochanter and shaft) in prepubertal girls receiving a 850-mg calcium supplement (dashed bars) or a placebo (open bars), for 1 year according to VDR 3’-end genotypes. Results are taken from Bonjour et al. [123] and Ferrari et al. [118]. Reproduced from Ferrari et al. [119].

VDR-3 polymorphisms, calcium intake, and puberty on BMD gain has been proposed to explain the relation between these genotypes and peak bone mass [119]. Interestingly, several investigators have also noted a significantly lower height among women and men with the VDR-3 BB compared to Bb or bb genotypes [118,120 – 122]. Considering the relationship between body size and bone size, as well as the influence of calcium intake on both body height and bone area during growth [123], it is tempting to speculate that VDR-3 alleles, together with environmental calcium, might exert an indirect and complex influence on peak bone mass by regulating skeletal growth. Altogether, these observations provide a possible physiological mechanism for the relationship between VDR gene polymorphisms and bone mass and emphasize the methodological limitations of earlier studies focusing on the association between VDR genotypes and BMD regardless of age and environmental factors. Moreover, other potential gene – environmental interactions, such as those involving physical exercise [124] as well as gene – gene interactions, might further modulate the relationship between VDR gene polymorphisms and bone mass, as for instance, an interaction between VDR and estrogen receptor (ER) gene polymorphisms. In summary, VDR-3 and 5alleles are possibly weak determinants of bone mineral density, their effects being easily confounded by the influence of many other genes and environmental factors. Hence, VDR gene polymorphisms alone are not clinically useful genetic markers of peak bone mass, but could be one significant factor to explain some of the variability observed in the population.

B. Physical Activity The responsiveness to either an increase or a decrease in mechanical strain is probably greater in growing than adult bones [45,125,126; see also Chapter 28]. Hence, the concept of public health programs aimed at increasing physical activity among healthy children and adolescents in order to maximize peak bone mass has been promoted. Several reports in children or adolescents involved in competitive sport or ballet dancing indicate that intense exercise is associated with an increase in bone mass accrual in weightbearing skeletal sites [127 – 132, 133, 134]. The question arises whether this increase in BMD/BMC resulting from intense exercise is translated into greater bone strength. A recent cross-sectional study in male-elite tennis players, using peripheral QCT and side-to-side arm comparison indicates that the increase in BMC reflected an increased bone size, which was associated with an augmentation in an index of bone strength. In contrast, no change in either cortical or trabecular vBMD was observed [135]. Whether the same type of beneficial structural change for bone strength is observed at other skeletal sites, such as vertebral bodies and proximal femur, in response to different kinds of intense exercise during childhood and adolescence, remains to be documented. In terms of general public health, observations made in elite athletes cannot be the basis of recommendations for the general population, as intense exercise is beyond the reach of most individuals. Much more relevant is information on the effect of moderate exercise on bone mass acquisition. Some [52,92,136], but not all

630

BONJOUR AND RIZZOLI

[16,53,137 – 139], cross-sectional studies have found a slightly positive association between physical activity and bone mass values in children and adolescents. However, the positive association found cross-sectionally was not confirmed by observational longitudinal studies relating bone mass gain to physical activity [50,137]. Measurements of the duration, intensity, and type of physical activity that are based on recall are not very precise, particularly in children. Therefore, it is possible that negative findings could be ascribed to poor validity in the methods used to estimate physical activity. Controlled prospective studies carried out in prepubertal girls [140] or boys [141] indicate that exercise programs undertaken in schools, and considered on the average as moderate, can increase bone mass acquisition. These studies indicate that the growing skeleton is certainly sensitive to exercise and suggest that prepuberty would be an opportune time for implementing physical education programs consisting of various moderate weightbearing exercises. Nevertheless, it remains uncertain to what extent the greater aBMD gain in response to moderate and readily accessible weight-bearing exercise is associated with a commensurate increase in bone strength [141]. The magnitude of benefit in terms of bone strength will depend on the nature of the structural change. An effect consisting primarily of an increased periosteal apposition and consecutive diameter will confer greater mechanical resistance than a response limited to the endosteal apposition rate leading essentially to a reduction in the endocortical diameter. There is a need for further studies aimed at examining the effects of mechanical loading components, such as magnitude and frequency of various types of exercise on the the mass and geometry of bones in children and adolescents [142]. Studies in adult elite athletes strongly indicate that increased bone mass gains resulting from intense physical activity during childhood and adolescence are maintained after training attenuates or even completely ceases [130,132,143 – 145]. Finally, the question whether the increased peak bone mass induced by physical exercise will be maintained into old age and confer a reduction in fracture rate remains open. A cross-sectional study of retired Australian-elite soccer players suggests that this may not be the case [146]. However, the lack of information on the peak bone mass values of these men does not allow one to draw firm conclusion about this observation.

C. Nutritional Factors Puberty is considered to be a period with major behavioral changes and alterations in lifestyle. It is also assumed that important modifications in food habits occur during pubertal maturation, particularly in affluent societies. However, there is still a lack of quantitative and qualitative

information regarding the evolution of both micro- and macronutrient intakes in relation to pubertal maturation. At the individual level, to what extent variations in the intakes of some nutrients in healthy, apparently well-nourished children and adolescents can affect bone mass accumulation, particularly at sites susceptible to osteoporotic fractures, has received increasing attention since the early 1990s. Most studies have focused on the intake of calcium. However, other nutrients such as protein should also be considered. It is usually accepted that increasing the calcium intake during childhood and adolescence will be associated with a greater bone mass gain and thereby a higher peak bone mass [147,148]. However, a survey of the literature on the relationship between dietary calcium and bone mass indicates that some [52,139,149 – 153], but not all, studies [7,50,136 – 138,154] have found a positive correlation between these two variables. As with physical activity, several sets of cross-sectional and longitudinal data, including our own unpublished results on dietary calcium intake and bone mass accrual in female and male subjects aged 9 to 19 years, are compatible with a “two threshold model.” On one side of the normal range one can conceive the existence of a “low” threshold, set at a total calcium intake of about 400 – 500 mg/day, below which a positive relationship can be found. Within this low range the positive effect of calcium would be explained merely by its role as a necessary substrate for bone mass accrual. On the other side of the normal range, there would be a “high” threshold, set at about 1600 mg/day, above which the calcium intake through another mechanism could exert a slightly positive influence on bone mass accrual. In addition, the levels of the two thresholds could vary according to the stage of pubertal maturation. In our own cross-sectional study [17], a significant positive relationship between total calcium intake as determined by two 5-day diaries was found in females in the pubertal subgroup P1 – P4, but not in the P5 subgroup. Furthermore, in the longitudinal study [39] when results were analyzed by taking into account the influence of age and pubertal maturation, the relationship between the absolute values of the calcium intake and the gain in BMD Z score suggested that calcium may be more important before than during pubertal maturation (Fig. 8). Several intervention studies have been carried out in children and adolescents [123,155 – 159]. Overall, these indicate greater bone mineral mass gain in children and adolescents receiving calcium supplementation over periods varying from 12 to 36 months. The benefit of supplemental calcium has been greater in the appendicular than in the axial skeleton [123,156]. Thus, in prepubertal children, calcium supplementation is more effective on cortical appendicular bone (radial and femoral diaphysis) than on axial trabecular rich bone (lumbar spine) or on the hip (femoral neck, trochanter) [123,156] (Fig. 9). The skeleton appears to be more

CHAPTER 25 Bone Acquisition in Adolescence

FIGURE 8 Relationship between calcium intake and change () in lumbar BMC in pre-, peri-, and postpubertal female and male adolescents. The mean calcium intake from dairy, vegetable, and mineral sources was recorded in two 5-day diet diaries at 1-year intervals. A positive correlation was found in prepubertal (P1), but not in peripubertal (P2 – P4) or postpubertal (P5) subjects. Each dot corresponds to the change in BMC adjusted for age and gender (Z score) in 193 subjects aged from 9 to 19 years. BMC data are from Theintz et al., J. Clin. Endocrin. Metab. 75, 1060 – 1065 (1992).

Influence of calcium-enriched foods on bone mineral accrual in prepubertal girls. Bars represent means  SEM of gains in aBMD measured at six skeletal sites in girls consuming food enriched (supplemented) or not (placebo) in calcium for 48 weeks. Calcium-enriched foods provided a supplement of 850 mg/day. # p  0.075; *p  0.037; **p  0.015; ***p  0.007. From Bonjour et al., J. Clin. Invest. 99, 1287 – 1294 (1997).

FIGURE 9

631

632 responsive to calcium supplementation before the onset of pubertal maturation [156]. As intuitively expected, this benefit may be particularly substantial in children with a relatively low calcium intake [123]. In 8-year-old prepubertal girls with a spontaneously low calcium intake, increasing the calcium intake from about 700 to 1400 mg augmented the mean gain in aBMD of six skeletal sites by 58% as compared to the placebo group after 1 year of supplementation [123]. This difference corresponds to a gain of 0.24 standard deviation (SD). If sustained over a period of 4 years, such an increase in the calcium intake could augment mean aBMD by 1 SD. Thus, milk calcium supplementation could modify the bone growth trajectory, thereby increasing peak bone mass. In this regard it is interesting to note that an intervention influencing calcium – phosphate metabolism and limited to the first year of life may also modify the trajectory of bone mass accrual. As a matter of fact, a 400 IU/day vitamin D supplementation given to infants for an average of 1 year was associated with a significant increase in aBMD measured at the age of 7 – 9 years [160]. The aBMD difference between the vitamin D-supplemented and nonsupplemented group was particularly significant at the femoral neck, trochanter, and radial metaphysis [160]. These observations are compatible with the “programming” concept, according to which environmental stimuli during critical periods of early development can provoke long-lasting modifications in structure and function [161,162]. Another aspect to consider is that the type of the supplemented calcium salt could modulate the nature of the bone response. Thus, the response to administration of a calcium – phosphate salt from milk extract appears to differ from those recorded with other calcium supplements. Indeed, the positive effect on aBMD was associated with an increase in the projected bone area at several sites of the skeleton and a slight increase in statural height [123]. This type of response was not observed when calcium was given as citrate malate salts [156,157], carbonate alone [158], or carbonate combined with gluconate lactate [159]. Interestingly, it was similar to the response to whole milk supplementation [163]. However in this study [163], the positive effect on bone size could be ascribed to other nutrients contained in whole milk, whereas in the other study the tested calcium-enriched foods had the same energy, lipid, and protein content as those given to the placebo group [123]. It is important to consider whether the gain resulting from the intervention will be lost after discontinuation of the calcium supplementation. The answer to this question remains uncertain. It could depend on the type of bone response observed, which could differ according to the type of the supplemented calcium salt. As mentioned earlier, with milk calcium – phosphate salt [123], the increase in aBMD was associated with an increase of bone size. One year after discontinuing the intervention, differences in the gain in aBMD and in the size of some bones were still de-

BONJOUR AND RIZZOLI

tectable at the limit of statistical significance [123,164]. We observed that this difference was still present 3.5 years after discontinuation of the supplementation [165]. These results need additional confirmation by long-term follow-up of the cohort, ideally until peak bone mass has been attained, as well as by other prospective studies. Nevertheless, they apparently differ from results obtained with other calcium salt supplements [156,166]. As a matter of fact, calcium given in other forms to pre- or peripubertal girls does not appear to modify bone size [156 – 159] or to induce a persistent effect after stopping the intervention [167,168]. This comparative analysis suggests that the positive effects observed on the aBMD or BMC gain with citrate malate salts [156,157] or carbonate alone [158] could be primarily related to an increment in the volumetric density resulting from an inhibition of bone remodeling. Despite a positive effect on mean aBMD gain, there is still wide interindividual variability in the response to calcium supplementation. As discussed earlier, it is possible that part of the variability in the bone gain response to calcium supplementation could be related to the VDR gene polymorphisms [118].

D. Nutrients Other Than Calcium and Bone Accumulation Among nutrients other than calcium, various experimental and clinical observations point to the existence of a relationship between the level of protein intake and either calcium – phosphate metabolism or bone mass, or even osteoporotic fracture risk [169,170]. Nevertheless, any long-term influence of dietary protein on bone mineral metabolism and skeletal mass so far has been difficult to identify. Apparently contradictory information suggests that either a deficient or an excessive protein supply could negatively affect the balance of calcium and the amount of bony tissue contained in the skeleton [169,170]. Despite these uncertainties, multiple animal and human studies indicate strongly that low protein intake per se could be particularly detrimental for both the acquisition of bone mass and the conservation of bone integrity with aging. During growth, undernutrition, including an inadequate supply of energy and protein, can severely impair bone development. Studies in experimental animals indicate that isolated protein deficiency leads to reduced bone mass and strength without histomorphometric evidence of osteomalacia [169,170]. Thus, an inadequate supply of protein appears to play a central role in the pathogenesis of the delayed skeletal growth and reduced bone mass observed in undernourished children. Low protein intake could be detrimental for skeletal integrity by lowering the production of IGF-I. Indeed, the hepatic production and plasma concentration of this growth

633

CHAPTER 25 Bone Acquisition in Adolescence

factor, which exerts several positive effects on the skeleton, are under the influence of dietary protein [171]. Protein restriction has been shown to reduce circulating IGF-I by inducing resistance to the hepatic action of the growth hormone [172]. In addition, protein restriction appears to decrease the anabolic actions of IGF-I on some target cells. In this regard, it is important to note that growing rats maintained on a low protein diet failed to restore growth when IGF-I was administered at doses sufficient to normalize its plasma concentrations. Variations in the production of IGF-I could explain some of the changes in bone and calcium – phosphate metabolism that have been observed in relation to the intake of dietary protein. Indeed, the plasma level of IGF-I is closely related to the growth rate of the organism. In humans, circulating IGF-I, of which the major source is the liver, rises progressively from 1 year of age to reach peak values during puberty. As described previously, this factor appears to play a key role in calcium – phosphate metabolism during growth by stimulating two kidney processes: Pi transport and the production of calcitriol [93,173]. IGF-I is considered an essential factor for bone longitudinal growth, as it stimulates proliferation and differentiation of chondrocytes in the epiphyseal plate [174]. It also plays a role on trabecular and cortical bone formation. In experimental animals, administration of IGF-I also affects bone mass positively [175], increasing the external diameter of long bone, probably by enhancing the process of periosteal apposition. Therefore, during adolescence a relative deficiency in IGF-I or a resistance to its action that could be due to an inadequate protein supply may result not only in a reduction in the skeletal longitudinal growth, but also in an impairment in widthwise or cross-sectional bone development. In well-nourished children and adolescents, the question arises of whether variations in the protein intake within the “normal” range can influence skeletal growth and thereby modulate the genetic potential in peak bone mass attainment. There is a positive relationship between protein intake, as assessed by two 5-day dietary diary methods with weighing most food intakes [170,176], and bone mass gain during pubertal maturation [170]. Because both bone mass and protein intake increase in both sexes during adolescence, it is not surprising to find a positive correlation between these two variables. However, we found that the correlation remained statistically significant even after correcting for the influence of either age or pubertal stage. The association between bone mass gain and protein intake was observed in both sexes at the lumbar spine, the proximal femur, and the femoral midshaft. The association appeared to be particularly significant from pubertal stage P2 to P4. However, these results should not be interpreted as evidence for a causal relationship between protein intake and bone mass gain. Indeed, it is quite possible that protein, which overall was related to

the amount of ingested calories in our cohort, is to a large extent determined by growth requirements during childhood and adolescence. As was the situation for other nutrients such as calcium, only prospective interventional studies will establish whether variations in protein intake within the range recorded in our Western “well-nourished” population can affect bone mass accumulation during growth. Such prospective intervention studies should delineate the crucial years during which modifications in nutrition would be particularly effective for bone mass accumulation in children and adolescents. This kind of information is of importance in order to make credible and well-targeted recommendations for osteoporosis prevention programs aimed at maximizing peak bone mass.

Acknowledgments Research by the authors of this chapter is supported by the Swiss National Science Foundation (Grant Nos. 32-49757.96, 32-58880.99, and 32-58962.99), Nestec Ltd., Cerin, Novartis, Institut Candia.

References 1. E. Seeman and J. L. Hopper, Genetic and environmental components of the population variance in bone density. Osteoporos. Int. 7(Suppl. 3), S10 – S16 (1997). 2. R. B. Mazess, Bone densitometry for clinical diagnosis and monitoring. In “Osteoporosis, Physiological Basis, Assessment, and Treatment” (H. F. DeLuca and R. Mazess, eds), pp. 63 – 85. Elsevier, New York, (1990). 3. P. Geusens, Photon absorptiometry in osteoporosis: Bone mineral measurements in animal models and in humans. Thesis submitted in partial fulfilment of the requirements for the degree of “Geaggregeerde voor het Hoger Onderwijs.” Orientaliste, Winksele, (1992). 4. B. L. Riggs and L. J. Melton III, “Osteoporosis: Etiology, Diagnosis and Management.” Raven Press, New York, (1988). 5. D. Uebelhart, F. Duboeuf, P. J. Meunier, and P. D. Delmas, Lateral dual-photon absorptiometry: A new technique to measure the bone mineral density at the lumbar spine. J. Bone Miner. Res. 5, 525 – 531 (1990). 6. D. O. Slosman, R. Rizzoli, A. Donath, and J. P. Bonjour, Vertebral bone mineral density measured laterally by dual-energy X-ray absorptiometry. Osteopor. Int. 1, 23 – 29 (1990). 7. D. K. Katzman, L. K. Bachrach, D. R. Carter, and R. Marcus, Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J. Clin. Endocrinol. Metab. 73, 1332 – 1339 (1991). 8. C. H. Turner, Toward a cure for osteoporosis: Reversal of excessive bone fragility. Osteopor. Int. 2, 12 – 19 (1991). 9. H. K. Genant, J. E. Block, P. Steiger, and C. C. Glueer, Quantitative computed tomography: Update 1989, In “Osteoporosis: Physiological Basis, Assessment, and Treatment” (H. F. DeLuca and R. Mazess, eds), pp. 87 – 98. Elsevier, New York, (1990). 10. P. Brinckmann, M. Biggemann, and D. Hillweg, Prediction of the compressive strength of human lumbar vertebrae. Spine 14, 606 – 610 (1989).

634 11. V. Gilsanz, M. L. Loro, T. F. Roe, J. Sayre, R. Gilsanz, and E. E. Schulz, Vertebral size in elderly women with osteoporosis: Mechanical implications and relationship to fractures. J. Clin. Invest. 95, 2332 – 2337 (1995). 12. M. Trotter and B. B. Hixon, Sequential changes in weight, density, and percentage ash weight of human skeletons from an early fetal period through old age. Anat. Rec. 179, 1 – 18 (1974). 13. A. Donath, P. Indermuhle, and R. Baud, Mineralometrie osseuse mesuree par l’absorption des photons d’une source d’ I125. Radiol. Clin. Biol. 43, 393 – 400 (1974). 14. S. Krabbe, C. Christiansen, P. Rodbro, and I. Transbol, Effect of puberty on rates on bone growth and mineralization. Arch. Dis. Child. 54, 950 – 953 (1979). 15. S. L. Hui, C. C. Johnston, and R. B. Mazess, Bone mass in normal children and serum testosterone. Growth 49, 34 – 43 (1985). 16. C. Glastre, P. Braillon, L. David, P. Cochat, P. J. Meunier, and P. D. Delmas, Measurement of bone mineral content of the lumbar spine by dual energy X-ray absorptiometry in normal children: Correlations with growth parameters. J. Clin. Endocrinol. Metab. 70, 1330 – 1333 (1990). 17. J. P. Bonjour, G. Theintz, B. Buchs, D. Slosman, and R. Rizzoli, Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991). 18. P. Geusens, F. Cantatore, J. Nijs, W. Proesmans, F. Emma, and J. Dequeker, Heterogeneity of growth of bone in children at the spine, radius and total skeleton. Growth Dev. Aging 55, 249 – 256 (1991). 19. V. Gilsanz, I. Boechat, T. F. Roe, M. L. Loro, J. W. Sayre, and W. G. Goodman, Gender differences in vertebral body sizes in children and adolescents. Radiology 190, 673 – 677 (1994). 20. A. L. Merz, M. Trotter, and R. R. Peterson, Estimation of skeleton weight in the living. Am. J. Phys. Anthropol. 14, 589 – 610 (1956). 21. H. E. Meema, Cortical bone atrophy and osteoporosis as a manifestation of aging. Am. J. Roentgenol. 89, 1287 – 1295 (1963). 22. S. Aharinejad, R. Bertagnoli, K. Wicke, W. Firbas, and B. Schneider, Morphometric analysis of vertebrae and intervertebral discs as a basis of disc replacement. Am. J. Anat. 189, 69 – 76 (1990). 23. E. Seeman, From density to structure: Growing up and growing old on the surfaces of bone. J. Bone Miner. Res. 12, 509 – 521 (1997). 24. P. Coupron, “Donnees histologiques quantitatives sur le vieillissement osseux humain.” M.D. thesis, Univ. Claude Bernard, Lyon (1972). 25. M. S. Dunnill, J. A. Anderson, and R. Whitehead, Quantitative histological studies on age changes in bone. J. Pathol. Bacteriol. 94, 275 – 291 (1967). 26. J. S Arnold, Amount and quality of trabecular bone in osteoporotic vertebral fractures. Clin. Endocrinol. Metab. 2, 221 – 238 (1973). 27. W. A. Kalender, D. Felsenberg, O. Louis, P. Lopez, E. Klotz, M. Osteaux, and J. Fraga, Reference values for trabecular and cortical vertebral bone density in single and dual-energy quantitative computed tomography. Eur. J. Radiol. 9, 75 – 80 (1989). 28. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechat, H. K. Huang, E. E. Schulz, C. R. Libanati, and C. C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). 29. V. Gilsanz, A. Kovanlikaya, G. Costin, T. F. Roe, J. Sayre, and F. Kaufman, Differential effect of gender on the sizes of the bones in the axial and appendicular skeletons. J. Clin. Endocrinol. Metab. 82, 1603 – 1607 (1997). 30. V. Gilsanz, D. L. Skaggs, A. Kovanlikaya, J. Sayre, M. Luiza Loro, F. Kaufman, and S. G. Korenman, Differential effect of race on the axial and appendicular skeletons of children. J. Clin. Endocrinol. Metab. 83, 1420 – 1427 (1998). 31. E. Seeman, Genetic and environmental determinants of variance in bone size, mass, and volumetric density of the proximal femur. In

BONJOUR AND RIZZOLI

32.

33.

34. 35. 36.

37.

38.

39.

40.

41.

42.

43.

44. 45. 46.

47. 48.

49.

50.

“The Genetics of Osteoporosis and Metabolic Bone Disease” (M. J. Econs, ed.), pp.1 – 27. Humana Press, Totowa (2000). V. Gilsanz, T. F. Roe, S. Mora, G. Costin, and W. G. Goodman, Changes in vertebral bone density in black girls and white girls during childhood and puberty. N. Engl. J. Med. 325, 1597 – 1600 (1991). N. H. Bell, J. Shary, J. Stevens, M. Garza, L. Gordon, and J. Edwards, Demonstration that bone mass is greater in black than in white children. J. Bone Miner. Res. 6, 719 – 723 (1991). V. Gilsanz, Phenotype and genotype of osteoporosis. Trends Endocrinol. Metab. 9, 184 – 190 (1998). E. Seeman, Osteoporosis in men. Osteopor. Int. 9(Suppl. 2), S97 – S110 (1999). Z. H. Han, S. Palnitkar, D. S. Rao, D. Nelson, and A. M. Parfitt, Effects of ethnicity and age or menopause on the structure and geometry of iliac bone. J. Bone Miner. Res. 11, 1967 – 1975 (1996). P. Ammann, R. Rizzoli, J. M. Meyer, and J. P Bonjour, Bone density and shape as determinants of bone strength in IGF-I and/or pamidronate-treated ovariectomized rats. Osteopor. Int. 6, 219 – 227 (1996). P. E. Fournier, R. Rizzoli, D. O. Slosman, B. Buchs, and J. P. Bonjour, Relative contribution of vertebral body and posterior arch in female and male lumbar spine peak bone mass. Osteopor. Int. 4, 264 – 272 (1994). G. Theintz, B. Buchs, R. Rizzoli, D. Slosman, H. Clavien, P. C. Sizonenko, and J. P. Bonjour, Longitudinal monitoring of bone mass accumulation in healthy adolescents: Evidence for a marked reduction after 16 years of age at the levels of lumbar spine and femoral neck in female subjects. J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992). S. Bass, P. D. Delmas, G. Pearce, E. Hendrich, A. Tabensky, and E. Seeman, The differing tempo of growth in bone size, mass, and density in girls is region-specific. J. Clin. Invest. 104, 795 – 804 (1999). P. E. Fournier, R. Rizzoli, D. O. Slosman, G. Theintz, and J. P. Bonjour, Asynchrony between the rates of standing height gain and bone mass accumulation during puberty. Osteopor. Int. 7, 525 – 532 (1997). L. A. Landin, Fracture patterns in children: Analysis of 8682 fractures with special reference to incidence, etiology and secular changes in a Swedish urban population 1950 – 1979. Acta Orthop. Scand. Suppl. 202, 1 – 109 (1983). D. A. Bailey, J. H. Wedge, R. G. McCullogh, A. D. Martin, and S. C. Bernhardson, Epidemiology of fractures of the distal end of the radius in children as associated with growth. J. Bone Jt. Surg. 71A, 1225 – 1230 (1989). P. A. Alffram and G. C. H. Bauer, Epidemiology of fractures of the forearm. J. Bone Jt. Surg. 44A, 105 – 114 (1962). A. M. Parfitt, The two faces of growth: Benefits and risks to bone integrity. Osteopor. Int. 4, 382 – 398 (1994). C. J. R. Blimkie, J. Lefevre, G. P. Beunen, R. Renson, J. Dequeker, and P. Van Damme, Fractures, physical activity, and growth velocity in adolescent Belgian boys. Med. Sci. Sports Exercise 25, 801 – 808 (1993). T. A. Einhorn, Bone strength: The bottom line. Calcif. Tissue Int. 51, 333 – 339 (1992). J. L. Hopper, R. M. Green, C. A. Nowson, D. Young, A. J. Sherwin, B. Kaymakci, R. G. Larkins, and J. D. Wark, Genetic, common environment, and individual specific components of variance for bone mineral density in 10- to 26-year-old females: A twin study. Am. J. Epidemiol. 147, 17 – 29 (1998). V. Matkovic, T. Jelic, G. M. Wardlaw, J. Z. Ilich, P. K. Goel, J. K. Wright, M. B. Andon, K. T. Smith, and R. P. Heaney, Timing of peak bone mass in caucasian females and its implication for the prevention of osteoporosis: Inference from a cross-sectional model. J. Clin. Invest. 93, 799 – 808 (1994). H. Kroger, A. Kotaniemi, L. Kroger, and E. Alhava, Development of bone mass and bone density of the spine and femoral neck: A

635

CHAPTER 25 Bone Acquisition in Adolescence

51.

52.

53.

54.

55. 56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

prospective study of 65 children and adolescents. Bone Miner. 23, 171 – 182 (1993). M. Gunnes, Bone mineral density in the cortical and trabecular distal forearm in healthy children and adolescents. Acta Paediatr. 83, 463 – 467 (1994). J. C. Ruiz, C. Mandel, and M. Garabedian, Influence of spontaneous calcium intake and physical exercise on the vertebral and femoral bone mineral density of children and adolescents. J. Bone Miner. Res. 10, 675 – 682 (1995). D. Young, J. L. Hopper, C. A. Nowson, R. M. Green, A. J. Sherwin, B. Kaymakci, M. Smid, C. S. Guest, R. G. Larkins, and J. D. Wark, Determinants of bone mass in 10- to 26-year-old females: A twin study. J. Bone Miner. Res. 10, 558 – 567 (1995). V. Gilsanz, D. T. Gibbens, M. Carlson, M. I. Boechat, C. E. Cann, and E. E. Schulz, Peak trabecular vertebral density: A comparison of adolescent and adult females. Calcif. Tissue Int. 43, 260 – 262 (1988). R. A. Bloom and J. W. Laws, Humeral cortical thickness as an index of osteoporosis in women. Br. J. Radiol. 43, 522 – 527 (1970). J. Andresen and H. E. Nielsen, Metacarpal bone mass in normal adults and in patients with chronic renal failure. In “Non-invasive Bone Measurements: Methodological Problems” (J. Dequeker and C. C. Johnson, Jr., eds.), pp. 169 – 173. IRL Press, (1982). B. Krolner and S. Pors Nielsen, Bone mineral content of the lumbar spine in normal and osteoporotic women: Cross-sectional and longitudinal studies. Clin. Sci. 62, 329 – 336 (1982). C. Christiansen, Bone mineral measurement with special reference to precision, accuracy, normal values, and clinical relevance, In “Noninvasive Bone Measurements: Methodological Problems” (J. Dequeker and C. C. Johnson, Jr., eds.), pp. 95 – 105. IRL Press, (1982). P. Geusens, J. Dequeker, A. Verstraeten, and J. Nijs, Age-,sex-, and menopause-related changes of vertebral and peripheral bone: Population study using dual and single photon absorptiometry and radiogrammetry. J. Nuclear. Med. 27, 1540 – 1549 (1986). B. L. Riggs, H. W. Wahner, L. J. Melton III, L. S. Richelson, H. L. Judd, and K. P. Offord, Rates of bone loss in the appendicular and axial skeletons of women. Evidence of substantial vertebral bone loss before menopause. J. Clin. Invest. 77, 1487 – 1491 (1986). R. B. Mazess, H. S. Barden, M. Ettinger, C. Johnston, B. DawsonHughes, D. Baran, M. Powell, and M. Notelovitz, Spine and femur density using dual-photon absorptiometry in U.S. white women. Bone Miner. 2, 211 – 219 (1987). J. E. Block, R. Smith, C. C. Glueer, P. Steiger, B. Ettinger, and H. K. Genant, Models of spinal trabecular bone loss as determined by quantitative computed tomography. J. Bone Miner. Res. 4, 249 – 257 (1989). L. R. Hedlund and J. C. Gallagher, The effect of age and menopause on bone mineral density of the proximal femur. J. Bone Miner. Res. 4, 639 – 642 (1989). S. Ortolani, C. Trevisan, M. L. Bianchi, M. P. Caraceni, F. M. Ulivieri, G. Gandolini, A. Montesano, and E. E. Polli, Spinal and forearm bone mass in relation to ageing and menopause in healthy Italian women. Eur. J. Clin. Invest. 21, 33 – 39 (1991). B. Buchs, R. Rizzoli, D. Slosman, V. Nydegger, and J. P. Bonjour, Densite minerale osseuse de la colonne lombaire, du col et de la diaphyse femoraux d’un echantillon de la population genevoise. Schweiz. Med. Wschr. 122, 1129 – 1136 (1992). D. O. Slosman, R. Rizzoli, C. Pichard, A. Donath, and J. P. Bonjour, Longitudinal measurement of regional and whole body bone mass in young healthy adults. Osteoporos. Int. 4, 185 – 190 (1994). J. D. Ringe, Precision and clinical application of peripheral single photon absorptiometry. In “Non-invasive Bone Measurements: Methodological Problems” (J. Dequeker and C. C. Johnston, Jr., eds.), pp. 47 – 54. IRL Press, Oxford, (1982). A. Rodin, B. Murby, M. A. Smith, M. Caleffi, I. Fentiman, M. G. Chapman, and I. Fogelman, Premenopausal bone loss in the lumbar

69.

70.

71.

72. 73.

74. 75. 76.

77.

78.

79. 80.

81.

81a.

82.

83.

84.

85.

86.

87.

spine and neck of femur: A study of 225 Caucasian women. Bone 11, 1 – 5 (1990). R. R. Recker, K. M. Davies, S. M. Hinders, R. P. Heaney, M. R. Stegman, and D. B. Kimmel, Bone gain in young adult women. JAMA 268, 2403 – 2407 (1992). S. M. Garn, C. G. Rohnmann, B. Wagner, and W. Ascoli, Continuing bone growth throughout life: A general phenomenon. Am. J. Phys. Anthrop. 26, 313 – 318 (1967). S. M. Garn, B. Wagner, C. G. Rohmann, and W. Ascoli, Further evidence for continuing bone expansion. Am. J. Phys. Anthrop. 28, 219 – 222 (1968). C. B. Ruff and W. C. Hayes, Sex differences in age-related remodeling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). J. P. Bonjour, J. Caverzasio, and R. Rizzoli, Homeostasis of inorganic phosphate and the kidney. In “Rickets, Nestle Nutrition Workshop Serie” (F. H. Glorieux, ed.), Vol. 21, pp. 35 – 46. Raven Press, New York, (1991). J. L. Rosen and R. W. Chesney, Circulating calcitriol concentrations in health and disease. J. Pediatr. 103, 1 – 17 (1983). J. Corvilain and M. Abramow, Growth and renal control of plasma phosphate. J. Clin. Endocrinol. 34, 452 – 459 (1972). J. M. Round, S. Butcher, and R. Steele, Changes in plasma inorganic phosphorus and alkaline phosphatase activity during the adolescent growth spurt. Ann. Hum. Biol. 6, 129 – 136 (1979). S. Krabbe, I. Transbol, and C. Christiansen, Bone mineral homeostasis, bone growth, and mineralisation during years of pubertal growth: A unifying concept. Arch. Dis. Child. 57, 359 – 363 (1982). J. Caverzasio and J. P. Bonjour, IGF-I, a key regulator of renal phosphate transport and 1,25-Dihydroxyvitamine D3 production during growth. News Physiol. Sci. 6, 206 – 210 (1991). L. E. Underwood, A. J. D’Ercole, and J. J. Van Wyk, Somatomedin C and the assessment of growth. Pediatr. Clin. North Am. 27, 771 (1980). S. Krabbe, C. Christiansen, P. Rodbro, and I. Transbol, Pubertal growth as reflected by simultaneous changes in bone mineral content and serum alkaline phosphatase. Acta Pediatr. Scand. 69, 49 – 52 (1980). S. Krabbe and C. Christiansen, Longitudinal study of calcium metabolism in male puberty. I. Bone mineral content and serum levels of alkaline phosphatase, phosphate and calcium. Acta Pediatr. Scand. 73, 745 – 749 (1984). S. Krabbe, L. Hummer, and C. Christiansen, Longitudinal study of calcium metabolism in male puberty. II. Relationship between mineralization and serum testosterone. Acta Pediatr. Scand. 73, 750 – 755 (1984). B. J. Riis, S. Krabbe, C. Christiansen , B. D. Catherwood, and L. J. Deftos, Bone turnover in male puberts: A longitudinal study. Calcif. Tissue Int. 37, 213 – 217 (1985). P. Szulc, E. Seeman, and P. D. Delmas, Biochemical measurements of bone turnover in children and adolescents. Osteopor. Int. 11, 281 – 294 (2000). P. D. Delmas, P. Chatelain, L. Malaval, and G. Bonne, Serum GLAprotein in growth hormone deficient children. J. Bone Miner. Res. 1, 333 – 338 (1986). J. S. Johansen, A. Giwercman, D. Hartwell, C. T. Nielsen, P. A. Price, C. Christiansen, and N. E. Skakkebak, Serum bone Gla-protein as a marker of bone growth in children and adolescents: Correlation with age, height, serum insulin-like growth factor I, and serum testosterone. J. Clin. Endocrinol. Metab. 67, 273 – 278 (1988). J. R. Ducharme and M. G. Forest, Développement pubertaire normal. In “Endocrinologie Pédiatrique, Hormones Gonadiques et Puberté” (J. Bertrand et al., eds.), pp. 315 – 335. Payot, Lausanne, (1982). J. L. Van den Brande, Régulation endocrinienne de la croissance. In “Endocrinologie Pédiatrique” (J. Bertrand et al., eds.), pp. 159 – 181. Payot, Lausanne, (1982).

636 88. J. Cadogan, A. Blumensohn, M. E. Barker, and R. Eastell, A longitudinal study of bone gain in pubertal girls: Anthropometric and biochemical correlates. J. Bone Miner. Res. 13, 1602 – 1612 (1998). 89. N. A. Pocock, J. A. Eisman, J. L. Hopper, M. G. Yeates, P. N. Sambrook, and S. Eberl, Genetic determinants of bone mass in adults: A twin study. J. Clin. Invest. 80, 706 – 710 (1987). 90. J. Dequeker, N. Nijs, A. Vestraeten, P. Geusens, and G. Gevers, Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8, 207 – 209 (1987). 91. C. W. Slemenda, J. C. Christian, C. J. Williams, J. A. Norton, and C. C. Johnston, Genetic determinants of bone mass in adult women: A reevaluation of the twin model and the potential importance of gene interaction on heritability estimates. J. Bone Miner Res. 6, 561 – 567 (1991). 92. C. W. Slemenda, J. Z. Miller, S. L. Hui, T. K. Reister, and C. C. Johnston Jr., Role of physical activity in the development of skeletal mass in children. J. Bone Miner. Res. 6, 1227 – 1233 (1991). 93. J. P. Bonjour, G. Theintz, F. Law, D. O. Slosman, and R. Rizzoli, Calcium-phosphate metabolism and bone mass accumulation during puberty. In “Fourth International Symposium on Osteoporosis and Consensus Development Conference”, pp. 205 – 207. Hong Kong, (1993). 94. E. Seeman, J. L. Hopper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554 – 558 (1989). 95. M. E. Cohen-Solal, C. Baudoin, M. Omouri, D. Kuntz, and M. C De Vernejoul, Bone mass in middle-aged osteoporotic men and their relatives: Familial effect. J. Bone Miner Res. 13, 1909 – 1914 (1998). 96. P. J. Kelly, N. A. Morrison, P. N. Sambrook, T. V. Nguyen, and J. A. Eisman, Genetic influences on bone turnover, bone density and fracture. Eur. J. Endocrinol. 133, 265 – 271 (1995). 97. T. V. Nguyen, G. M. Howard, P. J. Kelly, and J. A. Eisman, Bone mass, lean mass, and fat mass: Same genes or same environments? Am. J. Epidemiol. 147, 3 – 16 (1998). 98. E. Seeman, J. L. Hopper, N. R. Young, C. Formica, P. Goss, and C. Tsalamandris, Do genetic factors explain associations between muscle strength, lean mass, and bone density? A twin study. Am. J. Physiol. 270, E320 – E327 (1996). 99. F. A. Tylavsky, A. D. Bortz, R. L. Hancock, and J. J. B. Anderson, Familial resemblance of radial bone mass between premenopausal mothers and their college-age daughters. Calcif. Tissue Int. 45, 265 – 272 (1989). 100. E. A. Krall and B. Dawson-Hughes, Heritable and life-style determinants of bone mineral density. J. Bone Miner Res. 8, 1 – 9 (1993). 101. S. Ferrari, R. Rizzoli, D. Slosman, and J. P. Bonjour, Familial resemblance for bone mineral mass is expressed before puberty. J. Clin. Endocrinol. Metab. 83, 358 – 361 (1998). 102. P. Garnero, N. K. Arden, G. Griffiths, P. D. Delmas, and T. D. Spector, Genetic influence on bone turnover in postmenopausal twins. J. Clin. Endocrinol. Metab. 81, 140 – 146 (1996). 103. P. Garnero, O. Borel, E. Sornay-Rendu, and P. D. Delmas, Vitamin D receptor gene polymorphisms do not predict bone turnover and bone mass in healthy premenopausal women. J. Bone Miner Res. 10, 1283 – 1288 (1995). 104. P. Garnero, O. Borel, E. Sornay-Rendu, M. E. Arlot, and P. D. Delmas, Vitamin D receptor gene polymorphisms are not related to bone turnover, rate of bone loss, and bone mass in postmenopausal women: The OFELY study. J. Bone Miner Res. 11, 827 – 834 (1996). 105. F. Hustmyer, M. Peacock, S. Hui, C. C. Johnston, and J. Christian, Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J. Clin. Invest. 94, 2130 – 2134 (1994). 106. R. Kikuchi, T. Uemura, I. Gorai, S. Ohno, and H. Minaguchi, Early and late postmenopausal bone loss is associated with BsmI vitamin D receptor gene polymorphism in Japanese women. Calcif. Tissue Int. 64, 102 – 106 (1999).

BONJOUR AND RIZZOLI 107. J. M. Zmuda, J. A. Cauley, M. E. Danielson, R. L. Wolf, and R. E. Ferrell, Vitamin D receptor gene polymorphisms, bone turnover, and rates of bone loss in older African-American women. J. Bone Miner. Res. 12, 1446 – 1452 (1997). 108. A. M. Tokita, H. Matsumoto, N. A. Morrison, T. Tawa, Y. Miura, K. Fukamauchi, N. Mitsuhashi, M Irimoto, S. Yamamori, M. Miura, T. Watanabe, Y. Kuwabara, K. Yabuta, and J. A. Eisman, Vitamin D receptor alleles, bone mineral density and turnover in premenopausal Japanese women. J. Bone Miner Res. 11, 1003 – 1009 (1996). 109. A. G. Uitterlinden, H. A. P. Pols, H. Burger, Q. Huang, P. L. A. Van Daele, C. M. Van Duijn, A. Hofman, J. C. Birkenhäger, and J. P. T. M. Van Leeuwen, A large-scale population-based study of the association of vitamin D receptor gene polymorphisms with bone mineral density. J. Bone Miner Res. 11, 1241 – 1248 (1996). 110. H. L. Jørgensen, J. Schøller, J. C. Sand, M. Bjuring, C. Hassager, and C. Christiansen, Relation of common allelic variation at vitamin D receptor locus to bone mineral density and postmenopausal bone loss: Cross-sectional and longitudinal population study. Br. Med. J. 313, 586 – 590 (1996). 111. G. S. Cooper and D. M. Umbach, Are vitamin D receptor polymorphisms associated with bone mineral density? A meta-analysis. J. Bone Miner. Res. 11, 1841 – 1849 (1996). 112. C. Gross, T. R. Eccleshall, P. J. Malloy., M. L. Villa, R. Marcus, and D. Feldman, The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal Mexican-American women. Bone Miner. Res. 11, 1850 – 1855 (1996). 113. S. S. Harris, T. R. Eccleshall, C. Gross, B. Dawson-Hughes, and D. Feldman, The vitamin D receptor start codon polymorphism (FokI) and bone mineral density in premenopausal American black and white women. J. Bone Miner Res. 12, 1043 – 1048 (1997). 114. H. Arai, K. I. Miyamoto, Y. Taketani, H. Yamamoto, Y. Iemori, K. Morita, T. Tonai, T. Nishisho, S. Mori, and E. Takeda, A vitamin D receptor gene polymorphism in the translation initiation codon: Effect on protein activity and relation to bone mineral density in Japanese women. J. Bone Miner. Res. 12, 915 – 921 (1997). 115. T. R. Eccleshall, P. Garnero, C. Gross, P. D. Delmas, and D. Feldman, Lack of correlation between start codon polymorphism of the vitamin D receptor gene and bone mineral density in premenopausal French women: The OFELY study. J. Bone Miner Res. 13, 31 – 35 (1998). 116. S. Ferrari, R. Rizzoli, D. Manen, D. Slosman, and J. P. Bonjour, Vitamin D receptor gene start codon polymorphisms (FokI) and bone mineral density: Interaction with age, dietary calcium and 3’-end region polymorphisms. J. Bone Miner Res. 13, 363 – 370 (1998). 117. J. Sainz, J. M. Van Tornout, M. L. Loro, J. Sayre, T. F. Roe, and V. Gilsanz, Vitamin D-receptor gene polymorphisms and bone density in prepubertal American girls of Mexican descent. N. Engl. J. Med. 337, 77 – 82 (1997). 118. S. Ferrari, R. Rizzoli, D. Slosman, and J. P. Bonjour, Do dietary calcium and age explain the controversy surrounding the relationship between bone mineral density and vitamin D receptor gene polymorphisms? J. Bone Miner Res. 13, 363– 370 (1998). 119. S. Ferrari, J. P. Bonjour, and R. Rizzoli, The vitamin D receptor gene and calcium metabolism. Trends Endocrinol. Metab. 9, 259 – 265 (1998). 120. M. J. Barger-Lux, R. P. Heaney, J. Hayes, H. F. DeLuca, M. L. Johnson, and G. Gong, Vitamin D receptor gene polymorphisms, bone mass, body size, and vitamin D receptor density. Calcif. Tissue Int. 57, 161 – 162 (1995). 121. M. Lorentzon, R. Lorentzon, and P Nordstrom, Vitamin D receptor gene polymorphism is associated with birth height, growth to adolescence, and adult stature in healthy Caucasian men: A cross-sectional and longitudinal study. J. Clin. Endocrinol. Metab. 85, 1666 – 1670 (2000).

637

CHAPTER 25 Bone Acquisition in Adolescence 122. C. Tao, T. Yu, S. Garnett, J. Briody, J. Knight, H. Woodhead, and C. T. Cowell, Vitamin D receptor alleles predict growth and bone density in girls. Arch. Dis. Child. 79, 488 – 493 (1998). 123. J. P. Bonjour, A. L. Carrié, S. Ferrari, H. Clavien, D. Slosman, G. Theintz, and R. Rizzoli, Calcium enriched foods and bone mass growth in prepubertal girls: A randomized, double-blind, placebocontrolled trial. J. Clin. Invest. 99, 1287 – 1294 (1997). 124. E. W. Gregg, R. L. Wolf, A. M. Kriska, R. E. Ferrell, L. M. Salamone, D. Black, L. H. Kuller, and J. A. Cauley, Does the relationship between vitamin D receptor genotypes and bone parameters vary with physical activity level? J. Bone Miner Res. 11 (Suppl. 1), S210 (1996). 125. E. L. Smith and C. Gilligan, Physical activity effects on bone metabolism. Calcif. Tissue Int. 49, S50 – S54 (1991). 126. M. R. Forwood and D. B Burr, Physical activity and bone mass: Exercises in futility? Bone Miner. 21, 89 – 112 (1993). 127. P. Kannus, H. Haapasalo, M. Sankelo, H. Sievanen, M. Pasanen, A. Heinonen, P. Oja, and I. Vuori, Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann. Intern. Med. 123, 27 – 31 (1995). 128. A. Huddleston, D. Rockwell, D. Kulund, and B. Harrison, Bone mass in lifetime tennis athletes. JAMA 244, 1107 – 1109 (1980). 129. H. H. Jones, J. D. Priest, W. C. Hayes, C. C. Tichenor, and D. A. Nagel, Humeral hypertrophy in response to exercise. J. Bone Joint Surg. 59A, 204 – 208 (1977). 130. S. Bass, G. Pearce, M. Bradney, E. Hendrich, P. Delmas, A. Harding, and E. Seeman, Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepubertal and retired female gymnasts. J. Bone Miner. Res. 13, 500 – 507 (1998). 131. N. Young, C. Formica, G. Szmukler, and E. Seeman, Bone density at weight-bearing and non-weight-bearing sites in ballet dancers: The effects to exercise, hypogonadism, and body weight. J. Clin. Endocrinol. Metab. 78, 449 – 454 (1994). 132. M. K. Karlsson, O. Johnell, and K. J. Obrant, Is bone mineral density advantage maintained long-term in previous weight lifters? Calcif. Tissue Int. 57, 325 – 328 (1995). 133. K. M. Kahn, K. L. Bennell, J. L. Hopper, L. Flicker, C. A. Nowson, A. J. Sherwin, K. J. Crichton, P. R. Harcourt, and J. D. Wark, Self-reported ballet classes undertaken at age 10 – 12 years and hip bone mineral density in later life. Osteopor. Int. 8, 165 – 173 (1998). 134. F. L. Morris, W. R. Payne, and J. D. Wark, The impact of intense training on endogenous estrogen and progesterone concentrations and bone mineral acquisition in adolescent rowers. Osteopor. Int. 10, 361 – 368 (1999). 135. H. Haapasalo, S. Kontulainen, H. Sievanen, P. Kannus, M. Jarvinen, and I. Vuori, Exercise-induced bone gain is due to enlargement in bone size without a change in volumetric bone density: A peripheral quantitative computed tomography study of the upper arms of male tennis players. Bone 27, 351 – 357 (2000). 136. H. Kroger, A. Kotaniemi, P. Vainio, and E. Alhava, Bone densitometry of the spine and femur in children by dual-energy X-ray absorptiometry. Bone Miner. 17, 75 – 85 (1992). 137. S. K. Grimston, K. Morrison, J. A. Harder, and D. A. Hanley, Bone mineral density during puberty in western Canadian children. Bone Miner. 19, 85 – 96 (1992). 138. J. C. Cheng, N. Maffulli, S. S. Leung, W. T. Lee, J. T. Lau, and K. M. Chan, Axial and peripheral bone mineral acquisition: A 3-year longitudinal study in Chinese adolescents. Eur. J. Pediatr. 158, 506 – 512 (1999). 139. J. Z. Ilich, M. Skugor, T. Hangartner, A. Baoshe, and V. Matkovic, Relation of nutrition, body composition and physical activity to skeletal development: a cross-sectional study in preadolescent females. J. Am. Coll. Nutr. 17, 136 – 147 (1998). 140. F. L. Morris, G. A. Naughton, J. L. Gibbs, J. S. Carlson, and J. D. Wark, Prospective ten-month exercise intervention in premenarcheal

141.

142. 143.

144.

145.

146.

147.

148. 149.

150. 151.

152.

153.

154.

155.

156.

157.

158.

girls: Positive effects on bone and lean mass. J. Bone Miner. Res. 12, 1453 – 1462 (1997). M. Bradney, G. Pearce, G. Naughton, C. Sullivan, S. Bass, T. Beck, J. Carlson, and E. Seeman, Moderate exercise during growth in prepubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J. Bone Miner. Res. 13, 1814 – 1821 (1998). R. Marcus, Exercise: Moving in the right direction. J. Bone Miner. Res. 13, 1793 – 1796 (1998). C. Lindholm, K. Hagenfeldt, and H. Ringertz, Bone mineral content of young female former gymnasts. Acta Paediatr. 84, 1109 – 1112 (1995). E. M. Kirchner, R. D. Lewis, and P. O’Connor, Effect of past gymnastics participation on adult bone mass. J. Appl. Physiol. 80, 226 – 232 (1996). S. Kontulainen, P. Kannus, H. Haapasalo, A. Heinonen, H. Sievanen, P. Oja, and I. Vuori, Changes in bone mineral content with decreased training in competitive young adult tennis players and controls: A prospective 4-yr follow-up. Med. Sci. Sports Exerc. 31, 646 – 652 (1999). M. K. Karlsson, C. Linden, C. Karlsson, O. Johnell, K. Obrant, and E. Seeman, Exercise during growth and bone mineral density and fractures in old age. Lancet 355, 469 – 470 (2000). D. Teegarden, R. M. Lyle, W. R. Proulx, C. C. Johnston, and C. M. Weaver, Previous milk consumption is associated with greater bone density in young women. Am. J. Clin. Nutr. 69, 1014 – 1017 (1999). C. M. Weaver, Calcium requirements of physically active people. Am. J. Clin. Nutr. 72 (Suppl. 2), S579 – S584 (2000). R. B. Sandler, C. W. Slemenda, R. E. LaPorte, J. A. Cauley, M. M. Schramm, M. L. Barresi, and A. M. Kriska, Postmenopausal bone density and milk consumption in childhood and adolescence. Am. J. Clin. Nutr. 42, 270 – 274 (1985). G. Chan, Dietary calcium and bone mineral status of children and adolescents. Am. J. Dis. Child. 145, 631 – 634 (1991). J. M. Sentipal, G. M. Wardlaw, J. Mahan, and V. Matkovic, Influence of calcium intake and growth indexes on vertebral bone mineral density inp young females. Am. J. Clin. Nutr. 54, 425 – 428 (1991). F. A. Tylavsky, J. J. B. Anderson, R. V. Talmage, and T. N. Taft, Are calcium intakes and physical activity patterns during adolescence related to radial bone mass of white college-age females? Osteopor. Int. 2, 232 – 240 (1992). T. Lloyd, N. Rollings, M. B. Andon, L. M. Demers, D. F. Eggli, K. Kieselhorst, H. Kulin, J. R. Landis, J. K. Martel, G. Orr, and P. Smith, Determinants of bone density in young women.I. Relationships among pubertal development, total bodybone mass, and total body bone density in premenarchal females. J. Clin. Endocrinol. Metab. 75, 383 – 387 (1992). K. Uusi-Rasi, H. Haapasalo, P. Kannus, M. Pasanen, H. Sievanen, P. Oja, and I. Vuori, Determinants of bone mineralization in 8 to 20 year old Finnish females. Eur. J. Clin. Nutr. 51, 54 – 59 (1997). V. Matkovic, D. Fontana, C. Tominac, P. Goel, and C. H. Chesnut III, Factors that influence peak bone mass formation: A study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr. 52, 878 – 888 (1990). C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). T. Lloyd, M. B. Andon, N. Rollings, J. K. Martel, J. R. Landis, L. M. Demers, D. F. Eggli, K. Kieselhorst, and H. E. Kulin, Calcium supplementation and bone mineral density in adolescent girls. JAMA 270, 841 – 844 (1993). W. T. K. Lee, S. S. F. Leung, S. H. Wang, Y. C. Xu, W. P. Zeng, J. Lau, S. J. Oppenheimer, and J. C. Cheng, Double-blind, controlled calcium supplementation and bone mineral accretion in children

638

159.

160.

161. 162.

163.

164.

165.

166. 167.

BONJOUR AND RIZZOLI accustomed to a low-calcium diet. Am. J. Clin. Nutr. 60, 744 – 750 (1994). C. A. Nowson, R. M. Green, J. L. Hopper, A.J. Sherwin, D. Young, B. Kaymakci, C. S. Guest, M. Smid, R. G. Larkins, and J. D. Wark, A co-twin study of the effect of calcium supplementation on bone density during adolescence. Osteopor. Int. 7, 219 – 225 (1997). S. A. Zamora, R. Rizzoli, D. C. Belli, D. O. Slosman, and J. P. Bonjour, Vitamin D supplementation during infancy is associated with higher bone mineral mass in prepubertal girls. J. Clin. Endocrinol. Metab. 84, 4541 – 4544 (1999). D. J. Barker, Intrauterine programming of adult disease. Mol. Med. Today 1, 418 – 423 (1995). C. Cooper, C. Fall, P. Egger, R. Hobbs, R. Eastell, and D. Barker, Growth in infancy and bone mass in later life. Ann. Rheum. Dis. 56, 17 – 21 (1997). J. Cadogan, R. Eastell, N. Jones, and M. E. Barker., Milk intake and bone mineral acquisition in adolescent girls: Randomised, controlled intervention trial. Br. Med. J. 315, 1255 – 1260 (1997). P. Ammann, R. Rizzoli, D. Slosman, and J. P. Bonjour, Calcium supplements during childhood positively influence determinants of femoral neck bone strength. J. Bone Miner. Res. 12 (Suppl. 1), S172 (1997). J. P. Bonjour, D. Slosman, D. Barclay, and R. Rizzoli, A significant effect of calcium-enriched foods on bone mineral mass gain is maintained more than 3 years after discontinuation of intervention. J. Bone Miner. Res. 14 (Suppl. 1), S311 (1999). R. Rizzoli and J. P. Bonjour, Determinants of peak bone mass and mechanisms of bone loss. Osteopor. Int. 9 (Suppl. 2), S17 – S23 (1999). C. W. Slemenda, M. Peacock, S. Hui, L. Zhou, and C. C. Johnston, Reduced rates of skeletal remodeling are associated with increased

168.

169. 170.

171.

172.

173.

174. 175.

176.

bone mineral density during the development of peak skeletal mass. J. Bone Miner. Res. 12, 676 – 682 (1997). W. T. Lee, S. S. Leung, D. M. Leung, and J. C. Cheng, A follow-up study on the effects of calcium-supplement withdrawal and puberty on bone acquisition of children. Am. J. Clin. Nutr. 64, 71 – 77 (1996). E. S. Orwoll, The effects of dietary protein insufficiency and excess on skeletal health. Bone 13, 343 – 350 (1992). J. P. Bonjour and R. Rizzoli, Inadequate protein intake and osteoporosis possible involvement of the IGF system. In “Challenges of Modern Medicine” (P. Burckhardt and R. P. Heaney, eds.), Vol. 7, pp. 399 – 406. Ares Serono Symposia Publications, (1995). W. L. Isley, L. E. Underwood, and D. R. Clemmons, Dietary components that regulate serum somatomedin-C concentrations in humans. J. Clin. Invest. 71, 175 – 182 (1983). J. P. Thissen, S. Triest, M. Maes, L. E. Underwood, and J. M. Ketelslegers, The decreased plasma concentrations of insulin-like growth factor-I in protein-restricted rats is not due to decreased number of growth hormone receptors on isolated hepatocytes. J. Endocrinol. 124, 159 – 165 (1990). J. Caverzasio, C. Montessuit, and J. P. Bonjour, Stimulatory effect of insulin-like growth factor-1 on renal Pi transport and plasma 1,25-dihydroxyvitamin D3. Endocrinology 127, 453 – 459 (1990). E. R. Froesch, C. Schmid, J. Schwander, and J. Zapf, Actions of insulin-like growth factors. Annu. Rev. Physiol. 47, 443 – 467 (1985). P. Ammann, R. Rizzoli, K. Muller, D. Slosman, and J. P. Bonjour, IGF-I and pamidronate increase bone mineral density in ovariectomized adult rats. Am. J. Physiol. 265, E770 – E776 (1993).. H. Clavien, G. Theintz, R. Rizzoli, and J. P. Bonjour, Does puberty alter dietary habits in adolescents living in a western society? J. Adolesc. Health 19, 68 – 75 (1996).

CHAPTER 26

Genetics and Genomics of Osteoporosis ANDRÉ G. UITTERLINDEN, JOHANNES P. T. M. VAN LEEUWEN, AND HUIBERT A. P. POLS Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands

I. Introduction II. Top-down and Bottom-up Genomic Approaches III. Considerations and Complicating Factors

IV. The Map of Osteoporosis Genes V. Applications and Prospects References

I. INTRODUCTION

technology soon allowed the isolation of the genes responsible for diseases such as Duchenne muscular dystrophy, cystic fibrosis, Huntington’s disease, and several others. Currently, the chromosomal position of more than 500 disease genes of the estimated 3000 monogenic diseases has been determined, and close to 100 have been cloned and characterized. In the area of bone metabolism, accretion of knowledge on the molecular genetic nature of disease has also led to important discoveries. Among the cloned disease genes responsible for Mendelian bone disorders are the genes encoding collagen type I1 (located on chromosome 17q22) and collagen type I2 (located on chromosome 7q22.1), responsible for most forms of what is the best known and characterized genetic bone disease: osteogenesis imperfecta (OI). This inherited brittle-bone disorder predisposes a patient to easy fractures, even with little trauma, and to skeletal deformity (see Chapter 50). The condition involves either qualitative or quantitative alterations in type I collagen protein, which are the result of a variety of possible small point mutations or small deletions/duplications within one of the genes that encode the chains of the collagen type I

A. Genetic Diseases The concept of “genetic” diseases has evolved substantially in recent decades. This is not only due to new insights into the genetic nature of disease, but also based on the availability of methodology to identify and characterize genetic factors predisposing to disease. The importance of acquiring knowledge of such genetic risk factors lie in two possibilities (a) to determine a “risk profile” at a very early stage, through molecular genetic techniques, even before the disease presents clinically, and (b) to design therapeutic intervention strategies on the basis of knowledge of the molecular action of the proteins involved. Initially, genetic diseases were defined as single Mendelian traits, usually with early disease onset, relatively fast progression, and showing clear Mendelian inheritance patterns in families. Since it was recognized in 1980 that the genetic inheritance patterns of these monogenic diseases could be followed using naturally occuring DNA sequence variations [1], early molecular genetic

OSTEOPOROSIS, SECOND EDITION VOLUME 1

639

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

640

UITTERLINDEN, VAN LEEUWEN, AND POLS

protein. While bone fragility is common to all forms of OI, the clinical phenotypic presentation is remarkably variable, ranging from lethal perinatal forms to only a mild increase in fracture frequency in later-onset forms of the disease. Underlying this range of variation is socalled locus and allelic heterogeneity, i.e., the disease phenotype varies according to which gene (collagen type I1 or collagen type I2) is mutated and according to the type and location of the mutation.

factors. Epidemiological studies are needed to quantify the variability of a trait and identify potentially modifying environmental factors. Genetic epidemiological studies applying molecular genetic tools can then identify chromosomal regions harboring putative candidate genes. Finally, candidate gene studies will establish the contribution of particular gene variants in explaining the variation of the trait, also in relation to gene – environment interactions, and, last but not least, investigate the underlying molecular mechanisms.

B. Complex Traits C. Osteoporosis The characterization of the molecular genetic basis of osteogenesis imperfecta and other relatively simple genetic disorders (by current standards), is still changing our concept of disease. Analysis of such diseases not only illustrates the vast and devastating effects simple mutations can have, but also generates novel technological tools accelerating the process of gene discovery and mutation detection. Together, this provides the basis to tackle the more challenging problems of the common multifactorial diseases such as osteoporosis. Many of the most important medical conditions in the western world are usually not characterized by simple Mendelian inheritance patterns, early onset and straightforward diagnostic criteria. Most importantly, these complex diseases occur much more frequently in the population. Whereas cystic fibrosis, for example, has an estimated population incidence of 1 in 3000, the combined incidence of all forms of OI is about 1 in 10,000. Common diseases such as diabetes, hypertension, asthma, bipolar depression, and osteoporosis occur in 5 – 50% of the elderly population. In view of the increase in the maximum life expectancy of men and women in our society, these common diseases will increase in frequency even further and the search for the responsible genes has become a priority in medical research. Unlike the relatively straightforward genetics of the monogenic disorders, common diseases have a multifactorial nature (genetic and environmental conditions interact), are multigenic (multiple genes are involved), and usually have a late onset with variable clinical manifestations. It is therefore not surprising that these diseases are referred to as “complex traits.” However, due to the successful application of molecular genetic techniques to monogenic diseases, unraveling the etiology of complex traits by genetic means now has become a feasible mission [2]. In the field of bone metabolism, the main target disease is, of course, osteoporosis. The genetic dissection of complex traits follows similar analytical strategies for many of the common diseases, including osteoporosis. First, evidence from studies of twins and families is sought to demonstrate and estimate the heritability of the trait (or one or more of its composite features) and the influence of environmental

Osteoporosis is defined by decreased bone mineral density (BMD) and degenerative microarchitectural changes of bone tissue, and consequently an increased fracture risk. Naturally, in the absence of molecular insights into the cause of the disease, definitions of it remain vague and descriptive. The main emphasis in this definition is on aspects of bone, whereas the clinically relevant end point in osteoporosis is fracture. However, fracture risk is determined only in part by bone characteristics with other anthropometric and physiological parameters also contributing to fracture risk, such as cognitive function, body size, and muscle strength. Thus, the genetic analysis of osteoporosis will include the genetics of bone characteristics, such as BMD, but eventually must also address the genetics of cognition, muscle strength and so on. Of particular interest is BMD, which can be considered a quantitative trait. That is, in a population, BMD can assume a continuous variety of values that follow a normal distribution. Although “low” BMD can be defined as well as “high” BMD, such distinctions introduce considerable subjectivity as to what is a sensible threshold to distinguish cases from controls. This is of course in contrast with dichotomous traits such as fracture. Certain aspects of osteoporosis have strong genetic influences. This can be derived, for example, from genetic epidemiological analyses that showed that, in women, a maternal family history of fracture is related positively to recurrence risk [3]. Most evidence, however, has come from twin studies on BMD [4 – 8]. Monozygous (MZ) and dizygous (DZ) twins share 100 and 50% of their genome, respectively. If a trait is strongly influenced by genetic factors, one expects the variance between the two members of an MZ twin pair to be smaller than between members of a DZ twin pair. The difference in variance between the two twin types can be expressed as a heritability score between 0 and 100%. For BMD the heritability has been estimated to be high: from 50% up to 80%. This means that up to 80% of the variance in BMD values can be explained by genetic factors whereas the remaining 20% could reflect environmental factors. Twin studies, however, do not necessarily give

641

CHAPTER 26 Genetics and Genomics of Osteoporosis

an accurate estimate of heritability measures due to overestimation of the phenotypic similarity in MZ twins. Nevertheless, although the precise value of this figure is subject to discussion, it can be concluded that BMD has a strong genetic basis. In molecular terms the existence of considerable heritability for BMD as a phenotype means that there are “bone density” genes, variants of which will result in BMD levels that differ between individuals. Because they influence quantitative variation, these genes are referred to as quantitative trait loci (QTLs). From what we know about how BMD changes over time, these differences can become apparent in different ways, e.g., as peak BMD or as differences in the rates of bone loss at advanced age. In addition, the expression of the genetic influences on BMD can differ e.g., during periods of high bone turnover at puberty and menopause. Interestingly, the increased fracture risk associated with a positive family history of fracture persists after adjustment for BMD [3]. This indicated that there is also a genetic susceptibility to fracture that is mediated by additional factors other than only those predisposing to low BMD. One example includes hip axis length (HAL), a measure of femoral geometry. Twin studies suggest that 80% of the variation in HAL can be explained by genetic factors independent of BMD [7,8], whereas the same was suggested for ultrasound measurements of bone [7]. Thus, bone density and bone architecture will probably be influenced by shared but also by separate genetic factors. Composing a portfolio of genetic risk factors for “osteoporosis” will therefore necessitate determining on which subphenotype, i.e., which particular characteristic of osteoporosis, the factors of interest have the strongest influence. The heritability estimates of osteoporosis leave room for a considerable influence of environmental factors that can modify the effect of genetic predisposition. Some gene – environment interactions include diet, exercise, and exposure to sunlight. Environmental factors tend to change during the different periods of life, which can result in different “expression levels” of genetic susceptibility. Aging is associated with a general functional decline resulting in, for example, less exercise, less time spent outdoors, changes in diet, and so on. This can result in particular genetic susceptibilites being revealed only later in life after a period when they went unnoticed due to sufficient exposure to an environmental factor. Thus, genetic susceptibility can become more or less apparent in situations of stress. Taking all this into account it becomes evident that, not very surprisingly, osteoporosis is considered a “complex” genetic trait. This complex character is shared with other common and often age-related traits with genetic influences such as diabetes, schizophrenia, osteoarthritis, and cancer. “Complex” means that a trait is multigenic as well as multifactorial. Thus, genetic risk factors (i.e., certain alleles or

gene variants) will be transmitted from one generation to the next but their phenotypic expression will depend on the interaction with other gene variants and with environmental factors. A first step in the molecular dissection of genetic factors in osteoporosis is the “genomics” of osteoporosis. This involves determining chromosomal location (mapping), identification, and characterization of the set of genes, variants of which contribute to the genetic susceptibility for the different aspects (or subphenotypes) of osteoporosis. Finding the responsible gene for monogenic disorders has now become almost a routine exercise for specialized laboratories. However, the complex character of osteoporosis makes it quite resistant to the methods of analysis, which in the past decades have worked so well for monogenic diseases. Therefore, different and often more cumbersome approaches have to be applied (see, e.g., [2]). First, we will discuss different approaches and some more technical issues, followed by a review of some data obtained so far in the search for osteoporosis genes. Table 1 lists explanations of some terms used frequently in the following chapter when discussing genetic analyses.

II. TOP-DOWN AND BOTTOM-UP GENOMIC APPROACHES All of the analytical approaches to find “osteoporosis genes” are based on the observation that the genomic (and mitochondrial) DNA sequence between two individuals is not the same but will differ at certain positions. Finding DNA sequence variations between two individuals is not very difficult. This was demonstrated for the human lipoprotein lipase (LPL) gene [9] and the angiotensin converting enzyme (ACE) gene [10] by extensive sequence analysis of large stretches of DNA. In addition, coding and regulatory regions of large numbers of genes have been analyzed to find single nucleotide polymorphisms (SNPs) [11 – 14]. From these approaches it was estimated that there is, on average, one variant base pair every 500 bp (see Section III, B). The clue in genetic analysis of complex traits is therefore to find the ones that matter. To find these variations there are top-down and bottom-up approaches. In top-down approaches, large-scale genome searches are performed first, which indicate which chromosomal areas might contain osteoporosis genes. In an optimal setting, such searches are performed in hundreds of relatives (sibs, pedigrees, etc) with hundreds of DNA markers (mostly microsatellites) spread evenly over the genome. Genome searches are based on the assumption that relatives who share a certain phenotype will also share one or more chromosomal areas identical by descent (IBD; see later) containing one or more gene variants causing (to a certain extent) the phenotype of interest (e.g., low BMD). The

642

UITTERLINDEN, VAN LEEUWEN, AND POLS

TABLE 1 Brief Glossary of Genetic Terms Allele

One of several alternative forms of a DNA sequence at a specific chromosomal location (locus). Two alleles are present at each autosomal chromosomal locus in a cell: one inherited from the mother and the other from the father

DNA marker

A polymorphic DNA segment at a known chromosomal location.

Genetic map

The most likely order of DNA segments on the chromosome based on analysis of cosegregation of DNA markers in pedigrees

Genome search

The analysis of several hundred DNA markers (usually microsatellites), which are more or less evenly spread over all of the chromosomes, in collections of related individuals to look for linkage with a phenotype

Genotype

The combination of two alleles at a locus in an individual

Haplotype

A series of alleles found at linked loci on a single chromosome (phase)

IBD

Identity by descent. The situation where alleles in two or more individuals are identical because of common ancestry

IBS

Identity by state. The situation where alleles in two or more individuals are identical due to coincidence or to common ancestry

kbp

kilobase pairs (1  103 bp)

Linkage

The tendency of DNA sequences to be inherited together as a consequence of their close proximity on a chromosome

Linkage disequilibrium

Nonrandom association of alleles at linked loci

Locus

A unique chromosomal location defining the position of a particular DNA sequence

LOD score

Logarithm of the odds; measure of statistical likelihood that a genetic marker is associated through physical linkage with a gene causing or contributing to a particular phenotype

Mbp

Megabase pairs (1  106 bp)

Microsatellite

A locus consisting of tandemly repetitive sequence units the size of which is (arbitrarily) defined as 1 – 5 bp

Minisatellite

A locus consisting of tandemly repetitive sequence units the size of which is (arbitrarily) defined as 6 bp or more

Mutation

An alteration in the DNA sequence

Physical map

The order of DNA segments on a chromosome as determined by molecular analysis of (large) DNA segments

Polymorphism

The existence of two or more alleles at a frequency of at least 1% in the population

QTL

Quantitative trait locus; a gene that influences quantitative variation in a trait

RFLP

Restriction fragment length polymorphism

SNP

Single nucleotide polymorphism

Synteny

The location of loci on the same individual chromosome

UTR

Untranslated region

VNTR

Variable number of tandem repeats; a polymorphic micro- or minisatellite

gene is then said to be linked with the DNA marker used to “flag” a certain chromosomal region. Upon positive linkage, subsequent research will zoom in on identifying which one of the dozens of genes in the chromosomal area is the one involved in bone metabolism and then identify the particular sequence variant giving rise to aspects of a complex trait such as osteoporosis. The different steps in such a genome zooming approach are depicted in Fig. 1 (see also color plate). In contrast, the bottom-up approach builds on the known involvement of a particular gene in aspects of osteoporosis, e.g., bone metabolism, as established by, for example, cell biological and/or animal experiments. In this candidate gene, sequence variants have to be identified that are associated with differences in function of the encoded protein. Such variants will then be tested in association or linkage analyses to evaluate its contribution to the phenotype of interest. Such bottomup approaches have been pursued to find SNPs in cod-

ing and regulatory regions of candidate genes for blood pressure homeostasis [11], cardiovascular disorders [12,13], and endocrinological and neuropsychiatric disorders [14]. Similar approaches are also pursued for osteoporosis candidate genes (see later). Naturally, top-down and bottom-up approaches will meet each other somewhere down the line leading to maps of candidate osteoporosis genes and maps of genome areas containing putative osteoporosis genes, which will completely or partially overlap. This is illustrated in Fig. 2 and 3 (see also color plates), but will be discussed in more detail in Section IV.

A. Functional vs Anonymous DNA Polymorphisms Sequence analysis of a “candidate” osteoporosis gene in a number of different individuals will identify base pairs at certain positions that vary between individuals. Such

643

CHAPTER 26 Genetics and Genomics of Osteoporosis

FIGURE 1

Genome zooming involves the stepwise identification of sequence variation associated with aspects of a complex disease as depicted in this schematic flow diagram. (Left) The level of organization of the DNA molecules. (Right) The different analytical steps in the process. (See also color plate.)

sequence variants can be classified by definition as “mutations” when they occur at a frequency of less than 1% of the population and as “polymorphisms” when they occur at a frequency of at least 1%. Some of the sequence variants will be just polymorphic, whereas others will have functional consequences. These can include, for example, sequence variations leading to alterations in the amino acid composition of the protein, changes in the 5 promotor region leading to differences in expression, and/or polymorphisms in the 3 region leading to differences in mRNA degradation. The functional polymorphisms are of prime interest for further testing in association analyses to establish whether the candidate gene is a true osteoporosis gene. Because functional polymorphisms lead to meaningful biological differences in function of the encoded protein, this also makes the interpretation of association analyses using these variants quite straightforward. For functional polymorphisms it is expected that the same allele will be associated with the same phenotype in different populations. However, since it is clear that there is considerable work involved before such functional variations are identified as such, nonfunctional or anonymous polymorphisms are also of interest. In fact, because anonymous polymorphisms are so abundant (1 in 500), they are usually the first to be assayed in association analyses when a candidate gene has

been identified. The anonymous polymorphisms are then used as markers whereby the alleles “flag” different DNA sequence variants of (parts of) chromosomes between individuals rather than just the gene in or near which they are present. In the case of a positive association of one of the marker alleles with the phenotype of interest, one supposes then that the marker allele is in linkage disequilibrium (LD) with a truly functional polymorphism elsewhere in the gene. Usually it cannot be excluded that the LD extends into another nearby gene. In normal outbred populations this means that the other gene has to be in a region of approximately  1 Mbp near the marker. However, in more inbred populations (the Finns, inhabitants of islands such as Iceland), this area can be much larger. Modeling studies using population simulation to estimate the extent of LD surrounding common gene variants showed that “useful” association does not extend beyond 3 kb [15]. In genetic terms, this is extremely close and has considerable consequences, for example, for genome-wide genome searches using SNPs [16]. This estimate, however, is contradicted by other studies showing much larger distances over which LD can extend of up to several cM [17,18]. Thus, as was shown for the aldehyde dehydrogenase 2 gene [18], the pattern of LD in human populations is still difficult to predict. However, LD is an important population genetic phenomenon that

644

UITTERLINDEN, VAN LEEUWEN, AND POLS

FIGURE 2

A human karyotype indicating the position of chromosomal areas supposedly containing candidate bone mineral density genes. They have been identified in collections of sib pairs and/or large extended pedigrees or in mouse crosses by a total genome search (squares on the left of a chromosome) or by a partial genome search in the vicinity of candidate gene loci (circles on the right of a chromosome). Letters refer to references of the different studies: A [29], B [34], C [33], D [35], a [42], b [43], c [44], d [45]. (See also color plate.)

needs to be assessed empirically if the utility of genetic maps for population-based studies is to be put into perspective. To complicate LD matters further, dozens of genes can be present in an area of 500 kb, whereas functionally related genes also have a tendency to be localized near each other. Examples are the interleukin (IL)-1 cluster on chromosome 2q13, homeobox (HOX) genes (e.g., the HOX region 4 on chromosome 2q31-q32 containing six HOX genes within 70 kb of genomic DNA), and major histocompatibility complex (MHC) genes on chromosome 6p21. Also, in relation to bone and cartilage metabolism, some gene clusters have been found. These include, for example, a locus on chromosome 8q23 harboring BMP-1, the gene for hereditary multiple exocystoses (EXT) and osteoprotegerin (OPG). Also, on chromosome 12q13, the vitamin D receptor (VDR), COL2A1, and the 1-hydroxylase gene are located in the same vicinity.

Thus, a clear disadvantage of using anonymous polymorphisms is that interpretation is less straightforward (see Section III,A and Table 2). Yet, understandably, most association analyses so far have been done with anonymous polymorphisms (e.g., some of the steroid receptor polymorphisms), whereas few have been done with supposedly functional polymorphisms such as the collagen type 11 Sp1 polymorphism.

B. IBD vs IBS: Pedigrees, Sibs, and Populations DNA sequence variants shared between individuals can be identical by descent (IBD) and identical by state (IBS). IBD is due to the transmission of a certain allele through generations of relatives in pedigrees. Such an allele can be derived from a functional polymorphism or an anonymous polymorphism. In the latter case the marker allele is said to

645

CHAPTER 26 Genetics and Genomics of Osteoporosis

FIGURE 3

A human karyotype indicating the position of chromosomal areas containing candidate bone mineral density/osteoporosis genes as identified in association analyses. In addition, the position is indicated of selected chromosomal areas containing genes in which mutations give rise to skeletal dysplasias and bone disorders. Circles on the right indicate genes that have already been cloned, whereas triangles on the left indicate areas in which linkage has been demonstrated for some bone disorders and for which gene cloning efforts are currently underway. 1, Abers – Schönberg disease/osteopetrosis; 2, absorptive hypercalciuria with bone loss; 3, high bone mass (HBM) locus/osteoporosis pseudoglioma; 4, van Buchem’s disease/sclerosteosis; and 5, Paget’s disease/familial expansile osteolysis. (See also color plate.)

be “linked” to the functional allele when cosegregation of a certain phenotype and an allele of interest in such a pedigree structure is observed. Positive linkage (expressed as so-called LOD scores) gives convincing genetic evidence for the involvement of the locus being studied in the phenotype of interest. Linkage analysis in extended multigenerational pedigrees has been a tale of successes for the discovery of the causative gene in many so-called monogenetic diseases such as cystic fibrosis, Hungtington’s disease, neurofibromatosis, and Duchenne muscular dystrophy. The classical linkage approach in pedigrees, however, is less suitable to identify all the osteoporosis genes because (1) due to the multigenic nature of osteoporosis it rarely segregates as a monogenetic Mendelian disease (but with exceptions; see later) and (2) multigenerational pedigrees

are extremely difficult to collect for a late-onset disease such as osteoporosis. The IBD approach to identify osteoporosis genes is therefore mostly limited to sibling analysis including twins. Furthermore, it is most often limited to a subphenotype of osteoporosis, i.e., a risk factor such as low BMD, which can be measured at a younger age than the clinically relevant feature of osteoporosis: fractures. Sibs who share such a subphenotype (e.g., low BMD) are then also expected to share alleles at a causative locus more often than expected by chance. Because no knowledge is required of which genes are involved, the total human set of chromosomes (the human genome) has to be “searched” or “scanned” for linkage. This is done by analyzing very polymorphic (and, thus, informative with respect to segregation in almost every set of relatives analyzed) DNA markers (usually microsatellites) at regularly

646

UITTERLINDEN, VAN LEEUWEN, AND POLS

TABLE 2

Pitfalls in Genetic Association Studies

Epidemiological 1. Sample size is too small, leading to chance findings 2. Population is biased due to selection, admixture, inbreeding, etc. 3. Environmental factors differ among populations Genetic 1. Allelic heterogeneity: different alleles are associated in different populations 2. Locus heterogeneity: gene effects differ among populations due to genetic drift and founder effect 3. Linkage disequilibrium: one or more adjacent polymorphisms are the true susceptibility loci instead of the polymorphism being tested Molecular genetic 1. Low genetic resolution: unjustified grouping of alleles due to insufficient methodological discriminatory power 2. Anonymous polymorphisms: there is no known functional effect of the polymorphism to provide a direct biological explanation of the association

spaced positions along all chromosomes in large sets of sib pairs. Genome searches, using hundreds of DNA markers analyzed in collections of hundreds of sib pairs selected for a particular subphenotype, can therefore find chromosomal areas supposedly containing candidate genes. Although very useful, genome searches have the drawbacks that they can only identify relatively strong effects, are mostly limited to a few end points, cannot easily accommodate gene – gene and gene – environment effects, and, last but not least, result in a chromosomal area being identified rather than a particular gene. As the areas are usually quite large, it must be established whether any of the genes present in the identified chromosomal area have any bearing on the osteoporosis phenotype in the general population. This is accomplished by genetic epidemiological analyses of populations through IBS types of association analyses of candidate gene polymorphisms. Such studies will determine the contribution of one or more gene variants of the variance of the phenotype of interest and establish gene – gene and/or gene – environment interactions. In general, IBD analysis will identify gene variants with relatively strong effects, whereas IBS approaches are more likely to find relatively weaker effects and interactions. So far, candidate gene analyses have been popular as a “poor-man’s” alternative to genome searches for finding osteoporosis genes. However, because the conclusion of genome searches will be the analyses of a number of candidate genes in the areas identifed by genetic association studies, this direct approach of identifying “the SNPs that matter” is now growing in importance. This is of course further fueled by the drawbacks of genome searches (cost and sensitivity) and the plethora of genes being identified in the Human Genome Project. Quite soon at least the sequence of all human genes will be known and the comprehensive searches for polymorphisms in them will generate the ammunition for association studies on the one

hand and functional studies on the other. Although the approach of genetic association studies of candidate gene polymorphisms looks rather straightforward, the practice is not without controversies as is illustrated in Section III.

III. CONSIDERATIONS AND COMPLICATING FACTORS A. Epidemiological Pitfalls Table 2 lists several pitfalls in the analytical processes that have played (and are still playing) a role in the association analyses of candidate osteoporosis genes. Apart from these considerations, somewhat seemingly more trivial factors can also play a role. For example, the effect size, i.e., the actual difference in a certain measured end point (e.g., BMD or number of fractures) among genotypes, should not be confused with reliability of the conclusions (the confidence intervals around the point estimate) or their significance (the p value). Big effects (usually in small samples) that do not reach significance do not indicate that there is no relationship. It should impel the investigator to increase the sample size because the current number does not allow a straightforward conclusion. First line defenses against critique on this point usually include power calculations. However, power calculations are used frequently in cases where sample sizes are small (e.g., n  300) to demonstrate enough power ( 80%) to detect unrealistic large differences such as 1 SD in BMD or more in population analyses of BMD by genotype. Thus, association studies are best done with functional polymorphisms in large populations. Intuitively, it is clear that small differences require a large sample size to be able to detect them. It is therefore quite useless to reiterate associ-

647

CHAPTER 26 Genetics and Genomics of Osteoporosis

ation analyses (i.e., same polymorphism, same end point but in different populations) in samples of about the same size or even smaller than the original study population. In addition, there is a tendency in association studies following an original observation to analyze new, but still anonymous, polymorphisms. These replications will add to the confusion rather than solve it. It would be more informative to analyze functional polymorphisms and look at the relationship of this polymorphism with nearby SNPs such as occur in haplotypes of alleles of adjacent SNPs in a region of genomic DNA.

B. SNPs and Haplotypes From several studies it has become evident that, on average, 1 out of every 300 – 500 bp is variant in the population. This is highlighted, for example, by the comprehensive sequence analysis of 9.7 kb of genomic DNA encoding part of the human lipoprotein lipase gene in 71 different individuals [9]. In this relatively small stretch of genomic sequence, 88 different sites of sequence variation were found, with a mean heterozygosity of 20%. Of these, 79 were SNPs and 9 involved an insertion – deletion type of variation. Thus, candidate gene analyses will have to focus on which of the many variant nucleotides actually matter. That is, which sequence variation is functionally relevant by changing expression levels, changing codons in the expressed protein, or some other change. Given the average size of a gene and the relatively young age of human populations, it can be predicted that several sequence variations “that matter” will coexist in the same gene in a given sample from a study population. Indeed, several examples of osteoporosis candidate genes where this situation has been observed already exist, such as the interleukin 6 (IL-6), gene [19], the transforming growth factor- (TGF-) gene [20], and the vitamin D receptor (VDR) gene (see later). Therefore, a major task will be to distinguish the cis from the trans situation within a gene and find which other sequence variations will be in linkage with the one used in the analysis. This involves construction of haplotypes across the gene by resolving the relationship (or phase) among the different sequence variations and determination of which variant nucleotides frequently occur on the same chromosome as a result of LD (see Section II,A). This is especially relevant when association studies have used anonymous restriction fragment length polymorphisms (RFLPs) to detect association. If such an association is observed, the presumption is then that the associated allele is in close linkage with a truly functional variant elsewhere in the gene. However, sequential analysis of individual polymorphisms within a gene in an attempt to narrow the region in which a causative variant is thought to be present is fraught with problems of interpretation related to the number of sequence variations, their possible allelic combina-

tions (haplotypes), and their evolutionary relationship, which is shaped by recombination and mutation events. In the absence of knowledge of functionally, i.e., the exact nature of the quantitative variation, the complexity of the analytical problem should first be reduced. In this respect, haplotype-based analyses summarize the linkage disequilibrium relationship among several polymorphisms and provide a method of mapping causative alleles on a background of genetic variation. An example of such an approach are cladistic methodologies in which the evolutionary relationship among haplotypes is analyzed such as was done for the lipoprotein lipase (LPL) gene by Sing and colleagues [21,22]. These authors concluded that “nonrandom patterns of recombination and mutation suggest that randomly chosen SNPs may not be optimal for disequilibrium mapping of the LPL gene.” These observations have repercussions for the analysis of candidate genes in any complex trait, including osteoporosis. It means that analyzing one or few SNPs in a given candidate gene is not sufficient to implicate or exclude the gene as a susceptibility gene unless the SNP is truly functional, with well-documented consequences in molecular biological and cell biological terms. By combining this type of genetic analysis with truly functional studies it will become clear which of the polymorphisms are causative sequence variations (such as QTLs) as opposed to anonymous variants. On a somewhat larger scale the haplotypes of different adjacent genes are important to explain the amount of risk present for a given individual. This is especially important because many functionally related genes tend to be located in each others’s vicinity (see the examples in Section II,A).

C. Pleiotropic Effects When we consider risk factors for osteoporosis and fracture, factors other than characteristics of bone, such as BMD and bone architecture, must be taken into account. Such factors include fall frequency of individuals, cognitive ability and muscle strength among others. For comparisons of studies on “genetics of osteoporosis” it is therefore important to first define what the end point of the analysis is. Bone is, of course, a major target tissue in the genetic analysis of osteoporosis. However many, if not all, of the genes considered in the genetic analysis of bone density are expressed not exclusively in bone but also in several other tissues. For example, collagen type I1 is the most abundant bone matrix protein but is also present in vessel walls, the skin, and in other matrics. This phenomenon is referred to as pleiotropy, the involvement of a gene product in more than one metabolic pathway. Thus, genetic variations in pleiotropic genes will have influences on more than one end point and their effect could be missed and/or could influence the outcome if one is analyzing only a single end

648 point. Furthermore, proteins can be part of metabolic pathways that can be active at different levels and under different control at different parts of the life cycle. Finally, the hierarchical position of genes is of relevance, i.e., is there redundancy by other genes and are they upstream or downstream genes? Upstream genes tend to be master control genes (e.g., genes from the steroid receptor family encoding transcription factors) variations in which produce a cascade of effects in several pathways. Downstream genes will be expected to have a more limited repertoire of effects in view of their specialized nature. It is clear that a single sequence variation in a single gene often will not have a single effect. This makes the association analysis of sequence variations more troublesome but at the same time more realistic. Table 3 presents several examples of polymorphisms in known genes that have been considered candidate genes for one or more complex traits. Naturally, the discovery of pleiotropic effects is driven by the availability of the polymorphisms that have been described in one of the genes under study. However, they also reflect the inherent complexity of biological (disease) processes in that a single protein is involved in multiple metabolic pathways. One example is the methylene tetrahydrofolate reductase (MTHFR) gene, which has been implicated previously in atherosclerosis through homocysteine metabolism, but was also found to be associated with differences in BMD in a population of postmenopausal Japanese women [23]. The latter finding could reflect the involvement of the MTHFR gene in collagen cross-linking. Such a pleiotropic effect could then reflect the involvement of this enzyme in atherosclerosis as well as in bone metabolism. These biological processes share certain metabolic pathways encompassing matrix components and calcium deposition, for example. Thus, for genetic analyses such as linkage and association studies, the existence of pleiotropic effects could also be a reason to study the involvement of a particular gene variant that has been implicated in a particular disease process in another metabolic pathway. At a different level, pleiotropic effects can also be considered relevant for the aging process, given the fact that many complex traits show an onset at advanced age and have an increased prevalence with increasing age, that age is associated with a functional decline in many different pathways such as hormone metabolism, and that many complex diseases show interrelationships. Examples of such interrelationships in the field of bone and aging include the association of low BMD with stroke [24], the inverse relationship of plasma 25(OH)vitamin D with myocardial infarction [25], and the inverse relationship of osteoporosis and osteoarthritis [26]. In relation to the latter, genetic association studies have also indicated this to be the case when polymorphisms were studied of genes involved in bone metabolic pathways. Whereas VDR gene polymorphisms have been implicated in both osteo-

UITTERLINDEN, VAN LEEUWEN, AND POLS

porosis and osteoarthritis, and in particular osteophytosis [27], polymorphisms in the TGF- gene have also been associated with differences in BMD and with risk differences for osteophytosis [28]. These examples of pleitropic effects highlight the role that particular proteins play in bone metabolic pathways, which are involved in determining BMD as well as in the development of osteophytes in osteoarthritis.

IV. THE MAP OF OSTEOPOROSIS GENES A. Linkage Analysis Genome-wide searches for osteoporosis genes have already been inititated and are currently focusing on BMD genes. They are based on the analysis of concurrence of genetic similarity together with phenotypic similarity to identify chromosomal areas that are shared more often than expected by genetically related individuals who also share a phenotype. For example, two sisters have low BMD and share an allele for a DNA marker located at 11q12 among the 300 DNA markers tested and which are spread over all chromosomes. If such an observation is done in several such sib pairs at a frequency beyond what can be expected purely by chance, we conclude that this area contains a gene, variants of which determine differences in BMD. Thus, genome searches can involve the analysis of human sib pairs with low BMD but also the analysis of offspring of crosses of parental mouse strains that differ in BMD. Although the latter approach is powerful because there is extreme genetic flexibility to generate offspring, it also has considerable drawbacks. Obviously, mouse bone meatbolism is not identical to human bone metabolism. In addition, the genetic diversity of the inbred mice strains employed is also limited and does not necessarily reflect the genetic diversity of human populations. Results obtained so far in genome searches are not straightforward to interpret because the studies vary considerably according to the analytical strategies employed and, most importantly, with respect to the power they have to detect certain effects. Mouse studies have used different inbred strains of different genetic background and with different bone characteristics, whereas human studies are based on different inclusion criteria for probands and use subjects of different ethnic origin. In addition, they use sib pairs and/or extended multigenerational pedigrees and, as DNA markers, the studies use microsatellites spread through the genome, biallelic candidate gene SNPs (usually nonfunctional), and/or microsatellite markers, which are located in or nearby a selected group of candidate genes. Obviously, such diversity in approaches, methods, and types of subjects used in the analyses makes

649

CHAPTER 26 Genetics and Genomics of Osteoporosis

TABLE 3 Gene VDR

Pleiotropic Effects of Disease Gene Alleles

Polymorphisms a. BsmI-ApaI-Taql RFLPs

Risk alleles ?

Anonymous RFLPs in intron 8 and exon 9 (in LD with 3’UTR)

Phenotype/disease Calcium/bone metabolism Osteoporosis Pubertal growth, height Osteoarthritis, osteophytosis

b. Protein isoforms: M1/M4 (427 aa/423 aa)

T (M1/427 aa/“f”)

Calcium absorption

T to C at exon 2 eliminates first

Muscle strength

translation initiation codon;

Myocardial infarction

detected as FokI RFLP Immunomodulation c. G to A in Cdx2-binding site

A

in promotor area of exon 1a

Diabetes mellitus type 1 and 2 Crohn’s disease Infection susceptibility (TBC, leprosy) Sarcoidosis Multiple sclerosis Periodontal disease Cell proliferation, differentiation Prostate and breast cancer, melanoma Hyperparathyroidism Psoriasis

ER

Intron 2, Anonymous PvuII and

?

XbaI RFLPs

Osteoporosis Age of menopause, hysterectomy Breast cancer

ApoE

Protein isoform Cys112Arg;

4 (Arg/Arg)

Cys158Arg

Osteoporosis Ischemic, heart disease Alzheimers disease

IL-6

a. G-174C in 5’ promotor

C

b. 3’ AT-rich minisatellite VNTR

?

Osteoporosis Rheumatoid arthritis Atherosclerosis Alzheimer’s disease Systemic lupus erythematosus

TGF

a. Protein isoforms: Leu10Pro (T29C)

C (Pro)

b. C-509T in 5’ promotor

T

Osteoporosis Osteoarthritis, osteophytosis Myocardial Infarction

MTHFR

Protein isoforms: Ala222Val (C677T)

T (Val)

Osteoporosis Neural tube defect Coronary heart disease

Note: aa, amino acids.

it hard to establish and evaluate whether particular areas are identified more than once in relation to, for example, BMD. Although, initially, this will lead to an abundance of areas with (sub)significant linkage, it is expected that important areas will be identified more than once and, thus, indicate the position of important BMD genes. Figure 2 presents a summary of the chromosomal loci on the

human karyotype identified so far by the various linkage approaches. 1. HUMAN STUDIES a. Total Genome Searches The linkage approach using total genome searches in humans has identified several areas with more or less certainty that supposedly contain

650 candidate genes for low BMD. Devoto et al. [29] found suggestive evidence for loci predisposing to low hip or spine BMD on chromosome 1p, 2p, 4q, and 11q by using a genome search in a set of 149 members of seven large pedigrees (74 sib pairs) of French-Canadian, Greek, and Jewish extraction and in which low spinal BMD was segregating. In an additional collection of 64 sib pairs, the linkage to the 1p36.2-36.3 locus was extended and now reached a LOD score of 3.0 [30]. Interestingly, an independent association study (rather than a linkage analysis) of Scottish women using microsatellites across the 1p36 area also found this area implicated in determining BMD differences [31; see later]. Several interesting osteoporosis candidate bone genes are present in this area. Although Spotila et al. [32], reported a 3’UTR single nucleotide polymorphism in the tumor necrosis factor receptor type 2 gene to be associated with low BMD in unrelated individuals of these cohorts, this observation needs confirmation in larger studies. A study of the MTHFR gene, which is also located in this area, has found a well-known gene variant to be associated with low BMD [23; see Section III,C]. Although not an obvious osteoporosis candidate gene, the authors suggest that this protein is involved in cross-linking of collagen molecules and as such might contribute to variation in BMD. Niu and colleagues [33] reported a genome scan for genes predisposing for low forearm BMD, in 218 Chinese individuals from 96 nuclear families including 153 sib pairs. They reported areas on 2p and 13q to show evidence of linkage and noted that the 2p area was also reported by Devoto and colleagues. Preliminary results of a genome scan in 286 members of 10 large Mexican-American families for wrist, hip, and spine BMD identified low BMD loci on 7q31, 8q11, and 12q24 [34]. The largest genome screen so far was reported by Koller and colleagues [35]. It was done in two stages: first in a set of 429 Caucasian sister pairs with, subsequently, areas showing LOD scores higher than 1.8 being replicated in an expanded set of 464 Caucasian and 131 African-American sister pairs. They reported areas on 1q21, 5q33, 6p11, and 11q12 to show evidence of linkage. b. Locus Searches Another example of the linkage approach involves a partial genome search performed only in/near one or a limited number of “candidate gene” areas. Usually, microsatellites in or very near to the candidate gene of interest are analyzed in a collection of sib pairs similar to that used for genome searches. For example, Duncan and colleagues [36] analyzed 64 microsatellites in 23 candidate genes in a British Caucasian set of 165 sib pairs (115 families) and found suggestive linkage of low BMD with the parathyroid hormone receptor type 1 and moderate evidence for linkage of loci containing the epidermal growth factor gene, the collagen type I1 gene, the vitamin D receptor/collagen type II1 gene, the estrogen receptor 

UITTERLINDEN, VAN LEEUWEN, AND POLS

gene, the IL-1 gene, the IL-4 gene, and the IL-6 gene. In a similar analysis of 192 sib pairs (from 136 families) from Japan, Ota and colleagues [37] analyzed microsatellites in the IL-6, the IL-6 receptor, the calcium receptor, and the matrix Gla protein. Similar to Duncan and colleagues, moderate evidence was found for linkage of low BMD to the IL6 locus as indicated by increased allele sharing among sib pairs concordant for osteopenia (defined as a T score of 2.3). Finally, linkage approaches are also applied to evaluate single chromosomal loci of interest containing particular candidate genes. Examples of such approaches are the linkage analysis of the high bone mass (HBM) gene area on chromosome 11q12-q13 [38; and see Section IV,B,1]. Koller and colleagues [38] analyzed 374 sib ships of Caucasian and African-American extraction with seven microsatellite markers spanning 56 cM around the HBM region and found a maximum LOD score of 3.5 for D11S987 with femoral neck BMD. This same marker shows strong linkage to a number of bone disorders (see later), indicating that genes present in this area might play a role in explaining variation in BMD at the population level. Another example of this approach involves the microsatellite in the promotor region of the insulin-like growth factor type I gene [39]. Takacs and colleagues [39] analyzed 542 sibling pairs for linkage and association of BMD with this polymorphism but could not find evidence for involvement of this locus in explaining variation in BMD. As noted earlier, a special form of a locus search was introduced by Albagha and colleagues [31], who analyzed allele distribution of microsatellites in 54 women with high BMD and 54 women with low BMD. Using this association analysis (rather than linkage analysis), they found markers on 1p36 to be associated with differences in BMD, thereby presenting evidence of the involvement of this chromosomal area in determining BMD, independent of that shown by Devoto and colleagues. c. Major and Minor Effects So far, few of the chromosomal loci identified in the genome searches for low BMD have given convincing positive lod scores in more than one study. These can be considered to have major effects on BMD at several measured sites (e.g., femoral neck, lumbar spine.) and in different ethnic groups. On average, one can expect with a few hundred sib pairs to be able to detect genes with effects explaining roughly 20 – 30% of BMD. It is questionable whether these are realistic effects given the multigenic and multifactorial nature of the BMD trait. The linkage results must be regarded with caution, and larger studies are required to identify chromosomal loci harboring BMD genes. Thus, one might say that, if they exist, the major genes will be identified by this approach but minor genes will escape detection. These will have to be identified using association types of studies, which can accommodate

CHAPTER 26 Genetics and Genomics of Osteoporosis

gene – gene interactions and gene – environment interactions more easily. Association studies will be important anyhow because once linkage is observed, the area must be scrutinized for the presence of likely candidate genes, and polymorphism screening and association analysis will need to be performed to evaluate the candidate genes. 2. ANIMAL STUDIES Several animal models for osteoporosis have been described, including mouse, rat, rabbit, dog, sheep, and baboon. Of these, only the mouse is a useful model in genetic terms because of the highly developed genetic and physical maps of its genome and the availability of a large number of spontaneous and genetically engineered mutants. However, in view of its relatively close phenotypic (bone metabolism) and genetic (chromosomal synteny) similarity to humans, large efforts have been devoted to develop the osteoporosis baboon model, including the generation of maps of DNA markers used in linkage studies. A preliminary analysis of BMD QTLs in pedigreed baboons indicated the area homologous to human 11q12 and containing the HBM gene to also show linkage to BMD differences [40]. Total genome searches for BMD genes have been performed in mice making use of crosses of high and low BMD mice strains. Once one or more mouse BMD loci are identified by this approach, one can then try to find the paralogous human chromosomal BMD loci. This can be very difficult, however, because of (a) the capricious nature of the conservation of human – mouse synteny, i.e., the (dis)agreement in location on the same chromosome fragments between the two species [41], (b) the large size of the chromosomal regions identified, and (c) the substantial number of osteoporosis/BMD candidate genes present in the areas. Therefore, only a limited number of areas identified in mouse searches can be used directly to support findings in human searches awaiting the identification of the responsible mouse genes. The human homologue can then be scrutinized for its contribution to BMD variation at the population level. An overview of some of the chromosomal loci identified in mouse genome searches is presented in Table 4 and in Fig. 2. Klein and colleagues [42] performed a genome search on 24 recombinant inbred strains derived from C57BL/6 mice with a low BMD and DBA/2 mice with a high BMD and identified 10 QTLs linked to peak BMD in female mice. Shimizu and colleagues [43] analyzed F2 mice derived from crosses between SAMP6 mice, a mouse model of senile osteoporosis, with SAMP2 mice, which have a higher peak BMD, and identified 2 QTLs reaching significant linkage. Beamer and colleagues [44] analyzed progeny from a cross between C57BL/6J and CAST/EiJ mice and identified 4 QTLs for BMD. Benes and colleagues [45] also analyzed crosses of SAMP6 mice but used AKR1 as well as SAMR1 mice as high BMD strains and identified 6 QTL loci for BMD.

651 As shown in Table 4, several areas were identified in these genome searches of mouse BMD loci that overlap. The mouse chromosome 11 and 13 areas were identified by three out of four searches. The chromosome 11 area corresponds to the human 17q11-q22 area containing, for example, the COLIA1 gene and the vanBuchem/SCL gene. The chromosome 13 area corresponds to more than one human chromosomal area, including 1q42, 7p15, and 6p22. This also highlights one of the problems with the mouse genome search approaches that is eminent when mouse – human homology maps are compared [41]. In several cases, the evolutionary distance between human and mouse prevents a straightforward identification of paralogous human chromosomal areas. In addition, to what extent the mouse models chosen for the genetic crosses represent good models of characteristics of human osteoporosis remains to be established once the genes involved are identified. Mouse genome searches so far have focused on BMD. Nevertheless, in concert with the other approaches mentioned, this represents a productive way to contribute to the identification of BMD-regulating loci, variants of which could contribute to risk differences for osteoporosis in humans.

B. Association Analysis of Candidate Genes Any genetic analysis of osteoporosis will end up with a gene and one or more variants of it that can be used as risk predictors. Whereas the top-down approach, as discussed earlier, encompasses genome searches that will identify one or more chromosomal regions supposedly containing such candidate genes, the bottom-up approach involves the a priori choice of a particular, known gene to be searched for polymorphisms that might contribute to population variance of one or more risk factors for osteoporosis. The choice of such a candidate gene is guided by considerations that revolve around the involvement in bone biology. Several lines of evidence can be followed to establish such a role in bone biology of a particular gene product. For example, mutations in the gene of interest lead to a known Mendelian disease of bone or, vice versa, the gene responsible for a Mendelian bone disorder will be of interest to screen for polymorphisms and to be evaluated in association analysis. The involvement in bone biology can also be established when the gene of interest is knocked out in mice and a bone phenotype occurs, or simply when the gene product occurs specifically in bone tissue. Thus, an osteoporosis gene product will have characteristics all more or less in line with at least one of these considerations. In view of this rather wide definition, it can be expected that there are many potential osteoporosis candidate genes. Table 5 provides a list of such candidate genes, together with some characteristics, which at least to some extent can be considered as osteoporosis candidate genes. For each of

652

UITTERLINDEN, VAN LEEUWEN, AND POLS

TABLE 4

Genome Searches for BMD Loci in Micea Crosses

C57BL/6  C57BL/6J  SAMP6  SAMP6  DBA/2 CAST/EiJ SAMP2 AKR/J or SAMR1 Human chromosome position [42] [44] [43] [45] homology

Mouse cM chromosome 1

74 – 96



2

2 – 23



86





1q21-q42 

10p14-p15 20q11-q12



5

25

7

11 – 13

4p15-q32

44



11

45 – 60



13

10 – 22

14

2



15

43



16

9 – 28



18

48



18q21

19

53



10q23-q26

 11q13-q21 





17q22





1q42-43; 7p13-p15; 6p21-p25 3p14



11p13; 22q13 

3q27-28

a

A plus sign indicates significant linkage at the particular chromosomal position.

these, extensive information can be found at several web sites using the abbreviation of the gene’s name [46 – 48]. The genes are ranked by chromosomal position to allow for initial inspection of positional correspondence with chromosomal loci identified in linkage approaches by, e.g., genome searches in humans and mice. Examples will be discussed of the approaches and considerations that can lead to the choice of osteoporosis candidate genes to be studied further. In addition, a few particular candidate genes will be discussed in more detail because of the scrutiny they have undergone already. 1. LINKAGE ANALYSIS OF MONOGENIC BONE METABOLISM SYNDROMES Conventional linkage analysis in families in which a usually rare metabolic bone disorder or skeletal dysplasia is segregating as a Mendelian monogenic trait can lead to the discovery of genes playing a role in bone metabolism. Mutations in the coding sequence of such genes most often lead to the severe phenotypes characteristic of the syndrome. Whether polymorphic variants of such genes have possibly milder effects and are important for aspects of osteoporosis at the population level has to be determined by association analysis in large-scale epidemiological studies. Over 100 skeletal dysplasias have been described for which the responsible genes are known. A well-known example is osteogenesis imperfecta, which is caused by mutations in the bone matrix protein components collagen type I1 and

I2. Consequently, these genes have been searched for polymorphisms to be associated with osteoporosis, and not without success (see Section IV,B,4). In addition, genome searches are applied in “single gene” Mendelian bone disorders to discover the responsible genetic defect. The chromosomal location of some of these bone disorder genes is depicted in Fig. 3. Examples include human osteopetrosis, also known as Albers-Schönberg disease, which describes a group of hereditary disorders characterized by abnormal bone resorption. Three clinical forms exist, which have been mapped to different locations: autosomal recessive or infantile osteopetrosis (11q13), autosomal dominant osteopetrosis (1p21), and osteopetrosis with renal tubular acidosis, which is known to be caused by mutations in the carbonic anhydrase gene (8q22). A localization of interest in this respect includes the mapping on chromosome 11q12-q13 of a locus for high BMD containing the putative high bone mass (HBM) gene in an American Caucasian pedigree [49], for autosomal recessive osteopetrosis in two Bedouin pedigrees [50], and for the low BMD/osteoporosis pseudoglioma syndrome [51], all with high lod scores to the same DNA marker (D11S987). This locus has gained additional attention because the nonparametric QTL analysis of peak BMD in 374 sib ships suggested this locus to contain a gene influencing BMD variation in the general population [38; see earlier discussion]. Several candidate genes of interest are present in this area, including latent transforming growth factor 

TABLE 5 Gene name

Bone Genes by Chromosomal Location

Symbol

Cytogenetic location

Human diseasea 

Bone-specific alkaline phosphatase

ALPL

1p36.1

Tumor necrosis factor receptor 2

TNFR2

1p36.3

Methylenetetrahydrofolate reductase

MTHFR

1p36.3

Lysyl hydroxylase Collagen type 1X2 Colony-stimulating factor-1 Collagen type XI1

PLOD

1p36.3

COL9A2

1p33-p32.2

CSF-1

1p21-p13

COL11A1

1p21

Cathepsin K

CTSK

1q21

Thrombospondin III

THBS3

1q21-q24

Fibromodulin

FMOD

1q32

Osteocalcin

BGLAP

1q25-q31

Transforming growth factor 2

TGFB2

1q41

STK

2p23-p24

Serine threonine kinase

653

Pro-opiomelanocortin

POMC

2p23.3

Calmodulin 2

CALM2

2p21.3-p21.1

IL1A,  1B,  1RN

2q13

Collagen type III 1

COL3A1

2q24.3-q31

Collagen type V2

COL5A2

2q24.3-q31

Interleukin 1,  1,  1 receptor antagonist

Fibronectin 1

FN1

2q34

Thyroid hormone receptor 

THRB

3p24.3

Parathyroid hormone receptor 1

PTHR1

3p22-p21.1

Protein S

PROS1

3p11.1-q11.2

Calcium sensing receptor

CASR

3q13.3-q21

CLAPM1

3q28

2HS glycoprotein

AHSG

3q27-q29

Fibroblast growth factor receptor 3

FGFR3

4p16.3

GC

4q11-13

Albumin

ALB

Osteopontin

SPP1

Clathrin adaptor protein (AP50;AP2)

Vitamin D-binding protein





Association



20

10

8.5

3

50

50

1.5-4.5



7.5

 



 8

6

4q11-q13

17

15

4q21

8.2

7

1.6

15

7

2.0



Bone sialoprotein

IBSP

4q21

Dentin matrix acidic phosphoprotein

DMP1

4q21

Nuclear factor B, subunit 1

NFKB1

4q23-q24

EGF

4q25



Growth hormone receptor

GHR

5p13-p12



Arylsulfatase B

ARSB

5q11-q13

Epidermal growth factor

Mouse modelb



TABLE 5 Gene name

Symbol

Cytogenetic location

Versican

CSPG2

5q12-q14

Fibrillin 2

FBN2

5q23-q31

IL13, IL4, IL5, IL3, IL9

5q31.1

IL4

5q31.1

Interleukin 4 cytokine gene cluster Interleukin 4 (BSF-1) Interleukin 3 (M-CSF)

IL3

5q31.1

Sulfate transporter

DTDST

5q31-q34

Osteonectin

SPOCK

5q31-q33

Glucocorticoid receptor

NR3C1

5q31

Collagen type XI2 Human leukocyte antigen Core binding factor  subunit

COL11A2

6p21

HLA

6p21

CBFA1

6p21

COL10A1

6q21-22.3

Estrogen receptor

ESR1

6q25.1

Thrombospondin II

THBS2

6q27

Twist transcription factor

TWIST

7p21

IL6

7p21

Collagen type X1

654

Interleukin 6 Elastin

ELN

7q11.2

Calcitonin receptor

CALCR

7q21.3

Collagen type I2

COLIA2

7q21.1

3-adrenergic receptor

ADRB3

8p12-p11.2

(continued) Human diseasea



10/9/8

20

10

2.2/3.0

2

2

0.7

45

34

3.5













FGFR1

8p11



8q22



TNFRSF11B

8q24

Exostosin-1

EXT1

8q24.12



Tricho-rhino-phalangeal syndrome gene

TRPS1

8q24.12



ABL

9q34.1

Collagen type V1

COL5A1

9q34.2-q34.3

Collagen type V3

COL5A3

9q34.2-q34.3

Interleukin 1 receptor antagonist

IL1RA

10p15-p14

Nuclear factor B, subunit 2

NFKB2

10q24

Fibroblast growth factor receptor 2

FGFR2

10q26

PTH

11p15

Calcitonin

CALCA

11p15.2-p15.1

Exostosin 2

EXT2

11p12-p11

Hematopoetic transcription factor

PU.1

11p11.2

PU.1 (Spi1)

15



CA1,2,3

Parathyroid hormone

90



Carbonic anhydrase (1-3)

cAbl kinase

Association



Fibroblast growth factor receptor 1 Osteoprotegerin (OPG)

Mouse modelb





 







Vacuolar proton ATPase

V-ATPase



11q13.4-q13.5

(OC-116:TCIRG1:TIRC7:Atp6i) Interleukin 18

IL18

11q22.2

Matrix Gla protein

MGP

12p13.1-p12.3

TNFR1

12p13

VDR

12q13

Tumor necrosis factor  receptor 1 Vitamin D receptor

 



Collagen type II1

COL2A1

12q13

1-Hydroxylase

CYP27B1

12q13

Decorin

DCN

12q21-q23

Insulin-like growth factor I

IGF-I

12q22

 

Klotho-homologue, -glucosidase-like

KI

13q12

TNFSF11

13q14

Collagen type IV1

COL4A1

13q34

Collagen type IV2

COL4A2

13q34

Estrogen receptor

ESR2

14q23-q24.1

Bone morphogenic protein 4

BMP4

14q22-q23

Transforming growth factor 3

TGFB3

14q24

Osteoprotegerin ligand (OPGL, RANKL,





ODF, TRANCE)

655

c-fos oncogene

FOS

14q24.3

THBS1

15q15

Fibrillin 1

FBN1

15q21.1

Aromatase

CYP19

15q21.1

Thrombospondin I

Vitronectin



 7

5

110

65

10 1.7





VTN

17q11

4.5

8

THRA

17q11.2

27

10

Colony-stimulating factor 3 (GCSF)

CSF3

17q11.2-q12

Integrin 3

ITGB3

17q21

Thyroid hormone receptor 

Homeobox B cluster (9 genes A-I)

 

HOXB

17q21-q22

COLIA1

17q21.3-q22





Noggin

NOG

17q22





Growth hormone

GH1

17q22-q24



Collagen type I1

SRY-box 9 Receptor activator of nuclear factor

SOX9

17q24-q25

TNFRSF11A

18q21.2-q21.3

COMP

19p13



 

B (RANK) Cartilage oligomeric matrix protein Insulin receptor

INSR

19p13.2

TGFB1

19q13.1-13.3



Apolipoprotein E

ApoE

19q13



Growth differentiation factor 5

GDF5

20q11.2

Transforming growth factor 1



TABLE 5 Gene name Bone morphogenic protein 2 Oncogene src

Symbol

Cytogenetic location

BMP2

20q11-q12

SRC

20q12-q13

GNAS1

20q13

CBS

21q22.3

Arylsulfatase E

ARSE

Xp22.3

X-linked hypophosphatemia protein

PHEX

Xp22.2-p22.1

Guanine nucleotide protein,  subunit Cystathionine -synthetase

656

Androgen receptor

(continued) Human diseasea

Mouse modelb

Association

 

20

13



AR

Xq11

Cu2 transporting ATPase,  polypeptide

ATP7A

Xq12-Q13

150

23

Glypican-3

GPC3

Xq26

500

8

Biglycan

BGN

Xq28

a

2.1



Mutations in the candidate gene give rise to a known human (Mendelian) disease with a “bone” phenotype. A mouse model(s) exists that shows effects on bone, which are based on spontaneously occurring mutations of this gene and/or are based on transgenes for this gene and/or are based on knockout for this gene. c Association of polymorphic variants of this gene has been demonstrated with aspects of osteoporosis such as decreased BMD or increased fracture risk. b

CHAPTER 26 Genetics and Genomics of Osteoporosis

binding protein 2, fibronectin-like 2, and OC116, a vacuolar proton pump. The latter protein is of particular interest because this protein was identifed to be an osteoclast-specific proton pump, a mutation which is responsible for the oc/oc mouse model of osteopetrosis (see later). The final proof of identification of the HBM gene through further physical mapping and mutation screening is therefore eagerly awaited, whereas the subsequent polymorphism screening will be of interest for association analyses with variation in BMD and/or other aspects of osteoporosis. Further examples of the mapping of still unknown interesting bone disorder genes are the localization of the van Buchem’s disease/sclerosteosis gene, which is associated with systemic increased bone formation, to the 17q12-q21 area [52], the Albers – Schönberg autosomal dominant osteopetrosis gene to 1p21 [53], the gene for absorptive hypercalciuria with bone loss, which is associated with decreased BMD, to 1q24 [54], and the mapping of one of the Paget’s disease genes and the familial expansile osteolysis gene to 18q21-q22 [55]. The responsible gene in this latter area was shown to be the TNFRSF11A or RANK gene by demonstrating mutations in patients of four families, in the signal peptide of this protein, that is essential in osteoclast formation [56]. All of the responsible genes for these disorders play a role in bone metabolism and thus are of interest to be searched for polymorphisms and analyzed for association with aspects of osteoporosis. 2. MOUSE MODELS Another prolific source of osteoporosis candidate genes involves animal models in which one or more gene mutations are present, giving rise to bone phenotypes. A number of animal models, usually mouse models, have been described that mimic certain aspects of osteoporosis, but mainly osteopetrosis [57]. The models can be induced by operation (ovariectomy) or result from spontaneously arisen mutant strains or are based on genetically engineered strains such as transgenes or knockout models. Characterization of the underlying genetic defects will ultimately result in candidate genes, the human homologue of which can be analyzed in linkage and/or association studies to evaluate the contribution to differences in BMD and/or risk for osteoporosis. Indeed, several examples of such convergence of research approaches have been described. Several spontaneous mutations have occurred in mouse strains resulting in models of osteopetrosis, such as the op/op mouse, which is due to a mutation in the M-CSF1 gene [58], and the osteosclerotic mouse oc/oc. The latter model is of particular interest because the underlying mutation was shown to be a 1.6-kb deletion in the promotor region of the osteoclastspecific vacuolar proton pump ATPase subunit [59]. This genotype – phenotype relation is further supported by knocking out of this gene in -/- deficient Atp6i-mice, which also show an osteopetrosis phenotype [60]. Interestingly, this gene

657 has a human homologue (the OC-116 gene) that is located at chromosome 11q13, corresponding to the position where the HBM gene, the OP pseudoglioma gene, and the osteopetrosis gene were found to be located and where linkage to BMD was found in genome searches (see earlier discussion). This therefore seems to be a very promising locus as the reasons to implicate it in osteoporosis are coming from several different lines of evidence. Further osteopetrotic or osteosclerotic mouse models that were developed include knockout (KO) models for the c-src protooncogene [61], the c-fos gene [62], the NF-B1 and NF-B2 genes [63], the 3 integrin gene [64], and the cathepsin K gene [65], whereas a KO mouse model was also described for the c-Abl gene that leads to an osteoporotic phenotype [66]. Mouse models that mimic osteoporosis are still rather scarce. One frequently used approach is to induce osteoporosis by ovariectomy, but this has so far not been very helpful in identifying osteoporosis candidate genes. A set of spontaneous mutant mouse strains that develop osteoporosis are the so-called senescence-accelerated mouse (SAM) strains. The SAMP6 strain especially exhibits a lower BMD, which is thought to be due to a number of genetic variations. It is therefore used in crosses with high BMD strains to identify BMD genes in genome searches (see Section IV,A,2). Another example of a spontaneous osteoporosis mouse is the autosomal recessive Unhip (Unh) mouse, homozygotes of which develop bone mineralization defects leading to fractures. A genome scan has identified mouse chromosome 14;2 (corresponding to human chromosome 3p14) to harbor the mutated gene [67]. One of the first genetically engineered mouse models of osteoporosis was based on a transgene with increased expression of interleukin 4 [68]. Intriguingly, analysis of human sib pairs showed linkage of the human IL-4 gene to differences in serum IgE production [69], whereas it is known that osteoporosis is a common complication in patients with the hyper IgE syndrome. Further evidence to implicate this gene in osteoporosis comes from genome searches for BMD genes in which linkage to 5q31 was reported, the chromosomal area where the IL-4 cluster is located [35]. Another early example of a genetically engineered osteoporosis mouse is the biglycan-deficient KO mouse [70]. Mice deficient for this extracellular matrix proteoglycan are normal at birth but develop low bone mass, which becomes more obvious with age. By insertional mutagenesis of a novel mouse gene, called “klotho” (kl), a mouse model for aging was generated, including the development of osteopenia [71]. Although the accelerated aging phenotype is similar to SAM mouse models, the underlying defects are different. The klotho mouse mutation is a single gene variation while the gene shares sequence similarity with -glucosidase enzymes and has a human homologue on chromosome 13q12 [72]. Mouse models of osteoprotegerin have been generated, thereby providing strong evidence to implicate this gene in

658

UITTERLINDEN, VAN LEEUWEN, AND POLS

the regulation of bone mass. Whereas OPG -/- mice develop osteoporosis and increased incidence of fractures [73], transgenic mice overexpressing OPG develop osteopetrosis [74]. Mouse studies like those cited here are valuable because they can give molecular insight to the contribution of one or more genes to certain pathways in bone biology and to determining BMD and/or to risk differences for fracture. Together with the existence of a human disease in which the genes of interest are mutated and their presence in a chromosomal region showing linkage, the existence of osteoporotic or osteopetrotic mouse models makes the genes involved very likely to be prime candidate human osteoporosis genes. Although they can be supposed to be implicated in determining BMD variation, very few of these have actually undergone the scrutiny of association analyses in large population studies. That is, particular polymorphisms that have a functional consequence will have to be found in these genes and large-scale association analyses will have to be performed in several populations to evaluate their contribution to explaining osteoporosis risk at the population level. Only a few genes have undergone such scrutiny already, including the vitamin D receptor gene and the collagen type I1 gene. 3. VITAMIN D RECEPTOR GENE a. Association Studies Using BsmI, ApaI, and TaqI RFLPs The candidate gene that actually initiated the “molecular genetics of osteoporosis” is the vitamin D re-

FIGURE 4

ceptor gene. Three adjacent RFLPs for BsmI, ApaI, and TaqI, respectively, in intron 8/exon 9 at the 3’ end of the gene are most frequently studied (Fig. 4, see also color plate). Morrison et al. reported that the BsmI RFLP in the last intron of the VDR gene was related to serum osteocalcin concentration [75] and subsequently to BMD in a twin study and in postmenopausal women [76]. Although the initial observations on the twin study have been partially withdrawn [77], in the following years dozens of papers were published analyzing the same RFLP in relation to BMD. Some of these confirmed the observation, whereas others could not find an association or found another allele associated. In the largest study published so far and which analyzed 1782 Dutch elderly men and women, no effect of single RFLPs on BMD was observed, whereas a small effect was detected employing haplotypes constructed of the three adjacent 3’ RFLPs [78]. In line with this, a metaanalysis of 29 studies (excluding the Dutch cohort) on the relationship of VDR genotype with BMD [79] concluded that the VDR genotype is associated with BMD in elderly subjects but with only a 1 – 2% difference between extreme genotypes. In addition, Gong and colleagues analyzed 75 articles and abstracts on VDR genotype and BMD [80] and concluded that BMD is associated with the VDR genotype, especially in females before menopause, and that genetic heterogeneity and nongenetic factors play a role in finding the associations. This notion is supported by other studies that found evidence to suggest an influence on peak BMD

Position of polymorphisms in the vitamin D receptor gene. (See also color plate.)

CHAPTER 26 Genetics and Genomics of Osteoporosis

in younger subjects (age 7 – 29 years) [81,82], found a relation of VDR genotype with bone loss [83,84], and observed an influence of dietary calcium intake on the strength of the associations with bone density [85,86]. Interestingly, a study of American Caucasian women [87] and a preliminary report in Dutch Caucasian women [88] suggested that the VDR genotype is associated with increased fracture risk. This effect was mostly independent of (the small) genotype-related differences in BMD [88]. However, because other groups have not seen a relationship between fracture risk and VDR genotype [89,90] this relationship remains uncertain and needs further scrutiny. Of particular interest in this respect is the fact that the same risk allele is not always found associated with changes in bone parameters (but also with other end points; see later), preventing the straightforward interpretation of these associations. While the initial studies by Morrison et al. [75 – 77] suggested the “B” allele of the BsmI RFLP site to be the risk allele associated with low BMD, other studies could either confirm this, or did not find any effect, or found the (opposite) “b” allele to be the risk allele associated with low BMD. Such conflicting findings, which are not exclusive for the field of genetic association analysis of osteoporosis, could have any one or more of several reasons, which are shown in Table 2 and discussed in Section III,A. The most likely explanations for the conflicting results for the VDR gene are that, given the small effect on BMD, very often the statistical power is much too low and no conclusions on the presence or absence of a statistically significant effect can in fact be drawn. More important, however, is the fact that the 3’ Bsm-Apa-Taq RFLPs are not functional and that, consequently, differences in linkage disequilibrium (LD) between them and the truly functional allele can lead to altered associations. Furthermore, other functional polymorphisms will be present in the VDR gene, such as the FokI and the Cdx2 polymorphisms (see later), so that in fact unknown combinations of alleles across the gene (haplotypes) need to be tested. Finally, interactions between different genes and/or environmental factors play a role in the action of this important steroid hormone receptor transcription factor, and such interactions can of course differ between different populations. In addition, pleiotropic effects of this gene can play a role in influencing an association, but these will be discussed later. b. Other VDR Polymorphisms Alleles of the BsmI, ApaI, and TaqI polymorphisms in intron 8 and exon 9 are closely linked and haplotypes can be constructed over this 2.2-kb region [75,78]. The LD of these RFLPs extends into the 3’-untranslated region (UTR), which is a 3.2-kb sequence immediately adjacent to exon 9 [75,91,92] (see Fig. 4). More than 10 different sequence variations in the 3’UTR have been described, including a poly(A) repeat polymorphism. Analysis of the LD over this 5.5-kb region

659 at the 3’UTR of the VDR gene in different ethnic population groups indicated that the LD differed among populations [91]. A single RFLP, such as the BsmI RFLP, which is the most frequently used in association studies of the VDR gene, is therefore not a good marker for the LD with other sequence variations and, thus, the use of the BsmI RFLP might contribute to heterogeneity among association studies. As discussed before, this notion is strongly supported by the comprehensive sequence analysis of the LPL gene [21,22] Analysis of the genomic organization has shown that the VDR gene is quite large (approximately 80 kb) and has an extensive ( 40 kb) promotor region [94] capable of generating multiple tissue-specific transcripts [94]. In view of the genome-wide observed frequency of SNPs (see Section III,B), one can expect over 100 polymorphisms to be present in the VDR gene area alone, including areas that are functionally relevant, such as the promotor region. Indeed, more than 25 different polymorphisms are currently known to be present at the VDR locus (see Fig. 4 and Y. Fang et al., unpublished data), so far mostly near the 3’ end of the gene. However, toward the 5’ end of the gene in and near the promotor region other sequence variations have been reported as well. For example, a substitution (T to C) at exon 2 eliminates the first ATG translation initiation site and allows a second one 9 bp downstream to be used. Thus, two variant forms of the VDR protein can be translated that differ by three amino acids resulting in proteins of 427 (M1) and 424 (M4) amino acids. The existence of these two different forms of the VDR protein has been demonstrated, whereas the shorter form was found to give greater transcriptional activation [95,96]. The sequence change can be detected as a FokI RFLP (97), and the “f” allele (corresponding to M1, the longer protein) has been found associated with low BMD in several study populations [95,97,98], but this finding is not consistent [99,100]. The RFLP seems to be only in partial linkage disequilibrium with the 3’ polymorphisms [98,100]. It is therefore unlikely to explain the association results of the BsmI, ApaI, and TaqI polymorphisms. In view of the considerable distance between the two sites ( 40 kb) and the different nature of the polymorphisms, the FokI site should be treated as a different marker. This also holds true for the recently described G to A sequence variation in the Cdx-2-binding element just upstream of exon 1A [101,102]. Arai and colleagues [102] reported the G allele to have a decreased transactivation capacity and to be associated with a 10% decreased lumbar spine BMD in 123 Japanese women. It is likely that still more polymorphisms, including functional ones, will be discovered in this complex promotor region and that larger population studies will be necessary to document the LD over the region and to evaluate the associations with relevant end points such as BMD and fracture risk. In particular, studies should be undertaken

660 such as was done for the lipoprotein lipase gene, in which the VDR gene is scanned systematically for sequence variations. Haplotyping and cladistic analyses should be used to identify groups of SNPs linked together and, thus, simplify the association analyses. c. Pleiotropic Effects The vitamin D endocrine system has been shown to be involved in a number of endocrine pathways related to calcium metabolism, immune modulation, regulation of cell growth and differentiation (of keratinocytes, osteoblasts, cancer cells, T-cells), and so on (for a review, see Chapter 9) [103]. Thus, for a pleiotropic “master” gene, such as the VDR, one can expect to find associations of this gene with multiple traits and disease phenotypes. Indeed, the VDR gene has been found to be associated with a number of different phenotypes (see Table 3). The associations with osteoarthritis, hyperparathyroidism, cancer, and infection susceptibility are supported by several independent and large studies reporting similar associations. However, as for osteoporosis, different alleles are sometimes reported to be the risk allele and so the same considerations as described earlier (Section IV,B,3,a) should be taken into account. In addition, the potential confounding effects that arise from this pleiotropy can influence the associations observed. For example, VDR gene variants can influence calcium metabolism through differential absorption in the intestine and, at the same time, influence bone turnover and the occurrence of osteophytosis. These multiple actions taken together result in a net effect on BMD measured at a certain site, at a certain age, and in a subject consuming a certain diet. d. Functional Studies Interpretation of the VDR association studies is severely hindered by the fact that many of the polymorphisms considered are anonymous. The likely explanation for any observed association is then assumed to be the presence of a truly functional sequence variation elsewhere in the gene, which is, to a certain extent, in linkage with an allele of the anonymous polymorphism studied. Although the identification of these functional polymorphisms in the VDR gene is still eagerly awaited, several investigators have nevertheless analyzed multiple bioresponse parameters using the anonymous polymorphisms, including the BsmI and Bsm-Apa-Taq haplotypes, and a poly(A) tract in the 3’UTR as well as the FokI polymorphism. These studies include in vitro cell biological and molecular biological studies and in vivo measurements of biochemical markers and response to treatments with vitamin D, calcium, and even hormone replacement therapy (HRT) or bisphosphonates. In view of what has been discussed earlier, it is not very surprising that these studies have not shown one allele being consistently associated with all of the different parameters. Major caveats to these studies are the use of (a) anonymous rather than functional

UITTERLINDEN, VAN LEEUWEN, AND POLS

polymorphisms to group subjects and cells by genotype and (b) different types of bioresponses and different cell types and cell culture conditions in which the vitamin D response might not be evident. Therefore, the identification of a functional polymorphism and the use of different welldefined cell types will help in clarifying the molecular mechanisms underlying the associations observed. Part of the initial efforts to identify functional sequence variations have been focused on the 3’ regulatory region because this is close to the anonymous markers used so far in associations studies (see Fig. 4). While the BsmI, ApaI, and TaqI RFLPs are located near the 3’ end of the gene, the LD extends into the 3’ regulatory region containing the UTR. Morrison and colleagues earlier showed the 3’ UTR to contain sequence variations that were suggested to explain the observed associations [75] and provided evidence of differential luciferase actvity for the two UTRs that are linked to the two most frequent haplotypes baT and BAt. Durrin and colleagues [92] have shown certain parts of the UTR, so-called destabilizing elements, to be involved in determining stability of the VDR-mRNA. However, when eight individuals, selected by their poly(A) genotype, were sequenced, no polymorphisms were found in the destabilizing elements of the 3’UTR. Furthermore, UTRs linked to the two most common variants (the baT and BAt haplotype) were not found to differ with respect to mRNA stability [92]. However, only few individuals were sequenced so variations could have still been missed, whereas heterologous constructs (human VDR-UTR with a rabbit -globin gene) and cell types (mouse NIH3T3 cells) were also used to test for functionality. Because UTRs display cell-type specific effects on mRNA stability, this could be important in demonstrating functionality of sequence variations in the UTR. Taken together, these data indicate that multiple polymorphic variations exist in the VDR gene that could each have different types of consequences. Thus, 5’ promotor variations will affect mRNA expression patterns and levels, whereas 3’ UTR sequence variations will affect the mRNA stability. In combination, these genotypic differences are likely to affect the VDR protein levels and/or function, depending on the cell type, developmental stage, and activation status. Figure 5 presents an overview of the interacting factors that contribute to the control of action of 1,25(OH)2D3. These include other proteins involved, e.g., in synthesis and metabolism of the hormone and in DNA binding. Thus, the phenotypic variability as observed in action of the vitamin D endocrine system is likely to involve not only the VDR but also other proteins. Hence, polymorphic variations in the genes encoding such proteins will also contribute to genotype – phenotype relationships concerning the VDR genotype associations and might also contribute to heterogeneity between studies.

CHAPTER 26 Genetics and Genomics of Osteoporosis

661

FIGURE 5 Interactions in the vitamin D endocrine system. Different steps in the action of 1,25(OH)2D3 are synthesis, metabolism, binding to vitamin Dbinding protein, VDR binding, and DNA binding, which in turn are regulated by multiple signals. Because the genes coding for these signals can also be polymorphic, polymorphisms in the VDR gene, together with a number of polymorphisms in other genes, will determine the eventual vitamin D response. In view of interactions between these proteins, even individuals of identical VDR genotype can be expected to show variation in the association with a particular phenotype. In summary, one can conclude that VDR gene variants seem to influence a number of biological end points, including those related to osteoporosis. However, the associations have different magnitudes, with BMD being one of the weaker effects. In different study populations, different alleles of the anonymous RFLPs can be found associated with the same end point. This probably reflects the fact that linkage disequilibrium between the anonymous marker alleles and the causative alleles in (or very near) the VDR gene is likely to be different between populations. Finding functional sequence variants that matter, establishing the phase of alleles across the entire VDR gene, and defining haplotype patterns are therefore required to better understand the VDR associations. 4. THE COLLAGEN TYPE I1 GENE a. Association Studies Using the Sp1 Polymorphism Mutations in the genes encoding collagen type I1 and collagen type I2 cause the Mendelian disease osteogenesis

imperfecta. Thus, these genes were considered early on as candidate genes for osteoporosis. While no frequent allelic variants could be found in the coding region of these genes [104], Grant et al. [105] found a G to T substitution in intron 1 of the COLIA1 gene at a potential binding site for the Sp1 transcription factor. They observed the binding site to indeed bind the Sp1 transcription factor and the “T” allele to have a population frequency of about 18%, making this a polymorphism of potential functional significance. In an analysis of 205 predominantly postmenopausal British women, they reported decreased BMD and an increased fracture risk for carriers of the T allele. In a larger cohort of 1778 Dutch Caucasian elderly women, associations of the T allele with decreased BMD and increased fracture risk could be confirmed with evidence for a gene dose effect [106]. While the COLIA1 genotype-dependent fracture risk was strikingly independent of BMD [106], the BMD differences in this large cohort of elderly women increased with age, suggesting a relation to rates of bone loss [106]. This

662 notion is supported by the observation of increased rates of bone loss for subjects carrying the “T” allele in a 5-year follow-up analysis of 243 American men and women of 65 years and older [107]. Although some studies have not been able to confirm a relationship between this polymorphism and aspects of osteoporosis [108 – 111], the association of the “T” allele with decreased BMD and/or increased risk for fracture has been confirmed in several other studies in France [112], Denmark [113], the United Kingdom [114,115], and the Czech Republic [116]. Part of these negative results could be due to lack of power, especially in view of the low frequency of the TT homozygotes; ethnic differences in allele frequency, which have been demonstrated for this polymorphism [117,118]; analysis of the wrong end point, e.g., in view of the age dependency of the genotype effect and the different age distributions among populations tested; or hitherto unknown gene – gene or gene – environment interactions. For example, we have found an interaction between VDR genotype and COLIA1 genotype in determining the susceptibility to fracture whereby the risk increases further in carriers of both the VDR “baT” haplotype allele and the COLIA1 “T” allele over those carrying only one of these risk alleles [88]. In addition, if the Sp1 polymorphism would be considered simply an anonymous polymorphism, the causative sequence variation would be in linkage with this site and many factors controlling LD could explain discrepant association results. However, discrepancies due to different linkage disequilibria among populations seems an unlikely explanation for the different association results because it was shown in a Scottish case-control study that this polymorphism, rather than any of three other polymorphic sites at the COLIA1 gene in an area of 44 kb, determines susceptibility to fracture [119]. Interestingly, the “T” allele was also associated with decreased bone mass levels in a small cohort of young prepubertal Mexican-American girls as determined by computed tomography (CT) [120]. However, no association could be found with BMD levels as determined by dual energy Xray absorptiometry (DXA) measurement in prepubertal Caucasian children [121], but this could also be due to differences in measurement techniques to assess bone characteristics apart from the reasons mentioned earlier. In this regard, we also observed the Sp1 polymorphism to predict differences in heel ultrasound SOS measurements in the Rotterdam Study, independent of BMD differences (Kann et al., unpublished observations). The age dependency of the genotype-related differences suggests that the Sp1 polymorphism-related differences in bone characteristics (such as measured by DXA, CT, or ultrasound) could become evident in periods of increased collagen turnover during “stress periods,” such as the pubertal growth spurt, postmenopausal bone loss, and during aging. In line with this, we have observed the increased fracture

UITTERLINDEN, VAN LEEUWEN, AND POLS

risk to be most prominent among the TT homozygotes for “frailty” fractures of the hip, pelvis, and upper humerus rather than for nonfrailty fractures of the wrist and vertebra [122]. The association with the frailty phenotype is in line with the observations that the COLIA1 “T” allele is associated with differences in body frame size as measured by height [112] and weight [106,112]. b. Functional Studies Importantly, evidence suggests that the COLIA1 “T” allele has direct biological effects, which could explain the observed associations. The first report on the G to T polymorphism demonstrated the putative binding of the Sp1 transcription factor protein [105]. Subsequent preliminary reports suggested that the “T” allele binds the Sp1 protein twofold stronger and is associated with a threefold higher COLI1 mRNA level [123], and with increased COLI1 protein expression levels [124]. In cultured osteoblasts, such differences lead to altered COLI1/COLI2 protein ratios, very similar to what is seen for null mutations (allelic “knockouts”) in osteogenesis imperfecta patients, but to a much milder extent. On the basis of these so-called null mutations in OI patients, it can be speculated that an increased proportion of the COLIA1 homotrimer, such as could be the case in GT and TT subjects, would lead to more fragile bones. This notion is strongly supported by the observation that the “T” allele was found associated with decreased bone strength and that the yield strength of bone taken from the femoral neck was about half in “GT” heterozygotes compared to that of “GG” homozygotes [124]. This explanation of the COLIA1 Sp1 genotype effect is further supported by what is seen in the oim/oim mouse. In this naturally occurring mutant mouse strain, a COLIA1 homotrimer is produced due to a nonsense mutation in the COLIA2 gene. The phenotype of homozygous oim mice includes skeletal fractures, generalized osteopenia, and small body size [125], aspects of osteoporosis that are also observed in human TT homozygotes. Thus, in summary, a strong case is being built to implicate this polymorphism in osteoporosis. The overall effect on BMD in postmenopausal women appears to be small ( 2%), while there probably is a concomitant effect on bone structure and quality resulting in substantially increased fracture risk, mostly independent of BMD. The effects on bone and body size increase with age, which could contribute to frailty fractures being the predominant type of fracture found in carriers. Whether the Sp1 sequence variation is the only frequent functional polymorphism in this gene still remains to be established. Pleiotropic effects of this gene and interactions with environmental factors and other osteoporosis candidate genes have to be explored further. 5. OTHER GENES Although the VDR and the COLIA1 polymorphisms have received greatest attention so far, polymorphisms in

CHAPTER 26 Genetics and Genomics of Osteoporosis

several other candidate genes have also been studied. Mostly anonymous polymorphisms were studied in genes, including steroid receptor genes, cytokine genes, bone matrix proteins, and more exotic osteoporosis candidates, such as apolipoprotein E and the HLA genes. Although some of these showed associations with low BMD, increased fracture risk, or other skeletal phenotypes, the associations will need to be replicated in additional, preferably larger, populations to undergo the same scrutiny as applied to the VDR and COLIA1 gene polymorphisms. In addition, the identification of functional polymorphisms and description of the LD and haplotypes across the gene will clarify which SNP(s) contributes in what way to a particular phenotypic end point of interest. Table 5 lists “bone” genes, ordered by chromosomal location and with some of their characteristics, which by various means have been implicated in bone metabolism. For most of them an increased and/or specific expression in bone cells has been demonstrated. Additional lines of evidence also implicate them as osteoporosis candidate genes. For example, mutations in the gene of interest have been found in Mendelian diseases with a “bone” phenotype, some of them have been analyzed in genetically engineered mice, and some of them have been analyzed in association studies and found to be associated with differences in BMD and/or fracture risk. This list is not exhaustive and is certain to undergo changes over time when the Human Genome Project is completed and studies of the genetics of osteoporosis progress. However, by comparing Fig. 2 and 3, it becomes clear that some candidate gene loci have been identified already by multiple analytical approaches. For example, the COLIA1 locus (1) has been identified by genome searches (although with low LOD scores), (2) is a causative gene for the Mendelian bone disorder osteogenesis imperfecta, (3) leads to bone phenotypes in genetically engineered mice, e.g., in KO models as in Mov-13 mice or in spontaneous mutant strains such as oim/oim, and (4) by analyzing the Sp1-binding site polymorphism has been shown to be associated with osteoporosis end points such as decreased BMD and increased fracture risk. Such corroborative evidence makes genes such as the COLIA1 gene more likely to be true osteoporosis susceptibility genes. Thus, after genome searches, populationbased association analyses can determine to what extent the susceptibility gene determines variation in one or more of the parameters of osteoporosis, while, in addition, gene – gene and gene – environment interaction can be studied.

V. APPLICATIONS AND PROSPECTS We are now in the midst of an exciting era of molecular genetic studies of complex diseases such as diabetes, cancer, and osteoporosis. With the Human Genome Project approaching completion, a plethora of genes is ex-

663 pected to be identified, although without many clues to their function. Nevertheless, many genes will have to be scrutinized for their potential contribution to risk for osteoporosis. To this end, analytical strategies will probably develop along two parallel lines, including, on the one hand, functional studies such as protein expression studies, cell biological studies, and construction of animal models and, on the other hand, genetic studies, including linkage analysis in pedigrees, sib pair genome scans, and association analyses in populations. Thus, in the near future we can expect the molecular genetic scrutiny of a vast number of osteoporosis candidate genes. Their 5’ promoter, coding and 3’ regulatory regions will be analyzed for functional polymorphic variations. The functional effects of such variations will analyzed in cell culture models and animal transgenes and knockout models. This will give information on the molecular mechanisms and will also be valuable in the design of molecular interventions for therapeutic applications. Such genetic variations are also likely to be used as predictors of risk, possibly in combination with known and easily accessible risk factors such as age. In addition, some genetic markers will find applications in pharmacogenomics as predictors of response to treatment when particular medication will be applied in the treatment of osteoporosis or when particular diets are prescribed involving, for example, vitamin D or calcium. In complex diseases such as osteoporosis, interactions with environmental factors can determine the expression of genetic susceptibility to fracture. Therefore, there is considerable potential for influencing the susceptibility, for example, through dietary calcium and vitamin D intake and through exercise. However, in terms of treatment, it remains to be established to what extent it is actually possible to change life-long habits in the elderly. Clearly, the next phase in the genetics and genomics of osteoporosis will be to define the relative contribution of all these gene variants to differences in risk factors for osteoporosis. Thus, it will be likely that different polymorphisms in different genes are affecting different types of end points in different types of subjects under different types of circumstances. For example, some gene variants will have effects most prominently on vertebral fracture risk in women, through a mechanism independent of BMD, whereas other gene variants will specifically influence BMD differences in men. It is noteworthy that most of the approaches currently focus on identifying BMD genes. However, as emphasized earlier, low BMD is only one of several risk factors for the clinically most relevant end point in osteoporosis: fracture. Therefore, even given the attractive perspective of a “BMD gene map” of the human genome, much is still to be learned about what determines fracture susceptibility in molecular terms. This also touches on the aspect of pleiotropic effects of genes, which will increasingly influence the analysis of

664

UITTERLINDEN, VAN LEEUWEN, AND POLS

genetic disease risk. Many genes are involved in several metabolic pathways, and thus genetic variations are likely to affect a number of clinical end points (see also Table 3). Indeed, it can be expected that the map of osteoporosis genes will have considerable bearing on maps of susceptibility genes for many other age-related complex traits, such as osteoarthritis, diabetes, cardiovascular disease, and cancer. Together these collected complex gene maps will contribute to our insight in to several features of the aging process and longevity, with one prominent aspect being osteoporosis.

References 1. D. Botstein, D. L. White, M. Skolnik, and R. W. Davis, Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am. J. Hum. Genet. 32, 314 – 331 (1980). 2. E. S. Lander and N. J. Schork, Genetic dissection of complex traits. Science 265, 2037 – 2048 (1994). 3. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, et al., Risk factors for hip fracture in white women. N. Engl. J. Med. 332, 767 – 773 (1995). 4. D. M. Smith, W. E. Nance, K. W. Kang, J. C. Christian, and C. C. Johnston, Genetic factors in determining bone mass. J. Clin. Invest. 52, 2800 – 2808 (1973). 5. N. A. Pocock, J. A. Eisman, J. L. Hopper, G. M. Yeates, P. N. Sambrook, and S. Ebert, Genetic determinants of bone mass in adults: A twin study. J. Clin. Invest. 80, 706 – 710 (1987). 6. L. Flicker, J. L. Hopper, L. Rodgers, B. Kaymakci, R. M. Green, and J. D. Wark, Bone density determinants in elderly women: A twin study. J. Bone Miner. Res. 10, 1607 – 1613 (1995). 7. N. K. Arden, J. Baker, C. Hogg, et al., The heritablity of bone mineral density, ultrasound of the calcaneus and hip axis length: A study of postmenopausal twins. J. Bone Miner. Res. 11, 530 – 534 (1996). 8. L. Flicker, K. G. Faulkner, J. L. Hopper, R. M. Green, B. Kaymakci, C. A. Nowson, D. Young, and J. D. Wark, Determinants of hip axis length in women aged 10-89 years: A twin study. Bone 18, 41 – 45 (1996). 9. D. A. Nickerson, S. L. Taylor, K. M. Weiss, A. G. Clark, R. G. Hutchinson, J. Stengard, et al., DNA sequence diversity in a 9.7-kb region of the human lipoprotein lipase gene. Nature Genet. 19, 233 – 240 (1998). 10. M. J. Rieder, S. L. Taylor, A. G. Clark, and D. A. Nickerson, Sequence variation in the human angiotensin converting enzyme. Nature Genet. 22, 59 – 62 (1999). 11. M. K. Halushka, J. B. Fan, K. Bentley, L. Hsie, N. Shen, A. Weder, R. Cooper, R. Lipshutz, and A. Chakravarti, Patterns of single-nucleotide polymorphisms in candidate genes for blood-pressure homeostasis. Nature Genet. 22, 239 – 247 (1999). 12. F. Cambien, O. Poirier, V. Nicaud, S. Herrman, C. Mallet, S. Ricard, et al., Sequence diversity in 36 candidate genes for cardiovascular disorders. Am. J. Hum. Genet. 65, 183 – 191 (1999). 13. M. Cargill, D. Altshuler, J. Ireland, P. Sklar, K. Ardlie, N. Patil, et al., Characterization of single-nucleotide polymorphisms in coding regions of human genes. Nature Genet. 22, 231 – 238 (1999). 14. D. G. Wang, J.-B. Fan, C.-J. Siao, A. Berno, P. Young, R. Sapolsky, et al., Large-scale identification, mapping, and genotyping of singlenucleotide polymorphisms in the human genome. Science 280, 1077 – 1082 (1998). 15. L. Kruglyak, Prospects for whole-genome linkage disequilibrium mapping of common disease genes. Nature Genet. 22, 139 – 144 (1999).

16. A. Collins, C. Lonjou, and N. E. Morton, Genetic epidemiology of single-nucleotide polymorphisms. Proc. Natl. Acad. Sci. USA 96, 15173 – 15177 (1999). 17. A. C. Peterson, A. Di Rienzo, A.-E. Lehesjoki, A. de la Chapelle, M. Slatkin, and N. B. Freimer, The distribution of linkage disequilibrium over anonymous genome regions. Hum. Mol. Genet. 4, 887 – 894 (1995). 18. R. J. Peterson, D. Goldman, and J. C. Long, Effects of worldwide population subdivision on ALDH2 linkage disequilibrium. Genome Res. 9, 844 – 852 (1999). 19. C. F. Terry, V. Loukaci, and F. R. Green, Cooperative influence of genetic polymorphisms on interleukin-6 (IL-6) transcriptional regulation. J. Biol. Chem. 22, 2341 – 2349 (2000). 20. F. Cambien, S. Ricard, A. Troesch, C. Mallet, L. Généraz, A. Evans, et al., Polymorphisms of the transforming growth factor-1 gene in relation to myocardial infraction and blood pressure. Hypertension 28, 881 – 887 (1996). 21. A. G. Clark, K. M. Weiss, D. A. Nickerson, S. L. Taylor, A. Buchanan, J. Stengard, et al., Haplotype structure and population genetic inferences from nucleotide-sequence variation in human lipoprotein lipase. Am. J. Hum. Genet. 63, 595 – 612 (1998). 22. A. R. Templeton, A. G. Clark, K. M. Weiss, D. A. Nickerson, E. Boerwinkle, and C. F. Sing, Recombinational and mutational hotspots within the human lipoprotein lipase gene. Am. J. Hum. Genet. 66, 69 – 83 (2000). 23. M. Miyao, H. Morita, T. Hosoi, H. Kurihara, S. Inoue, S. Hosihino, et al., Association of Methylenetetrahydrofolate reductase (MTHFR) polymorphism with bone mineral density in postmenopausal Japanese women. Calcif. Tissue Int. 66, 190 – 194 (2000). 24. W. S. Browner, A. R. Pressman, M. C. Nevitt, J. A. Cauley, and S. R. Cummings, Association between low bone density and stroke in elderly women: The study of osteoporotic fractures. Stroke 24, 940 – 946 (1993). 25. R. Scragg, R. Jackson, I. M. Holdaway, T. Lim, and R. Beaglehole, Myocardial infarction is inversely associated with plasma 25-hydroxy vitamin D3 levels:a community-based study. Int. J. Epidemiol. 19, 559 – 563 (1990). 26. D. Hart, I. Mootoosamy, D. V. Doyle, and T. D. Spector, The relationship between osteoarthritis and osteoporosis in the general population: The Chingford study. Ann. Rheum. Dis. 53, 158 – 162 (1994). 27. A. G. Uitterlinden, H. Burger, Q. Huang, E. Odding, C. M. van Duijn, A. Hofman, et al., Vitamin D receptor genotype is associated with radiographic osteoarthritis at the knee. J. Clin. Invest. 100, 259 – 263 (1997). 28. Y. Yamada, H. Okuizumi, A. Miyauchi, Y. Takagi, K. Ikeda, and A. Harada, Association of transforming growth factor 1 genotype with spinal osteophytosis in japanese women. Arthritis Rheum. 43, 452 – 460 (2000). 29. M. Devoto, K. Shimoya, J. Caminis, J. Ott, A. Tenenhouse, M. P. Whyte, et al., First-stage autosomal genome screen in extended pedigrees suggests genes prediposing to low bone mineral density on chromosomes 1p, 2p, and 4q. Eur. J. Hum. Genet. 6, 151 – 157 (1998). 30. L. D. Spotila, M. Devoto, J. Caminis, R. Kosich, J. Korkko, J. Ott, A. Tenenhouse, and D. J. Prockop, Suggested linkage of low bone density to chromosome 1p36 is extended to a second cohort of sibpairs. Bone 23, S277 (1998). [Abstract T322] 31. O. M. E. Albagha, F. E. A. McGuigan, D. Reid, and S. H. Ralston, Association mapping of a locus for regulation of bone mass in the normal population using DNA pooling. J. Bone Miner. Res. 14, S142 (1999). [Abstract 1037] 32. L. D. Spotila, M. Devoto, A. Caminis, and A. Tenenhouse, Quantitative trait locus for bone mineral density on chromosome 1p36. Calcif. Tissue Int. 64, S36 (1999). [Abstract I-23]

CHAPTER 26 Genetics and Genomics of Osteoporosis 33. T. Niu, C. Chen, H. Cordell, J. Yang, B. Wang, Z. Wang, et al., A genome-wide scan for loci linked to forearm bone mineral density. Hum. Genet. 104, 226 – 233 (1999). 34. B. D. Mitchell, R. L. Bauer, R. Perez, S. H. Slifer, S. A. Cole, J. E. Hixson, and J. W. McCluer, Genome-wide scan for loci influencing bone density in Mexican Americans. Am. J. Hum. Genet. 63, A301 (1998). [Abstract 1741] 35. D. L. Koller, M. J. Econs, L. A. Rodriguez, J. C. Christian, S. L. Hui, P. Morin, et al., Genome screen for QTLs contributing to normal variation in bone mineral density and osteoporosis. J. Bone Miner. Res. 14(S1), S141 (1999). [Abstract 1036] 36. E. L. Duncan, M. A. Brown, J. Sinsheimer, J. Bell, A. J. Carr, B. P. Wordsworth, and J. A. H. Wass, Suggestive linkage of the parathyroid receptor type 1 to osteoporosis. J. Bone Miner. Res. 14, 1993 – 1999 (1999). 37. N. Ota, S. C. Hunt, T. Nakajima, T. Suzuki, T. Hosoi, H. Orimo, Y. Shirai, and M. Emi, Linkage of interleukin 6 locus to human osteopenia by sibling pair analysis. Hum. Genet. 105, 253 – 257 (1999). 38. D. L. Koller, L. A. Rodriguez, J. C. Christian, C. W. Slemenda, M. J. Econs, S. L. Hui, et al., Linkage of a QTL contributing to normal variation in bone mineral density to chromosome 11q12 – 13. J. Bone Miner. Res. 13, 1903 – 1908 (1999). 39. I. Takacs, D. L. Koller, M. Peacock, J. C. Christian, S. L. Hui, P. M. Conneally, et al., Sibling pair linkage and association studies between bone mineral density and the insulin-like growth factor I gene locus. J. Clin. Endocrinol. Metab. 84, 4467 – 4471 (1999). 40. M. C. Mahaney, P. Morin, L. A. Rodriguez, D. E. Newman, and J. Rogers, A quantitative trait locus on chromosome 11 may influence bone mineral density at several sites: Linkage analysis in pedigreed baboons. J. Bone Miner. Res. 12, S118 (1997). [Abstract 64] 41. www.informatics.jax.org 42. R. F. Klein, S. R. Mitchell, T. Phillips, J. K. Belknap, and E. S. Orwoll, Quantitative trait loci affecting peak bone mineral density in mice. J. Bone Miner. Res. 13, 1648 – 1656 (1998). 43. M. Shimizu, K. Higuchi, B. Bennett, C. Xia, T. Tsuboyama, S. Kasai, et al., Identification of peak bone mass QTL in a spontaneously osteoporotic mouse strain. Mamm. Genome 10, 81 – 87 (1999). 44. W. G. Beamer, K. L. Shultz, G. A. Churchill, W. N. Frankel, D. J. Baylink, C. J. Rosen, and L. R. Donahue, Quantitative trait loci for bone density in C57BL/6J and CAST/EiJ inbred mice. Mamm. Genome 10, 1043 – 1049 (1999). 45. H. Benes R. S. Weinstein, Z. Wenhui, J. J. Thaden, R. L. Jilka, S. C. Manolagas, and R. J. Shmookler, Chromosomal mapping of osteopenia-associated quantitative trait loci using closely related mouse strains. J. Bone Miner. Res. 15, 626 – 633 (2000). 46. www.ncbi.nlm.nih.gov/PubMed/ 47. www.ncbi.nlm.nih.gov/Omim/ 48. www.gdb.org/gdb/ 49. M. L. Johnson, G. Gong. W. Kimberling, S. M. Recker, D. B. Kimmel, and R. R. Recker, Linkage of a gene causing high bone mass to human chromosome 11 (11q12 – 13) Am. J. Hum. Genet. 60, 1326 – 1332 (1997). 50. C. Heany, H. Shalev, K. Elbedour, R. Carmi, J. B. Staack, V. C. Sheffield, and D. R. Beier, Human autosomal recessive osteopetrosis maps to 11q13, aposotion predicted by comparative mapping of the murine osteosclerosis (oc) mutation. Hum. Mol. Genet. 7, 1407 – 1410 (1998). 51. Y. Gong, M. Vikkula, L. Boon, J. Liu, P. Beighton, R. Ramesar, et al., Osteoporosis-pseudoglioma syndrome, a disorder affecting skeletal strength and vision, is assigned to chromosome region 11q12 – 13. Am. J. Hum. Genet. 59, 146 – 151 (1996). 52. W. Van Hul, W. Balemans, E. van Hul, F. G. Dikkers, H. Obee, R. J. Stokroos et al., van Buchem disease (hyperostosis corticalis generalisata) maps to chromosome 17q12-q21. Am. J. Hum. Genet. 62, 391 – 399 (1998).

665 53. W. Van Hul, J. Bollerslev, J. Gram, E. van Hul, W. Wuyts, O. Benichou, F. Vanhoenacker, and P. J. Willems, Localization of a gene for autosomal dominant osteopetrosis (Albers-Schönberg disease) to chromosome 1p21. Am. J. Hum. Genet. 61, 363 – 369 (1997). 54. B. Y. Reed, H. J. Heller, and C. Y. C. Gitomer, Pak, Mapping a gene defect in absorptive hypercalciuria to chromosome 1q23-3-q24. J. Clin. Endocrinol. Metab. 84, 3907 – 3913 (1999). 55. S. I. Haslam, W. van Hul, A. Morales-Piga, W. Balemans, J. L. SanMillan, K. Nakatsuka, P. Willems, N. E. Haites, and S. H. Ralston, Paget’s disease of bone: Evidence for a susceptibility locus on chromosome 18q and for genetic heterogeneity. J. Bone Miner. Res. 13, 911 – 917 (1998). 56. A. E. Hughes, S. H. Ralston, J. Marken, C. Bell, H. MacPherson, R. G. H. Wallace et al., Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nature Genet. 24, 45 – 48 (2000). 57. F. Lazner, M. Gowen, and I. Pavasovic, Kola, Osteopetrosis and osteoporosis: Two sides of the same coin. Hum. Mol. Genet. 8, 1839 – 1846 (1999). 58. H. Yoshida, S. Hayashi, T. Kunisada, M. Ogawa, S. Nishikawa, H. Okamura, T. Sudo, and L. D. Shultz, The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442 – 444 (1990). 59. J. C. Scimeca, A. Franchi, C. Trojani, H. Parrinello, J. Grosgeorge, C. Robert et al., The gene encoding the mouse homologue of the human osteoclast-specific 116-kDa V-ATPase subunit bears a deletion in osteosclerotic (oc/oc) mutants. Bone 26, 207 – 213 (2000). 60. Y. P. Li, W. Chen, Y. Liang, E. Li, and P. Stashenko, Atp6i-deficient mice exhibit severe ostepetrosis due to loss of osteoclast-mediated extracellular acidification. Nature Genet. 23, 447 – 451 (1999). 61. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src protooncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 62. A. E. Grigoriadis, Z. Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast-macrophage lineage determination and bone remodelling. Science 266, 443 – 448 (1994). 63. V. Iotsova, J. Caamano, J. Loy, Y. Yang, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NF-kB1 and NF-kB2. Nature Med. 3, 1285 – 1289 (1997). 64. K. P. McHugh, K. Hodivala-Dilke, M.-H. Zheng, N. Namba, J. Lam, D. Novack et al., Mice lacking 3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest. 105, 433 – 440 (2000). 65. P. Saftig, E. Hunziker, O. Wehmeyer, S. Jones, A. Boyde, W. Rommerskirch, J. D. Moritz, P. Schu, and K. von Figura, Impaired osteoclastic bone resorption leads to osteopetrosis in cathepsin-K-deficient mice. Proc. Natl. Acad. Sci. USA 95, 13453 – 13458 (1998). 66. B. Li, S. Boast, K. de los Santos, I. Schieren, M. Quiroz, S. L. Teitelbaum, M. M. Tondravi, and S. P. Goff, Mice deficient in Abl are osteoporotic and have defects in osteoblast maturation. Nature Genet. 24, 304 – 308 (2000). 67. G. Crawford-Sharpe, W. Beamer, C. Rosen, D. Baylink, C. Richman and J. Barker, A new mouse model for osteoporosis maps to mouse chromosome 14. J. Bone Miner. Res. 12, S119 (1997). [Abstract 67] 68. D. B. Lewis, H. D. Liggitt, E. L. Effmann, T. Motley, S. L. Teitelbaum, K. J. Jepsen, et al., Osteoporosis induced in mice by overproduction of interleukin 4. Proc. Natl. Acad. Sci. USA 90, 11618 – 11622 (1993). 69. D. G. Marsh, J. D. Neely, D. R. Breazeale, B. Ghosh, L. R. Freidhoff, E. Ehrlich-Kautzky et al., Linkage analysis of IL4 and other chromosome 5q31.1 markers and total serum immunoglobulin E concentrations. Science 264, 1152 – 1156 (1994).

666 70. T. Xu, P. Bianco, L. Fisher, G. Longenecker, E. Smith, S. Goldstein, et al., Targeted dieruption of thebiglycan gene leads to an osteoporosis-like phenotype in mice. Nature. Genet. 20, 78 – 82 (1998). 71. M. Kuro-o, Y. Matsumura, H. Aizawa, H. Kawaguchi, T. Suga, T. Utsugi, et al., Mutation of the mouse klotho gene leads to a syndrome resembling aging. Nature 390, 45 – 51 (1997). 72. Y. Matsumura, H. Aizawa, T. Shiraki-Ida, R. Nagai, M. Kuro-o, and Y.-I. Nabeshima, Identification of the human klotho gene and its two transcripts encoding membrane and secreted klotho protein. Biochem. Biophys. Res. Commun. 242, 626 – 630 (1998). 73. N. Bucay, I. Sarosi, C. R. Duncstan, S. Morony, J. Tarpley, C. Capparelli, S., Scull, et al., Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 74. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelly, M. S. Chang, R. Luthy, et al., Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 75. N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89, 6665 – 6669 (1992). 76. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 77. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles (correction). Nature 387, 106 (1997). 78. A. G. Uitterlinden, H. A. P. Pols, H. Burger, Q. Huang, P. L. A. van Daele, C. M. van Duijn, A. Hofman, J. C. Birkenhäger, and J. P. T. M. van Leeuwen, A large scale population based study of the association of vitamin D receptor gene polymorphisms with bone mineral density. J. Bone Miner. Res. 11, 1242 – 1248 (1996). 79. G. S. Cooper and D. M. Umbach, Are vitamin D receptor polymorphisms associated with bone mineral density? A meta-analysis. J. Bone Miner. Res. 11, 1841 – 1849 (1996). 80. G. Gong. H. S. Stern, S.-C. Cheng, N. Fong, J. Mordeson, H.-W. Deng, R. R. Recker, The association of bone mineral density with vitamin D receptor gene polymorphisms. Osteop. Int. 9, 55 – 64 (1999). 81. J. Sainz, J. M. van Tornhout, L. Loro, J. Sayre, T. F. Roe, and V. Gilsanz, Vitamin D receptor gene polymorphisms and bone density in prepubertal American girls of Mexican descent. N. Engl. J. Med. 337, 77 – 82 (1997). 82. L. A. Rubin, G. A. Hawker, V. D. Peltekova, L. J. Fielding, R. Ridout, and D. E. C. Cole, Determinants of peak bone mass in a young female canadian cohort. J. Bone Miner. Res. 14, 633 – 643 (1999). 83. R. W. Keen, P. J. Major, J. S. Lanchburry, and T. D. Spector, Vitamin D receptor gene polymorphism and bone loss. Lancet 345, 990 (1995). 84. A. Gough, P. Sambrook, J, Devlin, J. Lilley, A. Huisoon, J. Betteridge, et al., Effect of vitamin D receptor gene alleles on bone loss in early rheumatoid arthritis. J. Rheumatol. 25, 864 – 868 (1998). 85. E. A. Krall, P. Parry, J. B. Lichter, and B. Dawson-Hughes, Vitamin D receptor alleles and rates of bone loss: Influences of years since menopause and calcium intake. J. Bone Miner. Res. 10, 978 – 984 (1995). 86. D. P. Kiel, R. H. Myers, L. A. Cupples, X. F. Kong, X. H. Zhu, J. Ordovas, et al., The BsmI vitamin D receptor restriction fragment length polymorphism (bb) influences the effect of calcium intake on bone mineral density. J. Bone Miner. Res. 12, 1049 – 1057 (1997). 87. D. Feskanich, D. J. Hunter, W. C. Willett, S. E. Hankinson, B. W. Hollis, H. L. Hough, K. T. Kelsey, and G. A. Colditz, Vitamin

UITTERLINDEN, VAN LEEUWEN, AND POLS

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

D receptor genotype and the risk of bone fractures in women. Epidemiology 9, 535 – 539 (1998). A. G. Uitterlinden, F. Yue, J. P. T. M. van Leeuwen, and H. A. P. Pols, Polymorphisms in the vitamin D receptor- and collagen type I1 gene predict osteoporotic fracture in women. Am. J. Hum. Genet. 63, A223 (1998). [abstract 1279] L. A. Houston, S. F. A. Grant, D. M. Reid, and S. H. Ralston, Vitamin D receptor polymorphism, bone mineral density, and osteoporotic vertebral fracture: Studies in a UK population. Bone 18, 249 – 252 (1996). J. P. Berg, J. A. Falch, and E. Haug, Fracture rate, pre- and postmenopausal bone loss are not associated with vitamin D receptor genotype in a high-endemic area of osteoporosis. Eur. J. Endocrinol. 135, 96 – 100 (1996). S. A. Ingles, R. W. Haile, B. E. Henderson, L. N. Kolonel, G. Nakaichi, C.-Y. Shi, M. C. Yu, R. K. Ross, and G. A. Coetzee, Strength of linkage disequilibrium between two vitamin D receptor markers in five ethnic groups: Implications for association studies. Cancer Epi. Biom. Prev. 6, 93 – 98 (1997). L. K. Durrin, R. W. Haile, S. A. Ingles, and G. A. Coetzee, Vitamin D receptor 3 -untranslated region polymorphisms: Lack of effect on mRNA stability. Biochim. Biophys. Acta 1453, 311 – 320 (1999). K.-I. Miyamoto, R. A. Kesterson, H. Yamamoto, Y. Taketani, E. Nishiwaki, S. Tatsumi, et al., Structural organization of the human vitamin D receptor chromosomal gene and its promotor. Mol. Endocrinol. 11, 1165 – 1179 (1997). L. A. Crofts, M. S. Hancock, N. A. Morrison, and J. A. Eisman, Multiple promotors direct the tissue-specific expression of novel Nterminal variant human vitamin D receptor gene transcripts. Proc. Natl. Acad. Sci. USA 95, 10529 – 10534 (1998). H. Arai, K.-I. Miyamoto, Y. Taketani, H. Yamamoto, Y. Iemori, K. Morita et al., A vitamin D receptor gene polymorphism in the translation initiation codon: Effect on protein activity and relation to bone mineral density in Japanese women. J. Bone Miner. Res. 12, 915 – 921 (1997). P. W. Jurutka, L. S. Remus, G. K. Whitfield, P. D. Thompson, J. C. Hsieh, H. Zitzer, et al., The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol. Endocrinol., 14, 401 – 420 (2000). C. Gross, T. R. Eccleshall, P. J. Malloy, M. L. Villa, R. Marcus, and D. Feldman, The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal mexican-american women. J. Bone Miner. Res. 11, 1850 – 1855 (2000). L. Gennari, L. Becherini, R. Mansani, L. Masi, A. Falchetti, A. Morelli, et al., FokI polymorphism at translation inititation site of the vitamin D receptor gene predicts bone mineral density and vertebral fractures in postmenopausal italian women. J. Bone Miner. Res. 14, 1379 – 1386 (1999). T. R. Eccleshall, P. Garnero, C. Gross, P. D. Delmas, and D. Feldman, Lack of correlation between start codon polymorphism of the vitamin D receptor gene and bone mineral density in premenopausal french women: The OFELY study. J. Bone Miner. Res. 13, 31 – 35 (1998). S. Ferrari, R. Rizzoli, D. Manen, D. Slosman, and J.-P. Bonjour, Vitamin D receptor start codon polymorphisms (FokI) and bone mineral density: Interaction with age, dietary calcium, and 3 -end region polymorphisms. J. Bone Miner. Res. 13, 925 – 930 (1998). H. Yamamoto, K.-I. Miyamoto, L. Bailing, Y. Taketani, M. Kitano, Y. Inoue, et al., The caudal-related homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J. Bone Miner. Res. 14, 240 – 247 (1999). H. Arai, K. Miyamoto, M. Yoshida, M. Kubota, H. Yamamoto, Y. Taketani et al., The polymorphism in the caudal-related

CHAPTER 26 Genetics and Genomics of Osteoporosis

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

homeodomain protein Cdx-2 binding element in the human vitamin D receptor gene. J. Bone Miner. Res. 14, S191 (1999). [Abstract T084] S. Christakos, M. Raval-pandya, R. P. Wernyj, and W. Yang, Genomic mechanisms involved in the pleitropic actions of 1,25-dihydroxyvitamin D3. Biochem. J. 316, 361 – 371 (1996). L. D. Spotila, A. Colige, L. Sereda, C. Constantinou-Deltas, M. P. Whyte, B. L. Riggs et al., Mutation analysis of coding sequences for type I procollagen in individuals with low bone density. J. Bone Miner. Res. 9, 923 – 932 (1994). S. F. A. Grant, D. M. Reid, G. Blake, R. Herd, I. Fogelman, and S. H. Ralston, Reduced bone density and osteoporotic vertebral fracture associated with a polymorphic Sp 1 binding site in the collagen type I1 gene. Nature Genet. 14, 203 – 205 (1996). A. G. Uitterlinden, H. Burger, Q. Huang, F. Yue, F. E. A. McGuigan, S. F. A. Grant, A. Hofman, J. P. T. M. van Leeuwen, H. A. P. Pols, and S. H. Ralston, Relation of alleles at the collagen type I1 gene to bone density and risk of osteoporotic fractures in postmenopausal women. N. Engl. J. Med. 338, 1016 – 1021 (1998). S. S. Harris, M. S. Patel, D. E. C. Cole, and B. Dawson-Hughes, Associations of the collagen type I1 Sp 1 polymorphism with fiveyear rates of bone loss in older adults. Calcif. Tissue Int. 66, 268 – 271 (2000). M. Liden, B. Wilen, S. Ljunghall, and H. Melhus, Polymorphism at the Sp 1 binding site in the collagen type I1 gene does not predict bone mineral density in postmenopausal women in Sweden. Calcif. Tissue Int. 63, 293 – 295 (1998). F. G. Hustmeyer, G. Liu, C. C. Johnston, J. Christian, and M. Peacock, Polymorphism at an Sp 1 binding site of COLIA1 and bone mineral density in premenopausal female twins and elderly fracture patients. Osteop. Int. 9, 346 – 350 (1999). K. O. Han, I. G. Moon, C. S. Hwang, J. T. Choi, H. K. Yoon, H. K. Min, and I. K. Han, Lack of intronic Sp 1 binding-site polymorphism at the collagen type I1 gene in healthy Korean women. Bone 24, 135 – 137 (1999). A. M. Heegaard, H. L. Jorgensen, A. W. Vestergaard, C. Hassager, S. F. A. Grant, and S. H. Ralston, Lack of influence of Sp 1 polymorphism in the collagen type I1 gene on the rate of bone loss in postmenopausal women followed for 18 years. Bone 23, S373 (1998). [Abstract W303] P. Garnero, O. Borel, S. F. A. Grant, S. H. Ralston, P. D. Delmas, Collagen I1 Sp 1 polymorphism, bone mass, and bone turnover in healthy french premenopausal women: The Ofely study. J. Bone Miner. Res. 13, 813 – 817 (1998). B. L. Langdahl, S. H. Ralston, S. F. A. Grant, and E. F. Eriksen, An Sp 1 binding site polymorphism in the COLIA1 gene predicts osteoprotic fractures in both men and women. J. Bone Miner. Res. 13, 1384 – 1389 (1998).

667 114. R. W. Keen, K. L. Woodford-Richens, S. F. A. Grant, S. H. Ralston, J. S. Lanchbury, and T. D. Spector, Association of polymorphism at the type I collagen (COLIA1) locus with reduced bone mineral density, increased fracture risk, and increased collagen turnover. Arthritis Rheum. 42, 285 – 290 (1999). 115. G. Hampson, C. Evans, R. J. Petitt, W. D. Evans, S. J. Woodhead, J. R. Peters, and S. H. Ralston, Bone mineral density, collagen type I1 genotypes and bone turnover in premenopausal women with diabetes mellitus. Diabetology 41, 1314 – 1320 (1998). 116. M. Weichetova, J. J. Stepan, D. Michalska, T. Haas, H. A. P. Pols, and A. G. Uitterlinden, COLIA1 polymorphism contributes to bone mineral density to assess prevalent wrist fractures. Bone 26, 287 – 290 (2000). 117. S. Beavan, A. Prentice, B. Dibba, L. Yan, C. Cooper, and S. H. Ralston, Polymorphism of the collagen type I(alpha)1 gene and ethnic differences in hip-fracture rates. N. Engl. J. Med. 339, 351 – 352 (1998). 118. T. Nakajima, N. Ota, Y. Shirai, A. Hata, H. Yoshida, T. Suzuki, T. Hosoi, H. Orimo, and M. Emi, Ethnic difference in contribution of Sp 1 site variation of COLIA1 gene in genetic predisposition to osteoporosis. Calcif. Tissue Int. 65, 352 – 353 (1999). 119. F. E. A. McGuigan, D. M. Reid, and S. H. Ralston, The Sp 1 binding site polymorphism rather than other polymorphic sites for the COLIA1 locus determines susceptibility to osteoporotic fracture. Osteoporos. Int. 11, 338 – 343 (2000). 120. J. Sainz, J. M. van Tornhout, J. Sayre, F. Kaufman, and V. Gilsanz, Association of collagen type I1 gene polymorphism with bone density in early childhood. J. Clin. Endocrinol. Metab. 84, 853 – 855 (1999). 121. C. Tao, S. Garnett, V. Petrauskas, and C. T. Cowell, No association was found between collagen 1 type I gene and bone density in prepubertal children. J. Clin. Endocrinol. Metab. 84, 4293 – 4295 (1999). 122. A. G. Uitterlinden, F. Yue, S. H. Ralston, J. P. T. M. van Leeuwen, and H. A. P. Pols, The collagen type I alpha 1 Sp 1 polymorphism predicts hip fracture in women. Bone 23, S161 (1998). [Abstract 1050] 123. E. E. Hobson, S. F. A. Grant, and S. H. Ralston, The functional effects on Sp 1 binding and allele specific transcription of a collagen I(I) (COLIA1) polymorphism. Bone, 22, 10S (1998). [Abstract C36] 124. V. Dean, E. E. Hobson, R. M. Aspden, S. P. Robins, and S. H. Ralston, Relationship between COLIA1 Sp 1 alleles, gene transcription, collagen production and bone strength. Bone 23, S161 (1998). [Abstract 1051] 125. S. D. Chipman, H. O. Sweet, D. J. McBride, M. T. Davisson, S. C. Marks, A. R. Shuldiner et al., Defective pro2(I) collagen synthesis in a recessive mutation in mice: A model of human osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993).

CHAPTER 27

Nutrition and Risk for Osteoporosis ROBERT P. HEANEY

Creighton University, Omaha, Nebraska 68178

I. Introduction II. Problems in the Investigation of Nutritional Effects on Bone III. The Notion of a Nutrient Requirement IV. The Natural Intake of Calcium and Vitamin D

V. VI. VII. VIII.

whether hormones, exercise, or nutrition (as in this case). Thus, not only is nutrition just one of several interacting factors in any given fracture syndrome, but it may play quite different roles, or none at all, in certain of those syndromes, while being of greater importance in others. This was first suggested in the 1979 report by Matkovic and colleagues [2] from Croatia, which showed that a high calcium intake was associated with a strikingly reduced hip fracture risk, but not with an altered risk of distal forearm fracture in the same population. Nutrition affects bone health in two qualitatively distinct ways. Bone tissue deposition, maintenance, and repair are the result of cellular processes, and the cells of bone responsible for these functions are as dependent upon nutrition as are the cells of any other tissue. The production of bone matrix, for example, requires the synthesis and posttranslational modification of collagen and an array of other proteins. Nutrients involved in such synthesis include protein, the vitamins C, D, and K, and the minerals copper, manganese, and zinc. Phosphorus is also involved indirectly in these cellular activities. Additionally, the skeleton serves as a very large nutrient reserve for two minerals, calcium and phosphorus, and the size of that reserve (i.e., the

I. INTRODUCTION A. Nutrition in the Osteoporotic Fracture Context In 1990, osteoporosis was redefined, for the first time in nearly a century, as a condition of skeletal fragility due to decreased bone mass and to microarchitectural deterioration of bone tissue, with a consequent increased risk of fracture [1]. This definition was conceptually important because it both acknowledged and encouraged a shift in thinking about osteoporosis from an anatomic to a dynamic condition. Low bone mass became a risk factor for fracture rather than, as formerly, the defining characteristic of the disease. This redefinition accompanied a growing recognition that osteoporosis is not a single disorder but a group of more or less discrete fracture syndromes, multifactorial both in etiology and in pathogenesis. The recognition not only of a multiplicity of pathogenetic factors, but of disease heterogeneity adds another dimension of complexity, which must be considered when describing and assessing the role of any single factor,

OSTEOPOROSIS, SECOND EDITION VOLUME 1

Calcium Vitamin D Vitamin K Other Essential Nutrients References

669

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

670 massiveness of the skeletal structures) is dependent in part on the daily balance between absorbed intake and excretory loss of these two minerals. Bone mass is also dependent upon a variety of nonnutritional factors, such as genetics, mechanical loading, and hormonal status. These dependencies complicate the interpretation of low bone mass values because, while low bone mass always means a reduced calcium reserve, a simple reduction in bone mass does not necessarily mean that it had a nutritional cause. While several of the chapters in this volume describe in considerable detail the diversity of osteoporosis, it will be helpful here to recapitulate very briefly what is known of the complex domain of osteoporotic fragility. Only against that background will it be possible to situate nutrition adequately amidst the array of other pathogenetic influences. Factors involved in osteoporotic fractures can be organized hierarchically to include the injury itself, the strength of the bone, the mass and density of the bone, and the adequacy of nutrition as it affects bone mass. Hip fracture is perhaps the most serious of the fragility fractures, inasmuch as it carries an excess mortality, is expensive, and causes significant deterioration in the quality of life for most of its survivors. It is, as well, a good example of the many interacting factors that constitute this fracture domain and is used as such in this preliminary overview. Figure 1 illustrates, schematically, how the various contributing factors interact for hip fracture. It also highlights probable sites in this schema at which nutrition plays a role. 1. FRAILTY AND INJURY Almost all fractures, even those we term “low-trauma,” occur as a result of some injury, i.e., the application of more force to the bone than it is able to sustain (see Chapter 19). Usually this is a result of a fall or the application of bad body mechanics (e.g., bending forward to lift a heavy object). Although fracture incidence patterns differ somewhat

FIGURE 1 Schematic representation of the interplay of principal factors thought to be important in hip fracture. Asterisks denote factors with a recognized nutritional determinant. Copyright Robert P. Heaney (1995). Reproduced with permission.

ROBERT P. HEANEY

from site to site, the risk of virtually all fractures rises with age, and all fractures contribute to the burden of illness, disability, and expense, which the elderly (and society) bear. The first factor to consider is the fall itself. Normally, postural reflexes operate to get the arms into position to break the force of the fall or to swing the body so that it lands on the buttocks (or both). These reflexes are almost always effective in younger individuals, but they commonly fail in the elderly. As a result, young people rarely strike the lateral portion of the trochanteric region of the hip when they fall, whereas the fragile elderly do so more commonly. The force of the impact, when falling from standing height, may well be sufficient to break even a healthy femur if that force is concentrated in a small enough impact area [3] (see also Chapter 19). Additionally, hip fracture is a particularly serious problem in undernourished elderly individuals who have less muscle and fat mass around the hip, and therefore less soft tissue through which the force of the impact can be distributed to a larger area of the lateral surface of the trochanter. Nutrition enters into this region of the fracture domain both through its effect on propensity to fall [4] and on maintenance of the soft tissue mass. This latter factor, particularly, is the rationale for the development and successful deployment of hip pads as protection against hip fracture in the elderly [5]. In some cases, nutrition may also influence the central nervous system processing time or contribute to the general feebleness that predisposes to falling. The implication here is that we should attempt to improve general nutrition in the elderly or, failing that, attend to coexisting nutritional problems at the time of fracture repair. 2. INTRINSIC BONE STRENGTH AND FRAGILITY Strength in bone, as in most engineering structures, is dependent upon its mass density, upon the three-dimensional arrangement of its material in space, and upon the intrinsic strength of its component material (particularly, in bone, as that strength is influenced over long periods of use by the accumulation of unrepaired fatigue damage). All three factors play some role in most low trauma fractures, and it is not possible to say which may be the most important in any given case. Nevertheless, most of the investigative effort in this regard since the 1970s has been devoted to the measurement of bone mass and density, and hence much of what we know about bone strength in living individuals comes from our observation of this facet of the bone strength triad. There is, in fact, a general consensus that decreased bone mass produces a decrease in bone strength. However there clearly are other fragility factors as well, although there is less of a consensus as to how large a role they play [6]. Data of Ross et al. [7] show that prior spine fracture signifies the presence of fragility independent of,

671

CHAPTER 27 Nutrition and Risk for Osteoporosis

and at least as important, as the fragility due to low bone density. Similarly, Hui and colleagues [8] showed that the fracture risk gradient for age, holding density constant, was greater than the risk gradient for density itself. These effects, independent of bone mass, may be partly explained by structural and qualitative defects in bone. For example, individuals with compression fractures of the vertebrae have been found to have excessive loss of horizontal, cross-bracing trabeculae in their cancellous bone [9,10], whereas other individuals with the same overall degree of bone loss, but with the bracing trabeculae maintained intact, are less apt to fracture. This may be the basis for the predictive value of prior spine fracture [7]. It appears that women, particularly, are more prone to loss of horizontal trabeculae than men, and this fact is probably also the explanation for the 6:1 to 8:1 female:male sex differential in vertebral osteoporosis. Data of Eventov et al. [11] indicate the probable importance of repair of fatigue damage. Faulkner et al. [12] and Glüer et al. [13] have called attention to a probable role of geometric factors at the hip, specifically to hip axis length, and Gilsanz and associates [14] to the importance of vertebral body size.1 In summary, evidence from several quarters makes it clear that bony fragility has bases other than reduced bone density. Nevertheless, as noted elsewhere in this volume, fracture risk rises by a factor in the range of 1.5 – 2.5  for every drop in bone mass/density of 1 SD. Whatever the role of nonmass factors, it is an inescapable fact that most elderly individuals have bone mass values that are more than 2 SD below the young adult mean; hence they all can be said to be at a considerably increased risk for fragility fracture. Why some older persons do fracture and others do not appears to be explainable by a combination of random chance, differences in falling patterns, and the structural differences just described. Nutrition enters into this portion of the fracture domain predominantly through its influence on bone mass (or density). Because nonmass factors also influence bone strength, nutritional inadequacies can never explain more than a part of the problem, and nutritional interventions can never completely eliminate fragility fractures. It may be, also, that trace nutrients such as certain of the vitamins (e.g., C, D, and K) or minerals such as manganese, copper, and zinc (see Section VIII) directly influence the remodeling process and/or the character of the remodeled bone, and hence affect bone strength through their impact on the repair of inevitable fatigue damage. However, little is known about these possibilities in the adult skeleton. Hence, in most of what follows, the emphasis will be on the nutritional factors that influence bone mass. 1

Other things being equal, a long hip axis increases hip fracture risk, and a small cross-sectional area for vertebral bodies increases spine fracture risk.

3. BONE MASS/DENSITY Bone mass and density are themselves influenced by many factors. Holding body weight constant, the three most important — or at least the three most commonly found to be limiting in industrialized nations — are physical activity, gonadal hormones, and nutrition. In adults of industrialized nations the nutrients most critical for bone health are calcium and vitamin D. Calcium intake, specifically, may be inadequate for the straightforward reason that it is low; however, even when statistically “normal,” it may still be inadequate because of subnormal absorption [16] or greater than normal excretory loss [17,18]. Other nutrients are also essential for building a healthy skeleton, but, except for calcium, their effects are usually seen most clearly during growth. (Once built, the skeleton tends to be relatively insulated from many subsequent nutritional deficiencies.) (see Section II) In addition, a number of other factors also influence bone mass, such as smoking, alcohol abuse, and various drugs used to treat a variety of medical illnesses, as well as those illnesses themselves. The effects of each of these factors are largely independent. In other words, altering any one of them will not substitute for, or compensate for, adverse effects of the others. Thus, a high calcium intake will not prevent the loss of bone that occurs immediately following menopause in women or castration in men. Similarly, physical activity will not compensate for an inadequate calcium intake. Neither will a high calcium intake offset the effects of alcohol abuse or smoking. Much of the apparent confusion in the bone field since the mid-1970s could have been avoided if we had better understood that these factors, while interactive, are substantially independent. Finally, although much of the following discussion will focus on calcium, it is necessary to stress what should perhaps go without saying, that calcium is a nutrient, not a drug, and hence its beneficial effects will be confined to individuals whose intake of calcium is insufficient. Also, calcium is not an isolated nutrient; it occurs in foods along with other nutrients, and it has been shown that diets low in calcium tend also to be nutritionally poor in other respects as well [19,20]. Thus, while it is necessary to deal with nutrients one at a time in an analysis such as this, the disorders in our patients are likely to be more complex.

II. PROBLEMS IN THE INVESTIGATION OF NUTRITIONAL EFFECTS ON BONE There are significant problems for both observational and experimental approaches to the elucidation of nutrient effects on the skeleton, and failure to recognize or overcome them has led both to seemingly contradictory results

672

ROBERT P. HEANEY

among various studies and to substantial confusion about the role of nutrition in bone health. Some of these problems are nutrition specific; others are inherent in bone biology.

A. Nutrition-Specific Problems 1. ESTIMATION OF NUTRIENT INTAKE Two nutrients with clearly established effects on bone are vitamin D and calcium. For both there are substantial difficulties in estimating intake [21,22] (see also Chapter 64). Vitamin D is found naturally in very few foods (mostly fish oils and to a limited extent, egg yolks).2 For primitive humans, solar exposure would have been the principal source of vitamin D, as is still the case in rural cultures and in the young of even many urbanized societies. Vitamin D is added as a fortificant to fluid and dry milk in the United States and Canada (but not to most other dairy foods).3 Serum 25(OH)D levels have long been recognized as the best available indicator of vitamin D status. Even so, their significance is not fully clear [26]. Furthermore, because such levels are affected by season, no single value in any given individual adequately captures his or her year-round average. Such measurements are also sufficiently costly and invasive so as to be precluded in most epidemiological studies involving large numbers of subjects. Finally, while vitamins D2 and D3 have heretofor been considered equivalent in potency (and both measured and used as a fortificant interchangeably), Vieth [27] showed that D2 exhibits only 60% the potency of D3 in humans. Although a nutrient in the strict sense, calcium also presents serious difficulties to the investigator who would at-

2 One could argue that vitamin D should not be considered a nutrient at all because it is not a constituent of most foods and is naturally synthesized in abundance in our skin given adequate solar exposure. However, because it was lumped accidentally with the other vitamins (nutrients in the strict sense) in the early days of development of nutritional science, that is where we treat it and think about it today. That is not just historical curiosity. Nutritionists in the past have often held that one can get all the nutrients one needs from a balanced, varied diet and have been slow to embrace the notions of engineered foods and food fortification. That clearly is a misguided approach to an accidental nutrient such as vitamin D. As the species moved out of equatorial regions, humans did not hesitate to develop warm clothing and shelter as protection from a cold environment. For vitamin D, at least, the same approach would seem to be indicated now that the biochemistry involved is understood. Industrialized nations of the high latitudes have done that for infants and children for the past 60 years, and rickets is rare today in those countries, but vitamin D insufficiency is still common among adults, particularly among the elderly (see Section VI). 3 Unfortunately, quality control has been poor in the past [23 – 25], i.e., the level of fortification can be highly variable (ranging from near toxic levels to absent altogether in many skim milk samples, despite what may appear on the label). For both reasons, it has been extremely difficult to assess effective vitamin D intake by any sort of questionnaire.

tempt to estimate its intake. Food calcium content often varies widely from published food table values — sometimes by a factor of 2– 3, reflecting variations in soil mineral content and plant tissue hydration (among other factors). Even commercial milk exhibits 10 – 20 percent variability from dairy to dairy or state to state. In a chemical analysis of foods consumed in a series of metabolic balance studies, Charles [28] found that less than 70% of the actual variability in intake among a group of subjects was reflected in the calculated intakes derived from food table values for the foods consumed, despite the fact that the precise quantities of every food eaten were known with high accuracy. Outside of the metabolic ward environment, and particularly in epidemiological studies, there is the added uncertainty of portion size estimation and food item recall. Another problem is presented by large differences in bioavailability. The calcium of kale or collard greens is highly available [29], whereas that of spinach is nearly totally unavailable [30]. Thus actual intake and effective intake can differ substantially. Finally, there will be broad daily and seasonal variation in intake patterns. In this regard, Heaney and colleagues [31] showed, in a large series of 7-day consecutive diet records, that any random day picked out of the total record captured only 12.6% of the interday variance, and that the error of the estimate of the 7-day average from one of its days was 178 mg (which means that the 95% confidence interval covers a range of more than 700 mg!). The difficulty of estimating effective calcium intake is compounded by two further problems. First is the use of calcium salts as excipients or “inert” ingredients in many medications, or as nonnutritive additives to various bulk foods. In both cases their calcium content goes unrecognized and often unacknowledged on the product label. Second is the increased use of explicit calcium supplements since 1982. This should not, of itself, create a problem for estimating calcium intake. However, many tablets in the past exhibited highly variable pharmaceutical formulations [32,33] and hence unpredictable absorbability. Excipient calcium will not often produce major errors in intake estimates unless food source intakes are low (in which case undocumented medication calcium can easily account for half the actual calcium intake); nevertheless, Heaney et al. [31] reported several cases in which such unrecognized calcium contributed more than 1000 mg/day to the intake. In any event, both causes can lead to serious misclassification of individual intakes in observational or epidemiological studies, and therefore they will bias toward the null any investigation dependent on intake estimates. An illustration of the effect of this bias is found in a meta-analysis by Heaney [34] of 28 studies in late postmenopausal women published between 1988 and 1992. Twenty-three of these 28 studies reported a positive effect of calcium intake on bone mass, bone loss, or fracture. However, when they were subdivided according to whether the investigators

673

CHAPTER 27 Nutrition and Risk for Osteoporosis

controlled the calcium intake directly or relied on estimates of intake derived from questionnaires and food records, it turned out that all of the 12 studies in which investigators controlled the intake had demonstrated a significant calcium benefit, whereas all of the inconclusive studies had been those in which intake had been merely estimated. The difference is explainable by errors in intake estimates in the questionnaire-based studies. 2. MAGNITUDE OF NUTRIENT – NUTRIENT INTERACTIONS It is a commonplace of nutritional science that nutrients interact, thereby altering one another’s requirements.4 It is to be expected, therefore, that the nutrients important for bone health would also exhibit this sort of interaction. What may not have been expected is the very considerable magnitude of those interactions with regard to critical bone nutrients. Coingested nutrients alter both obligatory renal loss and intestinal absorption of calcium and phosphorus. While effects on absorption are comparatively modest, effects on obligatory loss can alter the minimum daily requirement for calcium very substantially. (These effects are covered in more detail later in this chapter.) For our purposes here, it is sufficient only to note that other nutrients, ingested within the normal range of human intakes, so alter the ability to maintain calcium equilibrium as to produce a fourfold difference between the lowest and the highest values for the minimum requirement. This is a quite extraordinary range and is virtually without parallel among other nutrients. It is for this reason that it is usually misleading to make comparisons among populations that may differ not only in calcium intake, but in intakes of protein and sodium particularly, as well as in the proportion of animal and vegetable food sources in the customary diet. It is likely that much of the seeming differences in the relationship of calcium to bone status across populations [35] can be attributed to differences in minimum requirement related to nutrient – nutrient interactions and, much of the apparent confusion surrounding this topic, to failure to give adequate consideration to the influence of these interactions.

B. Bone-Specific Problems 1. THE BONE REMODELING TRANSIENT The bone remodeling transient is dealt with in greater depth in Chapter 64, as well as elsewhere [36,37]. It is important to mention it briefly in this context because whenever bone remodeling is altered by an intervention (nutritional in this context), the changes in calcium balance or 4

Recommended dietary allowances (RDAs) are designed, in theory, to be generous enough to accommodate this food-related variability in requirement.

bone mass that follow will, for a period of 6 – 12 months, reflect not the effects (if any) of the intervention on steadystate bone balance, but shrinkage or expansion of the bone remodeling space caused by asynchrony of the changes produced in bone formation and resorption. This is a particular problem for calcium, as calcium alters endogenous parathyroid hormone (PTH) production, and PTH is the principal determinant of the amount of global skeletal remodeling. However, any other nutrient (such as vitamin D or phosphorus), which also alters PTH production (whether directly or indirectly), may produce qualitatively similar effects. Thus the classical nutritional stratagem of measuring balance in individuals on differing intake levels for periods of up to a few weeks, then giving them a short rest period, and then trying yet another intake for a few more weeks (and so forth) will not work for bone or its measurable surrogates. Unfortunately, there are no easy alternatives. Balance for nutrients that are bulk bone constituents can be assessed only under steady-state conditions, and for calcium that means either studying persons on their habitual intakes or deferring study for 6 – 12 months after altering the intake of a given nutrient. Both options severely limit what the investigator can do to test various hypotheses involving nutrition and bone status. 2. ISOLATION OF BULK BONE FROM CURRENT NUTRITIONAL INFLUENCES Bone is very much a living tissue, with its cells responding both to systemic influences and to strain patterns within the bony structure. Nevertheless, the mechanical properties of bone reside exclusively in the intercellular, nonliving, two-phase composite of fibrous protein and mineral. With the exception of use-related, accumulating fatigue damage, the inherent mechanical properties of this material are largely (though not entirely) determined at the time a unit of bone is formed. The entire skeleton is turned over at a rate of only 8 – 10% per year (and some regions much more slowly). Because only currently forming bone will be affected by current conditions, nutritional stresses have predictably small effects on current bone strength. The bulk of bone is, in effect, isolated from the systemic and environmental influences that can rapidly produce outspoken effects in soft tissues. This is not to say that there are no effects on bone. It is possible that bone cells, damaged by current nutritional problems, may die or otherwise fail in one or another of their monitoring functions. However, the effects of that failure may become evident only years in the future, and they are, accordingly, extremely difficult to study. 3. SLOW RESPONSE TIME OF BONE A corollary of the slow turnover of bone tissue is that bone mass changes relatively slowly in response to nutritional influences, either positive or negative. A gain or loss of bone amounting to at most 1 – 2% per year is

674 typically all that many interventions can produce in adults. Continued over many years, such a rate of change can have profound effects on skeletal strength, but it is a change that is hard to detect by absorptiometric methods in short-term investigations and is essentially impossible to detect reliably in individuals. While the gain or loss associated with a nutritional intervention may be real enough, its presence is dwarfed by the relatively huge mass of preexisting bone, and its detection tends to be swamped by the inevitable noise of measurement. Balance studies can sensitively detect much smaller changes (because the background bone mass is not reflected in the balance value), but they are subject to the problem of the remodeling transient discussed earlier, and sufficient time must be allowed for the system to come into equilibrium if they are to be useful. Serum and urine biomarkers (see Chapter 60) can sensitively signal qualitative changes in bone remodeling processes, but they are not sufficiently quantitative to tell us the size of any change in bone balance that may have been produced by an intervention. 4. LIFE PHASE SPECIFICITY OF BONY RESPONSE As will be developed in more detail later, the skeleton is the body’s reserve of the nutrient calcium. It is the largest reserve of all the nutrients, and one that has acquired an unrelated function in its own right, the mechanical support of our bodies (failure of which is the reason for this volume). Whereas bone strength is clearly an inverse function of bone density (see Section II), and any decrease in bone density must have mechanical consequences, nevertheless reserves, of their nature, are designed to be called upon, and it should not be surprising to find that there may be physiological circumstances in which the reserve will reduce some of its store of mineral, not always because the diet is insufficient to offset excretory and dermal losses, but precisely because the physiological situation demands it or because the body senses that some of the reserve is no longer needed. Lactation may be one such situation and menopause another. In any event, nutritional interventions should be expected to produce qualitatively different effects when they are deployed under such differing physiological circumstances. 5. COMMENT This discussion of investigative problems is, of necessity, brief. Its purpose has been to highlight the inherent difficulty involved in investigating problems at the interface of nutrition and bone status. Failure of bone biologists to recognize nutritional measurement problems and failure of nutritionists to reckon adequately with the complexities of bone biology will lead (and has led) to badly designed, inconclusive, or misleading investigations. This is a problem not only for investigators, but for those who attempt to make sense of what they report. It is not that easy alterna-

ROBERT P. HEANEY

tives are being overlooked. Rather, there are no easy alternatives. However, while the problems are difficult, they are not intractable.

III. THE NOTION OF A NUTRIENT REQUIREMENT Nutritional science was born about a century ago with the then revolutionary recognition that the absence of something could produce disease.5 Once nutritional deficiency was accepted as the cause of disease, the notion of a requirement centered on the intake needed to avoid the recognizable deficiency disease concerned. While the science of nutrition has advanced notably since its beginnings, particularly in understanding precisely what various nutrients do in the body, our definitions of a requirement are still often pegged to early 20th century ability to recognize and characterize disease. There is growing dissatisfaction with this disease-centered approach, and increasingly one reads that the field should redefine a requirement as the intake needed to produce optimal health. However, the main problem with the traditional approach to a requirement is not that it is negative (i.e., disease-centered) as that its definition of disease is primitive. It is centered on disorders that develop rapidly and have distinct clinical expression recognizable with the tools of 70 years ago. However, a deficiency that takes 10 years to develop or to make its presence evident is no less a deficiency than one which develops in 10 days.6 Vitamin K deficiency, for example, produces a bleeding disorder, which is the defining disease associated with the nutrient. Does absence of bleeding mean that vitamin K nutriture is adequate? We now recognize that vitamin K is necessary for -carboxylation of a large number of proteins in addition to the clotting factors, three of them involved in bone matrix (see Section VII). We also recognize that -carboxylation of these proteins can be very incomplete even when the clotting factors are normally carboxylated and that physiological vitamin K supplementation completely repairs this deficit. It is not known whether this undercarboxylation expresses itself as disease, but our ignorance in that regard does not guarantee that the absence of clotting disturbance means vitamin K sufficiency.

5

The prevailing notion at the time was that all disease was caused by infections or intoxications, i.e., by some noxious influence from outside the organism. 6 The first clearly identified deficiency disease, beriberi, typically develops from 30 – 90 days after the onset of thiamin deprivation and responds to treatment with roughly equal speed. One can speculate whether nutritional science would have developed at all if its disease states had typically had long latency periods.

CHAPTER 27 Nutrition and Risk for Osteoporosis

In what follows in this chapter, the author defines a requirement as the intake that ensures full expression of known functions of the nutrient concerned and presumes that any substantial deviation from full physiological expression is harmful until proved safe (see also Chapter 64).

IV. THE NATURAL INTAKE OF CALCIUM AND VITAMIN D It has only recently been recognized that both calcium and vitamin D were present in superabundance in the environment in which the human species evolved. It seems likely that over the millennia of evolution, human physiology developed mechanisms to protect the organism from getting too much of these important nutrients. In contrast, contemporary adult humans, living in industrialized nations at higher latitudes, have intakes of these nutrients often only a small fraction of what their primitive ancestors experienced, and our physiologies are, therefore, maladapted to what our environment currently provides. Vitamin D is produced normally in the skin by a photochemical reaction in which ultraviolet light from the sun changes 7-dehydrocholesterol into previtamin D. As the human species evolved in equatorial east Africa, with ample sunlight year round, two mechanisms coevolved that prevented accumulating an excess of vitamin D. One was skin pigmentation, which slowed the photochemical reaction, and the other was the fact that continued solar radiation degrades previtamin D to inert products before it is taken up into the circulating blood. As a result, vitamin D accumulation in the skin plateaus after a few minutes of sun exposure, with the time varying with skin pigmentation. Circulating levels of 25(OH) vitamin D (25OHD) under early conditions can be estimated from values observed in dark-skinned, outdoor laborers at tropical latitudes, which have been reported to be in the range of 150 nmol/liter [27] or  4 – 6 times what is typically measured in city dwellers at midlatitudes. As humans moved farther and farther north (away from the equator), and needed all the ultraviolet they could get, skin pigmentation became lighter and lighter. Still, in latitudes such as that of Boston and farther north, the sun is so low in the sky in winter that effectively none of the responsible ultraviolet rays gets through the atmosphere, even on a sunny day [38]. As a result, vitamin D tends to be a scarce nutrient at high latitudes, and without careful attention to maintaining adequacy, varying degrees of vitamin D insufficiency will be common. Just 75 years ago, more than 80% of the children in England showed evidences of rickets [39]. Thanks to nearly universal vitamin D prophylaxis in children, rickets is now a relatively rare disorder. Calcium was also present in abundance in the environment in which the human species evolved. The plant foods

675 eaten by hunter – gatherers provided a calcium intake that, adjusted for differences in body size, would have been in the range of 2000 – 4000 mg/day for 60 to 70 kg-adults [40,41]. (Contrast that figure with the median value for women 20 and over in the United States in the NHANES III study: in the range of 600 mg/day [42].) Sources available to our ancestors included a very large number of greens, tubers, roots, nuts, and berries, many of them with very high calcium nutrient densities [41]. Moreover, invertebrate and reptilian sources of animal protein typically have calcium-to-calorie ratios six-fold higher than fish or mammalian meats [43]. In contrast, cultivated cereal plants, legumes, and fruits — the plant foods modern humans mainly consume – exhibit augmented levels of carbohydrate and/or fat without a proportionate increase in minerals and vitamins; thus they almost always have lower calcium densities than do their wild cousins. The agricultural/pastoral revolution, which occurred from roughly 3000 to 10,000 years ago in various parts of the world, made it possible to feed vastly more people than the hunter – gatherer mode permitted. This was partly because of the increased energy content of polypoid cereal mutants (which occur spontaneously, but which, because of their greater seed weight, need human intervention for their efficient propagation). At the same time the agricultural revolution produced striking changes in micronutrient intake, generally for the worse. We see this reflected in modern times in the nutritional deficiencies that result when hunter – gatherers, such as the !Kung San people, are forced by restriction of their range to take up farming [44]. The effect on the calcium density of the diet is depicted in Fig. 2. Diets of hunter – gatherers would have been in the range of 70 – 90 mg Ca/100 kCal (somewhat higher if invertebrate protein sources featured prominently in the diet). Those who then domesticated animals and lived mainly off their milk (as do the Masai, the Ariaal, and, indeed, all pastoralist societies today) would have had a shift in diet to

FIGURE 2 Changes in calcium concentration of the diet associated with the agricultural/pastoral revolution. Copyright Robert P. Heaney (1995). Reproduced with permission.

676 somewhere in the neighborhood of 200 mg Ca/100 kCal.7 In contrast, those who settled on the land and subsisted mainly on cereal crops and legumes would, at least from these food sources alone, have had diets with calcium densities under 20 mg/100 kCal. While vegetable greens would have helped when available, calcium intakes based solely on cereals and legumes would probably not have been sufficient to sustain bone health. However, there are numerous, well-attested examples of peoples living in stable equilibrium with their environments who have developed nonfood ways of augmenting the meager calcium intake provided by a diet based on seed foods. The addition of lime to corn meal by indigenous peoples in Central America is one wellknown example. Less well known is the practice of pregnant southeast Asian women of drinking a liquid produced by soaking bones in vinegar [45]. Andean Indians have been reported to add both a particular plant ash and a heattreated rock powder to their cereal gruel [46]. All of these practices represent a kind of conscious addition of a substance that, de facto, augmented the calcium intake of a cereal-based diet. It seems likely that unconscious additions of the same sort were nearly universal among neolithic farming communities. The archaeological record has preserved numerous examples of stone mortars used for dehulling cereal grains and stone querns for grinding the seeds into flour [47]. In the fertile crescent at least, limestone would have been the most readily available and the most workable stone, and the hours of hand grinding of dehulled cereal grains would inevitably have added substantial calcium (as calcium carbonate) to the resulting flour.8 As technology advanced, and millstones were made of harder and harder rock (usually Si- and Al-based rather than Ca-based minerals), aggregate calcium intakes would have declined toward the lower line depicted in Fig. 2. Thus, the low calcium intakes that we take for granted today are relatively quite late arrivals on the human diet scene. Because hominid and early human diets were very rich in calcium, the human intestine either failed to develop effective absorptive transport mechanisms or actually de-

7 The Masai typically have calcium intakes in the range of 6000 – 7000 mg Ca/day, essentially all from milk. 8 The addition of calcium carbonate to bread flour in the United Kingdom during and after World War II, and in Japan in the postwar years, and the recent fortification of certain breads in the United States with calcium sulfate are but modern, conscious instances of what must have been an unwitting ancient practice. 9 It is instructive to compare the body’s handling of calcium with that of sodium, which was an environmentally scarce nutrient during hominid evolution. In contrast with calcium, essentially 100% of dietary sodium is absorbed, and dermal and renal sodium losses can be reduced to near zero.

ROBERT P. HEANEY

veloped an absorptive barrier to protect against too much calcium. Nor did mechanisms to conserve absorbed calcium develop. (Presumably, there would be little need to conserve in the face of environmental surfeit.) Humans typically absorb only about 25 – 35% of the calcium in contemporary diets [48] and put about 150 mg/day back into the gut in the digestive secretions [49]. Thus, net absorption of a dietary calcium increment is usually in the range of 10 – 15% even during growth when skeletal need is greatest [50]. Additionally, dermal losses are completely unregulated and renal conservation is limited as well. These are precisely the physiological patterns one would expect with an environmentally abundant nutrient.9 This is the background to why, despite a high standard of living and the potential to nourish ourselves at a level never previously achieved in the history of the race, civilized diets tend to be deficient in precisely these two critical nutrients, calcium and vitamin D.

V. CALCIUM A. The Skeleton as a Nutrient Reserve Throughout the course of vertebrate evolution, bone developed several times and has served many functions, such as dermal armor and internal stiffening [51]. Evidence from a variety of lines suggests that the most primitive function of the skeleton is actually to buffer the internal milieu for several essential minerals, notably calcium and phosphorus [52]. In some species, phosphorus would have been the critical element; in others, calcium. For both nutrients, the skeleton serves both as a source and as a sink, i.e., as a reserve to offset shortages and, to a limited extent, as a place for safely storing surpluses. We see this reserve feature of skeletal function expressed in diverse ways. For example, there is the long established fact that laboratory animals such as cats, rats, and dogs will reduce bone mass as needed to maintain near constancy of calcium levels in the extracellular fluid [53 – 55]. This activity is mediated by PTH and involves actual bone destruction, not leaching of calcium from bone. When calcium-deprived animals are parathyroidectomized, bone is spared, but severe hypocalcemia develops [56]. More physiologically, perhaps, deer temporarily increase bone resorption each year to meet the calcium and phosphorus demands of annual antler formation (which exceed the nutrient supply of late winter and spring foliage) [57]. Finally, we see the opposite side of the same function expressed in the now well-established fact that augmented calcium intake will slow or reduce age-related bone loss in humans (see later).10

677

CHAPTER 27 Nutrition and Risk for Osteoporosis

FIGURE 3 Schematic illustration of the relationship between health status and body depletion of a nutrient. (A) The depletion of body stores of a typical nutrient after placing the organism on a deficient intake. (B) The reserve (the pattern exhibited by calcium) is very large relative to the metabolic pool, but health, as reflected in skeletal strength, declines steadily as the reserve itself is depleted. Copyright Robert P. Heaney (1995). Reproduced with permission. While retaining this primitive, reserve function, bone in the higher, terrestrial vertebrates acquired a second role, namely internal stiffening and rigidity — what is today the most apparent feature of the skeleton. As such, calcium (or phosphorus) is the only nutrient with a reserve that possesses such a secondary function (with the possible exception of the thermal insulation provided by energy reserves). Figure. 3 presents this situation schematically and contrasts calcium with the bulk of other nutrients, depicting what happens for each when intake is curtailed. For typical nutrients, the reserve is first depleted, without a detectable impact on the health or functioning of the organism. Then, after the reserve is exhausted and the metabolic pool begins to be depleted, clinical disease expresses itself. For some nutrients (e.g., vitamin A or energy), the reserve can be quite large, and the latent period may last many months. However, for others (e.g., water-soluble vitamins), the reserve may be very small and detectable dysfunction develops soon after intake drops. With calcium, the reserve is vast relative to the cellular and extracellular metabolic pools of calcium. As a result, dietary insufficiency virtually never impairs biochemical functions that are dependent on calcium, at least in ways we can now recognize. However, because bone strength is a function of bone mass, it follows inexorably that any decrease whatsoever in the size of the calcium reserve — any decrease in bone mass — will produce a corresponding decrease in bone strength. We literally walk about on our calcium reserve. It is

10 That this reduction in bone loss is not simply a pharmacologic effect of calcium, as suggested by Kanis and Passmore [58], is indicated by two facts: (1) the effects are greater in those with low baseline calcium intakes and (2) even the augmented intakes employed in these studies are usually well below what primitive humans would have ingested, i.e., they are in the nutritional range, not the pharmacological range, of calcium intakes.

this unique relationship that is both the basis for the linkage of calcium nutriture with bone status and the explanation why reduction in the size of the reserve is the sole defining characteristic of the major human calcium deficiency syndrome.

B. Defining the Requirement for Calcium Unlike other nutrients, the requirement for calcium relates solely to this secondary function, i.e., to the size of the calcium reserve, in other words, to total skeletal and regional bone mass. However, unlike energy, which can be stored as fat without practical limit, the size of the calcium reserve is limited, even in the face of dietary surfeit, by genetic and mechanical factors (see later). As a result, calcium functions as a threshold nutrient, much as does iron. This means that, below some critical value, the effect (bone mass for calcium or hemoglobin mass for iron) will be limited by available supplies, while above that value, i.e., the “threshold,” no further benefit will accrue from additional intake. This biphasic relationship is depicted schematically in Fig. 4, in which the intake – effect relationship is depicted first schematically (A) and then (B) as exemplified by data derived from a growing animal model. In Fig. 4B, the effect of the nutrient is expressed directly as the amount of bone calcium an animal is able to accumulate from any given intake. However, if “effect” is broadened to mean “any change whatsoever,” then the diagram fits all life stages, even when bone may be undergoing some degree of involution. This generalized form of the threshold diagram is set forth in Fig. 5, which shows schematically what the intake/retention curves look like during growth, maturity, and involution. In brief, the plateau occurs at a positive value during growth, at zero retention in the mature individual, and sometimes at

678

ROBERT P. HEANEY

FIGURE 4

Threshold behavior of calcium intake. (A) Theoretical relationship of bone accumulation to intake. Below a certain value (the threshold), bone accumulation is a linear function of intake (the ascending line); in other words, the amount of bone that can be accumulated is limited by the amount of calcium ingested. Above the threshold (the horizontal line), bone accumulation is limited by other factors and is no longer related to changes in calcium intake. (B) Actual data from two experiments in growing rats showing how bone accumulation does, in fact, exhibit a threshold pattern. Redrawn from data in Forbes et al. [59].) Copyright Robert P. Heaney 1992. Reproduced with permission.

a negative value in the elderly. (Available evidence suggests that there are probably several involutional curves, with the plateau during involution at a negative value in the first 3 – 5 years after menopause, at zero for the next 10 years, and then at increasingly negative values with advancing age.) In Fig. 5B, which shows only a composite involutional curve, there are two points identified: one below (B) and one above (A) the threshold. At A, calcium retention is negative for reasons intrinsic to the skeleton, whereas at

FIGURE 5

B, involutional effects are compounded by inadequate intake, which makes the balance more negative than it needs to be. Point B (or below) is probably where most older adults in the industrialized nations would be situated today. The goal of calcium nutrition in this life stage is to move them to point A, thereby making certain that insufficient calcium intake is not aggravating any underlying bone loss. The functional indicator of nutritional adequacy for such a threshold nutrient is termed “maximal retention” and can

(A) Schematic calcium intake and retention curves for three life stages. Retention is greater than zero during growth, zero at maturity, and may be negative during involution. Asterisks represent minimum daily requirements. (B) The involution curve only. Points B designates an intake below the maximal calcium retention threshold, whereas point A designates an intake above the threshold. Copyright Robert P. Heaney (1998). Reproduced with permission.

679

CHAPTER 27 Nutrition and Risk for Osteoporosis

be located in Figs. 4A and 5A at the asterisks above the curves. The intake corresponding to this point represents the minimum daily requirement. Calcium retention in this sense is “maximal” only in that further intake of calcium will produce no further retention. (This is in contrast to treatment with hormones or drugs, which can sometimes produce further calcium retention.) This approach was used by the Food and Nutrition Board of the National Academy of Sciences for the first time in its development of recommended intakes for calcium in 1997 [60]. There has been much uncertainty and confusion in recent years about what the threshold intake may be for various ages and physiological states. With the 1994 Consensus Development Conference on Optimal Calcium Intake [61] and the report of the Panel on Calcium and Related Nutrients [60], the bulk of that confusion has been resolved. Evidence for the intakes recommended by the consensus panel is summarized both in the conference and panel reports and in recent reviews of the relationship of nutrition and osteoporosis [34,62] and will be summarized only briefly in ensuing sections of this chapter. It is worth noting, however, that the recommendations of the consensus conference, while expressed in quantitative terms, were basically qualitative: Contemporary calcium intakes in North America and northern Europe, by both men and women, are too low for optimal bone health in Caucasian individuals. The most persuasive of the evidence leading to this conclusion came in the form of several randomized controlled trials showing both a reduction in age-related bone loss and a reduction in fractures following augmentation of prevailing calcium intakes [63 – 72]. However, randomized controlled trials, at least as performed to date, are not well suited to dose ranging (largely because of the problem of the remodeling transient, (see earlier discussion). Hence, while the panel was convinced that prevailing intakes were too low, their recommended levels in several cases involved ranges,

TABLE 1 Age

and were clearly prudential judgments, centered of necessity on intakes employed in the trials concerned. It is instructive, therefore, to review all the recent recommendations in concert with the evidence from balance studies (to be discussed further later). Table 1 sets forth these various recommendations. As can be seen, while the 1994 National Institutes of Health (NIH) recommendations are, for most ages, substantially higher than the 1989 RDAs, they are actually quite close both to values derivable both from available balance studies and to the 1997 recommendations of the Food and Nutrition Board.

C. The Requirement at Various Life Stages There have been in excess of 140 studies published relating calcium intake to bone status, summarized and reviewed elsewhere [62]. In over 40% of this total, the investigators controlled calcium intake, and essentially all of these studies showed that calcium intakes above the then prevailing RDAs conferred a bone benefit. Even among the observational studies, in which calcium intake was not investigator controlled and could only be estimated, about 80% were positive. There is, thus, an overwhelming mass of evidence establishing the importance both of calcium for bone and of ensuring intakes higher than prevailing levels or former recommendations. What cannot be determined easily from controlled trials (as has already been noted) is the precise location of the intake threshold, i.e., the point where bony retention is maximal and further intake confers no further bony benefit. The following sections focus on estimating this intake level by age and physiological state. 1. GROWTH The human skeleton at birth contains approximately 25 g calcium and, in adult women, 1000 – 1200 g. All of

Various Estimates of Calcium Requirements in Women 1989 RDAa

NIHb

1997 RDAc

Balanced

1–5

800

800

—

1100

6 – 10

800

800 – 1200

960

1100

11 – 24

1200

1200 – 1500

1560

1600

Pregnancy/lactation

1200

1200 – 1500

1200 – 1560

—

24 – 50/65

800

1000

1200

800 – 1000

65

800

1500

1440

1500 – 1700

a

From Ref. [73] Recommendations for women as proposed by the Consensus Development Conference on Optimal Calcium Intake [61]. c The so-called “adequate intakes” of the new DRI values, multiplied by a factor of 1.2  to convert them into RDA format [60]. d Estimates derived from published balance studies [50]. b

680 this difference must come in by way of the diet. Further, unlike other structural nutrients such as protein, the amount of calcium retained is always substantially less than the amount ingested. This is because, as already noted, absorption efficiency is relatively low even during growth and because calcium is lost daily through shed skin, nails, hair, and sweat, as well as in urine and nonreabsorbed digestive secretions. The gap between calcium intake and calcium retention is larger than is generally appreciated. In the adult with a modest but repairable skeletal deficiency, only about 4 – 8% of ingested calcium is retained. While retention efficiency is generally higher during growth, even when bone accumulation is most rapid, less than half of the intake is actually retained — ranging from a high of 40% in term infants to 20% in young adults [50]. Even premature infants, with a permeable gut membrane and a relatively huge mineralization demand, exhibit net absorption of less than 60% [74], and dermal and urinary losses mean that they retain even less than that figure. This inefficient retention is not so much because the ability to build bone is limited but because, as noted elsewhere, human physiology is optimized to prevent calcium intoxication, not to cope with chronic shortage. Aside from the obvious fact that one cannot store what one does not ingest, how does suboptimal calcium intake limit bone mass accumulation? Except in unusual circumstances, it is not through limiting bone deposition. In most animal experiments, as well as human observations, a low calcium intake probably does not limit the growth in bone length or breadth. This is because bone-forming sites do not “see” the diet. They are exposed only to circulating levels of calcium, phosphorus, and the calciotrophic hormones, and even in the face of frank dietary calcium restriction, blood calcium levels change very little. An inadequate calcium intake does, however, result in a bone with a thinner cortex and fewer, thinner trabeculae. This comes about through modulation of the balance between the normal, ongoing processes of bone formation and bone resorption (see Chapters 12 and 15). To understand how dietary intake interacts with the modeling process, it is necessary to recall that bone reshapes itself continuously during growth. In growing long bones, new bone is deposited at the periosteal surface of the midshaft, at the endosteal surface of the submetaphyseal shaft, and at the growth plate. At the same time, bone is resorbed at the endosteal-trabecular surface and on the outer surface of the metaphyseal funnel. This produces concentric expansion of both external shaft diameter and medullary cavity diameter. This dual process reshapes bones so that they conform to growth in body size. The difference between the amount deposited and the amount resorbed is equal to the net bone gain (or loss). When ingested calcium is less than optimal, the endosteal-trabecular resorptive process increases, and the balance between formation and resorption, normally positive during growth, falls toward zero. This occurs because PTH

ROBERT P. HEANEY

augments bone resorption at the endosteal-trabecular surface in order to sustain the level of ionized calcium in the extracellular fluid. When the demands of mineralization at the periosteum and growth plates exceed the amount of calcium absorbed from the diet and released from growthrelated bone modeling, more PTH is secreted and resorption increases still further, until the balance becomes zero or even negative. If calcium is the only limiting nutrient, it is usually considered that growth in size continues normally, but that a limited quantity of mineral now has to be redistributed over an ever larger volume. Usually children’s diets in Western nations are not so calcium deprived as to preclude entirely any increase in bone mass, but occasional instances of severe restriction have been reported. Then, high levels of PTH drive phosphorus levels in the extracellular fluid so low that mineralization is inhibited and a rachitic type of lesion develops [75,76], even though the vitamin D status may be normal. In such circumstances, bone growth does slow.11 Short of such extreme situations, the principal perceptible effect of inadequate calcium intake during growth in developed countries is a skeleton of low mass — normally shaped and sized, but containing a smaller than normal amount of bone tissue. Having said that, it must be noted that there are at least two studies suggesting that augmented calcium intake may influence bone size as well as bone mass [77,78]. The first [77] is difficult to interpret because the supplementation included extra protein, phosphorus, and other key nutrients as well as calcium. However, the second [78], with better matching of other nutrient intakes, also showed a small effect of extra calcium on both bone mass and stature. Most of the periosteal expansion and growth in length, and much of the endosteal expansion during growth, are determined genetically and mechanically. Studies in twins have shown that a large fraction of the variability in peak bone mass is accounted for by the genetic program [79] (see also Chapter 26). However, as already noted, endosteal expansion can be increased in the face of insufficient calcium intake beyond what would be dictated by the genetic program. Thus, while an abundant diet will not produce more bone than the genetic program calls for, a deficient diet must restrict what a person is able to accumulate. Optimal peak bone mass for any given individual can be defined as a skeleton in which the balance between the concentric expansions of growth is solely determined by the individual’s genetic program and is not reduced by an exogenous shortage of calcium. Correspondingly, optimal calcium intake can be defined operationally as the intake that permits this full expression of the genetic program. 11 This issue is complicated by the fact that diets so severely deficient in calcium are commonly inadequate on other grounds as well, e.g., protein and calories. Consequently, the growth-stunting undoubtedly has multiple causes.

681

CHAPTER 27 Nutrition and Risk for Osteoporosis

As just noted, net bone accumulation will be greater as calcium intake increases, but only to the point where endosteal-trabecular resorption is due solely to the genetic program governing growth and is not being driven by body needs for calcium. Above that level, as seen in Figs. 4 and 5, further increases in calcium intake will produce no further bony accumulation. The intake required to achieve the full genetic program, and thus to assure peak bone mass, is the intake that corresponds to the beginning of the plateau region in Figs. 4 and 5. This value will be different for different stages of growth, in part because growth rates are not constant and also because, as body size increases, obligatory calcium losses through skin and excreta increase as well. The best approach to determine this value in humans lies, as with the laboratory animal, in testing various intakes for their influence on calcium retention, i.e., finding the plateau and locating its threshold. (In healthy individuals, calcium retention amounts to the same thing as bone tissue accumulation, as calcium is normally stored in the body only in the form of bone.) Over the past 75 years many such studies have been performed. When these reports are combined, it is possible to make out the pattern of plateau behavior found in laboratory animals and, from the aggregated data, to estimate the intake values that correspond to the threshold [50]. Figure 6 represents one example of the relationship between intake and retention, combining the results of many published studies of calcium balance. It is derived from a subset of the adolescents whose balances were assembled by Matkovic [80]. More recently, Jackman et al. [81] studied a series of adolescent girls (each at two intakes, varying from subject to subject) and reported an intake threshold at very nearly the same level as found by Matkovic and Heaney in their meta-

FIGURE 6 The relationship of calcium intake, on the horizontal axis, to calcium retention (balance), on the vertical axis, for a subset of the adolescents described by Matkovic and Heaney [50]. Note that despite the “noisiness” that is inevitable in measurements of balance in humans, there is clear evidence of an intake plateau, as observed in the animal experiments of Fig. 4. Note also that, for this age, the threshold of the plateau occurs at about 1500 mg Ca/day. Copyright Robert P. Heaney (1992). Reproduced with permission.

TABLE 2

Critical Values for Calcium Intake and Retention Efficiency by Agea

Intake threshold (mg/day)

Subthreshold retention efficiencyb

0–1

1090

0.407

13

2–8

1390

0.238

183

9 – 17

1480

0.356

320

18 – 30

957

0.200

732

Age

X-Axis intercept (mg/day)

a

Derived from analyses of published balance studies during growth [50]. Slope of the relationship of retention on intake.

b

analysis. Both approaches clearly show the plateau type of behavior that both animal studies and theoretical considerations predict. They also confirm that, at intakes less than the plateau threshold, daily storage is less than optimal, i.e., accumulation of bone is being limited by intake. Any such limiting intake must be considered inadequate. Table 2 and Fig. 7 summarize some of the relevant calculations flowing from the type of analysis of aggregated balance data exemplified in Fig. 6 for various stages of growth [50]. First are the threshold intake values, as judged from the assembled balance studies. In some instances these values are slightly higher than both the NIH figures and current DRIs (Table 1), but in general the various recent estimates are quite close to one another. Aside from the usefulness of these threshold values themselves, an especially notable feature of data in Table 2 is that even after linear growth has ceased (i.e., in young adults), calcium retention still occurs if the intake is high enough to support it. In other words, bony consolidation can continue after growth in stature has ceased. For this reason, calcium intake in young adults needs to be sufficient not only to maintain skeletal equilibrium, but to support this continuing augmentation of bone mass.

FIGURE 7 Regression lines for subthreshold regions of intake – balance relationships in infants, children, adolescents, and young adults from the data of Matkovic and Heaney [50]. Copyright Robert P. Heaney (1992). Reproduced with permission.

682 Figure 7 shows, for the four age groups delineated in Table 2, the best-fit regression lines for intake regions below the age-specific thresholds. This analysis of data reveals a number of interesting features. First, although the slopes are qualitatively similar, there are nevertheless some important quantitative differences among the age groups. The ability to make use of an increment in calcium intake is greater in infancy and adolescence (i.e., the slope is larger) when skeletal growth is most rapid and lower in childhood and the young adult years, when growth is slower, as would be expected. Perhaps of even greater interest is the rightward displacement of the regression lines in Fig. 6 with advancing age. This phenomenon, reflected in the values in Table 2 for the X-axis intercept, reflects the effect of age on obligatory loss. Whereas a zero balance is obviously not healthy for a growing organism, these zero-balance intake values are useful in that they reflect how much calcium an individual must ingest just to stay even (i.e., not to lose bone) at the ages concerned. As Fig. 7 shows, infants can reduce calcium loss to nearly zero on zero calcium intake. For older children and young adults, it takes larger and larger calcium intakes to sustain even zero balance. Most of this effect is accounted for by a rise in urine calcium with age. It is probably body size that is forcing the higher obligatory requirement since, in a multiple regression model of these data, body size continues to have an effect even after controlling for age [50]. At least 10 randomized controlled trials of calcium supplementation in children and adolescents have been published [e.g., 66,67,78,81 – 83], together with several longitudinal observational studies in young adults [e.g., 84]. All of the controlled trials were positive, as were three-fourths of the observational studies. As mentioned earlier, the bone remodeling transient contributes to the measured difference in these controlled trials. The relative size of its contribution remains uncertain; nevertheless, simulation of the remodeling transient indicates that the gain reported in these studies is greater than can be explained by that mechanism alone. In all studies, supplemental calcium elevated the children’s intakes above the 1989 RDA. The finding of greater bone gain in the supplemented children than in the control group underscores the inadequacy of the earlier RDA values. In other words, they indicate that the RDAs for 1989 and earlier lie on the ascending portion of the threshold curves of Figs. 4 and 5 rather than on the plateau. Hence, these studies reinforce the higher requirement values set forth in Table 2. In a third trial [80], one group of pubescent girls received approximately the 1989 RDA while the other group was held to a calcium intake of 450 mg per day (far less than the RDA but unfortunately not uncommon for girls of that age). As predicted, growth in stature was the same in both groups, but bone mass failed to increase

ROBERT P. HEANEY

in the low intake group, while it did in the high intake group. Matkovic et al. [85] had shown previously that intakes as low as 450 mg per day in adolescent girls did not support positive calcium balance, mainly because, despite intense skeletal demand at that life stage, the urinary conservation of calcium remained inefficient. While this third study does not specifically address the issue of what the intake requirement should be during adolescence, it does clearly document the deleterious effects of intakes well below the RDA. All of these intervention studies, as already noted, produce a remodeling transient. None was designed to evaluate steady-state changes, and hence their positive findings cannot be translated directly into a specific intake recommendation. However, the 4-year longitudinal study of young adults by Recker and colleagues [84] involved no alteration of calcium intake and hence avoided the problem of the transient. This study showed prospectively that bone augmentation continues well into the third decade. Bone mass gains in their subjects ranged from 0.5% per year for the forearm to 1.25% per year for total body bone mineral. The single most important correlate of the rate of bone accumulation in this Caucasian, middle-American group of women was calcium intake. The rate of accumulation was inversely proportional to age, with the best estimate of the age at which the rate reached zero being approximately 29 – 30 years. This suggests that the window of opportunity to achieve the full genetic program appears to remain at least partly open until about age 30. It is not certain that the gains observed in this study were as great as might have been possible, as subjects were studied on their habitual calcium intakes, which were below the threshold values of Table 2. Nevertheless, these observations of Recker and colleagues [84] provided qualitative confirmation of the analyses for this age group derived from the balance studies (Table 2), which showed that young adults were able to retain calcium if intakes were at or above the threshold. They also illustrate well how the data from observational studies can complement the data from controlled trials (see Chapter 64). The importance of ensuring full realization of the genetic potential for skeletal development lies in the fact that bone mass seems to track throughout life. Newton-John and Morgan [86] first noted this phenomenon over 30 years ago in cross-sectional data, and Matkovic et al. [87] showed very clearly in their study of two Croatian populations that those who had higher mass at age 30 remained higher than the others out to age 75, even though both groups were apparently losing bone with age. The same phenomenon has been seen in shorter term, longitudinal studies [88 – 90], both across puberty and in the postmenopausal years. Dertina et al. [88] have gone so far as to suggest that those most at risk for late life osteoporosis can be detected before puberty.

CHAPTER 27 Nutrition and Risk for Osteoporosis

2. MATURITY Once peak bone mass has been achieved, the principal force acting on the skeleton is no longer the impetus of growth, but the mechanical loads imposed in ordinary, everyday usage (see also Chapter 1). Skeletal structures, like all engineering structures, deform slightly under load. The skeleton senses that degree of deformation and attempts to adjust its mass (by controlling the balance between bone resorption and bone formation) so that this deformation remains on the order of 0.1 – 0.15% in any given dimension (see also Chapter 18). If a bone is loaded so heavily that it consistently bends more than that amount, then the balance between local formation and resorption is adjusted to favor formation, thus making that region stiffer. Conversely, if a bony segment is little used, and its bending is less than that critical amount, the skeleton senses that it has an excess of bone in the region concerned and adjusts the remodeling balance to remove some of the apparent surplus. This reference level of bending is one of the fascinating physiological constants of nature. Across the vertebrates, for all species and all bones studied to date, bone mass is regulated such that any given bone deforms by about that critical 0.1 – 0.15% in ordinary use. This reference level of bending is termed a set point, and the bone remodeling apparatus operates to minimize local deviations from this critical value. The cellular basis for the set point and the precise nature of the apparatus that detects departures from it remain unknown12; however, there is suggestive evidence that localizes this sensing apparatus to the network of osteocytes embedded in bone. For several years it has seemed likely that one of the principal determinants of the set point of this mass-regulating system is the level of gonadal hormones. Circumstantial evidence in support of this connection includes the facts that estrogen receptors in bone are concentrated in osteocytes [91] and that true bone density rises sharply at puberty [92] and declines by about the same amount at menopause (whether natural or artificial) [93]. These life-phase changes are what one would expect if estrogen influences the set point.13 While adjustments in mass around the set point presume an adequate calcium intake, it turns out that prevailing intakes tend to be closer to adequate during the ages 25 to 50 in women, as estrogen improves the efficiency of intestinal calcium absorption [48,93] and of renal calcium conservation [94,95]. Thus estrogen not only increases the reference level of bone density but it helps the body access and retain 12 Bone is not unique in this regard: the molecular basis for the set point in most biological feedback systems is not known. 13 The same level of bending sensed as tolerable in an estrogen-deprived state lies above the reference level when estrogen is present (and the set point is lower). The bone-remodeling apparatus responds by adding bone to reduce the size of the difference from the reference level of bending.

683 the mineral necessary to augment bone to that higher level. For this reason, except for the special circumstances of pregnancy and lactation (discussed later), the years from 25 to 50 are a time in life when a woman’s skeletal calcium need is at its lowest. She is no longer storing calcium, and her absorption and retention are operating at their adult peak efficiency. In a meta-analysis of 33 studies performed in adults between 18 and 50 years of age, Welten et al. [96] found a positive association between calcium intake and bone mass in this age group and noted that it seemed prudent to maintain an intake of 1500 mg/day during this life period. This is a higher figure than either the NIH recommendations or the 1997 DRIs of Table 1. Using balance methods in estrogen-replete women ingesting their habitual calcium intakes, Heaney et al. [94], found a mean intake for zero balance of slightly under 1000 mg/day, and Nordin [97], also using balance methods, arrived at a figure slightly above 800 mg/day. In a small prospective study of bone mass in premenopausal women, Recker et al. [98] found no detectable bone loss over a 2-year period on an estimated mean calcium intake of 651 mg. Studying women in their fifth decade, Baran et al. [99] found bone loss in a control group receiving 892 mg Ca/day. Loss occurred only from year two to year three, and not during the first 2 years of observation, and it is not clear from that paper whether the apparent loss in year three was related to the loss of sampling units that occurred between years two and three or whether there was actual loss in those who remained in the study. Requirement estimates based on balance studies make no provision for sweat loss of calcium, as by design, balance methods usually eliminate vigorous exercise. The importance of sweat loss has been highlighted by a study in male college athletes showing sweat calcium losses of over 200 mg in a single vigorous workout session [100]. Moreover, there was a perceptible loss of bone mineral density across a playing season that was preventable by adding calcium supplements to the athletes’ already good diets. It is likely that the lost bone would have been regained after the playing season was over (so long as the diet was adequate), i.e., that the skeleton was acting in its capacity as a calcium reserve during the playing season. Nevertheless, the study makes clear how large and important sweat losses can be. Undoubtedly such losses contribute to the low bone mass described in women athletes who often have less than generous calcium intakes. In conclusion, the bulk of the available evidence suggests that it is important to maintain an intake of 1000 – 1500 mg/day during the mature years. Moreover, there are other health reasons for maintaining a high calcium intake during this period [101], even if bone health can be supported adequately by an intake in the range of 800 – 1000 mg/day.

684 3. PREGNANCY AND LACTATION Pregnancy and lactation are circumstances in which the mother must provide for maintenance of her own skeleton as well as for construction of her child’s. Specifically, during the 9 months of pregnancy, she provides the fetus with 25 – 30 g calcium, and in her milk during the ensuing 9 months of lactation, another 50 – 75 g. This aggregate is in the range of 7 – 10% of her own total body calcium and would, presumably, produce a corresponding decrease in bone mass if she were not able to obtain some or all of the required quantities from ingested calcium. It has always seemed intuitively attractive, therefore, to recommend an increased calcium intake during these physiologically demanding life stages [73] (see also Table 1). Moreover, given the relatively low calcium intakes of modern industrialized societies, one might have expected that a history of multiparity and extended lactation would be associated with lower bone mass and an increased risk of osteoporosis. In general, however, epidemiological studies have found, if anything, the contrary. Most studies report a positive association between parity and bone mass/density [102 – 107], although occasional reports of negative associations can be found [108]. Much of the positive association turns out to be due to increased ponderosity, and after correcting for weight, the positive correlations tend to become statistically nonsignificant. Nevertheless, most are still on the positive side, and there is little or no hint in the available evidence that the calcium drain of pregnancy and lactation adversely affects the maternal skeleton. Bone remodeling accelerates in pregnancy [109 – 114], and maternal intestinal calcium absorption efficiency increases to the highest level since early infancy. Both changes begin well before significant fetal skeletal accumulation of calcium [109,115]. Both humans and rats show anticipatory storage of skeletal minerals prior to the onset of fetal skeletal mineralization [107,109], and Heaney and Skillman [109] estimated, from balance studies in pregnant women studied on their habitual calcium intakes, that cumulative calcium balance at term exceeded fetal needs and that the mother, therefore, went into lactation with a skeletal surplus. Brommage and Baxter [110] reported data consistent with a skeletal surplus in rats at delivery, and ultrasound methods suggest that the same occurs in pregnant mares [111]. However, this will not be possible if calcium intake is very low. Barger-Lux amd co-workers. [116] reported bone loss across pregnancy in young women with dietary calcium to protein ratios averaging 6.6 mg/g. During lactation the majority of reports indicate that some degree of bone loss regularly occurs [101,102,105 – 108,112,113,117 – 20], particularly in presumably reactive bony sites such as the centers of the vertebral bodies and the ultradistal radius [117,118]. However, this loss appears to reverse after weaning and may, therefore, represent to some extent a negative remodeling transient like what

ROBERT P. HEANEY

occurs in deer at the time of antler formation [57]. Immediately following delivery, absorption efficiency falls to or toward nonpregnant, nonlactating levels and remains at this relatively low level throughout lactation, despite the continuing drain of lactational calcium loss [109,113,121]; however, urinary calcium falls at the same time and remains low throughout lactation and for several months postweaning [112], while bone remodeling remains elevated [112,113]. This is a physiological situation conducive to replacement of lost bone.14 Lactating rats lose nearly one-third of their skeleton during milk production [119]. This loss doubles if the animals are placed on low calcium diets, but it does not diminish when the normally high calcium diet of a rat is increased as much as threefold [119]. It is likely that this bone loss represents an anticipatory phenomenon, i.e., rather than the calcium being drawn out of bone by the drain of lactation, the bone pumps calcium into the circulation for milk production. This is suggested by the reduced PTH levels during lactation [122], by the usually reported failure of increased calcium intake to reduce the loss, and by the high serum phosphorus levels during lactation.15 How this outpouring of skeletal mineral for the benefit of lactation occurs is less clear, although it is certainly plausible that the hypoestrin state of lactation would, like menopause or athletic amenorrhea, shift the bone set point and result in some downward reduction in bone density (thereby, effectively, releasing stored calcium and phosphorus). While Kalkwarf et al. [123] found no effect of calcium supplementation, a few reports suggest that even the modest reductions in bone mass normally found during human lactation can be reduced or eliminated by extra calcium [117,118]. The relatively slow growth of human infants (in comparison, for example, with rats) imposes a lower lactational burden on a human mother, and some of the differences between species may be attributable to quite significant differences in lactational demands for mineral. Given the concordance of balance data and epidemiological evidence, it seems likely either that adaptive mechanisms are usually sufficient to accommodate the calcium demands of pregnancy and lactation or that postweaning adjustments compensate for whatever bone may have been lost. As it turns out, physiological evidence indicates that both occur. Compensatory physiological adjustments surrounding pregnancy and lactation are more vigorous than at other life stages, and the current consensus is that a high calcium 14

The importance of urinary loss for balance is discussed in Section

V,D. 15 In this latter respect, phosphate is as necessary for milk production as calcium, and high serum P levels serve that important purpose. The contrary causal flow, i.e., lactation pulling calcium out of bone, would work against the lactational need for phosphorus, as PTH, mediating the response to all calcium needs, lowers serum phosphorus.

685

CHAPTER 27 Nutrition and Risk for Osteoporosis

intake makes less long term difference to a woman’s skeleton at this life stage than at most other times in her life. In summarizing the available literature, the panel responsible for the 1997 DRIs noted that there was no evidence on which to base a recommendation for a higher calcium intake during pregnancy and lactation than that considered optimal for other women of the same age. They did add, however, that the situation with adolescent pregnancy was problematic and inadequately studied, and that perhaps some increment above the adolescent recommendation in such individuals might be prudent. 4. MENOPAUSE It has been noted already that estrogen seems to adjust the bending set point of bone. Accordingly, whenever women lose ovarian hormones, either naturally at menopause or earlier as a result of anorexia nervosa or athletic amenorrhea, the skeleton seems to sense that it has more bone than it needs and hence allows resorption to carry away more bone than formation replaces. (Precisely the same change occurs when men lose testosterone for any reason.) This is equivalent to raising the bone-bending set point, as described previously. While varying somewhat from site to site across the skeleton, the downward adjustment in bone mass due to gonadal hormone lack amounts to approximately 10 – 15% of the bone a woman had in the lumbar spine and  6% at the total hip prior to menopause [93]. The importance of this phenomenon in a discussion of nutrient effects is to help distinguish menopausal bone loss from nutrient deficiency loss and to stress that menopausal loss, which is due to an absence of gonadal hormones, not to nutrient deficiency, cannot be substantially influenced by diet. Almost all of the published studies of calcium supplementation within 5 years following menopause failed to prevent bone loss [63,124,125]. Even Elders et al., [125], who employed a calcium intake in excess of 3000 mg/day, succeeded only in slowing menopausal loss, not in preventing it. However, Dutch women tend to be calcium replete because of high national dairy product consumption, and other studies have shown effects of calcium supplementation in the early menopausal years that are intermediate between placebo and estrogen [72,124,125]. It is likely that in any group of early menopausal women, there are some whose calcium intake is so inadequate that they are losing bone for two reasons (estrogen lack plus calcium insufficiency). Important as menopausal bone loss is, it is only a onetime, downward adjustment and, if nutrition is adequate, the loss continues for only a few years, after which the skeleton comes into a new steady state (although at a 5 – 15% lower bone mass). It is in this context that the importance of a high peak skeletal mass becomes apparent. One standard deviation for lumbar spine bone mineral content in normal women is about 10 – 15% of the young adult mean, and for total body bone mineral, about 12%. Hence a

woman at or above 1 SD above the mean can sustain the 15% menopausal loss and still end up with as much bone as the average woman has before menopause. In contrast, a woman at or under 1 SD below the young adult mean premenopausally drops to 2 SD below the mean as she crosses menopause and is therefore, by WHO criteria [127], already osteopenic and verging on frankly osteoporotic. As noted, the menopausal bone mass adjustment amounts to a loss at the spine and hip of 10 – 15% and  6%, respectively [128]. Hip bone change, both immediately before and after this menopausal downward adjustment, averages about  0.5% year, while, except for the menopausal loss, the spine curve is flat. However, this is so only so long as calcium intake is adequate. In this regard, it is important to recall the nonskeletal effects of estrogen described earlier, i.e., improvement of intestinal absorption and renal conservation [48,94,95]. Because of these effects, an estrogen- deficient woman has a higher calcium requirement, and unless she raises her calcium intake after menopause, she will continue to lose bone after the estrogen-dependent quantum has been lost, even if the same diet would have been adequate to maintain her skeleton before menopause. In other words, early in the menopausal period, her bone loss is mainly (or entirely) because of estrogen withdrawal, whereas later it is because of inadequate calcium intake. Figure 8 assembles, schematically, the set of factors contributing to bone loss in the postmenopausal period. Figure 8 shows both the self-limiting character of the loss due to estrogen deficiency and the usually slower, but continuing loss due to nutritional deficiency, when present. Unlike the estrogen-related loss, which mostly plays itself out in 3 – 6 years, an ongoing calcium deficiency loss will continue to deplete the skeleton indefinitely for the remainder of a woman’s life, i.e., unless the calcium intake is raised to a level sufficient to stop it. Furthermore, because both

FIGURE 8

Partition of age-related bone loss in a typical postmenopausal woman with an inadequate calcium intake. Based on a model described in detail elsewhere [93]. Copyright Robert P. Heaney (1990). Reproduced with permission.

686 absorption efficiency [48] and calcium intake [42] decline with age, the degree of calcium shortfall typically worsens with age. Thus it is important for a woman to increase her calcium intake after menopause, even though, for the first few years, doing so will not prevent estrogen-withdrawal bone loss. Both the 1984 NIH consensus conference on osteoporosis [129] and the 1994 Consensus Conference on Optimal Calcium Intake [61] recommended intakes of 1500 mg/day for estrogen-deprived postmenopausal women. It may be that the optimal intake is somewhat higher still (see later), but median intakes in the United States for women of this age are in the range of 500 – 600 mg/day [42], and if the bulk of them could be raised even to 1500 mg/day, the impact on skeletal health would be considerable. 5. SENESCENCE There is general agreement that bone is lost with aging. Early cross-sectional data suggested that spine loss began as early as age 30 – 35, but, except for the hip, longitudinal studies have not borne that out for most skeletal regions (e.g., spine, forearm) [98]. Significant loss probably does not begin until sometime in the sixth or seventh decade.16 This age-related bone loss occurs in both sexes, regardless of gonadal hormone levels. However, it is obscured at the commonly measured spine site in the years immediately following menopause in women by the substantially larger effect of estrogen withdrawal (see Fig. 8). It probably occurs, however, even in estrogen-treated women, at about the same rate as in men. This rate varies by skeletal region and is generally reported to be on the order of 0.5 – 1.0% per year by the seventh decade and accelerates with advancing age. Age-related loss involves both cortical and trabecular bone and can be due to several causes. These include disuse atrophy consequent upon reduced physical activity, an entropic kind of loss due to the accumulation of random remodeling errors, which, of their nature, tend to be irreversible,17 reduction in androgenic steroid levels, and finally nutritional deficiency loss. These types of bone loss are summarized in Fig. 8. While nutrient deficiency is clearly only a part of the problem, nevertheless it is common. Intestinal calcium absorption efficiency declines with age [48], at the same time as nutrient intake itself generally declines [42]; the result is that the diet of aging individuals becomes doubly inadequate. This inadequacy is clearly expressed, for example, in the rate of bone loss reported by Chapuy et al. [64] in the untreated control group of their large randomized trial of 16 For certain bony regions, density (BMD) may begin to decline earlier [130], but in most such instances there is a countervailing periosteal expansion, such that total regional bone mass remains constant and bone strength is, if anything, greater (see Chapter 19). 17 Examples include overlarge Haversian cavities, fenestrated trabecular plates, and severed trabecular spicules, which, once disconnected, become unloaded and hence are subject to rapid resorption [131].

ROBERT P. HEANEY

calcium and vitamin D supplementation. These women, with an average age of 84 and with calcium intakes that averaged 514 mg/day, were losing bone from the femur at rates of slightly more than 3% per year. That there was a causal connection between intake and bone loss is demonstrated by the fact that the loss was completely obliterated with calcium (and vitamin D) supplementation. It is in this age group that the most dramatic and persuasive evidence for the importance of a high calcium intake has been produced in recent years. This is primarily because most fragility fractures rise in frequency with age, and hence the opportunity to see a fracture benefit (if one exists) is greater then.18 Chapuy et al. [64] showed a reduction in hip fracture risk of 43% by 18 months after starting supplementation with calcium and vitamin D and a 32% reduction in other extremity fractures. In another study in elderly women, Chevalley et al. [68], resolved the question left unanswered in the study of Chapuy and co-workers (whether it was the calcium or the vitamin D that was responsible for the effect) by giving vitamin D to both controls and treated subjects, but calcium only to the treated group. They, too, found a reduction in femoral bone loss and in fracture incidence (vertebral in this case) in the calcium supplemented women. In a 4-year, randomized controlled trial in elderly women (mean age 73), Recker et al. [69] showed that a calcium supplement reduced both age-related bone loss and incident vertebral fractures. Their subjects had all received a multivitamin supplement containing 400 IU of vitamin D; hence most or all of the effect in the calciumsupplemented group can be attributed to the calcium alone. The studies of Chevalley et al. [68] and Recker et al. [69] should not be interpreted to mean that vitamin D is unimportant in this age group. It is likely that intakes of both calcium and vitamin D are commonly inadequate in the elderly (see later), and the high prevalence of combined deficiency has complicated study of the actual requirements of either nutrient in this age group. The importance of these studies lies in the fact that even after ensuring vitamin D repletion, there was still a calcium benefit, and hence presumptively calcium deficiency in this age group. Heikinheimo and associates [132] had shown earlier the converse in an elderly Finnish population. Vitamin D supplementation in this population (which tends to be calcium replete) significantly reduced all fractures, both in institutionalized and in free-living individuals. The calcium intake achieved in the Chapuy study was about 1700 mg/day; in the Chevalley study, 1400 mg/day, and in the Recker study, about 1600 mg/day. These values are in the range of the intake found by Heaney et al. 18 Reduction in bone loss is only presumptively beneficial. Until it can be shown that fracture incidence is reduced, bone mass effects are less persuasive, and despite the abundant theoretical underpinnings of why bone mass should be important, only fracture reduction is ultimately convincing.

687

CHAPTER 27 Nutrition and Risk for Osteoporosis

[94,133] to be the mean requirement for healthy estrogendeprived older women (1500 – 1700 mg/day). All these studies are, therefore, fully consistent with the more recent recommendations in the range of 1500 mg/day (Table 1). An important feature of these controlled trials in already elderly individuals was that bone mass was low in both treated and control groups at the start of the study, and while a significant difference in fracture rate was produced by calcium supplementation, even the supplemented groups would have to be considered as having an unacceptably high fracture rate. What these studies do not establish is how much lower the fracture rate might have been if a high calcium intake had been provided for the preceding 20 – 30 years of these women’s lives. Although not randomized trials, the studies of Matkovic et al. [4] and Holbrook et al. [134], strongly suggest that the effect may be larger than has been found with treatment started in the eighth and ninth decades of life. Both of these observational studies reported a hip fracture rate that was roughly 60% lower in elderly whose habitual calcium intakes had been high. While findings from observational studies such as these had not been considered persuasive in the absence of proof from controlled trials, the trials with fracture end points have now met that need. This is another instance of the point made in Chapter 64 that, when applied appropriately, data of observational studies and of controlled trials can complement one another in useful ways. Additional reinforcement comes from McKane et al. [135], who studied the effect of a large calcium supplement on PTH secretory dynamics in elderly women. In brief, a mean calcium intake of 2413 mg/day lowered PTH levels 40% to the young normal range, and normalized the abnormal PTH secretory dynamics typical of the elderly female. They concluded that the combination of declining oral calcium intake, deteriorating vitamin D status, reduced calcium absorption, and impaired renal conservation of calcium in the elderly led to parathyroid gland hyperactivity and increased bone resorption. Together, the aggregate of available studies underscores the importance of achieving at least the 1400 to 1500-mg target figure of the new recommendations for the elderly. At the same time it must be stressed, once again, that osteoporosis is a multifactorial condition, and that removing one of these factors (i.e., ensuring an adequate calcium intake) cannot be expected to eradicate all osteoporotic fractures.

Table 3

Food Factors and the Calcium Requirement Negative effects

Positive effects

Fiber

Fiber

Absorption Phytate

Food

Oxalate

Lactose

Caffeine

Carbohydrates

Fat

Lysine

Phosphorus

Fat

Excretion Protein

Phosphorus

Sodium

Alkaline ash

Chloride Acid ash [Aluminum hydroxide]

absorption [136 – 138] and typically exert relatively minor effects, whereas sodium and protein influence the urinary excretion of calcium [136,137] and can be of much greater significance for the calcium economy. Phosphorus and fat are sometimes mentioned in connection with calcium absorption, but their effect in humans seems minor to nonexistent. The basis for the importance of nutrients acting on absorption and excretion is illustrated in Fig. 9, which partitions the variance in calcium balance observed in 560 balances in healthy middle-aged women studied in the author’s laboratory. As Fig. 9 shows, only 11% of the variance in balance among these women is explained by differences in their calcium intakes, and absorption efficiency explains only another 15%. In contrast, urinary losses explain slightly more than 50%.19 The dominance in Fig. 9 of renal excretion would be trivial in primary bone-losing syndromes, but it is particularly noteworthy that it appears to be operative in conditions of health, for it means that

D. Nutritional Factors That Influence the Requirement

FIGURE 9 Partition of variance in calcium balance in normal women among the input – output processes involved in the calculation of balance. Copyright Robert P. Heaney (1994). Reproduced with permission.

There are several nutritional factors that influence or have been proposed to influence the calcium requirement (Table 3). The principal interacting nutrients are sodium, protein, caffeine, and fiber. Fiber and caffeine influence calcium

19 The author has already remarked upon the importance of urinary calcium loss in the context of the declining retention efficiency with age in growing children and has noted that the drop in urine calcium during lactation and postweaning helps compensate for lactational demands.

688 obligatory losses through the kidney pull calcium out of the skeleton. This is a concept for which we are particularly indebted to the work of Nordin and associates [139,140]. 1. INFLUENCES ON INTESTINAL ABSORPTION OF CALCIUM a. Fiber The effect of fiber is variable and generally small. In acute, single-meal absorption tests, many kinds of fiber have no influence at all on absorption, such as the fiber in green, leafy vegetables [30,34]. Moreover, fibers of the class termed nondigestible oligosaccharides (NDO), rather than interfering with absorption, have been shown in rats to increase both mucosal mass and calcium absorption [141], and there are reports in humans suggesting a similar effect, at least on absorption [142,143] (which is why fiber is listed in both columns of Table 3). The current theory is that volatile fatty acids produced in fermentation of the NDOs by colonic flora evoke gut hormone responses that regulate mucosal mass (thereby serving to match the metabolic cost of replacing the mucosa every 5 days to the level of food intake). In contrast, the fiber in wheat bran reduces the absorption of coingested calcium in single-meal tests, although except for extremes of fiber intake [144], the antiabsorptive effect is generally relatively small. Often lumped together with fiber are associated food constituents such as phytate and oxalate, both of which can reduce the availability of any calcium contained in the same food. For example, for equal ingested loads, the calcium of beans is only about half as available as the calcium of milk [145], whereas the calcium of spinach and rhubarb is nearly totally unavailable [30,146]. For spinach and rhubarb, the inhibition is mostly due to oxalate. For common beans, phytate is responsible for about half the interference, and oxalate, the other half. The effects of phytate and oxalate are highly variable. There is a sufficient quantity of both antiabsorbers in beans to complex all the calcium also present, and yet their combined absorptive interference is only half what might have been predicted. With the exception of bran, these interferences generally operate only on calcium contained in the same food. This is because the antiabsorber is usually already fully complexed with calcium in the ingested food. Thus, spinach does not typically interfere with the absorption of coingested milk calcium. b. Caffeine Often considered to have a deleterious effect on the calcium economy, caffeine actually has the smallest effect of the known interacting nutrients. A single cup of brewed coffee causes deterioration in calcium balance of 3 mg [137,138,147], mainly by reducing the absorption of calcium [137]. The effect is probably on active transport, although this is not known for certain. This effect is so small as to be more than adequately offset by a

ROBERT P. HEANEY

tablespoon or two of milk [137,147], and café au lait or caffe latte produces a substantial net calcium gain, despite their caffeine content. c. Fat Fat has also sometimes been presumed to reduce calcium absorption by a similar mechanism, i.e., formation of calcium soaps with unesterified fatty acids released in the chyme by intestinal lipases. However, in healthy adult humans no appreciable effect of fat intake on calcium absorption has been found. This is at least partly explained, as with phosphorus (see later), by the fact that the normal small intestine absorbs fat much more avidly than it does calcium. At intakes in the range of recommended levels, the feces contain a considerable stoichiometric excess of calcium relative to fatty acids. 2. INFLUENCES ON RENAL CONSERVATION OF CALCIUM a. Protein and Sodium As noted earlier, the effects of protein and of sodium are substantial [17,18,138,148]. Both nutrients increase urinary calcium loss across the full range of their own intakes, from very low to very high, so it is not a question of harmful effects of an excess of these nutrients. Sodium and calcium share the same transport system in the proximal tubule, and every 2300 mg sodium excreted by the kidney pulls 20 – 60 mg of calcium out with it. Every gram of protein metabolized in adults causes an increment in urine calcium loss of about 1 mg.20 This latter effect is probably due to excretion of the sulfate load produced in the metabolism of sulfur-containing amino acids (and is thus a kind of endogenous analog of the acid-rain problem). At low sodium and protein intakes, the minimum calcium requirement for skeletal maintenance for an adult female may be as little as 450 mg/day [139], whereas if her intake of both nutrients is high, she may require as much as 2000 mg/day to maintain calcium balance. A forceful illustration of the importance of sodium intake is provided by the report of Matkovic et al. [85] that urine calcium remains high in adolescent girls on calcium intakes too low to permit bone gain. The principal determinant of urinary calcium in such young women is sodium intake [149], not calcium intake. Differences in protein and sodium intake from one national group to another are part of the explanation of why studies in different countries have shown sometimes strikingly different calcium requirements [35]. At the same time, one usually finds a positive correlation between calcium intake and bone mass within the national range of intakes [150]. Hence, while sodium (and protein) intake

20 This protein effect would be predicted to be less during growth, particularly when growth is rapid, as in infancy. Then, much of the ingested protein is incorporated into tissue, whereas in adults, with no net tissue gain, protein catabolism matches protein intake.

689

CHAPTER 27 Nutrition and Risk for Osteoporosis

differences across cultures may obscure the calcium effect, they do not obliterate it. The acid/alkaline ash characteristic of the diet is also important, although the quantitative relationship of this diet feature to the calcium requirement is less completely developed. Nevertheless, it has clearly been shown that substitution of metabolizable anions (e.g., bicarbonate or acetate) for fixed anions (e.g., chloride) in various test diets will lower obligatory urinary calcium loss substantially [151,152]. This suggests that primarily vegetarian diets create a lower calcium requirement and provides a further explanation for the seemingly lower requirement in many nonindustrialized populations. However, it is not yet clear whether, within a population, vegetarians have higher bone mass values than omnivores [153], and such limited data as are available suggest, in fact, the contrary [154 – 156]. b. Phosphorus Phosphorus is commonly believed to reduce calcium absorption, but the evidence for that effect is scant to nonexistent, and there is much contrary evidence. Spencer and co-workers [157] showed no effect of even large increments in phosphate intake on overall calcium balance at low, normal, and high intakes of calcium. In adults, Ca:P ratios ranging from 0.2 to above 2.0 are without effect on calcium balance when studied under metabolic ward conditions and adjustments are made for calcium intake [138]. Still, phosphorus intake is not without effect on the calcium economy. It depresses urinary calcium loss and elevates digestive juice secretion of calcium by approximately equal amounts (which is why there is no net effect on balance [49,148]). While it is true that stoichiometric excesses of phosphate will tend to form complexes with calcium in the chyme, various calcium phosphate salts have been shown to exhibit absorbability similar to other calcium salts [158], and phosphate is, of course, a principal anion of the major food source of calcium (dairy products). In any case, phosphate itself is absorbed more readily than calcium (by a factor of at least 23), and at intakes of both nutrients in the range of their respective RDIs, absorption will leave a stoichiometric excess of calcium in the ileum, not the other way around. This explains the seeming paradox that high calcium intakes can block phosphate absorption (as in management of end-stage renal disease), whereas achievably high phosphate intakes have little or no effect on calcium absorption. c. Aluminum Although not in any sense a nutrient, aluminum, in the form of Al-containing antacids, also exerts significant effects on obligatory calcium loss in the urine [159]. By binding phosphate in the gut, these substances reduce phosphate absorption, lower integrated 24-h serum phosphate levels, and thereby elevate urinary calcium loss. This is the opposite of the more familiar hypocalciuric

effect of oral phosphate supplements. Therapeutic doses of Al-containing antacids can elevate urine calcium by 50 mg/day or more. 3. ENHANCERS OF CALCIUM ABSORPTION Relatively little work has been done on enhancers of calcium absorption. Lactose is said to improve absorption, but the effect may be confined to the rat. Human studies using various carbohydrates have generally shown some enhancement [160], but the effect may be confined to intestines damaged by disease or surgery, as it has been hard to find in healthy subjects [161]. Also, the effect of various carbohydrates may be nonspecific, due instead to alteration of the gastric emptying pattern associated with coingestion of other food constituents – what we have elsewhere characterized as the “meal effect” [162]. (That is the meaning of the entry “food” in Table 3.) Nevertheless, given the generally low absorbability of calcium, the prospect of finding substances that might improve calcium bioavailability has enticed many food processors. Various food fractions, such as casein phosphopeptide, derived from milk, have been found to improve calcium absorbability in certain experimental systems [163], although its effect in humans is probably small [164]. Likewise, certain amino acids, notably lysine, have been thought to enhance calcium absorption [165], but human evidence in their regard is sparse and inconsistent. Even fat might theoretically be viewed as an enhancer, as it is known to slow gastric emptying. However, we have been unable to find, using multiple regression methods, any effect of even large variations in fat intake on absorption fraction in our observational study of middle-aged women. 4. INTAKE VS INTERFERENCE For diets high in calcium, as would have been the case for our hunter – gatherer ancestors, high protein and possibly high sodium intakes could have been handled by the body without adverse effects. These nutrients create problems for the calcium economy of contemporary adult humans mainly because we typically have calcium intakes that are low relative to those of preagricultural humans. This is because at prevailing low intakes, compensatory adjustment mechanisms are already operating, and for many individuals, the capacity for further adaptation (e.g., increased absorption efficiency) is very limited. An increased demand for only 40 mg Ca/day would require a nearly 40% increase in intestinal absorption at intakes at the bottom quartile for North American and European women today, while the same demand can be met by an increase of only 1 – 2% in absorption efficiency at intakes such as prevailed during hominid evolution. The former is not possible, while the latter is easily accomplished. Thus, while there is some emphasis today among nutritionists on regulating the intake

690 of interfering nutrients, the real problem is not so much that the intakes of these other nutrients are high as that the calcium intake is too low to allow us to adjust to the inevitable nutrient – nutrient interactions that occur with any diet.

VI. VITAMIN D Vitamin D is discussed extensively in Chapter 9. This section focuses mainly on certain bone-related nutritional features of this multifacted substance. It has long been recognized that vitamin D is important for the absorption of calcium from the diet. Its role in that regard lies in facilitating active transport, probably by inducing the formation of a calcium-binding transport protein in intestinal mucosal cells. This function is particularly important for adaptation to low intakes. There is also, apparently, a second vitamin D-related absorption mechanism, transcaltachia [166], which is nongenomic in expression but nevertheless requires occupancy of the classical vitamin D receptor. Finally, absorption also occurs passively, probably mainly by way of paracellular diffusion. This route is not dependent on vitamin D and is not as well studied. The proportion of absorption by the three mechanisms varies with intake and is not well characterized in humans; at high calcium intakes (above 2000 mg/day), absorption fraction approaches that observed in anephric individuals (ca. 10 – 15% of intake). Under these circumstances it is likely that active transport contributes relatively little to the total absorbed load. Nevertheless, it is clear, at prevailing calcium intakes, that vitamin D status influences absorptive performance and that it thereby influences the minimum calcium requirement. A simple calculation suffices to establish the magnitude of this influence. Assume an intake of 1000 mg Ca/day. To that is added about 150 mg in the form of digestive secretions and sloughed off mucosa. If passive absorption is at a level of 12.5% of intake, net absorption would amount to 144 mg, leaving the individual in negative balance across the gut of 6 mg/day (and, of course, producing no calcium gain for the body to offset renal and dermal losses). If, however, vitamin D-mediated, active transport is operating, so that, for example, total absorption was 27.5% net absorption becomes 109 mg. The relationship of active transport to net absorption is shown graphically, for various intakes, in Fig. 10, which makes clear that meeting physiological demands for calcium would require very high calcium intakes in the absence of vitamin D. (That situation is depicted by the bottom line in Fig. 10, which is the net absorption contour for zero active absorption, as well as by the other lines depicting lower levels of active transport, reflecting, in turn, varying degrees of vitamin D insufficiency.) A principal storage form of the vitamin is 25-hydroxyvitamin D (25OHD), and its plasma level is generally regarded

ROBERT P. HEANEY

as the best clinical indicator of vitamin D status. Although usually considered to be about three orders of magnitude less potent than calcitriol in promoting active transport in animal receptor assays, there is growing evidence that it may possess physiological functions in its own right [167 –172]. In the only human dose –response studies performed to date, 25OHD was found to have a molar potency in the range of 1/125 to 1/400 that of 1,25(OH)2D3 [171 –173], not the 1/2000 figure usually considered to reflect relative 25OHD activity. Vitamin D status commonly deteriorates in the elderly, whose plasma 25OHD levels are generally lower than in young adults [174,175]. These elderly persons, without histological or biochemical evidence of osteomalacia, nevertheless exhibit high PTH levels, high serum alkaline phosphatase activity, and low absorptive performance, all of which move to or toward normal with physiological amounts of supplemental vitamin D [174 – 177]. The rate of age-related loss of bone has been found to be correlated inversely to dietary vitamin D [178]. Low dosage vitamin D supplementation of ostensibly healthy postmenopausal women significantly slows wintertime bone loss and reduces the annual parathyroid-mediated activation of the bone remodeling system that occurs in winter through late spring [175]. These changes all suggest relative vitamin D insufficiency. Low 25OHD levels in the elderly are partly due to decreased solar exposure, partly to decreased efficiency of skin vitamin D synthesis, and partly to a decreased intake of milk, the principal dietary source of the vitamin in North America. Moreover, the elderly exhibit other abnormalities of the vitamin D endocrine system that may further impair

FIGURE 10

Relationship of vitamin D-mediated, active calcium absorption, calcium intake, and net calcium gain across the gut. Each of the contours represents a different level of active absorption above a baseline passive absorption of 12.5%. (The values along each contour represent the sum total of passive and variable active absorption.) The horizontal and dashed lines indicate 0 and 200 mg/day net absorption, respectively. The former is the value at which the gut switches from a net excretory to a net absorptive mode, and the latter is the value needed to offset typical urinary and dermal losses in mature adults. Copyright Robert P. Heaney (1999). Reproduced with permission.

691

CHAPTER 27 Nutrition and Risk for Osteoporosis

their ability to adapt to reduced calcium intake. These include decreased responsiveness of the renal 1--hydroxylase to parathyroid hormone [179] and possibly, also, decreased mucosal responsiveness to calcitriol [180] (although available data do not permit distinguishing a decrease in mucosal responsiveness from a simple decrease in mucosal mass). For all these reasons there is a growing body of opinion that the requirement for vitamin D rises with age [27,176, 181 – 184], and a body of data that strongly suggests that relative vitamin D deficiency plays a role in several components of the osteoporosis syndrome. Perhaps most persuasive of all is the finding by Heikinheimo et al. [132] in a randomized, controlled trial of substantial reduction in all fractures in an elderly Finnish population given a single injection of 150,000 – 300,000 IU vitamin D each fall.21 Lips et al. [185], however, found no benefit from an additional 400 IU of vitamin D in a Dutch population. The foregoing studies (as well as others) lead inexorably to the conclusion that vitamin D insufficiency is prevalent in the middle-aged and elderly of northern Europe and North America. Moreover, in virtually none of the studies showing a benefit of supplemental vitamin D was frank osteomalacia a significant feature of the problem. Hence, as discussed earlier, this criterion for true vitamin D deficiency may well be much too strict to be nutritionally useful today. How the vitamin D requirement should to be defined is another matter. Holick [184] has presented data showing that it takes an intake of at least 600 IU/day, from all sources, to sustain serum 25(OH)D levels, and the doses of vitamin D used in the studies summarized earlier also suggest that an intake in the range of 500 – 800 IU/day is required for full expression of the known effects of vitamin D in adults. Vieth [27] presented evidence that the requirement may be higher still. What is not clear is how much of the effect of vitamin D in studies such as the fracture prevention trial of Heikinheimo and co-workers [132] is due to facilitating gut adaptation to marginal calcium intakes and how much may represent an extraintestinal effect of the vitamin in its own right, e.g., on muscle tone or coordination. Calcitriol receptors are widely distributed in many tissues, and calcitriol enhances PTH-mediated bone resorption and exhibits autocrine action in cell differentiation and in the immune response. Furthermore, calcitriol elicits a prompt and sizable increase in the osteoblast synthesis of osteocalcin [186]. Additionally, elevating serum 25OHD levels in the elderly improves the often incomplete -carboxylation of osteocalcin (see Section VII). Nevertheless, patients with vitamin D-dependent rickets type II, who lack functional calcitriol

21 This dose amounts to a daily average exposure of 410 – 820 IU and can hence be considered a physiological intake.

receptors, show essentially complete remission of most of their skeletal pathophysiology with intravenous calcium infusions alone [187]. Furthermore, while subtle impairment of immune function can be demonstrated in nutritional vitamin D deficiency, the defects appear to be sufficiently mild to be of little or no clinical consequence in most individuals. Hence the issue of the extraintestinal importance of vitamin D remains unclear.

VII. VITAMIN K The chemistry and physiology of vitamin K have been reviewed extensively elsewhere [186,188 – 191]. In brief, vitamin K is necessary for the -carboxylation of glutamic acid residues in a large number of proteins. Most familiar are those related to coagulation, in which seven vitamin K-dependent proteins are involved in one way or another. The -carboxyglutamic acid residues in the peptide chain bind calcium, either free or on the surface layers of crystals, and have been thought to function in various ways, including catalysis of the coagulation cascade, inhibiting mineralization (as in urine) [192], and serving as osteoclast chemotactic signals [193]. Vitamin K deficiency classically produces bleeding disorders, but the liver, where the clotting factors are produced, is highly efficient in extracting vitamin K from the circulation, and -carboxylation declines substantially in other tissues before the deficiency is severe enough to result in bleeding disorders. It may thus be that the bleeding tendencies that have been the hallmark of vitamin K deficiency are, in fact, the last manifestation of deficiency. If so, what the other clinical expressions of deficiency may be remain uncertain. Three vitamin K-dependent proteins are found in bone matrix: osteocalcin (bone gla protein; BGP), matrix glaprotein, and protein S. Only osteocalcin is unique to bone. There is also a kidney gla protein (nephrocalcin) [194], which may be involved in the renal conservation of calcium. Osteocalcin binds avidly to hydroxyapatite (but not to amorphous calcium phosphate) and is chemotactic for bone-resorbing cells. Originally thought to be synthesized and -carboxylated by osteoblasts as they deposit bone matrix, it now seems that osteocalcin is synthesized by osteocytes [195], particularly those newly embedded in forming bone matrix (see also Chapter 4). Roughly 30% of the synthesized osteocalcin is not incorporated into the matrix, but is released instead into the circulation, where, like alkaline phosphatase, it can be measured and used as an indicator of new bone formation (see Chapter 60). In vitamin K deficiency, such as would occur with coumarin anticoagulants, serum osteocalcin levels decline and the degree of carboxylation of the circulating

692 osteocalcin falls dramatically. Further, binding to hydroxyapatite of the osteocalcin produced under these conditions falls precipitously soon after starting anticoagulant therapy. It would seem, therefore, that vitamin K deficiency would have detectable skeletal effects. The problem is that they have been very hard to find. Rats reared and sustained to adult life under near total suppression of osteocalcin -carboxylation show only minor skeletal defects, mostly related to abnormalities in the growth apparatus [186]. Warfarin anticoagulation therapy in humans has generally not been found to be associated with decreased BMD or increased fractures (196,197). However, an osteocalcin knockout mouse exhibits a skeleton significantly more dense than normal [198], a finding compatible with osteocalcin playing a role in facilitating resorption. In aging humans, the problem of detecting skeletal abnormalities is compounded by the relative isolation of bone from current nutritional stresses, discussed briefly earlier. Various vitamin K-related abnormalities have been described in association with osteoporosis, but their pathogenetic significance remains unclear. Circulating vitamin K and menaquinone levels are low in hip fracture patients [199], but because these levels reflect only recent dietary intake [200,201], it is uncertain to what extent they reflect prefracture vitamin K status. Osteocalcin is undercarboxylated in osteoporotics, and this defect responds to physiological doses of vitamin K [202]. Finally, urine calcium has been reported to be high in some patients with osteoporosis and to fall in response to physiological doses of vitamin K [203,204]. In the same subjects, urine hydroxyproline was also found to be high and to fall on vitamin K treatment. The effect was confined to subjects with pretreatment hypercalciuria and could plausibly be explained as a defect first in a calcium transport protein, with a consequent renal leak of calcium, and a corresponding PTH-mediated increase in bone resorption (reflected in the increased hydroxyproline excretion). In a prospective study, Feskanich et al. [205] found lower hip fracture rates in those with highest vitamin K intakes, but the cohort was relatively young (mean age: 61), and the relevance of this finding to more typical hip fracture patients is uncertain. Whether or not vitamin K is important for bone health, serum vitamin K levels are indicators of general nutritional status, and it may simply be that the observation of low vitamin K levels in osteoporotics, especially in those with hip fracture, is a reflection mainly of the often poor nutrition of these individuals [206 – 208]. Manifestly, much about vitamin K and bone health remains unclear and more work must be done. Until such questions are resolved, it would seem prudent to ensure in the elderly a sufficient vitamin K intake to achieve full expression of the -carboxylation of all vitamin K-dependent proteins.

ROBERT P. HEANEY

VIII. OTHER ESSENTIAL NUTRIENTS A. Magnesium Magnesium is an essential intracellular cation, a cofactor of many basic cellular processes, particularly those involving energy metabolism. In the face of true magnesium deficiency, there is widespread cellular dysfunction, including the cells and tissues that control the calcium economy and bone remodeling, among others. While slightly more than half the body magnesium is contained in the mineral of bone, it is less certain whether it plays any role there or is, like zinc (see later), present simply accidentally, insofar as it was present in the ECF bathing the mineralizing site. However, magnesium alters the surface properties of calcium phosphate crystals, and its concentration in bone is sufficiently high to exert such an effect there. However, the physical – chemical equilibrium between bone crystals and the dissolved minerals in the ECF is itself poorly understood; hence any role of magnesium therein is correspondingly uncertain. Magnesium deficiency clearly occurs in humans of all ages, most often resulting from severe alcoholism or intestinal magnesium leaks, as from sprue or from ileostomy losses. One well-studied manifestation is hypocalcemia, now recognized to be due to refractoriness of the parathyroid glands to the hypocalcemic stimulus itself, coupled with refractoriness of the bone resorption apparatus to parathyroid hormone. Low bone mass is also a common feature in these situations. However individuals with magnesium deficiency commonly have calcium deficiency as well and for the same reasons, a varying combination of low intake, renal wastage, and intestinal leakage. One would expect, therefore, osteoporosis to be very common in such individuals, as is the case. How much of this bony deficit is due to the magnesium deficiency and how much to the calcium deficiency is unclear. (In a clinical sense the question is moot: both deficiencies need repairing.) Treating the underlying condition and replacing lost calcium increases bone density in these patients, but Rude et al. [209] showed that even when the underlying condition is controlled and serum magnesium is seemingly normal, additional magnesium supplementation will produce a further increase in bone mineral density. This latter observation highlights one of the difficulties besetting this field: the assessment of magnesium status. Serum magnesium is recognized not to be a reliable indicator of tissue magnesium repletion. Many investigators favor the magnesium tolerance test [210], i.e., measuring percentage retention of an intravenous infusion of magnesium. This is, of course, not practicable in clinical practice. Nevertheless, the observations of Rude et al. [209] highlight the fact that serum magnesium values within the “normal”

693

CHAPTER 27 Nutrition and Risk for Osteoporosis

reference range may mask a capacity to respond to further magnesium supplementation. This is precisely the point at which magnesium intersects the arena of the pathogenesis and treatment of common postmenopausal osteoporosis. Unfortunately, there is probably no segment of the osteoporosis field more beset with poorly designed, poorly executed, and inadequately powered studies than this one. For example, two small trials, one not randomized and, the other with a high loss of subjects during the trial, reported bone gain in postmenopausal women given a supplement containing magnesium [211,212]. Neither study constituted persuasive evidence of a magnesium effect. The upshot of these and many other even weaker studies is that it is simply not possible to say with any certainty what, if any, may be the role that magnesium plays in the pathogenesis or treatment of osteoporosis. One fact seems certain: in any unselected group of individuals with low bone mass, calcium and or vitamin D supplementation results in clear skeletal benefits (see earlier discussion), without using extra magnesium. Despite the fact that magnesium may be necessary for the functioning of such cells as those responsible for synthesizing 1,25(OH)2D [213], there is clear proof that supplemental magnesium does not enhance calcium absorption in ostensibly healthy older adults. Spencer et al. [214] more than doubled daily magnesium intake in a group of volunteers and could find no effect on calcium absorption, whether from low or normal calcium intakes. Similarly, the many randomized controlled trials demonstrating the efficacy of calcium supplementation in reducing age-related bone loss and fractures all achieved their effect without supplementing with magnesium. However, absence of proof is not the same as absence of effect. One cannot say, in the routine management of osteoporosis, that the results would not have been even better had extra magnesium been provided as well. Because sprue syndromes can be silent [215], a subtle magnesium deficiency could well exist in some individuals with otherwise typical osteoporosis (to mention only one potential cause of magnesium deficiency). Hence, lacking the ability easily to identify individuals with unrecognized magnesium deficiency, it is hard to argue against prudent attention to magnesium supplementation in individuals who have osteoporosis or are at high risk for fragility fractures.

B. Trace Minerals Several trace minerals, notably zinc, manganese, and copper, are essential metallic co-factors for enzymes involved in the synthesis of various bone matrix constituents. Ascorbic acid (along with copper) is needed for collagen cross-links. In growing animals, diets deficient in these nu-

trients produce definite skeletal abnormalities [216,217]. Additionally, zinc deficiency is well known to produce growth retardation and other abnormalities in humans. However, it is not known with certainty whether significant deficiencies of these elements develop in previously healthy adults or, if they do, whether such deficiencies contribute detectably to the osteoporosis problem. 1. COPPER Copper is of particular interest. The principal sources of copper in the diet are shellfish, nuts, legumes, whole grain cereals, and organ meats. A true dietary copper deficiency is considered to be rare and to be confined to special circumstances, such as with total parenteral nutrition or infants recovering from malnutrition. Recognized manifestations in humans have usually centered on disorders of hemopoiesis, mainly as an iron-refractory, hypochromic anemia and leukopenia. Osteoporosis or fragility fractures have not been generally considered to be a part of the syndrome. However, copper-deficient premature infants have underdeveloped, weak bones that fracture easily and respond to copper supplementation [218], and in one human with copper deficiency due to a copper transport defect, the patient’s morbidity included osteoporosis [219]. Copper is a necessary cofactor for lysyl oxidase, one of the principal enzymes involved in collagen cross-linking. These cross-links are important for connective tissue strength, both in tension and in compression, as they prevent the fibrils from sliding along one another’s length. Bone formed under conditions of lysyl oxidase inhibition is mechanically weak, independent of mass. Copper deficiency is reported to be associated with osteoporotic lesions in sheep, cattle, and rats [216,220]. Copper has not been studied much in connection with human osteoporosis, but in one study in which serum copper was measured, levels were correlated negatively with lumbar spine BMD, even after adjusting for body weight and dietary calcium intake [221]. In another [222] postmortem specimens of bone from osteoporotic individuals were reported to contain fewer cross-links than bone from age-matched controls. 2. ZINC Zinc is a known constituent of about 300 enzymes, including alkaline phosphatase, and it plays a role with other proteins, such as the estrogen receptor molecule. Its principal sources in the human diet are red meat, whole grain cereals, shellfish, and legumes. A 70-kg adult body contains 2 – 3 g zinc, about half in bone. Most of this bony zinc is located on the surfaces of the calcium phosphate crystals and probably has no metabolic significance. (Many cations present in the mineralizing environment adsorb to the oxygenrich phosphate groups on crystal surfaces and get stuck there as free water is displaced by new mineral deposition.) A fortuitous consequence of this situation is that urine zinc

694 reflects bone resorption. Thus Herzberg et al. [223] showed that urine zinc rises with age, is higher in patients with osteoporosis, and is reduced when postmenopausal women are given estrogen [224]. While some etiologic connection between zinc and osteoporosis cannot be ruled out, these observations are explained most easily as reflections of the enhanced bone resorption found in many patients with osteoporosis, the elevated resorption of the estrogen-deprived, postmenopausal state, and the well- known antiresorptive effect of estrogen. Urinary zinc excretion probably functions as a marker for bone resorption, rather than as a reflection of the underlying disease mechanisms. However, of known nutrients, zinc is the one most strongly related to serum IGF-1 [225], a growth factor known to be osteotrophic even in adults. In this connection, Schürch et al. [226] showed the importance of IGF-1 in recovery from hip fracture. In an observational study from Sweden, fracture risk was higher in individuals with low zinc intakes [227] and, after adjusting for other nutrients, the risk gradient showed the expected dose – response relationship. In a dietary survey of nearly 1000 British premenopausal women, New et al. [228] found high zinc intakes to be associated with higher bone density values at both spine and hip. Suggestive paleolithic evidence connecting zinc intake with bone status is provided by ancient skeletons discovered in Canary Island cave burials (where contamination by, or leeching of minerals into, groundwater is considered not to have occurred). Bones with a normal zinc content per unit ash, concentrated in one region of the islands, tended to be robust, whereas those with low zinc contents, on another island, were found to be osteoporotic [229]. The zinc content of bone, as suggested earlier, is determined by the circulating zinc levels when bone is mineralizing, and thus low bone zinc probably reflects low zinc intake throughout life. Whether this exposure played an etiologic role in the low bone mass of these skeletal remains is conjectural. 3. MANGANESE While manganese is also recognized as an essential nutrient, its precise role in nutrition is much less well characterized than that of copper and zinc. Although manganese deficiency is well recognized in both laboratory and farm animals, there is no generally recognized manganese deficiency syndrome in humans. Manganese is widely distributed in foods and is especially rich in tea. Bone manganese content is, like that of copper and zinc, a reflection mainly of serum levels prevailing at the time bone is formed, and thus a reflection of dietary manganese. Bone manganese probably has no other metabolic significance per se. Manganese is capable of activating many enzymes, but for most the effect is nonspecific. Manganese is, however, believed to be the preferred metal ion for certain glycosylation reactions involved in mucopolysaccharide

ROBERT P. HEANEY

synthesis. In this connection, manganese deficiency could interfere with both cartilage and bone matrix formation. Animals reared on manganese-deficient diets exhibit general growth retardation, but careful measurements indicate that long bone growth is disproportionately affected [230], possibly reflecting a specific problem with endochondral bone formation. There is also indication of delayed skeletal maturation, suggesting a role of manganese in chondrogenesis. Strause and co-workers [231] showed this quite nicely in a rat model in which demineralized bone powder was implanted subcutaneously. In control animals, cartilage formed around the powder implant and then osteogenesis occured. In manganese-deficient animals, neither development took place. In further work, Strause et al. [232] showed that manganese-deficient rats had both disordered regulation of calcium homeostasis and decreased bone mineral density. Because histology was not performed, it is not possible to say whether this represented impaired mineralization or osteoporosis. Finally, Reginster et al. [233] found low serum manganese in a group of 10 women with osteoporosis. What significance any of these findings may have for the bulk of human osteoporosis is uncertain. In one four-way, randomized intervention trial, a trace mineral cocktail including copper, zinc, and manganese slowed bone mineral loss in postmenopausal women when given either with or without supplemental calcium [234]. There appeared to be a small additional benefit from the extra trace minerals; however, the only statistically significant effect in this study was associated with the calcium supplement. This could mean that trace mineral deficiency plays no role in osteoporosis, but it could also mean that not all of the women treated suffered from such deficiency. In fact, because both osteoporotic and age-related bone loss are multifactorial, one would presume that only some of the subjects in such a study would be deficient, as there is no known way to select subjects for inclusion on the basis of presumed trace mineral need. Thus the suggestive findings of this study have to be considered grounds for further exploration of this issue.

References 1. Consensus Development Conference, Prophylaxis and treatment of osteoporosis. Am. J. Med. 90, 107 – 110 (1991). 2. V. Matkovic, K. Kostial, I. Simonovic, R. Buzina, A. Brodarec, and B. E. C. Nordin, Bone status and fracture rates in two regions of Yugoslavia. Am. J. Clin. Nutr. 32, 540 – 549 (1979). 3. S. N. Robinovitch, W. C. Hayes, and T. A. McMahon, Prediction of femoral impact forces in falls on the hip. J. Biomech. Eng. 113, 366 – 374 (1991). 4. B. Vellas, R. N. Baumgartner, S. J. Wayne, J. Concelcao, C. Lafont, J.-L. Albarede, and P. J. Garry, Relationship between malnutrition and falls in the elderly. Nutrition 8, 105 – 108 (1992). 5. J. B. Lauritzen, M. M. Petersen, and B. Lund, Effect of external hip protectors on hip fracture. Lancet 341, 11 – 13 (1993).

CHAPTER 27 Nutrition and Risk for Osteoporosis 6. R. P. Heaney, Is there a role for bone quality in fragility fractures? Calcif. Tissue Int. 53, S3 – S5 (1993). 7. P. D. Ross, J. W. Davis, R. S. Epstein, and R. D. Wasnich, Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann. Int. Med. 114, 919 – 923 (1991). 8. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). 9. M. Kleerekoper, L. A. Feldkamp, S. A. Goldstein, M. J. Flynn, and A. M. Parfitt, Cancellous bone architecture and bone strength. In “Osteoporosis” (C. Christiansen, J. S. Johansen, and B. J. Riis, eds.), pp. 294 – 300. 1987, Nørhaven A/S, Viborg, Denmark, (1987). 10. J. E. Aaron, N. B. Makins, and K. Sagreiya, The microanatomy of trabecular bone loss in normal aging men and women. Clin. Orthop. Rel. Res. 215, 260 – 271 (1987). 11. I. Eventov, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis and osseous remodeling in iliac crest and femoral neck biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 12, 1 – 6 (1991). 12. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C.-C. Glüer, and H. K. Genant, Simple measurement of femoral geomtry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 13. C. C. Glüer, S. R. Cummings, A. Pressman, J. Li, K. Glüer, K. G. Faulkner, S. Grampp, and H. K. Genant, Prediction of hip fractures from pelvic radiographs: The study of osteoporotic fractures. J. Bone Miner. Res. 9, 671 – 677 (1994). 14. V. Gilsanz, L. M. Loro, T. F. Roe, J. Sayre, R. Gilsanz, and E. E. Schulz, Vertebral size in elderly women with osteoporosis. J. Clin. Invest. 95, 2332 – 2337 (1995). 15. R. P. Heaney, Osteoporotic fracture space: A hypothesis. Bone Miner. 6, 1 – 13 (1989). 16. R. P. Heaney and R. R. Recker, Distribution of calcium absorption in middle-aged women. Am. J. Clin. Nutr. 43, 299 – 305 (1986). 17. B. E. C. Nordin, K. J. Polley, A. G. Need, H. A. Morris, and D. Marshall, The problem of calcium requirement. Am. J. Clin. Nutr. 45, 1295 – 1304 (1987). 18. B. E. C. Nordin, A. G. Need, H. A. Morris, and M. Horowitz, Sodium, calcium and osteoporosis. In “Nutritional Aspects of Osteoporosis” (P. Burckhardt and R. P. Heaney, eds.), Vol. 85 pp. 279 – 295. Raven Press, New York (1991). 19. M. J. Barger-Lux, R. P. Heaney, P. T. Packard, J. M. Lappe, and R. R. Recker, Nutritional correlates of low calcium intake. Clin. Appl. Nutr. 2, 39 – 44 (1992). 20. A. Devine R. L. Prince, and R. Bell, Nutritional effect of calcium supplementation by skim milk powder or calcium tablets on total nutrient intake in postmenopausal women. Am. J. Clin. Nutr. 64, 731 – 737 (1996). 21. E. Barrett-Connor, Diet assessment and analysis for epidemiologic studies of osteoporosis. In “Nutritional Aspects of Osteoporosis” (P. Burckhardt and R. P. Heaney, eds.), Vol. 85, pp. 91 – 98. Raven Press, New York (1991). 22. R. P. Heaney, Assessment and consistency of calcium intake. In “Nutritional Aspects of Osteoporosis” (P. Burckhardt and R.P. Heaney, eds.), Vol. 85 pp., 99 – 104. Raven Press, New York, (1991). 23. C. H. Jacobus, M. F. Holick, Q. Shao, T. C. Chen, I. A. Holm, J. M. Kolodny, G. E.-H. Fuleihan, and E. W. Seely, Hypervitaminosis D associated with drinking milk. N. Engl. J. Med. 326, 1173 – 1177 (1992). 24. M. F. Holick, Q. Shao, W. W. Liu, and T. C. Chen, The vitamin D content of fortified milk and infant formula. N. Engl. J. Med. 326, 1178 – 1181 (1992). 25. T. C. Chen, H. H. Heath III, and M. F. Holick, An update on the vitamin D content of fortified milk from the United States and Canada. N. Engl. J. Med. 329, 1507 (1993).

695 26. M. J. Barger-Lux, R. P. Heaney, S. J. Lanspa, J. C. Healy, and H. F. DeLuca, An investigation of sources of variation in calcium absorption efficiency. J. Clin. Endocrinol. Metab. 80, 406 – 411 (1995). 27. R. Vieth, Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety. Am. J. Clin. Nutr. 69, 842 – 856 (1999). 28. P. Charles, Metabolic bone disease evaluated by a combined calcium balance and tracer kinetic study. Dan. Med. Bull. 36, 463 – 479 (1989). 29. R. P. Heaney, C. M. Weaver, S. M. Hinders, B. Martin, and P. Packard, Absorbability of calcium from Brassica vegetables. J. Food Sci. 58, 1378 – 1380 (1993). 30. R. P. Heaney, C. M. Weaver, and R. R. Recker, Calcium absorbability from spinach. Am. J. Clin. Nutr. 47, 707 – 709 (1988). 31. R. P. Heaney, K. M. Davies, R. R. Recker, and P. T. Packard, Longterm consistency of nutrient intakes. J. Nutr. 120, 869 – 875 (1990). 32. C. J. Carr and R. F. Shangraw, Nutritional and pharmaceutical aspects of calcium supplementation. Am. Pharm. NS27: 49, 50, 54 – 57 (1987). 33. R. F. Shangraw, Factors to consider in the selection of a calcium supplement. In “Proceedings of the 1987 Special Topic Conference on Osteoporosis.” Public Health Reports S104, 46 – 50 (1989). 34. R. P. Heaney, Nutritional factors in osteoporosis. Annu. Rev. Nutr. 13, 287 – 316 (1993). 35. D. M. Hegsted, Calcium and osteoporosis, J. Nutr. 116, 2316 – 2319 (1986). 36. A. M. Parfitt, Morphologic basis of bone mineral measurements: Transient and steady state effects of treatment in osteoporosis. Miner. Electrolyte Metab. 4, 273 – 287 (1980). 37. R. P. Heaney, The bone remodeling transient: Implications for the interpretation of clinical studies of bone mass change. J. Bone Miner. Res. 9, 1515 – 1523 (1994). 38. A. R. Webb, L. W. Kline, and M. F. Holick, Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J. Clin. Endocrinol. Metab. 67, 373 – 378 (1988). 39. L. J. Harris, Vitamin D and Bone. In The Biochemistry and Physiology of Bone” (G. H. Bourne, ed.), pp. 581 – 622 Academic Press, New York, 1956. 40. S. B. Eaton and M. Konner, Paleolithic nutrition: A consideration of its nature and current implications. N. Engl. J. Med. 312, 283 – 289 (1985). 41. S. B. Eaton and D. A. Nelson, Calcium in evolutionary perspective. Am. J. Clin. Nutr. 54, 281S – 287S (1991). 42. K. Alaimo, M. A. McDowell, R. R. Briefel et al., Dietary intake of vitamins, minerals, and fiber of persons ages 2 months and over in the United States: Third National Health and Nutrition Examination Survey, Phase 1, 1988 – 1991. Advance data from vital and health statistics; No. 258. Hyattsville, Maryland: National Center for Health Statistics, 1994. 43. J. Brand Miller, K. W. James, and P. M. A. Maggiore, Tables of Composition of Australian Aboriginal Foods. Aboriginal Studies Press, Canberra, 1997. 44. F. J. Fernandes-Costa, J. Marshall, C. Ritchie, S. V. van Tonder, D. S. Dunn, T. Jenkins, and J. Metz, Transition from a hunter-gatherer to a settled lifestyle in the !Kung San: Effect on iron, folate and vitamin B12 nutrition. Am. J. Clin. Nutr. 40, 1295 – 1303 (1984). 45. A. Rosanoff and D. H. Calloway, Calcium source in Indochinese immigrants. N. Engl. J. Med. 306, 239 – 240 (1982). 46. P. T. Baker and R. B. Mazess, Calcium: Unusual sources in the highland Peruvian diet. Science 142, 1466 – 1467 (1963). 47. T. Molleson, The eloquent bones of Abu Hureyra. Sci. Ame. 271(2), 70 – 75 (1994). 48. R. P. Heaney, R. R. Recker, M. R. Stegman, and A. J. Moy, Calcium absorption in women: Relationships to calcium intake, estrogen status, and age. J. Bone Miner. Res. 4, 469 – 475 (1989).

696 49. R. P. Heaney and R. R. Recker, Determinants of endogenous fecal calcium in healthy women. J. Bone Miner. Res. 9, 1621 – 1627 (1994). 50. V. Matkovic and R. P. Heaney, Calcium balance during human growth: Evidence for threshold behavior. Am. J. Clin. Nutr. 55, 992 – 996 (1992). 51. M. R. Urist, The origin of bone. Discovery 25, 13 – 19 (1964). 52. M. R. Urist, The bone-body fluid continuum. Perspect. Biol. Med. 6, 75 – 115 (1962). 53. W. Bauer, J. C. Aub, and F. Albright, Studies of calcium and phosphorus metabolism. J. Exp. Med. 49, 145 (1929) 54. J. Gershon-Cohen and J. Jowsey, The relationship of dietary calcium to osteoporosis. Metabolism 13, 221 (1964). 55. M. Bodansky and V. B. Duff, Regulation of the level of calcium in the serum during pregnancy. JAMA 112, 223 – 229 (1939). 56. J. Jowsey and L. G. Raisz, Experimental osteoporosis and parathyroid activity. Endocrinology 82, 384 – 396 (1968). 57. W. J. Banks, Jr., G. P. Epling, R. A. Kainer, and R. W. Davis, Antler growth and osteoporosis. Anat. Rec. 162, 387 – 398 (1968). 58. J. A. Kanis and R. Passmore, Calcium supplementation of the diet — I and II. Br. Med. J. 298, 137 – 140, 205 – 208 (1989). 59. R. M. Forbes, K. E. Weingartner, H. M. Parker, R. R. Bell, and J. W. Erdman, Jr., Bioavailability to rats of zinc, magnesium and calcium in casein-, egg- and soy protein-containing diets. J. Nutr. 109, 1652 – 1660 (1979). 60. Dietary Reference Intakes for Calcium, Magnesium, Phosphorus, Vitamin D, and Fluoride. Food and Nutrition Board, Institute of Medicine. National Academy Press, Washington, DC, 1997. 61. NIH Consensus Conference: Optimal Calcium Intake. JAMA 272, 1942 – 1948 (1994). 62. R. P. Heaney, Calcium, dairy products, and osteoporosis. J. Am. Coll. Nutr. 19 (Suppl. 5), 835 – 995 (2000). 63. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, L. Sadowski, N. Sahyoun, and S. Tannenbaum, A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N. Engl. J. Med. 323, 878 – 883 (1990). 64. M. C. Chapuy, M. E. Arlot, F. Duboeuf et al., Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 65. I. R. Reid, R. W. Ames, M. C. Evans, G. D. Gamble, and S. J. Sharpe, Effect of calcium supplementation on bone loss in postmenopausal women. N. Engl. J. Med. 328, 460 – 464 (1993). 66. C. C. Johnston Jr., J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). 67. T. Lloyd, M. B. Andon, N. Rollings, J. K. Martel, J. R. Landis, L. M. Demers, D. F. Eggli, K. Kieselhorst, and H. E. Kulin, Calcium supplementation and bone mineral density in adolescent girls. JAMA 270, 841 – 844 (1993). 68. T. Chevalley, R. Rizzoli, V. Nydegger, D. Slosman, C.-H. Rapin, J.P. Michel, H. Vasey, and J.-P. Bonjour, Effects of calcium supplements on femoral bone mineral density and vertebral fracture rate in vitamin D-replete elderly patients. Osteopor. Int. 4, 245 – 252, (1994). 69. R. R. Recker, S. Hinders, K. M. Davies, R. P. Heaney, M. R. Stegman, D. B. Kimmel, and J. M. Lappe, Correcting calcium nutritional deficiency prevents spine fractures in elderly women. J. Bone Miner. Res. 11, 1961 – 1966 (1996). 70. B. Dawson-Hughes, S. S. Harris, E. A. Krall, and G. E. Dallal, Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. N. Engl. J. Med. 337, 670 – 676 (1997). 71. I. R. Reid, R. W. Ames, M. C. Evans, S. J. Sharpe, and G. D. Gamble, Determinants of the rate of bone loss in normal postmenopausal women. J. Clin. Endocrinol. Metab. 79, 950 – 954 (1994).

ROBERT P. HEANEY 72. J. F. Aloia, A. Vaswani, J. K. Yeh, P. L. Ross, E. Flaster, and F. A. Dilmanian, Calcium supplementation with and without hormone replacement therapy to prevent postmenopausal bone loss. Ann. Intern. Med. 120, 97 – 103 (1994). 73. “Recommended Dietary Allowances,” 10th Ed. National Acad. Press, Washington, DC, 1989. 74. F. Bronner, B. L. Salle, G. Putet, J. Rigo, and J. Senterre, Net calcium absorption in premature infants: Results of 103 metabolic balance studies. Am. J. Clin. Nutr. 56, 1037 – 1044 (1992). 75. J. M. Pettifor, P. Ross, J. Wang, G. Moodley, and J. Couper-Smith, Rickets in children of rural origin in South Africa: A comparison between rural and urban communities. J. Pediatr. 92, 320 – 324 (1978). 76. L. M. Oginni, C. A. Sharp, M. Worsfold, O.S. Badru, and M. W. J. Davie, Healing of rickets after calcium supplementation. Lancet 353, 296 (1999). 77. G. Leighton and P. L. McKinley, Milk consumption and the growth of school children. Dept. of Health for Scotland H.M.S.O. (1930). 78. J.-P. Bonjour, A.-L. Carrie, S. Ferrari, H. Clavien, D. Slosman et al., Calcium-enriched foods and bone mass growth in prepubertal girls: A randomized, double-blind, placebo-controlled trial. J. Clin. Invest. 99, 1287 – 1294 (1997). 79. J. C. Christian, P. L. Yu, C. W. Slemenda, and C. C. Johnson, Jr., Heritability of bone mass: A longitudinal study in aging male twins. Am. J. Hum. Genet. 44, 429 – 433 (1989). 80. V. Matkovic, Calcium metabolism and calcium requirements during skeletal modeling and consolidation of bone mass. Am. J. Clin. Nutr. 54, 245S – 260S (1991). 81. L. A. Jackman, S. S. Millane, B. R. Martin, O. B. Wood, G. P. McCabe et al., Calcium retention in relation to calcium intake and postmenarcheal age in adolescent females. Am. J. Clin. Nutr. 66, 327 – 333 (1997). 82. G. M. Chan, K. Hoffman, and M. McMurray, Effects of dietary products on bone and body composition in pubertal girls. J. Pediatr. 126, 551 – 556 (1995). 83. J. Cadogan, R. Eastell, N. Jones, and M. E. Barker, Milk intake and bone mineral acquisition in adolescent girls: Randomised, controlled intervention trial. Br. Med. J. 315, 1255 – 1260 (1997). 84. R. R. Recker, K. M. Davies, S. M. Hinders, R. P. Heaney, M. R. Stegman, and D. B. Kimmel, Bone gain in young adult women. JAMA 268, 2403 – 2408 (1992). 85. V. Matkovic, D. Fontana, C. Tominac, P. Goel, and C. H. Chesnut III, Factors that influence peak bone mass formation: A study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr. 52, 878 – 888 (1990). 86. H. F. Newton-John and B. D. Morgan, The loss of bone with age: Osteoporosis and fractures. Clin. Orthop. 71, 229 – 232 (1970). 87. V. Matkovic, K. Kostial, I. Simonovic, R. Buzina, A. Brodarec, and B. E. C. Nordin, Bone status and fracture rates in two regions of Yugoslavia. Am. J. Clin. Nutr. 32, 540 – 549 (1979). 88. D. Dertina, M. L. Loro, J. Sayre, F. Kaufman, and V. Gilsanz, Childhood bone measurements predict values at young adulthood. Bone 23, S288 (1998). 89. S. Ferrari, R. Rizzoli, D. Slosman, and J.-P. Bonjour, Familial resemblance for bone mineral mass is expressed before puberty. J. Clin. Endocrinol. Metab. 83, 358 – 361 (1998). 90. C. W. Slemenda, S. L. Hui, C. Longcope, H. Wellman, and C. C. Johnston, Jr, Predictors of bone mass in perimenopausal women. Ann. Intern. Med. 112, 96 – 101 (1990). 91. I. P. Braidman, L. K. Davenport, D. H. Carter, P. L. Selby, E. B. Mawer, and A. J. Freemont, Preliminary in situ identification of estrogen target cells in bone. J. Bone Miner. Res. 10, 74 – 80 (1995). 92. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechat, H. K. Huang, E. E. Schulz, C. R. Libanati, and C. C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988).

697

CHAPTER 27 Nutrition and Risk for Osteoporosis 93. R. P. Heaney, Estrogen-calcium interactions in the postmenopause: A quantitative description. Bone Miner. 11, 67 – 84 (1990). 94. R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in calcium balance performance. J. Lab. Clin. Med. 92, 953 – 963 (1978). 95. B. E. C. Nordin, A. G. Need, H. A. Morris, M. Horowitz, and W. G. Robertson, Evidence for a renal calcium leak in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 401 – 407 (1991). 96. D. C. Welten, H. C. G. Kemper, G. B. Post, and W. A. van Staveren, A meta-analysis of the effect of calcium intake on bone mass in females and males. J. Nutr. 125, 2802 – 2813 (1995). 97. B. E. C. Nordin, Calcium and osteoporosis. Nutrition 13, 664 – 686 (1997). 98. R. R. Recker, J. M. Lappe, K. M. Davies, and D. B. Kimmel, Change in bone mass immediately before menopause. J. Bone Miner. Res. 7, 857 – 862 (1992). 99. D. Baran, A. Sorensen, J. Grimes, R. Lew, A. Karellas, B. Johnson, and J. Roche, Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: A three-year prospective study. J. Clin. Endocrinol. Metab. 70, 264 – 270 (1989). 100. R. C. Klesges, K. D. Ward, M. L. Shelton, W. B. Applegate, E. D. Cantler et al., Changes in bone mineral content in male athletes. JAMA 276, 226 – 230 (1996). 101. R. P. Heaney, Calcium intake and the prevention of chronic disease. In “Nutritional Health: Strategies for Disease Prevention” (T. Wilson, and N. Temple, eds.), pp. 31 – 35. Humana Press, Totowa, NJ (2001). 102. C. A. Koetting and G. M. Wardlaw, Wrist, spine, and hip bone density in women with variable histories of lactation. Am. J. Clin. Nutr. 48, 1479 – 1481 (1988). 103. J. Byrne, M. R. Thomas, and G. M. Chan, Calcium intake and bone density of lactating women in their late childbearing years. J. Am. Diet. Assoc. 87, 883 – 887 (1987). 104. M. A. Hansen, K. Overgaard, B. J. Riis, and C. Christiansen, Potential risk factors for development of postmenopausal osteoporosis — examined over a 12-year period. Osteopor. Int. 1, 95 – 102 (1991). 105. K. Laitinen, M. Valimaki, and P. Keto, Bone mineral density measured by dual-energy X-ray absorptiometry in healthy Finnish women. Calcif. Tissue Int. 48, 224 – 231 (1991). 106. A. R. P. Walker, B. Richardson, and F. Walker, The influence of numerous pregnancies and lactations on bone dimensions in South African Bantu and Caucasian mothers. Clin. Sci. 42, 189 – 196 (1972). 107. D. Kritz-Silverstein, E. Barrett-Connor, and K. A. Hollenbach, Pregnancy and lactation as determinants of bone mineral density in postmenopausal women. Am. J. Epidemiol. 136, 1052 – 1059 (1992). 108. L. Lissner, C. Bengtsson, and T. Hansson, Bone mineral content in relation to lactation history in pre- and postmenopausal women. Calcif. Tissue Int. 48, 319 – 325 (1991). 109. R. P. Heaney and T. G. Skillman, Calcium metabolism in normal human pregnancy. J. Clin. Endocrinol. Metab. 33, 661 – 670 (1971). 110. R. Brommage and D. C. Baxter, Elevated calcium, phosphorus, and magnesium retention in pregnant rats prior to the onset of fetal skeletal mineralization. J. Bone Miner. Res. 3, 667 – 672 (1988). 111. M. J. McGlade, Effects of gestation, lactation, and maternal calcium intake on mechanical strength of equine bone. J. Am. Coll. Nutr. 12, 372 – 377 (1993). 112. M. A. Miller, T. H. Omura, and S. C. Miller, Increased cancellous bone remodeling during lactation in beagles. Bone 10, 279 – 285 (1989). 113. G. N. Kent, R. I. Price, D. H. Gutteridge, M. Smith, J. R. Allen et al., Human lactation: Forearm trabecular bone loss, increased bone turnover, and renal conservation of calcium and inorganic phosphate with recovery of bone mass following weaning. J. Bone Miner. Res. 5, 361 – 369 (1990). 114. N. A. Cross, L. S. Hillman, S. H. Allen, G. F. Krause, and N. E. Vieira, Calcium homeostasis and bone metabolism during pregnancy,

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129. 130.

131.

132.

133.

lactation and postweaning: A longitudinal study. Am. J. Clin. Nutr. 61, 514 – 523 (1995). G. N. Kent, R. I. Price, D. H. Gutteridge, K. J. Rosman, M. Smith, J. R. Allen, C. J. Hickling, and S. L. Blakeman, The efficiency of intestinal calcium absorption is increased in late pregnancy but not in established lactation. Calcif. Tissue Int. 48, 293 – 295 (1991). M. J. Barger-Lux, R. P. Heaney, K. M. Davies, and B. K. Chin BK, Bone mass before and after full-term pregnancy in young women with histories of dietary calcium deficiency. J. Bone Miner. Res. 14, S393 (1999). G. M. Chan, M. McMurry, K. Westover, K. Engelbert-Fenton, and M. R. Thomas, Effects of increased dietary calcium intake upon the calcium and bone mineral status of lactating adolescent and adult women. Am. J. Clin. Nutr. 46, 319 – 323 (1987). A. Morales, L. Tud-Tud Hans, M. Herber, A. K. Taylor, and D. J. Baylink, Lactation is associated with an increase in spinal bone density. J. Bone Miner. Res. 8, S156 (1993). R. Brommage and H. F. DeLuca, Regulation of bone mineral loss during lactation. Am. J. Physiol. Endocrinol. Metab. 11, E182 – E187 (1985). M. F. Sowers, G. Corton, B. Shapiro, M. L. Jannausch, M. Crutchfield, M. L. Smith, J. F. Randolph, and B. Hollis, Changes in bone density with lactation. JAMA 269, 3130 – 3135 (1993). B. L. Specker, N. E. Vieira, K. O. O’Brien, M. L. Ho, J. E. Heubi, S. A. Abrams, and A. L. Yergey, Calcium kinetics in lactating women with low and high calcium intakes. Am. J. Clin. Nutr. 59, 593 – 599 (1994). H. J. Kalkwarf, B. L. Specker, and M. Ho, Effects of calcium supplementation on calcium homeostasis and bone turnover in lactating women. J. Clin. Endocrinol. Metab. 84, 464 – 470 (1999). H. J. Kalkwarf, B. L. Specker, D. C. Bianchi J. Ranz, and M. Ho, The effect of calcium supplementation on bone density during lactation and after weaning. N. Engl. J. Med. 337, 523 – 528 (1997). B. Riis, K. Thomsen, and C. Christiansen, Does calcium supplementation prevent postmenopausal bone loss? N. Engl. J. Med. 316, 173 – 177 (1987). P. J. M. Elders, J. C. Netelenbos, P. Lips, F. C. van Ginkel, E. Khoe et al., Calcium supplementation reduces vertebral bone loss in perimenopausal women: A controlled trial in 248 women between 46 and 55 years of age. J. Clin. Endocrinol. Metab. 73, 533 – 540 (1991). R. R. Recker, P. D. Saville, and R. P. Heaney, The effect of estrogens and calcium carbonate on bone loss in postmenopausal women. Ann. Intern. Med. 87, 649 – 655 (1977). J. A. Kanis, L. J. Melton III, C. Christiansen, C. C. Johnston, and N. Khaltaev, The diagnosis of osteoporosis. J. Bone Miner. Res. 9, 1137 – 1141 (1994). R. R. Recker, J. M. Lappe, K. M. Davies, and R. P. Heaney. Characterization of perimenopausal bone loss: A prospective study. J. Bone Miner. Res. 15, 1965 – 1973 (2000). Osteoporosis Consensus Conference, JAMA 252, 799 – 802 (1984). V. Matkovic, T. Jelic, G. M. Wardlaw, J. Z. Ilich, P. K. Goel, J. K. Wright, M. B. Andon, K. T. Smith, and R. P. Heaney, Timing of peak bone mass in Caucasian females and its implication for the prevention of osteoporosis. J. Clin. Invest. 93, 799 – 808 (1994). Mosekilde Lis, Consequences if the remodeling process for vertebral trabecular bone structure: a scanning electron microscopy study (uncoupling of unloaded structures). Bone Miner. 10, 13 – 35 (1990). R. J. Heikinheimo, J. A. Inkovaara, E. J. Harju, M. V. Haavisto, R. H. Kaarela et al., Annual injection of vitamin D and fractures of aged bones. Calcif. Tissue Int. 51, 105 – 110 (1992). R. P. Heaney, The calcium controversy: A middle ground between the extremes. In “Proceedings of the 1987 Special Topic Conference on Osteoporosis.” Public Health Reports S104, 36 – 46 1989.

698 134. T. L. Holbrook, E. Barrett-Connor, and D. L. Wingard, Dietary calcium and risk of hip fracture: 14-year prospective population study. Lancet 2, 1046 – 1049 (1988). 135. W. R. McKane, S. Khosla, K. S. Egan, S. P. Robins, M. F. Burritt, and B. L. Riggs, Role of calcium intake in modulating age-related increases in parathyroid function and bone resorption. J. Clin. Endocrinol. Metab. 81, 1699 – 1703 (1996). 136. S. M. Pilch (ed.), Physiological effects and health consequences of dietary fiber. Prepared for the Center for Food Safety and Applied Nutrition, Food and Drug Administration under Contract No. FDA 223-84-2059 by the Life Sciences Research Office, Federation of American Societies for Experimental Biology. Available from FASEB Special Publications Office, Bethesda, MD, 1987. 137. M. J. Barger-Lux and R. P. Heaney, Caffeine and the calcium economy revisited. Osteopor Int. 5, 97 – 102 (1995). 138. R. P. Heaney and R. R. Recker, Effects of nitrogen, phosphorus, and caffeine on calcium balance in women. J. Lab. Clin. Med. 99, 46 – 55 (1982). 139. B. E. C. Nordin, A. G. Need, H. A. Morris, and M. Horowitz, The nature and significance of the relationship between urinary sodium and urinary calcium in women. J. Nutr. 123, 1615 – 1622 (1993). 140. B. E. C. Nordin, A. G. Need, H. A. Morris, M. Horowitz, B. E. Chatterton, and A. W. Sedgwick, Bad habits and bad bones. In “Nutritional Aspects of Osteoporosis ‘94” (P. Burckhardt, and R. P. Heaney, eds.), (Proceedings of 2nd International Symposium on Osteoporosis, Lausanne, May 1994). Challenges of Modern Medicine 7, 1 – 25, 1995. Ares-Serono Symposia Publications, Rome, Italy, 1994. 141. C. Demigne, M. A. Levrat, and C. Remesy, Effects of feeding fermentable carbohydrates on the cecal concentrations of minerals and their fluxes between the cecum and blood plasma in the rat. J. Nutr. 119, 1625 – 1630 (1989). 142. C. Coudray, J. Bellanger, C. Castiglia-Delavaud, C. Rémésy, M. Vermorel, and Y. Rayssignuier, Effect of soluble or partly soluble dietary fibre supplementation on absorption and balance of calcium, magnesium, iron and zinc in healthy young men. Eur. J. Clin. Nutr. 51, 375 – 380 (1997). 143. E.G. H. M. van den Heuvel, T. Muys, W. van Dokkum, and G. Schaafsma, Oligofructose stimulates calcium absorption in adolescents. Am. J. Clin. Nutr. 69, 544 – 548 (1999). 144. C. M. Weaver, R. P. Heaney, B. R. Martin, and M. L. Fitzsimmons, Human calcium absorption from whole wheat products. J. Nutr. 121, 1769 – 1775 (1991). 145. C. M. Weaver, R. P. Heaney, W. R. Proulx, S. M. Hinders, and P. T. Packard, Absorbability of calcium from common beans. J. Food Sci. 58, 1401 – 1403 (1993). 146. C. M. Weaver, R. P. Heaney, K. P. Nickel, and P. T. Packard, Calcium bioavailability from high oxalate vegetables: Chinese vegetables, sweet potatoes, and rhubarb. J. Food Sci. 62, 524 – 525 (1997). 147. E. Barrett-Connor, J. C. Chang, and S. L. Edelstein, Coffee-associated osteoporosis offset by daily milk consumption. JAMA 271, 280 – 283 (1994). 148. R. P. Heaney, Protein intake and the calcium economy. J. Am. Diet. Assoc. 93, 1261 – 1262 (1993). 149. V. Matkovic, J. Z. Ilich, M. B. Andon, L. C. Hsieh, M. A. Tzagournis, B. J. Lagger, and P. K. Goel, Urinary calcium, sodium, and bone mass of young females. Am. J. Clin. Nutr. 62, 417 – 425 (1995). 150. E. M. C. Lau, C. Cooper, and J. Woo, Calcium deficiency: A major cause of osteoporosis in Hong Kong Chinese. In “Nutritional Aspects of Osteoporosis” (P. Burckhardt and R. P. Heaney, eds.), Vol. 85, pp. 175 – 180. Raven Press; New York, (1991). 151. C. H. Berkelhammer, R. J. Wood, and M. D. Sitrin, Acetate and hypercalciuria during total parenteral nutrition. Am. J. Clin. Nutr. 48, 1482 (1988).

ROBERT P. HEANEY 152. A. Sebastian, S. T. Harris, J. H. Ottaway et al., Improved mineral balance and skeletal metabolism in postmenopausal women treated with potassium bicarbonate. N. Engl. J. Med. 330, 1776 (1994). 153. J. A. Reed, J. J. B. Anderson. F. A. Tylasky, and P. N. Gallagher, Jr., Comparative changes in radial-bone density in elderly female lactoovo-vegetarians and omnivores. Am. J. Clin. Nutr. 59, 1197S – 1202S (1994). 154. S. I. Barr, J. C. Prior, K. C. Janelle, and B. C. Lentle, Spinal bone mineral density in premenopausal vegetarian and nonvegetarian women: Cross-sectional and prospective comparisons. J. Am. Diet. Assoc. 98, 760 – 765 (1998). 155. J.-F. Chiu, S.-J. Lan, C.-Y. Yang, P.-W. Wang, W.-J. Yao, I.-H. Su, and C.-C. Hsieh, Long-term vegetarian diet and bone mineral density in postmenopausal Taiwanese women. Calcif. Tissue Int. 60, 245 – 249 (1997). 156. E. M. C. Lau, T. Kwok, J. Woo, and S. C. Ho, Bone mineral density in Chinese elderly female vegetarians, vegans, lacto-vegetarians and omnivores. Eur. J. Clin. Nutr. 52, 60 – 64 (1998). 157. H. Spencer, L. Kramer, D. Osis, and C. Norris, Effect of phosphorus on the absorption of calcium and on the calcium balance in man. J. Nutr. 108, 447 – 457 (1978). 158. R. P. Heaney, R. R. Recker, and C. M. Weaver, Absorbability of calcium sources: The limited role of solubility. Calcif. Tissue Int. 46, 300 – 304 (1990). 159. H. Spencer, L. Kramer, C. Norris, and D. Osis, Effect of small doses of aluminum-containing antacids on calcium and phosphorus metabolism. Am. J. Clin. Nutr. 36, 32 – 40 (1982). 160. S. E. Kelly, K. Chawla-Singh, J. H. Sellin, N. J. Yasillo, and I. H. Rosenberg, Effect of meal composition on calcium absorption: Enhancing effect of carbohydrate polymers. Gastroenterology 87, 596 – 600 (1984). 161. M. S. Sheikh, C. A. Santa Ana, N. J. Nicar, L. R. Schiller, and J. S. Fordtran, Calcium absorption: Effect of meal and glucose polymer. Am. J. Clin. Nutr. 48, 312 – 315 (1988). 162. R. P. Heaney, K. T. Smith, R. R. Recker, and S. M. Hinders, Meal effects on calcium absorption. Am. J. Clin. Nutr. 49, 372 – 376 (1989). 163. H. M. Mykkanen, and R. H. Wasserman, Enhanced absorption of calcium by casein phosphopeptides in rachitic and normal chicks. J. Nutr. 110, 2141 – 2148 (1980). 164. R. P. Heaney, Y. Saito, and H. Orimo, Effect of caseinphosphopeptide on absorbability of coingested calcium in normal postmenopausal women. J. Bone Miner. Metab. 12, 77 – 81 (1994). 165. D. Agnusdei, R. Civitelli, A. Camporeale, P. Hardi, and C. Gennari, Dietary L-lysine and calcium metabolism in humans. In “Nutritional Aspects of Osteoporosis ‘94” (Proceedings of 2nd International Symposium on Osteoporosis, Lausanne, May 1994). (P. Burckhardt and R. P. Heaney, eds.), Vol. 7, pp. 263 – 271. Challenges of Modern Medicine, Ares-Serono Symposia Publications, Rome, Italy, 1995. 166. A. W. Norman, Intestinal calcium absorption: A vitamin-D hormonemediated adaptive response. Am. J. Clin. Nutr. 51, 290 – 300 (1990). 167. N. H. Bell, S. Epstein, J. Shary, V. Greene, M. J. Oexmann, and S. Shaw, Evidence of a probable role for 25-hydroxyvitamin D in the regulation of calcium metabolism. J. Bone Miner. Res. 3, 489 – 95 (1988). 168. M. J. Barger-Lux, R. P. Heaney, S. J. Lanspa, J. C. Healy, and H. F. DeLuca, An investigation of sources of variation in calcium absorption physiology. J. Clin. Endocrinol. Metab. 80, 406 – 411, (1995). 169. R. M. Francis, M. Peacock, J. H. Storer, A. E. J. Davies, W. B. Brown, and B. E. C. Nordin, Calcium malabsorption in the elderly: The effect of treatment with oral 25-hydroxyvitamin D3. Eur. J. Clin. Invest. 13, 391 – 396 (1983). 170. C. A. Reasner, J. F. Dunn, D. Fetchick et al., Alteration of vitamin D metabolism in Mexican-Americans. J. Bone Miner. Res. 5, 793 – 794 (1990).

CHAPTER 27 Nutrition and Risk for Osteoporosis 171. I. H. Colodro, A. S. Brickman, J. W. Coburn, T. W. Osborn, and A. W. Norman, Effect of 25-hydroxy-vitamin D3 on intestinal absorption of calcium in normal man and patients with renal failure. Metabolism 27, 745 – 753 (1978). 172. A. Devine, I. M. Dick, S. Wilson, and R. L. Prince, Vitamin D status is unrelated to bone density or turnover in elderly postmenopausal women: A cross-sectional and longitudinal vitamin D intervention study. J. Bone Miner. Res. 14, S382 (1999). 173. R. P. Heaney, M. J. Barger-Lux, M. S. Dowell, T. C. Chen, and M. F. Holick, Calcium absorptive effects of vitamin D and its major metabolites. J. Clin. Endocrinol. Metab. 82, 4111 – 4116 (1997). 174. J. M. McKenna, R. Freaney, A. Meade, and F. P. Muldowney, Hypovitaminosis D and elevated serum alkaline phosphatase in elderly Irish people. Am. J. Clin. Nutr. 41, 101 – 109 (1985). 175. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, S. Harris, L. J. Sokoll, and G. Falconer, Effect of vitamin D supplementation on wintertime and overall bone loss in healthy postmenopausal women. Ann. Int. Med. 115, 505 – 12 (1991). 176. R. P. Heaney, Calcium, bone health, and osteoporosis. In “Bone and Mineral Research” (W. A. Peck, ed.), pp. 255 – 301. Elsevier Science, Amsterdam, 1986. 177. E. A. Krall, N. Sahyoun, S. Tannenbaum, G. E. Dallal, and B. Dawson-Hughes, Effect of vitamin D intake on seasonal variations in parathyroid hormone secretion in postmenopausal women. N. Engl. J. Med. 321, 1777 – 1783 (1989). 178. B. Lukert, J. Higgins, and M. Stoskopf, Menopausal bone loss is partially regulated by dietary intake of vitamin D. Calcif. Tissue Int. 51, 173 – 79 (1992). 179. D. M. Slovik, J. S. Adams, R. M. Neer, M. F. Holick, and J. T. Potts, Jr., Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 – 374 (1981). 180. R. M. Francis, M. Peacock, G. A. Taylor, J. H. Storer, and B. E. C. Nordin, Calcium malabsorption in elderly women with vertebral fractures: Evidence for resistance to the action of vitamin D metabolites on the bowel. Clin. Sci. 66, 103 – 107 (1984). 181. F. M. Gloth III, J. D. Tobin, S. S. Sherman, and B. W. Hollis, Is the recommended daily allowance for vitamin D too low for the homebound elderly? J. Am. Geriatr. Soc. 39, 137 – 141 (1991). 182. A. M. Parfitt, J. C. Gallagher, R. P. Heaney, C. C. Johnston, Jr., R. Neer, and G. D. Whedon, Vitamin D and bone health in the elderly. Am. J. Clin. Nutr. 36, 1014 – 1031 (1982). 183. P. M. Suter and R. M. Russell, Vitamin requirements of the elderly. Am. J. Clin. Nutr. 45, 501 – 512 (1987). 184. M. F. Holick, Sources of vitamin D: Diet and sunlight. In “Nutritional Aspects of Osteoporosis” (Proceedings of 2nd International Symposium on Osteoporosis, Lausanne, May 1994) (P. Burckhardt, R. P. Heaney, eds.), Vol. 7, pp. 289 – 309, Ares-Serono Symposia Publications, Rome, Italy, 1995. 185. P. Lips, W. C. Graafmans, M. E. Ooms, P. D. Bezemer, and L. M. Bouter, Vitamin D supplementation and fracture incidence in elderly persons. Ann. Intern. Med. 124, 400 – 406 (1996). 186. P. A. Price, Role of vitamin-K-dependent proteins in bone metabolism. Annu. Rev. Nutr. 8, 565 – 583 (1988). 187. M. Bliziotes, A. L. Yergey, M. S. Nanes, J. Muenzer, M. G. Begley et al., Absent intestinal response to calciferols in hereditary resistance to 1,25-dihydroxyvitamin D; documentation and effective therapy with high dose intravenous infusions. J. Clin. Endocrinol. Metab. 66, 294 – 300 (1988). 188. R. E. Olson, The function and metabolism of vitamin K. Annu. Rev. Nutr. 4, 281 – 337 (1984). 189. P. Szulc and P. D. Delmas, Is there a role for vitamin K deficiency in osteoporosis? In “Nutritional Aspects of Osteoporosis” (Proceedings of 2nd International Symposium on Osteoporosis, Lausanne, May 1994). (P. Burckhardt, R. P. Heaney, eds.), Vol. 7, pp. 357 – 366. Ares-Serono Publications, Rome, Italy, 1995.

699 190. C. Vermeer, M. H. J. Knapen, and K.-S. G. Jie, Increased vitamin K intake may retard postmenopausal loss of bone mass. In “Nutritional Aspects of Osteoporosis” (Proceedings of 2nd International Symposium on Osteoporosis, Lausanne, May 1994). (P. Burckhardt, and R. P. Heaney, eds.), Vol. 7, pp. 367 – 379 Ares-Serono Publications, Rome, Italy, 1995. 191. N. C. Binkley and J. W. Suttie, Vitamin K nutrition and osteoporosis. J. Nutr. 125, 1812 – 1821 (1995). 192. Y. Nakagawa, M. A. Ahmed, S. L. Hall, S. Deganello, and F. L. Coe, Isolation from human calcium oxalate renal stones of nephrocalcin, a glycoprotein inhibitor of calcium oxalate crystal growth. J. Clin. Invest. 79, 1782 – 1787 (1987). 193. P. V. Hauschka, J. B. Lian, D. E. C. Cole, and C. M. Gundberg, Osteocalcin and matrix gla protein: Vitamin K-dependent proteins in bone. Physiol. Rev. 69, 990 – 1047 (1989). 194. Y. Nakagawa, V. Abram, F. J. Kezdy, E. T. Kaiser, and F. L. Coe, Purification and characterization of the principal inhibitor of calcium monohydrate crystal growth in human urines. J. Biol. Chem. 258, 12594 – 12600 (1983). 195. R. Kasai, P. Bianco, P. G. Robey, and A. J. Kahn, Production and characterization of an antibody against the human bone GLA protein (BGP/osteocalcin) propeptide and its use in immunocytochemistry of bone cells. Bone Miner. 25, 167 – 182 (1994). 196. S. A. Jamal, W. S. Browner, D. C. Bauer, and S. R. Cummings, Warfarin use and risk for osteoporosis in elderly women. Ann. Intern. Med. 128, 829 – 832 (1998). 197. P. J. Caraballo, S. E. Gabriel, M. R. Castro, E. J. Atkinson, and L. J. Melton III, Changes in bone density after exposure to oral anticoagulants: A meta-analysis. Osteopor. Int. 9, 441 – 448 (1999). 198. P. Ducy, C. Desbois, B. Boyce, G. Pinero, B. Story et al., Increased bone formation in osteocalcin-deficient mice. Nature 382, 448 – 452 (1996). 199. S. J. Hodges, M. J. Pilkington, T. C. B. Stamp, A. Catterall, M. J. Shearer et al., Depressed levels of circulating menaquinones in patients with osteoporotic fractures of the spine and femoral neck. Bone 12, 387 – 89 (1991). 200. G. Ferland, J. A. Sadowski, and M. D. O’Brien, Dietary induced subclinical vitamin K deficiency in normal human subjects. J. Clin. Invest. 91, 1761 – 1768 (1993). 201. N. B. Roberts, J. D. Holding, H. P. J. Walsh, L. Klenerman, T. Helliwell et al., Serial changes in serum vitamin K1, triglyceride, cholesterol, osteocalcin and 25-hydroxyvitamin D3 in patients after hip replacement for fractured neck of femur or osteoarthritis. Eur. J. Clin. Invest. 26, 24 – 29 (1996). 202. L. C. Plantalech, M. C. Chapuy, M. Guillaumont, P. Chapuy, M. Leclerq, and P. D. Delmas, Impaired carboxylation of serum osteocalcin in elderly women: Effect of vitamin K1. In “Osteoporosis” (C. Christiansen, and K. Overgaard, eds.), pp. 345 – 347. Osteopress Ap, 1990. 203. M. H. J. Knapen, K. Hamulyak, and C. Vermeer, The effect of vitamin K supplementation on circulating osteocalcin (bone gla protein) and urinary calcium excretion. Ann. Int. Med. 111, 1001 – 1005 (1989). 204. C. Vermeer, and K. Hamulyak, Pathophysiology of vitamin K-deficiency and oral anticoagulants. Thromb Haemostas. 66, 153 – 59 (1991). 205. D. Feskanich, P. Weber, W. C. Willett, H. Rockett, S. L. Booth et al., Vitamin K intake and hip fractures in women: A prospective study. Am. J. Clin. Nutr. 69, 74 – 79 (1999). 206. G. Geinoz, C. H. Rapin, R. Rizzoli, R. Kraemer, B. Buchs et al., Relationship between bone mineral density and dietary intakes in elderly. Osteopor. Int. 3, 242 – 248 (1993). 207. H. Rico, M. Revilla, L. F. Villa, E. R. Hernandez, and J. P. Fernandez, Crush fracture syndrome in senile osteoporosis: A Nutritional consequence. J. Bone Miner. Res. 7, 317 – 19 (1992).

700 208. M. Delmi, C.-H. Rapin, J.-M. Bengoa, P. D. Delmas, H. Vasey, and J.-P. Bonjour, Dietary supplementation in elderly patients with fractured neck of the femur. Lancet 335, 1013 – 1016 (1990). 209. R. K. Rude and M. Olerich, Magnesium deficiency: Possible role in osteoporosis associated with gluten-sensitive enteropathy. Osteopor. Int. 6, 453 – 461 (1996). 210. E. Ryzen, N. Elbaum, F. R. Singer, and R. K. Rude, Parenteral magnesium tolerance testing in the evaluation of magnesium deficiency. Magnesium 4, 137 – 147 (1985). 211. G. E. Abraham, The importance of magnesium in the management of primary postmenopausal osteoporosis. J. Nutr. Med. 2, 165–178 (1991). 212. G. Stendig-Lindberg, R. Tepper, and I. Leichter, Trabecular bone density in a two-year controlled trial of peroral magnesium in osteoporosis. Magn. Res. 6, 155 – 163 (1994). 213. R. K. Rude, J. S. Adams, E. Ryzen, D. B. Endres, H. Niimi et al., Low serum concentrations of 1,25-dihydroxyvitamin D in human magnesium deficiency. J. Clin. Endocrinol. Metab. 61, 933 – 940 (1985). 214. H. Spencer, H. Fuller, C. Norris, and D. Williams, Effect of magnesium on the intestinal absorption of calcium in man. J. Am. Coll. Nutr. 13, 485 – 492 (1994). 215. S. Ott, J. R. Tucci, R. P. Heaney, and S. J. Marx, Hypocalciuria and abnormalities in mineral and skeletal homeostasis in patients with celiac sprue without intestinal symptoms. Endocrinol. Metab. 4, 201 – 206 (1997). 216. G. K. Davis and W. Mertz, Copper. In “Trace Elements in Human and Animal Nutrition” (W. Mertz, ed.), 5th Ed., Vol. I, pp. 301 – 364. Academic Press, San Diego, 1987. 217. W. Mertz (edr.) “Trace elements in Human and Animal Nutrition, 5th Ed. Academic Press, San Diego, 1987. 218. H. Schmidt, J. Herwig, and I. Greinacher, The skeletal changes in premature infants with copper deficiency. Rofo. Fortschritte aud dem Gebiete der Rontgenstrahlen und der Neuen Bildgebenden Verfahren 155, 38 – 42 (1991). 219. A. L. Buchman, C. L. Keen, H. V. Vinters, E. Harris, H. T. Chugani et al., Copper deficiency secondary to a copper transport defect: A new copper metabolic disturbance. Metabolism 43, 1462 – 1469 (1994). 220. J. J. Strain, A reassessment of diet and osteoporosis: Possible role for copper. Med. Hypotheses 27, 333 – 338 (1988). 221. G. Howard, M. Andon, M. Bracker, P. Saltman, and L. Strause, Low serum copper, a risk factor additional to low dietary calcium in postmenopausal bone loss. J. Trace Elements Exp. Med. 5, 23 – 31 (1992).

ROBERT P. HEANEY 222. H. Oxlund, L. Mosekilde, and G. Ørtoft, Reduced concentration of collagen reducible cross-links in human trabecular bone with respect to age and osteoporosis. Bone 19, 479 – 484 (1996). 223. M. Herzberg, J. Foldes, R. Steinberg, and J. Menczel, Zinc excretion in osteoporotic women. J. Bone Miner. Res. 5, 251 – 257 (1990). 224. M. Herzberg, A. Lusky, J. Blonder, and Y. Frenkel, The effect of estrogen replacement on zinc in serum and urine. Obstet. Gynecol. 87, 1035 – 1040 (1996). 225. A. Devine, C. Rosen, S. S. Mohan, D. J. Baylink, and R. L. Prince, Effects of zinc and other nutritional factors on IGF-1 and IGF binding proteins in postmenopausal women. Am. J. Clin. Nutr. 68, 200 – 206 (1998). 226. M.-A. Schirch, R. Rizzoli, D. Slosman, and J.-Ph. Bonjour, Protein supplements increase serum IGF-1 and decrease proximal femur bone loss in patients with a recent hip fracture. In “Osteoporosis” (S. E. Papapoulos, P. Lips, H. A. P. Pols, C. C. Johnston, and P. D. Delmas, eds.), pp. 327 – 329. Elsevier B.V. Science Amsterdam, The Netherlands, 1996. 227. S. Elmstahl, B. Gullberg, L. Janzon, O. Johnell, and B. Elmstahl, Increased incidence of fractures in middle-aged and elderly men with low intakes of phosphorus and zinc. Osteopor. Int. 8, 333 – 340 (1998). 228. S. A. New, C. Bolton-Smith, D. A. Grubb, and D. M. Reid, Nutritional influences on bone mineral density: A cross-sectional study in premenopausal women. Am. J. Clin. Nutr. 65, 1831 – 1839 (1997). 229. E. González-Reimers and M. Arnay-de-la-Rosa, Ancient skeletal remains of the Canary Islands: Bone histology and chemical analysis. Anthrop. Anz. 50, 201 – 215 (1992). 230. C. W. Asling and L. S. Hurley, The influence of trace elements on the skeleton. Clin. Orthop. 27, 213 – 264 (1963). 231. L. Strause, P. Saltman, and J. Glowacki, The effect of deficiencies of manganese and copper on osteoinduction and on resorption of bone particles in rats. Calcif. Tissue Int. 41, 145 – 150 (1987). 232. L. G. Strause, J. Hegenauer, P. Saltman, R. Cone, and D. Resnick, Effects of long-term dietary manganese and copper deficiency on rat skeleton. J. Nutr. 116, 135 – 141 (1986). 233. J. Y. Reginster, L. G. Strause, P. Saltman, and P. Franchimont, Trace elements and postmenopausal osteoporosis: A preliminary study of decreased serum manganese. Med. Sci. Res. 16, 337 – 338 (1988). 234. L. Strause, P. Saltman, K. T. Smith, M. Bracker, and M. B. Andon, Spinal bone loss in postmenopausal women supplemented with calcium and trace minerals. J. Nutr. 124, 1060 – 1064 (1994).

CHAPTER 28

Physical Activity and Osteoporosis BELINDA R. BECK JANET SHAW CHRISTINE M. SNOW

I. II. III. IV. V.

Griffith University, School of Physiotherapy and Exercise Science, Queensland 9726, Australia Department of Exercise and Sport Science, University of Utah, Salt Lake City, Utah 84112 Bone Research Laboratory, Oregon State University, Corvallis, Oregon 97331

Introduction Design Considerations Effects of Loading Bone Physical Activity and Bone Hormone Response to Chronic Exercise

VI. Meta-analyses and Reviews VII. Falling and Fracture VIII. Therapeutic Suggestions References

possible to control for all confounding factors and biases, their recognition and minimization will enhance the credibility of study results.

I. INTRODUCTION Since 1974, when Dalen and Olsson [1] reported that 50- to 59-year-old men with 25 years running experience had significantly greater bone mineral content than sedentary controls, the volume of literature supporting this contention has expanded enormously. The aim of this chapter is to examine the evidence that physical activity is beneficial to bone. We describe the response of normal bone tissue to mechanical loading and summarize the effect of physical activity on the bones of men and women of all ages. The influences of body mass, muscle strength, calcium supplementation, and hormone replacement therapy on this effect are also considered. Finally, we discuss the implications of the effect of exercise on the skeleton for falling and risk of fracture and present some exercise therapy suggestions.

A. Training Principles and Bone Mass Drinkwater [2] has emphasized the need to incorporate five principles into exercise study design: specificity, overload, reversibility, initial values, and diminishing returns. Although attention to these principles has become routine in the muscle and cardiovascular fields, they have only recently been applied to studies of bone. An exercise protocol should be designed to load the target bone, i.e., be specific to the site measured. Additionally, an exercise must overload bone in order to stimulate it. Lack of attention to this aspect is a common problem in published intervention studies and one that is difficult to overcome given the lack of information regarding the quantitative relationship between specific activities and the loads imposed at given skeletal sites. Reversibility refers to the reversal in bone response once a stimulus is removed. Initial values refer to the fact

II. DESIGN CONSIDERATIONS Attention to study design and execution is an important aspect of epidemiological research. While it is not always

OSTEOPOROSIS, SECOND EDITION

701

Copyright © 2001 by Academic Press.

702

BECK, SHAW, AND SNOW

that responses from bone are greatest when beginning mass is lower than average. Diminishing returns mean that once a given training level is achieved, further responses will be slow and of small magnitude. Attention to all five principles in research design will maximize the ability of investigators to observe an effect of physical activity on bone if such an effect truly exists. Another important consideration in the study of bone’s response to exercise is the time required to elicit a measurable response. An important distinction between the muscular/cardiovascular and the skeletal systems is the time over which a response to loading occurs. Whereas muscle responds within a few weeks to a training stimulus, at least 6 months is necessary for bone to initiate adaptation and complete one remodeling cycle.

B. Exercise Study Design Limitations Repeated observations of relatively high bone mass in athletes have fueled the assertion that physical activity is beneficial to bone. The dangers associated with the wholesale application of athlete observations to the general population warrant some attention to this topic. Cross-sectional studies, i.e., studies that compare bone mineral density (BMD) between already-exercising and nonexercising groups, contain an inherent limitation of subject selection bias. That is, individuals who choose to exercise may embody certain predisposing skeletal characteristics that influence their choice and ability to initiate and maintain regular physical activity. It is impossible to separate the influence of those characteristics from the influence of the activities on the skeleton. Another problematic methodological limitation of exercise cross-sectional studies is that instruments used to assess physical activity vary widely and few data exist to establish their validity or reliability. Many instruments commonly used were originally designed to assess aerobic work and do not reflect actual loads experienced by the skeleton. Even devices that enumerate steps taken generally do not record the intensity of impact and therefore do not fully describe skeletal loading. Attempts to relate bone mass to a characteristic with such a high degree of potential error risk the identification of erroneous relationships and consequently the generation of inappropriate conclusions. Comparability of cross-sectional studies is further complicated by the variety of subject groupings utilized to measure exercise effect: exercise vs sedentary, low-intensity activity vs high-intensity activity, sport vs sport, dominant side limb vs nondominant side limb, and sport vs retired from sport. With these limitations in mind, a brief review of cross-sectional data is presented later in the chapter to provide a historically comprehensive account of the available data and to illustrate the relative uniformity of findings from this study genre.

C. Other Methodological Concerns It is well recognized that randomizing a heterogeneous sample of previously sedentary subjects to either controlled exercise or no exercise intervention for the duration of a full sigma period is a more valid method of evaluating the effect of physical activity on bone than cross-sectional observation. Unfortunately, results from even the most rigorously randomized and controlled intervention trials may be confounded by other methodological problems that complicate data interpretation and comparisons between studies. The most important examples relate to the methods by which changes in skeletal status are measured. Variations among studies include skeletal site measured, measurement tools, precision error of similar measurement tools, and bone mineral element reported (content, density, and volumetric density). Other methodological shortcomings include the use of “inactive” control groups, which actually participate in nontrivial amounts of physical activity, reliance on subject recall for accounts of previous activity, and underrepresentation of ethnic minorities in study samples. For the reasons just listed, our review of exercise intervention trials focuses primarily on studies that utilized randomization to treatment groups, measured bone regions specifically loaded in the course of the intervention, and utilized instruments of bone mass measurement with the smallest degree of precision error.

III. EFFECTS OF LOADING BONE The primary purpose of a skeleton is to withstand the forces of gravity and muscle contraction to which it is routinely exposed. Bone exhibits a remarkable ability to adapt to changes in chronic mechanical loading in order to best withstand future loads of the same nature. Specifically, to optimize strength without unduly increasing weight, bones accommodate the loads that are habitually imposed upon them by undergoing alterations in mass, external geometry, and internal microarchitecture. This phenomenon, referred to as Wolff’s law, bears the name of a 19th century scientist who attempted a mathematical description of it [3]. While a complete understanding of the method by which bone perceives and responds to mechanical stimuli is yet to be achieved, certain elements of the process are known. The manner by which bone responds to mechanical loading provides the basis for understanding the effect of physical activity on the skeleton.

A. Stress and Strain Loads applied to the skeleton are generally described in terms of stress and strain. Stress is the force applied per unit area to an object. Strain is a measure of deformation in

703

CHAPTER 28 Physical Activity and Osteoporosis

response to the application of stress and can be calculated by dividing the change in an object’s length (in this case, bone) by its original length. One strain unit is equivalent to 0.1% deformation. Strain generates the adaptive response of bone to loading. Strain during bone loading can be measured by strain gauges adhered to the bone surface. Given the invasive nature of this procedure, it is normally only performed in animal experiments. Thus, in human studies, only an assumption of strain can be made when stress is applied to a bone.

B. Bone Adaptation The process of bone adaptation is achieved by temporary alterations in bone resorption and formation activity (see Chapter 1). These alterations persist until bone mass, geometry, and trabecular orientation approach a functional state. The functional parameter that is optimized is not completely understood, but is likely to be the amount of strain experienced per unit of bone. Support for this concept comes from measurements of strains from the bones of different species during physiological loading. Regardless of species or activity type, typical forms of loading produce bone surface strains between 2000 and 3000 microstrain [4]. A mechanostat theory, proposed to describe the adaptive response of bone to loading [5], hypothesized that skeletal responses differ selectively according to the magnitude of the engendered strain. The concept of bone modeling was introduced to refer to bone formation that is uncoupled from resorption. According to the mechanostat theory, when bone is loaded above 2500 microstrain, modeling occurs via increased bone formation, effecting periosteal expansion and reduced endosteal resorption. This adaptive combination creates a bone that is more resistant to deformation. However, when applied loads engender only 200 microstrain or less, modeling is inhibited and intracortical and endosteal remodeling is stimulated. Substantial reductions in chronic loading (such as would occur with immobilization or in microgravity) are thus associated with increased cortical bone porosity, expansion of the marrow cavity, thinning of the bone cortex, and, ultimately, bone that is less resistant to strain. Although control of bone adaptation is related to more complex load signals than simple peak strains, as even very low loads can induce bone formation if applied at sufficiently high frequencies (see Chapter 18), findings of animal studies support the general concept of bone as a mechanostat [6 – 9].

C. Bone Geometry Bone strength is influenced not only by its material properties, such as mineral density, but also by its macroand microscopic geometry. Macroscopic geometry refers to

the shape and dimension of a bone as a whole, whereas microscopic geometry pertains to the orientation of osteons and microporosities within bone tissue. Diaphyseal width contributes significantly to the ability of bone to resist bending loads. The cross-sectional moment of inertia (CSMI) is a measure of bone geometry that determines the resistance of bone to bending at a particular site. It is a function of cross-sectional area and the distribution of bone in that area relative to the point about which the bone bends (axis of rotation). The further the bone is distributed from the axis of rotation, the wider the bone and the more resistance it will have to bending. Martin and Burr [10] illustrated this concept stating that “if 100 mm2 is removed from the inner cortex of . . . bone . . . bending strength can be maintained by putting only approximately 30 mm2 back onto the outside surface” (p. 231). Until recently, very little attention had been given to the structural importance of bone geometry or to the effect of exercise on it. In fact, some of the losses in long bone strength that might be expected to occur with normal agerelated bone loss appear to be ameliorated by compensatory modifications in bone geometry. The effect of physical activity on bone geometry is addressed in Section IV,C,2,h.

D. Factors Influencing Bone Response to Loading As noted previously, the type of activity and overload are key to stimulating the skeletal system. While these variables have been carefully defined for the cardiovascular and muscular systems, they are less clear for bone. Avenues of overload include frequency, duration (repetitions), and intensity (the amount of force at a skeletal site). Of these, bone is most influenced by exercise intensity, an effect clearly illustrated in studies of gymnasts and runners. Gymnasts can experience forces at the ground of more than 12 times body weight, whereas ground reaction forces generated by runners, although high in repetition, are 3 – 5 times body weight. Accordingly, bone mass at both the hip and the spine is 30 – 40% higher in gymnasts than in runners [11]. Further, over a training season, gymnasts increase bone mass whereas runners do not [12], and during periods when gymnastics training is halted, bone mass decreases significantly [13]. By comparison, walking is a low-intensity activity, creating ground reaction forces of approximately 1 times body weight [14,15] and having little or no effect on bone mass. The importance of load magnitude (intensity) to bone stimulation has also been demonstrated in animal models [7,16,17]. The type of exercise is similarly an important contributor to bone response. First, as mentioned previously, the exercise must be site specific, i.e., it must load the target bone. Second, exercise incurring impact, i.e., that is, rapid

704

BECK, SHAW, AND SNOW

development of peak force, such as jumping, has consistently been shown to best stimulate bone mass accretion, particularly at the hip [18 – 20]. Bone formation and apposition have similarly been found to be linearly related to strain rate, independent of force [21,22]. Studies in the Bone Research Laboratory at Oregon State University of 1220 trials of two-footed jumps report an average time to peak force of 0.038 (Bauer et al., unpublished data, 1999). In contrast, peak forces in weight lifting develop more slowly. Hakkinen and associates [23,24] reported that the average time to peak force in a maximal isometric contraction of knee extensors was 0.176 and 0.103 in young untrained and trained men, respectively. A comparison of the bone mass of athletes participating in aerobic dance, squash, and speed skating indicated that while all athletes had higher weight-adjusted BMD than sedentary controls at the lumbar spine, femoral neck, proximal tibia, and calcaneus, squash players exhibited the highest. The high strain rates and high peak forces associated with playing squash may thus be more effective at stimulating bone formation than the large number of low-force repetitions that predominate in aerobics and speed skating [25].

E. Relationship of Body Mass to Bone Mass A strong positive relationship between body mass and bone mineral density or content has been reported by many [26 – 35], although a minority have found otherwise [36, 37]. This relationship is consistent with the tenets of Wolff’s law in that high body mass applies a greater daily gravitational load on the skeleton than low mass. It should be noted, however, that the phenomenon may also be attributable to a commonality of genetic control. That is, total body mass is likely to be related intrinsically to an individual’s predisposition for low or high bone mass. Weight loss may be a risk factor for loss of bone mass. Women aged 44 – 50 years on a dietary intervention who lost a mean of 3.2  4.7 kg over 18 months were observed to lose more bone mass at the hip and spine than controls who did not lose weight [38]. Moderate intensity aerobic activity intervention attenuated losses at the spine but not at the hip. Data are not entirely consistent in this area [39]. Discrepancies of results may be related to variations in the protocol intensity of exercise programs.

F. Relationship of Muscle to Bone Mass In 1970, Doyle et al. [40] found that psoas muscle weight was predictive of vertebral ash weight. Many now believe that the positive effects of exercise on BMD are due, at least in part, to the beneficial effects of exercise on

muscle, however, differences exist between study findings. That exercisers have greater muscle strength in most muscle groups than nonexercisers is intuitive and has been shown repeatedly [28,41]. It is also known that muscle mass is highly correlated with muscle strength and that lean body mass is positively correlated with bone mass and cross-sectional properties [42]. Muscle forces indeed confer substantial loads on bone. Some authors state that muscles place even greater loads on bones than those imposed by gravitational forces [43], accounting for more than 70% of bending moments [44]. Increased muscle mass therefore is likely to exert an effect on bone mass in two ways: (1) by increasing total body mass and the consequent magnitude of gravitational load on the skeleton and (2) by applying local strain by direct traction at sites of muscle origin and insertion. Hip, spine, whole body, and tibial BMD have all been positively correlated with back, bicep, quadricep, and grip strength. Back extensor muscle mass and strength in particular have been found to be the strongest, most robust predictors of BMD at many sites, particularly the spine and hip [28,37,45 – 49]. Such relationships between strength of certain muscle groups and nonlocal bone mass are likely to be an indication of a more indirect relationship among body, muscle, and bone mass. In fact, Foley and others [50] reported no relationship between maximum grip strength and proximal femoral BMD in men and only a weak relationship in women. Findings that grip and biceps strength correlate positively with forearm BMC [37,51], that quadriceps strength is a positive determinant of hip BMD [32,52], and that leg strength is positively correlated with hip BMD [53] are more in tune with the recurring concept of site specificity of skeletal response to loading. Some authors, however, have reported that quadricep strength is neither an independent predictor of BMD [54] nor related to distal femoral BMD at all [36]. Similarly, adolescent female horse riders exhibit significantly stronger thigh muscle strength than nonactive controls, but no difference in bone mass [55]. In some cases, the relationship between BMD and muscle strength may be weakened by the influence of high-impact activity [41]. Frost [5] stated that declining muscle strength in aging individuals decreases loads on bone previously adapted to strains generated by stronger, young-adult muscles. The resultant reduction in muscle-imposed bone strain constitutes a form of bone unloading so that disuse – osteopenia ensues. It is perplexing, however, that muscle strength gains following resistance training programs are only sometimes accompanied by gains in BMD [56 – 58], not always [59,60]. In order to fully understand the nature of the bone mass/muscle strength relationship, studies designed to primarily address this issue must be completed, with consideration given to initial BMD and muscle strength as well as removal of other mechanical influences.

705

CHAPTER 28 Physical Activity and Osteoporosis

IV. PHYSICAL ACTIVITY AND BONE A. Skeletal Unloading It is well known that chronic reductions in mechanical loading, such as occur following spinal cord injury, prolonged bed rest, and limb immobilization, precipitate generalized skeletal loss, particularly in bones that bear weight under normal conditions [61 – 64]. Paraplegic and quadriplegic patients may lose more than 2% of lower extremity bone mass for the first 4 – 6 months following injury, thereafter losing approximately 1% a month for the first year (see Chapter 46). Krølner and Toft [61] observed a 3.6% bone decrease in patients who were hospitalized with bed rest for an average of 27 days. Following an average 15 weeks of reambulation, a mean gain of 4.4% BMD was observed. Reductions in bone mass of up to 1% per month have been reported in astronauts exposed to microgravity [65]. These losses are not always entirely regained upon return to weight bearing in normal gravity.

B. Cross-Sectional Exercise Study Findings 1. CHILDREN AND ADOLESCENTS Studies in children and adolescents of various ethnic backgrounds generally support significant associations between physical activity and total body, hip, spine, and forearm BMD [30,34,54,66 – 71]. Evidence is accumulating to suggest that exercise confers the greatest long-term benefit when initiated in the prepubertal years [72,73]. Prepubertal gymnasts have greater BMD at weight-bearing sites than controls, an effect that strengthens as the duration of training increases [66,67,74,75]. Bailey and others [71] found that highly active children have a greater peak bone mineral accrual rate and overall bone mineral accumulation than inactive children for the 2 years around the time that bone mineral is accumulating the fastest (12.5 years for girls and 14.1 years for boys) (Fig. 1). This effect translated into 9 and 17% greater total body bone mineral content 1 year after peak bone mineral content velocity for active boys and girls, respectively. Kannus and others [73] found that female tennis and squash players who begin playing before menarche have twice as much bone in the humeral shaft as those who begin playing after puberty. Slemenda and associates [47], however, found no relationship between physical activity and BMD in peripubertal children and suggest that exercise exerts an influence on BMD before and after puberty but during puberty other factors, such as estrogen, become more influential on bone acquisition. This is supported by Haapasalo and others [76] who found no differences between spine BMD of athletic and control children until Tanner stages IV and V (average ages 13.5 and 15.5, respectively).

FIGURE 1

(A) Total body (TB), (B) lumbar spine (LS), and (C) femoral neck (FN) peak bone mineral accrual velocity (g/year) by inactive, average, and active physical activity groups for girls and boys. Means (SD bars) *significantly greater than inactive p 0.005 and **significantly greater than inactive p  0.001. From J. Bone Miner. Res. 14, 1672 – 1679 (1999), with permission of the American Society for Bone and Mineral Research.

Variations in the BMD response to different activities reflect the different loading patterns of each sport and the phenomenon of site specificity [70,77]. The effect is clearly demonstrated by the fact that dominant limbs have greater BMD than nondominant limbs [78], and athletes loading their dominant limbs preferentially while exercising develop even greater bilateral disparity [79,80]. BMC differences between playing and nonplaying arms in women squash and tennis players are about two times greater if participation in

706 the sport begins at or before menarche compared with afterward [73], although some have observed that the effect does not become evident until the adolescent growth spurt or Tanner stage III (mean age 12.6 years) [76]. 2. ADULTS Adults performing repetitive, weight bearing exercise at relatively high intensities have consistently greater BMD than nonexercisers or those exercising at low intensity. These differences have been observed in the whole body [28,35,41,81 – 86], spine, and/or proximal femur [28,33,35, 41,52 – 54,81,83,85 – 91], pelvis [41,84], distal femur [36], tibia [28,41,89,92,93], humerus [41], calcaneus [27,94], and forearm [89]. Broadband ultrasound attenuation and speed of sound transmission (parameters of bone quality) of the calcaneus are similarly higher in runners than in controls [83]. Again illustrating the phenomenon of site specificity, the high BMD of athletes is observed predominantly at the skeletal sites loaded during their respective activities [26,84,95,96], with some exceptions [82,88]. Bone mass differences between athletes participating in different sports (e.g., water polo and weight training [53]) or at different intensities of the same sport [93] are not as evident. Differences in bone mass observed between active and control populations may be due to differences in bone size rather than density [84,97]. Certain activities may not apply a sufficient stimulus to the skeleton to cause an adaptive response [51]. Athletes participating in moderate- to high-intensity impact activities such as running, jumping, and power lifting have greater bone mass than those performing low-intensity or non-weight-bearing activities [1,82,91,98]. Individuals who participate in nonweight-bearing activities such as swimming have similar BMD to nonexercisers [36,99], although some data in men exist to the contrary [100]. The degree of swimming participation may strongly influence the effect of the activity on bone density. Elite swimmers essentially unload their skeletons by spending extended periods of time in a reduced weight-bearing environment. Forces on the skeleton during swimming appear to be of insufficient magnitude to overcome the negative impact of substantially reduced daily weight bearing. Some have observed lower skull [81] and rib [29] BMD in athletes than in controls, raising the question of whether nonloaded bone actually suffers at the expense of BMD enhancement in loaded bone. In adults, as in children, the dominant arm exhibits greater total and cortical bone mass than the nondominant arm [78,101], and greater differences between right and left side limb bone masses are evident when the dominant limb is chronically overloaded [36,52,79,80]. The difference may be accounted for by an increased periosteal area and cortical thickness rather than BMD [101]. Predictably, rowers and triathletes do not demonstrate BMD sidedness [29],

BECK, SHAW, AND SNOW

given that rowing, running, swimming, and biking load bilateral limbs to a nearly equal degree. Occupational physical activity may influence BMD. Individuals who engage in load-bearing activity in the workplace reportedly have higher femoral BMD than those who predominantly sit [102]. Although exercise does not entirely prevent age-related bone loss [103], very active people generally have higher BMD than inactive people across all ages, indicating that the beneficial effect of habitual participation in physical activity on bone endures throughout life [31 – 33,67,81,104, 105]. While Greendale and colleagues [105] reported a significant linear trend in older men between both lifetime and current exercise and hip BMD, they found no relationship between osteoporotic fracture rate and exercise profile. This observation serves as a reminder that the maintenance of bone mass is arguably not a primary clinical or functional goal in and of itself. In fact, BMD maintenance is largely a method of achieving the more practical goal of minimizing risk of fracture. In conclusion cross-sectional studies, in general, provide support for the notion that a habitual athletic endeavor promotes superior bone mass compared with that accompanying a sedentary lifestyle. The magnitude of this difference is likely to depend on the nature and intensity of the activity, the age at which it was initiated, and the number of years spent in training. Exercise confers the most positive longterm benefits on the skeleton before growth has ceased.

C. Exercise Intervention Study Findings 1. ANIMAL TRIALS The use of animal models to evaluate the effect of physical activity on bone has several advantages. Subject heterogeneity, randomization, and exercise compliance can be tightly controlled, and additional interventions such as hormone replacement and nutritional supplementation or deprivation can be implemented without ethical complications. Some ovariectomized animals, such as rats, mimic the postmenopausal hypoestrogenic condition, such that application of findings to a human population at risk for osteoporosis is possible. There is a large body of literature reporting the effect of isolated or generalized mechanical loading on animal bones. Works by Hert and colleagues [106 – 110], Churches and others [111], Woo and associates [112], O’Connor and colleagues [21], Lanyon and others [113], and Burr and associates [114] preceded the landmark work of Rubin and Lanyon [16], which quantified the response of the functionally isolated avian ulna to various load regimes. The latter study demonstrated that removal of load bearing resulted in bone loss, that only 4 consecutive cycles of physiological strain magnitude (but altered strain distribution) were

CHAPTER 28 Physical Activity and Osteoporosis

necessary to prevent such loss, that 36 cycles per day (peak strain  2050 microstrain at 0.5 Hz) engendered substantial bone formation, and that additional increases in cycle number did not produce further gains in bone mass (Fig. 2) As bone turnover by Haversian remodeling is a common feature of turkey and human bone, the avian model is considered to be representative of changes expected to occur under similar conditions in the human skeleton. The same authors went on to report that strains of 1000 microstrain (applied via 100 consecutive load cycles at 1 Hz) prevented bone loss in the same model and that strains greater than 1000 microstrain were associated with a dose-dependent increase in bone mass, primarily accounted for by periosteal deposition [9] (Fig. 3). Since those reports, numerous studies have similarly concluded that unloading animal bone (via, for example, immobilization, detraining, spinal cord injury, microgravity) causes marked bone loss [115 – 119] and that chronic, repetitive, mechanical loading of animal bone in vivo (for periods of time dependent on the species) will increase bone mass [6 – 8,120 – 134]. Only one bout of loading may be necessary to stimulate a response from bone [135 – 138]. The absence of exercise-induced stimulation of bone cells in the rib following weight bearing exercise further substantiates the claim that bone adaptation is site specific [139]. Results of animal studies suggest that immature bone does not necessarily react as positively to chronic overload-

FIGURE 2

Percentage change in bone mineral content at the midshaft of the ulna preparation over a 6-week experimental period in bone subjected to 0 (), 4 (), 36 (), 360 (), or 1800 () consecutive cycles a day of an identical load regimen. From Rubin and Lanyon, J. Bone Joint. Surg. 66A(3); 397 – 402 (1984).

707

FIGURE 3 Change in cross-sectional area of the ulna midshaft derived from comparison between left (experimental) and right (intact) wings after an 8-week experimental period. Strains below 1000 microstrains are associated with areal reduction and those between 1000 and 4000 with a progressive increase. From Rubin and Lanyon (1985).

ing as mature bone. Young rats subjected to daily treadmill running demonstrated reduced periosteal and endosteal appositional growth resulting in reduced tibial stiffness [140]. Twelve months of running training increased the quantity of femoral diaphyseal bone in immature swine, but did not change its mechanical properties [112]. In contrast, the increased cortical area in mature roosters following 9 weeks of running was associated with optimized structural properties and increased maximal load [141]. Mosley and associates [142] reported that 2000 microstrain of axial loading superimposed on the normal activity of immature rats modified growth-related modeling, thereby reducing total new bone formation, whereas application of 4000 microstrain increased bone mass. There is some evidence to suggest that the nature of the skeletal response may change with age. Young female rats increased cortical mineralization in a site-specific manner following treadmill running, whereas older animals responded with a more generalized skeletal effect [143]. It has been argued that a reduction in bone mass with age in humans reflects the reduced habitual loading imposed by an increasingly sedentary lifestyle, rather than aging per se [144,145]. However, some contend the efficacy of the training stimulus diminishes with age [146 – 151]. Number and vigor of cell populations, concentrations of circulating growth factors, and production of bone matrix proteins all decline with advancing years [152,153] and may contribute to an age-related attenuation of the adaptive response.

708 However some animal studies have reported no age differences in the effect of exercise on bone. Raab and associates [154] found that training effects were not limited by age in female rats; relatively old male rats restored BMC to young adult values after initiation of a running program [155], and 4 h per week of running for 1 year prevented continued high bone turnover in adult rats with NH4Cl-induced osteoporosis [156]. Ovariectomized rats undergo substantial losses in trabecular bone volume and bone strength [157]. While not entirely preventing this loss, exercise has been shown to oppose it and almost completely prevents reductions in bone strength [157 – 160]. In one study, this effect was only seen when the exercise was of moderate (as opposed to high) intensity and duration [160]. In conclusion, animal experiments have thus enabled scientists to conclude that unloading bone precipitates loss of bone mass, that loads of large magnitude are more osteogenic than those of small magnitude, that load-related increases in mass of the immature skeleton may not be accompanied by increases in bone strength, that load effects on bone are site specific, and, finally, that exercise can reduce the loss of bone associated with estrogen withdrawal. 2. HUMAN TRIALS The inherent difficulties of subject recruitment and compliance associated with exercise intervention trials are reflected in the substantially reduced volume of quality data from this source. Additionally, as osteoporosis currently affects a greater proportion of women than men, exercise studies have a disproportionate emphasis on women. In the following section, studies are stratified according to subject age and, in the adult population, sex. a. Children and Adolescents The window of opportunity to increase peak bone mass as a strategy to reduce the risk of osteoporosis in later life is open only during the growing years. Exercise interventions for bone have only recently targeted the pediatric population so that few data exist for children. Of those, however, results are in general agreement with cross-sectional observations. Even in the very young, physical activity appears to promote the acquisition of bone. In a study of premature infants, five repetitions of range of motion, gentle compression, flexion, and extension exercises five times a week resulted in greater acquisition of BMD at 4 weeks in exercised babies than in controls [161]. Six-month-old infants participating in activity programs of either gross or fine motor skills for 18 months similarly exhibited enhanced bone accretion [162]. A strong influence of calcium intake was observed, with relatively low calcium being associated with reduced accretion. Areal BMD at the spine, legs, and whole body increased in prepubescent boys (mean age 10.4 years) compared to

BECK, SHAW, AND SNOW

controls following 8 months of weight-bearing activity that supplemented regular physical education [163]. A similar response was observed in young girls who performed resistance training plus jumping exercise for 10 months [164]. In a randomized study of 134 boys and girls (mean age  7.1 years), jumping at ground reaction forces of eight times body weight increased femoral neck BMC 5.6% compared with control children, an effect that was maintained 7 months after detraining [164a]. Similar results are found in children with disabilities. Eight months of daily physical activity resulted in increases in femoral BMD in children (average age 9 years) with cerebral palsy compared with losses in controls [165]. Four months of exercise intervention (40 min of treadmill, cycling, rowing, and games that elevated the heart rate above 150 beats per minute) increased the rate of whole body bone mass accretion in obese children aged 7 – 11 years [166]. To our knowledge, there is only one report of exercise intervention in adolescents. Following 8 months of plyometric and jumping exercise in an adolescent female cohort (age  14.2 years), no significant difference between groups was observed at any bone site, with the exception of an increase in trochanteric bone mineral content in the exercisers [167]. Because this trial was not randomized and controls were highly active, it is unclear whether the lack of response was due to the high level of activity in controls or the fact that adolescent bone does not respond as dramatically to increased loading as prepubertal bone. Figure 4 illustrates the findings of a selection of exercise intervention trials. b. Premenopausal Women Exercise training programs enhance the bone density of young women in a site-specific manner. Both resistance and weight-bearing endurance exercise programs have been observed to increase spine, femoral, and calcaneal BMD of young adult women [18, 19,168 – 171], although evidence from intervention trials supporting an effect at the femur is somewhat limited. Given the relatively strong effect of weight-bearing activity at the hip observed in cross-sectional studies, the effect of exercise intervention at this site warrants further investigation. Only a few studies have addressed the skeletal response to loading in the years just prior to menopause. Results have indicated that perimenopausal women who exercise will maintain BMD at loaded sites to a greater extent than those who do not [172,173]. c. Young Adult Men Army recruits completing 14 weeks of intensive physical training have been observed to increase leg BMC by up to 12.4% [92,174]. Recruits who begin training with the lowest bone density gain the greatest amount. Those who temporarily ceased training due to

709

CHAPTER 28 Physical Activity and Osteoporosis

FIGURE 4

Summary of bone mass changes observed in randomized controlled exercise intervention trials.

stress fracture also gained bone density, but to a lesser degree (5%). It is notable that 10% of recruits actually lost bone density. This effect was likely to stem from resorption-related remodeling porosity that had not yet been matched by formation due to the short time frame of the intervention. The discrepancy in results across recruits is indicative of the substantial individual variation that exists in the response of bone to exercise loading. Furthermore, the influence of training intensity on bone response becomes apparent when results of the army trials are compared with those of recreational athletes [1]. In the latter, subjects aged 25 – 52 failed to gain bone mass at the spine, humerus, femur, calcaneus or forearm following 3 months of either walking (3 km, 5 days a week) or running (5 km, 3 days a week). The bone mineral gains observed in the army recruits reflect the considerably more intense nature of exercise loading during basic army training, and/or perhaps the predominance of youthful subjects. Only one other male exercise intervention study has been reported. Following 9 months of marathon training, the amount of change in calcaneal BMC of runners significantly exceeded that of nonexercising controls [175]. The correlation between average distance run and percentage change in BMC indicated a strong positive relationship. d. Postmenopausal Women One of the strongest factors confounding outcomes of exercise studies of older women is that of menopause and the associated drop in levels of circulating estrogen, which induces rapid bone loss in the years immediately following menopause. Thus exercise interventions that combine both early and late postmenopausal women may fail to distinguish the factor imparting the greatest effect on bone. The two existing investigations specifically targeting early postmenopausal women concluded that resistance exercise benefited the

lumbar spine but provided insufficient stimulus to prevent hormone-related bone loss at other skeletal sites [176] and that both high-and low-impact exercise maintained spine BMD [177]. Weight training programs of 9 to 24 months duration in postmenopausal women not on hormone replacement generally report an increase or maintenance of BMD compared to losses in controls at the whole body [178], lumbar spine [58,176,178 – 180], proximal femur [58,180,181], and radius [181], although not without exception [59,182,183]. Weight-bearing endurance exercise trials of 7 to 18 months duration also generally report exercise benefits (increases or maintenance of BMD compared to losses in control subjects) at the whole body [178,184], lumbar spine [177,184 – 187], proximal femur [178,184], radius [187], and calcaneus [175,188]. Identification of forms of exercise that are appealing to the postmenopausal woman is a challenge and few studies have attempted to do so. Interestingly, 12 months of folkloristic dancing 3.2 h per week did not produce changes in radial or lumbar bone density in postmenopausal women with normal bone mass, however, subjects with osteoporosis exhibited a significant increase at the lumbar spine [189]. Not all forms of physical activity are beneficial to postmenopausal women, with low-intensity activities being particularly ineffective. There is a general consensus that walking alone does not improve or prevent loss of bone in postmenopausal women [190]. In a singular exception to this contention, Hatori and colleagues [186] reported that 7 months of three times per week walking above the anaerobic threshold increased lumbar spine BMD in this population. Further, 6 months of walking combined with resistance exercise increased vertebral trabecular BMC in osteoporotic, postmenopausal women compared to controls

710 who continued to lose bone [191]. Twelve months of waistdeep water exercise, however, did not prevent spine BMD loss nor change femoral BMD in osteoporotic postmenopausal women, despite changes in other functional fitness parameters [192]. Similarly, no change in femoral neck bone mass of post menopausal women was observed after 9 months of weight-bearing exercise wearing weighted vests [193]. Given numerous previous examples of site specificity, it is no surprise that weight bearing exercise does not increase forearm BMD in postmenopausal women [20,178]. In contrast, 5 months of diverse, dynamic, high force and high rate of loading forearm exercises increased forearm bone density significantly in osteoporotic, postmenopausal women [56,194]. There is some suggestion that the vitamin D receptor genotype influences the responsiveness to exercise of postmenopausal women [195], although, in premenopausal women, physical activity is reportedly beneficial to bone irrespective of genotype [196]. e. Hormone Replacement and Exercise Disruption of the normal hormonal milieu, such as occurs at menopause and in other estrogen-deficient states, precipitates bone loss due to increased osteoclast activation frequency and consequent increased bone resorption. Recent focus on hormone replacement therapy (HRT) of menopausal women has stimulated research concerning its efficacy in comparison to, and in combination with, exercise. Some findings suggest that exercise enhances the bone-conserving effect of HRT. For example, 1 year of resistance exercise increased spine, total body, and radial midshaft BMD significantly in estrogen-replaced, surgically menopausal women compared to estrogen-replaced, nonexercising controls who merely maintained BMD [197]. Similarly, whereas both 9 months of weight-bearing training (walking, jogging, stairs) and HRT increased total body and lumbar spine BMD in 60- to 72-year-old postmenopausal women supplemented to 1500 mg calcium/day, the combination of HRT plus exercise was more effective than exercise or HRT alone [198]. Other studies, however, have found otherwise. One hour of resistance exercise plus 2 h per week of walking or running for 1 year did not enhance the positive effect of estrogen supplementation on lumbar vertebral or femoral neck BMD in postmenopausal women [199]. Similarly, BMD loss at the lumbar spine and proximal femur was prevented by two different regimens of HRT in early postmenopausal women, but 3 h of exercise loading had no additional skeletal effect despite a positive effect of exercise on BMD in a hormone placebo group [60]. f. Older Men Results from exercise intervention trials with older men are meager and less compelling than crosssectional observations. In a study of only six men, Welsh and

BECK, SHAW, AND SNOW

Rutherford [184] found that trochanteric BMD increased significantly in 50- to 73-year-old men performing step and jumping exercises. No randomized, controlled endurance intervention trials have been reported for this population. g. Calcium and Exercise Some have reported that the effect of exercise on BMD is greatest when combined with supplemental calcium [200,201]. From a review of 17 trials, Specker [200] reported that physical activity affects BMD only when mean calcium intake exceeds 1000 mg per day. Many agree that the combination of calcium and exercise is more effective at increasing bone mass or, at least, reducing bone loss in postmenopausal women than calcium supplementation alone [185,202,203]. There is, however, some dissent on this issue. A crosssectional study of 422 women found that while both high levels of physical activity and high calcium intake were associated with a higher total body BMC than low- activity levels and low calcium intake, no significant interaction between the two variables was evident [85]. More pointedly, in a 2-year program of combined aerobics and weight training that increased the BMD of young women, calcium supplementation neither enhanced the exercise benefit nor improved BMD in the absence of exercise [168]. Thus, it appears that exercise provides a greater stimulus to bone than calcium, however, adequate calcium availability is undoubtedly necessary to provide the building blocks for exercise-induced gains in BMD. h. Exercise-Related Geometric Adaptation Improvements in bone strength related to altered geometric properties have been observed in the absence of gains in bone mineral content following exercise intervention in rats [204]. Femoral midshaft cortical thickness of young prepubertal boys was observed to increase following 8 months of weight-bearing activity [163]. Similarly, structural changes unrelated to increases in BMC, such as increased cortical bone area, have been reported in postmenopausal women subjected to site-specific exercises [205]. Dominant side limbs of athletes who preferentially load them normally exhibit expanded diaphyseal diameters in addition to increased BMD. Krahl and colleagues [206] observed significant differences in diameter and length of playing arm ulnae of tennis players compared to contralateral arms. The second metacarpals of playing hands were also wider and longer than in contralateral hands, whereas no differences were observed between limbs of controls. Similarly, Dalen and associates [79] observed a 27% difference in the cortical cross-sectional area between left and right humeri of tennis players compared to a nonsignificant 5% difference in controls. Significant differences between playing and nonplaying arm humeral cortical wall thickness, length, width, and cross-sectional moment of inertia have likewise been observed by others [80].

711

CHAPTER 28 Physical Activity and Osteoporosis

i. Diminishing Returns That only extremely sedentary individuals with very low bone mass exhibit noteworthy increases in skeletal mass following exercise intervention, and that introduction of even substantial new loading on normally active humans or animals increases bone mass by only a few percent, illustrates the principle of diminishing returns. That is, with increasing bone mass, physical activity is a progressively less efficient mechanism to stimulate bone gain. This phenomenon is intuitive given an understanding of Wolff’s law [3]. That is, a skeleton very low in mass will require relatively larger additions of bone to maintain functional stiffness under the imposition of novel exercise loading than one of normal mass. Furthermore, bone requires ongoing overload to continue to adapt. In conclusion, as the principle method of mechanically stimulating the skeleton, physical activity is a relatively effective method by which age-related skeletal loss may be halted or slowed. The most osteogenic activities are those of relatively high intensity and impact. Such loads are likely to be most effective when accompanied by adequate calcium consumption and, for hypoestrogenic women, hormone replacement. Individuals with low baseline values of bone density, and/or whose baseline levels of activity are lowest, are the most predisposed to gaining bone. The strength gains associated with exercise-induced changes in bone geometry may be profound, even in the absence of BMD increases. Studies performed in the future are bound to focus to a greater extent on this effect. Complete understanding of the relationship of exercise to bone mass in humans will be enhanced enormously with the development and validation of accurate, quantitative estimates of mechanical loading history.

V. HORMONE RESPONSE TO CHRONIC EXERCISE A. Women For some premenopausal women, chronic, intense, exercise training will induce amenorrhea. While the cause of exercise-associated amenorrhea was once thought to be low body fat, it is now thought that energy availability (dietary energy intake minus energy expended during exercise) and its effect on the hypothalamic – pituitary – thyroid (H-P-T) axis is the determining factor. Differences have indeed been observed in the H-P-T axis between women athletes with regular menstrual cycles and those with amenorrhea [207]. Reductions in total and free 3,5,3-triiodothyronine (T3) concentrations occur in women exercising at intensities that utilize slightly more energy than they consume, and increases in free thyroxine and reverse T3 concentration occur when energy intake is restricted to an even greater extent [208]. Loucks and associates [208] concluded that energy

deficiency suppresses reproductive function secondarily to thyroid function and thus athletic amenorrhea may be prevented or reversed by increasing energy availability through dietary reform, without moderating training. As with any condition that disrupts estrogen availability, the consequence of exercise-related amenorrhea is bone loss [209,209a,209b]. The term “female athlete triad” has been coined to describe the combined conditions of excessive dietary restraint, hormonal disturbance, and skeletal deficits in female athletes. The positive effect of exercise on bone mass appears to be insufficient to surmount the negative effects of inadequate energy intake and hyperintense exercise regimes in all but some athletes. The exceptions are gymnasts, athletes who load their bones at such high intensities that, despite a high prevalence of menstrual disturbance, have bone density values well above normal [11]. Long distance runners, however, who load their skeletons at much lower intensity, albeit for much greater duration, are not protected from amenorrhea-related bone loss. Although there are individual differences in this regard, loss of bone mass in female hyperendurance runners will place them at increased risk of stress fracture than their eumenorrheic running counterparts [210]. Oral contraceptives may be effective in stabilizing levels of circulating estrogen in athletes with menstrual dysfunction, thereby circumventing the female athlete triad; however, insufficient data exist to fully substantiate this effect [211]. Keen and Drinkwater [212] found that initiating oral contraceptive use approximately 8 years after athletic oligo- or amenorrhea did not improve bone mass, concluding that intervention would have to occur early to prevent negative consequences for bone.

B. Men The ability of chronic exercise to substantially modify hormone balance in men has not been shown. Investigators have found that male athletes exercising at a range of intensities appear to have serum concentrations of testosterone that lie within the normal range [27,29,93,213,214], including adolescents [215]. These generalizations notwithstanding, a degree of subtle hormonal perturbation may be evident in some athletes. Smith and Rutherford [29] found that, while in the normal range, serum total testosterone concentrations were significantly lower in triathletes than in controls, but not in rowers. Further, Wheeler and associates [216] found that total serum testosterone, non-sex hormone binding globulin (SHBG)-bound testosterone, and free testosterone concentrations in men running more than 64 km per week to be 83, 69.5, and 68.1% that of controls, respectively. Prolactin concentrations were also significantly lower in runners than in controls. Others have similarly observed that resting and

712 free testosterone concentrations of trained athletes are 68.8 and 72.6% that of controls [217]. Age may influence the effect. In contrast to younger athlete findings, elderly endurance athletes appear to have significantly greater levels of SHBG than controls [27,218]. The skeletal implications of exercise-related hormonal perturbation for men are unclear. Suominen and Rahkila [27] detected a negative correlation between BMD and SHBG in older endurance athletes but no relationship of BMD with testosterone. Further, intense body building training and self-administered anabolic steroids (testosterone: 193.75  147.82 mg/week) do not stimulate greater osteoblastic activity or bone formation than exercise alone [219]. Four months of progressive resistance exercise training 4 days/week, with or without growth hormone supplementation, did not appreciably increase whole body, spine, or proximal femur BMD in elderly men (mean age 67) with normal BMD [220]. Similarly, 6 months of resistance exercise training, with or without recombinant human growth hormone, increased muscle strength to an equivalent degree, but effected no change in BMD of older men [221,222] (Taaffe, personal communication). Given sparse data from long-term intervention trials, a connection among exercise, hormone status, and bone metabolism for men remains difficult to make.

VI. META-ANALYSES AND REVIEWS Many reviews and meta-analyses have been published regarding the effects of physical activity on bone mass in women [223 – 233]. The problems inherent in the study of exercise effects on bone become very apparent when a systematic review of the literature is attempted. To illustrate, whereas a meta-analysis of the effectiveness of physical activity for the prevention of bone loss in postmenopausal women drew 217 papers reporting prospective intervention trials from the literature, only 18 met inclusion criteria [224]. One author wrote, “Initially, it was hoped that a meta-analysis of the data could be feasible. This plan, however, had to be abandoned when it became clear that the trials were too heterogeneous for statistical pooling to be meaningful” (p. 359) [226]. Wolff and colleagues [233] also noted that the overall treatment effect found in their meta-analysis was almost twice as large for nonrandomized controlled trials than for randomized controlled trials, indicating a strong confounding influence of nonrandom allocation of subjects to groups. The consensus of exercise and bone literature reviews is that relatively few of the myriad published data have stemmed from studies suitably designed to measure such an effect, and that from those that were suitably designed, exercise can improve parameters of bone health or, at the very

BECK, SHAW, AND SNOW

least, reduce age-associated bone loss in a site-specific fashion. The position of the American College of Sports Medicine (ACSM) is that weight-bearing physical activity should be performed throughout life. Benefits include a positive effect of exercise on skeletal development, a reduction in age-related bone loss, and a reduced risk of falling. ACSM holds that exercise cannot match the protective effect of HRT on bone in the years immediately following menopause [234].

VII. FALLING AND FRACTURE Although a comprehensive discussion of this topic is presented in Chapter 32, the relevance of exercise to fall prevention warrants additional discussion here. Muscle weakness consistently emerges as an important antecedent factor for falls and as an independent risk factor for hip fracture [235]. There is general agreement that exercise occupies a central role in the maintenance of muscle strength, flexibility, and balance in the prevention of falls and, as such, is considered a logical approach to reducing osteoporosis-related fractures and promoting functional independence [225,236]. High-intensity progressive resistance training one, two, and three times per week for 24 weeks increased major upper and lower body muscle strength and neuromuscular performance significantly in communitydwelling men and women aged 65 – 79 years [237]. The lack of difference in improvement found between subjects regardless of number of training sessions per week suggests that enhanced strength and neuromuscular function in older adults is attainable with less commitment of time than previously thought. Most research examining the role of physical activity in decreasing the risk of falls and fracture has been conducted within an epidemiological context rather than through prospective intervention. The lack of prospective trials can be attributed to the large sample size and long follow-up period required to detect a significant effect for physical activity on either outcome measure. In general, most literature supports a protective effect of physical activity on the risk of fracture, especially hip fracture [238 – 241]. Specifically, the Study of Osteoporotic Fractures, a large, prospective, community-based, observational study of healthy, older, Caucasian women, found that moderately to vigorously active women had significant reductions in hip and vertebral fracture incidence compared to inactive women [238]. Individuals with normal mobility have been observed to improve bone mass, maximal aerobic capacity, well-being, stamina, mobility, and pain tolerance following exercise training, with no associated incidence of fracture [242]. However, those who have a limited ability to perform activities of daily living are likely to fall

CHAPTER 28 Physical Activity and Osteoporosis

more often when they begin exercising due to the higher probability of falls when moving more [243]. Most falls in these individuals lead to radial (Colles’) fractures due to the forward landing pattern associated with purposeful forward motion in walking [244 – 246]. Vertebral fractures are likely to occur with falls [247]; however, associations among falls, physical activity, and vertebral fracture are not well characterized [238] (see also Chapter 19). Given that the majority of hip fractures occur with a fall [248,249], it is important to know the links between physical activity and falls, especially injurious falls that are most likely to precipitate hip fracture [248,250,251]. Lateral instability, muscle weakness of the lower extremities, and inability to perform a tandem gait independently predict hip fracture and falls [235,252 – 254]. These intrinsic risk factors respond favorably to exercise [193,255,256]. Indeed, the convergence of risk factors for hip fracture and falls is not coincidental and intervention attempts to reduce their likelihood have been multifaceted. Most exercise interventions have focused on decreasing fall risk rather than reducing fracture incidence. These interventions have produced improvements in neuromuscular performance and characteristics related to fall risk. Specifically, dramatic improvements in muscle strength, performance on functional tasks, such as stair climbing and rising from a chair, gait speed, and even confidence in movement skills have been documented [237, 256 – 258]. Not all muscle-building interventions have associated strength gains with balance improvements [257]. Inconsistencies in data are probably due to a variation in exercise components employed. For example, a range of exercise intensities (low to high) and exercise modes (seated vs standing) exists in the literature, and such a variation probably contributes to deviation in the extent to which effects of training transfer to activities of daily living. Progressive, free-standing exercise using weighted vests has been well tolerated by community-dwelling postmenopausal women and by more frail elderly women living in a retirement residence [193,259]. In community-dwelling women, muscle strength and power gains over 9 months of training were predictive of improvements in lateral stability, whereas the more elderly group experienced significant improvements in gait parameters. Because falling to the side raises the risk of hip fracture six fold, improvements in lateral stability may reduce fracture risk significantly. Long-term interventions showing direct effects on incidence of falls and injurious falls are limited and the results are inconsistent. Lord and co-workers [260] implemented a 12-month, comprehensive exercise intervention to decrease falls and reported improvements in strength and balance but no change in incidence of falls. In contrast, Campbell et al. [255] implemented a multifactorial fall reduction intervention involving muscle-building activities plus walking exercise and reported a 40% reduction in all falls (injurious and

713 noninjurious). However, it is not known which component(s) of the program — muscle building, walking, or the two combined— was the most potent for reducing falls. Finally, data from the FICSIT trials (Frailty and Injuries: Cooperative Studies of Intervention Techniques) indicate that activities that are most beneficial for reducing incidence of falls include those that result in muscle strength gains and dynamic balance improvements [261]. In conclusion irrespective of the effect on bone, the strong positive effect of exercise on muscle strength at all ages suggests that exercise indirectly benefits skeletal health in older adults as a function of improved or maintained balance, coordination, and the related reduction in risk of falling.

VIII. THERAPEUTIC SUGGESTIONS Physical activity has the potential to curtail the development of osteoporosis and fragility fractures by increasing peak bone mass, maintaining or increasing adult bone mass, and reducing the risk and incidence of falls. Thus, the exercise prescription to benefit bone will differ across the life span according to the age and health of the participant. Exercise interventions of 6 – 18 months have demonstrated that activities of a high magnitude and a high loading rate promote bone gain in children and adults. These exercises include impact activities such as jumping and resistance activities such as weight training. Although walking at moderate intensity has not been shown to increase bone mass in these relatively short study interventions, a lifetime of walking is likely to be beneficial to the skeleton. In a recent report, individuals who engaged in lifetime weight-bearing activities such as walking had higher bone mass compared to those who were less active [262]. Whereas high-magnitude impact activities are recommended for increasing the bone mass of the younger, more robust skeleton, they are not recommended for elderly with skeletal fragility. Such individuals, with or without a history of vertebral compression fractures, should not engage in jumping activities or deep forward trunk flexion exercises such as rowing, toe touching, and sit-ups. Thus, before initiating a program of high-intensity activities, elderly individuals should consider undergoing a skeletal evaluation. This could include obtaining a history of fractures and an assessment of bone mineral density. Resistance training programs that promote muscle strength, balance, and bone maintenance will be of greatest skeletal value to osteoporotic individuals by reducing the risk of falling. Any individual undertaking a new program should begin slowly with careful attention to exercise form and appropriate progressions. Exercises that produce severe joint pain or muscle soreness that lasts more than 3 days should be discontinued until more moderate exercise can be tolerated.

714

BECK, SHAW, AND SNOW

To encourage life-long exercise compliance, as is clearly necessary for the maintenance of skeletal health and the prevention of falls, the most appropriate exercise prescription is one that takes into consideration enjoyment, cost, accessibility, and safety. While walking affords these benefits, individuals should additionally engage in strength-building exercises in order to overload, and thus strengthen, the muscular system.

References 1. N. Dalen and K. E. Olsson, Bone mineral content and physical activity Acta Orthop. Scand. 45, 170 – 174 (1974). 2. B. L. Drinkwater, 1994 C. H. McCloy, Research Lecture: Does physical activity play a role in preventing osteoporosis? Res. Q. Exerc. Sport 65, 197 – 206 (1994). 3. J. Wolff, “Das Gesetz der Transformation der Knochen.” Hirschwald Verlag, Berlin, 1892. 4. C. T. Rubin and L. E. Lanyon, Dynamic strain similarity in vertebrates: An alternative to allometric limb bone scaling. J. Theor. Biol. 107, 321 – 327 (1984). 5. H. M. Frost, Bone “mass” and the “mechanostat”. Anat. Rec. 219, 1 – 9 (1987). 6. R. A. Hillam and T. M. Skerry, Inhibition of bone resorption and stimulation of formation by mechanical loading of the modeling rat ulna in vivo. J. Bone Miner. Res. 10, 683 (1995). 7. K. M. Newhall, K. J. Rodnick, M. C. Van Der Meulen, D. R. Carter, and R. Marcus, Effects of voluntary exercise on bone mineral content in rats. J. Bone Miner. Res. 6, 289 – 296 (1991). 8. D. M. Raab, T. D. Crenshaw, D. B. Kimmel, and E. L. Smith, A histomorphometric study of cortical bone activity during increased weight-bearing exercise. J. Bone Miner. Res. 6, 741 – 749 (1991). 9. C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 10. R. B. Martin and D. B. Burr, “Structure, Function, and Adaptation of Compact Bone.” Raven Press, New York, 1989. 11. T. L. Robinson, C. Snow-Harter, D. R. Taaffe, D. Gillis, J. Shaw, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10, 26 – 35 (1995). 12. D. R. Taaffe, T. L. Robinson, C. M. Snow, and R. Marcus, High-impact exercise promotes bone gain in well-trained female athletes. J. Bone Miner. Res. 12, 255 – 260 (1997). 13. J. LaRiviere, C. M. Snow, and T. L. Robinson, Bone mass changes in female competitive gymnasts over two seasons. Med. Sci. Sports Exerc 28, S131 (1996). 14. H. Rohrle, R. Scholten, C. Sigolotto, W. Sollbach, and H. Kellner, Joint forces in the human pelvis-leg skeleton during walking. J. Biomech. 17, 409 – 424 (1984). 15. E. J. Bassey, J. J. Littlewood, and S. J. Taylor, Relations between compressive axial forces in an instrumented massive femoral implant, ground reaction forces, and integrated electromyographs from vastus lateralis during various “osteogenic” exercises. J. Biomech. 30, 213 – 223 (1997). 16. C. T. Rubin and L. E. Lanyon, Regulation of bone formation by applied dynamic loads. J. Bone Joint Sur. 66, 397 – 402 (1984). 17. H. E. van der Wiel, P. Lips, W. C. Graafmans, C. C. Danielsen, J. Nauta, A. van Lingen, and L. Mosekilde, Additional weight-bearing during exercsie is more important than duration of exercise for anabolic stimulus of bone: A study of running exercise in female rats. Bone 16, 73 – 80 (1995).

18. E. J. Bassey, M. C. Rothwell, J. J. Littlewood, and D. W. Pye, Preand postmenopausal women have different bone mineral density responses to the same high-impact exercise. J. Bone Miner. Res. 13, 1805 – 1813 (1998). 19. E. J. Bassey and S. J. Ramsdale, Increase in femoral bone density in young women following high-impact exercise. Osteopor. Int. 4, 72 – 75 (1994). 20. A. Heinonen, P. Kannus, H. Sievanen, P. Oja, M. Pasanen, M. Rinne, K. Uusi-Rasi, and I. Vuori, Randomised controlled trial of effect of high-impact exercise on selected risk factors for osteoporotic fractures. Lancet 348, 1343 – 1347 (1996). 21. J. A. O’Connor, L. E. Lanyon, and H. MacFie, The influence of strain rate on adaptive bone remodelling. J. Biomech. 15, 767 – 781 (1982). 22. C. H. Turner, I. Owan, and Y. Takano, Mechanotransduction in bone: Role of strain rate. Am. J. Physiol. 269, E438 – E442 (1995). 23. K. Hakkinen, P. V. Komi, and M. Alen, Effect of explosive type strength training on isometric force- and relaxation-time, electromyographic and muscle fibre characteristics of leg extensor muscles. Acta Physiol. Scand. 125, 587 – 600 (1985). 24. K. Hakkinen, M. Alen, and P. V. Komi, Changes in isometric forceand relaxation-time, electromyographic and muscle fibre characteristics of human skeletal muscle during strength training and detraining. Acta Physiol. Scand. 125, 573 – 585 (1985). 25. A. Heinonen, P. Oja, P. Kannus, H. Sievanen, H. Haapasalo, A. Manttari, and I. Vuori, Bone mineral density in female athletes representing sports with different loading characteristics of the skeleton. Bone 17, 197 – 203 (1995). 26. R. C. Hamdy, J. S. Anderson, K. E. Whalen, and L. M. Harvill, Regional differences in bone density of young men involved in different exercises. Med. Sci. Sports Exerc. 26, 884 – 888 (1994). 27. H. Suominen and P. Rahkila, Bone mineral density of the calcaneus in 70- to 80-yr-old male athletes and a population sample. Med. Sci. Sports Exerc. 23, 1227 – 1233 (1991). 28. C. Snow-Harter, R. Whalen, K. Myburgh, S. Arnaud, and R. Marcus, Bone mineral density, muscle strength, and recreational exercise in men. J. Bone Miner. Res. 7, 1291 – 1296 (1992). 29. R. Smith and O. M. Rutherford, Spine and total body bone mineral density and serum testosterone levels in male athletes. Eur. J. Appl. Physio. 67, 330 – 334 (1993). 30. D. C. Welten, H. C. G. Kemper, G. B. Post, W. Van Mechelen, J. Twisk, P. Lips, and G. J. Teule, Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 9, 1089 – 1096 (1994). 31. M. K. Karlsson, O. Johnell, and K. J. Obrant, Is bone mineral density advantage maintained long-term in previous weight lifters? Calci. Tissue Int. 57, 325 – 328 (1995). 32. N. W. Glynn, E. N. Meilahn, M. Charron, S. J. Anderson, L. H. Kuller, and J. A. Cauley, Determinants of bone mineral density in older men. J. Bone Miner. Res. 10, 1769 – 1777 (1995). 33. T. Sone, M. Miyake, N. Takeda, T. Tomomitsu, N. Otsuka, and M. Fukunaga, Influence of exercise and degenerative vertebral changes on BMD: A cross-sectional study in Japanese men. Gerontology 42, 57 – 66 (1996). 34. A. M. Boot, M. A. J. de Ridder, H. A. P. Pols, E. P. Krenning, and S. M. P. F. de Muinck Keizer-Schrama, Bone mineral density in children and adolescents: Relation to puberty, calcium intake, and physical activity. J. Clin. Endocrinol. Metab. 82, 57 – 62 (1997). 35. K. L. Madsen, W. C. Adams, and M. D. Van Loan, Effects of physical activity, body weight and composition, and muscular strength on bone density in young women. Med. Sci. Sports Exerc. 30, 114 – 120 (1998). 36. B. E. Nilsson, and N. E. Westlin, Bone density in athletes. Clin. Orthop. Relat. Res. 77, 179 – 182 (1971). 37. W. C. Bevier, R. A. Wiswell, G. Pyka, K. C. Kozak, K. M. Newhall, and R. Marcus, Relationship of body composition, muscle strength,

CHAPTER 28 Physical Activity and Osteoporosis

38.

39.

40. 41.

42.

43. 44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

and aerobic capacity to bone mineral density in older men and women. J. Bone Miner. Res. 4, 421 – 432 (1989). L. M. Salamone, J. A. Cauley, D. M. Black, et al., Effect of a lifestyle intervention on bone mineral density in premenopausal women: A randomized trial. Am. J. Clin. Nutr. 70, 97 – 103 (1999). A. S. Ryan, B. J. Nicklas, and K. E. Dennis, Aerobic exercise maintains regional bone mineral density during weight loss in postmenopausal women. J. Appl. Physiol. 84, 1305 – 1310 (1998). F. Doyle, J. Brown, and C. Lachance, Relation between bone mass and muscle weight. Lancet 1, 391 – 393 (1970). U. Pettersson, P. Nordstrom, and R. Lorentzon, A comparison of bone mineral density and muscle strength in young male adults with different exercise level. Calcif. Tissue Int. 64, 490 – 498 (1999). M. Moro, M. C. H. van der Meulen, B. J. Kiratli, R. Marcus, L. K. Bachrach, and D. R. Carter, Body mass is the primary determinant of midfemoral bone acquisition during adolescent growth. Bone 19, 519 – 526 (1996). F. Pauwels, “Gesammelte Abhandlungen zur Funktionellen Anatomie des Bewegungsapparates.” Springer, Berlin. T. W. Lu, S. J. Taylor, J. J. O’Connor, and P. S. Walker, Influence of muscle activity on the forces in the femur: An in vivo study. J. Biomech. 30, 1101 – 1106 (1997). M. Sinaki, M. C. McPhee, S. F. Hodgson, J. M. Merritt, and K. P. Offord, Relationship between bone mineral density of spine and strength of back extensors in healthy postmenopausal women. Mayo Clin. Proc. 61, 116 – 122 (1986). M. Sinaki and K. P. Offord, Physical activity in postmenopausal women: Effect on back muscle strength and bone mineral density of the spine. Arch. Phys. Med. Rehabil. 69, 277 – 280 (1988). C. W. Slemenda, T. K. Reister, S. L. Hui, J. Z. Miller, J. C. Christian, and C. C. Johnston, Jr., Influences on skeletal mineralization in children and adolescents: Evidence for varying effects of sexual maturation and physical activity. J. Pediatr. 125, 201 – 207 (1994). M. L. Villa, R. Marcus, R. Ramirez Delay, and J. L. Kelsey, Factors contributing to skeletal health of postmenopausal Mexican-American women. J. Bone Miner. Res. 10, 1233 – 1242 (1995). D. C. Bauer, W. S. Browner, J. A. Cauley, E. S. Orwoll, J. C. Scott, D. M. Black, J. L. Tao, and S. R. Cummings, Factors associated with appendicular bone mass in older women: The Study of Osteoporotic Fractures Research Group. Ann. Intern. Med. 118, 657 – 665 (1993). K. T. Foley, T. M. Owings, M. J. Pavol, and M. D. Grabiner, Maximum grip strength is not related to bone mineral density of the proximal femur in older adults. Calcif. Tissue Int. 64, 291 – 294 (1999). K. H. Myburgh, S. Charette, L. Zhou, C. R. Steele, S. Arnaud, and R. Marcus, Influence of recreational activity and muscle strength on ulnar bending stiffness in men. Med. Sci. Sports Exerc. 25, 592 – 596 (1993). J. A. Calbet, J. S. Moysi, C. Dorado, and L. P. Rodriguez, Bone mineral content and density in professional tennis players. Calcif. Tissue Int. 62, 491 – 496 (1998). J. E. Block, A. L. Friedlander, G. A. Brooks, P. Steiger, H. A. Stubbs, and H. K. Genant, Determinants of bone density among athletes engaged in weight-bearing and non-weight-bearing activity. J. Appl. Physiol. 67, 1100 – 1105 (1989). H. Duppe, P. Gardsell, O. Johnell, B. E. Nilsson, and K. Ringsberg, Bone mineral density, muscle strength and physical activity. A population-based study of 332 subjects aged 15 – 42 years. Acta Orthop. Scan. 68, 97 – 103 (1997). H. Alfredson, G. Hedberg, E. Bergstrom, P. Nordstrom, and R. Lorentzon, High thigh muscle strength but not bone mass in young horseback-riding females. Calcif. Tissue Int. 62, 497 – 501 (1998). A. Simkin, J. Ayalon, and I. Leichter, Increased trabecular bone density due to bone-loading exercises in postmenopausal osteoporotic women. Calcif. Tissue Int. 40, 59 – 63 (1987).

715 57. M. C. Beverly, T. A. Rider, M. J. Evans, and R. Smith, Local bone mineral response to brief exercise that stresses the skeleton.Br. Med. J. 299, 233 – 235 (1989). 58. M. E. Nelson, M. A. Fiatarone, C. M. Morganti, I. Trice, R. A. Greenberg, and W. J. Evans, Effects of high-intensity strength training on multiple risk factors for osteoporotic fractures: A randomized controlled trial. JAMA 272, 1909 – 1914 (1994). 59. L. A. Pruitt, D. R. Taaffe, and R. Marcus, Effects of a one-year high-intensity versus low-intensity resistance training program on bone mineral density in older women. J. Bone Miner. Res. 10, 1788 – 1795 (1995). 60. J. Heikkinen, E. Kyllonen, E. Kurttila-Matero, G. Wilen-Rosenqvist, K. S. Lankinen, H. Rita, and H. K. Vaananen, HRT and exercise: Effects on bone density, muscle strength and lipid metabolism. A placebo controlled 2– year prospective trial on two estrogen-progestin regimens in healthy postmenopausal women. Maturitas 26, 139 –149 (1997). 61. B. Krølner and B. Toft, Vertebral bone loss: An unheeded side effect of therapeutic bed rest. Clin. Sci. 64, 537 – 540 (1983). 62. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers, and D. E. McMillan, Effect of prolonged bed rest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 63. F. E. Tilton, J. J. Degioanni, and V. S. Schneider, Long-term followup of Skylab bone demineralization. Aviat. Space Environ. Med. 51, 1209 – 1213 (1980). 64. A. D. Leblanc, V. S. Schneider, H. J. Evans, D. A. Engelbretson, and J. M. Krebs, Bone mineral loss and recovery after 17 weeks of bed rest. J. Bone Miner. Res. 5, 843 – 850 (1990). 65. NASA, “The Effects of Space Travel on the Musculoskeletal System” NIH Publication, 93-3484 (1992). 66. D. Courteix, E. Lespessailles, S. L. Peres, P. Obert, P. Germain, and C. L. Benhamou, Effect of physical training on bone mineral density in prepubertal girls: A comparative study between impact-loading and non-impact- loading sports. Osteopor. Int. 8, 152 – 158 (1998). 67. S. Bass, G. Pearce, M. Bradney, E. Hendrich, P. D. Delmas, A. Harding, and E. Seeman, Exercise before puberty may confer residual benefits in bone density in adulthood: studies in active prepubertal and retired female gymnasts. J. Bone Miner. Res. 13, 500 – 507 (1998). 68. S. K. Grimston, N. D. Willows, and D. A. Hanley, Mechanical loading regime and its relationship to bone mineral density in children. Med. Sci. Sports Exerc. 25, 1203 – 1210 (1993). 69. M. Gunnes and E. H. Lehmann, Physical activity and dietary constituents as predictors of forearm cortical and trabecular bone gain in healthy children and adolescents: A prospective study. Acta Paediatr. 85, 19 – 25 (1996). 70. S. C. Tsai, C. H. Kao, and S. J. Wang, Comparison of bone mineral density between athletic and non-athletic Chinese male adolescents. Kaohsiung J. Med. Sci. 12, 573 – 580 (1996). 71. D. A. Bailey, H. A. McKay, R. L. Mirwald, P. R. Crocker, and R. A. Faulkner, A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: The University of Saskatchewan Bone Mineral Accrual Study. J. Bone Miner. Res. 14, 1672 – 1679 (1999). 72. K. M. Khan, K. L. Bennell, J. L. Hopper, L. Flicker, C. A. Nowson, A. J. Sherwin, K. J. Crichton, P. R. Harcourt, and J. D. Wark, Selfreported ballet classes undertaken at age 10-12 years and hip bone mineral density in later life. Osteopor. Int. 8, 165 – 173 (1998). 73. P. Kannus, H. Haapasalo, M. Sankelo, H. Sievanen, M. Pasanen, A. Heinonen, P. Oja, and I. Vuori, Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann. Intern. Med. 123, 27 – 31 (1995). 74. R. M. Daly, P. A. Rich, R. Klein, and S. Bass, Effects of high-impact exercise on ultrasonic and biochemical indices of skeletal status: A prospective study in young male gymnasts. J. Bone Miner. Res. 14, 1222 – 1230 (1999).

716 75. S. M. Nickols-Richardson, P. J. O’Connor, S. A. Shapses, and R. D. Lewis, Longitudinal bone mineral density changes in female child artistic gymnasts. J. Bone Miner. Res. 14, 994 – 1002 (1999). 76. H. Haapasalo, P. Kannus, H. Sievanen, M. Pasanen, K. Uusi-Rasi, A. Heinonen, P. Oja, and I. Vuori, Effect of long-term unilateral activity on bone mineral density of female junior tennis players. J. Bone Miner. Res. 13, 310 – 319 (1998). 77. P. Nordstrom, G. Nordstrom, K. Thorsen, and R. Lorentzon, Local bone mineral density, muscle strength, and exercise in adolescent boys: A comparitive study of two groups with different muscle strength and exercise levels. Calcif. Tissue Int. 58, 402 – 408 (1996). 78. H. Rico, M. Revilla, J. L. Cardenas, L. F. Villa, E. Fraile, F. J. Martin, and I. Arribas, Influence of weight and seasonal changes and radiogrammetry and bone densitometry. Calcif. Tissue Int. 54, 385 – 388 (1994). 79. N. Dalen, P. Laftman, H. Ohlsen, and L. Stromberg, The effect of athletic activity on the bone mass in human diaphyseal bone. Orthopedics 8, 1139 – 1141 (1985). 80. H. Haapasalo, H. Sievanen, P. Kannus, A. Heinenon, P. Oja, and I. Vuori, Dimensions and estimated mechanical characteristics of the humerus after long-term tennis loading. J. Bone Miner. Res. 11, 864 – 872 (1996). 81. M. K. Karlsson, R. Hasserius, and K. J. Obrant, Bone mineral density in athletes during and after career: A comparison between loaded and unloaded skeletal regions. Calcif. Tissue Int. 59, 245 – 248 (1996). 82. K. L. Bennell, S. A. Malcolm, K. M. Khan, S. A. Thomas, S. J. Reid, P. D. Brukner, P. R. Eberling, and J. D. Wark, Bone mass and bone turnover in power athletes, endurance athletes, and controls: A 12-month longitudinal study. Bone 20, 477 – 484 (1997). 83. H. Brahm, H. Strom, K. Piehl-Aulin, H. Mallmin, and S. Ljunghall, Bone metabolism in endurance trained athletes: A comparison to population-based controls based on DXA, SXA, quantitative ultrasound, and biochemical markers. Calcif. Tissue Int. 61, 448 – 454 (1997). 84. A. Wittich, C. A. Mautalen, M. B. Oliveri, A. Bagur, F. Somoza, and E. Rotemberg, Professional football (soccer) players have a markedly greater skeletal mineral content, density and size than age- and BMI-matched controls. Calcif. Tissue Int. 63, 112 – 117 (1998). 85. K. Uusi-Rasi, H. Sievanen, I. Vuori, M. Pasanen, A. Heinonen, and P. Oja, Associations of physical activity and calcium intake with bone mass and size in healthy women at different ages. J. Bone Miner. Res. 13, 133 – 142 (1998). 86. S. Tsuzuku, Y. Ikegami, and K. Yabe, Effects of high-intensity resistance training on bone mineral density in young male powerlifters. Calcif. Tissue. Int. 63, 283 – 286 (1998). 87. J. E. Block, H. K. Genant, and D. Black, Greater vertebral bone mineral mass in exercising young men. West. J. of Med. 145, 39 – 42 (1986). 88. L. A. Colletti, J. Edwards, L. Gordon, J. Shary, and N. H. Bell, The effects of muscle-building exercise on bone mineral density of the radius, spine, and hip in young men. Calcif. Tissue Int. 45, 12 – 14 (1989). 89. M. K. Karlsson, O. Johnell, and K. J. Obrant, Bone mineral density in weight lifters. Calcif. Tissue Int. 52, 212 – 215 (1993). 90. M. A. Mayoux-Benhamou, J. F. Leyge, C. Roux, and M. Revel, Cross-sectional study of weight-bearing activity on proximal femur bone mineral density. Calcif. Tissue Int. 64, 179 – 183 (1999). 91. A. G. Need, J. M. Wishard, F. Scopacasa, M. Horowitz, H. A. Morris, and B. E. Nordin, Effect of physical activity on femoral bone density on men. Br. Med. J. 310(6993), 1501 – 1502 (1995). 92. I. Leichter, A. Simkin, J. Y. Margulies, A. Bevas, R. Steinberg, M. Giladi, and C. Milgrom, Gain in mass density of bone following strenuous physical activity. J. Orthop. Res. 7, 86 – 90 (1989).

BECK, SHAW, AND SNOW 93. J. D. MacDougall, C. E. Webber, J. Martin, S. Ormerod, A. Chesley, E. V. Younglai, C. L. Gordon, and C. J. R. Blimkie, Relationship among running mileage, bone density, and serum testosterone in male runners. J. of Appl. Physio. 73, 1165 – 1170 (1992). 94. T. M. Hutchinson, R. T. Whalen, T. M. Cleet, J. M. Vogel, and S. B. Arnaud, Factors in daily physical activity related to calcaneal mineral density in men. Med. Sci. Sports Exerc. 27, 745 – 750 (1995). 95. J. F. Aloia, S. H. Cohn, T. Babu, C. Abesamis, N. Kalici, and K. Ellis, Skeletal mass and body composition in marathon runners. Metabolism 27, 1793 – 1796 (1978). 96. G. Flodgren, R. Hedelin, and K. Henriksson-Larsen, Bone mineral density in flatwater sprint kayakers. Calci. Tissue Int. 64, 374 – 379 (1999). 97. H. Brahm, H. Mallmin, K. Michaelsson, H. Strom, and S. Ljunghall, Relationships between bone mass measurements and lifetime physical activity in a Swedish population. Calcif. Tissue Int. 62, 400 – 412 (1998). 98. B. A. Michel, D. A. Bloch, and J. F. Fries, Weight-bearing exercise, overexercise, and lumbar bone density over age 50 years. Arch. Intern. Med. 149, 2325 – 2329 (1989). 99. D. R. Taaffe, C. Snow-Harter, D. A. Connolly, T. L. Robinson, M. D. Brown, and R. Marcus, Differential effects of swimming versus weight-bearing activity on bone mineral status on eumenorrheic athletes. J. Bone Miner. Res. 10, 586 – 593 (1995). 100. E. S. Orwoll, J. Ferar, S. K. Ovaitt, M. R. McClung, and K. Huntington, The relationship of swimming exercise to bone mass in men and women. Arch. Intern. Med. 149, 2197 – 2200 (1989). 101. N. Ashizawa, K. Nonaka, S. Michikami, T. Mizuki, H. Amagai, K. Tokuyama, and M. Suzuki, Tomographical description of tennisloaded radius: Reciprocal relation between bone size and volumetric BMD. J. Appl. Physiol. 86, 1347 – 1351 (1999). 102. M. Weiss, R. Yogev, and E. Dolev, Occupational sitting and low hip mineral density. Calcif. Tissue Int. 62, 47 – 50 (1998). 103. A. S. Ryan and D. Elahi, Loss of bone mineral density in women athletes during aging. Calcif. Tissue Int. 63, 287 – 292 (1998). 104. M. L. Pollock, L. J. Mengelkoch, J. E. Graves, D. T. Lowenthal, M. C. Limacher, C. Foster, and J. H. Wilmore, Twenty-year followup of aerobic power and body composition of older track athletes. J. Appl. Physiol. 82, 1508 – 1516 (1997). 105. G. A. Greendale, E. Barrett-Connor, S. Edelstein, S. Ingles, and R. Haile, Lifetime leisure exercise and osteoporosis: The Rancho Bernardo study. Am. J. Epidemiol. 141, 951 – 959 (1995). 106. J. Hert, M. Liskova, and B. Langat, Influence of long-term, continuous bending on the bone. Folia Morphol. 17, 389 (1969). 107. J. Hert, M. Liskova, and J. Landa, Reaction of bone to mechanical stimuli. 1. Continuous and intermittent loading of tibia in rabbit. Folia Morphol. 19, 290 – 317 (1971). 108. J. Hert, M. Sklenska, and M. Liskova, Reaction of bone to mechanical stimuli. 5. Effect of intermittent stress on the rabbit tibia after resection of the peripheral nerves. Folia Morphol. 19, 378 – 387 (1971). 109. J. Hert, F. Mertl, and A. Babicky, Reaction of bone to mechanical stimuli. Incorporation of 45Ca and 35S into rabbit tibia subjected to intermittent stress. Physiol. Bohemoslovaca 20, 575 – 581 (1971). 110. J. Hert, E. Pribylova, and M. Liskova, Reaction of bone to mechanical stimuli. 3. Microstructure of compact bone of rabbit tibia after intermittent loading. Acta Anat. 82, 218 – 230 (1972). 111. A. E. Churches, C. R. Howlett, K. J. Waldron, and G. W. Ward, The response of living bone to controlled time-varying loading: Method and preliminary results. J. Biomech. 12, 35 – 45 (1979). 112. S. L. Woo, S. C. Kuei, D. Amiel, M. A. Gomez, W. C. Hayes, F. C. White, and W. H. Akeson, The effect of prolonged physical training on the properties of long bone: A study of Wolff’s Law. J. Bone Joint. Surg. (Am) 63, 780 – 787 (1981).

717

CHAPTER 28 Physical Activity and Osteoporosis 113. L. E. Lanyon, A. E. Goodship, C. J. Pye, and J. H. MacFie, Mechanically adaptive bone remodelling. J. Biomech. 15, 141 – 154 (1982). 114. D. B. Burr, R. B. Martin, and P. A. Martin, Lower extremity loads stimulate bone formation in the vertebral column: Implications for osteoporosis. Spine 8, 681 – 686 (1983). 115. T. M. Skerry and L. E. Lanyon, Interruption of disuse by short duration walking exercise does not prevent bone loss in the sheep calcaneus. Bone 16, 269 – 274 (1995). 116. T. Thomas, L. Vico, T. M. Skerry, F. Caulin, L. E. Lanyon, C. Alexandre, and M.-H. Lafage, Architectural modifications and cellular response during disuse-related bone loss in calcaneus of the sheep. Am. J. of Physiol. 80, 198 – 202 (1996). 117. A. A. Biewener, N. L. Fazzalari, D. D. Konieczynski, and R. V. Baudinette, Adaptive changes in trabecular architecture in relation to functional strain patterns and disuse. Bone 19, 1 – 8 (1996). 118. P. Kannus, T. L. N. Jarvinen, H. Sievanen, M. Kvist, J. Rauhniemi, V.-M. Maunu, T. Hurme, L. Jozsa, and M. Jarvinen, Effects of immobilization, three forms of remobilization, and subsequent deconditioning on bone mineral content and density in rat femora. J. Bone Miner. Res. 11, 1339 – 1346 (1996). 119. N. E. Lane, A. J. Kaneps, S. M. Stover, G. Modin, and D. B. Kimmel, Bone mineral density and turnover following forelimb immobilization and recovery in young adult dogs. Calcif. Tissue Int. 59, 401 – 406 (1996). 120. D. B. Burr, M. B. Schaffler, K. H. Yang, M. Lukoschek, N. Sivaneri, J. D. Blaha, and E. L. Radin, Skeletal change in response to altered strain environments: Is woven bone a response to elevated strain? Bone 10, 223 – 233 (1989). 121. D. R. Young, D. M. Nunamaker, and M. D. Markel, Quantitative evaluation of the remodeling response of the proximal sesamoid bones to training-related stimuli in Thoroughbreds. Am. J. Vet. Res. 52, 1350 – 1356 (1991). 122. S. A. Feik, Early periosteal changes in translation-induced bone modeling. J. Anat. 182, 389 – 401 (1993). 123. M. R. Forwood and C. H. Turner, The response of rat tibiae to incremental bouts of mechanical loading: A quantum concept for bone formation. Bone 15, 603 – 609 (1994). 124. D. M. Raab-Cullen, M. P. Akhter, D. B. Kimmel, and R. R. Recker, Bone response to alternate-day mechanical loading of the rat tibia. J. Bone Miner. Res. 9, 203 – 211 (1994). 125. D. M. Raab-Cullen, M. P. Akhter, D. B. Kimmel, and R. R. Recker, Periosteal bone formation stimulated by externally induced bending strains. J. Bone Miner. Res. 9, 1143 – 1152 (1994). 126. C. T. Rubin, T. Gross, Y.-X. Qin, S. Fritton, F. Guilak, and K. McLeod, Differentiation of the bone-tissue remodeling response to axial and torsional loading in the turkey ulna. J. Bone Joint Surg. 78, 1523 – 1533 (1996). 127. G. C. Reilly, J. D. Currey, and A. E. Goodship, Exercise of young thoroughbred horses increases impact strength of the third metacarpal bone. J. Orthop. Res. 15, 862 – 868 (1997). 128. A. G. Torrance, J. R. Mosley, R. F. L. Suswillo, and L. E. Lanyon, Noninvasive loading of the rat ulna in vivo induces a strain-related modeling response uncomlicated by trauma or periosteal pressure. Calcif. Tissue Int. 54, 241 – 247 (1994). 129. C. H. Turner, M. R. Forwood, J.-Y. Rho, and T. Yoshikawa, Mechanical loading thresholds for lamellar and woven bone formation. J. Bone Miner. Res. 1, 87 – 97 (1994). 130. U. E. Pazzaglia, L. Andrini, and A. Di Nucci, The effects of mechanical forces on bones and joints. Experimental study on the rat tail. J. Bone Joint. Surg. Br. 79, 1024 – 1030 (1997). 131. Y. X. Qin, C. T. Rubin, and K. J. McLeod, Nonlinear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. J. Orthop. Res. 16, 482 – 489 (1998). 132. T. L. Jarvinen, P. Kannus, H. Sievanen, P. Jolma, A. Heinonen, and M. Jarvinen, Randomized controlled study of effects of sudden im-

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145. 146.

147.

148.

149.

150.

pact loading on rat femur. J. Bone Miner. Res. 13, 1475 – 1482 (1998). E. C. Firth, J. Delahunt, J. W. Wichtel, H. L. Birch, and A. E. Goodship, Galloping exercise induces regional changes in bone density within the third and radial carpal bones of Thoroughbred horses. Equine Vet. J. 31, 111 – 115 (1999). K. C. Westerlind, J. D. Fluckey, S. E. Gordon, W. J. Kraemer, P. A. Farrell, and R. T. Turner, Effect of resistance exercise training on cortical and cancellous bone in mature male rats. J. Appl. Physiol. 84, 459 – 464 (1998). M. D. Boppart, D. B. Kimmel, J. A. Yee, and D. M. Cullen, Time course of osteoblast appearance after in vivo mechanical loading. Bone 23, 409 – 415 (1998). R. A. Dodds, N. Ali, M. J. Pead, and L. E. Lanyon, Early loadingrelated changes in the activity of glucose 6-phosphate dehydrogenase and alkaline phosphatase in osteocytes and periosteal osteoblasts in rat fibulae in vivo. J. Bone Miner. Res. 8, 261 – 267 (1993). S. L. Dallas, G. Zaman, M. J. Pead, and L. E. Lanyon, Early strainrelated changes in cultured embryonic chick tibiotarsi parallel those associated with adaptive bone remodeling in vivo. J. Bone Miner. Res. 8, 251 – 259 (1993). M. J. Pead, T. M. Skerry, and L. E. Lanyon, Direct transformation from quiescence to bone formation in the adult periosteum following a single brief period of bone loading. J. Bone Miner. Res. 3, 647 – 656 (1988). L. J. Tommerup, D. M. Raab, T. D. Crenshaw, and E. L. Smith, Does weight-bearing exercise affect non-weight-bearing bone? J. Bone Miner. Res. 8, 1053 – 1058 (1993). M. R. Forwood and A. W. Parker, Repetitive loading, in vivo, of the tibiae and femora of rats: Effects of repeated bouts of treadmill running. Bone Miner. 13, 35 – 46 (1991). B. J. Loitz and R. F. Zernicke, Strenuous exercise-induced remodelling of mature bone: Relationships between in vivo strains and bone mechanics. J. Exp. Biol. 170, 1 – 18 (1992). J. R. Mosely, B. M. March, J. Lynch, and L. E. Lanyon, Strain magnitude related changes in whole bone architecture in growing rats. Bone 20, 191 – 198 (1997). R. McDonald, J. Hegenauer, and P. Saltman, Age-related differences in the bone mineralization pattern of rats following exercise. J Gerontol. 41, 445 – 452 (1986). R. K. Dishman and M. Steinhardt, Health locus of control predicts free-living, but not supervised, physical activity: A test of exercisespecific control and outcome- expectancy hypotheses. Res. Q. Exerc. Sport 61, 383 – 394 (1990). J. W. Rowe and R. L. Kahn, Human aging: Usual and successful. Science 237, 143 – 149 (1987). C. T. Rubin, S. D. Bain, and K. J. McLeod, Suppression of the osteogenic response in the aging skeleton. Calcif. Tissue Int. 50, 306 – 313 (1992). M. Silbermann, B. Bar-Shira-Maymon, R. Coleman, A. Reznick, Y. Weisman, E. Steinhagen-Thiessen, H. von der Mark, and K. von der Mark, Long-term physical exercise retards trabecular bone loss in lumbar vertebrae of aging female mice. Calcif. Tissue Int. 46, 80 – 93 (1990). E. Steinhagen-Thiessen, A. Reznik, and H. Hilz, Negative adaptation to physical training in senile mice. Mech. Aging. Dev. 12, 231 – 236 (1980). C. H. Turner, Y. Takano, and I. Owan, Aging changes mechanical loading thresholds for bone formation in rats. J. Bone Miner. Res. 10, 1544 – 1549 (1995). A. Z. Reznick, E. Steinhagen-Thiessen, B. Gellersen, and D. Gershon, The effect of short- and long-term exercise on aldolase activity in muscles of CW-1 and C57/BL mice of various ages. Mech. Aging. Dev. 23, 253 – 258 (1983).

718 151. A. Hoshi, H. Watanabe, M. Chiba, and Y. Inaba, Effects of exercise at different ages on bone density and mechanical properties of femoral bone of aged mice. Tohoku. J. Exp. Med. 185, 15 – 24 (1998). 152. J. D. Termine, Cellular activity, matrix proteins, and agin bone. Exp. Gerontol. 25, 217 – 221 (1990). 153. M. R. Benedict, S. Adiyaman, D. C. Ayers, F. D. Thomas, J. D. Calore, V. Dhar, and R. A. Richman, Dissociation of bone mineral density from age-related decreases in insulin-like growth factor-1 and its binding proteins in the male rat. J. Gerontol. 49, B224 – B230 (1994). 154. D. M. Raab, E. L. Smith, T. D. Crenshaw, and D. P. Thomas, Bone mechanical properties after exercise training in young and old rats. J. Appl. Physiol. 68, 130 – 134 (1990). 155. R. E. Beyer, J. C. Huang, and G. B. Wilshire, The effect of endurance exercise on bone dimensions, collagen, and calcium in the aged male rat. Exp. Gerontol. 20, 315 – 323 (1985). 156. K. H. Myburgh, T. D. Noakes, M. Roodt, and F. S. Hough, Effect of exercise on the development of osteoporosis in adult rats. J. Appl Physiol. 66, 14 – 19 (1989). 157. E. I. Barengolts, D. J. Curry, M. S. Bapna, and S. C. Kukreja, Effects of two non-endurance exercise protocols on established bone loss in ovariectomized adult rats. Calcif. Tissue Int. 52, 239 – 243 (1993). 158. Z. Peng, J. Tuukkanen, and H. K. Vaananen, Exercise can provide protection against bone loss and prevent the decrease in mechanical strength of femoral neck in ovariectomized rats. J. Bone Miner. Res. 9, 1559 – 1564 (1994). 159. J. Flieger, T. Karachalios, L. Khaldi, P. Raptou, and G. Lyritis, Mechanical stimulation in the form of vibration prevents postmenopausal bone loss in ovariectomized rats. Calcif. Tissue Int. 63, 510 – 514 (1998). 160. J. Iwamoto, T. Takeda, and S. Ichimura, Effects of moderate intensity exercise on tibial bone mass in mature ovariectomized rats: Bone histomorphometry study. Keio. J. Med. 47, 162 – 167 (1998). 161. L. Moyer-Mileur, M. Luetkemeier, L. Boomer, and G. M. Chan, Effect of physical activity on bone mineralization in premature infants. J. Pediatr. 127, 620 – 625 (1995). 162. B. L. Specker, L. Mulligan, and M. Ho, Longitudinal study of calcium intake, physical activity, and bone mineral content in infants 6-18 months of age. J. Bone Miner. Res. 14, 569 – 576 (1999). 163. M. Bradney, G. Pearce, G. Naughton, C. Sullivan, S. Bass, T. Beck, J. Carlson, and E. Seeman, Moderate exercise during growth in prepubertal boys: changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J. Bone Miner. Res. 13, 1814 – 1821 (1998). 164. F. L. Morris, G. A. Naughton, J. L. Gibbs, J. S. Carlson, and J. D. Wark, Prospective ten-month exercise intervention in premenarcheal girls: Positive effects on bone and lean mass. J. Bone Miner. Res. 12, 1453 – 1462 (1997). 164a. R. Fuchs, J. Bauer, and C. Snow, Jumping improves hip and lumbar spine bone mass in prepubescent children: A randomized controlled trial. J. Bone Miner. Res. 16(1), 148 –156 (2001). 165. K. E. Chad, D. A. Bailey, H. A. McKay, G. A. Zello, and R. E. Snyder, The effect of a weight-bearing physical activity program on bone mineral content and estimated volumetric density in children with spastic cerebral palsy. J. Pediatr. 135, 115 – 117 (1999). 166. B. Gutin, S. Owens, T. Okuyama, S. Riggs, M. Ferguson, and M. Litaker, Effect of physical training and its cessation on percent fat and bone density of children with obesity. Obes. Res. 7, 208 – 214 (1999). 167. K. Witzke and C. Snow, Effects of plyometric jump training on bone mass in adolescent girls. Med. Sci. Sports Exerc. 32, 1051 –1057 (2000).

BECK, SHAW, AND SNOW 168. A. L. Friedlander, H. K. Genant, S. Sadowsky, N. N. Byl, and C. C. Gluer, A two-year program of aerobics and weight-training enhances BMD of young women. J. Bone Miner. Res. 10, 574 (1995). 169. A. Heinonen, H. Sievanen, P. Kannus, P. Oja, and I. Vuori, Effects of unilateral strength training and detraining on bone mineral mass and estimated mechanical characteristics of the upper limb bones in young women. J. Bone Miner. Res. 11, 490 – 501 (1996). 170. T. Lohman, S. Going, R. Pamenter, et al., Effects of resistance training on regional and total bone mineral density in premenopausal women: A randomized prospective study. J. Bone Miner. Res. 10, 1015 (1995). 171. C. Snow-Harter, M. Bouxsein, B. T. Lewis, D. R. Carter, and R. Marcus, Effects of resistance and endurance exercise on bone mineral status of young women: A randomized exercise intervention trial. J. Bone Miner. Res. 7, 761 – 769 (1992). 172. T. M. Dornemann, R. G. McMurray, J. B. Renner, and J. J. Anderson, Effects of high-intensity resistance exercise on bone mineral density and muscle strength of 40 – 50 – year-old women. J. Sports Med. Phys. Fitness 37, 246 – 251 (1997). 173. A. Heinonen, P. Oja, H. Sievanen, M. Pasanen, and I. Vuori, Effect of two training regimens on bone mineral density in healthy perimenopausal women: A randomized controlled trial. J. Bone Miner. Res. 13, 483 – 490 (1998). 174. J. Y. Margulies, A. Simkin, I. Leichter, A. Bivas, R. Steinberg, M. Giladi, M. Stein, H. Kashtan, and C. Milgrom, Effect of intense physical activity on the bone-mineral content in the lower limbs of young adults. J. Bone Joint Surg. 68, 1090 – 1093 (1986). 175. J. A. Williams, J. Wagner, R. Wasnich, and L. Heilbrun, The effect of long-distance running upon appendicular bone mineral content. Med. Sci. Sports Exerc. 16, 223 – 227 (1984). 176. L. A. Pruitt, R. D. Jackson, R. L. Bartells, and H. J. Lehnhard, Weight-training effects on bone mineral density in early postmenopausal women. J. Bone Miner. Res. 7, 179 – 185 (1992). 177. K. A. Grove and B. R. Londeree, Bone density in postmenopausal women: High impact vs low impact exercise. Med. Sci. Sports Exerc. 24, 1190 – 1194 (1992). 178. W. M. Kohrt, A. A. Ehsani, and S. J. Birge Jr., Effects of exercise involving predominantly either joint-reaction or ground-reaction forces on bone mineral density in older women. J. Bone Miner. Res. 12, 1253 – 1261 (1997). 179. M. Revel, M. A. Mayoux-Benhamou, J. P. Rabourdin, F. Bagheri, and C. Roux, One-year psoas training can prevent lumbar bone loss in postmenopausal women: A randomized controlled trial. Calcif. Tissue Int. 53, 307 – 311 (1993). 180. G. L. Smidt, S. Y. Lin, K. D. O’Dwyer, and P. R. Blanpied, The effect of high-intensity trunk exercise on bone mineral density of postmenopausal women. Spine 17, 280 – 285 (1992). 181. D. Kerr, A. Morton, I. Dick, and R. Prince, Exercise effects on bone mass in postmenopausal women are site-specific and load-dependent. J. Bone Miner. Res. 11, 218 – 225 (1996). 182. M. Sinaki, H. W. Wahner, K. P. Offord, and S. F. Hodgson, Efficacy of nonloading exercises in prevention of vertebral bone loss in postmenopausal women: A controlled trial. Mayo Clin. Proc. 64, 762 – 769 (1989). 183. E. J. Bassey and S. J. Ramsdale, Weight-bearing exercise and ground reaction forces: A 12-month randomized controlled trial of effects on bone mineral density in healthy postmenopausal women. Bone 16, 469 – 476 (1995). 184. L. Welsh and O. M. Rutherford, Hip bone mineral density is improved by high-impact aerobic exercise in postmenopausal women and men over 50 years. Eur. J. Appl. Physiol. Occup. Physiol. 74, 511 – 517 (1996). 185. G. P. Dalsky, K. S. Stocke, A. A. Ehsani, E. Slatopolsky, W. C. Lee, and S. J. Birge Jr., Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann. Intern. Med. 108, 824 – 828 (1988).

CHAPTER 28 Physical Activity and Osteoporosis 186. M. Hatori, A. Hasegawa, H. Adachi, A. Shinozaki, R. Hayashi, H. Okano, H. Mizunuma, and K. Murata, The effects of walking at the anaerobic threshold level on vertebral bone loss in postmenopausal women. Calcif. Tissue Int. 52, 411 – 414 (1993). 187. B. Krolner, B. Toft, S. P. Nielsen, and E. Tondevold, Physical exercise as prophylaxis against involutional vertebral bone loss: A controlled trial. Clin. Sci. 64, 54 1– 546 (1983). 188. A. Rundgren, A. Aniansson, P. Ljungberg, and H. Wetterqvist, Effects of a training programme for elderly people on mineral content of the heel bone. Arch. Gerontol. Geriatr. 3, 243 – 248 (1984). 189. S. Kudlacek, F. Pietschmann, P. Bernecker, H. Resch, and R. Willvonseder, The impact of a senior dancing program on spinal and peripheral bone mass. Am. J. Phys. Med. Rehabil. 76, 477–481 (1997). 190. D. J. Cavanaugh and C. E. Cann, Brisk walking does not stop bone loss in postmenopausal women. Bone 9, 201 – 204 (1988). 191. G. Dilsen, C. Berker, A. Oral, and G. Varan, The role of physical exercise in prevention and management of osteoporosis. Clin. Rheumatol. 8 (Suppl. 2), 70 – 75 (1989). 192. G. Bravo, P. Gauthier, P. M. Roy, H. Payette, and P. Gaulin, A weight-bearing, water-based exercise program for osteopenic women: Its impact on bone, functional fitness, and well-being. Arch. Phys. Med. Rehabil. 78, 1375 – 1380 (1997). 193. J. M. Shaw and C. M. Snow, Weighted vest exercise improves indices of fall risk in older women. J. Gerontol. A Biol. Sci. Med. Sci. 53, M53 – 58 (1998). 194. J. Ayalon, A. Simkin, I. Leichter, and S. Raifmann, Dynamic bone loading exercises for postmenopausal women: Effect on the density of the distal radius. Arch. Phys. Med. Rehabil. 68, 280 – 283 (1987). 195. I. Tsuritani, K. S. Brooke-Wavell, S. S. Mastana, P. R. Jones, A. E. Hardman, and Y. Yamada, Does vitamin D receptor polymorphism influence the response of bone to brisk walking in postmenopausal women? Horm. Res. 50, 315 – 319 (1998). 196. T. L. Jarvinen, T. A. Jarvinen, H. Sievanen, et al., Vitamin D receptor alleles and bone’s response to physical activity. Calcif. Tissue Int. 62, 413 – 417 (1998). 197. M. Notelovitz, D. Martin, R. Tesar, F. Y. Khan, C. Probart, C. Fields, and L. McKenzie, Estrogen therapy and variable-resistance weight training increase bone mineral in surgically menopausal women. J. Bone Miner. Res. 6, 583 – 590 (1991). 198. W. M. Kohrt, A. A. Ehsani, and S. J. Birge Jr., HRT preserves increases in bone mineral density and reductions in body fat after a supervised exercise program. J. Appl. Physiol. 84, 1506 – 1512 (1998). 199. J. Heikkinen, E. Kurttila-Matero, E. Kyllonen, J. Vuori, T. Takala, and H. K. Vaananen, Moderate exercise does not enhance the positive effect of estrogen on bone mineral density in postmenopausal women. Calcif. Tissue Int. 49, S83 – 84 (1991). 200. B. L. Specker, Evidence for an interaction between calcium intake and physical activity on changes in bone mineral density. J. Bone Miner. Res. 11, 1539 – 1544 (1996). 201. B. Kanders, D. W. Dempster, and R. Lindsay, Interaction of calcium nutrition and physical activity on bone mass in young women. J. Bone Miner. Res. 3, 145 – 149 (1988). 202. J. Iwamoto, T. Takeda, T. Otani, and Y. Yabe, Effect of increased physical activity on bone mineral density in postmenopausal osteoporotic women. Keio J. Med. 47, 157 – 161 (1998). 203. R. Prince, A. Devine, I. Dick, A. Criddle, D. Kerr, N. Kent, R. Price, and A. Randall, The effects of calcium supplementation (milk powder or tablets) and exercise on bone density in postmenopausal women. J. Bone Miner. Res. 10, 1068 – 1075 (1995). 204. T. L. Jarvinen, P. Kannus, H. Sievanen, P. Jolma, A. Heinonen, and M. Jarvinen, Randomized controlled study of effects of sudden impact loading on rat femur. J. Bone Miner. Res. 13, 1475 – 1482 (1998).

719 205. S. Adami, D. Gatti, V. Braga, D. Bianchini, and M. Rossini, Sitespecific effects of strength training on bone structure and geometry of ultradistal radius in postmenopausal women. J. Bone Miner. Res. 14, 120 – 124 (1999). 206. H. Krahl, U. Michaelis, H.-G. Pieper, G. Quack, and M. Montag, Stimulation of bone growth through sports. A radiologic investigation of the upper extremities in professional tennis players. Am. J. Sports Med. 22, 751 – 757 (1994). 207. A. B. Loucks, G. A. Laughlin, J. F. Mortola, L. Girton, J. C. Nelson, and S. S. Yen, Hypothalamic-pituitary-thyroidal function in eumenorrheic and amenorrheic athletes. J. Clin. Endocrinol. Metab. 75, 514 – 518 (1992). 208. A. B. Loucks and E. M. Heath, Induction of low-T3 syndrome in exercising women occurs at a threshold of energy availability. Am. J. Physiol. 266, R817 – 823 (1994). 209. S. E. Tomten, J. A. Falch, K. I. Birkeland, P. Hemmersbach, and A. T. Hostmark, Bone mineral density and menstrual irregularities: A comparative study on cortical and trabecular bone structures in runners with alleged normal eating behavior. Int. J. Sports Med. 19, 92 – 97 (1998). 209a. B. L. Drinkwater, K. Nilson, C. H. Chestnut, W. J. Bremner, S. Shainholtz, and M. B. Southworth, Bone mineral content of amenorrheic and eumenorrheic athletes. New Eng. J. Med. 311(5), 277 – 281 (1984). 209b. R. Marcus, C. Cann, P. Madvig, J. Minkoff, M. Goddard, M. Bayer et al., Menstrual function and bone mass in elite women distance runners: Endocrine and metabolic features. Ann. Intern. Med. 102(2), 158 –163 (1985). 210. N. W. Constantini and M. P. Warren, Special problems of the female athlete. Baillieres Clin. Rheumatol. 8, 199 – 219 (1994). 211. M. Hartard, P. Bottermann, P. Bartenstein, D. Jeschke, and M. Schwaiger, Effects on bone mineral density of low-dosed oral contraceptives compared to and combined with physical activity. Contraception 55, 87 – 90 (1997). 212. A. D. Keen and B. L. Drinkwater, Irreversible bone loss in former amenorrheic athletes. Osteopor. Int. 7, 311 – 315 (1997). 213. S. E. MacConnie, A. Barkan, R. M. Lampman, M. A. Schork, and I. Z. Beitins, Decreased hypothalmic gonadotrophin-releasing hormone secretion in male marathon runners. N. Engl. J. Med. 315, 411 – 417 (1986). 214. M. L. Hetland, J. Haarbo, and C. Christiansen, Low bone mass and high bone turnover in male long distance runners. J. Clin. Endocrinol. Metab. 77, 770 – 775 (1993). 215. T. W. Rowland, A. H. Morris, J. F. Kelleher, B. L. Haag, and E. O. Reiter, Serum testosterone response to training in adolescent runners. Am. J. Dis. Child. 141, 881 – 883 (1987). 216. G. D. Wheeler, S. R. Wall, A. N. Belcastro, and D. C. Cumming, Reduced serum testosterone and prolactin levels in male distance runners. J. Am. Med. Assoc. 252, 514 – 516 (1984). 217. A. C. Hackney, W. E. Sinning, and B. C. Bruot, Reproductive hormonal profiles of endurance-trained and untrained males. Med. Sci. Sports Exerc. 20, 60 – 65 (1988). 218. C. S. Cooper, D. R. Taaffe, D. Guido, E. Packer, L. Holloway, and R. Marcus, Relationship of chronic endurance exercise to the somatotropic and sex hormone status of older men. Eur. J. Endocrinol. 138, 517 – 523 (1998). 219. C. E. Fiore, E. Cottini, C. Fargetta, G. di Salvio, R. Foti, and M. Raspagliesi, The effects of muscle-building exercise on forearm bone mineral content and osteoblast activity in drug-free and anabolic steroids self-administering young men. Bone Miner. 13, 77 – 83 (1991). 220. K. E. Yarasheski, J. A. Campbell, and W. M. Kohrt, Effect of resistance exercise and growth hormone on bone density in older men. Clin. Endocrinol. 47, 223 – 229 (1997). 221. D. R. Taaffe, L. Pruitt, J. Riem, R. L. Hintz, G. Butterfield, A. R. Hoffmen, and R. Marcus, Effect of recombinant human growth

720

222.

223. 224.

225. 226. 227. 228. 229. 230.

231. 232.

233.

234.

235.

236. 237.

238.

239. 240.

241.

242.

243.

244.

BECK, SHAW, AND SNOW hormone on the muscle strength response to resistance exercise in elderly men. J. Clin. Endocrinol. and Metab. 79, 1361 – 1366 (1994). D. R. Taaffe, I. H. Jin, T. H. Vu, A. R. Hoffman, and R. Marcus, Lack of effect of recombinant growth hormone (GH) on muscle morphology and GH-insulin-like growth factor expression in resistancetrained elderly men. J. Clin. Endocrinol. Metab. 81, 421 – 425 (1996). C. Snow-Harter and R. Marcus, Exercise, bone mineral density, and osteoporosis. Exerc. Sport Sci. Rev. 19, 351 – 388 (1991). A. Berard, G. Bravo, and P. Gauthier, Meta-analysis of the effectiveness of physical activity for the prevention of bone loss in postmenopausal women. Osteopor. Int. 7, 331 – 337 (1997). B. L. Drinkwater, Exercise in the prevention of osteoporosis. Osteopor. Int. 3, 169 – 171 (1993). E. Ernst, Exercise for female osteoporosis. A systematic review of randomised clinical trials. Sports Med. 25, 359 – 368 (1998). M. R. Forwood and D. B. Burr, Physical activity and bone mass: Exercises in futility? Bone Miner. 21, 89 – 112 (1993). B. Gutin and M. J. Kasper, Can vigorous exercise play a role in osteoporosis prevention? A review. Osteopor. Int. 2, 55 – 69 (1992). G. A. Kelley, Aerobic exercise and bone density at the hip in postmenopausal women: A meta-analysis. Prev. Med. 27, 798 – 807 (1998). G. A. Kelley, Exercise and regional bone mineral density in postmenopausal women: A meta-analytic review of randomized trials. Am. J. Phys. Med. Rehabil. 77, 76 – 87 (1998). J. M. Lane and M. Nydick, Osteoporosis: Current modes of prevention and treatment. J. Am. Acad. Orthop. Surg. 7, 19 – 31 (1999). R. Marcus, B. Drinkwater, G. Dalsky, J. Dufek, D. Raab, C. Slemenda, and C. Snow-Harter, Osteoporosis and exercise in women. Med. Sci. Sports Exerc. 24, S301 – 307 (1992). I. Wolff, J. J. van Croonenborg, H. C. Kemper, P. J. Kostense, and J. W. Twisk, The effect of exercise training programs on bone mass: A meta-analysis of published controlled trials in pre- and postmenopausal women. Osteopor. Int. 9, 1 – 12 (1999). American College of Sports Medicine Position Stand, Exercise and physical activity for older adults. Med. Sci. Sports Exerc. 30, 992 – 1008 (1998). R. H. Whipple, L. I. Wolfson, and P. M. Amerman, The relationship of knee and ankle weakness to falls in nursing home residents: An isokinetic study. J. Am. Geriatr. Soc. 35, 13 – 20 (1987). S. H. Allen, Exercise considerations for postmenopausal women with osteoporosis. Arthritis Care Res. 7, 205 – 214 (1994). D. Taaffe, C. Duret, S. Wheeler, and R. Marcus, Once-weekly resistance exercise improves muscle strength and neuromuscular performance in older adults. J. Am. Geriatr. Soci. 47, 1208 – 1214 (1999). E. W. Gregg, J. A. Cauley, D. G. Seeley, K. E. Ensrud, and D. C. Bauer, Physical activity and osteoporotic fracture risk in older women: Study of Osteoporotic Fractures Research Group. Ann Intern Med 129, 81 – 88 (1998). S. B. Jaglal, N. Kreiger, and G. Darlington, Past and recent physical activity and risk of hip fracture. Am. J. Epidemiol. 138, 107–118 (1993). S. B. Jaglal, N. Kreiger, and G. A. Darlington, Lifetime occupational physical activity and risk of hip fracture in women. Ann. Epidemiol. 5, 321 – 324 (1995). A. V. Schwartz, J. L. Kelsey, S. Sidney, and J. A. Grisso, Characteristics of falls and risk of hip fracture in elderly men. Osteopor. Int. 8, 240 – 246 (1998). R. Chow, J. Harrison, and J. Dornan, (1989). Prevention and rehabilitation of osteoporosis program: exercise and osteoporosis. Int J Rehabil Res 12, 49 – 56. J. A. Stevens, K. E. Powell, S. M. Smith, P. A. Wingo, and R. W. Sattin, Physical activity, functional limitations, and the risk of fall-related fractures in community-dwelling elderly. Ann. Epidemiol. 7, 54– 61 (1997). J. Grisso, E. Capezuti, and A. Schwartz, Falls as risk factors for fracture. In “Osteoporosis,” pp. 599 – 608. Academic Press, San Diego.

245. T. W. O’Neill, D. Marsden, J. E. Adams, and A. J. Silman, Risk factors, falls, and fracture of the distal forearm in Manchester, UK. J. Epidemiol. Community Health 50, 288 – 292 (1996). 246. M. J. Pavol, T. M. Owings, K. T. Foley, and M. D. Grabiner, The sex and age of older adults influence the outcome of induced trips. J. Gerontol. A Biol. Sci. Med. Sci. 54, M103 – 108 (1999). 247. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. D. Melton, Incidence of clinically diagnosed vertebral fractures A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 248. R. Norton, A. J. Campbell, T. Lee-Joe, E. Robinson, and M. Butler, Circumstances of falls resulting in hip fractures among older people. J. Am. Geriatr. Soc. 45, 1108 – 1112 (1997). 249. S. R. Cummings, D. M. Black, M. C. Nevitt, et al., Appendicular bone density and age predict hip fracture in women: The Study of Osteoporotic Fractures Research Group. JAMA 263, 665 – 668 (1990). 250. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 251. W. C. Hayes, E. R. Myers, J. N. Morris, T. N. Gerhart, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 252. A. Aniansson, C. Zetterberg, M. Hedberg, and K. G. Henriksson, Impaired muscle function with aging. A background factor in the incidence of fractures of the proximal end of the femur. Clin. Orthop. 193 – 201 (1984). 253. P. Dargent-Molina, F. Favier, H. Grandjean, C. Baudoin, A. M. Schott, E. Hausherr, P. J. Meunier, and G. Breart, Fall-related factors and risk of hip fracture: The EPIDOS prospective study [published erratum appears in Lancet 10, 348(9024),416 [1996] Lancet 348, 145 – 149. 254. B. E. Maki, P. J. Holliday, and A. K. Topper, A prospective study of postural balance and risk of falling in an ambulatory and independent elderly population. J. Gerontol. 49, M72 – 84 (1994). 255. A. J. Campbell, M. C. Robertson, M. M. Gardner, R. N. Norton, M. W. Tilyard, and D. M. Buchner, Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. Br. Med. J. 315, 1065 – 1069 (1997). 256. M. A. Fiatarone, E. C. Marks, N. D. Ryan, C. N. Meredith, L. A. Lipsitz, and W. J. Evans, High-intensity strength training in nonagenarians: Effects on skeletal muscle. JAMA 263, 3029 – 3034 (1990). 257. J. M. Chandler, P. W. Duncan, G. Kochersberger, and S. Studenski, Is lower extremity strength gain associated with improvement in physical performance and disability in frail, community-dwelling elders? Arch Phys. Med. Rehabil. 79, 24 – 30 (1998). 258. L. Wolfson, R. Whipple, C. Derby, J. Judge, M. King, P. Amerman, J. Schmidt, and D. Smyers, Balance and strength training in older adults: Intervention gains and Tai Chi maintenance. J. Am. Geriatr. Soc. 44, 498 – 506 (1996). 259. K. Protiva, S. Macdonald, K. Winters, and C. Snow, Effects of lower extremity resistance exercise on indices of fracture risk in frail elderly. J. Bone Miner. Res. 12, T567 (1997). 260. S. R. Lord, J. A. Ward, P. Williams, and M. Strudwick, The effect of a 12-month exercise trial on balance, strength, and falls in older women: A randomized controlled trial. J. Am. Geriatr. Soc. 43, 1198 – 1206 (1995). 261. L. Wolfson, J. Judge, R. Whipple, and M. King, Strength is a major factor in balance, gait, and the occurrence of falls. J. Gerontol. A Biol. Sci. Med. Sci. 50, 64 – 67 (1995). 262. C. M. Ulrich, C. C. Georgiou, D. E. Gillis, and C. M. Snow, Lifetime physical activity is associated with bone mineral density in premenopausal women. J. Womens Health 80, 365 – 375 (1999).

CHAPTER 29

Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis MARYFRAN SOWERS

I. II. III. IV. V. VI.

Department of Epidemiology, University of Michigan, School of Public Health, Ann Arbor, Michigan 48109

Introduction Pregnancy Age at First Pregnancy Parity and Nulliparity Lactation Ovarian Activity or Menstrual Cycle Characteristics and Bone Mass

I. INTRODUCTION Risk of fracture in older women is, in large part, related to the women’s bone mineral density (BMD) [1]. Although osteoporosis primarily affects older persons, predisposition to the condition is established in young adulthood. In the adult, bone tissue undergoes a continuous remodeling process consisting of periods of bone resorption and formation with maximal bone accretion, thought to occur in the early third decade, followed by relatively stable bone mineral density until the menopause transition [2]. Stochastic models developed by Horsman and Burkinshaw [3] suggested that two-thirds of the risk for fracture can be predicted based on premenopausal BMD. Therefore, in premenopausal women, the primary goal is to maximize or maintain bone mineral density. The World Health Organiza-

VII. VIII. IX. X. XI.

Dysfunctional Ovulation Oral Contraceptive Use Progestin-Injectible Contraceptives Oophorectomy Summary and Implications References

data from women aged 20 to 40 years. By the WHO [4] definition, osteoporosis exists when bone mineral density is 2.5 standard deviations (SD) below the mean values for women aged 20 to 40 years. Greater acquisition and longer maintenance of premenopausal bone mass establish a larger bone mineral reserve that could ultimately reduce the risk for osteoporosis and fracture following menopause. Identifying those factors related to the accrual, maintenance, or diminution of bone is important as there are, at present, no means of restoring lost bone. Reproductive activities and the hormones associated with reproduction may play a central role in bone mineral density levels during pre- and perimenopause. In this chapter, those endogenous and exogenous events that are related directly or indirectly to the capacity to reproduce will be considered for their importance to peak bone mass. In par-

722

MARYFRAN SOWERS

mechanisms of bone loss in lactation, the importance of injectible contraceptive, and the role of hormone variation in association with peak bone mass.

II. PREGNANCY Pregnancy and lactation are characterized by significant alterations in the maternal hormone environment, notably estrogen and prolactin concentrations. During the third trimester of pregnancy, estrogen levels rise as the placenta produces large quantities of estradiol to facilitate growth [5]. In marked contrast, lactation represents a hypoestrogenic state and elevated prolactin concentrations [6]. These events are associated with a substantial calcium transfer from the mother for redistribution to the fetus or infant. The total accumulation of calcium in a full-term neonate during pregnancy is approximately 30 g [7]. If maternal bone was the sole source of calcium, the mother’s skeleton would lose about 3% (30 g/1000 g) of its mineral per pregnancy. It has been unclear whether the size of the maternal bone depot is reduced during pregnancy. Typically, less than 20 – 30% of ingested dietary calcium is actually absorbed in an adult woman and the remaining calcium is excreted in the feces. If absorption efficiency could be doubled from 20 to 40% in women consuming moderate calcium intakes, the skeletal needs of the fetus could be met without extensively accessing the mineral stored in the maternal skeleton. Likewise, reducing maternal urinary calcium excretion could potentially also allow the demands of the mother and child to be met without influencing the maternal bone depot. A number of metabolic adaptations take place early in pregnancy to address the demands of the fetus. These adaptations include an increase in intestinal calcium absorptive capacity in response to an oral calcium load [8,9]; a slowing of gastric motility; an increase in renal function; an increase in the extracellular fluid volume; an increase in urinary calcium excretion [10,11] and a modest decline in serum calcium concentrations in the second trimester [11,12], apparently in parallel with the decline in serum albumin [10,12 – 14]. The collective response appears to be a state of positive calcium balance [15 – 17] even in adolescent pregnant women [15].

A. Studies of Bone Mass and Pregnancy Until recently, studies of bone mass and pregnancy suggested either no measurable bone mass loss with pregnancy [18 – 21] or bone loss in specific compartments (trabecular rather than cortical) [22] or only at selected bone sites [23]. With rare exception, these studies have substantial methodological and technical limitations. Typically, a sample size

X-ray absorptiometry technology to have sufficient power to detect a 3 – 4% difference in bone mass change, if it exists. Several of these studies were performed with insensitive techniques that would necessitate even larger sample sizes to detect the small bone change that would occur in a short time interval. Not only were earlier studies limited by methodological problems, there were a substantial number of related unanswered questions. For example, there are unanswered questions as to whether women who are culturally or racially diverse have similar bone change responses with pregnancy, particularly in a calcium-deficient maternal environment. Studies have not addressed the issue of age (adolescent pregnancy or pregnancy at obstetric maturity) and the potential for women in these groups to have different calcium needs and a different responsiveness of bone to the calcium demand of pregnancy on bone. There is a growing availability of bone ultrasound technology to measure bone status in pregnancy, a technology whose measurement was recently validated in a pregnant population [24]. This will facilitate new studies with large samples of culturally diverse women and allow the inclusion of adolescents in study samples. For example, Gambacciani et al. [25] reported a lower maternal bone mass in the last trimester with ultrasound, whereas Yamaga et al. [26] have published that there is no loss of bone mass by ultrasound among Japanese women.

B. Studies of Bone and Pregnancy Using Biochemical Markers The most frequently characterized bone turnover markers measured in pregnancy include circulating osteocalcin and alkaline phosphatase concentrations as indicators of bone formation. Markers of formation and resorption typically have not been reported simultaneously to more fully characterize the bone turnover experience. Serum concentrations of osteocalcin tend to be comparable to control values in the first trimester, decline in the second trimester of the pregnancy, and then recover in the third trimester to levels observed in normal nonpregnant controls. This has been observed in studies with repeated measures [27 – 28] or static comparisons [8,29]. Rodin and colleagues [28] observed that concentrations were within normal range within 48 h of delivery. Like osteocalcin, alkaline phosphatase has been evaluated as a marker of bone formation during pregnancy. Total serum alkaline phosphatase activity (ALK) increases gradually in the first and second trimesters with a rapid increase in the third trimester [12,14,28,30]. Rodin et al. [28] reported that both placental and bone-specific alkaline phosphatase isoenzyme patterns replicate the pattern seen in

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

Additionally, they documented that placental ALK declines to levels observed in the first trimester by 6 months postpartum; however, both total and bone-specific alkaline phosphatase are elevated at 6 weeks postpartum. In women who are lactating, activity remains elevated.

C. Studies of Pregnancy and Calciotropic Hormones There have been numerous studies of the calciotropic hormones [parathyroid hormone (PTH), 1,25-dihydroxyvitamin D, and calcitonin] during pregnancy, and these are reviewed by Verhaeghe and Bouillon [31], Chesney et al. [32], Sowers [33], and Kovacs and Kronenberg [34]. Parathyroid hormone promotes increased calcium mobilization from bone in response to lower levels of circulating calcium concentrations. Initially, pregnancy was regarded as a state of “physiologic hyperparathyroidism” as pregnancy was associated with an increase [29,30,35 – 37] in PTH concentrations. More recent studies, using more specific assays, challenged this concept and reported either no significant elevation of PTH with gestation [11] or a decrease in parathyroid hormone [37 – 41] relative to nonpregnant controls. The studies have generally not addressed dietary calcium intake, vitamin D status, or other factors that could, theoretically, influence PTH secretion. While the concept of “physiologic hyperparathyroidism” has been eclipsed at the present time, there is still the potential for functional hyperparathyroidism to exist in the absence of elevated PTH levels. Another agent, parathyroidrelated peptide (PTHrP), with sequence homology similar to PTH, has been described as higher in pregnant women as compared to nonpregnant controls. In pregnancy, PTHrP appears to play multiple roles, including promoting maternal – fetal calcium transfer and milk production [42 – 45]. 1,25-Dihydroxyvitamin D concentrations rise during pregnancy [12,14,27,38,46] and are believed to be responsible for the enhanced absorption of dietary calcium [37, 38]. Those factors that regulate the hormone during pregnancy are uncertain, although PTH, growth hormone, prolactin, and estrogen have all been suggested as candidates [47]. It is not known whether the increased 1,25-dihydroxyvitamin D levels arise from the placenta. Alterations in the levels of 1,25-dihydroxyvitamin D are not associated with a similar pattern in the levels of 25-hydroxyvitamin D [47]. Investigators have observed the same seasonal patterns in 25-hydroxyvitamin D levels in pregnant women that have been reported in nonpregnant women [14,36]. While it might be hypothesized that calcitonin concentrations should rise during pregnancy to protect the maternal skeleton from resorption, findings from the few studies

723

inconsistent. For example, Stevenson et al. [48] and Whitehead et al. [40], in cross-sectional studies, reported an increase in calcitonin in pregnant vs nonpregnant women. Pitkin and colleagues [13] reported at least six different calcitonin patterns when multiple measures were made on study participants. Stevenson and associates [48] observed no difference in calcitonin values between pregnant and lactating women. Synthesis of this information is difficult in that markedly different assays were used in these studies, limiting comparability. Additionally, there is now question as to how these assays relate to contemporary and specific calcitonin assays.

D. Summary and Implications Initial studies of bone change in pregnancy have not provided strong evidence of bone loss with pregnancy, although many of these studies had important design limitations that made consideration of their contribution quite tenuous. It is anticipated that future studies of bone and pregnancy will involve the use of bone ultrasound. This will expand the opportunity for studies using technology that no longer requires exposure of mother and infant to ionizing radiation. It is widely thought that pregnancy may not have a substantial impact on maternal calcium stores. The fetal demand is not extensive (30 g), and evidence shows that the calcium demands of the human fetus could be met through adaptive mechanisms, including higher circulating levels of 1,25-dihydroxyvitamin D and icreased intestinal absorption efficiency. Future work, using bone ultrasound technology, may challenge this assessment and suggest that there is indeed measurable bone loss with pregnancy. This would justify additional investigation to validate these findings and examine ultrasound measures into the postpartum period.

III. AGE AT FIRST PREGNANCY Excess bone resorption with pregnancy may not be a characteristic of the mature woman who has achieved full maximal bone mass. However, evidence shows that pregnancy at an earlier age, when the skeleton of both fetus and mother are maturing simultaneously, may result in lower bone density and increased risk for perimenopausal bone loss. Sowers et al. [49] observed cross-sectionally that a first pregnancy during adolescence was associated with lower premenopausal radial BMD. A subsequent longitudinal study [50] showed that parous women whose first pregnancy was before age 20 had significantly lower age-adjusted baseline radial BMD, lower follow-up radial BMD,

724

MARYFRAN SOWERS

confirmed by Fox et al. [51] in a cross-sectional study of about 1800 elderly women. The investigators speculated that the hormonal events of pregnancy during adolescence may jeopardize achieving the maximal peak in bone mineralization. Presently, studies are underway to measure bone in pregnant adolescents, taking into account whether the adolescent pregnant women are still growing or not. This greater specificity may help address the true underlying questions of biologic age and bone mineralization as compared to chronological age and bone mineralization. These studies will help establish the relevance and mechanisms of earlier age at first pregnancy as a risk factor for lower bone mineral density.

IV. PARITY AND NULLIPARITY The relationship of parity to bone mass is complex and poorly defined. Theoretically, bone mass may decrease because of the calcium demand of pregnancy. In contrast, bone mass may increase with the greater circulating estrogen levels in the third trimester of pregnancy and because of the increased bone loading that occurs with the weight increases in pregnancy. With the uncertain impact of parity on bone mass, it is a logical extension that the impact of parity on fracture is also ill-defined.

A. Studies of Parity and Bone Mass A number of studies have reported an increase in bone mass with parity, as measured with different measurement technologies at different bone sites with different parity classifications [51 – 55]. Other studies have found no association in studies of premenopausal [49,56,57] or postmenopausal women [58 – 61]. For example, in studies of Caucasian and Bantu women, Walker et al. [58] found no difference in metacarpal cortical area of women, aged 30 – 44, who had zero to one child as compared to those with more than six children. Likewise, Kritz-Silverstein et al. [62] reported no association with increasing number of pregnancies in women aged 60 – 89 years. Hreshchyshyn et al. [59] reported that BMD of the femoral neck declined with increasing number of live births, whereas there was no change in the lumbar spine. Studies of parity and bone mass may have inconsistent findings because at least three factors may differ from one study population to another. These include differences in the ability to conceive, differences in the ability to maintain a viable fetus to term, and differences in the amount of weight gained during and subsequently retained following pregnancy. Successful conception and pregnancy require distinct hormonal environments. To conceive, the hormonal

preparation of the endometrial bed and development of the ovum. Bone mass measured in nulliparous women may not be the appropriate comparison to bone mass in parous women. Nulliparous women include those who lack reproductive competence, those who do not have the opportunity to conceive, and those who do not want to conceive. The lack of reproductive competence may be related to lower bone mass. Likewise, among those who do not want to conceive, the use of selected contraceptive preparations may be associated with lower bone mass, particularly if their use was begun during adolescence and prior to reaching peak bone mineralization. Studies evaluating the bone density of nulliparous women reinforce the notion that they are inappropriate controls for studies of parity. In a longitudinal study of premenopausal women, Sowers et al. [50] found that nulliparity was highly predictive of reduced radial BMD, but not rate of change after controlling for age and body size. There was no relation between number of children and radial BMD when nulliparous women were not used as the referent group. Fox et al. [51] also identified that nulliparous women had significantly lower bone density of the distal radius in the postmenopausal women enrolled in the Baltimore center of the Study of Osteoporotic Fractures. The lower radial BMD in nulliparous women suggests that their risk may be associated with an inability to conceive or maintain a pregnancy. The number of “lost’” pregnancies can further complicate our understanding of the role for live births and bone mass. When pregnancies are terminated, either spontaneously or medically, before 26 weeks of gestation, the contribution of estrogen from the placenta is diminished. Hence, numerous early abortions (induced) may not affect bone status; however, the inability to continue a viable pregnancy (spontaneous abortion) may reflect a marginal hormonal environment that could be associated with lower bone mineral density. As such, careful interpretation of parity data is required if nulliparous women are an integral part of the reference population. Evaluation of parity in future studies should also include adjustment for confounders such as age of the mother and change in weight over time.

B. Studies of Parity and Fracture A longitudinal study [63] and a case-control study [64] provide evidence of a protective effect for parity in relation to hip fracture. In both studies, women with three or more children had an approximate 30 to 40% reduction in risk for fracture as compared to nulliparous women. While both studies addressed the contribution of other major potential confounders, the comparison groups, in both instances, were nulliparous women, groups whose biology may carry

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

have lower bone mineral density, they are likely to have a greater risk of fracture and a parous group using them as a reference would appear to have an inappropriately reduced risk for fracture. Numerous studies have shown no association of parity with fracture. The studies, in widely diverse populations, include hospital-based case-control studies in Connecticut [65] and Toronto [66], a population-based case-control study in Seattle, Washington [67], a case-control study of older women in southwest France [58], and a populationbased case-control study in Australia [69].

C. Summary and Implications It appears that if there is a protective effect of parity against fractures, mediated through greater bone mass, this effect is weak. A stronger case for a protective effect could be made for parity if the studies of both bone mass and fractures had used women with a single pregnancy as the comparison group and evaluated the likelihood of a “dose response” with succeeding numbers of children.

V. LACTATION A. Calcium Demand and Ovarian Suppression by Lactation At least two events that occur during lactation may have an impact on bone mass, including increased calcium demand and suppression of the hypothalamic – pituitary – ovarian (HPO) axis. There is substantial potential for significant calcium demand from the maternal skeleton. Mobilization of calcium from the maternal skeleton will be more highly variable than maternal skeletal mobilization in pregnancy, if it occurs, and the degree of calcium mobilization is dependent on the amount of breast milk produced and on the duration of the lactation period. An estimated cost to the maternal skeleton with 6 months of full lactation would be approximately 4 – 6% if no compensatory mechanism(s) existed for increasing calcium availability apart from mobilization of the skeletal depot. Calcium is transferred directly from serum to breast milk. It is estimated that approximately 600 ml/day of milk is produced at 3 months following parturition (168 mg calcium/day) and 1 liter of milk is produced per day at 6 months following parturition (280 mg calcium/day). The calcium concentration of milk is regulated and appears to be somewhat constant even in the face of variable maternal calcium intake. However, there is some debate as to the potential for the lower calcium content of breast milk in women with very low calcium intakes, as evidenced when

725

[70]. It was initially assumed that there was an increase in efficiency of calcium absorption in lactation, parallel to that observed in pregnancy. However, several studies, but not all [71], have reported that lactation is not associated with increased absorption efficiency [72,73]. In addition to the calcium demand with lactation, the hypothalamic – pituitary axis is suppressed in breast-feeding, as evidenced by the lack of luteinizing hormone release following administration of an estrogen challenge to lactating women [74]. Elevated prolactin concentrations associated with lactation inhibit pulsatile pituitary gonadotrophic hormone secretion, suppress the positive feedback effects of estrogens, interfere with ovarian steroidogenesis, and induce ovarian refractoriness to gonadotropic stimulation [75]. Women with prolactin-secreting adenomas also illustrate the negative impact of nonlactational-elevated prolactin on bone demineralization [76,77].

B. Studies of Bone Mass and Lactation Studies of bone mass published between 1960 and 1990 had mixed findings with respect to the impact of lactation. Various studies suggested bone loss with lactation, no significant negative effect of lactation on subsequent bone mass or fractures, and even a rise in bone density with lactation. However, findings from longitudinal studies and clinical trials [78] have consistently shown significant early losses of bone mineral density at the spine and hip in amounts of 5 – 7% of the total BMD [79 – 81]. The findings are also reported in animal studies [82]. Importantly, however, several of these studies have also documented that the bone mineral is largely restored in the 6- to 12-month period following weaning, as menses are reestablished [83]. Sowers et al. [84] have reported that women who have lost bone mass during lactation appear to continue recovery during a subsequent pregnancy occurring within 18 months of the previous pregnancy. These changes in calcium homeostasis appear to be independent of lifestyle, including dietary calcium intake and exercise. Bone loss and recovery experiences have been reported to occur in Gambian women with low calcium intakes [85] as well as in groups of white women with greater calcium intakes [85]. Additionally, Little and Clapp [86] reported that regular, self-selected, recreational exercise has no impact on early postpartum lactation-induced BMD loss. Caird et al. [87] reported that the bone loss of lactation is somewhat minimized by the use of oral progestogen-only contraception. Nonetheless, biochemical marker concentrations measured in women using the progestogen closely resemble those observed in lactating women using barrier contraceptive methods. The mechanism(s) that mediates rapid bone turnover and

726

MARYFRAN SOWERS

TABLE 1

Longitudinal Regression Models for Change in Spine and Femoral Neck Bone Mineral Density of Postpartum Women Time (months)

PTHrP (pmol/liter)

Prolactin (ng/ml)

I

(0.001)

(0.01)

(0.001)

—

II

(0.01)

(0.18)

—

(0.001)

III

(0.02)

(0.01)

(0.001)

—

IV

(0.06)

(0.03)

—

(0.001)

Model

Breast-feeding practice Fully Partially

Estradiol (pg/ml)

Menses resume

—

(0.001)

—

(0.97)

(0.001)

—

—

—

(0.001)

(0.15)

—

(0.01) —

Spine

Femoral neck I

(0.06)

(0.02)

(0.01)

—

—

(0.18)

II

(0.04)

(0.01)

—

(0.01)

(0.01)

(0.26)

—

III

(0.08)

(0.03)

(0.03)

—

—

—

(0.18)

IV

(0.08)

(0.01)

—

(0.01)

(0.68)

—

(0.06)

breast milk is controversial. At least two possible mechanisms may increase skeletal turnover in lactation. The calciotropic hormones, parathyroid hormone and 1,25-dihydroxyvitamin D, stimulate bone resorption. Thus, it was believed that the transfer of calcium and phosphate to breast milk would stimulate PTH and 1,25-dihydroxyvitamin D-induced bone resorption. However, these actions have not been well substantiated in studies of lactation. Indeed, in the rat, it has been reported that the bone loss of lactation is independent of both PTH and vitamin D concentrations [89,90]. In humans, studies have frequently observed little difference in concentrations of these calciotropic hormones between lactating women and controls [91]. The changes in BMD with lactation appear to be determined by the combined effects of lower estradiol concentrations and higher PTHrP which may be linked with the higher prolactin concentrations [83]. Data suggested that the changes in calcium homeostasis during lactation are not related to PTH, 1,25-dihydroxyvitamin D, or 25-hydroxyvitamin D concentrations or to the changes in the concentrations of these calciotropic hormones in the postpartum period [92]. Several lines of evidence suggest that PTHrP has a significant role in calcium metabolism in lactation. First, PTHrP was identified initially as the factor associated with the humoral hypercalcemia of malignancy that is expressed in multiple cancer types, but most notably with breast tumors [93 – 95]. Second, in animal studies, PTHrP has been shown to be synthesized in lactating mammary tissue [96, 97]; in rats, a temporal relation exists between elevations in serum prolactin levels and the local expression of PTHrP mRNA levels [98]. High concentrations of PTHrP have been described in the milk of a variety of mammals [99,100]. Sowers et al. [83] found that elevated PTHrP concentrations were significantly associated with breast-feeding status,

As shown with p-values in Table 1, PTHrP was the consistent and significant predictor in all four of the femoral neck BMD change models and 3 of the four longitudinal models for lumbar spine change, independent of the inclusion of serum prolactin or estradiol concentrations, time since resumption of menses, or breast feeding practice. Furthermore, PTHrP values were associated negatively and significantly with BMD change in the spine and femoral neck over time. The primary role of PTHrP in calcium metabolism during lactation may be more prominent in the early months following parturition. Consistent with a linkage of greater prolactin and detectable PTHrP values is the report by Stiegler et al. [101] that detectable concentrations of PTHrP were observed in approximately 50% of men and women with prolactin-secreting adenomas and osteopenia. The transitory elevation in PTHrP concentrations as women initiate weaning might even contribute to the BMD recovery observed between 6 and 18 months following parturition. Using tissue culture systems of fetal rat calvariae, Canalis et al. [102] demonstrated that continuous treatment with PTHrP reduced labeled proline incorporation into bone collagen by 50%. However, transient exposure to PTHrP actually doubled the increase in proline incorporation, an effect that the investigators attributed to enhancement of the local production of insulin-like growth factor I. Early in lactation, more constant PTHrP concentrations may be sustained by more frequent suckling. These sustained PTHrP concentrations, in turn, may minimize the amount of bone collagen formation and stimulate both bone resorption and formation. These actions would tend to assure a source of calcium and phosphate for incorporation into breast milk. Likewise, as lactation frequency subsides or as weaning behaviors are introduced, PTHrP secretion would become more episodic. Using as a paradigm the Canalis data as well as similar findings in studies conducted

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

mass recovery. The multiple lines of evidence and the temporality provide a compelling argument for a biological role for PTHrP in calcium transfer during lactation. While there appears to be bone loss and bone mineral recovery with extended lactation, unanswered questions still remain. The mechanisms by which loss and recovery occur need further elucidation. Studies are needed in specific subgroups, including adolescents, women of obstetric maturity who are lactating, and women with extended and repeated lactation. Finally, additional studies are needed to define the impact of heavy metal liberation (lead) that occurs concurrently with the rapid bone turnover observed in lactation [104].

C. Studies of Lactation and Fracture The likelihood that lactation is associated with subsequent fracture risk appears to be influenced by the duration of the lactation [67] and by whether the comparison group is based on parous or nulliparous women [64,69]. For example, a case-control study investigating risk factors for fracture in postmenopausal women found no overall greater fracture risk in women who had breast-fed vs women who had never breast-fed [67]. However, stratified analysis suggested that breast-feeding for less than 1 year might increase the risk whereas breast-feeding in excess of 1 year might decrease the risk. The case-control study by Krieger et al. [65] suggested a protective effect for breast-feeding. The importance of comparison group definition is demonstrated in the data of Hoffman et al. [64] as well as Cumming and Klineberg [69]. Hoffman et al. [64] reported a protective effect of breast-feeding in relation to hip fracture (with confidence intervals that included the null value); however, that association could not be reproduced when the comparison was limited to parous women. In contrast, a negative association was reported by Cumming and Klineberg [69] that persisted when the comparison was restricted to parous women; however, confidence intervals for the measure of association included the null value. A study conducted in southern France showed no association of breast-feeding with subsequent fracture [68].

D. Studies with Bone Turnover Markers Evidence supporting the observation of acute bone mineral loss and subsequent remineralization also comes from cross-sectional [79] and longitudinal [105] measurement of bone turnover markers and markers of calcium homeostasis [79,81]. Concentrations of osteocalcin [83,85] and bonespecific alkaline phosphatase reached their zenith in the early postpartum and subsequently decline. Profiles of the

727

E. Summary and Implications In summary, there appears to be little ultimate loss of mineral from the maternal skeleton with lactation of wellnourished women if, during or after lactation, menstrual cycling is reestablished. Current evidence indicates that extended lactation is associated with acute skeletal loss despite high dietary calcium intake. Variation in the calcium intake was not related to the amount of bone lost in either well-nourished or poorly nourished women. Likewise, calcium intake was not significantly associated with changes in bone turnover markers. The time to return of menses was consistently associated with time of bone mineral recovery. Presently, there is much to be learned about the mechanisms associated with the rapid loss during lactation as well as the rapid recovery of bone mineral that follows weaning. Investigations of PTHrP concentrations have been associated with the bone loss of lactation. Understanding these mechanisms could possibly be extended to other bone loss processes, including those associated with menopause, and potentially could serve as a model for facilitating bone mineral recovery.

VI. OVARIAN ACTIVITY OR MENSTRUAL CYCLE CHARACTERISTICS AND BONE MASS The endocrinology of the ovarian cycle and its physiological manifestation in the menstrual cycle have not been well studied in relation to bone mass. This section addresses the onset of the menstrual cycle and explores the effects of subclinical and clinical disruption of the ovarian cycle.

A. Age at Menarche The initiation of menses and accompanying estrogen surge may stimulate bone growth by increasing osteoblastic activity [5]. However, the role of age at menarche relative to bone mineral content could be defined more clearly if we understood whether age at menarche was related primarily to bone growth (and epiphyseal closure) or greater likelihood of mineralization as an adjunct to the increased likelihood of greater body size, or equally to both. Additionally, defining the initiation event for menarche, i.e., hormone sensitivity or critical body fat mass, would also allow greater understanding about the long-term impact of the age at menarche. Adolescents with early age at menarche establish

728

MARYFRAN SOWERS

demonstrated greater concentrations of estradiol and folliclestimulating hormone as compared to young women with later onset of menarche, while maintaining comparable body weights [106]. Two major hypotheses have emerged to explain the variation in age of menarche: one related to triggering of the pituitary – gonadal axis by maturation and the second associated with the achievement of a critical weight (body fat). Grumbach and colleagues [107] hypothesized that the onset of puberty is the result of decreasing hypothalamic sensitivity to gonadal steroids. The hypothesis postulates that the decreasing sensitivity results in increased output of the gonadal steroids (positive feedback), which ultimately brings about the morphologic and physiologic characteristics of sexual maturity. If this is the mechanism for menarche, it would imply that women with a delayed onset of menses might fail to establish higher concentrations of the gonadal steroids required for the feedback process. Such adolescents may have lower bone mineral density if there is continued failure throughout early adulthood to establish a “normal” menstrual cycle pattern. The early version of the critical weight (body fat) hypothesis elaborated by Frisch and Ravelle [108] proposed that menarche is achieved by attaining a critical body mass (as reflected by total weight). A secondary data analysis of three longitudinal growth studies suggested that the critical weight was 47.8 kg. The hypothesis was subsequently revised to suggest that the essential component of the body mass was in the fat compartment and that the critical fat level was 17% [109]. Frisch linked the critical fat hypothesis to hormone levels through the work of Nimrod and Ryan [110], who developed the concept of aromatization of androgens in body fat as sources of the estrogen, estrone. The hypothesis has been highly criticized for its methodological and empirical limitations (reviewed by Scott and Johnston [111]). Whether weight acts as the precipitating or secondary event in the initiation of menarche, lower weight (as a mechanical force) and lower body fat composition (that becomes a compromised secondary source of estrogens by the aromatization of androgens) have been suggested as risk factors for lower peak bone mass.

B. Studies of Bone Mineralization and Age at Menarche Numerous studies suggest that age at menarche is associated with bone growth and bone density. It has been observed that girls with an earlier onset of menarche are shorter, heavier, and have a shorter duration of bone growth than girls of usual age at menarche [112]. Conversely, girls with late age at menarche (14 years) are more likely to be taller, have lower body fat, and have lower bone density

[50,51,114] and for more rapid rate of premenopausal bone loss [50]. In the later study, there was no relation between age at menarche and radial BMD when nulliparous women were removed from analysis. Possibly, the hormonal environment that is associated with failure to conceive is the same environment associated with delayed puberty [50]. Age at menarche can be related to bone mineralization in at least two different ways. First, women with an earlier age at menarche are likely to have a longer time between menarche and the menopause (gynecological age), a time during which estradiol resources are available to support and maintain bone mineralization. Second, events that precipitate earlier menarche, including weight gain, may be associated with characteristics that have been reported to produce greater bone density and, by imputation, greater peak bone density.

C. Duration of Menstruation and Number of Menstrual Cycles One reason that disparities may exist in assessing the role of reproductive factors is that the various events markedly alter the likelihood of exposure to specific levels of hormones. For example, with pregnancy and lactation, the effect of the elevated estradiol levels of pregnancy followed by the suppressed levels during lactation may generate a cumulative influence on bone density quite different from the influence of each event alone. One approach to accommodate these normal fluctuations in hormone levels is to examine the number of menstrual cycles. Fox et al. [51] showed a positive association between radial bone density in postmenopausal women with each successive year of continued menstruation. Georgiou et al. [115] reported that bone mineral content in postmenopausal women was better explained by the total number of menstrual cycles than by the years since menopause or chronological age.

VII. DYSFUNCTIONAL OVULATION A. Marginal Hormone Status Although the prevalence of frank estrogen deficiency has been estimated to be approximately 2% in college-aged women, the prevalence in a general population, ages 20 to 40 years, is not well established. Furthermore, subclinical levels of estrogen insufficiency may be more common [116 – 118] and may influence bone density. Several studies have suggested that marginal hormone status is important in establishing variation in premenopausal bone mineral

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

Marginal hormone status associated with low premenopausal bone mass has been reported in two studies. Sowers et al. [115] described a nested case-control study in which significantly lower estradiol and testosterone concentrations, and higher luteinizing hormone values were found in the low BMD group than in the control group. In a subsequent study, Sowers et al. [116] showed that daily urinary hormone excretion patterns for women with lower peak bone mineral density differed from those of women with normal BMD. Healthy, menstruating women with low BMD from a large population-based study had significantly lower urinary sex steroid hormone concentrations during the luteal phase of menstrual cycles compared to hormone concentrations in premenopausal women with average BMD, even after considering the role of body size. Notably, LH peaks were lower and there was a muted progesterone response. These data suggest that subclinical decreases in circulating gonadal steroids may impair the attainment and/or maintenance of bone mass in otherwise reproductively normal women. Steinberg et al. [119] reported lower serum estradiol concentrations in perimenopausal women (mean age of 46) vs premenopausal women (mean age of 41). Free estrogen and free testosterone concentration were positively correlated with bone density. These hormone characteristics were observed in populations without anorexia nervosa or intense chronic physical activity.

B. Pronounced Events of Ovarian Dysfunction Two syndromes that include amenorrhea, chronic endurance exercise and anorexia nervosa, have been characterized relative to bone density. It has been assumed that amenorrhea in both of these syndromes arises from reductions in total body fat rather than from intrinsic disruption of the neuroendocrine system. Two other clinical entities, prolactin-secreting tumors and polycystic ovarian disease, are less extensively studied relative to bone mass and are assumed to have primary involvement of the neuroendocrine system.

C. Chronic Endurance Exercise Premenopausal athletes are typically characterized by low body fat, lean body mass, and greater bone mineral density than nonathletes. However, it has long been appreciated that pre- and perimenopausal women engaged in chronic endurance exercise, if accompanied by menstrual dysfunction, may be catabolic rather than anabolic for bone [120 – 128]. Reported menstrual cycle changes in women who

729

shortened luteal phase of the menstrual cycle [94], menstrual irregularities, oligomenorrhea, and amenorrhea [130,131]. Some investigators have suggested that the hypothalamic – pituitary – ovarian – adrenal axis is suppressed by rigorous physical activity; subsequently, bone mass is lower because of lower concentrations of estradiol [122,125,129,132] and progesterone [122] and higher concentrations of cortisol [133]. Another hypothesis is that lower BMD in female athletes is the consequence of repeated episodes of hyperprolactinemia [134], although increased basal prolactin values have not been identified consistently in amenorrheic athletes [125,133,135]. Other studies have shown that progesterone, prolactin, and testosterone concentrations all increase with strenuous physical activity [136 – 38]. Frisch [139] has argued that amenorrhea of exercise is due to a diminution of critical weight (fat mass). She further argues that there is a state associated with transitory weight recovery or moderate physical training that is accompanied by menstrual cycles that occur with shortened luteal phases or that are anovulatory. The critical weight hypothesis is not well supported in the literature, which indicates that both eumenorrheic and amemorrheic athletes may have similar amounts of body fat. For example, Myburgh et al. [140] found that amenorrheic athletes had lower BMD than controls, matched on age, body mass, and exercise quantity. This lower BMD was observed at the spine, proximal femur, and total body, but not at the midradial or tibial shafts. Linnell et al. [126] suggested that discrepancies observed in describing relationships between intense physical exercise and BMD may reflect the additive effect of low body fat and intrinsic ovarian dysfunction, indicating that these are not consistently simultaneous events. Prior et al. [116] concluded that decreases in spinal bone density among eumenorrheic women athletes correlated with asymptomatic disturbances of the ovulatory cycle and not with the degree of physical activity. Physical stress alone can influence menstrual cycling, regardless of body fat levels. While the catabolic effect of amenorrhea and strenuous endurance sports on bone mass in women is relatively consistently observed, demonstrating anabolic effects of fitness and moderate physical activity is more problematic. Potentially, fitness and moderate physical activity could be anabolic for bone by either hormonal mechanisms or increasing the mechanical loading on bone. A hormonal effect associated with physical fitness and body composition may be mediated through an increase in the secretion of growth hormone and thus somatomedin-C or insulin-like growth factor I. This hormone apparently stimulates the intermittent secretion of parathyroid hormone, collagen synthesis, and number of osteoblasts

730

MARYFRAN SOWERS

D. Amenorrhea of Anorexia Nervosa Osteoporosis is an established complication of anorexia nervosa [142 – 144]. Proposed mechanisms for osteopenia include estrogen deficiency, glucocorticoid excess [143], generalized malnutrition, and calcium intake deficiency, with the potential for more than one mechanism to be operating simultaneously. While some investigators have reported that the compulsive exercise frequently associated with anorexia nervosa was protective for bone loss [144], others have failed to observe this protective relation [143, 145 – 147]. This discrepancy may be related to the degree and intensity of exercise practiced by study participants. Because women with anorexia nervosa are frequently both underweight and amenorrheic, ascertaining the independent contributions of estrogen deficiency and decreased body mass to their osteopenia is difficult. However, in hyperprolactinemic amenorrhea, women with increased body weight are protected against osteopenia. This suggests the potential for independent contributions from both the underweight and hypoestrogenism [148]. Bachrach et al. [145] found body size, age at onset, and duration of anorexia nervosa, but not dietary calcium intake, physical activity level, or duration of amenorrhea to be correlated with BMD in adolescent girls. Dietary calcium supplementation has not promoted bone mineral maintenance; however, most studies acknowledge concerns about patient compliance with the therapy and short duration of therapy [143 – 145]. With rare exception [145], studies have failed to differentiate whether the subjects were women who had failed to acquire bone or women who had lost bone.

E. Hyperprolactinemia Gonadal suppression with prolactin-secreting tumors and other conditions associated with hyperprolactinemia may be an important contributor to low premenopausal bone mass and subsequent risk of osteoporosis in a limited number of women. It is estimated that hyperprolactinemia occurs in more than 25% of young adult women with amenorrhea. In a longitudinal study by Schlechte and associates [149], women with hyperprolactinemia had lower bone mass of the spine and radius at entry to the study. Over the 4.7-year follow-up, women with hyperprolactinemia did not lose bone mass, whereas healthy women had significant loss at the spine (but not radius). The investigators suggest that women with hyperprolactinemia may have retained bone mass in the face of decreased estradiol concentrations because of greater body mass (28 vs 24 kg/cm2) and higher testosterone concentrations. Restoration of gonadal function was not associated with normalization of the bone mineral [76,150]. Klibanski and Greenspan [148] also reported

prolactinemia but does not return bone density to the level observed in controls. However, it has been observed that hyperprolactinemic women who were eumenorrheic had greater bone density than hyperprolactinemic women who were amenorrheic [150], suggesting the potential for a differential response according to the duration of reduced estrogen stimulation.

F. Polycystic Ovarian Syndrome Polycystic ovarian syndrome (PCOS) is a heterogeneous group of conditions characterized by polyfollicular ovaries and a luteinizing hormone (LH)-dependent increase in androgen secretion. In addition to oligomenorrhea, this multifaceted syndrome may be accompanied by various degrees of virilization, obesity, hypertension, and diabetes. Di Carlo et al. [151] compared 188 women diagnosed with PCOS to a similar group of 142 patients with normal ovaries and reported that women with PCOS had significantly greater bone density of the lumbar spine (0.98 vs 0.87 g/cm2). The same group also reported higher serum concentrations of LH, prolactin, and, as expected, testosterone. The investigators speculated that several factors may be associated with the greater BMD in the face of amenorrhea in this group. The PCOS group had a greater body mass index (25.0 vs 22.9 kg/m2) than women with normal ovaries and had higher androgen levels.

G. Summary and Implications Among premenopausal women, there are variations in ovarian and gonadotropin hormones associated with variation in BMD. While the frank amenorrhea that may accompany chronic endurance physical activity, hyperprolactinemia, PCOS, and anorexia nervosa has long been recognized as being associated with lower BMD, the prevalence of these conditions is uncertain. Thus, it is difficult to ascertain the overall impact on peak bone mass and, by extension, osteoporosis and fracture. New studies now indicate that lower concentrations of hormones jeopardize BMD even when amenorrhea is not present. This suggests that as more is learned about the relationship between peak bone mass and osteoporosis risk, premenopausal hormone concentrations may become a more prominent source of intervention.

VIII. ORAL CONTRACEPTIVE USE If premenopausal hormone concentrations may become a more prominent source of intervention to reduce osteoporosis risk, then a logical area to evaluate is that of oral

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

bone mineral content has been of great interest as investigators have tried to determine parallels between oral contraceptives and hormone replacement therapy on bone density. The lines between oral contraceptives and estrogen therapy use have become increasingly blurred. Oral contraceptives are now approved for use in women over the age of 35, and estrogens are the common constituent frequently associated with both oral contraceptive products and hormone replacement therapy. However, there are major differences in the drug formulation, including the presence or absence and type of progestins, the dosage of active ingredients, and the regimens for their use. The impact of oral contraceptive use on BMD remains unresolved with studies reporting both no effect and a positive effect. In this review, both cross-sectional and longitudinal studies of oral contraceptive use and bone mineral content were examined and, when possible, dichotomized according to menopausal status. This dichotomy is useful for three reasons. First, formulations for oral contraceptives used by women prior to 1980 (who are now more likely to be periand postmenopausal) generally had significantly higher estrogen doses than the preparations to which most premenopausal women have been exposed. Second, there may have been different selection factors operating as to which women elected to use oral contraceptives in the 1960 – 1970s vs those currently using oral contraceptives.

A. Studies of Oral Contraceptive Use and Bone Mass Findings from studies of oral contraceptives and bone mass have been inconsistent, despite the substantial number of studies. The most consistent observation is that use of oral contraceptives has not been associated with lower BMD. Whether oral contraceptive preparations are associated with greater BMD remains debatable. Numerous cross-sectional studies [152 – 156] have reported a positive association between bone density and oral contraceptive use in various populations of premenopausal women. Likewise, several studies have reported a positive association of oral contraceptive use across a wide age range, including postmenopausal women [157,158]. In contrast, other crosssectional studies reported no association of oral contraceptives with BMD in premenopausal women [149,160] and among women across a wide age range [59,161 – 163]. In one of the few longitudinal studies, Recker et al. [164] reported a positive correlation between total body bone mineral and oral contraceptive use; however, they observed no association between oral contraceptive use and BMD of the forearm or lumbar spine. In a longitudinal study with 5 years of observation, Sowers et al. [50] reported that among 22 pre- and perimenopausal women who

731

a longer duration of use was associated with less radial BMD loss, after adjusting for age. In a study of 19 women who were administered 20 mcg of ethinyl estradiol and had bone measurements taken at the 3rd, 6th, and 12th cycle, the investigators reported a slight, but insignificant, rise in bone mass of the distal radius [165]. A major limitation of this study is the lack of a control group. In contrast, Gambacciani et al. [166] reported that 40- to 49-year-old women with oligomenorrhea using oral contraceptives (20 mcg of ethinyl estradiol) did not lose bone to the same degree as women with oligomenorrhea who did not use oral contraceptives. It appears that an underlying assumption of BMD studies is that oral contraceptive preparations increase the circulating estrogen concentrations. However, many of the current preparations provide hormone doses just adequate to suppress ovulation and not sufficient to generate the variation in physiologic ranges found throughout the menstrual cycle in women not using oral contraceptives. Most studies evaluating the potential effect of oral contraceptives on BMD reflect estrogen dosages of 35 mcg or greater. A study evaluating low-dose (20 mcg) oral contraceptives found BMD reduced in women using that pill [151]. There is good reason to believe that oral contraceptives may help promote bone mineralization in women with very low circulating hormones, amenorrhea, or oligomenorrha. There is less likelihood that BMD will be retained if the use of oral contraceptives actually lowers circulating estrogen concentrations in any particular woman. Indeed, a study by Garnero and colleagues [168] indicated that there was no overall difference in BMD between users and nonusers; however, oral contraceptive use was associated with a moderate decrease in bone turnover.

B. Studies of Oral Contraceptives and Fractures The number of studies of oral contraceptives and fracture is quite limited, in part because women who were of an age to use oral contraceptives in the 1950s and 1960s are just now achieving an age where fractures occur with sufficient frequency to make such a study efficient. Cooper et al. [169] examined the fracture experience of the 46,000 enrollees in the Royal College of General Practitioners Oral Contraception Study that began in 1974. The risk of subsequent fractures was significantly greater among oral contraceptive users than among nonusers.

C. Summary and Implications While there have been a substantial number of studies that relate bone mass and oral contraceptive use, ambiguity

732 may be difficult to synthesize for the following reasons. First, only some of the progestrogens are 19-nor-testosterone derivatives that have androgenic/anabolic properties. For example, Cundy et al. [170] reported that the degree of estrogen deficiency induced in women using depot medroxyprogesterone actetate (DMPA) for contraception may adversely affect bone density (see later). This is evidence for the importance of formulations of the particular oral contraceptive. Second, dose and duration of use may have a differential impact according to the chronological or gynecological age of the user. For example, the role of menopause may overshadow any impact of oral contraceptive use on BMD in postmenopausal women. The oral contraceptive effect may be different in adolescents still acquiring bone as compared to adult women who are more likely to be in a bone maintenance phase. Third, oral contraceptives are also used in the regulation of dysfunctional menstrual cycles. As such, the universe of oral contraceptive users may be quite heterogeneous and include women with conditions that include potential hormonal abnormalities, e.g., dysmenorrhea or irregular cycles, as well as women who use the hormones for contraception alone. Any future studies of oral contraceptive use should be undertaken in women in whom it can be determined if the hormonal preparation is being used for conception prevention or menstrual cycle regulation. Duration of use, as well as dose and type of the preparation, should also be addressed. In younger women, the issues of oral contraceptive use in bone acquisition vs bone maintenance should be addressed. In older women, the potential bone loss with age and menopausal status should be separated from the impact of duration of OPC use.

IX. PROGESTIN-INJECTIBLE CONTRACEPTIVES A progestin-only, injectable contraception, depot medroxyprogesterone acetate (Depo-Provera), given intramuscularly every 90 days, was approved for use in the United States in 1992. Worldwide, the contraceptive DMPA is used in more than 90 countries by an estimated 3.5 million women. There is a compelling physiological mechanism by which DMPA could compromise BMD. Contraception is achieved primarily through disruption of the hypothalamic – pituitary – ovarian axis. Because DMPA disrupts the HPO axis, it will suppress estrogen production, leading to a relative estrogen deficiency and a corresponding loss of BMD. If DMPA has an adverse effect on BMD, then a substantial cohort of young women may enter menopause with less bone mineral reserve and be at increased risk for the development of osteoporosis, fracture, and related morbidity following

MARYFRAN SOWERS

Approximately six studies have addressed potential BMD levels among users. The studies suggest that BMD values were approximately 3 – 7% lower than values in controls [170 – 173]. Studies also suggest that this BMD compromise resolves following the discontinuance of DMPA use. DMPA potentially offers a model of the importance of ovarian hormones in the achievement and maintenance of peak bone mineral density. Canterbury and Hatcher [174] found that mean serum estradiol concentrations in 207 women 3 months after DMPA injection were 51.6 pg/ml (SD not provided) compared to a mean estradiol concentration of 140 pg/ml reported for historical controls. Additionally, 25% of the women using DMPA (as compared to 15 % of nonusers) had estradiol levels less than 15 pg/ml, values similar to those of postmenopausal women. Furthermore, limited data document that DMPA disrupts the ovarian and menstrual cycles. During a normal menstrual cycle, the estradiol concentrations are at their lowest during the early follicular phase, rising to a peak at midcycle, followed by a decline and then presenting with a second peak during the midluteal phase [175]. Because DMPA inhibits the cyclical variation in estradiol concentrations, women using DMPA would experience a relative estrogen deficiency because of the absence of the increase in estradiol concentrations at the midcycle and during the luteal phase. Two studies, each with only three subjects, measured estradiol daily for 1 month prior to and up to 3 months after a single injection of DMPA of 150 mg [176,177]. Both studies observed the lack of the cyclical estradiol changes usually found during the normal menstrual cycle and estradiol concentrations that remained at levels consistent with the follicular phase of the menstrual cycle. Estimates of the frequency of amenorrhea vary by population and the duration of DMPA use. Large multinational studies report that between 35 and 66% of all women using DMPA will develop amenorrhea by 12 months, with 8 to 25% becoming amenorrheic after their initial injection. By the end of 5 years of use, approximately 70 to 80% of women will become amenorrheic [177 – 180]. Because DMPA is widely prescribed to adolescents, there is a concern that suppression of the HPO axis by DMPA could lead to a compromise in the attainment of peak BMD among these young women. A related issue is whether potential DMPA-related bone loss, particularly among adolescents, can be offset by a higher calcium intake. Reasonable evidence shows that calcium intake influences bone acquisition in pre-and peripubescent girls [181 – 183]. However, some studies suggest that calcium intake only influences the timing, not the magnitude, of peak bone mineralization and that any effect may be best realized only in those persons with extremely low ( 500 mg/day) calcium intakes [184]. The research evaluating

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

adults is limited and the few longitudinal studies specifically targeting women under age 40 are inconsistent. Recker and colleagues [164] found that calcium intake in 156 college women followed over 7 years was an independent predictor of bone gain over time. In contrast, dietary calcium intake, measured by both food frequency and 24-h dietary recall, did not predict change in radial BMD over 5 years in women aged 20 to 35 years. Similarly, no association was observed between calcium intake and change in bone mineral density of the femoral neck and lumbar spine in 153 Finnish women aged 13 – 27 years who were followed for 15 years [185]. Weight gain is an acknowledged side effect of DMPA that could affect BMD. Physiologically, weight gain with DMPA could be due to an increase in muscle mass related to the androgenic effect of the progestin or to an increase in fat mass resulting from the inhibition of the appetite control center in the hypothalamus [186]. Whether DMPA actually causes weight gain and the magnitude of that gain is controversial. Several international studies reported weight gains averaging 3 to 5 pounds at 12 months and 7 to 10 pounds at 24 months [187 – 190]. These studies included no control subjects; consequently, it is not known whether the weight gain was due to DMPA use or reflects typical population patterns. Other studies provide no evidence of a DMPA-related weight gain [191,192]. Weight gain could counteract the decline in BMD by providing an increase in the mechanical loading force on bone. Mechanical loading acts as a stimulus for osteoblast production, thereby increasing bone formation [193]. An alternative counterforce might arise from the conversion of androgens to estrogen in the peripheral adipose tissue. As adiposity increases, there is a greater adrenal production of androgens that increases the availability of the precursor hormone, as well as an accelerated rate of conversion from androstenedione [194,195]. Typically, extra-ovarian sources of estrogen account for a relatively small proportion of circulating estrogens and would most likely not affect BMD. However, in women with reduced ovarian estradiol production, these sources may become more important. In summary, the widespread and international use of DMPA could result in a systemic reduction in BMD by disruption of the HPO axis. While some evidence suggests that this reduction is recoverable over time, a series of questions remain about its long-term impact. The primary question relates to whether its use among still-growing adolescents will reduce the potential for those adolescents to achieve their genetic potential for bone mass. A second issue arises when DMPA is used as the contraceptive method of choice in women who have been lactating. Whether the coupling of these two events that are associated with BMD loss could result in an additive bone loss is

733

X. OOPHORECTOMY Oophorectomy is commonly cited as an example of hypoestrogenism with an impact on measures of calcium metabolism [196], fracture [65], and bone mineral content in both white [197 – 200] and Japanese women [201,202]. Richelson et al. [197] compared the BMD of the radius, femoral neck, and lumbar spine in women in their 5th decade who had undergone oophorectomy 20 years earlier, with the BMD measured at the same sites of women in their 7th decade who had undergone spontaneous menopause, also some 20 years earlier. BMD of the two groups of women were almost the same and suggested that estrogen is a factor as important as aging in determining the level of bone density. Indeed, Aitken et al. [198] reported that bone density, as measured by standard aluminum equivalents, was lowest in women who had undergone oophorectomy at earlier ages.

A. Studies of the Effect of Oophorectomy on Bone Several studies have attempted to define the nature of the rate of bone loss following oophorectomy. Cross-sectional data of Stepan et al. [203] suggested a mean loss of 2.8% of the metacarpal cortical area and 8% of the lumbar spine (by dual photon densitometry) in the first year following oophorectomy. Using statistical modeling of secondary data [204] of the cortical area, Reeves [204] projected that there was a doubling of bone resorption following oophorectomy. Genant et al. [205] estimated that annual bone mineral losses were approximately 8% in the vertebral spongiosum and about 2% in the peripheral cortex when evaluated by quantitative computed tomography. In contrast, Hreshchyshyn et al. [206] did not observe a more pronounced rate of change in women after oophorectomy as compared to naturally menopausal women. Bone turnover markers reflect higher bone resorption relative to bone formation in the period following oophorectomy [201,203]. Investigators observed an increase in serum osteocalcin concentrations beginning 1 – 2 months following oophorectomy and increasing up to 1 year of follow-up. Bone-specific alkaline phosphatase also rose, although at a slower rate than osteocalcin levels [203]. Ohta et al. [201,202] proposed that the bone loss in women after oophorectomy encompasses more than the diminution of estradiol. Oophorectomy also includes a marked reduction of estrone and androstenedione concentrations to values that are significantly lower than those concentrations measured in menopausal women. Ohta et al. [201] indicated that postmenopausal women retain some estrone secretion from ovarian interstitial cells that may not

734 A number of studies have reported that the calciotropic hormones, particularly PTH, do not undergo significant changes following oophorectomy [201,203,206], suggesting that the bone loss of oophorectomy is not dependent on the homeostatic regulation of serum calcium. Yet to be evaluated is the relationship to PTH, using a contemporary assay for intact PTH. While oophorectomy provides one model for the evaluation of ovarian hormone deprivation and BMD, that relationship may be confounded by those events that gave rise to the context for the oophorectomy. Oophorectomy with hysterectomy is performed for the treatment of malignancy, pelvic inflammatory disease, endometriosis, uterine fibroids, and other conditions that may influence BMD independently of the surgical procedure and its hormonal sequelae. Estrogen replacement is a frequently proposed strategy following oophorectomy, although data describing the frequency of its prescription, compliance, and duration of use appear to be unavailable for the general population. Aitken et al. [198] estimated that women lost approximately 8% of metacarpal bone mass in the first 2 years following oophorectomy in comparison to no measurable loss in women treated with mestranol (10 – 20 mcg/day). The same study also allotted a group of women to mestranol treatment who were 3 and 6 years postoophorectomy. While the women who were 3 years postoophorectomy maintained bone density, those women who were 6 years postsurgery continued to lose bone and manifested no responsiveness to the mestranol. The investigators interpreted this to mean that there is a limited window of time following oophorectomy when bone is most responsive to hormone replacement. A number of treatments for the bone loss associated with oophorectomy, apart from hormone replacement [207], have been evaluated. In a clinical trial of a synthetic flavonoid, Gambacciani et al. [208] demonstrated that women with oophorectomy/hysterectomy (n  16) acting as controls had significant loss of radial bone mass and elevated hydroxyproline levels in comparison to women in the treated group (n  16) 1 year following surgery. The prophylactic administration of salmon calcitonin in oophorectomized women apparently inhibited skeletal resorption as measured by radial bone mineral content and the behavior of Gla protein (osteocalcin) and hydroxyproline concentrations [209].

XI. SUMMARY AND IMPLICATIONS A primary focus of studies of the reproductive environment has focused on the role of pregnancy and lactation as risk factors for subsequent lower BMD and increased risk of fracture. Collectively, studies of pregnancy and lactation

MARYFRAN SOWERS

for sustained bone mass loss with a subsequent and adequate hormone environment that includes reestablishment of menses and the capacity to sustain continued reproduction. There is currently no evidence that multiple births interspersed with intensive lactation are deleterious to maintaining peak bone mass. Additionally, pregnancy and lactation provide models that can be examined to learn more of the biology associated with the maintenance of bone mass, particularly apart from calcium regulation by the calciotropic hormones. Lactation, in particular, offers the opportunity to understand the dynamics of bone loss as well as bone recovery. Studies of the reproductive environment and bone that have been reviewed support the concept that various hormone concentrations are important not only in their decline around the menopause, but also in the establishment and maintenance of peak bone mass. While studies of pregnancy and lactation suggest that they have little generalized impact on peak bone mass, the studies of age at first pregnancy, age at menarche, nulliparity, and subclinical ovarian hypofunction are intriguing. The overt amenorrhea of prolactin-secreting tumor and chronic endurance exercise consistently suggest the importance of an adequate estrogenic environment. This body of evidence suggests that a group of women could be identified who are at higher risk for compromise of their bone mass because of their hormone status. The evidence is sufficiently strong and the risk factors sufficiently potent to evaluate hormone and skeletal status in the period immediately prior to perimenopause.

References 1. R. B. Mazess, On aging bone loss. Clin. Orthop. Rel. Res. 165, 239 – 252 (1982). 2. L. Riggs, Overview of osteoporosis. West. J. Med., 154, 63 – 77 (1991). 3. A. Horsman and L. Burkinshaw, Stochastic models of femoral bone loss and hip fracture risk. In “Clinical Disorders of Bone and Mineral Metabolism” (M. J. Kleerkoper and S. M. Krane, eds.), pp. 253 – 263, 1989. 4. World Health Organization, Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: Report of a WHO study group. Technical report series 843 (1994). 5. R. B. Jaffe and S. Dell’Acqua, “The Endocrine Physiology of Pregnancy and the Peripartal Period.” Raven Press, New York (1985). 6. L. Speroff, R. H. Glass, and N. Kase, “Clinical Gynecology, Endocrinology and Infertility” Williams and Wilkins, Baltimore, 1989. 7. M. H. Givens and I. G. Macy, The chemical composition of the human fetus. J. Biol. Chem. 102, 7 – 17 (1933). 8. N. Kent, R. I. Price, D. H. Gutteridge, J. R. Allen, S. L. Blakeman, C. I. Bhagat, A. St. John, M. P. Barnes, M. Smith, and D. V. Evans, Acute effects of an oral calcium load in pregnancy and lactation: Findings on renal calcium conservation and biochemical indices of bone turnover. Miner. Electrolyte Metabol. 17, 1 – 7 (1991).

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

10.

11. 12.

13.

14.

15. 16. 17. 18. 19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

human pregnancy and lactation and after resumption of menses. Am. J. Clin. Nutr. 67, 693 – 701 (1998). E. B. Pedersen, S. Johannesen, S. Kristensen, A. B. Rasmussen, K. Emmertsen, J. G. Lauritsen, and M. Wohlert, Calcium, parathyroid hormone and calcitonin in normal pregnancy and preeclampsia. Gynecol. Obstet. Invest. 18, 156 – 164 (1984). R. K. Marya, S. Rathee, and M. Manrow, Urinary calcium excretion in pregnancy. Gynecol. Obstet. Invest. 23, 141 – 144 (1987). T. Markestad, M. Ulstein, L. Aksnes, and D. Aarskog, Vitamin D metabolites in vitamin D supplemented pregnant women. Acta. Obstet. Gynecol. Scand. 65, 63 – 67 (1986). R. M. Pitkin, W. A. Reynolds, G. A. Williams, and G. K. Hargis, Calcium metabolism in normal pregnancy: A longitudinal study. Am. J. Obstet. Gynecol. 133, 781 – 790 (1979). T. Kuoppala, R. Tuimala, M. Parviainen, T. Koskinen, and M. AlaHouhala, Serum levels of vitamin D metabolites, calcium, phosphorus, magnesium and alkaline phosphatase in Finnish women throughout pregnancy and in cord serum at delivery. Hum. Nutr. Clin. Nutr. 40C, 287 – 293 (1986). G. Duggin, N. E. Dale, R. C. Lyneham, R. A. Evans, and D. J. Tiller, Calcium balance in pregnancy. Lancet 2, 926 – 927 (1974). R. P. Heaney and T. G. Skillman, Calcium metabolism in normal human pregnancy. J. Clin. Endocrinol. 33, 661 – 670 (1971). S. Shenolikar, Absorption of dietary calcium in pregnancy. Am. J. Clin. Nutr. 23, 63 – 67 (1970). R. Frisancho, S. M. Garn, and W. Ascoli, Unaltered cortical area of pregnant and lactating women. Invest. Radiol. 6, 119 – 121 (1971). C. Christiansen, R. Rodbro, and B. Heinild, Unchanged total body calcium in normal human pregnancy. Obstet. Gynecol. Scand. 55, 141 – 143 (1976). M. F. Sowers, M. Crutchfield, M. Jannausch, et al., A prospective evaluation of bone mineral change in pregnancy. Obstet. Gynecol. 77, 841 – 845 (1991). N. Kent, R. I. Rice, D. H. Gutteridge, J. R. Allen, K. J. Rosman, M. Smith, C. I. Bhagat, S. G. Wilson, and R. W. Ratallack, Effect of pregnancy and lactation on maternal bone mass and calcium metabolism. Osteo. Int. 1 (Suppl.), S44 – S47 (1993). B. Lamke, J. Brundin, and P. Moberg, Changes of bone mineral content during pregnancy and lactation. Acta Obstet. Gynecol. Scand. 56, 217 – 219 (1977). B. L. Drinkwater and C. H. Chesnut III, Bone density changes during pregnancy and lactation in active women: A longitudinal study. Bone Miner. 14, 153 – 160 (1991). M. F. Sowers, M. Jannausch, T. Scholl, and J. Schall, The reproducibility of ultrasound bone measures in a triethnic population of pregnant adolescents and adult women. J. Bone Miner. Res. 13, 1768 – 1774 (1998). M. Gambacciani, A. Spinetti, R. Gallo, B. Cappagli, G. Teti, and V. Facchini, Ultrasonographic bone characteristics during normal pregnancy: Longitudinal and cross-sectional evaluation. Am. J. Obstet. Gynecol. 173, 890 – 893 (1995). A. Yamaga, M. Taga, H. Minaguchi, and K. Sato, Changes in bone mass as determined by ultrasound and biochemical markers of bone turnover during pregnancy and puerperium: A longitudinal study. J. Clin. Endocrinol. Metab. 81, 752 – 756 (1996). K. Seki, N. Makimura, C. Mitsui, J. Hirata, and I. Nagata, Calciumregulating hormones and osteocalcin levels during pregnancy: A longitudinal study. Am. J. Obstet. Gynecol. 164, 1248 – 1252 (1991). A. Rodin, A. Duncan, H. W. P. Quartero, G. Pistofidis, G. Mashiter, K. Whitaker, D. Crook, J. C. Stevenson, M. G. Chapman, and I. Fogelman, Serum concentrations of alkaline phosphatase isoenzymes and osteocalcin in normal pregnancy. J. Clin. Endocrinol. Metab. 68, 1123 – 1127 (1989).

30.

31. 32.

33.

34.

35.

36. 37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

735

tions during pregnancy and lactation: Implications for maternal mineral metabolism. J. Clin. Endocrinol. Metab. 64, 290 – 294 (1987). F. Okonofua, R. K. Menon, S. Houlder, M. Thomas, D. Robinson, S. O’Brien, and P. Dandona, Calcium, vitamin D and parathyroid hormone relationships in pregnant Caucasian and Asian women and their neonates. Ann. Clin. Biochem. 24, 22 – 28 (1987). J. Verhaeghe and R. Bouillon, Calciotropic hormones during reproduction. J. Steroid Biochem. Mol. Biol. 41, 469 – 477 (1992). R. W. Chesney, B. L. Specker, F. Mimouni, and S. McKay, Mineral metabolism during pregnancy and lactation. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 383 – 393. Raven Press, London (1992). M. F. Sowers, Pregnancy and lactation as risk factors for subsequent bone loss and osteoporosis J. Bone Miner. Res. 11, 1052 – 1060 (1996). C. S. Kovacs, and H. M. Kronenberg, Maternal-fetal calcium and bone metabolism during pregnancy, puerperium, and lactation. Endocr. Rev. 18, 832 – 72 (1997). W. G. Cushard, S. Creditor, J. M. Canterbury et al., Physiological hyperparathyroidism in pregnancy. J. Clin. Endocrinol. Metab. 34, 767 – 771 (1972). R. M. Pitkin, Calcium metabolism in pregnancy and the perinatal period: A review. Am. J. Obstet. Gynecol. 151, 99 – 109 (1985). O. K. Davis, D. S. Hawkins, L. P. Rubin, J. T. Posillico, E. M. Brown, and I. Schiff, Serum parathyroid hormone (PTH) in pregnant women determined by an immunoradiometric assay for intact PTH. J. Clin Endocrinol. Metab. 67, 850 – 852 (1983). J. M. Gertner, D. R. Coustan, A. S. Kliger, L. E. Mallette, N. Ravin, and A. E. Broadus, Pregnancy as a state of physiologic absorptive hypercalciuria. Am. J. Med. 81, 451 – 456 (1986). N. Rasmussen, A. Frolich, P. J. Hornnes, and L. Hegedus, Serum ionized calcium and intact parathyroid hormone levels during pregnancy and postpartum. Br. J. Obstet. Gynaecol. 97, 857 – 862 (1990). M. Whitehead, G. Lane, O. Young, S. Campbell, G. Abeyasekera, C. J. Hillyard, I. MacIntyre, K. G. Phang, and J. C. Stevenson, Interrelations of calcium-regulating hormones during normal pregnancy. Br. Med. J. 283, 10 – 12 (1981). A. Frolich, M. Rusnicki, W. Fischer-Rasmussen, and K. Olofsson, Serum concentrations of intact parathyroid hormone during late human pregnancy: A longitudinal study. Eur. J. Obstet. Gynecol. Repro. Biol. 42, 85 – 87 (1991). L. Kohlmeier and R. Marcus, Osteoporosis associated with pregnancy. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.), pp. 959 – 967 (1996). S. Bertelloni, G. Baroncelli, A. Pelletti et al, Parathyroid hormone related protein in healthy pregnant women. Calcif. Tissue Int. 54, 195 – 197 (1994). E. Bucht, M. Carlqvist, B. Hedlund et al, Parathyroid hormone related peptide in human milk measured by a mid-molecule radioimmunoassay. Metab. Clin. Exp. 41, 11 – 16 (1992). R. MacIsaac, J. Heath, C. Rodda et al., Role of the fetal parathyroid glands and parathyroid hormone-related protein in the regulation of placenta transport of calcium, magnesium and inorganic phosphate. Repaired. Fertile Div. 3, 447 – 457 (1991). S. G. Wilson, R. W. Retallack, J. C. Kent, G. K. Worth, and D. H. Gutteridge, Serum free 1,25-dihydroxyvitamin D and the free 1,25 dihydroxyvitamin D index during a longitudinal study of human pregnancy and lactation. Clin. Endocrinol. 32, 613 – 622 (1990). W. A. Reynolds, G. A. Williams, and R. M. Pitkin, Calciotropic hormone responsiveness during pregnancy. Am. J. Obstet. Gynecol. 139, 855 – 861 (1981). J. C. Stevenson, C. J. Hillyard, and I. A. MacIntyre, A physiological

736 49. M. F. R. Sowers, R. B. Wallace, and J. H. Lemke, Correlates of midradius bone density among premenopausal women: A community study. Prev. Med. 14, 585 – 596 (1985). 50. M. F. R. Sowers, M. K. Clark, B. Hollis et al., Radial bone mineral density in pre- and perimenopausal women: A prospective study of rates and risk factors for loss. J. Bone Miner. Res. 7, 647 – 657 (1992). 51. K. M. Fox, J. Magaziner, R. Sherwin, J. C. Scott, C. C. Plato, M. Nevitt, and S. Cummings, Reproductive correlates of bone mass in elderly women. J. Bone Miner. Res. 8, 901 – 908 (1993). 52. R. W. Smith, Dietary and hormonal factors in bone loss. Fed. Proc. 26, 1737 – 1746 (1967). 53. B. E. Nilsson, Parity and osteoporosis. Surg. Gynecol. Obstet. 129, 27 – 28 (1969). 54. L. J. Melton III, S. C. Bryant, H. W. Wahner, et al., Influence of breastfeeding and other reproductive factors on bone mass later in life. Osteopor. Int. 3, 76 – 83 (1993). 55. K. Laitinen, M. Valimaki, and P. Keto, Bone mineral density measured by dual-energy x-ray absorptiometry in healthy Finnish women. Calcif. Tissue Int. 48, 224 – 231 (1991). 56. M. L. Cox, S. A. Khan, D. W. Gau et al., Determinants of forearm bone density in premenopausal women: A study in one general practice. Br. J. Gen. Pract. 41, 194 – 196 (1991). 57. D. Picard, L. G. Ste-Marie, D. Coutu et al., Premenopausal bone mineral content relates to height, weight, and calcium intake during early adulthood. Bone Miner. 4, 299 – 309 (1988). 58. A. R. P. Walker, B. Richardson, and F. Walker, The influence of numerous pregnancies and lactations on bone dimensions in South African Bantu and Caucasian mothers. Clin. Sci. 42, 189 – 196 (1972). 59. M. M. Hreschychyn, A. Hopkins, and S. Zylstra, Associations of parity, breastfeeding, and birth control pills with lumbar spine and femoral neck bone densities. Am. J. Obstet. Gynecol. 159, 318 – 322 (1988). 60. J. C. Stevenson, B. Lelels, M. Davenport et al., Determinants of bone density in normal women: Risk factors for future osteoporosis. Br. Med. J. 298, 924 – 928 (1989). 61. D. C. Bauer, W. S. Browner, J. A. Cauley, et al., Factors associated with appendicular bone mass in older women. Ann. Intern. Med. 118, 657 – 665 (1993). 62. D. Kritz-Silverstein, E. Barrett-Connor, and K. A. Hollenbach, Pregnancy and lactation as determinants of bone mineral density in postmenopausal women. Am. J. Epidemiol. 136, 1052 – 1059 (1992). 63. A. Paganini-Hill, A. Chao, R. K. Ross et al., Exercise and other factors in the prevention of hip fracture: The Leisure World Study. Epidemiology 2, 16 – 25 (1991). 64. S. Hoffman, J. A. Grisso, J. L. Kelsey et al., Parity, lactation and hip fracture. Osteopor. Int. 3, 84 – 89 (1993). 65. N. Krieger, J. L. Kelsey, T. R. Holford et al., An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 141 – 148 (1982). 66. N. Krieger, A. Gross, and G. Hunter, Dietary factors and fracture in postmenopausal women: A case-control study. Int. J. Epidemiol. 21, 953 – 958 (1992). 67. B. W. Alderman, N. S. Weiss, J. R. Daling et al., Reproductive history and postmenopausal risk of hip and forearm fracture. Am. J. Epidemiol. 124, 262 – 267 (1986). 68. C. Ribot, F. Tremollieres, J. M. Pouilles et al., Risk factors for hip fracture. Bone 14, S77 – S80 (1993). 69. R. G. Cumming and R. J. Klineberg, Breastfeeding and other reproductive factors and the risk of hip fractures in elderly women. Int. J. Epidemiol. 22, 684 – 691 (1993). 70. M. A. Laskey, A. Prentice, J. Shaw, T. Zachou, S. M. Ceesay, L. Vasquez-Velasquez, and D. R. Fraser, Breast-milk calcium con-

MARYFRAN SOWERS 71. M. L. Thomas, Calcium uptake in duodenal epithelial cells increased during lactation. Proc. Soc. Exp. Biol. Med. 196, 214 – 217 (1991). 72. G. N. Kent, R. I. Price, D. H. Gutteridge, K. J. Rosman, M. Smith, J. R. Allen, C. J. Hickling, and S. L. Blakeman, The efficiency of intestinal calcium absorption is increased in late pregnancy but not in established lactation. Calcif. Tissue Int. 48, 293 – 295 (1991). 73. B. L. Specker, N. E. Vieira, K. O. O’Brien et al., Calcium kinetics in lactating women with low and high calcium intakes. Am. J. Clin. Nutr. 59, 593 – 599 (1994). 74. D. T. Baird, A. S. McNeilly, R. S. Sawers et al., Failure of estrogeninduced discharge of luteinizing hormone in lactating women. J. Clin. Endocrinol. Metab. 49, 500 – 506 (1979). 75. A. Zarate and E. S. Canales, Endocrine aspects of lactation and postpartum infertility. J. Steroid Biochem. 27, 1023 – 1028 (1987). 76. J. A. Schlechte, B. Sherman, and R. Martin, Bone density in amenorrheic women with and without hyperprolactinemia. J. Clin. Endocrinol. Metab. 56, 1120 – 1123 (1983). 77. A. Klibanski, R. M. Neer, I. Z. Beitins et al., Decreased bone density in hyperprolactinemic women. N. Engl. J. Med. 303, 1511 – 1514 (1980). 78. H. Kalkwarf, B. Specker, D. Bianchi, J. Ranz, and M. Ho, The effect of calcium supplementation on bone density during lactation and after weaning. N. Engl. J. Med. 337, 523 – 528 (1997). 79. G. N. Kent, R. I. Price, D. H. Gutteridge et al., Human lactation: Forearm trabecular bone loss, increased bone turnover, and renal conservation of calcium and inorganic phosphate with recovery of bone mass following weaning. J. Bone Miner. Res. 5, 361 – 369 (1990). 80. M. F. Sowers, G. Corton, B. Shapiro et al., Changes in bone density with lactation. JAMA 269, 3130 – 3135 (1993). 81. B. L. Specker, R. Tsang, and M. L. Ho, Changes in calcium homeostasis over the first year postpartum: Effect of lactation and weaning. Obstet. Gynecol. 78, 56 – 62 (1991). 82. C. J. Lees, C. P. Jerome, T. C. Register, and C. S. Carlson, Changes in bone mass and bone biomarkers of cynomolgus monkeys during pregnancy and lactation. J. Clin. Endocrinol. Metab. 83, 4298 – 302 (1998). 83. M. Sowers, B. Hollis, B. Shapiro, J. Randolph, C. Janney, D. Zhang, A. Schork, M. Crutchfield, F. Stanczyk, and M. Russel-Aulet, Elevated parathyroid hormone-related peptide associated with lactation and bone density loss. JAMA 276, 549 – 554 (1996). 84. M. F. Sowers, J. Randolph, B. Shapiro, and M. Jannausch, A prospective study of bone density and pregnancy after an extended period of lactation with bone loss. Obstet. Gynecol. 85, 285 – 289 (1995). 85. A. Prentice, L. Jarjou, D. Stirling, R. Buffenstein, and S. Fairweather-Tait, Biochemical markers of calcium and bone metabolism during 18 months of lactation in Gambian women accustomed to a low calcium intake and those consuming a calcium supplement. J. Clin. Endocrinol. Metab. 83, 1059 – 1066 (1998). 86. K. Little, J. Clapp, Self-selected recreational exercise has no impact on early postpartum lactation-induced bone loss. Med. Sci. Sports Exerc. 30, 831 – 836 (1998). 87. L. E. Caird, V. Reid-Thomas, W. J. Hannan, S. Gow, and A. F. Glasier, Oral progestogen-only contraception may protect against loss of bone mass in breast-feeding women. Clin. Endocrinol. 41, 739 – 745 (1994). 88. G. Saggese, G. Baroncelli, S. Bertelloni, and C. Cipolloni, Intact parathyroid hormone levels during pregnancy, in healthy term neonates and in hypocalcemic preterm infants. Acta Paeditr. Scand. 80, 36 – 41 (1991). 89. R. Brommage and H. F. DeLuca, Regulation of bone mineral loss during lactation. Am. J. Physiol. 248, E182 – E186 (1985).

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107. 108.

109.

110.

associated bone loss in normocalcemic rates. J. Bone Miner. Res. 5, 69 – 75 (1990). F. R. Greer, R. C. Tsang, J. E. Searcy, R. S. Levin, and J. J. Steichen, Mineral homeostasis during lactation relationship to serum 1,25-dihydroxyvitamin D, 25-hydroxyvitamin D, parathyroid hormone, and calcitonin. Am. J. Clin. Nutr. 36, 431 – 437 (1982). M. F. Sowers, D. Zhang, B. W. Hollis, B. Shapiro, C. A. Janney, M. Crutchfield, M. A. Schork, F. Stanczyk, and J. Randolph, Role of calciotrophic hormones in calcium mobilization of lactation. Am. J. Clin. Nutr. 67, 284 – 291 (1998). A. E. Broadus, M. Mangin, K. Ikeda et al., Humoral hypercalcemia of cancer: Identification of a novel parathyroid hormone-like peptide. N. Engl. J. Med. 319, 556 – 563 (1988). S. Wada, H. Kitamura, Y. Matsuura et al., Parathyroid hormonerelated protein as a cause of hypercalcemia in a B-cell type malignant lymphoma. Intern. Med. 31, 968 – 972 (1992). S. J. Vargas, M. T. Gillespie, G. J. Powell et al., Localization of parathyroid hormone-related protein mRNA expression in breast cancer and metastatic lesions by in situ hybridization. J. Bone Miner. Res. 7, 971 – 979 (1992). W. A. Ratcliffe, G. E. Thompson, A. D. Care, and M. Peaker, Production of parathyroid hormone-related protein by the mammary gland of the goat. J. Endocrinol. 133, 87 – 93 (1992). M. A. Thiede and G. A. Rodan, Expression of a calcium-mobilizing parathyroid hormone-like peptide in lactating mammary tissue. Science 242, 278 – 80 (1988). M. A. Thiede, The mRNA encoding a parathyroid hormone like peptide is produced in mammary tissue in response to elevations in serum prolactin. Mol. Endocrinol. 3, 1443 – 1447 (1989). A. A. Budayr, B. P. Halloran, J. C. King, D. Diep, R. A. Nissenson, and G. J. Stewler, High levels of a parathyroid hormone-like protein in milk. Proc. Natl. Acad. Sci. USA 86, 7183 – 7185 (1989). F. M. Law, P. J. Moate, D. D. Leaver et al., Parathyroid hormonerelated protein in milk and its correlation with bovine milk calcium. J. Endocrinol. 128, 21 – 26 (1991). C. Stiegler, G. Leb, R. Kleinert et al., Plasma levels of parathyroid hormone-related peptide are elevated in hyperprolactinemia and correlated to bone density status. J. Bone Miner. Res. 10, 751 – 759 (1995). E. Canalis, T. L. McCarthy, and M. Centrella, Differential effects of continuous and transient treatment with parathyroid hormone related peptide (PTHrp) on bone collagen synthesis. Endocrinology 126, 1806 – 1812 (1990). T. Uzawa, M. Hori, S. Ejiri et al., Comparison of the effects of intermittent and continuous administration of human parathyroid hormone (1 – 34) on rat bone. Bone 16, 477 – 484 (1995). B. L. Gulson, C. W. Jameson, K. R. Mahaffey, K. J. Mizon, M. J. Korsh, and G. Vimpani, Pregnancy increases mobilization of lead from maternal skeleton. J. Lab. Clin. Med., 130, 51 – 62 (1997). M. F. Sowers, D. Eyre, B. W. Hollis et al., Biochemical markers of bone turnovers in lactating and nonlactating postpartum women. J. Clin. Endocrinol. Metab. 80, 2210 – 2216 (1995). R. Vihko and D. Apter, Endocrine characteristics of adolescent menstrual cycles, Impact of early menarche. J. Steroid. Biochem. 20, 231 – 236 (1984). M. Grumbach, G. D. Grave, and F. Mayer, “Control of the Onset of Puberty.” Wiley, New York (1974). R. E. Frisch and R. Revelle, Height and weight at menarche and a hypothesis of critical body weight and adolescent events. Science 169, 397 – 398 (1969). R. E. Frisch and J. McArthur, Menstrual cycles: Fatness as a determinant of minimum weight for height necessary for their maintenance or onset. Science 185, 949 – 951 (1974). A. Nimrod and K. J. Ryan, Aromatization of androgens by human

737

111. E. C. Scott and F. E. Johnston, Critical fat, menarche, and the maintenance of menstrual cycles: A review. J. Adol. Health Care 2, 249 – 260 (1982). 112. S. M. Garn, M. LaVelle, K. R. Rosenberg, and V. M. Hawthorne, Maturational timing as a factor in female fatness and obesity. Am. J. Clin. Nutr. 43, 879 – 883 (1986). 113. D. A. Galuska and M. F. Sowers, “Predictors of Bone Density in Nulliparous, Young Women: The Role of Age of Menarche and Body Composition.” Doctoral dissertation, University Microfilms Ann Arbor, MI (1995). 114. D. I. Rosenthal, W. Mayo-Smith, C. W. Hayes et al., Age and bone mass in premenopausal women. J. Bone Miner. Res. 4, 533 – 538 (1989). 115. E. Georgiou, K. Ntalies, A. Papageorgiou, A. Korkotsidis, and C. Proukakis, Bone mineral loss related to menstrual history. Acta Orthop. Scand. 60, 192 – 194 (1989). 116. J. C. Prior, Y. M. Bigna, M. T. Schechter et al., Spinal bone loss and ovulatory disturbances. N. Engl. J. Med. 323, 1221 – 1227 (1990). 117. M. F. R. Sowers, B. Shapiro, M. A. Gilbraith et al., Health and hormonal characteristics of women with low premenopausal bone mass. Calcif. Tissue Int. 47, 130 – 135 (1990). 118. M. F. Sowers, M. Crutchfield, B. Shapiro, B. Zhang, M. La Pietra, J. F. Randolph, and M. A. Schork, Urinary ovarian and gonadotrophin hormone levels in premenopausal women with low bone mass. J. Bone Miner. Res. 13, 1191 – 1202 (1998). 119. K. K. Steinberg, L. W. Freni-Titulaer, E. G. DePuey et al., Sex steroids and bone density in premenopausal and perimenopausal women. J. Clin. Endocrinol. Metab. 69, 533 – 539 (1989). 120. E. Baker and L. Demers, Menstrual status in female athletes: Correlation with reproductive hormones and bone density. Obstet. Gynecol. 72, 683 – 687 (1988). 121. C. E. Cann, M. C. Martin, H. K. Genant et al., Decreased spinal mineral content in amenorrheic women. JAMA 251, 626 – 629 (1984). 122. B. L. Drinkwater, K. Nilson, C. H. Chesnut et al., Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311, 277 – 281 (1984). 123. B. L. Drinkwater, K. Nilson, S. Ott et al., Bone mineral density after resumption of menses in amenorrheic athletes. JAMA 245, 380 – 382 (1986). 124. C. B. Feicht, T. S. Johnson, B. J. Martin et al., Secondary amenorrhea in athletes. Lancet 1, 1145 – 1146 (1978). 125. J. S. Lindberg, W. B. Fears, M. M. Hunt et al., Exercise-induced amenorrhea and bone density. Ann. Int. Med. 101, 647 – 648 (1984). 126. S. L. Linnell, J. M. Stager, P. W. Blue et al., Bone mineral content and menstrual regularity in female runners. Med. Sci. Sports Exerc. 16, 343 – 348 (1984). 127. R. Marcus, C. Cann, P. Madvig et al., Menstrual function and bone mass in elite women distance runners. Ann. Int. Med. 102, 158 – 163 (1985). 128. M. P. Warren, J. Brooks-Gunn, L. H. Hamilton et al., Scoliosis and fractures in young ballet dancers. N. Engl. J. Med. 314, 1348 – 1353 (1986). 129. M. E. Nelson, E. C. Fisher, P. D. Catsos et al., Diet and bone status in amenorrheic runners. Am. J. Clin. Nutr. 43, 910 – 916 (1986). 130. B. L. Drinkwater, B. Bruemner, and C. H. Chestnut, Menstrual history as a determinant of current bone density in young athletes. JAMA 263, 545 – 548 (1990). 131. K. Henley and J. L. Vaitukaitis, Exercise-induced menstrual dysfunction. Annu. Rev. Med. 39, 443 – 451 (1988). 132. T. Lloyd, C. Myers, J. R. Buchanan et al., Collegiate women athletes with irregular menses during adolescence have decreased bone density. Obstet. Gynecol. 72, 639 – 642 (1988). 133. J.-H. Ding, C. B. Sheckter, B. L. Drinkwater et al., High serum corti-

738 134. M. M. Shangold, M. L. Gatz, and B. Thysen, Acute effect of exercise on plasma concentration of prolactin and testosterone in recreational women runners. Fertil. Steril. 35, 699 – 702 (1981). 135. F. E. Chang, W. G. Dodds, M. Sullivan, et al., The acute effects of exercise on prolactin and growth hormone secretion: Comparison between sedentary women and women runners with normal and abnormal menstrual cycles. J. Clin. Endocrinol. Metab. 62, 551 – 556 (1986). 136. J. E. Jurknowski, N. L. Jones, W. C. Walker, E. V. Younglais, and J. R. Sutton, Ovarian hormonal response to exercise. J. Appl. Physiol. Respirat. Environ. Exercise. Physiol. 44, 109 – 114 (1978). 137. G. R. Brisson, M. A. Volle, D. Decarufel, M. Desharnais, and M. Tanaka, Exercise-induced disassociation of the blood prolactin response in young women according to their sports habits. Horm. Metab. Res. 12, 201 – 205 (1980). 138. E. Dale, D. H. Gerlach, and A. L. Wilhite, Menstrual dysfunction of distance runners. Obstet. Gynecol. 54, 47 – 53 (1979). 139. R. E. Frisch, Body fat, menarche, fitness and fertility. Hum. Repaired. 2, 521 – 533 (1987). 140. K. H. Myburgh, L. K. Bachrach, B. Lewis, K. Kent, and R. Marcus, Low bone mineral density at axial and appendicular sites in amenorrheic athletes. Med. Sci. Sports Exerc. 25, 1197 – 1202 (1993). 141. P. J. Kelly, J. A. Eisman, M. C. Stuart et al., Somatomedin-C, physical fitness, and bone density. J. Clin. Endocrinol. Metab. 70, 718 – 723 (1990). 142. J. W. T. Ayers, G. P. Gidwani, I. M. V. Schmidt et al., Osteoporosis in hypoestrogenic young women with anorexia nervosa. Fertil. Steril. 41, 224 – 228 (1984). 143. B. M. K. Biller, V. Saxe, D. B. Herzog et al., Mechanisms of osteoporosis in adult and adolescent women with anorexia nervosa. J. Clin. Endocrinol. Metab. 68, 548 – 554 (1989). 144. N. A. Rigotti, S. R. Nussbaum, D. B. Herzog et al., Osteoporosis in women with anorexia nervosa. N. Engl. J. Med. 311, 1601 – 1606 (1984). 145. L. K. Bachrach, D. Guido, D. Katzman et al., Decreased bone density in adolescent girls with anorexia nervosa. Pediatrics 86, 440 – 447 (1990). 146. M. Myerson, B. Gutin, M. P. Warren et al., Total body bone density in amenorrheic runners. Obstet. Gynecol. 79, 973 – 978 (1992). 147. C. A. Raymond, Long-term sequelae pondered in anorexia nervosa. JAMA 257, 3324 – 3325 (1987). 148. A. Klibanski and S. L. Greenspan, Increase in bone mass after treatment of hyperprolactinemic amenorrhea. N. Engl. J. Med. 315, 542 – 546 (1986). 149. J. Schlechte, L. Walkner, and M. Kathol, A longitudinal analysis of premenopausal bone loss in healthy women and women with hyperprolactinemia. J. Clin. Endocrinol. Metab. 75, 6989 – 6703 (1992). 150. A. Klibanski, B. M. Biller, D. I. Rosenthal, D. A. Schoenfeld, and V. Saxe, Effects of prolactin and estrogen deficiency in amenorrheic bone loss. J. Clin. Endocrinol. Metab. 67, 124 – 130 (1988). 151. C. Di Carlo, A. Patel, Z. Shobam, M. L. Hall, J. MacDougall, and H. S. Jacobs, Polycystic ovaries as a relative protective factor for bone mineral loss in young women with amenorrhea. Fertil. Steril. 57, 314 – 319 (1992). 152. N. F. Goldsmith and J. O. Johnston, Bone mineral: Effects of oral contraceptives, pregnancy and lactation. J. Bone Joint Surg. Am. 57, 657 – 668 (1975). 153. R. Lindsay, J. Tohme, and B. Kanders, The effect of oral contraceptive use on vertebral bone mass in pre- and post-menopausal women. Contraception 34, 333 – 340 (1986). 154. E. Seeman, G. I. Szmukler, C. Formica, C. Tsalamandris, and R. Mestrovic, Osteoporosis in anorexia nervosa: The influence of peak

MARYFRAN SOWERS 155. H. Enzelsberger, M. Metka, G. Heytmanek, B. Schurz, C. H. Kurz, and M. Kusztrich, Influence of oral contraceptive use on bone density in climacteric women. Maturitas 9, 375 – 378 (1988). 156. R. Armamento-Villareal, D. T. Villareal, L. V. Avioli, and R. Civitelli, Estrogen status and heredity are major determinants of premenopausal bone mass. J. Clin. Invest. 90, 2464 – 2471 (1992). 157. M. Kleerekoper, R. S. Brienza, L. R. Schultz, and C. C. Johnson, Oral contraceptive use may protect against low bone mass. Arch. Intern. Med. 151, 1971 – 1976 (1991). 158. D. Kritz-Silverstein and E. Barrett-Conner, Bone mineral density in postmenopausal women as determined by prior oral contraceptive use. Am. J. Public. Health 83, 100 – 102 (1993). 159. R. B. Mazess and H. S. Barden, Bone density in pre-menopausal women: Effects of age, dietary intake, physical activity, smoking and birth-control pills. Am. J. Clin. Nutr. 53, 132 – 142 (1991). 160. T. Lloyd, J. R. Buchanan, G. R. Ursino, C. Myers, G. Woodward, and D. R. Halbert, Long-term oral contraceptive use does not effect trabecular bone density. Am. J. Obstet. Gynecol. 160, 402 – 404 (1989). 161. S. Murphy, K. T. Khaw, and J. E. Compston, Lack of relationship between hip and spine bone mineral density and oral contraceptive use. Eur. J. Clin. Invest. 23, 108 – 111 (1993). 162. J. A. Fortney, P. J. Feldblum, R. V. Talmage, J. Zhang, and S. E. Godwin, Bone mineral density and history of oral contraceptive use. J. Repaired. Med. 39, 105 – 109 (1994). 163. B. Berning, C. van Kuijk, H. E. Schutte, J. W. Kuiper, A. C. Drogendijk, and B. C. J. M. Fauser, Determinants of lumbar bone mineral density in normal weight, non-smoking women soon after menopause. A study using clinical data and quantitative computed tomography. Bone Miner. 21, 129 – 139 (1993). 164. R. R. Recker, M. Davies, S. M. Hinders et al., Bone gain in young adult women. JAMA 268, 2403 – 2408 (1992). 165. B. Mais, F. Fruzzetti, S. Ajossa, A. M. Paoletti, S. Guerriero, and G. B. Melis, Bone metabolism in young women taking a monophasic pill containing 20 mcg ethinyl estradiol: A prospective study. Contraception 48, 445 – 452 (1993). 166. M. Gambacciani, A. Spinetti, F. Taponeco, B. Cappagli, L. Piaggesi, and P. Fioretti, Longitudinal evaluation of perimenopausal vertebral bone loss: Effects of low-dose oral contraceptive preparation on bone mineral density and metabolism. Obstet. Gynecol. 83, 392 – 396 (1994). 167. F. Polatti, F. Perotti, N. Filippa, D. Gallina, and R. Nappi, Bone mass and long term monophasic oral contraceptive treatment in young women. Contraception 51, 221 – 224 (1995). 168. P. Garnero, E. Sornay-Rendu, and D. Delmas, Decreased bone turnover in oral contraceptive users. Bone 16, 499 – 503 (1995). 169. C. Cooper, P. Hannaford, P. Croft, and C. R. Kay, Oral contraceptive pill use and fractures in women: A prospective study. Bone 14, 41 – 45 (1993). 170. T. Cundy, M. Evans, H. Roberts, D. Wattie, R. Ames, and I. R. Reid, Bone density in women receiving depot medroxyprogesterone acetate for contraception. Br. Med. J. 303, 13 – 16 (1991). 171. T. Cundy, J. Cornish, H. Roberts, H. Elder, and S. Reed, Spinal bone density in women using depot medroxyprogesterone contraception. Obstet. Gynecol. 92, 569 – 573 (1998). 172. T. Cundy, J. Cornish, M. C. Evans, H. Roberts, and I. R. Reid, Recovery of bone density in women who stop using medroxyprogesterone acetate. Br. J. Med. 308, 247 – 248 (1994). 173. B. Gbolade, S. Elllis, and S. Randall, Bone density in long-term users of depot medroxyprogesterone acetate. Br. J. Obstet. Gynaecol. 105, 790 – 794 (1998). 174. C. Canterbury and R. Hatcher, Serum estradiol levels in women using depo-provera for more than a year. Proceedings from the Contraceptive Technology Conference, 68 – 79. Washington, DC (1997).

CHAPTER 29 Premenopausal Reproductive and Hormonal Characteristics and the Risk for Osteoporosis 176. D. Mishell, K. Kharma, and I. Thyornecroft, Estrogenic activity in women receiving an injectable progestogen for contraception. Am. J. Obstet. Gynecol. 113, 372 – 376 (1972). 177. E. M. Belsey, Vaginal bleeding patterns among women using one natural and eight hormonal methods of contraception. Contraception 38, 181 – 206 (1988). 178. A. Ortiz, F. Hiroi, Z. Stanczyk, Goebelsmann, and R. Mishell, Serum medroxyprogesterone acetate (MPA) concentrations and ovarian function following intramuscular injection of depo-MPA. J. Endocrinol. Metab. 44, 32 – 38 (1977). 179. H. Sangi-Haghpeykar, A. Poindexter, L. Bateman and R. Ditmore, Experiences of injectable contraceptive users in an urban setting. Obstet. Gynecol. 88, 227 – 233 (1996). 180. P. Schwallie, and R. Assenzo, Contraceptive use-efficiency study utilizing medroxyprogesterone acetate administered as an intramuscular injection once every 90 days. Fertil. Steril. 24, 331 – 9 (1973). 181. C. Johnson, J. Miller, C. Slemenda, T. Reister, S. Hui, J. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). 182. T. Lloyd, R. Buchannan, G. Ursinio, L. Myers, G. Woodard, and D. Hilbert, Long term oral contraceptive use does not affect trabecular bone density. Am. J. Obstet. Gynecol. 160, 402 – 404 (1989). 183. J. Nieves, A. Golden, E. Siris, J. Kelsey, and R. Lindsay, Teenage current calcium intake are related to bone mineral density of the hip and forearm in women aged 30 to 39 years. Am. J. Epidemiol. 15, 343 – 351 (1995). 184. J. Kristinsson, O. Valdimarsson, L. Steingrimsdottir, and G. Sigurdsson, Relation between calcium intake, grip strength and bone mineral density in the forearms of girls aged 13 and 15. J. Intern. Med. 236, 385 – 390 (1994). 185. M. Valimaki, M. Karkkainen, C. Lunberg-Allardt, K. Lantina, E. Alhara, J. Herkkma, O. Impivaara, P. Makela, J. Palmgra, and R. Seppura, Exercise, smoking, and calcium intake during adolescence and early determinants of peak bone mass. Br. Med. J. 309, 230 – 235 (1994). 186. H. D. Rees, R. W. Bonsall, and R. P. Michael, Localization of the synthetic progestin 3H-ORG 2058 in neurons of the primate brain: Evidence for the site of action of progestins on behavior. J. Comp. Neurol. 235, 336 – 342 (1985). 187. I. Fraser and E. Weisberg, A comprehensive review of injectable contraception with a special emphasis on depot medroxyprogesterone acetate. Med. J. Aust. 1, 1 – 20 (1981). 188. S. El Mahgoub, Body weight and cycle control of injectable contraceptives. J. Reprod. Med. 24, 119 – 126 (1980). 189. G. Leiman, Depo-medroxyprogesterone acetate as a contraceptive agent: Its effect on weight and blood pressure. Am. J. Obstet. Gynecol. 97 – 101 (1972). 190. M. Mukherjea, P. Mukherjee, and R. Biswas, Long-term contraception with Depo-Provera: A clinical evaluation. Intern. J. Fertil. 25, 122 – 126 (1980). 191. L. Moore, R. Valuk, C. McDougall, and W. Fink, A comparative study of one-year weight gain among users of medroxyprogesterone acetate, levonorgestrol implants, and oral contraceptives. Contraception 52, 215 – 220 (1995). 192. R. Mainwaring, H. Hales, K. Stevenson, H. Hataska, M. Poulson, K. Jones, and M. Peterson, Metabolic parameters, bleeding, and

193.

194.

195. 196.

197.

198.

199.

200. 201.

202.

203.

204. 205.

206.

207.

208.

209.

739

weight changes in U.S. women using progestin only contraceptives. Contraception 51, 149 – 153 (1994). A. Rodin, M. Murby, A. Smith, D. Caleffi, I. Fentiman, M. Chapman, and S. Fogelman, Premenopausal bone loss in lumbar spine and neck of femur: A study of 225 Caucasian women. Bone 11, 1 – 5 (1990). E. Samojlik, E. Kirshner, M. Samojlik, E. Drejka, G. Szmal, G. Schneider, and H. Ertel, Elevated production and metabolic clearance rates of androgen in morbidly obese women. J. Clin. Endocrinol. Metabol. 49, 949 – 954 (1990). P. Siiteri, Adipose tissue as a source of hormones. Am. J. Clin. Nutr. 45, 277 – 282 (1987). J. C. Gallagher, M. M. Young, and B. E. C. Nordin, Effects of artificial menopause on plasma and urine calcium and phosphate. Clin. Endocrinol. 1, 57 – 64 (1972). L. S. Richelson, H. W. Wahner, L. J. Melton III, and B. L. Riggs, Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N. Engl. J. Med. 311, 1273 – 1275 (1984). J. M. Aitken, D. M. Hart, R. Lindsay, J. B. Anderson, D. A. Smith, and G. M. Wilson, Prevention of bone loss following oophorectomy in premenopausal women. Ir. J. Med. Sci. 12, 607 – 614 (1976). S. Meema and H. E. Meema, Evaluation of cortical bone mass, thickness and density by z-scores in osteopenic conditions and in relation to menopause and estrogen treatment. Skel. Radiol. 8, 259 – 268 (1982). L. Svanberg, Effects of estrogen deficiency in women castrated when young. Acta Obstet. Gynecol. Scand. Suppl. 106, 11 – 15 (1982). H. Ohta, K. Makita, Y. Suda, T. Ikeda, T. Masuzawa, and S. Nozawa, Influence of oophorectomy on serum levels of sex steroids and bone metabolism and assessment of bone mineral density in lumbar trabecular bone by QCT-c value. J. Bone Miner. Res. 7, 659 – 665 (1992). H. Ohta, T. Masuzawa, T. Ikeda, Y. Suda, K. Makita, and S. Nozawa, Which is more osteoporosis-inducing, menopause or oophorectomy. Bone Miner. 19, 273 – 285 (1992). J. J. Stepan, J. Pospichal, J. Presl, and V. Pacovsky, Bone loss and biochemical indices of bone remodeling in surgically induced postmenopausal women. Bone 8, 279 – 284 (1987). J. Reeves, Bone turnover and trabecular plate survival after artificial menopause. Br. Med. J. 295, 757 – 760 (1987). H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gordan, Quantitative computed tomography of vertebral spongiosa: A sensitive method for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). M. M. Hreshchyshyn, A. Hopkins, S. Zylstra, and M. Anbar, Effects of natural menopause, hysterectomy, and oophorectomy on lumbar spine and femoral neck bone densities. Obstet. Gynecol. 72, 631 – 638 (1988). R. Lindsay, D. M. Hart, C. Forrest, et al., Prevention of spinal osteoporosis in oophorectomized women. Lancet 2, 1151 – 1154 (1980). M. Gambacciani, A. Spinetti, B. Cappagli, F. Taponeco, R. Felipetto, D. Parrini, N. Cappelli, and P. Fioretti, Effects of ipriflavone administration on bone mass and metabolism in ovariectomized women. J. Endocrinol. Invest. 16, 333 – 337 (1993). G. F. Mazzuoli, S. Tabolli, F. Bigi, C. Baltorta, S. Minisola, D. Diaciniti, L. Scarnecchia, G. Bianchi, M. Piolini, and S. Dell’Acqua, Effects of salmon calcitonin on the bone loss induced by ovariectomy. Calcif. Tissue Int. 47, 209 – 214 (1990).

CHAPTER 30

Postmenopausal Endogenous and Exogenous Hormones, Degree of Obesity, Thiazide Diuretics, and Risk of Osteoporosis JANE A. CAULEY AND LORAN M. SALAMONE University of Pittsburgh, Pittsburgh, Pennsylvania 15261

I. II. III. IV.

V. Obesity VI. Overall Summary References

Introduction Thiazide Diuretics Exogenous Estrogen Endogenous Estrogen, Bone Mass, and Fractures

I. INTRODUCTION

relationships to bone mass and fracture risk, it is critical that one consider the underlying factor of obesity.

This chapter reviews the relationship among thiazide diuretics, estrogen, both endogenous and exogenous, and the degree of obesity with bone mass and fractures among postmenopausal women. All of these factors may be interrelated to each other. For example, thiazide diuretic users tend to be more obese [1]. However, studies of selection factors for estrogen replacement use have consistently found that estrogen users are less obese than nonusers [2]. Moreover, obese women have higher levels of circulating estrogens, reflecting the aromatization of androstenedione to estrone in fat tissue [3]. Hence, when evaluating these

OSTEOPOROSIS, SECOND EDITION VOLUME 1

II. THIAZIDE DIURETICS A. Thiazide Diuretics and Bone Mass Cross-sectional studies have compared bone mass, either bone mineral content (BMC) or bone mineral density (BMD), of thiazide diuretic users and nonusers. Thiazide diuretic users tend to have about 4 to 5% higher appendicular bone mass than nonusers (Table 1) [1,4 – 10]. The studies in

741

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

TABLE 1 Study

Design

Subjects

A Summary of the Research on Thiazide Diuretics (TD) and Bone Mass Bone mass

TD duration/dose

TD prevalence

Results

Comments

Cross-sectional

1368 Men: Japanese ancestry: mean age  68 years

Distal radius and ulna; proximal radius and ulna calcaneus

7.25 years

n  323 (23.6%)

At all sites, TD users had 1.6 – 4.4% higher mean BMC than nonusers

Adjusted for BMI and age; no dosage effect; trend for increased BMC with duration but not significant

Sowers et al., 1985 [5]

Cross-sectional

324 White women ages 55 – 80

Mid-radius

N/A

n  99 (30.6%)

Mean BMD  0.647  0.008 in users, 0.618  0.005 in nonusers (P  0.01; 4.7% difference)

Adjusted for age, muscle area

Wasnich et al., 1986 [6]

Cross-sectional

n  993 women; mean age  63.4, all of Japanese origin

Distal and proximal radius, calcaneus, lumbar spine on a subset

9 Years; 46% 25 mg HCT; 35%, other HCT; 19%, other TD

n  260 (29.2%)

TD users had 4.9  5.4% higher BMC (p  0.01) than untreated subjects at all sites; ERT and TD users had 12 – 20% higher BMC than untreated

Adjusted for age, height, weight

Cauley et al., 1993 [1]

Cross-sectional

9704 Nonblack women age  65 Years; mean age  72 years

Distal and proximal radius, calcaneus

11.2 Years; median dose: 37.5 mg/day, 10 years; 50.0 mg/day,  10 years

n  2637 (27%)

Adjusted mean BMD 4 – 5% higher for current users than for never users; current users who used TD for  10 years had 1.2 – 2.4% higher BMD than current users, 10 years

Adjusted for age, weight, calcium intake, ERT

AdlandDavenport et al., 1985 [7]

Cross-sectional

Thiazide users, mean age  63.3; compared with 54 nonusers, mean age  59.8; all postmenopausal white women

Distal and mid-shaft radius

6.1  5.9 Years; mean HCT dose, 22.9 mg/ day

Mean BMC 2.4 – 6.1% and BMD 3.1 – 4.4% higher, for users than nonusers (NS)

Pair-matched on type of menopause, years postmenopausal, ERT, calcium and vitamin D intake, and BMI

Lindsay et al., 1987 [8]

Cross-sectional

182 White women with hypertension; 182 white women, no hypertension

Lumbar spine, femoral neck

N/A

n  105 (57.7%)

BMD 11 – 16% higher among TD users compared with controls; BMC 7 – 11% higher among TD users compared with hypertensives not treated with TD.

Age matched (abstract only)

Glynn et al., 1994 [9]

Cross-sectional

523 White men, all age50 years mean age  66 years

Lumbar spine, femoral neck, greater trochanter

N/A

n  115 (22%)

BMD 3.6 – 4.9% higher in TD users

Age adjusted

Morton et al., 1994 [10]

Cross-sectional

Rancho Bernado, 681 men; 1015 women ages 44 – 98 years, all white

Mid and distal radius, hip, spine

Men, 7.7 years; women, 9.2 years; mean dose, 36.7 mg/day

Current; men, n  103 (15%); women, n  240 (24%)

Women: TD users bad 3 – 8% higher BMD than never users Men: there was no effect of TD use on BMD

No association with dose; adjusted for age, BMI, smoking, and estrogen

Wasnich et al., 1990 [11]

Prospective; mean follow-up 5 years

1017 Men of Japanese ancestry; mean age  67.8 years at entry

Calcaneus, distal radius, proximal radius

11.9 Years; most common dose, 25 mg HCT

n  378 (32.0%)

Differences in bone loss over 5 years (TD users compared to nonusers): calcaneus p 49.2% (P  0.005); proximal radius p 45.3% (P  0.01); distal radius p28.8% (P  0.02)

Subjects taking nonthiazide anti-hypertensives had faster loss rates than nonusers

742

Wasnich et al., 1983 [4]

Prospective; 1-year follow-up

246 Women mean age  65 years for TD users; 61 years for nonusers

Whole-body and spinal BMD measured at baseline, 6 months, and 1 year

11  7 Years; HCT dose, 12.5 to 50 mg/day

n  25 (10.2%)

TD users had significantly higher BMD at both sites at baseline (13.6% higher spine BMD, 4.5% higher whole-body BMD); in winter/spring, bone loss reduced in TD users compared with nonusers; L2 – 4 0.46% vs  1.02%, P 0.02; whole body, 0.13% vs  0.67%, P 0.04, respectively

Adjusted for age, community, BMI; no group differences in physical activity or calcium and vitamin D intake; dose effect apparent for whole body BMD

Sowers el at., 1993 [13]

Prospective; 5-year follow-up

435 Caucasian women, ages 55 – 80 at baseline

Mid-radius

N/A

Current, n  21 (4.9%); Past, n  56 (15.5%)

Over 5 years current TD users lost significantly less bone mass than past or never users: (4.9%, current; 7.4%, past; and 7.2%, never) (P  0.004)

Adjusted for age, community, BMI, ERT, magnitude of effect on bone loss: ERT  TD

Christiansen et al.., 1980 [14]

Randomized trial

315 Northern European women in early postmenopause, mean age 50.1; mean duration of menopause 19.8 months; 26 women assigned to thiazide alone, 21 completed trial

Bilateral distal forearm

2 Years; Bendroflurnethiazide, 5 mg/day

N/A

Mean BMC p 1.5% over 2 years in TD only group; p 3.3% in placebo group (P  0.05)

Effect of TD primarily due to unchanged BMC during first 6 months; magnitude of effect on bone loss: ERT  TD

Transbot et al., 1982 [15]

Randomized trial

63 Women in early postmenopause, mean age  50.1; 54 completed 2 years, 38 completed 3 years

Bilateral distal forearm

2 Years treatment; 1 year placebo; Bendrofluemethiazide; 5 mg/day

Mean BMC values at end of trial: TD, 95.2% of baseline, placebo 94.1% (P  0.05)

Thiazide postponed BMC drop by 6 months; thereafter rate of loss was similar to placebo group

Giles et al., 1992 [16]

Randomized trial

32 Hypertensives assigned to either nitrendipine or HCT; mean age  62.9; 84% male; 72% black

Lumbar spine

1 Year; HCT, 50 mg/day

Lumbar BMD increased 5.6% and 6.0% in the TD group at 24 and 52 weeks, respectively (P  0.05); BMD unchanged in the nitrendipine group

Unblinded after 8 weeks

Wasnich et al. 1993 [17]

Randomized trial

113 Women enrolled in SHEP; all age  60

Calcaneus, distal radius, proximal radius

Follow-up over 4.3 years

BMD of TD users increased 1.6 to 4.1% compared to 0.7% to 1.10% in the placebo group

Data available in abstract form only

743

DawsonHughes and Harris, 1993 [12]

n  54, TD; n  59, placebo

Note. BMD, bone mineral density; TD, thiazide diuretic; p, decrease; BMC, bone mineral content; NS, not significant; ERT, estrogen replacement group; HCT, hydrochlorothiazide.

744 Table 1 have been organized by design. Lumbar spine BMD was found to be 5% higher in Japanese-American women [6] and 14% higher in Caucasian women [12] reporting current use of thiazide diuretics than nonusers. Hip BMD has also been shown to be higher in women [8,10] and men [9] using thiazide diuretics. Only one published study included whole body BMD and, consistent with the association between axial and appendicular BMD sites, whole body BMD was about 5% higher in women reporting current use of thiazide diuretics (TD) than in women not currently using them [12]. Most of these cross-sectional studies controlled for the confounding effects of age and the degree of obesity or body weight. Three observational prospective studies have been published that examined the effect of thiazide diuretics on rates of bone loss (Table 1) [11 – 13]. Wasnich et al. [11] studied 1017 men of Japanese ancestry. After 5 years, the annual rate of loss at the calcaneus was 49% lower for men on thiazide diuretics than that seen with non-TD users. Similar effects were observed for the proximal and distal radius. Dawson-Hughes and Harris [12] examined the effect of thiazide diuretics on seasonal changes in BMD. Winter/spring bone loss was reduced among TD users compared to that seen in nonusers: TD users experienced a 0.5% increase in spine BMD compared to a 1% loss among nonusers. Similar differences were observed for the whole body BMD. BMD tended to increase in the summer/spring at both the spine and the whole body, and in both TD users and nonusers. Sowers et al. [13] studied the determinants of loss in midradial bone mass over 5 years among 435 white older women ages 55 – 80 at baseline. Over 5 years, current TD users lost significantly less BMD (4.9%) than either past (7.4%) or never (7.2%) users. The effect of current TD use on bone loss was considerably less than that observed for long-term estrogen users. Hypercalciuria has been associated with osteoporosis, especially among older men. In a case series of five male patients, age 42 – 66 years with hypercalcuria and osteoporosis, treatment with 50 mg of hydrochlorothiazide for about 8 months was associated with an 8% increase in spine BMD and a 3% increase in hip BMD [14]. Low-dose hydrochlorothiazide treatment may be an effective treatment for hypercalciuria in older men. A randomized controlled clinical trial is the accepted gold standard for evaluating clinical interventions. With respect to thiazide diuretics, few clinical trials have been published [15 – 18]. Christiansen et al. [15] randomized 315 early postmenopausal women, all within 2 years of menopause, into 1 of 10 groups, including 7 treatment groups and 3 placebo groups. After 2 years, the bone mineral content of the distal forearm declined 1.5% in the TD group and 3.3% in the placebo group and increased 2.5% among the hormone users. Most of the effect of TD was limited to the first 6 months of treatment. Transbol et al. [16] randomized 63 women, at age 50 years, to receive thiazide diuretics versus placebo for 2

CAULEY AND SALAMONE

years, followed by 1 year of placebo in both groups. The sample of women in this study may be a subset of the clinical trial of Christiansen and associates [15]. No change in BMC was observed among the women randomized to TD for the first 6 months of the study. Thereafter, the rate of bone loss was similar in both the TD and the placebo despite a sustained decrease in the urinary calcium excretion among TD users. Giles et al. [17] randomized 32 hypertensive subjects to either 50 mg of hydrochlorothiazide or 10 mg of nitrendipine, a calcium antagonist. After 1 year, lumbar spine BMD did not change in the group receiving nitrendipine but increased about 5% in the hydrochlorothiazide group. As part of the Systolic Hypertension in the Elderly Project (SHEP), appendicular bone measurements were available on a subset of women at the Hawaiian SHEP clinic [18]. Over about 4 years, BMD of the TD users increased 1.6 to 4% compared to 0.7 to 1.1% among the placebo group, suggesting that TD could prevent bone loss, even in women over 60 years of age. Most of the research on TD and bone mass has focused on white postmenopausal women, but effects have generally been similar among white older men and men and women of Japanese ancestry. This suggests that TD may influence bone mass in a similar fashion across ethnic groups, although we have little data on TD and bone mass in black populations. The effect of TD on bone could differ according to when the TD was initiated with respect to menopause. In most of the cross-sectional studies, the subjects tended to be 60 years of age or older and well past menopause. It is conceivable that failure to see a more marked effect in the Danish clinical trial [15] reflected the fact that these women were all within 2 years of menopause. At this time, the effect of the estrogen deficiency may outweigh any benefit of TD. The effect of past use of TD on BMD has also been examined. In one study [1], past users had significantly higher appendicular BMD than never users, but the magnitude of the difference was considerably less than that observed for current TD users. For example, the mean distal radius BMD (g/cm2) was 0.359 among never users and 0.362 g/cm2 among past users, a less than 1% difference. In contrast, the BMD was 4 to 5% greater among current TD users than among never users. Prospectively, the rate of radial bone loss was similar among past and never users [13]. Few studies have examined the effect of duration of use of TD and bone mass. We compared BMD among current users who had used TD for more than 10 years with that among current TD users who reported use for 10 or fewer years [1]. The BMD of both short- and long-term users was significantly greater than the BMD of never users. Nevertheless, long-term users had about a 1 to 2% higher BMD than short-term users. Among men of Japanese ancestry, there was a trend for increasing BMC with duration of TD but results were not significant [4]. In most of the cross-sectional

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

studies, the duration of TD was quite long, ranging from 6 to 11 years. In contrast, most of the clinical trials have been considerably shorter in duration — 1 – 2 years [15 – 17]. It may take significantly longer than 2 years for TD to positively influence BMD. Information on the dose of TD was not often provided in the cross-sectional studies. In a few studies, the mean daily dose ranged from 25 to 50 mg hydrochlorothiazide [1,6,7,10 – 12,17]. Two studies have examined the effect of dose of TD on BMD. Wasnich et al. [4] reported no dosage effect in their study of Japanese men. These authors have suggested, however, that failure to see an effect of TD on bone loss in the Danish clinical trial reflects the relative low dosage of TD [12]. The subjects in the Danish clinical trials [15,16] were given a 5-mg dose of bendroflumethiazide, which is equivalent in terms of an effect on sodium excretion to a 6.4-mg dose of hydrochlorothiazide [12]. The usual indication for prescribing thiazide diuretics is hypertension. It is possible that hypertension by itself is associated with increased bone mass. However, it has been shown that among hypertensive women, BMD was higher among TD users than among nonusers, although the differences in BMD were greater comparing hypertensive TD users with normal controls [8]. Wasnich and colleagues [6] showed that the cross-sectional differences in BMD were similar comparing TD users with nonusers and with hypertensive nonusers: BMC was higher among TD users. Finally, prospectively, hypertensive men on other antihypertensive agents experienced an increased rate of loss, TD users experienced a significantly decreased rate of loss [18]. Results from the Rancho Bernardo Study were similar when hypertensive women were excluded from the analysis, suggesting that the effect of the TD was not due to hypertension [10]. Nevertheless, little information is available on TD among normotensive women. In most studies, the effect of TD on BMD was about 50% that observed for estrogen replacement therapy (ERT) [13]. Wasnich et al. [6] suggested that ERT and TD had additive effects on BMD. This was not supported by data from the Danish clinical trial, where the effect on BMC was similar among women randomized to ERT and women randomized to ERT plus thiazide [15]. It is not known if this reflects the lower dose of TD or the short duration of the trial.

B. Thiazide Diuretics and Fracture Data summarizing the effect of TD on fracture prevalence or incidence have been summarized in Table 2 [1,4 – 7,19 – 30]. The studies have been organized by design and chronologically. Two studies grouped all diuretics together [19,22]. Because the effects of TD and non-TD diuretics differ, the results of these two studies are difficult to interpret.

745 For all fractures (all nonspine or all fractures), most of the studies have failed to find a statistically significant difference between TD users and nonusers [1,4,5]. In one study the relative risk for self-reported nonspine fractures was lower among TD users than nonusers (RR  0.47), but the results were not statistically significant (P  .07) [6]. Similarly, Adland-Davenport et al. [7] examined the self-reported prevalence of fracture of the ribs, hip, wrist, and spine by TD use. TD users were found to have a 40% lower risk of these fractures than nonusers, but the results were not significant (P  0.45). In a prospective study of almost 10,000 older white women, there was no effect of TD on either the risk of any nonspine fracture (RR  0.95; 95% CI 0.83 to 1.09) or the risk of all osteoporotic fractures (RR  0.94; 0.81 to 1.10) [1]. No association was found between duration of use and the risk for either any nonspinal fracture or osteoporotic fracture [1]. The prevalence of vertebral deformities was compared among TD users and nonusers. Among men [4] and women [6] of Japanese origin, the prevalence of spinal fractures was reduced among TD users, but the results were not significant. Among Caucasian women in the United States, the prevalence of vertebral deformities as measured by morphometry among current users of TD who had taken these drugs for more than 10 years (25%; 95% CI, 19 to 30%) was similar to the prevalence among women who had never taken TD (24%; CI, 22 to 26%), past users (20%; CI, 15 to 25%), and current users who had used TD for 10 or fewer years (22%; CI, 18 to 26%) [1]. To our knowledge the effect of TD on vertebral fractures has not been examined prospectively. The relative risk of hip fractures by TD use varies from a relative risk (or odds ratio) of 0.68 [28] to 1.60 [25]. A recent meta-analysis suggested a 20% reduction in the risk of hip fracture (OR  0.82; 0.73 to 0.91) among current users of TD [31]. The case-control study of Heidrich et al. [25] deserves comment because it is the outlier among studies that examined TD and hip fractures. It was a large case-control study involving almost 500 hip fractures identified through medical records. Age- and gender-matched community controls were chosen. Pharmacy records were used for information on TD. The crude odds ratio (OR) for TD and hip fracture was 1.1 and not statistically significant. However, when they adjusted for the potential confounding effects of nursing home residence; previous hospitalizations; history of stroke, alcoholism or organic brain syndrome; body weight; leg paralysis; and use of phenobarbital, corticosteroids, or other diuretics, a significantly increased risk of hip fracture was found among women who reported current use of TD. All of the studies that found a protective association between TD and hip fractures have been observational, which allows the possibility that users of TD have a lower risk of hip fracture because of other confounding factors. The study by Heidrich et al. [25] was

TABLE 2

A Summary of the Research on Thiazide Diuretics (TD) and Risk of Fracture

746

Study

Design

TD duration/dose

TD prevalence

Wasnich et al., 1983 [4]

Cross-sectional

n  1368 men; mean age age  68 years, Japanese origin

Subjects

Nonspine, self-report since age 45; spine, lumbar films (50% sample)

Fracture ascertainment

7.25 Years

n  323 (23.6%)

No significant differences between users and nonusers in overall fracture prevalence; prevalence of lumbar spinal compression fractures 3.5% for users, 6.3% for nonusers (NS)

Results

Comments

Hale et al., 1984 [18]

Cross-sectional

n  3192; mean age  75.1

Self-report; any fracture in past 5 years

N/A

n  446 (29.6%)

For all diuretics: males, OR  1.81 (NS); females, OR  0.61 (0.03)

Results not stratified by type of diuretic

Sowers et al., 1985 [5]

Cross-sectional

324 Ambulatory white women aged 55 – 80

Self-report

N/A

n  99 (30.6%)

No difference in reported fracture between TD users and nonusers (actual data not provided)

Adjusted for age, muscle area

Wasnich et al., 1986 [6]

Cross-sectional

n  993 women; mean age  63.4, Japanese origin

Nonspine, self-report in past 5 years; spine, lumbar films (50% sample)

9 Years; 46%, 25 mg HCT; 35% other HCT; 19%, other TD

n  290 (29.2%)

For TD users vs nonusers: nonspine, RR  0.47 (P  0.07); spine, RR  0.79 (P  0.17)

Adjusted for age, height, weight, years of follow-up; ERT and TD  ERT alone  TD

AdlandDavenport et al., 1985 [7]

Cross-sectional

54 TD Users, mean age  63.3 compared with 54 non-TD users, mean age  59.8; all white postmenopausal women

Self-report: fracture of ribs, hip, wrist, vertebrae

6.1  5.9 years; mean dose; HCT 22.9 mg/ day

N/A

OR  0.60 (P  0.45)

Adjusted for type of menopause, years menopausal, duration of estrogen use, BMI, calcium and vitamin D

Paganini-Hill et al., 1981 [19]

Case-control

Cases: 83 female all  age 80; 166 community controls, matched to cases on age and race

Hip fractures: hospital discharge record review

Duration  5 years, cases n  9; controls, n  12;  5 years, cases n  11; controls n  27

Cases, 20 (24%); Controls, 39 (23%)

 5 yrs, OR  1.49, NS  5 yrs, OR  0.81, NS

Use of other antihypertensive agents associated with increased risk of hip fracture but it was not significant

Rashiq and Logan, 1986 [20]

Case-control

Cases: 102, all  age 60; age and sexmatched controls

Hip fractures: medical records review

N/A

Cases, n  5 (5.9%); Controls, n  19 (18.1%)

OR(95% CI)  0.28 (0.12, 0.66)

Adjusted for age, sex

Taggart, 1988 [21]

Case-control

Cases, 282 female, mean age  83 years; controls, 145 from general practice lists, mean age  81 years

Hip fractures; medical record

N/A

Cases, n  42 (15.0%); Controls, n  21 (14.5%)

All diuretics combined: RR  1.60 (0.05  P  0.10)

RR adjusted for age

747

Stevens and Mulrow, 1989 [22]

Case-control

Cases, 173, mean age 79.4 years; controls  134 emergency surgery, mean age  77.2 years

Hip fractures: medical records

N/A

Cases, n  35 (20.3%); Controls, n  25 (18.8%)

Current TD user vs not current: OR (95% CI)  0.95 (0.46, 1.97)

Adjusted for age, sex, body weight, cognition, stroke, Benzodiazepines, non-TD use, other drugs

Felson et al., 1991 [23]

Case-control

Cases, 116 female, mean age  77.2 years; controls, 672 age-matched, mean age  78.0 years

Hip fracture: ascertained by self-report, hospital records, and death records

N/A

Cases, n  46 (26.1%); Controls, n  172 (25.6%)

OR (95% CI): ever use TD, 1.33 (0.86 – 2.05); recent use TD, 0.69 (0.34, 1.40); recent use pure TD, 0.31 (0.11 – 0.88)

Adjusted for BMI, age at menopause, smoking, alcohol, estrogen use

Heidrich et al., 1991 [24]

Case-control

462 cases: mean age  75 years; 462 age and sex-matched community controls; mean age  74 years; 76% female

Hip fractures: ascertained through HMO records

N/A

Cases, n  199 (43%); controls, n  162 (35%)

Current vs nonusers, RR (95% CI): current, 1.6 (1.0 – 2.5); former, 1.2 (0.6 – 2.4)

Adjusted for alcohol, organic brain syndrome; history of stroke; leg paralysis; days hospitalized in previous year; nursing home residence; BMI; use of pheno barbitol, corticosteroids, or furosemide

Cumming and Klineberg, 1993 [25]

Case-control

209 Hip fracture cases; 207 community controls; age 65, nursing home residents oversampled

Hip fractures: identified by hospital record

N/A

15%

Current TD vs not current, OR (95% CI): all TD, 0.81 (0.46 – 1.45); combination TD, 0.85 (0.41 – 1.75); pure TD, 0.81 (0.34 – 1.89)

Adjusted for age, sex, type of residence, alcohol, BMI, cognitive status, physical activity, dairy intake, health status

Ray et al., 1989 [26]

Case-control

Cases, 905 hip fractures; controls, 5137 population controls; age  65

Hip fracture: identified from discharge records of 10 large hospitals

N/A

Cases, n  235 (26%); controls, n  1592 (31%)

Current users vs never, OR (95% CI):  2 years, 1.2 (0.9, 1.5); 2-5 years, 0.8 (0.7, 1.0); 6 years, 0.5 (0.3, 0.7); former users vs never users, 6 years: 0.7 (0.2, 2.3)

Adjusted for age, sex, nursing home or hospital admission, use of cardiovascular drugs or other anti-hypertensives

LaCroix et al., 1990 [27]

Prospective; 4 years follow-up

n  9518; mean age  74

n  242, hip fracture: identified at annual interviews

N/A

n  2570 (27%)

TD current users vs nonuser OR (95% CI)  0.68 (0.49 – 0.94)

Adjusted for community, age, sex, impaired mobility, BMI, smoking

Cauley et al., 1993 [1]

Prospective; average follow-up 3.32 years

9704 Nonblack women  65 years; mean age never-users  71.4, past users  71.4, current users  72.4

Prevalent vertebral fx’s by spine films; incident fractures confirmed by medical record review

11.2 years; median dose (mg/day):  10 years, TD:37.5,  10 years, TD: 50.0

n  2620 (27%)

For current, long-term ( 10 years) users compared with never users, RR (95% CI): nonspine, 1.06 (0.87, 1.29): osteoporotic, 1.06 (0.85, 1.32); hip, 0.63 (0.33, 1.20); wrist, 0.73 (0.45, 1.20); humerus, 1.43 (0.83, 2.48)

Adjusted for age, weight, functional status, calcium intake, ERT duration

Note. NA, not available; RR, Relative risk; HCT, hydrochlorothiazide; TD, thiazide diuretics; OR, odds ratio; ERT, estrogen replacement therapy; CI, confidence intervals.

748 unique in the number of medical, functional, and pharmacological variables controlled for in the analysis. This may have contributed to their finding of an increased risk of hip fracture associated with TD use. Past use of TD had no effect on hip fractures; pooled OR  1.06; 0.69 to 1.63 [31]. The results of TD on hip fracture were also similar in both cohort and case-control studies [31]. Duration of use has been examined for its effect on hip fractures in several studies [1,24 – 27,29]. In the metaanalysis by Jones et al. [31], long duration of TD use may be protective against hip fractures (OR: 0.82; 0.62 to 1.08) but not short duration (OR: 1.23; 0.99 – 1.54). Results from the Framingham study suggested that the dose of TD may be important [24]. No protective effect of TD on hip fracture was observed in individuals using agents that combined thiazides with other antihypertensive agents. In contrast, subjects who reported taking pure thiazide medications were 70% less likely to suffer a hip fracture. Because combination medications generally contain only 25 mg of hydrochlorothiazide, the authors suggested that the lack of protective effect among combination users reflected the lower dose. Cumming and Klineberg [26], however, reported a similar odds ratio for hip fractures among subjects reporting a combination TD (OR: 0.85; 0.41 to 1.75) and subjects reporting use of a pure TD (OR: 0.81; 0.34 to 1.89). The only study to directly examine TD dose and hip fracture was the case-control study of Heidrich et al. [25]. There was a trend of increasing risk of hip fracture with increasing dose [25]. There may be an interaction between TD use and calcium intake on hip fracture. Most of the studies did not control or stratify by calcium intake. Thiazide diuretics are known to decrease urinary excretion of calcium and to improve calcium balance [32]. The effect of TD on calcium balance might depend on calcium intake. Individuals with a low calcium intake may benefit more from TD, as TD will improve their calcium balance. Only one study addressed this issue and found that the age-adjusted risk of fractures in current TD users was similar in individuals with a low (M7>800 mg/day) and high (  800 mg/day) calcium intake [1]. Users of thiazide diuretics tend to be more obese than nonusers, and because the degree of obesity is inversely associated with the risk of hip fracture, the effect of thiazide diuretics on fracture risk may be obscured by inclusion of the extremely obese. To test this hypothesis, we excluded obese women (body mass index of 31.5 kg/m2, which corresponds to the 85th percentile) [1]. Results were similar to those derived from the analysis of the whole cohort. None of the relative risks was statistically significant, but they did show a trend toward a protective effect of thiazide diuretics against hip and wrist fracture, suggesting that inclusion of the very obese had little effect on the association between TD and fracture.

CAULEY AND SALAMONE

C. Underlying Mechanisms The mechanism by which TD may preserve bone mass and protect against hip fractures is not known. Thiazide diuretics have been shown to influence calcium metabolism. A reduction in the urinary excretion of calcium [14,32], increases in serum calcium [33], and decreased parathyroid hormone (PTH) [34] concentrations, accompanied by reduced serum concentrations of 1,25-dihydroxyvitamin D [(1,25(OH)2D)] and alterations in calcium absorption, have all been reported [35]. Thiazide diuretics may also decrease bone resorption [32] and bone turnover [36]. A recent animal study suggested that thiazide diuretics cause an increase in bone mineral crystallization [37]. Thiazide diuretics may decrease fracture risk by preserving bone mass. As reviewed earlier, cross-sectional studies have shown that TD users have slightly higher cortical and trabecular bone mass than nonusers. If thiazide diuretics reduce the risk for fracture by preserving bone mass, then a statistical adjustment for bone mass should attenuate the relation between thiazide use and the decreased risk for fractures. However, inclusion of the distal radius bone mass did not substantially change the relation between thiazide use and the reduced risk for most fractures, including hip fracture, despite the fact that distal radius bone mass predicts hip fracture and osteoporotic fractures [1]. For wrist fractures, a modest change was observed in the relative risk when the distal radius bone mass measured near the site of wrist fracture was included in the model. It is possible that the inclusion of bone density at the hip may have attenuated the effect of thiazide diuretics on hip fracture. However, TD could aggravate the risk for fractures by increasing the risk for falls. However, the association of diuretic use with falling is not consistent [38,39]. The Study of Osteoporotic Fractures found no association between TD use and risk of falls [1]. The age-adjusted odds ratio for experiencing two or more falls in the first year after baseline current users of TD compared with never users was 1.06; CI, 0.90 to 1.20.

D. Summary and Future Research Cross-sectionally, current thiazide diuretic users have higher axial and appendicular bone mass than either past or never users. This observation has been demonstrated consistently for both axial and appendicular bone sites and for both men and women. The magnitude of the effect of TD on bone mass is about 4 to 5%. Prospectively, thiazide diuretics preserve bone mass, although this observation is limited to appendicular bone mass. Over the short term, TD may abate the seasonal decrease in BMD observed in the fall and winter months.

749

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

The effect of TD on all fractures, including all nonspinal or osteoporotic, i.e., those fractures associated with low bone mass or prevalent vertebral fractures, has not been consistent. For hip fractures, thiazide diuretic users tend to be at reduced risk. The underlying mechanism(s) whereby TD preserve bone mass and prevent fractures has not been established. The effect of TD on hip fractures could not be explained by the higher bone mass of longterm users. There is a paucity of therapeutic agents currently approved for the prevention of osteoporotic fractures. Based on the literature it appears that thiazide diuretics may prevent hip fractures and preserve bone mass at least in elderly women. Jones et al. [31] have suggested that the results of the meta-analysis may preclude the necessity of doing a clinical trial on thiazide diuretics and hip fracture. However, we have very little data on the effect of TD on bone mass and fracture among normotensive women. We and others [1,40] believe that a randomized clinical trial on the effect of TD on the incidence of fractures is the only way to resolve the issue. It is conceivable that this trial could be done in osteopenic women with prevalent vertebral fractures. Focusing on these high-risk women would lead to a decrease in the required sample size and make the trial more feasible. Until such a trial is completed, it is premature to recommend thiazide diuretics for the prevention of osteoporosis and fractures.

III. EXOGENOUS ESTROGEN An estimated 31.7 million prescriptions for oral menopausal estrogens were dispensed in 1992, representing a greater than twofold increase since 1982 [41]. The use of transdermal estrogen also increased, although the absolute number of prescriptions was considerably lower (4.7 million in 1992). Prescriptions for medroxyprogesterone increased almost fivefold from 2.3 million in 1982 to 11.3 million in 1992. These trends demonstrate the vast exposure to postmenopausal estrogens and progestins. Despite the large number of prescriptions that are written, it is estimated that 20 – 30% of postmenopausal women never filled their prescriptions, 10% of users reported only intermittent use, and another 20% discontinued their therapy within 8 months [42] Compliance to HRT remains an issue. A more recent estimate is that fewer than 20% of postmenopausal women take HRT [43]. Estrogen is the cornerstone of preventive therapy for osteoporosis in menopausal women. It is endorsed by many expert panels as the first line of therapy for osteoporosis, including the National Osteoporosis Foundation [44]. More recently, however, the package insert was changed to reflect an approved indication for prevention and not treatment.

A. Estrogen and Bone Mass The effectiveness of estrogen in reducing bone loss is well established [45,46] (see also Chapter 69). Estrogens reduce bone loss by slowing bone turnover, which may be mediated by estrogen receptors in osteoblasts [47]. Most research has examined the effect of oral estrogens [45,46], but estrogens given transdermally [48,49] or by implants [50,51] are also effective. The estrogen dose, but not route of administration, influences the effectiveness of estrogens [45]. Combination (estrogen plus progestins) therapy and monotherapy (unopposed estrogens) have similar effects on bone mass [52]. Analyses of the correlates of baseline BMD among the 875 women recruited into the Postmenopausal Estrogen/Progestin Intervention Trial (PEPI) showed no additional benefit of progestins taken in conjunction with replacement estrogens [53]. The women in PEPI were more likely to have taken medroxyprogesterone acetate, and the authors point out that the effect of contraceptive androgenic progestins on BMD is unknown [53]. Progestins given alone have also been shown to reduce bone loss, particularly loss of cortical bone, but there is little evidence that combination therapy has additive or synergistic effects on bone mass [54]. The minimum duration of estrogen therapy for the prevention of bone loss is not known, but Lindsay [45] recommended a duration of 10 or more years or lifelong for those with established disease. Estrogen appears to be an effective therapy for women with established osteoporosis [55 – 59]. In one study the effect was greatest among those with the lowest bone mineral density [57]. The effect was similar for women on combination therapy [59] and in women receiving transdermal estrogen [58]. An important controversy regarding estrogen and bone mass is whether it is effective in very old women. Data from Framingham [60] showed little protective effect from an average of 10 years estrogen therapy on bone density among women 75 years of age and older. However, failure to see an effect of estrogen in the Framingham women may have reflected the fact that most of the estrogen use occurred in the past. Resnick and Greenspan [61] suggested that therapies that slow bone loss may be less effective among older women whose bone loss has slowed and whose bone is of poorer quality. Estrogen has been shown to be effective in preserving bone mass in elderly women [62,63]. As part of the Study of Osteoporotic Fractures (SOF), rates of bone loss increased among the very old women (age  80), but even among this group, rates were significantly lower in women who reported estrogen use [64]. There is little agreement on when to initiate hormone therapy. Initiation of hormone therapy at menopause effectively prevents bone loss; however, once a woman

750 discontinues her estrogen, bone loss quickly resumes. If a woman continues her estrogen, concern about breast cancer associated with long-term use may cause her to discontinue her therapy. Schneider et al. [65] demonstrated that women who initiated estrogen before age 60 and continued had the highest BMD. Of importance, however, women who initiated therapy after age 60 and continued had BMD that was significantly higher than women who had never used estrogen or discontinued estrogen. These data suggest that women can delay their use of estrogen and still receive a bone-sparing effect on BMD. This study, however, was observational and did not include data on fractures. 1. ESTROGEN AND BONE MASS: RANDOMIZED TRIALS The majority of studies of estrogen and BMD data published prior to 1990 relied on observational or uncontrolled studies. More recently, there have been nine randomized placebo-controlled trials of estrogen and BMD. The largest of these studies was the PEPI study, which enrolled 875 women, mean age 56 years, into a 3-year trial [66]. There were five treatment arms: placebo, conjugated equine estrogen (CEE) 0.625 mg/day, CEE with medroxyprogesterone acetate (MPA) given cyclically (5 mg, 12 days per month), CEE with MPA given continuously (2.5 mg/daily), and CEE with micronized progesterone (200 mg/day) given 12 days per month [66]. Women randomized to placebo lost hip and spine BMD. Hip BMD increased about 1.7% in all active treatment groups (all p0.05 versus placebo). Spine BMD also increased in all active treatment groups. The percentage increase in hip BMD was significantly greater for the CEE-MPA (cont) at 5% compared with an increase of 3.8% in the other treatment groups. Among a group of older women, all age  65 years, a lower dose of CEE (0.3 mg/day) was found to significantly increase spine BMD 3.5% over 3.5 years of observation in comparison to placebo [67]. All of the women had adequate calcium and vitamin D intake. Two randomized dose-ranging trials of estrone-sulfate (estropipate) in newly postmenopausal women (mean 51 years) demonstrated a significant increase in spine BMD measured by quantative computed tomograph (QCT) [68,69]. The increase in spine BMD was primarily observed in women who were randomized to higher doses of estrone-sulfate (0.625 or 1.25 mg/day). Oral micronized estradiol given to 63 postmenopausal women at three different doses resulted in a significant increase in spine BMD of 1% and hip BMD of 2% compared to placebo after 12 months of therapy [70]. There have been two randomized controlled trials (RCTs) of transdermal estradiol. Thirty-four women, followed for 18 months, were randomized to either placebo or 50 mg transdermal estradiol [71]. Forearm BMD increased about 4% in the estrogen arms and decreased 3.5% in the

CAULEY AND SALAMONE

placebo arm. In a dose-ranging study of transdermal estradiol, a minimum dose of 0.05 mg estradiol transdermal is needed to influence spine BMD [72]. A total of 406 women, mean age of 51 years, were randomized to ethinyl estradiol (Ee) — 0.3, 0.625, 1.25 mg, or placebo — and followed for 24 months. Of importance, all doses of Ee showed a significant increase in BMD compared to bone loss in the placebo group [73]. Combinations of Norethindrone acetate (NETA) and ethinyl estradiol were evaluated in 1265 women, mean age 52, for their effects on spine BMD (QCT) after 24 months [74]. The placebo group lost 7.4% of their BMD. All combination groups except the lowest dose group had lost significantly less BMD than the placebo group. An increase in BMD was observed for the combination 1.0 mg NETA and 5g Ee and 1.0 mg NETA and 10g Ee groups.

B. Estrogen and Fractures: Observational Data Current users of estrogen have a statistically significant decreased risk of hip, wrist, and spine fractures [75 – 86]. A recent meta-analysis suggested a 25% decrease in the risk of hip fracture in postmenopausal women who reported ever using estrogen [87]. Several studies reported the relative risk for hip and wrist fracture combined [75,76]. Examination of the effect of ERT on wrist fracture alone revealed about a 60% reduction in the risk of wrist fracture [81,84,86]. The relative risk of vertebral fractures among estrogen users was 0.60 (95% CI, 0.36 to 0.99) [84]. For all fractures, the relative risk was reduced by about 35% [86] to 50% [80] among current ERT users. Most studies have examined the relation between unopposed estrogen and fractures [75 – 79,81,82]. In one cohort study, women who reported using the combined estrogen and progestin regimen constituted about 40% of their users who initiated use before age 60 and 20% of users who initiated use after age 60 [85]. Results of this study showed about a 30% reduction in the risk of hip fracture among women using estrogen, suggesting that the high prevalence of combination users did not influence their results. This study did not however, compare the relative risks separately for unopposed estrogens and combination therapy. In the Study of Osteoporotic Fractures, we compared the relative risk of all nonspine fractures and wrist fractures between current users of unopposed estrogen and current users of estrogen plus progestin [86]. Among current users, the effect of unopposed estrogen was similar to that of estrogen plus progestin for wrist fracture and all nonspinal fractures. The multivariate-adjusted relative risk for wrist fracture among current unopposed estrogen users was 0.42 (CI, 0.25 to 0.70), and this risk among current users of estrogen plus progestin was 0.31 (CI, 0.11 to 0.84). For all nonspinal fractures, the multivariate-adjusted relative risk

751

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

was 0.69 (CI, 0.56 to 0.86) among users of unopposed estrogen and 0.51 (CI, 0.33 to 0.78) among users of estrogen plus progestin. We were unable to look specifically at hip fractures because the prevalence of use of the combined regimen was low in our cohort. The decrease in fracture risk associated with estrogen use is greatest among current or recent users, and the decreased risk tends to diminish with time after stopping estrogen [76,78,83]. In the Study of Osteoporotic Fractures cohort, we found no association between previous use of estrogen and the risk of either hip fracture or all nonspinal fractures [86]. Although previous users had a 20% decrease in the risk of wrist fracture, the confidence intervals included 1.0. Moreover, there was no association with fracture even among previous users who had used estrogen for 10 or more years. Previous users who initiated use within 5 years of menopause had a slightly lower risk of wrist and all nonspinal fractures, but these results were not statistically significant. These results imply that women need to continue estrogen therapy to reduce their risk of fracture. In most studies, a longer duration was more beneficial in reducing fracture than a short duration [76,77,79,86]. The protective effect of estrogen was more pronounced if estrogen was initiated around menopause [75,82,86]. In the large cohort study from Sweden, initiation of use after age 60 was not associated with a reduced risk of hip fracture, although these women tended to take less potent estrogens [85]. Initiation of use within 5 years of menopause among long-term users was associated with a 50% reduction in the risk for all nonspinal fracture, but no effect was noted if estrogen was initiated more than 5 years after menopause, even among long-term users [86]. The age-adjusted risk for all nonspinal fractures among current long-term users of estrogen who initiated estrogen within 5 years of menopause was 0.53 (CI, 0.37 to 0.74), and this risk was 0.97 (CI, 0.56 to 1.68) in current long-term users who initiated estrogen more than 5 years after menopause [86]. Relatively few studies have examined whether the dose of estrogen influenced the results on fractures. Three studies examined the effect of estrogen dose and fracture and found that the dose of estrogen had little effect on the degree of protection [20,76,83]. In the Leisure World Study, the age-adjusted relative risk of hip fracture was similar among women reporting a dose of  0.625 mg (RR  0.84; 0.58 to 1.21) and women reporting a dose of  1.25 mg (RR  0.91; 0.64 to 1.29) compared with never users [83]. Krieger et al. [78] combined dose with duration of use and computed a milligram month variable, which was defined as the total number of months of use times the milligrams of estrogen in that preparation [78]. Women who were exposed at a level of 50 mgmonth or greater were significantly less likely to fracture.

Several factors that could potentially modulate the effect of ERT on fractures have been examined: age, cigarette smoking, degree of obesity, type of menopause, and history of osteoporosis. 1. AGE In two prospective studies [82,85] and one case-control study [79] the protective effects of estrogen on fracture were greater in younger women (defined as either  age 80 [79] or  age 75 [82]) and weaker [79,82] among the older women. However in both the Framingham Study [82] and the Mediterranean Osteoporosis Study [79], the point estimate of the relative risk suggested a 20 – 30% reduced risk of hip fractures among the oldest women, but confidence intervals were wide and included 1.0. The failure to reach statistical significance may have reflected reduced power in the oldest age group, although it is possible that the degree of protection could still be reduced in older women. In the Study of Osteoporotic Fractures cohort, the association between estrogen use and risk for all nonspinal fractures was similar in those younger (RR  0.66, 0.53 to 0.84) and older than 75 years of age (RR  0.65; 0.43 to 0.97) [86]. The effect of estrogen on wrist fracture was also similar in younger and older women, but the confidence intervals were wider in the older age stratum because of the smaller number of women. We found an 80% decrease in the risk for hip fractures among women older than 75 years of age, and we found no effect on hip fracture in those 75 years age or younger. Once we had excluded women with a history of osteoporosis, the relative risk for hip fracture associated with current estrogen use was decreased in older and younger women, but confidence intervals included 1.0. 2. SMOKING Data from Framingham showed that among current smokers, estrogen use was not associated with a reduced risk of hip fracture (OR, 1.26; CI, 0.29 to 5.45) [88]. This is in contrast to an odds ratio for hip fracture among current nonsmoking estrogen users of 0.37 (CI, 0.19 to 0.75) [88]. In the Study of Osteoporotic Fractures, we did not confirm this finding. Current smoking did not attenuate the effect of estrogens on fractures [86]. Similarly, in the casecontrol study of Williams et al. [89], the benefit in preventing hip fractures was greatest in thin women who reported current cigarette smoking. 3. DEGREE OF OBESITY Estrogen users tend to be less obese than nonusers [2] and thin women are at increased risk of fracture [82]. Hence, there may be an interaction between estrogen use and obesity. The beneficial effects of estrogen on hip and wrist fractures varied by the level of obesity and smoking, with the greatest benefit achieved in thin nonsmoking women [89]. This is inconsistent with the observation of

752 the women recruited into the PEPI trial that the increase in BMD with increasing body mass index (BMI) was similar among estrogen users and nonusers. There was no interaction between estrogen and the degree of obesity as measured by the body mass index [53]. However, women with more than moderate obesity (BMI  35) were excluded from PEPI. 4. TYPE OF MENOPAUSE The case-control study of Paganini-Hill and colleagues [20] suggested that the effect of estrogen was greatest among oophorectomized women. The relative risk of hip fracture among oophorectomized women who reported using ERT for more than 5 years was 0.14 compared to 0.86 among women who had intact ovaries. To our knowledge, none of the other studies have been stratified by history of oophorectomy. These oophorectomy data were not adjusted for any other factors. Hence, this observation may reflect a failure to control for other differences between women with a history of oophorectomy and women with intact ovaries such as age at first use, or total duration of use. It is possible that women undergoing a surgical menopause initiate use soon after the surgery and hence earlier in their menopausal transition. The total duration of use may also be longer among oophorectomized women. 5. HISTORY OF OSTEOPOROSIS In The Study of Osteoporotic Fractures, we found that current use of estrogen was associated with similar reductions in the risk for wrist and all nonspinal fractures among women with and without a history of osteoporosis [86]. Current use of estrogen was not associated with a statistically significant reduction in the risk for hip fracture in women reporting a history of osteoporosis, although the reduction in risk in those without osteoporosis was statistically significant. Women with osteoporosis are more likely to take estrogen. Use of estrogen may be a marker of more severe osteoporosis, which could lead to an underestimation of the effect of estrogen in that group. 6. ESTROGEN AND FRACTURES: RANDOMIZED TRIALS There have been few randomized trials of hormone replacement and fractures (Table 3). Nachtigall et al. [90] followed 84 pairs of institutionalized women who were treated daily with placebo or conjugated estrogen 2.5 mg/day and medroxyprogesterone, 10 mg per day, 7 days per month for 10 years: there were seven fractures in the control group and none in the treatment group. Lindsay and colleagues [91] studied 100 women who were enrolled in a longitudinal study of bone loss following oophorectomy; about onehalf of the women were prescribed mestranol (average dose, 23 g/day) and half placebo. After a median followup of 9 years (range 6 – 12 years), estrogen treatment significantly reduced the incidence of vertebral compression.

CAULEY AND SALAMONE

A 1-year trial of 75 postmenopausal women with established osteoporosis reported a 40% reduction in the radiographic vertebral fracture rate in the group treated with 0.1 mg/day of transdermal estrogen compared with placebo [58]. However, a reanalysis of these data based on the number of subjects with at least one new vertebral fracture (not fracture rate) produced a nonsignificant result [92]. In a randomized trial of 464 nonosteoporotic Finnish women, HRT significantly reduced the risk of nonvertebral fractures by about 60%, after adjusting for BMD and history of previous fractures [93]. To our knowledge, this is the first randomized trial of HRT on clinical fractures. Nevertheless, the hormone regimen used in this study is not approved in the United States. The Heart and Estrogen/progestin Replacement Study (HERS) was a randomized, double-blind, placebo-controlled secondary prevention trial of daily use of conjugated equine estrogen plus medroxyprogesterone acetate (progestin) (Prempro) on the occurrence of nonfatal myocardial infarction or coronary heart disease death among women with documented heart disease [94]. Fractures were a secondary end point to the HERS trial. There was no evidence that 4 years of treatment with Prempro substantially reduced the incidence of fractures or height loss among these older, nonosteoporotic women who all had coronary heart disease [94,95]. How do we interpret the differences in the results of the randomized trials of HRT on fractures and observational studies? First, HERS is the largest RCT trial of HRT but was not designed to look at the primary effect of hormone therapy on fractures. Of importance, HRT may be more effective in preventing fractures among women who have osteoporosis. Second, the duration of the trials were relatively short,  4 years. A longer duration of HRT use may be needed to prevent fractures. Finally, the effect of HRT on vertebral fractures in HERS was limited to painful, clinical fractures, which represent only about one-third of the total number of fractures [95]. The earlier study of Lindsay et al. [91] did report an effect of estrogen on vertebral compression. The lack of randomized trials of HRT on fractures emphasize the need to clarify the effect of HRT on fractures in women, both those with and without osteoporosis. Nevertheless, given the data from the RCT of HRT and bone density, it seems reasonable to assume that HRT will preserve bone density, which should ultimately result in a reduction in fractures.

C. Summary and Future Research Current use of estrogen reduces the risk of all fractures; past use has little effect on fracture risk. Estrogen appears to reduce the risk of fracture among older women who smoke. The effects of unopposed estrogen and the

TABLE 3 Summary of Randomized Clinical Trials of Hormone Therapy and Fracture First author (year)

Intervention

Subject charactistics

Results

CEE 2.5mg daily and MPA 10 mg daily for 7 days (CYC) (n  67) Placebo (n  62) 10 years

Age matched 76.7% completed study 40%,  3 years since menopause 10 years

Number of fractures CEE and MPA 0 Placebo, 7

Lindsay (1980)

Mestranol average dose 23.3 g day (range 10 to 38 g) (n  58) Placebo (n  42) 9 years (range: 6 to 12)

Oophorectomized Subset of original trial: women who attended clinic visits consistently and for longest period of time Mean age about 48 years

Vertebral morphometry Mestranol: 4 – 7% greater anterior height (p  0.10) Mestranol: 34 – 56% smaller wedge angle (p  0.05) Mestranol: 9 – 13% greater ratio of central height to anterior height (p  0.06)

Lufkin (1992) Windeler (1995)

Estraderm 0.1 mg estradiol plus MPA 10 mg/day for 10 days (n  36) Placebo (n  39) 1 year

Osteoporotic women: prevalent vertebral fracture or BMD  10 percentile of normal premenopausal women All Caucasian Mean age 65 years (55 to 72)

Vertebral morphometry Number of new fractures PBO: 20 new fractures in 12 women E2 MPA: 8 new fractures in 7 women Fracture rate per 100 PY PBO: 58 E2 MPA: 23 (p  0.04) Number of women with  1 new fracture PBO: 12 E2 MPA: 7 (p  0.28)

Komulainen (1998)

E2Valerate 2 mg (days 1 – 21) and CPA 1 mg days 12 – 21 (n  116) Vitamin D 300 IU and 93 mg calcium (n  116) E2Valerate plus vitamin D (n  116) Placebo (n  116) 4.3 years

Subset of Kuopio Osteoporosis Study 6 – 24 months since last menstruation Nonosteoporotic: t score  2.5 Mean age 53 years (47 to 56)

All symptomatic nonvertebral fractures ITT RR (95% CI)a E2Valerate CPA 0.38 (0.15 – 0.99) Vitamin D 0. 64 (0.29 – 1.42) E2Valerate CPA vitamin D 0.48 (0.19 – 1.18) Placebo 1.0

Cauley (2000)

0.625 mg CEE in combination with 2.5 MPA daily (n  1380) Placebo (n  1383) 4.1 years

Documented Coronary Heart Disease 89% Caucasian Postmenopausal Intact uterus Mean age 67 years (44 to 79)

ITT RR (95% CI) All clinical fractures 0.94 (0.75 – 1.19) Hip 1.09 (0.51 – 2.31) Wrist 0.97 (0.58 – 1.62) Spine 0.69 (0.34 – 1.40)

753

Nachtigall (1979)

Notes: ITT, intention to treat; CEE, conjugated equine estrogen; MPA, medroxy progesterone acetate; PBO, lacebo; E2, estradiol; CPA, cyproterone acetate. a Adjusting for BMD and previous fractures.

754 combination of estrogen plus progestin are similar. For optimal protection from fractures, estrogen use should be initiated early in menopause and continued indefinitely. If women initiate estrogen at menopause and continue it indefinitely they may be exposed to it for about 30 years. Further studies are needed to evaluate the overall risks and benefits of such a prolonged use of estrogen. In addition, further studies are also needed on the effect of estrogen on fracture risk if estrogen is initiated much after menopause. Ettinger and Grady [96] have suggested that initiating estrogen at age 65 will result in a reduced risk of hip fracture. Some of the protective effects of estrogen observed among long-term users may reflect a compliance bias. These longterm users by definition are compliant women. Dr. Petititi has shown that among women assigned to receive a beta blocker or placebo, the mortality rate was lower among the women who had  75% compliance irrespective of whether they were assigned to either the study medication or the placebo [97]. Thus, hormone therapy,-especially long-term therapy,may be a marker of socioeconomic, clinical, or lifestyle factors that place estrogen users at a lower risk of fracture. Compliance to estrogen therapy is also important when considering its long-term effectiveness. The low compliance rate to estrogen therapy has been demonstrated in the United States [42,98] and elsewhere [99]. Furthermore, according to the Massachusetts Women’s Health survey of women receiving estrogen therapy for the first time, 20% stopped taking it within 9 months of initiation, 10% used it intermittently, and 20 – 30% never filled their prescriptions [98]. In a follow-up study of 1689 women in the Netherlands, 66% of women were no longer taking estrogen 1 year and 9 months after the initial survey [100]. In the Study of Osteoporotic Fractures cohort, a substantial proportion of women reported stopping estrogen because they felt they did not need it simply because their symptoms abated [101]. Bone loss occurs asymptomatically until a fracture occurs. Hence, women need to be educated about the long-term benefits and risks of estrogen. This education program may result in improved compliance. Finally, 1-year compliance rates were greater among women prescribed transdermal estrogens than among women taking oral estrogens [102]. Future research needs to identify the best hormonal regimen that will achieve maximum benefit, minimum risks, and maximum compliance.

IV. ENDOGENOUS ESTROGEN, BONE MASS, AND FRACTURES Almost any study of endogenous estrogen, bone density, or fracture depends on the measurement of hormonal milieu at a single point in time, revealing little about the exposure over the course of a lifetime, due to both the daily variability of such measurements during the menstrual cycle and

CAULEY AND SALAMONE

the lack of correlation of postmenopausal measures with premenopausal levels. Also, there is disagreement as to whether estrogen-responsive tissues respond to absolute levels of estrogen or some other parameter such as unbound estrogen, which estrogens are important (e.g., estrone versus estradiol), and how hormone binding and metabolism influence the biological effects on bone. In contrast to the well-documented protective effect of exogenous estrogens on bone, bone loss, and fracture, less is known about the relationship of endogenous estrogens to bone. Several approaches have been taken to examine the association among endogenous estrogens, BMD, and fracture: cross-sectional studies have examined serum estrogens and bone density; longitudinally, serum estrogens have been compared among fast and slow bone losers or correlated with rates of bone loss; case-control studies have compared serum estrogens in fracture cases and controls.

A. Cross-Sectional Studies The majority of the cross-sectional studies reported positive associations between various measures of bone density and serum estrogens [103 – 111]. One study reported positive correlations between radial bone mineral content and estrogen excretion [112]. Two studies that failed to report an association between serum estrogens and bone mass had very small sample sizes and may have had little power to see an effect [113,114]. Johnston et al. [115] studied 40 postmenopausal women with type II diabetes and, after controlling for obesity, age, and diabetic control, neither estrone nor estradiol was a significant predictor of radial bone mass. In this study, failure to see an association may have reflected the small sample size. In addition, the distribution of the degree of obesity was narrow and could have influenced the homogeneity of estrogen concentrations. In a cross-sectional analysis of a random subset of 274 women enrolled in the Study of Osteoporotic Fractures, women who had estradiol levels 10 – 25 pg/ml had 4.9, 9.6, 7.3, and 6.8% greater BMD at the total hip, calcaneous, radius, and spine than those with levels below 5 pg/ml [116]. After multiple adjustments, BMD differences remained statistically significant and corresponded to about 0.4 standard deviation limits. Bioavailable estradiol concentrations are also important determinants of BMD in men and women, with correlation coefficients of 0.31 to 0.38 in men and 0.38 to 0.45 in women [117].

B. Longitudinal Studies For the most part, an association between serum estrogens and rates of bone loss was reported in both peri- or

755

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

early postmenopausal women and older women [118 – 128]. Several studies compared estrogen concentrations in fast and slow users [119 – 122]. Two of these studies focused on early postmenopausal women and found higher estrogens among the slow losers [121,122]. One of the best studies in this group was the study by Slemenda et al. [124] in which 84 peri- and postmenopausal women were followed over 3 years. Both estrone and estradiol concentrations were significantly correlated with change in radial BMD, with correlation coefficients ranging from 0.24 to 0.36. No association was found between estrone or estradiol and bone loss in a larger population of Dutch women [125]. These women were much older than the women in the Slemenda et al. [124] studies and it is possible that estrogen levels may have a lower impact on very old women. However, it has been shown that lower levels of endogenous estrogens and higher sex hormone binding globulin concentrations were associated with more rapid bone loss from the calcaneus and hip [128]. The women in this study were, on average, 72 years old. Women with serum estradiol values 10 pg/ml averaged only 0.1% [95% confidence interval ( 0.7%, 0.5%)] annual hip bone loss, whereas women with values below  5 pg/ml averaged 0.8% (0.3, 1.2) hip bone loss per year. These associations were independent of age and body weight.

C. Fracture Studies Evaluation of the relationship between endogenous estrogen and fracture have inherent shortcomings. Fractures are caused by both bone-related and fall-related factors. Fall-related factors may be influenced by estrogen. Second, vertebral fractures are difficult to define, do not always come to medical attention, and, consequently, are not fully represented in many studies. Third, there may be metabolic and binding factors affecting bone that are not captured by a simple assessment of estrogen concentration. Finally, in many cases, measurement, of estrogen concentrations were taken after the occurrence of the fracture and may reflect postfracture changes in body weight, physical activity, and other factors. Comparison of estrogen concentrations in osteoporotic fracture cases and controls without fracture yields inconsistent results, with some studies reporting higher levels of estrogens among controls [129 – 134] and others reporting no difference [121,122,125,135 – 138]. Older women with estradiol concentrations  5 pg/ml were 60% (20% to 80%) less likely to have a prevalent vertebral deformity than women with lower levels [116]. In the Rancho Benardo cohort, low values estradiol were associated with prevalent vertebral fractures in older men, but not women [139]. Also, while a few of the studies reviewed here assess

estrogen binding and metabolism and, by implication, its availability to target tissues, there is not enough information from which to draw conclusions regarding these issues. While Bartizal et al. [136] found similar concentrations of estrogens in vertebral fracture cases and controls, estradiol binding was significantly higher in controls, consistent with the hypothesis that the apparent decrease in estrogen binding activity may reflect decreased tissue sensitivity to estrogen rather than a lower serum hormone level. More data are needed to support this hypothesis. Another important consideration when evaluating inconsistencies is uncontrolled confounding. Few of these studies addressed the issue of confounding by body weight, probably the most important confounder of the relationship between estrogen levels and fracture. For instance, Aloia and associates [134] acknowledged case-control differences in weight and other potential confounders, but the analysis does not control for them. Another difficulty in interpretation of these studies is the clinical nature of most of the samples. Many of the earlier studies (published before 1990) are difficult to interpret because of alterations in hormones after the fracture, nonrepresentative cases and controls, and use of insensitive assays. A much stronger design is a prospective study design, where hormones are measured prior to the incident of fracture. In a nested case-cohort study within SOF, women with undetectable serum estradiol (5 pg/ml) had a relative risk of 2.5 for subsequent hip fractures (95% CI, 1.4 – 4.6) and subsequent vertebral fracture (95% CI, 1.4 – 4.2) as compared to women with detectable serum estradiol concentrations [140].

D. Summary and Future Research Estrogen concentrations are related to BMD and to longitudinal changes in BMD. Postmenopausal women who have low levels of estradiol have lower BMD and experience faster rates of bone loss. These observations have been reported in both peri- and newly postmenopausal women as well as much older women. Women with low levels of estradiol are also 2.5 times more likely to experience a hip or vertebral fracture. These findings are remarkable. The associations between endogenous estradiol and fractures are much stronger than we observed between serum cholesterol and death from cardiovascular disease [141]. These findings raise the possibility that a single measurement of estradiol could be used to estimate a woman’s risk of suffering accelerated bone loss and fractures. Women with very low levels of estradiol could consider estrogen replacement therapy. The dose of estrogen replacement therapy could be titrated to the endogenous concentration of estrogen. A lower dose of estrogen could be considered [142]. It is also

756 possible that nonpharmacologic interventions that either reduce or raise serum estrogen concentrations could thereby influence the risk of disease.

V. OBESITY The epidemic of overweight in the United States is welldocumented. The third National Health and Nutrition Examination Survey [NHANES III (phase 1)] of 1988 – 1991 showed the age-specific prevalence of overweight, as defined by a body mass index (BMI) of 27.3 kg/m2, in women ages 50 – 59 years and 60 – 74 years of 52.0 and 41.3%, respectively [143]. This value represents approximately 120% of the desirable weight for women defined by the midpoint of the range of weights for a medium frame from the 1983 Metropolitan Height and Weight tables [144,145].

A. Degree of Obesity and Bone Mineral Density Evidence from both cross-sectional and longitudinal studies supports the generally held beliefs that obesity confers higher BMD and protects against postmenopausal bone loss. 1. CROSS-SECTIONAL STUDIES Numerous cross-sectional studies have demonstrated that a variety of measures of body size are positively correlated with BMD in postmenopausal women [146 –161]. In one of the earlier studies, Dalen et al. [160] found that obese women had an 11% greater metacarpal cortical area than the nonobese. Correlations with body size and either BMD or BMC measures range from 0.00 to 0.42 at the lumbar spine [152–156,158], 0.195 to 0.27 at the femoral neck [154–156,158], 0.00 to 0.44 at the radius [152–154,156,158], and 0.21 to 0.54 for the whole body [150,161]. Although no ideal weight-to-height ratio has been set for reducing osteoporosis and fracture risk, a higher BMI appears to confer protection; accordingly, a BMI 26 – 28 kg/m2 offers protection, whereas a slender figure of  22 – 24 kg/m2 increases risk [162]. A BMI of about 30 kg/m2 is associated with a 4 – 8% greater spine BMD, 8 – 9% greater hip BMD, and 25% greater radial BMD compared to a BMI of 20 kg/m2; furthermore, a BMI of 30 kg/m2 is associated with one-half the loss in spine BMD in early postmenopausal years than a BMI of 20 kg/m2 [148, 157, 163, 164]. By stratifying postmenopausal women by those 115% or above their ideal body weight (IBW) and those below 115%, Dawson-Hughes et al. [156] demonstrated that heavier women had significantly greater BMD at the spine, hip, and radius than normal weight women. Among the obese women, BMD declined only at the radius with increasing years since menopause, whereas declines in BMC were

CAULEY AND SALAMONE

shown at both the radius and the hip. The observation that hip BMC declined while hip BMD was maintained suggests that obese women may experience an adaptive bone remodeling response after menopause, translating into improved bone density. In another study conducted in obese peri- and postmenopausal women and age-matched normals, obesity was defined as an excess of body weight of more than 10% from normal weight, as calculated from Lorentz’s equation [157]. Obese postmenopausal women had significantly higher vertebral BMD than nonobese women (1.056 – 0.12 and 0.974 – 0.14 g/cm2, respectively, P  0.0001), whereas no significant differences in BMD were observed among perimenopausal women. This implies that obesity may be more instrumental in attenuating postmenopausal bone loss rather than in achieving higher peak BMD. From prospective cohort studies, such as the Framingham cohort’s biennial examination (1988 – 1989) of 693 postmenopausal women, both body weight and BMI explained a substantial proportion of the variance in BMD for both weight (lumbar spine, femoral neck) and nonweightbearing sites (radius), accounting for 8.9 to 19.8% of the total variance (P  0.01 for all) [165]. In the Rancho Bernardo population, which examined the association between BMD of the spine, hip, and radius and measures of body size, both body weight and BMI consistently contributed to BMD at all sites [166]. It is important to note that measures of body size explained a much larger percentage of the variance in BMD in the axial weight-bearing spine and hip (12.4 and 16.5%, respectively) than in the nonweight-bearing midshaft and ultradistal radial sites (8.1 and 3.9%, respectively). Similarly, in a Japanese-American population using multiple regression analysis, body weight was related more strongly to BMC at the os calcis than at nonweight-bearing radial sites [167]. It is well established that obesity is associated with higher BMD, at least from cross-sectional evidence. What remains unclear is whether lifetime body weight or weight change in adulthood is a significant determinant of BMD. It is logical that both the degree and the duration of obesity may be important to consider when evaluating the association between body weight and BMD. From a cross-sectional evaluation of the Rancho Bernardo cohort, current BMI explained more than 29% of the variance in BMD at the radial shaft and total hip, 11% at the total spine, and 17% at the ultradistal radius [147]. When weight change was considered from self-reported weight at age 18 and present weight, greater weight gain was associated with higher BMD, explaining approximately 26% of the variance at the total hip. Therefore, it is possible that weight change or weight fluctuation in adulthood may be a most important determinant of BMD. Longitudinal evaluations of obese and normal weight women are necessary to confirm these cross-sectional observations.

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

2. LONGITUDINAL STUDIES To date, few longitudinal studies have addressed the influence of body weight or degree of obesity on postmenopausal bone loss. The longitudinal rates of bone loss over 31 months in a population of overweight postmenopausal women (BMI  25.0, n  40) were compared to those of women of normal weight women (BMI  25.0, n  115) [148]. The annual rate of vertebral bone loss was reduced significantly among the overweight compared with normal weight women (0.54 – 1.1 and 1.46 – 1.6%, respectively, P  0.05). These results are in agreement with those of Harris et al. [149], who demonstrated that increased body weight up to 106% of IBW (corresponding to a BMI of 23.6 kg/m2) was protective against postmenopausal vertebral bone loss, but not against radial or femoral bone loss. Additional weight beyond 106% did not appear to provide any further protection from vertebral bone loss. From the Framingham cohort, the change in body weight from the biennial examination (1948 – 1951 to 1988 – 1989) in 693 women was the strongest explanatory factor for BMD among women at the spine, the femoral neck, and radius (10.7, 8.2, and 6.5%, respectively, of the total variance, P  0.01 for all) [165]. Of interest, the relationship between weight change and BMD was strongest in women not using estrogen.

B. Weight Loss and Bone Mineral Density The consequences of weight reduction on BMD in obese subjects remains, at present, unknown. Cross-sectional studies performed on obese subjects undergoing surgical procedures for weight reduction, including jejuno-ileal bypass, gastroplasty, and biliopancreatic bypass, are associated with some reduction in BMD, albeit not consistently [168 – 172]. Early longitudinal evidence was unable to demonstrate any significant declines in BMD following weight loss [171,172]. For example, in a study by Rickers et al., [171], there was no evidence of radial BMD loss in obese subjects at 12 months (n  11; 114% overweight) or 8 years after intestinal bypass (n  12; 38% overweight). These discordant results may be ascribed to varying amounts of dietary calcium and vitamin D provided, methodological problems in assessing bone in obese subjects, or perhaps inherent differences in cross-sectional and longitudinal study designs. More recently, changes in total and regional BMD, as measured by dual-energy X-ray absorptiometry (DXA), were assessed in obese patients undergoing rapid weight loss on a low-calorie diet for 2 weeks [mean  1.9 MJ (452 kcal) for women and 2.4 MJ (571 kcal) for men] and thereafter up to 4.2 MJ (1000 kcal) for women and 4.7 MJ (1119 kcal) for men [174]. After 15 weeks, whole body BMC was reduced significantly ( 5.9%), with the largest declines in

757 the trunk ( 6.9%) and the smallest in the arms ( 4.0%), suggesting greater losses in weight-bearing than in nonweight-bearing sites. In addition, losses in whole body BMC were correlated positively and strongly with fat mass losses (r  0.86, P  0.01) after 15 weeks and at 9 months (r  0.94, P  0.01), but not with lean mass (r  0.20, P  0.15). Compston et al. [175] found inconsistent losses in BMD of approximately 1 to 2% at weight-bearing sites in obese postmenopausal women participating in a low-calorie dietary intervention, but not at the radius [175]. It is of interest that after 10 months of follow-up, subjects returned to their baseline weight as well as baseline BMD values. Given the fact that body weight and BMD changed in the same direction during weight loss and gain suggests that observed changes in BMD may indeed be real. In contrast, Avenell et al. [176] did not observe parallel gains in weight and BMD in the second 6 months of a 1-year dietary study. These studies raise important questions about whether weight regain is accompanied by bone regain and whether the bone that is regained is of similar quality. More recently, Salamone et al. [177] examined the effect of a lifestyle intervention aimed at lowering dietary fat intake and increasing physical activity to produce modest weight loss or prevent weight gain on BMD in premenopausal women participating in the Women’s Healthy Lifestyle Project (WHLP). Of the 236 women enrolled, the intervention group experienced a mean weight loss of 3.2  4.7 kg over the 18-month study period (n  115) compared to a weight gain of 0.42  3.6 kg in the controls (n  111). The annualized rate of hip BMD loss was twofold higher in the intervention group (0.81  1.3%/year) than in the control group (0.42  1.1%/year); a similar although nonsignificant pattern was observed at the spine (0.70  1.4%/year vs 0.37  1.5%/year). Evidence from this study suggests that the overall risks and benefits of weight reduction among premenopausal women need to include effects on BMD and risks of osteoporosis. One would hypothesize that postmenopausal women would be at an even greater risk of osteoporosis given the additive effect of weight loss-induced bone loss and normal postmenopausal bone loss. Current studies are underway to determine the longer term repercussions of weight loss-induced bone loss at the completion of the 5-year intervention study. There is limited evidence to examine whether similar changes in BMD occur when exercise is included as a component of the intervention. In the WHLP study population, large increases in physical activity (4180 kJ/week) during the study attenuated spinal bone loss but had no impact on hip bone loss [177]. However, Svendsen et al. [178] compared an energy-restrictive dietary intervention with and without aerobic exercise and weight training and found no additional benefit of exercise on weight loss-induced changes in BMD. Further studies considering the type,

758

CAULEY AND SALAMONE

duration, and intensity of physical activity that confers protection against BMD during weight loss are needed. A few mechanisms have been postulated to explain the association between changes in weight and changes in BMD. First, it is possible that during weight loss and subsequent loss of fat tissue, there is a marked reduction in endogenous estrogen production. Such declines in fat mass with weight loss result in lower concentrations of androgen precursors for conversion to estrogens in peripheral tissues. Second, during weight loss, the rate of bone turnover may be enhanced such that changes in total body BMD are masked by larger ones in trabecular bone such as in the spine or hip [161], indicating that more region-specific measures are necessary to assess the clinical significance of weight loss on BMD. Third, changes in BMD may be artifactual. Because of the discordant results across studies on whether bone loss accompanies weight loss in obese subjects, it is possible that these differences reflect methodological issues in bone measures in obese subjects [173,179]. Earlier techniques to measure BMD, such as photon absorptiometry, might have been compromised by the influence of fat tissue, whereas the more recent DXA measurements, which are less dependent on adjacent fat tissue, may be more effective in monitoring changes in BMD. Fourth, with weight loss there is a decline in the mechanical stress on the skeleton that may influence bone remodeling. Next, nonsignificant reductions in markers of bone formation [178] and elevations in markers of bone resorption [180] have been reported among women consuming an energy-restrictive diet compared to controls. Similar results were shown by Salamone et al. [177] in which premenopausal women who lost the most weight ( 8%) experienced the largest increase in the percentage change in N-telopeptide, a marker of bone resorption. These parallel changes in bone resorption and BMD loss support the hypothesis that weight loss-induced bone loss may be mediated, at least in part, by alterations in bone remodeling. Finally, weight control programs may compromise the adequacy of nutrient intake, especially that of calcium and vitamin D – both of which are integral to bone health.

C. Methodological Issues Measurements of body composition and BMD by absorptiometric techniques such as dual photon absorptiometry (DPA) or DXA are generally limited by the basic assumption of a three-compartment model (bone, bone-free lean mass, and fat), which does not include water as a separate component [181]. Such compartmental measures assume uniform hydration, particularly of bone-free lean mass at 0.73 ml/g [182, 183], when indeed this value can vary by as much as 67 – 85% [183]. Because lean and fat mass are summed to give total weight, errors in lean mass

are propagated to the fat compartment, perhaps resulting in distorted values [184]. The extent to which absorptiometric measurements of soft tissue body composition are sensitive to variations in hydration status, particularly in the obese, is presently unknown. Many studies have examined both the accuracy and the precision of absorptiometric measures in populations of normal weight, although the validity of such measures is limited in the obese. In vitro studies have tried to simulate “fatness” by measuring aluminum phantoms at varying thicknesses. There was some suggestion of machine drift over time in the thickest phantom [185]. Similarly, evidence from in vitro phantoms scanned by DXA in water and simulated tissue (ground beef and fat) supports a significant effect of “thickness” on total body BMC, measured by DXA and a similar, although less conclusive, effect on BMD [186]. Others have shown no significant effect of increasing tissue equivalent thickness on DXA phantom measurements, whereas with DPA, significant reductions in phantom BMD with increasing thickness are observed [187]. Such differences in phantom measurements between DPA and DXA may explain, in part, differences observed in the strength of the relationship between body weight and BMD across studies. DXA measurements appear to be sensitive to the anteroposterior thickness of the body, as supported largely from in vitro studies. The influence of thickness on BMD measurements by DPA and DXA is reported as a consistent under- or overestimation [188 – 191]. Anterior – posterior bone measurements by DXA do not account for bone thickness, which is related to overall body size, leading to exaggerated results. This modest thickness effect may have the most significant implications for subjects having extreme truncal thicknesses or those who experience changes in thickness or body weight [184]. In theory, measurement errors by absorptiometry between thin and obese subjects may result from either the scanner’s misinterpretation of soft tissue pixels as bone, leading to reduced measures of BMD [186], or a reduction in the proportion of X-rays surviving attenuation, resulting in errors in bone edge delineation and subsequently, enlarged bone with reduced density. Svendsen et al. [192], as did Van Loan and Mayclin [193], suggested that large changes in body mass and composition may influence the ability of DXA to detect bone edges, thereby changing the estimation of bone area. Because of these limitations, DXA measures are truncated in women weighing up to 275 lbs. It is likely that systematic errors in BMD measurements by absorptiometry exist between thin and obese subjects, although the degree to which this occurs is not well established. The overall lack of instrument sensitivity in measuring changes in BMC and BMD has also been suggested. After a 15-week weight loss program, nonsignificant changes in BMC and bone area measurements occurred

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

in opposite directions, resulting in a significant decline in BMD without a loss in BMC [194]. These observations suggest that changes in BMD observed in weight loss studies may result from the lack of instrument sensitivity when body weight and body composition change and may indeed be an artifact and not a real physiologic change. Ritchie et al. [195] also showed that changes in BMD after pregnancy were the result of changes in the measurement of bone area, which in turn impacts the calculation of BMD. Clearly, there are significant methodological limitations in measuring bone density in the obese. Interpretation of such studies is further complicated by the fact that the definition of obesity is imprecise, often defined by different weight to height ratios.

D. Body Composition There is little disagreement that body weight plays a significant role in determining BMD, particularly at the axial skeleton. However, what remains less clear is what the relative contributions are of the fat and lean mass compartments of that body weight to BMD. While there is strong evidence to support an association between body weight and BMD, there is considerably less evidence relating the components of body weight, lean and fat mass, to bone mass. Among postmenopausal women, some studies attribute most or all of the weight effect at the weight-bearing skeleton to the fat component [188,196,197], whereas others assign a greater role to lean mass [153,198]. The percentage of variability of bone mass explained by fat-free mass ranged from 38 to 55% in premenopausal women and from 28 to 52% in postmenopausal women, whereas the percentage of variability explained by fat mass was only 1 to 9% in premenopausal women and 12 to 21% in postmenopausal women [198]. In postmenopausal women, Reid and colleagues [196] demonstrated that total fat mass, as measured by DXA, is a consistent predictor of spinal and femoral neck BMD. Similar correlations were reported between total fat mass and femoral neck BMD (r  0.38) as with body weight and BMD (r  0.39, P  0.05 for both) [196]. Analogous results were presented by Compston et al. [197], who showed that fat mass, after adjustment for lean mass and age, was significantly correlated with bone mineral measurements at the spine (r  0.30), femoral neck (r  0.31), and whole body BMD (r  0.48, P  0.01 for all). Low fat mass is considered to be a risk factor for osteoporotic fractures, as 11 – 15% of fracture patients had fat mass values, as measured by DPA, below the lowest value in the controls, whereas lean mass was essentially equal in both groups [188]. The fact that fast bone losers have a lower anthropometrically determined fat mass than slow

759 bone losers in early postmenopause further adds to the importance of fat mass in determining BMD in postmenopause [122]. Whether lean mass is a significant determinant of BMD in the postmenopausal years is, at present, unresolved. The importance of lean mass or muscularity to BMD is supported by both in vitro and in vivo studies. From human autopsy specimens, after controlling for age, weight, and height, there was a significant correlation between the ash weight of the third lumbar vertebrae and the weight of the left psoas muscle [199]. Sinaki et al. [200] reported a significant correlation between the isometric strength of the back extensor muscles and BMD of the lumbar vertebrae, accounting for 9% of the variance in spine density. In a study by Bevier et al. [153] conducted in women greater than 60 years of age, fat mass was calculated from skinfold thicknesses and bioelectrical impedance analysis (BIA), and lean mass was calculated by subtracting fat mass from body weight. Spinal density correlated with both fat and lean mass, but stepwise multiple regression analyses found that only lean mass contributed significantly to the prediction of spinal BMD. This finding was not confirmed by Aloia et al. [201], who found no significant correlations between total body potassium and trochanteric BMD or Ward’s triangle BMD. Indeed, this may be attributed to regional differences in the quantities of fat or lean mass, similar to regional differences in bone remodeling and architecture within the femur. More recent studies have shown that once bone thickness and body weight are taken into account, body composition appears to have little independent effect on BMD at spine, femoral neck, or total body. This observation is supported by Carter et al. [202], who introduced the term bone mineral apparent density by dividing BMC by an estimate of bone volume, and by Reid et al. [203], who divided BMD by height to adjust for differences in bone size. In a study by Harris and Dawson-Hughes [204], the association between nonfat soft tissue (NFST) and BMD was examined in 261 postmenopausal women. A height-dependent variable was calculated to correct for differences in bone thickness by dividing BMD by height [204]. Body weight was correlated positively with height-adjusted BMD at the spine, hip, and total body (r  0.22 – 0.26). However, the percentage of NFST was not associated with height-adjusted BMD at the spine or femoral neck and only weakly at total body (r  0.12). This finding supports the hypothesis that the protective effect of body weight is predominantly through its mechanical force on the skeleton; thus, previously reported associations between bone mineral and fat-free mass in postmenopausal women may be an artifact attributable to incomplete control for bone and body size and the inclusion of bone mineral in fat-free mass [205]. Mechanisms other than just the simple mechanical loading on the skeleton, however, have been postulated to

760 explain the association between body composition or the relative proportions of body fat or lean mass on BMD. Different mechanisms predominate at different stages of life. In premenopausal years, lean mass, a major component of weight, becomes more important in the maintenance of BMD. With increasing age and especially after menopause, fat mass and weight increase, while lean mass declines, changing the relative proportion of lean mass and fat mass to weight. Thus, the association between fat mass and bone mass increases with age as more of the weight loading on bones is attributable to fat mass [205]. Such increases in fat mass may promote greater estrogen production. Because the conversions of androgen to estrogen become a major source of estrogen after menopause, the endocrine effect of fat mass may be minimal during bone accretion. It is reasonable to conclude that lean mass may also be an important determinant of bone mass [206]. The positive association between lean mass and bone mass may indicate a potential genetic association between higher peak bone mass and higher lean tissue, a greater mechanical loading on the skeleton, or perhaps indicate higher levels of physical activity. Evidence from the Sydney Twin Study of Osteoporosis demonstrated that lean mass was a significant determinant of areal BMD, whereas fat mass was a significant determinant of volumetric BMD. From this study, it was concluded that lean mass, fat mass, and bone mass are under strong genetic control; although it is plausible that the same genes modulate bone mass, the association between bone mass and body composition appeared to be mediated primarily through common environmental factors [208]. Both fat and lean mass have independent influences on bone mass, but their relative contributions appear to vary by bone site based on trabecular content, physical mobility, and muscularity of the site [209]. It is possible that more trabecular bone sites are influenced more strongly by fat mass because they may respond more to the hormonal effects of fat mass and it is metabolically more active bone. In contrast, lean mass may act via mechanical stress of muscular contractions that result from physical activity. Lean mass may play a larger role at the femur, for example, because of the muscular activity in this portion of the body. Based on the available evidence mostly from cross-sectional studies, there is some consensus on the relative contributions of lean and fat mass to BMD in postmenopausal women. From cross-sectional data in postmenopausal women, lean mass explained a greater proportion of the variance than fat mass; longitudinally, however, annual changes in fat mass were associated with regional changes in BMD [206]. Whether loss of lean mass with aging contributes to loss of skeletal mass simply by reducing mechanical load on the skeleton is not established [210].

CAULEY AND SALAMONE

Further longitudinal studies on changes in BMD in the postmenopausal years are important in order to determine whether changes in BMD are sensitive to changes in body composition and, furthermore, to delineate the mechanism of such an association.

E. Degree of Obesity and Fractures The inverse association between body weight and risk of fractures has been documented extensively by cross-sectional [20,75,78,82,89] and case-control studies [211 – 219], although only three prospective studies have examined this association [213,215,216]. As discussed previously, obese women tend to have higher BMD than nonobese as well as lower rates of postmenopausal bone loss as compared to slender women [149]. Studies of hip fracture [75] and forearm fracture [75,213,217,218] find the thinnest women at the higher risk of fracture. From the Nurse’s Health Study, there was a 20% reduction in fracture incidence when women with a BMI  21 kg/m2 were compared to those  21 kg/m2 [217]. Similarly, in a study of hip fracture cases and trauma controls. From the Study of Osteoporotic Fractures, higher body weight was associated with a significant reduction in fractures of the distal forearm (RR  0.89; 95% CI, 0.79 to 1.01), but no real differences in fractures of the proximal humerus (RR  0.97; 95% CI, 0.82 to 1.15) [213]. Accordingly, heavy (as compared to thin) women appear less likely to develop osteoporosis and osteoporosis-related fractures. Kiel and colleagues [82] confirmed the strong inverse association between body weight and hip fracture using weight categories of Metropolitan weight tables (1959). With self-reported weight, Alderman et al. [218] demonstrated that the risk of hip or forearm fracture declined with increasing increments of weight. Another case-control study showed that the effect of obesity on fractures appears to be limited to women not taking replacement estrogens, with a reduced risk of hip fracture in the moderately obese groups, yet no protective effect in the most obese [20]. However,among obese women, as defined by a ponderal index between 9.6 and 12.5, with little or no use of estrogen replacement therapy (less than 1 year), the risk of hip fracture was greater in thin women than in obese women. Among the obese women, neither the risk of hip fracture nor that of forearm fracture was affected by the use of estrogens, with relative risks of 1.9; 0.7 to 5.4 and 0.8; 0.4 to 1.6, respectively [89]. From these results, it appears that the use of replacement estrogens changes fracture risk little, if any, in obese postmenopausal women, although as adiposity declines, the beneficial effect of estrogen therapy becomes more pronounced. More recently, Cumming and Klineberg [211]

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

considered the influence of both past weight (age 20 years) and recent weight (age 50 years) on the risk of hip fracture. Current high BMI was associated with a reduced risk of hip fracture, whereas high BMI at age 20 was associated with an increased risk of hip fracture. Comparing individuals who maintained or gained weight between age 50 and time of the fracture, the odds ratio for risk of fracture for those who lost weight was 1.9; 1.1 to 3.3, whereas for those who lost weight after age 20 was 3.4; 1.8 to 6.4, suggesting that weight loss may be associated with an increased risk of fracture. These results are consistent with those reported in the Study of Osteoporotic Fractures, indicating that weight loss after age 50 years is associated with reductions in BMD [220] and increased risk of fracture [221]. Cases of vertebral fractures between otherwise matched subjects were more likely seen in subjects with a BMI  24 kg/m2 than in subjects  26 kg/m2 [222,223]. Another study showed prevalences of vertebral fractures in women 60 – 80 years of 79, 48, and 27% for BMIs of 19, 22, and 28 kg/m2 [224]. Clearly, low body weight increases osteoporosis risk and a more generous body weight confers some protection against osteoporosis and related fracture risk to a point. Few studies, only one of which is prospectively designed [214], have examined the association between body composition and fracture risk [199,214,219]. Because body weight is composed of muscle and adipose tissue, it is reasonable that the components of this weight may offer some explanation of the association between body weight and fractures. In a Swedish study comparing urban and rural populations, lower muscle mass, as measured by BIA, was found in women experiencing fragility fractures after age 70. Muscle mass was correlated more strongly with forearm BMC than body weight, suggesting the importance of muscle mass as a primary determinant of bone mass and fracture risk [225]. There is additional evidence from the NHANES I cohort of 3595 women, ages 40 – 77 years, who were followed prospectively for 10 years, that muscle and adipose tissue contribute independently to the risk of hip fractures [214]. Of interest, women with hip fractures tended to weigh less and have lower triceps skinfold thickness and muscle arm area. These differences translated into a twofold reduction in the relative risk of hip fracture for BMI, tricep skinfolds, and muscle arm area (2.6; 1.8 to 3.9, 2.0; 1.4 to 2.9, and 1.7; 1.2 to 2.4), respectively, even after adjusting for potential confounders. Comparisons of subscapular measures and triceps skinfold thicknesses indicated that the risk of hip fractures was not associated with regional differences in adiposity. Whether body weight alone or the individual components of that weight are more significant in explaining the association between obesity and fracture risk has not been fully elucidated.

761

F. Mechanisms of the Association between Obesity and Bone Mineral Density Obesity, even if moderate, appears to protect against osteoporosis. The relationship between body size and BMD has not been fully explained, although several potential mechanisms have been presented. As shown, obese women tend to have higher BMD, particularly in the axial (weight-bearing) skeleton, than nonobese women [151 – 158]. Such differences in skeletal mass are largely ascribed to the increased mechanical load on the skeleton [210]. Indeed, if obesity persists from early adulthood, it is likely that excess body weight is a primary factor in both maximizing peak bone mass and maintaining BMD as an older adult. This mechanical hypothesis, however, does not entirely explain the association between obesity and BMD. In a study by Ribot et al. [157], vertebral BMD of slightly obese postmenopausal women was higher than that of normal weight women, although a similar trend was not demonstrated among perimenopausal women, suggesting the importance of hormonal factors in the relationship between obesity and BMD. In postmenopausal women, most circulating estrogens are derived from the peripheral conversion of androstenedione to estrone in adipose tissue (10 – 15%) [226 – 229] and muscle (25 – 30%) [230]. The adrenal production of androgens tends to be higher in obese than nonobese women, resulting in more available androgen precursors for conversion by the aromatase enzyme to estrogens in peripheral tissues. This peripheral aromatization seems to be accelerated in obese women [229], as shown by the increased percentage of androstenedione converted to estrone, ranging from 1 – 2% in normal weight women to 12 – 15% in women weighing between 300 and 400 lbs. [227]. Moreover, the conversion of androgens to estrogens is increased with age [230] and may be mediated by the type of obesity present (upper vs lower) [151,231,232]. Moreover, obese postmenopausal women may have lower concentrations of sex hormone-binding globulin (SHBG), which may result in a greater proportion of free, metabolically active estrogens as well as androgens [233]. Estrogen deficiency is integral in the development of postmenopausal osteoporosis and fractures. Because obese women have greater amounts of both adipose and muscle mass than the nonobese, it is possible that adiposity protects against osteoporosis by providing higher amounts of circulating estrogens, which may impact BMD through osteoclastic and osteoblastic activity, as well as in alterations in bone remodeling. This is supported by most [235], but not all [196], studies. Reid and colleagues [196] maintained that the conversion of adrenal androgens to estrogen in adipose tissue does not explain the relationship between adiposity and BMD, as this association was found to be independent of serum estrone levels. From Framingham data,

762 adiposity was correlated positively with estrogen levels in postmenopausal women, but related inversely to the rate of postmenopausal bone loss [165]. Moreover, women who never used replacement estrogens had a stronger association with BMD and body weight than women who used estrogen replacement therapy. Because estrogen users are generally less obese than nonusers [2], this lower association may rather reflect the narrow distribution of body weight among estrogen users. The observation that obese women experience attenuated rates of bone loss as compared to their normal weight counterparts might be explained by their enhanced bone formation, reduced bone resorption, or a combination of the two. Frumar et al. [236] demonstrated a negative correlation between fasting urinary calcium/creatinine ratio (Ca/Cr) and percentage ideal body weight (r   0.43, P  0.01) and a reduced Ca/Cr in obese than in nonobese postmenopausal women. As fat mass increases, bone resorption tends to decline, as measured by Ca/Cr, without a simultaneous decline in bone formation, measured by bone gla protein [237]. This adaptive mechanism in bone remodeling, which supports greater bone formation than bone resorption, might inhibit net bone loss. It is possible that calcium and vitamin D homeostasis is altered in obese women, resulting in reduced ionized calcium activity, increased PTH concentration, reduced 25-hydroxyvitamin D, and increased 1,25(-OH)2D [238 – 240]. Bell and colleagues [238] support alterations in the vitamin D endocrine system resulting from the greater mechanical strain on the axial skeleton. In obese subjects, greater than 30% or more above ideal body weight, 25-hydroxyvitamin D was significantly lower in the obese [238 – 241], whereas 1,25-dihydroxyvitamin D and serum PTH were significantly higher in obese than in age-matched, nonobese women [242]. Of interest, PTH and IBW were positively correlated (r  0.72, P  0.01). Given this, Bell et al. [238] asserted that the vitamin D endocrine system is altered in obese individuals; this alteration is characterized by secondary hyperparathyroidism, enhanced renal tubular reabsorption of calcium, and increased circulating 1,25(OH)2D. The consequent reduction in serum 25-hydroxyvitamin D is ascribed to a feedback inhibition on its hepatic synthesis, as determined by the increased concentration of 1,25-dihydroxyvitamin D [238]. Such alterations in the vitamin D endocrine system were confirmed by histomorphometric analyses of iliac crest biopsies in morbidly obese subjects [243].

G. Mechanisms for the Association between the Degree of Obesity and Fracture Risk Potential mechanisms for the protective effect of body weight on fracture risk have been postulated. First, as expected with higher body weight, there are greater amounts

CAULEY AND SALAMONE

of both muscle and adipose tissue, thereby increasing the mechanical stress on the weight-bearing skeleton. Greater weight-bearing stress present during early adulthood can maximize peak skeletal density and, in later life, may attenuate postmenopausal bone loss. Indeed, larger-boned women have more bone mass to lose than small-boned women before reaching critical fracture threshold levels. An alternative explanation may be that the obese have improved bone strength or bone quality, although to date, no studies have examined the role of obesity on bone quality. Second, heavy postmenopausal women have higher levels of endogenous estrogen because of aromatization of androstenedione to estrone in muscle and adipose tissue. In absence of ovarian function, the conversion of estrone to estradiol represents the primary source of circulating estradiol. In addition, because SHBG is higher in hip fracture patients than in age-matched controls [132], a difference largely ascribed to differences in body size, there is a lower concentration of biologically available estradiol and testosterone in hip fracture patients. Taken together, obese postmenopausal women have a significantly higher production of estradiol than their thinner counterparts, the availability of which is increased by lower concentrations of SHBG [244,245]. Accordingly, because estrogen therapy is known to prevent bone loss in postmenopausal women [246], an important mechanism by which obesity is protective against fracture risk is through both increased estrogen production and availability after menopause. Third, because obese women tend to be less active than thin women, they may be less likely to fall and sustain a fracture. However, in the event of a traumatic fall, higher amounts of muscle and adipose tissue in the obese may offer a cushioning protection for the skeleton [234]. In the event of bone trauma, some researchers have even advocated that slim elderly wear protective padding in the hip region to compensate for the loss of fat protection [247]. Finally, it is reasonable that obesity may protect against osteoporotic fracture through alterations in the vitamin D endocrine system, as suggested previously by Bell and colleagues [238].

H. Future Directions Although much progress has been made in elucidating the mechanisms associated with obesity, BMD, and fractures, much remains to be learned, particularly in the area of body composition and BMD. There is a major need for additional prospective cohort studies to examine changes in body weight and weight cycling in relation to BMD and fracture risk. Because the contribution of lifetime weight and weight change to BMD and fracture risk is unclear, further examination of both the duration and the degree of obesity on such

763

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones

end points is important. Of particular interest will be determining the relative importance of weight gain and weight loss on fracture risk among the elderly and whether weight change or fluctuation in early adulthood is an important determinant of BMD and fracture risk in later years. Additional studies are necessary to examine whether individual compartments of body weight, specifically lean and fat mass, are more important determinants of BMD and fracture risk than body weight alone. Finally, because of the substantially higher prevalence of overweight in African-American and Hispanic women as compared to Caucasian women [248], it is important to consider the potential effect of obesity on racial differences in BMD and fracture risk. Further explorations into genetic/environmental and their interaction contributions to BMD are integral to gaining a better understanding of the associations between obesity and both BMD and fractures.

3.

4.

5.

6.

7.

8.

9.

VI. OVERALL SUMMARY This chapter reviewed the relationship among three important determinants of osteoporotic risk. Estrogen replacement therapy has been consistently shown to preserve bone mass and prevent fractures. The biggest question surrounding estrogens relates to the risks and benefits of long-term use, as data suggest that women need to continue estrogens indefinitely. In addition, compliance to estrogen therapy is low; future studies are needed to identify ways to maximize compliance. Thiazide diuretics may prevent hip fractures, but the underlying mechanisms for this effect are not known. More data are needed in normotensives before thiazide diuretics can be advocated for the prevention of osteoporotic fractures. Data are needed not only on the efficacy of thiazide diuretics on fracture, but also careful documentation of the risks and side effects of this therapy. Obesity may be in the causal pathway among endogenous estrogens, bone mass, and fractures. In addition, there are other mechanisms for an effect of obesity on osteoporotic risk. Future research needs to identify this causal pathway to improve our understanding of the etiology of osteoporotic risk. In addition, prospective studies are needed to test the hypothesis that low serum estrogens predict individuals at risk for fracture, particularly hip fracture.

10.

11.

12. 13.

14.

15.

16.

17.

18.

19.

References 20. 1. J. A. Cauley, S. R. Cummings, D. G. Seeley, D. Black, W. Browner, L. H. Kuller, and M. C. Nevitt, Effects of thiazide diuretic therapy on bone mass, fractures, and falls. Ann. Intern. Med. 118, 666 – 673 (1993). 2. J. A. Cauley, S. R. Cummings, D. M. Black, S. R. Mascioli, and D. G. Seeley, Prevalence and determinants of estrogen replacement

21. 22.

therapy in elderly women. Am. J. Obstet. Gynecol. 163(5), 1438 – 1444 (1990). J. A. Cauley, J. P. Gutai, L. H. Kuller, and J. G. Powell, The relationship of endogenous sex steroid hormone concentrations to serum lipid and lipoprotein levels in postmenopausal women. Am. J. Epidemiol. 132, 884 – 894 (1990). R. D. Wasnich, R. J. Benfante, K. Yano, L. K. Heilburn, and J. M. Vogel, Thiazide effect on the mineral content of bone. N. Engl. J. Med. 309, 344 – 347 (1983). M. R. Sowers, R. B. Wallace, and J. H. Lemke, Correlates of mid-radius bone density among postmenopausal women: A community study. Am. J. Clin. Nutr. 41, 1045 – 1053 (1985). R. D. Wasnich, R. D. Philip, L. K. Heilbrun, J. M. Vogel, K. Yano, and R. J. Benefante, Differential effects of thiazide and estrogen upon bone mineral content and fracture prevalence. Obstet. Gynecol. 67, 457 – 462 (1986). P. Adland-Davenport, M. W. McKenzie, M. Notelovitz, L. C. McKenzie, and J. F. Pendergast, Thiazide diuretics and bone mineral content in postmenopausal women. Am. J. Obstet. Gynecol. 152, 630 – 634 (1985). R. Lindsay, B. Marshall, A. Haboubi, B. S. Herrington, and J. Tohme, Increased axial bone mass in women with hypertension: Role of thiazide therapy. J. Bone Miner. Res. 2, 223 (1987). N. W. Glynn, E. N. Meilahn, M. Charron, S. J. Anderson, L. H. Kuller, and J. A. Cauley, Determinants of bone mineral density in older men. J. Bone Miner. Res. 10, 1769 – 1777 (1995). D. J. Morton, E. L. Barrett-Connor, and S. L. Edelstein, Thiazides and bone mineral density in elderly men and women. Am. J. Epidemiol. 139, 1107 – 1115 (1994). R. Wasnich, J. Davis, P. Ross, and J. Vogel, Effects of thiazide on rates of bone mineral loss: A longitudinal study. Br. Med. J. 301, 1303 – 1305 (1990). B. Dawson-Hughes and S. Harris, Thiazides and seasonal bone change in healthy postmenopausal women. Bone Miner. 21, 41 – 51 (1993). M. R. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, Body size, estrogen use and thiazide diuretic use affect 5-year radial bone loss in postmenopausal women. Osteopor. Int. 3, 314 – 321 (1993). J. S. Adams, C. F. Song, and V. Kantorovich, Rapid recovery of bone mass in hypercalciuric men treated with hydrochlorothiazide. Ann. Intern. Med. 130, 658 – 660 (1999). C. Christiansen, M. S. Christiansen, P. McNair, C. Hagen, K. E. Stocklund, and I. Transbol, Prevention of early postmenopausal bone loss: Controlled 2-year study in 315 normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). I. Transbol, M. S. Christiansen, G. F. Jensen, C. Christiansen, and P. McNair, Thiazide for the postponement of postmenopausal bone loss. Metabolism 31, 383 – 386 (1982). T. D. Giles, G. E. Sander, L. E. Roffidal, A. C. Quiroz, and A. L. Mazzu, Comparative effects of nitrendipine and hydrochlorothiazide on calciotropic hormones and bone density in hypertensive patients. Am. J. Hypertens. 5, 875 – 879 (1992). R. D. Wasnich, J. W. Davis, Y. F. He, H. Petrovich, and P. D. Ross, Thiazides reduce bone loss: A randomized double-blind, placebo controlled trial. J. Bone Miner. Res. 8, 123 (1993). [Abstract] W. E. Hale, R. B. Steward, and R. G. Marks, Central nervous system symptoms of elderly subjects using antihypertensive drugs. J. Am. Geriatr. Soc. 32, 5 – 10 (1984). A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). S. Rashiq and R. F. Logan, Role of drugs in fractures of the femoral neck. Br. Med. J. 292, 861 – 863 (1986). H. M. Taggart, Do drugs affect the risk of hip fracture in elderly women? J. Am. Geriatr. Soc. 36, 1006 – 1010 (1988).

764 23. A. Stevens and C. Mulrow, Drugs affecting postural stability and other risk factors in the hip fracture epidemic-case-control study. Community Med. 11, 27 – 34 (1989). 24. D. T. Felson, D. Sloutskis, J. J. Anderson, J. M. Anthony, and D. P. Kiel, Thiazide diuretics and the risk of hip fracture. JAMA 265, 370 – 373 (1991). 25. F. E. Heidrich, A. Stergachis, and K. M. Gross, Diuretic drug use and the risk for hip fracture. Ann. Intern. Med. 115, 1 – 6 (1991). 26. R. G. Cumming and R. J. Klineberg, Psychotropics, thiazide diuretics and hip fractures in the elderly. Med. J. Aust. 158, 414 – 417 (1993). 27. W. A. Ray, W. Downey, M. R. Griffin, and L. J. Melton, Long-term use of thiazide diuretics and risk of hip fracture. Lancet April, 687 – 690 (1989). 28. A. Z. LaCroix, J. Wienpahl, L. R. White, R. B. Wallace, P. A. Scherr, L. K. George, J. Cornoni-Huntley, and A. M. Ostfeld, Thiazide diuretic agents and the incidence of hip fracture. N. Engl. J. Med. 322, 286 – 290 (1990). 29. D. Feskanich, W. C. Willett, M. J. Stampfer, and G. A. Colditz, A prospective study of thiazide use and fractures in women. Osteopor. Int. 7, 79 – 84 (1997). 30. Z. Guo, P. Wills, M. Vitanen, J. Fastbom and B. Winblad, Cognitive impairment, drug use, and the risk of hip fracture in persons over 75 years old: A community-based prospective study. Am. J. Epidemiol. 148, 887 – 92 (1998). 31. G. Jones, T. Nguyen, P. N. Sambrook, and J. A. Eisman, Thiazide diuretics and fractures:Can meta-analysis help? J. Bone Miner. Res. 10, 106 – 111 (1995). 32. S. Middler, C. Y. Pak, F. Murad, and F. Bartter, Thiazide diuretics and calcium metabolism. Metabolism 22, 139 – 146 (1973). 33. E. R. Yendt and M. Cohanim, Prevention of calcium stones with thiazides. Kidney Int. 13, 397 – 409 (1978). 34. P. N. Singhellakis, A. E. Nikov, C. Nicolou et al., Effect of thiazide diuretics on parathyroid function in humans. Acta Endocrinol. Suppl. 261, 32 – 35 (1983). 35. K. Sakhaee, M. J. Nicar, K. Glass, J. E. Zerwekh, and C. Y. Pak, Reduction of intestinal calcium absorption by hydrochlorothiazide in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 59, 1037 – 1043 (1984). 36. T. Steiniche, L. Mosekilde, M. S. Christensen, and F. Melsea, Histomorphometric analysis of bone in idiopathic hypercalciuria before and after treatment with thiazide. APMIS 97, 302 – 308 (1989). 37. R. O. Pierce and A. Perry, The effects of thiazide diuretics on bone mineral metabolism. J. Natl. Med. Assoc. 90, 46 – 50 (1998). 38. A. Z. LaCroix, Thiazide diuretic agents and prevention of hip fracture. Compr. Ther. 17, 30 – 39 (1991). 39. M. Nevitt, Falls in older persons: Risk factors and prevention. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, eds.). National Academy Press, Washington, DC, 1990. 40. W. A. Ray, Thiazide diuretics and osteoporosis: Time for a clinical trial? Ann. Intern. Med. 115, 64 – 65 (1991). 41. D. K. Wysowski, L. Golden, and L. Burke, Use of menopausal estrogens and medroxyprogesterone in the United States, 1982 – 1992. Obstet. Gynecol. 85, 6 – 10 (1995). 42. V. A. Ravnikar, Compliance with hormone therapy. Am. J. Obstet. Gynecol. 156, 1332 (1987). 43. S. J. Birge, Soy phytoestrogens: An adjunct to hormone replacement therapy? Maturitas 7, 209 – 212 (2000). 44. National Osteoporosis Foundation, Osteoporosis: Review of the evidence for prevention, diagnosis and treatment and cost-effectiveness analysis. Osteopor. Int. Suppl. 4, S1 – 86 (1998). 45. R. Lindsay, Hormone replacement therapy for prevention and treatment of osteoporosis. Am. J. Med. 95, 37S–39S (1993). 46. C. Christiansen, Hormonal prevention and treatment of osteoporosisstate of art 1990. J. Steroid Biochem. Mol. Biol. 37, 447 – 449 (1990).

CAULEY AND SALAMONE 47. E. F. Eriksen, D. S. Colvard, N. J. Berg et al., Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 48. S. L. Corson, A decade of experience with transdermal estrogen replacement therapy: Overview of key pharmacologic and clinical findings. Int. J. Fertil. 38(2), 79 – 91 (1993). 49. J. C. Stevenson, M. P. Cust, K. F. Gangar, T. C. Hillard, B. Lees, and M. I. Whitehead, Effects of transdermal versus oral hormone replacement therapy on bone density in spine and proximal femur in postmenopausal women. Lancet 335, 265 – 269 (1990). 50. E. F. Holland, A. T. Leather, and J. W. Studd, The effect of 25-mg percutaneous estradiol implants on the bone mass of postmenopausal women. Obstet. Gynecol. 83, 43 – 46 (1994). 51. S. J. S. Ryde, K. Bowen-Simpkins, P. Bowen-Simpkins, W. D. Evans, W. D. Morgan, and J. E. Compston, The effect of oestradiol implants on regional and total bone mass: A three-year longitudinal study. Clin. Endocrinol. 40, 33 – 38 (1994). 52. B. J. Riis, K. Thomsen, V. Strom, and C. Christiansen, The effect of percutaneous estradiol and natural progesterone on postmenopausal bone loss. Am. J. Obstet. Gynecol. 156, 61 – 65 (1987). 53. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral density in the postmenopausal estrogen/progestin interventions trial. J. Bone Miner. Res. 9, 1467 – 1476 (1994). 54. J. C. Gallagher, W. T. Kable, and D. Goldgar, Effect of progestin therapy on cortical and trabecular bone: Comparison with estrogen. Am. J. Med. 90, 171 – 178 (1991). 55. R. Civitelli, D. Agnusdei, P. Nardi, F. Zacchei, L. V. Avioli, and C. Gennari, Effects of one-year treatment with oestrogens on bone mass, intestinal calcium absorption and 25-hydroxyvitamin D-1 hydroxylase reserve in postmenopausal osteoporosis. Calcif. Tissue Int. 42, 77 – 86 (1988). 56. C. Christiansen and B. J. Riis, 17b-Oestradiol and continuous norethisterone: A unique treatment for established osteoporosis in elderly women. J. Clin. Endocrinol. Med. 71, 836 – 841 (1990). 57. R. Lindsay and J. F. Tohme, Oestrogen treatment of patients with established postmenopausal osteoporosis. Obstet. Gynecol. 76, 290 – 295 (1990). 58. E. G. Lufkin, H. W. Wahner, W. M. O’Fallon, S. F. Hodgson, M. A. Kotowicz, A. W. Lane, H. L. Judd, R. H. Caplan, and B. L. Riggs, Treatment of postmenopausal osteoporosis with transdermal oestrogen. Ann. Int. Med. 117, 1 – 9 (1992). 59. A. B. Grey, T. F. Cundy, and I. R. Reid, Continuous combined oestrogen/progestin therapy is well tolerated and increases bone density at the hip and spine in post-menopausal osteoporosis. Clin. Endocrinol. 40, 671 – 677 (1994). 60. D. T. Felson, Y. Zhang, M. T. Hannon, D. T. Kiel, P. W. F. Wilson, and J. J. Anderson, The effect of postmenopausal estrogen therapy on bone density in elderly women. N. Engl. J. Med. 329, 1141 – 1146 (1993). 61. N. M. Resnick and S. L. Greenspan, “Senile” osteoporosis reconsidered. JAMA 261, 1025 – 1029 (1989). 62. M. E. Quigley, P. L. Martin, A. M. Burnier, and P. Brooks, Estrogen therapy arrests bone loss in elderly women. Am. J. Obstet. Gynecol. 156, 1516 – 1523 (1987). 63. C. W. Marx, G. E. Dailey, C. Cheney, V. C. Vint, and D. B. Muchmore, Do estrogens improve bone mineral density in osteoporotic women over age 65? J. Bone Miner. Res. 7, 1275 – 1279 (1992). 64. K. E. Ensrud, L. Palermo, D. M. Black, J. Cauley, M. Jergas, E. S. Orwoll, M. C. Nevitt, K. M. Fox, and S. R. Cummings, Hip and calcaneal bone loss increase with advancing age: Longitudinal results from the Study of Osteoporotic Fractures. J. Bone Miner. Res. 11, 1778 – 87 (1995). 65. D. L. Schneider, E. L. Barrett-Connor, and D. J. Morton, Timing of postmenopausal estrogen for optimal bone mineral density. JAMA 277, 543 – 547 (1997).

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 66. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial Writing Group, Effects of hormone therapy on bone mineral density: Results from the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. JAMA 276, 1389 – 96 (1996). 67. R. R. Recker, M. Davies, R. M. Dowd, and R. P. Heaney, The effect of low-dose continuous estrogen and progesterone therapy with calcium and vitamin D on bone in elderly women: A randomized controlled trial. Ann. Intern. Med. 130, 897 – 904 (1999). 68. H. K. Genant, D. J. Baylink, J. C. Gallagher, S. T. Harris, P. Steiger, and M. Herber, Effect of estrone sulfate on postmenopausal bone loss. Am. J. Obstet Gynecol. 76, 579 – 84 (1990). 69. S. T. Harris, H. K. Genant, D. J. Baylink, J. C. Gallagher, S. K. Karp, M. A. McConnell et al., The effects of estrone (Ogen) on spinal bone density of postmenopausal women. Arch. Intern. Med. 151, 1980 – 4 (1991). 70. B. Ettinger, H. K. Genant, P. Steiger, and P. Madvig, Low-dosage micronized 17 -estradiol prevents bone loss in postmenopausal women. Am. J. Obstet. Gynecol. 6, 479 – 88 (1992). 71. S. Adami, R. Suppi, F. Bertoldo, M. Rossini, M. Residori, V. Maresca, and V. LoCascio, Transdermal estradiol in the treatment of postmenopausal bone loss. Bone Miner. 7, 79 – 86 (1989). 72. C. S. Field, S. J. Ory, H. W. Wahner, R. R. Herrmann, H. L. Judd, and B. L. Riggs, Preventive effects of transdermal 17 beta-estradiol on osteoporotic changes after surgical menopause: A two-year placebo-controlled trial. Am. J. Obstet. Gynecol. 168, 114 – 21 (1993). 73. H. K. Genant, J. Lucas, S. Weiss, M. Akin, R. Emkey, H. McNancyFlint, R. Downs, J. Mortola, N. Watts, H. M. Yang, N. Banav, J. J. Brennan, and J. C. Nolan, for the Estratab/Osteoporosis Study Group, Low-dose estrified estrogen therapy: Effects on bone, plasma estradiol concentrations, endometrium and lipid levels. Arch. Intern. Med. 157, 2609 – 15 (1997). 74. L. Speroff, J. Rowan, J. Symons, H. Genant, and W. Wilborn for the CHART study group, The comparative effect on bone density, endometrium and lipids of continuous hormones as replacement therapy (CHART study). JAMA 276, 1397 – 1403 (1996). 75. T. A. Hutchinson, S. M. Polansky, and A. R. Feinstein, Postmenopausal estrogens protect against fractures of hip and distal radius: A case-control study. Lancet 6, 705 – 709 (1979). 76. N. S. Weiss, C. L. Ure, J. H. Ballard, A. R. Williams, and J. R. Daling, Decreased risk of fractures of the hip and lower forearm with postmenopausal use of estrogen. N. Engl. J. Med. 303, 1195 – 1198 (1980). 77. R. E. Johnson and E. E. Specht, The risk of hip fracture in postmenopausal females with and without estrogen drug exposure. Am. J. Public Health 71, 138 – 144 (1981). 78. N. Krieger, J. L. Kelsey, T. R. Holford, and T. O’Connor, An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 141 – 148 (1982). 79. J. A. Kanis, O. Johnell, B. Gullberg, E. Allander, G. Dilsen, C. Gennari et al., Evidence for efficacy of drugs affecting bone metabolism in preventing hip fracture. Br. Med. J. 305, 1124 – 1128 (1992). 80. C. B. Hammond, F. R. Jelovsek, K. L. Lee, W. T. Creasman, and R. T. Parker, Effects of long-term estrogen replacement therapy. Am. J. Obstet. Gynecol. 122, 525 – 536 (1979). 81. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Intern. Med. 102, 319 – 324 (1985). 82. D. P. Kiel, D. T. Felson, J. J. Anderson, P. W. Wilson, and M. A. Moskowitz, Hip fracture and the use of estrogens in postmenopausal women: The Framingham Study. N. Engl. J. Med. 317, 1169 – 1174 (1987). 83. A. Paganini-Hill, A. Chao, R. K. Ross, and B. E. Henderson, Exercise and other factors in the prevention of hip fracture: The Leisure World Study. Epidemiology 2, 16 – 25 (1991).

765 84. P. Maxim, B. Ettinger, and G. M. Spitalny, Fracture protection provided by long-term estrogen treatment. Osteopor. Int. 5, 23 – 29 (1995). 85. T. Naessen, I. Persson, H. O. Adama, R. Bergstrom, and L. Bergkvist, Hormone replacement therapy and the risk for the first hip fracture: A prospective, population-based cohort study. Ann. Intern. Med. 113, 95 – 103 (1990). 86. J. A. Cauley, D. G. Seeley, K. Ensrud, B. Ettinger, D. Black, and S. R. Cummings, Estrogen replacement therapy and fractures in older women. Ann. Intern. Med. 122, 9 – 16 (1995). 87. D. Grady, S. M. Rubin, D. B. Petitti, C. S. Fox, D. Black, B. Ettinger et al., Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann. Intern. Med. 117, 1016 – 1037 (1992). 88. D. P. Kiel, J. A. Baron, J. J. Anderson, M. T. Hannan, and D. T. Felson, Smoking eliminates the protective effect of oral estrogens on the risk for hip fracture among women. Ann. Intern. Med. 116, 716 – 721 (1992). 89. A. R. Williams, N. S. Weiss, C. L. Ure, J. Ballard, and J. R. Daling, Effect of weight, smoking, and estrogen use on the risk of hip and forearm fractures in postmenopausal women. Obstet. Gynecol. 60, 695 – 696 (1992). 90. L. E. Nachtigall, R. H. Nachtigall, R. D. Nachtigall, and E. M. Beckman, Estrogen replacement therapy I: A 10-year prospective study in the relationship to osteoporosis. Am. J. Obstet. Gynecol, 53, 277 – 81 (1979). 91. R. Lindsay, D. M. Hart, C. Forrest and C. Baird, Prevention of spinal osteoporosis in ooporetctomized women. Lancet 2, 1151 – 53 (1980). 92. J. Windeler and S. Lange, Events per person year: A dubious concept. Br. Med. J. 310, 454 – 56 (1995). 93. M. H. Komulainen, H. Kroger, M. T. Tuppurainen, A. M. Heikkinen, E. Alhava, R. Honkanen, and S. Saarikowski, HRT and vitamin D in prevention of nonvertebral fractures in postmenopausal women: A 5year randomized trial. Maturitas 31, 45 – 54 (1998). 94. S. Hulley, D. Grady, T. Bush et al., Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women: Heart and Estrogen/progestin Replacement Study (HERS) research group. JAMA 380, 605 – 13 (1998). 95. J. A. Cauley, D. M. Black, E. Barrett-Connor, F. Harris, K. Shields, W. Applegate, and S. R. Cummings, Effects of hormone replacement therapy on clinical fractures and height loss: The Heart Estrogen/progestin Replacement Study (HERS). Submitted for publication. 96. B. Ettinger and D. Grady, Maximizing the benefit of estrogen therapy for prevention of osteoporosis. Menopause 1, 19 – 36(1994). 97. D. B. Petititi, Coronary heart disease and estrogen replacement therapy: Can compliance bias explain the results of observational studies? Ann. Epidemiol. 4, 115 – 118 (1994). 98. C. B. Hammond, F. R. Jelovsek, K. L. Lee, W. T. Creasman, and R. T. Parker, Effects of long-term estrogen replacement therapy. II. Neoplasia. Am. J. Obstet. Gynecol. 233, 537 – 547 (1979). 99. A. H. MacLennan, A. MacLennan, and D. Wilson, The prevalence of estrogen replacement therapy in South Australia. Maturitas 16, 175 – 183 (1993). 100. F. P. Groeneveld, F. P. Bareman, R. Barentsen, H. J. Dokter, A. C. Drogendijk, and A. W. Hoes, Determinants of first prescription of hormone replacement therapy: A follow-up study among 1689 women aged 45 – 60 years. Maturitas 20, 81 – 89 (1995). 101. L. M. Salamone, A. R. Pressman, D. G. Seeley, and J. A. Cauley, Estrogen replacement therapy: A survey of elderly women. Submitted for publication. 102. A. Cano, Compliance to hormone replacement therapy in menopausal women controlled in a third level academic centre. Maturitas 20, 91 – 99 (1995). 103. C. C. Johnston, S. L. Hui, R. M. Witt et al., Early menopausal changes in bone mass and sex steroids. J. Clin. Endocrinol. Metab. 61, 905 – 911 (1985).

766 104. J. A. Cauley, J. P. Gutai, R. V. Sandler et al., The relationship of endogenous estrogen to bone density and bone area in normal postmenopausal women. Am. J. Epidemiol. 124, 752 – 761 (1986). 105. J. A. Cauley, J. P. Gutai, L. H. Kuller, J. Scott, and M. C. Nevitt, Black-white differences in serum sex hormones and bone mineral density. Am. J. Epidemiol. 139, 1035 – 1046 (1994). 106. K. K. Steinberg, L. W. Freni-Titulaer, E. G. DePuey et al., Sex steroids and bone density in premenopausal and perimenopausal women. J. Clin. Endocrinol. Metab. 69, 533 – 539 (1989). 107. S. Rozenberg, H. Ham, D. Bosson et al., Age, steroids and bone mineral content. Maturitas 12, 137 – 143 (1990). 108. S. Murphy, K. Khaw, M. J. Sneyd et al., Endogenous sex hormones and bone mineral density among community-based postmenopausal women. Postgrad. Med. J. 68, 908 – 913 (1992). 109. G. M. Hall, L. A. Perry, and T. D. Spector, Depressed levels of dehydroepiandrosterone sulphate in postmenopausal women with rheumatoid arthritis but no relation with axial bone density. Ann. Rheum. Dis. 52, 211 – 214 (1993). 110. I. S. Fentiman, D. Y. Wang, D. S. Allen et al., Bone density of normal women in relation to endogenous and exogenous oestrogens. Br. J. Rheumatol. 33, 808 – 815 (1994). 111. T. Murakami, M. Shiraki, H. Orimo et al., Serum estradiol and radial mineral content in postmenopausal females. Endocrinol. Jpn. 26, 635 – 636 (1979). 112. M. H. Rae, P. A. Mole, and C. R. Paterson, Endogenous factors affecting bone mineral content in post-menopausal women. Maturitas 13, 319 – 324 (1991). 113. S. Deutsch, F. Benjamin, V. Seltzer et al., The correlation of serum estrogens and androgens with bone density in the late postmenopause. Int. J. Gynaecol. Obstet. 25, 217 – 222 (1987). 114. M. E. Nelson, C. N. Meredith, B. Dawson-Hughes et al., Hormone and bone mineral status in endurance-trained and sedentary postmenopausal women. J. Clin. Endocrinol. Metab. 66, 927 – 933 (1988). 115. C. C. Johnston, S. L. Hui, and C. Longcope, Bone mass and sex steroid concentrations in postmenopausal caucasian diabetics. Metabolism 34, 544 – 550 (1985). 116. B. Ettinger, A. Pressman, P. Sklarin, D. C. Bauer, J. A. Cauley, and S. R. Cummings, Associations between low levels of serum estradiol, bone density, and fractures among elderly women: The Study of Osteoporotic Fractures. J. Clin. Endocrinol. Metab. 83, 2239 – 2243 (1998). 117. S. Khosla, L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 118. R. Lindsay, D. M. Hart, A. Sweeney et al., Osteopenia and osteoporosis. Calcif. Tissue Res. 22 (Suppl.), 213 – 216 (1977). 119. J. A. Johnston, R. A. Khairi, and C. Longcope, Age-related bone loss. In “Osteoporosis II”, (U.S. Barzel, ed.), pp. 91 – 100. Grune and Stratton, New York, 1979. 120. S. C. Manolagas, D. C. Anderson, and R. Lindsay, Adrenal steroids and the development of osteoporosis in oophorectomised women. Lancet 3143, 597 – 600 (1979). 121. B. J. Riis, C. Christiansen, L. J. Deftos et al., The role of serum concentrations of estrogens on postmenopausal osteoporosis and bone turnover. In “Osteoporosis I” (C. Christiansen, ed.), pp. 333 – 336. Stifsbogtrykkeri Printers, Aalborg, 1984. 122. B. J. Riis, P. Rodbro, and C. Christiansen, The role of serum concentrations of sex steroids and bone turnover in the development and occurrence of postmenopausal osteoporosis. Calcif. Tissue Int. 38, 318 – 322 (1986). 123. J. F. Aloia, A. N. Vaswani, J. K. Yeh et al., Determinants of bone mass in postmenopausal women. Arch. Intern. Med. 143, 1700 – 1704 (1983).

CAULEY AND SALAMONE 124. C. Slemenda, S. L. Hui, C. Longcope et al., Sex steroids and bone mass: A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 125. A. M. van Hemert, J. C. Birkenhager, F. H. De Jong et al., Sex hormone binding globulin in postmenopausal women: A predictor of osteoporosis superior to endogenous oestrogens. Clin. Endocrinol. 31, 499 – 509 (1989). 126. S. Harris, G. E. Dallal, and B. Dawson-Hughes, Influence of body weight on rates of change in bone density of the spine, hip and radius in postmenopausal women. Calcif. Tissue Int. 50, 19 – 23 (1992). 127. E. C. H. Van Beresteijn, J. P. R. M. Van Laarhoven, and A. G. H. Smals, Body weight and/or endogenous estradiol as determinants of cortical bone mass and bone loss in healthy early postmenopausal women. Acta Endocrinol. 127, 226 – 230 (1992). 128. K. Stone, D. C. Bauer, D. M. Black, P. Sklarin, K. E. Ensrud, and S. R. Cummings, for the Study of Osteoporotic Fractures research group, Hormonal predictors of bone loss in elderly women: A prospective study. J. Bone. Miner. Res. 13, 1167 – 1174 (1998). 129. D. H. Marshall, R. G. Crilly, and B. E. C. Nordin, Plasma androstenedione and oestrone levels in normal and osteoporotic postmenopausal women. Br. Med. J. 2, 1177 – 1179 (1977). 130. B. Pelc, D. H. Marshall, P. Guha et al., The relation between plasma androstenedione, plasma oestrone and androstenedione to oestrone conversion rates in post-menopausal women with and without fractures. Clin. Sci. Mol. Med. 54, 125 – 131 (1978). 131. B. J. Davidson, B. L. Riggs, C. B. Coulam et al., Concentration of cytosolic estrogen receptors in patients with postmenopausal osteoporosis. Am. J. Obstet. Gynecol. 136, 430 – 434 (1980). 132. B. J. Davidson, R. K. Ross, A. Paganini-Hill et al., Total and free estrogens and androgens in postmenopausal women with hip fractures. J. Clin. Endocrinol. Metab. 54, 115 – 120 (1982). 133. L. R. Laufer, B. J. Davidson, R. K. Ross et al., Physical characteristics and sex hormone levels in patients with osteoporotic hip fractures or endometrial cancer. Am. J. Obstet. Gynecol. 145, 585 – 590 (1983). 134. J. F. Aloia, S. F. Cohn, A. Vaswani et al., Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 135. B. L. Riggs, R. J. Ryan, H. W. Wahner et al., Serum concentrations of estrogen, testosterone and gonadotropins in osteoporotic and nonosteoporotic postmenopausal women. J. Clin. Endocrinol. Metab. 36, 1097 – 1099 (1973). 136. F. J. Bartizal, C. B. Coulam, T. A. Gaffey et al., Impaired binding of estradiol to vaginal mucosal cells postmenopausal osteoporosis. Gynecol. Invest. 7, 330 – 336 (1976). 137. B. J. Davidson, B. L. Riggs, H. W. Wahner et al., Endogenous cortisol and sex steroids in patients with osteoporotic spinal fractures. Obstet. Gynecol. 61, 275 – 278 (1983). 138. C. Longcope, R. S. Baker, S. L. Hur et al., Androgen and estrogen dynamics in women with vertebral crush fractures. Maturitas 6, 309 – 318 (1984). 139. E. Barrett-Connor, J. E. Mueller, D. G. Von Mühlen, G. A. Laughlin, D. L. Schneider, and D. J. Sartoris, Low levels of estradiol are associated with vertebral fractures in older men, but not women: The Rancho Bernardo study. J. Clin. Endocrinol. Metab. 85, 219 – 223 (2000). 140. S. R. Cummings, W. S. Browner, D. Bauer, K. Stone, K. Ensrud, S. Jamal, and B. Ettinger for the Study of Osteoporotic Fractures research group, Endogenous hormones and the risk of hip and vertebral fractures among older women. N. Engl. J. Med. 339, 733 – 8 (1998). 141. N. F. Santoro, N. F. Col, M. H. Eckman, J. B. Wong, S. G. Pauker, J. A. Cauley, J. Zmuda, S. Crawford, C. B. Johannes, J. E. Rossouw, and C. N. Merz, Therapeutic controversy: Hormone replacement therapy: Where are we going? J. Clin. Endocrinol. Metab. 84(6), 1798 – 812 (1999).

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 142. J. C. Gallagher, Moderation of the daily dose of HRT: Prevention of osteoporosis. Maturitas 33, S57 – S63 (1999). 143. R. J. Kuczmarski, K. M. Flegal, S. M. Campbell, and C. L. Johnson, Increasing prevalence of overweight among US adults: The National Health and Nutrition Examination Surveys. JAMA 272, 205 – 211 (1994). 144. Metropolitan Life Insurance Company, Metropolitan height and weight tables. Stat. Bull. Metrop. Insur. Co. 64, 2 – 9 (1983). 145. National Institutes of Health, Health implications of obesity: National Institutes of Health Consensus Development Conference Statement. Ann. Intern. Med. 103, 1073 – 1077 (1985). 146. A. B. Grey, M. C. Evans, J. P. Stapleton, and J. R. Reid, Body weight and bone mineral density in postmenopausal women with primary hyperparathyroidism. Am. College Phys. 121, 745 – 749 (1994). 147. T. L. Holbrook and E. Barrett-Connor, The association of lifetime weight and weight control patterns with bone mineral density in an adult community. Bone Miner. 20, 141 – 149 (1993). 148. F. A. Tremollieres, J. M. Pouilles, and C. Ribot, Vertebral postmenopausal bone loss is reduced in overweight women: A longitudinal study in 155 early postmenopausal women. J. Clin. Endocrinol. Metab. 77, 683 – 686 (1993). 149. S. Harris, G. E. Dallal, and B. Dawson-Hughes, Influence of body weight on rates of change in bone density of the spine, hip, and radius in postmenopausal women. Calcif. Tissue Int. 50, 19 – 23 (1992). 150. R. Lindsay, F. Cosman, B. S. Herrington, and S. Himmelstein, Bone mass and body composition in normal women. J. Bone Miner. Res. 7, 55 – 63 (1992). 151. K. Kin, K. Kushida, K. Yamazaki, S. Okamoto, and T. Inoue, Bone mineral density of the spine in normal Japanese subjects using dualenergy X-ray absorptiometry: Effect of obesity and menopausal status. Calcif. Tissue Int. 49, 101 – 106 (1991). 152. E. C. H. van Beresteijn, P. R. M. Laarhoven, and A. G. H. Smals, Body weight and/or endogenous estradiol as determinants of cortical bone loss in healthy early postmenopausal women. Acta Endocrinol. 127, 226 – 230 (1992). 153. W. C. Bevier, R. A. Wiswell, G. Pyka, K. C. Kozak, K. M. Newhall, and R. Marcus, Relationship of body composition, muscle strength, and aerobic capacity to bone mineral density in older men and women. J. Bone Miner. Res. 4, 421 – 432 (1989). 154. N. Pocock, J. Eisman, T. Gwinn, P. Sambrook, P. Kelly, J. Freun, and M. Yates, Muscle strength, physical fitness and weight but not age predict femoral neck bone mass. J. Bone Miner. Res. 4, 441 – 447 (1989). 155. J. C. Stevenson, B. Lees, M. Devenport, M. P. Cust, and K. F. Ganger, Determinants of bone density in normal women: Risk factors for future osteoporosis? Br. Med. J. 298, 925 – 928 (1989). 156. B. Dawson-Hughes, C. Shipp, L. Sadowski, and G. Dallal, Bone density of the radius, spine and hip in relation to percent ideal body weight in postmenopausal women. Calcif. Tissue Int. 40, 310 – 314 (1987). 157. C. Ribot, F. Tremollieres, J. M. Pouilles, M. Bonneu, F. German, and J. P. Louvet, Obesity and postmenopausal bone loss: The influence of obesity on vertebral density and bone turnover in postmenopausal women. Bone 8, 327 – 331 (1988). 158. N. A. Pocock, J. A. Eisman, M. G. Yeates, P. N. Sambrook, and S. Eberl, Physical fitness is a major determinant of femoral neck and lumbar spine bone mineral density. J. Clin. Invest. 78, 618 – 621 (1986). 159. H. W. Daniell, Osteoporosis of the slender smoker. Arch. Intern. Med. 136, 298 – 304 (1976). 160. N. Dalen, D. Hallberg, and B. Lamke, Bone mass in obese subjects. Acta Med. Scand. 197, 353 – 355 (1975). 161. Y. Nishizawa, H. Koyama, T. Shoji, H. Aratani, S. Hagiwara, T. Miki, and H. Morii, Obesity as a determinant of regional bone mineral density. J. Nutr. Sci. Vitaminol. 37, S65 – S70 (1991).

767 162. G. M. Wardlaw, Putting body weight and osteoporosis into perspective. Am. J. Clin. Nutr. 63(Suppl.), 433S – 436S (1996). 163. S. M. Haffner and R. L. Bauer, The association of obesity and glucose and insulin ocncentrations with bone density in premenopausal and postmenopausal women. Metabolism. 42, 735 – 8 (1993). 164. M. Shiraki, H. Ito, H. Fujimaki, and T. Higuchi, Relation between body size and bone mineral density with special reference to sex hormones and calcium regulating hormones in elderly females. Endocrinol. Jpn. 38, 343 – 9 (1996). 165. D. T. Felson, Y. Zhang, M. T. Hannan, and J. J. Anderson, Effects of weight and body mass index on bone mineral density in men and women: The Framingham study. J. Bone Miner. Res. 8, 567 – 573 (1993). 166. S. L. Edelstein, and E. Barrett-Connor, Relation between body size and bone mineral density in elderly men and women. Am. J. Epidemiol. 138, 160 – 169 (1993). 167. K. Yano, R. D. Wasnich, J. M. Vogel, and L. K. Heilbrun, Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am. J. Epidemiol. 119, 751 – 764 (1984). 168. P. Dano and C. Christiansen, Calcium absorption and bone mineral content following intestinal shunt operation in obesity: A comparison of three types of operation. Scand. J. Gastroenterol. 9, 775 – 779 (1974). 169. J. E. Compston, S. Vedi, E. Giantta, G. Watson, D. Civalleri, and N. Scopinaro, Bone histomorphometry and vitamin D status after biliopancreatic bypass for obesity. Gastroenterology 87, 350 – 356 (1984). 170. J. M. Campbell, T. K. Hunt, J. H. Karam, and P. H. Forsham, Jejunoilieal bypass as a treatment of morbid obesity. Arch. Intern. Med. 137, 602 – 610 (1977). 171. H. Rickers, I. Balslev, H. Foltved, and P. Rodbro, Bone mineral content before and after intestinal bypass operation in obese patients. 209, 203 – 207 (1981). 172. H. Hey, B. Lund, O. H. Sorensen, B. Lund, and M. S. Christensen, Impairment of vitamin D and bone metabolism in patients with bypass operation for obesity. Acta Med. Scand. (Suppl.) 624, 73 – 78 (1979). 173. A. M. Parfitt, M. J. Miller, B. Frame, A. R. Villanueva, I. Olivier, and D. L. Thomson, Metabolic bone disease after intestinal bypass for treatment of obesity. Ann. Intern. Med. 89, 193 – 199 (1978). 174. L. B. Jensen, F. Quaade, and O. L. Sorensen, Bone loss accompanying voluntary weight loss in obese humans. J. Bone Miner. Res. 9, 459 – 463 (1994). 175. J. E. Compston, M. A. Laskey, P. I. Croucher, A. Coxon, and S. Keitzman, Effect of diet-induced weight loss of total body bone mass. Clin. Sci. 82, 429 – 432 (1992). 176. A. Avenell, P. R. Richmond, M. E. J. Jean, and D. M. Reid, Bone loss associated with a high-fibre weight reduction diet in postmenopausal women. Eur. J. Clin. Nutr. 48, 561 – 6 (1994). 177. L. M. Salamone, J. A. Cauley, D. M. Black, L. Simkin-Silverman, W. Lang, E. Gregg, L. Palermo, R. S. Epstein, L. H. Kuller, and R. Wing, Effect of a lifestyle intervention on bone mineral density in premenopausal women: A randomized trial. Am. J. Clin. Nutr. 70, 97 – 103 (1999). 178. O. Svendsen, C. Hassager, and C. Christiansen, Effects of an energy restrictive diet, with or without exercise on lean tissue mass, resting metabolic rate, cardiovascular risk factors and bone in overweight postmenopausal women. Am. J. Med. 95, 131 – 140 (1993). 179. B. Mazess, P. R. Judy, C. R. Wilson, and J. R. Cameron, Progress in clinical use of photon absorptiometry: Clinical aspects of metabolic bone disease. Excerpta Med. Amst. 37 – 41 (1973). 180. L. Hyldstrup, T. Andersen, P. McNair, L. Breum, and I. Transbol, Bone metabolism in obesity: Changes related to severe overweight and dietary weight reduction. Acta Endocrinol. 129, 393 – 398 (1993).

768 181. J. E. Pritchard, C. A. Nowson, B. J. Strauss, J. S. Carlson, B. Kaymakci, and J. D. Wark, Evaluation of dual energy x-ray absorptiometry as a method of measurement of body fat. Eur. J. Clin. Nutr. 47, 216 – 228 (1993). 182. R. Roubenoff and J. J. Kehayias, The meaning and measurements of lean body mass. Nutr. Rev. 46, 163 – 175 (1991). 183. S. B. Heymsfield and M. Waki, Body composition in humans: Advances in the development of multicompartment chemical models. Nutr. Rev. 49, 97 – 108 (1991). 184. R. Roubenoff, J. J. Kehayias, B. Dawson-Hughes, and S. Heymsfield, Use of dual-energy x-ray absorptiometry in body composition studies: Not yet a “gold standard.” Am. J. Clin. Nutr. 58, 589 – 591 (1993). 185. J. Johnson and B. Dawson-Hughes, Precision and stability of dualenergy x-ray absorptiometry measurements. Calcif. Tissue Int. 49, 174 – 178 (1991). 186. K. Buhl, S. B. Heymsfield, M. Russel-Aulet, J. Wang, R. N. Pierson, and S. Lichtman, Effect of tissue thickness on bone density and bone mineral by dual energy x-ray absorptiometry. FASEB J. 5, A924 (1991). [Abstract] 187. P. Martin, C. Verhas, C. Als, I. Geerts, J. Paternot, and P. Bergmann, Influence of patient’s weight on dual-photon absorptiometry and dual-energy x-ray absorptiometry measurements of bone mineral density. Osteopor. Int. 3, 198 – 203 (1993). 188. C. Hassager and C. Christiansen, Influence of soft-tissue body composition on bone mass and metabolism. Bone 10, 415 – 419 (1989). 189. H. Rickers, I. Balstev, H. Foldved, and P. Rodbro, Bone mineral content before and after intestinal bypass operation in obese patients. Acta Med. Scnd. 209, 203 – 207 (1981). 190. T. N. Hangartner and C. C. Johnston, Influence of fat on bone measurements with dual-energy absorptiometry. Bone Miner. 9, 71 – 81 (1990). 191. J. Haarbo, A. Godtfredsen, C. Hassager, and C. Christiansen, Validation of body composition by dual-energy x-ray absorptiometry (DEXA). Clin. Physiol. 11, 331 – 341 (1991). 192. O. L. Svendsen, C. Haarbo, Hassager, and C. Christiansen, Accuracy of measurements of body composition by dual energy x-ray absorptiometry in vivo. Am. J. Clin. Nutr. 57, 605 – 8 (1993). 193. M. D. Van Loan, and P. L. Mayclin, Body composition assessment: Dual energy x-ray absorptiometry (DXA) compared to reference methods. Eur. J. Clin. Nutr. 46, 125 – 30 (1992). 194. M. D. Van Loan, H. L. Johnson, and T. F. Barbieri, Effect of weight loss on bone mineral content and bone mineral density in obese women. Am. J. Clin. Nutr. 67, 734 – 8 (1998). 195. L. D. Ritchie, J. C. King, E. B. Fung, M. D. Van Loan, and C. E. Cann, Pregnancy and lactation cause spinal bone mineral loss in humans. FASEB J. 7, A67 (1993). (abstract) 196. I. R. Reid, R. Ames, M. C. Evans, S. Sharpe, G. Gamble, J. T. France, T. M. T. Lim, and T. F. Cundy, Determinants of total body and regional bone mineral density in normal postmenopausal women: A key role for fat mass. J. Clin. Endocrinol. Metab. 75, 45 – 51 (1992). 197. J. E. Compston, M. Bhambhani, M. A. Laskey, S. Murphy, and K. T. Khaw, Body composition and bone mass in postmenopausal women. Clin. Endocrinol. 37, 426 – 431 (1992). 198. J. F. Aloia, A. Vaswani, R. Ma, and E. Flaster, To what extent is bone mass determined by fat-free or fat mass? Am. J. Clin. Nutri. 61, 1110 – 1114 (1995). 199. F. Doyle, J. Brown, and C. Lachance, Relation between bone mass and muscle weight. Lancet 1, 391 – 393 (1970). 200. M. Sinaki, M. C. C. McPhee, S. F. Hodgson, J. M. Merrit, and M. S. Offord, Relationship between bone mineral density of spine and strength of back extensors healthy postmenopausal women. Mayo Clin. Proc. 61, 116 – 122 (1986). 201. J. F. Aloia, D. M. McGowan, A. N. Vaswani, P. Ross, and S. H. Cohn, Relationship of menopause to skeletal and muscle mass. Am. J. Clin. Nutr. 53, 1378 – 1383 (1991).

CAULEY AND SALAMONE 202. D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone. Miner. Res. 7, 137 – 145 (1992). 203. I. R. Reid, M. C. Evans, and R. W. Ames, Volumetric bone density of the lumbar spine is related to fat mass but not lean mass in normal postmenopausal women. Osteopor. Int. 4, 362 – 367 (1994). 204. S. S. Harris, and B. Dawson-Hughes, Weight, body composition, and bone density in postmenopausal women. Calcif. Tissue. Int. 59, 428 – 432 (1996). 205. R. N. Baumgartner, P. M. Stauber, K. M. Koehler, L. Romero, and P. J. Garry, Associations of fat and muscle masses with bone mineral in elderly men and women. Am. J. Clin. Nutri. 63, 365 – 72 (1996). 206. Z. Chen, T. G. Lohman, W. A. Stini, C. Ritenbaugh, and M. Aickin, Fat or lean tissue mass: Which one is the major determinant of bone mineral mass in healthy postmenopausal women? J. Bone Miner. Res. 12, 144 – 151 (1997). 207. R. B. Sandler, Muscle strength assessments and the prevention of osteoporosis. JAGS 37, 1192 – 1197 (1989). 208. T. V. Nguyen, G. M. Howard, P. J. Kelly, and J. A. Eisman, Bone mass, lean mass and fat mass: Same genes or same environment? Am. J. Epidemiol. 147, 3 – 16 (1998). 209. M. M. Hla, J. W. Davis, P. D. Ross, R. D. Wasnich, J. A. Yates, P. Ravn, D. J. Hosking, and M. R. McClung, A multicenter study of the influence of fat and lean mass on bone mineral content: Evidence for differences in their relative influence at major fracture sites. Am. J. Clin. Nutr. 64, 354 – 360 (1996). 210. G. A. Rodan, Mechanical loading, estrogen deficiency and the coupling of bone formation and bone resorption. J. Bone Miner. Res. 6, 527 – 530 (1991). 211. R. G. Cumming and R. J. Klineberg, Case-control study of risk factors for hip fractures in the elderly. Am. J. Epidemiol. 139, 493 – 503 (1994). 212. S. B. Jaglal, N. Kreiger, and G. Darlington, Past and recent physical activity and risk of hip fracture. Am. J. Epidemiol. 138, 107 – 118 (1993). 213. J. L. Kelsey, W. S. Browner, D. G. Seeley, M. C. Nevitt, and S. R. Cummings, Risk factors for fractures of the distal forearm and proximal humerus. Am. J. Epidemiol. 135, 477, 489 (1992). 214. M. E. Farmer, T. Harris, J. H. Madans, R. B. Wallace, J. CornoniHuntley, and L. R. White, Anthropometric indicators and hip fracture. The NHANES I epidemiologic follow-up study. J. Am. Geriatr. Soc. 37, 9 – 16 (1989). 215. P. Gardsell, O. Johnell, and B. E. Nilsson, Predicting fractures in women by using forearm densitometry. Calcif. Tissue Int. 44, 235 – 242 (1989). 216. M. E. Pruzansky, M. Turano, M. Luckey, and R. Senie, Low body weight as a risk factor for hip fracture in both black and white women. J. Orthop. Res. 7, 192 – 197 (1989). 217. D. Hemenway, G. A. Colditz, W. C. Willet, M. J. Stampfer, and F. E. Speizer, Fractures and lifestyle: Effect of cigarette smoking, alcohol intake, and relative weight on the risk of hip and forearm fractures in middle-aged women. Am. J. Public Health 78, 1554 – 1558 (1988). 218. B. W. Alderman, N. S. Weiss, J. R. Daling, C. L. Ure, and J. H. Ballard, Reproductive history and postmenopausal riskofhip and forearm fracture. Am. J. Epidemiol. 124, 262, 267 (1986). 219. R. Wooton, E. Breyson, U. Elsasser, H. Freeman, J. R. Green, and R. Hesp, Risk factors for fractured neck of femur in the elderly. Age Ageing 11, 160 – 168 (1982). 220. D. C. Bauer, W. S. Browner, J. A. Cauley et al., Factors associated with appendicular bone mass in older women. Ann. Int. Med. 118, 657 – 665 (1993). 221. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. A. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Engl. J. Med. 332, 767 – 773 (1995).

CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 222. R. Nuti, G. Martini, Measurements of bone mineral density by DXA total body absorptiometry in different skeletal sites in postmenopausal osteoporosis. Bone 13, 173 – 178 (1992). 223. H. Rico, M. Revilla, L. F. Villa, E. R. Hernandez, and J. P. Fernandez, Crush fracture syndrome in senile osteoporosis: A nutritional consequence? J. Bone Miner. Res. 7, 317 – 319 (1992). 224. M. E. Ooms, P. Lips, A. Van Lingen, and H. A. Valkenburg, Determinants of bone mineral density and risk factors for osteoporosis in healthy elderly women. J. Bone Miner. Res. 8, 66 – 75 (1993). 225. S. Elmstahl, P. Gardsell, K. Ringsberg, and I. Sernbo, Body composition and its relation to bone mass and fractures in an urban and rural population. Aging Clin. Exp. Res. 5, 47 – 54 (1993). 226. C. Longcope, H. Pratt, S. H. Schneider, and S. E. Fineberg, Aromatization of androgens by muscle and adipose tissue in vivo. J. Clin. Endocrinol. Metab. 46, 146 – 152 (1978). 227. P. Siiteri, Adipose tissue as a source of hormones. Am. J. Clin. Nutr. 45, 277 – 282 (1987). 228. J. M. Grodin, P. K. Siiteri, and P. C. MacDonald, Source of estrogen production in postmenopausal women. J. Clin. Endocrinol. Metab. 36, 207 – 214 (1973). 229. P. C. MacDonald, C. D. Edman, D. L. Hemsell, J. C. Porter, and P. K. Siiteri, Effect of obesity on conversion of plasma androstenedione to estrone in postmenopausal women with endometrial cancer. Am. J. Obstet. Gynecol. 130, 448 – 455 (1978). 230. C. Longcope and S. Baker, Androgen and estrogen dynamics: Relationships with age, weight and menopausal status. J. Clin. Endocrinol. Metab. 76, 601 – 604 (1993). 231. M. A. Kirshner, E. Samojlik, M. Drejka, E. Szmal, G. Schneider, and N. H. Ertel, Androgen-estrogen metabolism in women with upper body vs. lower body obesity. J. Clin. Endocrinol. Metab. 70, 473 – 478 (1990). 232. E. Samojlik, M. A. Kirschner, D. Silber, G. Schneider, and N. H. Ertel, Elevated production and metabolic clearance rates of androgen in morbidity obese women. J. Clin. Endocrinol. Metab. 49, 949 – 954 (1984). 233. D. Goodman-Gruen, and E. Barrett-Connor, Total but not bioavailable testosterone is a predictor of central adiposity in postmenopausal women. Int. J. Obesity 19, 293 – 298 (1995). 234. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 235. C. Christiansen and R. Lindsay, Estrogens, bone loss and preservation. Osteopor. Int. 1, 7 – 13 (1987).

769 236. A. M. Frumar, D. R. Meldrum, F. Geola, I. M. Shamonki, I. V. Tataryn, L. J. Deftos, and H. L. Judd, Relationship of fasting urinary calcium to circulating estrogen and body weight in postmenopausal women. J. Clin. Endocrinol. Metab. 50, 70 – 75(1980). 237. C. Hassager and C. Christiansen, Influence of soft-tissue body composition on bone mass and metabolism. Bone 10, 415 – 419 (1989). 238. N. H. Bell, S. Epstein, A. Greene, J. Shary, M. J. Oexmann, and S. Shaw, Evidence for alteration of the vitamin D endocrine system in obese subjects. J. Clin. Invest. 76, 370 – 373(1985). 239. T. Anderson, P. McNair, N. Fogh-Andersen, T. T. Nielson, L. Hyldstrup, and I. Transbol, Increased parathyroid hormone as a consequence of changed complex binding of plasma calcium in morbid obesity. Metabolism 35, 147 – 151 (1986). 240. Y. Liel, E. Ulmer, J. Shary, B. W. Hollis, and N. Bell, Low circulating vitamin D in obesity. Calcif. Tissue Int. 43, 199 – 201 (1988). 241. S. L. Teitelbaum, J. D. Halverson, M. Bates, L. Wise, and J. G. Haddad, Abnormalities of circulating 25OH vitamin D after jejunal bypass for obesity: Evidence of an adaptive response. Ann. Intern. Med. 86, 289 – 293 (1977). 242. R. L. Atkinson, W. T. Dahms, G. A. Bray, and A. A. Schwartz, Parathyroid hormone levels in obesity: Effects of intestinal bypass surgery. Miner. Electrolyte Metab. 1, 315 – 320 (1978). 243. J. E. Compston, Vedis, J. E. Ledger, J. C. Gazet, and R. E. Pilkington, Vitamin D status and bone histomorphometry in gross obesity. Am. J. Clin. Nutr. 34, 2359 – 2363 (1981). 244. J. A. Nisker, G. L. Hammond, B. J. Davidson, A. M. Frumar, N. K. Takaki, H. L. Judd, and P. K. Siiteri, Serum sex-hormone binding globulin capacity and the percentage of free estradiol in postmenopausal women with and without endometrial carcinoma. Am. J. Obstet. Gynecol. 138, 637 – 641 (1980). 245. B. J. Davidson, J. C. Gambone, L. D. Lagasse, T. W. Castaldo, G. L. Hammond, P. K. Siiteri, and H. L. Judd, Free estradiol in postmenopausal women with and without endometrial cancer. J. Clin. Endocrinol. Metab. 52, 404 – 408 (1981). 246. R. Lindsay, D. M. Hart, and D. M. Clark, The minimum effective dose of estrogen for the prevention and treatment of postmenopausal bone loss. Obstet. Gynecol. 63, 759 – 763 (1984). 247. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 248. D. F. Williamson, Descriptive epidemiology of body weight and weight change in U.S. adults. Ann. Intern. Med. 119, 646 – 649 (1993).

771

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

CHAPTER 31

Effects of Tobacco and Alcohol Use on Bone EGO SEEMAN

Austin and Repatriation Medical Centre, University of Melbourne, Heidelberg, Melbourne 3084, Australia

IV. Conclusions and Questions References

I. Introduction II. Tobacco and Bone III. Alcohol and Bone

I. INTRODUCTION

contradictory data are to be understood. Individuals rarely, if ever, differ in their exposure by only one factor. Confounding by other important covariates such as age, weight, exercise, nutritional factors, and drug therapy must be measured and taken into account in analyses (see Chapter 20). Thus, a difference in BMD between a smoker and nonsmoker may be due to differences in lifestyle factors other than tobacco consumption. Likewise, a true difference attributable to tobacco consumption may be obscured by differences in other factors unless these are taken into account in the analysis. Small sample sizes, biological variation, and difficulties in quantitation of the exposure “dose” of the risk factor reduce the power of studies to detect small but important effects. A true association between smoking and fracture rates is likely to be difficult to demonstrate in small studies because fractures are uncommon annual events in individuals. Fractures occur at a peak incidence of about 3 to 4 per 100 per year in women over 75 to 80 years of age. In younger groups, the incidence is 1 to 2 per 1000 or 10,000 per year depending on the decades studied. Thus, even if 20 – 30% of the risk for fracture was attributable to tobacco use, this risk may remain undetectable unless the samples sizes are very large or the prevalence of exposure was high. This problem is compounded further when end points are

To warrant study from a public health point of view, risk factors for osteoporosis should occur commonly and be amenable to intervention safely and at low cost. Modifying a risk factor that is uncommon and has a small effect on bone mineral density (BMD) is unlikely to influence fracture risk in the individual or the public health problem of fractures. Modifying a risk factor that is common and has a large effect (such as estrogen deficiency or falls) may help the individual and the public health problem. Modifying a risk factor that is uncommon but has a large effect (such as corticosteroid treatment) may help the individual, but not the public health problem. Modifying a risk factor that is common but has a small effect may not help the individual greatly, but may help the public health problem of osteoporosis. Tobacco and alcohol use fulfill several of these criteria. They are used by many individuals during a large proportion of their lives. The effect may be small in the short term but may become clinically important when exposure is prolonged. These are risk factors that can be modified at little, or no, expense. The difficulties in quantifying the effects of risk factors such as tobacco and alcohol should be acknowledged if

OSTEOPOROSIS, SECOND EDITION VOLUME 1

771

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

772

EGO SEEMAN

difficult to define such as a vertebral “fracture” or “deformity.” Similarly, hip fractures are the result of a fall as well as bone fragility so that identifying and measuring the risk conferred by tobacco use are difficult when multiple risk factors for falls and bone fragility are present in patients with hip fractures. Moreover, ascertainment bias may occur when patients with fractures are interviewed after the fracture. Circulating levels of some hormones exhibit diurnal variation or pulsatility. Estrogen concentrations vary from day to day by 50 to 200%, raising questions about the interpretation of negative studies and the usefulness of single measurements. If exposure to tobacco reduces estrogen concentration by 20 to 30%, this will be difficult to detect with single measurements taken at different times of day in different people. Rates of bone loss are around 1 to 2% per year, similar to the coefficient of variation of the methods used to measure BMD. In other words, the effect of exposure may be important but small in relation to the biological variation of the measurement and the precision of the technique. Thus, differences in BMD between a smoker and a nonsmoker or between a consumer of alcohol or an abstainer may be due to differences in concomitant lifestyle factors rather than any direct effect of tobacco or alcohol on bone. These covariates must be measured and their effects taken into account before an association is identified or concluded not to exist. Few studies pay meticulous attention to these methodological issues. Even when they are taken into account, the studies document associations, but causality can only be inferred from the observed association.

II. TOBACCO AND BONE A. The Problem Tobacco use is the single greatest preventable cause of premature death in the United States. Nearly 1 in 4 women age 18 years and over in the United States was a smoker in 1991. Thus, about 22 million women in the United States currently smoke cigarettes. Ninety percent of smokers begin to do so before the age of 20 years. Twenty-seven percent of women with 12 or fewer years of schooling were smokers compared to 12.5% with 16 or more years of schooling. Among girls aged 17 to 18 years, 33% who dropped out of school were current smokers compared to 17% who were still in high school or had graduated. Among adolescent girls aged 12 to 17 years, the reported smoking in the past month increased from 8.7% in 1990 to 9.4% in 1992. Smoking-related disease accounted for about 150,000 deaths among U.S. women in 1988. Lung cancer now surpasses breast cancer as a leading cause of cancer death

among women, accounting for 22% of female cancer deaths compared to 18% for breast cancer. Other tobaccorelated diseases include cancers of the oral cavity, esophagus, larynx, bladder, and pancreas, heart disease, stroke, emphysema, and bronchitis. Women smokers are at increased risk for cervical cancer, early menopause, complications of the oral contraceptive pill, and unfavorable pregnancy outcomes. There are 3000 new users daily in the United States; many of these are teenagers [1 – 3].

B. Fractures 1. HIP FRACTURES In most, but not all, studies, tobacco use is associated with an increased risk for hip fractures in women and in men. In general, the relative risks (RR) for hip fracture associated with tobacco use are about 1.2 to 1.5 with confidence intervals (CI) that include, or almost include, unity. The serious methodological problems associated with studying hip fracture pathogenesis should be recognized. In many studies, the sample sizes are small, around 100 cases or fewer. Survivors are interviewed whereas subjects with dementia and other serious comorbidity are omitted. This is a concern because a substantial proportion of the hip fracture cases will not be assessed. Data obtained cannot be verified and may be inaccurate given that the patients are elderly. In virtually all studies, exposure is assessed after the fracture has occurred. This may introduce ascertainment bias. In many studies, smoking was assessed dichotomously (ever versus never), without adjustment for age or of other potential confounding factors. Baker [4] found no association between smoking and hip fractures among 189 cases and 95 community controls. Paganini-Hill and colleagues [5] reported a RR of 1.5 (95% CI 0.8 to 3) among 91 cases less than 80 years old and 182 community controls. Williams et al. [6] reported a RR of 2.2 (95% CI, 0.6, 8.2) in 160 cases and 567 community controls. Kreiger and colleagues [7] reported a RR of 1.3 (95% CI, 0.8, 2.1) in 98 cases and 884 hospital controls. Boyce [8] reported a RR of 1.2 among 139 cases and 139 controls [8], whereas Alderman et al. [9] reported a RR of 1.2 (95% CI, 0.9, 1.6) in 160 cases aged 50 to 74 years and 567 controls. Cooper and Wickham [10] found that the RR for hip fracture in women was 1.7 (95% CI, 1.2, 2.4). This remained significant after adjusting for alcohol, body mass, activity, and calcium intake. The study was based on 240 cases (100 “ever,” 140 “never” smokers) compared with 480 controls. Among 60 men with hip fractures (45 “ever,” 15 “never” smokers, and 120 controls) the RR was 1.5 (95% CI, 0.7, 3.1). After adjusting for alcohol intake the RR was 0.7 (95% CI 0.3, 1.8). Johnell and colleagues [11] studied risk factors for hip fracture in 2086 women (mean age 78 years) with hip

773

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

fracture and 3532 controls from six countries. No detectable effect of tobacco use was identified. Alcohol consumption was protective in young adulthood. On multivariate analysis, late menarche, poor mental score, low body mass index (BMI) and physical activity, low exposure to sunlight, and low consumption of calcium and tea were independent risk factors, accounting for 70% of hip fractures. Cornuz and colleagues [12] followed 116,229 women aged 34 to 59 years for 12 years. At baseline, 31% were current smokers and 26% were current smokers and 26% former smokers; after 12 years, another 11% were former smokers. During follow-up there were 377 hip fractures, with an average age of 60 years. Risk increased linearly with cigarette consumption (RR 1.6; CI 1.1, 2.3). The age adjusted RR of hip fracture in current versus never smokers was 1.3 (95% CI 1.0, 1.7) and remained elevated, but not significantly so, after adjusting for menopausal status, use of estrogen, activity, and intakes of calcium, alcohol caffeine, and BMI. Melhus and colleagues [13] reported that an insufficient dietary intake of vitamins E and C may increase the risk of hip fracture in smokers. In 205 women aged 40 to 76 years who sustained a hip fracture (44 current and 161 never smokers) and in 746 controls (93 current and 653 never smokers), after adjustment for risk factors, the odds ratio (OR) for hip fracture among current smokers with a low vitamin E intake (below the median in controls) was 3.0 (95% CI 1.6, 5.4) and with a high vitamin E intake was 1.1 (CI 0.5, 2.4). Equivalent ORs for low and high vitamin C intakes were 3.0 (CI 1.6, 5.6) and 1.4 (CI 0.7, 3.0). The OR in current smokers with low intakes of both vitamin E and C was 4.9 (CI 2.2, 11.0). The influence of vitamin E and C intake was less pronounced in former than in current smokers. Thus, data support an association between tobacco use and hip fractures but confidence intervals often cross unity. There may be an interaction involving tobacco use, body weight, estrogen exposure, and fracture risk. Tobacco use may negate the protective effect of estrogen replacement therapy against hip fracture. Kiel et al. [14] observed 207 hip fractures among 34,700 women. Smoking was not associated with an increased risk of hip fracture (RR 1.22, 95% CI 0.76, 1.95). Data suggest that estrogen use was protective in nonsmokers (RR 0.37, CI 0.19, 0.75) but not in current smokers (RR 1.26, CI 0.29, 5.45). Forsen and colleagues [15] found that 421 hip fractures occurred during 1986 – 1989 among 34,856 adults over 50 years of age who had attended health screening between 1984 and 1986 (91% of the eligible population). For women who smoked, the RR of hip fracture was 1.5 (95% CI, 1.0, 2.4), 3.0 (95% CI, 1.8, 5.0) in thinner women (BMI  20kg/m2) and 1.8 (95% CI, 1.2, 2.9) in men (independent of body mass). Lack of physical activity increased the RR of hip fracture to 1.4 in women and 2.3 in men after adjustment for ill health.

Williams and colleagues [6] studied 353 women aged 50 – 74 with hip or forearm fractures and 567 communitybased controls. The RR of hip fracture was increased in thin women who smoked, particularly among nonusers of estrogen. The RR of hip fracture in women who were thin, never smoked, and did not use estrogens was 13.5 (95% CI, 2, 35.5) and 6.4 (95% CI, 2.1, 19.4) in those who had used estrogens for more than 1 year. Women who were obese were not at increased risk for hip fracture whether they smoked or not. The benefit of using estrogens in preventing hip fractures was greatest in the thin smoker. In contrast, Hemenway et al. [16] studied 96,508 middle-aged nurses 35 – 59 years of age. Hip or forearm fractures occurred in 925 of the women. No association with the risk for hip fracture and smoking was detected, perhaps because of the relatively young age of this cohort. Kreiger and colleagues [7] reported that of 98 hip fractures and 884 hospital controls, tobacco use was associated with an age adjusted RR of 1.27 (95% CI, 0.63, 2.56), an estimate that included unity, perhaps because of the small sample size. Law and colleagues [17] conducted a meta-analysis of 29 published cross-sectional studies including 2156 smokers and 9705 nonsmokers and 19 cohort and case control studies recording 3889 hip fractures. Risk of hip fracture was 17% greater at 60 years, 41% greater at 70 years, 71% greater at 80 years, and 108% greater at 90 years. An estimated 19% of current smokers and 12% nonsmokers would have a hip fracture by 85 years of age, and 37% of current smokers and 22% of nonsmokers by the age of 90 years. One hip fracture in eight was attributable to smoking in women (Fig. 1). 2. SPINE, FOREARM, AND OTHER FRACTURES Daniell [18] reported that 76% of 38 women with one or more vertebral fractures smoked at least 10 cigarettes per day for 5 or more years compared with 43% of 572 controls. Aloia et al. [19] reported that 58 postmenopausal women with crush fractures smoked twice as much as 58 age-matched controls. Seeman and colleagues [20] reported an odds ratio of 2.3 in men who were current smokers. Smoking, body weight, alcohol, and underlying comorbidity were independent predictors of vertebral fracture risk. The risk associated with smoking increased by 1.009 for each pack year of exposure. Seventy-nine percent of the cases and 63% of controls smoked (P  0.009), whereas 82% of cases and 70% of controls drank alcohol (P  0.02). The risk of osteoporosis increased by 1.007 for each ounce year of cumulative exposure (P  0.01). The relationship among these factors is shown in Fig. 2. In those with no underlying disease, the relative risk was 0.3 (Fig. 2) (hatched bar on the left), reaching 30 in an individual who drank alcohol, smoked, was not obese, and had an underlying illness.

774

EGO SEEMAN

FIGURE 1 Relative risk (95% confidence intervals) of hip fracture in smokers compared to nonsmokers in postmenopausal women according to age in cohort studies and case-control studies (open circles). Reproduced from Law and Hackshaw [17], with permission from the authors and publishers. The risk conferred by exposure to tobacco, alcohol, and underlying disease and the protective effect of obesity was examined in a multiple logistic model. The extent of the elevation in risk was age dependent. The effects of smoking and alcohol were not evident in persons under 60 years old, emerged in 60- to 69-year olds, and were strongest in those 70 years and over. In nonobese individuals over 70 years of age who drank alcohol, smoked, but had no dis-

FIGURE 2

ease, the relative risk was 20.2 (P  0.05). This risk was reduced to 6.9 in the presence of obesity. In a nonobese individual who drank, smoked, and had an underlying illness, the relative risk for vertebral fractures was 192.5 (P  0.05). In the study by Williams et al. [6] of women aged 50 – 74 with forearm fractures and 567 community-based controls, nonestrogen users had a RR for forearm fractures

Relative risk of spinal osteoporosis with vertebral fractures in men 60 to 69 years old with various combinations of tobacco use, alcohol use, obesity, and presence of underlying medical illness. From Seeman and colleagues [20]. Reproduced with permission from the authors and publisher.

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

of 5.4 (95% CI, 2.5, 11.3) if they were thin and ever smoked and 0.7 (95% CI, 0.2, 2.5) if they had used estrogens for more than 1 year. In estrogen users, there was no relationship between the risk of forearm fractures and a history of smoking and body weight. The benefit of using estrogens in preventing forearm fractures was greatest in the thin smoker. In the study by Hemenway et al. [16], no association between risk for forearm fracture and smoking was detected. Jensen [21] studied 285 women 70 years of age, of whom 77 were smokers and 103 were nonsmokers. A history of fractures was obtained in 40.3% of the smokers and 44.7% of the nonsmokers. The authors concluded that osteoporosis was of the slender, rather than the slender smoker, and that smokers have a reduced BMD appropriate for their slenderness. Thus, most studies suggest that tobacco use increases the risk for fracture, particularly in thinner individuals. This risk appears to emerge in advanced age and may reduce the protective effect of obesity and estrogen exposure for fracture, whereas higher body weight may protect from the effects of tobacco.

C. Bone Mineral Density The increased risk for fractures associated with tobacco use is likely to be partly conferred by a reduction in BMD. This effect may even be conferred before birth. Jones and colleagues [22] studied maternal smoking habits related to bone mass and growth in 330 children aged 8 years. For children born at term, smoking during pregnancy was associated with a lower height (1.53 cm; 95% CI 3.03, 0.03) and a trend to lower weight (1.35 kg; CI 2.75, 0.11). After adjustment for body size, BMD was lower at the lumbar spine (0.019 g/cm2; CI 0.033, 0.005) and femoral neck (0.018 g/cm2; CI 0.034, 0.002) but not total body (0.005 g/cm2; CI 0.015, 0.005) in the maternal smoking group. Placental weight was lower in smoking mothers (56 g; CI 95, 17), and adjustment for placental weight removed the effects of smoking on BMD and growth. In children born prior to 37 weeks, maternal smoking was not associated with deficits in BMD. The influence of smoking was not altered by adjustment for sports activity, dietary calcium intake, or sunlight exposure. Children who were breast-fed and had nonsmoking mothers had higher BMD at all sites (3.3 to 7.7%) than children who were not breast-fed and whose mothers smoked. The authors concluded that maternal smoking during pregnancy is associated with growth and bone mass in children at age 8 years. These associations may be mediated through placental size and function. Thus, the skeleton of offspring of smokers may be affected in early development.

775 Because tobacco users often start smoking by 10 to 12 years of age, the reduced peak BMD may contribute to any deficit in BMD in adulthood. Even though over 90% of peak BMD is achieved by about 15 to 18 years of age, exposure to tobacco may have important effects, as mineral accrual is very rapid, particularly in the peripubertal period, and if interrupted may result in biologically important deficits in BMD. No studies have been done to examine this possibility. Studies in children will be difficult to perform and interpret because information regarding exposure from children may be unreliable. In addition, socioeconomic factors must be taken into consideration when these studies are designed, as children from lower socioeconomic groups are more likely to take up smoking and may have lower BMD at the time that tobacco exposure begins. In addition, reduced bone size may contribute to any deficit in BMD because this measurement of “density” only partly takes bone size into account. There are no data available addressing the question of the effects of tobacco on stature and bone size in children or adults. Few prospective studies have examined the association between tobacco use and rates of bone loss. Krall and Dawson-Hughes [23] found that bone loss (percent per year) from the radius was greater in 34 smokers than in 278 nonsmokers ( 0.91  2.6 vs 0.004  2.57, P  0.05). Similar trends were found at the femoral neck (0.58  2.9 vs 0.16  2.84), oscalcis (1.42  3.32 vs 1.19  3.26), and spine (1.31  2.19 vs 0.97  2.16). More recently, Krall et al. [24] measured rates of change in BMD over 3 years in 402 men and women aged 65 years or above. One half of participants received daily calcium (500 mg elemental calcium) and cholecalciferol (700 IU) supplements and the remainder received placebo. The annualized rate of change in BMD was higher in the 32 smokers than in the nonsmokers at the femoral neck (0.71  0.29 vs 0.04  0.08%, P  0.02) and total body (0.36  0.10 vs 0.15  0.03%, P  0.05). Mean calcium absorption was lower in smokers after adjusting for gender, age, supplementation status, and dietary calcium and vitamin D intakes (12.9  0.8 vs 14.6  0.2%, P  0.05). Lowest absorption was in smokers of at least 20 cigarettes/day (12.1  1.1%). In the subgroup receiving calcium and cholecalciferol, urinary calcium/creatinine excretion was lower in smokers than in nonsmokers (44  12 vs 79  9%, P  0.05). In contrast, Slemenda et al. studied 84 peri- and postmenopausal women and found no association between rates of bone loss and tobacco consumption. However, there were 13 subjects who smoked over 20 pack years and 8 who smoked under 20 pack years. Increased rates of bone loss, determined by changes in metacarpal morphometry, have been reported in men [26]. Most studies have been cross-sectional and of these, most, but not all, report an association between tobacco exposure and reduced BMD. In general, no deficits in BMD

776

EGO SEEMAN

FIGURE 3 Differences (95% confidence intervals) as a proportion of 1 SD in BMD between female smokers and nonsmokers according to age and menopausal status. Open circles refer to two studies and solid circles refer to other studies cited. From Law and Hackshaw [17], reproduced with permission from the authors and publisher. are observed in smokers compared to premenopausal and perimenopausal “never” smokers [27 – 35]. As shown in Fig. 3, in the study by Law and Hackshaw [17], premenopausal smokers and nonsmokers had similar BMD and postmenopausal bone loss was greater in current smokers, diminishing 2% more for every 10 years increase in age with a difference of 6% at 80 years of age. By the age of 80 years, BMD was 0.45 SD lower in smokers than in nonsmokers. The estimate in men was 0.32 SD lower BMD in smokers relative to nonsmokers. The association was independent of body weight. Hansen [36] studied 249 healthy premenopausal women 39  6 years of age and found no association between BMD and tobacco use. Stevenson and colleagues [37] reported an association between BMD and tobacco use at the spine but not at the proximal femur, where McCulloch et al. reported lower BMD in heavy smokers. McDermott and Witte [39] reported no difference in the mid- and distal radius BMD in 35 smokers and 35 nonsmokers. Grainge et al. reported a negative correlation between BMD and months of smoking in 580 postmenopausal women aged 45 to 59 years. However, smoking duration accounted for only 1% of variance in BMD. May and colleagues [41] reported no association between past or present smoking and BMD among 453 men ages 65 to 76 years, after adjusting for age and weight. Previous analyses of women from the same population showed a dose – response relation between smoking and BMD. In a study of 186 women and 224 men aged 61 to 73 years, Egger et al. reported current male smokers had 7.3% lower lumbar spinal BMD than never smokers (95% CI 0.4, 14.2). For women, BMD was 7.7%

lower (CI 0.3, 15.6). Each decade of smoking reduced lumbar spinal BMD by 0.015 g/cm2. Smaller effects of smoking on BMD were observed at the femoral neck. Thus, deficits of about 0.5 to 1.0 SD are associated with prolonged exposure in postmenopausal smokers and in men. This is in accord with finding an increased relative risk for fractures in advanced age. Jones and Scott [43] compared 118 current smokers and 158 nonsmokers (mean ages 33 and 34 years, respectively). Smokers had lower BMD (femoral neck, 0.32 SD; 95% CI  0.60,  0.04; lumbar spine,  0.49 SD; CI  0.76,  0.22; and total body,  0.40 SD; CI  0.66,  0.14) among women with BMI  25 kg/m2, in whom there was a dose – response relation between cigarettes smoked and BMD. Relative to current smokers, women who ceased smoking between 1988 and 1996 (n  27) had higher BMI (by 2.5 kg/m2), body weight (7.2 kg), and lumbar BMD (0.52 SD). Smokers who had breast-fed at least one child had an additional deficit in BMD (femoral neck,  0.48 SD; CI  0.89,  0.07; lumbar spine,  0.39 SD; CI  0.80, 0.02; and total body,  0.37 SD; CI  0.77, 0.06). Smokers who participated in competitive sport had increments in bone mass (0.74 SD; CI 0.31, 1.17; 0.48 SD; CI 0.03, 0.93; and 0.42 SD; CI 0.00, 0.84, at respective sites). The cotwin control model overcomes several of the difficulties discussed in Section I by controlling for age, sex, and genetic composition, which are major determinants of BMD. To examine the effects of cigarette smoking on BMD, Hopper and Seeman [44] studied 41 female twin pairs (21 monozygotic), ages 27 to 73 years (mean 49), discordant for at least 5 pack years of smoking (mean 23,

777

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

FIGURE 4

Bone density (g/cm2) at the lumbar spine in the greater smoker and lesser smoker as a function of age. From E. Seeman, unpublished data.

maximum 64). Each twin was classified as either the “greater” or “lesser” smoker of a pair. The mean difference in smoking was 23 – 15 pack years (range 5 to 64). The utility of this model is illustrated in Fig. 4; BMD in the greater smoker (closed symbol) and lesser smoker (open symbol) diminish as age advances, but the difference in BMD between them is not readily apparent because genetic and other environmental factors also influence an individual’s BMD. As shown in Fig. 5, when the within-pair difference in BMD between the greater and the lesser smoking twin is plotted as a function of increasing discordancy in pack years smoking, the effect of smoking is apparent. As the difference in pack years between greater and lesser smoker increases, the BMD deficit increases so that after about 20 pack years of discordancy, few data points are above the x axis. The BMD deficit in the smoker is about 5 to 10%, or over a half a standard deviation, with over 20 pack years of difference in tobacco exposure. For the 20 most discordant pairs, BMD was lower in the greater smoking twin at the lumbar spine by 9.3 to 3.1% (P  0.008), at the femoral neck by 5.8 to 2.9% (P  0.06), and at the femoral shaft by 6.5 to 3.2% (P  0.05). Across all 41 twin pairs, for each 10 pack years of smoking, the deficit in BMD increased by 2% at the lumbar spine and by about 1% at the femoral sites. Women who smoke one pack of cigarettes each day through adult life will, by menopause, have an average deficit in BMD of 0.5 to 0.8 of a population standard deviation or a 5 to 8% deficit, an amount sufficient to increase their risk for fracture. In vitro, bone strength decreases threefold with a 10% decrease in mineral content. A 10% difference in BMD is equivalent to almost one population standard deviation, 10 years of age-related bone loss, and about one-half the diminution found in postmenopausal women relative to premenopausal women. A decrease of

FIGURE 5

Within-pair difference (greater minus less smoker) in bone density expressed as a percentage of the pair mean and plotted as a function of pack years discordancy at the lumbar spine, femoral neck, and femoral shaft. Monozygotic pairs are represented by a closed symbol and dizygotic pairs by an open symbol. From J. L. Hopper and E. Seeman [44]. Reproduced with permission from the authors and publisher.

10% over 10 years may confer a 44% increase in hip fracture risk. The greater smoker consumed more coffee (P  0.02), and among the 20 most discordant pairs, the twin who smoked more walked less (P  0.004). Of the 16 postmenopausal pairs, menopausal age was 1.5 – 1.2 years earlier in the greater smoker (P  0.20). Allowing for life style factors resulted in only minor changes in the strengths of the association between BMD and smoking. At the lumbar spine, the difference in BMD per 10 pack years decreased from 2.0 to 1.7% after adjusting for a positive association between that difference in BMD and the difference on the walking scale (P  0.03). In the 20 most discordant

778

EGO SEEMAN

pairs, after adjusting for discordancy in estrogen replacement therapy, the difference in BMD increased from 9.3 to 10.4% (P  0.004). Further multivariate analyses that allowed for differences in other measured life style factors, including alcohol and coffee consumption, did not influence the strength of association with smoking. The relationships between deficits in BMD at all sites and difference in pack years smoking were not eliminated by allowing for biochemical factors. Syversen and associates [45] reported that female rats aged 2 months exposed to nicotine vapor (  100 ng/ml) for 20 days, 5 days/week for 2 years had 10% lower body weight than controls throughout the study but no differences in femoral BMC or BMD; femoral length; ultimate bending moment, ultimate energy absorption, stiffness, or deflection in the femoral shaft and femoral neck; or midshaft cortical area and stress.

D. Bone Loss: Increased Bone Resorption Bone loss occurs if there is an imbalance between the amount of bone resorbed and the amount of bone formed. The evidence available examining whether one or both of these mechanisms contributes to the bone loss associated with smoking is limited. De Vernejoul and colleagues [46] reported histomorphometric changes in 11 men ages 35 – 50 years with idiopathic osteoporosis who were mild alcoholics and heavy smokers. In patients vs controls, trabecular bone volume was 15.5  3.4% vs 24.3  5.4% (mean  SD). Mean wall thickness was 50.6  6.7 mm vs 59.9  5.6 mm, and mean trabecular plate thickness was 159  27.4 mm vs 193  37.2 mm. There was no evidence of osteomalacia. Bone resorption was not increased relative to controls. These data are consistent with reduced bone formation accounting for the deficit in bone volume. Hansen [36] studied 249 healthy premenopausal women 39  6 years of age and found that serum osteocalcin was reduced in smokers compared to nonsmokers (P  0.01). Leino et al. studied 519 women and found with multivariate analysis that serum osteocalcin was associated with alcohol consumption but not tobacco use; alkaline phosphatase and total ionized calcium were associated with cigarette smoking. In the cotwin control study [44], while no evidence for reduced bone formation was detected, there was evidence for increased bone resorption. Serum calcium concentrations were higher in the greater smoker in 17 of the 26 pairs (P  0.08). For the 11 most discordant pairs, serum phosphate and serum alkaline phosphatase activity were higher in the greater smoker (P  0.05, P  0.006, respectively). Pair differences in serum calcium were positively associated with pair differences in urinary hydroxyproline excre-

tion (r  0.8, P  0.001) and urine pyridinoline to creatinine ratio (measured in 17 pairs) (r  0.8, P  0.001). At the lumbar spine, the lower BMD in the greater smoker was associated with higher serum calcium and urine pyridinoline values, consistent with increased bone resorption. As the within-pair smoking difference increased, the differences in serum parathyroid hormone concentrations decreased by 5.0  2.5% for each 10 pack years (P  0.05). Pair difference in serum parathyroid hormone was negatively associated with pair difference in BMD at the lumbar spine (r   0.60; P  0.005), femoral neck (r   0.75; P  0.001), and femoral shaft (r   0.53; P  0.01). Allowing for pack years of smoking and serum parathyroid hormone, which together accounted for 70% of the variation in pair differences in BMD at the lumbar spine, the deficit in spinal BMD increased with pair difference in serum calcium (P  0.05) and, in the 17 pairs measured, with the pair difference in urine pyridinoline (P  0.06). Smoking was associated with higher serum folliclestimulating hormone (P  0.02) and luteinizing hormone (P  0.03) concentrations and lower serum parathyroid hormone concentrations (P  0.05). Several of these relationships are shown in Fig. 6.

E. Bone Loss: Decreased Bone Formation Reduced bone formation may, in part, be due to a reduction in osteoblast numbers. Fang et al. [48] colleagues examined the effects of nicotine on UMR (University, Melbourne, Repatriation) 106-01 rat osteosarcoma cell proliferation and activity. Nicotine produced a dose-dependent suppression of thymidine incorporation with maximum suppression seen at 10 mM. One micromolar nicotine decreased cell number by 20 – 30%. There was a dose-dependent increase in alkaline phosphatase activity with a maximum stimulation of 189% relative to control at 1mM, suggesting that nicotine may promote the differentiation of osteoblasts. In contrast, Lenz and colleagues [49] reported that nicotine stimulated DNA synthesis in osteoblast-like cells from embryonic chick calvaria and inhibited collagen synthesis and alkaline phosphatase activity. Moreover, Galvin et al. [50] reported that smokeless tobacco extract did not affect thymidine incorporation and inhibited collagen synthesis and alkaline phosphatase activity in chick embryo calvarial cells. These disparate observations may be related to species, cell culture conditions, or the type of osteoblast model used.

F. Effects of Tobacco on Estrogen Metabolism The increased bone resorption is, in part, likely to be mediated by a reduction in circulating estrogen in women. This reduction in circulating estrogen is probably due to

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

FIGURE 6

Within-pair difference (greater minus lesser smoker) in serum-luteinizing hormone, follicle-stimulating hormone, parathyroid hormone, and calcium expressed as a percentage of the pair mean and plotted as a function of pack years discordancy. Monozygotic pairs are represented by a closed symbol and dizygotic pairs by an open symbol. From J. L. Hopper and E. Seeman [44]. Reproduced with permission from the authors and publisher.

decreased production and increased degradation of circulating estrogen. The inferences are based mainly on crosssectional studies. Postmenopausal smokers have lower estrogen concentration than nonsmokers. Smoking is associated with an earlier menopause. Pregnant women who smoke have lower estrogen levels than nonsmokers. McMahon and col-

779 leagues [51] showed that premenopausal women who smoke have lower luteal phase excretion of estrone, estradiol, and estriol than nonsmokers. There were no differences detected in the follicular phase in premenopausal women or in any measures in postmenopausal women. Barbieri and colleagues [52] examined the effects of constituents of tobacco on estrogen production by human choriocarcinoma cells and placental microsomes. Nicotine, cotinine, and anabasine inhibited the conversion of androstenedione to estrogen in a dose-dependent fashion. The inhibition of aromatase was reversible and competitive as supraphysiological doses of androstenedione block the inhibition of aromatase by these tobacco products. Figure 7 shows the effects of nicotine, cotinine, and anabasine on estrogen and progesterone accumulation in choriocarcinoma cell cultures. The inhibition of estrogen production was specific. There was no change in progesterone accumulation. Figure 8 shows that nicotine inhibition of aromatase activity in human placental microsomes was competitive. The mechanism probably reflects inhibition of the aromatase enzyme system. It is likely that nicotine interacts directly with placental microsomal cytochrome P450 and may reversibly alter the function of active sites of the cytochrome P450 component of the aromatase enzyme system. Aminoglutethamide inhibits that aromatization of testosterone and shares structural similarities to nicotine. Increased degradation of endogenous and exogenous estrogens may contribute to the lower levels of circulating estrogen. Estradiol is reversibly oxidized to estrone, which is then irreversibly hydroxylated in the C-2 position to form 2-hydroxyestrone and 2-methoxestrone or is hydroxylated in the 16 position to form 16-hydroxyestrone and -estriol. Michnovicz and colleagues [53,54] studied 14 premenopausal women who smoked 15 to 30 cigarettes per day and 13 premenopausal nonsmokers aged 21 – 44 years. Increased 2-hydroxylation of estrogens was observed. Total estrogen excretion did not differ in smokers and nonsmokers; however, the metabolite 2-hydroxyestrone constituted 53.6  2.2% vs 35.1  1.8% of the total estrogen excretion (P  0.02). As shown in Fig. 9, the extent of 2-hydroxylation was increased. The metabolite 2-hydroxyestrone was elevated in smokers (17.2  2.4 mg/g vs 9.4  1.2 mg/g creatinine, P  0.02) and there was a reduction in urinary estriol in smokers (11.9  1.2 mg/g vs 35.2  10.7 mg/g creatinine). Jensen and colleagues [55] studied 136 postmenopausal women. Of these, 63 received 4 mg estradiol, 42 received 2 mg, 23 received 1 mg, and 23 received placebo. Serum estradiol did not differ at baseline. Figure 10 shows that both circulating estrone and estradiol were lower in smokers than in nonsmokers, being half that of nonsmokers in the high dose group. In the 114 women receiving hormone therapy, there was an inverse relationship between the number of cigarettes smoked and the change in estrone (r  0.34, P  0.001)

780

EGO SEEMAN

FIGURE 7

Nicotine, cotinine, and anabasine reduced estrogen accumulation but not progesterone accumulation in choriocarcinoma cell cultures. From Barbieri and colleagues [52]. Reproduced with permission from the authors and publisher.

and estradiol (r  0.31, P  0.001). Body weight was 59.2 kg (range 39.5 to 86.5) in smokers and 67.3 kg (range 46 to 110) in nonsmokers. Whether this may have confounded the observations is uncertain. Data are consistent with increased metabolism of the administered estrogen. Cassidenti and colleagues [56] studied the baseline estrogen status of 13 postmenopausal smokers and 12 nonsmokers. There was no difference in baseline measurements of estrone, estradiol, estrone glucuronide, or sulfate. The unbound estradiol concentration was lower in smokers than in nonsmokers 8 h after the administration of micronized estradiol (13.3  1.4 pg/ml vs 25.9  4.8 pg/ml, P  0.02) with the higher dose (2 mg) of micronized estradiol. The same pattern was present with lower doses in the smokers at 4 and 8 h after administration of estrogen. Sex hormone-binding globulin was higher in smokers than in nonsmokers with the 1-mg (136.7  8.2 nmol/liter vs 61.6  8.3 mol/liter, P  0.0001). Serum estrone sulfate levels were higher smokers receiving the larger dose of estrogen. Data are compatible with tobacco increasing the hepatic synthesis of sex hormone binding globulin. Not all investigators document an association between tobacco consumption and estrogen. Cauley and colleagues studied 176 white postmenopausal women mean age 58 years. These investigators found the degree of obesity to be a major determinant of estrone and estradiol. The estrone was 40% higher than in nonobese, Smokers had higher androstenedione than nonsmokers estrone (82.1.  52.3 vs 65.9  42.1, mean  SD, P  0.11), but there was little

FIGURE 8

Nicotine inhibition of aromatase activity in human placental microsomes using testosterone substrate concentrations of 4.9, 7.4 and 12.5. From Barbieri and colleagues [52]. Reproduced with permission from the authors and publisher.

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

781

FIGURE 9

The extent of estradiol 2-hydroxylation as determined by a radiometric assay in 14 premenopausal female smokers and 13 nonsmoking controls. Bars indicate means  ISEM. From Michnovicz and colleagues [54]. Reproduced with permission from the authors and publisher.

difference in serum estrogens between the 26 smokers and 150 nonsmokers. The small sample size may have limited the power of the study to detect a difference. Crawford et al., found 33% lower ethinyl estradiol levels in smokers versus nonsmokers 16 – 19 h after ingestion of that drug. This difference did not reach statistical significance. Geilser and colleagues, reported that plasma estrone, estradiol, and estrone sulphate were 40 – 70% lower in smokers when estrogen replacement therapy (ERT) was given orally but not transdermaly. Most results are consistent with the notion that tobacco use reduces circulating estrogens, leading to increased serum concentrations of follicle-stimulating hormone and luteinizing hormone and to increased bone resorption. The latter results in increased circulating calcium and a subsequent lowering of serum parathyroid hormone concentration and in increased urine hydroxyproline and prydinoline excretion. There is independent support for an effect of smoking on the production and degradation of estrogens.

G. Effects of Tobacco on Testosterone and Other Steroids Testosterone has been reported to be elevated, unchanged, or decreased in men who smoke. For example, in a population-based study of 590 men ages 30 to 79

FIGURE 10 Serum concentrations of estrone (E1) and estradiol (E2) in smokers and nonsmokers after treatment with low, medium, and high doses of estradiol. Values are means  ISEM. From Jensen and colleagues [55]. Reproduced with permission from the authors and publisher.

years, serum testosterone was modestly increased in current smokers. Current smokers also had higher endogenous androstenedione, estrone, and estradiol levels than nonsmokers [60,61]. The increased estrogens were explained by neither adiposity nor alcohol consumption. In contrast, Handelsman et al. [62] studied 71 nonsmokers and 23 smokers ages 16 to 47 years and found no difference in plasma testosterone, follicle-stimulating hormone (FSH), leuteinizing hormone (LH), or prolactin concentration. Total sperm output, sperm motility, total motile sperm, and oval sperm density were reduced relative to nonsmokers. Briggs [63] reported changes in plasma testosterone after 7 days of abstinence in men smoking 30 cigarettes per day. Smokers showed a rise of 1.65  0.5 ng/ml after 7 days of abstinence.There was no change in plasma testosterone in controls. Briggs suggested that carbon monoxide may inhibit testosterone formation due to the blockade of 17 hydroxylation of progesterone. The enzyme catalyzing

782 this reaction, NADP: pregnane-17-oxo-reductase, located in the microsomal fraction, requires a cofactor related to cytochrome P450. Carbon monoxide inhibits microsomal cytochrome P450. Meikle and colleagues [64] studied the effects of nicotine and cotinine on testosterone metabolism. Both competitively inhibit 3-hydroxysteroid dehydrogenase, an enzyme that converts dihydrotestosterone to 3-androstanediol in dog prostate microsomes. Microsomal fractions incubated for 1 h with nicotine and cotinine resulted in elevated dishydrotestosterone and suppressed 3-androstanediol. (The 5reductase activity is unaffected by these metabolites of tobacco.) The concentrations used in these experiments were similar to those achieved in humans who smoke. Elevated testosterone levels have also been reported in women smokers. Friedman and colleagues [35] studied 9 postmenopausal smokers and 16 nonsmokers and found elevated testosterone (360  111.7 pg/ml versus 244.9  128.9 pg/ml, mean  SD, P  0.05) and nonsignificantly higher dihydrotestosterone (135.6  81.9 pg/ml versus 88.4  41.2 pg/ml). Progesterone levels were elevated (107.8  59.9 pg/ml versus 65.1  34.8 pg/ml, P 0.05) as were those of 17-hydroxyprogesterone (708.4  418.7 pg/ml versus 197.8  74.1 pg/ml, P  0.0005) and androstenedione (844.8  503.6 pg/ml versus 486.8  287.8 pg/ml, P  0.05). The differences in dehydroepiandrostenedione (DHEA) sulfate failed to reach statistical significance (1622.2  1503.02 pg/ml versus 864.6  428.4pg/ml). Serum cortisol was higher (169.2  35.2 ng/ml versus 115.3  27.6 ng/ml. P  0.0001). In contrast, Cauley et al. [57] found no relationship between testosterone and cigarette smoking. Smokers had higher androstenedione levels (82.1  52.3 versus 65.9  42.1, P  0.11) (mean  SD) than nonsmokers. This was independent of obesity, alcohol consumption, and other confounder. Higher circulating progesterone may protect from osteoporosis whereas hypercortisolism may increase the risk for osteoporosis in women who smoke. Adrenal cortisol and androgens increase due to the stimulation of ACTH release. Nicotine inhibits 11-hydroxylase and 21hydroxylase, causing the increase in DHEA and androstenedione. Higher androstenedione may also be the result of aromatase inhibition. Inhibition of desmolase also reduces estrogen and progestational hormones. 1. CESSATION OF SMOKING There are scant data concerning the effects of cessations of tobacco use on the skeleton. In the study by Cornuz and colleagues [12], the risk of fracture in former smokers did not decrease until 10 years after quitting: at  5 years, RR 1.1 (CI 0.8, 1.7); at 5 – 9 years, RR 1.1 (CI 0.7, 1.7): and at 10 years, RR 0.7 (CI 0.5, 0.9). Adjustment for BMI reduced the benefit of quitting smoking at 10 years (RR 0.8; CI 0.5, 1.0).

EGO SEEMAN

H. Summary Tobacco is used by many individuals during a large proportion of their lives and apears to be associated with an increased risk of fracture of the axial and appendicular skeleton in women and in men. This risk emerges in advanced age and reduces the skeletal protective effects of obesity and estrogen exposure. The effect is mediated, in part, by a reduction in BMD, which is probably due primarily to increased bone loss. Decreased bone formation and increased bone resorption are responsible for bone loss. Increased bone resorption associated with smoking is, in part, due to a reduction in the production and an acceleration in the degradation of estrogen. The mechanism of reduced bone formation is uncertain. The effects of tobacco use on BMD are usually undetectable until late in adulthood because smoking one pack per day on average results in a deficit in spinal BMD of 2% per decade. However, a deficit of 0.5 to 0.8 standardized deviations directly attributable to tobacco use may be incurred over the three decades from age 20 to 50 years, a change that may double fracture risk.

III. ALCOHOL AND BONE A. Fractures 1. ALCOHOL ABUSE Alcohol abuse appears to confer a high risk for fracture in women and men. When present, it is commonly associated with fractures of the axial and appendicular skeleton. However, in women, only a small proportion of fractures are attributable to alcohol abuse. In contrast, alcohol abuse should be strongly suspected in men presenting with fractures. Johnell and colleagues [65] found that 25% of all men and 37% of men over 30 years of age admitted to the hospital for lower extremity fractures had a history of alcohol abuse, [65]. This was found in only 4% of women admitted for lower limb fractures. Most of the studies of fracture prevalence in alcoholic men are based on small series, many of which are uncontrolled. Nevertheless, the majority suggest an unusually high prevalence of fractures, particularly given the age distribution of subjects. De Vernejoul and colleagues [46] reported that 11 men with osteoporosis had a history of heavy alcohol and tobacco use. Bikle [66] reported that 25% of alcoholic subjects under 45 years of age had fractures of the spine. Lindsell and colleagues [67] found rib fractures in 28% of 72 patients with alcoholic cirrhosis, 1% of 77 with non-alcohol-related liver disease, and 7% of 149 controls. Israel and colleagues observed that 29% of 198 male alcoholics had rib or spine fractures compared to 2% of 218 controls. The authors suggest that the presence of a verte-

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

bral fracture in men should be regarded as a sign of alcohol consumption. Diamond et al. [69] found that 30% of the 40 subjects who drank alcohol had a history of one or more fractures (wrist, 6; femoral neck, 4; ribs, 4; others, 7). Only 6 (15%) had radiological evidence of crush fractures. Unilateral and bilateral femoral neck fractures occur with increased frequency in alcohol abusers [71 – 74]. 2. MODERATE ALCOHOL CONSUMPTION Evidence for an association between moderate alcohol intake and fractures is contradictory. Men with spine fractures report a higher prevalence of alcohol and tobacco consumption. Seeman et al. [10] reported that moderate alcohol use was associated with a relative risk of 2.4 in 105 men with spine fractures and controls. The risk of osteoporosis increased by 1.007 for each ounce year of cumulative exposure (P  0.01). The relationship between these factors is shown in Fig. 2. The risk was age dependent. The effects of alcohol were not evident in men under 60 years old, emerged in 60 to 69 years old, and were strongest in those 70 years and over. In nonobese individuals over 70 years who drank alcohol and smoked but had no underlying disease, the relative risk was 20.2 (P  0.05). In a nonobese individual who drank, smoked, and had an underlying illness, the relative risk for osteoporosis and vertebral fractures was 192.5 (P  0.05). In contrast, Hemenway and co-workers [74] observed 271 wrist fractures in a study of 51,529 men aged 50 to 75 years during 271,552 person years of observation. The risk of fracture was unrelated to alcohol or tobacco onsumption, age, height, or weight. Paganini-Hill et al. [5] observed a trend for higher risk for fractures with an increasing number of “shots” of liquor per week after menopause. Women drinking more than eight shots per week had a relative risk of 1.85 compared to nondrinkers. Hernandez-Avila and colleagues [75] surveyed 84,484 women in the United States aged 34 – 59 years and found that 5934 forearm and hip fractures occurred during 6 years. A daily consumption of 25 g alcohol (one to two whiskies) was associated with a risk of 2.33 (95% CI, 1.18, 4.57) for hip fractures and 1.38 (95% CI, 1.09, 1.74) for forearm fractures. Significant trends were found for beer and liquor, but not wine. Hemenway and colleagues [74] studied 96,508 nurses aged 35 – 59 years. There was an interaction between body weight and alcohol intake. Approximately 30% took no alcohol, 12% drank between 0 and 1.4 g per day, 20% drank 1. 5 – 4.9 g per day, 19% drank 5 – 15 g per day, and 13% drank more than 15 g per day. Under 1% of respondents reported a fracture during the 4 years (925 of 96,508). Increased fracture risk was found in women who drank more than 15 g of alcohol per day and had a relative weight of less than 21 kg/m2. The increased risk in lean women was confined to those over 50 years of age. Neither factor was independently associated with fracture risk. Women who

783 drank more than 15 g of alcohol per day were not at increased risk unless they were thin and women who were thin were not at increased risk unless they drank more than 15 g of alcohol per day. Those with low body weight who drank most heavily had an age-adjusted risk of 1.73 (95% CI, 1.3, 2.29). In terms of absolute risk, heavy drinkers with low body weight accounted for less than 4% of the population but had 6% of the fractures. That is, only a small proportion of fractures in women are attributable to alcohol. It is important to note that the study cohort was young with only 17% in the oldest age bracket of 55 – 60 years. Felson and colleagues [76] examined the association between alcohol consumption and hip fractures using a retrospective cohort design. During 117,224 person years, 217 hip fractures occurred (174 in women, 43 in men). In the women, the relative risks were 1, 1.34 (95% CI, 0.91, 1.95), and 1.54 (95% CI, 0.95% CI, 0.92, 2.58) for light, moderate (2 to 6 oz. per week), and heavy (more than 7 oz. per week) consumption categories. In men, the age-adjusted relative risks were 1.0 with light, 0.78 (95% CI, 0.34, 1.78) with moderate, and 1.26 (95% CI, 0.62, 2.55) with heavy intake. All confidence intervals included unity. For the entire group, the relative risk was 1.28 per 7 oz. per week of consumption. In those less than 65 years of age, the relative risk with heavy alcohol consumption was 1.4 (95% CI, 1.07, 1.84). In those aged 65 years and over, the relative risk was 1.17 (95% CI, 0.9, 1.53). Stratification acording to drinking, age, and gender resulted in small numbers of cases in each cell and wide confidence inervals. Moderate alcohol intake (4 – 12 drinks per week) increased the risk for hip fracture in women under 65 years of age. Holdrup and collegues [77] recorded alcohol intake in 17,868 men and 13,917 women in three studies between 1964 and 1992. During follow-up, there were 500 first hip fractures in women and 307 in men. Low to moderate alcohol intake (1 – 13 drinks/week in women; 1 – 27 in men) was not associated with hip fracture. In women who drank 14 – 27 drinks/week, the age-adjusted RR of hip fracture was 1.44 (95% CI 1.03, 2.03) and 1.32 (CI 0.92, 1.87) after adjustment for confounder. In men, the RR of hip fracture increased as intake increased above 28 drinks/week: from 1.75 (CI 1.06, 2.89) for 28 – 41 drinks to 1.84 (CI 1.00, 3.41) for 42 – 69 drinks to 5.28 (CI 2.60, 10.70) for 70 or more drinks, after adjustment for confounder. RR of hip fracture in men and women was higher for those preferring beer (1.46; CI 1.11, 1.91) than wine (0.77; CI 0.58, 1.03) and spirits (0.82; CI 0.58, 1.14). The authors conclude different thresholds exist for the harmful effects of drinking on hip fracture between men and women. Navez Diaz and colleagues [78] reported no detectable association between frequency of alcohol intake and vertebral deformity in 14,237 men and women aged 50 years or greater from 19 European countries. Vertebral

784

EGO SEEMAN

deformity was present in 809 men and 884 women. On stratification by age, women 65 years and over who drink alcohol more than 5 days a week had a reduced risk of vertebral deformity; adjusted for age, BMI smoking, current physical activity, and previous fractures, the OR was 0.65 (95% Ci 0.43, 0.99). Thus, moderate alcohol consumption may be a risk factor for fractures, but available data are inconclusive. A risk associated with alcohol consumption may be seen primarily among those with lower body weight. In women, only a small proportion of the fractures are explained by alcohol exposure.

B. Bone Mineral Density 1. ALCOHOL ABUSE There is great deal of evidence for an association between reduced BMD and alcoholism. Alcoholism is associated with other risk factors for osteoporosis and fractures. Poor nutrition, leanness, liver disease, malabsorption, vitamin D deficiency, hypogonadism, hemosiderosis, parathyroid dysfunction, and tobacco use may contribute to the pathogenesis of bone disease in alcoholism. Saville [79] reported reduced trabecular bone volume in iliac crest samples from 39 alcoholics. Diez and colleagues studied 20 men and 6 women 47.7  11.7 years of age without liver disease. An intake of at least 150 g alcohol per day for 8 years or more was associated with decreased bone volume (RR 0.06; 95% CI, 0.01, 0.34). Dalen and Lamke [81] found that well nourished alcoholics had a deficit in BMD of 4.7% at the proximal femur and 8.2% at the calcaneus, with more rapid bone loss at these sites during 40 months of follow-up. Chon et al. [81] reported reduced BMD by about 0.5 to 0.7 SD in chronically alcoholic, but otherwise healthy, men abstinent from alcohol for a median of 4 months. Reduced BMD is not always reported. Harding and colleagues [83] studied alcoholics ages 20 to 40 years and found no reduction in BMD. Laitinen and colleagues found no deficit in BMD at the spine or proximal femur in 27 eugonadal noncirrhotic alcoholic men. The man duration of drinking was 17 years, ranging from 6 to 30 years. Those with longer history of alcohol consumption had lower BMD after adjusting for age and weight. Peris and associates [85] investigated the effect of trauma and/or ethanol induced osteopenia in 76 chronic male alcoholics and in 62 matched controls. Twenty-seven alcoholics (36%) had 41 vertebral fractures and 46 (61%) had a history of nonvertebral fractures. Lumbar spinal BMD was lower in alcoholics than controls (1.12  0.2 vs 1.19  0.1g/cm2, (P  0.009) but did not differ between those with and without fractures. By densitometric criteria, 22 alcoholics (29%) and 5 controls (8%) had osteoporosis. Eight of the former had vertebral fractures and 5 had a lumbar BMD below the vertebral threshold. Previous trauma

was reported by 24 (89%) alcoholics with vertebral fracture and by 28 (57%) without. Alcoholic men frequently have vertebral fractures despite a normal BMD. 2. MODERATE ALCOHOL CONSUMPTION Data on the association between moderate alcohol consumption and BMD are difficult to interpret. In one study of 142 men and 220 women, BMD was measured 12 years after documentation of alcohol intake by a questionnaire [86]. Intake was recorded as low (less than 87.3 g per week or 19.1 g per week or 19.1 g per day), medium (87. 4 – 180.9 g per week or 19. 2 – 41.1 g per day), or high (greater than 181 g per week or 41.2 g per day). Increasing alcohol consumption was associated with higher BMD at the proximal femur in men and at the spine in women. In relation to 1 week of alcohol intake, femoral neck BMD in men and lumbar spine BMD in women were higher in those with a higher alcohol intake (P  0.01). In relation to 24-h intake, BMD at the radius and spine increased in women, not men. In the study by Grainge and colleagues [40] alcohol consumption was not associated with BMD in postmenopausal women. Heavy beer drinkers had lower BMD than non/moderate drinkers; heavy wine drinkers appeared to have higher BMD, while for spirits the pattern was unclear. Hansen and colleagues [87] reported a decreased rate of bone loss in 121 postmenopausal women with moderate alcohol consumption followed for 12 years. Angus and colleagues [88] and Laitinen and colleagues [89] have also reported higher BMD in association with moderate alcohol consumption. Gonzalez-Galkin et al. [90] studied 26 heavy drinkers (more than 100 g ethanol per day for more than 10 years), 13 moderate drinkers (60 – 100 g ethanol per day), and 19 controls. BMD was reduced and correlated with cumulative alcohol intake. These subjects had no underlying cirrhosis. BMD was reduced by about 1 SD at the lumbar spine and 0.6 SD at the femoral neck in 51 heavy drinkers and by about 0.5 SD in 51 in moderate drinkers. Slemenda and colleagues [91] studied 111 U.S. male veterans and found that those drinking more than 1.5 drinks per day lost more bone than nondrinkers during 16 years of follow-up. May and associates [41] studied 458 men with a mean age of 69.1 years (range 64 – 76). BMD at the proximal femur was higher in drinkers than in nondrinkers before and after adjusting for age and weight. After further adjustment for tobacco use, caffeine intake, and activity, the differences remained significant at the trochanter only. The authors classified intake into units, with one unit being equivalent to 9 g of alcohol. The mean alcohol intake was 10 units ranging from 1 to 100 units per week. Among the 335 drinkers, BMD was 0.79  0.13 g/cm2 and among the 123 nondrinkers, BMD was 0.75  0.12 g/cm2 at the femoral neck (P  0.008). The adjusted differences were nonsignificant. At the trochanter the values were 0.75  0.13 g/cm2 versus 0.71  0.12 g/cm2 (P  0.007). This remained significant

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

after adjustment. BMD at Ward’s triangle in drinkers was 0.54  0.13 g/cm2 versus 0.50  0.12 g/cm2 (P  0.006). This was not significant after adjustment. Differences in BMD in the lumbar spine were not significant. Men in the highest tertiles of alcohol intake (more than 11 units per week) had higher glutamyl transferase levels and a lower caffeine intake and were younger. Alcohol intake of one to two drinks per day did not appear to have a detrimental effect on BMD. The results expressed as a function of increasing alcohol consumption showed a trend for higher BMD in those consuming a greater number of units per alcohol per week but at no site were these increments significant after adjusting for covariates. Although the authors concluded that alcohol intake is associated with higher BMD, the evidence for this is not compelling. Thus, available data suggest that moderate alcohol intake is unlikely to be associated with lower BMD. When an association is found, the deficits appear to be explained by confounding factors. The association between moderate alcohol intake and higher BMD remains a possibility, but should be interpreted with caution. Although these results were adjusted for covariates, moderate alcohol intake may be associated with higher socioeconomic groups with better nutrition and lifestyle during growth and adulthood.

C. Histomorphometry before and after Abstention from Alcohol Osteoporosis is common in alcoholics. Osteomalacia may be found but is less common. Arlot and colleagues studied 77 French patients with aseptic necrosis of the hip. Sixty-eight had low trabecular bone volume and of these, 29 were alcoholics. Only 9 had osteomalacia and of these, 4 were alcoholics. Bikle reported a high bone turnover osteoporosis in younger alcoholics. Low turnover osteoporosis is common in chronic alcoholism. Saville [79] reported reduced trabecular bone volume in iliac crest samples from 39 alcoholics. Diez and colleagues [80] studied 20 men and 6 women 47. 7 11.7 years of age without liver disease. An intake of at least 150 g alcohol per day for 8 years or more was associated with decreased bone volume (RR 0.06; 95% CI, 0.01, 0.34). Static and dynamic histomorphometric measures show that the reduction in BMD is due to reduced bone formation with a contribution of increased bone resorption. In the study by Diez et al. [80] data were reported relative to biopsies from 26 normal subjects (kidney donors), 8 male, 18 female, mean age 44.8  16 years (SD). The following parameters of bone formation were reported: decreased bone formation rate (0.023  0.028 mm3/mm2/day vs 0.0409  0.020 mm3/mm2/day, P  0.013), decreased mineralized surfaces (3.75  4.4% vs 5.9  2.9%, P  0.01), decreased osteoid maturation rate (2.4 

785 2.8%/day vs 3.1  0.1%/day, P  0.02), increased mineralization lag time (175.5  405.5 days vs 34.5  8.1 days, P  0.01, P  0.008), and reduced mineral appositional rate (0.309  0.269 mm/day vs 0.771  0.529 mm/day, P  0.03). Increased resorption surface and increased osteoclast number were observed. In a study of female rats by Diez and associates [93], intraperitoneal ethanol (2 g/kg BW) or saline was administered followed by sacrifice at 1, 4, 8, and 24. Ethanol decreased osteoid surface with osteoblasts (42.86  15.61 vs 64.57  6.24%, P 0.05) and osteoclast number (0.05  0.02 vs 0.17  0.09 n/mm2, P 0.05). Osteoclast surfaces decreased at 4 (0.129  0.09 vs 0.425  0.26%, P 0.05), with an increase at 8 (0.765  0.24 vs 0.226  0.17, P 0.01). The osteoblast surface was unchanged. Thus, measures of bone formation are reduced, with a reduction in remodeling consistent with a reduction in bone turnover. Chappard and co-workers [94] reported reduced trabecular bone volume (14.2  4.6% vs 18.8  4.8%, P  0.01), mean wall thickness (39.6  8.1 mm vs 50.2  8.7 mm, P  0.001), and bone formation rate as assessed by double tetracycline labelling in 20 patients 59.1  10.1 years of age with alcoholic cirrhosis. The mineral appositional rate was reduced by about 50%, where mineralizing surfaces were reduced and the bone formation rate was half that found in controls (0.009  0.001 mm3/mm2/year vs 0.0175  0.0125 mm3/mm2/year, P  0.001). Trabecular thickness was reduced (106  25.2 mm vs 131  19 mm, P  0.05). There was no change in trabecular number. Eroded surfaces were increased (8.1  5.2% vs 3.7  1.1%, P  0.001). Osteoclast numbers were not increased. Schnitzler and associates [95] examined the contributions of alcohol and iron in osteoporosis associated with hemosiderosis. Iliac crest bone biopsies were taken from 53 black male drinkers: 38 with and 15 without iron overload. Bone volume and trabecular thickness were lower in both groups than in matched controls and were attributed to alcohol. Mineralization lag time was increased in 34% of patients with iron overload and in 27% of those without, combined with low normal or subnormal osteoid thickness; this evidence of osteoblast dysfunction is also attributed to alcohol. Erosion depth (r  0.373), trabecular number (r  0.295), and trabecular separation (r  0.347) correlated with iron granule number in marrow and are attributed mainly to iron overload. Crilly and colleagues [94] examined the effects of abstention from alcohol on bone histomorphometry in men (16 current drinkers and 9 abstainers). Osteoid seams were increased in abstainers (11.4  0.68 mm vs 7.95  0.48 mm, P  0.001). The osteoid surface with osteoblasts was 20.1  4.3% vs 5.5  1.7% (P  0.01). Mean wall thickness was higher but not significantly so in abstainers (41.0  2.8 mm vs 34.7  2.3 mm, NS). Likewise, seams with double labels were higher but not significantly so in

786 abstainers (0.58  0.07 vs 0.37  0.10, NS). Mineralization rate was 0.52  0.1 mm/day vs 0.26  0.07 mm/day (P  0.04). Mineralization lag time was 28  4 days vs 62  10 days, (P  0.006). Resorption parameters were no different. Diamond and associates [97] studied 28 current drinkers and 12 subjects who had not taken alcohol for at least 6 months. Thirty-five controls were studied without clinical or biochemical evidence of liver disease. Reduced BMD was found in both drinkers and abstainers, but abstainers had higher osteoblast perimeters, mineralizing perimeters, and bone formation rate with a shorter mineralization lag time than patients who continued to drink. Serum testosterone concentrations were higher in abstainers than in drinkers (18.5  3.1 pg/ml vs 14.7  1.7 pg/ml) and did not differ from controls. Lindholm and co-workers [98] studied the reversibility of the effects of alcohol abuse on bone histomorphometry in men who had abstained from alcohol for at least 2 years. There were eight abstainers and nine drinkers. Mineralizing surfaces, bone formation rate, osteoid width and wall thickness, adjusted appositional rate, and formation period were greater in abstainers than in drinkers. There were no significant differences in resorptive indices between drinkers and abstainers. Median activation frequency was higher in abstainers than in drinkers. Although there was no difference in 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D was higher in abstainers (111.6 vs 56.4, median values, P  0.01). Data were consistent with reversible histological changes, although ideally paired biopsies would have provided more convincing evidence, as disease severity may have been less in those able to abstain from alcohol. There was no evidence of osteomalacia. In growing animals, alcohol appears to reduce growth in size and volumetric density, effects that are partially restored by resumption of growth with cessation of alcohol. For example, Sampson and colleagues [99] studied 4-weekold female rats fed a liquid diet containing 35% ethanol-derived calories, a liquid control diet, or standard chow for 2 or 4 weeks. The alcohol group gained less weight than chow controls at 2 and 4 weeks. Relative to liquid controls, the alcohol group had lower femoral length, diameter, volume, wet weight, dry weight, ash weight, BMD, and percentage mineral at 4 weeks. Chow-fed animals had similar or greater values for these parameters than liquid controls. Tibial trabecular area and thickness in the alcohol group were lower than in chow controls but not in liquid controls. Serum insulin-like growth factor-I (IGF-I) was reduced at 2 and 4 weeks by the alcohol diet relative to chow controls, with intermediate levels in the liquid controls. In animals fed alcohol for 2 weeks then chow for 2 weeks, relative to the group alcohol fed for 4 weeks, the recovery group had greater weight; greater femoral length, volume, wet weight, dry weight, and ash weight; lower tib-

EGO SEEMAN

ial trabecular area; and greater tibial trabecular number and thickness. Femoral density, percentage mineral, and percentage water did not differ. The recovery group was below pair-fed controls for femoral length, wet weight, dry weight, ash weight, and density. The recovery group had IGF-I levels close to, but below, controls. The authors concluded that animals removed from the alcohol diet improved incompletely in all parameters, probably due to the continued growth of young bones rather than to regaining bone lost during alcohol consumption. Kidder and colleagues [100] studied 6-month-old ovariectomized rats receiving a liquid diet supplemented isocalorically with 13 or 35% ethanol for 2 months. Ethanol at both concentrations reduced the ovariectomy-related increase in body weight. The 35% ethanol group increased tibial cortical medullary area (1.31  0.04 vs 1.08  0.07 mm2) relative to ovariectomy alone. Cortical bone area and periosteal perimeter were unchanged. Dynamic parameters in cortical bone and static and dynamic parameters in cancellous bone were unaltered. The authors concluded that a chronic ingestion of high doses of alcohol does not accentuate bone loss in ovariectomized rats. Thus, the main histomorphometric abnormality is a reduction in parameters of bone formation, although eroded surfaces may be increased. A reduction in bone formation may be a direct toxic effect of alcohol causing a reduction in osteoblast life span or a reduction in the activity of osteoblasts. A reduction in serum testosterone concentration may contribute, as there is an association between bone volume and serum testosterone in alcoholic cirrhosis [100].

D. Biochemical Measures of Bone Remodeling In general, measurement of biochemical markers of bone remodeling confirms that bone loss is due to a reduction in bone formation rather than increased bone resorption. In the vast majority of studies, circulating osteocalcin, a marker of bone formation, is reduced and increases with abstention from alcohol. Jaouhari and co-workers [101] studied 40 chronic alcoholics before and after 3 weeks of ethanol withdrawal and compared the results with 50 nonalcoholic controls matched for age and gender. Plasma osteocalcin was reduced in the cases (3.0  2.6 mg/liter vs 4.7  2.8 mg/liter). After 21 days of withdrawal, osteocalcin levels increased (5.8  3.5 mg/liter) and were no different from those of controls. The hydroxyapatite-binding capacity of the plasma osteocalcin before and after withdrawal was also not different from that of controls. These features support the view that the reduction in bone formation may be reversible. Pepersack and colleagues [102] studied the effects of chronic alcoholism on bone turnover markers in 12 alco-

787

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

holic men before and during a 2-week period of alcohol withdrawal. Osteocalcin was reduced and urinary excretion of hydroxyproline was increased, suggesting an imbalance in bone turnover. An increased renal threshold for phosphate (TmP/GFR) was present without a change in serum (PTH). Following alcohol withdrawal, serum osteocalcin increased from 1.8  0.3 to 2.6  0.3 ng/ml, no different from the control value of 2.6  0.2 ng/ml. TmP/GFR was 3.8  0.3 mg/dl and increased to 4.3  0.2 mg/dl after 2 weeks. Fasting hydroxyproline excretion was elevated compared to that of controls (19.9  1.7 mg/mg versus 13.4  1.8 mg/mg creatinine) and did not fall to the normal range after 2 weeks. No changes were observed in serum concentrations of calcium, phosphate, magnesium, albumin, (PTH), 25-hydroxyvitamin D, or 1,25-dihydroxyvitamin D during the observation period. Preedy and colleagues [103] studied the urinary excretion of collagen degradation products using a rat model of alcoholic bone disease and suggested that there is a reduction in bone resorption. Six weeks of ethanol feeding showed that there was a reduction in total and conjugated deoxypyridinoline, a marker found in only type I collagen of bone and dentine. There was a 20% reduction in free deoxypyridinoline (NS). The 24-h urinary excretion of total, free, and conjugated deoxypyridinoline was reduced by 25 – 55%. Total, free, and conjugated pyridinoline were unaltered. There was no effect on the 24-h urinary excretion of pyridinoline. (Pyridinoline is found in cartilage and bone and some other tissues.) These observations are consistent with low bone remodeling associated with alcohol consumption. Hydroxyproline may be increased in current users of alcohol, consistent with increased bone resorption [104]. However, hydroxyproline may not be from bone. The increase in hydroxyproline excretion reported by Diamond and colleagues [97] may reflect a change in nutritional status, as hydroxyproline excretion is not only determined by bone resorption. Pyridinoline is found largely in type II and type IX collagens of cartilage and to a lesser extent, in type I collagen of bone. Deoxypyridinoline is found only in type I collagen of bone and dentine. Neither cross-links are found in the skin. Laitinen and co-workers [89] studied 27 noncirrhotic male alcoholics in a hospital for 2 weeks. Osteocalcin was reduced by 28% at the time of admission. The procollagen 1 carboxyterminal propeptide was reduced by 17% and normalized with abstention. Ionized calcium increased. PTH was unchanged. Intestinal calcium absorption measured by stable strontium was 37% higher than controls and decreased over time. Rico et al. [104] studied 15 patients with acute alcohol intoxication. Serum osteocalcin was lower than in controls of similar age and sex (2.7  0.9 ng/ml vs 6.6  0.8 ng/ml). There was no correlation between serum osteocal-

cin and alkaline phosphatase. However, there was a correlation between osteocalcin and glutamyl transferase activity (r   0.78, P  0.001).

E. Alcohol and Cellular Function There is evidence based on calvarial cultures that alcohol or acetaldehyde may reduce bone formation. Hurley et al. [105] measured bone formation using tritiated proline incorporation into collagenase-digestible protein and noncollagen protein. One percent ethanol decreased tritiated thymidine incorporation by 31% at 24 h whereas 0.1% ethanol increased tritiated thymidine incorporation by 22%, collagenase digestible protein by 73%, and noncollagen protein by 67% at 24 h. Prostaglandin release was decreased by 88 and 75% using 1 and 0.3% ethanol, respectively. Acetaldehyde inhibited tritiated thymidine and proline incorporation and inhibited PTH-stimulated bone resorption whereas ethanol had no effect. The authors concluded that ethanol has little effect on bone formation or resorption; however, acetaldehyde is a potent inhibitor of both. Giuliani and colleagues [106] reported that ethanol and acetaldehyde inhibit osteoblastogenesis of bone marrow cells, an effect that may contribute to the reduced bone formation found in alcoholics. Farley and colleagues [107] showed that ethanol reduced tritiated thymidine incorporation into monolayer cultures of calvarial cells in a time and dose-dependent manner. Ethanol also reduced the mitogenic action of human skeletal growth factor, sodium fluoride, and PTH. Alkaline phosphatase activity per cell was decreased. Tritiated hydroxyproline synthesis from proline was unaffected by ethanol, but the effect of PTH and sodium fluoride on hydroxyproline synthesis was inhibited with 0.2% ethanol. These features are consistent with the view that ethanol may reduce bone formation directly and indirectly. Ethanol increased cAMP and prostaglandin E2, (PG2E) production by calvarial cells and increased skeletal collagen resorption reflected in the release of tritiated hydroxyproline from intact embryonic chick tibiae prelabeled with tritiated proline in vitro. These authors also provided evidence to suggest that ethanol affects bone cells by changing membrane fluidity. Friday and Howard [108] showed that the synthesis of DNA measured by tritium-labeled thymidine incorporation was reduced in human osteoblasts derived by the collagenase digestion of trabecular bone after exposure to 0.0 5 – 1% ethanol. Protein synthesis measured by tritiated proline incorporation into trichloroacetic acid-precipitable material was reduced. Human bone cell protein concentrations and alkaline phosphatase activity were reduced after exposure to 1% ethanol but not lower doses. Cheung and associates [109] studied the direct effects of ethanol on osteoclasts from long bones of 19-day

788 prehatched chicks. The osteoclasts were seeded onto dentine slices and cultured with and without ethanol. Resorption pits in dentine were measured using confocal laser reflection microscopy. Increased pit numbers, areas, volumes, and volume/area ratios were observed with 0.001 and 0.01% ethanol. Greatest mean volume and area resorbed per pit and number of pits occurred with 0.01% ethanol. Volume/area (mean depth) per pit was greatest with 0.001% ethanol. These are ethanol concentrations encountered in social drinkers. Osteoclasts are derived from a pluripotent stem cell most likely to be a colony-forming unit from the granulocyte – macrophage series (CFU-GM). Ethanol and its metabolites impair GFU-GM formation. Tisman and Herbet [110] showed that ethanol inhibits the proliferation of granulocyte – macrophage progenitor cells. Meagher and colleagues [111] showed that erythroid progenitor cells were suppressed by 0. 0 5 – 0.2% ethanol and 0. 001 – 0.01% acetaldehyde. Granulocyte – macrophage progenitor cell suppression required 3% ethanol or 0.03% acetaldehyde. These suppressive effects were partly reversed by folinic acid or pyridoxine. The lowest concentration of acetaldehyde that inhibited colony formation was 0.001% for burst-forming units – erthyroid (BFU-E) and colony-forming units – erthyroid (CFU-E) and 0.03% for CFU-GM. Balsinde and co-workers [112] showed that ethanol inhibits the mouse peritoneal macrophage superoxide anion response to phorbol myristate acetate, perhaps through disorganization of the plasma membrane. These investigators studied the interaction of ethanol and exogenous arachadonic acid in the generation of extracellular messengers by mouse peritoneal macrophages. Ethanol caused a dose-dependent decrease in the production of cyclooxygenase and lipooxygenase metabolites. Hydroxy-eicostetranoic acid metabolites and 6-ketoprostaglandinl levels were reduced in the presence of ethanol. The effect of ethanol on the metabolism of exogenous arachadonic acid to 5 hydroxy-eicostetranoic acid by macrophages was dose dependent. The failure of generation of oxygenated metabolites of arachadonic acid may be due to the impairment of access of arachadonic acid into the cell. Gilhus and Matre [113] studied mononuclear cells from 10 healthy blood doners. Incubation with ethanol at 37°C and 10 liter reduced active E rosette-forming cells. The percentage of cells with phagocytic capacity was reduced from 11.7  5.4 to 7.0  5.0 (P  0.01). Incubation at 1.0 g/liter had similar effects, but not at 0.1 g/liter. Ethanol reduces the mobilization and phagocytic capacity of macrophages in the lung, in the peritoneal cavity, and in the lining of vascular sinusoids. Fixed macrophages in intoxicated patients have depressed phagocytic activity, but the function returns to normal after abstention. Ethanol impairs the phagocytic ability of monocytes and macrophages and enhances the ability to produce superox-

EGO SEEMAN

ide anions and ruffled borders in hepatic macrophages [94 – 96]. Whether these models apply to the cells of bone is uncertain. Changes in membrane fluidity have been reported with ethanol in vivo and in vitro. Dimethyl sulfoxide, ethylene glycol, and lethicin increase membrane fluidity and mimic the effects of ethanol on bone cell proliferation. The changes in membrane fluidity may contribute to changes in membrane-bound enzymes such as skeletal alkaline phosphatase, changes in bone cell responsiveness to parathyroid hormone, and changes in cyclic AMP production [114 – 116].

F. Gonadal Function Gonadal dysfunction in men may be due to concurrent direct effects of alcohol on testicular function, hypothalamic pituitary function, altered clearance by the liver due to increased hepatic testosterone A ring reductase activity, increased hepatic blood flow, and interference with retinonic acid metabolism. Clarification of whether one or more of these mechanisms is responsible will require studies designed specifically to address this question. Alcoholics may have reduced serum testosterone concentrations. Diamond and associates [69,97] studied 115 consecutive ambulant patients with histologically proven chronic liver disease and 113 age- and gender-matched controls. The 40 alcoholic patients had a higher prevalence of peripheral fractures than patients with other liver disorders. The presence of hypogonadism was an important risk factor. Among 81 eugonadal subjects, 12 had fractures. Among 34 hypogonadal subjects, 21 had fractures. Van Thiel and colleagues [117,118] studied 40 men with alcoholic liver disease and reported reduced serum testosterone with more severe derangements of liver histology. FSH and LH were elevated, consistent with a primary abnormality of testicular function. Sex hormone-binding globulin was elevated eightfold. However, the response to clomiphene stimulation was diminished, suggesting that both a hypothalamic – pituitary and a local testicular defect were responsible for the hypogonadism in these subjects. Mendelson and colleagues [119] showed that acute alcohol intoxication in 16 healthy nonalcoholic males was associated with a reduction in serum testosterone and a rise, not fall, in LH concentration. A rapid fall in testosterone during acute alcohol administration may be the result if hepatic blood flow and liver metabolism of the testosterone increase. Rubin et al. [120] report increased hepatic metabolism via testosterone A ring reductase. Gordon and colleagues [121] found increased metabolic clearance and reduced production of testosterone in 11 healthy men given ethanol for 4 weeks. The male volunteers were ages 20 – 40 years with no evidence of liver dis-

789

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

ease. Nine were social drinkers consuming no more than 70 g per week of alcohol and 2 were chronic alcoholics. Subjects were given alcohol for 4 weeks, with liver biopsy before and at the end of the experiment in 5 of the subjects. Hepatic testosterone A ring reductase activity increased in the 5 subjects tested. The mean plasma testosterone concentration decreased in the 6 subjects tested, based on 49 observations in each of 4 subjects and 25 observations in each of 2 subjects per 24 h. The metabolic clearance rate of testosterone increased, while the production rate decreased in 3 of the 4 subjects. Inconsistent results were obtained in measured gonadotropins. Testosterone-binding capacity fell during the study. Gordon and co-workers [122] showed that hepatic aromatase activity is increased in rats exposed to ethanol, and a reduction in plasma testosterone may be due to an increased conversion to estrogen. Plasma testosterone decreased by 55% (318  48 ng/dl vs 144  22 ng/dl), Whereas plasma estrogen increased by 60% (1.5  0.1 ng/dl vs 2.4  0.4 ng/dl). There were no significant differences in testosterone within the testis (assessed in vitro). Hepatic aromatase activity increased with testosterone substrate in the assay (4.8  0.5 pmol/h/10 mg protein vs 6.3  0.7 pmol/h/10 mg protein). Van Thiel and colleagues [123] showed that alcohol-fed rats had a marked reduction in testicular mass compared to controls and that this was caused by a reduction in the mean seminiferous tubular diameter and a reduction in the amount of germinal epithelium. Serum testosterone decreased compared to isocaloric controls (291.0  38.1

FIGURE 11

pg/ml vs 1619.3  283.3 pg/ml). There were no differences in the histology of the pituitary gland in the two groups of animals. Microsomal reductase activity more than doubled in ethanol-fed ratscompared to that seen in controls (43.4  3.1 vs 22.3  1.9, P  0.005). Vitamin A is essential for spermatogenesis [124]. It is ingested as retinol and oxidized to the active retinal by alcohol dehydrogenase. Alcohol dehydrogenase metabolizes ethanol to acetaldehyde. Van Thiel and colleagues [123] showed that ethanol inhibited the oxidation of retinol by testicular homogenates containing alcohol dehydrogenase. Ethanol inhibits testicular retinal formation. Complete inhibition of retinol oxidation occurred at molar concentrations of ethanol greater than 2  105. The hypothesis was not directly tested by measuring either spermatogenesis or testosterone metabolism. Ginsburg et al. [125] studied 12 healthy postmenopausal women aged 54.5  4.0 years receiving estradiol (1 mg/day) and progestin and 12 women aged 61.7  6.4 years not receiving hormone replacement therapy (HRT). Each took 0.7 g/kg alcohol and isoenergetic placebo in random sequence on consecutive days. After alcohol ingestion, circulating alcohol levels were similar in both groups. In the non-HRT group, circulating estradiol was unchanged by alcohol. Circulating estradiol increased three fold in the HRT group (from 297 to 973 pmol/l) and remained elevated for 5 h (Fig. 11). Estradiol and alcohol levels correlated during the ascending and descending phases of the blood alcohol curve (r  0.9, both). Acute alcohol may thus result in sustained elevations in estradiol.

Estradiol levels after alcohol and placebo remain low and not different in women not receiving estrogen. In women receiving estrogen, estradiol increased and remained elevated after alcohol but not placebo. From Ginsburg and colleagues [125], with permission from the authors and publisher.

790

EGO SEEMAN

G. Alcohol and Mineral Metabolism Vitamin D deficiency may result from a lack of sunlight exposure, malabsorption due to gastrectomy, pancreatitis, liver disease, or protein malnutrition. Reduced vitamin D and 1,25-dihydroxyvitamin D levels are reported in alcoholics with or without liver disease. Circulating free hormones are usually normal [66]. The deficits are reversible with the cessation of alcohol [66,97,98] and are partly due to reduced production of the vitamin D-binding proteins. Serum calcium concentrations may be reduced in alcoholics. Ionized calcium activity is usually normal or slightly reduced but may be increased. For example, Diez and colleagues [80] found increased plasma calcium and decreased PTH after adjustment for age and gender in 20 men and 6 women (47.7  11.7 years of age) without liver disease and with a daily intake of at least 150 g alcohol for 8 years or more. In this study, the ethanol group showed reduced plasma calcium (8.51  0.23 vs 9.10  0.29 mg/dl, P  0.01) and reduced PTH (23.51  5.72 vs 76.39  11.66 pg/ml, P  0.001) at 1 h and reduced osteocalcin at 24 h (36.98  2.21 vs 45.77  5.05 ng/ml, P  0.05). Magnesium deficiency is common in alcoholics and is due to malabsorption, renal wasting, and the use of diuretics [93]. It is associated with hypocalcemia, hypoparathyroidism, and PTH resistance with an impaired cAMP response and 1,25-dihydroxyvitamin D response to PTH. Magnesium deficiency may reduce PTH secretion and contribute to a rapid fall of serum calcium with acute alcohol administration [93]. Phosphate deficiency occurs in alcoholics, and reduced muscle phosphate may contribute to the myopathy. PTH may be normal, reduced, or increased [93]. PTH may decrease in response to acute alcohol consumption. This may precede, rather than follow, a fall in ionized calcium activity, so the effect may be direct. Elevated PTH values have been found in alcoholic men [98], but Diamond and colleagues [97] found no difference between drinkers and abstainers. The reasons for the variable observations are uncertain.

cell, nutritional deficiency, depressed glycolytic enzyme activity, direct inhibition of muscle carbohydrate metabolism, and potassium deficiency [126]. Myopathy is characterized by a decreased diameter of type II muscle fibers (fast twitch). The II B fibers, which have no or few mitochondria, are more affected than the type II A fibers. Type I fibers, which are rich in mitochondria and are slow twitch, with aerobic or oxidative metabolism, are less sensitive and may show compensatory hypertrophy. The decrease in type II fibers is responsible for the loss of muscle mass, which may at least in part contribute to the frequency of falls, difficulties in gait, proximal muscle weakness, and perhaps muscle cramps. The acute form of rhabdomyolysis occurs in less than 1% of alcoholics. The pathogenesis is likely to be due to free radical damage. Type I fibers have a higher antioxidant capacity than type II fibers. Alcoholics have a reduced antioxidant status. A deficiency in antioxidants is associated with myopathy in animal models [127,128].

I. Summary Alcohol abuse appears to be associated with an increased risk for fracture in women and men. However, few fractures in women are attributable to alcohol abuse, wheras alcohol abuse with or without concomitant hypogonadism is reported to be much more common in men presenting with fractures. This increased risk may be at least, in part, attributable to a reduction in BMD, which is due to a reduction in bone formation. Increased bone resorption probably contributes, but evidence for this is less compelling. Moderate alcohol consumption may be associated with a higher risk for fractures, but data are not consistent. Some evidence suggests that moderate alcohol consumption may be associated with a higher BMD. A concomitant lifestyle or socioeconomic factors may explain this association.

IV. CONCLUSIONS AND QUESTIONS H. Alcohol and Myopathy Alcohol myopathy is characterized by acute muscle tenderness, swelling, pain, and myoglobinuria. Type I myopathy is a subclinical disorder manifested by biochemical changes only (elevated creatinine phosphokinase). Type II disease is an acute variety characterized by muscle cramps, diffuse muscle weakness, rhabdomyolysis, and myoglobinuria. Type III myopathy is a chronic form associated with proximal muscle wasting and weakness. Several mechanisms are involved in the pathogenesis, including prolonged ischemia, direct injury to the sarcolemmal membrane, toxic inhibition of active transport by the muscle

Tobacco use is an important risk factor for osteoporosis in women. This information should be used by health organizations as a means of dissuading women from taking up smoking or continuing to smoke. Bone loss is likely to be responsible for the low BMD in smokers. The effects of tobacco use on mineral accrual during growth are unknown. Tobacco use is probably a risk factor for osteoporosis in men as well. Alcohol excess appears to be a cause of fractures, but from a public health point of view, this is a problem in men, not in women. Specific studies should be designed to address the question of whether moderate alcohol consumption is a protective factor against fractures, an observation that has been made by several groups. Whether

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone

confounding by socioeconomic and other factors explain this observation is uncertain. If moderate alcohol use is associated with higher BMD, then the mechanism of this effect needs to be understood.

References 1. V. L. Ernster, Women and smoking. Am. J. Public Health 83(9), 1202 – 1204 (1993). 2. CDC, “Reducing the Health Consequences of Smoking: 25 Years of Progress – A Report of the Surgeon General,” DHHS publication No. (C.D.C.) 89-8411. U.S. Department of Health and Human Services, Public Health Service, Rockville, Md, 1989. 3. A.C. King, C. B. Taylor, and W. L. Haskell, Smoking in older women. Is being a female a “risk factor” for continued cigarette use? Arch. Intern. Med. 150, 1841 – 1846 (1990). 4. M. R. Baker, “The Epidemiology and Etiology of Femoral Neck Fracture,” M.D. thesis, University of Newcastle-upon-Tyne, 1980. 5. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 6. A. R. Williams, N. S. Weiss, C. L. Ure, J. Ballard, and J. R. Daling, Effect of weight, smoking and estrogen use on the risk of hip and forearm fractures in postmenopausal women. Obstet. Gynaecol. 60, 695 – 699 (1982). 7. N. Kreiger, J. L. Kelsey, T. R. Holford, and T. O’Connor, An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 141–148 (1982). 8. W. J. Boyce, “Fracture of the Proximal Femur: Studies of the Distribution, Risk Factors, Care and Outcome in Oxford, 1983 – 1984,” MFCM thesis, Royal College of Physicians, London, 1985. 9. B. W. Alderman, N. S. Weiss, J. R. Daling, C. L. Ure, and J. H. Ballard, Reproductive history and postmenopausal risk of hip and forearm fracture. Am. J. Epidemiol. 124, 262 – 267 (1986). 10. C. Cooper and C. Wickham, “Smoking and Hormone-Related Disorders” (N. Wald and J. Baron, eds.), Chapter 8, pp. 93 – 99. Oxford Univ. Press, 1990. 11. O. Johnell, B. Gullberg, J. A. Kanis, E.Allander, L. Elffors, J. Dequeker, G. Dilsen, C. Gennari, A.L. Vaz, G.Lyritis, G. Mazzuoli, L.Miravet, M. Passeri, R.P. Cano, A. Rapado, and C. Ribot, Risk factors for hip fracture in European women: The MEDOS study. J. Bone Miner. Res. 10, 1802 – 15 (1995). 12. J. Cornuz, D. Feskanich, W. C. Willett, and G. A. Colditz, Smoking, smoking cessation, and risk of hip fracture in women. Am. J. Med, 106, 311 – 4 (1999). 13. H. Melhus, K. Michaelsson, L. Holmberg, A. Wolk, and S. Ljunghall, Smoking, antioxidant vitamins, and the risk of hip fracture. J. Bone Miner. Res. 14, 129 – 35 (1999). 14. D. P. Kiel, J. A. Baron, J. J. Anderson, M. T. Hannan, and D. T. Felson, Smoking eliminates the protective effect of oral estrogens on the risk for hip fracture among women. Ann. Int. Med. 116, 716 – 721 (1992). 15. L. Forsen, A. Bjorndal, K. Bjartveit, T. H. Edna, J. Holmen, V. Jessen, and G. Westberg, Interaction between current smoking, leanness, and physical inactivity in the prediction of hip fracture. J. Bone Miner. Res. 9, 1671 – 1678 (1994). 16. D. Hemenway, G. A. Colditz, W. C. Willett, M. J. Stampfer, and S. E. Steizer, Fractures and lifestyle: Effect of cigarette smoking, alcohol intake, and relative weight on the risk of hip and forearm fractures in middle aged women. Am. J. Public Health 78, 1554 – 1558 (1988). 17. M. R. Law and A. K. Hackshaw, A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: Recognition of a major effect. Br Med. J. 315, 841 – 846 (1997).

791 18. H. W. Daniell, Osteoporosis of the slender smoker: Vertebral compression fractures and loss of metacarpal cortex in relation to postmenopausal cigarette smoking and lack of obesity. Arch. Intern. Med. 136, 298 – 304 (1976). 19. J. Aloia, S. H. Cohn, A. Vaswani, J. K. Yeh, and K. Ellis, Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 20. E. Seeman, L. J. Melton III, T. O’Fallon, and B. L. Riggs, Risk factors for spinal osteoporosis in men. Am. J. Med. 75, 977 – 983 (1983). 21. G.F. Jensen, Osteoporosis of the slender smoker revisited by epidemiologic approach. Eur. J. Clin. Invest. 16, 239 – 242 (1986). 22. G. Jones, M. Riley, and T. Dwyer, Maternal smoking during pregnancy, growth, and bone mass in prepubertal children. J. Bone Miner. Res. 14, 146 – 51 (1999). 23. E. A. Krall and B. Dawson-Hughes, Smoking and bone loss among postmenopausal women. J. Bone Miner. Res. 6, 331 – 338 (1991). 24. E. A. Krall and B. Dawson-Hughes, Smoking increases bone loss and decreases intestinal calcium absorption. J. Bone Miner. Res. 14, 215 – 20 (1999). 25. C. W. Slemenda, S. L. Hui, C. Longcope, and C. Johnston, Jr., Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4, 737 – 741 (1989). 26. D. Sparrow, N. I. Beausoleil, A. J. Garvey, B. Rosner, and J. E. Silbert, The influence of cigarette smoking and age on bone loss in men. Arch. Environ. Health 37, 246 – 249 (1982). 27. A. Rundgren and D. Mellstrom, The effect of tobacco smoking on the bone mineral content of the ageing skeleton. Mechanisms Ageing Dev. 28, 272 – 277 (1984). 28. O. Johnell and B. E. Nilsson, Life-style and bone mineral mass in perimenopausal women. Calcif. Tissue Int. 35, 354–356 (1984). 29. O. Lindquist and C. Bengtsson, Menopausal age in relation to smoking. Acta Med. Scand. 205, 73 – 77 (1979). 30. I. Hollo, I. Gergely, and M. Boross, Influence of heavy smoking upon the bone mineral content of the radius of the aged and effect of tobacco smoke on the sensitivity to calcitonin of rats. Aktuel. Gerontol. 9, 365 – 368 (1979). 31. R. Lindsay, The influence of cigarette smoking on bone mass and bone loss. “Osteoporosis: Recent Advances in Pathogenesis and Treatment” (H. F. Deluca, H. M. Frost, C.C. Johnston, Jr., and A. M. Parfitt, eds.), p. 481. Univ. Park Press, Baltimore, MD, 1981. 32. M. R. Sowers, R. B. Wallace, and J. H. Lemke, Correlates of mid-radius bone density among postmenopausal women: A community study. Am. J. Clin. Nutr. 41, 1045 – 1053 (1985). 33. H. Suominen, E. Heikkinen, P. Vainio, and T. Laitinen, Mineral density of calcaneous in men at different ages: A population study with special references to life-style factors. Age Ageing 13, 273 – 281 (1984). 34. P. McNair, M. S. Christiansen, S. Madsbad, C. Christiansen, C. Binder, and I. Transbol, Bone loss in patients with diabetes mellitus: Effects of smoking. Miner. Electrolyte Metab. 3, 94 – 97 (1980). 35. A. J. Friedman, V. A. Ravnikar, and R. L. Barbieri, Serum steroid hormone profiles in postmenopausal smokers and nonsmokers. Fertil. Steril. 47(3), 398 – 401 (1987). 36. M. A. Hansen, Assessment of age and risk factors on bone density and bone turnover in healthy premenopausal women. Osteopor. Int. 4, 123 – 128 (1994). 37. J. C. Stevenson, B. Lees, M. Devenport, M. P. Cust, and K. F. Ganger, Determinants of bone density in normal women: Risk factors for future osteoporosis? Br. Med. J. 298, 924 – 928 (1989). 38. R. G. McCulloch, D. A. Bailey, C. S. Houston, and B. L. Dodd. Effects of physical activity, dietary calcium intake, and selected lifestyle factors on bone density in young women. Can. Med. Assoc. J. 142, 221 – 227 (1990). 39. M. T. McDermott and M. C. Witte, Bone mineral content in smokers. South. Med. J. 81, 477 – 480 (1988).

792 40. M. J. Grainge, C. A. C. Coupland, S. J. Cliffe, C. E. D. Chilvers, and D. J. Hosking, Cigarette smoking, alcohol and caffeine consumption, and bone mineral density in postmenopausal women. Osteopo. Int. 8, 355 – 63 (1998). 41. H. May, S. Murphy, and K. T. Khaw, Alcohol consumption and bone mineral density in older men. Gerontology 41, 152 – 158 (1995). 42. P. Egger, S. Duggleby, R. Hobbs, C. Fall, and C. Cooper, Cigarette smoking and bone mineral density in the elderly. J. Epidemiol. Community Health 50, 47 – 50 (1996). 43. G. Jones and F. S. Scott, A cross-sectional study of smoking and bone mineral density in premenopausal parous women: Effect of body mass index, breastfeeding, and sports participation. J. Bone Miner. Res. 14, 1628 – 33 (1999). 44. J. L. Hopper and E. Seeman, Bone density in twins discordant for tobacco use. N. Engl. J. Med. 330, 387– 392 (1994). 45. U. Syversen, L. Nordsletten, J. A. Falch, J. E. Madsen, O. G. Nilsen and H. L. Waldum, Effect of lifelong nicotine inhalation on bone mass and mechanical properties in female rat femurs. Calcif. Tissue Int. 65, 246 – 9 (1999). 46. M. C. DeVernejoul, J. Bielakoff, M. Herve, J. Gueris, M. Hott, D. Modrowski, D. Kuntz, L. Miravet, and A. Ryckewaert, Evidence of defective osteoblastic function. A role for alcohol and tobacco consumption in osteoporosis in middle-aged men. Clin. Orthop. Rel. Res. 179, 1107 – 1115 (1983). 47. A. Leino, J. Jorvisalo, O. Impivaara, and M. Kaitsaari, Ovarian hormone status, life-style factors, and markers of bone metabolism in women aged 50 years. Calcif. Tissue Int. 54, 262–267 (1994). 48. M. A. Fang, P. J. Frost, A. Lida-Klein, and T. J. Hahn, Effects of nicotine on cellular function on UMR 106-01 osteoblast-like cells. Bone 12, 283 – 286 (1991). 49. L. G. Lenz, R. G. S. Galvin, and W. K. Ramp, Nicotine inhibits collagen synthesis and alkaline phosphatase activity but stimulates DNA synthesis in osteoblast-like cells. J. Bone Miner. Res. 5, S159 (1990). 50. R. J. S. Galvin, L. G. Lenz, and W. K. Ramp, Effects of smokeless tobacco extracts on DNA and collagen synthesis in fibroblasts and osteoblasts. J. Bone Miner. Res. 5, S159 (1990). 51. B. MacMahon, D. Trichopoulos, P. Cole, and J. Brown, Cigarette smoking and urinary estrogens. N. Engl. J. Med. 307, 1062–1065 (1982). 52. R. L. Barbieri, J. Gochberg, and K. J. Ryan, Nicotine, cotinine, and anabasine inhibit aromatase in human trophoblast in vitro. J. Clin. Invest. 77, 1727 – 1733 (1986). 53. J. Michnovicz, H. Naganuma, R. J. Hershcopf, H. L. Bradlow, and J. Fishman, Increased urinary catechol estrogen excretion in female smokers. Steroids 52, 69 – 83 (1988). 54. J. Michnovicz, R. J. Herschcopf, H. Naganuma, H. L. Bradlow, and J. Fishman, Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cigarette smoking. N. Engl. J. Med. 315, 1305 – 1309 (1986). 55. J. Jensen, C. Christiansen, and P. Rodbro, Cigarette smoking, serum estrogens, and bone loss during hormone-replacement therapy early after menopause. N. Engl. J. Med. 313, 973 – 975 (1985). 56. D. L. Cassidenti, A. G. Vijod, M. A. Vijod, F. Z. Stanczyk, and R. A. Lobo, Short-term effects of smoking on the pharmacokinetic profiles of micronized estradiol in postmenopausal women. Am. J. Obstet. Gynecol. 163, 1953 – 1960 (1990). 57. J. A. Cauley, J. P. Gutai, L. H. Kuller, D. LeDonne, and J. G. Powell, The epidemiology of serum sex hormones in postmenopausal women. Am. J. Epidemiol. 124, 752 – 761 (1989). 58. F. E. Crawford, D. J. Back, M. L’E. Orme, and A. M. Breckenridge. Oral contraceptive steroid plasma concentrations in smokers and non-smokers. Br. Med. J. 282, 1829 – 1830 (1981). 59. J. Geisler, I.H. Omsjo, S.I. Helle, D. Ekse, T. Silsand, and PE Lonning, Plasma ostrogen fractions in postmenopausal women receiving hormone replacement therapy: Influence of route of ad-

EGO SEEMAN

60.

61.

62. 63. 64.

65.

66.

67.

68.

69.

70. 71. 72. 73.

74.

75.

76.

77.

78.

79. 80.

81.

ministration and cigarette smoking. J. Endocrinol. 162, 265 – 270 (1999). E. Barrett-Connor and Kay-Tee Khaw, Cigarette smoking and increased endogenous estrogen levels in men. Am. J. Epidemiol. 126, 187 – 192 (1987). K-T. Khaw, S. Tazuke, and E. Barrett-Connor, Cigarette smoking and levels of adrenal androgens in postmenopausal women. N. Engl. J. Med. 318, 1705 – 1709 (1988). D. J. Handelsman, J. Conway, L. M. Boylan, and J. R. Turtle, Int. J. Androl. 7, 369 – 382 (1984). M. H. Briggs, Med. J. Aust. March 24 (1973). A. W. Meikle, X. H. Liu, G. N. Taylor, and J. D. Stringham, Nicotine and cotonine effects on 3 alpha hydroxysteroid dehydrogenase in canine prostate. Life Sci. 43, 1845 – 1850 (1998). O. Johnell, H. Kristenson, and I. Redlund-Johnell, Lower limb fractures and registration for Alcoholism. Scand J. Soc. Med. 13, 95–97 (1985). D. D. Bikle, Alcohol-induced bone disease. “Osteoporosis: Nutritional Aspects” (A. P. Simopoulous and C. Galli, eds.), Vol. 73, pp. 53 – 79. Basel, Karger, 1993. D. R. M. Lindsell, A. G. Wilson, and J. D. Maxwell, Fractures on the chest radiograph in detection of alcoholic liver disease. Br. Med. J. 285, 597 – 599 (1982). Y. Israel, H. Orrego, S. Holt, D. W. Macdonald, and H. E. Meema, Identification of alcohol abuse: Thoracic fractures on routine chest xrays as indicators of alcoholism. Alcohol. Clin. Exp. Res. 4, 420 – 422 (1980). T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, J. Roche, and S. Posen, Osteoporosis and skeletal fractures in chronic liver disease. Gut 31, 82 – 87 (1990). B. E. Nilsson, Conditions contributing to fracture of the femoral neck. Acta Chir. Scand. 136, 383 – 384 (1970). N. Dalen and A. L. Feldreich, Osteopenia in alcoholism. Clin. Orthop. 99, 201 – 202 (1974). C. M. Schnizler and L. Solomon, Bone changes after alcohol abuse. S. Afr. Med. J. 66, 730 (1984). I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Vaz Lopez, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. Cano Perez, A. Rapado, and C. Ribot. The variable incidence of hip fracture in southern Europe: The MEDOS study. Osteopor Int. 4, 253 – 263 (1994). D. ‘Hemenway, D. R. Azrael, E. B. Rimm, D. Feskanich, and W. C. Willett, Risk factors for wrist fracture: Effect of age, cigarettes, alcohol, body height, relative weight, and handedness on the risk for distal forearm fractures in men. Am. J. Epidemiol. 140, 361 –367 (1994). M. Hernandez-Avila, G. A. Colditz, M. J. Stampfer, B. Rosner, F. E. Speizer, and W. C. Willet, Caffeine, moderate alcohol intake, and risk of fractures of the hip and forearm in middle-aged women. Am. J. Clin. Nutr. 54, 157 – 163 (1991). D. T. Felson, D. P. Kiel, J. J. Anderson, and W. B. Kannel, Alcohol consumption and hip fractures: The Framingham Study. Am. J. Epidemiol. 128, 1102 – 1110 (1988). S. Holdrup, M. Gronbaek, A. Gottschau, J.B. Lauritzen, M. Schroll, Alcohol intake, beverage preference, and risk of hip fracture in men and women. Am. J. Epidemiol. 149, 993 – 1001 (1999). M. Navez Diaz, T.W. O’Neill and A.J. Silman, The influence of alcohol consumption on the risk of vertebral deformity. Osteopor Int., 7, 65 – 71 (1997). P. D. Saville, Changes in bone mass with age and alcoholism. J. Bone J. Surg. 47, 492 – 499 (1965). A. Diez, J. Puig, S. Serrano, M-L. Marinoso, J. Bosch, J. Marrugat. et al., Alcohol-induced bone disease in the absence of severe chronic liver damage. J. Bone Miner. Res. 9, 825 – 831 (1994). N. Dalen and B. Lamke, Bone mineral losses in alcoholics. Acta Orthop. Scand. 47, 469 – 471 (1976).

CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone 82. K. S. Chon, D. J. Sartoris, S. A. Brown, and P. Clopton, Alcoholismassociated spinal and femoral bone loss in abstinent male alcoholics, as measured by dual x-ray absorptiometry. Skel. Radiol. 21, 431 – 436 (1992). 83. A. Harding, J. Dunlap, S. Cook, A. Mattalino, F. Azar, M. O’Brien, and M. Kester, Osteoporotic correlates of alcoholism in young males. Orthopedics 11, 279–282 (1988). 84. K. Laitenen, C. Lamberg-Allardt, R. Tunninen, M. Harkonen, and M. Valimaki, Bone mineral density and abstention-induced changes in bone and mineral metabolism in noncirrhotic male alcoholics. Am. J. Med. 93, 642 – 650 (1992). 85. P. Peris, N. Guanabens, A. Pares, F. Pons, L. del Rio, A. Monegal, X. Suris, J. Caballeria, J. Rodes and J. Munoz-Gomez, Vertebral fractures and osteopenia in chronic alcoholic patients. Calcif, Tissue Int. 57, 111 – 4 (1995). 86. T. L. Holbrook and E. Barrett-Connor, A prospective study of alcohol consumption and bone mineral density. Br. Med. J. 306, 1506 – 1509 (1993). 87. M. A. Hansen, K. Overgaard, B. J. Riis, and C. Christiansen, Potential risk factor for development of postmenopausal osteoporosis examined over a 12-year period. Osteopor. Int. 1, 95 – 102 (1991). 88. R. M. Angus, P. N. Sambrook, and N. A. Pocock, Dietary intake and bone mineral density. Bone Miner. 4, 265 – 277 (1988). 89. K. Laitinen, M. Valimaki, and P. Keto, Bone mineral density measured by dual-energy x-ray absorptiometry in healty Finnish women. Calcif. Tissue Int. 48, 224 – 231 (1991). 90. J. L. Gonzales-Calvin, A. Garcia-Sanchez, V. Bellot, M. Munoz-Torres, E. Raya-Alvares, and D. Salvatierra-Rios, Mineral metabolism, osteoblastic function and bone mass in chronic alcoholism. Alcohol Alcoholism 28, 571 – 579 (1993). 91. C. W. Slemenda, S. L. Hui, C. Longcope, and C. Johnston, Jr., Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4, 737 – 741 (1989). 92. M. E. Arlot, M. Bonjean, and P. M. Chavassieux, Bone histology in adults with aseptic necrosis. J. Bone J. Surg. 65A, 1319 – 1327 (1983). 93. A. Diez, S. Serrano, J. Cucurull, L. L. Marinoso, J. Bosch, J. Puig, X. Nogues, J. Aubia, Acute effects of ethanol on mineral metabolism and trabecular bone in Sprague-Dawley rats. Calcif-Tissue Int. 61, 168 – 71 (1997). 94. D. Chappard, B. Plantard, M. Petitjeqan, C. Alexandre, and G. Riffat, Alcoholic cirrhosis and osteoporosis in men: A light and scanning electron microscopy study. J. Stud. Alcohol 52, 269 – 274 (1991). 95. C. M. Schnitzler, A. P. MacPhail, R. Shires, E. Schnaid, J. M. Mesquita, and H. J. Robson, Osteoporosis in African hemosiderosis: role of alcohol and iron. J. Bone Miner, Res. 9, 1865 – 1873 (1994). 96. R. G. Crilly, C. Anderson, D. Hogan, and L. Delaquerriere-Richardson, Bone histomorphometry, bone mass, and related parameters in alcoholic males. Calcif. Tissue Int. 43, 269 – 276 (1988). 97. T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, and S. Posen, Ethanol reduces bone formation and may cause osteoporosis. Am. J. Med. 86, 282 – 288 (1989). 98. J. Lindholm, T. Steiniche, E. Rasmussen, G. Thamsborg, I. Nielsen, H. Brockstedt-Rasmussen, T. Storm, L. Hyldstrup, and C. Schou, Bone disorder in men with chronic alcoholism: A reversible disease? J. Clin. Endocrinol. Metab. 73, 118 – 124 (1991). 99. H. W. Sampson and H. Spears, Osteopenia due to chronic alcohol consumption by young actively growing rats is not completely reversible. Alcohol Clin Exp. Res. 23, 324 – 327 (1999). 100. L. S. Kidder and R. T. Turne, Dietary ethanol does not accelerate bone loss in ovarietomized rats. Alcohol Clin. Exp. Res. 22, 2159 – 2164 (1998). 101. J. Jaouhari, F. Schiele, P. Pirollet, E. Lecomte, F. Paille, and Y. Artur, Concentration and hydroxyapatite binding capacity of plasma osteocalcin in chronic alcoholic men: Effect of a three-week withdrawal therapy. Bone Miner. 21, 171 – 178 (1993).

793 102. T. Pepersack, M. Fuss, J. Otero, P. Bergmann, J. Valsamis, and J. Corvilain, Longitudinal study of bone metabolism after ethanol withdrawal in alcoholic patients. J. Bone Miner. Res. 7, 383, 387 (1992). 103. V. R. Preedy, R. A. Sherwood, C. I. O. Akpoguma, and D. Black, The urinary excretion of the collagen degradation markers pyridinoline and deoxypyridinoline in an experimental rat model of alcoholic bone disease. Alcohol Alcohol. 26, 191–198 (1991). 104. H. Rico, J. A. Cabranes, J. Cabello, F. Gomez-Casresana, and E. R. Hernandez, Low serum osteocalcin in acutre alcohol intoxication: A direct toxic effect of alcohol on osteoblasts. Bone Miner. 2, 221 – 225 (1987). 105. M. M. Hurley, D. L. Martin, B. E. Kream, and L. G. Raisz, Effects of ethanol and acetaldehyde on collagen synthesis, prostaglandin release, and resorption of fetal rat bone in organ culture. Bone 11, 47 – 51 (1990). 106. N. Giuliani, G. Girasole, P. P. Vescovi, G. Passeri, M. Pedrazzoni, Ethanol and acetaldehyde inhibit the formation of early osteoblast progenitors in murine and human bone marrow cultures. Alcohol Clin. Exp. Res. 23, 381 – 385 (1999). 107. J. R. Farley, R. Fitzsimmons, A. K. Taylor, U. M. Jorch, and K-H. W. Lau, Direct effects of ethanol on bone resorptionand formation in vitro. Arch. Biochem. Biophys 238, 305 – 314 (1985). 108. K. E. Friday and G. A. Howard, Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40, 562 – 565 (1991). 109. R. C-Y. Cheung, C. Gray, A. Boyde, and S. J. Jones. Effects of ethanol on bone cells in vitro resulting in increased resorption. Bone 16, 143 – 147 (1995). 110. G.Tisman and V. Herbert, In vitro myelosuppression and immunosupression by ethanol. J. Clin. Invest. 52, 1410 – 1414 (1973). 111. R. C. Meagher, F. Sieber, and J. L. Spivak, Suppression of hematopoietic-progenitor-cell proliferation by ethanol and acetaldehyde. N. Engl. J. Med. 307, 845 – 849 (1982). 112. J. Balsinde, A. Schueller, and E. Diez, The interaction of ethanol and exogenous arachidonic acid in the generation of extracellular messengers by mouse peritoneal macrophages. Biochem. Biophys. Acta 970, 83 – 89 (1988). 113. N. E. Gilhus and R. Matre, In vitro effect of ethanol on subpopulations of human blood mononuclear cells. Int. Arch. Allergy Appl. Immunol. 68, 382 – 386 (1982). 114. A. Zuiable, E. Wiener, and S. N. Wickramasinghe, In vitro effects of ethanol on the phagocytic and microbial killing activities of normal human monocytes and monocyte-derived macrophages. Clin. Lab. Haematol. 14, 137 – 147 (1992). 115. A. Castro, D. L. Lefkowitz, and S. S. Lefkowitz, The effects of alcohol on murine macrophage function. Life Sci. 52, 1585 – 1593 (1993). 116. S. Yamada, S. Mochida, A. Ohno, K. Hirata, I. Ogata, Y. Ohta, and K. Fujiwara, Evidence for enhanced secretory function of hepatic macrophages after longterm ethanol feeding in rats. Liver 11, 220 – 224 (1991). 117. D. H. Van Thiel, J. S. Gavaler, R. Lester, and M. D. Goodman, Alcohol-induced testicular atrophy. An experimental model for hypogonadism occurring in chronic alcoholic men. Gastroenterology 69, 326 – 332 (1975). 118. D. H. Van Thiel, R. Lester, and R. J. Sherins, Hypogonadism in alcoholic liver disease: Evidence for a double defect. Gastroenterology 67, 1188 – 1199 (1974). 119. J. H. Mendelson, N. K. Mello, and J. Ellingboe, Effects of acute alcohol intake on pituitary – gonadal hormones in normal human males. J. Pharmacol. Exp. Ther. 202, 676 – 682 (1977). 120. E. Rubin, C. S. Lieber, K. Altman, G. G. Gordon, and A. L. Southren, Testosterone metabolism in the liver. Science 191, 563 – 564 (1976).

794 121. G. G. Gordon, K. Altman, A. L. Southren, E. Rubin, and C. S. Lieber, Effect of alcohol (ethanol) administration on sex-hormone metabolism in normal men. N. Engl. J. Med. 295, 793 – 797 (1976). 122. G. G. Gordon, A. L. Southren, J. Vittek, and C. S. Lieber, The effect of alcohol ingestion on heptic aromatase activity and plasma steroid hormones in the rat. Metabolism 28, 20 – 24 (1979). 123. D. H. Van Thiel, J. Gavaler, and R. Lester, Ethanol inhibition of vitamin A metabolism in the testes: Possible mechanism for sterility in alcoholics. Science 186, 941 – 942 (1974). 124. J. L. Leathem, In “The Testes” (A.D. Johnson, W. R. Gomes, and N. L. Vandermark, eds.) Vol. 3, pp. 188 – 190. Academic Press, New York, 1970.

EGO SEEMAN 125. E. S. Ginsburg, N. K. Mello, J. H. Mendelson, R. L. Barbieri, S. K. Teoh, M. Rothman, X. Gao, and J. W. Sholar, Effects of alcohol ingestion on estrogens in postmenopausal women. JAMA 276, 1747 – 51 (1996). 126. S. Feitberg, S. Epstein, F. Ismail, and C.D’Amanda, Deranged bone mineral metabolism in chronic alcoholism. Metabolism 36, 322 – 326 (1987). 127. J. P. Knochel, G. L. Bilbrey, T. J. Fuller, and N. W. Carter, The muscle cell in chronic alcoholism: The possible role of phosphate depletion in alcoholic myopathy. Ann. N.Y. Acad. Sci. 252, 274 – 286 (1975). 128. V. R. Preedy and T. J. Peter, Alcohol and muscle disease. J. R. Soc. Med. 87, 188–189 (1994).

795

CHAPTER 32

Falls as Risk Factors for Fractures CHAPTER 32

Falls as Risk Factors for Fractures ANN V. SCHWARTZ ELIZABETH CAPEZUTI JEANE ANN GRISSO

Department of Epidemiology and Biostatistics, University of California, San Francisco, School of Medicine, San Francisco, California 94143 Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, Georgia 30322 Center for Clinical Epidemiology and Biostatistics, Division of General Internal Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104

IV. Prevention of Falls and Fall-Related Fractures V. Summary and Directions for Future Research References

I. Introduction II. Risk Factors for Falls III. Risk Factors for a Fall-Related Injury

I. INTRODUCTION Falls are a common event among the elderly. Population-based surveys and prospective studies indicate that around 30% of community-dwelling elderly fall one or more times each year. Of those who report falling, about half report more than one fall during the year [1 – 10]. In residential institutions, the proportion of those who fall is higher, about 40 – 50% [11 – 13]. Fall-related injuries are the leading cause of mortality due to unintentional injuries among adults 65 and older in the United States [14]. In 1995, 11,057 deaths among the elderly were attributed to falls [14]. This figure, which is derived from death certificates, is likely to underestimate the extent to which falls play a role in fatalities. About one-third of all deaths from falls occur among those 85 years of age or older [15]. Falls are also the leading cause of nonfatal injuries among the elderly [16]. An estimated 5 – 10% of those over age 75 visit a hospital emergency department each year for

OSTEOPOROSIS, SECOND EDITION VOLUME 1

treatment of a fall-related injury; about one-third are subsequently hospitalized [17 – 20]. Fall-related injuries include fractures and other serious injuries (dislocated joints, subdural hematomas, and lacerations requiring sutures) and minor injuries (bruises, abrasions, certain sprains, and other soft tissue injuries). Fractures alone account for 30 – 40% of fall injuries among the elderly [19,21 – 23]. One-fourth to one-third of these fractures are of the hip, resulting in over 250,000 hospital admissions for hip fracture each year in the United States [24]. Hip fractures in particular are associated with increased mortality and disability. Of those who can walk without assistance at the time of hip fracture, half cannot resume this level of independence afterward [25]. Several types of fractures among the elderly are usually due to a fall. It is well documented that over 90% of hip fractures in the elderly occur as a result of a fall. Other fractures that are largely due to falls include those of the distal forearm (Colles’ or wrist fracture), pelvis, humerus, rib, leg, hand, patella, ankle, elbow, and face [26 – 30].

795

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

796

SCHWARTZ, CAPEZUTI, AND GRISSO

Although fall-related injuries can be devastating, most falls do not result in serious injury. In prospective studies of community-dwelling elderly, about 10% of falls result in an injury requiring medical attention. About 3 – 5% of falls in the elderly result in a fracture; about 1% result in hip fracture [23]. However, even falls that do not result in serious injury may have long-term health consequences for the elderly. Fear of falling can lead to decreased independence and mobility [31]. Multiple falls are often a reason for admission to a nursing home [32]. The percentage of the elderly who fall increases with age, more steeply after age 75. Women are somewhat more likely to fall than men until age 85 when the percentages who report a fall are nearly equal [33]. These estimates of fall frequency are based on studies among mainly non-Hispanic Caucasian women. Investigation of fall rates in other groups is limited. Mexican-American women appear to fall at a similar rate [34]. A study of older Japanese-American women found that their age-adjusted fall rate was about half the rates reported for white women [35]. Two prospective studies indicate that fall frequency may be somewhat lower among black women than among white women [5,8].

II. RISK FACTORS FOR FALLS A. Introduction Most falls in the elderly are probably due to both intrinsic (host) and extrinsic (environmental) factors. Although there has been a great deal of progress in the identification of intrinsic risk factors, investigation of extrinsic factors remains more limited. Environmental factors are thought to be particularly important in falls among the more active elderly, whereas intrinsic factors may play more of a role among the frail elderly [33]. In studies of risk factors for falls, a fall is usually defined as “an event that results in a person coming to rest inadvertently on the ground or other lower level.” Falls due to an overwhelming force or event, such as a motor vehicle accident or loss of consciousness, are usually excluded [32].

B. Intrinsic (Host) Factors

TABLE 1 Intrinsic Risk Factors for Falls among the Elderlya Risk factor

Evidence for association

Demographic characteristics Older age



Gender, Women



Race, White



Functional level ADL/IADL



Cane/Walker use



History of falls



Gait, balance, strength Walking speed



Lower extremity strength



Upper extremity strength



Postural sway



Impaired reflexes



Sensory Vision



Lower extremity sensory perception



Chronic illnesses Heart disease

/

Parkinson’s disease



Other neuromuscular disease



Stroke



Urinary incontinence



Arthritis



Acute illness



Medications, alcohol No. medications used



Hypnotics



Sedatives



Antipsychotics



Antidepressants



Antiparkinson drugs



Cardiac

/

Diuretics

/

Antihypertensives

/

Alcohol

/

Mental status Cognitive impairment Depression

 

 , strong; , moderate; /, inconsistent.

a

Falls among the elderly are generally associated with frailty and poor health. Table 1 summarizes major intrinsic factors that have been identified as risk factors for falls. A person’s chances of falling increase with age but the functional level appears to be a better predictor of falls than age alone [5,8,36]. Limitations in both activities of daily living (ADL) and instrumental activities of daily living (IADL)

are associated with an increased risk of falling [5,8,13,36 – 38]. Functional or performance-based measures indicating deficits in balance, gait, and strength are strongly associated with falls [2,4,5,8,9,13,36,38 – 41], with evidence that fall risk increases with the number of disabilities [38]. These

CHAPTER 32

797

Falls as Risk Factors for Fractures

measures include tests such as standing up from a chair, walking speed, step length, and postural sway. Reduced grip and lower extremity strength are also correlated with falls [1,2,5,13,36,39,42,43]. Maintaining postural control and avoiding environmental obstacles depend on proprioceptive, vestibular, and visual input translated into appropriate motor responses [44]. Both impaired visual acuity [5,8,12,37,38,45] and diminished sensory function in the lower extremities [8,13,42,46] have been shown to be associated with falls in many, but not all [2,4,5,13,36,47], studies. It may be that visual or somatosensory impairment is most important for those who have other impairments, such as gait abnormalities or muscle weakness [23]. The ability to translate perceptions into appropriate motor responses diminishes with age [48]. Laboratory-based studies indicate that older adults have more difficulty maintaining balance when their attention is divided [49]. However, detailed measurements of central neurological processes have not been included in larger epidemiological studies of risk factors for falls. Reaction time has been assessed in a few studies with varying results [5,50,51]. Global cognitive impairment is associated with an increased risk of falls [2,4,7,8,37,52]. Impaired cognitive functioning is a significant predictor of fall-related injury in both nursing home residents [22,53] and community-residing older adults [54,55]. Hypothesized causal connections include a neurologically based reduction in the ability to maintain balance and behavioral changes, such as wandering, associated with impaired judgement [52,56]. Depression is also found in association with increased falls [5,8,37,40,57]. Depression is associated with disability and poor health and may lead to decreased attention to environmental hazards. In addition, antidepressant medications are associated with falls [8,58 – 60]. The presence of certain chronic medical conditions has been found to increase the risk of a fall, including Parkinson’s disease [2,5,12], urinary incontinence [5,8,13], dementia [12,52], history of stroke [2,7,61], and arthritis [1,2,5,58,62]. A period of acute illness may also lead to a greater risk of a fall [8,38,63]. Use of a greater number of medications is associated with falls, although it has been difficult to distinguish an effect of medications from the chronic condition that is being treated [64]. Specific medications may increase the risk of falls through depressed psychomotor function, reduced alertness, greater fatigue, or postural hypotension [58]. For psychotropic medication [65,66], an increase in the risk of falls seems to be well established. A recent meta-analysis found that among 14 cardiac and analgesic drugs or drug groups, only diuretics, digoxin, and type IA antiarrhythmic agents significantly increased the risk of falling [64]. Alcohol use has been postulated to increase the risk of falls and has been evaluated in a range of studies [67].

However, most studies, including community-based prospective studies, have not found an association [2,5,7,8,68].

C. Extrinsic Factors Environmental or extrinsic risk factors include situational or activity related factors as well as environmental hazards. Most falls occur when older persons are performing their usual activities, such as rising from a chair or ambulating [8,9,11,38,69 – 71]. The relationship between environmental hazards and falls for community-residing older adults is most frequently ascertained by self-report [5,72,73] with subjects implicating environmental factors in one-third to one-half of their falls [32,74,75]. Home hazards include inappropriately placed furniture or objects, scatter rugs, and slippery surfaces [74,76,77]. Environmental hazards outside the home include irregular sidewalk surfaces, ill-repaired stairs, and traffic lights that do not allow sufficient time to cross the street [74]. Table 2 lists potential hazards that can be found in the home. Observational studies have not identified strong associations between environmental hazards and falls, but this may be the result of inherent difficulties in study design [5,8,78 – 80]. Assessing environmental influences is difficult because (i) environmental conditions may change over time, (ii) only a fraction of falls are likely to occur in any particular location, and (iii) there are no standardized methods developed to use in attributing a fall to a specific environmental hazard. Environmental hazards may be more important for healthy and mobile elders [81]. In a prospective study of falls in 325 community-residing older adults, Northridge et al. [82] found an association between falls and the home environment for vigorous but not frail participants. Certain functional characteristics (notably arthritis and poor depth perception) increased the likelihood of falls in the presence of environmental hazards [83]. Although observational studies have not provided strong evidence for a role of environmental hazards in falls, a prevention trial focused on home hazard reduction demonstrated some success in decreasing falls (see Section IV,A,5) [76]. For hospitalized and institutionalized elderly, traveling to the bathroom is the most frequently reported activity-related risk for falls [84 – 88]. In the hospital, disorienting effects of an unfamiliar environment are considered a major causative factor for increased fall risk. Similarly, for nursing home residents, high fall rates are observed most often in the first week after admission or transfer to a different unit in the same nursing home [70,89]. Similar to the community setting, general environmental factors contributing to falls in nursing homes have not been well studied. One review of the falls literature proposes an

798

SCHWARTZ, CAPEZUTI, AND GRISSO

TABLE 2

Environmental Home Hazards

Ground surfaces Throw rugs Loose carpets (not taped or tacked down) Slippery floors Cords and wires on the floor Low-lying objects, e.g., toys Stairs with rugs or in poor repair Furniture

injury [22]. Another study examining restraint use in confused ambulatory nursing home residents found an increased risk of falling among restrained residents compared to nonrestrained residents [93]. Accurate information on whether restraints were being used at the time of a fall is difficult to obtain [96,97] so studies to date have not considered whether restraints have a direct effect on falls and injuries. Most importantly, studies examining the effect of reducing restraint use in nursing homes report no significant increase in falls or fall-related injuries [98 – 101].

Clutter, especially if there is insufficient space for unobstructed mobility Unstable furniture Low-lying furniture, e.g., coffee table

III. RISK FACTORS FOR A FALL-RELATED INJURY

Low chairs without armrest support or seat backs Beds that are too high or too low Cabinets that are either too high or too low Lighting Glare from unshielded windows or lamps or highly polished floors Low or dim compounded by dark-colored walls Absence of night-lights Bathroom Low toilet seats and/or no secure grab bars

Early research on the identification of fall risk factors has been focused on all falls, despite the fact that the majority of falls do not result in injury. Some researchers suggest that fall-related serious injury, not falls per se, is the significant outcome measure and, thus, research and preventive actions should focus on risk factors for injurious falls [70,102,103]. However, it is plausible that efforts to reduce falls in general would also reduce injuries [104]. The most effective approach remains to be clearly established.

Absence of slip-resistant, strongly secured grab bars Absence of nonslip surfaces or assistive devices (e.g., tub chairs in bathtub) Door jambs Other Poorly maintained walking aids and equipment Improper shoes (not slip-resistant, high-heeled, too large)

association between long trouser legs and ill-fitting shoes and increased fall risk [90]. Other environmental factors are generally regarded as less problematic for nursing home residents, as nursing home regulations mandate specific environmental adaptations meant to prevent falls. Physical restraints, including side rails, have been considered an environmental risk factor for falls due to a significant incidence of falls in restrained elders [22,91 – 94]. While restraint use was traditionally viewed as a preventive strategy, observational studies have indicated that restraint use does not reduce fall risk and may instead contribute indirectly to an increase in falls and injuries. Physical restraint use, by immobilizing older persons, can result in reduced muscle mass, strength, joint flexibility, and vasomotor stability [22,95]. In a prospective study, nursing home residents who were restrained for any amount of time during followup had an increased risk of fall-related serious injury even after controlling for risk factors highly associated with

A. Differentiating Risk Factors for Falls from Risk Factors for Injurious Falls Table 3 lists risk factors for injurious falls. Many of the risk factors are the same as those in Table 1 for falls in general, including slow gait, lower and upper extremity weakness, and use of psychotropic medications. However, there are some risk factors specific for injurious falls. Circumstances of the fall, including the height of the fall and orientation of the faller, influence the risk of injury. Certain host factors such as low bone density increase the risk of injury in a fall. Recurrent fallers are, not surprisingly, more likely to sustain a fall-related injury [22,36,105 – 108]. In a population-based case-control study, a strong association was found between reported numbers of falls in the past year and risk of hip fracture, especially among men [109]. Lipsitz et al. [36] contended that “because recurrent fallers are most likely to experience injury from repeated episodes, they constitute an important group to target for diagnostic and preventive efforts.” There is, however, also evidence that factors beyond the greater frequency of falls among the elderly are associated with a greater risk of fall injuries. Speechley and Tinetti found that although “frail” elderly fell more often, they had a lower rate of serious injury per fall than the “vigorous” elderly, leading to a similar rate of serious injury due to

CHAPTER 32

799

Falls as Risk Factors for Fractures

TABLE 3

Risk Factors for Injurious Falls

Risk factor

Evidence for association

Orientation of fall Slow gait



Falling sideways



Falling on/near hip



Falling while turning



Falling while reaching



Falls from a standing height



Protective responses Neurological disease



Upper extremity weakness



Lower extremity weakness



Use of ambulatory aids



Syncopal fall or “drop attack”



Long-acting hypnotic-anxiolytic (benzodiazepines) use



Tricyclic antidepressant use



Antipsychotic use



Impaired cognition



Physical restraint-induced deconditioning



Local shock absorbers Lower body weight/skinfold thickness



Weakness of hip abductors



Falls on hard surface



Bone strength Low bone mineral density/osteoporosis



Caucasians



Women



falls in these two groups [81]. Older women who experience a rapid increase in falls appear to be at higher risk of fracture, independent of the total number of falls (A. Schwartz, personal communication). Cummings and Nevitt [110] noted that the risk of hip fracture among elderly white women increases very steeply with age, whereas the risk of falling rises only moderately.

B. Risk Factors for Injurious Falls 1. HIP FRACTURE MODEL The most comprehensive model for causes of hip fracture identifies and categorizes risk factors according to orientation of the fall, protective responses, local shock absorbers, and bone strength [110]. It is hypothesized “that risk factors for each step in the sequence will

interact in a multiplicative fashion” [110]. Although hip fractures (as opposed to other injuries) are the focus of this model, it is appropriate since the most devastating consequence of injurious falls is hip fracture [108] (see also Chapter 19). 2. ORIENTATION OF THE FALL The points of impact of a fall influence the type and extent of injury. The elderly may be more likely to land on the hip in a fall, thus increasing their risk of hip fracture, whereas, most injurious falls in young and middle-aged adults result in wrist fractures. This is most likely a result of the more rapid gait speed in the young in comparison to that in older persons, which tends to propel the faller forward, resulting in principal impact on the wrist (Fig. 1). In contrast, injurious falls in older persons generally occur while the individual is standing still, during transfer, or while walking slowly with little forward momentum [2,110]. Risk of hip fracture is reported as significantly increased in women falling sideways, straight down, or those who fall on or near the hip; falling backward may be associated with a decreased risk of hip fracture [111 – 113]. Falling during a displacing activity, such as turning, is more likely to result in a serious injury than falling while walking in one direction [81,110]. Certain circumstances are more likely to increase loading forces on the proximal femur, thus resulting in hip fracture. Falls from a standing height or higher are more likely to result in injury [114,115]. Similarly, Tinetti’s study of nursing home residents reported that those who fell while rising from a chair were unlikely to be injured [102]. The risk of fall-related injury in community-residing elderly is higher during stair climbing and when turning around or reaching for objects [55,115]. Risk of hip fracture is increased in taller subjects [112], presumably due to falling from a greater height as well as longer hip axis length [116]. 3. PROTECTIVE RESPONSES Several reflexes and postural responses are initiated during a fall, which potentially prevent or reduce injury. These responses are protective if they can “change the orientation of the faller or reduce the energy of a fall” [110]. The effectiveness of reflex actions depends on the speed of execution and the strength of the muscles initiating the protective movement [110]. Research links impaired protective responses with an increased likelihood of hip fracture. These factors include preexisting neurologic conditions, both upper and lower limb dysfunction or weakness, and use of ambulatory aids [22,50,54,55,102,106,112,117 – 121]. Fallers who initiated protective responses, such as grabbing or hitting an object before landing and those who

800 landed on their hand, were less likely to sustain a hip fracture than those who did not [112]. During syncopal falls, the faller is unable to initiate protective responses, which contributes to a greater likelihood of fractures [19,55]. Likewise, during a “drop attack” a person is unable to recognize the sensation of falling and thus will not display an appropriate protective response [106]. The sedating effect of certain drugs can impair protective responses. Thus, in addition to increasing the risk of falls generally, long-acting hypnotics – anxiolytics (including benzodiazepines), tricyclic antidepressants, and antipsychotics may increase the risk of a fall-related injury [65]. 4. LOCAL SHOCK ABSORBERS The impact of falls can be potentially absorbed by surrounding soft tissue (skin, fat, and muscles), thereby reducing the likelihood of an injury. A lower prevalence of injury occurs in fallers with more skinfold thickness [106] or overall higher body weight [52,54,112,122 – 124]. The apparent protective effect of higher body weight on reducing the risk of hip fracture may be due to increased bone mineral density, local shock-absorbing capacity of muscle and fat, or both. The strength of hip abductors, the largest group of muscles surrounding the hip, is postulated to affect absorption on impact [110]. The risk of fall-related serious injury in older adults is increased when falls occur on a hard surface [55,112,123]. Grisso et al. [125] reported falling on a hard surface as significantly associated with hip fracture; however, this was more likely in women 55 to 74 years of age than in those 75 or older. Flooring specifically designed to reduce the impact of a fall may reduce the likelihood of hip fracture [126]. 5. BONE STRENGTH Bone strength is considered the “last defense” against hip fractures. A fracture will occur if the residual energy of the fall applied to the proximal femur exceeds the strength of the bone [110]. Bone mineral density is the usual measure of bone strength, although other factors, such as geometry and microarchitecture of the bone, may also be important [116]. Nevitt and colleagues [112] found substantial increases in hip and wrist fractures in those with lower bone density, including a nearly sevenfold increase in the risk of a hip fracture when a woman with low bone density falls on the hip. Numerous studies indicate a direct relationship between fracture risk and osteoporosis [27,127,128]. Race, size, and gender, indirect indicators of osteoporosis, also are frequently cited [19,22,27,55,120,129]. The Cummings et al. model provides a useful framework for understanding how numerous risk factors may contribute to injurious falls [110].

SCHWARTZ, CAPEZUTI, AND GRISSO

IV. PREVENTION OF FALLS AND FALL-RELATED FRACTURES A. Prevention of Falls 1. MULTIFACTORIAL INTERVENTIONS Randomized trials have demonstrated the possibility of reducing the rate of falls among the elderly. The most promising results to date have been obtained with multifactorial interventions [130]. Tinetti et al. [131] randomized elderly members of an HMO to either interventions for a range of identified problems or social visits with research staff. The interventions targeted postural hypotension, use of sedative – hypnotics, multiple prescription drugs, environmental hazards, balance and gait impairments, and decreased muscle strength [131]. Results showed a significantly lower fall rate among those persons receiving the intervention (35%) than among the social visit group (47%) [132]. Targeted risk factors were also significantly reduced, with the intervention group showing decreases in the use of multiple medications as well as improvement in measures of balance, gait, and functional impairment [133]. In a later study, Close and co-workers [134] followed communitydwelling older patients who had presented to an emergency room (ER) with a fall. They demonstrated a reduced risk of falling after the ER visit (OR  0.39; 95% CI 0.23 – 0.66) in the group that received a detailed medical and physical therapy assessment with referral as indicated compared to usual care. The benefits of a comprehensive individual assessment were also demonstrated in a nursing home setting by Ray et al. [135], who found a 19.1% (95% CI 2.4 – 35.8%) reduction in the proportion of re-current fallers at intervention facilities. 2. EXERCISE Exercise alone has been shown to reduce falls in three studies using different exercise programs: Tai Chi [136], strength and endurance training [137], and strength and balance exercises combined with walking [138]. A metaanalysis of FICSIT (Frailty and Injuries: Cooperative Studies of Intervention Techniques) studies found that the exercise component of the tested interventions reduced falls by about 10% [139]. In other studies, however, exercise has not been shown to be effective in reducing falls [140 – 143]. To date, clinical trials of exercise with falls as an outcome have not targeted those over 75 years of age. However, evidence shows that exercise programs are effective in improving strength and balance, and thus reducing the risk factors for falls, among older individuals. In the Boston FICSIT study, Fiatarone and co-workers [144] demonstrated, in a nursing home population whose mean age was 87 years, that a progressive resistance exercise training

CHAPTER 32

801

Falls as Risk Factors for Fractures

intervention significantly increased muscle strength, gait velocity, the ability to climb stairs, and the general level of physical activity. 3. MEDICATIONS Campbell and colleagues [145] conducted a trial of withdrawal of psychotropic medication among communitydwelling elderly and reported a 66% reduction in fall risk for the intervention group. Some multifactorial interventions have also included a component designed to reduce medication use. Reducing inappropriate drug prescribing and prescribing of multiple medications may be a particularly promising way of preventing falls for several reasons. First, the problem is extremely common. In a recent analysis of a national survey of community-dwelling persons 65 years or older, 25 to 32% of those surveyed were taking potentially inappropriate medications [146]. This means that up to 9 million older persons may be at risk of adverse effects of drugs, including falls. Second, it is relatively easy to identify individuals who are taking these medications through clinical practices and nursing homes. Extensive screening procedures are not required. Third, interventions targeting physician prescribing patterns could potentially affect large numbers of older persons, given that most older persons have primary care physicians whom they visit regularly. However, changing physician prescribing patterns is not as simple as might be supposed. Studies have documented that simply providing information is not adequate [147]. Promising methods include training “opinion leaders” who then influence their colleagues, providing feedback to individual physicians, and instituting administrative interventions that include incentives or disincentives. Two controlled trials have demonstrated substantial reductions in the use of psychoactive drugs. These studies, however, were conducted in nursing homes where it may be easier to affect prescribing patterns. Ray et al. [148] reported that an intervention including in-service education of administrative, physician, and nursing staff resulted in a 72% reduction in the days of antipsychotic use compared with only a 13% reduction in the control nursing homes (P  0.001). Avorn et al. [149] conducted a randomized controlled trial of an educational intervention in which a clinical pharmacist conducted multiple interactive visits with the physicians who commonly prescribed psychoactive drugs. Psychoactive drug use declined by 27% in experimental nursing homes compared with 8% in control homes (P  0.02) [149]. There has also been documentation that regulations or administrative procedures can result in decreased antipsychotic drug use [150,151]. Shorr and co-workers [150] reported that antipsychotic drug use decreased by 27% (P  0.001) after implementation of the Nursing Home Reform Act [Omnibus Budget Reconciliation Act (OBRA) of 1987]. This federal

legislation was designed in part to reduce unnecessary antipsychotic drugs through establishing clinical practice guidelines and follow-up inspections of Medicaid/ Medicare certified nursing homes. In addition, permanent withdrawal from certain medications may be difficult to achieve. In the clinical trial of withdrawal from psychotropic medications reported by Campbell et al. [145], the authors noted that about half of the participants who withdrew from psychotropics had restarted medication within a month of the trial’s completion. 4. VISION Visual impairment has been found to be a risk factor for falls and hip fracture [8,45,120,127,152,153]. Although correction of vision problems has been included in some multifactorial interventions [134,154,155], it has not been the specific focus of a prevention trial to date. Numerous age-related physiological changes occur in the eye that commonly affect vision in older persons, including cataracts, glaucoma, diabetic retinopathy, and macular degeneration [8]. Many of these conditions, if detected early, can be treated and improved by measures ranging from providing appropriate eyewear to surgery for cataracts. In addition, it is also important to develop interventions that help older persons adapt to functional limitations resulting from decreased vision, such as visual rehabilitation programs, or environmental adaptations, such as adequate lighting, placement of bright-colored tape on steps, and other home modifications. Studies are needed to evaluate the impact of interventions to improve visual function on the risk of falls. 5. EXTRINSIC FACTORS Although fall prevention programs usually advocate environmental modifications, few studies have evaluated the impact of environmental hazard reduction on fall occurrence. Hornbrook et al. [76] conducted a randomized trial of a falls prevention program that included home safety as part of the intervention. The study demonstrated a reduction in the proportion of fallers of about 15% (P  0.05), although the number of falls was only reduced by 7%. Home safety hazards were identified using a standard hazards protocol. Participants were provided information on procedures for obtaining technical and financial assistance in making safety repairs and were invited to attend fall prevention classes. Sixty-two percent of intervention households received financial assistance to make safety repairs, and some modifications were achieved in every intervention household (including removal of throw rugs, objects in pathways, installation of night-lights, and installation of bathtub nonskid mats or strips). Schwarz and colleagues [156] conducted a controlled trial in low-income

802

SCHWARTZ, CAPEZUTI, AND GRISSO

innercity neighborhoods that reported a statistically significant reduction in home hazards. However, no reduction in fall injury rates was observed through ongoing surveillance of emergency department records (D. Schwarz, personal communication).

B. Prevention of Fall-Related Injuries Although preventing fall-related injuries and subsequent disability is most important, to date no published studies have addressed the impact of preventive measures on injurious falls as a primary outcome. One of the reasons is that serious injuries are less common, and thus large sample sizes would be required to detect clinically important effects. Nevertheless, many interventions designed to prevent fall occurrence may also be effective in preventing fall-related injuries. A pooled estimate from five multifactorial interventions indicated a possible reduction of 30% in falls requiring medical care (OR 0.70; 95% CI 0.47 – 1.04) [130]. One of the most disabling of fall outcomes is the occurrence of a fracture. The probability that a given fall results in fracture is, in part, dependent on the level of bone mass and perhaps the quality of the bone structure. Several medications have been demonstrated to prevent loss of bone mass and reduce fracture risk, including alendronate [157,158], estrogen [159], and raloxifene [160] (see also Chapters 69 and 72). Clinical trials have demonstrated that calcium supplementation is somewhat effective in preserving bone mass, particularly among older women with low calcium intakes (no greater than 400 mg/day) [161]. Vitamin D supplementation may also be a valuable preventive measure. A randomized double-blind, placebo-controlled clinical trial of vitamin D and calcium supplements demonstrated a 26% reduction in the incidence of nonvertebral fractures [162] (see also Chapters 67 and 68). Extrinsic factors may also be effective in reducing the likelihood that a fall will result in serious injury. In laboratory simulations of a fall on the hip, polyurethane foam pads can reduce the peak impact force by nearly 20%. Although the peak femoral force remains greater than the force needed to cause a femoral fracture for the currently available hip pads, at least one trial has demonstrated a 53% reduction in hip fractures among persons randomized to receive external hip protectors [163]. Of note, during this trial no hip fracture was sustained by any of the six residents wearing a hip protector at the time of a fall, and none of the eight residents in the intervention group who had a hip fracture was wearing a hip protector at the time of the fracture. However, the authors found that only 24% of residents given hip protectors wore them regularly. Improvements in the design of hip protectors may increase compliance as well as effectiveness [164].

FIGURE 1

(A) fall occurring while standing still, walking slowly, or slowly descending a step has little forward momentum. With little forward momentum, the principal point of impact will be near the hip. (B) A fall occurring during rapid walking has enough forward momentum to carry the faller onto the hands or knees instead of the hip.

Other environmental modifications are also promising. Experts in biomechanics and engineering have developed a new kind of flexible flooring that may reduce the force of the impact from a fall by as much as 30% [165]. Beds that can be adjusted electrically from 23 inches down to 7 inches above the floor have become readily available [166]. A pilot study demonstrated less risk of falls, recurrent falls, and injuries among frail nursing home residents using a low-height bed (10 inches above the floor) compared to residents using a standard bed (21 inches above the floor) [167]. In addition, traditional environmental modifications have not been evaluated but could assist in protective responses (such as grab bars in bathrooms or second hand rails on staircases) or diminish the force of the impact of a fall (such as wall-to-wall carpeting or flexible flooring). Finally, interventions should be considered on a community or governmental level. Although investigators and health care providers had been trying to reduce the use of psychotropic medications in nursing homes for years, the implementation of the federal regulations promulgated by the Nursing Home Reform Act (OBRA-87) resulted in significant reductions in a very short period of time. In addition, implementation of the OBRA-87 legislation reduced the use of physical restraints without an increase in fall-related injuries [101]. Similarly, instituting safe access

CHAPTER 32

Falls as Risk Factors for Fractures

for older persons and the disabled in public areas may be effective in preventing falls and fall-related injuries that occur outside the home.

V. SUMMARY AND DIRECTIONS FOR FUTURE RESEARCH In recent years, we have learned a great deal about the importance of falls in relation to the risk of osteoporotic fractures. There have been significant advances in knowledge concerning which fall risk factors are important determinants of osteoporotic fractures. We have learned that multiple factors affect fall risk and that the most effective interventions will probably need to address several types of risk factors. Although we have made significant progress in understanding the pathogenesis of falls and fall-related injuries, there is much still to be learned. Future research to test the effectiveness of various interventions in reducing fall injuries is especially needed.

References 1. A. J. Blake, K. Morgan, and M. J. Bendall, Falls by elderly people at home: Prevalence and associated factors. Age Ageing 17, 365 – 372 (1988). 2. A. J. Campbell, M. J. Borrie, and G. F. Spears, Risk factors for falls in a community-based prospective study of people 70 years and older. J.Gerontol. 44, M112 – M117 (1989). 3. W. A. Hale, M. J. Delaney, and W. C. McGaghie, Characteristics and predictors of falls in elderly patients. J. Fam. Pract. 34, 577 – 581 (1992). 4. S. R. Lord, R. D. Clark, and I. W. Webster, Physiological factors associated with falls in an elderly population. J. Am. Geriatr. Soc. 39, 1194 – 1200 (1991). 5. M. C. Nevitt, S. T. Cummings, and S. Kidd, Risk factors for recurrent nonsyncopal falls: A prospective study. JAMA 261, 2663 – 2668 (1989). 6. B. C. Perry, Falls among the elderly living in high-rise apartments. J. Fam. Pract. 14, 1069 – 1073 (1982). 7. D. Prudham, Factors associated with falls in the elderly: A community study. Age Ageing 10, 141 – 146 (1981). 8. M. E. Tinetti, M. Speechley, and S. F. Ginter, Risk factors for falls among elderly persons living in the community. N. Engl. J. Med. 319, 1701 – 1707 (1988). 9. A. K. Topper, B. E. Maki, and P. J. Holliday, Are activity-based assessments of balance and gait in the elderly predictive of risk of falling and/or type of fall? J. Am. Geriatr. Soc. 41, 479 – 487 (1993). 10. C. Wickham, C. Cooper, and B. M. Margetts, Muscle strength, activity, housing and the risk of falls in elderly people. Age Ageing 18, 47 – 51 (1989). 11. J. H. Gurwitz, M. T. Sanchez-Croos, M. A. Eckler, and J. Matulis, The epidemiology of adverse and unexpected events in the longterm care setting. J. Am. Geriatr. Soc. 42, 33 – 38 (1994). 12. P. O. Jantti, V. I. Pyykko, A. L. J. Hervonen, V. I. Pyykko, and A.L. Hervonen, Falls among elderly nursing home residents. Public Health 107, 89 – 96 (1993). 13. A. S. Robbins, L. Z. Rubenstein, and K. R. Josephson, Predictors of falls among elderly people: Results of two population-based studies. Arch. Intern. Med. 149, 1628 – 1631 (1989).

803 14. National Safety Council, “Accident Facts.” National Safety Council, Itasca, IL (1998). 15. S. P. Baker, B. O’Neill, and R. S. Karpf, “The Injury Fact Book.” Oxford Univ. Press, New York (1992). 16. D. P. Rice, E.J. MacKenzie, et al., “Cost of Injury in the United States: A Report to Congress, 1989.” Johns Hopkins Univ. Baltimore MD (1989). 17. D. l. Covington, J. G. Maxwell, and T. V. Clancy, Hospital resources used to treat the injured elderly at North Carolina trauma centers. J. Am. Geriatr. Soc. 41, 847 – 852 (1993). 18. D. F. Schwarz, J. A. Grisso, C. G. Miles, et al., A longitudinal study of injury morbidity in an African-American population. JAMA 271, 755 – 760 (1994). 19. R. W. Sattin, D. A. L. Humber, C. A. deVito, et al., The incidence of fall injury events among the elderly in a defined population. Am. J. Epidemiol. 131, 1028 – 1037 (1990). 20. D. Fife, J. I. Barancik, and B. F. Chatterjee, Northeastern Ohio trauma study. II. Injury rates by age sex, and cause. Am. J. Public Health 74, 473 – 478 (1984). 21. J. A. Grisso, D. F. Schwarz, V. Wolfson, et al., The impact of falls in an inner-city elderly African-American population. J. Am. Geriatr. Soc. 40, 673 – 678 (1992). 22. M. E. Tinetti, W. L. Liu, and S. F. Ginter, Mechanical restraint use and fall-related injuries among residents of skilled nursing facilities. Ann. Intern. Med. 116, 369 – 374 (1992). 23. M. C. Nevitt, Falls in the elderly: Risk factors and prevention. In “Gait Disorders of Aging: Falls and Therapeutic Strategies” (J. C. Masdeu, L. Sudarsky, and L. Wolfson, eds.), pp. 13 – 36. LippincottRaven, Philadelphia (1997). 24. National Center for Health Statistics, National hospital discharge survey: Annual summary, 1991. In “Vital and Health Statistics, Series 13, No. 114.” Department of Health and Human Services, Hyattsville (1993). 25. Institute of Medicine, Osteoporosis. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, Eds.). National Academy Press, Washington, DC (1990). 26. J. A. Cauley, S. R. Cummings, D. G. Seeley, et al., Effects of thiazide diuretic therapy on bone mass, fractures, and falls. Ann. Intern. Med. 118, 666 – 673 (1993). 27. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 28. M. R. Griffin, W. A. Ray, R. L. Fought, and L. J. I. Melton, Blackwhite differences in fracture rates. Am. J. Epidemiol. 136, 1378 – 1385 (1992). 29. J. L. Kelsey, W. S. Browner, D. G. Seeley, et al., Risk factors for fractures of the distal forearm and proximal humerus. Am. J. Epidemiol. 135, 477 – 489 (1992). 30. D. G. Seeley, W. S. Browner, M. C. Nevitt, et al., Which fractures are associated with low appendicular bone mass in elderly women? Ann. Intern. Med. 115, 837 – 842 (1991). 31. M. E. Tinetti, C. F. Mendes de Leon, J. T. Doucette, et al., Fear of falling and fall-related efficacy in relationship to functioning among community-living elders. J. Gerontol. 49, M140 – M147 (1994). 32. Kellogg International Work Group on the Prevention of Falls by the Elderly, The prevention of falls in later life. Dan. Med. Bull. 34 (Suppl. 4), 1 – 24 (1987). 33. M. C. Nevitt, Falls in older persons: Risk factors and prevention. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, eds.), National Academy Press, Washington, DC (1990). 34. A. V. Schwartz, M. L. Villa, M. Prill, J. A. Kelsey, J. A. Galinus, R. R. Delay, M. C. Nevitt, D. A. Bloch, R. Marcus, and J. L. Kelsey, Falls in older Mexican-American women. J. Am. Geriatr. Soc. 47, 1 – 8 (1999).

804 35. J. W. Davis, P. D. Ross, M. C. Nevitt, and R. D. Wasnich, Incidence rates of falls among Japanese men and women living in Hawaii. J. Clin. Epidemiol. 50, 589 – 594 (1997). 36. L. A. Lipsitz, P. V. Jonsson, M. M. Kelley, and J. S. Koestner, Causes and correlates of recurrent falls in ambulatory frail elderly. J. Gerontol. 46, M114 – M122 (1991). 37. A. J. Campbell, J. Reinken, and B. C. Allan, Falls in old age: A study of frequency and related clinical factors. Age Ageing 10, 264 – 270 (1981). 38. M. E. Tinetti, T. F. Williams, and R. Mayewski, Fall risk index for elderly patients based on number of chronic disabilities. Am. J. Med. 80, 429 – 434 (1986). 39. G. M. Gehlsen and M. H. Whaley, Falls in the elderly. II. Balance, strength, and flexibility. Arch. Phys. Med. Rehabil. 71, 739 – 741 (1990). 40. B. E. Maki, P. J. Holliday, and A. K. Topper, A prospective study of postural balance and risk of falling in an ambulatory and independent elderly population. J. Gerontol. 49, M72 – M84 (1994). 41. C. Ring, U. S. L. Nayak, and B. Isaac, Balance function in elderly people who have and who have not fallen. Arch. Phys. Med. Rehabil. 69, 261 – 264 (1988). 42. G. S. Sorock and D. M. Labiner, Peripheral neuromuscular dysfunction and falls in an elderly cohort. Am. J. Epidemiol. 136, 584 – 591 (1992). 43. R. H. Whipple, L. I. Wolfson, and P. M. Amerman, The relationship of knee and ankle weakness to falls in nursing home residents: An isokinetic study. J. Am. Geriatr. Soc. 35, 13 – 20 (1987). 44. N. B. Alexander, Postural control in older adults. J. Am. Geriatr. Soc. 42, 93 – 108 (1994). 45. R. Q. Ivers, R. G. Cumming, P. Mitchell, and K. Attebo, Visual impairment and falls in older adults: The Blue Mountains Eye Study. J. Am. Geriatr. Soc. 46, 58 – 64 (1998). 46. J. K. Richardson, C. Ching, and E. A. Hurvitz, The relationship between electromyographically documented peripheral neuropathy and falls. J. Am. Geriatr. Soc. 40, 1008 – 1012 (1992). 47. W. C. Graafmans, M. E. Ooms, H. M. Hofstee, P. D. Bezemer, L. M. Bouter, and P. Lips, Falls in the elderly: A prospective study of risk factors and risk profiles. Am. J. Epidemiol. 143, 1129 – 1136 (1996). 48. M. H. Woollacott and P. F. Tang, Balance control during walking in the older adult: Research and its implications. Phys. Ther. 77, 646 – 660 (1997). 49. L. A. Brown, A. Shumway-Cook, and M. H. Woollacott, Attentional demands and postural recovery: The effects of aging. J. Gerontol. A Biol. Sci. Med. Sci. 54, M165 – M17 (1999). 50. S. R. Lord, J. A. Ward, P. Williams, and K. J. Anstey, Physiological factors associated with falls in older community-dwelling women. J. Am. Geriatr. Soc. 42, 1110 – 1117 (1994). 51. S. M. Woolley, S. J. Czaja, and C. G. Drury, An assessment of falls in elderly men and women. J. Gerontol. A Biol. Sci. Med. Sci. 52, M80 – M87 (1997). 52. D. M. Buchner and E. B. Larson, Falls and fractures in patients with alzheimer-type dementia. JAMA 257, 1492 – 1495 (1987). 53. J. L. Melton III, C. M. Beard, E. Kokmen, et al., Fracture risk in patients with Alzheimer’s disease. J. Am. Geriatr. Soc. 42, 614 – 619 (1994). 54. J. A. Grisso, G. Y. Chiu, G. Maislin, W. C. Steinmann, and J. Portale, Risk factors for hip fractures in men: A preliminary study. J. Bone. Miner. Res. 6, 865 – 868 (1991). 55. M. C. Nevitt, S. R. Cummings, and E. S. Hudes, Risk factors for injurious falls: A prospective study. J. Gerontol. 46, M164 – M170 (1991). 56. E. H. Franssen, L. E. Souren, C. L. Torossian, and B. Reisberg, Equilibrium and limb coordination in mild cognitive impairment and mild Alzheimer’s disease. J. Am. Geriatr. Soc. 47, 463 – 469 (1999). 57. M. A. Whooley, K. E. Kip, J. A. Cauley, K. E. Ensrud, M. C. Nevitt, and W. S. Browner, Depression, falls, and risk of fracture in older

SCHWARTZ, CAPEZUTI, AND GRISSO

58.

59.

60.

61.

62.

63.

64.

65. 66.

67. 68.

69.

70.

71. 72. 73. 74.

75. 76.

77. 78.

79.

80.

women: Study of Osteoporotic Fractures Research Group. Arch. Intern. Med. 159, 484 – 490 (1999). W. A. Ray and M. R. Griffin, Prescribed medications, falling, and fall-related injuries. In “Reducing Frailty and Falls in Older Persons” (R. Weindruch, E. C. Hadley, and M. G. Ory, eds.), Charles C. Thomas, Springfield, III (1991). R. G. Cumming, J. P. Miller, J. L. Kelsey, et al., Medications and multiple falls in elderly people: The St. Louis OASIS study. Age Ageing 20, 455 – 461 (1991). P. B. Thapa, P. Gideon, T. W. Cost, A. B. Milam, and W. A. Ray, Antidepressants and the risk of falls among nursing home residents. N. Engl. J. Med. 339, 875 – 882 (1998). N. E. Mayo, N. Korner-Bitensky, and R. Becker, Predicting falls among patients in a rehabilitation hospital. Am. J. Phys. Med. Rehabil. 68, 139 – 146 (1989). E. Granek, S. P. Baker, H. Abbey, et al., Medications and diagnoses in relation to falls in a long-term care facility. J. Am. Geriatr. Soc. 35, 503 – 511 (1987). J. Mahoney, M. Sager, N. C. Dunham, and J. Johnson, Risk of falls after hospital discharge. J. Am. Geriatr. Soc. 42, 269 – 274 (1994). R. M. Leipzig, R. G. Cumming, and M. E. Tinetti, Drugs and falls in older people: A systematic review and meta-analysis. II. Cardiac and analgesic drugs. J. Am. Geriatr. Soc. 47, 40 – 50 (1999). R. G. Cumming, Epidemiology of medication-related falls and fractures in the elderly. Drugs Aging 12, 43 – 53 (1998). R. M. Leipzig, R. G. Cumming, and M. E. Tinetti, Drugs and falls in older people: A systematic review and meta-analysis. I. Psychotropic drugs. J. Am. Geriatr. Soc. 47, 30 – 39 (1999). J. E. Carlson, Alcohol use and falls. J. Am. Geriatr. Soc. 41, 346 – 347 (1993). D. E. Nelson, R. W. Sattin, J. A. Langlois, et al., Alcohol as a risk factor for fall injury events among elderly persons living in the community. J. Am. Geriatr. Soc. 40, 658 – 661 (1992). W. P. Berg, H. M. Alessio, E. M. Mills, and C. Tong, Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 26, 261 – 268 (1997). P. T. van Dijk, O. G. Meulenberg, H. J. van de Sande, and J. D. Habbema, Falls in dementia patients. Gerontologist 33, 200 – 204 (1993). H. Nickens, Intrinsic factors in falling among the elderly. Arch. Intern. Med. 145, 1089 – 1093 (1985). E. M. Lau and S. P. Donnan, Falls and hip fracture in Hong Kong Chinese. Public Health 104, 117 – 121 (1990). R. Tideiksaar, Geriatric falls in the home. Home Healthc. Nurse 4, 14 – 23 (1986). L. Z. Rubenstein, A. S. Robbins, B. L. Schulman, J. Rosado, D. Osterweil, and K. R. Josephson, Falls and instability in the elderly. J. Am. Geriatr. Soc. 36, 266 – 278 (1988). B. E. Maki, P. J. Holliday, and A. K. Topper, Fear of falling and postural performance in the elderly. J. Gerontol. 46, M123 – M131 (1991). M. C. Hornbrook, V. J. Stevens, D. J. Wingfield, J. F. Hollis, M. R. Greenlick, and M. G. Ory, Preventing falls among communitydwelling older persons: Results from a randomized trial. Gerontologist 34, 16 – 23 (1994). R. Tideiksaar, “Falling in Old Age: Prevention and Management.” Springer, New York (1997). L. Clemson, R. G. Cumming, and M. Roland, Case-control study of hazards in the home and risk of falls and hip fractures. Age Ageing 25, 97 – 101 (1996). S. Studenski, P. W. Duncan, J. Chandler, et al., Predicting falls: The role of mobility and nonphysical factors. J. Am. Geriatr. Soc. 42, 297 – 302 (1994). A. J. Campbell, M. J. Borrie, G. F. Spears, S. L. Jackson, J. S. Brown, and J. L. Fitzgerald, Circumstances and consequences of

CHAPTER 32

81.

82.

83.

84.

85. 86. 87.

88. 89. 90. 91.

92.

93.

94. 95.

96.

97.

98.

99.

100.

101.

Falls as Risk Factors for Fractures

falls experienced by a community population 70 years and over during a prospective study [published erratum appears in Age Ageing 19(5), 345 – 346 (1990)]. Age Ageing 19, 136 – 141 (1990). M. Speechley and M. Tinetti, Falls and injuries in frail and vigorous community elderly persons. J. Am. Geriatr. Soc. 39, 46 – 52 (1991). M. E. Northridge, M. C. Nevitt, J. L. Kelsey, and B. Link, Home hazards and falls in the elderly: The role of health and functional status. Am. J. Public Health 85, 509 – 515 (1995). M. E. Northridge, M. C. Nevitt, and J. L. Kelsey, Non-syncopal falls in the elderly in relation to home environments. Osteopor. Int. 6, 249 – 255 (1996). M. J. Ashley, C. I. Gryfe, and A. Amies, A longitudinal study of falls in an elderly population II. Some circumstances of falling. Age Ageing 6, 211 – 220 (1977). N. Coyle Cancer audit. A problem-focused approach to nursing audit: Patient falls. Cancer Nurs. 2, 389 – 391 (1979). J. Foerster, A study of falls: The elderly nursing home resident. J. N. Y. State Nurses Assoc. 12, 9 – 17 (1981). A. Halpert and J. P. Connors, Prevention of patient falls through perceived control and other techniques. Law Med. Health Care 14, 20 – 4, 12 (1986). N. G. Rainville, Effect of an implemented fall prevention program on the frequency of patient falls. QRB Qual. Rev. Bull. 10, 287–291 (1984). G. M. Tinker, Accidents in a geriatric department. Age Ageing 8, 196 – 198 (1979). M. E. Tinetti and M. Speechley, Prevention of falls among the elderly. N. Engl. J. Med. 320, 1055 – 1059 (1989). Y. T. Gross, Y. Shimamoto, C. L. Rose, and B. Frank, Why do they fall? Monitoring risk factors in nursing homes. J. Gerontol. Nurs. 16, 20 – 25 (1990). L. C. Mion, J. D. Frengley, C. A. Jakovcic, and J. A. Marino, A further exploration of the use of physical restraints in hospitalized patients [published erratum appears in J. Am. Geriatr. Soc. 38(2), 181 (1990)]. J. Am. Geriatr. Soc. 37, 949 – 956 (1989). E. Capezuti, L. Evans, N. Strumpf, and G. Maislin, Physical restraint use and falls in nursing home residents. J. Am. Geriatr. Soc. 44, 627 – 633 (1996). E. Capezuti, N. Strumpf, L. Evans, and G. Maislin, Gerontologist 38(Spec. No.1), 226 (1998). [Abstract] L. K. Evans and N. E. Strumpf, Tying down the elderly: A review of the literature on physical restraint. J. Am. Geriatr. Soc. 37, 65 – 74 (1989). D. M. Buchner, M. E. Cress, E. H. Wagner, B. J. de Lateur, R. Price, and I. B. Abrass, The Seattle FICSIT/MoveIt study: The effect of exercise on gait and balance in older adults. J. Am Geriatr. Soc. 41, 321 – 325 (1993). D. N. Kanten, C. D. Mulrow, M. B. Gerety, M. J. Lichtenstein, C. Aguilar and J. E. Cornell, Falls: An examination of three reporting methods in nursing homes. J. Am. Geriatr. Soc. 41, 662 – 666 (1993). F. K. Ejaz, J. A. Jones, and M. S. Rose, Falls among nursing home residents: An examination of incident reports before and after restraint reduction programs. J. Am. Geriatr. Soc. 42, 960 – 964 (1994). L. K. Evans, N. E. Strumpf, S. L. Allen-Taylor, E. Capezuti, G. Maislin, and B. Jacobsen, A clinical trial to reduce restraints in nursing homes. J. Am. Geriatr. Soc. 45, 675 – 681 (1997). E. Capezuti, N. Strumpf, L. Evans, and G. Maislin, Outcomes of nighttime physical restraint removal for severely impaired nursing home residents. Am. J. Alzheim. Dis. 14, 157 – 164 (1999). E. Capezuti, N. E. Strumpf, L. K. Evans, J. A. Grisso, and G. Maislin, The relationship between physical restraint removal and falls and injuries among nursing home residents. J. Gerontol. A Biol. Sci. Med. Sci. 53, M47 – M52 (1998).

805 102. M. E. Tinetti, Factors associated with serious injury during falls by ambulatory nursing home residents. J. Am. Geriatr. Soc. 35, 644 – 648 (1987). 103. S. F. Ginter and L. C. Mion, Falls in the nursing home: Preventable or inevitable? J. Gerontol. Nurs. 18, 43 – 48 (1992). 104. M. E. Tinetti, J. Doucette, E. Claus, and R. Marottoli, Risk factors for serious injury during falls by older persons in the community. J. Am. Geriatr. Soc. 43, 1214 – 1221 (1995). 105. C. I. Gryfe, A. Amies, and M. J. Ashley, A longitudinal study of falls in an elderly population. I. Incidence and morbidity. Age Ageing 6, 201 – 210 (1977). 106. J. C. Brocklehurst, A. N. Exton-Smith, B. S. Lempert, L. P. Hunt, and M. K. Palmer, Fracture of the femur in old age: A two-centre study of associated clinical factors and the cause of the fall. Age Ageing 7, 2 – 15 (1978). 107. G. Berry, R. H. Fisher, and S. Lang, Detrimental incidents, including falls, in an elderly institutional population. J. Am. Geriatr. Soc. 29, 322 – 324 (1981). 108. J. M. Mossey, Social and psychologic factors related to falls among the elderly. Clin. Geriatr. Med. 1, 541 – 553 (1985). 109. R. G. Cumming, and R. J. Klineberg, Fall frequency and characteristics and the risk of hip fractures. J. Am. Geriatr. Soc. 42, 774 – 778 (1994). 110. S. R. Cummings and M. C. Nevitt, A hypothesis: The causes of hip fractures. J. Gerontol. 44, M107 – M111 (1989). 111. W. C. Hayes, E. R. Myers, J. N. Morris, T. N. Gerhart, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 112. M. C. Nevitt, S. R. Cummings, et al., Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). 113. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 114. J. A. Grisso, A. R. Wishner, D. F. Schwarz, B. A. Weene, J. H. Holmes, and R.L. Sutton, A population-based study of injuries in inner-city women. Am. J. Epidemiol. 134, 59 – 68 (1991). 115. M. E. Tinetti, J. T. Doucette, and E. B. Claus, The contribution of predisposing and situational risk factors to serious fall injuries. J. Am. Geriatr. Soc. 43, 1207 – 1213 (1995). 116. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The Study of Osteoporotic Fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 117. T. E. Cohn and D. J. Lasley, Visual depth illusion and falls in the elderly. Clin. Geriatr. Med. 1, 601 – 620 (1985). 118. C. Cooper, D. J. Barker, and C. Wickham, Physical activity, muscle strength, and calcium intake in fracture of the proximal femur in Britain. Br. Med. J. 297, 1443 – 1446 (1988). 119. D. T. Felson, J. J. Anderson, M. T. Hannan, R. C. Milton, P. W. Wilson, and D. P. Kiel, Impaired vision and hip fracture. The Framingham Study. J. Am. Geriatr. Soc. 37, 495 – 500 (1989). 120. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, and S. Hoffman, Risk factors for hip fracture in black women: The Northeast Hip Fracture Study Group. N. Engl. J. Med. 330, 1555 – 1559 (1994). 121. J. A. Grisso, J. L. Kelsey, L. A. O’Brien, C. G. Miles, S. Sidney, G. Maislin, K. LaPann, D. Moritz, and B. Peters, Risk factors for hip fracture in men. Hip Fracture Study Group. Am. J. Epidemiol. 145, 786 – 793 (1997). 122. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 123. R. Wootton, E. Bryson, U. Elsasser, H. Freeman, J. R. Green, R. Hesp, E. A. Hudson, L. Klenerman, T. Smith, and J. Zanelli, Risk

806

124.

125.

126.

127.

128. 129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

SCHWARTZ, CAPEZUTI, AND GRISSO factors for fractured neck of femur in the elderly. Age Ageing 11, 160 – 168 (1982). A. V. Schwartz, J. L. Kelsey, S. Sidney, and J. A. Grisso, Characteristics of falls and risk of hip fracture in elderly men. Osteopor. Int. 8, 240 – 246 (1998). J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, and F. Kaplan, Risk factors for falls as a cause of hip fracture in women: The Northeast Hip Fracture Study Group. N. Engl. J. Med. 324, 1326 – 1331 (1991). B. E. Maki and G. R. Fernie, Impact attenuation of floorcoverings in simulated falling accidents. Appl. Ergonom. 21, 107 – 114 (1990). S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women: Study of Osteoporotic Fractures Research Group. N. Engl. J. Med. 332, 767 – 773 (1995). L. J. Melton and B. L. Riggs, Risk factors for injury after a fall. Clin. Geriatr. Med. 1, 525 – 536 (1985). M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public. Health 74, 1374 – 1380 (1984). L. D. Gillespie, W. J. Gillespie, R. Cumming, S. E. Lamb, and B. H. Rowe, Interventions for preventing falls in the elderly (Cochrane Review). In “The Cochrane Library,” 3, Ed., Update Software, Oxford, (1999). M. E. Tinetti, D. I. Baker, P. A. Garrett, M. Gottschalk, M. L. Koch, and R. I. Horwitz, Yale FICSIT: Risk factor abatement strategy for fall prevention. J. Am. Geriatr. Soc. 41, 315 – 320 (1993). M. E. Tinetti, D. I. Baker, G. McAvay, E. B. Claus, P. Garrett, M. Gottschalk, M. L. Koch, K. Trainor, and R. I. Horwitz, A multifactorial intervention to reduce the risk of falling among elderly people living in the community . N. Engl. J. Med. 331, 821 – 827 (1994). M. E. Tinetti, G. McAvay, and E. Claus, Does multiple risk factor reduction explain the reduction in fall rate in the yale ficsit trial? Frailty and injuries cooperative studies of intervention techniques. Am. J. Epidemiol. 144, 389 – 399 (1996). J. Close, M. Ellis, R. Hooper, E. Glucksman, S. Jackson, and C. Swift, Prevention of falls in the elderly trial (PROFET): A randomised controlled trial. Lancet 353, 93 – 97 (1999). W. A. Ray, J. A. Taylor, K. G. Meador, P. B. Thapa, A. K. Brown, H. K. Kajihara, C. Davis, P. Gideon, and M. R. Griffin, A randomized trial of a consultation service to reduce falls in nursing homes. JAMA 278, 557 – 562 (1997). S. L. Wolf, H. X. Barnhart, G. L. Ellison, and C. E. Coogler, The effect of Tai Chi Quan and computerized balance training on postural stability in older subjects. Atlanta FICSIT Group. Frailty and Injuries: Cooperative Studies on Intervention Techniques. Phys. Ther. 77, 371 – 81; discussion 382 – 384 (1997). D. M. Buchner, M. E. Cress, B. J. de Lateur, P. C. Esselman, A. J. Margherita, R. Price, and E. H. Wagner, The effect of strength and endurance training on gait, balance, fall risk, and health services use in community-living older adults. J. Gerontol. A Biol. Sci. Med. Sci. 52, M218 – M224 (1997). A. J. Campbell, M. C. Robertson, M. M. Gardner, R. N. Norton, M. W. Tilyard, and D. M. Buchner, Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. Br. Med. J. 315, 1065 – 1069 (1997). M. A. Province, E. C. Hadley, M. C. Hornbrook, L. A. Lipsitz, J. P. Miller, C. D. Mulrow, M. G. Ory, R. W. Sattin, M. E. Tinetti, and S. L. Wolf, The effects of exercise on falls in elderly patients: A

140.

141.

142.

143.

144.

145.

146.

147. 148.

149.

150.

151.

152.

153.

154.

155.

preplanned meta-analysis of the FICSIT Trials. Frailty and Injuries: Cooperative Studies of Intervention Techniques. JAMA 273, 1341 – 1347 (1995). M. E. McMurdo, P. A. Mole, and C. R. Paterson, Controlled trial of weight bearing exercise in older women in relation to bone density and falls. Br. Med. J. 314, 569 (1997). S. R. Lord, J. A. Ward, P. Williams, and M. Strudwick, The effect of a 12-month exercise trial on balance, strength, and falls in older women: A randomized controlled trial. J. Am. Geriatr. Soc. 43, 1198 – 1206 (1995). S. Reinsch, P. MacRae, P. A. Lachenbruch, and J. S. Tobis, Attempts to prevent falls and injury: A prospective community study. Gerontologist 32, 450 – 456 (1992). C. D. Mulrow, M. B. Gerety, D. Kanten, J. E. Cornell, L. A. DeNino, L. Chiodo, C. Aguilar, M. B. O’Neil, J. Rosenberg, and R. M. Solis, A randomized trial of physical rehabilitation for very frail nursing home residents. JAMA 271, 519 – 524 (1994). M. A. Fiatarone, E. F. O’Neill, N. D. Ryan, K. M. Clements, G. R. Solares, M. E. Nelson, S. B. Roberts, J. J. Kehayias, L. A. Lipsitz, and W. J. Evans, Exercise training and nutritional supplementation for physical frailty in very elderly people. N. Engl. J. Med. 330, 1769 – 1775 (1994). A. J. Campbell, M. C. Robertson, M. M. Gardner, R. N. Norton, and D. M. Buchner, Psychotropic medication withdrawal and a homebased exercise program to prevent falls: A randomized, controlled trial. J. Am. Geriatr. Soc. 47, 850 – 853 (1999). S. M. Willcox, D. U. Himmelstein, and S. Woolhandler, Inappropriate drug prescribing for the community-dwelling elderly. JAMA 272, 292 – 296 (1994). P. J. Greco and J. M. Eisenberg, Changing physicians’ practices. N. Engl. J. Med. 329, 1271 – 1273 (1993). W. A. Ray, J. A. Taylor, K. G. Meador, M. J. Lichtenstein, M. R. Griffin, R. Fought, M. L. Adams, and D. G. Blazer, Reducing antipsychotic drug use in nursing homes: A controlled trial of provider education. Arch. Intern. Med. 153, 713 – 721 (1993). J. Avorn, S. B. Soumerai, D. E. Everitt, D. Ross-Degnan, M. H. Beers, D. Sherman, S. R. Salem-Schatz, and D. Fields, A randomized trial of a program to reduce the use of psychoactive drugs in nursing homes . N. Engl. J. Med. 327, 168 – 173 (1992). R. I. Shorr, R. L. Fought, and W. A. Ray, Changes in antipsychotic drug use in nursing homes during implementation of the OBRA-87 regulations . JAMA 271, 358 – 362 (1994). E. L. Siegler, E. Capezuti, G. Maislin, M. Baumgarten, L. Evans, and N. Strumpf, Effects of a restraint reduction intervention and OBRA ‘87 regulations on psychoactive drug use in nursing homes . J. Am. Geriatr. Soc. 45, 791 – 796 (1997). M. C. Hornbrook, V. J. Stevens, and D. J. Wingfield, Seniors’ Program for Injury Control and Education. J. Am. Geriatr. Soc. 41, 309 – 314 (1993). P. Dargent-Molina, F. Favier, H. Grandjean, C. Baudoin, A. M. Schott, E. Hausherr, P. J. Meunier, and G. Breart, Fall-related factors and risk of hip fracture: The EPIDOS prospective study. epidemiologie de l’osteoporose. Lancet 348, 145 – 149 (1996). L. Z. Rubenstein, A. S. Robbins, K. R. Josephson, B. L. Schulman, and D. Osterweil, The value of assessing falls in an elderly population: A randomized clinical trial. Ann. Intern. Med. 113, 308 – 316 (1990). E. H. Wagner, A. Z. Lacroix, L. Grothaus, S. G. Leveille, J. A. Hecht, K. Artz, K. Odle, and D. M. Buchner, Preventing disability and falls in older adults: A population-based randomized trial. Am. J. Public Health 84, 1800 – 1806 (1994).

CHAPTER 32

Falls as Risk Factors for Fractures

156. D. F. Schwarz, J. A. Grisso, C. Miles, J. H. Holmes, and R. L. Sutton, An injury prevention program in an urban African-American community. Am. J. Public Health 83, 675 – 680 (1993). 157. D. M. Black, S. R. Cummings, D. B. Karpf, J. A. Cauley, D. E. Thompson, M. C. Nevitt, D. C. Bauer, H. K. Genant, W. L. Haskell, R. Marcus, S. M. Ott, J. C. Torner, S. A. Quandt, T. F. Reiss, and K. E. Ensrud, Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures: Fracture Intervention Trial Research Group. Lancet 348, 1535 – 1541 (1996). 158. S. R. Cummings, D. M. Black, D. E. Thompson, W. B. Applegate, E. Barrett-Connor, T. A. Musliner, L. Palermo, R. Prineas, S. M. Rubin, J. C. Scott, T. Vogt, R. Wallace, A. J. Yates, A. Z. Lacroix, Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: Results from the Fracture Intervention Trial. JAMA 280, 2077 – 2082 (1998). 159. M. H. Komulainen, H. Kroger, M. T. Tuppurainen, A. M. Heikkinen, E. Alhava, R. Honkanen, and S. Saarikoski, HRT and Vit D in prevention of non-vertebral fractures in postmenopausal women; a 5 year randomized trial. Maturitas 31, 45 – 54 (1998). 160. B. Ettinger, D. M. Black, B. H. Mitlak, R. K. Knickerbocker, et al., Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: Results from a 3-year randomized clinical trial. JAMA 282, 637 – 645 (1999)

807 161. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, L. Sadowski, N. Sahyoun, and S. Tannenbaum, A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N. Engl. J. Med. 323, 878 – 883 (1990). 162. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in the elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 163. J. B. Lauritzen, M. M. Petersen, and B. Lund, Effect of external hip protectors on hip fractures. Lancet 341, 11 – 13 (1993). 164. S. N. Robinovitch, W. C. Hayes, and T. A. McMahon, Energy-shunting hip padding system attenuates femoral impact force in a simulated fall. J. Biomech. Eng. 117, 409 – 413 (1995). 165. P. Cavanagh, Saving frequent fallers. The New York Times Magazine (1994). 166. E. Capezuti, K. A. Talerico, N. Strumpf, and L. Evans, Individualized assessment and intervention in bilateral siderail use. Geriatr. Nurs. 19, 322 – 330 (1998). 167. E. Capezuti, K. A. Talerico, L. Wagner, N. Strumpf, H. Becker, and G. Maislin, The effect of a low-height bed intervention on nightfalls among frail nursing home residents. Gerontologist 39 (Special issue 1), 196.

CHAPTER 33

Assessing Fracture Risk CHARLES W. SLEMENDA,† C. CONRAD JOHNSTON, AND SIU L. HUI Department of Medicine, Indiana University School of Medicine and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202

I. II. III. IV.

Introduction Skeletal Elements of Fracture Risk Nonskeletal Elements of Fracture Risk Assessment Measuring Bone Loss for Risk Assessment: Necessary or Useful?

V. Monitoring Bone Mass VI. The Expression of Risk VII. Summary References

I. INTRODUCTION

Thus, there will always be fractures that occur among people evaluated as being at low risk, and some at the highest risk will avoid fracture (primarily through the avoidance of falls). Second, there are events, some identifiable and some not (e.g., rapid bone loss, neurological disorders), that will substantially increase a patient’s risk after the initial assessment, but the high frequency of monitoring of bone mass and other risk factors necessary to detect such changes is impractical and potentially very expensive. It is also, for the majority of patients, uninformative. Third, the acceptable balance between the costs associated with diagnosis and treatment of osteoporosis and the prevention of the suffering and costs associated with osteoporotic fractures is variable, both as perceived by patients and as perceived by providers of health care. Whereas measurements of all skeletal sites every few months might provide virtually perfect understanding of this aspect of risk, it would be prohibitively expensive and would improve fracture prevention marginally, if at all. Also, it would increase cost:benefit ratios greatly. What follows is an evaluation of both skeletal and nonskeletal factors that influence risk for osteoporotic fractures, and of the consequences associated with various courses of action.

The identification of patients at highest risk of suffering osteoporotic fractures is an issue of both clinical and scientific importance. Risk factors for fracture include not only low bone mass and associated defects of macroand microarchitecture, but also factors that affect risk of falling and other as yet poorly understood mechanisms. From a scientific point of view, all of these elements deserve attention, whereas for clinical purposes a more narrow approach addressing factors influencing the interpretation of alterable risks may be appropriate. However, the divergence between these approaches is small, as will be shown. Inherent in the clinical approaches to identification of risks and the prevention of osteoporotic fractures are several factors that make this aspect of the practice of clinical medicine difficult. First, there is no level of bone mass and no identifiable phenotype that reduces fracture risk to zero.

†

Deceased.

OSTEOPOROSIS, SECOND EDITION VOLUME 1

809

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

810

SLEMENDA, JOHNSTON, AND HUI

II. SKELETAL ELEMENTS OF FRACTURE RISK A. Bone Mass Low bone mass increases the risk of osteoporotic fractures, and this risk is independent of the risk associated with increasing age [1 – 3]. The earliest prospective studies of bone mass and fractures showed that appendicular measurements of both bone mineral content (BMC) and bone mineral density (BMD) could identify those women at the highest risk of fractures and that the risk of nonspine fractures increased approximately 1.5- to 2-fold for each standard deviation (SD) decline in bone mass at the radius or calcaneus [3 – 6]. Furthermore, there was no evidence for a threshold effect because risk increased progressively with declining BMD. Later studies of axial skeletal sites, e.g., the hip and spine, demonstrated similar predictive value for these measurements [3,7,8]. A comprehensive review [9] concluded that all measurement sites had similar predictive abilities (relative risk, RR, about 1.5 per SD decrease in BMD) except for spine measurements predicting spine (relative risk of 2.3) [7] and hip measurements predicting hip (relative risk of 2.6) [8]. A measurement of BMD at any skeletal site provides information regarding bone mass and associated fracture risk that cannot be otherwise obtained. Numerous studies have shown that clinical risk factors, although significantly associated with bone mass, misclassify more than one-third of women with low bone mass [10], although some risk factors may influence fracture risk through mechanisms other than bone mass. In the absence of bone mass measurements, risk factor information may be useful, but this is an inferior approach to the classification of fracture risk. For the identification of patients with low bone mass for whom therapies to preserve bone mass are being considered, only direct measurement of bone mass provides accurate and specific data for assessment of this risk. It is also necessary to distinguish between risk factors for low bone mass and risk factors for fractures, which may differ, and in rare cases work in opposite directions, as discussed later.

B. Which Skeletal Site(s) Should Be Measured? Bone mass can be measured at central sites (spine and hip) or peripheral sites (forearm, heel, etc.). Almost any measurement predicts fractures at any site, but generally measurements at a given site best predict fractures at the same site. The ideal bone mass measurement would produce the largest gradient of fracture risk from high to low bone mass and would classify patients with the least error. If only one measurement could be made, a scan of the proximal femur appears to approach most closely this ideal for

hip fractures. First, the higher relative risk of hip fractures associated with declines in BMD at this site (RR = 2.6 per SD difference in BMD) confers a considerable advantage in terms of risk gradient compared with measurements of other sites for which relative risks have been found to be 1.5 – 2.0. For example, given these relative risks, at any given age, a woman in the 10th percentile of proximal femur BMD (RR = 2.6) has 11.5 times the fracture risk of a woman with BMD in the 90th percentile (1.28 SD above the mean vs 1.28 SD below the mean); for the lower RR of 2 associated with measurements at other sites, this gradient is only six-fold. Second, although for cross-sectional measurements the error at the hip is slightly higher than at the spine, this is trivial relative to the increased risk gradient. Further, spine measurements are associated with other problems, such as the possibility of deformed of fractured vertebrae in the scan area, the development of osteophytes, and other artifacts among the elderly. Fortunately, there are fewer problems with artifacts in measurements of the spine for women between 50 and 65. Clinically, spine BMD is therefore most useful in perimenopausal women who might experience rapid bone loss due to the abrupt changes in sex hormone concentrations. Peripheral measurements have been gaining popularity because of the lower cost and portability of the instruments. Generally, a low peripheral measurement would trigger a central measurement for consideration of treatment and establishing a baseline for subsequent monitoring. However, the trigger point for optimizing cost effectiveness has not been established for any of the peripheral devices. Besides the choice of skeletal sites, the choice of instrument for measuring bone mass is important. Dual-energy X-ray absorptiometry (DXA) is the most popular method for measuring central sites; its high precision and ability to predict fractures are well established. Recently, more portable DXA scanners for measuring peripheral sites are also available; these instruments demonstrate adequate precision but their ability to predict fractures needs to be established. Also gaining popularity are quantitative ultrasound (QUS) measurements of the calcaneus; the scanners are portable and do not involve radiation. QUS has been shown to predict hip fractures almost as well as hip measurements [11 – 13]. In addition, it improves fracture prediction even after accounting for BMD measured by DXA. This observation has led some people to believe that QUS may provide a measure of bone “quality” not captured by the quantitative measurement of BMD. Finally, advocates of quantitative computed tomography (QCT) prefer this measurement because it measures volumetric density in the vertebrae and, more recently, at the hip [14]. However, data for predicting fracture based on QCT measurements are not as extensive as for DXA measurements. Furthermore, the higher level of radiation makes QCT less desirable in clinical settings.

811

CHAPTER 33 Assessing Fracture Risk

Is the measurement of a single site adequate for decisions regarding the initiation of therapy? Yes and no. Extensive clinical experience and many studies have shown that the correspondence in BMD among skeletal sites is modest, and thus some patients with adequate bone mass at one site will have low bone mass at another. Given the lack of prospective data for predicting fractures, low BMD at a peripheral site should be corroborated by a central measurement. The moderate correlation between these two sites [15,16] would suggest that measurements of both the spine and the hip are needed. It has been argued that, from a practical point of view, the incremental costs associated with a second measurement of a patient already on a scanning table are very small, and thus these measurements should be done then. However, for the measurement of a second site to have value it should be shown to improve fracture prediction. Data from the Study of Osteoporotic Fractures (SOF) have important implications in this regard [17]. Using a cut off value of 2.5 SD from the mean for young normal women, a hip measurement alone yielded a higher gradient of risk (the ratio of fractures in the low relative to the high bone mass group equaled about 5) than using both hip and spine measurements (i.e., for either site below 2.5 SD, the fracture ratio was about 2.6). These data probably reflect in part the increasing prevalence of arthritis-related artifacts in spine scans of this older, mostly 65- to 74-yearold, population, which diminishes the value of spine measurements for older people. In contrast, this and other studies, such as the EPIDOS study in Europe, find that the combination of DXA and QUS measurements best predicts fractures [18]. For which patients is one bone mass measurement enough? First, those with low bone mass detected in their first scan to not need a second baseline scan. These individuals will be treated, and for the treatment to be considered effective it will require that this first site respond to the chosen therapy. It should be noted that some therapies may have differing effects depending on skeletal site; few data exist, however, regarding clinical approaches to this issue. It is obvious that patients should be measured at least at the site expected to respond favorably to therapy. Second, those with very high proximal femur bone mass, e.g.,  /   1 SD compared with young normal values, have a very low probability of having low BMD at the spine, probably do not need a second baseline scan at some other site, and probably will not reach the treatment threshold until late in life. Third, individuals with intermediate BMD values but very poor risk factor profiles, as discussed later, can also probably be treated without further measurements. The fact that there are a number of factors that increase risk independently of BMD makes this feasible. There will then remain a small group with marginally low BMD values for the proximal femur and without an adequately informative risk factor profile to make a well-informed clinical

judgment. For these, a second measurement will be informative when the second site measured is below the treatment threshold. Other practical issues regarding the approach to this problem have been raised. Some assessment of the patients’ risk factors will have to be made prior to scanning so that the technicians performing scans will know whether one or two scans are needed. The use of more data than the scan itself will also require additional planning, as described later.

C. At What Level of Bone Mass Should Treatment Be Considered? Choosing a point at which to institute therapy to prevent fractures consists of balancing risks, costs, and benefits. This balance will surely differ depending on the patients’ and physicians’ views, and the largest differences between individual’s perceptions of the risk:benefit ratio will probably be in the perception of risks. Regarding benefits, from the standpoint of an individual patient, long-term (5 – 10 years or more) hormone replacement therapy will diminish fracture risk by about half [19], but if fracture risk is very low this may be perceived as a trivial change. From a public health perspective the most efficient use of therapy is in the treatment of the highest risk subjects. The risks associated with hormone replacement and other therapies can be viewed both objectively and as a matter of personal perspective. For hormone replacement therapy as an example, in addition to the clear benefits in terms of reduced fracture risk, there is the disputed benefit of reduced risk for coronary heart disease. There are side effects of treatment, such as bleeding and breast tenderness for some women, but the major concerns relate to cancer risks. In particular, breast cancer risks remain poorly defined but hold great importance for many women. Even small increases in breast cancer risk may be unacceptable for many patients. Thus, the recommendations for treatment cut offs that follow are obviously guidelines to be altered as patients and their physicians choose. The World Health Organization defines osteopenia as hip BMD of 1 to 2.5 SD relative to young normals and osteoporosis as more than 2.5 SD [20] below young normals. The number of standard deviations below young normals has now been widely accepted as the T score. The National Osteoporosis Foundation (NOF) guidelines recommend treatment for women with BMD T scores below 2 in the absence of osteoporosis risk factors (to be discussed later), and with T scores below 1.5 otherwise [21], Implementation of these guidelines, however, is not always straightforward because the same individual may have very different T scores based on different measurements of BMD.

812

SLEMENDA, JOHNSTON, AND HUI

In the United States, most manufacturers have adopted normal reference values for the hip from the Third National Health and Nutrition Survey (NHANES III) based on a large representative sample of the U. S. population. The means and standard deviations of DXA measurements at the hip between ages 20 and 29 years in the survey are adopted as young normal values. Although these values are available for men and women, white, black, and MexicanAmericans, it is not clear whether gender- and race-specific norms should be used for calculating T scores in assessing fracture risk. Fracture risk assessment based on other bone measurements is even more problematic because data from nationally representative samples do not exist. For the spine it appears that the manufacturers have similar databases to define “young normal.” The databases for establishing norms for peripheral measurements are generally smaller and more subject to biased selection of subjects. There is an initiative underway in the bone research community for a normative study in which a large number of representative subjects are measured on most of the instruments in the market. However, even if there is a large representative sample for establishing young normals for different instruments, inherent problems exist in defining T scores for other instruments or other skeletal sites in the same way as for DXA measurements at the hip. It has been shown that the prevalence of osteoporosis (based on T  2.5) in any age group varies greatly depending on which instrument the T score is based [22]. For example, at age 60, the prevalence of osteoporosis in white women ranges from 3% based on peripheral DXA at the heel to 50% based on spine QCT. This also implies that two equal T scores based on two instruments and/or skeletal sites do not reflect the same risk of fracture. In order to correct this problem, the National Osteoporosis Foundation, the International Society for Clinical Densitometry, and the American Society for Bone and Mineral Research have joined forces to derive T score equivalents for most densitometers in the market. In the near future, clinicians will be able to obtain measurements reported as T score equivalents on different instruments, and the same T score equivalents will reflect the same level of fracture risk. In the meantime, it is probably prudent to obtain a central measurement of BMD to establish a high risk of fracture prior to starting treatment.

D. Skeletal Structure Personal history of fracture as an adult is a strong predictor of future fractures [23 – 25]. The existence of a previous fracture since age 40 (or 50) confers a significant increase in risk (RR  1.5 to 2.0) above and beyond that associated with low bone mass [26]. In particular, those with existing vertebral deformities have an increased risk of

fractures of other vertebral bodies [23]. It is almost certain that part of this increase in risk results from local changes in load bearing due to the initial fracture, but the presence of a fracture at any site also increases the risk for fractures elsewhere, again independent of BMD. One implication of these data may be that a fracture also indicates structural defects not captured by bone mass, e.g., flaws in skeletal microarchitecture. Unfortunately, no technique can measure the quality of bone microarchitecture in vivo, other than possibly QUS. Therefore, at the time of assessment of fracture risk, a question regarding fracture history since age 40 (or 50) is essential and that an affirmative response should be considered to increase risk approximately twofold, similar to a 1 SD further decline in BMD; i.e., a patient with a positive fracture history would reach the treatment threshold earlier than a patient without. Studies have addressed the role of skeletal geometry, particularly in the proximal femur, in the etiology of hip fractures. Faulkner and colleagues [27] observed that a longer distance from the inner edge of the pelvic brim to the outer edge of the greater trochanter, as measured along the axis of the femoral neck, called hip axis length (HAL), was associated with an increased risk of hip Fractures, Duboeuf et al. [28] later found that HAL is associated with cervical but not trochanteric fractures of the hip. Other measurements have also been suggested [30]. Although broadly consistent, these studies differed in the elements of femoral geometry that contributed to the alteration of fracture risk [30]. Each of these will require further study before their practical application to hip fracture risk assessment is possible. It is also unclear whether hip geometry can be altered by any intervention.

III. NONSKELETAL ELEMENTS OF FRACTURE RISK ASSESSMENT A. Lifestyle Factors Although much research has focused on risk factors for hip fractures, available evidence suggests that the same risk factors apply to fractures at other sites. Large prospective studies, including SOF and EPIDOS, have published a number of risk factors for hip fractures [26,31]. While most studies have focused on hip fractures in old women, some studies have included all fractures in perimenopausal women and men. The elements of lifestyle considered generally include smoking, alcohol intake, caffeine and calcium consumption, and work and physical activities. Although moderate alcohol consumption is harmless or even beneficial, caffeine consumption has been associated with a small increase in fracture risk. The combined evidence from multiple studies suggests that smoking from menopause onward increases the risk of hip fracture by

813

CHAPTER 33 Assessing Fracture Risk

about half, and this increase cannot be totally explained by the earlier onset of menopause or by the reduction in body weight or circulating estrogen [32]. Physical activities generally are shown be protective against fractures, but there is no strong relationship between fractures and calcium intake beyond adolescence [33,34].

B. Health and Medications Besides the well-known risk factors of menopause and low estrogen, self-rated health has been shown to be a strong independent predictor of hip fractures [26]. A selfreported history of a prior fracture or a family history of fractures also increases the risk of future fractures. In multiple studies [31], fracture can be predicted by various functional assessments, including elevated resting pulse rates, inability to rise from a chair, slower gait speed, and difficulty in tandem walk. Visual acuity and depth perception, as well as dementia, were also associated with significant increases in fracture incidence in different studies. The use of long-acting benzodiazepines and other psychotropic drugs has been shown repeatedly to be associated with increased risk of falls [35] and, as a consequence, fractures [26].

C. Body Size The stereotype of the patient with osteoporosis is a frail, thin, short woman, and this description contains an element of truth, but fracture risk assessment requires closer examination. Body weight, apart from its effects on bone mass, has not been shown to affect fracture risk independently, despite plausible mechanisms for fatness to exert additional protective effects (e.g., soft tissue energy absorption during falls) [36]. The part of the weight made up of lean body mass may be correlated with stronger muscles that help maintain balance and prevent falls. Increase in weight since age 25 has also been suggested as protective, whereas weight loss predicts fractures [37]. The mechanism for this effect is unclear, although it is possible that women who lose weight also loss bone, and there are resulting microarchitectural deficits that weaken the femur. It is also possible that loss of muscle mass during weight loss weakens the balance and increases the propensity to fall.

IV. MEASURING BONE LOSS FOR RISK ASSESSMENT: NECESSARY OR USEFUL? A separate issue in the decision regarding how many and which sites to measure is the role of rapid bone loss,

particularly from the spine, in early postmenopausal women. Obviously, without a spine measurement, assessment of bone loss from the spine is impossible. The genesis of this concept arises both from theoretical considerations, which suggest that rapid bone loss may lead to perforations of trabeculae, and from the data of Christiansen and colleagues [38] suggesting that those women with the most rapid early postmenopausal bone loss have an increased risk of vertebral crush fractures and deformities, although without an increase in risk of nonspine fractures. The issues of assessing rates of bone loss, monitoring the effectiveness of therapies, and remeasurement of patients with initial measurements above but near the threshold for treatment are related and are discussed further later.

A. Bone Loss and Fracture Risk It is obvious that bone loss contributes to fracture risk through its effect on bone mass, but the importance of estimating rates of bone loss depends on whether loss itself contributes to risk beyond its effects on bone mass. Before proceeding, it should be noted that this discussion does not address whether a patient will ever require a second measurement of bone mass; this is generally true and is considered further, later in this chapter. The primary issue here is whether it is necessary to identify that group of women with adequate bone mass, but very rapid bone loss, who might, as a result of this loss, produce structural defects in bone that make it weaker than would be estimated based on its mass alone. In other words, apart from the interpretation of later bone mass measurements themselves, is there a reason to interpret the rate of bone loss in assessing fracture risk? Danish data showing that rates of loss predict vertebral fractures independent of bone mass are one source of evidence supporting this concept [38], and the finding that nonspine fractures predict spine fractures independently may also be thought to support such an idea [23]. If it is assumed that these data are critical in the assessment of spine (and perhaps other) fracture risks, how should this be approached? With the improved precision of densitometers it can be expected that with excellent quality control the standard deviation of repeated measurements of an individual at the same site will approach 0.01 g/cm2, i.e., something near a 1% coefficient of variation (SD/mean of repeated measures); for those with lower bone mass, precision is probably worse (due to problems with edge detection in low bone mass regions, and perhaps other factors). To address the issue of how measurement precision affects the estimation of rates of bone loss, assume, for example, that a group of women all are losing bone from some skeletal site at exactly 3% per year (relatively rapid bone loss of

814 the sort that might produce microarchitectural deficiencies). If two measurements of that site taken 1 year apart, 95% of the apparent rates of bone loss would be between 0.2 and 5.8% per year [based on var (A-B)  varA  varB], and thus, virtually all of these women would be classified correctly as losing bone mass, although half would have apparent rates of loss less than 3% per year and approximately 16% would have rates less than 1.5% per year. Thus, if one could be sure that only those people with substantial rates of loss were being measured, rates of bone loss could be estimated quite well with measurements after 1 year. Of course, if one had this knowledge, measurements would not be necessary. For more modest rates of loss or for machines with poorer precision (or when software changes occur), the misclassification errors would be markedly higher. For example, for a machine with 1.5% error (CV) the range of estimated rates of loss (for the 3% per year group) would be 1.2% to 7.2% per year. That is, even with substantial bone loss and very good instrument precision (1.5% CV), some women losing 3% per year would appear to be gaining bone. Where the errors in estimates of rates of bone loss become especially important is when no bone loss is occurring, e.g., with effective therapies. With zero bone loss, half of women will still be seen as losing bone due to variability in estimates of bone loss rates and some of these rates will be substantial (with 1.5% CV, approximately one-sixth of the women will appear to have bone loss greater than 2% per year at 1 year), but these errors, when expressed as loss per year, will be smaller with longer intervals between measurements. In contrast to the less realistic example of a sample of women where all are losing bone at 3% per year, measurements at 1-year intervals appear less valuable for the vast majority of patients. The ultimate implication of these data are that for women losing substantial amounts of bone (3% per year), measurements at 1-year intervals (with 1% CV) would allow most to be recognized as losing bone mass. Unfortunately, half of the women losing no bone at all would also be classified as losing bone mass. In practice, patients are a mix of these various groups that are very difficult to separate in the short term. With measurements at 2-year intervals, the overlap between these groups would be much smaller and would be smaller yet with greater intervals between measurements. The greatest danger in misclassification is probably in the failure to treat rapid bone loss, but with current techniques it is unlikely that measurements at intervals of 2 years or less would provide adequately precise data for this purpose. It should also be recognized that although there are a few women with rapid, persistent bone loss, rapid short-term bone loss is a poor predictor of who will lose bone rapidly

SLEMENDA, JOHNSTON, AND HUI

over a long period of time [39]. Thus, there will be far more women not losing bone who would be misclassified than the reverse. Although rapid bone loss is a problem deserving clinical attention, identification of these patients is very difficult. Hence, for women not initially treated, bone loss estimates are probably not necessary. Except in very rare circumstances, subsequent measurement of bone mass in these women should be interpreted simply on the basis of whether or not they fall below the treatment cutoff.

B. Biochemical Measurements for the Identification of Rapid Bone Losers and Nonresponders to Therapy It has been suggested that biochemical measurements may be useful in either monitoring responses to therapy or identifying those with rapid rates of bone loss. Generally, these biochemical markers have included those which reflect aspects of skeletal metabolism. Although some are thought to reflect primarily bone formation or osteoblast activity (e.g., osteocalcin) and others bone resorption or osteoclast activity [e.g., tartrate-resistant acid phosphatase (TRAP) or collagen cross-links], their primary value is in estimating rates of skeletal turnover. More rapid turnover, reflected by higher concentrations of any of these markers, is generally associated with more rapid rates of bone loss in perimenopausal women [40,41] and in subjects under treatment [42,43]. However, the existence of significant correlations with rates of bone loss does not necessarily imply clinical utility. A comprehensive review by the NOF concludes that bone turnover markers are valuable in research on metabolic bone diseases and in the clinical management of Paget’s disease. However, these measurements cannot be recommended for clinical use until substantial positive and negative predictive values can be established from more studies with fracture end points [41].

V. MONITORING BONE MASS A. Bone Mass: Whom Not to Measure Patients whose femoral neck bone mass is more than 1 SD above the young normal mean would need to lose bone at 0.03 g/cm2 for nearly 9 years to achieve a bone mass at least 1 SD below the young normal mean. This rate of loss (almost 4% per year) for a sustained period of time is an unlikely occurrence. These individuals are likely to remain in the lowest age-specific risk group and the efficiency both of therapies and of monitoring bone mass will be very poor.

815

CHAPTER 33 Assessing Fracture Risk

With the development of new drugs, therapeutic options are likely to increase in the next few years, although there will remain a group of individuals for whom there are no options available or acceptable. Although it has been suggested that bone mass measurements may be useful in motivating patients to more appropriate decisions regarding elements of fracture risk and the adherence to therapy, this has not been proven and would require further study before bone mass measurements could be recommended on this basis. Therefore, for patients without available therapeutic options, repeated bone mass measurements are not indicated.

B. Whom to Monitor and When Among the untreated with initial bone mass measurements above the treatment threshold, second and subsequent measurements should be made based on the probability that the patient would have lost enough bone to require therapy. However, the reassessment would not be done to estimate bone loss, but only to determine whether bone mass was now below the treatment level. If the chosen treatment level is some value relative to young normals, then this process is somewhat easier. Assuming mean rates of bone loss from the hip of 1 – 1.5% per year, approximately half of the patients with BMD 1 SD above the treatment threshold would reach this threshold in 7 – 10 years. Of course some would achieve this level sooner and some would never reach it. Those closer to the threshold would require second measurements sooner. However, although bone mass measurements must be the cornerstone of assessment when considering therapies that work through the preservation of bone mass, many other data remain that are of value in assessing fracture risk. If an untreated patient suffers a fracture, this would alter risk assessment in favor of earlier treatment. Similarly, some acute changes should prompt earlier reassessment. For example, prolonged illness associated with bed rest or forced immobility should be considered reasons for earlier reassessment, as would the use of medications known to influence fracture risk. In contrast, changes in risk factors that would diminish fracture risk should increase the interval between measurements.

C. Monitoring Therapy The quantitative analysis in Section IV,A also applies to the monitoring of therapy. Over the short term, it is difficult to determine whether an individual patient is responding to therapy because of the measurements error. Given misinformation on a patient’s rate of bone loss, the clinician can

actually be misled into making the wrong clinical decision, such as stopping or switching the treatment.

VI. THE EXPRESSION OF RISK A. Absolute and Relative Risk Fracture risk, as discussed thus far, is expressed as relative to young normals with no risk factors, but several other ways of expressing risk may be equally or more useful. Absolute risk, i.e., fractures per 1000 women per year, or lifetime fracture risk may be preferred. The probability of any event, e.g., a fracture, may be expressed in many ways. With 300,000 hip fractures among 250,000,000 Americans each year, the probability of a randomly chosen person suffering such a fracture is 0.0012, 12 chances in 10,000 each year. Improving on this estimate of risk was the purpose of the preceding discussion. Women of age 72 with bone mass near the median for their age and three to four additional risk factors suffered hip fractures at a rate of 56/10,000 per year in SOF. Extrapolating from the SOF data [24], women with similar bone mass who walked for exercise (relative hazard  0.7 compared with nonwalkers) would be expected to suffer fractures at a rate of 39/10,000 per year. Of course, this does not address the efficacy of recommending that women begin walking to change their risk profile, but it gives a rough estimate of how absolute risk differs depending on risk profile and bone mass. Absolute risks are small over any short time period. However, cumulative risk over a patient’s remaining life span can be considerable.

B. Lifetime Risk Melton and colleagues [44] explored the concept of lifetime fracture risk, and several other groups have published data on the probability of a fracture in one’s remaining lifespan. Approximately 40% of women suffer some sort of fracture between ages 50 and 85, and roughly one woman in six will experience a hip fracture [44]. How BMD and other potential risk factors affect lifetime fracture risk has been explored less thoroughly because of the lack of studies of sufficient duration and the assumptions required to extrapolate from short-term risk to lifetime risk. In one study where some subjects were followed for up to 15 years, 44% of those with radius BMD below 0.6 g/cm had experienced a nonspine fracture by 7.5 years follow-up, in contrast to 15% of those with bone mass above 0.6 g/cm. Melton and associates [45] have explored potential models for estimating lifetime fracture risk, but the lifetime risk estimates require

816 numerous assumptions [e.g., remaining years of life, longterm effects of risk factors that might change (e.g., BMD)] and the development of appropriate models. As an example, the projected decrease in future mortality increases the lifetime risk of hip fracture substantially [46]. Perhaps more importantly, the application of multiple risk factor data to lifetime risk estimates is more complex (although possible).

VII. SUMMARY Assessment of the risk for osteoporotic fractures is a complex issue. The prevention of fractures through selective application of therapeutic interventions aimed at preserving bone mass is intuitively appealing, but difficult to apply. However, several issues are clear. First, if therapy to prevent bone loss is being considered, then bone mass measurements are probably necessary for several reasons: (i) it is inappropriate to treat someone for low bone mass if bone mass is high, (ii) it is impossible to monitor the effectiveness of therapy without a baseline bone mass measurement, and (iii) there is currently no method of identifying those at either high or low risk of suffering fractures due to osteopenia without direct measurement of bone mass. Second, numerous factors appear to predict fracture risk independently of bone mass. Most of these factors have been studied in relation to hip fractures, but several seem to be associated with fractures at multiple sites. For example, a fracture after age 40 (or 50) is associated with a higher risk of other fractures later in life. In the absence of bone mass measurements, these additional risk factors could provide some basis for risk assessment, but their ideal use is in the context of a complete clinical assessment with bone mass measurements as a cornerstone. In this same vein, some women will have bone mass that is so low that interventions aimed at the preservation of bone mass may be of little value without additional protection (e.g., pads for the hips), and bone mass measurements are the only way of identifying such women. Finally, there is general agreement that fracture risk can be altered substantially with pharmacological intervention, especially with the newest generation of drugs coming onto the market. Precise application of bone mass measurements and additional risk factors will benefit from further study, but this should not cause delay in the use of these data now. It is inevitable that physicians and their patients will interpret the risks and benefits of therapies in ways appropriate to their own circumstances. However, there are now adequate data to move the assessment of fracture risk onto firmer scientific ground and to quantify some of the elements of risk. Whether a physician chooses to recommend treatment at 1 SD compared to young normal BMD or at the bottom one-third of patients at any given age, it is certain that those at highest risk are being

SLEMENDA, JOHNSTON, AND HUI

treated. Moreover, the increased information now available regarding risk factors can also be used to identify those individual at highest risk of suffering osteoporotic fractures with greater success.

References 1. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). 2. S. R. Cummings, D. M. Black, M. C. Nevitt et al., Appendicular bone density and age predict hip fracture in women. JAMA 262, 665 – 668 (1989). 3. L. J. Melton, E. J. Atkinson, W. M. O’Fallan, H. W. Wahner, and B. L. Riggs, Long-term fracture prediction by bone mineral assessed at different skeletal sites. J. Bone Miner. Res. 10, 1227 – 1233 (1993). 4. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Baseline measurement of bone mass predicts fractures in white women, Ann. Intern. Med. 111, 355 – 361 (1989). 5. R. Wasnich, P. D. Ross, L. K. Heibrun, and J. M. Vogel, Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am. J. Obstet. Gynecol. 153, 745 – 751 (1985). 6. P. Gardsell, O. Johnell, B. E. Nilsson, and B. Gullberg, Predicting various fragility fractures in women by forearm bone densitometry: A follow-up study. Calcif. Tissue Int. 52, 348 – 353 (1993). 7. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture osteopororis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1998). 8. S. R. Cummings, D. M. Black, M. C. Nevitt et al., Bone density at various sites for the prediction of hip fractures. Lancet 341, 72 – 75 (19993). 9. D. Marshall,O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312, 1254 – 1259 (1996). 10. C. W. Slemenda, S. L. Hui, C. Longcope, H. Wellman, and C. C. Johnston, Predictors of bone mass in perimenopausal women: A prospective study of clinical data using photon absorptiometry. Ann. Intern. Med. 112, 96 – 101 (1990). 11. D. T. Baran, A.M. Kelly, A. Karellas, M. Gionet, M. Price, D. Leahey, S. Steuterman, B. Mcsherry, and J. Roche, Ultrasound attenuation of the os calsis in women with osteoporosis and hip fractures. Calcif. Tissue Int. 43, 138 – 142 (1998). 12. D. Hans, P. Dargent-Molina, A. M. Schoot, J. L. Sebert, C. Cormier, P. O. Kotzki, P. D. Delmas, J. M. Pouilles, G. Breart, and P. J. Meunier, Ultrasonoraphic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 308, 511 – 514 (1996). 13. C. F. Njeh, C. M. Boivin, and C. M. Langton, The role of ultrasound in the assessment of osteoporosis: A review. Osteopor. Int. 7(1), 7 – 22 (1997). 14. T. F. Lang, J. H. Keyak, M. W. Heitz, P. Augat, Y. Lu, A. Mathur, and H. K. Genant, Volumetric quantitative computed tomography of the proximal femur: Precision and relation to bone strength. Bone 21(1), 101 – 108 (1997). 15. K. Lai, M. Rencken, B. L. Drinkwater, and C. H. Chestnut III, Site of bone density measurement may affect therapy decision. Calcif. Tissue Int. 53, 225 – 228 (1993). 16. P. J. Ryan, G. M. Blake, R. Herd, J. Parker, and I. Foegelman, Spine and Femur BMD by DXA in patients with varying severity spinal osteoporosis. Calcif. Tissue Int. 52, 263 – 268 (1993).

CHAPTER 33 Assessing Fracture Risk 17. S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, K. Ensrud, H. K. Genant, L. Palermo, J. Scott, and T. M. Vogt, Bone density at various sites for prediction of hip fractures. Lancet 341, 72 – 75 (1993). 18. D. M. Black, S. R. Cummings, H. K. Genant, M. C. Nevitt, L. Palermo, and W. Browner, Axial and appendicular, bone density predict fractures in older women. J. Bone & Miner. Res. 7, 633 – 638 (1992). 19. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 20. J. A. Kanis and WHO study group, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis,” WHO technical report series 843, 1994. 21. NOF guideline, National Osteoporosis foundation, Washington, DC, 1998. 22. K. G. Faulkner, E. V. Stetten, and P. Miller, Discordance in patient classification using t-scores. J. Clin. Desitomet. 2, 343 – 350 (1999). 23. P. D. Ross, J. W. Davis, R. Epstein, and R. D. Wasnich, Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 24. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Engl. J. Med. 322, 767 – 773 (1995). 25. O. Johnell, Risk factors, for osteoporosis fractures. In “Bone Densitometry and Osteoporosis” H. K. Genant, G. Guglielmi, M. Jergas, Springer Verlag, Berlin, 1998. 26. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Eng1. J. Med. 332, 767 – 773 (1995). 27. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant. simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 28. F. Duboeuf, D. Hans, A. M. Schott, P. O. Kotzki, F. Favier, C. Marcelli, P. F. Meunier, and P. D. Delmas, Different morphometric and densitometric parameters perdict cervical and trochanteric hip fracture: The EPIDOS study. J. Bone Miner. Res. 12(11), 1995 – 1902 (1997). 29. T. Yoshikawa, C. H. Turner, M. Peacock, C. W. Slemenda, C. M. Weaver, D. Teegarden, P. Markwardt, and D. B. Burr, Geometric structure of the femoral neck measured using dula-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 1053 – 1064 (1994). 30. C. C. Gluer, S. R. Cummings, A. Pressman, J. Li, K. Gluer, K. G. Faulkner, S. Grampp, and H. K. Genant, Prediction of hip fractures from pelvic radioraphs: The study of osteoporotic fractures. J.Bone Miner. Res. 9, 671 – 677 (1994). 31. P. Dargent-Molina, F. Favier, H. Grandjean, C. Baudoin, A. M. Schott, E. Hausherr, P. J. Meunier, and G. Breart, Fall-related factors and risk of hip fracture: The EPIDOS prospective study. Lancet 348, 416 (1996).

817 32. M. R. Law and A. K. Hackshaw, A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: Recognition of major effects. Br. Med. J. 315, 841 – 846 (1997). 33. D. Feskanich, W. C. Willett, M. J. Stampfer, and G. A. Colditz, Milk dietary calcium and bone fractures in women: A 12-year prospective study. Am. J. Public Health 87, 992 – 997 (1997). 34. R. G. Cumming, S. R. Cummings, M. C. Nevitt, J. Scott, K. E. Ensrud, T. M. Vogt, and K. Fox, Calcium intake and fracture risk: Results from the study of osteoporotic fractures. Am. J. Epidemiol. 145, 926 – 934 (1997). 35. W. A. Ray, M. R. Griffen, and W. Downey, Benzodiazepines of long and short elimination half-life and the risk of hip fracture. JAMA 262, 3303 – 3307 (1989). 36. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral, density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 37. K. E. Ensrud, J. Cauley, R. Lipschutz, and S. R. Cummings, Weight change and fractures in older women. Arch. Intern. Med. 157, 857 – 863 (1997). 38. C. Christiansen, M. A. Hansen, K. Overgaard, and B. J. Riis, Prediction of future fracture risk, In “Proceedings of the Fourth International symposium on Osteoporosis and Concensus Development Conference” (C. Christiansen and B.J. Riis, eds.), pp. 52 – 54. Handelstrykkeriet Aalborg Aps, Aalborg, Denmark, 1993. 39. S. L. Hui, C. W. Slemenda, and C. C. Johnston, The contribution of bone loss to postmenopausal osteoporosis. Osteopor. Int. 1, 30 – 34 (1990). 40. C. W. Slemenda, S. L. Hui, C. Longcope, and C. C. Johnston, Sex steroids and bone mass: A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 41. A. C. Looker, D. C. Bauer, C. H. Chestnut III, C. M. Gundberg, M. C. Hochberg, G. Klee, M. Kleerekoper, N. B. Watts, and N. H. Bell, Clinical use of biochemical markers of bone remodeling. Current status and furture direction. Osteopor. Int. 11(6), 467 – 480 (2000). 42. P. Ravan, M Bidstrup, C. Christiasen, and H. Lawaetz, Prediction of long term response in bone mass by markers: 4-year results from the Danish cohort of the EPIC study. J. Bone Miner. Res. 14, S370 (1999). 43. R. Branton and D. A. Percival, Measurement of urinary CTx using osteosal in patients on long term fosamax therapy. J. Bone Miner. Res. 14, S162 (1999). 44. L. J. Melton et al., How many women have osteoporosis. J. Bone Miner. Res. 7, 1005 – 1010 (1992). 45. L. J. Melton, S. H. Kan, H. W. Wahner, and B. L. Riggs, Lifetime fracture risk: An approach to hip fracture risk assessment based on bone mineral density and age. J. Clin. Epidemiol. 41, 985 – 994 (1998). 46. A. Oden, A. Dawson, W. Dere, O. Johnell, B. Jonsson, and J. A. Kanis, Lifetime risk of hip fractures is underestimated. Osteopor. Int. 8(6), 599 – 603 (1998).

CHAPTER 34

Outcomes of Osteoporotic Fractures

I. II. III. IV.

GAIL A. GREENDALE

Department of Medicine, Division of Geriatrics, University of California, Los Angeles, School of Medicine, Los Angeles, California 90095

ELIZABETH BARRETT-CONNOR

Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California 90293

V. Outcomes of Hip Fractures VI. Conclusions References

Introduction Definitions of Functional Outcomes Related to Fracture Outcomes of Wrist Fractures Outcomes of Vertebral Fractures

I. INTRODUCTION

tasks cover a wide range of activities from simple self-care to complex occupational duties. Functional tasks are generally classified by the level of difficulty into basic (BADL), intermediate or instrumental (IADL), and advanced (AADL) activities of daily living. BADLs are the rudiments of personal care, such as eating, dressing, and bathing. IADLs constitute those activities required to maintain independent living, e.g., cooking, shopping, and transportation [2 – 4]. AADLs are elective, vary by individual, and may be important components of the maintenance of personal satisfaction and well-being [5,6]. Examples of AADLs include recreational and intellectual pursuits. Deterioration of functional capability at any of these levels often follows osteoporotic fracture. Other outcomes of osteoporotic fracture include declines in physical capabilities (e.g., walking, bending, and other movements required in daily life), increased social support requirements, worsened economic status, and diminished quality of life, including depression and deterioration in perceived health. These outcomes can have an impact on

The consequences of osteoporotic fracture, particularly of the hip, are often calculated in economic terms. In the United States alone, the annual hospital, nursing home, and lost wage cost attributable to hip fracture may be as high as $10 billion [1]. However, there are other important ramifications of osteoporotic fractures. The human costs of osteoporosis include diminution in functional status, health status, and independence. From the individual’s perspective, quality of life is threatened by these declines; from a societal point of view, loss of functional independence is a major determinant of the need for in-home assistance or institutionalization.

II. DEFINITIONS OF FUNCTIONAL OUTCOMES RELATED TO FRACTURE Functional competence is defined as the ability to complete functional tasks and to discharge social roles. These

OSTEOPOROSIS, SECOND EDITION VOLUME 1

819

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

820

GREENDALE AND BARRETT-CONNOR

TABLE 1 Design

Case Series of Outcome after Wrist Fracture Sample size

Results

Reference

Patients treated with closed reduction under anesthesia 10 years postfracture

55 of 100 treated cases

Gartland and Werley rating: 47% excellent, 38% good, 11% fair, 4% poor 44% hand pain by analog scale 40% weaker in fractured than nonfractured hand 27% one or more components of algodystrophy

[8]

Patients treated at a university-based orthopedic clinic 2 – 6 weeks after cast removal

59 of 60 treated cases

29% subjective hand pain 39% tender by dolorimetry 5% hand swelling; 15% finger swelling 41% symptoms of vasomotor instability

[7]

Patients treated at a university-based orthopedic clinic 10 years postfracture

55 of 100 treated cases (85% of 10-year survivors)

Gartland and Werley rating: 49% excellent, 36% good, 11% fair, 4% poor 27% one or more features of algodystrophy

[9]

Case series of fractures treated nonsurgically between 1977 and 1980 at a university-based orthopedic clinic 1.5 – 6 years postfracture

297 of 640 cases

Gartland and Werley rating: 38% excellent, 49% good, 11.5% fair, 1.5% poor

[11]

Case series of casted or reduced fractures at a university-based orthopedic clinic 3 and 6 month follow-up

215 of 235 at 3 months 209 of 235 at 6 months

36% diminished grip strength 36% radio-ulnar joint pain Patients perceived outcome, 48% excellent, 32% good, 18% fair, 2% poor.

functional competence, but are not synonymous with it. For example, a wrist fracture may profoundly limit the job performance of a craftsman, but have minimal occupational consequence to an executive. This chapter summarizes information about the outcomes of wrist, vertebral, and hip fracture because they are common and have been studied with respect to functional impact. The majority of included studies did not specify the level of trauma associated with the fracture; however, because the preponderance of subjects were older men and women, it is likely that most fractures were related to skeletal fragility.

III. OUTCOMES OF WRIST FRACTURES Most studies of function after wrist fracture are clinical case series (Table 1). The length of follow-up has ranged from weeks [7] up to a decade [8,9]. Thus some short-term and some longer term data are available on the prevalence of symptoms and disabilities in select groups of patients.

Gartland and Werley rating: 41% excellent or good at 3 months 69% excellent or good at 6 months

[13]

A surprising proportion of patients report symptoms after wrist fracture. Across series, the prevalence of hand pain ranges from 29 to 44% and hand weakness between 36 and 40%. Algodystrophy (also termed Sudeck’s atrophy or reflex sympathetic dystrophy) also occurs after wrist fracture. While the definition of algodystrophy is not uniform, it usually is composed of hand pain, limited finger movement, and vasomotor instability [7]. The reported prevalence of postfracture algodystrophy varies widely, in part due to differing definitions, ranging from 0.1 to 47% [7,8]. Table 1 shows results from studies of wrist fracture outcome classified by the Gartland and Werley [10] scoring system. This scale records self-reported symptoms, such as pain and limitation of movement or function, and physician-assessed objective measures, such as range of motion and muscle strength. Items included in the Gartland and Werley scale are reproduced in modified form in Table 2. By these criteria, most patients with wrist fracture are classified as having a good or excellent outcome [8,9,11 – 13]. However, two major shortcomings of this scale must be noted. First, good results are overestimated; e.g., a patient with hand pain and diminished grip strength (and no other symptom or sign) would be

821

CHAPTER 34 Outcomes Of Osteoporotic Fractures

TABLE 2 Items Included in the Gartland and Werley Rating System for Evaluation of Healed Wrist Fracturesa Item

Demeritsb

Subjective evaluation (by patient) No pain, disability, or limitation of motion

0

Occasional pain, slight limitation of motion, no disability

2

Occasional pain, some limitation of motion, feeling of weakness in wrist, no particular disability if careful; activities slightly restricted

4

Pain, limitation of motion, disability, activities more or less markedly restricted

6

Objective evaluation (by physician) Dorsiflexion 45°

5

Palmar flexion 30°

1

Ulnar deviation 15°

3

Radial deviation 15°

1

Supination 50°

2

Pronation 50°

2

Circumduction

1

Pain in distal radionavicular joint

1

Grip strength: 60% or less of opposite side

1

Residual deformity (by X-ray) Prominent ulnar styloid Residual dorsal tilt

1 2

Radial deviation of hand

2 – 3c

Finger stiffness

1 – 2c

Nerve complications

1 – 2 – 3c

Arthritis

1 – 2 – 3c

a

Modified from Gartland and Werley [10]. Demerits are given for each sign or symptom. Scores are excellent (0 – 2), good (3 – 8), fair (9 – 20), and poor (21). c Can give higher demerits according to degree of severity. b

classified in the “good” outcome category. Second, there are no reported validation studies of the scale, constructed in 1951. One predictor of poor (Gartland and Werley) outcome is the appearance of radial shortening by X-ray [9,11,14,15]. A shortened radius may reduce the mechanical function of extensor tendons [14]. The presence of postfracture algodystrophy also correlates with a poorer Gartland and Werley outcome [7,8,12]. Impairments in activities of daily living are also related to wrist fracture. In one population-based cohort, after controlling for other diseases, women with a history of wrist fracture were three times more likely to report difficulty with shopping for groceries or clothing, nine times more likely to have difficulty cooking, and two to three times more likely to have difficulty getting into and out of a car or descending stairs than women who have never fractured their wrists [16].

IV. OUTCOMES OF VERTEBRAL FRACTURES A. Pain and Functional Outcomes Studies of pain and functional outcomes after vertebral fracture can be classified into two major categories: those concerned with the consequences of vertebral deformities and those that focus on the effects of clinical fractures. Vertebral deformity studies examine the associations between radiographic evidence of vertebral fracture (loss of vertebral height), only 30% of which are severe enough to be recognized clinically [17], and clinical outcomes. Studies of patients with clinically evident fractures investigate the functional impairments, physical limitations, and symptoms of clinically diagnosed osteoporotic vertebral fracture; patients presenting for the evaluation of a problem thought to be related to vertebral fracture (e.g., back pain) are typically the subjects of these studies. Outcomes associated with radiographically determined vertebral deformity are summarized in Table 3; overall, cross-sectional surveys find that symptoms and disabilities are more pronounced when the degree of deformity is high. Women with moderate to severe prevalent vertebral deformities (see grading criteria, Table 3) report more back pain, general disability, disability specifically attributed to back problems, poorer self-rated health, and greater embarrassment about their appearance than women without vertebral deformity [18,19,20]. Physical limitation in at least one of six movements (e.g., bending) due to back pain, pain-associated activity limitations, and doctor visits for back problems occur with increased frequency among women with severe prevalent vertebral deformity compared to those without such deformity [21]. Longitudinal studies suggest that relatively recent vertebral deformities, rather than remote ones, produce negative health outcomes [22,23]. The Honolulu Osteoporosis Study classified recent deformities as those that occurred in the past 4 years. The number of recent vertebral deformities was significantly related to the odds of functional impairment (difficulty in performing three or more basic or intermediate level activities of daily living) [22]. In contrast, deformities that occurred more than 4 years previously were unrelated to functional limitations. Using a similar time frame (about 4 years) to define incident fractures, the Study of Osteoporotic Fractures (SOF) also found a differential impact of incident compared to prevalent deformities [23]. Even severe (greater than 4 SD) prevalent vertebral deformity was unrelated to pain or limited function. However, one or more incident deformities significantly increased the odds of back pain, back-related disability, annual days of bed rest, and number of limited activity days. Adverse health outcomes may be associated with clinical vertebral fractures, but data are sparse (Table 4). In one

822

GREENDALE AND BARRETT-CONNOR

TABLE 3

Pain and Functional Outcomes Associated with Radiographically Determined Vertebral Deformity

Design/subjects

Sample size

Vertebral deformity grading

Results

Reference

Cross-sectional study of volunteers 55 – 75 years of age

204

Normal:  15% anterior height loss Minimal 15 – 20% anterior height loss Mid 20 – 25% anterior height loss Moderate:  25% anterior height loss; mid or posterior losses with end plate deformity Severe: Marked crush fracture Score: Fractures weighted by severity and summed

Minimal and mild deformities were not associated with physical, functional, or emotional limitations Moderate to severe deformity was associated with higher general disability scores, some specific disability attributed to back problems, and embarrassment

[18]

Same subjects as Ettinger et al. [18]

204

Same vertebral X-ray readings and scoring method as original Ettinger study with one additional classification: women were assigned to a severity category based on the highest grade of fracture they manifested

Risk of back pain and disability attributed to back problems was associated with higher fracture summary score and higher fracture severity classification

[19]

Cross-sectional from a population-based recruitment 65 – 70 years of age

2,992

Digitized computer reading used to grade each vertebra between T2 and L4 Severity was graded by standard deviation departures from normative values for each vertebral level in the cohort Classification based on (i) worst deformity by vertebral level. (ii) number of severe deformities, and (iii) worst deformity by wedge, end plate, and crush deformity

No increase in frequency of back pain was evident until grade of vertebral deformity reached 4 SD If at least one 4 SD deformity was present, more difficulty performing activities related to back function and a higher back disability score was found

[21]

Community-based cohort of men and women from 19 European countries 50 – 79 years of age

15,570

McKloskey method and Eastell and Melton method: Grade 1   3 SD and grade 2   4 SD compared to normative values

Limitations in back-related activities of daily living, back pain, and poorer self-related health associated with one prevalent deformity. More severe deformity and greater number of deformities associated with increasingly worse function, greater pain, and poorer health status.

[20]

Community-based cohort of Japanese-American women in Hawaii 55 – 93 years of age

569

Prevalent: 3 SD deformity Incident:  20% height loss in one dimension

Number of recent (within 4 years) vertebral deformities positively associated with greater impairment in activities of daily living and with more physician visits for back pain.

[22]

Community-based cohort of white women  65 years of age

7,223

Prevalent: 3 SD deformity compared to normal mean values for anterior, mid, or posterior dimensions Prevalent, severe: 4 SD deformity Incident:  20% and at least 4 mm vertebral height loss

Greater number of incident deformities positively related to greater back-related disability, annual number of beddays, and annual number of limited activity days

[23]

(continues)

823

CHAPTER 34 Outcomes Of Osteoporotic Fractures

TABLE 3 Design/subjects

Sample size

Volunteers for study of osteoporosis risk factors recruited from general practitioner’s offices 50 – 82 years of age

222

(continued )

Vertebral deformity grading Wedge (anterior) concavities and compressions (posterior) graded within vertebral level by comparing posterior height to inferior-antero-posterior dimension Vertebrac also graded relative to the adjacent superior and inferior vertebra (except T4 and L4)

Results

Reference

No association was found between number of deformities and back pain Grade 2 or 3 deformity and low BMD was not associated with more back pain

[61]

A higher prevalence of back pain was not reported by the women with severe deformity

[62]

Grade 1: 2 Sd deformity (N  120) Grade 2: 3 SD deformity (N  27) Grade 3: 4 SD deformity (N  8) 77% of 1307 age-eligible members of a Londonbased general practice 45 – 69 years of age

1035

McCluskey method of grading vertebral deformity based on up to four adjacent vertebrac Mild 2-2.99 SD deformity Severe 3 SD  deformity (N  20)

series of patients hospitalized for evaluation and treatment of vertebral fracture, 64% had pain and 70% had difficulty bending and rising [24]. Even within this highly selected, symptomatic sample, the number and severity of vertebral deformities (assessed by the Spinal Deformity index, Table 4) correlated modestly (0.29 to 0.44) with pain intensity, dysphoric mood, and degree of limitation in six physical movements (e.g., bending, lifting). In another study in the outpatient setting, women with one or more vertebral fractures had poorer self-rated function and measured physical performance (time to stand from a seated position) compared to a comparison group of patients with low back pain who did not have fractures [25]. The prevalence of symptoms, functional impairments, and negative emotional consequences in a group of 100 women, all of whom had chronic pain attributed to osteoporosis and at least one clinical vertebral fracture, is summarized in Table 5 [26]. The average time since fracture diagnosis in this study was 3.8 years, suggesting that some women with fracture have prolonged pain and disability; whether this is attributable to the fracture cannot be determined on the basis of these data. No correlation between

any quality of life domain and the number of vertebral fractures (median 2, range 1 – 11) was found. Physical performance impairments in trunk strength, walking speed, range of motion, and functional reach are also more common among women with several vertebral fractures than among nonfractured controls matched on age and comorbidity [27]. Instrumental activities of daily living may also be curtailed by vertebral fracture. In a cross-sectional population survey, women clinically diagnosed with vertebral fractures an average of 6 years previously were three times more likely to have difficulty cooking and shopping than comparable women without clinical vertebral fractures [16].

B. Illness, Hospitalization, and Mortality Outcomes Although the association between vertebral fractures and hospitalization has received little attention, one U.S. study found that hospitalization for vertebral fractures is not rare [28]. The sex-and race-specific incidence of

824

GREENDALE AND BARRETT-CONNOR

TABLE 4

Outcomes Associated with Clinically Diagnosed Vertebral Fractures

Design/subjects Cross-sectional clinical sample of patients admitted to hospital for evaluation or treatment 51 women, 61 ( 11) years 19 men, 52 ( 12) years

Cross-sectional survey selected from osteoporosis practice

Inclusions (i) postmenopausal, over 50; (ii) at least one vertebral fracture; (iii) clinical diagnosis of osteoporosis back pain Exclusions. (i) severe concomitant disease, (ii) unable to complete questionnaire Of 122 contacted women. 100 (82%) agreed to participate, average age 69 ( 8) years

Case-control study selected from a geriatric practice Cases: 10 women with documented vertebral fractures, 82 ( 6) years

Vertebral deformity grading

Results

Reference

Spinal deformity index (SDI): T3 to L5 anterior, posterior, and central heights calculated relative to T4, expressed as SD units SDI calculated by summing deviations between T3 and L5

Patients reported the following symptoms pain with activity (64%), limited bending (71%), and limited rising (70%) 41% needed help in self-care Correlation between SDI and physical limitations, 0.44

[24]

Vertebral fracture minimum criterion was 15% reduction in anterior compared to posterior height Anterior vertebral height loss graded as minimal, mild, moderate, or severe (grading criteria not given)

High prevalence of quality of life impairment (see Table 5)

[26]

No association between number of severity of fractures and severity of symptoms

Fracture criterion was a minimum of 30% or greater anterior wedge deformity Cases were required to have at least two vertebral fractures

Cases performed more poorly on all physical performance measures (e.g., lumbar spine motion, functional reach) Cases had more difficulty with activities of daily living and had more pain with activity

[27]

Height reduction of  20% in at least one vertebra

Cases were more dependent in self care, had more limitations in basic ADLS, and had poorer measured lower extremity strength

[25]

Controls: nonfractured female patients, 80 ( 6) years Major exclusions: inability to ambulare independently, poor visual acuity, failed cognitive status screen

Case-control study from an endocrinology practice Cases: 63 women with vertebral fracture, aged 65 ( 7.9) years Controls: 77 women with chronic low back pain and no spinal fracture, aged 56 ( 6.5) years

hospitalization due to osteoporotic vertebral fracture (1986 – 1989) was estimated using Medicare principal diagnosis of vertebral fracture, excluding persons under 65, and fracture diagnostic codes consistent with injury. Rates were calculated using the 1985 census as the de-

nominator. In women, there were 111,999 hospitalizations for vertebral fracture annually, a rate of 17.1 per 10,000. The annual rate in men was 3.7 per 10,000. Age-, gender-, and race-related differences in hospitalization rates were similar to those observed with hip fracture.

825

CHAPTER 34 Outcomes Of Osteoporotic Fractures

TABLE 5

Quality of Life in 100 Women with Osteoporotic Fracturesa Frequencyb

Importancec

Impactd

Pain

95

3.40

323.00

Pain when standing

85

3.47

295.00

Fatigue

73

3.40

248.00

pain when carrying

75

3.27

245.00

Pain when sitting

73

3.15

230.00

Difficulty carrying

87

3.23

281.00

Difficulty lifting

81

3.20

259.00

Difficulty walking

60

3.50

210.00

Difficulty bending

70

2.97

208.00

68

2.97

202.00

Vacuuming

81

2.95

239.00

Housework

70

3.19

223.00

Shopping for food

60

3.38

203.00

Shopping for clothes

61

3.30

201.00

Cleaning a bathtub

61

3.21

196.00

Afraid of falling

82

3.32

272.00

Afraid of fractures

74

3.57

264.00

Frustration

66

3.15

208.00

Anger

53

3.40

180.00

Overwhelmed

49

3.12

153.00

Traveling

57

3.30

188.00

Vacationing

41

3.41

140.00

Sports

43

3.07

132.00

Dancing

43

3.05

131.00

Attending church

35

2.94

103.00

Item Symptoms

Physical functioning

Difficulty finding comfortable chair Difficulty with activities of daily living

Emotions

Difficulty with leisure/social activities

a At least one clinically diagnosed vertebral fracture. From D. J. Cook, Arthritis Rheum. 36, 750–756 (1993). b The number of patients who cited the item as being a problem they experienced as a result of osteoporosis. c Rated by each patient, using a scale of 1 – 5, where 1 represents not at all important and 5 represents very important. d The product of frequency values times mean importance values (maximum possible score 500).

The consequences of vertebral fracture may include compromised respiratory function. In a consecutive case series of 132 women referred to a Canadian osteoporosis clinic, measured lung function was inversely related to the

number of prevalent vertebral fractures; forced vital capacity declined by about 10% for each prevalent thoracic anterior wedge deformity [29]. Independent of fracture number, the degree of kyphosis (assessed by Cobb’s angle) was associated with poorer lung function. (The correlation between Cobb’s angle and presence of vertebral fracture was only 0.5, suggesting that nonvertebral fracture-related kyphosis is also an important determinant of lung volumes [29].) Similarly, in a case-control study, women with prevalent vertebral fractures demonstrated forced expiratory flows (FEV1) that were 80% of predicted values (when standardized to FEV1 expected for their 25-year-old height) [30]. A comparison group of low back pain patients had significantly higher FEV1 values (92% of predicted) [30]. A decrease in survival has been noted after a clinically diagnosed vertebral fracture [31]. In one population-based cohort, the survival of patients with a clinical vertebral fracture was 61% compared to the expected population survival rate of 76% (relative survival 61/76, or 0.81). The survival rate of individuals with vertebral fractures diverged steadily from the expected rate over the course of follow-up. In the same population, the relative survival after hip fracture was similar (0.82), but the excess mortality was concentrated in the first 5 years after fracture, suggesting the excess early mortality after hip fracture likely represents the effect of fracture complications and comorbidity. Gradual widening over time between observed and expected mortality following vertebral fracture suggests that vertebral fracture is a marker for other conditions that increase the risk of death. Browner and colleagues [32] have shown that low bone mineral density (without fracture) predicts higher mortality, further supporting the concept that comorbidity and frailty often accompany osteoporosis. Women with radiographic vertebral deformities also experience higher all-cause mortality rates and are at higher risk of death due to pulmonary disease and cancer [33]. In SOF, compared to women without vertebral deformity, those with one or more fractures had a 23% greater ageadjusted mortality; the risk of death increased with the number of fractures. The significance of the relation between vertebral fractures and mortality was maintained after adjustment for numerous potential confounders, including smoking, bone density, self-reported health status, physical activity, and estrogen use (relative risk for one or more fractures 1.16) [33]. A large, multisite European cohort study found similar associations between vertebral deformity and mortality in both women and men [34]. The age-adjusted relative risk of death was 1.9 and 1.3 in women and men, respectively. After multiple adjustments for alcohol use, smoking, body mass index, self-rated health, and steroid use, relative risks or mortality was only slightly reduced 1.6 and 1.2 in women and men [34].

826

GREENDALE AND BARRETT-CONNOR

TABLE 6

Studies of Functional Outcome of Hip Fracturea

Design

Sample size/subjects

Type of fracture

Reference

Consecutive case series of subjects admitted to a rehabilitation ward whose hip fractures had occurred within 6 months or less

118 women 12 men Age range: 50 – 98 years

56% trochanteric 44% femoral neck

[35]

Consecutive patients over 54 years old with new hip fractures admitted to the university hospital

50 women 25 men

68% intertrochanteric 22% femoral neck

[37]

84% admitted from home

Age not reported

Consecutive case series of newly diagnosed hip fractures at one county and two private hospitals

63 women 29 men

51% femoral neck 45% intertrochanteric

[38]

Able to speak English; not disoriented

Mean age: 71 years

4% greater trochanter

Consecutive cases of fracture admitted to seven Baltimore area hospitals

442 women 94 men mean age: 78 years

53% extracapsular

[36]

a

Results of studies are shown in Table 7.

V. OUTCOMES OF HIP FRACTURES Functional competence in basic and intermediate activities of daily living and physical functioning is markedly diminished after hip fracture. Most studies have assessed postfracture declines in function compared to patient or proxy recall of function prior to the fracture [35 – 38]. One report measured function prior to and after hip fracture [39]. As shown in Tables 6 and 7, return to prefracture competence in ADL occurs in less than 50% of patients by 6 months after fracture; little further improvement in ADL is made by 1 year postfracture. Hip fractures also have a devastating effect on IADL: at 6 months postfracture, approximately one-fourth of patients regain their prefracture functional status, with no further recovery evident by 1 year after the fracture event. A similarly modest return to prior social/role function is obtained by most hip fracture TABLE 7

patients, with only 26% returning to premorbid levels of function. In these instances, recovery refers to a resumption of prior function, not to attainment of independent function. Patients who were independent in ADL and IADL prior to their hip fracture also suffer marked deterioration in functional status after fracture. For example, about half of those who dressed independently before fracture regained this ability and only one-third of patients resumed independent transferring (i.e., the ability to move from bed to chair or from chair to upright posture) [39]. Considering patients who reported independence in several ADL and basic mobility (bed to chair and toilet transfers, putting on socks and shoes, and indoor walking), 33% recovered independence in all functions after hip fracture. At 1 year after fracture, 40% of patients were independent in all BADL, compared to 70% prior to fracture. Physical performance deteriorates significantly after hip fracture. One study that recorded functional status prior to

Results of Studies Summarized in Table 6a

Walking (%)b

Basic ADL (%)b

6 months

1 year

6 months

1 year

24

40

43

48

53 (intertrochanteric) 79 (subcapital)

Intermediate ADL (%)b 6 months

1 year

[35]

33 (all fractures)

21

65 62 a

64

[37]

[38] 46

48

27

29

Marottoli et al. [39] study not included, as results are reported only for return to independent status. Percentages are those returning to prefracture levels, not necessarily to independent function.

b

Reference

[36]

827

CHAPTER 34 Outcomes Of Osteoporotic Fractures

hip fracture occurrence reported poorer physical function outcomes than those studies that estimated prefracture function by recall: at 6 months, only 15% of subjects with hip fracture could walk across a room independently compared to 75% at baseline. While 41% could walk one-half mile at baseline, 6% could do so 6 months after hip fracture. The ability to climb stairs was regained by only 8%; prior to fracture, 63% had been able to perform this activity. Other studies that employed recall of function prior to fracture found less severe, but quite substantial losses of mobility. By 66 months after fracture, between 24 and 62% regained prior capability in walking; these figures improved to 40 – 79% at 1 year. Hip fracture results in functional limitations in both intermediate and advanced activities of daily living. These limitations include diminished competence in money management, cooking, performance of housework, use of transportation, grocery shopping, carrying bundles, taking medications, visiting friends and engaging in community activities [16,35 – 37,40 – 42].

A. Predictors of Recovery after Hip Fracture Elders who fracture their hips tend to have other diseases and functional limitations prior to the occurrence of the fracture. Thus, the hip fracture is often not the sole factor leading to functional decline; rather, it is maybe the “last straw” effect of the hip fracture in the setting of other comorbidities and limitations that leads to the profound decrease in functional capability. Factors that predict better recovery from hip fracture are consistent with the concept that the outcome of hip fracture depends largely on the prior condition of individuals who suffer the fractures. Recovery of prefracture functional status and/or return to living at home is more common in patients who were younger [36,37,43], in better general health [37,41,45], not demented [36,43 – 45], and had larger social networks [36,44]. In some [36,37,47], but not all [47], studies intertrochanteric fractures were associated with better functional outcomes, fewer postoperative complications, and lower mortality than femoral neck fractures.

from 5 to 9% compared to 1 to 3% for similarly aged women [48]. The effect of race on mortality varies by gender; white men have a slightly higher mortality rate than black men, while the converse is observed for women [53]. Nonoperative treatment of hip fracture is associated with higher mortality [49,54], but the frailest patients are those most likely to be treated in this fashion. Length of hospitalization for hip fracture has declined substantially in the United States over the past decade, from 20 days in 1981 to 13 days in 1990 [48]. In parallel, in-hospital hip fracture mortality has diminished from 11% in 1970 to 3 – 4% in 1991, perhaps reflecting a shift in deaths to the nursing home setting, as well as improved medical and surgical care of fractures. Nevertheless, 1-year mortality after hip fracture in the United States has remained stable at 20 – 25% [48]. Similar to the pattern seen with in-hospital deaths, older age, male sex, and comorbidity are associated with higher mortality [36]. Fracture treatment also appears to be related to longterm survival; however, underlying differences in patients chosen for each operative procedure and for nonoperative treatment make the interpretation of differences difficult. Neither acute nor long-term mortality rates after hip fracture are attributable solely to hip fracture, as they represent all-cause mortality estimates. In general, age-specific mortality during the year after hip fracture exceeds by 6 – 14% all-cause mortality in comparable age groups. Estimates of the time after fracture during which mortality exceeds age-specific population norms range from 6 months to 4 years [36,55,56,47]. If hip fracture leads to excess mortality, would its prevention save lives? A recent analysis from SOF suggests that the answer to this question may be no [47]. After adjustment for many other predictors of mortality (such as age, health status, smoking, weight, exercise) women with hip or pelvic fracture were 2.4 times more likely to die compared to nonfracture controls. However, a detailed review of death certificates and hospital records of the 64 cases of hip or pelvic fracture revealed that only 14% of deaths were caused or hastened by the fracture. More commonly, the fractures were markers of underlying chronic diseases.

C. Institutionalization B. Mortality from Hip Fracture Early mortality occurring during the acute hospitalization for hip fracture is relatively uncommon. According to the U.S. National Hospital Discharge Surveys of 1988 and 1991, between 3 and 4% of patients died during hospital admission for hip fracture [48]. Older age, male sex, and poorer general health are related to higher in-hospital mortality [49 – 52]. Death rates for men ages 50 to 99 range

Wide variation in practice patterns and availability of services by region and country make it difficult to estimate the proportion of patients who are transferred to nursing homes after acute hospitalization for hip fracture. In the United States, this figure varies from one-fourth to three-fourths of hospital discharges after hip fracture [57,58]. In England, up to one-third of the hospital stay may be accounted for by the unavailability of nursing

828

GREENDALE AND BARRETT-CONNOR

TABLE 8

Sex

No.

Discharge Status and Destination for People with a Hip Fracture Treated in Short-Stay Nonfederal Hospitals in 1991a

Died in hospital (%)

Left against Discharged to home (%)

Discharged to another medical advice (%)

Discharged to a long-term care short-stay hospital (%)

Discharged alive; institution (%)

Destination not stated (%)

Discharged status not stated (%)

Both sexes

281,685

10,535 (4)

90,889 (32)

1390 (1)

36,278 (13)

108,756 (39)

29,079 (10)

4758 (2)

Men

68,541

5114 (7)

22,888 (33)

224 (1)

8866 (13)

24,467 (36)

6454 (9)

528 (1)

Women

213,144

5421 (3)

68,001 (32)

1166 (1)

27,412 (13)

84,289 (40)

22,625 (11)

4230 (2)

a

Modified from U. S. Congress, OTA, 1994.

home placements; it is possible that longer hospitalization would lead to fewer nursing home admissions, but this has received limited study. Using the 1988 and 1991 National Hospital Discharge Surveys, the U.S. Office of Technology Assessment [48] has compiled overall estimates of discharges to nursing homes from nonfederal hospitals (Table 8). The Post Acute Care Study estimated that the duration of nursing home placements after hip fracture is as follows: 1 month, 24%; 2 months, 8%; 3 months, 8%; 4 months, 8%; 5 months, 8%; 6 months, 10%; and 1 year or more, 34% [48]. However, characteristics other than hip fracture, such as old age, ADL and IADL dependence, and impaired mental status, predict a long nursing home stay [59,60]. Therefore, to estimate the use of services and expenditures related to hip fracture, the Office of Technology Assessment concluded that a maximum of 1 year’s nursing home stay should be attributed to hip fracture and that subsequent time in the nursing home is largely due to other factors such as dementia, comorbidity, frailty, and lack of social support.

VI. CONCLUSIONS Mortality and nursing home placement are well-recognized outcomes of osteoporotic hip fracture. The consequences of other types of osteoporotic fractures are less well studied and not widely appreciated. Wrist and vertebral fractures are associated with pain and limitations in multiple activities of daily living. Additional studies are required to further delineate the functional sequelae of nonhip fractures. The development of strategies to prevent loss of independence and function after osteoporotic fracture is an important goal for future research.

References 1. M. J. Parker, J. W. Myles, J. K. Anand et al., Cost – benefit analysis of hip fracture treatment. J. Bone J. Surg. 74B, 261 – 264 (1992). 2. S. Katz, T. D. Downs, H. R. Cash et al., Progress in development of the index of ADL. Gerontology, 10, 20 (1970). 3. M. P. Lawton and E. M. Brody, Assessment of older people: Selfmaintaining and instrumental activities of daily living measure. Gerontology 9, 179 (1969). 4. G. G. Fillebaum, Screening the elderly: A brief instrumental activities of daily living measure. J. Am. Geriatr. Soc. 33, 698 (1985). 5. D. B. Reuben and D. H. Solomon, Assessment in geriatrics: Of caveats and names. J. Am. Geriatr. Soc. 37, 570 – 572 (1989). 6. D. B. Reuben, D. L. Weiland, and L. Z. Rubenstein, Functional assessment in older persons: Concepts and implications. Facts Res. Gerontol. 7, 231 – 240 (1993). 7. R. M. Atkins, T. Duckworth, and J. A. Kanis, Features of algodiptrophy after Colles’ fracture. J. Bone J. Surg. 72(8), 105 – 110 (1990). 8. J. Field, D. Warwick, C. G. Bannister et al., Long-term prognosis of displaced Colles’ fracture: A 10-year prospective review. Injury 23, 529 – 532 (1992). 9. D. Warwick, J. Field, D. Prothero et al., Function after ten years after Colles’ fracture. Clin. Orthop. Rel. Res. 295, 270 – 274 (1993). 10. J. J. Gartland and C. W. Werley, Evaluation of healed Colles’ fractures. J. Bone J. Surg. 33A, 895 – 907 (1951). 11. M. Altissimi, R. Antenucci, C. Fiacca et al., Long-term results of conservative treatment of fractures of the distal radius. Clin. Orthop. Rel. Res. 206, 202 – 210 (1986). 12. R. H. Roumen, W. L. E. M. Hesp, and E. D. M. Bruggink, Unstable Colles’ fracture in elderly patients. J. Bone J. Surg. 73B, 307 – 311 (1991). 13. H. D. Stewart, A. R. Innes, and F. D. Burke, Factors affecting the outcome of Colles’ fracture: An anatomical and functional study. Injury 16, 289 – 295 (1985). 14. M. McQueen and J. Caspers, Colles’ fracture: Does the anatomical result affect the final function? J. Bone J. Surg. 70B, 649 – 651 (1988). 15. R. N. Villar, D. Marsh, N. Rushton et al., Three years after Colles’ fracture. J. Bone J. Surg. 69B, 635 – 638 (1987). 16. G. A. Greendale, E. Barrett-Connor, S. Ingles et al., Late physical and functional effects of osteoporotic fractures in women: The Rancho Bernardo Study. J. Am. Geriatr. Soc. 43, 955 – 961 (1995).

CHAPTER 34 Outcomes Of Osteoporotic Fractures 17. C. Cooper, G. Campion, L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992). 18. B. Ettinger, J. E. Block, R. Smith et al., An examination of the association between vertebral deformities, physical disabilities and psychosocial problems. Maturitus 10, 283 – 296 (1988). 19. P. D. Ross, B. Ettinger, J. W. Davis et al., Evaluation of adverse health outcomes associated with vertebral fractures. Osteopor. Int. 1, 134 – 140 (1991). 20. C. Mathis, U. Weber, T. W. O’Neill et al., Health impact associated with vertebral deformities: Resulta from the European vertebral osteoporosis study (EVOS). Osteopor. Int. 8, 364 – 372 (1988). 21. B. Ettinger, D. M. Black, M. C. Nevitt et al., Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). 22. C. Huang, P. D. Ross, and R. D. Wasnich, Vertebral fracture and other predictors of physical impairment and health care utilization. Arch. Intern. Med. 156, 2469 – 2475 (1996). 23. M. C. Nevitt, B. Ettinger, D. M. Black et al., The association of radiographically detected vertebral fractures with back pain and function: A prospective study. Ann. Intern. Med. 128, 793 – 800 (1998). 24. G. Leidig, H. W. Minne, P. Sauer et al., A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. 8, 217 – 229 (1990). 25. G. Leidig-Bruckner, H. W. Minne, C. Schlaich et al., Clinical grading of spinal osteoporosis of life components and spinal deformity in women and chronic low back pain and women with vertebral osteoporosis. J. Bone Miner. Res. 12, 663 – 675 (1997). 26. D. J. Cook, G. H. Guyatt, J. D. Adachi et al., Quality of life issues in women with vertebral fractures due to osteoporosis. Arthritis Rheum. 36, 750 – 756 (1993). 27. K. W. Lyles, D. T. Gold, K. M. Shipp et al., Association of osteoporotic vertebral compression fractures with impaired functional status. Am. J. Med. 94, 595 – 601 (1993). 28. S. J. Jacobsen, C. Cooper, M. S. Gottlieb et al., Hospitalization with vertebral fracture among the aged: A national population-based study, 1986 – 89. Epidemiology 3, 515 – 518 (1992). 29. J. A. Leech, C. Dulberg, S. Kellie et al., Relationship of lung function to severity of osteoporosis in women. Am. Rev. Respir. Dis. 141, 68 – 71 (1990). 30. C. Schlaich, H. W. Minne, T. Bruckner et al., Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteopor. Int. 8, 261 – 267 (1998). 31. C. Cooper, E. S. Atkinson, S. J. Jacobsen et al., Population-based study of survival after osteoporotic fractures. Am. J. Epidemiol. 137(9), 1001 – 1005 (1993). 32. W. S. Browner, D. G. Seeley, T. M. Vogt et al., Non-trauma mortality in elderly women with low bone mineral density. Lancet 338, 355 – 358 (1991). 33. D. M. Kado, W. S. Browner, L. Palermo et al., Vertebral fractures and mortality in older women: A prospective study. Arch. Intern. Med. 159, 1215 – 1220 (1999). 34. A. A. Ismail, T. W. O’Neill, C. Cooper et al., Mortality associated with vertebral deformity in men and women: Results from the European prospective osteoporosis study (EPOS). Osteopor. Int. 8, 291 – 297 (1998). 35. S. Katz, K. Heiple, T. Downs et al., Long-term course of 147 patients with fracture of the hip. Surg. Gynecol. Obstet. 124, 1219 – 1230 (1967). 36. J. Magaziner, E. M. Simonsick, T. M. Kashner et al., Predictors of functional recovery one year following hospital discharge for hip fracture: A prospective study. J. Gerontol. 45(3), M101 – M107 (1990). 37. A. M. Jette, B. A. Harris, P. D. Cleary et al., Functional recovery after hip fracture. Arch. Phys. Med. Rehabil. 68, 737 – 740 (1987).

829 38. S. R. Cummings, S. L. Phillips, M. E. Wheat et al., Recovery of function after hip fracture: The role of social supports. J. Am. Geriatr. Soc. 36(9), 801 – 806 (1988). 39. R. A. Maritolli, L. F. Berkman, and L. M. Cooney, Decline in physical function following hip fracture. J. Am. Geriatr. Soc. 40, 861–866 (1992). 40. G. Jarnlo, Hip fracture patients: Background factors and function. Scand. J. Rehabil. Med. Suppl. 24, 1 – 31 (1990). 41. L. Ceder, K. Svensson, and K. G. Thorngren, Statistical prediction of rehabilitation in elderly patients with hip fractures. Clin. Orthop. Rel. Res. 152, 185 – 190 (1980). 42. A. A. Guccione, D. T. Felson, J. J. Anderson et al., The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. Am. J. Public Health 84, 351 – 358 (1994). 43. J. M. Mossey, E. Mutran, K. Knott et al., Determinants of recovery 12 months after hip fracture: The importance of psychosocial factors. Am. J. Public Health 79(3), 279 – 296 (1989). 44. J. A. Van der Sluijs and G. H. I. M. Walenkamp, How predictable is rehabilitation after hip fracture? Acta Orthop. Scand. 62, 567 – 572 (1991). 45. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 46. Keene (1993). 47. W. S. Browner, A. R. Pressman, M. C. Nevitt et al., Mortality following fractures in older women. Arch. Intern. Med. 156, 1521 – 1525 (1996). 48. U. S. Congress Office of Technology Assessment, “Hip Fracture Outcomes in People Age 50 and Over: Background Paper,” OTA-BP-H120. U. S. Government Printing Office, Washington, DC, 1994. 49. A. H. Myers, E. G. Robinson, M. L. Van Natta et al., Hip fractures among the elderly: Factors associated with in-hospital mortality. Am. J. Epidemiol. 134(10), 1128 – 1137 (1991). 50. T. I. Davidson and W. N. Bodey, Factors influencing survival following fractures of the upper end of the femur. Injury 17, 12 – 14 (1986). 51. T. B. Young and A. C. C. Gibbs, Prognosis factors for the elderly with proximal femoral fracture. Arch. Emerg. Med. 1(4), 215 – 224 (1984). 52. J. G. Crane and C. B. Kernek, Mortality associated with hip fractures in a single geriatric hospital and residential health faciltiy: A ten-year review. J. Am. Geriatr. Soc. 31, 472 – 475 (1983). 53. S. E. Kellie and J. A. Brody, Sex-specific and race-specific hip fracture rates. Am. J. Public Health. 80(3), 326 – 328 (1990). 54. L. Matheny, T. F. Scott, C. M. Craythorne et al., Hospital mortality in 342 hip fractures. WV Med. J. 76(8), 188 – 190 (1980). 55. E. S. Fisher, J. A. Baron, D. J. Malenka et al., Hip fracture incidence and mortality in New England. Epidemiology 32(2), 116 – 122 (1991). 56. J. E. Kenzora, R. E. Mc Carthy, and J. D. Lowell, Hip fracture mortality. Clin. Orthop. Rel. Res. 186, 45 – 56 (1984). 57. J. F. Fitzgerald, L. F. Fagan, W. M. Tierney et al., Changing patterns of hip fracture care before and after implementation of the prospective payment system. JAMA 258(2), 218 – 221 (1987). 58. M. B. Gerety, V. Soderholm-Difatte, and C. H. Winograd, Impact of prospective payment and discharge location on the outcome of hip fracture. J. Gen. Intern. Med. 4, 388 – 391 (1989). 59. K. L. Kahn, E. B. Keeler, and M. J. Sherwood, Comparing outcomes of care before and after implementation of the DRG-based prospective payment system. JAMA 264(15), 1984 – 1988 (1990). 60. E. A. Chrischilles, C. D. Butler, C. S. Davis et al., A model of lifetime osteoporosis impact. Arch. Intern. Med. 151, 2026 – 2032 (1991). 61. P. H. Nicholson, M. J. Haddaway, M. W. Davie et al., Vertebral deformity, bone mineral density, back pain and height loss in unscreened women over 50 years. Osteopor. Int. 3, 300 – 307 (1993). 62. T. D. Spector, E. V. McCloskey, D. V. Doyle et al., Prevalence of vertebral fracture in women and the relationship with bone density and symptoms: The Chingford Study. J. Bone Miner. Res. 8, 817 – 822 (1993).

CHAPTER 35

The Nature of Osteoporosis ROBERT MARCUS SHARMILLA MAJUMDER

Veterans Affairs Medical Center, Palo Alto, California 94304 Department of Radiology, University of California, San Francisco, San Francisco, California 94143

I. Defining Osteoporosis II. The Nature of Osteoporotic Bone

III. Conclusions References

I. DEFINING OSTEOPOROSIS

related to menopausal estrogen loss and the other to aging. This concept has been elaborated upon by Riggs and associates [2] who suggested the terms Type I osteoporosis, to signify a loss of trabecular bone after menopause, and Type II osteoporosis, to represent a loss of cortical and trabecular bone in men and women as the end result of age-related bone loss. Whereas the Type I disorder directly results from lack of endogenous estrogen, Type II osteoporosis reflects the composite influences of long-term remodeling efficiency, adequacy of dietary calcium and vitamin D, intestinal mineral absorption, renal mineral handling, and parathyroid hormone (PTH) secretion. This model is discussed in detail in Chapter 38. Although there is heuristic value to defining subsets of patients in this manner, compelling validation of the model remains to be offered. Iliac crest biopsies do not show a characteristic histomorphometric profile of patients whose clinical status suggests type I osteoporosis. Most importantly, the model suffers by adhering to a bone loss paradigm. Postmenopausal women with low bone mass are assumed to have achieved that status because they experienced a drastic menopausal loss of bone. However, we understand that bone mass at any time in adult life reflects the peak investment in bone mineral at skeletal maturity minus that which has been subsequently lost. A woman who experienced interruption of menses, extended bed rest, eating disorder, or systemic illness during her

This chapter introduces a series of contributions dealing with osteoporosis in a wide variety of clinical settings. Its purpose is to consider the definition of osteoporosis and to discuss the nature of osteoporotic bone, its mass and distribution, its microscopic architecture, and other aspects of its quality and strength. Osteoporosis is a condition of generalized skeletal fragility in which bone strength is sufficiently weak that fractures occur with minimal trauma, often no more than is applied by routine daily activity. “Primary” osteoporosis is by tradition a skeletal disorder of postmenopausal women (“postmenopausal” osteoporosis) or of older men and women (“senile” osteoporosis). The term “secondary” osteoporosis refers to bone loss resulting from specific, defined clinical disorders, such as thyrotoxicosis or hyperadrenocorticism. Because of the intimate relationship of reproductive hormone deficiency to the development of postmenopausal osteoporosis, several conditions that could legitimately be considered secondary forms of osteoporosis are commonly treated as variants of primary osteoporosis. These include osteoporosis resulting from exercise-related amenorrhea and from prolactinsecreting tumors. Albright and Reifenstein [1] proposed in 1948 that primary osteoporosis consists of two separate entities, one

OSTEOPOROSIS, SECOND EDITION VOLUME 2

3

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

4 adolescent growth years might enter adult life having failed to achieve the bone mass that would have been predicted from her genetic or constitutional profile. If she then underwent a perfectly normal rate of bone loss, her skeleton would still be in jeopardy simply due to the deficit in peak bone mass. Thus, for the present, it may be most appropriate to consider osteoporosis the consequence of a stochastic process, that is, multiple genetic, physical, hormonal, and nutritional factors acting alone or in concert to diminish skeletal integrity. Although historical artifacts show that characteristic deformities of vertebral osteoporosis were recognized in antiquity [3], broad awareness of this condition has come about only during the past few decades, catalyzed in particular by the work of Albright and colleagues [1]. Unfortunately, because traditional radiographic techniques cannot distinguish osteoporosis until it is severe, confirmation of the diagnosis remains problematic. Until recently, diagnosis was by necessity clinical, requiring a history of one or more low-trauma fractures. Although highly specific, such a grossly insensitive diagnostic criterion offers no assistance to physicians who hope to identify and treat affected individuals who have been fortunate not yet to have sustained a fracture. The introduction of accurate noninvasive bone mass measurements afforded the opportunity to make an early diagnosis of osteoporosis. Bone mineral density (BMD) of patients with osteoporotic fractures was generally found to be lower than that of age-matched nonfractured controls. However, it soon became evident that substantial overlap exists in the distribution of BMD of patients with and without osteoporotic fracture [2,4 – 6] (Fig. 1) and that BMD does not accurately predict the presence of osteoporotic fractures (Fig. 2) [6].

MARCUS AND MAJUMDER

Several factors underlie the poor ability of BMD measurements to predict fracture prevalence. The normative data against which BMD comparisons are most often made have been determined for Caucasian men and women, and do not necessarily apply to other ethnic groups. BMD is clearly related to body weight, yet routine clinical bone mass assessments are not weight-adjusted. Various features of bone geometry that affect bone strength and fracture risk are not generally considered in the clinical interpretation of bone mass measurements. These include bone size, the distribution of bone mass around its bending axis (moments of inertia), and some derivative functions, such as the hip axis length [7]. Moreover, bone mass determinations cannot distinguish individuals with low mass and intact microarchitecture from those with equal mass who have trabecular disruption and cortical porosity. They also cannot distinguish other aspects of bone quality. Finally, some patients with skeletal fragility have enjoyed the good fortune not to experience a fracture by the time of their bone mass measurement. These individuals, who would be designated “nonfractured,” may nonetheless be in severe jeopardy for fracture in the near future. Thus, as a means of diagnosing the presence or absence of osteoporotic fracture, BMD assessment is not useful.

FIGURE 2 FIGURE 1

Individual lumbar spine BMD values for 84 women with one or more vertebral compression fractures compared to age-predicted mean values (regression line) and 90% confidence interval (shaded region). Results show substantial BMD overlap of fracture patients with the normal range. Reproduced from The Journal of Clinical Investigation (2) by copyright permission of the Society for Clinical Investigation.

Receiver-operated curve (ROC) relating the sensitivity and specificity of bone density measurement techniques for diagnosis of prevalent fractures. DPA, dual photon absorptiometry; QCT, quantitative computed tomography; TBC, total body calcium; SPA, single photon absorptiometry. A perfect diagnostic instrument would give a line that reaches the top left-hand corner of the box. Reprinted from Ott et al., with permission (6).

5

CHAPTER 35 Nature of Osteoporosis

A very different conclusion is reached when bone mass is used to assess an individual’s prospective long-term risk for fracture. This topic is fully discussed in Chapters 33 and 59. Briefly stated, large prospective studies have shown that a reduction in BMD of 1 standard deviation from the mean value for an age-specific population confers a two- to three fold increase in long-term fracture risk [8 – 11]. In a manner similar to that by which serum cholesterol concentration predicts risk for heart attack or blood pressure predicts risk for stroke, BMD measurements can successfully identify subjects at risk of fracture and can help physicians select those individuals who will derive greatest benefit for initiation of therapy. In 1994, a group of senior investigators in this field offered a working definition of osteoporosis based exclusively on bone mass [12]. The reasoning behind this proposal, made on behalf of the World Health Organization (WHO), was that the clinical significance of osteoporosis lies exclusively in the occurrence of fracture, that bone mass predicts long-term fracture risk, and that selection of rigorous diagnostic criteria would minimize the number of patients who are incorrectly diagnosed. The authors suggested a cutoff BMD value of 2.5 standard deviations below the average for healthy young adult women. Using this value, approximately 30% of postmenopausal women would be designated as osteoporotic (Fig. 3), which gives a realistic projection of lifetime fracture rates. In addition, Kanis et al. [12] proposed that BMD values of 1 – 2 standard deviations below the young adult mean be designated as “osteopenic.” Such values identify individuals at increased risk for fracture, but for whom a diagnosis of osteoporosis would not be justified since it would mislabel far more individuals than would actually be expected ever to fracture. This approach has proven useful for clinical management, but has several limitations, all of which were acknowledged by the authors. Although a significant relationship between BMD and fracture is likely to hold for men, no evidence yet suggests that the cutoff of 2.5 SD carries the same risk as for women. The applicability of this criterion to young people prior to their acquisition of peak bone mass would be inappropriate. The BMD measurement is itself subject to several confounding factors, including bone size and geometry. As BMD correlations among skeletal sites are not strong, designating a person “normal” based on a single site, for example the lumbar spine, necessarily overlooks individuals with low bone density elsewhere, such as the hip. It seems reasonable to suppose that adjustment of bone density readings for such factors as body size, bone geometry, and ethnic background might improve the accuracy of this technique. It should be evident that the WHO proposal [12] will be of limited use to investigators whose interest is the nature and causes of osteoporosis. Knowledge of a low bone density at a particular point in

FIGURE 3

Bone mineral density in women at different ages and the prevalence of osteoporosis, defined as a BMD value 2.5 standard deviations below that for a healthy 25-year-old woman. Reprinted from Kanis et al., with permission (12).

time offers no information regarding the adequacy of peak bone mass attained, the amount of bone that may have been lost, or the quality of bone that remains.

II. THE NATURE OF OSTEOPOROTIC BONE At a recent consensus development conference [13], osteoporosis was defined as “a disease characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to enhanced bone fragility and a consequent increase in fracture risk.” Implicit in this definition is the view that osteoporosis results from deficits in normally composed bone and that the residual bone is defective in amount and distribution, but not in matrix composition or mineralization. In contrast, osteomalacic bone matrix is grossly undermineralized. Moreover, for many years the prevailing view has been that osteoporosis develops

6

MARCUS AND MAJUMDER

through excessive loss of bone. Only recently has attention been drawn to abnormalities in bone acquisition as a basis for subsequent bone fragility (see Chapter 25). This latter issue notwithstanding, the dominant model of osteoporosis among workers in the field has, until recently, emphasized only the amount and distribution of bone substance. However, the great overlap in bone density between individuals with and without fracture indicates the limitations of such a model to account adequately for individual differences in fracture susceptibility. In other words, additional properties of bone quality may contribute to skeletal fragility. Several lines of evidence support such a view. McCabe et al. [14] used a technique called Mechanical Resistance Tissue Analysis to show an age-related decrease in ulnar bending stiffness in women that was associated with, but not completely accounted for by changes in either bone mineral or gross bone geometry. Kann et al. [15] demonstrated an age-related decrease in the resonance frequency of cortical bone in women. Transmission velocity of ultrasound, a property related to bone material and structural properties, degrades with age, and has a relationship to fracture risk that is not fully accounted for by bone mass itself [16]. In 1993, the National Institute on Aging reported the proceedings of a multidisciplinary review of bone quality in osteoporosis [17], from which one may safely conclude that very little was known in this area. During the past several years, considerable effort has been focused on adapting histomorphometric analysis as well as developing noninvasive methods to the study of bone quality. In the remainder of this chapter, we review several aspects of this problem (Table 1). At the outset, however, we emphasize that such qualitative abnormalities, like decreased bone mass, are commonly associated with normal human aging and do not appear to be specific markers for osteoporosis.

A. Alterations in Bone Composition Considerable evidence indicates that the mineral composition of bone is neither homogeneous nor constant throughout life, and substantial heterogeneity in the mineralization of bone matrix can easily be shown using microra-

TABLE 1

Aspects of Bone Quality in Osteoporosis

Cortical porosity Undermineralized matrix Cement line accumulation Trabecular thinning, perforation and disruption Fatigue accumulation

FIGURE 4

Microradiograph of a section from the femoral shaft of a 62year-old man demonstrating microheterogeneity of matrix mineralization. Different gray levels reflect differences in mineralization, white representing the highest level. From Grynpas (18) with permission.

diography [18] (Fig. 4). As early as 1960, Jowsey [19] reported that the abundance of “low density” osteons was increased in biopsies obtained from subjects older than 50 years. This observation, as well as others [20], suggests either that a greater fraction of the total osteonal pool is actively engaged in remodeling, so that at any given time more remodeling units are encountered early in the mineralization process, or that completion of osteonal mineralization somehow deteriorates with age. As discussed by Crofts et al. [21], the modest rise in remodeling activation frequency observed after age 50 does not adequately explain the higher prevalence of lower-density osteons. 1. DECREASED MATRIX MINERALIZATION The first systematic effort to assess the composition of human osteoporotic bone was that of Burnell et al. [22], who compared iliac crest biopsies from osteoporotic postmenopausal women with vertebral compression fractures to biopsies from normal controls. As expected, osteoporotic bone was less dense. However, the fraction of mineral per gram of bone tissue was also reduced. Moreover, within the mineral phase, carbonate and the calcium-to-phosphorus ratio were decreased, while sodium and magnesium content were increased, yet the same biopsies gave no hint of osteomalacia. Although these results describe average values for the entire study cohort, they reveal considerable heterogeneity in bone composition, even within this group of clinically homogeneous patients. Most patients had normal results; one quarter showed undermineralized matrix, and only a few showed decreased matrix but normal mineralization. The authors also found that the subjects with decreased mineral fraction were those who also had an increased content of sodium and magnesium in the

7

CHAPTER 35 Nature of Osteoporosis

mineral phase, suggesting the presence of skeletal calcium deficiency. What is the additional mechanical consequence of undermineralization to an already porotic bone? While no published clinical fracture data address this question, mineralization is known to contribute importantly to bone structural strength [23], and mineral content strongly affects fundamental bone material properties, such as Young’s modulus of elasticity. In fact, a modest 7% increase in bone mineral content is associated with a threefold increase in bone stiffness and a doubling in breaking strength [24]. Thus, it seems inescapable that undermineralization would promote bone fragility. In a follow-up study, Burnell et al. [25] showed that daily supplements of calcium and vitamin D improved the calcium content of bone tissue while simultaneously decreasing its sodium content. Given prevailing views about pathogenesis of osteoporosis, in which simultaneous turnover of matrix and mineral is the consequence of bone remodeling, it is not clear how alterations in remodeling activity alone would eventuate in a relatively undercalcified bone matrix. The authors offered two explanations for this finding. First, they proposed that bone formation occurring in an environment of restricted calcium supply could lead to failure of bone to achieve its full mineralized potential. Alternatively, given the extensive hydroxyapatite surface area that reaches chemical equilibrium with bone interstitial and marrow fluids, a deficiency in dietary calcium or a decrease in calcium absorption efficiency, perhaps brought about by marginal vitamin D status, could slightly decrease bone fluid Ca2, with a shift in chemical equilibrium that favors calcium loss. To maintain electrical neutrality, bone calcium would then be replaced by sodium or magnesium. There is at present no basis for favoring either of these models, and the results of the calcium intervention study [25] are compatible with both. In fact, two other mechanisms also merit consideration. To provide additional carbonate and phosphate buffer to the extracellular fluid, the labile bone mineral pool is mobilized during sustained metabolic acidosis, a compensation that passively increases urinary calcium excretion. The concept that skeletal depletion reflects lifelong buffering of diets rich in animal protein (so-called “acid-ash diets”) persists as a minority view regarding the pathogenesis of osteoporosis [26]. It is possible that the undermineralized matrix in some of the patients of Burnell et al. [22] reflects such a buffering process. Alternatively, in the presence of an overall increase in remodeling activity, newly remodeled osteons might not have sufficient time to become fully mineralized prior to the next wave of remodeling at the same site. If that were the case, the average mineral content of each completed osteon would be decreased, and if this process continued for a sustained period, the average degree of matrix mineralization would be reduced. With calcium and vitamin D therapy

remodeling would be suppressed and newly completed osteons would have the chance to become fully mineralized. 2. SPATIAL HETEROGENEITY OF MINERALIZATION The work of Crofts et al. [21] emphasizes the analytical complexity that must be brought to compositional studies, particularly of cortical bone, and indicates heterogeneous mineralization within the femur, and indeed, within individual osteons, that may affect bone material properties and strength. Using backscattered electron imaging, the authors described age- and locational differences in osteonal mineralization among different regions of femoral cortex. They noted a 12% decrease in ash content of bones from an older group compared to those from young adults, and reported that mineralization decreased with distance from the central Haversian canal. These results indicate a decrease in mineral content of equivalent magnitude to that shown to have substantial mechanical consequences [24]. Moreover, they support previous work [27,28] demonstrating a consistent pattern in cortical bone of relatively increased mineralization of regions immediately subtending central Haversian canals. These results have major, although imperfectly understood, implications regarding the site within bone where fractures are likely to be initiated [29]. 3. FLUORIDE ACCUMULATION IN BONE CRYSTAL Incorporation of fluoride into the hydroxyapatite crystal increases crystal brittleness, so lifelong fluoride exposure is potentially another influence on age-related changes in bone mineral properties. The fluoride content of potable water varies from almost 0 to as much as 3 – 4 mg/liter. Communities that regulate the fluoride content of drinking water to minimize dental carries provide a level of ~1 mg/liter (1 ppm). Concentrations of  3 mg/liter may be associated with mottling of dental enamel. When fluoride content is higher, the risk of skeletal fluorosis increases, particularly in hot environments where increased daily water consumption may be required. The relationship of lifelong fluoride exposure to osteoporosis is not certain. The earliest epidemiological studies suggested that higher exposure to fluoride in water was associated with fewer consequences of osteopenia [30]. Subsequent work has not confirmed this view [31], and even suggests higher fracture rates in populations exposed to high [32] or even apparently optimally regulated fluoride exposure [33]. Richards et al. [34] evaluated vertebral trabecular bone mass, mechanical strength, mineral content, and fluoride burden in cadaveric bone samples. As one might predict, bone mass and strength both decreased with age. Bone fluoride content, however, more than tripled from age 20 to 80 years. However, no independent effect of bone fluoride content could be shown on mechanical strength. The authors concluded that long-term fluoride consumption does not independently affect bone quality once the effects of age and

8

MARCUS AND MAJUMDER

sex on bone mass are taken into account. Discussion of the effect of fluoride on bone mass and rates of loss, as well as its therapeutic effects at pharmacologic amounts is presented elsewhere in Chapters 27, 74, and 75. As pointed out above, traditional descriptions of osteoporosis reflect the belief that no systematic differences can be consistently demonstrated between osteoporotic and normal bone. From the evidence described here it can be seen that although this view may be true when bone specimens are grossly compared, careful examination of biopsy material reveals mineralization to be spatially heterogeneous and variable with age and demonstrates subtle, but clinically meaningful, alterations in bone composition for at least a subgroup of osteoporotic patients.

B. Loss of Trabecular Connectivity Normal trabecular bone resembles a honeycomb. It consists of a highly connected network of vertical and horizontal plates, called trabeculae. By contrast, osteoporotic trabecular bone shows apparent replacement of plates by rods, and obvious trabecular disruption (Fig. 5), giving rise to the view that loss of connectivity is an important component of skeletal fragility in osteoporosis. In this section we review trabecular bone structure, the methods by which connectivity has been traditionally assessed, attempts to evaluate the role of connectivity in osteoporosis, and new developments in the three dimensional analysis of trabecular bone. Human trabecular bone is generally anisotropic with respect to its mechanical properties and architecture. That is

FIGURE 5

to say, the bone does not look or behave identically when held in one position compared to its appearance and behavior following rotation by 90°. Trabecular anisotropy reflects the manner in which gravitational stresses are transmitted through the skeleton. Because human locomotion is bipedal and upright, vertical trabeculae are thicker (approximately 200 m) than horizontal trabeculae [35]. By contrast, arboreal primates show isotropic trabecular bone; i.e., the bone appears identical regardless of which side is up. Trabecular distribution is heterogeneous within each vertebral body, so that central regions are dominated by thinner vertical plates. Although the vertebral shell is called cortical, it is not histologically identical to compact bone and may more properly be considered a condensation of trabecular bone [35]. Although horizontal trabeculae are shorter and thinner than vertical trabeculae, they make an important contribution to trabecular strength. Figure 6 illustrates the difference in ultimate breaking strength (called the Euler buckling load) of supported and unsupported columns of similar dimension. In this example, a single horizontal connecting element confers a fourfold increase in load-bearing capacity [36]. 1. ASSESSMENT OF TRABECULAR CONNECTIVITY ANATOMIC SPECIMENS

IN

Several laboratories have brought to bear multiple techniques for characterizing normal trabecular architecture and its changes with age. Most work has focused on vertebral microarchitecture, because vertebrae constitute an important site of clinical fracture, and because they, along with

Scanning electron micrograph of normal (left) and osteoporotic (right) vertebral trabecular bone. Note transformation of trabecular plates to rods and trabecular perforations (photograph by Dr. Jon Kosek, reproduced with permission).

CHAPTER 35 Nature of Osteoporosis

FIGURE 6

Effect of horizontal trabeculae on resistance to buckling. The pillar on the left is not stabilized by connecting elements and is assigned a critical buckling load (Pcr) of 1.0. The pillar on the right has a single horizontal connection approximately halfway along its length. Buckling strength is four-fold greater (Pcr  4.0). Reproduced from Snow-Harter and Marcus (36) with permission.

the pelvis and proximal femur, are the sole repository of red marrow in adults, bringing their trabecular surfaces into close proximity to the cells that initiate bone remodeling. With age, characteristic changes first appear in the vertebral centrum and spread radially. The trabecular network, the superior and inferior vertebral endplates, and the cortical rim all attenuate. This, together with progressive loss of trabecular connections, results in a severe loss of mechanical competence for individual vertebral bodies. The loadbearing capacity of a vertebral body approximates 1000 kg in young adults, but is markedly reduced in older people to about 100 kg [37]. Evidence has been presented that loss of horizontal trabeculae occurs earlier and to a greater extent than vertical trabeculae [38,39], a view that dominates thinking in this area. However, some reports have not confirmed preferential loss of horizontal trabeculae [40,41]. For example, Vogel et al. [41] showed that the number of vertical trabeculae is approximately twice that of horizontal trabeculae throughout adult life, and that the slopes of loss for each are parallel. Although these studies appeared to be fastidiously carried out, the number of measurements was insufficient to permit definite conclusions. Changes in trabecular structure with aging include thinning and loss of contiguity. The first systematic approach to analyzing the relative contributions of these two processes was reported by Parfitt et al. [42], who found that trabecular thickness did not decrease with age in women, forcing the conclusion that loss of trabecular connections was the major event. Subsequent work from Weinstein and Hutson [43] indicated that thinning and disruption are both important aspects of bone loss, whereas that from other laboratories [44,45] suggested that trabecular thinning occurs with age in men, but not in women. The weight of evidence, therefore, indicates the primary feature in women to be loss of entire trabecular elements [42,44 – 46].

9 Stereologic assessment of trabecular connectivity is not a trivial undertaking and involves a series of assumptions regarding isotropy and obliquity of tissue sections. Vesterby and colleagues [47 – 50] pointed out difficulties with these assumptions and proposed a new variable called the Star Volume, which they define as “the mean volume of all the parts of an object which can be seen unobscured from a random point inside the object in all possible directions.” It is thus a stereological estimate that applies without the aforementioned assumptions regarding isotropy that limit earlier methods. The mean star volume is obtained from a sufficient number of measurements to provide statistical power, generally about 200. Very long values for individual measurements are obtained only when there is no trabecular bone to obliterate the view. Using a limited number of specimens, Vesterby et al. [49] showed a striking agerelated increase in marrow space star volume of both the lumbar vertebrae and the iliac crest in men and women. Using a different approach, Vedi et al. [51] employed a technique of strut analysis that is based on the classification of trabeculae, or struts, and on the recognition of junctions between three or more struts (called nodes) and of free ends. By this method, the number of free ends is inversely related to trabecular connectivity, whereas nodes are indicative of trabecular connections. Vedi et al. [51] showed impressive correlations between the density of nodes and struts and trabecular bone volume on iliac crest biopsies. All methods discussed to this point require extrapolation of 2-D biopsy information to 3-D. Vogel et al. [41] pointed out that results will vary by up to 100% if trabeculae resemble rods as opposed to plates. They also introduced an ingenious direct approach to measure both 2-D and 3-D architecture in the same biopsy samples. The results of a small number of samples indicate that age-related decreases in trabecular bone volume are due primarily to transformation of trabecular plates into rods by multiple perforations, a process discussed by Parfitt [52]. Direct visualization of perforations in the 3-D specimens showed them to be confined to trabecular plates at younger ages, but present also in rods at older ages. Another novel technique, fractal geometry, has been widely used to characterize topological features of objects in astronomy, chemistry, and other physical sciences, and has recently been applied to bone structure [53 – 55]. Fractal analysis describes objects with rough surface features, such as seacoasts, and may be ideally suited to assess the microstructure of trabecular bone. Similarity of surface topology over a range of magnification is a fundamental property of fractal objects, that is, any piece of a fractal, appropriately magnified, resembles the whole. The surface ruggedness of a fractal may be defined by its fractal dimension (D), a mathematical term describing the manner in which the object fills space. Buckland-Wright et al. [54] combined high-resolution radiographs and fractal analysis

10

MARCUS AND MAJUMDER

to assess lumbar spine structure in a large group of postmenopausal women who were stratified into high and low BMD groups. In radiographs of healthy vertebral bodies, the presence of large vertical plates reduced the number of visible structures, so that the derived fractal dimension differed from that of porotic vertebrae, where vertical plates were reduced to narrow rods, intertrabecular space was enlarged, and previously hidden posterior structures, such as fragmented trabeculae, became increasingly visibile. The results were compatible with the view that trabecular bone loss consists of focal fenestration of vertical plates into a lattice of bars and rods [42]. Gundersen et al. [56] took fundamental issue with the validity of even attempting to estimate the connectivity of three-dimensional objects by anything less than a threedimensional sample. They adapted the Euler number (a topological feature based on the number of holes and connected elements in an object) and a method for calculating the Euler number in thin sections (the Conneuler) to skeletal analysis. Although this technique shows promise, no data concerning age- or osteoporosis-related changes have yet been presented. 2. RECENT DEVELOPMENTS TRABECULAR STRUCTURE

IN

THE

ANALYSIS

OF

As described above, evidence suggesting an important role for trabecular structure in determining bone strength coupled with ambiguities in the application of stereologic methods to this problem has led a number of investigators to invest considerable effort in developing 3-D imaging techniques to assess trabecular bone microarchitecture. In this section we summarize recent progress in this area. Three-dimensional evaluation of trabecular bone structure has received great impetus from the advent of CT scanners. These machines enable imaging of trabecular bone specimens at resolutions of 14 – 50 m, and in vivo scanners produce images at resolutions of 100 m. Figure 7 represents a three-dimensional image from a human iliac crest biopsy. As seen from the image, apart from obtaining routine measures of histomorphometry, such images provide a volumetric rendition of the three-dimensional structure. Peripheral computed tomography equipment under development in a research setting [57] has also led to the generation of in vivo images of trabecular bone architecture in the distal radius (Fig. 8). Goldstein and colleagues [58,59] prepared 8-mm cubic specimens of cadaveric trabecular bone for microCT scanning followed by mechanical testing. Results showed highly significant relationships of both the trabecular plate number and connectivity with the trabecular bone volume and that the bone volume explained 90% of the variance in bone strength. Consequently, although they found connectivity to be an important feature of skeletal integrity, its contribution to bone strength was contained within the information

FIGURE 7

A microcomputed tomography image of a specimen from a human iliac crest. The image is a three-dimensional rendering showing the rod-and-plate-like structure of the iliac crest. Image courtesy of Andres Laib, data acquired in laboratory of Dr. Peter Ruegsegger (ETH Zurich). Reproduced with permission.

provided by bone volume itself. One must keep in mind that these interesting results were based solely on material from four skeletons, none of which were osteoporotic. In recent years, the noninvasive and nonionizing nature of magnetic resonance (MR) has also been exploited to produce images of trabecular bone both in vitro and in vivo. MR techniques have been used in vitro to obtain images at resolutions as high as 50 µm isotropic [60 – 62]. Hipp et al.

11

CHAPTER 35 Nature of Osteoporosis

FIGURE 8

A microcomputed tomography image of the distal radius in vivo. The image is a three-dimensional rendering showing the cortical shell and trabecular bone. Image courtesy of Andres Laib, data acquired in laboratory of Dr. Peter Ruegsegger (ETH Zurich). Reproduced with permission.

[63] compared MR-derived stereology measures, determined from isotropic image voxels of 50 m, to those obtained using optical imaging and found good correlations, while Chung et al. [61] found good correlations of the estimated apparent trabecular bone volume (BV/TV) determined from MR images to those determined using displacement techniques. Antich et al. [60] confirmed that application of MR techniques to iliac crest biopsy specimens may be used to monitor changes in trabecular bone structure after fluoride therapy. In a recent study involving a set of 94 trabecular bone cubes from 13 cadavers and specimens from the calcaneus , distal femur, proximal femur, and vertebral bodies, the relationship between MRderived 3D measures of trabecular architecture, bone mineral density and biomechanical properties have been demonstrated [64].

Shown in Fig. 9, reverse gray scale images of two specimens from the vertebral body show clear visual differences in structure and density. The spatial resolution (117  117  300 m) of these images is typically greater than, or on the order of, trabecular bone dimensions. This gave rise to partial volume effects, but although the MR image derived 3D measures of trabecular architecture differ from those measured using 20-m optical images, there is good correlation between the two sets of measures [65]. The correlation coefficients are 0.69, 0.89, and 0.78 (P 0.01) for the apparent BV/TV, trabecular spacing (Tb.Sp), and trabecular number (Tb.N), respectively. The correlation between the trabecular thickness measure was poor, R  0.06 (P  0.84) [65]. The lower correlations specifically for measures such as the apparent Tb.Th, are expected based on the limited spatial resolution of the MR images compared to the trabecular sizes [66,67].

12

MARCUS AND MAJUMDER

FIGURE 9

Two vertebral bodies imaged using magnetic resonance imaging at a resolution of 117  117  300 m using a 1.5 T scanner.

Recent introduction of new micro-finite element (microFE) techniques should permit calculatation of cancellous bone mechanical properties directly from high-resolution images of its internal architecture. Computer reconstructions made from MR images were converted to micro-FE models from which bone elastic properties have been calculated [68]. The results of the MR-based FE models correlated well to those of the more accurate micro-CT based models, thus emphasizing the ability not only to compute architectural features from these images, but also to translate them into biomechanical parameters. In human MR studies, the spatial image resolution has ranged from 78 to 195 m in plane to 300 to 1000 m in the slice orientation depending on the anatomical site being examined [66,69 – 71]. Figure 10 illustrates representative calcaneal images, clearly depicting the

FIGURE 10

spatial heterogenity of trabecular bone microarchitecture. By selecting regions of interest extending from the joint line and 3 cm into the shaft in the radius and the posterior region in the calcaneus differences in trabecular architecture between patients with and without hip fractures have been demonstrated [72]. 3. RELATIONSHIP OF TRABECULAR CONNECTIVITY OSTEOPOROTIC FRACTURE

TO

Limited information is available concerning the contribution of trabecular connectivity to osteoporotic fractures. Kleerekoper et al. [73] compared indices of connectivity on iliac crest biopsy specimens from patients with osteoporotic fracture to those from nonfracture controls who had approximately the same trabecular and cortical bone mass.

Representative radius and calcaneus images. Representative radiograph (lateral projection) of a calcaneus reflects the heterogenity in bone content at this skeletal site (left). (Right) A sagittal MR section (500 m thick, 195  195 in plane resolution) showing trabecular bone microarchitecture in the calcaneus. The heterogeneity in trabecular structure seen as thicker, and a greater number of trabeculae in the subtalar portion is visually seen from the MR image.

13

CHAPTER 35 Nature of Osteoporosis

The fracture group had a 20% decrease in trabecular plate density and increases of equal magnitude in both the separation and thickness of trabecular plates. In other words, the negative effect of having fewer, more sparsely placed trabeculae was not counteracted by the relative thickness of the residual trabeculae. Recker [74] has reported a similar comparison. In his study, biopsies from fracture patients showed an 11% decrease in trabecular number, a 37% decrease in trabecular connecting nodes, and a 37% increase in trabecular free ends compared to the control biopsies. A report by Khosla et al. [75] suggests that trabecular disruption characterizes younger patients with osteoporosis as well. The authors evaluated histomorphometric features of iliac crest biopsies from 18 patients with idiopathic osteoporosis and 18 normal controls, average age 35 years. Although differences in the respective trabecular bone volumes for the two groups (16% vs 24%) were highly significant, differences in average trabecular wall thickness (40.0 mm vs 46.5 mm) were much smaller, implying that at least some of the deficit in trabecular bone mass must be due to dropout of entire trabecular units, i.e., a degradation in trabecular connectivity. Weinstein and Majumdar [53] applied fractal analysis to photographs of iliac crest biopsies from control subjects and from patients with osteoporosis. Although groups were not separable by traditional two-dimensional measures of microarchitecture, the fractal dimension, D, differed significantly. Taken together, results from these studies support a role for connectivity as a contributor to fracture. A dissenting view was registered by Compston et al. [76], who found no greater trabecular disruption in osteoporotic patients than in controls and suggested that structural changes in primary osteoporosis do not differ qualitatively from those of age-related bone loss. Microarchitectural comparisons have now been extended to the hip. Ciarelli et al. [77] prepared cubes of trabecular bone from the proximal femurs of women who had sustained hip fractures and from cadaveric nonfracture control bones. Control specimens had higher trabecular bone volume, trabecular number, and connectivity than did specimens from fracture patients. No differences were observed in mean trabecular thickness. Mechanical testing showed greater ultimate stress tolerance in the control bones. Furthermore, fracture specimens showed greater anisotropy of trabecular orientation, with fewer trabecular elements lying in a plane that was transverse to the primary load axis. It remains premature to offer a firm conclusion regarding an independent role for trabecular connectivity in osteoporotic fracture. Reasons for this uncertainty are multiple and include an apparently enormous bone-to-bone heterogeneity (variations up to 100%!) in trabecular volume and structure throughout the axial skeleton [78] as well as inadequacies of stereological methods. Ultimate resolution of this question will likely require more extensive application of the 3D techniques described above.

C. Accumulation of Cement Lines Secondary Haversian bone has poorer tensile strength and material properties than does primary lamellar bone [79,80]. In part, this reduced tensile strength may reflect the long-term accumulation of cement lines, which are the visible residua of previous bone remodeling events. Cement lines are observed as thin ribbons of loosely woven collagen fibers, distinguished easily under light microscopy from the surrounding lamellar bone (Fig. 11). They demarcate the area of deepest bone resorption and form the scaffold upon which new bone is deposited. Following completion of a remodeling cycle, the new, apparently pristine lamellar bone is interrupted by an area of structurally weaker woven bone. Carter and Hayes [81] have demonstrated debonding and disruption of bone at the cement line as a consequence of fatigue microdamage, indicating that cement lines represent loci of structural least resistance. As a consequence of many years of bone remodeling, both cortical and trabecular bone show a plethora of cement lines as well as evidence for previously remodeled areas

FIGURE 11

Cement lines. Reproduced from Carter and Hayes (81) with permission. (Bottom) Magnified view of boxed area in top panel illustrating debonding of the cement line following application of mechanical stress.

14

MARCUS AND MAJUMDER

that have been partly to completely removed (Fig. 12). For any given bone density, such highly remodeled bone is structurally weaker than is primary lamellar bone of younger adults. Choi and Goldstein [82] compared the fatigue and mechanical properties of single trabeculae to those of cortical bone specimens of equal size. The trabecular specimens had significantly lower elastic moduli and lower fatigue strength than cortical specimens, despite having higher mineral density. The authors proposed that cement line accumulation had decreased the mechanical properties of the trabecular bone.

D. Increased Cortical Porosity Strictly speaking, porosity is a measure of the prevalence and size of holes within bony cortex. Such holes represent Haversian canals, osteocyte lacunae, and new cutting cones that have been produced by alterations in systemic or local factors favoring resorption over bone forma-

tion (Fig. 13). Cortical porosity is difficult to assess, since most cortical holes are smaller than the resolution capability of the measuring instruments. Even with biopsy, this has been a difficult problem, since much of the process appears as “trabecularization” of the endocortex. In other words, porosity is sufficiently great that the histomorphometrist may read a highly porous endocortex as a simple extension of the trabecular bone. When bone is acquired during growth, primary Haversian canals constitute the major, if not exclusive, source of cortical porosity. Later, as a consequence of continuous remodeling, secondary Haversian systems gradually accumulate. Martin [83] used a rib model to analyze the relationships among the components of cortical porosity and age in humans. Total porosity actually decreased between ages 10 and 40, but rose progressively thereafter. The contribution of secondary Haversian canals to total porosity increased dramatically until age 40 years, following which it remained stable. Progressive cortical porosity after age 40 was attributable largely to an expanded remodeling space due to an increase in remodeling activation rate. Thus, increased cortical porosity is a feature of normal skeletal aging. Increased porosity is also characteristic of PTHdependent bone resorption. Therefore, to the extent that

FIGURE 12

Extensive cortical remodeling in trans-ileal biopsy specimen from an elderly woman (photograph courtesy of R. R. Recker with permission).

FIGURE 13

Porosity in human cortical bone (photograph courtesy of D. R. Carter with permission).

15

CHAPTER 35 Nature of Osteoporosis

hypersecretion of PTH in older individuals promotes increased remodeling activity, increased cortical porosity due to expanded remodeling space will occur. In theory, interventions designed to suppress PTH secretion and constrict the remodeling space should constrain the accumulation of cortical porosity with age.

E. Bone Fatigue Accumulation of fatigue with sustained use is a fundamental property of all materials. Although many investigators treat the terms “fatigue damage” and “microscopic damage” interchangeably, this is a simplistic construction, since cyclic loading produces functional, yet invisible, changes over time. Repetitive loading of compact bone leads to progressive deterioration in the modulus of elasticity, and ultimately to structural failure. Indeed, Carter and Hayes [84] showed that bone has relatively poor fatigue properties compared to a number of common building materials. Fatigue accumulates over time. Assuming that the intensity of loading remains constant, there is little difference in cumulative fatigue if materials are subjected to 10,000 cycles each day for 5 days or to 50,000 cycles at a single session. The importance of fatigue as a contributor to compact bone failure has been questioned by Schaffler et al. [85], who subjected standardized bone plugs to many loading cycles at relatively low, physiologic strains. A 5% deterioration in modulus of elasticity was observed after about 106 loading cycles, but no further change was observed even after 20  106 cycles, leading the authors to conclude that, under physiological conditions of low-magnitude repetitive strain, compact bone shows greater resistance to fatigue than would be predicted from the work of Carter and Hayes [84] and others, who carried out tests at higher strain magnitudes. Schaffler et al. [85] calculated that at the strains typically measured during running, 1200 – 1500 microstrain, cyclic loads equivalent to 11,000 miles of running could be sustained before repair processes would be necessary. Nonetheless, the relatively high frequency of clinical “stress” or fatigue fractures with overuse, particularly when habitual loading has been markedly and precipitously increased, certainly argues for the clinical relevance of fatigue accumulation in vivo. Schaffler et al. [86] suggest that vigorous activity might subject bone to brief episodes of very high strain, which would drastically reduce the number of cycles necessary to produce failure. Furthermore, if repetitive loading in vivo were to stimulate remodeling activity, the initial resorption phase would temporarily increase the remodeling space, thereby decreasing bone strength. Continued loading of such a bone, even at relatively low intensity, might then increase the risk of fracture. Recent evidence indicates that bone fatigue is as-

sociated with osteocytic apoptosis, which appears to be an important step in targeting bone for a subsequent remodeling event [87]. The contribution of fatigue to osteoporotic fracture is therefore complex. Certainly, it is important to the extent that by stimulating bone remodeling fatigue promotes the visible deterioration in cortical and trabecular microarchitecture described elsewhere in this chapter. Furthermore, cortical bones with extensive secondary Haversian remodeling, such as occurs normally with age, show fatigue earlier than does primary lamellar bone [86,88]. Beyond that, however, even prior to the emergence of visible damage, changes in fundamental bone material properties accumulate due to repetitive loading over time, and these subtle changes are likely contributors to overall bone fragility.

III. CONCLUSIONS At the beginning of this chapter we discussed the limitations of a bone mass-based diagnosis of osteoporosis [12]. A primary difficulty with such a definition is that its sensitivity to factors known collectively as “bone quality” has not been clarified, and it is tempting to attribute the diagnostic ambiguities of BMD measurements to their failure to account for these features. Although these concerns persist, the fact that information contained in the BMD estimate accounts in part for some of the important geometric, material, and microarchitectural properties, solidifies its rationale as a diagnostic criterion. Certainly, any substantial degree of matrix undermineralization would be reflected in a lower BMD, and the data of Goldstein et al. [59] suggest that trabecular disruption of sufficient magnitude to be mechanically important would also register as a bone mineral deficit, and therefore as a lower BMD. Qualitative features that would not be included in a BMD assessment include cement lines, unremodeled fatigue damage, and fluoride accumulation. The question remains whether osteoporosis should be viewed as one or more unique diagnostic entities, as is the case for Paget’s disease, or whether it is more useful to consider it a condition of skeletal fragility resulting from a stochastic process, in which contributory factors include age, body size, adequacy of peak bone mass, degree of adult bone loss, and accumulation of qualitative impairments. Since the overall trajectory over time of adolescent bone acquisition and adult bone loss appears to be universal, the only basis for considering osteoporosis one or more distinct entities would be a demonstration that its qualitative abnormalities, such as those discussed above, are restricted to those patients who have suffered a fragility fracture. Although evidence remains incomplete, it seems unlikely that such specificity will be validated for most of these abnormalities.

16

MARCUS AND MAJUMDER

References 1. F. Albright and E. C. Reifenstein, Jr., “The Parathyroid Glands and Metabolic Bone Disease: Selected Studies,” p. 162. Williams and Wilkins, Baltimore, 1948. 2. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton III, Changes in bone mineral density of the proximal femur and spine with aging. Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982). 3. T. Appelboom and J-J Body, The antiquity of osteoporosis: More questions than answers. Calcif. Tissue Int. 53, 367–369 (1993). 4. B, Krølner and S.P. Nielsen, Bone mineral content of the lumbar spine in normal and osteoporotic women: Cross-sectional and longitudinal studies. Clin. Sci. 62, 329 – 336 (1982). 5. R. Pacifici, N. Susman, P. L. Carr, S. J. Birge, and L. V. Avioli, Single and dual energy tomographic analysis of spinal trabecular bone. A comparative study in normal and osteoporotic women. J. Clin. Endocrinol. Metab. 64, 209 – 214 (1987). 6. S. M. Ott, R. F. Kilcoyne, and C. H. Chesnut III, Ability of four different techniques of measuring bone mass to diagnose vertebral fractures in postmenopausal women. J. Bone. Miner. Res., 2, 201 – 210 (1987). 7. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C-C Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1218 (1993). 8. R. D. Wasnich, P. D. Ross, L. K. Heilbrun, and J. M. Vogel, Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am. J. Obstet. Gynecol. 153, 745 – 751 (1985). 9. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Baseline measurement of bone mass predicts fracture in white women. Ann. Intern. Med. 111, 355–361 (1989). 10. S. R. Cummings, D. M. Black, M. C. Nevitt, W. S. Browner, J. A. Cauley, H. K. Genant, S. R. Mascioli, J. C. Scott, D. G. Seeley, P. Steiger, and T. Vogt, for the SOF Research Group, Appendicular bone density and age predict hip fracture in women. JAMA 263, 665 – 668 (1990). 11. L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, H. W. Wahner, and B. L. Riggs, Long-term fracture prediction by bone mineral assessed at different skeletal sites. J. Bone Miner. Res. 8, 1227 – 1234 (1993). 12. J. A. Kanis, L. J. Melton III, C. Christiansen, C. C. Johnston, Jr., and N. Haltaev, Perspective: The diagnosis of osteoporosis. J. Bone Miner. Res. 9, 1137 – 1142 (1994). 13. Consensus Development Conference,1991, Prophylaxis and treatment of osteoporosis. Am. J. Med. 90, 107–110 (1991). 14. F. McCabe, L.-J. Zhou, C. R. Steele, and R. Marcus, Non-invasive assessment of ulnar bending stiffness in women. J. Bone Miner. Res. 6, 53 – 59 (1991). 15. P. Kann, S. Graeben, and J. Beyer, Age-dependence of bone material quality shown by measurement of frequency of resonance in the ulna. Calcif. Tissue Int. 54, 96 – 100 (1994). 16. G. H. Brandenburger, Clinical determination of bone quality: Is ultrasound an answer? Calcif. Tissue. Int. 53 (Suppl. 1), S151 – S156 (1993). 17. S. Sherman, R. P. Heaney, A. M. Parfitt, E. C. Hadley, and C. Dutta, NIA Workshop on aging and bone quality. Calcif. Tissue Intl. 53, (Suppl. 1), S1 – S180 (1993). 18. M. Grynpas, Age and disease-related changes in the mineral of bone.Calcif. Tissue. Int. 53 (Suppl. 1), S57–S64 (1993). 19. J. Jowsey, Age changes in human bone. Clin. Orthop. 17, 210 – 218 (1960). 20. D. J. Ortner, Aging effects on osteon remodeling. Calcif. Tissue Res. 18, 27 – 36 (1975). 21. R. D. Crofts, T. M. Boyce, and R. D. Bloebaum, Aging changes in osteon mineralization in the human femoral neck. Bone 15, 147– 52 (1994).

22. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, M. W. Mathews, and E. J. Teubner, Bone matrix and mineral abnormalities in postmenopausal osteoporosis. Metabolism 31, 1113 – 1120 (1982). 23. G. P. Vose and A. L. Kubala, Jr.,. Bone strength-its relationship to x-ray determined ash content. Hum. Biol. 31, 261 – 270 (1959). 24. J. D. Currey, The mechanical consequences of variation in the mineral content of bone. J. Biomech. 2, 1 – 11 (1969). 25. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, and E. J. Teubner, The role of skeletal calcium deficiency in postmenopausal osteoporosis. Calcif. Tissue Int. 38, 187 – 92 (1986). 26. U. Barzel, The effect of excessive acid feeding on bone. Calcif. Tissue Res. 4, 94 – 100 (1969). 27. R. Amprino and A. Engstrom, Studies on x-ray absorption and diffraction of bone tissue. Acta Anat. 15, 1 – 22 (1952). 28. D. J. Ortner and D. W. Von Endt, Microscopic and electron microprobe characterization of the sclerotic lamellae in human osteons. Isr. J. Med. Sci. 7, 480 – 482 (1971). 29. T. M. Boyce and R. D. Bloebaum, Cortical aging differences and fracture implications for the human femoral neck. Bone, 14, 769 –778 (1993). 30. N. C. Leone, C. A. Stevenson, T. F. Hilbish, and M. C. Sosman, A roentgenologic study of a human population exposed to high-fluoride domestic water (a 10-year study). Am. J. Roentg 74, 874 – 885 (1955). 31. J. Avorn and L. C. Niessen, Relationship betwen long bone fractures and water fluoridation. Gerodontology 2, 175 – 179 (1986). 32. M. R. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, A prospective study of bone mineral content and fracture in communities with differential fluoride exposure. Am. J. Epidemiol. 133, 649 – 660 (1991). 33. C. Cooper, C. A. C. Wickham, D. J. R. Barker, and S. J. Jacobsen, Water fluoridation and hip fracture. JAMA 266, 513 – 514 (1991). 34. A. Richards, Li. Mosekilde, and C. H. Søgaard, Normal age-related changes in fluoride content of vertebral trabecular bone-relation to bone quality. Bone 15, 21–26 (1994). 35. L. Mosekilde, Vertebral structure and strength in vivo and in vitro. Calcif. Tissue Res. 53 (Suppl. 1), S121 – S126 (1993). 36. C. Snow-Harter, and R. Marcus, Exercise, bone density, and osteoporosis. Exercise Sport Sci. Rev. 19, 351 – 388 (1991). 37. Li. Mosekilde and Le. Mosekilde, Sex differences in age-related changes in vertebral size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 38. P. J. Atkinson, Variation in trabecular structure of vertebrae with age. Calcif. Tissue Res. 1, 24 – 32 (1967). 39. Li. Mosekilde, Age-related changes in vertebral trabecular bone architecture-assessed by a new method. Bone 9, 247 – 250 (1988). 40. B. D. Snyder, S. Piazza, W. T. Edwards, and W.C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue Int. 53 (Suppl. 1), S14 – S22 (1993). 41. M. Vogel, M. Hahn, and G. Delling, Relation beween 2- and 3-dimensional architecture of trabecular bone in the human spine. Bone 14, 199 – 203 (1993). 42. A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationship between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis: Implications for the microanatomic and cellular mechanisms of bone loss. J. Clin. Invest. 72, 1396 – 1409 (1983). 43. R. S. Weinstein and M. S. Hutson, Decreased trabecular width and increased trabecular spacing contribute to bone loss with ageing. Bone 8, 137 – 142 (1987). 44. J. E. Aaron, N. B. Makins, and K. Sagreiya, The microanatomy of trabecular bone loss in normal aging in men and women. Clin. Orthop. Rel. Res. 215, 260 – 271 (1987). 45. R. W. E. Mellish, N. J. Garrahan, and J. E. Compston, Age-related changes in trabecular width and spacing in human iliac crest biopsies. Bone and Miner. 6, 331 – 338 (1989). 46. D. H. Birkenhäger-Frenkel, P. Courpron, E. A. Hüpscher, E. Clermonts, M.F. Coutinho, P. I. M. Schmitz, and P.J. Meunier,

17

CHAPTER 35 Nature of Osteoporosis

47.

48.

49.

50.

51.

52. 53. 54.

55. 56.

57. 58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

Age-related changes in trabecular bone structure. Bone and Miner. 4, 197 – 216 (1988). A. Vesterby, H. J. G. Gundersen, and F. Melsen, Star volume of marrow space and trabeculae of the first lumbar vertebra: Sampling efficiency and biological variation. Bone 10, 7 – 13 (1989). A. Vesterby, H. J. G. Gundersen, F. Melsen, and L. Mosekilde, Normal postmenopausal women show iliac crest trabecular thickening on vertical sections. Bone, 10, 333 – 339 (1989). A. Vesterby, Star volume of marrow space and trabeculae in iliac crest: Sampling procedure and correlation to star volume of first lumbar vertebrae. Bone 11, 149 – 55 (1990). A. Vesterby, H. J. G. Gundersen, F. Melsen, and L. Mosekilde, Marrow space star volume in the iliac crest. Decreases in osteoporotic patients after continuous treatment with fluoride, calcium, and vitamin D2 for 5 years. Bone 12, 33 – 7 (1991). S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of bone biopsies from the iliac crest of normal British subjects. Metab. Bone Dis and Relat. Res. 4, 231 – 236 (1982). A. M. Parfitt, Trabecular bone architecture in the pathogenesis and prevention of fracture. Am. J. Med. 82(Suppl. 1B), 68 – 72 (1987). R. S. Weinstein and S. Majumdar, Fractal geomery and vertebral compression fractures. J. Bone Miner. Res. 9, 1797 – 1802 (1994). J. C. Buckland-Wright, J. A. Lynch, J. Rymer, and I. Fogelman, Fractal signature analysis of macroradiographs measures trabecular organization in lumbar vertebrae of postmenopausal women. Calcif. Tissue Int. 54, 106 – 112 (1994). P. Caligiuri, M. L. Giger, and M. Favus, Multifractal radiographic analysis of osteoporosis. Med. Phys. 21, 503 – 508 (1994). H. J. G. Gundersen, R. W. Boyce, J.R. Nyengaard, and A. Odgaard, The conneulor: Unbiased estimation of connectivity using physical disectors under projection. Bone 14, 217 – 222 (1993). R. Muller, T. Hildebrand, H. J. Hauselmann, and P. Ruesegger, J. Bone Miner. Res. 11, 1745 – 1750 (1996). L.A. Feldkamp, S. A. Goldstein, A. M. Parfitt, G. Jesion, and M. Kleerekoper, The direct examination of three– dimensional bone architecture in vitro by computed tomography. J. Bone Miner. Res. 4, 3 – 11 (1989). S. A. Goldstein, R. Goulet, and D. McCubbrey, Measurement and significance of three-dimensional architecture to the mechanical integrity of trabecular bone. Calcif. Tissue Int. 53 (Suppl.), S127 – S133 (1993). P. P. Antich, R. P. Mason, R. McColl, J. Zerwech, and C. Y. C. Pak, Trabecular architecture studies by 3D MRI microscopy in bone biopsies. J. Bone Miner. Res. 9S1, 327 (1994). H. W. Chung, F. W. Wehrli, J. L. Williams, and S. L. Wehrli, Three dimensional nuclear magnetic resonance micro-imaging of trabecular bone J. Bone Miner. Res. 10, 1452 – 1461 (1995). R. D. Kapadia, W. B. High, H. A. Soulleveld, D. Bertolini, and S. K. Sarkar, Magnetic resonance microscopy in rat skeletal research Magn. Reson. Med. 30, 247 – 250 (1993). J. A. Hipp, A. Jansujwicz, C. A. Simmons, and B. Snyder, Trabecular bone morphology using micro-magnetic resonance imaging. J. Bone Miner. Res. 11, 286 – 292 (1996). S. Majumdar, M. Kothari, P. Augat, D. C. Newitt, T. M. Link, J. C. Lin, T. Lang, Y.Lu, and H. K. Genant, High resolution magnetic resonance (MR): Three dimensional trabecular bone architecture and biomechanical properties. Bone May 22, 445 – 454 (1998). M. Kothari,T. Keaveny, J. Lin, D. C. Newitt, H. K Genant, and S. Majumdar, Impact of spatial resolution on the prediction of trabecular architecture parameters. Bone 22, 437 – 443 (1998). S. Majumdar, D. C. Newitt, A. Mathur, et al., Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with x-ray tomographic microscopy and biomechanics. Osteoporosis Int. 6, 376 – 385 (1996). R. Muller, M. Hahn, M. Vogel, G. Delling, and P. Ruegsegger, Morphometric analysis of non-invasively assessed bone biopsies:

68.

69.

70.

71.

72.

73.

74. 75.

76.

77.

78.

79. 80.

81. 82.

83. 84. 85.

86. 87.

88.

Comparison of high resolution computed tomography and histologic sections. Bone , 8, 215 – 220 (1996). B. van Rietbergen, S. Majumdar, W. Pistoia, D. C. Newitt, M. Kothari, A. Laib, and P. Rüegsegger, Assessment of cancellous bone mechanical properties from micro-Fe models based on micro-Ct, Pqct and Mr images. Stud. Technol. Healthcare 6(5 – 6), 413 – 20 (1998). H. Jara, F. W. Wehrli, H. Chung, and J. C. Ford, High-resolution variable flip angle 3D MR imaging of trabecular microstructure in vivo. Magn. Reson. Med. 29, 528 – 539 (1993). B. Kuehn, B. Stampa, M. Heller, and C. Glueer, In vivo assessment of trabecular bone structure of the human phalanges using high resolution magnetic resonance imaging. Osteoporosis Int. 7, 291 (1997). X. Ouyang, K. Selby, P. Lang, et al., High resolution MR imaging of the calcaneus: Age-related changes in trabecular structure and comparison with DXA measurements. Calcif. Tissue Int. 60, 139 – 147 (1997). T. Link, S. Majumdar, J. Lin, et al., Can texture analysis of high resolution MRI of calcaneus be used to predict osteoporotic hip fractures? J. Bone Miner. Res. (2001) (in press). M. Kleerekoper, A. R. Villanueva, J. Stanciu, D. S. Rao, and A. M. Parfitt, The role of three-dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif. Tissue Int 37, 594 – 597 (1985). R. R. Recker, Architecture and vertebral fracture. Calcif. Tissue Int., 53 (Suppl.), S139 – S142 (1993). S. Khosla, E. G. Lufkin, S. F. Hodgson, L. A. Fitzpatrick, and L. J. Melton III, Epidemiology and clinical features of osteoporosis in young individuals. Bone 15, 551 – 555 (1994). J. E. Compston, N. J. Garrahan, P. I. Croucher, C. D. P. Wright, and K. Yamaguchi, Quantitative analysis of trabecular bone structure: Studies in patients with osteoporosis and in women treated with tamoxifen. Bone 13, A8 (1992). T. E. Ciarrelli, D. P. Fyhrie, M. B. Schaffler, and S. A. Goldstein, Variations in three-dimensional cancellous bone architecture of the proximal femur in female hip fractures and in controls. J. Bone Miner. Res. 15, 32 – 40 (2000). H Ritzel, M. Amling, M. Vogel, M. Posl, M. Hahn, M. Werner, and G. Delling, Spongiosa structure and polyostotic heterogeneity in osteoporosis. Mechanism of bone transformation, morphology, clinical significance. Pathologe (German) 17, 68 – 77 (1996). S. Saha and W. C. Hayes, Relations between tensile impact properties and microstructure of compact bone. Calcif. Tissue Res. 24, 65 – 72 (1977). R. Vincentelli and M. Grigorov, The effect of Haversian remodeling on the tensile properties of human cortical bone. J. Biomech. 18, 201 – 207 (1985). D. R. Carter and W. C. Hayes, Compact bone fatigue damage. A microscopic examination. Clin. Orthop. Relat. Res. 127, 265 – 274 (1977). K. Choi and S. A. Goldstein, A comparison of the fatigue behavior of human trabecular and cortical bone tissue. J. Biomech. 25, 1371 – 1381 (1992). R. B. Martin, On the signifiance of remodeling space and activation rate changes in bone remodeling. Bone 12, 391 – 400 (1991) D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two phase porous structure. J. Bone Joint Surg. 59A, 954 – 962 (1977). M. B. Schaffler, E. L. Radin, and D. B. Burr, Long-term fatigue behavior of compact bone at low strain magnitude and rate. Bone 11, 321 – 326 (1990). M. B. Schaffler and D. B. Burr, Stiffness of compact bone: Effects of porosity and density. J Biomech, 21, 13 – 16 (1988). O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Miner. Res. 15, 60 – 67 (2000). D. B. Burr, M. B. Schaffler, and R. G. Frederickson, Composition of the cement line and its mechanical role as an interface in human compact bone. J. Biomech. 21, 939–945 (1988).

CHAPTER 36

Local and Systemic Factors in the Pathogenesis of Osteoporosis LAWRENCE G. RAISZ

I. II. III. IV. V. VI.

Department of Endocrinology and Metabolism, University of Connecticut Health Center, Farmington, Connecticut 06032

VII. Colony-Stimulating Factors VIII. Further Considerations of Interactions of Systemic Hormones and Local Mediators IX. Conclusions References

Introduction Limited Role of Systemic Hormones Local Factors Cytokines Prostaglandins Growth Factors

I. INTRODUCTION

The action of estrogen on cytokine production has been reviewed elsewhere in this volume (Chapters 10, 13, and 41), as have the effects of calcium regulating and other systemic hormones. In this chapter, I will summarize current evidence that supports a role for local factors acting both directly and as mediators of systemic hormones. Syndromes involving cancellous bone loss, which include vertebral crush fractures and Colles’ fractures, as well as fractures of the pelvis and trochanteric fractures of the hip, are most likely to involve local factors [5,6]. These comprise a substantial proportion of osteoporotic fractures and often occur with minimal or no trauma. Fractures involving cortical bone, although usually associated with some trauma, are also more likely to occur in individuals with bone loss. Thus, high bone turnover is associated with cortical bone loss and an increased risk of hip fracture [7,8]. The relative importance of high bone turnover with increased remodeling and increased resorption, as opposed to low bone turnover with decreased formation in the

The concept that local bone factors play an important role in the pathogenesis of osteoporosis has developed from several lines of evidence [1 – 4]. Among these are the following: (i) it has been difficult to demonstrate relevant differences in the production of systemic hormones between osteoporotic patients and matched controls, (ii) many local factors which regulate bone metabolism have been identified, and (iii) estrogen deficiency, as well as changes in other systemic hormones which are thought to play pathogenetic roles in osteoporosis, have marked effects on local factors. Since this chapter was written for the first edition of Osteoporosis there have been many new observations which increase our understanding of the interaction between local and systemic factors and expand our concepts of the cellular mechanisms by which these interactions can result in decreased bone mass and increased bone fragility.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

19

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

20 pathogenesis of osteoporosis, is still not fully resolved, although a number of studies indicate that rapid bone loss is likely to be associated with high turnover [8,9].

II. LIMITED ROLE OF SYSTEMIC HORMONES There is evidence to support a primary role for either calcium regulating or systemic hormones other than sex hormones in the pathogenesis of cancellous bone loss. Agerelated calcium deficiency, which involves decreased intake of both calcium and vitamin D, as well as decreased formation of, and responsiveness to, calcitriol, results in secondary hyperparathyroidism, and this probably plays a role in cortical bone loss with age [10]. However, in vertebral crush fracture patients, serum parathyroid hormone (PTH) concentrations are not increased. On the contrary, there is a blunted PTH response to hypocalcemia [11]. It is plausible that this blunted response, reflecting the enhancement of bone resorption by local factors, moves calcium from bone to blood without requiring PTH, and hence parathyroid responsiveness is decreased. Calcitonin deficiency has not been demonstrated in vertebral crush fracture patients [12]. Circulating IGF-I and IGF-I-BP3 levels decrease with age [13]. These changes could aggravate bone loss, but have not been found to be greater in most osteoporotic patients, although they have been implicated in idiopathic osteoporosis in men [14]. Glucocorticoid excess can clearly produce osteoporosis, largely by inhibiting bone formation [15]. There is little evidence that glucocorticoid excess is important in primary osteoporosis, although high trough glucocorticoid levels in the evening have been associated with decreased bone mass in older men [16]. Sex hormones are critically important in the development and maintenance of the skeleton, but their mechanisms of action are still not clear. Estrogen is important in both sexes [17]. Males with a defective estrogen receptors or lack of aromatase, which converts androgens to estrogen, show failure of epiphyseal closure and high bone turnover with low bone mass, despite full responsiveness to androgen [18,19]. In male rats with androgen resistance, there is bone loss after orchiectomy, presumably due to loss of estrogen formed from androgen by aromatization [20]. In men with aromatase deficiency treatment with estrogen can decrease bone turnover and increase bone mass [19]. There is also evidence that androgens have direct effects on bone [21]. Thus, the sex hormones appear to work in concert, probably through separate but overlapping pathways. A generalization, consistent with current data, is that androgens increase bone mass indirectly by increasing muscle mass and directly by stimulating bone formation, particularly in the periosteum, while estrogens prevent bone loss by decreasing trabecular and endocortical bone resorption and decreasing turnover.

LAWRENCE G. RAISZ

A role for progesterone in the skeleton has been proposed but is not established. Progesterone is a potent mitogen in bone cell cultures [22,23], but there is little direct evidence for an anabolic action in adult humans [24]. The possibility that other reproductive hormones, including gonadotropins, inhibin, and prolactin, have skeletal effects has not been adequately explored.

III. LOCAL FACTORS The interaction of local and systemic factors in regulating bone metabolism is essential for the skeleton to play its dual roles as a structure for locomotion and protection of internal organs and as a storehouse for mineral. The structure of the skeleton is determined by mechanical forces. Recent studies on the effects of loading and unloading on the skeleton have identified a number of local factors which may mediate cellular responses, including nitric oxide, prostaglandins insulin-like growth factor-I (IGF-I) and glutamate [25 – 27]. Production of these factors may be initiated by fluid shear stress exerted on the osteocyte-osteoblast canalicular network [28]. Local regulation also involves a complex interplay between cells in the marrow and in bone. Both hematopoietic precursors for osteoclasts and mesenchymal precursors for osteoblasts are present in the marrow. Marrow cells may produce some of the local factors which act on bone cells [29,30]. Other tissues, including the vasculature, cartilage, muscle, tendon, and synovium, as well as blood-borne elements, including platelets and leukocytes, can produce factors that regulate bone metabolism [31]. There is as yet no evidence for a role for extraskeletal sources of local factors in the pathogenesis of osteoporosis, but this has not been adequately explored. Since the previous version of this chapter was written there have been exciting new findings on the mechanism of the interaction between cells of the osteoblastic and osteoclastic lineage in regulating bone resorption. As discussed elsewhere in this volume (Chapters 3, 12, and 13), osteoclasts and their precursors express receptor activator of NFB (RANK) for which the ligand, variously termed RANK ligand (RANKL), osteoclast differentiation factor (ODF), TRANCE, or OPGL, is expressed on stromal cells of the osteoblast lineage. This interaction is critical for the formation and activity of osteoclasts. It can be inhibited by a decoy receptor osteoprotegerin (OPG), which is produced by many cells. Both RANKL and OPG are regulated by both the systemic and local factors which influence bone resorption [32]. In reviewing the role of local factors, it is important to recognize that the data are derived largely from animal models and that there is little information on humans. Nevertheless, the fact that different portions of the skeleton are

CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis

differentially affected in osteoporosis and that there are marked local changes in cancellous bone structure in osteoporosis supports a role for local factors. Further support comes from the clinical observation that a disease in which bone loss is clearly due to the secretion of local factors, multiple myeloma, can produce a rapidly progressive vertebral crush fracture syndrome which resembles severe primary osteoporosis. Multiple local factors are probably responsible for the intense osteoclastic activity produced by myeloma [33]. Other marrow disorders can also produce bone loss, presumably by local mechanisms.

21

In humans, increased IL-1 and IL-1ra production by circulating macrophages has been described in postmenopausal women, with inhibition by estrogen and greater persistence in osteoporotic patients, but not all studies confirm this observation [51,52]. Studies of cytokine production by marrow cultures have also produced variable results [53]. Direct measurement in bones from healthy and affected humans will be necessary before the roles of individual cytokines can be fully defined. However, the current availability of specific antagonists such as IL-1-receptor antagonist and TNF binding protein for clinical trials may also provide clues as to the role of these factors.

IV. CYTOKINES V. PROSTAGLANDINS The role of cytokines in bone resorption was first suggested by the finding that mononuclear leukocytes could produce an osteoclast activating factor [34]. The bone resorbing activity in the supernatants of mitogen or antigenactivated leukocytes ultimately turned out to be interleukin(IL)-1 [35,36]. However, other cytokines are also active. Tumor necrosis factor (TNF) is a potent bone resorber [37]. IL-6 and its soluble receptor may be cofactors for osteoclast generation [38]. IL-7 and IL-11 can increase bone resorption [39,40]. IL-4, IL-13, and interferon gamma can inhibit bone resorption in part by decreasing prostaglandin production [41,42]. However, IL-4 overexpression in mice produces osteoporosis associated with decreased osteoblast function [43]. Direct evidence for involvement of cytokines in the pathogenesis of osteoporosis is derived largely from studies in ovariectomized rodents. Estrogen can inhibit IL-6 production [44], and the increase in bone turnover after orchidectomy is inhibited by IL-6 antibodies [45]. Bone loss following ovariectomy in rats can be abrogated by administration of a combination of the IL-1 receptor antagonist and TNF, soluble binding protein [46]. Moreover, animals in whom the IL-1 activating receptor (IL-1-R1) has been knocked out do not show bone loss after ovariectomy [47]. Resorptive factors may be derived from the marrow. Bone marrow supernatant fractions from oophorectomized animals stimulate bone resorption and increase prostaglandin production in bone by a mechanism which can be blocked by IL-1ra as well as indomethacin [30,38]. However, it is not clear that IL-1 itself is the agonist, since the concentrations measured by immunoassay are not increased in the marrow supernatant fractions from ovariectomized animals compared to those from sham-operated controls. One possibility is that the effects of estrogen deprivation results in reciprocal increases in the activating receptor, IL-1-R1, and decreases in the decoy receptor IL-1-R2 [48]. Another possibility is that estrogen deficiency results in a decrease in OPG production since estradiol can decrease OPG levels in osteoblastic cell cultures [49].

Prostaglandins are potent, multifunctional regulators of bone metabolism, and their production by both hematopoietic cells and bone cells is abundant and highly regulated [54 – 56]. Most of the systemic hormones, cytokines, and growth factors which affect bone metabolism also affect prostaglandin production; however, the importance of this prostaglandin production in the response to hormones and local factors varies greatly. Agents that stimulate bone resorption also stimulate prostaglandin production, while inhibitors of bone resorption, such as glucocorticoids, IL-4, and interferon gamma, can inhibit prostaglandin production. In cell cultures which can produce osteoclasts in response to various stimuli, inhibition of prostaglandin synthesis decreases osteoclast formation regardless of the stimulator, although to a variable degree. The role of prostaglandin is probably dependent upon its synthesis by osteoblasts and this in turn appears to depend on the de novo synthesis of inducible prostaglandin GH synthase or cyclooxygenase-2 (COX-2) which has been shown to be highly regulated in osteoblastic cells [56]. Prostaglandins stimulate bone resorption in organ culture and in vivo [57,58]. This effect is probably mediated through cyclic AMP and prostaglandins of the E series are most potent. However, prostaglandins can also produce transient direct inhibition of the function of isolated osteoclasts [59]. In vivo, prostaglandins of the E series are potent stimulators of both endosteal and periosteal bone formation [60]. High concentrations of prostaglandins can inhibit collagen synthesis in vitro. This effect was greatest with prostaglandins of the F series and was mimicked by activators of protein kinase C [61]. Prostaglandin F2 may also stimulate bone resorption in part due to its ability to increase endogenous prostaglandin production [62]. This latter effect of prostaglandins, which we have termed “autoamplification’’ can occur with many prostanoids and may be important in maintaining and enhancing small signals such as fluid shear stress [63,64]. There is evidence that impact loading can stimulate prostaglandin production and that prostaglandin production

22

LAWRENCE G. RAISZ

is required for the increase in bone formation that occurs in response to mechanical forces [25]. As noted above this may be due to fluid shear stress on the syncytium of osteoblasts and osteocytes in bone [28]. Small strains can affect fluid flow in the canaliculi connecting osteocytes to each other and to surface osteoblasts and lining cells. The effects of fluid shear stress may be initiated by release of arachidonic acid and prostaglandin production by constitutive cyclooxygenase (COX-1) and sustained and amplified by induction of COX-2 [64]. The initial bone loss that occurs after immobilization may also be prostaglandin dependent, since it is inhibited by nonsteroidal anti-inflammatory drugs (NSAIDs) [65,66]. However, a sustained decrease in bone formation occurs after immobilization which is not reversed by indomethacin [67]. As noted above, marrow supernatant fractions from oophorectomized animals can stimulate bone resorption by a prostaglandin-dependent mechanism [48]. This effect is associated with a induction of COX-2. NSAIDs can also inhibit bone resorption in postmenopausal women [68]. Epidemiologic studies have demonstrated an increase in bone mineral density in some patients on nonselective NSAIDs [69], but these drugs are seldom taken continuously at high doses. The recent development of COX-2 selective inhibitors may make it possible to test the role of endogenous prostaglandins in human bone remodeling. In addition to prostaglandins produced by the cyclooxygenase pathway, metabolism of arachidonic acid by the lipoxygenase pathway may produce leukotrienes which can stimulate osteoclastic resorption [70]. Another potential regulator is nitric oxide (NO). NO has complex effects on bone resorption, although its most prominent effect is direct inhibition of osteoclastic activity [71].

VI. GROWTH FACTORS The families of insulin-like growth factors and of transforming growth factor  (TGF-) and related bone morphogenetic proteins (BMPs) have been extensively studied for their role in bone growth [72 – 74] (see Chapter 14). It is possible that deficiency in the production of, or response to, these growth factors is important in the pathogenesis of decreased bone formation in osteoporosis. Studies of the insulin-like growth factor (IGF) family are complicated by the fact that the binding proteins (IGFBPs) are also regulated and have both inhibitory and stimulatory effects on IGF responses [75]. IGFBP-5 is present in bone matrix and is regulated by PTH and prostaglandins [76]. Both IGFBP5 and the inhibitory binding protein, IGFBP-4 may play a role in bone loss [77,78]. Transforming growth factor beta (TGF-) is a multifunctional regulator of bone, which can both stimulate and inhibit resorption, but is largely stimulatory for bone

formation [74]. The inhibition of osteoclast activity by estrogen has been attributed to an increase in TGF- which enhances osteoclast apoptosis [79]. Paradoxically overexpression of TGF- in osteoblasts results in osteoporosis-like phenotype in young mice [80], while diminished TGF activity has been associated with spontaneous bone loss in old male mice [81]. Bone morphogenetic proteins are likely to be involved in osteoporosis, but have not been studied extensively [73]. Among the heparin binding growth factors, basic fibroblast growth factor (FGF) has been shown to increase bone formation in vivo although it inhibits collagen synthesis in vitro [82,83]. FGF can increase bone resorption by both prostaglandin-dependent and independent mechanisms [84,85]. Platelet-derived growth factor is also a potent mitogen in bone and a stimulator of bone resorption [86,87].

VII. COLONY-STIMULATING FACTORS The critical role of macrophage colony-stimulating factor (M-CSF or CSF-1) in bone resorption was first suggested by the fact that mice deficient in this factor were osteopetrotic. Subsequent studies showed that M-CSF could restore osteoclastic activity in such animals [88]. These studies have been amplified by gene knock-out experiments in which animals lacking c-fos, a critical transcription factor for macrophages, also show osteopetrosis, which could be reversed by transplanting marrow cells from wild-type mice [89]. M-CSF is probably critical for the replication of osteoclast precursors and when added with RANKL can stimulate osteoclast production in hematopoietic cell cultures devoid of osteoblasts or their precursors [90]. Granulocyte/monocyte-CSF (GM-CSF) is also important in the regulation of bone resorption. Its affects appear to be inhibitory, probably due to a stimulation of cells along alternative pathways such as macrophage formation, which prevents them from differentiating into osteoclasts. Addition of GM-CSF to cell cultures can block osteoclast formation. IL-18 probably acts by increasing GM-CSF production in such cultures [91].

VIII. FURTHER CONSIDERATIONS OF INTERACTIONS OF SYSTEMIC HORMONES AND LOCAL MEDIATORS A number of interactions of systemic hormones, particularly sex hormones, with local factors have already been discussed. Our understanding in this area is still quite fragmentary, but it is clear that multiple systemic hormones have effects on multiple local factors. Thus, knock-out,

CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis

overexpression, or inhibition experiments involving a single factor may not fully elucidate the role of that factor because other factors can substitute for its action. The major systemic mediators of bone resorption, PTH and 1,25-dihydroxyvitamin D stimulate prostaglandin production in bone and their ability to stimulate osteoclastogenesis in cell culture is partially prostaglandin dependent [92]. However, these hormones are not dependent on prostaglandins for their ability to stimulate bone resorption in organ culture. Indeed, the only evidence for a prostaglandindependent resorption pathway for a systemic hormone is that of thyroid hormone in mouse calvariae [93]. PTH and thyroid hormone, as well as IGF-I and prostaglandin E2, can increase interleukin-6 production [94,95]. Bone resorbing hormones, as well as local factors that stimulate bone resorption, can increase TGF- activity in organ cultures [96]. This is probably largely the result of activation and release from the matrix rather than new synthesis [97]. Activation of TGF- could play an important role in limiting resorption and initiating the coupled formation response. Glucocorticoids have complex interactions with local factors [15]. Glucocorticoids can inhibit production of interleukin-1, interleukin-6, prostaglandins, insulin-like growth factors, and IGF binding proteins from hematopoietic or bone cells. However, glucocorticoids can also increase the sensitivity to local factors, particularly prostaglandins and IGFs. The diurnal rhythm of glucocorticoid secretion is critical for normal skeletal metabolism. A decrease in glucocorticoid secretion in the afternoon and night may be permissive for the nocturnal increase in bone turnover. Small doses of glucocorticoids given in the evening can prevent the nocturnal rise in osteocalcin, and a block of the morning increase in glucocorticoid secretion can result in a sustained increase in this marker of osteoblastic activity [98,99]. One area of interaction between systemic and local factors which needs further exploration is the growth hormone/IGF-I system. Growth hormone can affect IGF production in skeletal tissue as well as in the liver. Since circulating IGF-I produces feedback inhibition of growth hormone secretion, alterations in hepatic production could have important indirect effects on skeletal production. Oral estrogen therapy produces a decrease in hepatic IGF-I production which could result in an increased growth hormone secretion which in turn would increase local IGF-I production [100]. In contrast, transdermal estrogen produces no change or an increase in circulating IGF-I and could have the opposite effect.

IX. CONCLUSIONS Even though the available data are limited, it seems likely that differences in the production of, or response to, local factors will be important in the pathogenesis of

23

osteoporosis. Since these local factors often act in concert and are often coordinately regulated, it seems likely that the relevant differences will involve multiple local factors in most patients. This hypothesis is particularly attractive because there is so little difference in the levels of systemic hormones in osteoporotic patients compared to that seen in age-matched controls. A plausible hypothesis would be that the production or activity of local factors changes to varying degrees with age and estrogen deficiency and that the individuals with severe osteoporosis are those who have (i) large increases in bone resorbing factors, (ii) loss of inhibitors of resorption, (iii) large increases in inhibitors of formation, or (iv) loss of stimulators of bone formation [101,102]. Another critical factor in the pathogenesis of osteoporosis is peak bone mass. It is quite possible that the genetic determinants of peak bone mass involve effects on local factors, particularly growth factors. Moreover, the effect of weight-bearing activity on peak bone mass is likely to be mediated by local factors. Regulation involves contributions from marrow cells of both the hematopoietic and the mesenchymal lineages as well as bone cells themselves. There is some evidence for a decrease in the number of osteogenic stem cells in the marrow with age, and this could account in part for age-related decreases in bone formation [103 – 105]. On the other hand, the differentiation of cells of the osteoclastic lineage may be maintained with age, thus maintaining high rates of bone resorption (J. A. Lorenzo, unpublished observations). An interaction between marrow and bone could explain the fact that bone loss in osteoporosis largely involves cancellous bone and endosteal surfaces of cortical bone [48,106]. Finally it is likely that the pathogenesis of severe osteoporosis will be heterogeneous; that is, there will be patients with different patterns of abnormalities involving different systemic hormones and local factors. The division into Type I, postmenopausal osteoporosis, and type II, senile osteoporosis, represent a simplified version of this concept. The broad range of values for bone resorption and formation in bone biopsies in osteoporosis supports the concept of heterogeneity [107]. Thus, rather than a division into high and low turnover forms, it is more likely that there will be a spectrum of relative contributions of abnormalities of resorption and formation in different patients. Moreover, these changes may vary over the course of the disease. Confirmation of the hypothesis that local factors play a role in osteoporosis and identification of specific factors should now be possible through the application of newer methods of molecular and cellular biology as well as new pharmacologic approaches. A combination of the use of microarray technology and quantitative reverse-transcriptasepolymerase chain reaction methodology should allow us to identify and quantify specific local factors and their receptors. It may be more difficult to prove that they play a role in the pathogenesis. Selective inhibitors of cytokines,

24

LAWRENCE G. RAISZ

prostaglandins, and other local factors are being developed and could be tested in appropriately selected patients. Such an approach will not only improve our understanding of the pathogenetic mechanisms in osteoporosis, but might also lead to a more specific preventive and therapeutic measures in this disorder.

References 1. B. L. Riggs and L. J. Melton III, Involutional osteoporosis. N. Engl. J. Med. 314, 1676–1686 (1986). 2. L. G. Raisz, Local and systemic factors in the pathogenesis of osteoporosis. N. Engl. J. Med. 318, 818 – 828 (1988). 3. L. G.Raisz, The osteoporosis revolution. Ann. Intern. Med. 126, 458 – 462 (1997). 4. J. Center and J. Eisman, The epidemiology and pathogenesis of osteoporosis. Baillieres Clin. Endocrinol. Metab. 11, 23 – 62 (1997). 5. T. Steiniche, P. Christiansen, A. Vesterby, C. Hasling, R. Ullerup, L. Mosekilde, and F. Melsen, Marked changes in iliac crest bone structure in postmenopausal osteoporotic patients without any signs of disturbed bone remodeling or balance. Bone 15, 73 – 80 (1994). 6. J. A. P. Jayasinghe, S. J. Jones, and A. Boyde, Three dimensional photographic study of cancellous bone in human fourth lumbar vertebral bodies. Anat. Embryol. 189, 259 – 274 (1994). 7. P. Garnero, P. Dargent-Molina, D. Hans, A. M. Schott, G. Breart, P. J. Meunier, and P. D. Delmas, Do markers of bone resorption add to bone mineral density and ultrasonographic heel measurement for the prediction of hip fracture in elderly women? The EPIDOS prospective study. Osteoporosis Int. 8, 563 – 569 (1998). 8. P. Garnero, E. Sornay-Rendu, F. Duboeuf, and P. D. Delmas, Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: The OFELY study. J. Bone Miner. Res. 14, 1614 – 1621 (1999). 9. M. A. Hansen, K.Overgaard, B. J. Riis, and C. Christiansen, Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 Year study. Br. Med. J. 303, 961 – 964 (1991). 10. S. Boonen, P. Broos, G. Verbeke, J. Aerssens, E. Van Herck, I. Jans, J. Dequeker, and R. Buillon, Calciotropic hormones and markers of bone remodeling in age-related (type II) femoral neck osteoporosis: Alterations consistent with secondary hyperparathyroidism-induced bone resorption. J. Gerontol. A. Biol. Sci. Med. Sci. 52, M286 – M293 (1997). 11. S. J. Silverberg, E. Shane, L. de la Cruz, et al., Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D3 formation in women with osteoporosis. N. Engl. J. Med. 320, 277– 281 (1989). 12. R. D. Tiegs, J. J. Body, H. W. Wahner, J. Barta, B. L. Riggs, and H. Heath III, Calcitonin secretion in postmenopausal osteoporosis. N. Engl. J. Med. 312, 1097 – 1100 (1985). 13. C. A. Benbassat, K. C. Maki, and T.G. Unterman, Circulating levels of insulin-like growth factor (IGF) binding protein-1 and -3 in aging men: Relationships to insulin, glucose, IGF, and dehydroepiandrosterone sulfate levels and anthropometric measures. J. Clin. Endocrinol. Metab. 82, 1484 – 1491 (1997). 14. E. S. Kurland, F. K. Chan, C. J. Rosen, and J. P. Bilezikian, Normal growth hormone secretory reserve in men with idiopathic osteoporosis and reduced circulating levels of insulin-like growth factor-I. J. Clin. Endocrinol. Metab. 83, 2576 – 2579 (1998). 15. B. P. Lukert, and L. G. Raisz, Glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 84, 3058 – 3063 (1994).

16. E. Dennison, P. Hindmarsh, C. Fall, S. Kellingray, D. Barker, D. Phillips, and C. Cooper, Profiles of endogenous circulating cortisol and bone mineral density in healthy elderly men. J. Clin. Endocrinol. Metab. 33, 3058 –3063 (1999). 17. S. Khosla and L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 18. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and Korach, Estrogen resistance caused by  mutation in the estrogen receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 19. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 20. D. Vanderschueren, E. Vanherck, P. Geusens, A. Suiker, W. Visser, K. Chung, and R. Bouillon, Androgen resistance and deficiency have different effects on the growing skeleton of the rat. Calcif. Tissue Int. 55, 198 – 203 (1994). 21. L. C. Hofbauer and S. Khosla, Androgen effects on bone metabolism: Recent progress and controversies. Eur. J. Endocrinol. 140, 271 – 286 (1999). 22. F. A. Tremollieres, D. D. Strong, D. J. Baylink, and S. Mohan, Progesterone and promegestone stimulate human bone cell proliferation and insulin-like growth factor – 2 production. Acta Endocrinol. 126, 329 – 337 (1992). 23. D. L. Manzi, C. C. Pilbeam, and L. G. Raisz, The anabolic effects of progesterone on fetal rat calvaria in tissue culture. J. Soc. Gynecol. Invest. 1, 302 – 309 (1994). 24. Z. Ikram, L. Dulipsingh, and K. M. Prestwood, Lack of effect of short-term micronized progesterone on bone turnover in postmenopausal women. J. Women’s Health & Gender-Based Med. 8, 973–978 (1999). 25. J. W. Chow, S. W. Fox, M. J. Lean, and T. J. Chambers, Role of nitric oxide and prostaglandins in mechanically induced bone formation. J. Bone Miner. Res. 13, 1039 – 1044 (1998). 26. J. M. Lean, C. J. Jagger, T. J. Chambers, and J. W. Chow, Increased insulin-like growth factor I mRNA expression in rat osteocytes in resonse to mechanical stimulation. Am. J. Physiol. 268, E318 – E327 (1995). 27. T. M. Skerry, Identification of novel signaling pathways during functional daptation of the skeleton to mechanical loading: The role of glutamate as a paracrine signaling agent in the skeleton. J. Bone Miner. Res. 17, 66 – 70 (1995). 28. E. J. Burger and J. Klein-Nulend, Mechanotransduction in bone–– Role of the lacuno-canalicular network. FASEB J. 13, S101 – 112 (1999). 29. S. V. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryotic Gene Expression. 8, 1 – 17 (1998). 30. H. Kawaguchi, C. C. Pilbeam, S. J. Vargas, E. E. Morse, J. A. Lorenzo, and L. G. Raisz, Ovariectomy enhances and estrogen replacement inhibits the activity of bone marrow factors that stimulate prostaglandin production in cultured mouse calvariae. J. Clin. Invest. 96, 539 – 548 (1995). 31. P. Collin-Osdoby, Role of vascular endothelial cells in bone biology. J. Cell. Biochem. 55, 304 – 309 (1994). 32. T. Suda, N. Takahashi, N. Udagawa, E. Jimmi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocrinol. Rev. 20, 345 – 357 (1999). 33. G. D. Roodman, Mechanisms of bone lesions in multiple myeloma and lymphoma. Cancer 80, 1557 – 1563 (1997). 34. L. G. Raisz, R. A. Luben, G. R. Mundy, J. W. Dietrich, J. E. Horton, and C. L. Trummel, Effect of osteoclast activating factor from

CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

human leukocytes on bone metabolism. J. Clin. Invest. 56, 408 – 413 (1975). F. E. Dewhirst, P. P. Stashenko, J. E. Mole, and T. Tsurumachi, Purification and partial sequence of human osteoclast-activating factor: Identity with interleukin 1b. J. Immunol. 135, 2562 – 2568 (1985). J. A. Lorenzo, S. L. Sousa, C. Alander, L. G. Raisz, and C. A. Dinarello, Comparison of the bone-resorbing activity in the supernatants from phytohemagglutinin-stimulated human peripheral blood mononuclear cells with that of cytokines through the use of an antiserum to interleukin 1. Endocrinology 121, 1164 – 1170 (1987). D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 319, 516 – 518 (1986). C. Miyaura, K. Kusano, T. Masuzawa, O. Chaki, Y Onoe, M. Aoyagi, T. Sasaki, T. Tamura, Y. Koishihara, Y. Ohsugi, et al., Endogenous bone-resorbing factors in estrogen deficiency: Cooperative effects of IL–1 AND IL–6. J. Bone Miner. Res. 10, 1365 – 1373 (1995). C. Miyaura, Y. Onoe, M. Inada, K. Maki, K. Ikuta, M. Ito, and T. Suda, Increased B-lymphopoiesis by interleukin 7 induces bone loss in mice with intact ovarian function: Similarity to estrogen deficiency. Proc. Natl. Sci. USA 94, 2581 – 1591 (1996). E. Romas, N. Udagawa, H. Zhou, T. Tamura, M Saito, T. Taga, D. J. Hilton, T. Suda, K. W. Ng, and T. J. Martin, The role of gp 130-mediated signals in osteoclast development: Regulation of interleukin 11 production by osteoblasts and its distribution of its receptor in bone marrow cultures. J. Exp. Med. 183, 2581 – 2591 (1996). Y. Onoe, C. Miyaura, T. Kaminakayashiki, Y. Nagai, K. Noguchi, Q. R. Chen, H. Seo, H. Ohta, S. Nozawa, I. Kudo, and T. Suda, IL-4 and IL-13 inhibit bone resorption by suppressing cyclooxygenase – 2-dependent prostaglandin synthesis in osteoblasts. J. Immunol. 156, 758 – 764 (1996). O. Hoffman, K. Klaushofer, H. Gleispach, H. J. Leis, T. Luger, K. Koller, and M. Peterlik, ?Interferon inhibits basal and interleukin 1induced prostaglandin production and bone resorption in neonatal mouse calvaria. Biochem. Biophys. Res. Commun. 143, 38 – 43 (1987). D. B. Lewis, H. D. Liggitt, E. L. Effmann, S. T. Motley, S. L. Teitelbaum, K. J. Jepsen, S. A. Goldstein, J. Bonadio, J. Carpenter, and R. M. Perlmutter, Osteoporosis induced in mice by overproduction of interleukin-4. Proc. Natl. Acad. Sci. USA 90, 11618 – 11622 (1993). S. T. Pottratz, T. Bellido, H. Mocharlia, D. Crabb, and S. C. Manolagas, 17b-Estradiol inhibits expression of human interleukin – 6 promoterreporter constructs by a receptor dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). T. Bellido, R. L. Jilka, B. F. Boyce, G. Girasole, H. Broxmeyer, S. A. Dalrymple, R. Murray, and S. C. Manolagas, Regulation of interleukin – 6, osteoclastogenesis, and bone mass by androgens. The role of the androgen receptor. J. Clin. Invest. 95, 2886 – 2895 (1995). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin – 1 and tumor necrosis factor is required to completely prevent bone loss in the early postovariectomy period. Endocrinology 136, 3054 – 3061 (1995). J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, et al., Mice lacking the type I interleukin – 1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). T. Sunyer, J. Lewis, and P. Collin-Osdoby, Estrogen’s bone-protective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and

50.

51.

52.

53.

54.

55. 56.

57. 58.

59.

60. 61.

62.

63.

64.

65.

66.

67.

68.

25

protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370. (1999). R. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurtry, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). R. Pacifici, J. L. Vannice, L. Rifas, and R. B. Kimble, Monocytic secretion of interleukin – 1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). W. McKane, S. Khosla, J. M. Peterson, K. Egan, and B. L. Riggs, Circulating levels of cytokines that modulate bone resorption: Effects of age and menopause in women. J. Bone Miner. Res. 9(8), 1313 – 1318 (1994). M. Kassem, S. Khosla, T. C. Spelsberg, B. L. Riggs, Cytokine production in the bone marrow microenvironment: Failure to demonstrate estrogen regulation in early postmenpausal women. J. Clin. Endocrinol. Metal. 81, 513 – 518 (1996). H. Kawaguchi, C. C. Pilbeam, J. R. Harrison, and L. G. Raisz, The role of prostaglandins in the regulation of bone metabolism. Clin. Orthop. Rel. Res. 313, 36 – 46 (1995). L. G. Raisz, Prostaglandins and bone: Physiology and pathophysiology. Osteoarthritis Cartilage 7, 419 – 421 (1999). C. C. Pilbeam, H. Kawaguchi, Y. Hakeda, O. Voznesensky, C. B. Alander, and L. G. Raisz, Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3E1 cells. J. Biol. Chem. 268, 25643 – 25649 (1993). D. C. Klein and L. G. Raisz, Prostaglandins: Stimulation of bone resorption in tissue culture. Endocrinology 86, 1436 – 1440 (1970). B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interleukin – 1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989). K. Fuller and T. J. Chambers, Effects of arachidonic acid metabolites on bone resorption by isolated rat osteoclasts. J. Bone Miner. Res. 4, 209 – 215 (1989). W. J. Jee and Y. F. Ma, The in vivo anabolic actions of prostaglandins in bone. Bone 21, 297 – 304 (1997. P. M. Fall, D. T. Breault, and L. G. Raisz, Inhibition of collagen synthesis by prostaglandins in the immortalized rat osteoblastic clonal cell line, Py1a: Structure activity relations and signal transduction mechanisms. J. Bone Miner. Res. 9, 1935 – 1943 (1994). L. G. Raisz, C. B. Alander, P. M. Fall, and H. A. Simmons, Effects of prostaglandin F2 on bone formation and resorption in cultured neonatal mouse calvariae: Role of prostaglandin E2 production. Endocrinology 126, 1076 – 1079 (1990). K. M. Reich and J. A. Frangos, Protein kinase-C mediates flowinduced prostaglandin E2 production in osteoblasts. Calcif. Tissue Int. 52, 62 – 66 (1993). J. Klein-Nulend, E. H. Burger, C. M. Semeins, L. G. Raisz, and C. C. Pilbeam, Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J. Bone Miner Res. 12, 45 – 51 (1997). D. D. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner. Res. 3, 409 – 414 (1988). D. J. Waters, D. D. Caywood, G. J. Trachte, R. T. Turner, and S. F. Hodgson, Immobilization increases bone prostaglandin E. Effect of acetylsalicylic acid on disuse osteoporosis studied in dogs. Acta Orthop. Scand. 62, 238 – 243 (1991). M. Weinreb, G. A. Rodan, and D. D. Thompson, Osteopenia in the immobilized rat hind limb is associated with increased bone resorption and decreased bone formation. Bone 10, 187 – 194 (1989). N. H. Bell, B. W. Hollis, J. R. Shary, D. R. Eyre, R. Eastell, A. Colwell, and R. G. G. Russell, Diclofenac sodium inhibits bone

26

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

LAWRENCE G. RAISZ resorption in postmenopausal women. Am. J. Med. 96, 349 – 353 (1994). D. J. Morton, E. L. Barrett-Connor, D. L. Schneider, Nonsteroidal anti-inflammatory drugs and bone mineral density in older women: The Rancho Bernardo study. J. Bone Miner. Res. 13, 1924 – 1931 (1998). M. A. Flynn, M. Qiao, C. Garcia, M. Dallas, and L. F. Bonewald, Avian osteoclast cells are stimulated to resorb calcified matrices by and possess receptors for leukotriene B–4. Calcif. Tissue Intl. 64, 154 – 159 (1999). R. J. van’t Hof, S. H. Ralston, Cytokine-induced nitric osxide inhibits bone resorption by inducing apoptosis of osteoclast progenitors and suppressing osteoclast activity. J. Bone Miner. Res. 12, 1797 – 1804 (1997). G. Crawford-Sharpe and C. J. Rosen, Insulin-like growth factor-I and the skeleton: New perspectives. Endocrinologist 9, 81 – 86 (1999). J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). L. F. Bonewald and S. L. Dallas, Role of active and latent transforming growth factor b in bone formation. J. Cell. Biochem. 55, 350 – 357 (1994). C. A. Conover Regulation and physiological role of insulin-like growth factor binding proteins. Endocrinol J. 43, S43 – S48 (1996). Y. Hakeda, H. Kawaguchi, M. Hurley, C. C. Pilbeam, C. Abreu, T. A. Linkhart, S. Mohan, M. Kumegawa, and L. G. Raisz, Intact insulin-like growth factor binding protein-5 (IGFBP-5) associates with bone matrix and soluble fragments of IGFBP-5 accumulate in culture medium of neonatal mouse calvariae treated with parathyroid hormone or prostaglandin E2. J. Cell. Physiol. 166, 370 – 379 (1996). C. Rosen, L. R. Donahue, S. Hunter, M. Holick, H. Kavookjian, A. Kirschenbaum, S. Mohan, and D. J. Baylink, The 24/25-kDa serum insulin-like growth factor-binding protein is increased in elderly women with hip and spine fractures. J. Clin. Endocrinol. Metab. 74, 24 – 27 (1992). S. Mohan and D. J. Baylink, Serum insulin-like growth factor binding protein (IGFBP)-4 and IGFBP-5 levels in aging and age-associated diseases, Endocrinology 7, 87 – 91 (1997). D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nat. Med. 2, 1132 – 1136 (1996). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell. Biol. 132, 195 – 210 (1996). D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF-beta. J. Cell Biochem. 70, 478 – 488 (1998). C. R. Dunstan, R Boyce, I. R. Garrett, E. Izbicka, W. H. Burgess, and G. R. Mundy, Systemic administration of acidic fibroblast growth factor (FGF-1) prevents bone loss and increases new bone formation in ovariectomized rats, J. Bone Miner. Res. 14, 953 – 959 (1999). M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type-I collagen gene expression in osteoblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). H. A. Simmons and L. G. Raisz, Effects of acid and basic fibroblast growth factor and heparin on resorption of cultured fetal rat long bones. J. Bone Miner. Res. 6, 1301 – 1305 (1991). M. M. Hurley, S. K. Lee, L. G. Raisz, P. Bernecker, and J. Lorenzo, Basic fibroblast growth factor induces osteoclast formation in murine bone marrow cultures, Bone 22, 309 – 316 (1998).

86. J. M. Hock and E. Canalis, Platelet-derived growth factor enhances bone cell replication, but not differentiated function of osteoblasts, Endocrinology 134, 1423 – 1428 (1994). 87. Z. Zhang, J. Chen, and D. Jin, Platlet-derived growth factor (PDGF)BB stimulates osteoclastic bone resorption directly: The role of rector beta, Biochem. Biophys. Res. Commun. 251, 190 – 194 (1998). 88. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage colony stimulating factor restores in vivo bone resorption in the OP/OP osteopetrotic mouse. Endocrinology 127, 2592 – 2594 (1990). 89. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast – macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 90. J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 91. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, and M. Kurimoto, Interleukin 18 inhibits osteoclast formation via T cell production of pranulocyte macrophage colony-stimulating factor J. Clin Invest. 101, 595 – 603 (1998). 92. H. Inoue, T. Tsujisawa, T. Fukuizumi, S. Kawagishi, and C. Uchiyama, SC-19220, a prostaglandin E2 antagonist, inhibits osteoclast formation by 1,25-dihydroxyvitamin D3 in cell cultures. J. Endocrinol. 161, 231 – 236 (1999). 93. K. Klaushofer, O. Hoffman, H. Gleispach, et al., Bone-resorbing activity of thyroid hormones is regulated to prostaglandin production in cultured neonatal mouse calvaria. J. Bone Miner. Res. 4, 305 – 312 (1989). 94. I. Holt, M. W. J. Davie, I. P. Braidman, and M. J. Marshall, Prostaglandin E (2) stimulates the production of interleukin – 6 by neonatal mouse parietal bones. Bone Miner. 25, 47 – 58 (1994). 95. M. C. Slootweg, W. W. Most, E. Vanbeek, L. P. C. Schot, S. E. Papapoulos, and C. W. G. M. Lowik, Osteoclast formation together with interleukin-6 production in mouse long bones is increased by insulinlike growth factor-I. J. Endocrinol. 132, 433 – 438 (1992). 96. J. Pfeilschifter and G. R. Mundy, Modulation of type b transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 97. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardzik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of latent transforming growth factor b (TGFb) complexes in osteoblastlike cell lines — Production of a latent complex lacking the latent TGFb-binding protein. J. Biol. Chem. 269, 6815 – 6822 (1994). 98. H. K. Nielsen, P. Charles, and L. Mosekilde, The effect of single oral doses of prednisone on the circadian rhythm of serum osteocalcin in normal subjects. J. Clin. Endocrinol. Metab. 67, 1025 – 1030 (1988). 99. H. K. Nielsen, K. Brixen, M. Kassem, P. Charles, and L. Mosekilde, Inhibition of the morning cortisol peak abolishes the expected morning decrease in serum osteocalcin in normal males — Evidence of a controlling effect of serum cortisol on the circadian rhythm in serum osteocalcin. J. Clin. Endocrinol. Metab. 74, 1410 – 1414 (1992). 100. K. K. Y. Ho and A. J. Weissberger, Impact of short-term estrogen administration on growth hormone secretion and action: Distinct routedependent effects on connective and bone tissue metabolism. J. Bone Miner. Res. 7, 821 – 827 (1992). 101. M. E. Cohensolal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. Devernejoul, Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. 77(6), 1648–1653 (1993). 103. J. Pfeilschifter, I. Diel, U. Pilz, K. Brunotte, A. Naumann, and R. Ziegler, Mitogenic responsiveness of human bone cells in vitro to hormones and growth factors decreases with age. J. Bone Miner. Res. 8, 707 – 718 (1993).

CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis 104. R. L. Jilka, R. S. Weinstein, K, Takahashi, A. M. Parfitt, and S. C. Manolagas, Linkage of decreased bone mass with impaired osteoclastogenesis in a murine model of accelerated senescence. J. Clin Invest. 97, 1732 – 1740 (1996). 105. G. D’Ippolito, P. C. Schiller, C. Ricordi, B. A. Roos, and G. A. Howard, Age-related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J Bone Miner. Res. 14, 1115 – 1122 (1999).

27

106. H. Bismar, I. Diel, R. Ziegler, and J. Pfeilschifter, Increased cytokine secretion by human bone marrow cells after menopause or discontinuation of estrogen replacement. J. Bone Miner. Res. 9, S158 (1994). 107. E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodelling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990).

CHAPTER 37

Animal Models for in Vivo Experimentation in Osteoporosis Research DONALD B. KIMMEL

Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131

I. Introduction II. A Perspective on in Vivo Animal Experimentation III. Criteria for Animal Models in Osteoporosis: A Time for Compromise

IV. The Criteria V. Animal Models for Human Osteoporosis References

of animals that are either losing bone or that have become osteopenic following ovariectomy because of the relationship of human osteoporosis to estrogen depletion. The FDA requires data from both the rat and a larger species. Studies of all rats must be of 12 months duration, but those in other species must be of 16 months duration, ostensibly equivalent to 4 years of human treatment, and include histologic evaluation. The guidelines also suggest measuring not only bone density and biochemical markers of bone turnover, but also bone strength by biomechanical testing, as surrogates for the propensity to develop fragility fractures. These guidelines are based principally on past experiments with current bone-active agents. Since these agents mainly slow bone turnover (e.g., estrogens, calcitonin, and bisphosphonates), the requirements are structured to find relatively modest changes over sustained periods. When published experience with agents that stimulate bone formation accumulates, it is likely that the FDA guidelines for testing such agents will evolve appropriately.

I. INTRODUCTION This chapter presents criteria for evaluating animal models of osteoporosis and then applies them using today’s facts. The criteria are based on: (i) knowledge about human osteoporosis, (ii) fundamental understanding of human and animal skeletons, (iii) experiments that use the same agent in different species, and (iv) recognition of unmet needs in osteoporosis research. Both the criteria and their evaluation should evolve with time as data about osteoporosis, each animal model, and new evaluation methods accumulate.

A. FDA Recommendations for Animal Models of Osteoporosis The Food and Drug Administration (FDA) has established guidelines for using animals in preclinical testing of agents intended to treat osteoporosis [1]. It recommends use

OSTEOPOROSIS, SECOND EDITION VOLUME 2

29

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

30

II. A PERSPECTIVE ON IN VIVO ANIMAL EXPERIMENTATION In vivo animal experimentation is overused in osteoporosis research. Though this may signal nothing more than an imbalance between the numbers of bench and clinical investigators, it may also mark the existence of something more alarming, misconceptions about efficient experimentation. In vivo adult human experimentation is desirable because its results require minimal extrapolation to predict the outcome in afflicted adult humans. Epidemiologic data bases that contain information about human osteoporosis should be mined exhaustively [2 – 6]. Convenient opportunities to add critical information to those data bases should be leveraged. When natural products or agents already approved for human use have a sound basis for efficacy in new clinical situations, a thorough examination of existing human data and a Phase I human trial, not an animal experiment, are likely to be the next logical steps. When an understanding of the inheritance of bone characteristics in humans is desired, studies of human pedigrees or affected human sib-pairs are the most efficient way of establishing the linkage of specific skeletal phenotypes to genetic loci [7 – 9]. In summary, despite the apparent emphasis of this chapter, skeletal investigators should heed the adage, “When possible, do it in humans,” more often. However, when experimental designs or new therapeutic agents that pose overt short- or long-term risks are to be evaluated, in vivo animal experimentation is the main option. For instance, when critical questions about the effects of agents on turnover, structure, and cellular activity in the sites in humans most prone to osteoporotic fracture (e.g., vertebrae, hip, and wrist) arise, it is apparent that human experimentation will not generate suitable samples. An appropriately designed animal experiment is the only approach. When animal experimentation is indicated and an in vivo animal model that accurately reproduces human data is available, an animal experiment is the best solution. The lack of such a strategy, including the availability and proper use of animal models in the late 1970s, might have played a role in the unfavorable impression left by an osteoporosis treatment trial involving sodium fluoride [10 – 12]. That outcome led to the FDA’s cautious and somewhat onerous requirement that investigators of future osteoporosis treatments demonstrate both bone mass increases and eventual anti-fracture efficacy in clinical trials to achieve full approval.

III. CRITERIA FOR ANIMAL MODELS IN OSTEOPOROSIS: A TIME FOR COMPROMISE Expecting full parallelism of human symptoms with in vivo animal models is unrealistic. These criteria are flexible, occasionally creatively applied, and clearly able to

DONALD B. KIMMEL

evolve as new evidence or needs appear. They place the highest value on animal models that match the clinically apparent behavior of osteoporosis, treating the detailed match of mechanisms as a fine point. All bone researchers can and should participate in the refinement of these criteria. Clinical investigators can facilitate the development of in vivo animal models by both providing details of the behavior of the human osteoporotic skeleton when new data become available and ranking the importance of various clinical characteristics of osteoporosis. In vivo animal investigators can help by doing experiments that show how their models fit important human symptoms. The recent history of the ovariectomized rat model is a textbook example of bench scientists doing experiments to validate a highly relevant preclinical in vivo animal model [13 – 16]. That effort, largely completed in the 1980s, was made possible by the bone mass and metabolic data collected about human osteoporosis during the 1970s. Recent efforts to develop standardized tests of bone fragility in small and large animals as surrogates for osteoporotic fracture are another example [17 – 20]. In vivo animal scientists can also help by developing more animal models that employ relevant physiologic conditions that can be readily applied by capable investigators. For example, when the goal is to characterize better an adult disease process, choosing a growing animal, simply to see quicker or more marked changes, makes little sense. Doing nerve resection to study disuse [21,22] may really model only the motionless, denervated limb, not the motionless limb. Combining extreme calcium deprivation with estrogen depletion to accelerate (or even just create) bone loss, may produce a confused situation that bears little resemblance to the clinical picture [23]. For osteoporosis research, using relevant methods in an adult animal to produce a consistent, albeit incomplete set of symptoms, is more acceptable than using nonphysiologic circumstances to develop a full set of symptoms. A model that gives sporadic results in the hands of numerous investigators or requires convoluted manipulations is not likely to become widely accepted.

A. Summary Today’s knowledge of animal models and osteoporosis means that using an irrelevant animal model in osteoporosis research is inexcusable and wasteful. Imposing nonphysiologic conditions to create desired symptoms is undesirable and always decreases a model’s relevance. Using a relevant, though costly model is occasionally necessary when unique information is needed. The continuous evaluation of new animal models for relevance and cost-effectiveness benefits all.

31

CHAPTER 37 Animal Models for in Vivo Experimentation

IV. THE CRITERIA Application of in vivo animal models for the skeleton has been reviewed elsewhere [24 – 27]. This chapter not only emphasizes many of their important points, but also widens the perspective by encompassing additional animal models. Its fundamental approach is to find animal models that match tissue behaviors that can be measured in vivo in osteoporotic humans, rather than to assure identical cellular mechanisms.

A. Existence of Growth and Adult Phases An in vivo animal model of osteoporosis should exhibit both growing and adult skeletal phases of meaningful duration. Peak bone mass is a concept in osteoporosis etiology that is receiving increased attention [28 – 30]. Its importance is clear because a single bone mass measurement at menopause is the best predictor of future fracture in healthy persons [31 – 33]. That value approximates peak bone mass, because bone mass changes before menopause are minor [34 – 38]. Growth processes, principally bone modeling and its modifiers like nutrition, physical activity, and heredity [30,39 – 55], determine peak bone mass. Adult phase skeletal processes (predominantly bone remodeling, but some bone modeling to effect shape changes in response to changing physical activity patterns) determine bone mass after attainment of peak bone mass [56]. The best animal models of osteoporosis would have both growth and adult phases of duration that allow both accuracy and time frame compression.

B. Menstrual/Estrus Cyclicity Humans not only have a menarche and regular, frequent ovulatory cycles, but also generally experience bone loss at cessation of ovarian function. The linkage of these two facts may seem weak, but a strong hint of the importance is the low bone mass that exists long term or transiently in amenorrheic individuals [58 – 64] and the bone accumulation that occurs upon resumption of normal menses [63 – 65]. Mammals (and humans) that have regular, frequent ovulatory cycles with high peaks of estradiol may be the only ones that suffer estrogen-depletion bone loss. Some hold that regularly cycling female mammals accumulate an estrogen-related component of bone that is integrated into the skeleton [66,67] and lost summarily at menopause, as if estrogen regulates bone mass [57]. It would follow that animals with infrequent cycles and low estradiol peaks might develop only a small estrogen-related bone compartment and show little estrogen depletion bone loss.

C. Natural Menopause Most women experience a natural menopause of 2 – 7 years duration [68]. Only 25 – 30% of women experience surgical menopause and many of those have preserved ovarian function or prompt estrogen replacement [69]. Most animal skeletal models of estrogen depletion invoke surgical or medical oophorectomy [13,70,71]. No meaningful differences in bone behavior between surgical and natural menopause are known [72,73]. Nonetheless, animals with a natural menopause with hormone changes like that of humans might be the best models for human osteoporosis. For instance the gradually increasing intermittency of estrogen peaks over a long period may elicit a different bone adaptive response than the precipitous removal of ovaries without estrogen replacement.

D. Bone Loss and Rise in Turnover Rate after Estrogen Depletion Following estrogen depletion, bone loss accelerates [74 – 77] for a time in multiple sites [78] and then decelerates and enters a semiplateau phase [79,80]. These changes are most pronounced in cancellous regions and at endocortical surfaces [81,82]. Estrogen status plays a much more important role in determining bone quantity in the aged female skeleton than does age [83 – 85]. Estrogen depletion changes in intracortical (Haversian) remodeling in humans are poorly documented. This cancellous and endocortical bone loss is accompanied by an increase in bone turnover [86,87] and a marked, transient negative calcium balance [88]. Histomorphometric changes of increased turnover across menopause are readily demonstrable within individual humans [89]. These behaviors should be easily confirmed in an accurate animal model of postmenopausal osteoporosis. It is even reasonable that the same measuring techniques applied in humans be expected to work for animals.

E. Skeletal Reponse to Estrogen Replacement Oophorectomized or menopausal women given prompt estrogen replacement experience a smaller rise in turnover [86,90], less bone loss [91 – 93], and fewer fractures [91,94 – 96] than those who receive no estrogen replacement [97]. This response is demonstrated well by histomorphometric technqiues in humans and animals [16,90,98]. This response is so stereotyped in adult women that one should expect an accurate ovariectomized animal model to experience an identical response to estrogen replacment.

32

DONALD B. KIMMEL

F. Development of Osteoporotic Fractures and Steady-State Osteopenia

H. Remodeling 1. CANCELLOUS

In its strictest definition, osteoporosis is marked by the occurrence of low trauma fractures of the spine and hip [99 – 101]. Because of the excellent ability to quantitate bone mass [102], the World Health Organization has recently provided a numeric definition of osteoporosis, as all women with bone mass at one or more bone sites 2.5 standard deviations below the young adult normal [103]. In the broadest sense, this is rational because women with osteoporotic fractures tend to be osteopenic [104 – 106]. While the WHO definition of osteoporosis should prove helpful for patients, clinicians, and clinical researchers alike, it is unlikely to help in animal experimentation. Though one might imagine experiments on osteopenic subpopulations of animals, the population-based studies necessary to screen for appropriatelyosteopenic animals for a WHO-like definition of osteoporosis are impractical. Enrolling only extremely osteopenic animals in trials, while appealing as a means of maximizing the chance of fragility fracture, is simply not feasible. Along those lines, it would be extremely helpful to have an accurate animal model that developed fragility fractures after estrogen depletion, because preclinical trials of the ability of agents to halt fragility fracture could be done. Despite the existence of good animal models for postmenopausal bone loss, none exhibit low trauma fragility fracture. This fact may speak to the importance of humans’ propensity to have both low peak bone mass and late-life bone loss, as necessary ingredients in the development of an osteoporotic fracture syndrome. Animals, in whom latelife bone loss can be readily created, may simply lack the contribution of low peak bone mass that is necessary to put them below their fracture threshold. The one animal model with low peak bone mass, the SAM/P6 mouse, has fragility fractures [107].

G. Bone Loss and Decreased Formation after Decreased Mechanical Usage Older humans experience loss of both cancellous and cortical bone as well as declines in bone formation unrelated to estrogen depletion [108,109]. These changes come during a life phase when a generalized decline in physical activity also occurs. While extreme physical inactivity, as during bedrest or paraplegia, causes marked bone loss [110 – 112], the impact of long-term, mild inactivity is neither well-understood nor easily assessed. This bone behavior should also be easily demonstrable in an accurate animal model of osteoporosis.

Cancellous bone remodeling, the in situ removal and replacement of aged bone tissue with new bone tissue, is an important process [56]. It has frequently been mentioned that bone surfaces adjacent to marrow cells experience high remodeling rates. This process is also expressed at endocortical surfaces, quite possibly initiated from marrow cavities [82]. An accurate animal model would display such activity in its skeleton. 2. CORTICAL Cortical bone plays a dominant role in skeletal structure. Adult humans have Haversian or intracortical bone remodeling, the process of in situ removal and replacement of aged cortical bone tissue [56]. Though the extent to which this process is affected by estrogen depletion has never been clearly defined, postmenopausal osteoporosis is normally characterized by only minimal cortical porosity. Nonetheless, an accurate animal model should display considerable steady-state Haversian remodeling, because Haversian remodeling is extremely important to the maintenance of cortical bone strength. Adverse changes in Haversian remodeling caused by agents being considered for treatment of osteoporosis can and should be revealed first in animal studies, not in Phase III or Phase IV human studies.

I. Time Frame Compression In adult, estrogen-depleted women, the phase of accelerated estrogen-depletion bone loss lasts 5 – 8 years. The time from attainment of peak bone mass until the development of fragility fractures is 30 or more years. An effective animal model known to experience peak bone mass followed by postovariectomy bone loss should compress both times by an order of magnitude or more.

J. Convenience Convenience for animal models is denominated as cost of purchase, availability, housing requirements, and handling difficulties and the necessity for designing/implementing/validating new analysis procedures. Using an animal model with the highest degree of accuracy can be so inconvenient that it is not worthwhile. It may occasionally be worse than doing a human study. For example, if an intervention requires active subject cooperation, animals may not be capable of that cooperation. On the other hand, in animal experiments, recruitment, lost sampling units, and compliance to pharmaceutical regimens, all nagging problems in clinical research, are nonissues.

33

CHAPTER 37 Animal Models for in Vivo Experimentation

Animals that seem convenient to some investigators because of specialized facilities and expertise at their organization are not the best choice for others lacking those tools. In today’s rapidly changing scientific environment, investigators tempted to maintain familiar, but outdated techniques often impede their own progress in the name of “convenience.” New enterprises, with the opportunity to allocate resources freshly, enjoy both the advantages and the burdens of the chance to define their own level of convenience. When an organization needs information that can be obtained only from experimental animals that require specialized housing and care, it often makes sense to contract externally for that phase of the project in the same manner that expert laboratories are asked to perform specialized analysis of various tissues. However one views it, the worst scenario is to have wasted scarce resources on an irrelevant model in the name of convenience.

V. ANIMAL MODELS FOR HUMAN OSTEOPOROSIS Animals that should receive initial consideration in osteoporosis research are bird, mouse, rat, rabbit, dog, pig, sheep, and nonhuman primate species. Because the above criteria are directed at clinical symptoms and outcomes in humans, the animal models will generally be evaluated as to how they duplicate those outcomes. Dissimilarities from the human condition in detailed mechanisms of development of one or more of the conditions may well exist. The

TABLE 1 Attribute

fit of each animal model to the above criteria is summarized in Table 1.

A. Avian Birds have growing and adult skeletal phases. Adult female birds have daily egg-laying cycles that correspond to alternating deposition and resorption of medullary bone during cyclic oviposition and egg calcification [113]. The bone accumulation phase occurs with rising serum estradiol and the removal phase accompanies falling serum estradiol. Estradiol treatment of male birds also causes medullary bone accumulation [114]. The bird skeleton experiences localized bone loss during immobilization and increased bone mass during applied mechanical loads [115 – 118]. Though avian models have broken new ground in understanding bone cell origin [119,120] and bone responses to the mechanical environment, they are not now known to be models for studying the estrogen/fracture-centered disease of osteoporosis. Furthermore, current data, mostly observational in nature, suggest that birds normally have little cancellous or Haversian remodeling. The interesting bone response to estrogen in birds provides many opportunities for experiments that bear on bone biology [121], but the estrogen-related bone buildup suggests a fundamental dissimilarity to adult mammalian physiology. While it can be loosely inferred that hypoestrogenemia in birds is associated with medullary bone loss, just as

Summary of in Vivo Animal Models for Osteoporosis

Human

Avian

Mouse

Rat

Dog

Pig

Sheep

Primate Yes

Growth and adult phases?

Yes

OK

OK

OK

Yes

Yes

Yes

Menstrual/estrus cyclicity

28 Days

Daily

Inducible

4 – 5 Days

205 Days

21 Days

21 Days 21 – 28 seasonal Days

Natural menopause

Yes

No

Yes*

Yes*

No

?

?

Yes

Bone loss after estrogen depletion

Yes

?

Probably

Yes

Not consistent

Weak

Weak

Yes

Response to estrogen

Turnover z4

Formation z3

Formation z3

Turnover z4

Not consistent

?

?

Turnover z4

Development of osteoporotic fractures

Yes

No

No

No

No

?

?

No

Cancellous remodeling

Yes

No

No

Some

Yes

Yes

Yes

Yes

Haversian remodeling

Present (study site difficult)

No

No

Low levels; inducible

Yes

Yes

Yes

Yes

Time frame compression

No

?

Yes

Yes

No

Some

Some

Some

*

Convenience

OK

Yes

Yes

Yes

Weak

Poor

Poor

Depends

Drug dose range like humans?

Yes

?

No

No (1/100)

Close

?

?

Yes

Cost effectiveness

Yes*

No

Yes

Yes

Weak

?

?

?

34

DONALD B. KIMMEL

estrogen depletion in mammals is associated with osteopenia, the course of bone mass following oophorectomy, an event more relevant to osteoporosis and osteopenia, is not known. The estrogen-related bone accumulation may actually prove important in understanding the accumulation of peak bone mass in pubertal humans [66,67,122 – 124], but seems likely to hinder the proper interpretation of experiments about osteoporosis, an adult disease. Birds are an example of a model that is convenient with low cost, but is largely irrelevant for osteoporosis research, because their skeletal behavior does not mimic features associated with adult human osteoporosis.

B. Mouse Though past success with using mice in physiologic aspects of skeletal research is encouraging, the future is even brighter because of the relative ease with which the mouse can be applied in genetic studies. It has been spectacular as a model for the human in osteopetrosis [125 – 129], osteoclast and stromal cell ontogeny [130], and cytokine and marrow studies [131 – 134]. The mouse has become even more popular for the ease with which its genome can be manipulated [135 – 138] and investigated. For example, C57Bl/6J and C3H/HeJ mice have low and high peak bone mass, respectively, which lend them to genetic investigations using F2 cross and F1 backcross studies of extremes of a continuous phenotype [139]. It is logical that the mouse be considered as an animal model that can solve new problems in the osteoporosis field. However, data that would validate the mouse as an in vivo model for osteoporosis research are in short supply. The few existing studies suggest that, much like in the rat, cancellous [138,140], but not cortical [141], bone loss occurs soon after ovariectomy. Genetically hypogonadal female mice are osteopenic [142]. Estrogen-depletion bone loss appears to be prevented by estrogen replacement [140]. Ectopic bone ossicles are preserved by estradiol treatment [143]. Mice experience cancellous and cortical bone loss in the vertebrae [144,145] and femur [146] during the second year of life. However, increased bone formation with new woven bone deposition after estrogen administration, as in birds, is a routine finding [140,141,147 – 150] and may depend upon the presence of the uterus [149]. The mouse’s inducible estrus cycle, rather than a hypothalamic – pituitary-regulated cycle, is another important dissimilarity to humans. Validation of the mouse as a model for osteoporosis will take some targeted experimental work. The few available reports suggest that the mouse suffers estrogen-depletion bone loss which is stopped by estrogen replacement. Its time course and the site specificity for the development of estrogen-depletion osteopenia must be established, perhaps

in a strain-specific fashion, as it has been for the rat [13,14]. The aberrant formation response to estradiol at endocortical surfaces must be considered. This unique response may be age-related. Until it is shown either that an older mouse has no such anabolic response to estradiol or that estradiol can elicit a bone formation response in larger adult mammals or adult humans, the chances for an acceptable murine model for osteoporosis seem remote. Furthermore, the similar life span of mice and rats, the smaller, though adequate, bone specimens provided by the mouse, and the overall similar cost for experimentation suggests that it will take an application of the mouse not possible in the rat, like gene manipulation or genetic marker studies, to induce investigators to choose the mouse. 1. THE SAM MOUSE The SAM/P6 (senescence-accelerated mouse) mouse has low peak bone mass and develops fractures in middle and old age [107,152 – 154]. It is the only experimental animal with documented fragility fractures of aging. The SAM mouse needs a full genetic [155], hormonal [156], and biomechanical characterization, including site specificity for fractures. If it does not suffer from collagen defects like those in osteogenesis imperfecta [135,157], it may, when combined with standard osteopenic prevention approaches like estrogen or bisphosphonate treatment, provide an opportunity to study the role of low peak bone mass in causing late-life fractures. It may provide further opportunity to study treatments that would enhance peak bone mass. Such studies must include testing of the material properties of the bone tissue itself, as has been done for MOV-13 osteogenesis imperfecta mice [137]. In the near term, considering the availability of SAM/P6, C57Bl/6J, and C3H/HeJ mice and the existing knowledge of the mouse genome [158], using proper crossing and back-crossing techniques [9], and cDNA probing of previously mapped genetic loci should be fruitful. A whole genome search should be able to identify polymorphisms at one or more genetic loci that are linked to low or high bone mass. The disclosure of such markers can lead, at the very least, to an increased understanding of peak bone mass control in one species, the mouse. Paradigms developed in small animals can subsequently be tested in larger animals or humans.

C. Rat The rat has a long history of providing accurate fundamental data that apply to the adult human skeleton. It gave the first evidence that osteoclasts ingest bone substance [159] and early evidence about osteoclast origin [160 – 163]. The rat skeleton was once held unsuitable as an adult human skeletal model because many epiphyseal

CHAPTER 37 Animal Models for in Vivo Experimentation

growth cartilages in male rats remain open past age 30 months [164]. The past unintentional focus on male rats masked important sex differences in epiphyseal closure times. In recent osteoporosis research projects using female rats, it became apparent that anatomically identical growth cartilages close earlier in females than in males [165 – 168]. Studies of the effects of gonadal hormones on growing male and female rats also suggest that intact females cease growing earlier than males [169]. When studies of rat models of male osteoporosis are undertaken, the results may be more difficult to interpret than in female rats [25]. Bone elongation ceases and effective epiphyseal closure ensues at important sampling sites in female rats by age 6 – 9 months, an age after which considerable useful experimental lifespan remains [170]. Periosteal expansion continues until about age 10 months, marking the age of peak bone mass in the adult female rat [171,172]. The mean healthy life span for these rats is 21 – 24 months, after which the incidence of estrogen-dependent mammary tumors rises to unacceptable levels. Thus the female rat has an appreciable life span both before and after attainment of peak bone mass. The lengthy bone accumulation period in the rat presents ample untapped opportunities for the study of both internal and external determinants of peak bone mass. Even those holding strong opinions about the inappropriateness of the rat as a model of adult human skeletal disease because of its “continuous growth and lack of remodeling,” have relented. They now only caution to use female rats of at least 6 – 9 months age and avoid studying Haversian remodeling [173]. Adult female rats have a regular estrus cycle in which E2 levels spike for 18 h every 4 days [174]. During the second year of life, the fraction of rats found in constant diestrus rises gradually [175], and cancellous bone loss is frequently observed [171]. While this is not a true menopause, spikes in E2 cease as bone loss occurs, making a linkage of rat “menopause” to cancellous bone loss possible. Following ovariectomy, loss of cancellous bone mass and strength occurs, then decelerates in a site-specific fashion, to enter a plateau phase [13,14,176 – 181]. This cancellous bone loss is accompanied by an increase in the rate of bone turnover [14,182 – 183]. These features mimic well the bone changes that follow oophorectomy or natural menopause in humans. Not all cancellous bone sites in the rat experience such bone loss [184], further tightening the parallel of the rat and the human skeletons, since osteoporotic fragility fractures and osteopenia are limited to a few sites in humans [78,185]. Dual energy X-ray absorptiometry (DXA), the current state-of-the-art for measuring bone mass in humans, can also be readily applied for rats [179,186 – 188]. Ovariectomized rats given prompt estrogen replacement experience no rise in turnover [16,98,189] and no bone loss [16,190 – 192]. This response fully parallels that seen in

35 postmenopausal women receiving estrogen replacement. Agents like bisphosphonates [15,180,193,194] and calcitonin [195,196] also block the rise in turnover and bone loss in rats, just as they do in humans [197 – 203]. Though interpretations of data from one laboratory suggest that high doses of estradiol stimulate bone formation in the rat [204,205], this contention has been largely discounted by others using similar experimental designs who offer alternative explanations that correctly consider the use of growing rats and an unusually long fluorochrome interval [206]. The rat, like other experimental animal models of osteoporosis, has no fragility fractures associated with the development of osteopenia. This apparent shortcoming can be overcome by mechanical tests of various bone through servohydraulic test systems. Such assays now exist for the vertebral body [17,18,191], femoral shaft [207], and proximal femur [18,20,208,209]. Prompt estrogen replacement in 1year-old ovariectomized rats preserves vertebral body strength in rats [210], much as it reduces fracture incidence in older women [6,91,92,94,96,214,215]. Using rat vertebral body strength as a surrogate for human vertebral body strength (implied from fracture likelihood) is likely to enable additional preclinical evaluation end points for agents being considered as treatments that can decrease osteoporotic fracture incidence. Such preclinical testing may avoid the problems encountered during the investigation of sodium fluoride as a treatment for osteoporosis [10,11]. Rats lose bone following immobilization. Several methods of permanent and temporary immobilization are available [21,22,216 – 220]. Acute phase [218], chronic phase [219], and recovery phase [220,221,223] bone changes related to disuse are easily studied. Rats are a weak model of glucocorticoid osteopenia, because they exhibit decreased bone formation, but do not consistently develop osteopenia [224,226]. Adult rats have adequate amounts of cancellous bone remodeling to permit useful experiments [173,227,228]. Controversy exists as to the relative frequency of modeling and remodeling activity. Classic reversal lines at the base of cancellous osteons are not present in 4-month-old rats [204], suggesting that modeling predominates at this age. While modeling would indeed be expected to be most frequent in such young rats, studies of cement line morphology in older rats are needed to confirm that bone activity in rats does undergo the expected transition to remodeling activity during adulthood [56]. In spite of the lack of studies of cement line morphology, indirect reasoning strongly supports the existence of remodeling in rat cancellous bone. Many cancellous bone regions in adult rats maintain stable bone mass for long periods of time while showing abundant bone formation and resorption. This neutral balance for resorption and formation [13,178] suggests the presence of remodeling. Whether the resorption and formation activity is locally or nonlocally linked, the regional tendency

36 toward mass preservation is more suggestive of remodeling than modeling, which is usually associated with changes in bone mass and shape [56]. In most of its cortical bone, the rat displays low levels of Haversian remodeling, often not detectably different from zero. However, processes resembling intracortical remodeling are induced by strong anabolic agents [229] or stressful metabolic conditions [230,235]. Unfortunately, it is not known whether these same agents and conditions also accelerate Haversian remodeling in humans. Current data suggest that the rat has such low levels of Haversian remodeling that it is impractical to use it for analysis of Haversian remodeling behavior, especially when evaluating agents that suppress Haversian remodeling. These data come from cross sections taken at the tibio – fibula junction or mid-femur [231,232]. However, cortical bone regions that surround cancellous bone, as in long bone metaphyses, are a reasonable, and currently uninvestigated, site to check to find higher levels of Haversian remodeling. In 3-month-old ovariectomized rats, the phase of accelerated estrogen-depletion bone loss lasts 3 – 4 months in the proximal tibial metaphysis [13], a 20-fold time frame compression when compared to that of estrogen-deplete women. The rat reaches peak bone mass by age 10 months, a 30-fold time frame compression when compared to that of the adult human. The rat is also among the most convenient of experiment animals to handle and house. In executing experiments with ovariectomized rats, it is usually wise to make some attempt at food restriction to limit weight gain in the ovariectomized groups. This may take the form of pair feeding to a sham group (conservative) or weight restriction to a sham group (aggressive). Either type of food restriction will accentuate bone loss [233,234], possibly creating more reliable cortical bone loss and certainly quickening the development of osteopenia. When selecting aged female rats for study, investigators are often tempted to choose “retired breeders” because of their low cost. However, the highly variable skeletal status of retired breeders is likely to influence experimental outcomes. Most female rats are bred for the first time at age 3 – 4 months, before bone elongation is completed and long before peak bone mass is achieved. They often eat a diet that is 0.6% calcium and 0.55% phosphorus. The calcium demands of pregnancy and lactation inhibit normal patterns of skeletal growth. Thus, rats delivered to investigators as retired breeders at age 6 or 7 months have usually had two or three litters plus calcium-stressful lactation periods [235,236], causing them to have abnormally immature skeletons with long bone metaphyses that are relatively depleted of cancellous bone. When they are taken from breeding and placed in an experiment, they experience a period of “catch-up” growth during which cancellous bone of the metaphysis is repleted by resumption of endochondral processes and considerable cortical bone enhancement

DONALD B. KIMMEL

occurs, such that they reach a usual peak bone mass by age 10 – 12 months. This period of catch-up growth might be confused with the anabolic effect of an agent. Thus, the retired breeder female rat is somewhat acceptable when experiments to add bone to the osteopenic skeleton are planned, because its skeleton is fundamentally immature and osteopenic. The interaction of delayed growth processes with potentially anabolic agents is likely to be a problem for quantitative interpretation. However, when doing prevention experiments, the retired breeder female rat is generally unreliable because of both the (likely) already osteopenic condition of the skeleton, from which little more bone could be lost, and the period of catch-up growth. Virgin female rats that are more skeletally mature and not osteopenic in their long bone metaphyses, are much more acceptable for prevention studies, because they have considerably more bone in their long bone metaphyses that can be lost and whose loss can be prevented after ovariectomy [237]. 1. SUMMARY The FDA has made a wise choice in requiring rat experiments during osteoporosis research. The ovariectomized rat is an excellent model that correctly emulates the most important clinical features of the estrogen-deplete adult human skeleton. Its site-specific development of cancellous osteopenia is one of the most certain physiologic responses in skeletal research. Ample time exists for experimental designs that either preventestrogen-depletion bone loss or restore bone lost after estrogen depletion. Its response to estrogen replacement closely parallels that seen in the human. The rat’s low levels of Haversian remodeling present little immediate problem when testing agents for their ability to prevent the loss of cancellous bone or rebuild lost cancellous bone. Investigators can compensate for the lack of fragility fractures by biomechanical testing. Rats are convenient for most investigators; unbred females ages 6 – 10 months are the optimal choice. Existing laboratory measurement tools of biochemistry, histomorphometry, densitometry, and mechanical testing are readily applicable.

D. Guinea Pig, Rabbit, Ferret, and Cat Though occasional reports using guinea pigs, rabbits, ferrets, and cats in osteoporosis research have appeared [238 – 240], few studies of estrogen depletion bone loss exist. Too few experiments generally exist to assess properly their validity. Seven-month-old ovariectomized guinea pigs do not lose bone by 4 months postsurgery [239]. Adult rabbits have active Haversian remodeling and might serve as a model for testing Haversian remodeling after treatment with agents that have strong anabolic effects on cancellous bone. Their reproductive cycle bears dissimilarity to that of

37

CHAPTER 37 Animal Models for in Vivo Experimentation

humans. The success of the rabbit and dog [238,241], animals with significant Haversian remodeling, as models for glucocorticoid-induced osteopenia, coupled with the failure of the rat, an animal with mimimal Haversian remodeling, suggests that glucocorticoid osteopenia is a disease of deranged Haversian remodeling. The ferret, weighing less than 1 kg, has Haversian remodeling [242]. Its normal skeletal physiology, including the accumulation of estrogen-dependent bone that seems to accompany normal cyclicity in other mammals, is dependent on a regular light cycle [243]. It exhibits expected changes in bone remodeling rate and bone volume during treatment with parathyroid hormone [244,245]. Though a relatively small animal with appreciable quantities of Haversian remodeling would be welcome, the general appeal of other animal models with both Haversian remodeling and already proven human-like endocrine characteristics cannot be overlooked.

E. Dog The adult dog is a reliable model for the adult human skeleton that is generally similar in both metabolic and structural characteristics. Studies of 239Pu-injected beagles not only first proved the existence of adult cancellous bone remodeling [159], but also contributed one of the earliest indications of the hematogenous origin of osteoclasts [246]. The ratio of cortical to cancellous bone is similar to that in humans [247 – 249]. Haversian and cancellous osteons remodel with similar morphology, though more rapidly in dogs [250,251]. Skeletal responsiveness of the adult dog parallels that of the adult human for corticosteroids [241,252], uremia [253,254], bisphosphonates [255,256], disuse [110 – 112,257 – 260], and parathyroid hormone excess [245,261]. In contrast to all other applications for adult beagles as a model of the adult human skeleton, the oophorectomized beagle is controversial. Many individual studies lack significant findings, but a metaanalysis of the overall data [262,263] suggests that estrogen-depletion bone loss of perhaps 8 – 10% annually occurs in oophorectomized beagles. This conclusion is based on the three most prominent types of studies: (i) those that find a significant decline [264 – 266,270 – 273], (ii) those that find a nonsignificant decline [71,267,268,274], and (iii) those that find no change in bone mass or strength [23,269,275 – 282]. Increases rarely occur [283,284], usually in experiments of very small sample size (N  3). One strongly positive paper showing bone loss [264] uses spinal densitometry, a tool of high precision [285]. A mildly positive paper found some decline in vertebral trabecular strength, using mechanical testing, a tool with good specificity, though weak precision for trabecular bone evaluation [267,268].

The temporal pattern in bone formation after oophorectomy suggests an early rise with a later return to baseline or subbaseline levels. Work from one group [270 – 272] suggests that formation falls rapidly to 50% of baseline, with no increased turnover phase. This suggests the possibility of a marked dissimilarity to histomorphometric findings in transmenopausal humans, where the early rise in both formation and individual cell activity [89] has been well documented [86]. OF

1. THE CANINE ESTRUS CYCLE COMPARED TO THOSE HUMANS AND OTHER ANIMALS

E2 levels are usually very low in the dog, rising twice yearly for several weeks [286 – 288]. In rats, E2 spikes for 18 h every 4 days [174]. In women, E2 spikes for 1 or 2 days monthly [289,290]. The estrus cycle in monkeys has a similar frequency to that in humans, but reaches E2 peaks only about half as high [291]. Integrated estrogen exposure in dogs, though only marginally less than in rats, is only onefourth that in humans. It is similar to that in primates, except during the peak periods. E2 peaks in canines are one-sixth as frequent and one-sixth as high as in humans. This difference could contribute to the dog’s developing only a small estrogen-dependent compartment of cancellous bone. Because of the inconsistency of the ovariectomized dog model, predicting the outcome of experiments with it is difficult. The data suggest that estrogen treatment suppresses turnover, but that its bone mass sparing effects are uncertain [279 – 282]. 2. SUMMARY The adult beagle is an excellent model of the adult human skeleton except for estrogen depletion. Its principal advantage over smaller animals is its Haversian remodeling. Though the oophorectomized beagle has estrogen-depletion osteopenia, poor interlaboratory reproducibility has caused skepticism. The main problem has been that most of the individual studies contain insufficient power to detect the expected bone loss [265 – 284]. Beagles are less estrogen-replete than women and may have a smaller estrogendependent compartment of bone in their skeleton. Despite its inconsistency for developing estrogen-depletion bone loss, the dog remains an excellent model for testing the effects on Haversian remodeling of agents that have strong anabolic effects on cancellous bone. However, a general strategy of using one animal model displaying both Haversian remodeling and consistent estrogen-depletion bone loss has high appeal.

F. Pig The pig has both growing and adult skeletal phases. The oophorectomized pig has been tested a few times

38

DONALD B. KIMMEL

[292 – 294]. In one study, though minor structural deterioration was noted, no differences in bone mass, either by densitometry or histomorphometry, were seen. In a second study, minor bone loss from the fourth lumbar vertebra was noted at 3 months postovariectomy. The pig has a regular estrus cycle that is somewhat shorter than the human menstrual cycle. Pigs have been used successfully to study fluoride and exercise effects on the skeleton [295,296]. It appears that the age of peak bone mass in the pig is older than 2 – 3 years, a factor that has confounded experimental efforts thus far and seems destined to introduce logistical difficulty into designing experiments with the estrogen-deplete pig. More work will be necessary before the pig can gain acceptance as a model of estrogen-depletion bone loss.

G. Sheep The ewe has both growing and adult skeletal phases, but the age of peak bone mass is not known. Some species have a regular estrus cycle during the short days of winter and experience anestrus when days are longer [297,298]. The ewe has also been used to study fluoride effects in the adult skeleton. The findings seem to parallel histologic changes in humans with frequent signs of increased formation, sluggish mineralization, and toxicity to bone forming cells [26,299 – 301]. Skeletal behavior over oophorectomy has only recently been addressed [302]. Early data suggest a postoophorectomy picture that is considerably less straightforward than that seen in postmenopausal women, rats, or monkeys. Glucocorticoid data seem consistent with findings in other animal models and humans [252,304]. Sheep can be housed readily at most vivariums and pose little problem for handling. More data are needed to determine the age of peak bone mass, and the role of seasonal variation in bone mass, before the usefulness of the adult ewe as an in vivo model of osteoporosis can be validated.

H. Nonhuman Primate The nonhuman primate has both growing and adult skeletal phases. Peak bone mass occurs around age 10 – 11 years in cynomolgous and rhesus monkeys and baboons [305 – 307]. All nonhuman primates have a regular menstrual cycle with an approximate duration of 28 days, an excellent analog of adult women. Nonhuman primates experience a natural menopause near the end of the second decade of life. Nonhuman primates in captive environments are ideal candidates for family pedigree analysis, the most powerful way of establishing genetic linkage to phenotypic traits [9]. The ability to study DNA polymorphisms in nonhuman primates has also been demonstrated [308,309]. Combining

genetic studies of the mouse and nonhuman primate offers a multispecies approach to understanding the genetic control of peak bone mass that can lead to the development of concepts that could be applied in humans. Nonhuman primates experience decreased bone mass and bone strength with increased turnover [310], after ovariectomy [70,310 – 315], premature menopause [316], or GnRH agonist treatment [70]. The response to estrogen replacement has not been well-characterized. Nonhuman primates experience bone loss with age [317 – 318] and after prolonged immobilization [319]. Histomorphometric studies of primates and humans yield remarkably similar values [306 – 310,320,321]. Postovariectomy changes in bone mass are frequently masked by the selection of animals of an age that they are still acquiring peak bone mass [322 – 324]. Late life spinal pathology in baboons [326,327] and rhesus monkeys [318,325] is mostly osteoarthritis. Baboons experience osteopenia [317] with an age-related decline in anterior vertebral height that bears more similarity to that accompanying osteoarthritis than to vertebral crush fractures [326 – 328]. This means that nonhuman primates are not likely to be a model of fragility fractures, because in humans osteoarthritis and osteoporosis tend to be mutually exclusive conditions [328 – 330]. In addition, spinal bone mass measurements in older primates may be affected because the osteophytes that occur in osteoarthritis have enough calcium to change spinal bone mass values meaningfully and obscure proper interpretation [331 – 334]. Like the dog, nonhuman primates offer cancellous and Haversian remodeling that is directly comparable to that found in the adult human skeleton. The time course of bone loss after ovariectomy has not been well-characterized, but appears to offer significant time frame compression when compared to that in humans. Like all skeletal experiments intended to yield information about osteoporosis, care should be exercised in choosing animals that have achieved peak bone mass. Extreme requirements for housing and care of nonhuman primates limit their use to relatively small numbers of facilities. However, when handled by experienced staff in an appropriate environment, they present few care problems.

I. Summary of All in Vivo Animal Models When an in vivo osteoporosis research project cannot be done in humans, the 10-month-old female rat is the first animal model of choice. It has reached peak bone mass and can be manipulated to accurately simulate most clinical findings of osteoporosis in the adult female skeleton. Methods like serum biochemistry, histomorphometry, and densitometry that are routinely used in humans, are applicable in rats. Like all animal models of osteoporosis, the rat

CHAPTER 37 Animal Models for in Vivo Experimentation

develops no fragility fractures. However, mechanical testing of rat bones substitutes well as an accurate predictor of bone fragility. However, the rat’s low levels of Haversian remodeling do not permit accurate evaluation of intracortical bone behavior, making studies of larger animals a must. The dog is generally an accurate model of the adult human skeleton, but has yielded inconsistent results in the acutely estrogen-deplete state that differ from those in humans. Data on all other oophorectomized species except nonhuman primates, are scarce. Estrogen-deplete nonhuman primates are the large animal of choice when Haversian remodeling outcomes are to be studied. Avians have a metabolically active skeleton whose behavior does not seem directly relevant to osteoporosis research. Data about estrogen-deplete mice now seem promising, but the high state of development of the ovariectomized rat model suggests that developing an ovariectomized mouse model as an alternative is not urgent. When studied before age 10 months, female rats and mice have skeletal growth and maturation phases that are useful for experiments about peak bone mass. Mice are likely to be useful in revealing both genetic markers that control peak bone mass and gene manipulations that affect both bone mass and structure. The rabbit is the animal model of choice for studying glucocorticoid osteoporosis.

References 1. “Guidelines for Preclinical and Clinical Evaluation of Agents Used in the Prevention or Treatment of Postmenopausal Osteoporosis.” Division of Metabolism and Endocrine Drug Products, Food and Drug Administration, Washington, DC, 1994. 2. K. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1218 (1993). 3. S. B. Soroko, E. L. Barrett-Conner, S. L. Edelstein, and D. Kritz-Silverstein, Family history of osteoporosis and bone mineral density at the axial skeleton: The Rancho Bernardo study. J. Bone Miner. Res. 9, 761 – 770 (1994). 4. I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Lopes-Vaz, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. Perez Cano, A. Rapado, and C. Ribot, The variable incidence of hip fracture in Southern Europe: The MEDOS study. Osteoporosis Int. 4, 253 – 263 (1994). 5. G. Jones, T. Nguyen, P. N. Sambrook, P. J. Kelly, C. Gilbert, and J. A. Eisman, Symptomatic fracture incidence in elderly men and women: The Dubbo osteoporosis epidemiology study (DOES). Osteoporosis Int. 4, 277 – 282 (1994). 6. D. P. Kiel, D. T. Felson, J. J. Anderson, P. W. F., and Wilson, M. A. Moskowitz, Hip fracture and the use of estrogens in postmenopausal women. The Framingham Study. N. Engl. J. Med. 317, 1169 – 1174 (1987). 7. B. K. Suarez, J. Rice, and T. Reich, The generalized sib pair IBD distribution: Its use in the detection of linkage. Ann. Hum. Genet. 42, 87 – 92 (1978). 8. N. Risch, Linkage strategies for genetically complex traits. II. The power of affected relative pairs. Am. J. Hum. Genet. 46, 229–241 (1990).

39 9. E. S. Lander and N. J. Schork, Genetic dissection of complex traits. Science 263, 2037 – 2048 (1994). 10. M. Kleerekoper, E. L. Peterson, D. A. Nelson, E. Phillips, M. A. Schork, B. C. Tilley, and A. M. Parfitt, A randomized trial of sodium fluoride as a treatment for postmenopausal osteoporosis. Calcif. Tissue Int. 49, 155 – 161 (1991). 11. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. Chao, H. W. Wahner, J. M. Muhs, S. L. Cedel, and L. J. Melton, Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). 12. B. L. Riggs, W. M. O’Fallon, A. Lane, S. F. Hodgson, H. W. Wahner, J. Muhs, E. Chao, and L. J. Melton III, Clinical trial of fluoride therapy in postmenopausal osteoporotic women: Extended observations and additional analysis. J. Bone Miner. Res. 9, 265 – 275 (1994). 13. T. J. Wronski, L. M. Dann, K. S. Scott, and M. Cintron, Long-term effects of ovariectomy and aging on the rat skeleton. Calcif. Tissue Int. 45, 360 – 366 (1989). 14. T. J. Wronski, M. Cintron, and L. M. Dann, Temporal relationship between bone loss and increased bone turnover in ovariectomized rats. Calcif. Tissue Int. 42, 179 – 183 (1988). 15. T. J. Wronski, L. M. Dann, K. S. Scott, and L. R. Crooke, Endocrine and pharmacological suppressors of bone turnover protect against osteopenia in ovariectomized rats. Endocrinology 125, 810 – 816 (1989). 16. T. J. Wronski, M. Cintron, A. L. Doherty, and L. M. Dann, Estrogen treatment prevents osteopenia and depresses bone turnover in ovariectomized rats. Endocrinology 123, 681 – 686 (1988). 17. Li Mosekilde, C. H. Sogaard, C. C. Danielson, O. Torring, and M. H. L. Nilsson, The anabolic effects of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: A comparison study between hPTH-(1 – 34) and hPTH-(1 – 84) Endocrinology 129, 421 – 428 (1991). 18. B. C. Toolan, M. Shea, E. R. Myers, R. E. Borchers, J. G. Seedor, H. Quartuccio, G. Rodan, and W. C. Hayes, Effects of 4-amino-1-hydroxybutylidene bisphosphonate on bone biomechanics in rats. J. Bone Miner. Res. 7, 1399 – 1406 (1992). 19. M. Kasra and M. D. Grynpas, Effect of long-term ovariectomy on bone mechanical properites in young female cynomolgus monkeys. Bone 15, 577 – 561. (1994). 20. C. H. Sogaard, T. J. Wronski, J. E. McOsker, and Li Mosekilde, The positive effect of PTH on femoral neck bone strength in ovariectomized rats is more pronounced than that of estrogen or bisphosphonate. Endocrinology 134, 650 – 657 (1994). 21. J. A. Svesatikoglou and S. Mattson, Changes in composition and metabolic activity of the skeletal parts of the extremity of the adult rat following resection of the sciatic nerve. Acta Chir. Scand. (Suppl. 476), 16 – 20 (1976). 22. G. K. Wakley, B. L. Baum, K. S. Hannon, and R. T. Turner, The effects of tamoxifen on the osteopenia induced by sciatic neurotomy in the rat: A histomorphometric study. Calcif. Tissue Int. 43, 383 – 388 (1988). 23. P. Geusens, L. P. C. Schot, J. Nijs, and J. Dequeker, Calcium-deficient diet in ovariectomized dogs limits the effects of 17-b-estradiol and nandrolone decanoate on bone. J. Bone Miner. Res. 6, 791 – 798 (1991). 24. J. B. Rodgers, M. C. Monier-Faugere, and H. H. Malluche, Animal models for the study of bone loss after cessation of ovarian function. Bone 14, 369 – 377 (1993). 25. D. Vanderschueren, E. Van Herck, P. Schot, E. Rush, T. Einhorn, P. Geusens, and R. Bouillon, The aged male rat as a model for human osteoporosis: Evaluation by nondestructive measurements and biomechanical testing. Calcif. Tissue Int. 53, 342 – 347 (1993). 26. P. Chavassieux, Bone effects of fluoride in animal models in vivo. A review and a recent study. J. Bone Miner. Res. 5 (Suppl. 1), S95 – S100 (1990).

40 27. M. Cesnjaj, A. Stavljenic, and S. Vukicevic, In vivo models in the study of osteopenias. Eur. J. Clin. Chem. Clin. Biochem. 29, 211 – 219 (1991). 28. C. H. Chesnut, Theoretical overview: Bone development, peak bone mass, bone loss, and fracture risk. Am. J. Med. 91 (Suppl. 5B), 2S – 4S (1991). 29. V. Matkovic, D. Fontana, C. Tominac, P. Goel, and C. H. Chesnut III, Factors that influence peak bone mass formation: A study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr. 52, 878 – 888 (1990). 30. V. Matkovic, Calicum intake and peak bone mass. N. Engl. J. Med. 327, 119 – 120 (1992). 31. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Baseline measurement of bone mass predicts fracture in white women. Ann. Int. Med. 111, 355 – 361 (1989). 32. P. Gardsell, O. Johnell, and B. E. Nilsson, Predicting fractures in women by using forearm bone densitometry. Calcif. Tissue Int. 45, 235 – 242 (1989). 33. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture spinal osteoporosis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1988). 34. R. R. Recker, J. M. Lappe, K. M. Davies, and D. B. Kimmel, Change in bone mass immediately before menopause. J. Bone Miner. Res. 7, 857 – 862 (1992). 35. A. Rodin, B. Murby, M. A. Smith, M. Caleffi, I. Fentimen, M. G. Chapman, and I. Fogelman, Premenopausal bone loss inthe lumbar spine and femoral neck: A study of 225 Caucasian women. Bone 11, 1 – 15 (1990). 36. J. Schlechte, L Walkner, and M. Kathol, A longitudinal analysis of premenopausal bone loss in healthy women and women with hyperprolactinemia. J. Clin. Endocrinol. Metab. 75, 698 – 703 (1992). 37. R. Marcus, J. Kosek, A. Pfefferbaum, and S. Horning, Age-related loss of trabecular bone in premenopausal women: A biopsy study. Calcif. Tissue Int. 35, 406 – 409 (1983). 38. H. E. Meema and S. Meema, Longitudinal microradioscopic comparisons on endosteal and juxtaendosteal bone loss in premenopausal and postmenopausal women, and in those with end stage renal disease. Bone 8, 343 – 350 (1988). 39. C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87. (1992). 40. H. H. Jones, J. D. Priest, W. C. Hayes, C. C. Tichenor, and D. A. Nagel, Humeral hypertrophy in response to exercise. J. Bone Jt. Surg. 59A, 204 – 208 (1977). 41. H. Haapasalo, P. Kannus, H. Sievanen, A. Heinonen, P. Oja, and I. Vuori, Long-term unilateral loading and bone mineral density and content in female squash players. Calcif. Tissue Int. 54, 249 – 255 (1994). 42. T. J. Beck, C. B. Ruff, W. W. Scott, C. C. Plato, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29. 43. P. J. Kelly, L. Twomey, P. N. Sambrook, and J. A. Eisman, Sex differences in peak adult bone mineral density. J. Bone Miner. Res. 5, 1169 – 1175 (1990). 44. Y. Liel, J. Edwards, J. Shary, K. M. Spicer, L. Gordon, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip, and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66, 1247 – 1250 (1988). 45. D. P. McCormick, S. W. Ponder, H. D. Fawcett, and J. L. Palmer, Spinal bone mineral density in 335 normal and obese children and adolescents: Evidence for ethnic and sex differences. J. Bone Miner. Res. 6, 507 – 525 (1991). 46. M. Moller, A. Horsman, B. Harvald, M. Hauge, K. Henningsen, and B. E. C. Nordin, Metacarpal morphometry in monozygotic and dizygotic elderly twins. Calcif. Tissue Res. 25, 197 – 201 (1978).

DONALD B. KIMMEL 47. J. Dequeker, J. Nijs, A. Verstraeten, P. Geusens, and G. Gevers, Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8, 207 – 209 (1987). 48. D. M. Smith, W. E. Nance, W. K. Ke, J. C. Christian, and C. C. Johnston, Genetic factors in determining bone mass. J. Clin. Inv. 52, 2800 – 2808 (1973). 49. N. A. Pocock, J. A. Eisman, J. L. Hopper, M. G. Yeates, P. N. Sambrook, and S. Ebert, Genetic determinants of bone mass in adults: A twin study. J. Clin. Inv. 80, 706 – 710 (1991). 50. E. Seeman, J. L. Hopper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554 – 558 (1989). 51. R. A. Evans, G. M. Marel, E. K. Lancaster, S. Kos, M. Evans, and S. Y. P. Wong, Bone mass is low in relatives of osteoporotic patients. Ann. Int. Med. 109, 870 – 873 (1988). 52. F. A. Tylavsky, A. D. Bortz, R. L. Hancock, and J. J. B. Anderson, Familial resemblance of radial bone mass between premenopausal mothers and their college-age daughters. Calcif. Tissue Int. 45, 265 – 272 (1989). 53. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 54. N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal phsyiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89, 6665 – 6669 (1992). 55. F. G. Hustmyer, M. Peacock, S. Hui, C. C. Johnston, and J. Christian, Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J. Clin. Inv. 94, 2130 – 2134 (1994). 56. H. M. Frost, The skeletal intermediary organization. Metab. Bone Dis. Rel. Res. 4, 281 – 290 (1983). 57. H. M. Frost, The mechanostat: A proposed pathogenic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone Miner. 2, 73 – 85 (1987). 58. B. L. Drinkwater, K. Nilson, C. H. Chesnut, W. J. Brenner, S. Shainholtz, and M. B. Southworth, Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311, 277 – 281 (1984). 59. R. Marcus, C. Cann, P. Madvig, J. Minkoff, M. Goddard, M. Bayer, M. Martin, L. Gaudiani, W. Haskell, and H. Genant, Menstrual function and bone mass in elite women distance runners. Ann. Intern. Med. 102, 158 – 163 (1985). 60. C. E. Cann, M. C. Martin, H. K. Genant, and R. B. Jaffe, Decreased spinal mineral content in amenorrheic women. J. Am. Med. Assoc. 251, 626 – 629 (1984). 61. B. M. K. Biller, H. B. A. Baum, D. I. Rosenthal, V. C. Saxe, P. M. Charpie, and A. Klibanski, Progressive trabecular osteopenia in women with hyperprolactinemic amenorrhea. J. Clin. Endocrinol. Metab. 75, 692 – 697 (1992). 62. H. Rico, F. Arnanz, M. Revilla, S. Perera, M. Iritia, L. F. Villa, and I. Arribas, Total and regional bone mineral content in women treated with GnRH agonists. Calcif. Tissue Int. 52, 354 – 357 (1993). 63. A. Klibanski and S. L Greenspan, Increase in bone mass after treatment of hyperprolactinemic amenorrhea. N. Engl. J. Med. 315, 542 – 546 (1986). 64. Y. M. Dawood, V. Lewis, and J. Ramos, Cortical and trabecular bone mineral content in women with endometriosis: Effect of gonadotropin-releasing hormone agonist and danazol. Fertil. Steril. 52, 21 – 27 (1989). 65. B. L. Drinkwater, K. Nilson, and S. Ott, Bone mineral density after resumption of menses in amenorrheic athletes. J. Am. Med. Assoc. 256, 380 – 382 (1986). 66. S. M. Garn, “The Early Gain and Later Loss of Cortical Bone. Thomas, Springfield, IL, 1970. 67. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechart, H. K. Huang, E. E. Schulz, C. R. Libanati, and C. C. Cann,

CHAPTER 37 Animal Models for in Vivo Experimentation

68. 69. 70.

71.

72.

73.

74.

75.

76.

77.

78.

79. 80.

81.

82.

83.

84. 85. 86. 87.

88.

89.

Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). A. E. Treloar, Menstrual cyclicity and the pre-menopause. Maturitas 3, 249 – 264 (1981). B. McMahon and B. Worcester, “Age at Menopause,” Series 11:19. National Center for Health Statistics, Washington, DC, 1966. D. R. Mann, K. G. Gould, and D. C. Collins, A potential primate model for bone loss resulting from medical oophorectomy or menopause. J. Clin. Endocrinol. Metab. 71, 105 – 110 (1990). R. B. Martin, R. L. Butcher, L. L. Sherwood, P. Buckendahl, R. D. Boyd, D. Farris, H. Sharkey, and G. Dannucci, Effects of ovariectomy in beagle dogs. Bone 8, 23 – 31 (1987). C. E. Cann, H. K. Genant, B. Ettinger, and G. S. Gordan, Spinal mineral loss in oophorectomized women. J. Am. Med. Assoc. 244, 2056 – 2059 (1980). D. Hartwell, B. J. Riis, and C. Christiansen, Changes in vitamin D metabolism during natural and medical menopause. J. Clin. Endocrinol. Metab. 71, 127 – 132 (1990). B. E. C. Nordin, A. Horsman, R. Brook, and D. A. Williams, The relationship between oestrogen status and bone loss in postmenopausal women. Clin. Endocrinol. 5, 353S – 361S (1976). O. Lindquist, C. Bengtsson, T. Hansson, and R. Jonsson, Changes in bone mineral content of the axial skeleton in relation to aging and the menopause. Scand. J. Clin. Lab. Invest. 43, 333 – 338 (1983). J. C. Gallagher, D. Goldgar, and A. Moy, Total bone calcium in normal women: Effect of age and menopause status. J. Bone Miner. Res. 2, 491 – 496 (1987). L. R. Hedlund and J. C. Gallagher, The effect of age and menopause on bone mineral density of the proximal femur. J. Bone Miner. Res. 4, 639 – 642 (1989). R. Wasnich, K. Yano, and J. Vogel, Postmenopausal bone loss at multiple skeletal sites: Relationship to estrogen use. J. Chron. Dis. 36, 781 – 790 (1983). R. Lindsay, D. M. Hart, and A. MacLean, Bone response to termination of estrogen treatment. Lancet 1, 1325 – 1327 (1978). A. Horsman, M. Simpson, P. A. Kirby, and B. E. C. Nordin, Non-linear bone loss in oophorectomized women. Br. J. Radio 50, 504 – 507 (1977). L. Nilas, J. Borg, and C. Christiansen, Different rates of loss of trabecular and cortical bone after the menopause. In “Osteoporosis” (C. Christiansen, C. D. Arnaud, B. E. C. Nordin, A. M. Parfitt, W. A. Peck, and B. L. Riggs, Eds.), pp. 161 – 163. Glostrup Hospital, Copenhagen, 1984. N. M. Keshawarz and R. R. Recker, Expansion of the medullary cavity at the expense of cortex in postmenopausal osteoporosis. Metab. Bone Dis. Relat. Res. 5, 223 – 228 (1984). L. S. Richelson, H. W. Wahner, L. J. Melton, and B. L. Riggs, Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N. Engl. J. Med. 311, 1273 – 1275 (1984). J. Dequeker and P. Geusens, Contributions of aging and estrogen deficiency to postmenopausal bone loss. N. Engl. J. Med. 313, 453 (1985). L. Nilas and C. Christiansen, Bone mass and its relationship to age and the menopause. J. Clin. Endocrinol. Metab. 65, 697 – 702. (1987). R. R. Recker, R. P. Heaney, and P. D. Saville, Menopausal changes in remodeling. J. Lab. Clin. Med. 92, 964 – 971 (1978). J. J. Stepan, J. Pospichal, J. Presl, and V. Pacovsky, Bone loss and biochemical indices of bone remodeling in surgically-induced postmenopausal women. Bone 8, 279 – 284 (1987). R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in calcium balance performance. J. Lab. Clin. Med. 92, 953 – 963 (1978). T. Coble, D. B. Kimmel, J. M. Lappe, and R. R. Recker, Association of bone markers with cancellous bone behavior in normal premanopausal women. J. Bone Miner. Res. 9 (Suppl. 1), S260[B122] (1994).

41 90. T. Steiniche, C. Hasling, P. Charles, E. F. Eriksen, L. Mosekilde, and F. Melson, A randomized study on the effects of estrogen/gestagen or high dose oral calcium on trabecular bone remodeling in postmenopausal osteoporosis. Bone 10, 313 – 320 (1989). 91. Deleted in proof. 92. C. Christiansen, M. S. Christensen, P. McNair, C. Hagen, K. E. Stocklund, and I. Transbol, Prevention of early postmenopausal bone loss: Controlled 2-year study in 315 normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). 93. S. Meema and H. E. Meema, Menopausal bone loss and estrogen replacement. Isr. J. Med. Sci. 12, 601 – 606 (1976). 94. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Int. Med. 102, 319 – 324 (1985). 95. R. Lindsay, D. M. Hart, and J. M. Aitken, Long-term prevention of osteoporosis by estrogen. Lancet 1, 1038 – 1041 (1976). 96. Deleted in proof. 97. C. Christiansen and R. Lindsay, Estrogens, bone loss and preservation. Osteoporosis Int. 1, 7 – 13 (1990). 98. R. T. Turner, J. J. Vandersteenhover, and N. H. Bell, The effects of ovariectomy and 17-b-estradiol on cortical bone histomorphometry in growing rats. J. Bone Miner. Res. 2, 115 – 122 (1987). 99. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 100. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 101. G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedus, and I. Transbol, Epidemiology of postmenopausal spinal and long bone fractures. Clin. Orthop. Relat. Res. 166, 75 – 81 (1982). 102. C. C. Johnston, C. W. Slemenda, and L. J. Melton, Clinical use of bone densitometry. N. Engl. J. Med. 324, 1105 – 1109 (1991). 103. J. A. Kanis, L. J. Melton, C. Christiansen, C. C. Johnston, and N. Khalttaev, The diagnosis of osteoporosis. J. Bone Miner. Res. 9, 1137 – 1141 (1994). 104. B. E. Nilsson and N. E. Westlin, Bone mineral content and fragility fractures. Clin. Orthop. Rel. Res. 125, 196 – 199 (1977). 105. J. F. Aloia, A. Vaswani, D. McGowan, and P. Ross, Preferential osteopenia in women with osteoporotic fractures. Bone Miner. 18, 51 – 63 (1992). 106. I. Eventon, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodeling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 11, 1 – 6 (1991). 107. M. Matsushita, T. Tsyboyama, R. Kasai, H. Okumura, T. Yamamuro, K. Higuchi, K. Higuchi, A. Coin, T. Yonezu, A. Utani, M. Umezawa, and T. Takeda, Age-related changes in bone mass in the senescenceaccelerated mouse (SAM). Am. J. Pathol. 125, 276 – 283 (1986). 108. O. Schaadt and H. Bohr, different trends of age-related diminution of bone mineral content in the lumbar spine, femoral neck, and femoral shaft in women. Calcif. Tissue Int. 42, 71 – 76 (1988). 109. B. L. Riggs and L. J. Melton III, Heterogeneity of involutional osteoporosis: Evidence for two osteoporosis syndromes. Am. J. Med. 75, 899 – 901 (1983). 110. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers and D. E. McMillan, Effect of prolonged bedrest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 111. P. Minaire, P. J. Meunier, C. Edouard, J. Bernard, P. Courpron, and J. Bourret, Quantitative histologic data on disuse osteoporosis. Calcif. Tissue Res. 17, 57 – 73 (1974). 112. B. Krolner and B. Toft, Vertebral bone loss: An unheeded side effect of therapeutic bed rest. Clin. Sci. 64, 537 – 540 (1983). 113. W. Bloom, M. A. Bloom, and F. C. McLean, Calcification and ossification: Medullary bone changes in the reproductive cycle of female pigeons. Anat. Rec. 79, 443 – 466 (1941).

42 114. S. C. Miller and B. M. Bowman, Medullary bone osteogenesis following estrogen administration to mature male Japanese quail. Dev. Biol. 87, 52 – 63 (1981). 115. C. T. Rubin and L. E. Lanyon, Regulation of bone formation by applied dynamic loads. J. Bone Jt. Surg. 66A, 397 – 402 (1984). 116. C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 117. J. J. Matsuda, R. K. Zernicke, A. C. Vailas, V. A. Pedvini,A. PedriniMille, and J. A. Maynard, Structural and mechanical adaptation of immature bone to strenuous exercise. J. Appl. Physiol. 60, 2028–2034 (1986). 118. L. E. Lanyon, C. T. Rubin, and G. Baust, Modulation of bone loss during calcium insufficiency by controlled dynamic loading. Calcif. Tissue Int. 38, 209 – 216 (1986). 119. A. J. Kahn and D. J. Simmons, Investigation of cell lineage in bone using a chimera of chick and quail embyronic tissue. Nature 258, 325 – 326 (1975). 120. F. V. Jotereau and N. M. Ledouarin, The developmental relationship between osteocytes and osteoclasts: A study using the quail-chick nuclear marker in endochondral ossification. Dev. Biol. 63, 263 – 265 (1978). 121. S. C. Miller, Osteoclast cell-surface changes during the egg-laying cycle in Japanese quail. J. Cell Biol. 75, 104 – 118 (1977). 122. K. M. Attie, N. R. Ramirez, F. A. Conte, S. L. Kaplan, and M. M. Grumbach, The pubertal growth spurt in eight patients with true precocious puberty and growth hormone deficiency: Evidence for a direct role of sex steroids. J. Clin. Endocrinol. Metab. 71, 975 – 983 (1990). 123. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechart, H. K. Huang, E. E. Schulz, C. R. Libanati,and C. C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). 124. J. P. Bonjour, G. Theintz, B. Buchs, D. Slosman, and R. Rizzoli, Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991). 125. D. G. Walker, Congenital osteopetrosis in mice cured by parabiotic union with normal siblings. Endocrinology 91, 916 – 920 (1972). 126. D. G. Walker, Osteopetrosis cured by temporary parabiosis. Science 180, 75 (1973). 127. D. W. H. Barnes, J. F. Loutit, and J. M Sansom, Histocompatible cells for the resolution of osteopetrosis in microophthalmic mice. Proc. R. Soc. London 188, 501 – 505 (1975). 128. D. G. Walker, Bone resorption restored in osteopetrotic mice by transplants of normal bone marrow and spleen cells. Science 190, 784 – 785 (1975). 129. D. G. Walker, Spleen cells transmit osteopetrosis in mice. Science 190, 785 – 786 (1975). 130. J. E. Lennon and H. S. Micklem, Stromal cells in long-term murine bone marrow culture: FACS studies and origin of stromal cells in radiation chimeras. Exp. Hematol. 14, 287 – 292 (1986). 131. S. Kyoizumi, C. M. Baum, H. Kaneshima, J. M. McCune, E. J. Yee, and Namikawa, Implantation and maintenance of functional human bone marrow in SCI-hu mice. Blood 79, 1704, 1711 (1992). 132. N. A. Nicola, C. G. Begley, and D. Metcalf, Identification of the human analogue of a regulator that induces differentiation in murine leukemic cells. Nature 314, 625 – 627 (1985). 133. J. F. Krowka, S. Sarin, R. Namikawa, J. M. McCune, and H. Kaneshima, Human T cells in the SCID-hu mouse are phenotypically normal and functionally competent. J. Immunol. 146, 3751 – 3757 (1991). 134. C. B. Shoemaker and L. D. Mitsock, Murine erythropoietin gene: Cloning, expression, and human gene homology. Mol. Cell Biol. 6, 849 – 856 (1986). 135. J. P. Cassella, R. Pereira, J. S. Khillan, D. J. Prockop, N. Garrington, and S. Y. Ali, An ultrastructural, micoranalytical, andspectroscopic study of bone from a transgenic mouse with aCOLAL1.A1 pro-Alpha-1 mutation. Bone 15, 611 – 619 (1994).

DONALD B. KIMMEL 136. S. C. Marks and P. W. Lane, Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 67, 11 – 18 (1976). 137. R. A. Altschuler, D. F. Dolan, M. Ptok, G. Gholizadeh, J.Bonadio, and J. E. Hawkins, An evaluation of otopathology in the MOV-13 transgenic mutant mouse. Ann. NY Acad. Sci. 630, 249 – 252 (1991). 138. R. Balena, F. Costantini, M. Yamamoto, A. Markatos, R. Cortese, G. A. Rodan, and V. Poli, Mice with IL-6 gene knock-out do not lose cancellous bone after ovariectomy. J. Bone Miner. Res. 8 (Suppl. 1), S130 (1993). 139. E. S. Lander and D. Botstein, Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185 – 199 (1989). 140. H. K. Suzuki, Effects of estradiol-17-b-n-valerate on endosteal ossification and linear growth in the mouse femur. Endocrinology 60, 743 – 747 (1958). 141. M. W. Edwards, S. D. Bain, M. C. Bailey, M. M. Lantry, and G. A. Howard, 17-b-Estradiol stimulation of endosteal bone formation in the ovariectomized mouse: An animal model for the evaluation of bone-targeted estrogens. Bone 13, 29 – 34 (1992). 142. L. R. Donahue, C. J. Rosen, and W. G. Beamer, Reduced bone density in hypogonadal mice. J. Bone Miner. Res. 9 (Suppl. 1), S193 (1994). 143. J. Hashimoto, K. Takaoka, H. Yoshikawa, S. Miyamoto, S. Suzuki, and K. Ono, Preservation of ectopically-induced bone in the mouse by estradiol. Bone 12, 249 – 255 (1991). 144. B. Bar-Shira-Maymon, R. Coleman, A. Cohen, E. SteinhagenThiessen, and M. Silbermann, Age-related bone loss in lumbar vertebrae of CW-1 female mice: A histomorphometric study. Calcif. Tissue Int. 44, 36 – 45 (1989). 145. B. Bar-Shira-Maymon, R. Coleman, E. Steinhagen-Thiessen, and M. Silbermann, Correlation between alkaline and acid phosphatease activities and age-related osteopenia in murine vertebrae. Calcif. Tissue Int. 44, 99 – 207 (1989). 146. A. Weiss, I. Arbell, E. Steinhagen-Thiessen, and M. Silbermann, Structural changes in aging bone: Osteopenia in the proximal femurs of female mice. Bone 12, 165 – 172 (1991). 147. M. R. Urist, A. M. Budy, and F. C. McLean, Endosteal bone formation in estrogen-treated mice. J. Bone Jt. Surg. 32A, 143 – 162 (1950). 148. S. D. Bain, M. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 149. S. D. Bain, M. C. Bailey, and M. W. Edwards, The anabolic effect of estrogen on endosteal bone formation in the mouse is attenuated by ovariohysterectomy: A role for the uterus in the skeletal response to estrogen? Calcif Tissue Int. 51, 223 – 228 (1992). 150. S. D. Bain, E. Jensen, D. L. Celino, M. C. Bailey, M. M. Lantry, and M. W. Edwards, High-dose gestagens modulate bone resorption and formation and enhance estrogen-induced endosteal bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 291 – 230 (1993). 151. Deleted in proof. 152. T. Tsuboyama, K. Takahashi, M. Matsushita, H. Okumura, T. Yamamuro, M. Umezawa, and T. Takeda, Decreased endosteal formation during cortical bone modeling in SAM-P/6 mice with a low peak bone mass. Bone Miner. 7, 1 – 12 (1989). 153. T. Tsuboyama, M. Matsushita, H. Okumura, T. Yamamuro, K. Hanada, and T. Takeda, Modification of strain-specific femoral bone density by bone marrow chimerism in mice: A study on the spontaneously osteoporotic mouse (SAM-P/6). Bone 10, 269 – 277 (1989). 154. K. Takahashi, T. Tsuboyama, M. Matsushita, R. Kasai, H. Okumura, T. Yamamuro, Y. Okamoto, K. Toriyama, K. Kitagawa, and T. Takeda, Modification of strain-specific femoral bone density by

CHAPTER 37 Animal Models for in Vivo Experimentation

155.

156.

157.

158.

159.

160. 161.

162.

163. 164. 165.

166.

167.

168.

169.

170.

171.

172.

173. 174.

bone marrow-derived factors administered neonatally: A study on the spontaneously osteoporotic mouse, SAMP6. Bone Miner. 24, 245 – 255 (1994). T. Tsuboyama, K. Takahashi, T. Tamamuro, M. Hosokawa, and T. Takeda, Cross-mating study on bone mass in the spontaneously osteoporotic mouse (SAM-P/6) Bone Miner. 23, 57 – 64 (1993). K. Takahashi, T. Tsuboyama, M. Matsushita, R. Kasai, H. Okumura, T. Yamamuro, Y. Okamoto, K. Kitagawa, and T. Takeda, Effective intervention of low peak bone mass and bone modeling in the spontaneous murine model of senile osteoporosis, SAM-P/6, by calcium supplement and hormone treatment. Bone 15, 209 – 215 (1994). L. D. Spotila, C. D. Constantinou, L. Sereda, A. Ganguly, B. L. Riggs, and D. J. Prockop, Mutation in a gene for type I procollagen (COL1A2) in a woman with postmenopausal osteoporosis: Evidence for phenotypic and genotypic overlap withmild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423 – 5427. (1991). M. Ringwald, R. Baldock, J. Bard, M. Kaufman, J. T. Eppig, J. E. Richardson, J. H. Nadeau, and D. Davidson, A databasefor mouse development. Science 265, 2033 – 2034 (1994). J. S. Arnold and W. S. S. Jee, Bone growth and osteoclastic activity as indicated by radioautographic distribution of 239Pu. Am. J. Anat. 101, 367 – 417 (1957). G. Gothlin and J. L. E. Ericsson, On the histogenesis of the cells in fracture callus. Virchows Arch. Abt. B Zellpath. 12, 318 – 329 (1973). H. Andersen and M. E. Matthiessen, The histiocyte in human fetal tissues: Its morphology, cytochemistry, origin, function, and fate. Z. Zellforsch. 72, 193 – 211 (1966). S. C. Marks and G. B. Schneider, Evidence for a relationship between lymphoid cells and osteoclasts: Bone resorption restored in ia (osteopetrotic) rats by lymphocytes, monocytesand macrophages from a normal littermate. Am. J. Anat. 152, 331 – 342 (1978). K. Buring, On the origin of cells in heterotopic bone formation. Clin. Orthop. Relat. Res. 110, 293 – 302 (1973). A. B. Dawson, The age order of epiphyseal union in the long bones of the albino rat. Anat. Rec. 31, 1 – 17 (1925). R. M. Acheson, M. N. MacIntyre, and E. Oldham, Techniques in longitudinal studies of the skeletal development of the rat. Br. J. Nutr. 13, 283 – 296 (1959). E. E. Joss, E. H. Sobel, and K. A. Zuppinger, Skeletal maturation in rats with special reference to order and time of epiphyseal closure. Endocrinology 72, 117 – 122 (1963). C. Spark and A. B. Dawson, The order and appearance of centers of ossification in the fore and hind limbs of the albino rat, with reference to the possible influence of the sex factor. Am. J. Anat. 41, 411 – 445 (1928). H. E. Swanson and J. J. van der Werff Ten Bosch, Sex differences in growth of rats, and their modification by a single injection of testosterone propionate shortly after birth. J. Endocrinol. 26, 197 –207 (1963). R. T. Turner, K. S. Hannon, L. M. Demers, J. Buchanan, and N. H. Bell, Differential effects of gonadal function on bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). D. B. Kimmel, Quantitative histologic changes in the proximal tibial epiphyseal growth cartilage of aged female rats. Cells Mater. 1 (Suppl.), 11 – 18 (1992). D. Schapira, R. Laton-Miller, D. Barzilai, and M. Silbermann, The rat as a model for studies of the aging skeleton. Cells Mater. (Suppl.) 1, 181 – 188 (1992). X. J. Li, W. S. S. Jee, H. Z. Ke, S. Mori, and T. Akamine, Age related changes of cancellous and cortical bone histomorphometry in female Sprague – Dawley rats. Cells Mater. (Suppl.) 1, 25 – 37 (1992). H. M. Frost and W. S. S. Jee, On the rat model of human osteopenias and osteoporoses. Bone Miner. 18, 227 – 236 (1992). R. L. Butcher, W. E. Collins, and N. W. Fugo, Plasma concentration of LH, FSH, prolactin, progesterone, and estradiol-17-b throughout the four day cycle of the rat. J. Endocrinol. 94, 1704 – 1708 (1974).

43 175. K. H. Lu, B. R. Hopper, T. M. Vargo, and S. S. C. Yen, Chronological changes in sex steroid, gonadotropin, and prolactin secretion in againg female rats displaying different reproductive states. Biol. Reprod. 21, 193 – 203 (1979). 176. D. N. Kalu, The ovariectomized rat as a model of postmenopausal osteopenia. Bone Miner. 15, 175 – 191 (1991). 177. T. J. Wronski, The ovariectomized rat as an animal model for postmenopausal bone loss. Cells Mater. (Suppl. 1)1, 69 – 74 (1992). 178. T. J. Wronski, L. M. Dann, and S. L. Horner, Time course of vertebral osteopenia in ovariectomized rats. Bone 10, 295 – 301 (1989). 179. D. B. Kimmel and T. J. Wronski, Non-destructive measurement of bone mineral in femurs from ovariectomized rats. Calcif. Tissue Int. 46, 101 – 110 (1990). 180. J. G. Seedor, H. A. Quartuccio, and D. D. Thompson, The bisphosphonate alendronate (MK-217) inhibits bone loss due to ovariectomy in rats. J. Bone Miner. Res. 6, 339 – 346 (1991). 181. Li Mosekilde, C. C. Danielsen, and U. B. Knudsen, The effect of aging and ovariectomy on the vertebral bone mass and biomechanical properties of mature rats. Bone 14, 1 – 6 (1993). 182. T. J. Wronski, C. C. Walsh, and L. A. Ignaszewski, Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone 7, 119 – 124 (1986). 183. D. Black, C. Farquharson, and S. P. Robins, Excretion of pyridinium cross-links of collagen in ovariectomized rats as urinary markers for increased bone resorption. Calcif. Tissue Int. 44, 343 – 347 (1989). 184. H. Ito, H. Z. Ke, W. S. S. Jee, and T. Sakou, Anabolic responses of an adult cancellous bone site to prostaglandin E2 in the rat. Bone Miner. 21, 219 – 236 (1993). 185. S. R. Cummings, Epidemiologic studies of osteoporotic fractures: Methodologic issues. Calcif. Tissue Int. 49 (Suppl.), S15 – S20 (1991). 186. P. Amman, R. Rizzoli, D. Slosman, and J. P. Bonjour, Sequential and precise in vivo measurement of bone mineral density in rats using dual energy x-ray absorptiometry. J. Bone Miner. Res. 7, 311 – 317 (1992). 187. M. G. Griffin, R. Kimble, W. Hopfer, and R. Pacifici, DXA of the rat: Accuracy, precision, and measurement of bone loss. J. Bone Miner. Res. 8, 795 – 800 (1993). 188. B. H. Mitlak, D. Schoenfeld, and R. M. Neer, Accuracy, precision, and utility of spine and whole-skeleton mineral measurements by DXA in rats. J. Bone Miner. Res. 9, 119 – 126 (1994). 189. D. N. Kalu, C. C. Liu, E. Salerno, B. W. Hollis, R. Echon, and M. Ray, Skeletal response of ovariectomized rats to low and high doses of 17-b-estradiol. Bone Miner. 14, 175 – 187 (1991). 190. T. J. Wronski, C. F. Yen, and K. S. Scott, Estrogen and diphosphonate treatment provide long-term protection against osteopenia in ovariectomized rats. J. Bone Miner. Res. 6, 387 – 394 (1991). 191. P. T. Beall, L. K. Misra, R. L. Young, H. J. Spjut, H. J Evans, and A. LeBlanc, Clomiphene protects against osteoporosis in the mature ovariectomized rat. Calcif. Tissue Int. 36, 123 – 125 (1984). 192. S. C. Garner, J. J. B. Anderson, M. H. Mar, and I. Parikh, Estrogens reduce bone loss in the ovariectomized, lactating rat model. Bone Miner. 15, 19 – 31 (1992). 193. R. Schenk, W. A. Merz, H. A. Fleisch, R. C. Muhlbauer, and R. G. G. Russell, Effects of ethane-1-hydroxy-1,1-diphosphonate (EHDP) and dichloromethylene diphosphonate (C12MDP) on the calcification and resorption of cartilage and bone in thetibial epiphysis and metaphysis of rats. Calcif. Tissue Res. 11, 196 – 214 (1973). 194. R. Schenk, P. Eggli, H. Fleisch, and S. Rosini, Quantitative morphometric evaluation of the inhibitory activity of new aminobisphosphonates on bone resorption in the rat. Calcif. Tissue Int. 38, 342 – 349 (1986). 195. K. Nakatsuka, Y. Nishizawa, S. Hagiware, H. Koyama, T. Miki, H. Ochi, and H. Morii, Effect of calcitonin on total body bone mineral contents of experimental osteoporotic rats determined by dual photon absorptiometry. Calcif. Tissue Int. 47, 378 – 382 (1990).

44 196. G. F., Mazzuoli, S. Tabolli, F. Bigi, C. Valtorta, S. Minisola, D. Diacinti, L. Scarnecchia, G. Bianchi, M. Piolini, and S. Dell’Acqua, Effects of salmon clacitonin on the bone loss induced by ovariectomy. Calcif. Tissue Int. 47, 209 – 214 (1990). 197. S. T. Harris, B. J. Gertz, H. K. Genant, D. R. Eyre, T. T. Survill, J. N. Ventura, J. DeBrock, E. Ricerca, and C. H. Chesnut, The effect of short term treatment with alendronate on vertebral density and biochemical markers of bone remodeling in early postmenopausal women. J. Clin. Endocr. Metab. 76, 1399 – 1406 (1993). 198. N. B. Watts, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Miller, R. D. Jackson, A. A. Licata, P. Ross, G. C. Woodson, M. J. Yanover, W. J. Mysiw, L. Kohse, M. B. Rao, P. Steiger, B. Richmond, and C. H. Chesnut, Intermittent cyclical etidronate treatment of postmenopausal osteoporosis. N. Engl. J. Med. 323, 73 – 79 (1990). 199. G. A. Fromm, E. Vega, L. Plantalech, A. M. Galich, and C. A. Mautalen, Differential action of pamidronate on trabecular and cortical bone in women with involutional osteoporosis. Calcif Tissue Int. 49, 129 – 133 (1991). 200. M. L. Smith, I. Fogelman, D. M. Hart, E. Scott, J. Bevan, and I. Leggate, Effect of etidronate disodium on bone turnover following surgical menopause. Calcif. Tissue Int. 44, 74 – 79 (1989). 201. J. Y. Reginster, R. Deroisy, D. Denis, J. Collette, M. P. Lecart, N. Sarlet, D. Ethgen, and P. Franchimont, Prevention of postmenopausal bone loss by tiludronate. Lancet 2, 1469 – 1470 (1989). 202. I. MacIntyre, M. I. Whitehead, L. M. Banks, J. C. Stevenson, S. J. Wimalawansa, and M. J. R. Healy, Calcitonin for prevention of postmenopausal bone loss. Lancet 1, 900 – 901 (1988). 203. J. Szucs, C. Horvath, E. Kollin, M. Szathmari, and I. Hollo, Threeyear calcitonin combination therapy for postmenopausal osteoporosis with crush fractures of the spine. Calcif. Tissue Int. 50, 7 – 10 (1992). 204. J. M. Lean, J. W. M. Chow, and T. J. Chambers, Estrogen induces bone formation on non-resorptive surfaces in the rat. Bone 14, 297 – 302 (1993). 205. J. W. M. Chow, J. M. Lean, and T. J. Chambers, 17-b-Estradiol stimulates cancellous bone formation in female rats. Endocrinology 130, 3025 – 3032 (1992). 206. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocrinol. Rev. 15, 275 – 300 (1994). 207. J. L. Ferretti, R. D. Tessaro, C. J. Delgado, C. E. Bozzini, R. M. Alippi, and A. C. Barelo, Biomechanical performance of diaphyseal shafts and bone tissue of femurs from protein-restricted rats. Bone Miner. 4, 329 – 339 (1988). 208. C. H. Sogaard, C. C. Danielsen, E. B. Thorling, and Li Mosekilde, Long-term exercise of young and adult female rats: Effect on femoral neck biomechanical competence and bone structure. J. Bone Miner. Res. 9, 409 – 416 (1994). 209. Z. Peng, J. Tuukkanen, and H. K. Vaananen, Exercise can provide protection against bone loss and prevent the decrease in mechanical strength of femoral neck in ovariectomized rats. J. Bone Miner. Res. 9, 1559 – 1564 (1994). 210. T. Okano, M. P. Akhter, T. Howard, and D. B. Kimmel, Estrogen effects on bone mass, bone strength, and bone turnover in aged ovariectomized rats. J. Bone Miner. Res. 10 (Suppl. 1), S456, (T407) (1995). 211. Deleted in proof. 212. Deleted in proof. 213. Deleted in proof. 214. L. E. Nachtigall, R. H. Nachtigall, and R. D. Nachtigall, Estrogen replacement therapy. I. A 10-year prospective study in the relationship to osteoporosis. Obstet. Gynecol. 53, 277 – 281 (1979). 215. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 216. J. A. Svesatikoglou and S E. Larsson, Changes in composition and metabolic activity of the skeletal parts of the extremity of the adult

DONALD B. KIMMEL

217.

218.

219.

220. 221. 222.

223. 224.

225.

226.

227. 228.

229.

230. 231.

232. 233.

234.

235.

236.

237.

238.

rat following below-knee amputation. Acta Chir. Scand. (Suppl. 476), 9 – 15 (1976). V. T. Thomaidis and T. S. Lindholm, The effect of remobilization on the extremity of the adult rat after short-term immobilization in a plaster cast. Acta Chir. Scand. (Suppl. 476), 36 – 39 (1976). D. D. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner. Res. 3, 409 – 414 (1988). X. J. Li, W. S. S. Jee, S. Y. Chow, and D. M. Woodbury, Adaptation of cancellous bone to aging and immobilization in the adult rat: A single photon absorptiometry and histomorphometry study. Anat. Rec. 227, 12 – 24 (1990). J. U. Lindgren and S. Mattson, The reversibility of disuse osteoporosis. Calcif. Tissue Res. 23, 179 – 184 (1977). H. Maeda, D. B. Kimmel, N. Lane, and D. Raab, The musculoskeletal response to immobilization and recovery. Bone 14, 153 – 159 (1993). N. Lane, H. Maeda, D. M. Cullen, and D. B. Kimmel, Cancellous bone behavior in hindlimb immobilized rats during and after naproxen treatment. Bone Miner. 26, 43 – 59. (1994). S. Mattson, The reversibility of disuse osteoporosis. Acta Orthop. Scand. (Suppl. 144), 59 – 88 (1972). R. H. Young and W. A. J. Crane, Effect of hydrocortisone on the utilization of tritiated thymidine for skeletal growth in the rat. Ann. Rheum. Dis. 23, 163 – 167 (1964). D. J. Simmons and A. S. Kunin, Autoradiographic and biochemical investigations of the effect of cortisone on the bones of the rat. Clin. Orthop. Rel. Res. 55, 201 – 215 (1967). J. L. Ferretti, S. O. Vazquez, C. J. Delgado, R. Capozza, and G. Cointry, Biphasic dose-response curves of cortisol effects on rat diaphyseal bone biomechanics. Calcif. Tissue Int. 50, 49 – 54 (1992). A. Vignery and R. Baron, Dynamic histomorphometry of alveolar bone remodeling in the adult rat. Anat. Rec. 196, 191 – 200 (1980). R. Baron, R. Tross, and A. Vignery, Evidence of sequential modeling in rat trabecular bone: Morphology, dynamic histomorphometry, and changes during skeletal maturation. Anat. Rec. 208, 137 – 145 (1984). W. S. S. Jee, S. Mori, X. J. Li, and S. Chan, Prostaglandin E2 enhances cortical bone mass and activates intracortical bone remodeling in intact and ovariectomized female rats. Bone 11, 253 – 266 (1990). E. Ruth, An experimental study of the Haversian-type vascular channels. Anat. Rec. 112, 429 – 455 (1953). D. Baylink, M. Stauffer, J. Wergedal, and C. Rich, Formation, mineralization, and resorption of bone in vitamin D-deficient rats. J. Clin. Invest. 49, 1122 – 1134 (1970). E. R. Morey and D. J. Baylink, Inhibition of bone formation during spaceflight. Science 201, 1138 – 1141 (1978). T. J. Wronski, P. A. Schenck, M. Cintron, and C. C. Walsh, Effect of body weight on osteopenia in ovariectomized rats. Calcif. Tissue Int. 43, 179 – 183 (1987). R. E. Roudebush, D. E. Magee, D. N. Benslay, A. M. Bendele, and H. U. Bryant, Effect of weight manipulation on bone loss due to ovariectomy and the protective effects of estrogen in the rat. Calcif. Tissue Int. 53, 61 – 64 (1993). F. R. deWinter and R. Steendijk, The effect of a low-calcium diet in lactating rats: Observations on the rapid development and repair of osteoporosis. Calcif. Tissue Res. 17, 303 – 316 (1975). S. C. Miller, J. G. Shupe, E. H. Redd, M. A. Miller, and T. H Omura, Changes in bone mineral and bone formation rates during pregnancy and lactation in rats. Bone 7, 283 – 287 (1986). N. Binkley and D. Kimmel, Effect of age and parity on skeletal response to ovariectomy in rats. J. Bone Miner. Res. 9 (Suppl. 1), S197 (1994). H. Duncan, C. A Hanson, and A. Curtiss, The different effects of soluble and crystalline hydrocortisone on bone. Calcif. Tissue Res. 12, 159 – 168 (1973).

CHAPTER 37 Animal Models for in Vivo Experimentation 239. D. Vanderschueren, E. Van Herck, A. M. Suiker, K. Allewaert, W. J. Visser, P. Geusens, and R. Bouillon, Bone and mineral metabolism in the adult guinea pig: Long-term effects of estrogen and androgen deficiency. J. Bone Miner. Res. 7, 1407 – 1415 (1992). 240. J. Jowsey and J. Gershon-Cohen, Effect of dietary calcium levels on production and reversal of experimental osteoporosis in cats. Proc. Soc. Exp. Biol. Med. 116, 437 – 441 (1964). 241. S. Jett, K. Wu, H. Duncan, and H. M. Frost, Adrenalcorticosteroid and salicylate actions on human and canine Haversian bone formation and resorption. Clin. Orthop. Relat. Res. 68, 310 – 315 (1970). 242. M. S. Mackey, M. L. Stevens, X. J. Li, D. L Tressler, K. S. Combs, and J. E. McOsker, The ferret: A small animal model with BMUbased remodeling for skeletal research. J. Bone Miner. Res. 9 (Suppl. 1), S319 (1994). 243. X. J. Li, M. S. Mackey, M. L. Stevens, D. L. Tressler, K. S. Combs, and J. E. McOsker, The ferret: Skeletal response to estrogen deficiency. J. Bone Miner. Res. 9 (Suppl.1), S397 (1994). 244. X. J. Li, M. L. Stevens, M. S. Mackey, K. S. Combs, D. L. Tressler, and J. E. McOsker, The ferret: Skeletal responses to treatment with PTH. J. Bone Miner. Res. 9 (Suppl. 1), S258 (1994). 245. A. M. Parfitt, The actions of PTH on bone — Relation to bone remodeling and turnover, calcium homeostasis, and metabolic bone disease, PTH and osteoblast relationships between bone turnover and bone loss, and the state of bones in primary hyperparathyroidism. Metabolism 25, 1033 – 1069 (1976). 246. W. S. S. Jee and P. D. Nolan, Origin of osteoclasts from the fusion of phagocytes. Nature 200, 225 – 226 (1963). 247. J. K. Gong, J. S. Arnold, and S. H. Cohn, Composition of trabecular and cortical bone. Anat. Rec. 149, 325 – 331 (1964). 248. L. C. Johnson, Composition of trabecular and cortical bone in humans. In “Bone Biodynamics” (H. M. Frost, Ed.), pp. 543 – 654. Little, Brown, Boston, 1964. 249. N. J. Parks, W. S. S. Jee, R. B. Dell, and G. E. Miller, Assessment of cortical and trabecular bone distribution in the beagle skeleton by neutron activation analysis. Anat. Rec. 215, 230 – 250 (1986). 250. H. M. Frost, Tetracycline-based histological analysis of bone remodeling. Calcif. Tissue Res. 3, 211 – 239 (1969). 251. D. B. Kimmel and W. S. S. Jee, A quantitative histologic study of bone turnover in young adult beagles. Anat. Rec. 203, 35 – 51 (1982). 252. C. Bressot, P. J. Meunier, M. C. Chapuy, E. Lejeune, C. Edouard, and A. J. Darby, Histomorphometric profile, pathophysiology and reversibility of corticosteroid-induced osteoporosis. Metab. Bone Dis. Relat. Res. 1, 303 – 311 (1979). 253. E. Ritz, B. Krempien, O. Mehls, and H. H. Malluche, Skeletal abnormalities in chronic renal insufficiency before and during maintenance hemodialysis. Kidney Int. 4, 116 – 127 (1973). 254. H. H. Malluche, M. C. Faugere, R. M. Friedler, C. Matthews, and P. Fanti, Calcitriol, parathyroid hormone, and accumulation of aluminum in dogs with renal failure. J. Clin. Inv. 79, 754 – 761 (1987). 255. P. D. Delmas, S. Charhon, M. C. Chapuy, E. Vignon, D. Briancon, C. Edouard, and P. J. Meunier, Long-term effect of dichloromethylene disphosphonate on skeletal lesions in multiple myeloma. Metab. Bone Dis. Relat. Res. 4, 163 – 168 (1982). 256. L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. In “Bone Histomorphometry III” (W. S. S. Jee and A. M. Partiff, Eds.), pp. 389 – 407. Armour Montagu, Paris, 1981. 257. H. K. Uhthoff, G. Sekaly, and Z. F. G. Jaworski, Effect of long-term nontraumatic immobilization on metaphyseal spongiosa in young adult and old beagle dogs. Clin. Orthop. Relat. Res. 192, 279 – 283 (1985). 258. H. K. Uhthoff and Z. F. G. Jaworski, Bone loss in response to longterm immobilisation. J. Bone Jt. Surg. 60B, 420 – 429 (1978). 259. Z. F. G. Jaworski, M. Liskova-Kiar, and H. K Uhthoff, Effect of long-term immobilisation on the pattern of bone loss in older dogs. J. Bone Jt. Surg. 62B, 104 – 110 (1980).

45 260. Z. F. G. Jaworski and H. K. Uhthoff, Reversibility of nontraumatic disuse osteoporosis during its active phase. Bone 7, 431 – 439 (1986). 261. R. D. Podbesek, C. Edouard, P. J. Meunier, J. A. Parsons, J. Reeve, R. W. Stevenson, and J. M Zanelli, Effect of two treatment regimes with synthetic human parathyroid hormone fragment on bone formation and the tissue balance of trabecular bone in greyhounds. Endocrinology 112, 1000 – 1006 (1983). 262. D. B. Kimmel, The oophorectomized beagle as an experimental model for estrogen-depletion bone loss in the adult human. Cells Mater. (Suppl.) 1, 75 – 84 (1992). 263. R. G. Cumming, Calcium intake and bone mass: A quantitative review of the evidence. Calcif. Tissue. Int. 47, 194 – 201 (1990). 264. M. K. Drezner and T. Nesbitt, Role of calcitriol in prevention of osteoporosis: Part I. Metabolism 39 (Suppl. 1), 18 – 23 (1990). 265. A. Barbier, J. P. Bonjour, P. Geusens, M. C. deVernejoul, and F. Lacheretz, Tiludronate: A bisphosphonate with a positive effect on bone quality in experimental models. J. Bone Min. Res. 6 (Suppl. 1), S217 (1991). 266. J. A. Shaw, S. C. Wilson, E. M. Paul, and A. Bruno, Vertebral trabecular bone volume in dogs: Comparison of oophorectomized females, sham females, and males. Trans. Orthop. Res. Soc. 17(135) (1992). 267. W. R. Lynch, S. A. Goldstein, A. M. Schulick, and C. A. Mase, The effects of calcitonin on torsional properties of the long bones of ovariohysterectomized beagles. Trans. 37th Annu. Meeting Orthop. Res. Soc. 16, 430 (1991). [Abstract] 268. D. A. McCubbrey, R. W. Goulet, S. A. Goldstein, and A. M. Shulick, Vertebral body architecture and mechancial properties alterations due to ovariectomy in the beagle. Trans. 37th Annu. Meeting Orthop. Res. Soc. 16, 284 (1991). [Abstract] 269. R. W. Boyce, A. F. Franks, M. L. Jankowsky, C. M. Orcutt, A. M. Piacquadio, J. M. White, and J. M. Bevan, Sequential histomorphometric changes in cancellous bone from ovariohysterectomized dogs. J. Bone Miner. Res. 5, 947 – 953 (1990). 270. H. H. Malluche, M. C. Faugere, M. Rush, and R. M. Friedler, Osteoblastic insufficiency is responsible for maintenance of osteopenia after loss of ovarian function in experimental beagle dogs. Endocrinology 119, 2649 – 2655 (1986). 271. M. C. Faugere, P. Fanti, R. M. Friedler, and H. H. Malluche, Bone changes occurring early after cessation of ovarian function in beagle dogs: A histomorphometric study employing sequential biopsies. J. Bone Miner. Res. 5, 263 – 272 (1990). 272. H. H. Malluche, M. C. Faugere, R. M. Friedler, and P. Fanti, 1,25Dihydroxyvitamin D3 corrects bone loss but suppresses bone remodeling in ovariohysterectomized beagle dogs. Endocrinology 122, 1988 – 2006 (1988). 273. T. Nakamura, Y. Nagai, H. Yamato, K. Suzuki, and H. Orimo, Regulation of bone turnover and prevention of bone atrophyin ovariectomized beagle dogs by the adminsitration of 24R,25(OH)2D3. Calcif. Tissue Int. 50, 221 – 227 (1992). 274. G. A. Dannucci, R. B. Martin, and P. Patterson-Buckendahl, Ovariectomy and trabecular bone remodeling in the dog. Calcif. Tissue Int. 40, 194 – 199 (1987). 275. K. K. Karambolova, G. R. Snow, and C. Anderson, Differences in periosteal and corticoendosteal bone envelope activities in spayed and intact beagles: A histomorphometric study. Calcif. Tissue Int. 41, 665 – 668 (1985). 276. V. Shen, D. W. Dempster, R. Birchman, R. W. E. Mellish, E. Church, D. Kohn, and R. Lindsay, Lack of changes in histomorphometric, bone mass, and biochemical parameters in ovariohysterectomized dogs. Bone 13, 311 – 316 (1992). 277. K. K. Karambolova, G. R. Snow, and C. Anderson, Surface activity on the periosteal and corticoendosteal envelopes following continuous progestogen supplementation in spayed beagles. Calcif. Tissue Int. 38, 239 – 243 (1986).

46 278. K. K. Karambolova, G. R. Snow, and C. Anderson, Effects of continuous 17-b-estradiol administration on the periosteal and corticoendosteal envelope activity in spayed beagles. Calcif. Tissue Int. 40, 12 – 15 (1987). 279. G. R. Snow and C. Anderson, The effects of 17-b-estradiol and progestagen on trabecular bone remodeling in oophorectomized dogs. Calcif. Tissue. Int. 39, 198 – 205 (1986). 280. G. R. Snow and C. Anderson, The effects of continuous estradiol therapy on cortical bone remodeling activity in the spayed beagle. Calcif. Tissue Int. 37, 437 – 440 (1985). 281. G. R. Snow, M. A. Cook, and C. Anderson, Oophorectomy and cortical bone remodeling in the beagle. Calcif. Tissue Int. 36, 586 – 590 (1984). 282. G. R. Snow and C. Anderson, The effects of continuous progestogen treatment on cortical bone remodeling activity in beagles. Calcif. Tissue Int. 37, 282 – 286 (1985). 283. T. Koyama, H. Uno, A. Nagaami, T. Makita, T. Konno, and H. Takahashi, Effects of active vitamin D3 derivatives on bone remodeling in ovariectomized beagle dogs — Two year study — II Histomorphometric analysis of trabecular bone. Jpn. J. Bone Miner. 2, 268 – 275 (1984). 284. T. Koyama, A. Nagaami, T. Makita, T. Konno, H. Akazawa, and H. Takahashi, Effect of 1-,24(R)-dihydroxyvitamin D3 on trabecular bone remodeling in adult beagle dogs. Jpn. J. Bone Miner. 2, 249 – 257 (1984). 285. C. C. Shipp, P. S. Berger, M. S. Deehr, and B. Dawson-Hughes, Precision of dual photon absorptiometry. Calcif. Tissue Int. 42, 287 – 282 (1988). 286. P. W. Concannon, W. Hansel, and W. J. Visek, The ovarian cycle of the bitch: Plasma estrogen, LH and progesterone. Biol. Reprod. 13, 112 – 121 (1975). 287. G. E. Jones, A. R. Boyns, E. H. D. Cameron, E. T. Bell,D. W. Christie, and M. F. Parkes, Plasma oestradiol, luteinizing hormone and progesterone during the oestrous cycle in the beagle bitch. J. Endocrinol. 57, 331 – 332 (1973). 288. E. T. Bell, D. W. Christie, and E. V. Younglai, Plasma oestrogen levels during the canine oestrous cycle. J. Endocrinol. 51, 225 – 226 (1971). 289. D. T. Baird and A. Guevara, Concentration of unconjugated estrone and estradiol in peripheral plasma in nonpregnant women throughout the menstrual cycle, castrate and postmenopausal women and in men. J. Clin. Endocrinol. 29, 149 – 156 (1969). 290. M. J. Reed, P. A. Beranek, and V. H. T. James, Oestrogen production and metabolism in perimenopausal women. Maturitas 8, 29 – 34 (1986). 291. C. Longcope, L. Hoberg, S. Steuterman, and D. Baran, The effect of ovariectomy on spine bone mineral density in rhesus monkeys. Bone 10, 341 – 344 (1989). 292. Li Mosekilde, S. E. Weisbrode, J. A. Safron, H. F. Stills, M. L. Jankowsky, D. C. Ebert, C. C. Danielsen, C. H. Sogaard, A. F. Franks, M. L. Stevens, C. L. Paddock, and R. W. Boyce, Calcium-restricted ovariectomized Sinclair S-1 minipigs: An animal model of osteopenia and trabecular plate perforation. Bone 14, 379 – 382 (1993). 293. M. L. Stevens, N. Sacco-Gibson, K. S. Combs, C. Lowry, and X. J. Li, Evaluation of skeletal parameters in the Sinclair S-1 minipig: Histomorphometric assessment of skeletal changes at 3 months postovariectomy. J. Bone Miner. Res. 9(Suppl. 1), S258 (1994). 294. A. F. Franks, W. E. Wyder, X. J. Li, M. W. Lundy, M. L. Pfennig, N. Sacco-Gibson, R. M. Roberts, W. F. Trout, W. R. Myers, and G. A. Bouchard, Evaluation of skeletal parameters in the Sinclair S-1 minipig: Longitudinal analysis of ovariectomy-induced skeletal changes. J. Bone Miner. Res. 9 (Suppl. 1), S191 (1994). 295. J. Kragstrup, A. Richards, and O. Fejerskov, Effects of fluoride on cortical bone remodeling in the growing domestic pig. Bone 10, 421 – 424 (1989).

DONALD B. KIMMEL 296. D. M. Raab, T. Crenshaw, D. B. Kimmel, and E. L. Smith, Cortical bone response in adult swine after exercise. J. Bone Miner. Res. 6, 741 – 749 (1991). 297. R. Webb, G. Baxter, D. McBride, M. Ritchie, and A. J. Springbett, Mechanism controlling ovulation rate in ewes in relation to seasonal anoestrus. J. Reprod. Fertil. 94, 141 – 151 (1992). 298. B. Malpaux and F. J. Karsch, A role for short days in sustaining seasonal reproductive activity in the ewe. J. Reprod. Fertil. 90, 555 – 562 (1990). 299. E. F. Eriksen, L. Mosekilde, and F. Melsen, Effect of sodium fluoride, calcium, phosphate, and vitamin D2 on trabecular bone balance and remodeling in osteoporotics. Bone 6, 381 – 389 (1985). 300. P. Chavassieux, P. Pastoureau, G. Boivin, M. C. Chapuy, P. D. Delmas, and P. J. Meunier, Dose effects on ewe bone remodeling of short-term sodium fluoride administration — A histomorphometric and biochemical study. Bone 12, 421 – 427 (1991). 301. P. Chavassieux, P. Pastoureau, G. Boivin, M. C. Chapuy, P. D. Delmas, G. Milhaud, and P. J. Meunier, Fluoride-induced bone changes in lambs during and after exposure to sodium fluoride. Osteoporosis Int. 2, 26 – 33 (1991). 302. S. B. Hornby, S. L. Ford, N. J. Timpson, A. K. Ancill, and G. P. Evans, Skeletal changes in the ewe after ovariectomy. J. Bone Miner. Res. 9 (Suppl. 1), S258 (1994). 303. C. Bressot, P. J. Meunier, M. C. Chapuy, E. Lejeune, C. Edouard, and A. J. Darby, Histomorphometric profile, pathophysiology and reversibility of corticosteroid-induced osteoporosis. Metab. Bone Dis. Rel. Res. 1, 303 – 311 (1979). 304. P. Chavassieux, P. Pastoureau, M. C. Chapuy, P. D. Delmas, and P. J. Meunier, Glucocorticoid-induced inhibition of osteoblastic bone formation in ewes: A biochemical and histomorphometric study. Osteroporosis Int. 3, 97 – 102 (1993). 305. M. J. Jayo, D. S. Weaver, S. E. Rankin, and J. R. Kaplan, Accuracy and reproducibility of lumbar bone mineral status determined by DPA in live male cynomolgous macaques. Lab. Anim. Sci. 40, 266 – 269 (1990). 306. N. S. Pope, K. G. Gould, D. C. Anderson, and D. R. Mann, Effects of age and sex on bone density in the rhesus monkey. Bone 10, 109 – 112 (1989). 307. M. J. Jayo, S. E. Rankin, D. S. Weaver, C. S. Carlson, andT. B. Clarkson, Accuracy and precision of lumbar bone mineral content by DXA in live female monkeys. Calcif. Tissue Int. 49, 438 – 440 (1991). 308. J. Rogers and K. K. Kidd, Nuclear DNA polymorphisms in a wild population of yellow baboons (Papio hamadryas cynocephalus) from Mikumi National Park Tanzania. Am. J. Phys. Anthropol. 90, 477 – 486 (1993). 309. J. Rogers, G. Guano, and K. K. Kidd, Variability in nucelar DNA among non-human primates: Application of molecular genetic techniques to intro- and inter-species genetic analyses. Am. J. Primatol. 27, 93 – 105 (1992). 310. C. Jerome, D. B. Kimmel, J. A. McAlister, and D. S. Weaver, Effects of ovariectomy on iliac trabecular bone in baboons (Papio anubis). Calcif. Tissue Int. 39, 206 – 208 (1986). 311. C. Miller and D. Weaver, Bone loss in ovariectomized monkeys. Calcif. Tissue Int. 38, 62 – 65 (1986). 312. L. C. Miller, D. S. Weaver, J. A. McAlister, and D. R. Koritnik, Effects of ovariectomy on vertebral trabecular bone in the cynomolgus monkey. Calcif. Tissue Int. 38, 62 – 65 (1986). 313. K. Lundon and M. Grynpas, The long term effect of ovariectomy on the quality and quantity of cortical bone in the young cynomolgus monkey: A comparison of density fractionation and histomorphometric techniques. Bone 14, 389 – 395 (1993). 314. K. Lundon, M. Dumitriu, and M. Grynpas, The long-term effect of ovariectomy on the quality and quantity of cancellous bone in young macaques. Bone Miner. 24, 135 – 149 (1994).

CHAPTER 37 Animal Models for in Vivo Experimentation 315. D. R. Mann, C. G. Rudman, M. A. Akinbami, and K. G. Gould, Preservation of bone mass in hypogonadal female monkeys with recombinant human growth hormone administration. J. Clin. Endocrinol. Metab. 74, 1263 – 1269 (1992). 316. C. A. Shively, M. J. Jayo, D. S. Weaver, and J. R. Kaplan, Reduced vertebral bone mineral density in socially subordinate female cynomolgus macaques. Am. J. Primatol. 24, 135 – 138 (1991). 317. T. B. Aufdemorte, W. C. Fox, D. Miller, K. Buffum, G. R. Holt, and K. D. Carey, A non-human primate model for the study of osteoporosis and oral bone loss. Bone 14, 581 – 586 (1993). 318. M. D. Grynpas, C. B. Huckell, K. J. Reichs, C. J. Derousseau, C. Greenwood, and M. J. Kessler, Effect of age and osteoarthritis on bone mineral in rhesus monkey vertebrae. J. Bone Miner. Res. 8, 909 – 917 (1993). 319. D. R. Young, W. J. Niklowitz, and C. R. Steele, Tibial changes in experimental disuse osteoporosis in the monkey. Calcif. Tissue Int. 35, 304 – 308 (1983). 320. C. M. Schnitzler, U. Ripamonti, and J. M. Mesquita, Histomorphometry of iliac crest trabecular bone in adult male baboons in captivity. Calcif. Tissue Int. 52, 447 – 454 (1993). 321. R. R. Recker, D. B. Kimmel, A. M. Parfitt, K. M. Davies, N. Keshawarz, and S. Hinders, Static and tetracycline-based bone histomorphometric data from 34 normal post-menopausal females. J. Bone Miner. Res. 3, 133 – 144 (1988). 322. C. P. Jerome, C. J. Lees, T. C. Register, and R. P. Holmes, Bone loss in skeletally mature female monkeys fed minimal requirement of calcium is accelerated by ovariectomy. J. Bone Miner. Res. 9 (Suppl. 1), S259 (1994). 323. D. S. Weaver, R. A. Power, C. P. Jerome, B. M. Phifer, and T. C. Register, The effects of immediate and delayed treatment of ovariectomized cynomolgus monkeys with nandrolone decanoate on densitometry, bone biomarkers, sex steroids and body weight. J. Bone Miner. Res. 9 (Suppl. 1), S393 (1994).

47 324. R. A. Power, D. S. Weaver, C. P. Jerome, B. M. Phifer, and T. C. Register, Histomorphometry and serum bone biomarkersin ovariectomized nandrolone-treated cynomolgus monkeys. J. Bone Miner. Res. 9 (Suppl. 1), S393 (1994). 325. C. S. Carlson, R. F. Loeser, M. J. Jayo, D. S. Weaver, M. R. Adams, and C. P. Jerome, Osteoarthritis in cynomolgus macaques: A primate model of naturally occurring disease. J. Orthop. Res. 12, 331 – 339 (1994). 326. D. B. Kimmel, N. E. Lane, C. M. Kammerer, M. R. Stegman, K. S. Rice, and R. R. Recker, Spinal pathology in adult baboons. J. Bone Miner. Res. 8 (Suppl. 1), S279 (1993). 327. K. P. Hughes, D. B. Kimmel, C. M. Kammerer, K. M. Davies, K. S. Rice, and R. R. Recker, Vertebral morphometry in adult female baboons. J. Bone Miner. Res. 9 (Suppl. 1), S209 (1994). 328. J. Dequeker, The relationship between osteoporosis and osteoarthritis. Clin. Rheum. Dis. 11, 271 – 296 (1985). 329. H. Pogrund, M. Rutenberg, M. Makin, G. C. Robin, R. Steinberg, and R. Bloom, Osteoarthritis of the hand and osteoporosis. Clin. Orthop. Relat. Res. 203, 239 – 243 (1986). 330. K. C. Ng, P. A. Revell, M. Beer, B. J. Boucherf, R. D. Cohen, and H. L. F. Currey, Incidence of metabolic bone disease in the rheumatoid arthritis and osteoarthritis. Ann. Rheum. Dis. 43, 370 – 377 (1984). 331. A. Hopkins, S. Zylstra, M. M. Hreshchyshyn, and M. Anbar, Normal and abnormal features of the lumbar spine observed in dual energy absorptiometry scans. Clin. Nucl. Med. 14, 410 – 414 (1989). 332. E. S. Orwoll, S. K. Oviatt, and T. Mann, The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J. Clin. Endocrinol. Metab. 70, 1202 – 1207 (1990). 333. I. R. Reid, M. C. Evans, R. Ames, and D. J. Wattie, The influence of osteophytes and aortic calcification on spinal mineral density in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 1372 – 1374 (1991). 334. P. J. Drinka, A. A. DeSmet, S. F. Bauwens, and A. Rogot, The effect of overlying calcification on lumbar bone densitometry. Calcif. Tissue Int. 50, 507 – 510 (1992).

CHAPTER 38

The Type I/Type II Model for Involutional Osteoporosis Update and Modification Based on New Observations B. LAWRENCE RIGGS,* SUNDEEP KHOSLA,* AND L. JOSEPH MELTON III,*† * Department of Internal Medicine, Division of Endocrinology and Metabolism, and †Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905

I. Introduction II. Validating Evidence III. Tests of Validity

IV. Conceptual Problems V. Summary and Conclusions References

I. INTRODUCTION

genders and is strongly related to age, we prefer to use the term involutional osteoporosis in place of primary osteoporosis. Riggs and Melton in 1983 [1], and in subsequent publications [2,3], proposed that involutional osteoporosis could be subdivided into two distinct syndromes — type I osteoporosis and type II osteoporosis — that differed with respect to changes in regional bone mineral density (BMD), pattern of fractures, hormonal changes, and causal mechanisms (Table 1). In this chapter, we review this model and update and modify it based on new observations, especially the findings that estrogen deficiency is a major cause of bone loss in elderly women and, perhaps also, in aging men [4].

A. Background Osteoporosis often is divided into primary osteoporosis and secondary osteoporosis syndromes, depending on whether or not the patient has a recognizable disease or is using a drug that can cause bone loss. Additionally, there is a general agreement that the rare syndromes of juvenile osteoporosis that occur in children near puberty and idiopathic osteoporosis that occurs in young adults of both sexes should be considered as separate entities. Because it occurs in both

OSTEOPOROSIS, SECOND EDITION VOLUME 2

49

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

50

RIGGS, KHOSLA, AND MELTON

TABLE 1

Characterization of the Two Main Types of Involutional Osteoporosis Type I

Type II

Age (years)

51 – 75

 70

Sex ratio (F:M)

6:1

2:1

Type of bone loss

Mainly trabecular

Trabecular and cortical

Rate of bone loss

Accelerated

Not accelerated

Major fracture sites

Vertebrae (crush) and distal radius

Vertebrae (multiple wedge) and hip

Parathyroid function

Decreased

Increased

Estrogen effects

Mainly skeletal

Mainly extraskeletal

Main causes

Menopause plus individual predisposing factor(s)

Factors related to aging including late effects of estrogen deficiency

B. Clinical Characteristics Type I osteoporosis typically affects women within 15 to 20 years after menopause. It is characterized by fractures occurring at sites that contain relatively large amounts of cancellous bone, such as the vertebral body, distal forearm, and ankle. The mandible and the maxilla also contain substantial amounts of cancellous bone, and, therefore, increased tooth loss occurs in type I osteoporosis. The vertebral fractures are of the crush or collapse type and usually are associated with a reduction of more than 25% of vertebral height. These are commonly acutely painful [5], and the back pain may require up to 6 months to subside. Type II osteoporosis occurs in both sexes but is twice as common in women as in men. Although type II osteoporosis can occur at any age, it is the predominant form of osteoporosis in women and men over the age of 70 years. Fractures associated with type II osteoporosis occur at sites that contain both cancellous and cortical bone. The most typical type II osteoporotic fracture is hip fracture. Also, fractures of the pelvis, proximal humerus, and proximal tibia are commonly associated with type II osteoporosis. Typically, a specific type of vertebral fracture can be included as part of the syndrome of type II osteoporosis. Rather than the acutely painful crush or collapse fractures observed in type I osteoporosis, there usually is a gradual and progressive deformation of the vertebrae, leading to dorsal kyphosis, often referred to as “the dowager’s hump.” Such fractures are painless or are associated with minimal aching pain, are manifest as anterior wedge deformities usually with less than a 25% reduction in vertebral height, are located almost exclusively in the mid-thoracic area, and generally occur in a series of several adjacent vertebrae.

C. Relationship to Patterns of Age-Related Bone Loss The manifestations of type I and type II osteoporosis are closely related to underlying patterns of age-related bone

loss. Based on cross-sectional and longitudinal bone densitometric studies [6 – 9], two distinct phases of age-related bone loss have been recognized — a slow age-related phase that occurs in both sexes and an accelerated phase that occurs only in postmenopausal women and, more rarely, in hypogonadal men. These are shown diagrammatically in Fig. 1. The slow phase of bone loss begins about age 40 and continues throughout life, has a similar rate in both sexes, and results in loss of similar amounts of cortical and cancellous bone. A transient accelerated phase beginning at the menopause is superimposed in women and results in a loss of disproportionately more cancellous bone than cortical bone. The accelerated postmenopausal loss lasts about 10 years, although most of the bone is lost in the first 3 to 4 years, and it declines exponentially with time until it

FIGURE 1

Changes in bone mass with aging in men and women showing patterns of bone loss. (I) Peak bone mass, (II) rapid phase of bone loss seen in women around the menopause, (III) is the age-related bone loss which is similar in both men and women. Modified from B. L. Riggs and L. J. Melton III, Involutional osteoporosis. In “Oxford Textbook of Geriatric Medicine, (J. G. Evans and T. F. Williams, eds.), pp 405 – 411. Oxford University Press, Oxford.

51

CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis

merges asymptotically with the continuing slow phase of loss.

II. VALIDATING EVIDENCE The type I/type II model of involutional osteoporosis is supported by four different types of data: — differences in fracture patterns as assessed by epidemiologic studies, patterns of bone loss as assessed by bone densitometry, differences in parathyroid function, and differences in hormonal mechanisms of bone loss. That these data are largely independent of each other and yet lead to the same conclusion is, we believe, strongly supportive of the central hypothesis.

A. Differences in Fracture Pattern Striking differences have been noted with regard to the distribution by gender, frequency of occurrence, and location of fractures. As the epidemiological characteristics of a disease reflect its underlying pathophysiology, these observed differences in fracture pattern suggest that the pathophysiological mechanisms responsible for type I and type II osteoporosis are different. The female-to-male incidence ratio is about 6:1 for vertebral fractures occurring between ages 51 and 65 years, whereas it is only 2:1 for hip fractures occurring after 75 years [1,2]. Among women, the incidence of relatively “pure” fractures of type I osteoporosis, such as fracture of the distal forearm, rises soon after menopause, continues to

FIGURE 2

rise for another 10 to 15 years, and then plateaus. In contrast, the incidence of fractures of the femoral neck, a relatively “pure” type II osteoporotic fracture, increases exponentially throughout life. These contrasting patterns are shown in Fig. 2. In contrast to these clearly different fracture patterns for the hip and wrist, the pattern for vertebral fracture appears to be intermediate between the type I and type II patterns. Like type I fractures, the incidence of vertebral fractures increases soon after menopause and by age 65 only 2% of women will have had a hip fracture, whereas 11% will have had a vertebral fracture. Thereafter, the ratio of incidence for vertebra and hip fractures declines steadily until, among women who are 85 years of age or older, it is 1:1 [1,2]. We believe that this intermittent pattern occurs because vertebral fractures are, in fact, an admixture of two different types — crush fractures due to excessive osteoclastic activity and perforative resorption of cancellous plates occurring soon after menopause and wedge fractures due to trabecular thinning from impaired osteoblastic activity later in life. Direct evidence for such an admixture is supported by the study of Jensen et al. [10] who found in a random sample of 70-year-old Danish women, onefifth of whom had vertebral fractures, that 20% of them were crush fractures whereas 80% of them were wedge deformities. Finally, the proportional content of cancellous bone determines the location of fracture in the two osteoporotic syndromes. In patients with type I osteoporosis, fractures occur at sites that contain 50 to 75% cancellous bone, whereas, in patients with type II osteoporosis, fractures occur at sites that contain only 25 to 50% cancellous bone.

Incidence rates for a “pure” type I osteoporotic fracture (Colles’ fracture of distal forearm) and for a “pure” type II osteoporotic fracture (fracture of proximal femur) plotted as a function of age at the time of the fracture. Data are for women from the community population of Rochester, Minnesota (L. J. Melton and B. L. Riggs, unpublished data).

52

RIGGS, KHOSLA, AND MELTON

TABLE 2 Clinical syndrome Location pattern

Age-Related Fractures: Types of Patterns

Type I osteoporosis

Type II osteoporosis

Distal forearm

Proximal femur

Distal tibia

Proximal humerus

Vertebrae (crush)

Traumatic only Shafts of limb bones

Proximal tibia Pelvis Vertebrae (multiple wedge)

Sex ratio (F:M)

6:1

Bone composition

 85% Trabecular

2:1

1:1

Cortical 50 – 70%

 85% Cortical

Trabecular 30 – 50% Note. F, female; M, male.

These complex relationships are shown in Table 2. A third pattern is exhibited by fractures of the shafts of the long bones that are composed almost entirely of cortical bone. These fractures show neither a definite female preponderance nor an increase with age, and they occur almost entirely in association with severe trauma. Thus, this third pattern of fractures does not seem to be directly related to osteoporosis.

B. Differences in Patterns of Bone Loss Women with vertebral crush fractures due to type I osteoporosis have a mean BMD for lumbar spine, a site containing predominately cancellous bone, that is about 2 standard deviations (SD) below the mean for age-matched normal women, and about half of the individual values are below the 10th percentile of normal. However, for the radius shaft, a site containing predominately cortical bone, the mean BMD is only about 0.5 SD below the normal mean, and only a few individual values are below the 10th percentile [7]. In contrast, elderly patients with hip fracture due to type II osteoporosis have only small decreases below the age-adjusted normal mean for BMD at any site, although individual values tend to cluster in the lower part of the range [8]. These data suggest that patients with type I osteoporosis have lost disproportionately more cancellous bone than cortical bone, whereas patients with type II osteoporosis have lost similar amounts of both types of bone (Fig. 3).

C. Differences in Parathyroid Function An important difference that is qualitative, rather than quantitative, is parathyroid function as it changes in opposite directions in the two syndromes. As previously reviewed [4], during the accelerated phase of bone loss

early after menopause and in established osteoporosis, parathyroid function is normal or slightly decreased due to compensatory suppression to maintain normal levels of serum ionized calcium activity in the presence of the primary increased osteoclastic activity. When bone resorption is decreased by estrogen treatment, however, there are significant increases in circulating PTH, both in the early postmenopausal accelerated phase of bone loss and in established type I osteoporosis [11], consistent with a partial suppression of parathyroid function. In contrast, PTH concentrations increase with aging and its levels are higher in some patients with hip fracture than their age-matched controls [12]. Moreover, elderly women have abnormal PTH secretory dynamics that are consistent with functional parathyroid gland hyperplasia. Therefore, in type I osteoporosis, decreased PTH secretion may be the result of the increase in bone loss, whereas, in type II osteoporosis, increased PTH secretion may be driving the bone loss.

D. Differences in Causal Mechanisms The data thus far cited show that there are major differences in the presentation of the two syndromes suggesting that there are major differences in pathophysiologic mechanisms. We will review below the evidence that such differences do exist. 1. TYPE I OSTEOPOROSIS The predilection for women and the temporal proximity to menopause implicate estrogen deficiency as the etiologic agent for this type of osteoporosis. The relatively rapid and large decrease in estrogen secretion at menopause leads to increased bone turnover, and bone resorption increases to a greater extent than does bone formation [4,12], leading to rapid bone loss. However, although all postmenopausal women are estrogen deficient, type I osteoporosis occurs in only about 10 to 20% of them. To reconcile this discrepancy,

53

CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis

FIGURE 3

Bone mineral density (BMD) levels for vertebrae and femoral neck, plotted as a function of age for 111 patients with vertebral fractures () and 49 patients with hip fractures (). The line represents the regression on age; the cross-hatched area shows the 90% confidence limits for 166 normal women. Note that the fracture threshold (90th percentile of the measurements for patients with fractures) is about 1.0 g/cm2 and is independent of age. Data are expanded from those reported by B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton, III, Changes in bone mineral density of the proximal femur and spine with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982).

we have suggested that the type I osteoporosis syndrome requires not only estrogen deficiency, but some additional factor or factors that operate only in the presence of estrogen deficiency and lead to an exacerbation and prolongation of the rapid phase of bone loss [1 – 3]. In normal women, this rapid phase lasts only from 4 to 8 years, whereas in women who develop type I osteoporosis it may last for 15 to 20 years. It is unclear at present what are the factors enhancing bone loss in this subset of early postmenopausal women. Perhaps a genetically determined increased responsiveness of bone in the presence of estrogen deficiency may predispose some postmenopausal women, but not others, to excessive bone loss. Much has been learned in the past decade about the paracrine mediators of the estrogen effect on bone. A number of proinflammatory cytokines, including interleukin-1 (IL-1), IL-6, tumor necrosis factor (TNF), and prostaglandin E2, increase the abundance of osteoclast precursors in bone marrow; and the presence of macrophage colony-stimulating factor(M-CSF) and newly discovered members of the TNF and TNF receptor superfamilies, osteoprotegerin and receptor activator of NF-B(RANK) ligand [12], allow these to be differentiated into active mature osteoclasts. Estrogen appears to act at multiple levels in this regulatory system to reduce osteoclast formation. Polymorphisms of genes coding for these paracrine effectors could lead to enhanced resorptive activity when estrogen is deficient. Genetic polymorphisms resulting in differences in the

number or the function of estrogen receptors could also augment the effect of estrogen deficiency on bone cell activity, although this has not yet been convincingly demonstrated. Another potential predisposing factor is the presence of impaired renal tubular calcium transport leading to chronic renal calcium losses in women with type I osteoporosis [14]. The skeletal and renal abnormalities would not be mutually exclusive and both could result from local increases of the same cytokine(s) in bone and kidney. 2. TYPE II OSTEOPOROSIS IN WOMEN This form of osteoporosis is caused by age-related factors occurring throughout the entire population of aging women and men. Of these, the two most important proximate causes appear to be secondary hyperparathyroidism and an age-related impairment in osteoblast function. In addition, some elderly housebound men and women develop nutritional deficiency of vitamin D [15], which exacerbates the secondary hyperparathyroidism and bone loss. Serum intact PTH concentrations and indices of bone turnover increase pari passu in aging women, and these increases correlate directly with each other, even after adjusting for the effect of age. As previously reviewed [4], suppression of PTH secretion by intravenous calcium infusion abolished the differences in bone resorption markers between young and elderly women. This suggests strongly that the increase in bone resorption in aging women is PTH-dependent. Concentrations of serum PTH and bone

54 resorption markers also increase in aging men and the pattern of increase closely resembles that in aging women. Thus, a considerable body of data implicates secondary hyperparathyroidism as a major cause of the slow, age-related phase of bone loss in both genders. Although the original 1983 hypothesis attributed the increase in bone resorption and secondary hyperparathyroidism to a primary age-related defect in intestinal calcium absorption and renal calcium conservation, this part of the hypothesis now must be modified to accommodate the new findings of the continued effects of estrogen deficiency during the slow phase of bone loss in both aging women and men [4]. Both McKane et al. [16], in experimental studies, and Khosla et al. [17], in population-based observational studies, demonstrated that the increases in both serum PTH and in bone resorption in postmenopausal women could be restored to the level found in premenopausal women by estrogen replacement therapy. Moreover, it has been recently demonstrated that, in addition to direct effects on bone cells, estrogen has potent extraskeletal effects on calcium homeostasis. The intestine contains estrogen receptors. Estrogen acts through these to increase intestinal calcium absorption [18,19], possibly by enhancing the responsiveness of the intestine to 1,25-dihydroxyvitamin D [19]. Estrogen also acts to increase renal calcium conservation [20,21] by enhancing tubular reabsorption of calcium through a PTHindependent mechanism [21]. In the presence of estrogen deficiency, the loss of these extraskeletal effects leads to calcium wasting and to a substantial increase in the level of dietary calcium required to prevent negative calcium balance. That the effects of estrogen in postmenopausal women are mainly extraskeletal, rather than skeletal, was further demonstrated by McKane et al. [22] who demonstrated that chronic ingestion of a very high calcium intake, 2400 mg/day, also restored bone resorption and serum PTH to premenopausal levels. Thus, the increases in bone resorption in elderly women can be normalized either by estrogen replacement (which restores the extraskeletal calcium fluxes to premenopausal levels) or by large increases in calcium intake (which offsets the net calcium losses induced by postmenopausal abnormalities in extraskeletal calcium fluxes). The cause of the impaired osteoblast function in elderly women is unclear. This abnormality exacerbates bone loss by preventing a compensatory increase in bone formation to offset the postmenopausal increases in bone resorption. As previously reviewed [4], histomorphometric data have demonstrated decreased bone formation at the cellular level with aging, which is decreased even more in patients with osteoporosis than in age- matched normals. These changes generally have been attributed to age-related factors, particularly to decreased paracrine production of growth factors or to decreased circulating levels of growth hormone and insulin-like growth factor-I (IGF-I).

RIGGS, KHOSLA, AND MELTON

It is also possible that this late occurring defect in osteoblast function could be partly or entirely the result of estrogen deficiency. Several observations are consistent with this possibility. Estrogen treatment increases production of IGF-I [23] and transforming growth factor- [24] by osteoblastic cells in vitro and acts to decrease osteoblast apoptosis [25]. Also, as previously reviewed [4], an effect of estrogen to stimulate skin collagen synthesis by fibroblasts has been repeatedly demonstrated and, as with fi broblasts, collagen production by osteoblasts represents 90% of cell biosynthetic capacity. However, estrogen treatment also has been reported both to stimulate and to inhibit bone formation in experimental animals and to stimulate and inhibit proliferation of human osteoblastic cells in vitro, so the issue remains unresolved. 3. TYPE II OSTEOPOROSIS IN MEN Men also lose bone with aging and have a third as many fractures as women. After accounting for their lack of the rapid phase of early postmenopausal bone loss, men exhibit the same pattern of slow bone loss and similar increases in bone resorption and serum PTH as aging women do [26]. These observations suggest that the causal mechanism for slow bone loss in aging men may be the same as or very similar to that of aging women. Recently reported data suggest that estrogen deficiency is the main cause for the secondary hyperparathyroidism and bone loss in elderly women [4]. However, because only a few aging men develop overt hypogonadism, how can estrogen deficiency be implicated? Recent studies in which either free or bioavailable sex steroids have been measured show that aging men do in fact have substantial decreases in both bioavailable estrogen and testosterone [4,26] (Table 3). In contrast to women, in whom the major cause of the decrease is a sharp

TABLE 3 Changes in Sex Steroids and Related Factors in Men and Women over Life in Random Age-Stratified Samples of 350 Women and 350 Men Men % change

Women % change

27**

45**

Bioavailable estrogen

47**

83**

Bioavailable testosterone

64**

28*

Sex hormone binding globulin

124**

1

Luteinizing hormone

285**

731**

Follicle stimulating hormone

505**

1805**

Lateral spine BMD Serum

Note. Data are from Khosla et al. (J. Clin. Endocrinol. Metab. 83 2266 – 2274, 1998) *P  0.05 **P  0.005.

55

CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis

drop in their production rate due to menopause, the major cause of the decreases in both estrogen and testosterone in men is a large age-related increase in sex hormone binding globulin [26], which binds the sex steroids in an inactive complex. Recent clinical reports that young adult men with either null mutations of the estrogen receptor- gene (who cannot respond to estrogen) [27] or the aromatase gene (who cannot synthesize estrogen) [28,29] have osteoporosis despite normal serum testosterone concentrations suggest that estrogen plays a major role in maintaining bone mass in men. As previously reviewed [4], this is further corroborated by four large, population-based observational studies that have shown that serum free or bioavailable estrogen correlates better with BMD in men than does free or bioavailable testosterone. Nonetheless, these findings do not exclude a substantial effect of the decrease of bioavailable testosterone in mediating bone loss in aging men. Further studies are needed to quantify the relative effects of reduced levels of both sex steroids.

III. TESTS OF VALIDITY For any hypothesis to be useful, it should lead to predictions that can be tested experimentally. The limited number of tests that have been made thus far have all supported the hypothesis and are reviewed below. First, Owen et al. [30] and Kotowicz et al. [31] carried out epidemiologic studies of the hypothesis that fracture occurrence in involutional osteoporosis is the result of two distinct syndromes. Using the resources of the Rochester Epidemiology Project, they followed subjects through time by reviewing their medical records to determine whether distal forearm fractures or vertebral crush fractures occurring in women below age 65 would be subsequently associated with a small or a large increase in hip fractures late in life. If the null hypothesis that there is but one osteoporotic syndrome is correct, women having early postmenopausal fractures would have the most severe disease and should have a large subsequent increase in hip fractures late in life as compared to the overall Rochester population, possibly a 5- to 10-fold increase. In contrast, the increase in hip fractures late in life would be relatively small if the hypothesis of separate syndromes were correct. Using this populationbased retrospective cohort study design, these investigators found that the relative risk of hip fracture in women with previous distal forearm fractures was 1.2 and for women with vertebral fractures was 1.8. Thus, these results strongly support the type I/type II model for osteoporosis. Second, if involutional osteoporosis is a single entity, there should be no differences in the type of bone lost in patients with typical type I and type II fractures. This was tested experimentally by Johnston et al. [32] who studied

iliac bone biopsies from 32 women with vertebral fractures (type I osteoporosis) and compared them with biopsies obtained from 27 patients with hip fractures (type II osteoporosis). Patients with vertebral fractures had decreased trabecular bone volume compared to age-matched controls, and patients with hip fractures had a deficit of both cortical and trabecular bone compared to perimenopausal women, but not different from age-matched controls. These findings of histological differences in the proportion of cortical and frabicular bone type I and type II osteoporosis is consistent with different pathophysiological mechanisms.

IV. CONCEPTUAL PROBLEMS Since the type I/type II model of osteoporosis was proposed, questions and concerns have been raised. We feel that they can be answered and resolved by existing data and by a clearer understanding of the hypothesis. These are discussed below.

A. Role of Falls It has been suggested that both type I and type II osteoporosis are characterized by osteopenia and that apparent differences in fracture patterns might be explained by agerelated differences in the frequency and types of falls [33]. While the increased propensity of the elderly to fall contributes to fracture risk, it is the difference in the relative proportions of cancellous and cortical bone mass that mainly determine the differences in fracture pattern between the two types of disorders. Thus, it is clear that falls onto the hip lead to an increased risk of hip fractures [34] but comparable age-related increases in distal femur fracture [35] cannot be explained in this manner. Although certain falls predispose to certain types of fractures [33], the risk of fracture is proportional to the bone mass at the fracture site, and this in turn is a function of the overall pattern of cancellous and cortical bone loss.

B. Role of Peak Bone Mass The type I/type II model for involutional osteoporosis explains pathogenesis mainly in terms of patterns of agerelated bone loss rather than of differences in peak bone mass. Most estimates are that the variance in BMD for 70year-old women is approximately equally due to peak bone mass and to subsequent rates of bone loss [12,36 – 38]. Thus, the rank order of individuals in the Gaussian distribution of BMD values at the onset of involutional bone loss greatly influences their risk of falling below the threshold for fracture as age-related bone loss ensues. This is

56

RIGGS, KHOSLA, AND MELTON

supported by the demonstration of reduced bone mass in premenopausal daughters of osteoporotic women [39]. For women with type I osteoporosis, however, the greater bone loss rather than a greater BMD at the onset of the bone loss clearly is playing the dominant role. There is large variability in the rate of bone loss in the early years after menopause [38,40]. Also, women with established type I osteoporosis have higher turnover than their peers matched for years after menopause [41,42], and these differences are reflected by their much lower BMD values (Fig. 3). However, the contribution of peak bone mass to fracture risk probably is greater for type II osteoporosis because high variability in the rates of the slow phase of bone loss has not been demonstrated and because there is a large overlap between individuals with and without fractures in this syndrome. These observations suggest that the rank order of individuals in the Gaussian distribution of BMD values at the onset of involutional bone loss may be the main determinant of fracture risk in type II osteoporosis, consistent with the hypothesis proposed by Newton-John and Morgan [43] over 40 years ago.

C. Overlap of Causal Processes in Both Syndromes Another objection is that the menopause in women and the aging processes in both sexes, the proposed causes of type I osteoporosis and type II osteoporosis respectively, are universal processes. Therefore, how could these result in two separate syndromes? As illustrated in Fig. 4, specificity occurs because type I osteoporosis is the result of not just menopause but also of one or more additional factors that are present only in some postmenopausal women and that amplify and extend the bone loss induced by estrogen deficiency. Thus, women with type I osteoporosis may represent a separate population, distinct from their postmenopausal peers, who are losing or have lost bone at a more rapid rate or for a longer duration. In contrast, the processes causing type II osteoporosis appear to involve the entire population of aging men and women and, as the slow phase of bone loss progresses, an increasing number of them will have BMD values below the fracture threshold. The occurrence of an accelerated phase of bone loss after menopause in all women, however, including those destined to develop type II osteoporosis, explains why elderly women have a twofold greater increase in hip fracture than elderly men even though the rates of slow bone loss are similar in both sexes.

D. Causal Role of Estrogen Deficiency Because we now consider estrogen deficiency to be the dominant mechanism for both the rapid accelerated phase

FIGURE 4

Model for development of type I osteoporosis and type II osteoporosis. Dark line, represents regression of BMD on age for women. Degree of fracture risk for levels of BMD are shown in inset. Slow bone loss begins in the fourth decade and accelerates transiently after menopause. Light line, represents regression of BMD on age for the subset of women in whom type I osteoporosis develops. In these, an additional factor(s) causes the postmenopausal acceleration of bone loss to be exaggerated and prolonged so that BMD falls below the fracture threshold within 15 to 20 years after menopause. In contrast, patients with type II osteoporosis represent the entire population of aging women or men (Gaussian distribution shown by top-shaped figure). As slow bone loss progresses, increasing proportions of the general population fall below the fracture threshold late in life. Although all eventually will become at risk for fractures (distribution of fracture patients shown by small solid circles), those with the lowest BMD are at the greatest risk (From B. L. Riggs and L. J. Melton, III Clinical heterogeneity of involutional osteoporosis: Implications for prevention therapy. J. Clin. Endocrinol. Metab 70, 1229 – 1232 (1990)).

of bone loss in early postmenopausal women and for the slow phase of bone loss in elderly women and aging men [4], how is it possible for a single process to produce different clinical manifestations? New information suggests that estrogen-deficiency causes bone loss both through the loss of the direct action of estrogen on bone cells (that restrains bone turnover) and through the loss of the action of estrogen on the intestine and kidney (that maintain extraskeletal calcium fluxes) [4]. In the rapid phase of bone loss in early postmenopausal women, the contribution of both processes are balanced with a slight predominance of the loss of the direct effects on bone cells that restrain their activity. This slight imbalance leads to a net transfer of calcium from bone into the extracellular fluids that causes minimal suppression of parathyroid function. As the response of bone cells to estrogen deficiency wanes, the rapid phase of bone loss subsides over 4 to 8 years. Thereafter, the effect of estrogen loss on extraskeletal calcium homeostasis leads to secondary hyperparathyroidism that then becomes the predominant mechanism driving bone loss. The presence of estrogen-independent causes of decreased osteoblast function, if these exist, and nutritional vitamin D

CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis

deficiency in some elderly persons could further aggravate this loss.

V. SUMMARY AND CONCLUSIONS Evidence from multiple sources suggests that involutional osteoporosis can be subdivided into distinct syndromes of type I (postmenopausal) osteoporosis and type II (age-related) osteoporosis. Although estrogen deficiency is the dominant cause of bone loss in both types of osteoporosis, the mechanisms by which it induces bone loss differs. The type I/type II model has proved to be useful in organizing the large amount of information that has developed on osteoporosis during the past 15 years and has been helpful in examining the pathogenesis of the disease and in evaluating therapeutic results. We believe that it still provides these useful functions. Undoubtedly, however, further modifications will be required as new information is accrued on pathogenesis, particularly that relating to underlying genetic and molecular mechanisms.

Acknowledgments This work was supported by NIH Grants AG – 04875 and AR – 27065.

References 1. B. L. Riggs and L. J. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1983). 2. B. L. Riggs and L. J. Melton. Medical progress series: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 3. B. L. Riggs and L. J. Melton, Clinical heterogeneity of involutional osteoporosis: Implications for preventive therapy. J. Clin. Endocrinol. Metab. 70, 1229 – 1232 (1990). 4. B. L. Riggs, S. Khosla, and L. J., Melton III, A unitary model for involutional osteoporosis: Estrogen deficiency causes both Type I and Type II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13, 763 – 773 (1998). 5. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J., Melton III, The incidence of clinically diagnosed vertebral fractures: A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 6. R. Lindsay, J. M. Aitken, J. B. Anderson, D. M. Hart, E. B. MacDonald, and A. C. Clarke. Long-term prevention of postmenopausal osteoporosis by oestrogen: Evidence for an increased bone mass after delayed onset of oestrogen treatment. Lancet 1, 1038 – 1041 (1976). 7. B. L. Riggs, H. W. Wahner, W. L. Dunn, R. B. Mazess, K. P. Offord, and L. J., Melton III, Differential changes in bone mineral density of the appendicular and axial skeleton with aging: Relationship to spinal osteoporosis. J. Clin. Invest. 67, 328 – 335 (1981). 8. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J., Melton III, Changes in bone mineral density of the proximal femur with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982).

57 9. J. M. Pouilles, F. Tremollieres, and C. Ribot. The effects of menopause on longitudinal bone loss from the spine. Calcif. Tissue Int. 52, 340 – 343 (1993). 10. G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedus, and I. Transbøl. Epidemiology of postmenopausal spinal and long bone fractures: A unifying approach to postmenopausal osteoporosis. Clin. Orthop. Relat. Res. 166, 75 – 81 (1982). 11. B. l. Riggs, C. D. Arnaud, J. Jowsey, R. S. Goldsmith, and P. J. Kelly, Parathyroid function in primary osteoporosis. J. Clin. Invest. 52, 181 – 184 (1973). 12. P. Garnero, Sornay-E. Rendu, M. Choppy, and P. D. Delmas, Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J. Bone Miner. Res. 11, 337 – 349 (1996). 13. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 14. H. M. Heshmati, S. Khosla, M. F. Burritt, W. M. O’Fallon, and B. L. Riggs , Primary defect in renal calcium conservation may contribute to the pathogenesis of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 83, 1916 – 1920 (1988). 15. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 16. R. W. McKane, S. Khosla, J. Risteli, S. P. Robins, J. M. Muhs, and B. L. Riggs, Role of estrogen deficiency in pathogenesis of secondary hyperparathyroidism and increased bone resorption in elderly women. Proc. Assoc. Am. Physicians 109, 174 – 180 (1997). 17. S. Khosla, E. J. Atkinson, L. J. Melton, III, and B. L. Riggs, Effects of age and estrogen status on serum parathyroid hormone levels and biochemical markers of bone turnover in women: A population-based study. J. Clin. Endocrinol. Metab. 82, 1522 – 1527 (1997). 18. J. C. Gallagher, B. L. Riggs, and H. F. DeLuca, Effect of estrogen on calcium absorption and serum vitamin D metabolites in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 51, 1359 – 1364 (1980). 19. C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D3 in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). 20. B. E. C. Nordin, A. G. Need, H. A. Morris, M. Horowitz, and W. G. Robertson, Evidence for a renal calcium leak in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 401 – 407 (1991). 21. W. R. McKane, S. Khosla, M. F. Burritt, P. C. Kao, D. M. Wilson, S. J. Ory, and B. L. Riggs, Mechanism of renal calcium conservation with estrogen replacement therapy in women in early postmenopause: A clinical research center study. J. Clin. Endocrinol. Metab. 80, 3458 – 3464 (1995). 22. W. R. McKane, S. Khosla, K. S. Egan, S. P. Robins, M. F. Burritt, and B. L. Riggs, Role of calcium intake in modulating age-related increases in parathyroid function and bone resorption. J. Clin. Endocrinol. Metab. 81, 1699 – 1703 (1996). 23. M. Ernst, J. K. Heath, and G. A. Rodan, Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bones. Endocrinology 125, 825 – 833 (1989). 24. M. J. Oursler, C. Cortese, P. E. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor production in normal human osteoblast-like cells by 17-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). 25. S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocrinol. Rev. (in press).

58 26. S. Khosla, L. J. Melton, III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 27. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 28. C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effect of testosterone and estradiol in a man with aromatase deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). 29. A. Morishima, M. M. Grumbach, and J. P. Bilezikian, Estrogen markedly increases bone mass in an estrogen deficient young man with aromatase deficiency. J. Bone Miner. Res. 12, S126. 30. R. A. Owen, L. J. Melton, III, D. M. Ilstrup, K. A. Johnson, and B. L. Riggs, Colles’ fracture and subsequent hip fracture risk. Clin. Orthop. 171, 37 – 44 (1982). 31. M. A. Kotowicz, L. J. Melton III, C. Cooper, E. J. Atkinson, W. M. O’Fallon, and B. L. Riggs, Risk of hip fracture in women with vertebral fracture. J. Bone Miner. Res. 9, 599 – 605 (1994). 32. C. C. Johnston, J. Norton, M. R. A. Khairi, C. Kernek, C. Edouard, M. Arlot, and P. J. Meunier, Heterogeneity of fracture syndromes in postmenopausal women. J. Clin. Endocrinol. Metab. 61, 551 – 556 (1985). 33. M. C. Nevitt, S. R. Cummings, and the Study of Osteoporotic Fractures Research Group, Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993).

RIGGS, KHOSLA, AND MELTON 34. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 35. T. J. Arneson, L. J. Melton III, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin. Orthop. 234, 188 – 194 (1988). 36. D. E. Meier, E. S. Orwoll, and J. M. Jones, Marked disparity between trabecular and cortical bone loss with age in healthy men. Ann. Intern. Med. 101, 605 – 612 (1984). 37. J. E. Block, R. Smith, C. Glueer, P. Steiger, B. Ettinger, and H. K. Genant, Models of spinal trabecular bone loss as determined by quantitative computed tomography. J. Bone Miner. Res. 4, 249 – 257 (1989). 38. M. A. Hansen, K. Overgaard, B. J. Riis, and C. Christiansen, Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 Year study. Br. Med. J. 303, 961 – 964 (1991). 39. E. Seeman, J. L. Hopper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554 – 558 (1989). 40. A. Laib, H. J. Hauselmann, and P. Ruegsegger, In vivo high resolution 3D-QCT of the human forearm. Technol. Health Care 6, 329–337 (1998). 41. E. F. Eriksen, S. F. Hodgson, R. Eastell, W. F. O’Fallon, and B. L. Riggs, Trabecular bone formation and resorption rates in type I (postmenopausal) osteoporosis. J. Bone Miner. Res. 3, S203 (1988). 42. R. Eastell, S. P. Robins, T. Colwell, A. M. A. Assiri, B. L. Riggs, and R. G. G. Russell, Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporos. Int. 3, 255 – 260 (1993). 43. H. F. Newton-John and D. B. Morgan, Osteoporosis: Disease or senescence? Lancet 1, 232 – 233.

CHAPTER 39

Bone Remodeling Findings in Osteoporosis ROBERT R. RECKER AND M. JANET BARGER-LUX Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131

I. Introduction II. Background III. Histomorphometric Findings in Osteoporosis

IV. Bone Remodeling at the Whole-Organism Level V. Summary References

I. INTRODUCTION

II. BACKGROUND

This chapter will consider the role of bone remodeling — chiefly as disclosed by histomorphometry of transilial bone biopsies from human subjects — in the pathogenesis of osteoporosis. The following pages will examine the findings that are plausible concomitants of osteoporosis, understood here as the loss of skeletal adaptation to normal mechanical usage. A recent consensus conference on osteoporosis [1] has redefined osteoporosis as “a skeletal disorder characterized by compromised bone strength [emphasis added] predisposing to an increased risk of fracture.” By focusing on compromised bone strength, not simply bone mass, this consensus definition has converged with the focus of this chapter. Comparisons of findings from healthy subjects and subjects with osteoporosis will be affected by the fact that “osteoporotic” specimens have been collected after the disease has become identifiable, usually after fragility fractures have occurred. The processes that led to bone fragility (whether approached via histomorphometric, kinetic, or biochemical methods) may no longer be present.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Other chapters in this volume describe bone anatomy, the bone remodeling system, the remodeling transient, and markers of bone turnover in detail. The following section is limited to those features that are particularly relevant to the present discussion.

A. Functional Organization of Bone 1. BONE ENVELOPES In a recent commentary and review [2], Seeman reminded researchers that “bone biology is the study of the behavior of the surfaces of the skeleton [emphasis added].” Each of five bone envelopes identifies one of these surfaces, layers in which groups of cells perform a similar function within a similar relative location. Each bone envelope has characteristic metabolic activities and characteristic systems for organizing the function of its bone cells. Bone gain is mediated by the outermost (periosteal) bone envelope, and bone loss occurs in the innermost (transitional and endosteal) envelopes. The following paragraphs

59

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

60

RECKER AND BARGER-LUX

identify the bone envelopes, from the outer to the inner margin of bone tissue. a. Periosteal Envelope The periosteal envelope or periosteum covers most of the skeleton on its outer margin. It includes a sheath of fibrous connective tissue and an underlying layer of undifferentiated cells (cambium). Periosteum is absent from the attachment points of tendons and ligaments, areas covered by articular cartilage, the neck of the femur, the subscapular area, and sesamoid bones. Expansion of the periosteal envelope corresponds to changes in skeletal dimensions that occur throughout life. The periosteal envelope functions in fracture healing and in adaptation to intense mechanical usage [3 – 5]. b. Haversian Envelope The cells and surfaces of cortical bone delineate the Haversian envelope [6]. It includes the Haversian systems, their vessels and nerves, the osteocytes and their cytoplasmic extensions, the adjacent mineralized tissue, and the walls of the Volkmann’s canals. Thinning of the Haversian envelope (i.e., cortical thinning) corresponds to movement of the cortico-endosteal envelope toward the periosteal envelope [7]. c. Cortico-endosteal Envelope This envelope delimits the outermost boundary of the medullary canal, where trabeculae connect to the inner wall of the cortex. In adolescent girls, this envelope moves centrally toward the marrow cavity for a brief period; otherwise, the cortico-endosteal envelope expands throughout life [8]. d. Transitional Envelope This envelope is immediately adjacent to the cortico-endosteal envelope. Keshawarz and Recker identified this area as the transitional zone [7]. Though most workers have included it within the endosteal envelope, the transitional envelope can be considered separately because it is there that trabeculation of cortical bone has occurred. e. Endosteal Envelope This is the interface between bone marrow and trabecular (cancellous) bone [6]. The ratio of endosteal surface to trabecular bone volume is high compared to surface-to-volume ratios of the other envelopes. 2. INTERMEDIARY ORGANIZATION: REMODELING Frost [9] coined the term “intermediary organization” (IO) of the skeleton to describe the regulation of bone cell activity. He recognized that bone cells do not function individually and that the end product of their group activity is not disrupted by interventions that change the work or life span of a single class of bone cells such as osteoclasts. He inferred that bone cell activity was regulated by an order of control higher than a single cell (osteoclast or osteoblast) or a single function (resorption or formation). Four functional

subdivisions of the IO paradigm can be recognized: growth, modeling, remodeling, and fracture healing. Each has its own discreet IO but utilizes the same bone cells to accomplish its work. This discussion will focus on the remodeling IO. a. Activation, Resorption, and Formation Remodeling is the predominant metabolic activity of the skeleton in adult human life. It functions to remove and replace bone. At each remodeling site, the ARF sequence of activity takes place, with activation of the remodeling process, resorption of existing bone, and formation of new bone [10]. Osteoclasts excavate resorption tunnels (in cortical bone) and surface pits (in trabecular bone), osteoblasts fill the tunnels and pits with matrix, and the matrix mineralizes to become new bone. A complete cycle takes about 6 months in trabecular bone [11] and longer in cortical bone. The group of cells which carry out the work of a remodeling site has been called the basic multicellular unit (BMU), and the quantum of bone formed by a BMU has been called the basic structural unit (BSU) [12] (see Figs. 1 – 3). b. Coupling of Resorption and Formation The tight link between osteoclast and osteoblast function in bone remodeling is referred to as “coupling” [13]. In Frost’ paradigm, coupling is part of the remodeling IO. Both resorption and formation occur at the same site and, absent problems (e.g., calcium deficiency or a reduction in physical activity), the amount formed is almost always very nearly equal to the amount resorbed. There has been little success in finding ways of uncoupling resorption and formation. Antiresorptive agents, rather than uncoupling resorption and formation, function by inhibiting osteoclast activation. c. The Remodeling Transient After administration of an antiresorptive agent, bone mass increases until the remodeling projects already underway have been completed. The gain in bone mass that corresponds to filling of the remodeling space with new bone is the remodeling transient [14]. This phenomenon does not indicate “cure” of a skeletal disease. To determine whether a treatment can raise bone mass beyond filling the remodeling space requires very long periods of observation (as long as several years in humans): after a complete remodeling cycle has passed, one must observe the new steady state for two or more cycles.

B. Sources of Bone Fragility 1. LOW BONE MASS Low bone mass has long been recognized as a major component of skeletal fragility, and techniques for measuring bone mass in vivo have established this link [15].

CHAPTER 39 Bone Remodeling Findings in Osteoporosis

Basic multicellular unit (BMU). Photomicrograph (Goldner stain,  325 original magnification), of trabecular resorbing surface from a thin section of an undemineralized human transilial biopsy spoecimen. The scalloped resorption surface is occupied on the left by a multinucleated osteoclast. (Copyright, Dr. Robert R. Recker, used with permission.)

FIGURE 1

FIGURE 2 Basic multicellular unit (BMU). Photomicrograph (Goldner stain,  325, original magnification), of trabecular forming surface from a thin section of an undemineralized human transilial biopsy specimen. The plump osteoblasts with eccentric nuclei are lined up on a dark osteoid surface. (Copyright, Dr. Robert R. Recker, used with permission.)

61

62

RECKER AND BARGER-LUX

Basic structural unit (BSU). Photomicrograph (Toluidine blue stain,  150 original magnification), of completed remodeling sites from a thin section of an undemineralized human transilial biopsy specimen. Dark stained cement lines demarcate the limit of resorption and the start of formation. (Copyright, Dr. Robert R. Recker, used with permission.)

FIGURE 3

a. Earlier Acquisition vs Later Conservation Low bone mass can originate from failure to acquire sufficient bone mass during skeletal growth and consolidation and/or failure to conserve bone later on [1]. However, much of the literature has uncritically attributed low bone mass to bone loss, sometimes by using the two terms almost interchangeably. In a young subject, there usually is no basis for attributing a finding of low bone mass to loss. It is, of course, well-established that bone loss occurs in association with menopause, disease, and the multiple concomitants of aging. In an older individual, however, low peak bone mass earlier in adult life may well exaggerate the bone deficit now attributed to loss. Bone loss in adults occurs via the bone remodeling mechanism. Failure to conserve bone during adult life could be due to intrinsic defect(s) in the remodeling IO, its physiologic functioning (e.g., in response to calcium deficiency or skeletal disuse), or some combination of the two. b. The Bone Deficit in Osteoporosis Cortical bone comprises about 80% of the mineralized skeleton of adults, with the remaining 20% present as trabeculae. Disappearance of half the trabeculae in the entire skeleton, therefore, would decrease total-body bone mineral by only about 10%. (In whole human vertebrae, the trabecular proportion is only modestly higher, about 26% [16].) However, bone mass measurements from patients with osteoporosis and

fragility fractures are typically 25 to 40% lower than corresponding values from ostensibly healthy young women [17]. These findings suggest that much of bone deficit in osteoporosis is a deficit of cortical bone. 2. MICROARCHITECTURAL DETERIORATION OF BONE Skeletal fragility out of proportion to low bone mass has been attributed to microarchitectural defects such as loss of trabecular connectivity or accumulation of microdamage. a. Loss of Trabecular Connectivity Loss of trabecular elements would reduce the number of interconnections between trabeculae. This loss of connectivity would explain some of the exaggerated fragility of the skeleton in patients with osteoporosis, because buckling strength varies with the fourth power of the distance between cross-connections in a structure. b. Microdamage in Bone Loading of any structural material is countered by stress, that is, the bend-resisting force within the material. Repeated submaximal loading produces microdamage, which weakens the material. The gravitational environment of Earth, the material properties of bone, the loads placed on the skeleton, and the number of load cycles together indicate that microdamage does occur in the human skeleton during life. Frost has estimated that, without repair of microdamage by the remodeling IO, total skeletal failure would ensue within 2 years [9].

63

CHAPTER 39 Bone Remodeling Findings in Osteoporosis

Figure 4 shows the expected appearance of microdamage in bone: cracks that have extended and propagated according to the direction of stress and the microstructure of the area. However, the study of microdamage in bone has been quite difficult, and we do not know how much of the

skeletal fragility of osteoporosis is due to accumulation of microdamage. c. Ultrasound Evaluation of Bone Bone densitometry measured by the “gold standard” method, dual-energy

FIGURE 4 Light photomicrograph (A) and SEM (B,C) of progressively higher magnification of a microcrack in bovine cortical bone. The specimen had been loaded (flexed) in cycles until there was change in mechanical properties, but a complete fatigue had not yet occurred. The crack propagated along the cement line and was trapped by the oseon.

64

RECKER AND BARGER-LUX

FIGURE 4

X-ray absorptionetry (DXA), cannot yield information on the structural factors that contribute to bone strength (i.e., trabecular connectivity, microdamage, or molecular aspects of bone matrix). Early work with quantitative ultrasound (QUS) of bone suggested that this technology might permit assessment of all the determinants of bone strength [18]. In a comparative study of subjects matched for BMD of the spine or forearm, patellar ultrasound velocity was significantly lower in osteoporotic patients than in normals (see Table 1) [19]. QUS has since been used to evaluate fracture risk in elderly women [20] and to screen for osteoporosis in perimenopausal women [21]. In biomechanical terms, QUS is related to the elastic properties of bone, which should bear some relationship to its microscructure. To date, however, only one published

TABLE 1

(continued )

clinical study has combined ultrasound and histomorphometry to examine trabecular connectivity [22]. In that study, the investigators cut sections from methylmethacrylateembedded specimens and tested the remaining block with a specialized ultrasound device. They reported relationships between in vitro QUS (of trabecular plus cortical bone) and two indices of trabecular connectivity.

C. Bone Histomorphometry Histomorphometry of undecalcified bone biopsies has been the principal method used to study bone remodeling as it relates to osteoporosis. Prior to reviewing the findings, however, two caveats are in order:

Ultrasound Velocitya in Osteoporotic and Normal Subjects Matched for Bone Density Matched by BMD at the

N pairs

Osteoporotic subjects mean  SD (m/s)

Normal subjects mean  SD (m/s)

P value

BV/Tb.N

Spineb

38

1800  15

1854  10

0.005

BV/Sr.V

Forearmc

46

1783  12

1827  14

0.02

a

As AVU (apparent velocity of ultrasound) measured at the patella. Spine BMD by DPA  0.706  0.025 g/cm2 c Forearm BMD by SPA  0.768  0.023 g/cm. b

65

CHAPTER 39 Bone Remodeling Findings in Osteoporosis

1. Studies of iliac crest biopsies have generated nearly all of the published histomorphometric data. While there are correlations between iliac and spinal trabecular bone [23], these relationships are far from perfect. Extrapolation of iliac crest findings to bone at other sites must be made with caution (though, to estimate and illustrate, we do so in the section that follows). 2. Nearly all published data from patients with osteoporosis are cross-sectional. Attempts to understand the longitudinal course of the disease from cross-sectional data will almost certainly be confounded by cohort differences [24]. Despite these cautions, it is worthwhile to look at histomorphometric findings in search of clues.

III. HISTOMORPHOMETRIC FINDINGS IN OSTEOPOROSIS A. Findings in Transilial Bone Biopsies 1. CORTICAL BONE a. Thinner Cortices Table 2 presents results of a comparative study of women with osteoporosis and ostensibly healthy women [25]. The women in the osteoporotic group were older (67 vs 60 years); they were also somewhat shorter (1.56 vs 1.60 m), and they weighed about 20 % less (55.9 vs 70.2 kg; figures are medians). The core width (CW, the distance between the inner and outer periosteal surfaces of the ilium) was less in the women with osteoporosis than in the healthy women (8.4  2.3 mm vs 9.4  1.7 mm, P  0.02), indicating smaller skeletons in the osteoporotic group. In the women with osteoporosis, inner and outer cortices (shown as cortical width Ct.W1 and Ct.W2,

respectively) were about 30 and 37% narrower [25]. (Extrapolated to the entire skeleton, 33% less cortical bone translates into 26% less total-body bone mineral.) This finding is consistent with the statement that low bone mass in osteoporosis represents chiefly a smaller amount of cortical bone. b. Trabeculation of Cortical Bone The pattern of bone loss from the cortex in osteoporosis is from the corticoendosteal surface, where the medullary canal enlarges at the expense of the inner cortex [7]. Bone loss does not occur at the periosteal surface. In transilial biopsies from patients with osteoporosis, resorption cavities within the subendosteal area enlarge and coalesce, resulting in net loss of bone tissue; progressive trabeculation of the cortex through this process creates the transitional envelope identified earlier. The resorption spaces in this zone are not typical Haversian systems. They do not run parallel to the long axis of the bone; instead, from the two-dimensional perspective of microscopic sections, they appear to be oriented haphazardly. The ARF pattern is operative at the corticoendosteal surface; however, formation, though not absent, is incomplete. In the emerging transitional envelope, therefore, the net effect of remodeling is bone loss. As the process continues, the inner cortex becomes completely trabeculated, and the new cortico-endosteal surface is closer to the periosteum. The pattern of cortical thinning and trabeculation of the inner cortex resembles that seen in disuse in the adult remodeling animal [26] (see Fig. 5). The major difference is that the process proceeds at a pace slower than that seen after acute disuse. 2. TRABECULAR BONE Together, the findings described below indicate that the major structural feature of trabecular bone loss in

TABLE 2 Structural Data Osteoporotic subjects N  90a (mean  SD) BV/TV

13.9  4.5

13.42

(mean  SD) 21.2  4.90

P value

20.5

0.001

8.6

0.02

0.586  0.264 (85)

0.517

0.984  0.386

0.861

0.001

Ct.W2

0.383  0.174 (79)

0.315

0.600  0.251 (31)

0.526

0.001

Cn.Wi

7.4  2.3 (79)

7.1

NS

W.Th

28.0  4.44

Tb.Th

124  27

Tb.N

1.12  0.24

Tb.Sp

928  0.218

8.3

7.4 28.2 119 1.09 894

9.4  1.7 (31)

(median)

Ct.W1

C.W

8.4  2.3 (79)

(median)

Normal subjects N  34a

7.8  1.8 (31) 32.1  4.13 138  29 1.55  0.26 654  139

Note. Modified from [25] a Ns which vary from the totals of 90 or 34 are shown in parentheses.

31.9 132 1.59 600

0.001 0.02 0.001 0.001

66

RECKER AND BARGER-LUX

FIGURE 5

Pattern of bone loss in disuse. Unstained, undemineralized cross sections of both humeri of adult beagle. (Left) Cross section from the left forelimb which was immobilized for 40 weeks in a plater cast. (Right) Cross section from the right forelimb which was not immobilized. Note the enlargement of the medullary canal on the left, with no significant difference in the outer circumference. (Courtesy of Prof. J. W. Jaworski, Ottawa.)

osteoporosis is disappearance of entire trabecular elements [25,27]. a. Unremarkable Dynamic Measures Studies from at least three centers have failed to uncover any important differences in the remodeling dynamics of trabecular bone between healthy subjects and patients with established osteoporosis [25,28,29]. As suggested earlier, absence of dynamic evidence of the processes leading to low bone mass means simply that such evidence was no longer present at the time of biopsy. b. Reduction in Trabecular Bone Mass Trabecular bone volume (BV/TV), averages about 35% lower in patients with osteoporosis, as compared to healthy individuals [25]; but there is considerable overlap. However, BV/TV does not give the complete picture of the effects of osteoporosis on trabecular bone. c. Reduction in Trabecular Connectivity Trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular separation (Tb.S) are measures that describe the configuration of trabeculae in space. In the study cited earlier [25], investigators

from our laboratory compared these variables in transilial biopsies from normal subjects and from patients with osteoporosis (see Table 2). Tb.Th was only about 10% lower in the patients, while Tb.N was 28% lower and Tb.S was 30% higher. Specimens taken at autopsy from the vertebral bodies of osteoporotic patients have also shown loss of trabecular connectivity [30]. We also compared the spatial arrangement of trabeculae in osteoporotic patients and normals matched by BV/TV (see Table 3) [31]. The patients were about 10% lower in Tb.N and about 10% higher in Tb.Sp and Tb.Th. However, star volume (Sr.V), a more sensitive measurement of connectivity [32], was about 36% higher in the patients. A recent paper utilized a direct method (“three-dimensional histology”) for studying trabecular discontinuity: the technique involves superficial staining of a thick (300 m) section from each transilial biopsy. By this technique, artifactual trabecular termini (created by transecting trabeculae during sectioning) on the surface take up the stain, but real termini within the section do not. The investigators compared osteopenic, postmenopausal women with and without vertebral fractures, and the fracture group had almost four times as many termini (as number per mm2 of section

67

CHAPTER 39 Bone Remodeling Findings in Osteoporosis

TABLE 3 Spatial Arrangement of Trabeculae in Osteoporotic and Normal Subjects Matched for BV/TV Osteoporotic subjects (mean  SD)

Normal subjects (mean  SD)

P value

BV/Tb.N

1.33  0.21

1.49  0.27

 0.001

BV/Tb.Sp

753  117

686  171

 0.05

BV/Tb.Th

144  27

130  26

 0.05

9.661  1.053

7.128  0.825

 0.03

BV/Sr.V

Note. N  23 pairs; BV/TV  19.1  4.2.

surface) as did the nonfracture group [33]. Conventional histomorphometry of the same biopsies failed to disclose a structural distinction between the fracture and nonfracture groups, which were similar in trabecular mass [34]. The finding that osteoporotic patients with vertebral fractures are lower in trabecular connectivity than other individuals with the same trabecular bone mass [33] probably does explain, at least in part, why some persons with low spine BMD sustain vertebral fractures and others do not. d. Trabecular Thinning Uniform loss of bone from all trabecular surfaces would be relatively easy to document from biopsy data: there would be general thinning of trabeculae. Normal values for wall thickness (W.Th) would suggest excessively deep erosion by osteoclasts. Reduced values for wall thickness would suggest incomplete replacement of eroded bone by the osteoblasts. A combination of these two mechanisms would be harder to detect, but still detectable. Wall thickness is reduced in biopsies from patients with osteoporosis [35], but trabecular thinning is a relatively minor part of the bone loss. Erosion depth, measured directly, is reportedly not increased in patients with osteoporosis [36,37]; however, the measurement itself is technically challenging, and other laboratories have had difficulty obtaining consistent results. Because resorption depth is apparently normal in osteoporosis but wall thickness is reduced, osteoblast work must be reduced. Parfitt et al. [35] have attributed reduced osteoblast work in osteoporosis to defective recruitment of osteoblast teams. 3. REPAIR OF MICRODAMAGE Certain changes in remodeling seen in osteoporotic patients imply inefficient or delayed repair of normally occurring microdamage. These changes might be associated with reduced mechanical competence of bone by allowing the accumulation of an excessive amount of unrepaired microdamage. a. Prolonged Remodeling Periods Prolonged remodeling periods might well be associated with inefficient or

delayed repair of microdamage. Accumulation of microdamage would probably compound skeletal weakness associated with low bone mass and/or deranged microstructure. Some patients with fragility fractures (and some normal persons, as well) do have prolonged remodeling periods combined with normal appositional rates [17]. This combination of findings results from excessive off time, that is, interruption of a remodeling cycle during bone formation; eventually the process resumes and the BMU completes its work. Local changes in remodeling may be even more important. For example, Eventov [38] examined biopsies obtained intraoperatively from the femoral neck of hip fracture patients; the numbers of osteoclasts and osteoblasts in these biopsies were extremely low compared to the numbers in iliac crest biopsies taken at the same time from the same patients. The paper concluded that slow remodeling rates had allowed unrepaired microdamage to accumulate at the hip, contributing to the hip fractures that later occurred. b. Inefficient Detection? It seems likely that the network of living osteocytes and their processes are involved in detection of microdamage. If so, loss of osteocytes or defects in their function could permit unrepaired microdamage to accumulate, undetected. Elucidation of the damage detection system, as well as identification of defects in that system, awaits further research. c. Suppression of Repair? A recent animal study [39] examines the question of whether suppression of bone turnover allows microdamage to accumulate. The investigators gave daily injections of etidronate (0.5 or 5 mg/kg) or vehicle to skeletally mature dogs for 12 months, with X-rays later in the treatment period and harvesting of bone samples from ribs, pelvis, femurs, and spine at its conclusion. At the lower dose, etidronate reduced trabecular bone turnover; at the high dose, activation frequency was zero in both cortical and trabecular bone. Fractures of ribs and/or thoracic spinous processes occurred in nearly all of the dogs in the high-dose group; however, microdamage (as numerical density of microcracks) was not greater at their sites of fracture. The investigators attributed the fractures that occurred to inhibition of mineralization (increased osteoid volume, especially in trabecular bone), rather than to accumulation of microdamage.

B. Findings in Rib Biopsies In comparative studies of rib biopsies from osteoporosis patients and normals, Frost [6,40] found clear, significant differences. In the osteoporotic specimens, appositional and bone formation rates were depressed, and remodeling

68

RECKER AND BARGER-LUX

periods were prolonged. These abnormalities were consistent with inhibited repair of microdamage in cortical bone [42], a situation that could be an important determinant of excess fragility in the appendicular skeleton. Because their cortical shells may be an important determinant of the compressive strength of whole vertebral bodies [37], microdamage in cortical bone could also contribute to excess vertebral fragility.

IV. BONE REMODELING AT THE WHOLE-ORGANISM LEVEL

B. Markers of Resorption and Formation Over the past decade, a number of biochemical markers of bone resorption and bone formation have been studied [e.g., 46,47]. Unlike histomorphometric studies, bone markers reflect remodeling at the whole-organism level; unlike bone kinetics studies, they do not require the extended presence of subjects or the in vivo use of radioactive materials. In a recent study [48], investigators measured the following bone markers in healthy pre- and postmenopausal women and in osteoporotic women before and during treatment with alendronate: Markers of bone resorptiona

Data that reflect bone remodeling at the level of the total skeleton complement the results of histomorphometric studies. Histomorphometric data come from small, local samples that cannot represent the average of all skeletal sites. Calcium kinetics studies and bone markers estimate remodeling rates for the skeleton as a whole, but they cannot provide information about the structural changes that remodeling activities effect.

A. Calcium Kinetics By following the disappearance of 45Ca given intravenously, resorption and accretion of bone mineral for the total skeleton can be calculated. The method requires diets matched to the habitual calcium intake of each subject and complete collection of excreta for an 8 to 12-day in-patient stay. In a recent study employing this method, Heaney and Draper [43] reported that, among 33 healthy women in early postmenopause, resorption exceeded accretion by a mean value of 66 mg/day. In 1968, Heaney [44] had reported results of calcium kinetics studies from a longitudinal, observational study of healthy women. They examined changes in accretion and resorption rates in the 115 subjects who were tested at entry and again about 5 years later and reported that: • When menopause occurred during the interval between studies, changes averaged 56 mg/day in accretion and 80 mg/day in resorption; however, • Changes among subjects who were postmenopausal for both tests were similar to those of subjects who were premenopausal for both tests. The latter finding is evidence that the process that produced the bone loss (here, menopause-related) could no longer be demonstrated. Other work published in 1978 by Heaney et al. [45] was also compatible with the same general conclusion; whereas variation in bone turnover was greater among women with established osteoporosis, their mean values were indistinguishable from those of normals.

CTx NTx fDpd tDpd

C-terminal telopeptide of type I collagen N-terminal cross-linked telopeptide free deoxypyridinoline total deoxypyridinoline

Markers of bone formationb bAP Oc PINP PICP

bone-specific alkaline phosphatase osteocalcin N-terminal propeptide of type I procollagen C-terminal propeptide of type I procollagen

a CTx can be measured in serum or urine; other resorption markers require urine specimens. b All these formation markers require serum specimens.

By measuring the above markers in specimens collected from the premenopausal group on 4 successive days, the investigators calculated both individual coefficient of variation (iCV) and least significant change (LSC) for each assay. The investigators assessed the usefulness of the above markers to classify and follow individual patients by use of these calculations. They reported higher iCVs for the resorption markers (14.6 to 22.3%) than the formation markers (7.2 to 14.4%), a finding that is probably associated with the specimen requirements of the assays. In 15 of 16 patients (one was a false positive) treated with alendronate, a drop in serum CTx (of at least its LSC) after 4 months of treatment correctly predicted a significant gain in BMD (i.e., at least 27 mg/cm2) at 12 months. This study identified serum CTx as “most effective of the markers tested” for longitudinal monitoring of individuals [48]. Many studies have used bone markers to examine differences between groups; however, this paper helps to resolve some uncertainty about the practical usefulness of bone markers to classify and follow individuals.

V. SUMMARY The present state of information allows us to conclude that: • Low bone mass in osteoporosis represents primarily

69

CHAPTER 39 Bone Remodeling Findings in Osteoporosis

a deficit of cortical bone. Bone is lost from the corticoendosteal surface, and eventually the inner cortex becomes completely trabeculated. • Important differences in the remodeling dynamics of trabecular bone between healthy subjects and patients with established osteoporosis are usually absent at the time of biopsy. However, a number of structural findings are of interest. • Trabecular bone mass is probably reduced in most women with osteoporosis, and trabecular thinning is relatively minor. However, osteoporotic patients with vertebral fractures are lower in trabecular connectivity than are others with the same trabecular bone mass. • It is unclear whether inefficient detection of microdamage contributes to fragility fractures. However locally inefficient repair of microdamage (e.g., markedly reduced numbers of osteoclasts and osteoblasts) may well be an important factor. The repair of microdamage is an important issue in both trabecular and cortical bone. • Calcium kinetics studies have shown clearly that, in association with menopause (i.e., estrogen deprivation), the relationship between bone resorption and bone accretion changes in favor of resorption. • Data that describe within-individual magnitude of change (both day-to-day and in response to treatment) permit evaluation of the utility of bone markers to classify and follow individuals.

References 1. Osteoporosis prevention, diagnosis, and therapy (March 27 – 29, 2000). NIH Consensus Statement Online 17(2), 1 – 36 (cited July 27, 2000). 2. E. Seeman, From density to structure: Growing up and growing old on the surfaces of bone. J. Bone Miner. Res. 12, 509 – 521 (1997). 3. M. P. Akhter, D. M. Raab, C. H. Turner, D. B. Kimmel, and R. R. Recker, Characterization of a non-invasive method of in vivo strain application. J. Biomech. 25, 1241 – 1246 (1992). 4. D. B. Burr, M. B Schaffler, K. H. Yang, M. Lukoschek, N. Sivaneri, J. D. Blaha and E. L. Radin, Skeletal change in response to altered strain environments: Is woven bone a response to elevated strain? Bone 10, 223 – 233 (1989). 5. C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 6. H. M. Frost, “Bone Remodeling and Its Relationship to Metabolic Bone Diseases,” Vol. 3. Thomas, Springfield, IL, 1973. 7. N. M. Keshawarz and R. R. Recker, Expansion of the medullary cavity at the expense of cortex in postmenopausal osteoporosis. Metab. Bone Dis. Relat. Res. 5, 223 – 228 (1984). 8. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone in Nutritional Perspective” Thomas, Springfield, IL, (1970). 9. H. M. Frost, “Intermediary Organization of the Skeleton,” Vols. 1 and 2 CRC Press, Boca Raton, FL, (1986). 10. H. M. Frost, Dynamics of bone remodeling. In “Bone Biodynamics” (H. M. Frost, ed.), pp. 315 – 333. Little, Brown, Boston, (1964). 11. R. R. Recker, D. B. Kimmel, A. M. Parfitt, K. M. Davies, N. Keshawarz, and S. Hinders, Static and tetracycline-based bone histomorphometric

12.

13.

14.

15.

16.

17.

18.

19. 20.

21.

22.

23.

24. 25.

26. 27.

28.

29.

30.

31.

data from 34 normal postmenopausal females. J. Bone Miner. Res. 3, 133 – 144 (1988). H. M. Frost, Analysis of trabecular bone dynamics. In “Bone Histomorphometry: Techniques and Interpretation” (R. R. Recker, ed.), pp. 109 – 142. CRC Press, Boca Raton, FL (1983). J. L. Ivey and D. J. Baylink, Postmenopausal osteoporosis: Proposed roles of defective coupling and estrogen deficiency. Metab. Bone Dis. Relat. Res. 3, 3 – 7 (1981). R. P. Heaney, The bone-remodeling transient: Implications for the interpretation of clinical studies of bone mass change. J. Bone Miner. Res. 9, 1515 – 1523 (1994). D. M. Black, S. R. Cummings, H. K. Genant, M. C. Nevitt, L. Palermo and W. Browner, Axial and appendicular bone density predict fractures in older women. J. Bone Miner. Res. 7, 633 – 638 (1992). S. Y. Nottestad, J. J. Baumel, D. B. Kimmel, R. R. Recker, and R. P. Heaney, The proportion of trabecular bone in human vertebrae. J. Bone Miner. Res. 2, 221 – 229 (1987). B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L.G. Melton, III, Changes in bone mineral density of the proximal femur and spine with aging. J. Clin. Invest. 70, 716 – 723 (1982). R. P. Heaney, L. V. Avioli, C. H. Chesnut, J. Lappe, R. R. Recker, and G. H. Brandenburger, Osteoporotic bone fragility: Detection by ultrasound transmission velocity. J. Am. Med. Assoc. 261, 2986 – 2990 (1989). D. B. Kimmel, Unpublished observations (1993). C. C. Gluer and The International Quantitative Ultrasound Consensus Group, Quantitative ultrasound techniques for the assessment of osteoporosis: Expert agreement on current status. J. Bone Miner. Res. 8, 1280 – 1288 (1997). J. Nairus, S. Ahmadi, S. Baker, and D. Baran, Quantitative ultrasound: An indicator of osteoporosis in perimenopausal women. J. Clin. Densitom. 2, 141 – 147 (2000). J. E. Zerwekh, P. O. Antich, S. Mehta, K. Sakhaee, F. Gottschalk, and C. Y. Pak, Reflection ultrasound velocities and histomorphometric and connectivity analyses: Correlations and effect of slow release sodium fluoride. J. Bone Miner. Res. 12, 2068 – 2075 (1997). L. Mosekilde and L. Mosekilde, Iliac crest trabecular bone volume as predictor for vertebral compressive strength, ash density and trabecular bone volume in normal individuals. Bone 9, 195 – 199 (1988). R. R. Recker and R. P. Heaney, Letter to Editor. J. Bone Miner. Res. 5, 307 – 308 (1990). D. B. Kimmel, R. R. Recker, J. C. Gallagher, A. S. Vaswani, and J. F. Aloia, A comparison of iliac bone histomorphometric data in postmenopausal osteoporotic and normal subjects. Bone Miner. 11, 217 – 235 (1990). Z. F. G. Jaworski and H. K. Uhthoff, Reversibility of nontraumatic disuse osteoporosis during its active phase. Bone 7, 432 – 439 (1986). A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. J. Clin. Invest. 7, 1396 – 1409 (1983). M. E. Arlot, P. D. Delmas, D. Chappard, and P. J. Meunier, Trabecular and endocortical bone remodeling in postmenopausal osteoporosis: Comparison with normal postmenopausal women. Osteoporosis Int. 1, 41 – 49 (1990). T. Steiniche, P. Christiansen, A. Vesterby, C. Hasling, R. Ullerup, L. Mosekilde, and F. Melsen, Marked changes in iliac crest bone structure in postmenopausal osteoporotic patients without any signs of disturbed bone remodeling or balance. Bone 15, 73 – 79 (1994). A. Vesterby, Star volume of marrow space and trabeculae in iliac crest: Sampling procedure and correlation to star volume of first lumbar vertebra. Bone 11, 149 – 155 (1990). R. R. Recker, Architecture and vertebral fracture. Calcif. Tissue Int. 53, S139 – S142 (1993).

70 32. A. Vesterby, H. J. Gundersen, and F. Melsen, Star volume of marrow space and trabeculae of the first lumbar vertebra: Sampling efficiency and biological variation. Bone 10, 7 – 13 (1989). 33. J. E. Aaron, P. A. Shore, R. C. Shore, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. II. Three-dimensional histology. Bone 27, 277 – 282 (2000). 34. L. Hordon, M. Raisi, J. E. Aaron, J. E. Paxton, S. K. Beneton, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. I. Twodimensional histology. Bone 27, 271 – 276 (2000). 35. A. M. Parfitt, A. R. Villanueva, J. Foldes, and S. D. Rao, Relations between histologic indices of bone formation: Implications for the pathogenesis of spinal osteoporosis. J. Bone Miner. Res. 10, 466 – 473 (1995). 36. E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodeling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption, and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990). 37. M. E. Cohen-Solal, M-S. Shih, M. W. Lundy, and A. M. Parfitt, A new method for measuring cancellous bone erosion depth: Application to the cellular mechanisms of bone loss in postmenopausal osteoporosis. J. Bone Miner. Res. 6, 1331 – 1338 (1991). 38. I. Eventov, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodelling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 12, 1 – 6 (1991). 39. T. Hirano, C. H. Turner, M. R. Forwood, C. C. Johnston, and D. B. Burr, Does suppression of bone turnover impair mechanical proper-

RECKER AND BARGER-LUX

40.

41.

42.

43.

44. 45. 46.

47.

48.

ties by allowing microdamage accumulation? Bone 27, 13 – 20 (2000). S. Jett, K. Wu, and H. M. Frost, Tetracycline-based histological measurement of cortical-endosteal bone formation in normal and osteoporotic rib. Henry Ford Hosp. Med. J. 15, 325 – 344 (1967). H. M. Frost, Transient steady state phenomena in microdamage physiology: A proposed algorithm for lamellar bone. Calcif. Tissue Int. 44, 367 – 381 (1989). A. Vesterby, L. Mosekilde, H. J. Gundersen, F. Melsen, L. Mosekilde, and K. Holme, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991). R. P. Heaney and M. W. Draper, Raloxifene and estrogen: Comparative bone-remodeling kinetics. J. Clin. Endocrinol. Metab. 82, 3425 – 3429 (1997). R. P. Heaney, Kinetic studies of calcium in metabolic bone disease. Clin. Endocrinol. 2, 309 – 323 (1968). R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in bone remodeling. J. Lab. Clin. Med. 92, 964 – 970 (1978). P. Garnero, W. J. Shih, E. Gineyts, D. B. Karpf, and P. D. Delmas, Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J. Clin. Endocrinol. Metab. 79, 1693 – 1700 (1994). R. Eastell, S. P. Robins, T. Colwell, A. M. A. Assiri, B. L. Riggs, and R. G. G. Russell, Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporosis Int. 3, 255 – 260 (1993). E. Fink, C. Cormier, P. Steinmetz, C. Kindermans, Y. Le Bouc, and J-C. Souberbielle, Differences in the capacity of several biochemical bone markers to assess high bone turnover in early menopause and response to alendronate therapy. Osteoporosis Int. 11, 295 – 303 (2000).

CHAPTER 40

The Role of Parathyroid Hormone and Vitamin D in the Pathogenesis of Osteoporosis JOHN P. BILEZIKIAN* AND SHONNI J. SILVERBERG† Departments of Medicine*† and Pharmacology,* College of Physicians and Surgeons, Columbia University, New York, New York 10032

I. Introduction II. Vitamin D and Osteoporosis III. Parathyroid Function in Osteoporosis

IV. Summary References

I. INTRODUCTION

relationship between bone resorption and formation [1]. Any imbalance favoring resorption over formation will lead to bone loss. The time line of this slow but inexorable process predicts that if a woman or man lives long enough, she or he will experience the disorder we call osteoporosis. Viewed in this context, osteoporosis is an intrinsic outcome of the aging process. Also intrinsic to the aging process are changes in the synthesis, metabolism, and responsiveness of vitamin D and parathyroid hormone. It is possible that the age-associated changes in vitamin D and parathyroid hormone are causally related to the age-associated changes in bone mass. On the other hand, some of the hormonal changes may be adaptive, serving to protect the aging skeleton from further weakening. Superimposed upon the age-related decline in bone mass is a set of other pathophysiological challenges, such as

Many factors contribute to the bone loss that characterizes the syndrome of osteoporosis. This chapter focuses primarily upon the two major calcium-regulating hormones, parathyroid hormone and vitamin D. Evidence has accumulated slowly in support of their potential involvement in the development of osteoporosis. This chapter reviews provocative literature that could accomodate hypotheses implicating these two hormones. Among the panopoly of functions that parathyroid hormone and vitamin D serve (see Chapters 7 and 9) none is more important than to maintain normal bone remodeling. After peak bone mass has been achieved and after a short period of neutral bone balance, most adult life is believed to be associated with a loosening of a rather tightly coupled

OSTEOPOROSIS, SECOND EDITION VOLUME 2

71

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

72

BILEZIKIAN AND SILVERBERG

estrogen deficiency, that add to the risk of osteoporosis in a given individual. Altered relationships between the calcium-regulating hormones and bone resulting from estrogen deficiency, for example, could lead to very different effects of these hormones on bone balance. Thus, the actions of parathyroid hormone on bone might be influenced by the presence or absence of estrogens. This chapter considers potential roles of vitamin D and parathyroid hormone in the context of the aging process per se and in association with selected other changes (e.g., estrogen deficiency). The classical metabolic bone diseases associated with gross deficiences or excesses of vitamin D or parathyroid hormone are not covered in this chapter.

II. VITAMIN D AND OSTEOPOROSIS Impaired calcium absorption from the gastrointestinal tract due to abnormalities at any point in the synthesis and metabolism of, or responsiveness to, vitamin D underlie the putative role of vitamin D in the development of osteoporosis. Possible abnormalities include vitamin D deficiency, abnormal production of 25-hydroxyvitamin D (25OHD) or 1,25-dihydroxyvitamin D (1,25(OH)2D), the active form of the vitamin, altered intestinal vitamin D receptor number or sensitivity, or acquired resistance to vitamin D (see Chapter 9). Individuals deficient in calcium as a result of these possible abnormalities will be at greater risk for an imbalance in the dynamic interplay between bone formation and bone resorption. The clinical consequences of these abnormalities in the context of osteoporosis differ from the classical disorder associated with overt vitamin D deficiency, namely osteomalacia. Most, but not all studies [2,3] report a fall in the circulating concentration of 1,25(OH)2D with advancing age [4 – 7]. Osteoporotic individuals have circulating 1,25(OH)2D levels that are even lower [4,8]. Reduced concentration of calcitriol thus is implicated as a cause of age-associated reductions in bone mass and of further reduction in bone mass in osteoporosis. Since 1,25(OH)2D concentrations reflect the availability of precursors and metabolic processes that lead to its formation and metabolism, a number of specific factors could account for reductions in circulating levels (Table 1).

A. Vitamin D Deficiency The early literature describing vitamin D deficiency came mainly from the United Kingdom where subclinical vitamin D deficiency has long been recognized to be a factor in the development of some forms of osteoporosis. In the past 25 years, reports have come from around the globe as well [9]. Inadequate supply of vitamin D from its two sources, diet and sunlight, leads to vitamin D deficiency. Vitamin D is

TABLE 1 Pathogenesis of Osteoporosis: Possible Roles of Vitamin D I. Vitamin D deficiency A. Dietary deficiency B. Insufficient ultraviolet B exposure II. Altered vitamin D metabolism A. Defect in the renal 1-hydroxylase B. Intestinal vitamin D receptor abnormality 1. Decreased receptor number 2. End organ resistance

derived from dietary sources and by supplementation in milk in the United States. It is not found as a supplement in other dairy products. Adult Americans obtain approximately 75 – 100 IU per day in their diets [10], a value that is below the former guidelines of 400 IU per day. Even this figure, which is generally not met, probably underestimates physiological requirements set at the current time by the U.S. Food and Nutrition Board to be 600 – 800 IU per day (11). In addition to possible reductions in dietary vitamin D intake with advancing years, its absorption from the gastrointestinal tract is affected by age. In women, a decline of up to 40% in absorption of vitamin D in the distal ileum has been reported to occur with advancing age [12]. Whether women who develop osteoporosis demonstrate even further reductions in dietary vitamin D absorption than expected for their age is not established (13 – 16). The other source of vitamin D is the skin, where ultraviolet B energy of 290 – 315 nm converts 7-dehydrocholesterol to previtamin D [17]. 7-Dehydrocholesterol levels in the skin fall by approximately 50% between 20 and 80 years of age [18]. It is not known whether the reduced quantity of skin substrate is exaggerated in those affected by osteoporosis. It is also not known whether the thermal-dependent conversion of previtamin D to vitamin D is impaired with aging or whether osteoporotic subjects are more deficient in this step. In addition to reduced levels of 7-dehydrocholesterol with aging, a comparison of vitamin D status from several global regions shows seasonal variation in concentrations of 25-hydroxyvitamin D. If dietary sources of and absorption of vitamin D do not vary substantially over a short period of time, from year to year, seasonal variations of 25-hydroxyvitamin D could reflect changes in UV exposure. In winter months at northern latitudes, decreased exposure of skin to UV sunlight of correct wavelength can lead to vitamin D insufficiency. It has long been assumed [19,20] that vitamin D deficiency was unique to individuals [21 – 26] living in northern latitudes in whom dietary sources and food supplements were inadequate. This was not considered to be common in the United States because of its generally lower latitude compared to countries where vitamin D deficiency has

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D

been a problem and because of the fortification of milk with vitamin D. More recent studies have suggested that subclinical vitamin D deficiency may indeed play a role in bone loss in postmenopausal women in the United States. In Maine (45.5ø northern latitude) and Boston, Massachusetts (42.2ø northern latitude), no synthesis of previtamin D occurs in the skin during the late fall and winter months [27,28]. Bone density also exhibited a steep decline during those months, rising again during the summer [28]. The rise in bone density was associated with resolution of an accompanying secondary hyperparathyroidism. It did not, however, compensate for the bone loss in winter. Subclinical vitamin D deficiency has been shown in subpopulations of osteoporotic American women. Vallareal et al. reported low concentrations (less than 38 nmol/liter) of 25hydroxyvitamin D in 9% (49/539) of women referred for osteoporosis screening in St. Louis, Missouri [29]. These women had no symptoms of vitamin D deficiency, but they did have lower mean vertebral bone mineral density, lower serum calcium and phosphorus concentrations, and higher circulating parathyroid hormone and urinary calcium excretion than vitamin D-sufficient patients. Multivariate analysis suggested that the secondary hyperparathyroidism induced by the vitamin D-deficient state was a predominant factor leading to low bone density. Studies that have attempted to implicate subclinical vitamin D deficiency in the development of osteoporosis have depended upon measurement of circulating 25-hydroxyvitamin D. Since this is believed to be an accurate representation of the storage form of the vitamin, such measurements are useful. However, Peacock et al. suggested that a 25hydroxyvitamin D value of 50 nmol/L may be necessary to ensure sufficient 1,25(OH)2D production [30]. Thus, the standard published values for the “normal range” (25 – 125 nmol/L or 10 – 50 ng/mL) includes individuals who should really be regarded as having subnormal levels of 25hydroxyvitamin D. In an older population entering a New York nursing home, 86% had 25-hydroxyvitamin D levels below 50 nmol/L. In 41% of the group, 25-hydroxyvitamin D concentrations were below 25 nmol/L [31]. The bone densities among subjects with lower levels of 25hydroxyvitamin D were below the mean for age- and gender-matched controls. Gloth et al. also reported that vitamin D deficiency is found commonly in the elderly with estimates ranging from 38 to 54% (32,33), and Thomas reported that among an unselected free-living population living in the metropolitan area of Boston, over 50% had evidence for vitamin D deficiency (34).

B. Altered Vitamin D Metabolism There is no evidence that the liver loses its rather impressive capacity to convert vitamin D to 25-hydroxyvitamin D

73 with aging or in osteoporosis. When deficiencies in 25hydroxyvitamin D are observed, they are due either to limited availability of the substrate, vitamin D, to drug-related altered vitamin D metabolism [35,36], or to severe, advanced liver disease. One exception in this regard is primary biliary cirrhosis, a chronic liver disease in which there appears to be an impaired 25-hydroxylating ability that is disproportionately greater than the loss of functioning hepatic mass [37]. In contrast to the rather limitless capacity of the liver to form 25-hydroxyvitamin D, the kidney controls the formation of the active metabolite, 1,25(OH)2D, exquisitely well. Even in the setting of vitamin D toxicity, the kidney effectively controls excessive production of this active metabolite [38]. Factors that help to regulate the renal enzyme responsible for the formation of 1,25(OH)2D are parathyroid hormone, phosphorus, calcium, and 1,25(OH)2D itself (see Chapter 9). A decline in the ability of the kidney to form this active metabolite occurs with age and has been implicated as a possible mechanism for age-related osteoporosis [39]. Slovik et al. demonstrated that older individuals respond to the stimulating effects of parathyroid hormone on 1,25(OH)2D production less well than young normal controls [40]. This landmark study did not compare agematched subjects with and without osteoporosis and thus could not implicate a specific defect in osteoporosis. Nevertheless, the observations did support the notion that renal responsiveness to the stimulatory effects of parathyroid hormone is reduced with age. A possible specific defect in renal hydroxylating capacity was studied by Riggs et al., who infused 400 IU of parathyroid hormone on 3 consecutive days to osteoporotic and age-matched nonosteoporotic subjects. The results showed no differences between the two groups in the 1,25(OH)2D response to parathyroid hormone infusion [41]. Tsai et al. extended these observations by studying four different groups of women in a protocol that tested the possibility of an age-related decline in renal hydroxylating capability as well as a specific defect in osteoporotic individuals [7]. Older subjects had lower baseline 1,25(OH)2D concentrations and responded less well to the stimulatory effects of parathyroid hormone on formation. The data correlated best with diminishing responsiveness to parathyroid hormone as a function of declining renal function. Osteoporotic women responded less well than the age-matched nonosteoporotic group but their renal function was also substantially lower. Thus, it was not possible to conclude that the osteoporotic women had a significantly more sluggish response to parathyroid hormone than their nonosteoporotic counterparts. These studies support the generally accepted notion that the renal capacity to form 1,25(OH)2D diminishes with age. They do not support the idea that in osteoporotic subjects this age-related decline in 1—hydroxylating capacity is further compromised.

74

BILEZIKIAN AND SILVERBERG

The decline in renal hydroxylating capacity associated with advancing age could be due to the aging process per se or to declines in renal function. Halloran et al. [42] elegantly addressed this question, at least in men, by showing that the impaired responsiveness of the renal enzyme is due exclusively to a decline in renal function. Only in the sense that advancing age and declining renal function are mutually associated can the decline in renal enzymatic activity be said to be a function of age. In Halloran’s study of healthy 70-year-old men with completely normal endogenous creatinine clearance, the ability of the kidney to convert 25-hydroxyvitamin D to 1,25(OH)2D was normal. It can be concluded, therefore, that one pathophysiologic mechanism that accounts for reduced synthesis of 1,25(OH)2D is an age-related decline in renal function. One could advance the idea that those individuals who are able to maintain renal responsiveness do so either by being fortunate enough to maintain normal renal function or, more likely, by developing mechanisms to accomodate this agerelated decline (see below).

C. Altered Vitamin D Sensitivity In addition to possible changes in the quantity of vitamin D precursors and the capacity to generate the active metabolite of vitamin D, other hypotheses related to the role of vitamin D in the pathogenesis of osteoporosis focus on the reduced sensitivity to 1,25(OH)2D in the small intestine. Such Intestinal resistance is suggested to be primary alteration [2]. One theory proposes that osteoporosis is associated with an acquired alteration in binding of 1,25(OH)2D to its receptors in the small intestine [43]. Another attributes the apparent reduction in intestinal sensitivity to an age-related reduction in intestinal vitamin D receptor concentration, which has been observed in intestinal biopsies obtained from women spanning the wide age range of 20 – 87 years [44]. If vitamin D receptor concentrations are limiting, lower receptor abundance should lead to impaired calcium absorption. Presumably, this mechanism could not be satisfactorily overcome by increased calcium intake. In support of this hypothesis, Gennari et al. have shown impaired 1,25(OH)2D stimulation of fractional calcium absorption [45]. More recently, Pattanaungkul et al. revisited this issue in a protocol that correlated 1,25(OH)2D concentrations in young and elderly subjects with fractional fasting calcium absorption [46]. In the young subjects (mean age 28.7  5.3 years), fractional calcium absorption correlated significantly with the free 1,25(OH)2D index (r  0.63, P  0.01). In the elderly group, however (mean age 72.5  3.0 years), fractional calcium absorption was not significantly related to the free 1,25(OH)2D level. Moreover, the slopes of the relationship between these two indices were significantly greater in the young as than in

the elderly. These observations, therefore, support the idea that the elderly are relatively resistant to the physiological actions of 1,25(OH)2D on intestinal calcium absorption.

D. Genetic Variations in the Vitamin D Receptor (VDR) Genetic polymorphisms of VDR have been implicated in the development of osteoporosis [47] (see Chapter 26). In several studies a particular allele of the vitamin D receptor, as characterized by several restriction enzyme patterns, was associated with reduced bone mass [47,48]. Homozygous carriers of the allele designated “B” were found to have lower bone mass than homozygous carriers of the allele designated “b”. These studies remain controversial because they have not been confirmed in other investigations (49). Moreover, it has been very difficult to “link” identified VDR polymorphisms to a physiological abnormality in a system regulated by 1,25(OH)2D. These studies are nevertheless important because a predisposition to osteoporosis on the basis of vitamin D genetics may still be confirmed. More important, however, is the fact that these studies literally spawned an explosive search for genes that might contribute to osteoporosis. It is clear from these other studies that osteoporosis is a complex genetic trait, with the likelihood that a number of genes are going to be shown to be important in the overall disposition to this disease (50).

E. Overall View of Mechanisms of Vitamin D Alterations in Osteoporosis The importance of the various mechanisms by which alterations in the vitamin D system can lead to osteoporosis is unclear at this time. Any or all of the aforementioned potential pathophysiologic considerations could contribute to bone loss with aging and could be further deranged in osteoporosis. Hypotheses relating alterations in vitamin D to osteoporosis are based upon the idea that calcium absorption is ultimately impaired. There are other potential roles for vitamin D in bone health that could be important in the development of osteoporosis. If vitamin D is ultimately demonstrated to be important in the events leading to osteoporosis, its role is unlikely to be similar to its role in the events leading to osteomalacia, the principal clinical result of gross vitamin D deficiency. Alternatively, abnormalities in the vitamin D system, some of which have been demonstrated to occur with aging, may not be causative in the development of osteoporosis. Rather, these alterations in the vitamin D system could require compensatory adaptations in other systems, such as those related to parathyroid hormone. Osteoporosis could

75

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D

result if the adaptive responses are inadequate, or if they themselves promote increased bone turnover. One possible compensatory mechanism is parathyroid hormone responsiveness, a subject that is covered in the next section of this chapter.

III. PARATHYROID FUNCTION IN OSTEOPOROSIS There is now general agreement that parathyroid function increases with advancing age. Initially this view was somewhat controversial due, in part, to the fact that declining renal function with age leads to an expected secondary increase in circulating parathyroid hormone. These secondary increases stemming directly from declining renal function are believed to be caused by several factors. Declining 1,25(OH)2D concentration as a function of impaired renal hydroxylating ability (see above) would relieve the inhibitory effects of this metabolite on the parathyroid hormone gene. They would also lead to a reduction in calcium absorption. Second, a very small but physiologically significant reduction in serum calcium concentration would be associated with slight increases in those of phosphorus resulting from declining renal function. These factors together would stimulate parathyroid hormone secretion. Although this view is entirely reasonable, it was hard to prove that parathyroid hormone concentrates actually increase because early generation assays for the hormone detected primarily circulating inactive hormone fragments. Such fragments normally accumulate when renal function declines. Thus, the increases in detectable parathyroid hormone may not have been due to the metabolic sequences described above, but merely to the reduced clearance of inactive hormone fragments. Immunoradiometric (IRMA) and immunochemiluminometric (ICMA) assays, which selectively measure circulating intact parathyroid hormone, as well as a large, inactive fragment [51] have helped to settle the point [52]. Biologically active concentrates of parathyroid hormone increase normally with age. Although this increase clearly reflects declining renal function, elevated values also occur even in older individuals who have no apparent decline in renal function. This point is illustrated well by Halloran et al. [42], who studied young and elderly men with normal renal function. Despite normal serum ionized calcium activity, serum 1,25(OH)2D, and urinary calcium excretion, basal serum parathyroid hormone was 1.5-fold higher in the elderly men than in the younger men. The increase in parathyroid hormone may be associated with a similar rise in bone turnover as assessed both by bone markers (53 – 56) and by histomorphometric indices (56 – 58). On the other hand, Gallagher et al. [59] could not show any increase in parathyroid hormone with age in a study of over 700 subjects.

FIGURE 1

Serum parathyroid hormone concentrations as a function of age among an age-stratified sample in Rochester, Minnesota. Men (solid line, squares) and women (dashed line, circles) both show an increase with age. The correlation with age was 0.30 for men and for women (P  0.001). Adapted from Khosla et al. [60] with permission.

Most studies have demonstrated the increases in serum parathyroid hormone concentration as a continuous relationship with age (60; Fig. 1) but more detailed analysis has shown that the major increase in circulating parathyroid hormone occurs in women greater than 70 years old. In fact, Koh et al. [61] showed a 1.9% decline per decade between 55 and 69 years. Between 70 and 81 years, there was a 3.8% increase per decade. Similarly, Prince et al. [62] showed that the age-related increase in parathyroid hormone becomes evident about 20 years after the menopause (Fig. 2), approximately 70 years of age. There are two fundamental hypotheses to explain the increase in parathyroid hormone concentrations with age. One links, in a causal way, the increase in parathyroid hormone to age-related bone loss (Table 2). An alternative hypothesis links the increase in hormone concentration to protection against age-related bone loss. This section will review the data in support of each of these hypotheses.

FIGURE 2

Serum parathyroid hormone concentrations in women as a function of time since menopause. Results with different letters are significantly different (P  0.05). Adapted from Prince et al. [62] with permission.

76

BILEZIKIAN AND SILVERBERG

TABLE 2 Changes in Serum PTH and in Biochemical Markers of Bone Turnover in 304 Women Residents of Rochester, MN, from the Third into the Tenth Decade of Life Spearman correlation coefficients

Increase with age (%)

vs age

vs PTH

PTH

54

0.354*

1.00

BSAP

38

0.329*

0.192†

OC

64

0.392*

0.206†

76

*

0.203†

*

0.190†

Variable

fPYD NTx

86

0.505

0.344

Note. BSAP, bone specific alkaline phosphatase; OC, osteocalcin; fPYD, urinary free pyridinoline; NTx, urinary N-telopeptide of type I collagen. Adapted from Riggs et al. [76] with permission. * P  0.0001. † P  0.001.

A. Parathyroid Hormone as a Contributing Factor to Osteoporosis 1. ALTERED PARATHYROID HORMONE SENSITIVITY The accelerated rate of bone loss in the early postmenopausal years has been explained by the local release of bone-resorbing cytokines [63,64]. Riggs and Melton [65] proposed that such local factors and the ensuing rapid loss of skeletal calcium could actually lead to suppression of parathyroid hormone. Reduced parathyroid hormone concentrations in osteoporotic women in the face of higher bone turnover could reflect enhanced skeletal sensitivity [66,67] to these local bone-resorbing factors. Although some studies have shown parathyroid hormone suppression in the early postmenopausal years, more often parathyroid hormone concentrations have not differed from those of age-matched controls. Despite normal circulating concentrations, parathyroid hormone could still induce bone loss if sensitivity is heightened in the postmenopausal state. One point in support of this idea comes from the literature on primary hyperparathyroidism, a clinical disorder seen often in women during the first decade after the menopause [68]. It seems that the hyperparathyroid process is unmasked in these women as a function of declining estrogen status. The reduced estrogen levels could also enhance skeletal sensitivity to parathyroid hormone. However, there are no prospective data showing that in these women, primary hyperparathyroidism existed in a subclinical form before menopause. When postmenopausal women with primary hyperparathyroidism are given estrogen, it is generally accepted that calcium concentrations decline somewhat without any change in the circulating concentration of parathyroid

hormone [69,70]. In the setting of the normocalcemic individual, however, McKane et al. [71] have demonstrated that within 6 months of estrogen replacement therapy in early postmenopausal women, and despite no change in circulating serum calcium concentration, parathyroid hormone concentrations increase by 38%. In the aggregate, these observations argue for an inhibitory effect of estrogen on parathyroid hormone action in bone, for enhanced sensitivity to the skeletal effects of parathyroid hormone when women become estrogen deficient, and for a physiological need for increased circulating parathyroid hormone when estrogen is provided to the early postmenopausal woman. The notion of enhanced skeletal sensitivity to parathyroid hormone is not new. Since the mid-1960s, this mechanism has been considered a possible explanation for the development of postmenopausal osteoporosis [72 – 74]. Kotowicz et al. [67] obtained histomorphometric data suggesting that in postmenopausal osteoporosis, the resorptive effects of parathyroid hormone are enhanced. For each picomole per liter rise in circulating parathyroid hormone, osteoporotic women had higher activation frequency (1.3% per year), bone resorption rate (3.9% per year), and cancellous bone loss (2.8% per year). On the other hand, enhanced sensitivity to parathyroid hormone has not been universally demonstrated. Tsai et al. [75] reported no difference between osteoporotic and normal women in urinary excretion of calcium or hydroxyproline in response to 400 units per day of bovine parathyroid hormone. Interpretation of the studies by Tsai et al. was limited by the pharmacologic amounts of hormone used and the relative insensitivity of the markers that were used to monitor the effect. Nevertheless, Ebeling et al. [66] reported similar results when calcium deprivation was used to stimulate endogenous parathyroid hormone secretion. The concept of parathyroid hormone as a pathophysiological culprit in postmenopausal osteoporosis has been promulgated even more boldly by Riggs and associates [76]. They recently proposed a “unitary model for involutional osteoporosis” according to which not only is parathyroid hormone involved in the early postmenopausal effects of estrogen deficiency (a relative suppression) but increases in parathyroid hormone also become important later in the setting of age-related bone loss. They proposed that in the second phase of bone loss (type II or age-related bone loss), the actions of estrogen deficiency predominate in nonskeletal sites, such as the intestine and kidney. Such extraskeletal actions would promote increased urinary calcium loss and reduce gut calcium absorption. The resulting secondary increase in parathyroid hormone is key in this explanation of the continued loss of bone mass with aging. Although Riggs et al. cite a number of studies in support of this hypothesis [77 – 80], our knowledge is admittedly still limited. We know little of the extraskeletal actions of estrogens with

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D

regard to renal calcium handling and to gastrointestinal absorption of calcium. Also vexing is the hypothesis that only late-term estrogen deficiency is associated with an increase in parathyroid hormone. If one can account for these two separate actions of estrogens (skeletal vs extraskeletal] that are dichotomous in time, the secondary increase in parathyroid hormone occurring in this later phase would help to account for the more apparent cortical bone loss and an amelioration of the earlier rapid cancellous bone loss. Such observations gain support from data on primary hyperparathyroidism in which increased parathyroid hormone concentrations are associated with preservation of cancellous bone at the expense of cortical bone [81]. In fact, this and other observations have led many investigators to consider parathyroid hormone a potential therapy for postmenopausal osteoporosis [82] (see Chapter 77). Other confounding elements to the hypothesis put forth by Riggs et al. are noted by Bilezikian [83]. Nevertheless, circumstantial evidence does implicate parathyroid hormone in the pathogenesis of age-related osteoporosis. Positive correlations have been made between increases in parathyroid hormone and markers of bone turnover in elderly women [84,85]. As shown in Table 2, these correlations are modest, at best, with correlation coefficients generally in the neighborhood of 0.20, well below the more robust association of bone markers to age, per se. In support of a causal relationship, Ledger et al. showed that urinary collagen N-telopeptide excretion is suppressed to the young normal range after a 24-h calcium infusion [86]. Further evidence that argues for a negative effect of rising parathyroid hormone concentration with age comes from attempts to correlate parathyroid hormone with bone loss. Using peripheral quantitative computed tomography to distinguish cancellous from cortical elements [87], Boonen et al. showed a negative correlation between cortical bone loss and rising parathyroid hormone levels. Ledger et al. also showed that elevated parathyroid hormone concentrations in the elderly can be reduced to levels seen in young normals by administration of 1,25(OH)2D [88]. While these data in the aggregate argue for a role for parathyroid hormone in the pathogenesis of age-related osteoporosis, their indirect nature argues for caution in establishing a causal link at this time.

B. Parathyroid Hormone as a Protective Influence 1. ALTERED PARATHYROID HORMONE RESPONSIVENESS The observations reviewed above are consistent with a negative effect of parathyroid hormone on bone. Increased normal secretion or enhanced sensitivity to parathyroid

77 hormone could promote to bone loss on this basis. In contrast, it is possible that parathyroid hormone responsiveness is important to maintain bone health and that in osteoporosis this responsiveness is lost. In this context, which presents a completely different view, reduced responsiveness of the parathyroid glands contributes to the development of osteoporosis. Altered responsiveness could underlie changes in circadian rhythmicity of parathyroid hormone secretion. Daily parathyroid hormone secretion follows a biphasic profile with peaks at approximately 1800 and 0200 hours [89 – 91]. Presumably, the larger nocturnal peak represents a compsensation for mild hypocalcemia induced by night-time fasting. Calvo et al. reported that women exhibited a blunted parathyroid hormone peak concentration relative to that of men and, subsequently, a less dramatic decline in night-time urinary calcium excretion. Night-time urinary calcium excretion declined in men by 34%, whereas in women, it decreased by only 17%. This nocturnal calcium wasting could, over the years, contribute to the greater bone loss seen in women. Postmenopausal osteoporotic women showed a further blunting of their nocturnal parathyroid hormone peak, with no change in nocturnal fractional excretion of calcium [92]. The inefficient renal calcium conservation thus documented could contribute to the osteoporotic process. The cause of this blunted parathyroid hormone response to nocturnal fasting is unknown. More sophisticated pulsatility studies by Prank and colleagues [93] have shown that osteoporotic women demonstrate poorly predictable time series of pulses and patterns of parathyroid hormone secretion. Creating a discriminating statistic by fitting a time series model to pooled data from normal subjects, normal and osteoporotic subjects could be distinguished from each other [93,94]. In contrast, Samuels and colleagues [95] could not demonstrate any differences in amplitude or frequency, or pulsatile parathyroid hormone secretory parameters between osteoporotic and normal subjects. The lack of a difference was not influenced by the presence of estrogens. Further evidence for abnormalities in parathyroid hormone secretion in osteoporosis comes from the work of Silverberg et al. [96,97]. These studies were based on the premise that a mild hypocalcemic challenge should lead to age-appropriate increases in parathyroid hormone concentration. Oral phosphate was used to induce the hypocalcemic challenge. The first studies were conducted with two distinct groups of younger and older subjects who had no evidence for osteoporosis. In both cases, the serum phosphorus concentration rose and the serum calcium level fell to the same extent. Young subjects showed a 43% increase in parathyroid hormone concentration over baseline values, whereas older women showed a much more exuberant response to the same hypocalcemic stimulus, with a 2.5-fold increase over baseline levels. This protocol set up two

78 opposing stimuli with respect to 1,25(OH)2D phosphorus as an inhibitor and parathyroid hormone as a stimulus. In both cases, the opposing regulators were neutralized and 1,25(OH)2D concentration did not change. These data were interpreted to suggest that older, normal subjects require more parathyroid hormone for a given hypocalcemic challenge to maintain 1,25 (OH)2D status. Such a formulation is consistent with the reduced renal capacity to form this metabolite with age. It is also possible that the aging skeleton requires a greater amount of parathyroid hormone to achieve effects that are seen at lower levels in younger subjects. The same protocol was applied to a group of postmenopausal women with osteoporosis [97]. After phosphate administration, these women experienced the same increase in serum phosphorus concentration and the same reduction in serum calcium concentration that was observed for the young subjects and the age-matched older women. In contrast to the marked increase in parathyroid hormone in their age-matched counterparts, the osteoporotic women demonstrated only a modest 43% increase (Fig. 3). Although this was sufficient in younger individuals to prevent the inhibitory effects of phosphorus on 1,25(OH)2D production, it did not suffice in these osteoporotic women as 1,25(OH)2D concentrations fell by 50% (Fig. 4). These observations are consistent with the presence of an abnormality in parathyroid secretory function in osteoporosis. Osteoporotic women thus have both reduced ability to form 1,25(OH)2D and a superimposed deficiency in parathyroid responsiveness. The need for more parathyroid hormone with age could be achieved by altering the calcium set point. For any given serum calcium concentration, the parathyroid hormone concentration is higher in the elderly. This could account for the age-related increase in parathyroid hormone concentration in the absence of any change in circulating calcium.

BILEZIKIAN AND SILVERBERG

FIGURE 4 The effect of a hypocalcemic stimulus on the parathyroid hormone response (Fig. 3) and on 1,25(OH)2 D concentrations in osteoporotic and nonosteoporotic postmenopausal women and in young normal subjects. The details of the experimental protocol are given in the text (reprinted from [97] with permission). The means in patients with osteoporosis are represented by hatched bars, those in age-matched controls by open bars, and those in a comparison group of young, healthy subjects by solid bars. An asterisk denotes a significant change from baseline (P  0.05).

When Ledger et al. [88] studied this point with a provocative challenge, no age-related increase in the set point for parathyroid hormone secretion could be demonstrated. When postmenopausal women with osteoporosis were studied, however, differences did emerge. Cosman et al. [98] used infusions of the synthetic peptide, human parathyroid hormone(1 – 34) to assess suppressibility of endogenous parathyroid hormone secretion. It was possible to distinguish between exogenous human PTH (1 – 34) and endogenous human parathyroid hormone (1 – 84) by use of the immunoradiometric assay, which does not detect the 1 – 34 exogenously administered peptide. The data were consistent with a higher calcium set point in osteoporotic women. Similarly, Portale et al. showed in elderly men that the set point of parathyroid hormone responsiveness to calcium was “shifted” to the right [99]. Such results are consistent with a protective effect of parathyroid hormone in the pathogenesis of osteoporosis. 2. RACIAL DIFFERENCES IN THE PARATHYROID HORMONE–VITAMIN D AXIS

FIGURE 3 The effect of a hypocalcemic stimulus on the parathyroid hormone response in osteoporotic and nonosteoporotic postmenopausal women and in young normal subjects. The details of the experimental protocol are given in the text. Reprinted from [97] with permission.

The major difference in bone mass between Caucasian and Black individuals has provided another opportunity to assess whether or not increased exposure to parathyroid hormone preserves bone. Black individuals enjoy approximately 10% higher bone mass than Caucasians throughout adult life. Moreover, fracture incidence is lower in Blacks [100,101] (see Chapter 22). Higher peak bone mass is certainly a major reason for the relatively protected skeleton in Black individuals. The observation by Modlin and Bell et al. [102,103] that Blacks have lower urinary

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D

calcium excretion than Caucasians has led to additional ethnic comparisons of the parathyroid – vitamin D axis. Although there is uniform agreement that urinary calcium excretion is lower in Blacks than in Whites, variable results have been reported concerning the hormones of mineral metabolism. Most, but not all, investigators have reported increased concentrations of parathyroid hormone, 1,25(OH)2D, or both in Black subjects. Once again, comparisons have been difficult due to use in some studies of parathyroid hormone assays that do not accurately reflect biologically active material [103 – 106]. Similarly, most investigators have reported that Blacks have lower serum levels of 25-hydroxyvitamin D. A prevailing hypothesis to explain this observation [105,106] is based on the fact that the increased skin pigment in Blacks leads to decreased dermal vitamin D synthesis [107,108]. This, in turn, leads to a secondary hyperparathyroidism with a resulting increase in 1,25(OH)2D concentration. The actions of the secondary hyperparathyroid state on renal tubular function leads to urinary calcium conservation. Bell postulated that increased 1,25(OH)2D and decreased urinary calcium could account for higher bone density observed in black individuals [105]. There is no difference between Black and White subjects with respect to calcium absorption from the gastrointestinal tract. These observations tend to support the idea that parathyroid hormone contributes to bone mass in Black individuals and is consistent with ideas about the anabolic properties of this hormone. Clearly, the mechanisms leading to greater bone mass in Blacks are more complex than differences in ambient concentrations of calcium-regulating hormones. For example, many studies have confirmed that bone turnover as assessed by histomorphometric analysis is reduced in Blacks. Osteocalcin [104 – 106], urinary hydroxyproline [104], and bonespecific alkaline phosphatase concentrations [104] are lower than in Caucasians, raising the possibility that high bone density in the face of increased parathyroid hormone and reduced bone turnover could reflect skeletal parathyroid hormone insensitivity [109]. Recent data in support of skeletal resistance to parathyroid hormone in Blacks come from El-Haji Fuleihan et al., who assessed responsiveness to hypo- and hypercalcemia [106]. They demonstrated higher maximal and minimal parathyroid hormone responses in Black subjects with no alteration in the set point or slope of the calcium – parathyroid hormone curve. However, just as baseline parathyroid hormone concentrations were higher and those of osteocalcin were lower in Blacks, hypocalcemia led to a less exuberant rise in osteocalcin in Blacks despite a vigorous rise in circulating parathyroid hormone. Although the differences in bone mass and susceptibility to osteoporosis among Blacks and Whites are incompletely understood, the possible role of the parathyroid hormone – vitamin D axis certainly deserves further investigation. A

79 recent report demonstrating by QCT analysis 40% more cancellous bone density at the lumbar spine in American Blacks than in Caucasians raises questions about racial differences in selected types of bone [104]. If, in fact, differences in the parathyroid – vitamin D axis underlie the relative protection from bone loss seen in Black subjects, differential effects in cancellous and cortical bone would not be surprising.

C. Parathyroid Hormone: A Positive or Negative Factor in Age-Related Bone Loss? The data reviewed in this chapter argue that parathyroid hormone can be viewed as either a negative or a positive factor in preservation of the postmenopausal skeleton. The age-related increase in parathyroid hormone may be adaptive or maladaptive. More information will be needed to sort out these diametrically opposite views. However, we already have abundant information about the skeleton in the classic condition of parathyroid hormone excess, primary hyperparathyroidism. In this disorder, excess parathyroid hormone leads to relative protection against bone loss in the lumbar spine [110]. By histomorphometric analysis, static, dynamic, and structural indices of the cancellous skeleton reveal maintenance of bone mass [111,112] and are contrasted dramatically with such indices in age-matched postmenopausal women with osteoporosis [113]. Moreover, longitudinal data confirm that the protective effect persists over time [114]. The data in primary hyperparathyroidism differ so strikingly from those in osteoporosis that one must conclude that the bone diseases of hyperparathyroidism and osteoporosis reflect two completely different disorders. In view of the protection accorded the very site that is at early risk for postmenopausal bone loss, namely the cancellous bone of the vertebral spine, parathyroid hormone’s actions on the aging skeleton are best viewed as protective, not deleterious. The hormone’s apparent anabolic effects on skeletal sites at risk in postmenopausal osteoporosis has led to studies designed to use it as a therapeutic agent. Clinical trials in postmenopausal women and in men have confirmed the hypothesis that parathyroid hormone increases bone mineral density of the lumbar spine [115,116] (see Chapter 77). Studies by Neer [117], Lindsay [118], Fujita [119], Kurland [120], Roe [121], Lane [122] and their colleagues have shown impressive effects of parathyroid hormone to increase bone density of the lumbar spine and hip without deleterious effects on cortical bone (Fig. 5). In other situations such as the bone loss due to the use of GnRH agonist [123] and when PTH is used as a combination therapy with bisphosphonate [124], parathyroid hormone has shown impressive anabolic potential. It is of interest that these increases are, in general, associated with enhanced bone turnover during the first 6 to 12 months of therapy.

80

BILEZIKIAN AND SILVERBERG

IV. SUMMARY This chapter has reviewed the evidence for different views of the parathyroid — vitamin D axis in postmenopausal osteoporosis. It is clear that age-related changes occur in this system, and it is likely that, to a certain extent, such changes are beneficial to an older individual who requires a greater amount of parathyroid hormone than is needed to achieve the same effect that can be reached at lower levels in youth. It is also evident that either insufficient or overexuberant adaptation could exert negative effects on the skeleton. Many other factors play a role in the common phenotype that we recognize as postmenopausal osteoporosis, and they are covered elsewhere in this book.

References

FIGURE 5

Changes in bone density after administration of parathyroid hormone (1 – 34) to men with idiopathic osteoporosis. The group receiving PTH is shown by the dark symbols; those receiving placebo are shown by the open symbols. The data are shown as percentage changes from baseline  SEM for (A) lumbar spine, (B) femoral neck, and (C) distal (1/3) radius. Significant between group comparisons are indicated by the asterisk (P  0.05) and within group comparisons are indicated by the  symbol (P  0.005). Reprinted from Kurland et al. (120) with permission.

The negative side of this hypothesis is that parathyroid hormone can have deleterious effects on cortical bone. In the mild form of hyperparathyroidism seen today, such demonstrable reductions are not severe [114,124]. In more severe forms of primary hyperparathyroidism, cortical bone loss is obvious. At some point in the spectrum of parathyroid hormone excess, markedly elevated hormone concentrations will lead to bone loss from the vertebral spine [125].

1. S. M. Ott, Bone remodeling and mineral homeostasis: Theoretical and methodological approach. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds), pp. 231 – 242. Academic Press, San Diego, 1996. 2. R. Eastell, A. L. Yergey, N. E. Viera, S. L. Cedel, R. Kumar, and B. L. Riggs, Interrelationship among vitamin D metabolism, true calcium absorption, parathyroid function and age in women: Evidence of an age-related intetinal resistance to 1,25-dihydroxyvitamin D action. J. Bone Miner. Res 6, 125 – 132, (1991). 3. S. Epstein, G. Bryce, J. Hinman, O. N. Miller, B. L. Riggs, S. L. Hui, and C. C. Johnston, Jr., The effect of age on bone mineral regulating hormones. Bone 7, 421 – 425 (1986). 4. J. Gallagher, B. Riggs, J. Eisman, A. Hamstra, S. Arnaud, and H. DeLuca, Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. J. Clin. Invest. 66, 729 (1979). 5. L. Avioli, J. McDonald, and S. Lee, The influence of age on the intestinal absorption of 47ca in women and its relation to calcium absorption in postmenopausal osteoporosis. J. Clin. Invest. 44, 1960 (1965). 6. O. Sorenson, B. Lumholtz, B. Lund, I. L. Hjelmshand, L. Mosekilde, F. Melsen, J. E. Bishop, and A. W. Norman, Acute effects of parathyroid hormone on vitamin D metabolism in patients with the bone loss of aging. J. Clin. Endocrinol. Metab. 54, 1258 – 1261 (1981). 7. K. Tsai, H. Heath, R. Kumar, and B. Riggs, Impaired vitamin D metabolism with aging in women. J. Clin. Invest. 3, 1668 – 1672 (1984). 8. J. F. Aloia, S. Cohn, and A. Vaswani, Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 9. M. J. McKenna, Differences in vitamin D status between countries in young adults and the elderly. Am J. Med. 93, 69 – 77 (1992). 10. B. Dawson-Hughes, Regulation of parathyroid hormone by dietary calcium and vitamin D. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine eds.), pp. 55 – 65. Raven Press, New York, 1994. 11. “National Academy of Science Recommended Dietary Allowances, 10th revised edition. The National Academy Press, Washington, DC, 1997. 12. J. M. Barragry, M. W. France, D. Corless, S. P. Gupta, S. Switala, B. J. Boucher, and R. D. Cohen, Intestinal cholecalciferol absorption in the elderly and in young adults. Clin. Sci. Mol. Sci. 55, 213 – 220 (1978). 13. M. J. Favus, Vitamin D. In “The Aging Skeleton” (C. J. Rosen, J. Glowacki, J. P. Bilezikian, eds), pp. 613 – 622. Academic Press, San Diego, CA, 2000.

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D 14. O. H. Sorenson, B. Lumholtz, B. Lund, I. L. Hjelmstrand, L. Mosekilde, F. Melsen, J. E. Bishop, and A. W. Norman, Acute effects of parathyroid hormone on vitamin D. metabolism in patients with the bone loss of aging. J. Clin. Endocrinol. Metab. 54, 1258 – 1261 (1982). 15. B. L. Riggs, A. Hamstra, and H. F. Deluca, Assessment of 25hydroxyvitamin D 1alpha-hydroxylase reserve in postmenopausal osteoporosis by administration of parathyroid extract. J. Clin. Endocrinol. Metab. 53, 833 – 835 (1981) 16. K. S. Tsai, H. Heath III, R. Kumar, and B. L. Riggs, Impaired vitamin D metabolism with aging in women. Posssible role in pathogenesis of senile osteoporosis. J. Clin. Invest. 73, 1668 – 1672 (1984). 17. T. L. Clemens and J. L. H. O’Riordan, Vitamin D. In “Principles and Practice of Endocrinology and Metabolism” (K. L. Becker, ed.), 3rd ed. Lippincott, Philadelphia, 2001. 18. J. MacLaughlin and M. F. Holick, Aging decreases the capacity of human skin to produce vitamin D3. J. Clin. Invest. 76, 1536 – 1538 (1985). 19. B. Riggs and L. Melton, Medical progress: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 20. A. R. Webb, C. Pilbeam, N. Hanafin, and M. Holick, An evaluation of the relative contributions of exposure to sunlight and of diet to the circulating concentrations of 25 hydroxyvitamin D in an elderly nursing home population in Boston. Am. J. Clin. Nutr. 51, 1075 – 1081 (1990). 21. H. M. Hodkinson, B. R. Stanton, P. Round, and C. Morgan, Sunlight, vitamin D and osteomalacia in the elderly. Lancet 1, 404 – 406 (1973). 22. A. R. Webb and M. F. Holick, The role of sunlight in the cutaneous production of vitamin D3. Annu. Rev. Nutr. 8, 375 – 399 (1988). 23. R. A. Bouillon, J. H. Auwers, W. D. Lissens, and W. K. Pelemans, Vitamin D status in the elderly: Seasonal substrate deficiency causes 1,25-dihyroxycholecaciferol deficiency. Am. J. Clin. Nutr. 45, 755 – 763 (1987). 24. T. C. B. Stamp and J. M. Round, Seasonal changes in human plasma levels of 25-hydroxyvitamin D. Nature 247, 1563 (1974). 25. M. A. Preece, S. Tomlinson, C. A. Ribot, J. Pietrek, H. T. Korn, D. M. Davies, J. A. Ford, M. G. Dunnijan, and I. L. O’Riordan, Studies of vitamin D deficiency in man. Q. J. Med. 44, 575 (1975). 26. M. F. Holick, The photobiology of vitamin D in man. In “Vitamin D. Basic and Clinical Aspects” (R. Kumar, ed.). Martinus Nijhoff, Boston, 1984. 27. A. R. Webb, L. Kline, and M. F. Holick, Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J. Clin. Endocrinol. Metab. 67, 373 – 378 (1988). 28. C. J. Rosen, A. Morrison, H. Zhou, D. Storm, S. J. Hunter, K. Musgrave, T. Chen, W. Wei, and M. F. Holick, Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner. 25, 83 – 92 (1994). 29. D. Villareal, R. Civitelli, A. Chines, and L. Avioli, Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J. Clin. Endocrinol. Metab. 72, 628 – 634 (1991). 30. M. Peacock, P. L. Selby, R. M. Francis, W. B. Brown, and L. Hordon, Vitamin D deficiency, insufficiency, sufficiency and intoxication: What do they mean? In “Vitamin D: Chemical, Biological and Clinical Update: 6th Workshop on Vitamin D” (A. W. Norman et al., eds.), pp. 569 – 570. Walter de Gruyter, New York, 1985. 31. L. Komar, J. Nieves, F. Cosman, A. Rubin, V. Shen, and R. Lindsay, Calcium homeostasis of an elderly population upon admission to a nursing home. J. Am. Geriatr. Soc. 41, 1057 – 1064 (1993). 32. F. M. Gloth, Vitamin D. In “The Aging Skeleton” (C. R. Rosen, J. Glowacki, and J. P. Bilezikian, eds), pp. 185 – 194. CA Academic Press, San Diego, 1999.

81 33. F. M. Gloth, C. M. Gundberg, B. W. Hollis, J. G. Haddad, J. D. Tobin, The prevalence of vitamin D deficiency in a cohort of homebound elderly subjects compared to a normative matched population in the United States. JAMA 274, 1683 – 1686 (1995). 34. M. K. Thomas, D. M. Lloyd-Jones, R. I. Thadhani, A. C. Shaw, D. J. Deraska, B. T. Kitch, E. C. Vamakis, I. M. Dick, R. L. Prince, and J. S. Finkelstein, Hypovitaminosis D in medical inpatients. N. Engl. J. Med. 338, 777 – 783 (1998). 35. D. S. Rao and M. Honasoge, Metabolic bone disease in gastrointestinal, hepatobiliary and pancreatic. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), 3rd ed., pp. 306 – 310. Lippinocott-Raven, New York, 1996. 36. L. M. Siegel and J. P. Bilezikian, Metabolic bone diseases and disorders of the gastrointestinal tract. In “Proceedings of the Falk Symposium: Gastrointestinal Tract and the Endocrine System.” Walters Kluwers, The Netherlands, 1995. 37. H. C. Mitchison, A. J. Malcolm, M. F. Bassendine, and O. F. W. James, Metabolic bone disease in primary biliary cirrhosis at presentation. Gastroenterology 94, 463 – 470 (1988). 38. D. P. DeSimone and N. H. Bell, Hypercalcemia and abnormal vitamin D metabolism. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 435 – 442. Raven Press, New York, 1992. 39. J. Gallagher, The pathogenesis of osteoporosis. Bone Miner. 9, 215 – 227 (1990). 40. D. M. Slovik, J. S. Adams, R. M. Neer, M. F. Hollick, and J. T. Potts, Jr., Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 – 374 (1981). 41. B. Riggs, A. Hamstra, and H. DeLuca, Assessment of 25-hydroxyvitamin D 1-alpha hydroxylase reserve in postmenopausal osteoporosis by administration of parathyroid extract. J. Clin. Endocrinol. Metab. 53, 833 – 835 (1981). 42. B. P. Halloran, A. Portale, E. T. Lonergan, and R. C. Morris, Jr., Production and metabolic clearance of 1,25-dihydroxyvitamin D in men: Effect of advancing age. J. Clin. Endocrinol. Metab. 70, 3180 – 3323 (1990). 43. R. Francis, M. Peacock, G. Taylor, J. Storer, and B. Nordin, Calcium malaborption in elderly women with vertebral fractures: Evidence for resistance to the action of vitamin D metabolites on the bowel. Clin. Sci. 66, 103 – 107 (1984). 44. P. Ebeling, M. Sandgren, E. DiMango, A. Lane, H. DeLuca, and B. Riggs, Evidence of an age-related decrease in intestinal responsiveness to vitamin D: Relationship between serum 1,25-dihydroxyvitamin D and intestinal vitamin D receptor concentrations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). 45. C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). 46. S. Pattanaungkul, B. L. Riggs, A. L. Yergey, N. E. Vieira, W. M. O’Fallon, and S. Khosla, Relationship of intestinal calcium absorption to 1,25-dihydroxyvitamin D levels in young versus elderly women: Evidence for age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J. Clin. Endocrinol. Metab. 85, 4023–4027 (2001). 47. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature (London) 367, 384 (1994). 48. N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89, 6665 – 6669 (1992). 49. F. G. Hustmyer, M. Peacock, S. Hui, C. C. Johnston, and J. Christian, Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J. Clin. Invest. 94, 2130 – 2134 (1994).

82 50. T. V. Nguyen and J. A. Eisman, Genetics of fracture: Challenges and opportunities. J. Bone Miner. Res. 15, 1253 – 1256 (2000). 51. S. J. Silverberg, I. N. Brown, J. P. Bilezikian, and L. J. Deftos, A new highly sensitive assay for parathyroid hormone in primary hyperparathyroidism. J. Bone Miner. Res. 15(Suppl. I), S167 (2000). 52. S. R. Nussbaum and J. T. Potts, Jr., Advances in Immunoassays for parathyroid hormone: Clinical applications in skeletal disorders of bone and mineral metabolism. In The Parathyroids (J. P. Bilezikian, R-Marcus, and M. Levine, eds), pp. 157 – 170. Raven Press, New York, 1994. 53. P. D. Delmas, D. Stenner, H. W. Wahner, K. G. Mann, and B. L. Riggs, Increase in serum BGP with aging in women: Implications for the mechanism of age-related bone loss. J. Clin. Invest. 71, 1316 – 1321 (1983). 54. R. J. Duda, J. F. O’Brien, J. A. Katzmann, J. M. Peterson, and K. G. Mann, B. L. Riggs, Concurrent assays of circulating bone GLA protein and bone alkaline phosphatase: Effects of age, sex and metabolic bone disease. J. Clin. Endocrinol. Metab. 66, 951 – 957 (1988). 55. R. Eastell, P. D. Delmas, S. F. Hodgson, E. F. Eriksen, K. G. Mann, and B. L. Riggs, Bone formation rate in older normal women: Concurrent assessment with bone histomorphometry, calcium kinetics, and biochemical markers. J. Clin. Endocrinol. 67, 741 – 748 (1988). 56. S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of bone biopsies form the iliac crest of normal British subjects. Metab. Bone Dis. Relat. Res. 4, 231 – 236 (1982). 57. S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of dynamic parameters of trabecular bone formation in the iliac crest of normal British subjects. Metab. Bone Dis. Relat. Res. 5, 69 – 74 (1983 – 1984). 58. S. Epstein, J. Poser, R. McClintock, C. C. Johnston, G. Bryce, and S. Hui, Differences in serum bone GLA protein with age and sex. Lancet 1, 307 – 310 (1987) 59. J. C. Gallagher, H. K. Kinyamu, S. E. Fowler et al., Calciotropic hormones and bone markers in the elderly. J. Bone Miner. Res. 13, 475 – 482 (1988). 60. S. Khosla, L. J. Melton III, and B. L. Riggs, Parathyroid function in the normal aging process. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.). Academic Press, San Diego, in press. 61. L.K. H. Koh, D. C. Bauer, B. A. Forsyth, L. R. Gore, M. T. Vogt, and S. R. Cummings, PTH in postmenopausal osteoporotic women. J. Bone Miner. Res. 12, S166 (1997). 62. R. L. Prince, I. Dick, A. Devine, R. I. Price, D. H. Gutteridge, D. Kerr, A. Criddle, P. Garcia-Webb, and A. St. John, The effects of menopause and age on calcitropic hormones: A cross-sectional study of 655 healthy women aged 35 – 90. J. Bone Miner. Res. 10, 835 – 842 (1995). 63. R. G. H. Jilka, G. Girasole, G. Passeri, D. Williams, J. Abrams, B. Boyce, H. Broxmeyer, and S. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257, 88 – 91 (1992). 64. S. C. Manolagas, R. L. Jilka, T. Bellido, C. A. O’Brien, and A. M. Parfitt, Interleukin-6-type cytokines and their receptors. In “Principles of Bone Biology” (J. P. Bilezikian, G. R. Raisz, and G. A. Rodan, Eds), pp. 701 – 714. Academic Press, San Diego, 1996. 65. B. Riggs and L. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1983). 66. P. Ebeling, J. Jones, M. Burritt, C. Duerson, A. Lane, C. Hassager, R. Kumar, and B. Riggs, Skeletal responsiveness to endogenous parathyroid hormone in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 75, 1033 – 1038 (1992). 67. M. Kotowicz, G. Klee, P. Kao, S. Hodgson, S. Cedel, E. Eriksen, D. Gonchoroff, H. Judd, and B. L. Riggs, Relationship between serum intact PTH and bone remodeling in Type I osteoporosis: Evidence that skeletal sensitivity is increased. Osteoporosis 1, 14 – 20 (1990).

BILEZIKIAN AND SILVERBERG 68. J. P. Bilezikian, S. J. Silverberg, F. Gartenberg, T. S. Kim, T. Jacobs, E. Siris, and E. Shane, Clinical presentation of primary hyperparathyroidism. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.), pp. 457 – 470. Raven Press, New York, 1994. 69. R. Marcus, P. Madvig, M. Crim, A. Pont, and J. Kosek, Conjugated estrogens in the treamtent of postmenopausal women with hyperparathyroidism. Ann. Int. Med. 100, 633 – 640 (1984). 70. P. Selby and M. Peacock, Ethinyl estradiol and norethindrone in the treatment of primary hyperparathyroidism in postmenopausal women. N. Engl. J. Med. 314, 1481 – 1485 (1986). 71. W. R. McKane, S. Khosla, M. F. Burritt, P. C. Kao, D. M. Wilson, S. J. Ory, and B. L. Riggs, Mechanism of renal calcium conservation with estrogen replacement therapy in women in early post menopause, J. Clin. Endocrinol. 80, 3458 – 3464 (1995). 72. R. Heaney, A unified concept of osteoporosis. Am. J. Med. 39, 377 – 380 (1965). 73. C. Jasani, B. Nordin, D. Smith, and I. Swanson, Spinal osteoporosis and the menopause. Proc. R. Soc. Med. 58, 441 – 444 (1965). 74. B. L. Riggs and L. J. Melton III, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 (1983). 75. K. Tsai, P. Ebeling, and B. Riggs, Bone responsiveness to parathyroid hormone in normal and osteoporotic postmenopausal women. J. Clin. Endocrinol. Metab. 69, 1024 – 1027 (1989). 76. B. L. Riggs, S. Khosla, and L. J. Melton III, A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men, J. Bone Miner. Res. 13, 763 – 773 (1998). 77. H. M. Heshmati, S. Khosla, M. F. Burritt, W. M. O’Fallon, and B. L. Riggs, A defect in renal calcium conservation may contribute to the pathogenesis of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 83, 1916 – 1920 (1998). 78. S. Khosla, E. J. Atkinson, L. J. Melton III, and B. L. Riggs, Effects of age and estrogen status on serum parathyroid hormone levels and biochemical markers of bone turnover in women: A populationbased study. J. Clin. Endocrinol. Metab. 82, 1522 – 1527 (1997). 79. W. R. McKane, S. Khosla, J. Risteli, S. P. Robins, J. M. Muhs, and B. L. Riggs, Role of estrogen deficiency in pathogenesis of secondary hyperparathyroidism and increased bone resorption in elderly women. Proc. Assoc. Am. Phys. 109, 174 – 180 (1997). 80. B. L. Riggs, W. M. O’Fallon, J. Muhs, M. K. O’Conner, R. Kumar, and L. J. Melton, Long-term effects of calcium supplementation on serum parathyroid hormone level, bone turnover, and bone loss in elderly women. J. Bone Miner. Res. 13, 168 – 174 (1998). 81. M. Parisien, D. W. Dempster, E. Shane, and J. B. Bilezikian, Histomorphometric analysis of bone in primary hyperparathyroidism. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.), pp. 567 – 574. Raven Press, New York, 1994. 82. J. S. Finkelstein, Pharmacological mechanisms of therapeutics: Parathyroid hormone. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 993 – 1006 Academic Press, San Diego, 1996. 83. J. P. Bilezikian, Estrogens and postmenopausal osteoporosis: Was Albright right after all? J. Bone Miner. Res. 13, 774 – 776 (1998). 84. J. P. Bilezikian and S. J. Silverberg, Parathyroid hormone: Does it have a role in the pathogenesis of osteoporosis. Clin. Lab. Med. 20, 559 – 567 (2000). 85. R. Eastell, A. L. Yergey, N. E. S. L. Veiria, R. Kumar, and B. L. Riggs, Interrelationships among vitamin D metabolism, true calcium absorption, parathyroid function, and age in women: Evidence of an age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J Bone Miner. Res. 6, 125 – 132 (1991). 86. G. A. Ledger, M. F. Burritt, P. C. Kao, W. M. O’Fallon, B. L. Riggs, and S. Khosla, Role of parathyroid hormone in mediating nocturnal and age-related increases in bone resorption. J. Clin. Endocrin. Metab. 80, 3304 – 3310 (1995).

83

CHAPTER 40 Role of Parathyroid Hormone and Vitamin D 87. S. Boonen, X. G. Cheng, J. Nijs, P. H. Nicholson, G. Verbeke, E. Lesaffre, J. Aerssens, and J. Dequeker, Factors associated with cortical and trabecular bone loss as quantified by peripheral computed tomography (pQCT) at the ultradistal radius in aging women. Calcif. Tissue Int. 60, 164 – 170 (1997). 88. G. A. Ledger, M. F. Burritt, P. C. Kao, W. M. O’Fallon, B. L. Riggs, and S. Khosla, Abnormalities of parathyroid hormone secretion in elderly women that are reversible by short term therapy with 1,25dihydroxyvitamin D3. J. Clin. Endocrinol. Metab. 79, 211 – 216 (1994). 89. J. F. Tohme, F. Cosman, and R. Lindsay, Osteoporosis. In “Principles and Practice of Endocrinology and Metabolism” (K. L. Becker, ed.), 2nd ed. Lippincott, Philadelphia, 1995. 90. M. S. Calvo, R. Eastell, K. P. Offord, E. J. Bergsalh, and M. F. Burritt, Circadian variation in ionized calcium and intact parathyroid hormone: Evidence for sex differences in calcium homeostasis. J. Clin. Endocrinol. Metab. 72, 77 – 82 (1991). 91. M. E. Markowitz, S. Arnaud, J. F. Rosen, M. Thorpy, and S. Laximinarayan, Temporal relationships between the circadian rhythms of serum PTH and calcium concentrations. J. Clin. Endocrinol. Metab. 67, 1068 – 1073. 92. R. Eastell, M. S. Calvo, M. F. Burritt, K. P. Offord, R. G. G. Russell, and B. L. Riggs, Abnormalities in circadian patterns of bone resorption and renal calcium conservation in type I osteoporosis. J. Clin. Endocrinol. Metab. 74, 487 – 494 (1992). 93. K. Prank, S. Nowlan, H. Harms, et al., Time series prediction of plasma hormone concentration: Evidence for differences in predictability of parathyroid hormone secretion between osteoporotic patients and normal controls. J. Clin. Invest. 95, 2910 – 2919 (1995). 94. L. J. Deftos and S. J. Schiff, Predicting PTH pulses and patterns in osteoporosis. J. Clin. Invest. 95, 2433 – 2434 (1995). 95. M. H. Samuels, J. D. Veldhuis, P. Kramer, et al., Episodic secretion of parathyroid hormone in postmenopausal women: Assessment by deconvolution analysis and approximate entropy. J. Bone Miner. Res. 12, 616 – 623 (1997). 96. S. J. Silverberg, E. Shane, T. L. Clemens, D. W. Dempster, G. V. Segre, R. Lindsay, and J. P. Bilezikian, The effects of oral phosphate on major indices of skeletal metabolism in normal subjects. J. Bone Miner. Res. 1, 383 – 338 (1986). 97. S. J. Silverberg, E. Shane, L. de la Cruz, G. V. Segre, T. L. Clemens, and J. P. Bilezikian, Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D3 formation in women with osteoporosis. N. Engl. J. Med. 320, 277 – 281 (1989). 98. F. Cosman, V. Shen, B. Herrington, and R. Lindsay, Response of the parathyroid gland to infusion of human parathyroid hormone(1 – 34). J. Clin. Endocrinol. Metab. 73, 1345 – 1351 (1991). 99. A. A. Portale, E. T. Lonergan, D. M. Tanney, and B. P. Halloran, Aging alters calcium regulation of serum concentration of parathyroid hormone in healthy men. Am. J. Physiol. 35, E139 – E146 (1997). 100. Y. Liel, H. Edwards, J. Shary, K. M. Spicer, L. Gordan, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66, 1247 – 1250 (1988). 101. M. E. Farmer, L. R. White, and J. A. Brody, Race and sex differences in hip fracture incidence. Am. J. Public Health 44, 1374 – 1380 (1984). 102. M. Modlin, Urinary calcium in normal adults and in patients with renal stones: An interracial study. Invest. Urol. 5, 49 – 57 (1967). 103. N. H. Bell, A. Greene, S. Epstein, M. J. Oexmann, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in blacks. J. Clin. Invest. 76, 470 – 473 (1985). 104. M. Kleerekoper, D. Nelson, E. L. Peterson, M. J. Flynn, A. S. Pawluszka, G. Jacobson, and P. Wilson, Reference data for bone mass, calcitropic hormones and biochemical markers of bone remodeling in

105.

106.

107.

108.

109. 110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

older (55 – 75) postmenopausal white and black women. J. Bone Miner. Metab. 9, 1267 – 1276 (1994). D. E. Meier, M. M. Luckey, S. Wallenstein, T. L. Clemens, E. S. Orwoll, and C. I. Waslien, Calcium, vitamin D, and parathyroid hormone status in young white and black women: Association with racial differences in bone mass. J. Clin. Endocrinol. Metab. 72, 703 – 710 (1991). G. El-Haji Fuleihan, C. M. Gundberg, R. Gleason, E. M. Brown, M. E. Stronski, F. D. Grant, and P. R. Conlin, Racial differences in parathyroid hormone dynamics. J. Clin. Endocrinol. Metab. 79, 1642 – 1647 (1994). T. L. Clemens, S. L. Henderson, J. S. Adams, M. F. Holick, Increased skin pigment reduces the capacity of skin to synthesize vitamin D3. Lancet 1, 74 – 76 (1982). M. F. Holick, J. A. MacLaughlin, and S. H. Doppelt, Regulation of cutaneous previtamin D photosynthesis in man: Skin pigment is not an essential regulator. Science 211, 590 – 593 (1981). R. S. Weinstein and N. H. Bell, Diminished rates of bone formation in normal black adults. N. Engl. J. Med. 319, 1698 – 1701 (1988). S. J. Silverberg, E. Shane, L. de la Cruz, D. W. Dempster, F. Feldman, D. Seldin, T. P. Jacobs, E. S. Siris, M. Cafferty, M. V. Parisien, R. Lindsay, T. L. Clemens, and J. P. Bilezikian, Skeletal disease in primary hyperparathyroidism. J. Bone Miner. Res. 4, 283 – 291 (1989). M. Parisien, S. J. Silverberg, E. Shane, L. de la Cruz, R. Lindsay, J. P. Bilezikian, and D. W. Dempster, The histomorphometry of bone in primary hyperparathyroidism: Preservation of cancellous bone. J. Clin. Endocrinol. Metab. 70, 930 – 938 (1990). M. Parisien, R. Mellish, S. J. Silverberg, E. Shane, R. Lindsay, J. P. Bilezikian, and D. W. Dempster, Maintenance of cancellous bone connectivity in primary hyperparathyroidism: Trabecular and strut analysis. J. Bone Miner. Res. 7, 913 – 920 (1992). M. Parisien, F. Cosman, R. W. R. Mellish, M. Schnitzer, J. Nieves, S. J. Silverberg, E. Shane, D. Kimmel, R. R. Recker, J. P. Bilezikian, R. Lindsay, and D. W. Dempster, Bone structure in postmenopausal hyperparathyroid, osteoporotic and normal women. J. Bone Miner. Res. 10, 1393 – 1399 (1995). S. J. Silverberg, F. Gartenberg, T. P. Jacobs, E. Shane, E. Siris, R. B. Staron, D. McMahon, and J. P. Bilezikian, Longitudinal measurements of bone density and biochemical indices in untreated primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 80, 723 – 728 (1995). D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. T. Botts, M. A. Daly, J. A. Campbell, and R. M. Neer, Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1 – 34) and 1,25-dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 3811 (1986). J. Reeve, J. N. Bradbeer, M. Arlot, U. M. Davies, J. R. Green, L. Hampton, C. Edouard, R. Hesp, P. Hulme, and J. P. Ashby, hPTH1-34 treatment of osteoporosis with added hormone replacement therapy: Biochemical, kinetic and histological responses. Osteoporosis Int. 1, 162 – 170 (1991). R. Neer, C. Arnaud, J. R. Zanchetta, R. Prince, G. A. Gaich, and J. Y. Reginster, “Recombinant Human PTH [rhPTH(1 – 34)] Reduces the Risk of Spine and Non-spine Fractures in Postmenopausal Osteoporosis. 82nd Annual Meeting of the Endocrine Society, S193, 2000. [Abstract] R. Lindsay, J. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, D. Dempster, and F. Cosman, Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550 – 555 (1997) T. Fujita, T. Inoue, M. Morii, R. Morita, H. Norimatsu, H. Orimo, H. E. Takahashi, K. Yamamoto, and M. Fukunaga, Effect of intermittent weekly dose of human parathyroid hormone (1 – 34) on osteoporosis: A randomized double-masked prospective study using three dose levels. Osteoporosis Int. 9, 296 – 306 (1999).

84 120. E. S. Kurland, F. Cosman, D. J. McMahon, C. J. Rosen, R. Lindsay, and J. P. Bilezikian, Therapy of idiopathic osteoporosis in men with parathyroid hormone: effects on bone mineral density and bone markers. J. Clin. Endocrinol. Metab. 85, 3069 – 3076 (2000). 121. E. B. Roe, S. D. Shanchez, G. A. del Puero, E. Pierini, P. Bacchetti, C. E. Cann, and C. D. Arnaud, Parathyroid hormone 1 – 34 (hPTH 1 – 34) and estrogen produce dramatic bone density increases in postmenopausal osteoporosis – Results from a placebo-controlled randomized trial. J. Bone Miner. Res. 14, S137 (1999). 122. N. E. Lane, S. Sanchez, G. W. Modin, H. K. Genant, E. Pierini, and C. D. Arnaud, Parathyroid hormone treatment can reverse corticosteroid-induced osteoporosis: Results of a randomized controlled clinical trial. J. Clin. Invest. 102, 1627 – 1633 (1998). 123. J. S. Finkelstein, A. Klibanski, E. H. J. Schaefer, M. D. Hornstein, I. Schiff, and R. M. Neer, Parathyroid hormone for the prevention of

BILEZIKIAN AND SILVERBERG bone loss induced by estrogen deficiency. N. Engl. J. Med. 331, 1618 – 1623 (1994). 124. R. S. Rittmaster, M. Bolognese, M. P. Ettinger, D. A. Hanley, A. B. Hodsman, D. L. Kendler, and C. J. Rosen, Enhancement of bone mass in osteoporotic women with parathyroid hormone followed by alendronate. J. Clin. Endocrinol. Metab. 85, 2129 – 2134 (2000). 125. D. S. Rao, R. J. Wilson, M. Kleerekoper, and A. M. Parfitt, Lack of biochemical progression or continuation of accelerated bone loss in mild asymptomatic primary hyperparathyroidism: Evidence for biphasic disease course. J. Clin. Endocrinol. Metab. 676, 1294 – 1298 (1988). 126. S. J. Silverberg, F. Locker, and J. P. Bilezikian, Vertebral osteopenia: A new indication for surgery in primary hyperparathyroidism, J. Clin. Endocrinol. Metab. 81, 4007 – 4012 (1996).

CHAPTER 41

Postmenopausal Osteoporosis How the Hormonal Changes of Menopause Cause Bone Loss ROBERTO PACIFICI

Division of Bone and Mineral Diseases, Washington University, St. Louis, Missouri 63110

I. Steroid Biosynthesis and Menopause II. Mechanism of Action of Estrogen in Bone

III. Summary and Conclusions References

I. STEROID BIOSYNTHESIS AND MENOPAUSE

postmenopausal period the major source of estradiol becomes the peripheral conversion of estrone and testosterone. This conversion takes places at many extraglandular sites, but mainly in the adipose tissue. The latter pathway is enhanced with aging and obesity. Due to the depletion of responsive follicles after menopause, the range of progesterone concentrations resembles that observed in premenopausal women during the proliferative phase. Small amounts of progesterone continue to be made by the adrenal glands, which also become the main source of androstenedione, the most abundant androgen in postmenopausal women [2]. The ovaries continue to account for about 20% of total androstenedione production. In contrast, ovarian production of testosterone does not decrease significantly after the menopause [2]. Dehydroepiandrosterone (DHEA) and DHEA-sulfate are mostly produced by the adrenal gland. The production of these androgens declines after age 30, independent of ovarian function. Additionally, it should be underlined that menopause is also the cause of an increased production of the pituitary

Menopause represents a critical life step characterized by complex endocrine changes which affect the musculoskeletal system and its neurological control. The hallmark of the menopausal transition is the cessation of menses. However, the hormonal changes that signal decreased ovarian function begin to occur in the decade prior to the development of frankly irregular cycles. Pathognomonic of the cessation of ovarian function is a marked decline in the 17-estradiol concentrations (Table 1), the major estrogen in women of reproductive age [1]. Menopause is also characterized by a marked decrease in estrone serum concentration. However, estrone produced as a result of the peripheral conversion of androstenedione and testosterone of adrenal and ovarian origins [1] becomes the most abundant estrogen after menopause. To a lesser extent estrone results from the hydrolysis of estrone sulfate and this represents a large and stable pool of estrogen in the body. In the

OSTEOPOROSIS, SECOND EDITION VOLUME 2

85

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

86

ROBERTO PACIFICI

TABLE 1

Serum Levels of Sex Steroid Hormones after Menopause

Hormone

Postmenopausal serum level

17 Estradiol (pg/ml) Estrone (pg/ml)

5 – 20 10 – 20

Progesterone (ng/ml) Androstenedione (pg/dl)

0–6 400 – 1100

DHEA (g/ml)

Major source after menopause Peripheral conversion of estrone and testosterone Peripheral conversion of androstenedione and testosterone Ovary and Adrenal Adrenal

0–3

Adrenal

DHEA-sulfate (g/ml)

0.82 – 3.38

Adrenal

Testosterone

144 – 252

Ovary

hormones follicle-stimulating hormone (FSH) and lutenizing hormone (LH). Although the role of these substances in maintaining bone health is unknown, evidence is beginning to emerge indicating that pituitary hormones may modulate the effects of estrogen in bone [3], perhaps by activating the estrogen receptor (ER) [4].

II. MECHANISM OF ACTION OF ESTROGEN IN BONE A. Introduction Postmenopausal osteoporosis is a heterogeneous disorder characterized by a progressive loss of bone tissue which begins after natural or surgical menopause and leads to fracture within 15 – 20 years from the cessation of the ovarian function [5]. Although suboptimal skeletal development (“low peak bone mass”) and age-related bone loss may be contributing factors, a hormone-dependent increase in bone resorption and accelerated loss of bone mass in the first 5 or 10 years after menopause appears to be the main pathogenetic factor [6,7] of this condition. That estrogen deficiency plays a major role in postmenopausal bone loss is strongly supported by the higher prevalence of osteoporosis in women than in men [8], the increase in the rate of bone mineral loss detectable by bone densitometry after artificial or natural menopause [9 – 11], the existence of a relationship between circulating estrogen and rates of bone loss [12 – 14], and the protective effect of estrogen replacement with respect to both bone mass loss and fracture incidence [15 – 17]. The potential fracture risk for any postmenopausal female depends on the degree of bone turnover, the rate and extent of bone loss, associated disease processes which induce bone loss, age of menarche and menopause, and the bone mass content achieved at skeletal maturity. The latter depends on the extent of estro-

gen exposure, habitual physical activity, quantity of calcium intake, and genetic predisposition (see Chapter 25). The manner with which the genetic “signal” conditions those biological mechanisms which are essential to achieve peak bone mass in adolescence is still unknown, although evidence is beginning to accumulate that low peak bone mass may be linked to a particular vitamin D receptor genotype [18] (see Chapter 26). The bone-sparing effect of estrogen is mainly related to its ability to block bone resorption [19], although stimulation of bone formation is likely to play a contributory role [20,21]. Estrogen-dependent inhibition of bone resorption is, in turn, due to both decreased osteoclastogenesis and diminished resorptive activity of mature osteoclasts. However, inhibition of osteoclast formation is currently regarded as the main mechanisms by which estradiol (E2 prevents bone loss [19,22].

B. Cells and Cytokines Which Regulate Osteoclast Formation Osteoclasts arise by cytokine-driven proliferation and differentiation of monocyte/macrophage precursors, a process facilitated by bone marrow stromal cells (Fig. 1) These cells provide a physical support for nascent osteoclasts and produce soluble and membrane-associated factors essential for the proliferation and/or the differentiation of osteoclast precursors [23]. During inflammation, activated T cells, a population of lymphocytes which does not participate in the regulation of physiologic bone turnover, assumes a key role in stimulating osteoclast formation and does so by producing potent membrane-bound and soluble cytokines [24]. Another cell lineage that may have an important role in the regulation of osteoclastogenesis is that of B cells although their exact role remains controversial. For example, B-cell-deficient mice have been found to display decreased trabecular area and increased bone resorption, as

87

CHAPTER 41 Postmenopausal Osteoporosis

FIGURE 1

Cell and cytokines critical for osteoclast formation. Estrogen decreases osteoclast formation by downregulating the monocytic production of IL-1 and TNF and the stromal cell production of M-CSF and IL-6.

compared to B-cell-replete mice of the same strain [25], suggesting that B cells inhibit bone resorption and osteoclastogenesis. In contrast, other studies have shown that estrogen deficiency up regulates B-lymphopoiesis in the bone marrow [26,27], suggesting that cells of the B lineage may contribute to the increased osteoclast [OC] production characteristic of estrogen deficient animals. In humans B cells are an important source of transforming growth factor (TGF), a factor that inhibits osteoclast formation by inducing apoptosis of early and late osteoclast precursors and mature osteoclasts. Among the cytokines involved in the regulation of osteoclast formation are (RANKL) (also known as OPGL, TRANCE, or ODF) and macrophage colony-stimulating factor (M-CSF) (Table 2) [28 – 33] (see Chapters 3, 6, 12, and 13). These factors are produced primarily by stromal cells and osteoblasts. However, T cells are an important source of both factors, especially in stimulated conditions [24]. RANKL is a member of the tumor necrosis factor (TNF) family which exists in a membrane-bound and in a soluble form. RANKL binds to the transmembrane receptor RANK, which is expressed on the surface of osteoclasts and osteoclast precursors of the monocytic lineage [28]. RANKL binds TABLE 2

also to osteoprotegerin (OPG), a soluble decoy receptor produced by numerous hematopoietic cells. Thus, OPG, by sequestering RANKL and preventing its binding to RANK, functions as a potent anti-osteoclastogenic cytokine [34]. In the presence of M-CSF, RANKL induces the differentiation of monocytic cells into osteoclasts [28] by activating the MAP kinase Jun terminal Kinase (JNK), an enzyme which enhances the production of two essential osteoclastogenic transcription factors: c-Fos and c-Jun [35]. RANKL binding to RANK also activates NFB, a family of transcription factors essential for osteoclast formation and survival. Under physiological conditions M-CSF and RANKL are the only factors produced in the bone marrow in an amount sufficient to induce osteoclast formation. Thus, MCSF and RANKL are regarded as the true essential physiologic osteoclastogenic cytokines. The critical role of each of these cytokines in the osteoclastogenic process is demonstrated by the finding that deletion of either gene prompts osteopetrosis due to absence of osteoclasts, a circumstance reversed by administration of the relevant cytokine [29,36,37]. M-CSF induces the proliferation of early osteoclast precursors, the differentiation of more mature osteoclasts, the fusion of mononucleated preosteoclasts, and it increases the survival of mature osteoclasts [38 – 40]. RANKL does not induce cell proliferation, but promotes the differentiation of osteoclast precursors from an early stage of maturation to fully mature multinucleated osteoclasts. RANKL is also capable of activating mature osteoclasts, thus rendering these cells capable of resorbing bone. While consensus exists that RANKL stimulates bone resorption in organ cultures, the effect of M-CSF on bone resorption is controversial, as both inhibitory and stimulatory effects on bone resorption have been reported [38 – 43]. Monocytes, stromal cells, osteoblasts, and lymphocytes produce inflammatory cytokines which have direct proosteoclastogenic effects. Among these factors are interleukin (IL)-1, IL-6, IL-11, and TNF [44 – 53]. These factors stimulate osteoclast formation by increasing the stromal cell

Effects of Cytokines on Bone Resorption

Cytokine

Stimulate osteoclast formation

Stimulate resorption activity of mature osteoclasts

Stimulate in vivo bone resorption under conditions of estrogen deficiency

RANKL

Yes

Yes

No

M-CSF

Yes

?

Yes

IL-1

No

Yes

Yes

IL-6

Yes

No

No

TNF

Yes

Yes

Yes

GM-CSF

Yes

No

No

88

ROBERTO PACIFICI

production of RANKL [30,54,55] and M-CSF [56,57]. Another factor relevant for osteoclastogenesis is TGF-. This cytokine stimulates OPG production [58], thus inhibiting osteoclast formation. While in physiological conditions IL-1, IL-6, and TNF are produced in the bone marrow at low concentration, their bone marrow levels increase both during inflammation and in conditions of estrogen deficiency [19,22]. Thus, IL-1, IL-6, and TNF play a critical role in enhancing osteoclast production, survival, and activity in pathological conditions. Recently Kobayashi et al. [59] have demonstrated that TNF, in the presence of M-CSF, induces the differentiation of monocytes into mature multinucleated osteoclasts, which are, however, incapable of resorbing bone. Neither IL-1 nor IL-6 are capable of directly promoting the differentiation of osteoclast precursors into multinucleated osteoclasts. However, the addition of IL-1 to cultures of osteoclasts generated using TNF and M-CSF induces in these cells the capacity of resorbing bone and increases their survival. Thus, TNF is a true osteoclastogenic cytokine which can induce osteoclast formation when present at high concentrations. In contrast, IL-1 is incapable of inducing osteoclast formation, although it promotes osteoclast activation and survival. Since M-CSF and RANKL are present in the bone marrow in physiological conditions, osteopetrosis is not a feature of transgenic mice lacking the capacity of producing and/or responding to either IL-1, IL-6, or TNF [60 – 62]. Thus, IL-1, IL-6, and TNF stimulate osteoclastogenesis in pathological conditions, but are not essential for baseline osteoclastogenesis. In summary, the accumulated data demonstrate that RANKL and M-CSF are the only two factors known at the present time which are absolutely critical for osteoclast formation in physiological conditions. In contrast, the inflammatory cytokines IL-1, IL-6, and TNF are not essential for the maintenance of baseline osteoclastogenesis, although they are key for enhancing osteoclast formation and osteoclast activity during inflammation [40,63] and in conditions of E2 deficiency [19,22].

C. Effects of Estrogen on the Production of Osteclastogenic Cytokines It is now recognized that estrogen downregulates the production of several proosteoclastogenic factors, including IL-1, IL-6, TNF, M-CSF, and Prostaglandin E2 (PGE2) In addition, estrogen stimulates the production of important anti-osteoclastogenic factors, including IL-1ra [64], OPG [65], and TGF- [66]. The expression of estrogen receptors (both  and ) has been demonstrated in monocytes [67,68], osteoclasts [69], stromal cells, osteoblasts [70], and T cells [71]. Thus, all the major populations involved in the osteoclastogenesis

are a target of the sex steroid (see Chapter 10 for further discussion of estrogens and Chapter 13 for cytokines). The cytokines first recognized to be regulated by estrogen were IL-1 and TNF. This observation was prompted by the finding that monocytes of patients with “high turnover” osteoporosis, the histological hallmark of postmenopausal osteoporosis, secrete increased amounts of IL-1 [72]. Crosssectional and prospective comparisons of pre- and postmenopausal women revealed that monocytic production of IL-1 and TNF increases after natural and surgical menopause and is decreased by treatment with estrogen and progesterone [73,74]. Subsequent observations showed that the postmenopausal increase in IL-1 activity results from an effect of estrogen on the production of both IL-1 and IL-1ra [64]. Studies in normal women undergoing ovariectomy (ovx] [75,76] revealed that estrogen withdrawal is associated with an increased production of IL-1 and TNF, and also of granuloayte-macrophage colony-stimulating factor (GMCSF). The changes in these cytokine levels occur in a temporal sequence consistent with a causal role of IL-1, TNF, and GM-CSF in the pathogenesis of ovx-induced bone loss [75]. Estrogen and progesterone have been shown to decrease the secretion of IL-1 from peripheral blood and bone marrow monocytes and to decrease the steady-state expression of IL-1 mRNA in monocytes [77]. However, the exact molecular mechanism by which E2 decreases IL-1 production remains to be determined. Estrogen has been shown to increase the expression of the decoy type II, IL-1 receptor in bone marrow cells and osteoclasts [78]. Thus upregulation of cell responsiveness to IL-1 via downregulation of IL11RII is also likely to be a key mechanism by which estrogen deficiency induces bone loss. Recently we have clarified how estrogen downregulates the monocytic production of TNF. The genomic effects of E2 are mediated by a ligand-inducible transcription factor, known as ER. Following the discovery of a second ER (ER) [79 – 81], the first identified ER is now known as ER [79]. ER is expressed primarily in the uterus, testis, ovary, and pituitary, while ER is expressed mostly in the prostate, ovary, lung, bladder, brain, uterus, and testis [82]. Although ER and ER have similar DNA binding domains, their A/B domains and activation-function region 1 (AF-1) and quite different, suggesting that they may differentially regulate ER responsive genes [82]. Recent studies have indeed demonstrated that ER and ER respond differently to ligands, leading to opposite effects on AP-1-induced gene expression [83]. Specifically, while the ER mediated effects of E2 lead to stimulation of AP-1 induced gene expression, E2 acts as repressor of AP-1-induced transcription when bound to ER [83]. Cells of the monocytic lineage are known to express both ER and ER. Estrogen binding to ER leads to decreased activation of the Jun terminal kinase (JNK), a phenomenon which leads to decreased production of c-Jun

CHAPTER 41 Postmenopausal Osteoporosis

and JunD, two members of the AP-1 family of transcription factors [84]. Decreased AP-1 production results in decreased AP-1-induced TNF gene expression and lower TNF production [84, 85]. Studies conducted to determine if estrogen regulates the production of IL-6 revealed that in murine stromal and osteoblastic cells IL-6 production is inhibited by the addition of estrogen [49] and stimulated by estrogen withdrawal [51]. In vivo studies also revealed that the production of IL-6 is increased in cultures of bone marrow cells from ovx mice [50]. This effect is mediated, at least in the mouse, by an indirect effect of estrogen on the transcription activity of the proximal 225-bp sequence of the IL-6 promoter [86,87]. Interestingly, although studies with human cell lines demonstrated an inhibitory effect of estrogen on the human IL-6 promoter [88], three independent groups have failed to demonstrate an inhibitory effect of estrogen on IL-6 production from human bone cells and stromal cells expressing functional estrogen receptors [89 – 91]. These data raise the possibility that the production of human IL-6 protein does not increase under conditions of estrogen deficiency. This is further supported by a report that, in humans, surgical menopause is not followed by an increase in IL-6, although it causes an increase in soluble IL-6 receptor [92]. Recent studies have unveiled that one of the key mechanism by which estrogen regulates osteoclastogenesis is by modulating the stromal cell production of M-CSF. Under conditions of E2 deficiency the high bone marrow levels of IL-1 and TNF lead to the expansion of a stromal cell population which produces larger amounts of soluble M-CSF [93]. These high M-CSF producing stromal cells have an increased capacity to support osteoclastogenesis (Fig. 2). Interestingly, estrogen has no direct regulatory effects on the production of soluble M-CSF as it regulates M-CSF secretion exclusively by conditioning the differentiation of stromal cells toward a phenotype characterized by a lower production of M-CSF. The high M-CSF producing stromal cells found in estrogen deficient mice are characterized by increased phosphorylation of the transcription factor Egr-1. While Egr-1 binds and sequesters the nuclear protein Sp-1, phosphorylated Egr-1 does not bind to Sp-1. As a result, cells from estrogen-deficient mice are characterized by increased levels of free Sp-1. This protein binds to the M-CSF promoter and stimulates M-CSF gene expression [94] (Fig. 3). In addition to an indirect effect on soluble M-CSF, E2 has been shown to decrease the production of membrane-bound M-CSF via a direct effect on bone marrow cells [95,96]. However, the source of membrane-bound M-CSF which is under estrogen regulation remains to be defined. Regardless of the specific cell involved, estrogen regulates this key osteoclastogenic cytokine with at least two distinct mechanisms. Little information on the effects of menopause on the production of RANKL is currently available. However, the promoter region of the RANKL gene does not contain

89

FIGURE 2 Estrogen regulates the differentiation of stromal cell precursors and leads to the formation of “low” M-CSF producing stromal cells.

regions known to be repressed (directly or indirectly) by estrogen [97]. Therefore it is likely that future studies will confirm the preliminary observation available at the moment, which indicates that estrogen does not regulate RANKL. In contrast, estrogen has been shown to increase the production of OPG in osteoblastic cells [65]. Thus, estrogen enhancement of OPG secretion by osteoblastic cells is likely to represent another major mechanism in explaining the antiresorptive action of estrogen on bone. Another possible intermediate in estrogen action is TGF. This growth factor is a multifunctional protein that is produced by many mammalian cells including osteoblasts and has a wide range of biological activities (see Chapter 14). TGF is a potent osteoblast mitogen [98]. Under specific experimental conditions TGF- decreases both osteoclastic resorptive activity and osteoclast recruitment. Oursler et al. have reported that estrogen increases the steady-state level of TGF- mRNA and release of TGF- protein [66]. This mechanism provides the first example of “positive” effects of estrogen in bone which may result in decreased bone turnover.

D. Effects of Menopause on the Production of Bone Resorbing Cytokines An abundance of in vitro studies that demonstrated the potent effects of IL-1, TNF, and IL-6 on bone prompted a

90

ROBERTO PACIFICI

FIGURE 3 Mechanism by which stromal cells from estrogen deficient mice produce low levels of M-CSF. Stromal cells from estrogen deficient mice exhibit increased CKII-dependent phosphorylation of the nuclear protein Egr-1. Phosphorylated Egr-1 binds less avidly to the transcriptional activator Sp-1 and the resulting higher levels of free Sp-1 stimulates M-CSF gene expression. series of investigations on the relationship between bone remodeling, cytokine production, and osteoporosis (see Chapter 13). These studies were conducted using cultures of peripheral blood monocytes because these cells, when cultured in polystyrene plates with ordinary tissue culture medium (which contains small amounts of (LPS)), express IL-1 and TNF mRNA and secrete small quantities of IL-1 and TNF protein [99 – 101]. Another reason that prompted investigators to select this model is that the secretion of cytokines from peripheral blood monocytes reflects the secretory activity of bone marrow mononuclear cells [101]. This is not surprising because the in vitro production of cytokines from cultured monocytes is a reflection of phenotypic characteristics acquired in response to local stimuli during their maturation in the bone marrow, and these characteristics are maintained after release into the circulation [102]. This phenomenon is thought to play an important role in providing the basis for tissue and functional specificity. Consequently, monocyte cytokine secretion is relevant to postmenopausal bone loss as it mirrors cytokine secretion from marrow resident cells of the monocyte macrophage lineage or monocytes that have homed to bone [103]. This hypothesis was first proved correct by studies of Pioli et al. showing that in Pagetic patients the secretion of IL-1 from blood monocytes correlates with that from bone marrow mononuclear cells [104] as well as from observations in rats and mice, where ovariectomy and estrogen replacement have been found to regulate the bone marrow mononuclear cell production of IL-1 and TNF [101,105]. It is also important to recognize that monocytes are the major source of IL-1 and TNF in the bone marrow [100]. Moreover, the anatomical proximity of mononuclear cells to remodeling loci, the capacity to secrete numerous products

all recognized for their effects on bone remodeling, and the expression of integrin receptors [106], which make these cells capable of adhering to the bone matrix, make them likely candidates as participants in skeletal remodeling. Investigation of the monocytic production of IL-1 led to the discovery that monocytes of patients with “highturnover” osteoporosis secrete higher IL-1 activity than those from both patients with “low turnover” osteoporosis and, indeed, those from normal subjects [72]. Since increased bone turnover is characteristic of the early postmenopausal period, these data suggested the hypothesis that the bone sparing effect of estrogen is related to its ability to block the production of IL-1 from cells of the monocytic lineage. This hypothesis was first tested in studies designed to investigate the effect of natural menopause and estrogen/progesterone replacement on the monocytic production of IL-1. The results showed that IL-1 activity increases after menopause in both normal and osteoporotic women. However, whereas IL-1 activity in normal women spontaneously returned to premenopausal levels within 7 years after menopause, in osteoporotic subjects the increase in IL-1 activity lasted up to 15 years after menopause [73]. As a result, the finding of increased IL-1 activity 8 – 15 years after menopause is characteristic of women with postmenopausal osteoporosis. The data also showed that treatment of women in both the early and the late postmenopausal periods with estrogen and progesterone normalizes IL-1 activity within the first month of treatment. Similar effects of menopause have also been documented for TNF and GM-CSF. The latter is a cytokine recognized as a potent stimulator of osteoclastogenesis [74]. The increased production of cytokines associated with estrogen withdrawal occurs in a secular fashion consistent with

91

CHAPTER 41 Postmenopausal Osteoporosis

a direct causal role of these factors on postmenopausal bone loss. This was demonstrated by analyzing the time course of changes in cytokine secretion and markers of bone turnover in normal women undergoing bilateral ovariectomy. Using this strategy it was demonstrated that the monocytic secretion of GM-CSF increases within 1 week after ovariectomy.

This is followed by a marked increase in TNF and IL-1 at 2 weeks postsurgery. The increase in the latter two cytokines is associated with a concurrent increase in biochemical indices of bone resorption [75] (Fig. 4). Initiation of estrogen replacement therapy at 1 month after ovariectomy results in the rapid normalization of cytokine production [75]. Subsequent

FIGURE 4 Effect of ovariectomy and subsequent estrogen/progesterone replacement therapy on the monocytic production of cytokines and biochemical indices of bone turnover.

92

ROBERTO PACIFICI

studies confirmed that natural and surgical menopause are associated with an increased production of IL-1 and TNF from peripheral blood and bone marrow monocytes [76,107 – 109]. An increased mononuclear cell production of IL-6 has also been reported after ovariectomy [108]. Since IL-1 and TNF are powerful stimulators of IL-6 production [90,110], the latter is likely to reflect the impact of the higher levels of IL-1 and TNF induced by ovariectomy. That the increased monocytic production of cytokines plays a direct role in inducing bone resorption was later demonstrated by Cohen-Solal et al. [111] by examining the bone resorption activity of monocyte supernatants obtained from pre- and postmenopausal women. Using this approach it was found that the culture media of monocytes obtained from postmenopausal women has a higher in vitro bone resorption activity than that from either premenopausal women or estrogen treated postmenopausal women. The increased bone resorption activity of media from postmenopausal subjects is blocked by the addition of IL-1ra and anti-TNF antibody. Recent studies of bone marrow supernatants from estrogen deficient mice by Suda and coworkers have also indicated that IL-1 plays a dominant role in mediating the impact of estrogen withdrawal on bone resorption. Antibodies to IL-1a [the dominant IL-1 species in mice] but not antibodies to many other cytokines completely blocked the bone resorption activity of the monocyte conditioned medium. Antibodies to IL-1b, IL-6, and IL-6 receptor resulted in a partial neutralization of bone resorption activity. Thus, it is likely that IL-1 and TNF account for most of the resorption activity produced by cultured monocytes. Still to be determined is whether this effect is direct or mediated by other factors produced in response to IL-1 and TNF. More direct evidence in favor of cause – effect relationship between increased production of IL-1 and TNF (and IL-6) and postmenopausal osteoporosis is also provided by the findings of Ralston demonstrating that IL-1, TNF, and IL-6 mRNAs are expressed more frequently in bone cells from untreated postmenopausal women than in those from women on estrogen replacement [90].

E. Effect of Menopause on the Production of IL-1 Receptor Antagonist IL-1 bioassays are based on the measurement of the proliferation activity of IL-1-dependent cell lines. Thus, IL-1 bioactivity is stimulated by IL-1 and inhibited by IL-1ra. Therefore, IL-1 bioactivity reflects closely the IL-1/IL-1ra ratio. Since mammalian cells secrete IL-1ra along with IL-1, the regulatory effects of ovarian steroid on IL-1 bioactivity may involve both IL-1 ( or ) and IL-1ra. Studies have addressed this issue and revealed that estrogen and progesterone downregulate the production of both IL-1 and IL-1ra [64]. In contrast, estrogen and progesterone have no inhibitory effects on the secretion of IL-1. Interestingly, in

normal women the decrease in IL-1 bioactivity that accompanies the passage of time since menopause is associated with a parallel increase in the secretion of IL-1ra. Thus, in normal women the increasing production of IL-1ra which accompanies the passage of time since menopause is likely to help restore normal monocytic IL-1 bioactivity after the menopause. IL-1 is a powerful autocrine factor. In fact, the IL-1 produced by monocytes binds to IL-1 receptors expressed on the monocyte surface and further stimulates IL1 secretion [112]. Since this process is inhibited by IL-1ra, the progressive postmenopausal increase in IL-1ra secretion observed in nonosteoporotic women may also help to explain the parallel decrease in the secretion of IL-1b observed in these subjects as time elapses from menopause. As discussed above, in osteoporotic women the production of IL-1 bioactivity is increased for a length of time twice as long as in normal women. This is associated with an increased secretion of IL-1 which persists as long as the increase in IL-1 bioactivity. Interestingly, the levels of IL-1ra measured in osteoporotic women are higher than those of normal women, but do not change with the passage of time since menopause [64]. Thus, in osteoporotic women IL-1 bioactivity appears to be primarily regulated by changes in the production of IL-1. Since only a small fraction of the cytokines released into the bone microenvironment escape into the systemic circulation, studies based on the measurement of serum cytokine levels have been, for the most part, unrewarding. However, the recent development of supersensitive cytokine assays has made it possible to document that the serum IL-1/IL1ra ratio is significantly higher in women with postmenopausal osteoporosis than in their normal counterparts [113]. The use of these sensitive assays has also led to the demonstration that the rate of bone loss in osteoporotic women correlates inversely with serum IL-1ra levels [114]. Taken together, these data demonstrate that a modulatory action of estrogen and progesterone on the secretion of IL1ra contributes to the events of the menopause and the effects of hormone replacement on IL-1 bioactivity. The molecular mechanism by which estrogen and menopause regulate the monocytic production of IL-1ra remains to be defined. The local microenvironment is known to condition the production of IL-1ra. For example, alveolar macrophages from patients with interstitial lung disease produce more IL-1ra than those from normal controls [115]. It is likely, therefore, that the increased bone resorption induced by IL-1 and other cytokines after the menopause may lead to the release in the bone microenvironment of factors which, in turn, stimulate the secretion of IL-1ra. One such a factor is TGF- [116], a constituent of the bone matrix released locally upon activation of osteoclastic bone resorption [117,118]. The differences in the secretory pattern of IL-1ra observed between normal and osteoporotic women could, indeed, result from the more intense bone resorption and the resulting more

93

CHAPTER 41 Postmenopausal Osteoporosis

TABLE 3

Effect of Cytokine Inhibition on Bone Mass and Bone Turnover in Ovariectomized Rats and Mice in the First Month after Surgery (Early Postovariectomy Period) TNFbp

IL-1ra  TNFbp

Anti-IL-6 Ab

Anti-M-CSF Ab





















IL-1ra Prevents ovx-induced bone loss Blocks osteoclast formation Blocks mature osteoclasts











Stimulate bone formation











abundant release of TGF- which characterize the postmenopausal period of women with osteoporosis [119]. Since altered T4/T8 lymphocyte ratio [120,121] and abnormal mixed leukocyte reactions have been reported in osteoporotic patients [122], it is conceivable that specific monocyte phenotypes characterized by the ability to produce constitutively high amounts of IL-1ra may be preferentially expressed in osteoporotic patients. Should this be the case, the difference in IL-1ra levels observed between normal and osteoporotic patients could be related to intrinsic differences in the prevailing monocyte population.

F. Cytokine Inhibitors and Transgenic Mice: Tools For Investigating the Contribution of Candidate Factors to Ovariectomy-Induced Bone Loss Since several cytokines are under hormonal control and exhibit overlapping biological effects, analysis of cytokine

expression and secretion in bone and bone marrow cells is unlikely to provide definite evidence in favor of a cause – effect relationship between increased cytokine production and postmenopausal bone loss. However, direct demonstration that cytokines mediate the impact of estrogen deficiency on bone can be achieved with the use of genetic models and specific cytokine antagonists, such as the IL-1 antagonist, IL-1ra, and the TNF antagonist, TNF binding protein (TNFbp). Lorenzo et al. have shown that mice insensitive to IL-1 due to the lack of IL-1 receptor type I are protected against ovx-induced bone loss [60]. These findings confirmed earlier studies conducted by treating ovariectomized rats with IL-1ra beginning either at the time of surgery (early postovariectomy period) or 4 weeks later (late postovariectomy period) [107]. These experiments revealed that the functional block of IL-1 has distinct effects in both periods (see Fig. 5 and Tables 3 and 4). In fact, in the second month after ovariectomy, treatment with IL-1ra completely blocked bone loss, duplicating the effect of estrogen. In contrast, in the first month after ovariectomy, bone loss

Effect (mean  SEM] of IL-1ra treatment on the bone mineral density (BMD) of the distal femur in ovx rats. Results are expressed as percentage change from baseline. When treatments were started at the time of ovariectomized (left) rats treated with IL-1ra () had a smaller decrease (*P 0.005) in BMD than rats treated with IL-1ra vehicle (). However, a complete prevention of bone loss was obtained with estrogen () but not IL-1ra. When treatments were started 4 weeks after ovariectomy (right) both IL-1ra and estrogen were equally effective in completely preventing additional bone loss.

FIGURE 5

94

ROBERTO PACIFICI

TABLE 4

Effect of Cytokine Inhibition on Bone Mass and Bone Turnover in Ovariectomized Rats and Mice in the Second Month after Ovariectomy (Late Postovariectomy Period) IL-1ra

TNFbp

IL-1ra  TNFbp

Anti-IL-6 Ab

Anti-M-CSF Ab

Prevents ovx-induced bone loss



 a

 a

?

?

Blocks osteoclast formation



 a

 a

?

?

Blocks mature osteoclasts







?

?

Stimulate bone formation







?

?

a

Preliminary unpublished data.

was completely prevented by estrogen replacement therapy and decreased by about 40% by IL-1ra treatment. These findings indicated that cytokines produced independently of IL-1 contribute to induce bone loss in the early postovariectomy period. Since IL-1 and TNF have powerful additive and synergistic effects in many systems, TNF appeared to be the most likely candidate factor. That TNF contributes to bone loss in the early postovariectomy period was demonstrated by treating ovariectomized rats with IL-1ra, TNFbp, and a combination of the two inhibitors for 2 weeks starting at the time of surgery [123]. This critical experiment demonstrated that while treatment with either IL-1ra or TNFbp alone partially prevented ovariectomy-induced bone loss, complete bone sparing was achieved when ovariectomized rats were treated simultaneously with IL-1ra and TNFbp (Fig. 6). The mechanism responsible for the bone sparing effects of IL-1ra and TNFbp was elucidated by analyzing biochemical and histomorphometric indices of bone turnover in the early and late postovariectomy period. In the immediate postovariectomy period, urinary excretion of deoxypyridinium (DPD), a marker of bone resorption which reflects the number and the functional activity of mature osteoclasts [124], was decreased by both IL-1ra and TNFbp. However, an inhibitory effect comparable to that of estrogen was achieved only when ovariectomized rats were treated simultaneously with both inhibitors [123]. In contrast, in the second month after ovariectomy, treatment with IL-1ra alone was sufficient to normalize DPD excretion [107]. While trabecular bone loss prevails in the late postovariectomy period, cortical bone is preferentially lost in mature rats in the early postovariectomy period. This is typically associated with an increase in the number of osteoclasts adhering to endocortical, but not trabecular surfaces. This increase is completely prevented by both IL-1ra and TNFbp [123]. However, there is no additive effects of the two inhibitors, thus suggesting that inhibition of osteoclastogenesis is not the mechanism accounting for the more potent bone sparing effect of combined IL-1ra  TNFbp treatment.

Since IL-1ra and TNFbp have significant additive antiresorptive effects, but have no additive effects in blocking osteoclastogenesis, inhibition of osteoclast activation appears to be the main mechanism of the potent antiresorptive effects of combined treatment with IL-1ra and TNFbp in the early postovariectomy period. In contrast, in the late postovariectomy period treatment with IL-1ra alone is sufficient for completely arresting bone loss and for normalizing the number of adherent osteoclasts [107]. Similarly, since inhibition of both IL-1 and TNF is required to prevent the effects of ovariectomy on bone

FIGURE 6 Effect [mean  SEM] of IL-1ra and TNFbp treatment on the bone mineral density (BMD) of the distal femur in ovx rats. Results are expressed as percentage change from baseline. ****P  0.05 compared to baseline. *P  0.05 compared to BSA treated ovx rats. **P  0.05 compared to BSA-, TNFbp-, and IL-1ra-treated ovx rats.

CHAPTER 41 Postmenopausal Osteoporosis

resorption, IL-1 and TNF likely have independent and redundant stimulatory effects on osteoclast activation, the primary mechanism by which bone resorption acutely increases early after ovariectomy. Inhibition of either IL-1 or TNF is sufficient to block the postovariectomy increase in osteoclastogenesis. Thus, IL-1 and TNF must possess either synergistic or sequential effects on osteoclastogenesis, the primary mechanism underlying long-term elevations of bone resorption [125,126]. Analysis of changes in bone density which follow 1 month treatment with either estrogen or IL-1ra provides further insight on the mechanism of action of IL-1ra [126]. Cessation of estrogen therapy is followed by a rapid resumption of bone loss. In contrast, ovariectomized rats treated with IL-1ra maintain stable bone density for the first 4 weeks after the completion of the treatment. In these rats bone loss resumes only 6 weeks after discontinuation of the IL-1ra treatment. Estrogen regulates both osteoclastogenesis and osteoclast activation, although the short time between estrogen withdrawal and the onset of bone loss points to a “downstream” event, such as osteoclast activation, as the functionally dominant mechanism by which estrogen blocks bone resorption. In contrast, the long interval between stopping IL-1ra treatment and induction of bone loss suggests that in the late postovariectomy period the functional block of IL-1 with IL-1ra affects mainly “upstream” events, such as the proliferation and differentiation of osteoclast precursors. Consequently, it appears that in the last postovariectomy period, IL-1ra is more potent in blocking osteoclastogenesis than in inhibiting osteoclast activation. Since the active life span of rat osteoclasts is only 2.5 days [127], a constant supply of new precursor cells is needed to replenish aged osteoclasts at each resorptive site [128]. Thus, in ovariectomized rats treated with IL-1ra the persistent activation of mature osteoclasts is likely to lead to an impoverishment of those osteoclast precursors downstream from the IL-1-dependent step(s). As a result, the time necessary for reexpansion of the osteoclastic pool is likely to account for the lag time between discontinuation of IL-1ra treatment and the resumption of bone resorption. Histomorphometric studies also demonstrated important effects of IL-1ra and TNFbp on bone formation [107, 129].Two weeks after surgery there were no significant differences in trabecular and cortical bone formation rate between ovariectomy and sham operated rats. Interestingly, however, treatment with either IL-1ra or TNFbp induced a marked increase in bone formation rate in ovariectomized but not in sham-operated rats. This suggests that inhibition of endocortical bone formation resulting from high levels of IL-1 and TNF (characteristic of the early postovariectomy period) counteracts and masks direct stimulatory effects of ovariectomy on bone formation. In contrast, in the late postovariectomy period, bone formation is increased in

95 the trabecular but not in the cortical bone and in this time period neither IL-1ra nor TNFbp have significant effects on this index. Thus, when taken together, the data support the hypothesis that estrogen deficiency modulates bone resorption via an IL-1/TNF-dependent pathway and bone formation via a complex mechanism which involves an IL-1/TNF-independent stimulatory effect and an IL-1/TNF-mediated inhibitory effect. Early after ovariectomy the dominant phenomena mediated by IL-1 and TNF are the stimulation of osteoclast activity and the inhibition of bone formation. As time progresses from ovariectomy, the IL-1- and TNFdependent inhibition of bone formation subsides while the most important effect of these factors becomes the induction of osteoclastogenesis. These initial observations about the causal role of TNF were confirmed by Ammann et al., who reported that transgenic mice insensitive to TNF due to the overexpression of soluble TNF receptor, are also protected against ovxinduced bone loss [62]. Finally, an orally active inhibitor of IL-1 and TNF production was also shown to completely prevent bone loss in ovx rats [130]. Although the finding that functional block of either IL-1 or TNF is sufficient to prevent ovx-induced bone loss may appear to be difficult to reconcile, it should be emphasized that in most biological systems IL-1 and TNF have potent synergistic effects. Thus, the functional block of one of these two cytokines elicits biological effects identical to those induced by the block of both IL-1 and TNF. The longterm stimulation of bone resorption which follows ovx is sustained primarily by an expansion of the osteoclastic pool. Since OC formation is synergistically stimulated by IL-1 and TNF [101], it is not surprising that long-term inhibition of either IL-1 or TNF results in complete prevention of ovx induced bone loss. While studies with transgenic mice and inhibitors of IL-1 and TNF have consistently demonstrated that IL-1 and TNF are key inducers of bone loss in ovx animals, investigations aimed at assessing the contribution of IL-6 to ovxinduced bone loss have yielded conflicting results. In favor of a causal role for IL-6 in ovx induced bone loss is the report of Poli et al., indicating that IL-6 knock out mice are protected against the loss of trabecular bone induced by ovx [61]. By contrast are studies demonstrating that osteoporosis is not a feature of transgenic mice overexpressing IL-6 [131]. Studies have also been conducted by injecting an antibody neutralizing IL-6 in ovx mice. Neutralizing IL-6 prevents the increase in OC formation induced by estrogen deficiency [50,129] but does prevent ovx-induced bone loss and does not decrease in vivo bone resorption [129]. These findings confirm that IL-6 contributes to the expansion of the osteoclastic pool induced by ovx. However, this cytokine does not appear to be the dominant factor in inducing bone loss in estrogen deficient mice.

96

FIGURE 7 Treatment with the anti M-CSF Ab 5A1 Ab prevents ovx induced bone loss. Results (mean  SEM) are expressed as percentage change from baseline *P  0.05 compared to baseline and to any other group. Recent studies have also demonstrated that the functional block of M-CSF by the anti M-CSF antibody 5A1 completely prevents ovx-induced bone loss in mice (Fig. 7) [132]. That M-CSF is another cytokine which plays a key role in ovx induced bone loss was further demonstrated by examining mice lacking the transcription factor Egr-1 [133]. Egr-1-deficient mice produce maximal amounts of M-CSF both in the presence and in the absence of estrogen [94]. Thus, ovx does not stimulate osteoclast formation in these mice, as it fails to further stimulate M-CSF production. Importantly, Egr-1 deficient mice are completely protected against ovx-induced bone loss, a finding that confirms the relevance of M-CSF [132]. No studies have been conducted to determine the effects of ovx in mice insensitive to RANKL, although one would predict that these animal will sustain significant bone loss due to the stimulated production of TNF. The role of IL-1, IL-6, TNF, M-CSF, and RANKL in osteoclastogenesis has been directly investigated using murine bone marrow cultures obtained from ovariectomized mice. Ovariectomy not only increases the number of bone marrow cells, but also increases the number of osteoclasts generated by ex vivo cultures of bone marrow cells [134]. IL-1ra and TNFbp both completely prevent the increase in osteoclastogenesis induced by ovariectomy [129]. Neither IL-1ra nor TNFbp decrease osteoclast formation in

ROBERTO PACIFICI

sham-operated mice. Osteoclast formation is also decreased, in part, by the anti-IL-6 antibody 20F3. However, the anti-IL-6 antibody is less effective than IL-1ra and TNFbp and, more importantly, decreases osteoclastogenesis in both ovariectomized and sham-operated mice [129], indicating that the contribution of IL-6 to osteoclastogenesis does not increase with estrogen deficiency. The formation of osteoclasts in ex vivo cultures of bone marrow cells from ovariectomized mice is also blocked by in vitro treatment with IL-1ra or TNFbp. In contrast, in vitro treatment with the anti-IL-6 antibody 20F3 has no effects in cultures from either ovariectomized or sham-operated mice. Another important difference between these inhibitors is that in vivo treatment with IL-1ra and TNFbp also decreases the urinary excretion of DPD in a manner similar to that of estrogen, whereas the anti-IL-6 antibody does not. In contrast, when in vitro bone resorption is evaluated by examining the effects of the three inhibitors on the formation of resorption lacunae, it appears that IL-1ra, anti-IL-6 antibody, and TNFbp all inhibit the formation of resorption pits [101]. Since the regulatory role of IL-6 is limited to the initial steps of the osteoclast differentiation process [135], it could be that the block of IL-6 in vivo is insufficient to prevent the complete maturation and activation of those cells which are downstream with respect to the IL-6 -dependent steps According to this hypothesis, the lack of change in DPD excretion with anti-IL-6 antibody treatment would reflect the maintenance of an unaltered pool of active, mature osteoclasts. Conversely, the decreased pit formation observed with the IL-6 block is likely to reflect the decreased bone marrow content of osteoclast precursors and the resulting decrease in the number of cells which reach functional maturity in vitro [32]. From these data it appears reasonable to hypothesize that inhibition of IL-1 and TNF blocks bone resorption in vivo and in vitro because, at least in rodent, these cytokines regulate early and late steps of osteoclast maturation. In vivo treatment of ovx mice with anti M-CSF antibody prevents the effects of ovx on ex-vivo osteoclast formation and bone resorption [132], in a manner similar to treatment with IL-1 and TNF antagonists. These data are consistent with the notion that estrogen deficiency increases M-CSF production indirectly, via an IL-1 and TNF-mediated mechanism [93, 94]. Recent studies have demonstrated that the presence of severe osteopetrosis due to complete lack of osteoclasts in mice lacking either RANKL or the RANKL receptor RANK. However, the effects of ovariectomy and/or estrogen deficiency in these animals remain to be investigated.

III. SUMMARY AND CONCLUSIONS The mechanism(s) of the bone sparing effects of estrogen appears to be particularly complex as it involves the

CHAPTER 41 Postmenopausal Osteoporosis

regulated production of cytokines from hematopoietic cells and bone cells [103,136] and the responsiveness of stromal cells to these cytokines. In addition, the contribution of specific factors to postmenopausal bone loss appears to vary as the system adapts over time to the hormonal withdrawal. Although many details of this process remain to be defined, it is now clearly established that estrogen downregulates the production of proosteoclastogenic and antiosteoclastogenic factors. Among them are IL-1, IL-6, TNF, M-CSF, OPG, and TGF-. Uncertainty remains over the role [if any] of RANKL in the mechanism by which estrogen prevents bone loss. The exact contribution of IL-6 also remains unclear because of insufficient data demonstrating that the block of IL-6 decreases bone resorption in vivo and bone loss in a bona fide experimental model of postmenopausal osteoporosis. However, the exact role of RANKL and IL-6 is likely to be defined in the near future. At the present time RANKL and M-CSF should be regarded as the factors responsible for osteoclast renewal under unstimulated conditions. The enhanced osteoclastogenesis and the increased osteoclastic bone resorption leading to postmenopausal bone loss results from stimulated production of inflammatory cytokines. These factors increase the production of M-CSF and RANKL and, in the case of TNF, directly synergize with RANKL to maximize osteoclast formation. Remarkable progress has been accomplished in clarifying the mechanism of the bone sparing effect of estrogen in animal models. A more challenging task will be to demonstrate the relevance of the mechanisms described above in human subjects.

References 1. G. A. Greendale and H. L. Judd, The menopause: Health implications and clinical management. Geriatrics Soc. 41, 426–436 (1993) [Review]. 2. R. J. Chang and H. L. Judd, The ovary after menopause. Clin. Obstet. Gynecol. 24, 181 – 191 (1981). 3. H. W. Cole, S. Chandrasekhar, H. M. Hsuing, M. L. Heiman, and M. Sato, Beneficial effects of ethynyl estradiol (EE2) and raloxifene in the ovariectomized (ovx) rat are dependent on an intact pituitary. J. Bone Miner. Res. 9, Abs 57 (1994). 4. R. F. Power, S. K. Mani, J. Codina, O. M. Conneely, and B. W. O’Malley, Dopaminergic and ligand-independent activation of steroid hormone receptors. Science 254, 1636 – 1639 (1991). 5. National Institutes of Health Consensus Conference. J. Am. Med. Assoc. 252, 799 – 802 (1984). 6. B. L. Riggs and L. J. Melton, Medical progress: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1684 (1986). 7. B. L. Riggs and L. J. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1986). 8. L. Nilas and C. Christiansen, Bone mass and its relationship to age and the menopause. J. Clin. Endocrinol. Metab. 65, 697 – 702 (1987). 9. B. L. Riggs, H. W. Wahner, W. L. Dunn, R. B. Mazess, K. P. Offord, and L. J. Melton, Differential changes in bone mineral density of the appendicular and the axial skeleton with aging. J. Clin. Invest. 67, 328 – 335 (1981).

97 10. C. Slemenda, S. L. Hui, C. Longcope, and C. C. Johnston, Sex steroids and bone mass. A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 11. H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gordan, Quantitative computed tomography of vertebral spongiosa: A sensitive method for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). 12. C. C. Johnston, Jr., S. L. Hui, R. M. Witt, R. Appledorn, R. S. Baker, and C. Longcope, Early menopausal changes in bone mass and sex steroids. J. Clin. Endocrinol. Metab. 61, 905 – 911 (1985). 13. H. Ohta, K. Makita, Y. Suda, T. Ikeda, T. Masuzawa, and S. Nozawa, Influence of oophorectomy on serum levels of sex steroids and bone metabolism and assessment of bone mineral density in lumbar trabecular bone by QCT-C value. J. Bone Miner. Res. 7, 659 – 665 (1992). 14. H. Ohta, T. Ikeda, T. Masuzawa, K. Makita, Y. Suda, and S. Nozawa Differences in axial bone mineral density, serum levels of sex steroids, and bone metabolism between postmenopausal and age- and body size-matched premenopausal subjects. Bone 14, 111 – 116 (1993). 15. B. Ettinger, H. K. Genant, and E. C. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Intern. Med. 102, 319 – 324 (1985). 16. R. Lindsay, D. M. Hart, C. Forrest, and C. Baird, Prevention of spinal osteoporosis in oophorectomised women. Lancet 2, 1151 – 1154 (1980). 17. A. Horsman, M. Jones, R. Francis, and B. E. C. Nordin The effect of estrogen dose on postmenopausal bone loss. N. Eng. J. Med. 309, 1405 – 1407 (1983). 18. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 19. S. C. Manolagas, and R. L. Jilka, Bone marrow, cytokines, and bone remodeling. N. Engl. J. Med. 332, 305 – 311 (1995). 20. J. Chow, J. H. Tobias, K. W. Colston, and T. J. Chambers, Estrogen maintains trabecular bone volume in rats not only by suppression of bone resorption but also by stimulation of bone formation. J. Clin. Invest. 89, 74 – 78 (1992). 21. S. D Bain, S. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High-dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 22. R. Pacifici, Estrogen, cytokines and pathogenesis of postmenopausal osteoporosis. J. Bone Miner. Res. 11, 1043 – 1051 (1996). 23. G. D. Roodman, Advances in bone biology: the osteoclast. Endocr. Rev. 17, 308 – 332 (1996). 24. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 25. I. R. Dissanayake, G. R. Goodman, E. R. Baker, S. Pun, L. Ciui, Y. A. Ma, S. S. Jee.W, S. N. Popoff, and S. Epstein, The role of the B lymphocyte in estrogen deficient and normal bone. J. Bone Miner. Res. 12 (Suppl. 1), F467-(Abs.) (1997). 26. R. G. Erben, S. Raith, J. Eberle, and M. Stangassinger, Ovariectomy augments B lymphopoiesis and generation of monocytemacrophage precursors in rat bone marrow. Am. J. Physiol. 274, E476 – 483 (1998). 27. T. Masuzawa, C. Miyaura, Y. Onoe, K. Kusano, H. Ohta, S. Nozawa, and T. Suda. Estrogen deficiency stimulates B lymphopoiesis in mouse bone marrow. J. Clin. Invest. 94, 1090 – 1097 (1994). 28. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu,

98

29.

30.

31.

32. 33.

34.

35.

36.

37.

38.

39.

40.

41.

ROBERTO PACIFICI J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, RANKL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998). B. R. Macdonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long-term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 232 (1986). T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endocr. Rev. 13, 66 – 80 (1992). S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin. Invest. 91, 257 – 263 (1993). W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997) [see comments]. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci USA 96, 3540 – 3545 (1999) [in process citation]. R. Felix, M. G. Cecchini, and H. Fleish, Macrophage colonystimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 127, 2592 – 2597 (1990). H. Kodama, M. Ose, S. Niida, and A. Yamasaki, Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J. Exp. Med. 173, 1291 – 1294 (1991). K. Fuller, J. M. Owens, C. J. Jagger, A. Wilson, R. Moss, and T. J. Chambers, Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J. Exp. Med. 178, 1733 – 1744 (1993). U. Sarma and A. M. Flanagan, Macrophage colony-stimulating factor induces substantial osteoclast generation and bone resorption in human bone marrow cultures. Blood 88, 2531 – 2540 (1996). T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). G. Hattersley, E. Dorey, M. A. Horton, and T. J. Chambers, Human macrophage colony-stimulating factor inhibits bone resorption by osteoclasts disaggregated from rat bone. J. Cell. Physiol. 137, 199 – 203 (1988).

42. M. Edwards, U. Sarma, and A. M. Flanagan, Macrophage colonystimulating factor increases bone resorption by osteoclasts disaggregated from human fetal long bones. Bone 22, 325 – 329 (1998). 43. R. L. Lees and J.N. Heersche, Macrophage colony stimulating factor increases bone resorption in dispersed osteoclast cultures by increasing osteoclast size. J. Bone Miner. Res. 14, 937 – 945 (1999). 44. M. Gowen, D. D. Wood, E. J. Ihrie, M. K. B. McGuire, and R. G. G. Russell, An interleukin-1-like factor stimulates bone resorption in vitro. Nature 306, 378 – 380 (1983). 45. J. A. Lorenzo, S. L. Sousa, C. Alander, L. G. Raisz, and C. A. Dinarello, Comparison of the bone-resorbing activity in the supernatants from phytohemaglutinin-stimulated human peripheral blood mononuclear cells with that of cytokines through the use of an antiserum to interleukin 1. Endocrinology 121, 1164 – 1170 (1987). 46. M. Gowen, D. D. Wood, and R. G. G. Russell, Stimulation of the proliferation of human bone cells in vitro by human monocyte products with interleukin-1 activity. J. Clin. Invest. 75, 1223 – 1229 (1985). 47. E. Canalis, Interleukin-1 has independent effects on deoxyribonucleic acid and collagen synthesis in cultures of rat calvariae. Endocrinology 118, 74 – 81 (1986). 48. P. Stashenko, F. E. Dewhirst, M. L. Rooney, L. A. Desjardins, and J.D. Heeley, Interleukin-1  is a potent inhibitor of bone formation in vitro. J. Bone Miner. Res. 2, 559 – 565 (1987). 49. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblasts in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 50. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: mediation by Interleukin6. Science 257, 88 – 91 (1992). 51. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993). 52. D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumor necrosis factor. Nature 319, 516 – 518 (1986). 53. B. M. Thomson, G. R. Mundy, and T. J. Chambers, Tumor necrosis factor  and  induce osteoblastic cells to stimulate osteoclastic bone resorption. J. Immunol. 138, 775 – 779 (1987). 54. L. C. Hofbauer, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, B. L. Riggs, and S. Khosla, Interleukin-1beta and tumor necrosis factoralpha, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells Bone 25, 255 – 259 (1999). 55. C. A. O’Brien, I. Gubrij, S. C. Lin, R. L. Saylors, and S. C. Manolagas, STAT3 activation in stromal/osteoblastic cells is required for induction of the receptor activator of NF-kappaB ligand and stimulation of osteoclastogenesis by gp130-utilizing cytokines or interleukin-1 but not 1,25-dihydroxyvitamin D3 or parathyroid hormone. J. Biol. Chem. 274, 19301 – 19308 (1999). 56. C. Thery, E. R. Stanely, and M. Mallat, Interleukin 1 and tumor necrosis factor-a stimulate the production of colony-stimulating factor 1 by murine astrocytes. J. Neurochem. 59, 1183 – 1186 (1992). 57. W. E. Fibbe, J. Van Damme, A. Billiau, H. M. Goselink, P. J. Vogt, P. R. Van Eeden, B. W. Altroch, and J.H.F. Falkenburg Intereleukin-1 induces human marrow stromal cells in long term culturs to produce G-CSF and M-CSF. Blood 71, 431 – 435 (1988). 58. H. Takai, M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, and Y. Yamada, Transforming growth factor-beta stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells. J. Biol. Chem. 273, 27091 – 27096 (1998).

99

CHAPTER 41 Postmenopausal Osteoporosis 59. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL – RANK interaction. J. Exp. Med. 191, 275 – 286 (2000). 60. J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, G. Gronowicz, J. Kalinowski, and C. C. Pilbeam, Mice lacking the type I interleukin-1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). 61. V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Ciliberto, G. A. Rodan, and F. Costantini. Interleukin6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994). 62. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vassalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1703 (1997). 63. P. Isomaki and J. Punnonen, Pro- and anti-inflammatory cytokines in rheumatoid arthritis. Ann. Med. 29, 499 – 507 (1997). 64. R. Pacifici, J. L. Vannice, L. Rifas, and R. B, Kimble. Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: effect of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). 65. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370 (1999). 66. M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). 67. J. J. Weusten, M. A. Blankenstein, F. H. Gmelig-Meyling, H. J. Schuurman, L. Kater, and J. H. Thijssen. Presence of estrogen receptors in human blood mononuclear cells and thymocytes. Acta Endocrinol. 112, 409 – 414 (1986). 68. S. Gulshan, A. B. McCruden, and W. H. Stimson, Estrogen receptor in macrophages. Scand. J. Immunol. 31, 691 – 697 (1990). 69. M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Natl. Acad. Sci. USA 88, 6613 – 6617 (1991). 70. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 71. J. H. M. Cohen, L. Danel, G. Cordier, S. Saez, and J.-P. Revillard, Sex steroid receptor in peripheral T cells: Absence of androgen receptors and restriction of estrogen receptor to CD8 positive cells. J. Immunol. 131, 2767 – 2772 (1983). 72. R. Pacifici, L. Rifas, S. Teitelbaum, E. Slatopolsky, R. McCracken, M. Bergfeld, W. Lee, L. V. Avioli, and W. A. Peck, Spontaneous release of interleukin 1 from human blood monocytes reflects bone formation in idiopathic osteoporosis. Proc. Natl. Acad. Sci. USA 84, 4616 – 4620 (1987). 73. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurtry, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). 74. R. Pacifici, C. Brown, L. Rifas, and L. V. Avioli, TNF and GM-CSF secretion from human blood monocytes: Effect of menopause and estrogen replacement. J. Bone Miner. Res. 5, Abs 145 (1990). 75. R. Pacifici, C. Brown, E. Puscheck, E. Friedrich, E. Slatopolsky, D. Maggio, R. McCracken, and L. V. Avioli, Effect of surgical menopause and estrogen replacement on cytokine release from

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

human blood mononuclear cells. Proc. Natl. Acad. Sci. USA 88, 5134 – 5138 (1991). C. E. Fiore, Falcidia.E, R. Foti, M. Motta, and C. Tamburino, Differences in the time course of the effects of oophorectomy in women on parameters of bone metabolism and Interleukin-1 levels in the circulation. Bone Miner. 20, 79 – 85 (1993). M. L. Polan, A. Daniele, and A. Kuo, Gonadal steroids modulate human monocyte interleukin-1 (IL-1) activity. Fertil. Steril. 49, 964 – 968 (1988). T. Sunyer, J. Lewis, and P. Osdoby, Estrogen decreases the steady state levels of the IL-1 signaling receptor (type I) while increasing those of the IL-1 decoy receptor (type II)] in human osteoclast-like cells. J. Bone Miner. Res. 12(Suppl. 1), Abs131 (1997). G. G. Kuiper, E. Enmark, M. Pelto-Huikko, S. Nilsson, and J. A. Gustafsson, Cloning of a novel receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. U.S.A. 93, 5925 – 5930 (1996). S. Mosselman, J. Polman, and R. Dijkema, ER beta: Identification and characterization of a novel human estrogen receptor. FEBS Lett. 392, 49 – 53 (1996). G. B. Tremblay, A. Tremblay, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, F. Labrie, and V. Giguere, Cloning, chromosomal localization, and functional analysis of the murine estrogen receptor beta. Mol. Endocrinol. 11, 353 – 365 (1997). J. F. Couse, J. Lindzey, K. Grandien, J. A. Gustafsson, and K. S. Korach, Tissue distribution and quantitative analysis of estrogen receptor- alpha (ERalpha) and estrogen receptor-beta (ERbeta) messenger ribonucleic acid in the wild-type and ERalpha-knockout mouse. Endocrinology 138, 4613 – 4621 (1997). K. Paech, P. Webb, G. G. Kuiper, S. Nilsson, J. Gustafsson, P. J. Kushner, and T. S. Scanlan, Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites [see comments]. Science 277, 1508 – 1510 (1997). S. Srivastava, M. N. Weitzmann, S. Cenci, F. P. Ross, S. Adler, and R. Pacifici. Estrogen decreases TNF gene expression by blocking JNK activity and the resulting production of c-jun and junD. J. Clin. Invest. 104, 503 – 513 (1999). J. An, R. C. Ribeiro, P. Webb, J. A. Gustafsson, P. J. Kushner, J. D. Baxter, and D. C. Leitman, Estradiol repression of tumor necrosis factor-alpha transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc. Natl. Acad. Sci. USA 96, 15161 – 15166 (1999). S. T. Pottratz, T. Bellido, H. Mocharia, D. Crabb, and S. Manolagas, 17-Estradiol inhibits expression of human interleukin-6 promoterreporter constructs by a receptor-dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). A. Ray, K. E. Prefontaine, and P. Ray, Down-modulation of interleukin-6 gene expression by 17 beta-estradiol in the absence of high affinity DNA binding by the estrogen receptor. J. Biol. Chem. 269, 12940 – 12946 (1994). B. Stein and M. X. Yang, Repression of the interleukin-6 promoter by estrogen receptor is mediated by NF-kappa-B and C/EBP-BETA. Mol. Cell. Biol. 15, 4971 – 4979 (1995). D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumour necrosis factor but not interleukin 6 from adult human osteoblasts in vitro. Osteoporosis Int. 2, 94 – 102 (1992). L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 and tumor necrosis factor-: lack of regulation by 17-estradiol. Endocrinology 130, 2528 – 2534 (1992). L. Rifas, J. S. Kenney, M. Marcelli, R. Pacifici, L. L. Dawson, S. Cheng, and L. V. Avioli, Production of interleukin-6 in human osteoblasts and human bone marrow stromal cells: Evidence that induction by interleukin-1 and tumor necrosis factor-  is not regulated by ovarian steroids. Endocrinology 136, 4056 – 4067 (1995).

100 92. G. Girasole, M. Pedrazzoni, N. Giuliani, G. Passeri, and M. Passeri, Increased serum soluble interleukin-6 receptors levels are induced by ovariectomy, prevented by estrogen replacement and reversed by alendronate administration. J. Bone Miner. Res. 10, A86 (1995). 93. R. B. Kimble, S. Srivastava, F. P. Ross, A. Matayoshi, and R. Pacifici, Estrogen deficiency increases the ability of stromal cells to support osteoclastogenesis via an IL-1 and TNF mediated stimulation of M-CSF production. J. Biol. Chem. 271, 28890 – 28897 (1996). 94. S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and iots interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). 95. U. Sarma, M. Edwards, K. Motoyoshi, and A. M. Flanagan, Inhibition of bone resorption by 17beta-estradiol in human bone marrow cultures. J. Cell Physiol. 175, 99 – 108 (1998). 96. C. K. Lea, U. Sarma, and A. M. Flanagan. Macrophage colony stimulating-factor transcripts are differentially regulated in rat bone-marrow by gender hormones. Endocrinology 140, 273 – 279 (1999). 97. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/RANKL/ODF gene. Biochim. Biophys. Acta. 1445, 134 – 141 (1999). 98. M. J. Oursler, M. J. Osteoclast synthesis and secretion and activation of latent transforming growth factor beta. J. Bone Miner. Res. 9, 443 – 452 (1994). 99. R. C. Fuhlbrigge, D. D. Chaplin, J. -M. Kiely, and E. R. Unanue, Regulation of interleukin 1 gene expression by adherence and lipopolysaccharide. J. Immunol. 138, 3799 – 3802 (1987). 100. C. A. Dinarello, Interleukin-1 and its biologically related cytokines. Adv. Immunol 44, 153 – 205 (1989). 101. R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). 102. A. L. Witsell and L. B. Schook, Macrophage heterogeneity occurs through a developmental mechanism. Proc. Natl. Acad. Sci. USA 88, 1963 – 1967 (1991). 103. M. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). 104. G. Pioli, G. Girasole, M. Pedrazzoni, P. Sansoni, L. Davoli, G. Ciotti, A. Mantovani, and M. Passeri, Spontaneous release of Interleukin 1 (IL-1) from medullary mononuclear cells of Pagetic subjects. Calcified Tissue International 45, 257 – 259 (1989). 105. R. B. Kimble, A. B. Matayoshi, J. L. Vannice, and R. Pacifici, Long-term treatment with IL-1 receptor antagonist (IL-1ra) blocks bone loss in ovariectomized rats. J. Bone Miner. Res. 55, 260 – 265 (1994). 106. R. O. Hynes, Integrins; Versatility, modulation, and signaling in cell adhesion, Cell 69, 11 – 25 (1992). 107. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 108. G. Pioli, G. Basini, M. Pedrazzoni, G. Musetti, V. Ulietti, D. Bresciani, P. Villa, A. Bacci, D. Hughes, G. Russell, and M. Passeri, Spontaneous release of Interleukin-1 and interleukin-6 by peripheral blood monocytes after ovariectomy. Clin. Sci. 83, 503 – 507 (1992). 109. T. Matsuda, K. Matsui, Y. Shimakoshi, Y. Aida, and S. Hukuda, 1-Hydroxyethilidene-1, 1-bisphosphonate decreases the postovariectomy-enhanced interleukin 1 production by peritoneal macrophages in adult rats. Calc. Tissue Int. 49, 403 – 406 (1991). 110. D. L. Lacey, E. L. Grosso, S. A. Moser, J. Erdmann, H. L. Tan, R. Pacifici, and D. T. Villareal, IL-1 induced murine osteoblast IL-6

ROBERTO PACIFICI

111.

112.

113.

114.

115.

116. 117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

production is mediated by the type I IL-1receptor and is increased by 1,25 dihydroxyvitamin D3. J. Clin. Invest. 91, 1731 – 1742 (1993). M. E. Cohen-Solal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. de Vernejoul, Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. Metab. 77, 1648 – 1653 (1993). C. L. Dinarello, T. Ikejima, S. J. C. Warner, S. F. Orencole, G. Lonneman, J. G. Cannon, and P. Libby, Interleukin-1 induces interleukin-1. I. Induction of circulating interleukin-1 in rabbits in vivo and in human mononuclear cells in vitro. J.Immunol. 139, 1902 – 1908 (1991). S. Khosla, J. M. Peterson, K. Egan, J. D. Jones, and B. L. Riggs. Circulating cytokine levels in osteoporotic and normal women, J. Clin. Endocrinol. Metab. 79, 707 – 711 (1994). R. A. Hannon, A. Blumshon, H. Galatti, H. Fenner, R. G. G. Russell, and R. Eastell, Cytokine receptors, receptor antagonists and postmenopausal bone loss. J. Bone Miner. Res. 8, 143A (1993). B. Galve-de Rochemonteix, L. P. Nicod, A. F. Junod, and J. M. Dayer, Characterization of a specific 20- to 25-kD interleukin-1 inhibitor from cultured human lung macrophages. Am. J. Respir. Cell Mol. Biol. 3, 355 – 361 (1992). W. P. Arend, Interleukin 1 receptor antagonist: A new member of the interleukin 1 family. J. Clin. Invest. 88, 1445 – 1451 (1991). J. Pfeilschifter, B. M. Seyedin, and G. R. Mundy, Transforming growth factor beta inhibits bone resorption in fetal rat long bone cultures. J. Clin. Invest. 82, 680 – 685 (1988). R. O. C. Oreffo, G. R. Mundy, S. M. Seyedin, and L.F. Bonewald, Activation of the bone derived TGF beta complex by isolated osteoclasts. Biochem. Biophys. Res. Commun. 158, 817 – 823 (1989). P. D. Delmas, Biochemical markers of bone turnover in osteoporosis. In “Osteoporosis” (B. L. Riggs and L. J. Melton, eds.) pp. Raven Press, New York, 297 – 316 (1988). C. J. Rosen, K. Usiskin, M. Owens, C. O. Barlascini, M. Belsky, and R. A. Adler, T lymphocyte surface antigen markers in osteoporosis. J Bone. Miner. Res. 5, 851 – 855 (1990). Y. Imai, T. Tsunenari, M. Fukase, and T. Fujita, Quantitative bone histomorphometry and circulating T lymphocyte subsets in postmenopausal osteoporosis. J. Bone Miner. Res. 5, 393 – 399 (1990). J. S. Duke-Cohan, H. Weinberg, R. Sharon, and D. Nahor, Immunological function in osteoporosis. Clin. Immunol. Immunopathol. 35, 125 – 129 (1985). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin1 and tumor necrosis factor is required to completely prevent bone loss in the early post-ovariectomy period. Endocrinology 136, 3054 – 3061 (1995). D. Uebelhart, A. Schlemmer, J. S. Johansen, E. Gineyts, C. Christiansen, and P. D. Delmas, Effect of menopause and hormone replacement therapy on the urinary excretion of pyridinium crosslinks. J. Clin. Endocrinol. Metab. 72, 367–373 (1992). T. J. Wronski, L. M. Dann, H. Qi, and C. F. Yen, Skeletal effects of withdrawal of estrogen and diphosphonate treatment in ovariectomized rats. Calcif. Tissue Int. 53, 210 – 216 (1993). R. B. Kimble, J. L. Vannice, C. Brownfield, and R. Pacifici, Persistent bone-sparing effect of interleukin-1receptor antagonist: A hypothesis on the role of IL-1 in ovariectomy induced bone loss. Calcif. Tissue Int. 55, 260 – 265 (1994). R. Baron, L. Neff, P. Tran Van, J. R. Nefussi, and A. Vignery, Kinetic and cytochemical identification of osteoclast precursors and their differentiation into multinucleated osteoclasts. Am. J. Pathol. 122(2), 36 3 – 378 (1986). A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures.

CHAPTER 41 Postmenopausal Osteoporosis In “Osteoporosis” (B. L. Riggs ed.), Raven Press, New York, pp. 45 – 93 (1988). 129. R. Kimble, S. Bain, and R. Pacifici, The functional block of TNF but not of IL-6 prevents bone loss in ovariectomized mice. J. Bone Miner. Res. 12, 935 – 941 (1997). 130. J. N. Bradbeer, S. J. Stroup, J. C. Hoffman, and M. Gowen, An orally active inhibitor of cytokine synthesis prevents bone loss in the ovariectomized rat. J. Bone Miner. Res. 11 (Suppl. 1), Abs123 (1996). 131. H. Kitamura, H. Kawata, F. Takahashi, Y. Higuchi, T. Furuichi, and H. Ohkawa, Bone marrow neutrophilia and suppressed bone turnover in human Interleukin-6 transgenic mice. Am. J. Pathol. 147, 1682 – 1692 (1995).

101 132. S. Cenci, M. N. Weitzmann, and R. Pacifici, M-CSF neutralization and Egr-1 deficiency prevent ovariectomy induced bone loss. J. Clin. Invest. 105, 1279 – 1287 (2000). 133. T. E. Wilson, M. L. Day, T. Pexton, K. A. Padgett, M. Johnston, and J. Milbrandt, In vivo mutational analysis of NGFI-A zinc fingers. J. Biol. Chem. 267, 3718 – 3724 (1992). 134. D. N. Kalu, Proliferation of tartrate-resistant acid phosphatase positive multinucleate cells in ovariectomized animals. Proc. Soc. Exp. Biol. Med. 195, 70 – 74 (1990). 135. G. D. Roodman, Interleukin-6: An osteotropic factor? J. Bone Miner. Res. 7, 475 – 478 (1992). 136. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994).

CHAPTER 42

Osteoporosis in Men Epidemiology, Pathophysiology, and Clinical Characterization ERIC S. ORWOLL AND ROBERT F. KLEIN Bone and Mineral Research Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201

I. II. III. IV. V. VI.

VII. VIII. IX. X.

Introduction Fractures in Men The Major Determinants of Skeletal Health in Men Osteoporosis The Evaluation of Osteoporosis in Men Therapy

I. INTRODUCTION

II. FRACTURES IN MEN A. The Incidence of Fractures

Although osteoporosis has long been considered a disease of women, the earliest reports of the epidemiology of fractures associated with osteoporosis clearly showed that the classical age-related increase in fractures seen in women was also evident in men. In the past few years it has been recognized that the problem of osteoporosis in men represents an important public health issue [1] as well as a huge personal burden for those men who are affected [2]. It also presents a unique array of scientific challenges and opportunities [3 – 5]. Here we examine the issue of osteoporosis in men and compare its pathophysiology and clinical presentation to parallel processes in women.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Hypogonadism Alcoholism Tobacco Renal Stone Disease References

The incidence of all fractures is higher in men than women from adolescence through middle life [6 – 8] (Fig.1), and the personal and economic impact of these early life fractures is enormous. The average number of hospitalizations for fractures in men between the ages of 18 and 44 years was 181,000 in the United States (1985 – 1988), and the annual number of lost work days for men due to fractures was 17,543,000 [9]. Despite the importance of early life fractures in men, little has been done to understand their causation. Many result from serious trauma, but to some extent relative bone fragility may

103

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

104

ORWOLL AND KLEIN

FIGURE 1 Average annual fracture incidence rate in males and females per 10,000 population, by age group [512].

contribute to fracture risk during this period. For instance, recent long-term follow-up of men who had sustained traumatic tibial or forearm fractures in early mid-life revealed that they were at much greater risk for later hip fracture [10]. At about age 40 – 50 years there is a reversal of this trend, with fractures in general, and in particular those of the pelvis, humerus, forearm, and femur, becoming much more common in women. However, the incidence of fractures due to minimal-to-moderate trauma (particularly hip and spine) also increases rapidly with aging in men (Fig. 2) and reflects an increasing prevalence of skeletal fragility. 1. PROXIMAL FEMUR The proximal femur is the most important site of osteoporotic fracture, about which the most complete epidemio-

FIGURE 2

logical data are available. The incidence of hip fracture rises exponentially in men with aging, as it does in women. However, the age at which the increase begins is slightly older (5 – 10 years) in men [11]. In U.S. men older than 65 years, the incidence of hip fracture is 4 – 5/1000 [12,13], compared to 8 – 10/1000 in similarly aged U.S. women. A 2 – 3:1 female:male ratio has also been reported in northern Europe and Australia, although in other geographic areas the ratio has been noted to be much lower [8,14]. In southern Europe and other areas, the incidence of hip fracture is relatively lower in both sexes, and men have as many hip fractures as do women [15 – 17], and in Asian populations the male:female incidence may actually be quite low or reversed [18,19]. Since there are fewer older men than women, the absolute number of hip fractures tends to be proportionately less in men (of those experiencing their first hip fracture 65 years or older, 165,000 in men vs 580,000 in women in the United States, in 1984 – 1987, or 22% of the total in men) [13]. In the United States (Rochester) the lifetime risk from age 50 onward of a hip fracture has been calculated to be 6% in men and 17.5% in women [20] and 2.4% in men and 9% in women in Canada (Saskatchewan and Manitoba) [21]. It is estimated that approximately 30% of hip fractures worldwide will occur in men [22]. Unfortunately, the number of hip fractures is projected to increase dramatically as the elderly population expands [23 – 25]. In some populations increasing fracture incidence in men has been noted (United States and northern Europe), whereas in many areas the rate of hip fracture in women and men appears to have stabilized, or even declined [15, 24,26 – 29]. Perhaps as a result of a higher prevalence of concomitant disease [30], mortality associated with a hip fracture in elderly men (75 years) is considerably higher than in

Age-specific incidence rates of hip, vertebral, and Colle’s fracture in Rochester, Minnesota, men and women [513].

105

CHAPTER 42 Osteoporosis in Men

FIGURE 3

Observed and expected race- and sex-specific survival following fracture of the hip, all ages combined [514].

women [31 – 34] (Fig. 3). In Europe the incidence of fracture is at least twice as great in women, but the death rates for femoral neck fractures are approximately equal, again suggesting a greater mortality risk in men [35]. Ethnic differences in the incidence of hip fracture in men are substantial. For instance, African-American men experience hip fractures at a rate only half that of Caucasians [13]. Interestingly, whereas African-American women are at significantly lower risk for hip fracture than Caucasian women, African-American men are less protected [11,36]. There are not extensive comparative data concerning other ethnicities, but Asian men have considerably lower incidence rates than in Caucasian populations [19,37]. Geography influences hip fracture rates for unclear reasons. The incidence of fracture is higher in urban than in rural men [38], and there are great variations in the incidence and sex ratio of hip fractures in southern Europe [39]. These differences are presumably the results of a mix of genetic and environmental factors. 2. VERTEBRAE Vertebral fracture is also an important sequel of osteoporosis. As in women, the presence of vertebral fracture in men is associated with loss of height, kyphosis, increased risk of other fractures, and increased disability [40, 41]. Since diagnostic criteria for a vertebral fracture are unsettled, and vertebral fracture infrequently results in hospitalization, consistent epidemiological information is somewhat limited. Previously considered uncommon in men, recent information suggests that the incidence of osteoporotic vertebral fracture in U.S. men is about half that in women (similar to hip fractures) [37,42 – 44]. Until about age 65, the prevalence of vertebral fracture is actually higher in men than women [45,46], and to some extent this

represents an increase in the occurrence of early life trauma in men [47]. In fact, vertebral and femoral bone mineral density values are lower in men with vertebral fractures than in nonfractured controls [42,48], indicating that vertebral fracture in men is not merely the result of a higher rate of trauma, but is also related to a low bone mass. Fractures are primarily in low thoracic vertebrae in men, but are found at all levels. Most fractures are anterior compression in type [42], with vertebral crush fractures occurring less frequently than in women. Vertebral epiphysitis (Scheuermann’s disease) is an uncommon cause of significant vertebral deformity in men [42]. As in women, the presence of vertebral deformity has adverse consequences on mortality [32], and the occurrence of a vertebral fracture indicates a much higher likelihood of sustaining other osteoporotic fractures [29]. 3. OTHER FRACTURES Other fractures (radius/ulna, humerus, pelvis, femoral shaft) share a common epidemiological pattern. Men experience more of these in youth, but with unusual exceptions (e.g., humerus) the incidence remains relatively stable during mid-life, while rising markedly in women [26,49 – 51]. It is only later (75 years) that the incidence of limb fractures begins to rise in men, and it then does so rapidly [49]. This increase is due primarily to an increase in lower limb fractures, while upper extremity fractures do not change as much. In older men, as in women, the likelihood of underlying bone pathology or propensity to fall (e.g., alcoholism) increased markedly [49]. Importantly, the occurrence of a distal forearm fracture [52,53] or a tibial fracture [10] in a man indicates a considerably increased risk of subsequent hip fracture, presumably as a result of low bone mass and/or an increased risk of falling.

106

ORWOLL AND KLEIN

B. The Determinants of Fracture 1. BONE MASS

FIGURE 4

Os calcis bone mineral content of men by 5-year age groups and grade of osteoporosis (n  821). Each point and bar represents the mean and SE, respectively. Grade 0, normal (no radiographic evidence of vertebral demineralization or fracture). Grades 1 – 3, progressively severe vertebral demineralization and fracture [58].

FIGURE 5

In women, bone mineral density (BMD) is clearly related to fracture risk in both retrospective case-controlled studies and prospective trials. There are few data available in men, but the available evidence is consistent with a similar inverse relationship of bone mass to fracture. For instance, in men with spinal fractures, measures of femoral cortical area and Singh index [54], proximal femoral dual photon absorptiometry (DPA) [42], vertebral quantitative computed tomography (QCT) [55], vertebral DPA [56,57], calcaneal SPA [58], and spine and hip dual energy X-ray absorptiometry (DXA) [48,59] have all revealed lower mean values than in control men (Fig. 4). In addition, the incidence of fracture is higher in men as bone density falls [60] (Fig. 5). In addition to lower BMD, men with vertebral fracture seem to have smaller vertebrae [61], probably reflecting the importance of size on biomechanical strength. Chevalley et al. and Karlsson et al. observed that hip and spine BMD are clearly reduced in a series of men with hip fracture compared to age-matched controls, and more recently several studies have documented the relationship between low BMD and increased risk of fracture at the hip and other appendicular sites [62 – 67]. In fact, the degree to which low BMD increases fracture risk appears to be similar between men and women. Although ultrasound measures are becoming more commonly used,

Incidence of new vertebral fractures (1981 – 1994) among men and women in the HOS, by quartiles of baseline calcaneus BMD [60].

107

CHAPTER 42 Osteoporosis in Men

only sparse data relate ultrasound measures to fracture risk in men. As in women, there is a clear overlap of bone density in men with fractures and nonfracture control subjects, indicating that bone density is not the sole determinant of vertebral fracture risk. Fracture is a somewhat chance event, and factors other than BMD (bone size, strength, propensity to fall, etc.) are also important variables [68]. When a man suffers a fracture, it implies a much higher risk of subsequent fracture [52,69,70], a relationship that may be even stronger in men than in women [71]. Even men who experience a traumatic fracture in mid-life are at higher risk of a later osteoporotic fracture [10,53]. 2. FALLS In addition to bone mass, the risk of falling has been identified as a major determinant of fracture in women. Although men have a somewhat lower risk of falls than women [72 – 74], a variety of factors indirectly related to risk of falling are associated with fracture. This is important, as the incidence of falls in the elderly appear to be increasing [75]. For instance, Nguyen et al. found that men who had experienced a nontraumatic fracture exhibit more body sway and lower grip strength (as well as lower bone density) than nonfracture controls [68]. In studies of men with hip fractures [76,77] a number of factors associated with falls were found to be more prevalent than in controls. These included receiving a disability pension, neurological disease, confusion, “ambulatory problems,” and alcohol use. The severity of falls, and the site of impact affect the likelihood of fracture [78 – 80]. Hemenway et al. reported that taller, heavier men have more hip fractures, possibly related to the force of impact involved in a fall [81]. As in women, the use of several classes of psychotropic drugs is associated with hip fracture risk in men [82,83]. Men with hip fracture weigh less, have lower fat and lean body mass, and more commonly live alone or are not married than control subjects [62,77,84]. These differences suggest a body habitus and lifestyle more conducive to falls and injury, as well as the possibility of other interacting risk factors (nutritional deficiencies, comorbidities). Finally, the characteristics of falls may be different in men and women, which in turn may influence the kinds if fractures that result [85]. 3. WHY ARE FRACTURES LESS COMMON IN MEN WOMEN?

THAN IN

The cause of the greater fracture rate in women is complex. First, accumulation of skeletal mass during growth, particularly in puberty, is greater in men than in women, resulting in larger bones. In tubular bones, there is a greater total width (20% in the second metacarpal) [86] and greater cortical width in early adulthood [87]. This difference persists throughout life. Since resistance to fracture in tubular bones is related both to total diameter and cortical thickness

it follows that long bone fractures should be less common in men [88]. Gender differences in the dimensions of axial bones may also contribute substantially to differences in mechanical competence [88]. For instance, compressive strength is strongly related to vertebral end plate area [89 – 91], and when bone density and body size are taken into account, fractures are more common in individuals with smaller vertebrae [92] (Fig. 6). From puberty on, mean vertebral cross-sectional area is 25% greater in men [92]. Moreover, in men vertebrae increase in cross-sectional area by 25 – 30% with aging as a result of periosteal apposition [93]. This process also occurs in women [94], but it may be more accentuated in men [95], thus amplifying the biomechanical advantage. Interestingly, the girth of the femur and other long bones increases with age in men more than in women [95 – 99]. These differences at the proximal femur and vertebrae may help explain the lower hip and vertebral fracture rates in men, particularly since the relative gender difference in peak hip and spine mineral density and in the rate of age-related decline in density are small. Other gender differences in skeletal anatomy may also provide a male advantage. Second, women lose more bone mineral with aging than do men, a phenomenon most apparent in long bones. Cortical porosity increases more in women [100], and women lose more at the endosteal surface and gain less periosteally than do men and thus accrue less biomechanical benefit [101 – 103]. At the proximal femur, as well, there is evidence that men lose bone less quickly than do women [104 – 106]. Moreover, as discussed above, a gender difference in the character of age-related changes in trabecular bone structure probably contributes to a greater fracture risk in women. Whereas in men the age-related fall in mineral density at trabecular sites (which is almost as impressive as in women) is the result of generalized trabecular thinning with some loss of trabeculae [96,107], in women there is a more marked loss of trabecular elements. Finally, elderly men fall less frequently than do women, reducing the risk of trauma as a cause of fracture [72 – 74,108].

III. THE MAJOR DETERMINANTS OF SKELETAL HEALTH IN MEN A. Peak Bone Mass Development In early childhood, there are few discernible differences between the skeletons of boys and girls [109]. During adolescence both sexes exhibit dramatic increases in bone mass which are closely related to pubertal stage and is almost complete when puberty ends [110]. In boys the achievement of peak bone mass is later, not only because puberty is later in onset but also because boys accrete bone for a

108

ORWOLL AND KLEIN

FIGURE 6

Axial cross-sectional area, transverse diameter, anteroposterior diameter, and vertebral body volume of the first (L1), second (L2), and third (L3) lumbar vertebral bodies in 12 pairs of men and women matched for age, vertebral body height, and bone density. Values are mean 1 SD, *P  0.001. Values in men are shown by the open bars; values in women are shown by shaded bars [515].

longer time during this crucial period [109] (Fig. 7). The adolescent development of adult bone mass depends upon changes in both density and size, with increases in size being quantitatively much more important [87,88,109]. During pubertal skeletal maturation obvious sexual differences in adult skeletal morphology emerge, and sexual differences appear to be related only to differences in size. Virtually all skeletal dimensions in men are larger than those in women. For example, radial width and cortical thickness are considerably larger in men [101], femoral neck cross-sectional area is larger [97], and vertebral body cross-sectional area is larger [111]. As a result, total body bone mineral is greater in men (3100 – 3500 g in young men vs 2300 – 2700 g in young women) [112,113]. The development of peak bone mass in boys is influenced by a variety of factors, among the most important being exercise, nutrition (calcium and vitamin D, protein), the adequacy and timing of puberty, and the presence of adverse medical events and lifestyles (e.g., smoking) [114 – 119]. The acquisition of peak bone mass in men, as in women, is also strongly influenced by genetic factors. Krall and Dawson-Hughes estimated heritability to be 40 – 83% at

several measurement sites in men [120]. Men with a family history of osteoporosis or fractures have lower bone mass and/or greater fracture risk than those without [121 – 124]. The specific genes responsible have yet to be identified, although several candidates have been evaluated, including the estrogen receptors, androgen receptor, vitamin D receptor, collagen type I1, insulin-like growth factor 1, aromatase, etc.

B. Age-Related Bone Loss 1. BONE MASS a. Appendicular Bone Cross-sectional studies suggest that age is associated with a fairly linear decrease in cortical bone mass [56,101,113,125 – 127], but some also indicate the BMD:age slope becomes more negative in men after 50 years [56,101,126,128]. This slope is not quite as steep as that in women [101,129], thereby accentuating the gender differences in cortical mass present in early adulthood. However, the rate of cortical bone loss in men as reported in longitudinal studies is considerably more rapid (5 – 10%/decade)

109

CHAPTER 42 Osteoporosis in Men

FIGURE 7 Bone mass gain at the lumbar spine during adolescence. The yearly increase in lumbar spine BMD and BMC is depicted in males and females (mean  SE) [516]. [56,127,128,130] than previously estimated from crosssectional studies (1 – 3%/decade) [125,126,131]. b. Axial Bone The decline in axial bone density was initially considered to be relatively slow in men, primarily

FIGURE 8

because of cross-sectional studies using techniques that assess total spinal bone mass (DPA). Vertebral bone density as measured by QCT, however, suggested a much faster rate of bone loss with aging in normal men [132]. Subsequently, results derived from DPA were shown to be influenced by artifacts in measurement introduced by extravertebral calcifications. If men with such calcifications are excluded, the relationship of spinal bone density to age is similar in men and women [42]. Longitudinal studies verify a more rapid rate of vertebral bone loss with aging in normal men [56]. Moreover, bone volume in the iliac crest declines at very similar rates in both men and women. In cross-sectional studies the slope of density with age at proximal femoral sites is significantly negative in men, albeit somewhat less than that in women [125,131,133,134]. In longitudinal studies, femoral neck bone loss clearly occurs in men, at a rate approximately the same as that in women [104]. In both sexes, the rate of femoral bone loss accelerates with increasing age [104,105]. In the large Rotterdam Study Burger et al. noted that older women lose more bone between ages 55 and 70 (presumably reflecting the effects of menopause) but thereafter the rate of bone loss accelerates in both genders [105] (Fig. 8). Although the end result is a greater cumulative bone loss in older women, it is important to recognize the dramatic effect of aging on bone loss in men as well. Finally, cross-sectional studies using ultrasound measures of bone (calcaneus) also indicate a change with aging. Broadband ultrasound attenuation and speed of sound both decline with advancing age in men [135], at a rate clearly less than that in women. The interpretation of ultrasound measurements is yet somewhat uncertain, but they may reflect not only bone mass but also structural or material properties of bone. 2. BONE ARCHITECTURE a. Appendicular Bone In cortical bone, men experience an increase in porosity with aging, although at a rate somewhat slower than that seen in women [100,136]. This

Mean yearly rate of change in bone mineral density (BMD) and 95% confidence interval according to age group and sex, the Rotterdam Study, the Netherlands, 1990 – 1995. P values are for linear trends [105].

110

ORWOLL AND KLEIN

results in a reduction in density and mechanical strength [99] and probably increases fracture risk. Although men of greater weight and lean body mass have larger appendicular cortical areas, this does not protect them from age-related loss [101]. However, the decline in cortical mass is to some extent compensated by changes in cortical dimensions [137]. In both sexes, there is an age-related increase in cortical width, and since fracture resistance is so dependent on geometry, this change is beneficial. In a two-decade study Garn et al. found that the rate of metacarpal cortical loss in men was very similar in both men and women, but periosteal apposition was somewhat greater in men (and endocortical loss somewhat less), mitigating the loss of thickness and overall mass [101]. This gender difference can be observed in other long bones as well [99] (Fig. 9). The increased periosteal apposition rate and the somewhat lesser rate of cortical loss with age in men are in accord with the fracture patterns observed in the elderly, in whom the rate of appendicular fractures is less in men than in women. b. Axial Bone There are microscopic changes in axial bone architecture with aging that in all likelihood influence fracture risk independent of changes in bone mass. A decline in vertebral trabecular number and thickness with age is associated with a reduction in compressive strength [138], and men with vertebral and femoral fractures have a lower trabecular plate density [139]. In men and women there is a generalized loss of trabeculae, but loss of horizontal elements (number and thickness) is particularly marked, in turn resulting in less support to vertical, load bearing trabeculae [140]. Similar changes in trabecular structure in other locations (e.g., proximal femur) probably also contribute to fracture risk. In fact, the quantitation of proximal femoral trabecular patterns reveals a definite loss of trabeculation with age in men, and men with osteo-

FIGURE 10

FIGURE 9

Age-related changes in the calculated failure moment of male and female human femoral shafts in bending. The slope of the female data is significantly different from 0 [99].

porotic fractures have less trabeculation than control men [141]. In addition to trabecular loss, the appearance of microfractures increases with age and may also contribute to fracture risk [142]. Despite the basic similarities of these processes in men and women, the nature of trabecular bone loss may have gender differences. Using histomorphometry to analyze vertebral bone, Mosekilde found that while bone density is not particularly different between older men and women, the microarchitectural pattern of trabecular loss is distinct. Women tend to experience both trabecular thinning and trabecular loss (particularly horizontal elements) while men experience trabecular thinning with less trabecular dropout [140]. Similar results have been described in iliac crest biopsy specimens [107,139,143,144] (Fig. 10). Gender

Changes in trabecular number (solid lines) and trabecular width (dashed lines) with age in the iliac crest of men and women [143].

111

CHAPTER 42 Osteoporosis in Men

[160,161]. Ljunghall et al. [157] and Kurland et al. [162] noted lower IGF-I concentrations in men with idiopathic osteoporosis. Similarly, the skeletal content of important growth factors, including IGF-I and TGF- , decline with age in men and may herald major changes in growth factor action [156]. Some of the gender differences in bone biology may be due to the growth factor axis. The relationship between IGF-I and BMD may be different in men and women [161,163] and there are apparently gender differences in the skeletal response to growth hormone therapy [164,165]. 2. CALCIUM NUTRITION AND BONE LOSS FIGURE 11

The ratio of urinary pyridinoline to urinary creatinine excretion (mean  SE) in 440 adults expressed as a percentage of values in young adults [517].

differences in the nature of structural changes with aging may have important biomechanical consequences, as load bearing of vertebral specimens appears to be better preserved in men than in women [140].

C. Causes of Age-Related Bone Loss Aging in normal men is associated with detectable appendicular and substantial axial bone loss. The cause of this loss in unknown but is speculated to be related to a number of factors. Histomorphometric techniques demonstrate a reduction in bone formation (mean wall thickness) in both sexes [143,145 – 149] that probably contributes to the decline in bone mass. An additional age-related increase in bone resorption in men is not apparent using these methods [145,148]. However, markers of bone remodeling increase with age in men [150 – 152] (Fig. 11), raising the possibility of an acceleration in bone turnover that contributes to bone loss. In addition to these putative influences, several other processes contribute to the pathophysiology of senile bone loss, including nutritional deficiencies, inactivity, and loss of gonadal function. 1. GROWTH FACTOR ACTION Growth factor concentrations decline with age [153, 154], and there are several reasons to link age-related bone loss in men to changes in growth factor or cytokine physiology [155 – 157]. Recently, Johannson et al. found a surprising correlation between insulin-like growth factor binding protein 3 (IGFBP-3) and bone mineral density in men [158], and similar relationships between insulin-like growth factor-I (IGF-I) and BMD have been noted by others [159]. Since there is a relationship between IGF-I and sex steroid concentrations, an effect of growth factor action and bone may be in part related to sex steroid action as well

Riggs and Melton [166] have suggested that senile (Type II) osteoporosis in men and women is due, at least in part, to alterations in calcium economy. The average level of dietary calcium necessary to maintain mineral balance is relatively low in young men (400 – 600 mg/day) but the range is large and there are data that suggest a higher requirement in older men [167,168]. Although U.S. men achieve a mean dietary calcium intake considerably greater than that of women (800 vs 500 mg/day in the 1978 NHANES survey) these data still indicate that about one half of men ingest less than the recommended daily allowance (800 mg), and many ingest much less. In addition, aging in men has been associated with increased parathyroid hormone (PTH) concentrations [169,170], reduced circulating 25-hydroxyvitamin D [171], and (in some studies) subnormal 1,25-dihydroxyvitamin D concentrations [172 – 175]. In the Baltimore Longitudinal Study of Aging [176,177], lower radial bone density in men was related to higher PTH and lower 25-hydroxyvitamin D concentrations. Halloran and Bikle [178] summarized the data relating age related changes in calcium homeostasis and bone health in men. Several reports have linked dietary calcium intake to levels of bone density in men, but the evidence is not yet conclusive. In a study of 222 subjects, Kroger and Laitinen found that men in the highest tertile of calcium intake (1200 mg/day) have higher proximal femoral BMD (but not spinal BMD) than those in the lowest tertile (800 mg/day) [177]. Similarly, in a cross-sectional study of a small group of men, Kelly and Pocock found that measures of axial BMD correlate with dietary calcium intake, but appendicular radial BMD do not [179]. Other groups who have examined appendicular bone mass in adult men in longitudinal studies have also found no clear relationship to calcium intake [101,104,130,180]. These results suggest that calcium intake may play a role in the determination of axial, but not (or to a lesser extent) appendicular bone mass. However, in the only published controlled trial of calcium supplementation in adult men, no beneficial effects were found on the rate of bone mineral loss from either spinal or radial sites [56], despite the fact that urine calcium

112 excretion increased and PTH concentrations were suppressed. Osteocalcin concentrations were not altered. The results of this trial are somewhat muted by the relatively large dietary calcium intake of the subjects prior to supplementation (1100 mg/day), and supplementation in a less calcium-replete population may prove to be more effective. In a longitudinal study of the Rotterdam population, Burger et al. found lower calcium intake to be associated with higher rates of bone loss [105]. There have been no studies of the relationship between calcium intake and skeletal structure (e.g., cortical thickness, trabecular architecture, remodeling rates, material properties). A variety of studies have examined the relationship between dietary calcium intake and hip fracture in men, with inconsistent results. In a small case-controlled study in Hong Kong, Lau et al. found that a very low calcium intake (75 mg/day) was associated with fracture risk [181]. In a British case-control trial, Cooper et al. found those men with the highest calcium intakes (1041 mg/day) to be significantly protected [182]. In several longitudinal observational trials (including the NHANES I follow-up study [183], the Rancho Bernardo study [184], and a study of eight communities in Britain [185], hip fracture risk in men was strongly suggested to be related to dietary calcium intake, but the relationships did not reach statistical significance. However, two other very large studies [68,186] found no relation between calcium intake and hip fracture risk in men. Looker et al. [183] have pointed out the pitfalls inherent in trials of this sort, including low power, difficulties in estimating calcium intake and the effects of confounding variables. In general, evaluations of the relationship between calcium nutrition and hip fracture in men are suggestive of a beneficial effect, but remain inconclusive. Moreover, there have been no attempts to examine the effects of calcium intake on other fractures, in particular vertebral fractures. Albeit incomplete, the data are probably consistent with a limited role for dietary calcium insufficiency in the determination of the rate of bone loss and fractures in men. 3. WEIGHT AND PHYSICAL ACTIVITY Mechanical force exerts major effects on bone mass in men [187], and it is probably one of the fundamental variables responsible for the gender dimorphism in bone mass and structure. In cross-sectional studies, bone mass is greater in physically active men [183,188 – 193], an effect that can be demonstrated at both the regional (i.e., the particular anatomic region affected) and the systemic levels. Muscle strength and lean body mass in men also correlate with bone density both regionally and systemically [183,188,194]. Furthermore, muscle strength is related to bone bending stiffness in men, an index of strength independent of mass, suggesting that mechanical force has effects not only on bone mass but also quality [195]. Longi-

ORWOLL AND KLEIN

tudinal studies tend to corroborate the effect of mechanical force on skeletal mass in men [104,196] but are very few in number. Finally, exercise has been strongly related to a reduction in hip fracture rates in men [185,186] an effect that may also relate to a reduced risk of falls. Unfortunately, the fairly consistent finding of positive correlations between exercise history and/or strength and bone mass in crosssectional studies have not been confirmed in longitudinal investigations. As in women, body weight is itself highly correlated with bone density in men [132], an effect that could be related to the mechanical effect exerted by mass alone or to a particular aspect of body composition (i.e., lean vs fat mass, adipose distribution). Reid et al. [197] suggested that there are gender differences in the relative effects of body composition on skeletal morphology. In their studies bone mass was associated with fat mass in women, but not in men (lean mass was not associated with bone mass in either sex). They speculated that androgens may contribute to the lack of fat – bone correlation in men, as androgen action is associated with an increase in bone mass but a fall in adiposity. Low body weight is also associated with increased rates of bone loss in men [105,198,199]. In sum, the available data strongly suggest a powerful effect of weight and mechanical force on the male skeleton. In view of the clear decline in physical activity and muscle strength with aging [200,201], bone loss in men may, in part, relate to a diminution of the trophic effects of mechanical force on skeletal tissues. Certainly, the character of agerelated bone loss closely mimics that of chronic disuse [202], but this tentative conclusion requires confirmation in longitudinal studies. 4. CHANGES IN GONADAL FUNCTION Aging in men is associated with changes in the hypothalamic – pituitary – gonadal axis that result in notable declines in total and free testosterone levels [203,204]. These changes have given rise to considerable speculation as to whether several of the concomitants of aging, are the result at least in part of the decline in testosterone levels [205,206]. For instance, the well documented declines in muscle strength and bone mass with aging have been suggested to be potential sequelae [207]. Indeed, there are several lines of evidence firmly linking androgen action to skeletal mass in men (vide infra), and there have been several attempts to link bone mass to testosterone levels. Kelly and Pocock [179] found that free testosterone levels correlated with ultra distal bone density (but not with a variety of other densitometric measurement sites) in a group of men ages 21 – 79 even after the effects of age were considered. Similarly, in several other studies androgen concentrations were found to correlate with (albeit weakly) bone mineral density [208 – 212]. However, these findings have not been corroborated by other investigators [158,

113

CHAPTER 42 Osteoporosis in Men

TABLE 1

Effect of T Administration on Biochemical Parameters Related to Bone Baseline

T

Placebo

Serum calcium (nmol/liter)

2.28  0.02

2.23  0.02

2.24  0.03

Serum phosphate (nmol/liter)

0.98  0.04

0.91  0.05

0.96  0.04

Creatinine clearance (ml/min)

104  5

110  8

103  4

54  4

51  3

52  3

Variable

Serum AP (U/liter) Osteocalcin (ng/ml) Calcium/GFR (mmol/dl)

4.3  0.6

4.3  0.7

4.1  0.7

4.87  0.52

4.43  0.59

4.59  0.49

HPro/GFR (mol/dl)

151  10

108  8*

142  13

TmP (mg/dl)

2.47  0.18

2.26  0.09

2.40  0.14

1,25-Dihydroxyvitamin D (pg/ml)

23.8  1.8

27.8  1.7

31.5  2.4

PTH (pg/ml)

216  17

188  18

200  19

Note. Values are the mean SE. AP, alkaline phosphatase; HPro, hydroxyproline. Reproduced with permission for J. S. Tenover [217]. *P  0.01.

213 – 216]. In an attempt to test the hypothesis that relative androgen deficiency has a skeletal impact in older men, Tenover reported in a small study (13 men) that parenteral testosterone supplementation reduced urinary hydroxyproline excretion [217] (see Table 1) and Snyder et al. found that testosterone administration increased spinal BMD slightly in older men [218]. Clarke et al. suggested that changes in testosterone are not as important as the skeletal effects of age-related declines in adrenal androgens [219]. The study of this and similar issues is made particularly difficult by the inability to adequately assess the long-term, integrated level of androgen action on bone with crosssectional or relatively short-term study designs. In general the correlations between serum androgen concentrations and bone mass in adult men have been weak or insignificant, but the role of androgen action on the skeleton remains unclear. The effects of estrogen on the male skeleton have become of great interest [220,221]. Several men with abnormalities of estrogen action, including one man with an abnormal estrogen receptor [222] and two others with aromatase deficiency [223,224], have presented as young adults with failure of skeletal maturation and low bone mass. Therapy with estrogen prompted a major increase in bone mineral density in those with aromatase deficiency [224,225]. Clearly, estrogen is important in male skeletal development, and these cases further raise the issue of whether estrogen is the more important sex steroid [226]. Additionally, estrogen has sometimes reported to be positively correlated with bone mineral density [227,228] or

fractures [229] in older men, even to a greater extent than is testosterone. These findings raise important issues related to the role and regulation of aromatase activity and the relative importance of estrogens vs androgens in skeletal homeostasis. Nevertheless, it is important to note that androgens clearly influence the skeleton, and that they probably play an independent, and coordinated, role with estrogens [230]. An elucidation of these interactions will be crucial for the understanding of bone biology in both genders. In addition, the appropriate use of estrogen measurements in the evaluation of men with osteoporosis and the use of estrogens or selective estrogen receptor modulators in the management of men with low bone mass require clarification. Obviously the issue of the importance of gonadal insufficiency in the genesis of senile bone loss in men remains unresolved. In addition to the decline in androgen levels associated with aging, overt gonadal insufficiency (reviewed below) would be expected to contribute to any aging effects in an additive fashion.

IV. OSTEOPOROSIS Osteoporosis in men is a heterogeneous condition, encompassing a wide variety of etiologies and clinical presentations. In practice, it is common to uncover several potential explanations for bone loss and fractures in a single patient.

A. Age-Related Osteoporosis Bone loss that occurs with aging is an important feature of osteoporosis in men and women (see above). In some men, age-related bone loss may alone suffice to cause nontraumatic fractures. Even when other causes of bone loss are present (i.e., hypogonadism, alcoholism), the universal loss of bone that accompanies aging unquestionably contributes to the eventual propensity for fractures.

B. Idiopathic (Primary) Osteoporosis Osteoporosis in men has been termed idiopathic if no known cause can be identified on clinical and laboratory grounds. Although metabolic bone disease in men has been traditionally considered to be more commonly related to “secondary” causes [166,231 – 234], this impression is difficult to substantiate. In fact, the frequency with which osteoporosis in men has been found to be idiopathic is significant. In large series of osteoporotic men, many patients were considered to have bone disease of unknown etiology (70 of 105 subjects [231], 40 of 94 subjects [235], and 60 of 95 subjects [232]).

114 The age of men with primary, or idiopathic, osteoporosis varies widely (23 – 86 years) with an average in the mid-60s. This age range overlaps that of “senile” osteoporosis, and differentiation of idiopathic and senile osteoporosis is somewhat arbitrary. Riggs and Melton [166] defined senile (or Type II) osteoporosis as occurring in either sex after age 70, but this definition obviously does not exclude the potential for pathophysiological overlap between older and younger patients. The universal decline in bone mass that happens as a concomitant of aging has the potential for eventually producing clinical osteoporosis in all individuals, and some idiopathic osteoporosis may represent this age-related process or its premature onset. Once again, it is important to emphasize that broad classifications of osteoporosis are of limited value in considering an individual patient, in whom several pathogenetic mechanisms (sometimes occult) may be operative. The character of idiopathic osteoporosis in men is relatively indistinct. After major secondary contributors to bone loss have been eliminated, more detailed biochemical and histomorphometric analyses of men with idiopathic disease fail to reveal consistent features [54,232,234,236]. Some patients have slightly increased serum alkaline phosphatase activity [54,232]. Reduced intestinal calcium absorption has been reported in the presence of lowered 1,25-dihydroxyvitamin D concentrations [54], but calcium balance has not been systematically examined in osteoporotics males. Osteoblastic dysfunction may contribute to osteoporosis in men [234,236]. Bordier et al. reported studies of a series of 11 patients with idiopathic osteoporosis, 10 of whom were men [237]. In those subjects histomorphometric parameters clearly suggested that a defect in bone formation contributed to loss of bone [235] Osteoblastic dysfunction, however, is not a consistent finding in idiopathic male osteoporosis [54,235]. Nordin et al. suggested that accelerated resorption may also be a primary mediator [238]. These apparently discrepant results may in part be explicable by findings reported by Aaron et al. [239], who noted that young men ( 49 years) with idiopathic osteoporosis had reductions in bone-forming parameters (osteoid surface, mean wall thickness), while older men had forming parameters similar to age-matched controls but evidence of slightly increased resorption. Khosla et al. also found the young patients with idiopathic osteoporosis (both men and women) had histomorphometric characteristics, suggesting a defect in bone formation [240]. Moreover, Marie et al. found that most men with eugonadal osteoporosis had evidence of decreased bone formation. When osteoblasts from these patients were compared to normal controls or to osteoporotics without reduced bone formation, a lower proliferation capacity was found [241] (Fig. 12). Thus the histomorphometric

ORWOLL AND KLEIN

FIGURE 12 The time course of osteoblastic cell proliferation evaluated by DNA synthesis in bone samples from osteoporotic men with low (open circles) or normal (open triangles) bone formation as determined by the extent of double-labeled surface, compared to normal bone cells (solid squares) (mean  SE) [241]. mechanisms of early-onset primary osteoporosis may differ from those operating when the disorder appears later in life. The late-onset form appears to resemble the bone loss that occurs with normal aging, albeit at a more accelerated pace. A possible explanation for the osteoblastic defect postulated to be important in idiopathic osteoporosis is a reduction in growth factor action [242]. Several reports suggest that men with idiopathic osteoporosis have relatively low IGF-I, or IGFBP-3, levels [157,162,243], a finding that seems to relate to lower indices of osteoblast work. Other etiologies for idiopathic osteoporosis have been suggested, including abnormalities in cortisol dynamics [199]. It is likely that at least a fraction of the men who present with idiopathic osteoporosis have genetic underpinnings of their disorder. In fact, IGF-I concentrations are related to the presence of polymorphisms in the IGF-I gene [244]. Other genes have been implicated [245 – 249], but more definitive studies are needed. Certainly this is an area that needs additional development. Finally, the microarchitectural features of idiopathic osteoporosis in men have not been defined. Francis et al. [54] and Aaron et al. [239] did report that men with primary osteoporosis have a reduction in iliac crest bone volume and surface primarily because of a reduction in trabecular number rather than a decline in thickness. The issue of trabecular connectivity was not formally evaluated.

115

CHAPTER 42 Osteoporosis in Men

C. Osteoporosis Secondary to Other Disorders The pathophysiologic character of osteoporosis in men is minimally explored, but several reports have examined the risk factors present for bone disease in small patient populations. In the available series [54,231,232,250], 30 – 60% of men evaluated for vertebral fractures had “secondary” causes (underlying illness) contributing to the presence of bone disease. Most of these studies were of selected subjects – most commonly men presenting for health care in the United States or Great Britain because of vertebral fracture. Other series are small with heterogeneous patient groups [251]. Hence the findings may not accurately represent a generalizable spectrum of disease. There have not been adequate studies of the character of the bone disease present in men sustaining femoral or other fractures. The principal conditions found in men with reduced bone mass are shown in Table 2. Prominent are glucocorticoid excess, hypogonadism, alcoholism, gastrectomy and other gastrointestinal disorders, and hypercalciuria. These disorders will be the focus of this discussion. They and others have been recently reviewed [252]. Similar attempts to examine the contributing factors in osteoporotic women suggest that the spectrum of disorders differs somewhat [26,54,166,253], but glucocorticoid excess, premature hypogonadism, and gastrointestinal disease are prominent in women as well. It has been suggested that the number of men with “secondary” osteoporosis is higher than in women [166], but in other objective evaluations [253] the

TABLE 2

Osteoporosis in Men

I. Primary Senile Idiopathic II. Secondary Hypogonadism Glucocorticoid excess Alcoholism Gastrointestinal disorders Hypercalciuria Smoking Anticonvulsants Thyrotoxicosis Immobilization Osteogenesis imperfecta Homocystinuria Systemic mastocytosis Neoplastic diseases Rheumatoid arthritis

proportion of women with major illnesses contributing to the development of bone disease is actually very similar to that observed in male osteoporotics. 1. GLUCOCORTICOID EXCESS In the largest series of men evaluated for spinal osteoporosis, glucocorticoid excess (particularly exogenous) is the most prominent of the secondary causes identified, accounting for 16 – 18% of the men evaluated [54,231]. Commonly, glucocorticoid use is but one of several risk factors present in patients with chronic medical problems. For instance, there is widespread clinical recognition of the frequency of osteoporosis in older men with chronic obstructive pulmonary disease treated with glucocorticoids, often in the presence of tobacco and alcohol abuse. The pathophysiology of glucocorticoid-induced osteoporosis, although incompletely understood, is presumably similar in men and women (see also Chapter 44). The primary mechanism by which glucocorticoids cause bone loss appears to be via a direct receptor-medicated inhibitory effect on osteoblast activity. Physiologic concentrations of glucocorticoids enhance the function of differentiated osteoblasts [254,255], but extended exposure to supraphysiologic concentrations results in inhibition of osteoblastic synthetic function [256,257]. Recent evidence suggests that inhibition of osteoblastic collagen synthesis is mediated through glucocorticoid-induced inhibition of local synthesis of factors that promote collagen synthesis (e.g., insulin-like growth factor I, prostaglandin E2) [258,259]. Another skeletal effect that may be important in the pathogenesis of glucocorticoid-induced osteoporosis may be impairment of osteoblast recruitment [260]. Additionally, certain systemic effects of glucocorticoids can exert profound influence on skeletal health. Muscle weakness due to muscle fiber atrophy is a frequent concomitant to glucocorticoid administration [261 – 263] and a clear-cut association between steroid myopathy and osteoporosis has been observed [264]. The relative immobility that is imposed by such muscle weakness is likely to contribute to the bone loss in this setting. Most patients receiving pharmacologic doses of glucocorticoids exhibit impaired intestinal absorption of calcium and elevated urinary calcium excretion, resulting in a negative calcium balance [265]. Finally, glucocorticoids can alter gonadal hormone production. Exogenous glucocorticoids markedly reduce testosterone levels in men [266 – 268] by mechanisms that have not been fully defined, but which may include central inhibition of GnRH release, suppression of pituitary sensitivity to GnRH, and direct antagonism of testicular steroidogenesis [267 – 270]. Impotence and loss of libido frequently occur in the clinical settings in which glucocorticoids are administered and are attributed to the effects of the chronic illness. However, these symptoms

116

ORWOLL AND KLEIN

may actually be due to glucocorticoid-induced hypogonadism, which in turn may contribute substantially to the resulting low bone mass. Clinicians who care for men with osteoporosis should be aware of this phenomenon and recognize it as an important cause of a low serum testosterone value. Furthermore, because administration of testosterone to hypogonadal men improves bone mass, such therapy may be useful to prevent and treat glucocorticoid-induced osteoporosis in men. 2. HYPOGONADISM Sex steroids have major influences on the regulation of bone metabolism. The obvious importance of menopause to osteoporosis drew early attention to the role of estrogen, and both clinical and basic observations have also highlighted the importance of androgens in bone physiology in both sexes. There is an expanding understanding of the cellular effects of androgens on bone remodeling (see Chapter 11). Androgen receptors are present in physiologically relevant concentrations in osteoblastic cells [271 – 273], and androgens affect a variety of osteoblastic functions, including proliferation, growth factor and cytokine production, and bone matrix protein production (collagen, osteocalcin, osteopontin) [274 – 281]. Hence, there is an excellent foundation in basic research for the precept that androgens are active in bone. On the other hand, estrogen has assumed an even more important role as its importance in the control of male skeletal function has emerged. a. Pubertal Hypogonadism Since adolescence is so important for skeletal maturation, disorders of puberty have the potential to impair peak bone acquisition and thus to influence fracture risk throughout adulthood [282]. With adolescence, bone accretion (in both cancellous and cortical bone compartments) in both sexes is closely related to gonadal maturation [110,283] (Fig. 13). Testosterone has major effects on calcium kinetics and balance in boys [284,285] (Fig. 14). It is not known whether adrenarche affects the rate of skeletal maturation [286]. Strongly supporting the importance of androgen action in the achievement of peak bone mass in men is the fact that genetic males with complete androgen insensitivity (testicular feminization) experience increased pubertal growth but achieve a bone mass less than expected of androgen replete men [287 – 289]. Reduced bone mass is found in men who experienced an abnormal puberty (Klinefelter’s and Kallman’s syndromes) [290 – 292]. In Klinefelter’s syndrome radial bone mass is related to serum testosterone concentrations, and patients have lower circulating osteocalcin and higher rates of hydroxyproline excretion [293]. It is proposed that the failure to acquire peak bone mass with puberty is the primary abnormality in these forms of early-onset hypogonadism [290]. In fact, constitutionally delayed puberty is associated with permanent reductions in bone density

FIGURE 13 The relationships between bone mineral density of the lumbar spine, femoral neck, and femoral shaft and pubertal stages in male and female subject (*P  0.05) [110]. [294 – 296] (Fig. 15), although some have suggested that the effect may be more on bone size than density [297]. An apparent emphasis on cortical bone mass in males suggests that androgen action is especially important in the modeling

117

CHAPTER 42 Osteoporosis in Men

FIGURE 14 Changes in Va (dietary calcium absorption), Vbal (net calcium retention), and Vo (rate of bone accretion) in prepubertal children treated with testosterone (n  6). Solid bars, before testosterone; hatched bars, after testosterone. *P  0.05 [284].

drifts that add periosteal new bone during growth, and the animal model of testicular feminization supports this postulate in that cortical mass is disproportionately reduced [298]. A reasonable hypothesis is that androgens increase trabecular bone formation at epiphyseal areas, and strongly promote the addition of cortical thickness through periosteal and endosteal growth — processes that then are impaired in the presence of hypogonadism.

FIGURE 15

Radial bone mineral density in 23 men with a history of delayed puberty and 21 normal men [296].

Until recently there were no clear examples of estrogen deficiency in men, and the direct role of estrogen in skeletal metabolism in men remains controversial. As described above, there are now several examples of men with inadequate estrogen action (but adequate androgen effects) during development. From these individuals, it is clear that estrogen is needed for adequate epiphyseal function as well as bone mass development. The histological or biochemical nature of the skeletal effects of prepubertal hypogonadism in humans is unknown. Some animal studies have indirectly investigated the effects of prepubertal hypogonadism. Hock and Gera [299] examined trabecular and cortical bone mass in rats castrated (or sham operated) just before sexual maturation (4 weeks). In both compartments, bone mass in the hypogonadal rats continued to increase but at a rate considerably slower than in the controls, a finding consistent with the hypothesis that hypogonadism impairs the pubertal stimulus to bone mass accumulation. Another study of adolescent rats based on histomorphometric analysis of cortical bone suggests that periosteal bone formation and cortical thickness is decreased in adolescent males following castration [300], indicating that a reduction in bone formation contributes to low bone mass. Certainly the fact that male rats have markedly greater cortical mass than do females, and that the differences are eliminated by castration [300], strongly support a prominent role for androgens in this phase of cortical bone accumulation. b. Postpubertal Hypogonadism Androgens also appear essential for the maintenance of bone mass in adult men, as the development of hypogonadism in mature men is associated with low bone mass. Hypogonadism is present

118 in 5 – 33% of men evaluated for vertebral fractures and osteoporosis [54,231,233], and hip fractures in elderly men apparently occur more commonly in the setting of hypogonadism [301]. Reduced bone mass and fractures are associated with many forms of hypogonadism, including castration, hyperprolactinemia, anorexia, and hemachromatosis [302 – 305]. A very important group of men with hypogonadism are those treated with castration and/or GnRH agonists for prostate disease. Here bone loss is rapid and the development of osteoporosis at an accelerated rate can be expected [306 – 309]. Vertebral and appendicular bone mass are both reduced in hypogonadal men, but in adult-onset hypogonadism vertebral loss is relatively more pronounced. The degree of reduction in bone density has been correlated with levels of serum testosterone in some series [293,310,311], but in other groups no association between the two variables was detected [312]. There may be a threshold level of serum testosterone below which skeletal health is impaired, but at present it is not possible to establish that hypothesis. In addition to the link between primary testicular dysfunction and low bone mass, reduction in gonadal function secondary to several other conditions is now postulated to contribute to the development of bone loss. For instance, hypogonadism is suspected to contribute to the reduced bone mass associated with glucocorticoid excess, renal insufficiency, and other conditions [313]. The histological pattern of hypogonadal bone loss in adult men is inadequately described. A single report examines skeletal metabolism in the period immediately after gonadal failure. Stepan and Lachman [302] studied a small group of men in the years immediately following castration. The subjects lost bone rapidly (approximately 7%/year) and had clear biochemical indications of increased bone remodeling (increased serum osteocalcin levels and urinary hydroxyproline excretion). Similar evidence of an increase in remodeling was reported by Goldray et al. in studies of GnRH agonist administration [309]. Unfortunately, no direct histomorphometric analyses were reported. An early increase in remodeling after androgen withdrawal is also consistent with recent reports of the biochemical and cellular events that are associated with androgen action (a suppression of cytokine production and osteoclast formation) [314]. The histomorphometric data (primarily from patients with long standing hypogonadism), and the better documented sequence of events that follows gonadal hypofunction in menopause, suggest that this period of increased remodeling is followed by a subsequent phase of reduced turnover, possibly accompanied by a decline in bone formation. Several reports have described the histomorphometric character of hypogonadal, osteopenic men, but they are uncontrolled, and are from subjects in whom hypogonadism was of varied causation (both early and late onset) and of

ORWOLL AND KLEIN

long duration. For instance, in a study of 13 men with longstanding hypogonadism Francis and Peacock found that bone remodeling and formation were reduced, and 1,25-dihydroxyvitamin D concentrations were low in those with fractures [315]. With testosterone therapy, 1,25-dihydroxyvitamin D values increased and there was some indication of an increase in “formation” parameters. Similarly, Delmas et al. reported decreased rates of formation in a small group of hypogonadal men [316], and formation rates were low in a single case reported by Baran et al. [317] (unfortunately vitamin D metabolite levels were not reported in these studies). These data raise the issue of whether androgens provide an important stimulus to bone formation. In contrast, Jackson et al. [318] reported histomorphometric analyses of a small group of osteoporotic, chronically hypogonadal men with normal vitamin D status. In these patients no apparent defect in mineralization was observed, and the authors speculated that nutritional vitamin D defi ciency may have been a factor in the European studies of Francis and Delmas. Jackson et al. found a slight increase in mean remodeling rate and concluded that androgen defi ciency induces a remodeling defect similar to that of estrogen deficiency in the postmenopausal period. In both the series from Jackson and Francis [315,318], trabecular number was reduced in the hypogonadal men, indicating that trabecular loss is a feature of androgen deficiency as it is in estrogen deficiency [139]. There are no reports of cortical remodeling dynamics. Actually, the remodeling character of all these study populations was heterogeneous, and in view of the of the variability in the small groups, the presence of other confounding clinical conditions, and the lack of adequate controls, no firm conclusions can be drawn concerning the remodeling defect induced by hypogonadism in men. Animal studies provide some additional information (see also Chapter 11), but the current model of skeletal metabolism in hypogonadism remains quite incomplete. 3. ALCOHOLISM It is well established that long-term alcohol consumption can result in a host of abnormal clinical, biochemical, and physiologic findings that stem from the toxic effects of ethanol on the liver, gonads, marrow, heart, and brain. The fact that prolonged abuse of alcohol is also detrimental to skeletal integrity in men has only recently been recognized [319 – 321] (see Chapter 31). Numerous studies over the past quarter century have demonstrated a reduction in bone mass in alcoholics, especially in the iliac crest, calcaneus, vertebral column, and hip [231,234,322 – 335 ] — all areas with a high proportion of metabolically active, trabecular bone. It should be pointed out, however, these studies can be criticized for their small size and for being poorly controlled. In addition, recent reports suggest that modest alcohol intake may actually be associated with increases in bone density [336].

119

CHAPTER 42 Osteoporosis in Men

More extensive studies have attempted to estimate the prevalence of skeletal fracture in the alcoholic population. Reduced bone mass is evident on routine radiographs in a significant percentage of individuals (25 – 50%) whose drinking habits have prompted them to seek medical help [328 – 331]. The degree to which bone disease is present in the entire population of alcoholics remains uncertain, and determination of the true incidence of alcohol-induced reductions in bone mass in men must await a survey of a large number of cases. However, the habitual consumption of alcoholic beverages is clearly recognized as a significant negative determinant of bone mass in epidemiological surveys of men [130,231,330,334] and has been shown in longitudinal studies to be associated with increased rates of bone loss [130]. Thus, the link between alcohol abuse and bone disease is well established. Although there is a definite relationship between alcohol abuse and bone disease, the mechanisms by which alcohol induces bone disease remain unclear. An association between alcoholism and accidental injury is well recognized. However, emerging evidence now suggests that in addition to the increased incidence of trauma, the high incidence of fracture in this population may also stem from generalized skeletal fragility. The in vivo rate of skeletal protein synthesis of young male rats is reduced by 25 – 30% in response to acute ethanol exposure [337]. Furthermore, the chronic administration of ethanol to rats produces ultrastructural changes in bone morphology that compromise mechanical strength [338,339]. Bone histomorphometric studies have provided further insight into the specific nature of the skeletal disorder induced by ethanol. Iliac crest biopsy usually reveals significant reductions in trabecular bone

volume, osteoid matrix, number of osteoblasts, mineral apposition and overall rate of bone formation [331 – 333, 340 – 342] (Table 3). Marrow fibrosis is uncommon and there is generally no evidence of osteomalacia [234,332, 333,342], except in patients who have previously undergone gastric surgery [343]. Parameters reflecting osteoclast activity (e.g., eroded surface, resorption depth, and resorption period) are, for the most part, spared [331 – 333, 340 – 342] (Table 3). Since the osteoblast is the cell responsible for bone formation, the histomorphometric findings in alcoholic patients with bone disease suggest that the osteoblast may be specifically targeted by alcohol. Nutritional deficiencies are common in alcoholics. However, poor nutrition alone does not induce osteoporosis in experimental animals [344] and none of the histomorphometric studies cited above demonstrated any evidence for nutritional deficiency. Mild hypocalcemia, hypophosphatemia, and hypomagnesemia are frequently present in ambulatory alcoholic men because of poor dietary intake, malabsorption, and increased renal excretion [341, 345 – 349]. Hypocalcemia, if severe enough, could result in low bone mass by inducing a state of secondary hyperparathyroidism. However, evidence for hyperparathyroidism with accelerated bone remodeling is not seen on bone biopsies of affected patients [331 – 333,340]. Recently, acute alcohol intoxication has been reported to reduce PTH concentrations [350]. The relevance of this finding is unclear as the reduction in PTH lasts only a few hours, and low bone mass is not observed in hypoparathyroid patients. Early studies found circulating levels of vitamin D metabolites to be low [234,325,351], but subsequent investigation has excluded vitamin D deficiency as a major

TABLE 3 Abstainers (n  9)

Drinkers (n  16)

Significance (P value)

Parameters of Bone Formation Osteoid volume (%)

0.68  0.17

0.38  0.08

NS

Osteoid seam width (m)

11.4  0.68

7.95  0.48

0.001

Osteoid surface with osteoblasts (%)

20.1  4.3

5.5  1.7

0.01

Osteoblasts/10 cm surface

143.5  18.3

51.5  18

0.003

Parameters of bone resorption 6.9  1.03

5.5  0.7

NS

Extent of surface with osteoclasts (5)

0.76  0.24

1.11  0.22

NS

Osteoclasts/10 cm surface

17.1  3.4

21.3  5.1

NS

Mineralization rate (M/day)

0.52  0.1

0.26  0.07

0.04

Mineralization lag time (days)

28  4

62  10

0.006

Osteon remodeling time (days)

140  20

423  78

0.003

Extent of surface with lacunae (%)

Parameters of Mineralization

Note. Adapted from Crilly et al. [33].

120 cause of alcohol-induced bone disease by demonstrating normal vitamin D absorption [352] and conversion to 25hydroxyvitamin D [353] in alcoholic individuals and, more directly, by the measurement of normal free concentrations of 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D in patients with alcoholic cirrhosis and alcoholic bone disease [333,354,355]. These findings do not exclude the possibility of an alcohol-induced vitamin D resistant state, but again the lack of histomorphometric evidence of osteomalacia in vitamin D-replete osteopenic alcoholic subjects [332] argues strongly against such a possibility. Hypogonadism is clearly a risk factor for osteoporosis in men (see above), and chronic alcoholic men suffer from impotence, sterility, and testicular atrophy [356]. Although most studies of alcoholic men with bone disease report normal androgen values [333,335,348], reduced serum free testosterone concentrations in alcoholic subjects with osteoporosis were reported by Diamond [332]. The testosterone levels were on average lower than those of the male control subjects but still fell within the normal range for the general male population overall. Recent studies suggest that ethanol also exerts toxic effects directly at the cellular level in bone. Ethanol induces a dose-dependent reduction in cellular protein and DNA synthesis in human osteoblasts in vitro [357,358]. Further evidence implicating a direct effect of ethanol on osteoblast activity comes from studies examining circulating bone Gla protein (BGP, osteocalcin) levels in alcoholic subjects. BGP is a small peptide synthesized by active osteoblasts, a portion of which is released into the circulation. BGP values are positively correlated with histomorphometric parameters of bone formation in normal individuals [359] and in patients with metabolic bone disease [360]. Chronic alcoholic patients exhibit significantly lower BGP concentrations than age-matched controls [361,362]. Moreover, alcohol has a dose-dependent suppressive effect on circulating BGP levels [350,363,364]. The consumption of 50 g of ethanol (equivalent to four “shots” of scotch whisky) over 45 min results in a 30% decrease in serum BGP concentration that is detectable 2 h later. Beyond these fragmentary attempts at characterization, however, little is known about phenotypic regulation by ethanol in the osteoblast. 4. TOBACCO USE Tobacco use is associated with lowered bone mass and fractures in women [365 – 367]. Tobacco was also linked to an increased prevalence of vertebral fractures in men in the cohort studies of Seeman [231], in which the relative risk of vertebral fracture in smokers was 2.3. This risk was independent of alcohol consumption, and in fact the risk imparted by the two variables was multiplicative. Hip fracture rates are higher in currently smoking men, particularly those smoking more heavily [77,186]. In support of the adverse effects of smoking on bone health in men, Slemenda

ORWOLL AND KLEIN

FIGURE 16 Bone loss in members of male twin pairs who differed in cigarette smoking behavior. Points above the line indicated pairs in which the twin who smoked more lost more bone [130].

et al. [130] found that the rate of bone loss from the radius was significantly greater (140%) in subjects who smoked than in nonsmokers (Fig. 16). In fact, there was a correlation between the number of cigarettes smoked and the rapidity of bone loss and the smokers lost more bone than did their nonsmoking twins. In this study the adverse effects of smoking and alcohol use were independent. A variety of other studies have noted the relationship between smoking and low bone mass, bone loss, and fractures in men [51,77,198,368 – 371] (see Chapter 31). The mechanism by which smoking affects bone in men is unclear. In women, tobacco use has been associated with lower weight, calcium absorption, and estrogen levels, all of which are negatively associated with bone mass [365,366,372]. Alternatively, smoking may impair respiratory function and hence bone metabolism, or a direct toxic effect of smoking on bone metabolism may exist. There are no data available concerning the effects of tobacco use in other forms (chewing, snuff). Whether smoking adversely affects androgen levels or other potential effectors of bone remodeling in men is unknown. 5. RENAL STONE DISEASE Several reports have linked hypercalciuria or nephrolithiasis in men to a reduction in BMD [231,373 – 377] and this issue was recently reviewed by Zerwekh [378]. It is not clear whether this apparent increase in bone disease results from a greater impact of nephrolithiasis in males [379] or perhaps reflects the fact that hypercalciuria is more than twice as common in men than women [380]. In fact,

121

CHAPTER 42 Osteoporosis in Men

hypercalciuria has also been linked to osteoporosis in women [381]. The etiology of the low bone mass observed in hypercalciuric patients is unclear, but has been postulated to involve an alteration in mineral metabolism. Somewhat surprisingly, osteopenia has been reported in patients with absorptive hypercalciuria [379,382 – 384]. In this setting, increased serum concentration of 1,25-dihydroxyvitamin D may increase bone resorption, but also the correlation of urinary sodium and sulfate levels with BMD in these patients implicate a contribution of dietary protein and sodium as well [382]. In renal hypercalciuria, a negative calcium balance with secondary hyperparathyroidism (including increased 1,25-dihydroxyvitamin D levels) is potentially important [383,385,386]. This hypothesis is supported by the finding that lower dietary calcium intakes in men with renal hypercalciuria are associated with further reductions in bone mass [387]. However, in some patients hypercalciuria may be only part of a more diffuse metabolic abnormality that affects bone metabolism in other ways. For instance, renal tubular acidosis may be present with hypercalciuria and low bone mass in the presence of complex abnormalities of mineral and bone metabolism [388], hypercalciuria has been linked to phosphate-wasting disorders causing low bone mass [389], medullary sponge kidney is not uncommonly associated with increased urinary calcium excretion and disordered parathyroid function [390 – 392], and Dent’s disease preferentially affects males and presents with metabolic bone disease, hypercalciuria and several other renal tubular abnormalities [393]. In most patients with idiopathic hypercalciuria, the bone deficit is quite modest, and of itself unlikely to result in clinically significant bone disease. On the other hand, renal lithiasis has been associated with symptomatic osteoporosis in men [231]. In a small series of relatively young men (five subjects ages 27 – 57 years), Perry et al. [394] found osteoporosis in association with moderate hypercalciuria in the absence of other risk factors for bone loss. In all patients, calcium hyperabsorption appeared to contribute to the hypercalciuria. Histomorphometric analysis revealed an increased rate of bone remodeling in the face of no apparent alteration in mineral metabolism. In a somewhat different experience, Zerwekh et al. [384] reported on 16 men (mean age 50  11) referred for evaluations of osteoporosis nine were hypercalciuric without any other obvious cause of bone disease. Further examination showed that all of the hypercalciuric group had evidence of an element of absorptive hypercalciuria, and four actually had increased gastrointestinal calcium absorption and 1,25-dihydroxyvitamin D concentrations. In the hypercalciuric subgroup, bone formation rates were depressed (reduced bone formation rate, increased mineralization lag time) in comparison to normals or normocalciuric osteoporotics, with no differences in indices of bone resorption. Similar findings were

reported in men with idiopathic hypercalciuria [395], and in a mixed population of absorptive and renal hypercalciurics [379]. In other groups of men with unexplained osteoporosis, some have been reported to be hypercalciuric [234,251, 396]. Thus, there are suggestive, but still preliminary data linking hypercalciuria to bone loss and osteoporosis in men. The specific pathophysiology involved and the clinical spectrum of resultant bone disease remain somewhat unclear. Although the relationship between hypercalciuria and osteopenia is relatively strong, and the pathophysiology at least superficially intuitive, some intriguing data suggest there may be other factors which are also important. Jaeger et al. studied a large group of renal stone formers and found low BMD not only in the hypercalciurics but also in the normocalciuric patients [377]. In their subjects, BMD was correlated with other factors which influence stone forming potential, including urinary sulfates, sodium, uric acid, and pH, raising the issue of whether the cause of osteopenia in stone formers is related to aspects of renal function aside from, or in addition to, calcium handling. 6. MISCELLANEOUS DISORDERS A variety of other illnesses or medications have been associated with bone loss or fractures in men, including anticonvulsant use [54,231,397], thyrotoxicosis [398], immobilization [399], liver and renal disease [232], homocystinuria [54], and others. However, there is little evidence to suggest that the skeletal abnormalities induced by these conditions affect men any differently (qualitatively or quantitatively) than women.

V. THE EVALUATION OF OSTEOPOROSIS IN MEN Guidelines for the efficient, cost-effective approach for the evaluation of patients with low bone mass, or patients suspected of having low bone mass, are poorly validated for either sex. Current recommendations are therefore based on existing knowledge of the epidemiology and clinical characteristics of osteoporosis [400,401] rather than upon models that have been carefully tested in prospective studies. Within these constraints it is possible to formulate an approach to the male osteoporotic (Fig. 17). Of necessity it derives from the more mature knowledge base available concerning osteoporosis in women, but may depart in several key areas.

A. The Diagnosis of Osteoporosis In some men (as in women) the diagnosis of an osteopenic metabolic bone disease can be made with basic

122

ORWOLL AND KLEIN

FIGURE 17

Scheme for the diagnosis and evaluation of osteoporosis in men [518].

clinical information. Most important is a clear history of low trauma fractures in the absence of evidence of a focal pathological process (malignancy, infection, Paget’s disease, etc.). There are several clinical situations in which the presence of osteoporosis cannot be confidently determined, but should be considered likely. In these circumstances further diagnostic steps are appropriate. These situations include the presence of suspicious fractures, the radiographic presence of low bone mass, and conditions known to be associated with increased risk of bone loss. 1. FRACTURES The presence of a low trauma fracture should raise the probability of metabolic bone disease and prompt further

evaluation (i.e., densitometry). Certainly the occurrence of classic osteoporotic fractures (vertebral, proximal femoral) in the absence of focal pathology should raise immediate concern. The incidental finding of vertebral deformity in men warrants comment, as their prevalence is relatively high [42,402]. It is frequently assumed that many of these result from excessive trauma, or developmental deformity (Scheuermann’s disease), and hence should not be considered the consequence of low bone mass. Davies et al. [403] reported that the prevalence of vertebral deformity is high in men, but does not increase with aging as it does in women, suggesting that these deformities may not be related to changes in bone mass or structure. However, other

123

CHAPTER 42 Osteoporosis in Men

studies show that men with even relatively small degrees of vertebral deformity (vertebral height reduction 2 standard deviations) generally have mean BMD values significantly below those of subjects without deformity [42] (Fig. 18). These observations argue that the finding of vertebral abnormality (for instance on chest radiography) should raise the concern that an osteopenic disorder is present, and further evaluation should be considered.

2. RADIOGRAPHIC DETECTION OF REDUCED BONE MASS Low bone mass detected radiographically, even in the absence of a fracture, is of concern because the loss of bone must be well advanced before it is detectable with routine x-ray procedures. If, in fact, a generalized reduction in bone mass is noted radiographically, it indicates the presence of clinically significant depression of bone mass, and should prompt a diagnostic evaluation. Although useful to observe when present, radiographic estimates of reduced bone mass are qualitative, and the actual quantitation of bone density provides a more definitive estimate of disease severity and a baseline from which to gauge subsequent change. 3. CLINICAL CONDITIONS ASSOCIATED LOW BONE MASS

WITH

As discussed above, it is apparent that there are a number of causes of increased risk for osteoporosis in men, including glucocorticoid excess, alcoholism, and hypogonadism (discussed above). The presence of one, or particularly several, of these conditions should prompt concern, and the consideration for further characterization of skeletal status. 4. BONE MINERAL DENSITY MEASUREMENTS

FIGURE 18

Lumbar spine (a) and femoral neck (b) bone mineral density in male subjects with and without vertebral deformity. Four grades of vertebral deformity are illustrated (anterior vertebral/posterior vertebral height ratio of 0.85 or 0.80, and anterior vertebral heights 2 SD or 3 SD of a population mean) [42].

In men who present with findings that suggest the presence of metabolic bone disease (low trauma fractures, radiographic criteria indicating the presence of a reduction in bone mass, or conditions associated with bone loss), the measurement of BMD should be strongly considered. These measurements can be useful in several ways, including cementing the diagnosis of low bone mass and gauging the severity of the process. Generalized screening of older men with bone mass measures is worth evaluating as a strategy, but should not yet be routinely adopted. For reasons that have been previously elaborated [404,405], bone density measures provide valuable data that cannot be deduced from other clinical information, and can solidify the diagnosis of low bone mass. Although this contention is derived from studies in women, the basic tenets should be applicable in men as well. Specifically, (i) low bone mass is related to fracture, (ii) bone mass measures predict fracture risk, (iii) bone mass can be accurately measured, (iv) an understanding of bone mass may influence the therapeutic approach, and (v) treatment of osteoporosis affects fracture risk. The diagnostic criteria that should be used to identify men with high fracture risk, and thus in need of intervention, is uncertain. Although it is clear that there is an inverse relationship between bone density and fracture risk [64,66], the relationship between bone mineral density and fracture risk is not well established in men [406]. Some have suggested that the relationship between the absolute

124

ORWOLL AND KLEIN

level of bone density and future fracture risk should be the same in men and women [60,66,407] while others have noted gender differences [408]. Until the fracture risk associated with any given level of bone density is established in prospective trials, it may be reasonable to utilize reference ranges based on young normal male values.

B. Differential Diagnosis The intent at this stage of the evaluation should be to determine with reasonable certainty the histological cause of the osteopenic disorder, and to identify the etiologic factors contributing to it. In women, the vast majority of patients with osteopenic fractures have histological osteoporosis, but a small proportion are found to be osteomalacic [409 – 411]. Similarly, a fraction of men with fracture have osteomalacia [409 – 411]. Osteomalacia is estimated to be present in 4 to 47% of men with femoral fractures, with most reports being 20% [409 – 413]. Since some foods are fortified with vitamin D, occult osteomalacia may be less frequent in the United States than in other areas (e.g., northern Europe). Increasing age is associated with an increasing prevalence of osteomalacia [411]. Some have suggested that women with femoral fracture are more frequently osteomalacic than men [409, 410], but others report no distinction [411]. Thus far, the only patients who have been carefully surveyed are those with proximal femoral fractures, and it is not known whether populations with other fractures (vertebral) would include similar proportions of osteoporotic and osteomalacic individuals. Although the exact magnitude of the problem presented by osteomalacia in men is uncertain, it is clear that any differential diagnosis of low bone mass and fractures in men must consider the possibility. This becomes particularly imperative because the treatment for osteomalacia differs considerably from that of osteoporosis [414]. 1. INITIAL EVALUATION: HISTORY, PHYSICAL, AND ROUTINE BIOCHEMICAL MEASURES The history, physical, and routine biochemical profile can be very helpful in directing a focused evaluation of a man with low bone mass. Several approaches for the differential diagnosis of low bone mass have been suggested using standard clinical and biochemical information [253,414,415]. The goals of this stage of the evaluation should be to determine the specific diagnosis (what is the cause of the low bone mass – osteoporosis or osteomalacia?) and to identify contributing factors in the genesis of the disorder. Of particular importance in the history and physical examination, therefore, are signs of genetic, nutritional/environmental, social (alcohol, tobacco), medical, or pharmacological factors that may be present. Routine laboratory testing should include serum creatinine, calcium,

phosphorus, alkaline phosphatase, and liver function tests, as well as a complete blood count. If, on the basis of these tests, there is evidence for medical conditions associated with bone loss (alcoholism, hyperparathyroidism, malignancy, Cushing’s syndrome, thyrotoxicosis, malabsorption, etc.), a definitive diagnosis should be pursued with appropriate testing. 2. EVALUATION OF THE PATIENT WITH “IDIOPATHIC” OSTEOPOROSIS In men with reduced bone mass in whom no clear pathophysiology is identified by the routine methods above, it has been considered appropriate to be diagnostically aggressive, primarily because the potential for occult “secondary” causes of osteoporosis may be higher in men. However, the incidence of occult causes of osteoporosis in men, or whether it is greater than in women, is poorly studied. The diagnostic yield and cost effectiveness of extensive biochemical studies in the man with apparently “idiopathic” osteoporosis is unknown. Nevertheless, lacking this information, a reasonable evaluation of the man without an obvious etiology for osteoporosis might include: • 24-h urine calcium and creatinine to identify idiopathic hypercalciuria • 24-h urine cortisol excretion to identify Cushing syndrome • serum 25-hydroxyvitamin D concentration • serum testosterone and LH 3. HISTOMORPHOMETRIC CHARACTERIZATION Transiliac bone biopsy is a safe and effective means of assessing bone histology and remodeling [416]. Some have suggested that a transiliac bone biopsy is indicated in those men in whom a thorough biochemical evaluation has failed to reveal an etiology for osteoporosis [233]. The rationale for this approach is based on the need to accomplish several objectives: (i) ensure that occult osteomalacia is not present; (ii) identify unusual causes of osteoporosis that may be revealed only by histological analysis, such as mastocytosis [417,418]; and (iii) to yield information concerning the remodeling rate, which in turn may further direct the differential diagnosis (e.g., unappreciated thyrotoxicosis or secondary hyperparathyroidism suggested by the presence of increased turnover) or may be helpful in designing the most appropriate therapeutic approach. However, considerable histologic heterogeneity exists among men with osteoporosis. Whether distinct histologic patterns represent different stages of a single disease entity, separate subtypes of the disease, or simply an arbitrary subdivision of a normal distribution of remodeling rates is unknown. Vigorous attempts have been made to substitute sensitive and specific biochemical markers of bone turnover for histomorphometric estimates of bone turnover. Serum

125

CHAPTER 42 Osteoporosis in Men

levels of osteocalcin and procollagen peptides, and urinary collagen crosslink excretion correlate well with remodeling rates [152], at least when patients with overt metabolic bone disorders are included in the analyses. In normals and most osteoporotic patients the usefulness of these measures in predicting bone turnover rates is less clear. Serum osteocalcin concentrations may be elevated in men with osteomalacia, but appear to offer no added sensitivity for making the diagnosis of osteomalacia beyond that of conventional biochemical tests (alkaline phosphatase, 25-hydroxyvitamin D) [419]. Similarly, urinary pyridinoline excretion may be increased in osteomalacia [420]. With this information, a reasonable approach to the evaluation of remodeling dynamics in men with idiopathic osteoporosis (no etiology apparent from non-invasive testing) may be to combine the advantages of the biochemical markers of bone turnover with those of bone biopsy. An initial biochemical assessment of bone turnover should provide an understanding of remodeling rate. In the presence of an increase in biochemical indices of remodeling (osteocalcin, pyridinoline), a bone biopsy may be appropriate to identify unusual causes of high-turnover osteoporosis (e.g., mastocytosis). Bone biopsy may also be particularly helpful if there is any clinical concern for occult osteomalacia. Although alkaline phosphatase activity is usually increased in osteomalacia [414], even in this situation, a bone biopsy can reveal unanticipated osteomalacia, particularly in older men [411]. Unfortunately, the diagnostic yield or clinical impact of the bone biopsy is unknown. There is concern that it may be low, thus detracting from its clinical applicability. Essentially, the decision to utilize a bone biopsy is not well codified, and remains a matter of expert judgment (see Chapter 63).

VI. THERAPY Therapy of osteoporotic disorders in men is relatively unexplored. There have been very few trials of osteoporosis therapies performed specifically in male populations, although some men with osteoporosis have been included in mixed populations treated with a variety of agents [421]. In general it is very difficult to assess independently the success of these approaches in the male subjects.

A. Calcitonin There has been one trial of calcitonin therapy in a small group of men with idiopathic osteoporosis [422] in which total body calcium tended to increase during a 24-month treatment interval (100 IU administered subcutaneously each day with a calcium and vitamin D supplement). However, the change was not significantly different from that

observed in the control groups (receiving calcium plus vitamin D supplements, or vitamin D alone), and there were no changes in radial bone mass. In another uncontrolled, 12month trial of subcutaneously administered cyclical calcitonin (100 IU three times per week for three months, followed by three months without calcitonin) in men with vertebral osteoporosis small benefits were noted in spinal and proximal femoral bone density (compared to baseline) [423]. Men have been included in several other trials of calcitonin therapy, but the results in men are not separable from those in women subjects. There are no published studies of the effectiveness of intranasal calcitonin in men. Although there are few data, from a theoretical perspective calcitonin should be useful in reducing osteoclastic activity in at least some patients with osteoporosis or in those at risk of continuing bone loss. Pain following vertebral fracture has been reported to be alleviated with calcitonin, and some reports of this benefit have included men [424]. Whether men can be expected to respond differently than women is unknown.

B. Bisphosphonates There have been few trials of bisphosphonates performed exclusively in men, and many have been reported only in preliminary form [421]. Nevertheless, there is no conceptual barrier to the use of bisphosphonates in men, and recent reports describe positive results. Male patients with osteoporosis have been included in mixed patient populations, and have seemed to experience beneficial effects on calcium balance and lumbar spine bone density during treatment with pamidronate [425]. Men were specifically reported to benefit (increased vertebral bone density, with no change in femoral density) from etidronate treatment in a 12-month study [426]. An uncontrolled observational experience with intermittent cyclical etidronate (with calcium supplementation) in men with idiopathic osteoporosis and vertebral fractures [427] recently found small increases in lumbar spine and proximal femoral bone mass (3.2 and 0.7% per year, respectively) (Fig. 19), but no data were presented concerning fracture occurrence. There was no change in alkaline phosphatase activity. In another uncontrolled trial, Geusens et al. reported a somewhat more robust response to cyclical etidronate in osteoporotic men [428]. In the first large controlled trial of a bisphosphonate (or any therapy) in men with primary osteoporosis, alendronate had positive results on bone mass and reduced the rate of vertebral fracture [429]. These results provide considerable support for the effectiveness of bisphosphonates in men with osteoporosis. Of interest, the increase in bone mass resulting from alendronate was as great as was previously reported in postmenopausal women with osteoporosis, and was as great

126

ORWOLL AND KLEIN

FIGURE 19

Men with idiopathic osteoporosis appeared to have a positive response to therapy with intermittent cyclical etidronate, especially subjects of greater age. The scatterplot is of change in lumbar spine bone mineral density (BMD) against age at first scan. The average change in BMD per year of treatment was 0.024 g/cm2, or 3.2% [427].

in men who began the trial with low free testosterone levels as in those with normal levels. Bisphosphonates have also been examined in men with secondary causes of osteoporosis, especially in the context of glucocorticoid therapy. Men were included in some studies that indicated a positive effect of etidronate in glucocorticoid-treated patients [430]. For instance, in a large trial of alendronate in men receiving glucocorticoids, positive effects were noted in lumbar spine BMD [431]. Similar results have been reported in trials with other bisphosphonates (etidronate, risedronate) [432,433] in which the increase in BMD at several skeletal sites, and the tendency toward a reduction in fracture risk, was similar in men and women. There are a variety of other situations in which bisphosphonates may be useful, but little experience is yet available. For instance, inhibitors of bone resorption have been considered attractive in states of immobilization and in inflammatory conditions (e.g., rheumatoid arthritis). Men who receive anti-androgen therapy for prostate carcinoma are at risk for bone loss, and anti resorptive therapy should provide some protection for those patients. In fact, Diamond et al. reported that intermittent cyclic etidronate therapy (plus calcium) reversed bone loss initially experienced in men following long acting gonadotropin-releasing hormone agonist plus androgen antagonist therapy [307]

(Table 2). Some early reports are available in other conditions [434], and more can be expected as the effects of bisphosphonates in men are further explored.

C. Thiazide Diuretics Evidence supports a beneficial effect of thiazide administration on bone mass, rates of bone loss, and hip fracture risk in men [435 – 437]. For instance, in case controlled trials the use of thiazides reduced the rate of loss in calcaneal bone density by 49% compared to controls [438] and the relative risk of hip fracture was halved by exposure to thiazides for more than 6 years [439]. In a trial of similar design, thiazide use in men was associated with an adjusted odds ratio of femur fracture of 0.2 (95% CI 0.1 – 0.7) [440]. Other diuretics did not seem to impart the same benefits. Unfortunately, none of the available studies has been randomized or controlled, so a confident estimate of the magnitude of the protective effect is not possible. Moreover, the available literature does not allow a comparison of the relative benefits in men and women [441]. The mechanism for the positive effect is unclear, but it has been postulated to stem from the hypocalciuric effects of thiazides. Although probably not appropriately considered a primary treatment modality, a thiazide is probably the diuretic of

127

CHAPTER 42 Osteoporosis in Men

Concentrations of osteocalcin (g/liter), carboxyl-terminal cross-linked telopeptide of type I collagen (ICTP; g/liter), carboxyl-terminal propeptide of type I procollagen (PICP, g/liter  10 – 1) and bone-specific alkaline phosphatase (b-ALP; kat/L  10 – 1) in 21 men and 15 women with GHD before and after 9 months of treatment with recombinant human growth hormone. Results are shown as the mean  SD. **P  0.01; ***P  0.001. The P values in letters refer to the difference in response to treatment between men and women [164].

FIGURE 20

choice in osteoporotic patients (other considerations not withstanding).

D. Parathyroid Hormone Parathyroid hormone administration to osteoporotic subjects has been shown to increase trabecular bone formation and bone volume in concert with an increase in calcium balance [442 – 444]. Slovik et al. reported that in a small group of men with idiopathic osteoporosis, combined PTH and 1,25-dihydroxyvitamin D administration increased trabecular (spinal) bone mass and improved intestinal calcium absorption [443]. Recently, PTH has been shown to exert striking increases in BMD and decreases in fracture incidence in postmenopausal osteoporotic women (see Chapter 77). Although its ultimate role in the treatment of osteoporosis, either alone or in concert with other agents [444], remains unclear, the potential of PTH appears similar in men and women.

E. Growth Hormone Growth hormone (or other growth factors) [155] may have anabolic actions on the skeleton in the elderly and in subjects with osteoporosis, but the available data are inconclusive [445]. Low levels of IGF-I have been reported to be

present in men with idiopathic osteoporosis, and in a study of healthy older men with low IGF-I levels, Rudman et al. [446] found that in addition to positive effects on lean mass, fat mass, and skin thickness, vertebral bone mass was increased slightly (1.6%) by the administration of growth hormone for 6 months. Radial and proximal femoral densities were unaffected. There are a number of reports that growth hormone administration may improve bone mass in growth hormone deficient adults [447,448], and the treatment of adults with growth hormone provokes an increase in biochemical markers of bone remodeling. Men have been reported to be more responsive to replacement therapy with growth hormone than women (Fig. 20) [164,165], raising interesting questions of the relative importance of growth factors in men vs women and the role of sex steroids in growth factor action [449]. The potential benefits of growth hormone on body composition (increased lean mass and perhaps muscle strength) [450] have been suggested to have additional benefits in patients with osteoporosis, but the functional importance of those changes has been questioned [451,452]. In adults with growth hormone deficiency, growth hormone replacement may have a more promising role. In that context bone mass has been noted to increase in most [453 – 455] but not all [456] studies. The response might be dependent on sex steroid action [449]. Despite interesting preliminary findings, the use of growth hormone is fraught with a variety of uncertainties, the benefits remain inconsistent and experimental results have been

128

ORWOLL AND KLEIN

difficult to interpret [445]. In either sex, growth hormone therapy is thus of potential, but as yet unproven usefulness (see Chapter 78).

ularly those at risk of vertebral fracture (see Chapters 74 and 75).

G. Specific Therapeutic Approaches F. Fluoride The use of fluoride in the therapy of osteoporosis remains controversial. Consistent and sometimes dramatic increases in vertebral bone mass can be achieved with supplemental fluoride, but the effectiveness of fluoride therapy in reducing fracture rates is uncertain. Nevertheless, there is an active interest in refining the formulation and dose of fluoride in the hope of taking better advantage of its anabolic properties. In fact, a recent evaluation of cyclically administered slow-release form of fluoride was reported to increase bone mass and reduce fracture rates in older women [457]. As with many of the other therapies discussed, there have been few specific trials of fluoride administration in men. In some studies, osteoporotic men have been included in the treatment groups but it is difficult to ascertain whether responses were in any way sex-specific. Ringe et al. recently described a 36-month, controlled trial of intermittent monofluorophosphate and calcium (vs calcium alone) in a large group on men with idiopathic osteoporosis [458]. Whereas the calcium-treated men lost bone density at the spine, radius, and proximal femur, those treated with fluoride experienced increases. The most marked change was seen at the lumber spine (8.9%), with more modest increases at the other sites. Importantly, there were fewer patients who experienced vertebral fractures in the fluoride treated group (10% vs 40%, P  0.008) (Fig. 21), and back pain was reduced (P  0.0003). There were fewer non-vertebral fractures in the fluoridetreated men as well, but the difference was not significant. These results are similar to parallel studies in women and suggest that fluoride may have some benefit in men, partic-

FIGURE 21 The percentage of osteoporotic men who experience new vertebral fractures during 3 years of treatment with monofluorophosphate plus calcium (MFP/Ca) or calcium alone (Ca) [458].

As discussed above, there is some information that suggests that specific treatment of underlying conditions associated with low bone mass can be effective in stabilizing or improving skeletal mass in osteoporotic men, but even those data are scarce, and the effects of these therapies on fracture risk are unknown for most agents. A suggested approach for the treatment of osteoporosis in men is outlined in Figure 22. 1. PREVENTION OF AGE-RELATED BONE LOSS Although it has recently become apparent that bone loss and fractures with aging in men is an important public health issue, there is very little information available concerning its prevention. Reasonable guidelines can be developed on the basis of current pathophysiologic models and on experience in women, but these approaches lack validation. a. Exercise Whereas an exercise prescription is diffi cult to generate with currently available information, activity is probably beneficial in several ways. Reductions in strength and coordination contribute to fracture via an increased risk of falling [459]. In addition, inactivity is associated with bone loss, and exercise may increase or maintain bone mass. Specific exercise prescriptions to accomplish these goals have not been confirmed in men or women, although it is clear that strength can be dramatically increased, and risk of falls reduced, in the elderly with achievable levels of exercise [459 – 461]. That fracture rates are lower in elderly men who exercise modestly buttresses this contention [186]. Beck and Marcus have recently reviewed the issue of exercise, men, and skeletal health [187] (see Chapter 28). b. Calcium An area of obvious interest is the influence of calcium and vitamin D nutrition [462]. Calcium intake is probably important in the achievement of optimal peak bone mass in boys [119], as well as the prevention and therapy of osteoporosis later in life. Calcium absorption declines with aging in men as in women, particularly after the age of 60, and well-documented changes in mineral metabolism occur concomitantly with age in men [178]. These data suggest both that optimal levels of calcium intake may change with age and that inadequate calcium nutrition can have an adverse effect on skeletal mass. However, the level of calcium intake that should be recommended is unclear, as few prospective studies have addressed this issue. No bone density benefit was observed from calcium/ vitamin D supplementation in a very well nourished population (mean dietary calcium intake 1000 mg/day) [56], and no anti-fracture benefit was observed in a large trial of

129

CHAPTER 42 Osteoporosis in Men

FIGURE 22

Algorithm for the choice of men at risk of fracture who may benefit from pharmacological therapy [519].

vitamin D supplementation in older men and women with relatively high baseline intakes [463]. Dietary calcium intake was not found to be related to fracture rate in the men followed as part of the Health Professionals Follow-up Study [464]. On the other hand, an improvement in bone density was noted in healthy older men in response to a calcium and vitamin D supplement, while placebo treated men lost bone [465]. On the basis of the available information, and the likelihood of a high degree of safety, the U.S. Institute of Medicine recently recommended that men should have a calcium intake of 1200 mg/day, and a vitamin D intake of 800 IU. A reasonable approach, therefore, is to suggest a calcium intake of at least 1200 mg/day in both preventative as well as therapeutic situations. A NIH Consensus Development Conference has suggested the somewhat higher calcium intake of 1500 mg/day in men after 65 years [466]. Although these recommendations for supplemental calcium and vitamin D are reasonable, some attention to individual differences is probably important. For instance, the use of an invariant level of vitamin D supplementation

(e.g., 800 IU/day) may result in inadequate effects in some patients, especially those who have low levels of vitamin D at baseline. In a study of the effects of vitamin D (and calcium supplementation) in men, Orwoll et al. found that the average increase in 25(OH) vitamin D concentration in response to 25 g (1000 IU) per day of cholecalciferol was 30 nmol/liter (12 ng/dl). However, the increase was no greater in those who started with reduced 25(OH) vitamin D levels (Fig. 23), with the result that men who start with low vitamin D status could be inadequately treated with conventional amounts of supplement. Certainly vitamin D insufficiency is common in older men [467 – 470], and adjustments in the dose of supplements based on initial vitamin D values may be useful. The use of follow-up vitamin D measurements should provide assurance that adequate vitamin D status has been achieved. Similarly, in some special situations (e.g., glucocorticoid excess, malabsorption) dietary calcium requirements may be somewhat increased over those routinely recommended. At a given level of sodium excretion, elderly men were found to have a greater calcium excretion than women (despite similar dietary

130

ORWOLL AND KLEIN

FIGURE 23.

The relationship between baseline and follow-up 25(OH)D levels in normal men treated with 25 g (1000 IU) per day of cholecalciferol or placebo for 1 year. The regression line for each group is shown [520].

calcium intakes) [471], suggesting that excess dietary sodium intake should be especially avoided in men at risk for osteoporosis. One theoretical concern regarding dietary calcium/vitamin D supplementation has been the precipitation of calcium renal stones in susceptible individuals. Recent data suggest that dietary calcium intake actually correlates negatively with the risk of nephrolithiasis in men [472], potentially by increasing gastrointestinal oxalate binding. The Institute of Medicine recommendations suggest that intakes below 2000 mg/day are safe. 2. IDIOPATHIC OSTEOPOROSIS Idiopathic osteoporosis in men is a perplexing and difficult disorder for the clinician. In addition to the lack of knowledge of its pathophysiology, there have been essentially no attempts to define appropriate treatments. In the absence of contributory factors that can be addressed, basic issues of nutritional adequacy (calcium and vitamin D) and physical activity (both for its trophic effects on the skeleton as well as the desire to maintain strength and coordination to prevent falls) should be addressed. As low bone mass must have its genesis in a remodeling imbalance, it may be appropriate to consider antiresorptive agents (bisphosphonates, calcitonin). The efficacy of alendronate in the therapy of men with osteoporosis (without secondary causes) was recently demonstrated [429], and in that light bisphos-

phonates should be considered a mainstay of therapy. Since many men with idiopathic osteoporosis appear to have impaired osteoblastic function, the development of boneforming therapies (e.g., PTH) is very attractive. 3. GLUCOCORTICOID EXCESS The current clinical management of glucocorticoid-induced osteoporosis is based on limited data, not only with regard to the efficacy of preventive and therapeutic regimens, but also in terms of our limited understanding of the pathophysiology of the disease. Various therapies including, calcium, vitamin D, calcitonin, bisphosphonates, sex steroids, and fluoride have been examined, but usually in open studies of limited subjects measuring effects on bone mass rather than large scale investigations evaluating fracture risk (see Chapter 44). Certainly management of patients receiving long-term glucocorticoid therapy should include minimally effective doses, at all times; discontinuation of the drug, when practical; and topical administration, if possible. Although alternate-day glucocorticoid dosing preserves normal function of the hypothalamic – pituitary – adrenal axis, there is no evidence that such a regimen offers any advantage in terms of preventing bone loss [473,474]. Calcium supplements diminish indices of bone resorption [475] and thiazide diuretics combined with reduced dietary sodium intake improve gastrointestinal absorption

131

CHAPTER 42 Osteoporosis in Men

of calcium and attenuate urinary calcium losses [476,477]. Pharmacologic doses of vitamin D have been widely used to treat glucocorticoid-induced osteoporosis in the past. Such therapy is not justified on the basis of vitamin D deficiency [478] and has not consistently shown therapeutic benefit [479 – 484]. However, in some studies which included male patients, vitamin D therapy appeared beneficial. For instance, Sambrook et al. [484] demonstrated a preservation of lumbar spine (but not femoral or radial) BMD with calcium and calcitriol. In this study, the addition of nasal calcitonin to calcium and calcitriol had no additional benefit. Vitamin D toxicity frequently accompanies use of pharmacological doses of vitamin D [479,481,482, 484,485]. Supplementation with lower doses (800 – 1000 IU/day) is certainly safe. Because long-term glucocorticoid therapy reduces serum testosterone levels and administration of testosterone to hypogonadal men improves bone mass, such therapy may be helpful, but has not been adequately evaluated. Sodium fluoride stimulates replication and function of osteoblasts and, as such, might be particularly useful in overcoming the primary inhibitory effects of glucocorticoids on the osteoblast. Several open and uncontrolled studies have revealed variable responses to treatment. One study of only 6 months duration found no effect

on the rate of glucocorticoid-induced bone loss [483], while another study examining long-term fluoride therapy demonstrated marked histologic improvement in indices of bone formation and trabecular mass [486]. Agents that inhibit bone resorption, such as calcitonin and bisphosphonates, have also been shown to be of therapeutic benefit [484, 487 – 491]. However, the efficacy of these agents appears to be greatest when administered in a preventive fashion from the time of initial exposure to glucocorticoids [484,487].

VII. HYPOGONADISM A. Androgen Replacement in Hypogonadal Adult Men Androgen therapy in hypogonadal men has been shown to positively affect bone mass, at least in most patient groups [305,312,492,493]. For instance, Katznelson et al. recently reported an increase in spinal BMD of 5 – 6% in a group of adult men with hypogonadism treated with testosterone for 18 months [494], although there was an insignificant increase in radial BMD (Fig. 24). As in the experience reported by Katznelson et al., the increase in density

FIGURE 24 Changes in percentage body fat and BMD in hypogonadal men receiving testosterone replacement therapy. (a) Percentage body fat determined by bioelectric impedance analysis. (b) AP spinal BMD determined by dual-energy X-ray absorptiometry. (c) Radial BMD determined by single photon absorptiometry. Data are represented as the mean  SEM percentage of the baseline. Statistical significance for analysis of the mean slope is shown in the bottom right-hand corner of each figure [494].

132

ORWOLL AND KLEIN

FIGURE 25

Increase in spinal BMD during long-term testosterone substitution therapy up to 16 years in 72 hypogonadal patients. Circles indicate hypogonadal patients with first quantitative computed tomography (QCT) measurement before institution of therapy, squares show those patients already receiving therapy at the first measurement. The dark shaded area indicates the range of high fracture risk, the top area shows the range without significant fracture risk, and the middle area indicates the intermediate range where fractures may occur [495].

following testosterone replacement generally appears to be most apparent in cancellous bone (e.g., lumbar spine), although the literature is not particularly consistent in this regard. Most reports indicate that the increase in bone mass with testosterone therapy can be expected to be modest in the short term (up to 24 months), but Behre et al. noted an increase in spinal trabecular BMD of 20% in the first year of testosterone therapy in a group of hypogonadal men, and further increases thereafter [495] (Fig. 25). The most marked increases were observed in those with the lowest testosterone levels before therapy. In men treated for at least 3 years, bone density was found to be at levels normally expected for their ages. Although the experience remains small, there is a suggestion that in older men with hypogonadism the response to therapy can be expected to be similar to that in younger adult patients [495,496]. The cellular mechanisms responsible for improvements in bone mass are unclear. As discussed above, in the early phases of androgen deficiency (e.g., following castration) there appears to be a phase of increased remodeling and resorption, so that therapy may be beneficial because of an inhibitory effect on osteoclastic activity. However, in most available clinical studies, treated patient populations have had well-established hypogonadism and were characterized by an array of remodeling states. In these subjects the cellular effects of androgen replacement are not well known. In

some reports testosterone therapy appeared to result in an increase in cancellous bone formation [315,317], but in other series there appeared to be no clear remodeling trend induced by therapy [492]. Most recently, several groups have reported that biochemical indices of remodeling decline in response to testosterone replacement [494,497], which is what might be predicted if sex steroid deficiency results in an increase in remodeling, and bone loss on that basis. Interestingly, some reports also suggest that osteocalcin levels may increase with androgen therapy [496,498], perhaps signaling an increase in bone formation. In addition to the generally positive effects of androgen replacement therapy in hypogonadal men, additional benefits may be gained from the increases that have been noted in strength and lean body mass in these patients [494,496,499,500]. Since lean body mass and strength have been correlated with bone mass and a reduced propensity to fall, they may further serve to promote bone health and reduce fracture risk. Despite the generally positive tenor of most studies of the skeletal effects of testosterone replacement, in some patient groups, for instance those with Kleinfelter’s syndrome, the advantage associated with androgen therapy is questionable, as the available studies report very mixed results [501,502]. This may be because the level of androgen deficiency in Kleinfelter’s (as in the case of some other

133

CHAPTER 42 Osteoporosis in Men

causes of hypogonadism) is quite variable. These findings suggest the need to carefully consider the potential benefits of androgen replacement in each patient individually. The most efficacious doses and routes of androgen administration for the prevention/therapy of bone loss in men remain uncertain. The specific testosterone values necessary for an optimal effect have not been defined, but current practice is to attempt to ensure testosterone concentrations similar to those of normal young men. Moreover, whether the pulsatile pattern of testosterone exposure characteristic of intramuscular administration is more or less conducive to skeletal health than the more stable pattern produced by transdermal administration is unknown. In some studies, transdermal testosterone therapy appeared to be as effective as was intramuscular administration in promoting bone mass [495]. Current oral preparations of androgens are not appropriate in view of the higher incidence of adverse effects associated with their use. However, newer androgenic compounds may obviate this concern. The follow-up of hypogonadal men treated with testosterone, although not well codified, should certainly include careful monitoring for adverse effects. The risk of prostate disease in androgen treated men is unknown, but regular prostate evaluations are necessary to ensure that any development of benign or malignant disease is detected early in its course. The development of errythrocytosis is not uncommon, particularly with intramuscular testosterone administration, and complete blood counts at 6 – 12 month intervals are useful to detect its appearance. Other problems that have been postulated to be of concern in androgentreated men are hyperlipidemia and sleep apnea [503]. In terms of skeletal disease, therapeutic success may be assessed via follow-up bone mass measures. In view of recent reports, increases in bone density can be anticipated in the average patient. Although the role of biochemical markers of remodeling is controversial, the available data suggest that an adequate androgen effect should be accompanied by a fall in indices of bone resorption, an effect that should be especially useful if resorption markers are increased at baseline. Markers of bone formation may be more difficult to use at present in routine clinical situations, as some reports suggest that increases follow therapy while others support a decline. The response may depend on the specific marker. Clinicians deciding on a follow-up strategy must be aware of the uncertainty currently inherent in the field and the vagaries of using the available tools (i.e., issues of measurement precision). There remain many additional unresolved issues concerning the role of androgen treatment in the prevention/ therapy of osteoporosis in hypogonadal men, including: • The degree of hypogonadism (level of testosterone) at which adverse skeletal effects begin to occur is undefined, and hence it is difficult to decide upon the usefulness of

therapy in many men with borderline levels of serum testosterone. • Because hypogonadism in men results in deficiencies of estrogen as well as testosterone, and since testosterone therapy results in increases in serum estrogen (as well as androgen) levels, the relative roles of estrogen vs testosterone in affecting skeletal health in hypogonadal men are unclear. It is unknown whether it is useful to assess estrogen concentrations in the diagnosis of hypogonadal bone disease in men, or whether using estrogen levels to monitor the success of testosterone therapy is beneficial. • In general, the available treatment studies are of relatively short duration, and it is unclear how long any increases in bone mass can be sustained and what eventual treatment effect can be expected. • As of yet, the increase in bone mass that appears to accompany testosterone therapy is of uncertain usefulness in preventing fractures. • Whether pretreatment age, duration of hypogonadism, degree of osteopenia, remodeling character, and associated medical conditions affect the therapeutic response is relatively unknown. • Potential adverse effects of androgen therapy (e.g., prostate, lipid) are not well delineated.

B. Androgen Therapy in Eugonadal Men It has been hypothesized that androgens may have positive effects on bone formation and resorption. The threshold level of androgens necessary to provide maximal skeletal benefits is unknown, and some have speculated that testosterone supplementation would benefit osteoporotic men even in the face of normal testosterone levels. The experience with this approach has been very limited, but Anderson et al. recently found in an uncontrolled trial that testosterone supplementation was associated with an increase in bone density, and a reduction in biochemical markers of remodeling, in a group of osteoporotic, eugonadal men [504]. This approach remains very much of uncertain benefit, and until its advantages are documented in controlled trials it cannot be recommended. This is particularly true in view of the lack of knowledge concerning the potential adverse effects that may be associated with testosterone supplementation.

C. Androgen Replacement in Adolescence Because adolescence is such a critically important part of the process of attaining optimal peak bone mass, it is also especially vulnerable to disruption by alterations in gonadal function. Even constitutional pubertal delay is associated with a reduction in peak bone mass development,

134

ORWOLL AND KLEIN

despite eventual full gonadal development [294, 296]. The impairment in bone mass in adolescence with organic hypogonadism (hypogonadotropic hypogonadism) is similar to that in patients with this form of hypogonadism studied later in life, suggesting that the detrimental effect suffered in adolescence is the major cause of osteopenia [290]. In view of the major effects of androgens on the skeleton during growth (whether direct or indirect, as discussed above), the response to therapy of gonadal dysfunction during this time would be expected to be brisk. Although studies are few, this would appear to be the case [505]. Finkelstein et al. reported that treatment of hypogonadal men with testosterone elicited the most robust skeletal response in those who were skeletally immature (open epiphyses) [492]. In young men considered to have constitutional delay of puberty, testosterone therapy results in a clear increase in bone mass, but whether this provides a solution to the problem of low peak bone mass in these patients is not yet known [506]. All this information suggests that the diagnosis of frank hypogonadism during childhood or adolescence carries with it the risk of impaired skeletal development, and that there is an opportunity to improve bone mass with androgen therapy. In fact, from a skeletal perspective, it appears that therapy should be initiated before epiphyseal closure to maximize bone mass accumulation. Issues that are unresolved include whether bone mass can be normalized with therapy, the most appropriate doses and timing of therapy, and the source of the beneficial effects (androgen vs estrogen, growth factor stimulation, etc.).

the levels (threshold concentrations) that are associated with adverse effects on bone.

E. Androgen Therapy in Secondary Forms of Metabolic Bone Disease A variety of system illnesses and medications are associated with lowered testosterone levels [508], and it has been postulated that relative hypogonadism may contribute to the bone loss that also accompanies many of these conditions. For instance, renal insufficiency, glucocorticoid excess, post-transplantation, malnutrition, and alcoholism are all associated with osteopenia and with low testosterone concentrations. Although there is little experience with testosterone supplementation in these patients, there may be advantages to skeletal health as well as to other tissues (muscle, red cells, etc.). In a randomized study of crossover design, Reid et al. [509] reported that testosterone therapy apparently improved bone density (and body composition) in a small group of men receiving glucocorticoids (Fig. 26). Similarly, testosterone therapy apparently improved forearm bone mass in a small group of men with hemochromatosis (treated simultaneously with venesection) [305]. The number of patients affected by conditions associated with low testosterone levels is potentially quite large, and more information is needed to understand the role of androgen replacement in the prevention/therapy of concomitant bone loss.

D. Androgen Replacement in Aging Men Old age is associated with a panoply of physical changes in men, many of which have been speculated to be related, either directly or indirectly, to the decline in androgens that accompanies aging [507]. A few small trials of androgen administration in older men have suggested that there may be beneficial effects (increased strength and improved body composition) [217,496,500], and some reports indicate that bone mass or biochemical indices of remodeling may improve [217,218,496]. Whether androgen replacement therapy can prevent or reverse bone loss in aging men is of enormous importance, but until more definitive data are available concerning both advantages and disadvantages, testosterone replacement should not be utilized in elderly patients unless there is convincing evidence for androgen deficiency. This decision is difficult in many older men who have symptoms that can be associated with androgen deficiency but which are also common in the aged regardless of gonadal status (weakness, loss of libido or sexual ability, etc.). The identification of hypogonadism in this group is made especially challenging by the expected decline in androgen levels with age and the dearth of data concerning

FIGURE 26

Rate of change in BMD of lumbar spine during control or testosterone treatment periods (each of 6 months duration) in men receiving glucocorticoid therapy. Data are given as the mean  SEM. There was a significant difference between groups (P  0.05). The asterisk indicates a significant difference from 0 (P  0.005) [509].

135

CHAPTER 42 Osteoporosis in Men

F. New Development in Sex Steroid Therapy in Men Selective estrogen receptor modulators have considerably altered the concept of estrogen replacement therapy in postmenopausal women (see Chapter 70). Considerable interest has developed concerning the relevance of these compounds in the treatment of men as well. Animal studies suggest that selective estrogen receptor modulator may have encouraging effects in males [510], and clinical trials are underway in men. Finally, selective androgen receptor modulators are being developed, and promise to be useful as osteoprotective agents while reducing adverse effects on prostate, lipids, etc. [511].

reported. The toxic effects of alcohol and fluoride on the gastrointestinal tract may likely preclude its use in individuals that continue to drink.

IX. TOBACCO The ideal approach to osteoporosis associated with tobacco use is smoking cessation. Whether cessation leads to reduced rates of bone loss or to a gain in bone mass is unknown. In the studies by Slemenda et al. [130] there was apparently a protective effect of heavy physical activity on the bone loss induced by smoking. Other interventions (nutritional supplements, antiresorptive drugs) might potentially reduce the incidence of low bone mass in men who did not abstain, although these possibilities are untested.

VIII. ALCOHOLISM Osteoporosis should be suspected in every chronic alcohol abuser, and patients with “idiopathic” osteoporosis should be routinely and thoroughly questioned about drinking habits. Once the diagnosis of alcohol-induced bone disease has been established, a number of measures are recommended. Aggressive medical and psychiatric treatment should be pursued in the hopes of interrupting the cycle of chronic alcohol ingestion and thereby diminish the risk of further skeletal deterioration. A careful dietary history should be followed by an adequate well-balanced diet rich in calcium-containing products. Evidence that calcium supplementation will improve the bone disease of alcoholics has not been reported, but it is reasonable to minimize other potential risk factors for bone loss if possible. Adequate vitamin D nutrition and physical exercise should be encouraged. Tobacco use and excessive consumption of phosphate-binding antacids should be discouraged. Presumably, the cessation of alcohol intake will stop further progression of bone loss, but data are scant. Moreover, no evidence has been reported that bone, once lost, will be restored when alcohol abuse is discontinued. Studies on alcohol abstainers have demonstrated a rapid recovery of osteoblast function (as assessed histomorphometrically and by biochemical parameters of bone remodeling) within as little as 2 weeks after cessation of drinking, but no significant differences in bone mineral content were observed between abstainers and actively drinking men [302,326,332, 348,361]. The relatively short period of abstinence however, makes these results inconclusive. The challenge in alcohol-induced bone disease is to stimulate bone formation. Most drugs currently used to treat other forms of osteoporosis work primarily by inhibiting osteoclastic bone resorption. Agents such as fluoride, parathyroid hormone or growth hormone may stimulate bone formation, but such regimens remain investigational and no therapeutic trials in alcoholic men have been

X. RENAL STONE DISEASE Attempts to treat low bone mass associated with idiopathic hypercalciuria are not yet well developed, but have been summarized recently [378]. As the most likely cause of defects in bone remodeling result from the mineral abnormalities induced by the renal calcium disturbance, it seems prudent to prevent hypercalciuria. In patients with either renal or absorptive hypercalciuria, thiazide diuretics would be appropriate. It is unknown whether therapy of the metabolic disturbances present in some patients with hypercalciuric renal stones (e.g., acidosis) is associated with any benefit.

References 1. N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton, Medical expenditures for the treatment of osteoporotic fractures in the united states in 1995: Report from the national osteoporosis foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). 2. D. T. Gold, Outcomes and the personal impact of osteoporosis. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 51 – 63. Academic Press, San Diego, 1999. 3. C. Niewoehner, Osteoporosis in men: Is it more common than we think? Postgrad. Med. 93, 59 – 60, 63 – 70 (1993). 4. E. Seeman, Osteoporosis in men: Epidemiology, pathophysiology, and treatment possibilities. Am. J. Med. 95, 22S – 28S (1993). 5. A. C. Scane, A. M. Sutcliffe, and R. M. Francis, Osteoporosis in men. Clin. Rheum. 7, 589 – 601 (1992). 6. T. J. Arneson, L. J. I. Melton, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin. Orthop. Relat. Res. 234, 188 – 194 (1988). 7. L. J. I. Melton and S. R. Cummings, Heterogeneity of age-related fractures: Implications for epidemiology. J. Bone Miner. Res. 2, 321 – 331 (1987). 8. K. M. Sanders, E. Seeman, A. M. Ugoni, J. A. Pasco, T. J. Martin, B. Skoric, G. C. Nicholson, and M. A. Kotowicz, Age- and genderspecific rate of fractures in Australia: A population-based study. Osteoporosis Int. 10, 240 – 247 (1999).

136 9. A. A. o. Orthopaedic Surgeons, Musculoskeletal injuries. In “Musculoskeletal conditions in the United States” (A. A. o. O. Surgeons, ed.), pp. 85 – 124. American Academy of Orthopaedic Surgeons, Park Ridge, IL, 1992. 10. M. Karlsson, R. Hasserius, and E. J. Obrant, Individuals who sustain nonosteoporotic fractures continue to also sustain fragility fractures. Calcif. Tissue. Int. 53, 229 – 231 (1993). 11. M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public Health 74, 1374 – 1380 (1984). 12. W. E. Bacon, G. S. Smith, and S. P. Baker, Geographic variation in the occurrence of hip fractures among the elderly white US population. Am. J. Public Health 79, 1556 – 1558 (1989). 13. S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Hip fracture incidence among the old and very old: A population-based study of 745,435 cases. Am. J. Public Health 80, 871 – 873 (1990). 14. J. D. Gallagher, L. J. Melton, and B. L. Riggs, Epidemiology of fracture of the proximal femur in Rochester, MN. Clin. Orthop. Relat. Res. 150, 163 – 171 (1980). 15. J. A. Kanis, The incidence of hip fracture in Europe. Osteoporosis Int. 1, S10 – S15 (1993). 16. J. Chalmers and K. C. Ho, Geographical variations in senile osteoporosis: The association with physical activity. J. Bone Jt. Surg. 52B, 667 – 675 (1970). 17. L. Solomon, Osteoporosis and fracture of the femoral neck in the South African Bantu. J. Bone Jt. Surg. 50B, 2 – 13 (1968). 18. P. C. N. Wong, Fracture epidemiology in a mixed southeastern Asian community (Singapore). Clin. Orthop. 45, 55 – 61 (1966). 19. A. V. Schwartz, J. L. Kelsey, S. Maggi, M. Tuttleman, S. C. Ho, P. V. Jonsson, G. Poor, J. A. Sisson de Castro, L. Xu, C. C. Matkin, L. M. Nelson, and S. P. Heyse, International variation in the incidence of hip fractures: Cross-national project on osteoporosis for the world health organization program for research on aging. Osteoporosis Int. 9, 242 – 253 (1999). 20. L. J. Melton, III and E. A. Chrischilles, Perspective: How many women have osteoporosis? J. Bone Miner. Res. 7, 1005 – 1010 (1992). 21. A. D. Martin, K. G. Silverthorn, C. S. Houston, S. Bernhardson, A. Wajda, and L. L. Roos, The incidence of fracture of the proximal femur in two million Canadians from 1972 to 1984. Clin. Orthop. Rel. Res. 266, 111 – 118 (1991). 22. C. Cooper and G. Campion, Hip fractures in the elderly: a worldwide projection. Osteoporosis Int. 2(6), 285 – 289 (1992). 23. E. L. Schneider and J. M. Guralnik, The aging of America: impact on health care costs. JAMA 263, 2335 – 2340 (1990). 24. T. M. Huusko, P. Karppi, V. Avikainen, H. Kautiainen, and R. Sulkava, The changing picture of hip fractures: Dramatic change in age distribution and no change in age-adjusted incidence within 10 years in central Finland. Bone 24, 257 – 259 (1999). 25. B. Gullberg, O. Johnell, and J. A. Kanis, World-wide projections for hip fracture. Osteoporosis Int. 7, 407 – 413 (1997). 26. L. J. Melton, III, W. M. O’Fallon, and B. L. Riggs, Clinical investigations: Secular trends in the incidence of hip fractures. Calcif. Tissue Int. 41, 57 – 64 (1987). 27. J. A. Falch, T. S. Kaastad, and G. Bohler, Secular increase and geographical differences in hip fracture incidence in Norway. Bone 14, 634 – 645 (1993). 28. J. Parkkari, P. Kannus, S. Niemi, M. Pasanen, M. Jarvinen, P. Luthje, and I. Vuori, Increasing age-adjusted incidence of hip fractures in Finland: The number and incidence of fractures in 1970 – 1991 and prediction for the future. Calcif. Tissue Int. 55, 342 – 345 (1994). 29. L. J. Melton, C. S. Crowson, and W. M. O’Fallon, Fracture incidence in Olmsted County, Minnesota: Comparison of urban with rural rates and changes in urban rates over time. Osteoporosis Int. 9, 29 – 37 (1999).

ORWOLL AND KLEIN 30. G. Poor, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton, Determinants of reduced survival following hip fractures in men. Clin. Orthop. Relat. Res. 319, 260 – 265 (1995). 31. L. J. Melton, III and B. L. Riggs, Epidemiology of age-related fractures. In “The Osteoporotic Syndrome” (L. V. Avioli, ed.), pp. 45 – 72. Grune & Stratton, New York, 1983. 32. J. R. Center, T. V. Nguyen, D. Schneider, P. N. Sambrook, and J. A. Eisman, Mortality after all mamor types of osteoporotic fracture in men and women: An observational study. Lancet 353, 878 – 882 (1999). 33. E. S. Fisher, J. A. Baron, D. J. Malenka, J. A. Barrett, W. D. Kniffin, F. S. Whaley, and T. A. Bubolz, Hip fracture incidence and mortality in new england. Epidemiology 2, 116 – 122 (1991). 34. I. Diamond and A. S. Gordon, Cellular and molecular neuroscience of alcoholism. Physiol. Rev. 77, 1 – 20 (1997). 35. S. P. Heyse, Epidemiology of hip fractures in the elderly: A crossnational analysis of mortality rates for femoral neck fractures. Osteoporosis Int. 1, S16 – S19 (1993). 36. J. A. Baron, M. Karagas, J. Barrett, W. Kniffin, D. Malenka, M. Mayor, and R. B. Keller, Basic Epidemiology of fractures of the upper and lower limb among Americans over 65 years of age. Epidemiology 7, 612 – 618 (1996). 37. P. D. Ross, H. Norimatsu, J. W. Davis, K. Yano, R. D. Wasnich, S. Fujiwara, Y. Hosoda, and L. J. I. Melton, A comparison of hip fracture incidence among native Japanese, Japanese Americans, and American Caucasians. Am. J. Epidemiol. 133, 801 – 809 (1991). 38. R. Madhok, L. J. Melton, E. J. Atkinsom, W. M. O’Fallon, and D. G. Lewallen, Urban vs. rural increase in hip fracture incidence. Acta. Orthop. Scand. 64, 543 – 548 (1993). 39. I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Lopes Vaz, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. P. Cano, A. Rapado, and C. Ribot, The variable incidence of hip fracture in Southern Europe: The MEDOS study. Osteoporosis Int. 4, 253 – 263 (1994). 40. H. Burger, P. L. A. Van Daele, K. Grashuis, A. Hofman, D. E. Grobbee, H. E. Schutte, J. C. Birkenhager, and H. A. P. Pols, Vertebral deformities and functional impairment in men and women. J. Bone Miner. Res. 12, 152 – 157 (1997). 41. A. C. Scane, A. M. Sutcliffe, and R. M. Francis, The sequelae of vertebral crush fractures in men. Osteoporosis Int. 4, 89 – 92 (1994). 42. T. Mann, S. K. Oviatt, D. Wilson, D. Nelson, and E. S. Orwoll, Vertebral deformity in men. J. Bone Miner. Res. 7, 1259 – 1265 (1992). 43. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton, Incidence of clinically diagnosed vertebral fractures: A population-based study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 44. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 45. T. W. O’Neill, D. Felsenberg, J. Varlow, C. Cooper, J. A. Kanis, and A. J. Silman, The prevalence of vertebral deformity in european men and women: The eurpoean vertebral osteoporosis study. J. Bone Miner. Res. 11, 1010 – 1018 (1996). 46. K. M. Davies, M. R. Stegman, R. P. Heaney, and R. R. Recker, Prevalence and severity of vertebral fracture: The saunders county bone quality study. Osteoporosis Int. 6, 160 – 165 (1996). 47. A. J. Silman, T. W. O’Neill, C. Cooper, J. Kanis, and D. Felsenberg, Influence of physical activity on vertebral deformity in men and women: Results from the European vertebral osteoporosis study. J. Bone Miner. Res. 12, 813 – 819 (1997). 48. M. Lunt, D. Felsenberg, J. Reeve, L. Benevolenskaya, J. Cannata, J. Dequeker, C. Dodenhof, J. A. Falch, P. Masaryk, H. A. P. Pols, G. Poor, D. M. Reid, C. Scheidt-Nave, K. Weber, J. Varlow, J. A. Kanis, T. W. O’Neill, and A. J. Silman, Bone density variation and its effects on risk of vertebral deformity in men and women studied in thirteen european centers: The EVOS study. J. Bone Miner. Res. 12, 1883 – 1894 (1997).

CHAPTER 42 Osteoporosis in Men 49. W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. I. Frequency and distribution. Mayo Clin. Proc. 54, 701 – 707 (1979). 50. A. J. Buhr and A. M. Cooke, Fracture patterns. Lancet 1, 531 – 536 (1959). 51. D. Hemenway, D. R. Azrael, E. B. Rimm, D. Feskanich, and W. C. Willett, Risk factors for wrist fracture: Effect of age, cigarettes, alcohol, body height, relative weight and handedness on the risk for distal forearm fractures in men. Am. J. Epidemiol. 140, 361 – 367 (1994). 52. M. T. Cuddihy, S. E. Gabriel, C. S. Crowson, W. M. O’Fallon, and L. J. Melton, Forearm fractures as predictors of subsequent osteoporotic fractures. Osteoporosis Int. 9, 469 – 475 (1999). 53. H. Mallmin, S. Ljunghall, I. Persson, T. Naessen, U. B. Krusemo, and R. Bergstrom, Fracture of the distal forearm as a forecaster of subsequent hip fracture: A population-based cohort study with 24 years follow-up. Calcif. Tissue Int. 52, 269 – 272 (1993). 54. R. M. Francis, M. Peacock, D. H. Marshall, A. Horsman, and J. E. Aaron, Spinal osteoporosis in men. Bone Miner. 5, 347 – 357 (1989). 55. H. K. Genant, G. S. Gordan, and P. G. J. Hoffman, Osteoporosis Part I. Advanced radiologic assessment using quantitative computed tomography — Medical Staff Conference, University of California, San Francisco. West J. Med. 139, 75 – 84 (1983). 56. E. S. Orwoll, S. K. Oviatt, M. R. McClung, L. J. Deftos, and G. Sexton, The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann. Int. Med. 112, 29 – 34 (1990). 57. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton, Changes in bone mineral density of the proximal femur and spine with aging. J. Clin. Invest. 70, 716 – 723 (1982). 58. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture spinal osteoporosis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1988). 59. R. C. Hamdy, S. Moore, V. Cancellaro, K. Whalen, C. Landy, D. Chi, S. Reynolds, and G. Cambron, “Osteoporosis in Men — Prevalence and Biochemical Parameters.” 1992 Program and Abstracts of the 14th Annual Meeting of the American Society for Bone and Mineral Research, Liebert, Mary Ann, Inc., Minneapolis, MN, 1992. 60. P. D. Ross, A. Lombardi, and D. Freedholm, The assessment of bone mass in men. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 505 – 525. Academic Press, San Diego, 1999. 61. E. Vega, C. Mautalen, H. Gomez, A. Garrido, L. Melo, and A. O. Sahores, Bone mineral density in patients with cervical and trochanteric fractures of the proximal femur. Osteoporosis Int. 1, 81 – 86 (1991). 62. M. E. Mussolino, A. C. Looker, J. H. Madans, J. A. Langlois, and E. S. Orwoll, Risk factors for hip fracture in white men: The NHANES I epidemiologic follow-up study. J. Bone Miner. Res. 13, 918 – 924 (1998). 63. C. E. D. H. De Laet, B. A. van Hout, H. Burger, A. Hofman, and H. A. P. Pols, Bone density and risk of hip fracture in men and women: Cross sectional analysis. Br. Med. J. 315, 221 – 225 (1997). 64. T. V. Nguyen, J. A. Eisman, P. J. Kelly, and P. N. Sambrook, Risk factors for osteoporotic fractures in elderly men. Am. J. Epidemiol. 144, 258 – 261 (1996). 65. H. Kroger, M. Lunt, J. Reeve, J. Dequeker, J. E. Adams, J. C. Birkenhager, M. Diaz Curiel, D. Felsenberg, L. Hyldstrup, P. Kotzki, A.-M. Laval-Jeantet, P. Lips, O. Louis, R. Perez Cano, C. Reiners, C. Ribot, P. Ruegsegger, P. Schneider, P. Braillon, and J. Pearson, Bone density reduction in various measurement sites in men and women with osteoporotic fractures of spine and hip: The european quantitation of osteoporosis study. Calcif. Tissue Int. 64, 191 – 199 (1999).

137 66. C. E. D. H. De Laet, B. A. Van Hout, H. Burger, A. E. A. M. Weel, A. Hofman, and H. A. P. Pols, Hip fracture prediction in elderly men and women: Validation of the Rotterdam Study. J. Bone Mineral. Res. 13, 1587 – 1593 (1998). 67. L. J. Melton, E. J. Atkinson, M. K. O’Connor, W. M. O’Fallon, and B. L. Riggs, Bone density and fracture risk in men. J. Bone Miner. Res. 13, 1915 – 1923 (1998). 68. T. Nguyen, P. Sambrook, P. Kelly, G. Jones, S. Lord, and J. Freund. Prediction of osteoporotic fractures by postural instability and bone density. Br. Med. J. 307, 1111 – 1115 (1993). 69. C. M. Klotzbuecher, P. D. Ross, P. B. Landsman, T. A. Abbott, and M. Berger, Patients with prior fractures have an increased risk of future fractures: A summary of the literature and statistical synthesis. J. Bone Miner. Res. 15, 721 – 739 (2000). 70. S. Fujiwara, F. Kasagi, M. Yamada, and K. Kodama, Risk factors for hip fracture in a Japanese cohort. J. Bone Miner. Res. 12, 998 – 1004 (1997). 71. L. J. Melton, E. J. Atkinson, C. Cooper, W. M. O’Fallon, and B. L. Riggs, Vertebral fractures predict subsequent fractures. Osteoporosis Int. 10, 214 – 221 (1999). 72. A. M. Tromp, J. H. Smit, D. J. H. Deeg, L. M. Bouter, and P. Lips, Predictors for falls and fractures in the Longitudinal Aging Study Amsterdam. J. Bone Miner. Res. 13, 1932 – 1939 (1998). 73. W. C. Graafmans, M. E. Ooms, H. M. A. Hofstee, P. D. Bezemer, L. M. Bouter, and P. Lips, Falls in the elderly: A prospective study of risk factors and risk profiles. Am. J. Epidemiol. 143, 1129 – 1136 (1996). 74. K. Aoyagi, P. D. Ross, J. W. Davis, R. D. Wasnich, T. Hayashi, and T. Takemoto, Falls amont community-dwelling elderly in japan. J. Bone Miner. Res. 13, 1468 – 1474 (1998). 75. P. Kannus, J. Parkkari, S. Koskinen, S. Niemi, M. Palvanen, M. Jarvinen, and I. Vuori, Fall-induced injuries and deaths among older adults. JAMA 281, 1895 – 1899 (1999). 76. J. A. Grisso, G. Y. Chiu, G. Maislin, W. C. Steinmann, and J. Portale, Risk factors for hip fractures in men: a preliminary study. J. Bone Miner. Res. 6, 865 – 868 (1991). 77. H. E. Meyer, A. Tverdal, and J. A. Falch, Risk factors for hip fracture in middle-aged Norwegian women and men. Am. J. Epidemiol. 137, 1203 – 1211 (1993). 78. A. V. Schwartz, J. L. Kelsey, S. Sidney, and J. A. Grisso, Characteristics of falls and risk of hip fracture in elderly men. Osteoporosis Int. 8, 240 – 246 (1998). 79. W. C. Hayes, E. R. Myers, J. N. Morris, T. N. Gerhardt, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 80. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factor for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 81. D. Hemenway, D. R. Azrael, and D. Feskanich, “The bigger they are . . . ”: The effect of height, relative weight, alcohol intake and cigarette smoking on the risk of hip fracture in middle-aged men. Am. J. Public Health (1994). 82. W. A. Ray, M. R. Griffin, W. Schaffner, D. K. Baugh, and L. J. Melton III, Psychotropic drug use and the risk of hip fracture. N. Engl. J. Med. 316, 363 – 370 (1987). 83. W. A. Ray, M. R. Griffin, and W. Downey, Benzodiazepines of long and short elimination half-life and the risk of hip fracture. JAMA 262, 3303 – 3307 (1989). 84. M. K. Karlsson, O. Johnell, B. E. Nilsson, I. Sernbo, and K. J. Obrant, Bone mineral mass in hip fracture patients. Bone 14, 161 – 165 (1993). 85. T. W. O’Neill, J. Varlow, A. J. Silman, J. Reeve, D. M. Reid, C. Todd, and A. D. Woolf, Age and sex influences on fall characteristics. Ann. Rheum. Dis. 53, 773 – 775 (1994).

138 86. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone in Nutritional Perspective.” Charles C. Thomas, Springfield, IL, 1970. 87. V. Gilsanz, Accumulation of bone mass during childhood and adolescence. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 65 – 85. Academic Press, San Diego, 1999. 88. E. Seeman, Bone size, mass, and volumetric density: The importance of structure on skeletal health. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 87 – 109. Academic Press, San Diego, 1999. 89. A. Vesterby, L. Mosekilde, and H. J. G. Gundersen, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991). 90. P. Brinckmann, M. Biggemann, and D. Hilweg, Prediction of the compressive strength of human lumbar vertebrae. Spine 14, 606 – 610 (1989). 91. S. D. Rockoff, E. Sweet, and J. Bleustein, The relative contribution of trabecular and cortical bone to the strength of human lumbar vertebrae. Calcif. Tissue Res. 3, 163 – 175 (1969). 92. V. Gilsanz, E. E. Shultz, and L. Loro, Smaller cross-sectional area of intact vertebral bodies in women with vertebral fractures. J. Bone Miner. Res. 8, S326 (1993). 93. L. Mosekilde and L. Mosekilde, Normal vertebral body size and compressive strength relations to age and to vertebral and iliac trabecular bone compression strength. Bone 7, 207 – 212 (1986). 94. V. Matkovic, T. Jelic, and G. M. Wardlaw, Timing of peak bone mass in caucasian females and its implication for the prevention of osteoporosis: Inference from a cross-sectional model. J. Clin. Invest. 93 (1994). 95. L. Mosekilde and L. Mosekilde, Sex differences in age-related changes in vertebral body size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 96. R. W. J. Smith and R. R. Walker, Femoral expansion in aging women: implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 97. T. J. Beck, C. B. Ruff, W. W. J. Scott, C. C. Plato, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 98. T. J. Beck, C. B. Ruff, and K. Bissessur, Age-related changes in female femoral neck geometry: Implications for bone strength. Calcif. Tissue Int. 53, S41 – S46 (1993). 99. B. Martin, Aging and strength of bone as a structural material. Calcif. Tissue Int. 53 (Suppl 1), S34 – S40 (1993). 100. H. Brockstedt, M. Kassem, and E. F. Eriksen, Age- and sex-related changes in iliac cortical bone mass and remodeling. Bone 14, 681 – 691 (1993). 101. S. M. Garn, T. V. Sullivan, S. A. Decker, F. A. Larkin, and V. M. Hawthorne, Continuing bone expansion and increasing bone loss over a two-decade period in men and women from a total community sample. Am. J. Hum. Biol. 4, 57 – 67 (1992). 102. C. B. Ruff and W. C. Hayes, Sex differences in age-related remodeling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). 103. R. B. Martin and P. J. Atkinson, Age- and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 104. G. Jones, T. Nguyen, P. Sambrook, P. J. Kelly, and J. A. Eisman, Progressive loss of bone in the femoral neck in elderly people: Longitudinal findings from the Dubbo osteoporosis epidemiology study. Br. Med. J. 309, 691 – 695 (1994). 105. H. Burger, C. E. D. H. de Laet, P. L. A. van Daele, A. E. A. M. Weel, J. C. M. Wittman, A. Hofman, and H. A. P. Pols, Risk factors for increased bone loss in an elderly population. Am. J. Epidemiol. 147, 871 – 879 (1998).

ORWOLL AND KLEIN 106. H. Burger, P. L. A. van Daele, D. Algra, F. A. van den Ouweland, D. E. Grobbee, A. Hofman, C. van Kuijk, H. E. Schutte, J. C. Birkenhager, and H. A. P. Pols, The association between age and bone mineral density in men and women aged 55 years and over: The Rotterdam Study. Bone Miner. 25, 1 – 13 (1994). 107. J. E. Compston, R. W. E. Mellish, P. Croucher, R. Newcombe, and N. J. Garrahan, Structural mechanism of trabecular bone loss in men. Bone 8, 339 – 350 (1989). 108. D. Prudham and J. G. Evans, Factors associated with falls in the elderly: A community study. Age Ageing 10, 141 – 146 (1981). 109. J.-P. Bonjour, G. Theintz, F. Law, D. Slosman, and R. Rizzoli, Peak bone mass. Osteoporosis Int. 1, S7 – S13 (1994). 110. J.-P. Bonjour, G. Theintz, B. Buchs, D. Slosman, and R. Rizzoli, Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991). 111. V. Gilsanz, M. I. Beochat, and T. Roe, Gender differences in vertebral dimensions in children and adolescents. Radiology 190, 673 – 677 (1993). 112. H. Rico, M. Revilla, E. R. Hernandez, L. F. Villa, and M. Alvarez del Buergo, Sex differences in the acquisition of total bone mineral mass peak assessed through dual-energy x-ray absorptiometry. Calcif. Tissue Int. 51, 251 – 254 (1992). 113. A. Gotfredsen, A. Hadberg, L. Nilas, and C. Christiansen, Total body bone mineral in healthy adults. J. Lab. Clin. Med. 110, 362 – 368 (1987). 114. C. M. Weaver, M. Peacock, and C. C. Johnston, Adolescent nutrition in the prevention of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 84, 1839 – 1843 (1999). 115. G. Jones and T. Dwyer, Bone mass in prepubertal children: Gender differences and the role of physical activity and sunlight exposure. J. Clin. Endorinol. Metab. 83, 4274 – 4279 (1998). 116. M. J. Valimaki, M. Karkkainen, C. Lambergallardt, K. Laitinen, E. Alhava, J. Heikkinen, O. Impivaara, P. Makela, J. Palmgren, R. Seppanen, I. Vuori, H. K. Akerblom, M. Uhari, J. Viikari, M. Dahl, E. Pesonen, M. Pietikainen, M. Salo, A. Aromaa, L. Kannas, L. Keltikangasjarvinen, V. Kuusela, K. Pyorala, T. Ronnemaa, L. Rasanen, S. Sama, A. Seppanen, and R. Telama, Exercise, smoking, and calcium intake during adolescence and early adulthood as determinants of peak bone mass. Br. Med. J. 309, 230 – 235 (1994). 117. A. M. Fehily, R. J. Coles, W. D. Evans, and P. C. Elwood, Factors affecting bone density in young adults. Am. J. Clin. Nutr. 56, 579 – 586 (1992). 118. D. C. Welten, H. C. G. Kemper, G. B. Post, W. Van Mechelen, J. Twisk, P. Lips, and G. J. Teule, Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 9, 1089 – 1096 (1994). 119. C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). 120. E. A. Krall and B. Dawson-Hughes, Heritable and life-style determinants of bone mineral density. J. Bone Miner. Res. 8, 1 – 9 (1993). 121. M. N. Diaz, T. W. O’Neill, and A. J. Silman, The influence of family history of hip fracture on the risk of vertebral deformity in men and women: The european vertebral osteoporosis study. Bone 20, 145 – 149 (1997). 122. M. E. Cohen-Solal, C. Baudoin, M. Omouri, D. Kuntz, and M. C. de Vernejoul, Bone mass in middle-aged osteoporotic men and their relatives: Familial effect. J. Bone Miner. Res. 13, 1909 – 1914 (1998). 123. R. A. Evans, G. M. Marel, E. K. Lancaster, S. Kos, M. Evans, and S. Y. P. Wong, Bone mass is low in relatives of osteoporotic patients. Ann. Intern. Med. 109, 870 – 873 (1988).

CHAPTER 42 Osteoporosis in Men 124. S. B. Soroko, E. Barrett-Connor, S. L. Edelstein, and D. KritzSilverstein, Family history of osteoporosis and bone mineral density at the axial skeleton: The Rancho Bernardo study. J. Bone Miner. Res. 9, 761 – 769 (1994). 125. M. T. Hannan, D. T. Felson, and J. J. Anderson, Bone mineral density in elderly men and women: Results from the Framingham osteoporosis study. J. Bone Miner. Res. 7, 547 – 553 (1992). 126. R. B. Mazess, H. S. Barden, P. J. Drinka, S. F. Bauwens, E. S. Orwoll, and N. H. Bell, Influence of age and body weight on spine and femur bone mineral density in U.S. white men. J. Bone Miner. Res. 5, 645 – 652 (1990). 127. J. W. Davis, P. D. Ross, J. M. Vogel, and R. D. Wasnich, Age-related changes in bone mass among japanese-american men. Bone Miner. 15, 227 – 236 (1991). 128. J. D. Tobin, K. M. Fox, and M. L. Cejku, Bone density changes in normal men: A 4 – 19 year longitudinal study. J. Bone Miner. Res. 8, 102 (1993). 129. B. A. Blunt, M. R. Klauber, E. L. Barrett-Connor, and S. L. Edelstein, Sex differences in bone mineral density in 1653 men and women in the sixth through tenth decades of life: The Rancho Bernardo Study. J. Bone Miner. Res. 9, 1333 – 1338 (1994). 130. C. W. Slemenda, J. C. Christian, T. Reed, T. K. Reister, C. J. Williams, and C. C. J. Johnston, Long-term bone loss in men: Effects of genetic and environmental factors. Ann. Intern. Med. 117, 286 – 291 (1992). 131. B. L. Riggs, H. W. Wahner, W. L. Dunn, R. B. Mazess, K. P. Offord, and L. J. I. Melton, Differential changes in bone mineral density of the appendicular and axial skeleton with aging. J. Clin. Invest. 67, 328 – 335 (1981). 132. D. E. Meier, E. S. Orwoll, and J. M. Jones, Marked disparity between trabecular and cortical bone loss with age in healthy men: Measurement by vertebral computed tomography and radial photon absorptiometry. Ann. Intern. Med. 101, 605 – 612 (1984). 133. J. R. Elliott, N. L. Gilchrist, J. E. Wells, J. G. Turner, E. Ayling, W. J. Gillespie, R. Sainsbury, A. Hornblow, and R. A. Donald, Effects of age and sex on bone density at the hip and spine in a normal Caucasian New Zealand population. NZ Med. J. 103, 33 – 37 (1990). 134. T. Yoshikawa, C. H. Turner, M. Peacock, C. W. Slemenda, C. M. Weaver, D. Teegarden, P. Markwardt, and D. B. Burr, Geometric structure of the femoral neck measured using dual-energy X-ray absorptiometry. J. Bone Miner. Res. 9, 1053 – 1064 (1994). 135. P. L. A. Van Daele, H. Burger, D. Algra, A. Hofman, D. E. Grobbee, J. C. Birkenhager, and H. A. P. Pols, Age-associated changes in ultrasound measurements of the calcaneus in men and women: The Rotterdam Study. J. Bone Miner. Res. 9, 1751 – 1757 (1994). 136. A.-M. Laval-Jeantet, C. Bergot, R. Carroll, and F. Garcia-Schaefer, Cortical bone senescence and mineral bone density of the humerus. Calcif. Tissue Int. 35, 268 – 272 (1983). 137. R. B. Martin, Aging and changes in cortical mass and structure. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 111 – 128. Academic Press, San Diego, 1999. 138. L. Mosekilde, A. Viidik, and L. Mosekilde, Correlation between the compressive strength of iliac and vertebral trabecular bone in normal individuals. Bone 6, 291 – 295 (1985). 139. A. M. Parfitt and H. E. Mathews, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. J. Clin. Invest. 72, 1396 – 1409 (1983). 140. L. Mosekilde, Sex differences in age-related loss of vertebral trabecular bone mass and structure — Biomechanical consequences. Bone 10, 425 – 432 (1989). 141. M. Singh, B. L. Riggs, J. W. Beabout, and J. Jowsey, Femoral trabecular-pattern index for evaluation of spinal osteoporosis. Ann. Int. Med. 77, 63 – 67 (1972). 142. H. M. Frost, Suggested fundamental concepts in skeletal physiology. Calcif. Tissue Int. 52, 1 – 4 (1993).

139 143. J. E. Aaron, N. B. Makins, and K. Sagreiya, The microanatomy of trabecular bone loss in normal aging men and women. Clin. Orthop. Rel. Res. 215, 260 – 271 (1987). 144. J. E. Compston, R. W. E. Mellish, and N. J. Garrahan, Age-related changes in iliac crest trabecular microanatomic bone structure in man. Bone 8, 289 – 292 (1987). 145. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr. Rev. 7, 379 – 408 (1986). 146. E. F. Eriksen, L. Mosekilde, and F. Melsen, Trabecular bone resorption depth decreases with age: Differences between normal males and females. Bone 6, 141 – 146 (1985). 147. P. I. Croucher, R. W. E. Mellish, S. Vedi, N. J. Garrahan, and J. E. Compston, The relationship between resorption depth and mean interstitial bone thickness: Age-related changes in man. Calcif. Tissue Int. 45, 15 – 19 (1989). 148. F. Melsen, B. Melsen, L. Mosekilde, and S. Bergmann, Histomorphometric analysis of normal bone from the iliac crest. Acta. Pathol. Microbiol. Scand. 86, 70 – 81 (1986). 149. J. Kragstrup, F. Melsen, and L. Mosekilde, Thickness of bone formed at remodeling sites in normal human iliac trabecular bone: Variations with age and sex. Metab. Bone Dis. Relat. Res. 5, 17 – 21 (1983). 150. E. S. Orwoll, N. H. Bell, M. S. Nanes, K. A. Flessland, M. B. Pettinger, J. S. Mallinak, and D. F. Cain, Collagen N-telopeptide excretion in men: The effects of age and intra-subject variability. J. Clin. Endocrinol. Metab. 83, 3930 – 3935 (1998). 151. E. S. Orwoll and L. J. Deftos, Serum osteocalcin (BGP) levels in normal men: A longitudinal evaluation reveals an age-associated increase. J. Bone Miner. Res. 5(3), 259 – 262 (1990). 152. P. D. Delmas, Clinical use of biochemical markers of bone remodeling in osteoporosis. Bone 13, S17 – S21 (1992). 153. J. Pfeilschifter and R. Ziegler, Relationship between IGF-I and skeletal aging. Eur. J. Endocrinol. 138, 617 – 618 (1998). 154. D. Fatayerji and R. Eastell, Age-related changes in bone turnover in men. J. Bone Miner. Res. 14, 1203 – 1210 (1999). 155. C. J. Rosen, Insulin-like growth factors and bone: Implications for the pathogenesis and treatment of osteoporosis. In “Osteoporosis in men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 157 – 178. Academic Press, San Diego, 1999. 156. V. Nicolas, A. Prewett, P. Bettica, S. Mohan, R. D. Finkelman, D. J. Baylink, and J. R. Farley, Age-related decreases in insulin-like growth factor-I and transforming growth factor-B in femoral cortical bone from both men and women: Implications for bone loss with aging. J. Clin. Endocrinol. Metab. 78, 1011 – 1016 (1994). 157. S. Ljunghall, A. G. Johansson, P. Burman, O. Kampe, E. Lindh, and F. A. Karlsson, Low plasma levels of insulin-like growth factor 1 (IGF-1) in male patients with idiopathic osteoporosis. J. Intern. Med. 232, 59 – 64 (1992). 158. A. Johansson, A. Forslund, and L. Hambraeus, Growth-hormonedependent insulin-like growth factor binding protein is a major determinant of bone mineral density in healthy men. J. Bone Miner. Res. 9, 915 – 921 (1994). 159. J. A. Langlois, C. J. Rosen, M. Visser, M. T. Hannan, T. Harris, P. W. Wilson, and D. P. Kiel, Association between insulin-like growth factor I and bone mineral density in older women and men: The Framingham Heart Study. J. Clin. Endocrinol. Metab. 83, 4257 – 4262 (1998). 160. J. Pfeilschifter, C. Scheidt-Nave, G. Leidig-Bruckner, H. W. Woitge, W. F. Blum, C. Wuster, D. Haack, and R. Ziegler, Relationship between circulating insulin like growth factor components and sex hormones in a population-based sample of 50- to 80-year-old men and women. J. Clin. Endocrinol. Metab. 81, 2534 – 2540 (1996). 161. E. Barrett-Connor and D. Goodman-Gruen, Gender differences in insulin-like growth factor and bone mineral density association in

140

162.

163.

164.

165.

166. 167. 168. 169.

170.

171.

172.

173.

174. 175.

176.

177.

178.

179.

180.

ORWOLL AND KLEIN old age: The Rancho Bernardo Study. J. Bone Miner. Res. 13, 1343 – 1349 (1998). T. S. Kurland, C. J. Rosen, F. Cosman, D. McMahon, F. Chan, E. Shane, R. Lindsay, D. Dempster, and J. P. Bilezikian, Insulin-like growth factor-I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 82, 2799 – 2805 (1997). J. A. M. J. L. Janssen, H. Burger, R. P. Stolk, D. e. Grobbee, F. H. de Jong, S. W. J. Lanberts, and H. A. P. Pols, Gender-specific relationship between serum free and total IGF-I and bone mineral density in elderly men and women. Eur. J. Endocrinol. 138, 627 – 632 (1998). P. Burman, A. G. Johansson, A. Siegbahn, B. Vessby, and A. Karlsson, Growth hormone (GH)-deficient men are more responsive to GH replacement therapy than women. J. Clin. Endocrinol. Metab. 82, 550 – 555 (1997). A. G. Johansson, B. E. Engstrom, S. Jjunghall, F. A. Karlsson, and P. Burman, Gender differences in the effects of long term growth hormone (GH) treatment in bone in adults with GH deficiency. J. Clin. Endocrinol. Metab. 84, 2002 – 2007 (1999). B. L. Riggs and L. J. Melton, Medical Progress: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). M. I. Irwin and E. W. Kienholz, A conspectus of research on calcium requirements of man. J. Nutr. 103, 1019 – 1095 (1973). J. E. Morley, Nutritional status of the elderly. Am. J. Med. 81, 679 – 695 (1986). G. Young, R. Marcus, J. R. Minkoff, L. Y. Kim, and G. V. Segre, Age-related rise in parathyroid hormone in man: The use of intact and midmolecule antisera to distinguish hormone secretion from retention. J. Bone Miner. Res. 2, 367 – 374 (1987). D. B. Endres, C. H. Morgan, P. J. Garry, and J. L. Omdahl, Agerelated changes in serum immunoreactive parathyroid hormone and its biological action in healthy men and women. J. Clin. Endocrinol. Metab. 65, 724 – 731 (1987). E. S. Orwoll and D. E. Meier, Alterations in calcium, vitamin D, and parathyroid hormone physiology in normal men with aging: Relationship to the development of senile osteopenia. J. Clin. Endocrinol. Metab. 63, 1262 – 1269 (1986). D. M. Slovik, J. S. Adams, R. M. Neer, M. F. Holick, and J. R. Potts, Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 (1981). S. C. Manolagas, F. L. Culler, and J. E. Howard, The cytoreceptor assay for 1,25-dihydroxyvitamin D in elderly osteoporotic patients. J. Clin. Endocrinol. Metab. 56, 751 – 760 (1983). S. Epstein, G. Bryce, and J. W. Hinman, The influence of age of bone mineral regulating hormones. Bone 7, 421 – 425 (1986). B. P. Halloran, A. A. Portale, E. T. Lonergan, and R. C. J. Morris, Production and metabolic clearance of 1,25-dihydroxyvitamin D in men: Effect of advancing age. J. Clin. Endocrinol. Metab. 70, 318 – 323 (1990). S. S. Sherman, J. D. Tobin, B. W. Hollis, C. M. Gundberg, T. A. Roy, and C. C. Plato, Biochemical parameters associated with low bone density in healthy men and women. J. Bone Miner. Res. 7, 1123 – 1130 (1992). H. Kroger and K. Laitinen, Bone mineral density measured by dualenergy x-ray absorptiometry in normal men. Eur. J. Clin. Invest. 22, 454 – 460 (1992). B. P. Halloran and D. D. Bikle, Age-related changes in mineral metabolism. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 179 – 195. Academic Press, San Diego, 1999. P. J. Kelly, N. A. Pocock, P. N. Sambrook, and J. A. Eisman, Dietary calcium, sex hormones, and bone mineral density in men. Br. Med. J. 300, 1361 – 1364 (1990). V. Matkovic, K. Kostial, I. Simonovic, R. Buzina, A. Brodarec, and B. E. C. Nordin, Bone status and fracture rates in two regions of Yugoslavia. Am. J. Clin. Nutr. 32, 540 – 549 (1979).

181. E. Lau, S. Donnan, D. J. P. Barker, and C. Cooper, Physical activity and calcium intake in fracture of the proximal femur in Hong Kong. Br. Med. J. 297, 1441 – 1443 (1988). 182. C. Cooper, D. J. Barker, and C. Wickham, Physical activity, muscle strength, and calcium intake in fracture of the proximal femur in Britain. Br. Med. J. 297(6661), 1443 – 1446 (1988). 183. A. C. Looker, T. B. Harris, J. H. Madans, and C. T. Sempos, Dietary calcium and hip fracture risk: The NHANES I epidemiologic followup study. Osteoporosis. Int. 3, 177 – 184 (1993). 184. T. L. Holbrook, E. Barrett-Connor, and D. L. Wingard, Dietary calcium and risk of hip fracture: 14-Year prospective population study. Lancet 2, 1046 – 1049 (1988). 185. C. A. C. Wickham, K. Walsh, C. Cooper, D. J. P. Barker, B. M. Margetts, J. Morris, and S. A. Bruce, Dietary calcium, physical activity, and risk of hip fracture: a prospective study. Br. Med. J. 299, 889 – 892 (1989). 186. A. Paganini-Hill, A. Chao, R. K. Ross, and B. E. Henderson, Exercise and other factors in the prevention of hip fracture: The Leisure World study. Epidemiology 2, 16 – 25 (1991). 187. B. Beck and R. Marcus, Skeletal effects of exercise in men. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 129 – 155. Academic Press, San Diego, 1999. 188. C. Snow-Harter, R. Whalen, K. Myburgh, S. Arnaud, and R. Marcus, Bone mineral density, muscle strength, and recreational exercise in men. J. Bone Miner. Res. 7, 1291 – 1296 (1992). 189. L. A. Colletti, J. Edwards, L. Bordon, J. Shary, and N. H. Bell, The effects of muscle-building exercise on bone mineral density of the radius, spine and hip in young men. Calcif. Tissue Int. 45, 12 – 14 (1989). 190. J. E. Block, H. K. Genant, and D. Black, Greater vertebral bone mineral mass in exercising young men. West. J. Med. 145, 39 – 42 (1986). 191. J. E. Block, A. L. Friedlander, G. A. Brooks, P. Steiger, H. A. Stubbs, and H. K. Genant, Determinants of bone density among athletes engaged in weight-bearing and non-weight-bearing activity. J. Appl. Physiol. 67, 1100 – 1105 (1989). 192. H. H. Jones, J. D. Priest, W. C. Hayes, C. C. Tichenor, and D. A. Nagel, Humeral hypertrophy in response to exercise. J. Bone Jt. Surg. 59-A, 204 – 208 (1977). 193. E. S. Orwoll, J. Ferar, S. K. Oviatt, M. R. McClung, and K. Huntington, The relationship of swimming exercise to bone mass in men and women. Arch. Int. Med. 149, 2197 – 2200 (1989). 194. R. C. Hamdy, J. S. Anderson, K. E. Whalen, and L. M. Harvill, Regional differences in bone density of young men involved in different exercises. Med. Sci. Sports. Exercise 26, 884 – 888 (1994). 195. K. H. Myburgh, L.-J. Zhou, C. R. Steele, S. Arnaud, and R. Marcus, In vivo assessment of forearm bone mass and ulnar bending stiffness in healthy men. J. Bone Miner. Res. 7, 1345 – 1350 (1992). 196. J. A. Williams, J. Wagner, R. Wasnich, and L. Heilbrun, The effect of long-distance running upon appendicular bone mineral content. Med. Sci. Sports. Exercise 16, 223 – 227 (1984). 197. I. R. Reid, L. D. Plank, and M. C. Evans, Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J. Clin. Endocrinol. Metab. 75, 779 – 782 (1992). 198. M. T. Hannan, D. T. Felson, B. Dawson-Hughes, K. L. Tucker, L. A. Cupples, P. W. F. Wilson, and D. P. Kiel, Risk factors for longitudinal bone loss in elderly men and women: The framingham osteoporosis study. J. Bone Miner. Res. 15, 710 – 720 (2000). 199. E. Dennison, P. Hindmarsh, C. Fall, S. Kellingray, D. Barker, D. Phillips, and C. Cooper, Profiles of endogenous circulating cortisol and bone mineral density in healthy elderly men. J. Clin. Endocrinol. Metab. 84, 3058 – 3063 (1999). 200. L. Larsson, B. Sjodin, and J. Karlsson, Histochemical and biochemical changes in human skeletal muscle with age in sedentary males, age 22 – 65 years. Acta Physiol. Scand. 103, 31 – 39 (1978).

141

CHAPTER 42 Osteoporosis in Men 201. A. Aniansson, and E. Gustaffson, Physical training in elderly men with special reference to quadriceps muscle strength and morphology. Clin. Physiol. 1, 89 – 98 (1981). 202. H. M. Frost, The role of changes in mechanical usage set points in the pathogenesis of osteoporosis. J. Bone Miner. Res. 7, 253 – 261 (1992). 203. A. Vermeulen, Clinical review 24: Androgens in the aging male. J. Clin. Endocrinol. Metab. 73, 221 – 224 (1991). 204. A. Vermeulen and J. M. Kaufman, Editorial: Role of the hypothalamo – pituitary function in the hypoandrogenism of healthy aging. J. Clin. Endocrinol. Metab. 75, 704 – 706 (1992). 205. R. M. Francis, The effects of testosterone on osteoporosis in men. Clin. Endocrinol. 50, 411 – 414 (1999). 206. F. H. Anderson, R. M. Francis, P. L. Selby, and C. Cooper, Sex hormones and osteoporosis in men. Calcif. Tissue Int. 62, 185 – 188 (1998). 207. C. W. Bardin, R. S. Swerdloff, and R. J. Santen, Special Article — Androgens: Risk and benefits. J. Clin. Endocrinol. Metab. 73, 4 – 17 (1991). 208. S. Murphy, K.-T. Khaw, A. Cassidy, and J. E. Compston, Sex hormones and bone mineral density in elderly men. Bone Miner. 20, 133 – 140 (1993). 209. D. Rudman, P. J. Drinka, C. R. Wilson, D. E. Mattson, F. Scherman, M. C. Cuisinier, and S. Schultz, Relations of endogenous anabolic hormones and physical activity to bone mineral density and lean body mass in elderly men. Clin. Endocrinol. 40, 653 – 661 (1994). 210. S. Boonen, X.-G. Cheng, J. Nijs, P. H. F. Nicholson, G. Verbeke, E. Lasaffre, J. Aerssens, and J. Dequeker, Factors associated with cortical and trabecular bone loss as quantified by peripheral computed tomography (pQCT) at the ultradistal radius in aging women. Calcif. Tissue Int. 1997, 164 – 170 (1997). 211. A. M. Kenny, J. C. Gallagher, K. M. Prestwood, C. A. Gruman, and L. G. Raisz, Bone density, bone turnover, and hormone levels in men over age 75. J. Gerontol. 53A, M419 – M425 (1998). 212. B. Ongphiphadhanakul, R. Rajatanavin, L. Chailurkit, N. Paiseu, K. Teerarungsikul, R. Sirisriro, S. Komindr, and G. Puavilai, Serum testosterone and its relation to bone mineral density and body composition in normal males. Clin. Endocrinol. 43, 727 – 733 (1995). 213. P. J. Drinka, J. Olson, S. Bauwens, S. Voeks, I. Carlson, and M. Wilson, Lack of association between free testosterone and bone density separate from age in elderly males. Calcif. Tissue Int. 52, 67 – 69 (1993). 214. D. E. Meier, E. S. Orwoll, E. J. Keenan, and R. M. Fagerstrom, Marked decline in trabecular bone mineral content in healthy men with age: Lack of association with sex steroid levels. J. Am. Geriatr. Soc. 35, 189 – 197 (1987). 215. F. Nyquist, P. Gardsell, I. Sernbo, J. O. Jeppsson, and O. Johnell, Assessment of sex hormones and bone mineral density in relation to occurrence of fractures in men: A prospective population-based study. Bone 22, 147 – 151 (1998). 216. A. Rapado, F. Hawkins, L. Sobrinho, M. Diaz-Curiel, A. GalvaoTelles, S. Arver, J. Melo Gomes, N. Mazer, J. Garcia e Costa, C. Horcajada, E. Lopez-Gavilanes, M. Mascarenhas, K. Papaietro, M. B. Lopez Alvarez, M. C. Periera, G. Martinez, I. Valverde, J. J. Garcia, J. J. Carballal, and I. Garcia, Bone mineral density and androgen levels in elderly males. Calcif. Tissue. Int. 65, 417 – 421 (1999). 217. J. S. Tenover, Effects of testosterone supplementation in the aging male. J. Clin. Endocrinol. Metab. 75, 1092 – 1098 (1992). 218. P. J. Snyder, H. Peachey, P. Hannoush, J. A. Berline, L. Loh, J. H. Holmes, A. Dlewati, J. Staley, J. Santanna, S. C. Kapoor, M. F. Attie, J. G. Haddad, and B. L. Strom, Effect of testosterone treatment on bone mineral density in men over 65 years of age. J. Clin. Endocrinol. Metab. 84, 1966 – 1972 (1999). 219. B. L. Clarke, P. R. Ebeling, and J. D. Jones, “Increased Bone Turnover with Aging in Men is Not Due to Testosterone Deficiency.”

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

233. 234.

235.

236.

Endocrine Society 75th Annual Meeting, Program & Abstracts, Endocrine Society, Bethesda, MD, Las Vegas, NV, 1993. [Abstract] M. M. Grumbach and R. J. Auchus, Estrogen: Consequences and implications of human mutations in synthesis and action. J. Clin. Endocrinol. Metab. 84, 4677 – 4694 (1999). P. M. Doran, R. T. Turner, B. L. Riggs, and S. Khosla, Estrogens and bone health. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 275 – 298. Academic Press, San Diego, 1999. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). A. Morishima, M. M. Grumbach, E. R. Simpson, C. Fisher, and K. Qin, Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J. Clin. Endocrinol. Metab. 80, 3689 – 3698 (1995). J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). V. Rochira, M. Faustini-Fustini, A. Balestrieri, and C. Carani, Estrogen replacement therapy in a man with congenital aromatase deficiency: Effects of different doses of transdermal estradiol on bone mineral density and hormonal parameters. J. Clin. Endocrinol. Metab. 85, 1841 – 1845 (2000). B. L. Riggs, S. Khosla, and L. J. Melton, A unitary model for involutional osteoporosis: Estrogen deficiency causes both Type I and Type II osteoporosis in women and contributes to bone loss in men. J. Bone Miner. Res. 13, 763 – 773 (1998). C. W. Slemenda, C. Longscope, L. Zhou, S. Hui, M. Peacock, and C. C. Johnston, Sex steroids and bone mass in older men. Positive associations with serum estrogens and negative associations with androgens. J. Clin. Invest. 100, 1755 – 1759 (1997). S. Khosla, L. J. Melton, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). E. Barrett-Connor, J. E. Mueller, D. G. Von Muhlen, G. A. Laughlin, D. L. Schneider, and D. J. Sartoris, Low levels of estradiol are associated with vertebral fractures in older men, but not women: The Rancho Bernardo Study. J. Clin. Endocrinol. Metab. 85, 219 – 223 (2000). D. Vanderschueren, S. Boonen, and R. Bouillon, Action of androgens versus estrogens in male skeletal homeostasis. Bone 23, 391 – 394 (1998). E. Seeman, L. J. Melton, III, W. M. O’Fallon, and B. L. Riggs, Risk factors for spinal osteoporosis in men. Am. J. Med. 75, 977 – 983 (1983). H. Resch, P. Pietschmann, W. Woloszczuk, E. Krexner, P. Bernecker, and R. Willvonseder, Bone mass and biochemical parameters of bone metabolism in men with spinal osteoporosis. Eur. J. Clin. Invest. 22, 542 – 545 (1992). J. A. Jackson and M. Kleerekoper, Osteoporosis in men: Diagnosis, pathophysiology, and prevention. Medicine 69, 137 – 152 (1990). M. C. DeVernejoul, J. Bielakoff, and M. Herve, Evidence for defective osteoblastic function. A role for alcohol and tobacco consumption in osteoporosis in middle-aged men. Clin. Orthop. 179, 107 – 115 (1983). A. M. Parfitt and H. Duncan, Metabolic bone disease affecting the spine. In “The Spine” (R. Rothman, ed.), 2nd ed., pp. 775 – 905. Saunders, Philadephia, PA, 1982. E. Hills, C. R. Dunstan, S. Y. P. Wong, and R. A. Evans, Bone histology in young adult osteoporosis. J. Clin. Pathol. 42, 391 – 397 (1989).

142 237. P. J. Bordier, L. Miravet, and D. Hioco, Young adult osteoporosis. Clin. Endocrinol. Metab. 2, 277 – 292 (1973). 238. B. E. C. Nordin, J. Aaron, R. Speed, R. M. Francis, and N. Makins, Bone formation and resorption as the determinants of trabecular bone volume in normal and osteoporotic men. Scott. Med. J. 29, 171 – 175 (1984). 239. J. E. Aaron, R. M. Francis, M. Peacock, and N. B. Makins, Contrasting microanatomy of idiopathic and corticosteroid-induced osteoporosis. Clin. Orthop. Rel. Res. 243, 294 – 305 (1989). 240. S. Khosla, E. G. Lufkin, S. F. Hodgson, L. A. Fitzpatrick, and L. J. I. Melton, Epidemiology and clinical features of osteoporosis in young individuals. Bone 15, 551 – 555 (1994). 241. P. J. Marie, M. C. de Vernejoul, D. Donnes, and M. Hott, Decreased DNA synthesis by culture osteoblastic cells in eugonadal osteoporotic men with defective bone formation. J. Clin. Invest. 88, 1167 – 1172 (1991). 242. S. Khosla, Editorial: Idiopathic osteoporosis — Is the osteoblast to blame? J. Clin. Endocrinol. Metab. 82, 2792 – 2794 (1997). 243. A. G. Johansson, E. F. Eriksen, E. Lindh, B. Langdahl, W. F. Blum, A. Lindahl, O. Ljunggren, and S. Ljunghall, Reduced serum levels of the growth hormone-dependent insulin-like growth factor binding protein and a negative bone balance at the level of individual remodeling units in idiopathic osteoporosis in men. J. Clin. Endocrinol. Metab. 82, 2795 – 2798 (1997). 244. C. J. Rosen, E. S. Kurland, D. Vereault, R. A. Adler, P. J. Rackoff, W. Y. Craig, S. Witte, J. Rogers, and J. P. Bilezikian, Association between serum insulin growth factor-I (IGF-I) and a simple sequence repeat in IGF-1 gene: Implications for genetic studies of bone mineral density. J. Clin. Endocrinol. Metab. 83, 2286 – 2290 (1998). 245. S. S. Papiha, L. C. Allcroft, R. M. Kanan, R. M. Francis, and H. K. Datta, Vitamin D binding protein gene in male osteoporosis: Association of plasma DBP and bone mineral density with (TAAA)n-Alu polymorphism in DBP. Calcif. Tissue Int. 65, 262 – 266 (1999). 246. R. M. Francis, F. Harrington, E. Turner, S. S. Papiha, and H. K. Datta, Vitamin D receptor gene polymorphism in men and its effect on bone density and calcium absorption. Clin. Endocrinol. 46, 83 – 86 (1997). 247. S. Ferrari, D. Manen, J.-P. Bonjour, D. Slosman, and R. Rizzoli, Bone mineral mass and calcium and phosphate metabolism in young men: Relationships with vitamin D receptor allelic polymorphisms. J. Clin. Endocrinol. Metab. 84, 2043 – 2048 (1999). 248. I. Braidman, C. Baris, L. Wood, P. Selby, J. Adams, A. Freemont, and J. Hoyland, Preliminary evidence for impaired receptor-a protein expression in osteoblasts and osteoclasts from men with idiopathic osteoporosis. Bone 26, 423 – 427 (2000). 249. L. R. Cardon, C. Garner, S. T. Bennett, I. J. Mackay, R. M. Edwards, J. Cornish, M. Hegde, M. A. F. Murray, I. R. Reid, and T. Cundy, Evidence for a major gene for bone mineral density in idiopathic osteoporotic families. J. Bone Miner. Res. 15, 1132 – 1137 (2000). 250. N. Kelepouris, F. Kaplan, and M. Attie, Severe osteoporosis in men: a nine year follow-up study. J. Bone Miner. Res. 8, S332 (1993). 251. A. A. Licata, T. W. Bauer, and B. Stulberg, Chemical and histological features of men with osteoporosis. Cleveland Clin. J. Med. 60, 56 – 59 (1993). 252. P. R. Ebeling, Secondary causes of osteoporosis. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 483 – 504. Academic Press, San Diego, 1999. 253. B. E. Johnson, B. Lucasey, R. G. Robinson, and B. P. Lukert, Contributing diagnoses in osteoporosis. Arch. Intern. Med. 149, 1069 – 1072 (1989). 254. T. J. Hahn, S. L. Westbrook, and L. R. Halstead, Cortisol modulation of osteoblast metabolic activity in cultured neonatal rat bone. Endocrinol. 104, 715 – 721 (1984).

ORWOLL AND KLEIN 255. G. L. Wong, Basal activities and hormone responsiveness of osteoclast-like and osteoblast-like bone cells are regulated by glucocorticoids. J. Biol. Chem. 254, 6337 – 6340 (1979). 256. E. M. Canalis, Effect of glucocorticoids on type I collagen synthesis, alkaline phosphatase activity, and deocyribonucleic acid content in cultured rat calvariae. Endocrinology 112, 931 – 939 (1983). 257. J. W. Dietrich, E. M. Canalis, and D. M. Maina, Effect of glucocorticoids in fetal rat bone collagen synthesis in vitro. Endocrinology 104, 715 – 721 (1979). 258. T. L. McCarthy, M. Centrella, and E. Canalis, Cortisol inhibits the synthesis of insulin-like growth factor-I in skeletal cells. Endocrinology 126, 1569 – 1575 (1990). 259. L. G. Raisz and H. a. Simmon, Effects of parathyroid hormone and cortisol on prostaglandin production by neonatal rat calvaria in vitro. Endocrinol. Res. 11, 59 – 74 (1985). 260. Y. S. Chuyn, B. E. Kream, and L. G. Raisz, Cortisol decreases bone formation by inhibiting periosteal cell proliferation. Endocrinol. 114, 477 – 480 (1984). 261. R. S. Williams, Triamcinolone myopathy. Lancet 1, 698 – 701 (1959). 262. J. B. Harman, Muscular wasting and corticosteroid therapy. Lancet 1, 887 (1959). 263. B. Danneskjold-Samsoe and G. Grimby, The influence of prednisone on the muscle morphology and muscle enzymes in patients with rheumatoid arthritis. Clin. Sci. 71, 693 – 701 (1986). 264. A. Askari, P. J. Vignos, and R. W. Moskowitz, Steroid myopathy in connective tissue disease. Am. J. Med. 61, 485 – 492 (1976). 265. B. P. Lukert and J. S. Adams, Calcium and phosphorus homeostasis in man: Effect of corticosteroids. Arch. Int. Med. 136, 1249 – 1253 (1976). 266. P. Doerr and K. M. Pirke, Cortisol-induced suppression of plasma testosterone in normal adult males. J. Clin. Endocrinol. Metab. 43, 622 – 629 (1976). 267. M. R. MacAdams, R. H. White, and B. E. Chipps, Reduction of serum testosterone levels during chronic glucocorticoid therapy. Ann. Int. Med. 104, 648 – 651 (1986). 268. J. D. Veldhuis, G. Lizarralde, and A. Iranmanesh, Divergent effects of short term glucocorticoid excess on the gonadotropic and somatotropic axes in normal men. J. Clin. Endocrinol. Metab. 74, 96 – 102 (1992). 269. M. Sakakura, K. Takebe, and S. Nakagawa, Inhibition of luteinizing hormone secretion induced by synthetic LRH by long-term treatment with glucocorticoids in human subjects. J. Clin. Endocrinol. Metab. 40, 774 – 779 (1975). 270. D. Evain, A. M. Morera, and J. M. Saez, Glucocorticoid receptors in interstitial cells of the rat testis. J. Steroid. Biochem. 7, 1135 – 1139 (1976). 271. D. S. Colvard, E. F. Eriksen, P. E. Keeting, E. M. Wilson, D. B. Lubahn, F. S. French, B. L. Riggs, and T. C. Spelsberg, Identification of androgen receptors in normal human osteoblast-like cells. Proc. Natl. Acad. Sci. USA 86, 854 – 857 (1989). 272. C. H. Kasperk, J. E. Wergedal, J. R. Farley, T. A. Linkhart, R. T. Turner, and D. J. Baylink, Androgens directly stimulate proliferation of bone cells in vitro. Endocrinology. 124(3), 1576 – 1578 (1989). 273. E. S. Orwoll, L. Stribrska, E. E. Ramsay, and E. Keenan, Androgen receptors in osteoblast-like cell lines. Calcif. Tissue Int. 49, 182 – 187 (1991). 274. A. Carracosa, L. Audi, M. A. Ferrandez, and A. Ballabriga, Biological effects of androgens and identification of specific dihydrotestosterone-binding sites in cultured human fetal epiphyseal chondrocytes. J. Clin. Endocrinol. Metab. 70, 134 – 140 (1990). 275. S. P. Kapur and A. H. Reddi, Influence of testosterone and dihydrotestosterone on bond-matrix induced endochondral bone formation. Calcif. Tissue Int. 44, 108 – 113 (1989). 276. C. C. Pilbeam and L. G. Raisz, Effects of androgens on parathyroid hormone and interleukin-1-stimulated prostaglandin production in cultured neonatal mouse calvariae. J. Bone Miner. Res. 5, 1183 – 1188 (1990).

143

CHAPTER 42 Osteoporosis in Men 277. C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). 278. S. Fukayama and H. J. Tashjian, Direct modulation by androgens of the response of human bone cells (SaOS-2) to human parathyroid hormone (PTH) and PTH-related protein. Endocrinology 125, 1789 – 1794 (1989). 279. R. Vaishnav, J. N. Beresford, J. A. Gallagher, and R. G. G. Russell, Effects of the anabolic steroid stanozolol on cells derived from human bone. Clin. Sci. 74, 455 – 460 (1988). 280. D. J. Benz, M. R. Haussler, M. A. Thomas, B. Speelman, and B. S. Komm, High-affinity androgen binding and androgenic regulation of a1(I)-procollagen and transforming growth factor-B steady state messenger ribonucleic acid levels in human osteoblast-like osteosarcoma cells. Endocrinology 128, 2723 – 2730 (1991). 281. M. T. Corvol, A. Carracosa, L. Tsagris, O. Blanchard, and R. Rappaport, Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: Influence of age and sex. Endocrinology 120, 1422 – 1429 (1987). 282. J. D. Wilson and C. Roehrborn, Long-term consequences of castration in men: Lessions from the skoptzy and the eunuchs of the Chinese and Ottoman courts. J. Clin. Endocrinol. Metab. 84, 4324 – 4331 (1999). 283. S. Krabbe, L. Hummer, and C. Christiansen, Longitudinal study of calcium metabolism in male puberty. Acta. Poediatr. Scand. 73, 750 – 755 (1984). 284. N. Mauras, M. W. Haymond, D. Darmaun, N. E. Vieira, S. A. Abrams, and A. L. Yergey, Calcium and protein kinetics in prepubertal boys. Positive effect of testosterone. J. Clin. Invest. 93, 1014 – 1019 (1994). 285. N. Mauras, V. Y. Hayes, N. E. Vieira, A. L. Yergey, and K. O. O’Brien, Profound hypogonadism has significant negative effects on calcium balance in males: A calcium kinetic study. J. Bone Miner. Res. 14, 577 – 582 (1999). 286. L. N. Parker, Adrenarche. Endocr. Metab. Clin. N. Am. 20, 71 – 83 (1991). 287. S. G. Soule, G. Conway, G. M. Prelevic, M. Prentice, J. Ginsburg, and H. S. Jacobs, Osteopenia is a feature of the androgen insensitivity syndrome. Clin. Endocrinol. 43, 671 – 675 (1995). 288. M. Munoz-Torres, E. Jodar, M. Quesada, and Escobar-Jimenez Bone mass in androgen-insensitivity syndrome: Response to hormonal replacement therapy. Calcif. Tissue Int. 57, 94 – 96 (1995). 289. R. Marcus, D. Leary, D. L. Schneider, E. Shane, M. Favus, and C. A. Quigley, The contribution of testosterone to skeletal development and maintenance: Lessons from the androgen insensitivity syndrome. J. Clin. Endocrinol. Metab. 85, 1032 – 1037 (2000). 290. J. S. Finkelstein, A. Klibanski, R. M. Neer, S. L. Greenspan, D. I. Rosenthal, and W. F. J. Crowley, Osteoporosis in men with idiopathic hypogonadotropic hypogonadism. Ann. Int. Med. 106, 354 – 361 (1987). 291. D. A. S. Smith and M. S. Walker, Changes in plasma steroids and bone density in Kleinfelter’s syndrome. Calcif. Tissue Int. 22 (Suppl.), 225 – 228 (1977). 292. O. Arisaka, M. Arisaka, A. Hosaka, N. Shimura, and K. Yabuta, Effect of testosterone on radial bone mineral density in adolescent male hypogonadism. Acta. Paediatr. Scand. 80, 378 – 380 (1991). 293. M. Horowitz, J. M. Wishart, P. D. O’Loughlin, H. A. Morris, A. G. Need, and B. E. C. Nordin, Osteoporosis and Kleinefelter’s syndrome. Clin. Endocrinol. (Oxf.) 36, 113 – 118 (1992). 294. J. S. Finkelstein, A. Klibanski, and R. M. Neer, A longitudinal evaluation of bone mineral density in adult men with histories of delayed puberty. J. Clin. Endocrinol. Metab. 81, 1152 – 1155 (1996). 295. M. N. Moreira-Andres, F. J. Canizo, F. J. de la Cruz, A. Gomez-de la Camara, and F. G. Hawkins, Bone mineral status in prepubertal chil-

296.

297.

298.

299.

300.

301.

302.

303.

304.

305.

306.

307.

308.

309.

310.

311.

312.

313. 314.

dren with constitutional delay of growth and puberty. Eur. J. Endocrinol. 139, 271 – 275 (1998). J. S. Finkelstein, R. M. Neer, B. M. K. Biller, J. D. Crawford, and A. Klibanski, Osteopenia in men with a history of delayed puberty. N. Engl. J. Med. 326, 600 – 604 (1992). S. Bertelloni, G. I. Baroncelli, M. Ferdeghini, G. Perri, and G. Saggese, Normal volumetric bone mineral density and bone turnover in young men with histories of constitutional delay of puberty. J. Clin. Endocrinol. Metab. 83, 4280 – 4283 (1998). D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Scot, K. Chung, R. S. Lucas, T. A. Einhorn, and R. Bouillon, Bone and mineral metabolism in the androgen-resistant (testicular feminized) male rat. J. Bone Miner. Res. 8, 801 – 809 (1993). J. M. Hock, I. Gera, J. Fonseca, and L. G. Raisz, Human parathyroid hormone (1 – 34) increases bone mass in ovariectomized and orchidectomized rats. Endocrinology 122, 2899 – 2904 (1988). R. T. Turner, K. S. Hannon, L. N. Demers, J. Buchanan, and N. Bell, Differential effects of gonadal function in bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). H. L. Stanley, B. P. Schmitt, R. M. Poses, and W. P. Deiss, Does hypogonadism contribute to the occurrence of a minimal trauma hip fracture in elderly men? J. Am. Geriatr. Soc. 39, 766 – 771 (1991). J. J. Stepan, M. Lachman, J. Zverina, V. Pacovsky, and D. J. Baylink, Castrated men exhibit bone loss: Effect of calcitonin treatment on biochemical indices of bone remodeling. J. Clin. Endocrinol. Metab. 69, 523 – 527 (1989). S. L. Greenspan, R. M. Neer, E. C. Ridgway, and A. Klibanski, Osteoporosis in men with hyperprolactinemic hypogonadism. Ann. Int. Med. 104, 777 – 782 (1986). S. L. Greenspan, D. S. Oppenheim, and A. Klibanski, Importance of gonadal steroids to bone mass in men with hyperprolactinemic hypogonadism. Ann. Int. Med. 110, 526 – 531 (1989). T. Diamond, D. Stiel, and S. Posen, Effects of testosterone and venesection on spinal and peripheral bone mineral in six hypogonadal men with hemochromatosis. J. Bone Miner. Res. 6, 39 – 43 (1991). M. F. Townsend, W. H. Sanders, R. O. Northway, and S. D. Graham, Bone fractures associated with luteinizing hormone-releasing hormone agonists used in the treatment of prostate carcinoma. Cancer 79, 545 – 550 (1997). T. Diamond, J. Campbell, C. Bryant, and W. Lynch, The effect of combined androgen blockade on bone turnover and bone mineral densities in men treated for prostate carcinoma. Cancer 83, 1561 – 1566 (1998). J. F. Maillefert, J. Sibilia, F. Michel, C. Saussine, R. M. Javier, and C. Tavernier, Bone mineral density in men treated with synthetic gonadotropin-releasing hormone agonists for prostatic carcinoma. J. Urol. 161, 1219 – 1222 (1999). D. Goldray, Y. Weisman, N. Jaccard, C. Merdler, J. Chen, and H. Matzkin, Decreased bone density in elderly men treated with the gonadotropin-releasing hormone agonist decapeptyl (D-Trp6GnRH). J. Clin. Endocrinol. Metab. 76, 288 – 290 (1993). C. Foresta, G. Ruzza, R. Mioni, A. Meneghello, and C. Baccichetti, Testosterone and bone loss in Klinefelter’s syndrome. Horm. Metab. Res. 15, 56 – 67 (1983). G. Isaia, M. Mussetta, F. Pecchio, A. Sciolla, M. Di Stefano, and G. M. Molinatti, Effect of testosterone on bone in hypogonadal males. Maturitas 15, 47 – 51 (1992). J. P. Devogelaer, S. De Cooman, and C. N. de Deuxchaisnes, Low bone mass in hypogonadal males. Effect of testosterone substitution therapy, a densitometric study. Maturitas 15, 17 – 23 (1992). B. P. Lukert and L. G. Raisz, Glucocorticoid-induced osteoporosis: Pathogenesis and management. Ann. Int. Med. 112, 352 – 364 (1990). G. Girasole and G. Passeri, Upregulation of osteoclastogenic potential of the marrow is induced by orchidectomy and is reversed by

144

315.

316.

317.

318.

319.

320. 321. 322. 323. 324. 325.

326. 327.

328.

329.

330.

331.

332.

333.

334.

335.

336.

337.

ORWOLL AND KLEIN testosterone replacement in the mouse. J. Bone Miner. Res. 7, S96 (1992). R. M. Francis and M. Peacock, Osteoporosis in hypogonadal men: Role of decreased plasma 1,25-dihydroxyvitamin D, calcium malabsorption, and low bone formation. Bone 7, 261 – 268 (1986). P. Delmas and P. J. Meunier, L’osteoporose au cours du syndrome de Klinefelter. Donnees histologiques osseuses quantitatives dans cinq cas. Relation avec la carence hormonale. Nouv. Presse Med. 10, 687 (1981). D. T. Baran, M. A. Bergfeld, S. L. Teitelbaum, and L. V. Avioli, Effect of testosterone therapy on bone formation in an osteoporotic hypogonadal male. Calcif. Tissue Res. 26, 103 – 106 (1978). J. A. Jackson, M. Kleerekoper, A. M. Parfitt, D. Sukhaker Rao, A. R. Villanueva, and B. Frame, Bone histomorphometry in hypogonadal and eugonadal men with spinal osteoporosis. J. Clin. Endocrinol. Metab. 65, 53 – 58 (1987). D. D. Bikle, Alcohol-induced bone disease. In “Osteoporosis: Nutritional Aspects” (A. P. Simopoulos and C. Gallis, ed.), pp. 53 – 79. Karger, Basel, 1993. K. Laitinen and M. Valimaki, Bone and the “comforts of life.” Ann. Med. 25, 413 – 425 (1993). C. Moniz, Alcohol and bone. Br. Med. Bull. 50, 67 – 75 (1994). P. D. Saville, Changes in bone mass with age and alcoholism. J. Bone J. Surg. 47A, 492 – 499 (1965). N. Dalen and B. Lamke, Bone mineral losses in alcoholics. Acta Orthop. Scand. 47, 469 – 471 (1976). B. E. Nilsson, Conditions contributing to fracture of the femoral neck. Acta Chir. Scand. 136, 383 – 384 (1970). B. C. Lalor, M. W. France, D. Powell, P. H. Adams, and T. C. Counihan, Bone and mineral metabolism and chronic alcohol abuse. Q. J. Med. 59, 497 – 511 (1986). S. Feitelberg, S. Epstein, and F. Ismail, Deranged bone mineral metabolism in chronic alcoholism. Metabolism 36, 322 – 326 (1987). R. G. Crilly and L. Richardson-Delaquerriere, Current bone mass and body weight changes in alcoholic males. Calcif. Tissue Int. 46, 169 – 172 (1990). H. Spencer, N. Rubio, E. Rubio, M. Indreika, and A. Seitam, Chronic alcoholism. Frequently overlooked cause of osteoporosis in men. Am. J. Med. 80, 393 – 397 (1986). D. R. Lindsell, A. G. Wilson, and J. D. Maxwell, Fractures on the chest radiograph in detection of alcoholic liver disease. Br. Med. J. 285, 597 – 599 (1982). Y. Israel, H. Orrego, S. Holt, D. W. Macdonald, and H. E. Meema, Identification of alcohol abuse: Thoracic fractures on routine chest x-rays as indicators of alcoholism. Alcoholism 4, 420 – 422 (1980). R. G. Crilly, C. Anderson, and D. Hogan, Bone histomorphometry, bone mass and related parameters in alcoholic males. Calcif. Tissue Int. 43, 269 – 276 (1988). T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, and S. Posen, Ethanol reduced bone formation and may cause osteoporosis. Am. J. Med. 86, 282 – 288 (1989). D. D. Bikle, H. K. Genant, C. E. Cann, R. R. Recker, B. P. Halloran, and G. J. Strewler, Bone disease in alcohol abuse. Ann. Int. Med. 103, 42 – 48 (1985). M. Hernandez-Avila, G. A. Colditz, and M. J. Stampger, Caffeine, moderate alcohol intake and risk of fractures of the hip and forearm in middle-aged women. Am. J. Clin. Nutr. 54, 157 – 163 (1991). J. L. Gonzalez-Calvin, A. Garcia-Sanchez, and V. Bellot, Mineral metabolism, osteoblastic function and bone mass in chronic alcoholism. Alcohol Alcohol. 28, 571 – 579 (1993). T. L. Holbrook and E. Barrett-Connor, A prospective study of alcohol consumption and bone mineral density. Br. Med. J. 306, 1506 – 1509 (1993). V. R. Preedy, J. S. Marway, and J. R. Salisbury, Protein synthesis in bone and skin of the rat are inhibited by ethanol: Implications for

338.

339. 340.

341.

342.

343. 344. 345.

346.

347.

348.

349. 350.

351.

352.

353.

354.

355.

356.

357. 358.

whole body metabolism. Alcohol Clin. Exp. Res. 14, 165 – 168 (1990). T.-C. Peng, R. P. Kusy, P. F. Hirsch, and J. R. Hagaman, Ethanolinduced changes in the morphology and strength of femurs of rats. Alcohol Clin. Exp. Res. 12, 655 – 659 (1988). R. O. Pierce and A. Perry, The effect of ethanol on bone mineral. J. Natl. Med. Assoc. 83, 505 – 508 (1991). J. Lindholm, T. Steiniche, and E. Rasmussen, Bone disorder in men with chronic alcoholism: A reversible disease? J. Clin. Endocrinol. Metab. 73, 118 – 124 (1991). D. D. Bikle, A. Stesin, and B. Halloran, Alcohol-induced bone disease: Relationship to age and parathyroid hormone levels. Alcohol Clin. Exp. Res. 17, 690 – 695 (1993). A. Diez, J. Puig, and S. Serrano, Alcohol-induced bone disease in the absence of severe chronic liver damage. J. Bone Miner. Res. 9, 825 – 831 (1994). O. Johnell, B. E. Nilsson, and P. E. Wikluind, Bone morphometry in alcoholics. Clin. Orthop. 165, 253 – 258 (1982). P. D. Saville and C. S. Lieber, Increases in skeletal calcium and femur cortex thickness in undernutrition. J. Nutr. 99, 141 – 144 (1969). J. M. Kalbfleisch, R. D. Lindeman, and H. E. Ginn, Effects of ethanol administration on urinary excretion of magnesium and other electrolytes in alcoholic and normal subjects. J. Clin. Invest. 42, 1471 – 1475 (1963). D. H. Avery, J. E. Overall, and H. M. Calil, Plasma calcium and phosphate during alcohol intoxication. J. Stud. Alcohol. 44, 205 – 214 (1983). R. Anderson, M. Cohen, and R. Haller, Skeletal muscle phosphorus and magnesium deficiency in alcoholic myopathy. Min. Elect. Metab. 4, 106 – 112 (1980). K. Laitinen, C. Lamberg-Allardt, and R. Tunninen, Bone mineral density and abstention-induced changes in bone and mineral metabolism in noncirrhotic male alcoholics. Am. J. Med. 93, 642 – 650 (1992). M. C. Territo and K. R. Tanaka, Hypophosphatemia in chronic alcoholism. Arch. Int. Med. 134, 445 – 447 (1974). K. Laitinen, C. Lamberg-Allardt, and R. Tunninen, Transient hypoparathyroidism during acute alcohol intoxication. N. Engl. J. Med. 324, 721 – 727 (1991). M. Z. Verbanck, J. Verbanck, J. Brauman, and J. P. Mullier, Bone histology and 25OH-vitamin D plasma levels in alcoholics without cirrhosis. Calcif. Tissue Res. 22, 538 – 541 (1977). O. H. Sorensen, B. Lund, and M. Hilden, 25-Hydroxylation in chronic alcoholic liver disease. In “Vitamin D: Biochemical, Chemical and Clinical Aspects Related to Calcium Metabolism” (A. W. Norman, K. Schaefer and J. W. Coburn, eds.), pp. 843 – 845. Walter de Guyter, Hawthorne, NY, 1977. D. B. Posner, R. M. Russell, and S. Ansood, S. Effective 25-hydroxylation of vitamin D in alcoholic cirrhosis. Gastroenterology 74, 866 – 870 (1978). D. D. Bikle, B. P. Halloran, E. Gee, E. Ryzen, and J. G. Haddad, Free 25-hydroxyvitamin D levels are normal in subjects with liver disease and reduced total 25-hydroxyvitamin D levels. J. Clin. Invest. 78, 748 – 752 (1986). D. D. Bikle, E. Gee, and B. Halloran, Free 1,25-dihydroxyvitamin D levels in serum from normal subjects, pregnant subjects and subject with liver disease. J. Clin. Invest. 74, 1966 – 1971 (1984). M. Valimaki, M. Salaspuro, and R. Ylikahri, Liver damage and sex hormones in chronic male alcoholics. Clin. Endocrinol. 17, 469 – 477 (1982). K. Friday and G. A. Howard, Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40, 562 – 565 (1991). P. Chavassieux, C. M. Serre, P. Vernaud, P. Delmas, and P. Meunier, In vitro evaluation of dose-effects of ethanol on human osteoblastic cells. Bone Miner. 22, 95 – 103 (1993).

CHAPTER 42 Osteoporosis in Men 359. M. Garcia-Carrasco, M. Gruson, and C. deVernejoul, Osteocalcin and bone histomorphometric parameters in adults without bone disease. Calcif. Tissue Int. 42, 13 – 17 (1988). 360. P. D. Delmas, L. Malaval, and M. E. Arlot, Serum bone Gla-protein compared to bone histomorphometry in endocrine diseases. Bone 6, 339 – 341 (1985). 361. J. Jaouhari, F. Schiele, and P. Pirollet, Concentration and hydroxyapatite binding capacity of plasma osteocalcin in chronic alcoholic men: Effect of a three-week withdrawal therapy. Bone Miner. 21, 171 – 178 (1993). 362. M. Labib, M. Abdel-Kader, and L. Ranganath, Bone disease in chronic alcoholism: The value of plasma osteocalcin measurement. Alcohol Alcoholism 24, 141 – 144 (1989). 363. H. Rico, J. A. Cabranes, and J. Cabello, Low serum osteocalcin in acute alcohol intoxication: A direct effect of alcohol on osteoblasts. Bone Miner. 2, 221 – 225 (1987). 364. H. K. Nielsen, L. Lundby, and K. Rasmssen, Alcohol decreases serum osteocalcin in a dose-dependent way in normal subjects. Calcif. Tissue Int. 46, 173 – 178 (1990). 365. E. A. Krall and B. Dawson-Hughes, Smoking and bone loss among postmenopausal women. J. Bone Miner. Res. 6, 331 – 338 (1991). 366. J. Jensen and C. Christiansen, Effects of smoking on serum lipoproteins and bone mineral content during postmenopausal hormone replacement therapy. Am. J. Obstetr. Gynecol. 159, 820 – 825 (1988). 367. H. W. Daniell, Osteoporosis of the slender smoker. Arch. Intern. Med. 136, 298 – 304 (1976). 368. J. Kanis, O. Johnell, B. Gullberg, E. Allander, L. Elffors, J. Ranstam, J. Dequeker, G. Dilsen, C. Gennari, A. Lopes Vaz, G. Lyritis, G. Mazzuoli, L. Miravet, M. Passeri, R. Perez Cano, A. Rapado, and C. Ribot, Risk factors for hip fracture in men from southern Europe: The MEDOS study. Osteoporosis Int. 9, 45 – 54 (1999). 369. J. M. Vogel, J. W. Davis, A. Nomura, R. D. Wasnich, and P. D. Ross, The effects of smoking on bone mass and the rates of bone loss among elderly japanese-american men. J. Bone Miner. Res. 12, 1495 – 1501 (1997). 370. T. V. Nguyen, P. J. Kelly, P. N. Sambrook, C. Gilbert, N. A. Pocock, and J. A. Eisman, Lifestyle factors and bone density in the elderly: Implications for osteoporosis prevention. J. Bone Miner. Res. 9, 1339 – 1346 (1994). 371. D. P. Kiel, Y. Zhang, M. T. Hannan, J. J. Anderson, J. A. Baron, and D. T. Felson, The effect of smoking at different life stages on bone mineral density in elderly men and women. Osteoporosis Int. 6, 240 – 248 (1996). 372. J. Jensen, C. Christiansen, and P. Rodbro, Cigarette smoking, serum estrogens and bone loss during hormone-replacement therapy early after menopause. N. Engl. J. Med. 313, 973 – 975 (1985). 373. P. Bordier, A. Ryckewart, and A. Gueris, On the pathogenesis of socalled hypercalciuria. Am. J. Med. 63, 398 – 409 (1977). 374. E. M. Alhava, M. Juuti, and P. Karjalainen, Bone mineral density in patients with urolithiasis. Scand. J. Urol. Nephrol. 10, 154 – 156 (1976). 375. S. Lawoyin, S. Sismilich, and R. Browne, Bone mineral content in patients with calcium urolithiasis. Metabolism 28, 1250 – 1254 (1979). 376. D. R. Wilson, J. Barkin, and A. L. Strauss, Bone mineral content in calcium urolithiasis. Urol. Res. 12, 37 (1984). 377. P. Jaeger, K. Lippuner, J. P. Casez, B. Hess, D. Ackermann, and C. Hug, Low bone mass in idiopathic renal stone formers: Magnitude and significance. J. Bone Miner. Res. 9, 1525 – 1532 (1994). 378. J. E. Zerwekh, Hypercalciuria and bone disease. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 463 – 482. Academic Press, San Diego, 1999. 379. I. P. Heilberg, L. A. Martini, V. L. Szejnfeld, A. B. Carvalho, S. A. Draibe, H. Ajzen, O. L. Ramos, and N. Schor, Bone disease in calcium stone forming patients. Clin. Nephrol. 42, 175 – 182 (1994).

145 380. L. H. Smith, The medical aspects of urolithiasis. J. Urol. 141, 707 – 710 (1989). 381. K. Sakhaee, M. J. Nicar, K. Glass, and C. Y. C. Pak, Postmenopausal osteoporosis as a manifestation of renal hypercalciuria with secondary hyperparathyroidism. J. Clin. Endocrinol. Metab. 61, 368 – 373 (1985). 382. F. Pietschmann, N. A. Breslau, and C. Y. C. Pak, Reduced vertebral bone density in hypercalciuric nephrolithiasis. J. Bone Miner. Res. 7(12), 1383 – 1388 (1992). 383. R. A. L. Sutton and V. R. Walker, Bone resorption and hypercalciuria in calcium stoneformers. Metabolism 35, 485 – 488 (1986). 384. J. E. Zerwekh, K. Sakhaee, N. A. Breslau, F. Gottschalk, and C. Y. C. Pak, Impaired bone formation in male idiopathic osteoporosis: Further reduction in the presence of concomitant hypercalciuria. Osteoporosis Int. 2, 128 – 134 (1992). 385. F. L. Coe, J. M. Canterbury, J. J. Firpo, and E. Reiss, Evidence for secondary hyperparathyroidism in idiopathic hypercalciuria. J. Clin. Invest. 52, 134 – 142 (1973). 386. J. J. Lemann, E. M. Worcester, and R. W. Gray, Hypercalciuria and stones. Am. J. Kidney. Dis. XVII, 386 – 391 (1991). 387. M. Fuss, T. Pepersack, J. Van Geel, J. Corvilain, J.-C. Vandewalle, P. Bergmann, and J. Simon, Involvement of low-calcium diet in the reduced bone mineral content of idiopathic renal stone formers. Calcif. Tissue Int. 46, 9 – 13 (1990). 388. R. J. Brenner, D. B. Spring, A. Sebastian, E. M. McSherry, H. K. Genant, A. J. Palubinskas, and R. C. J. Morris, Incidence of radiographically evident bone disease, nephrocalcinosis, and nephrolithiasis in various types of renal tubular acidosis. N. Engl. J. Med. 307, 217 – 221 (1982). 389. M. Tieder, D. Modai, U. Shaked, R. Samuel, R. Arie, A. Halabe, K. Maor et al., “Idiopathic” hypercalciuria and hereditary hypophosphatemic rickets. N. Engl. J. Med. 316, 125 – 129 (1987). 390. D. S. Rao, B. Frame, and M. A. Block, Primary hyperparathroidism: A cause of hypercalciuria and renal stones in patients with medullary sponge kidney. JAMA 237, 1353 – 1355 (1977). 391. P. W. Dlabal, R. M. Jordan, and S. G. Dorfman, Medullary sponge kidney and renal-leak hypercalciuria. JAMA 241, 1490 – 1491 (1979). 392. M. O’Neill, N. A. Breslau, and C. Y. C. Pak, Metabolic evaluation of nephrolithiasis in patients with medullary sponge kidney. JAMA 245, 1233 – 1236 (1981). 393. O. M. Wrong, A. G. W. Norden, and T. G. Feest, Dent’s disease: A familial proximal renal tubular syndrome with low-molecular-weight proteinuria, hypercalciuria, nephrocalcinosis, metabolic bone disease, progressive renal failure and a marked male predominance. Q. J. Med. 87, 473 – 493 (1994). 394. H. M. I. Perry, M. D. Fallon, M. Bergfeld, S. L. Teitelbaum, and L. V. Avioli, Osteoporosis in young men — A syndrome of hypercalciuria and accelerated bone turnover. Arch. Intern. Med. 142, 1295 – 1298 (1982). 395. H. H. Malluche, W. Tschoepe, E. Ritz, W. Meyer-Sabellek, and S. G. Massry, Abnormal bone histology in idiopathic hypercalciuria. J. Clin. Endocrinol. Metab. 50, 654 – 658 (1980). 396. W. P. U. Jackson, Osteoporosis of unknown cause in younger people. J. Bone J. Surg. 40B, 420 – 441 (1958). 397. T. J. Hahn, Drug-induced disorders of vitamin D and mineral metabolism. Clin. Endocrinol. Metab. 9, 107 – 129 (1980). 398. M. Bornemann, J. R. Saxon, and G. S. I. Kidd, Osteoporosis unmasked by hyperthyroidism in a young man with osteogenesis imperfecta. Arch. Intern. Med. 147, 1947 – 1948 (1987). 399. C. L. Donaldson, S. B. Hulley, and J. M. Vogel, Effect of prolonged bed rest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 400. R. Eastell and B. L. Riggs, Diagnostic evaluation of osteoporosis. Endocrinol. Metab. Clin. North. Am. 17, 547 – 571 (1980). 401. J. M. Lane and V. J. Vigorita, Osteoporosis. Orthop. Clin. North. Am. 15, 711 – 728 (1984).

146 402. T. W. O’Neill, J. Valow, and C. Cooper, Differences in vertebral deformity indices between 3 European populations. J. Bone Miner. Res. 8, S149 (1993). 403. K. M. Davies, M. R. Stegman, and R. R. Recker, Preliminary vertebral deformity analysis for a rural population of older men and women. J. Bone Miner. Res. 8, S331 (1993). 404. C. C. J. Johnston, Clinical indications for bone mass measurements: A report from the scientific advisory board on the national osteoporosis foundation. J. Bone Miner. Res. 4, 1 – 28 (1989). 405. C. C. J. r. Johnston, C. W. Slemenda, and L. J. I. I. I. Melton, Clinical use of bone densitometry. N. Engl. J. Med. 324, 1105 – 1109 (1991). 406. L. J. Melton, Editorial: The prevalence of osteoporosis. J. Bone Miner. Res. 12, 1769 – 1771 (1997). 407. S. Cheng, H. Suominen, R. Sakari-Rantala, P. Laukkanen, V. Avikainen, and E. Heikkinen, Calcaneal bone mineral density predicts fracture occurrence: A five-year follow-up study in elderly people. J. Bone Miner. Res. 12, 1075 – 1082 (1997). 408. P. L. Selby, M. Davies, and J. E. Adams, Do men and women fracture bones at similar bone densities? Osteoporosis Int. 11, 153 – 157 (2000). 409. G. A. Campbell, D. J. Hosking, J. R. Kemm, and R. V. Boyd, How common is osteomalacia in the elderly? Lancet 2, 386 – 388 (1984). 410. J. E. Aaron, L. Stasiak, J. C. Gallagher, E. B. Longton, M. Nicholson, J. Anderson, and B. E. C. Nordin, Frequency of osteomalacia and osteoporosis in fractures of the proximal femur. Lancet 1, 229 – 233 (1974). 411. L. D. Hordon and M. Peacock, Osteomalacia and osteoporosis in femoral neck fracture. Bone Miner. 11, 247 – 259 (1990). 412. L. Sokoloff, Occult osteomalacia in America (USA) patients with fracture of the hip. Am. J. Surg. Pathol. 2, 21 – 30 (1978). 413. T. J. Wilton, D. J. Hosking, E. Pawley, A. Stevens, and L. Harvey, Osteomalacia and femoral neck fractures in the elderly patient. J. Bone J. Surg. 69-B, 388 – 390 (1987). 414. G. M. Marel, M. J. McKenna, and B. Frame, Osteomalacia. In “Bone and Mineral Research/4” (W. A. Peck, ed.), Vol. 4, pp. 335 – 413. Elsevier Science, Amsterdam, 1986. 415. F. R. Singer, Metabolic bone disease. In “Endocrinology and Metabolism” (P. Felig, J. D. Baxter, A. E. Broadus and L. A. Frohman, eds.), pp. 1454 – 1499. McGraw-Hill, New York, 1987. 416. R. F. Klein and M. Gunness, The transiliac bone biopsy: When to get it and how to interpret it. Endocrinologist 2, 158 – 168 (1992). 417. A. Chines, R. Pacifici, L. A. Avioli, P. E. Korenblat, and S. L. Teitelbaum, Systemic mastocytosis and osteoporosis. Osteoporosis Int. 1, S147 – S149 (1993). 418. A. Chines, R. Pacifici, L. V. Avioli, S. L. Teitelbaum, and P. E. Korenblat, Systemic mastocytosis presenting as osteoporosis: A clinical and histomorphometric study. J. Clin. Endocrinol. Metab. 72, 140 – 144 (1991). 419. B. Demiaux, M. E. Arlot, M.-C. Chapuy, P. J. Meunier, and P. D. Delams, Serum osteocalcin is increased in patients with osteomalacia: Correlations with biochemical and histomorphometric findings. Clin. Endocrinol. Metab. 74, 1146 – 1151 (1992). 420. S. P. Robins, D. Black, C. R. Paterson, D. M. Reid, A. Duncan, and M. J. Seibel, Evaluation of urinary hydroxypyridinium crosslink measurements as resorption markers in metabolic bone disease. Eur. J. Clin. Invest. 21, 310 – 315 (1991). 421. R. Eastell, I. T. Boyle, J. Compston, C. Cooper, I. Fogelman, R. M. Francis, D. J. Hosking, D. W. Purdie, S. Ralston, J. Reeve, D. M. Reid, R. G. G. Russell, and J. C. Stevenson, Management of male osteoporosis: Report of the UK consensus group. Q. J. Med. 91, 71 – 92 (1998). 422. R. Agrawal, S. Wallach, S. Cohn, M. Tessier, R. Verch, M. Hussain, and I. Zanzi, Calcitonin treatment of osteoporosis. In “Calcitonin” (A. Pecile, ed.), pp. 237. Exerpta Medica, Amsterdam, 1981.

ORWOLL AND KLEIN 423. L. Erlacher, J. Kettenbach, H. Kiener, W. Graninger, F. Kainberger, and P. Pietschmann, Salmon calcitonin and calcium in the treatment of male osteoporosis: The effect on bone mineral density. Wien Klin. Wochenschr. 109, 270 – 274 (1997). 424. G. P. Lyritis, I. Paspati, T. Karachalios, D. Loakimidis, G. Skarantavos, and P. G. Lyritis, Pain relief from nasal salmon calcitonin in osteoporotic vertebral crush fractures. Acta. Orthop. Scand. 68 (Suppl. 275), 112 – 114 (1997). 425. R. Valkema, F.-J. F. E. Vismans, S. E. Papapoulos, E. K. J. Pauwels, and O. L. M. Bijvoet, Maintained improvement in calcium balance and bone mineral content in patients with osteoporosis treated with the bisphosphonate APD. Bone Miner. 5, 183 – 192 (1989). 426. S. M. Orme, M. Simpson, S. P. Stewart, B. Oldroyd, C. F. Westmacott, M. A. Smith, and P. E. Belchetz, Comparison of changes in bone mineral in idiopathic and secondary osteoporosis following therapy with cyclical disodium etidronate and high dose calcium supplementation. Clin. Endocrinol. 41, 245 – 250 (1994). 427. F. H. Anderson, R. M. Francis, J. C. Bishop, and D. J. Rawlings, Effect of intermittent cyclical disodium etidronate therapy on bone mineral density in men with vertebral fractures. Age Ageing 26, 359 – 365 (1997). 428. P. Geusens, J. Van Hoof, J. Raus, J. Dequeker, J. Nijs, and J. Joly, Treatment with etidronate for men with idiopathic osteoporosis. Ann. Rheum. Dis. 56, 280 (1997). 429. E. Orwoll, M. Ettinger, S. Weiss, P. Miller, D. Kendler, J. Graham, S. Adami, K. Weber, R. Lorenc, P. Pietschmann, K. Vandormael, and A. Lombardi, Alendronate treatment of osteoporosis in men. N. Engl. J. Med. 343, 604 – 610 (2000). 430. P. Pitt, F. Li, P. Todd, D. Webber, S. Pack, and C. Moniz, A double blind placebo controlled study to determine the effects of intermittent cyclical etidronate on bone mineral density in patients on long-term oral corticosteroid treatment. Thorax 53, 351 – 356 (1998). 431. K. G. Saag, R. Emkey, T. J. Schnitzer, J. P. Brown, F. Hawkins, S. Goemaere, G. Thamsborg, U. A. Liberman, P. D. Delmas, M.-P. Malice, M. Czachur, and A. G. Diafotis, Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N. Engl. J. Med. 339, 292 – 299 (1998). 432. J. D. Adachi, W. G. Bensen, J. Brown, D. Hanley, A. Hodsman, R. Josse, D. L. Kendler, B. Lentle, W. Olszynski, L.-G. Ste.-Marie, A. Tenenhouse, and A. A. Chines, Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N. Engl. J. Med. 337, 382 – 387 (1997). 433. D. M. Reid, R. A. Hughes, R. F. J. M. Laan, N. A. Sacco-Gibson, D. H. S. A. Wenderoth, R. A. Eusebio, and J. P. Devogelaer, Efficacy and safety of daily residronate in the treatment of corticosteroidinduced osteoporosis in men and women: A randomized trial. J. Bone Miner. Res. 15, 1006 – 1013 (2000). 434. I. P. Heilberg, L. A. Martini, S. H. Teixeira, V. L. Szejnfeld, A. B. Carvalho, R. Lobao, and S. A. Draibe, Effect of etidronate treatment on bone mass of male nephrolithiasis patients with idiopathic hypercalciuria and osteopenia. Nehpron 79, 430 – 437 (1998). 435. A. Z. LeCroix, J. Wienpahl, L. R. White, R. B. Wallace, P. A. Scherr, L. K. George, J. Cornoni-Huntley, and A. M. Ostfeld, Thiazide diuretic agents and the incidence of hip fracture. N. Engl. J. Med. 322, 286 – 290 (1990). 436. R. D. Wasnich, R. J. Benfante, K. Yano, L. Heilbrun, and J. M. Vogel, Thiazide effect on the mineral content of bone. N. Engl. J. Med. 309, 344 – 347 (1983). 437. D. J. Morton, E. L. Barrett-Connor, and S. L. Edelstein, Thiazides and bone mineral density in elderly men and women. Am. J. Epidemiol. 139, 1107 – 1115 (1994). 438. R. Wasnich, J. Davis, P. Ross, and J. Vogel, Effect of thiazide on rates of bone mineral loss: a longitudinal study. Br. Med. J. 301, 1303 – 1305 (1990).

147

CHAPTER 42 Osteoporosis in Men 439. W. A. Ray, M. R. Griffin, W. Downey, and L. J. I. Melton, Long-term use of thiazide diuretics and risk of hip fracture. Lancet 1, 687 – 690 (1989). 440. R. M. C. Herings, B. H. C. Stricker, A. de Boer, A. Bakker, F. Sturmans, and A. Stergachis, Current use of thiazide diuretics and prevention of femur fractures. J. Clin. Epidemiol. 49, 115 – 119 (1996). 441. G. Jones, T. Nguyen, P. N. Sambrook, and J. A. Eisman, Thiazide diuretic and fractures: Can meta-analysis help? J. Bone Miner. Res. 10, 106 – 111 (1995). 442. J. Reeve, P. J. Neunier, J. A. Parsons, M. Bernat, O. L. M. Bijvoet, P. Courpron, C. Edouard, L. Klenerman, R. M. Neer, J. C. Renier, D. Slovik, F. J. F. E. Vismans, and J. R. J. Potts, Anabolic effect of human parathyroid hormone fragment on trabecular bone in involutional osteoporosis: A multicentre trial. Br. Med. J. 7 June, 1340 – 1344 (1980). 443. D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. R. J. Potts, M. A. Daly, J. A. Campbell, and R. M. Neer, Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1 – 34) and 1,25-dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 381 (1986). 444. J. Reeve, J. N. Bradbeer, M. Arlot, U. M. Davies, J. R. Green, L. Hampton, C. Edouard, R. Hesp, P. Hulme, J. P. Ashby, J. M. Zanelli, and P. J. Meunier, hPTH 1 – 34 treatment of osteoporosis with added hormone replacement therapy: Biochemical, kinetic and histological responses. Osteoporosis Int. 1, 162 – 170 (1991). 445. R. Marcus and A. R. Hoffman, Growth hormone as therapy for older men and women. Annu. Rev. Pharmacol. Toxicol. 38, 45 – 61 (1998). 446. D. Rudman, A. G. Feller, S. Hoskote, S. Nagraj, G. A. Gergans, P. Y. Lalitha, A. F. Goldberg, R. A. Schlenker, L. Cohn, I. W. Rudman, and D. E. Mattson, Effects of human growth hormone in men over 60 years old. N. Engl. J. Med. 323, 1 – 6 (1990). 447. H. Kotzmann, M. Riedl, P. Bernecker, M. Clodi, F. Kainberger, A. Kaider, W. Woloszczuk, and A. Luger, Effects of long term growth hormone substitution therapy on bone mineral density and parameters of bone metabolism in adult patients with growth hormone deficiency. Calcif. Tissue Int. 62, 40 – 46 (1998). 448. R. Marcus, Skeletal effects of growth hormone and IGF-1 in adults. Endocrine 7, 53 – 55 (1997). 449. J. P. T. Span, G. F. F. M. Pieters, C. G. J. Sweep, A. R. M. M. Hermus, and A. G. H. Smals, Gender differences in insulin-like growth factor I response to growth hormone (GH) treatment in GHdeficient adults: Role of sex hormone replacement. J. Clin. Endocrinol. Metab. 85, 1121 – 1125 (2000). 450. D. Rudman, A. G. Feller, L. Cohn, K. R. Shetty, I. W. Rudman, and M. W. Draper, Effects of human growth hormone on body composition in elderly men. Horm. Res. 36 (Suppl. 1), 73 – 81 (1991). 451. M. A. Papadakis, D. Grady, D. Black, M. J. Tierney, G. A. W. Gooding, M. Schambelan, and C. Grunfeld, Growth hormone replacement in healthy older men improved body composition but not functional ability. Ann. Intern. Med. 124, 708 – 716 (1996). 452. D. R. Taaffe, L. Pruitt, J. Reim, R. L. Hinta, G. Butterfield, A. R. Hoffman, and R. Marcus, Effect of recombinant human growth hormone on the muscle strength response to resistance exercise in elderly men. J. Clin. Endocrinol. Metab. 79, 1361 – 1366 (1994). 453. H. B. A. Baum, B. M. K. Biller, J. S. Finkelstein, K. B. Cannistraro, D. S. Oppenheim, D. A. Schoenfeld, T. S. Michel, H. Wittink, and A. Klibanski, Effects of physiologic growth hormone therapy on bone density and body composition in patients with adult-onset growth hormone deficiency. Ann. Intern. Med. 125, 883 – 890 (1996). 454. Y. J. H. Janssen, N. A. T. Hamdy, H. M. Frolich, and F. Roelfsema, Skeletal effects of two years of treatment with low physiological doses of recombinant growth hormone (GH) in patients with adult-onset GH deficiency. J. Clin. Endocrinol. Metab. 83, 2143 – 2148 (1998). 455. A. G. Johansson, E. Lindh, W. F. Blum, G. Kollerup, O. H. Sorensen, and S. Ljunghall, Effects of growth hormone and insulin-

456.

457.

458.

459.

460.

461.

462.

463.

464.

465.

466. 467.

468.

469.

470.

471.

472.

like growth factor I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 81, 44 – 48 (1996). K. Brixen, T. B. Hansen, E. Hauge, N. Vahl, J. O. L. Jorgensen, J. S. Christiansen, L. Mosekilde, C. Hagen, and F. Melsen, Growth hormone treatment in adults with adult-onset growth hormone deficiency increases iliac crest trabecular bone turnover: A 1-year, double-blind, randomized, placebo-controlled study. J. Bone Miner. Res. 15, 293 – 300 (2000). C. Y. C. Pak, K. Sakhaee, V. Piziak, R. D. Peterson, N. A. Breslau, P. Boyd, J. R. Poindexter, J. Herzog, A. Heard-Sakhaee, S. Haynes, B. Adams-Huet, and J. S. Reisch, Slow-release sodium fluoride in the management of postmenopausal osteoporosis: A randomized controlled trial. Ann. Int. Med. 120, 625 – 632 (1994). J. D. Ringe, A. Dorst, C. Kipshoven, L. C. Rovati, and I. Setniker, Avoidance of vertebral fractures in men with idiopathic osteoporosis by a three year therapy with calcium and low-dose intermittent monofluorophosphate. Osteoporosis Int. 8, 47 – 52 (1998). L. Z. Rubenstein and K. R. Josephson, Causes and prevention of falls in elderly people. In “Falls, Balance and Gait Disorders in the Elderly” (B. Vellas, M. Toupet, L. Rubenstein, J. L. Albarede and Y. Christen, eds.), Elsevier, Paris, 1992. M. A. Fiatarone, E. O’Neill, N. D. Ryan, K. M. Clements, G. R. Solares, M. E. Nelson, S. B. Roberts, J. J. Keyhayias, L. A. Lipsitz, and W. J. Evans, Exercise training and nutritional supplementation for physical frailty in very elderly people. N. Engl. J. Med. 330, 1769 – 1775 (1994). M. E. Tinetti, D. I. Baker, G. McAvay, E. B. Claus, P. Garrett, M. Gottschalk, M. L. Koch, K. Trainor, and R. I. Horwitz, A multifactorial intervention to reduce the risk of galling among elderly people living in the community. N. Engl. J. Med. 331, 821 – 827 (1994). B. Dawson-Hughes, Calcium and vitamin D Nutrition. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 197 – 209. Academic Press, San Diego, 1999. P. Lips, W. C. Graafmans, M. E. Ooms, P. D. Bezemer, and L. M. Bouter, Vitamin D supplementation and fracture incidence in elderly persons. Ann. Intern. Med. 124, 400 – 406 (1996). W. Owusu, W. C. Willett, D. Feskanich, A. Ascherio, D. Spiegelman, and G. A. Colditz, Calcium intake and the incidence of forearm and hip fractures among men. J. Nutr. 127, 1782 – 1787 (1997). B. Dawson-Hughes, S. S. Harris, E. A. Krall, and G. E. Dallal, Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. N. Engl. J. Med. 337, 670 – 702 (1997). NIH “NIH Consensus Statement,” Vol. 12, 4th ed. National Institutes of Health, Bethesda, MD. M. K. Thomas, D. M. Lloyd-Jones, R. I. Thadhani, A. C. Shaw, D. J. Deraska, B. T. Kitch, E. C. Vamvakas, I. M. Dick, R. L. Prince, and J. S. Finkelstein, Hypovitaminosis D in medical inpatients. N. Engl. J. Med. 338, 777 – 783 (1998). R. P. J. van der Wielen, M. R. H. Lowik, H. van den Berg, L. C. P. G. M. de Groot, J. Haller, O. Moreiras, and W. A. van Staveren, Serum vitamin D concentrations among elderly people in Europe. Lancet 346, 207 – 10 (1995). E. E. Delvin, A. Imbach, and M. Copti, Vitamin D nutritional status and related biochemical indices in an autonomous elderly population. Am. J. Clin. Nutr. 48, 373 – 378 (1988). P. Lips, F. C. van Ginkel, M. J. M. Jongen, F. Rubertus, W. J. F. van der Vijgh, and J. C. Netelenbos, Determinants of vitamin D status in patients with hip fracture and in elderly control subjects. Am. J. Clin. Nutr. 46, 1005 – 1010 (1987). B. Dawson-Hughes, S. E. Fowler, G. Dalsky, and C. Gallagher, Sodium excretion influences calcium homeostasis in elderly men and women. J. Nutr. 126, 2107 – 2112 (1996). G. C. Curhan, W. C. Willett, E. B. Rimm, and M. J. Stampfer, A prospective study of dietary calcium and other nutrients and the

148

473.

474.

475.

476.

477.

478.

479.

480.

481.

482.

483.

484.

485.

486.

487.

488.

489.

490.

491.

ORWOLL AND KLEIN risk of symptomatic kidney stones. N. Engl. J. Med. 328, 833 – 838 (1993). E. Nordborg, T. Hansson, and R. Jonson, Glucocorticosteroidinduced osteoporosis in giant-cell arteritis (GCA). Scand. J. Rheum. 85, 7 (1990). O. S. Gluck, W. A. Murphy, and T. J. Hahn, Bone loss in adults receiving alternate day glucocorticoid therapy: A comparison with daily therapy. Arthritis. Rheum. 24, 833 – 838 (1985). I. R. Reid and H. K. Ibbertson, Calcium supplements in the prevention of steroid-induced osteoporosis. Am. J. Clin. Nutr. 44, 287 – 290 (1986). J. S. Adams, T. O. Wahl, and B. P. Luker, Effects of hydrochlorothiaize and dietary sodium restriction on calcium metabolism in corticosteroid treated patients. Metabolism 30, 217 – 221 (1981). Y. Suzuki, Y. Ichikawa, and E. Saito, Importance of increased urinary calcium excretion in the development of secondary hyperparathyroidism of patients under glucocorticoid therapy. Metabolism 151 – 156 (1983). D. D. Bikle, B. Halloran, and L. Fong, Elevated 1,25-dihydroxyvitamin D levels in patients with chronic obstructive pulmonary disease treated with prednisone. J. Clin. Endocrinol. Metab. 76, 456 – 461 (1993). J. R. Condon, J. R. Nassim, and C. E. Dent, Possible prevention and treatment of steroid-induced osteoporosis. Postgrad. Med. J. 54, 249 – 252 (1978). J. J. Braun, D. H. Birkenhager-Frenkel, and A. H. Reitveld, Influence of 1-(OH)D3 administration on bone mineral metabolism in patients on chronic glucocorticoid treatment: A double-blind controlled study. Clin. Endocrinol. (Oxf) 19, 265 – 273 (1983). T. R. Dykman, K. M. Haralson, and O. S. Gluck, Effect of oral 1,25-dihydroxyvitamin D and calcium on glucocorticoid-induced osteopenia in patients with rheumatic disease. Arthritis Rheum. 27, 1336 – 1341 (1984). J. W. Bijlsma, J. A. Raymakers, and C. Mosch, Effect of oral calcium and vitamin D on glucocorticoid-induced osteopenia. Clin. Exp. Rheumatol. 6, 113 – 119 (1988). H. Rickers, A. Deding, and C. Christiansen, Mineral loss in cortical and trabecular bone during high-dose prednisone treatment. Calcif. Tissue Int. 36, 269 – 273 (1984). P. Sambrook, J. Birmingahm, and P. Kelly, Prevention of corticosteroid osteoporosis - a comparison of calcium, calcitriol and calcitonin. N. Engl. J. Med. 328, 1747 – 1752 (1993). M. S. Schwartzman and W. A. Frank, Vitamin D toxicity complicating the treatment of senile, postmenopausal. Am. J. Med. 82, 224 – 230 (1987). P. J. Meunier, D. Birancon, and P. Chavassieux, Treatment with fluoride: Bone histomorphometric findings. In “Osteoporosis” (C. Christiansen, J. S. Johansen, and B. J. Riis, ed.), pp. 824 – 828. Osteopress, Copenhagen, Denmark, 1987. L. Montemurro, G. Schiraldi, P. Fraioli, G. Tosi, A. Riboldi, and G. Rizzato, Prevention of corticosteroid-induced osteoporosis with salmon calcitonin in sarcoid patients. Calcif. Tissue Int. 49, 71 – 76 (1991). I. R. Reid, C. J. Alexander, A. R. King, and H. K. Ibbertson, Prevention of steroid-induced osteoporosis with (3-amino-1-hydroxypropylidene)-1, 1-bisphosphonate(APD). Lancet January 23, 143 – 146 (1988). M. Luengo, C. Picado, and L. Del Rio, Treatment of steroid-induced osteoporosis with calcitonin in corticosteroid-dependent asthma. Am. Rev. Respir. Dis. 142, 104 – 107 (1990). T. Nishioka, H. Kurayama, and T. Yasuda, Nasal administration of salmon calcitonin for prevention of glucocorticoid-induced osteoporosis in children with nephrosis. J. Pediatr. 118, 703 – 707 (1991). J. R. Reid, S. W. Heap, and A. R. King, Two-year followup of bisphosphonate (APD) treatment in steroid osteoporosis. Lancet 2, 1144 (1988).

492. J. S. Finkelstein, A. Klibanski, R. M. Neer, S. H. Doppelt, D. I. Rosenthal, G. V. Serge, and W. F. Crowley, Increases in bone density during treatment of men with idiopathic hypogonadotropic hypogonadism. J. Clin. Endocrinol. Metab. 69, 776 – 783 (1989). 493. E. S. Orwoll and R. F. Klein, Osteoporosis in men. Endocrine. Rev. 16, 87 – 116 (1995). 494. L. Katznelson, J. S. Finkelstein, D. A. Schoenfeld, D. I. Rosenthal, E. J. Anderson, and A. Klibanski, Increase in bone density and lean body mass during testosterone administration in men with acquired hypogonadism. J. Clin. Endocrinol. Metab. 81, 4358 – 4365 (1996). 495. H. M. Behre, S. Kliesch, E. Leifke, T. M. Link, and E. Nieschlag, Long-term effect of testosterone therapy on bone mineral density in hypogonadal men. J. Clin. Endocrinol. Metab. 82, 2386 – 2390 (1997). 496. J. E. Morley, H. M. Perry, III, F. E. Kaiser, D. Kraenzel, J. Jensen, K. Houston, M. Mattammal, and H. M. Perry, Jr., Effects of testosterone replacement therapy in old hypogonadal males: A preliminary study. J. Am. Geriatr. Soc. 41, 149 – 152 (1993). 497. C. Wang, D. R. Eyre, D. Clark, C. Kleinberg, C. Newman, A. Iranmanesh, J. Veldhuis, R. E. Dudley, N. Berman, T. Davidson, T. J. Barstow, R. Sinow, G. Alexander, and R. S. Swerdloff, Sublingual testosterone replacement improves muscle mass and strength, decreases bone resorption, and increases bone formation markers in hypogonadal men-A clinical research center study. J. Clin. Endocrinol. Metab. 81, 3654 – 3662 (1996). 498. C.-Y. Guo, H. Jones, and R. Eastell, Treatment of isolated hypogonadotropic hypogonadism effect on bone mineral density and bone turnover. J. Clin. Endocrinol. Metab. 82, 658 – 665 (1997). 499. C. Wang, G. Alexander, N. Berman, B. Salehian, T. Davidson, V. McDonald, B. Steiner, L. Hull, C. Callegari, and R. S. Swerdloff, Testosterone replacement therapy improves mood in hypogonadal men-A clinical research center study. J. Clin. Endocrinol. Metab. 81, 3578 – 3583 (1996). 500. R. Sih, J. E. Morley, F. E. Kaiser, H. M. I. Perry, P. Patrick, and C. Ross, Testosterone replacement in older hypogonadal men: A 12month randomized controlled trial. J. Clin. Endocrinol. Metab. 82, 1661 – 1667 (1997). 501. A. Kubler, G. Schulz, U. Cordes, J. Beyer, and U. Krause, The influence of testosterone substitution on bone mineral density in patients with Klinefelter’s syndrome. Exp. Clin. Endocrinol. 100, 129 – 132 (1992). 502. F. H. W. Wong, K. K. Pun, and C. Wang, Loss of bone mass in patients with Klinefelter’s syndrome despite sufficient testosterone replacement. Osteoporosis Int. 7, 281 – 287 (1993). 503. R. S. Swerdloff and C. Wang, Androgen deficiency and aging in men. West J. Med. 159, 579 – 585 (1993). 504. F. H. Anderson, R. M. Francis, R. T. Peaston, and H. J. Wastell, Androgen supplementation in eugonadal men with osteoporosis: Effects of six months’ treatment on markers of bone formation and resorption. J. Bone Miner. Res. 12, 472 – 478 (1997). 505. O. Arisaka, M. Arisaka, Y. Nakayama, S. Fujiwara, and K. Yabuta, Effect of testosterone on bone density and bone metabolism in adolescent male hypogonadism. Metabolism 44, 419 – 423 (1995). 506. S. Bertelloni, G. I. Baroncelli, R. Battini, G. Perri, and G. Saggese, Short-term effect of testosterone treatment on reduced bone density in boys with constitutional delay of puberty. J. Bone Miner. Res. 10, 1488 – 1495 (1995). 507. S. W. J. Lamberts, A. W. van den Beld, and A. van der Lely, The endocrinology of aging. Science 278, 419 – 424 (1997). 508. A. Gray, H. A. Feldman, J. B. McKinlay, and C. Longcope, Age, disease and changing sex hormone levels in middle-aged men: Results of the Massachusetts male aging study. J. Clin. Endocrinol. Metab. 73, 1016 – 1025 (1991). 509. I. R. Reid, D. J. Wattie, M. C. Evans, and J. P. Stapleton, Testosterone therapy in glucocorticoid-treated men. Arch. Intern. Med. 156, 1173 – 1177 (1996).

CHAPTER 42 Osteoporosis in Men 510. H. Z. Ke, H. Qi, D. T. Crawford, K. L. Chidsey-Frink, H. A. Simmons, and D. D. Thompson, Lasofoxifene (CP-336,156), a selective estrogen receptor modulator, prevents bone loss induced by aging and orchidectomy in the adult rat. Endocrinology 141, 1338 – 1344 (2000). 511. A. Negro-Vilar, Selective androgen receptor modulators (SARMs): A novel approach to androgen therapy for the new millennium. J. Clin. Endocrinol. Metab. 84, 3459 – 3462 (1999). 512. L. J. Donaldson, A. Cook, and R. G. Thomson, Incidence of fractures in a geographically defined population. J. Epi. Commun. Health 44, 241 – 245 (1990). 513. C. Cooper and L. J. I. Melton, Epidemiology of osteoporosis. Trends Endocrinol. Metab. 3, 224 – 229 (1992). 514. S. J. Jacobson, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Race and sex differences in mortality following fracture of the hip. Am. J. Public Health 82, 1147 – 1150 (1992). 515. V. Gilsanz, M. I. Boechat, R. Gilsanz, M. L. Loro, T. F. Roe, and W. G. Goodman, Gender differences in vertebral sizes in adults: Biomechanical implications. Radiology 190, 678 – 682 (1994).

149 516. G. Theintz, G. Buchs, R. Rizzoli, D. Slosman, H. Clavien, P. C. Sizonenko, and J.-P. Bonjour, Longitudinal monitoring of bone mass accumulation in healthy adolescents: Evidence for a marked reduction after 16 years of age at the levels of lumbar spine and femoral neck in female subjects. J. Clin. Endocrinol. Metab. 75, 1060 – 1065 (1992). 517. J. A. Kanis and S. Adami, Bone loss in the elderly. Osteoporosis Int. 1, 59 – 65 (1994). 518. E. S. Orwoll, The clinical evaluation of osteoporosis in men. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 527 – 552. Academic Press, San Diego, 1999. 519. E. S. Orwoll, The prevention and therapy of osteoporosis in men. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 553 – 569. Academic Press, San Diego, 1999. 520. E. S. Orwoll, R. M. Weigel, S. K. Oviatt, M. R. McClung, and L. J. Deftos, Calcium and cholecalciferol: Effects of small supplements in normal men. Am. J. Clin. Nutr. 48, 127 – 130 (1988).

CHAPTER 43

Osteoporosis in Childhood and Adolescence LAURA K. BACHRACH

I. II. III. IV.

Department of Pediatrics, Stanford University School of Medicine, Stanford, California 94305

Introduction Anorexia Nervosa Reproductive Disorders Other Endocrine Disorders

V. Systemic Disease VI. Idiopathic Juvenile Osteoporosis VII. Summary and Future Directions References

I. INTRODUCTION

mineral. Heritable factors account for an estimated 60 – 80% of the variability in peak bone mass [3,4]. The genes determining osteoporotic risk remain elusive and gene therapy for those at greatest risk is not available. The remaining differences in bone mass are explained by modifiable factors including body mass, nutrition, physical activity, and endocrine function [5 – 9]. In healthy youth, manipulation of these factors has proven successful in augmenting bone mineral accrual. Increasing calcium intake or weightbearing physical activity has been shown to increase rates of bone mineral acquisition by 1 – 10% in short term studies [10 – 13]. If sustained, gains in peak bone mass of this magnitude might reduce the incidence of osteoporosis. The risk of fragility fracture has been estimated to decrease by 40% for each gain of 5% in peak bone mass [14]. For children with a variety of chronic disorders, preserving bone health is a more challenging and immediate concern. Bone mineral accrual may be compromised by malnutrition, reduced physical activity, endocrine dysfunction, cytokine overproduction, or exposure to medications that interfere with bone metabolism. Without recognition and treatment, patients with these disorders are at risk for osteopenia (low bone mineral for age) or osteoporosis

Osteoporosis is recognized increasingly as a pediatric concern. This interest reflects the growing awareness that the bone mineral accrued by early adulthood is a major determinant of the lifetime risk of bone fragility. The magnitude of peak bone mass, achieved by early adulthood, explains an estimated 60% of the risk of osteoporosis, with subsequent bone loss accounting for the remainder [1]. Adolescence is a particularly critical period for establishing bone health. The bone mineral acquired normally during the teen years typically equals the amount of bone mineral lost throughout adult life [2]. The tempo and determinants of bone growth and mineral acquisition from infancy through adolescence are described in detail in Chapters 24 and 25. This chapter will focus on several of the disorders of childhood that may compromise these critical gains. Discussion will be limited to the common causes of acquisitional osteopenia, a term derived to emphasize failure of bone accrual. Genetic disorders of bone are reviewed in Chapter 50 and osteopenia of the premature infant is discussed in Chapter 24. To address skeletal disorders of childhood and adolescence requires an understanding of the determinants of bone

OSTEOPOROSIS, SECOND EDITION VOLUME 2

151

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

152 (a more profound reduction in bone mineral with disruption of microarchitecture sufficient to permit atraumatic fractures). Chronic malnutrition also may result in osteomalacia, in which mineralization of the osteoid is reduced. The risk of childhood osteopenias is influenced by the age at onset and duration of illness and by the number of associated risk factors. The timing of illness is important because bone growth and mineral accrual proceed at variable rates throughout the skeleton [15]. A greater proportion of appendicular growth is completed before puberty, whereas more spinal growth occurs during adolescence under the influence of sex steroids. Trabecular bone acquisition is completed before cortical bone. The severity and site of bone mineral deficits will reflect the developmental stage when illness begins and the length of time that accrual is interrupted. Bone densitometry, biochemical bone markers, and fracture rates have been employed to evaluate skeletal health in children although they are more complex to interpret than in adults. Dual energy X-ray absorptiometry (DXA) is generally considered the method of choice for assessing bone mass because of its speed, precision, availability, and low radiation exposure [16]. Although most DXA software programs include limited pediatric norms, reference data for Hologic [17,18], Lunar [19,20], and other equipment have been published. The limitations of quantitative computed tomography for children include a 10- to 100-fold greater radiation dose [16] and fewer pediatric reference data [21] than seen with DXA. Quantitative ultrasound (QUS) of the heel, tibia, patella, and phalanges is being tested in children because this technique is rapid, inexpensive, and free of ionizing radiation [16,22]. QUS remains less attractive than DXA at present because ultrasound is less precise and pediatric norms are more limited [22]. Interpretation of DXA results is far more challenging in children than in adults because of the changes in bone size and geometry that accompany growth and pubertal development. In adults, observed bone mineral is generally reported as a t score, the standard deviations from an expected reference value for healthy young adults. In childhood, the definition of “normal” is a moving target. Appropriate interpretation of the data must consider pubertal or skeletal maturation, ethnicity, and body size of the patient. Bone mineral accrual is highly correlated with peak height velocity [2] and puberty. Variability in timing of these events creates a broad spectrum of normal values for bone size and mineral throughout adolescence. For this reason, bone mineral data from pediatric patients should be compared with reference data from youth at a similar stage of maturity. On average, males mature later than females and failure to use gender-specific norms can result in an overestimation of “osteopenia” in males [23]. Blacks have greater bone mass than nonblacks [18,21] and ethnic-specific data should be used if available.

LAURA K. BACHRACH

Correcting for bone size remains the most controversial and challenging aspect in the analysis of DXA data. Bone mass is reported in terms of bone mineral content (BMC, g) and density (BMD, g/cm2), both of which are influenced by bone size. Although BMD adjusts for the area of bone scanned, it does not correct for differences in bone thickness. As a result, true bone density is underestimated in smaller bones and overestimated in larger ones. To address this problem, several models have been developed to estimate bone volume [20,24,25], which correlate with true bone volume in vitro [20]. To correct whole body bone mineral content for bone size, Molgaard et al. have suggested taking both height and bone area into consideration [27]. Despite the lack of consensus about the best method to adjust for bone size, skeletal age, and pubertal stage, these factors must be considered when interpreting pediatric densitometry studies. The interpretation of bone markers is also more complex because concentrations vary during normal development. Mean values and interindividual variability for both formation and resorption markers are several-fold higher in children than in adults [28,29]. Biochemical markers reach maximal values in early adolescence (Tanner stage II) and decline thereafter, despite continued gains in bone size and mineral density [28,29]. In fact, concentrations of formation markers correlate inversely with bone density while resorption markers are inversely related to bone area and spine bone density [29]. The broad range of normal values and the need to adjust for developmental stage limit the value of bone markers in defining turnover as normal or abnormal. Use of other skeletal health indicators has been hampered by a paucity of pediatric reference data. The recent publication of histomorphometry parameters from 58 healthy children will be useful in interpreting bone biopsy data in the future [29a]. To date, the degree of osteopenia that defines short-term fracture risk has not been established for children or adolescents. This reflects the fact that most studies of pediatric osteopenias have lacked the power to determine if the number of fractures exceeded the expected incidence for gender and age [30]. Treatment for acquisitional osteopenia is directed at correcting the nutritional, activity, and hormonal risk factors contributing to the deficits. Beyond these general measures, there is a limited therapeutic armamentarium for treating childhood osteopenia. The safety or efficacy of agents used to treat osteoporosis in adults have not been established in children. Furthermore, most available agents act by inhibiting resorption rather than stimulating bone formation. Although increased bone loss may occur in some disorders, all childhood osteopenias represent a failure of acquisition. A safe and effective anabolic agent is needed to reverse this process. The paucity of therapeutic options increases the importance of early recognition and correction of risk factors.

CHAPTER 43 Osteoporosis in Childhood and Adolescence

II. ANOREXIA NERVOSA Osteopenia occurs commonly in anorexia nervosa, a psychiatric condition of self-induced malnutrition and amenorrhea. Mean BMD for young adult patients has been reported to be 1 SD below normal, but bone densities below -2 SD have been reported in up to 50% of patients in some series [31]. In younger patients, osteoporosis is frequent and may occur early in the course of the illness. One study found that 12 of 18 adolescents (mean age 16) had BMD values more than 2 S.D. below expected; half of those with significant osteopenia had been diagnosed within the previous 1.5 years [32]. Osteopenia cannot be dismissed as artifact of small bone size, because reduced volumetric bone density has been observed by quantitative computed tomography (QCT) [33]. The increased incidence of pathologic fractures in chronic anorexia nervosa is further evidence of reduced bone mass [31]. The frequency and severity of osteopenia in anorexia nervosa may be explained by the timing of onset and myriad risk factors. This eating disorder begins typically in mid-adolescence, a critical period for bone mineral accrual. Intake of calories, protein, and calcium is inadequate and deficits may be compounded by increased urinary calcium loss [31,34]. Hypogonadotropic hypogonadism and estrogen deficiency develop secondary to hypothalamic dysfunction and malnutrition. Elevated endogenous glucocorticoid production and abuse of cathartics and diuretics may contribute further to osteopenia [31,35]. Correcting the nutritional deficiencies effectively reduces osteopenia in patients with anorexia nervosa. Measurable increases in BMD accompany weight gain in most [31,36,37] but not all studies [38]. Changes in BMD with weight gain must be interpreted with caution because DXA measurements of bone mass are influenced by changes in body composition; BMD increases an estimated 5% for each centimeter increase in local fat mass [39]. The benefits of weight rehabilitation for osteopenia cannot be dismissed as artifact, however, because gains in bone mass have been documented by QCT as well [33]. Furthermore, bone histomorphometry indicated normal turnover rates in two teens who maintained body weights at or above 84% of expected [40]. Since gains in bone mineral precede the return of menses, weight rehabilitation per se appears to be effective therapy for osteopenia. The value of other general therapeutic measures is less certain. Calcium supplementation has not proven sufficient to prevent osteopenia [32,38,41]. It is appropriate, however, to ensure the recommended daily intake of 1300 mg [42]. Physical activity has been found to protect against osteopenia at weight-bearing [41,43] and non-weight-bearing sites [44] in some studies while others have failed to confirm this benefit [32]. Until further data are available, it seems reasonable to avoid extremes of physical activity. Bed rest is

153 likely to contribute to bone loss while intensive activity increases the risk of pathological fractures, slower weight gain, and prolonged amenorrhea. The greatest controversy surrounds the value of pharmacologic agents for osteopenia in anorexia nervosa. The safety and efficacy of bisphosphonates has not been established and the effects of estrogen therapy on bone are debated. Estrogen therapy failed to improve bone mass [45,46] and to restore bone apposition to normal [40] in several studies. By contrast, Seeman et al. observed that patients receiving oral contraceptives had greater spine BMD than patients not supplemented with estrogen, although the bone mass in both groups remained significantly below normal for age [41]. Estrogen therapy had no protective effect at the femoral neck [41]. In the only prospective, randomized study to date, estrogen-treated and control subjects had similar changes in vertebral bone density (measured by QCT) after 1.5 years (Fig. 1) [33]. Sex hormone therapy provided significant protection against further bone loss only in patients weighing less than 70% of ideal. These data suggest that estrogen replacement therapy is not a substitute for nutritional support, although this treatment may be a valuable adjunct in severely malnourished patients. Use of hormone therapy has the disadvantage of masking the return of spontaneous menses, an indication of clinical recovery. Furthermore, estrogen accelerates epiphyseal closure and should be avoided in young patients who have not reached final height. Even with effective therapy, recovery from osteopenia may be incomplete. In short-term longitudinal studies, reduced spine BMD and increased fracture rates persisted in many patients despite weight gain, return of menses, sex steroid or calcium supplementation, and physical activity

FIGURE 1 Change in vertebral trabecular bone mineral density (mg/cc of K2HPO4) in young adult women with anorexia nervosa without (control) or with hormone replacement therapy (estrogen). Shaded area represents the normal mean BD  1 SD. Modified with permission from Klibanski et al. [33], © The Endocrine Society.

154

LAURA K. BACHRACH

[33,36,38]. The largest long-term study to date examined bone mass of 51 women an average of 11.7 years after their first admission for anorexia nervosa [37]. Mean radial and lumbar spine BMD was 2.18 and 1.73 SD below expected respectively in patients who continued to have malnutrition and amenorrhea. The skeletal status of women who had recovered from their eating disorder was better, but mean spine and radial BMD Z-scores remained 0.26 and 0.65, respectively. Whether earlier recognition or more aggressive nutritional and/or hormonal support can restore peak bone mass to normal remains to be determined.

III. REPRODUCTIVE DISORDERS Estrogen plays an essential role in the completion of bone maturation and mineral accrual in both males and females [9]. Alteration in the timing or magnitude of sex steroid production during adolescence may have transient or sustained effects on bone mass. The deleterious effects of estrogen deficiency may be compounded by malnutrition.

A. Exercise-Associated Amenorrhea The relationship between physical activity and skeletal health appears to be a double-edged sword. Weight-bearing exercise enhances acquisition of bone mineral activity and may influence bone size in children and adolescents [2,6,8,12,13]. Intensive activity can reach a point of diminishing returns, however, especially if accompanied by marginal nutrition, low body mass, and reproductive dysfunction [47 – 49]. Low bone mass and increased fracture rates may result [47,50]. The risk of osteopenia varies by skeletal site and type of activity. For distance runners, osteopenia occurs most commonly at the lumbar spine and hip but develops in the appendicular skeleton as well [48,51]. By contrast, bone mass in dancers [52] and gymnasts [53] may be preserved at weight bearing sites despite low body mass and menstrual disorders. For prepubertal gymnasts, annual gains in BMD are greater than in controls and this increased bone mass is sustained as long as 20 years following retirement [54]. These findings suggest that more intensive mechanical loading may counteract the adverse effects of hypogonadism, at least at some sites. Intensive training has also been linked to reduced bone growth in younger athletes. Theintz et al. observed that teen-aged gymnasts had shortened leg lengths and failed to exhibit the expected growth spurt during puberty [55]. They predicted that continuing gymnastics through adolescence would reduce final adult height. These conclusions have been challenged by Bass et al. who found that leg

length was reduced in girls at the time they began the sport, suggestive that this finding reflected a selection bias [56]. Upper segment (spinal) growth was slower in active gymnasts than in controls but there was evidence of catch-up growth in this region after retirement from the sport. As in anorexia nervosa, both nutritional and hormonal factors have been implicated as causes for osteopenia [49]. In fact, the incidence of overt eating disorders among intensive athletes is increased [57]. Intake of calories, fat, protein, and calcium may be inadequate in athletes who restrict their diet to maintain the svelte physique desirable for their sport [49]. Increased concentrations of endogenous glucocorticoids have been observed in some cases [49]. Intensive training during childhood or adolescence may be associated with delayed onset and completion of puberty. This may reflect a selection bias favoring dancers, gymnasts, or runners who mature later [56]. Alternatively, each year of intensive training may delay menarche by 5 months [58]. For elite athletes who begin their sport after menarche, an estimated 5 – 50% will develop secondary oligomenorrhea or amenorrhea [58]. Milder menstrual disturbances, such as a shortened luteal phase, may also develop but not be clinically apparent [59]. Management of athletic adolescents with delayed puberty or amenorrhea should address all risk factors for osteopenia [49,60]. Alternative causes of amenorrhea including pregnancy, thyroid disorders, hyperprolactinemia, or ovarian failure should be excluded. Bone densitometry is appropriate to evaluate bone mass. Demonstration of significant osteopenia may motivate the athlete to accept therapeutic recommendations. Intervention should include dietary counseling to ensure adequate calcium and caloric intake and weight gain in underweight individuals. Reducing the intensity of activity and/or weight gain are likely to restore normal reproductive function within months [61]. As in anorexia nervosa, the efficacy of estrogen replacement to preserve or increase bone mass has not been established. Two small retrospective studies observed that amenorrheic female runners who had taken estrogen/progestins had greater gains in BMD [62] and a history of fewer stress fractures [47] than women not taking replacement therapy. A single randomized study compared changes in BMD in female runners given hormone replacement therapy and 1000 mg/day calcium, calcium alone, or no treatment [63]. After 1 year, BMD at the hip and spine was 3.9 – 5.8% higher in runners with spontaneous or hormonereplacement therapy (HRT)-induced menses than in controls. Intention to treat analysis indicated an improvement of only 1.5%, however. The spontaneous return of menses in controls and withdrawals from treatment make it more difficult to determine with certainty if estrogen therapy benefits bone health in this disorder. Even with the return of spontaneous menses, deficits may not be fully restored [64].

155

CHAPTER 43 Osteoporosis in Childhood and Adolescence

Further studies are needed to define the factors that confer risk and benefit in younger athletes.

B. Delayed and Precocious Puberty Sex steroids are required to complete epiphyseal maturation and bone mineral accrual [9]. Whether delayed or early exposure to sex hormones influences peak bone mass remains controversial. Studies by Finkelstein et al. suggest that a delay in the onset of puberty is sufficient to compromise peak bone mass [65,66]. Subjects included 23 healthy young men who had began puberty after age 15. Mean BMD at the spine, femoral neck, and radius was significantly lower than that of males with a normal tempo of puberty. Nearly half of the men with delayed puberty had BMD values which were 1 SD or more below the mean for control subjects. This deficit could not be attributed to delayed acquisition, since subjects were compared with younger controls (to control for duration of exposure to sex steroids). Furthermore, there were no significant further gains in bone mass over the subsequent 2 years [66]. Two subsequent studies have concluded that idiopathic delayed puberty does not compromise peak bone mass. Moore et al. found that spine, femoral neck, and femoral shaft bone mass in 38 men (mean age 24.5) with a history of late puberty did not differ from age-matched controls [67]. Bertelloni et al. also found no difference in final height or estimated volumetric bone density in young adult men with a history of late puberty and controls who matured at the expected age [68]. Although testosterone therapy for late puberty has been shown to stimulate short-term gains in radius BMD [69], final height [68] and peak bone mass [67,68] are not significantly different between treated and untreated males. These data suggest that the decision to treat delayed puberty with exogenous testosterone can be based upon personal preference, since there appears to be no benefit or adverse effects on bone health. Fewer data address the effects of idiopathic delayed puberty in females. In most series to date, late onset of menarche has been attributable to intensive physical training or eating disorders in which other risk factors are present [60]. In one study of young adult women, peak bone mass at the spine was found to be inversely correlated with age at menarche [70]. Data on treatment suggest that hormonal replacement for hypogonadotropic amenorrhea has marginal to no benefits for bone mineral [71]. In contrast to delayed puberty, early exposure to sex steroid results in accelerated bone growth and advanced skeletal maturation. Children with precocious puberty have significantly increased BMD for chronological age at the spine [72 – 74]; femoral neck BMD has been reported as increased [72] or normal [73]; and whole body

BMD [73] is at expected levels for chronologic age. Increased spine BMD can be attributed to the advanced skeletal maturation and increased bone size accompanying early pubertal development. BMD corrected for skeletal age is normal [72,74] or reduced [73] and estimated volumetric bone mineral density is normal [72 – 74] for chronologic age. Gonadotropin releasing hormone (GnRH) analogs are used to halt early sexual maturation and allow for greater linear growth. Because these agents cause rapid bone loss in adults [75,76], their effects of pediatric bone mass has been scrutinized. Although early studies reported a decrease in radius BMD with GnRH treatment [77], most subsequent studies have found no significant declines in absolute or age-adjusted BMD at the spine [72,73], femoral neck [72], or whole body [73]. Antoniazzi et al. found that calcium supplementation prevented a decrease in areal and volumetric spine bone mineral density during GnRH treatment [74]. Whether or not GnRH analogs are used, peak bone mass appears to be normal despite a history of central precocious puberty [78,79]. Discontinuing GnRH therapy at a bone age of 11.5 years or younger may allow significantly greater gains in spine BMD [78].

C. Turner Syndrome The skeletal health of females with Turner syndrome has been studied for more than 35 years. Despite this scrutiny, the prevalence and etiology of osteopenia in this disorder remain in question [80 – 86]. The phenotypic features of Turner syndrome include short stature, anomalies of the kidneys, heart, nails, and soft tissue, and ovarian failure. Most investigators have observed a high prevalence of osteopenia in Turner patients, estimating that as many as 60 – 80% of children and adolescents have decreased bone mass for age [80 – 86]. Mean spine and radius BMC and BMD has averaged 14 – 34% below expected in Turner subjects who have not been treated with human growth hormone (hGH) or estrogen therapy. Osteopenia has been reported at all ages, although one study found significant deficits only in Turner females older than 14 years [81]. The apparent osteopenia can be explained largely by delayed skeletal maturation and small bone size typical of Turner patients. When Turner females are matched to younger controls of similar bone age or height, differences in BMC or BMD are reduced or eliminated [80,84,86]. Estimated volumetric bone density (BMAD) in 19 adolescents (mean age 14.3 years) with Turner syndrome also did not differ from controls (Fig. 2) [84]. Despite the reassuring data on bone mass, one study found a higher than expected rate of fractures in girls with Turner

156

FIGURE 2

Spine BMD and BMAD of Turner females (solid circles) and controls (open circles) plotted by chronologic age. The bold line represents the mean values for normal females. Reprinted with permission from Neely et al. [84], © The Endocrine Society.

syndrome, suggestive that bone quality might not be normal [81]. Most Turner females are treated with recombinant human growth hormone (hGH) to augment linear growth and sex steroid therapy to promote pubertal development. Both agents have been associated with gains in bone mineral although these increases are confounded by changes in bone size and maturation. Growth hormone therapy has been shown to increase BMD in patients treated for 1 year or longer [86]. BMD remained below expected values for chronological or bone age in some [80] but not all [84] studies. Estrogen replacement without or with hGH also contributes to increased bone mineral accrual [80,84,87 – 89]. The optimal timing to initiate sex steroids has been debated. Mora et al. found that initiation of estrogen after age 12 resulted in significantly lower radius BMC than earlier replacement. However, the timing of treatment was not randomized, no pretreatment bone mineral data were available, and no correction was made for bone size in this study [85]. By contrast, two studies have concluded that there is no significant osteopenia in Turner girls treated with hGH but not given estrogen through mid-adolescence [81,84]. Peak bone mass may be reduced in Turner women despite estrogen replacement [87 – 89]. Osteopenia has been observed at the calcaneus, radius, hip and spine in many Turner adults [87 – 89]; in one study, mean spine BMC Z-scores were 4.5 and 2.3 for patients without or with estrogen

LAURA K. BACHRACH

replacement, respectively [87]. Vertebral and radius bone mass have been shown to correlate the duration of estrogen therapy but not with the age when hormone replacement was initiated [88]. Despite the apparent osteopenia, only one study found an apparent increase in the incidence of low impact fractures in adults with Turner syndrome [89]. It is possible that osteoporosis has been underreported since few centers follow Turner syndrome patients through adulthood. Alternatively, osteopenia may have been overestimated since BMD was not corrected for bone size. Alterations in bone geometry such as a shortened hip axis length might also serve to protect against fractures [90]. The failure of estrogen replacement therapy (ERT) and growth hormone to restore bone mineral to normal in some patients has led investigators to postulate that loss of X chromosome material causes an intrinsic skeletal defect [87]. Saggese et al. observed that females with Turner syndrome had a blunted rise in 1,25-dihydroxyvitamin D in the face of calcium restriction [91]. Countering the theory of a primary skeletal disorder is the observation that the severity of osteopenia in Turner adults is not different from that in other forms of primary amenorrhea [89]. Furthermore, karyotype (XO versus other partial delections) does not predict the severity of osteopenia in Turner patients [88]. Finally, Turner patients with spontaneous menarche achieve normal spine and radius bone mass [89]. Based upon current knowledge, it is prudent to optimize hormonal therapy and lifestyle to foster bone mineral accrual. Decisions regarding the timing and administration of hGH or estrogen should be based upon statural issues and personal preference. Activity and calcium intake should be encouraged as for all youth. The potential benefit of supplementation with 1,25-dihydroxyvitamin D requires further study. Finally, additional longitudinal studies of adult females with Turner syndrome to evaluate volumetric bone mass, hip axis length, and the lifetime risk of fracture are needed.

IV. OTHER ENDOCRINE DISORDERS A. Growth Hormone Deficiency Growth hormone (GH) and the insulin-like growth factors (IGFs) are essential for normal bone growth, while the contribution of these hormones to bone mineral accrual is less well defined [92 – 96]. Osteopenia has been described in children [94,96,97] and in adults [98] with GH deficiency (GHD) and GH resistance [99] and children with idiopathic short stature [100]. Deficits appear greatest in the appendicular skeleton where cortical bone predominates while the spine and hip are less affected. Osteopenia can be explained in part by the reduced bone size and delayed skeletal maturation, which are characteristic of GH deficiency. Deficits in

157

CHAPTER 43 Osteoporosis in Childhood and Adolescence

bone mass remain significant, however, after adjusting for bone age and size in GH-deficient patients [94]. By contrast, volumetric bone density and bone histomorphology appear to be normal in GH-resistant adults [99]. The osteopenia seen in GHD patients has been attributed to inadequate bone mineral accrual rather than increased bone loss [101]. This conclusion is consistent with the proposed mechanisms of action for GH. In vitro, GH appears to act directly by stimulating osteoblast proliferation and production of bone matrix factors [92,93]. GH also stimulates bone accrual indirectly by enhancing calcium absorption. GH increases renal hydroxylation of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D and increases the sensitivity of intestinal cells to parathyroid hormone [94,95]. Therapy with hGH stimulates bone mineral accretion if given at an adequate dose and frequency. When hGH is administered 6 – 7 days a week at a dose of 0.5 – 0.6 IU/kg (0.16 – 0.20mg/kg), deficits in bone mineral are reduced or eliminated [94,97,100]. Whether normal peak bone mass can be achieved in GHD remains questionable [101,102]. The mean BMD Z-scores of adults treated for childhood onset GHD were 1.6 and 1.2 at the lumbar spine and hip, respectively; 33% of the subjects had spine BMD Z scores more than 2 SD below expected [102]. Estimates of volumetric bone density were also significantly below expected (Z scores of 0.90 for spine and 0.74 for hip) and 14% had values more than 2 SD below normal [102]. Radius bone mass (corrected for bone width) was 20% below expected adult values as compared to a 9% reduction in spine BMD, indicating the greater effect of GHD on the appendicular skeleton. Several factors may have contributed to the persistent deficits in bone mass observed in these adults. The most likely explanation is that GH replacement was suboptimal, since the supply of GH was limited prior to the release of recombinant hGH in 1985. Alternatively, other pituitary hormone deficiencies found in more than half of the GH-deficient adults could have contributed to the osteopenia [101,102]. This explanation seems unlikely because the bone mass of patients with isolated growth hormone deficiency and those with generalized hypopituitarism did not differ significantly. Finally, GH replacement may be necessary following completion of linear growth to achieve and maintain normal peak bone mass. Several studies are underway to address the need for hGH in growth hormone-deficient adults to optimize skeletal health. Preliminary data indicate that low-dose hGH therapy in adult patients (ages 16 – 29) increases spine and forearm bone density [98].

than 50 years of study [103 – 106]. Osteopenia is observed more commonly at appendicular than at axial sites. Low forearm bone mass has been observed in 13 – 70% of young patients with IDDM and remained low after correcting for delays in skeletal maturation [104,107 – 110]. Cortical and trabecular bone mass were estimated to be 8 and 14% below expected, respectively [104]. By contrast, deficits of the axial skeleton have been mild or absent. Roe et al. found that vertebral trabecular bone mineral (measured by QCT) was similar in 48 patients with IDDM and controls matched for race, sex, and age; cortical bone density in this region was slightly (3.5%) but significantly lower in the diabetics [111]. Ponder et al. reported normal lumbar spine BMD in 56 children with IDDM [112]. Boys who had IDDM for more than 1 year had significantly reduced weight and height for age, but BMD corrected for height was normal. In summary, it would appear that mild osteopenia is a common finding in children and adolescents with IDDM, but that the magnitude of the deficit increases little or not at all over time. The clinical significance of this observation is uncertain because there is no apparent increase in fracture incidence in children or adults with diabetes [104 – 106]. The risk factors for osteopenia in IDDM remain controversial. Some studies observe lower bone mass in male [107] or female diabetics [108, 113] while most report no gender difference [104,111,112,114]. Osteopenia can be present at the time of diagnosis [104,109,110] and the deficit may worsen or remain stable over the next 2 – 5 years [104,109 – 111,113]. Metabolic control (as assessed by hemoglobin A1C) has not been shown to predict bone mineral status in several studies [111 – 113]. Animal models suggest that diabetic bone disease represents a failure of bone formation. Decreased osteoblast recruitment resulting in reduced bone formation, glycosylation of collagen, reduced GH and IGFs, and abnormalities in vitamin D production and receptors have been identified [105]. The relevance of these animal data to IDDM in children is not established, particularly since duration of illness and metabolic control failed to predict bone mineral status in a consistent manner. Reduced serum osteocalcin concentrations were observed in diabetic children at the time of diagnosis but normalized within 15 days of insulin therapy, suggesting a transient decrease in bone formation [115]. However, elevated calcitonin concentrations were found in another study, suggestive of increased bone resorption [116]. The alterations in insulin-like growth factors and their binding proteins seen in patients with IDDM warrant further investigation in the pathogenesis of diabetic bone disease.

B. Diabetes Mellitus

C. Hyperthyroidism

The prevalence and etiology of bone deficits in insulindependent diabetes (IDDM) remain unsettled despite more

Thyroid hormone excess disrupts normal mineral homeostasis and may result in significant bone loss [117 – 119].

158

LAURA K. BACHRACH

In the thyrotoxic state, accelerated bone resorption leads to increased serum levels of calcium and phosphorus. This, in turn, suppresses PTH and 1,25-dihydroxyvitamin D levels, which reduces gastrointestinal absorption and increases urinary loss of calcium. Negative calcium balance ensues and contributes to deficits in bone mass. The risk of osteoporosis increases with the severity and duration of hyperthyroidism and the presence of other risk factors, such as estrogen deficiency [117 – 119] (see also Chapter 47). Both endogenous hyperthyroidism and excessive thyroxine replacement have been associated with increased bone resorption and osteopenia in children [119 – 120]. Children with Graves disease may develop profound deficits and atraumatic fractures. The former appear to be reversible with successful anti-thyroidal drug therapy [119]. Reduced proximal forearm BMC has also been observed in children and adolescents treated with high-dose L-thyroxine (120 g/M2/day) for goiter or cancer [120]. Children with congenital hypothyroidism may also be at increased risk for osteopenia because of the relatively large doses of thyroxine recommended to foster intellectual development [121]. Until the effects of high dose thyroid hormone replacement on bone mineral accrual is examined systematically, it appears prudent to avoid suppressive doses of L-thyroxine, except in cases of thyroid cancer.

D. Glucocorticoid Excess The osteopenia associated with glucocorticoid excess is the most severe and challenging of the endocrine-induced skeletal disorders in children. The adverse effects of glucocorticoids in adults are reviewed in Chapter 44. In children, steroid-induced bone loss is compounded by inhibition of bone growth and mineral accrual [122]. Bone loss is generally most rapid in the first 6 months of therapy, but the risk of bone disease appears to increase with cumulative dose of glucocorticoids used [123,124]. Deficits in bone mineral are most dramatic in regions with increased trabecular bone content such as the hip and spine. Reliance on areal bone density (BMD) may overestimate the apparent deficit in bone mineral in these smaller bones. However, osteopenia persists after correcting for bone volume and skeletal delay [123] and the increased occurrence of low trauma fractures, most frequently in the ribs and vertebrae, confirms the abnormal nature of the bone. Avascular necrosis of the femoral head is a less common complication of excessive steroid exposure. Glucocorticoid excess contributes to osteopenia through several mechanisms [125]. High steroid concentrations interfere with both intestinal calcium absorption and renal tubular calcium reabsorption, resulting in negative calcium balance. Compensatory elevations in parathyroid hormone can ensue, stimulating increased bone modeling. The normal repair of

bone resorptive sites is hampered by the direct inhibitory effect of glucocorticoids on osteoblast function. Glucocorticoid excess may also act indirectly to reduce bone growth and mineral acquisition through inhibition of sex steroid and growth hormone production. Since Cushing’s disease is extremely rare in childhood [122], steroid-induced bone disease is usually the result of prolonged glucocorticoid therapy for asthma, rheumatologic disorders, inflammatory bowel disease, malignancies, or organ transplantation [123,126]. The adverse effects of glucocorticoids are typically compounded by additional risk factors associated with the underlying disease for which steroids are prescribed. Immobilization, poor intake or absorption of vitamin D, calcium, or other nutrients, and elevated cytokine levels are among the variables likely to contribute to osteopenia Treatment for steroid-induced osteopenia in children includes the same general measures used in adults [125]. Prompt surgery or radiation therapy reduces excessive steroid production in Cushing’s disease, leading to rapid improvement in bone mineral [122,127]. Whether bone mass returns to normal remains uncertain [127]. For the majority of patients with steroid-induced skeletal deficits, glucocorticoids are required to treat underlying disorders. In this situation, the minimally effective steroid dose should be prescribed [125]. Alternate-day steroid dosing has not proven to cause less bone loss than daily administration [128]. Inhaled glucocorticoids have fewer adverse effects on bone than oral or intravenous administration, but they can suppress bone growth and mineral acquisition and induce bone loss in high dose [129]. Newer steroid derivatives, such as deflazacort, may have bone-sparing effects in children [130]. Adequate vitamin D and calcium intake should be maintained to offset increased urinary calcium excretion. Calcium intake should meet or exceed the recommended daily allowance for age [42]. Vitamin D supplementation at 400 IU of vitamin D is generally sufficient, but the adequacy of vitamin D stores can be determined by measurement of serum 25-hydroxyvitamin D concentrations. In adults, calcitriol combined with calcium protected against steroid-induced bone loss at the spine, but not at the hip or radius [131]. Calcitriol has not been tested systematically in children. Additional risk factors associated with the underlying disease, such as decreased mobility, malnutrition, malabsorption, or sex steroid deficiency, should be addressed as part of the therapeutic plan. These general measures are frequently not enough to prevent osteopenia in the face of chronic glucocorticoid excess. Studies in adult subjects indicate that bisphosphonates effectively reduce bone loss and fracture incidence [132,133]. Unfortunately, use of bisphosphonates or other anti-resorptive agents in children has been very limited. Calcitonin has been shown to improve bone mineral status in small numbers of glucocorticoid-treated children [134]. Controlled trials of bisphosphonates for steroid-induced

CHAPTER 43 Osteoporosis in Childhood and Adolescence

osteoporosis in children are underway. To date, case reports indicate that etidronate and pamidronate have no major adverse effects in children treated for osteogenesis imperfecta [135] and a variety of other conditions [136,137]. Recombinant human growth hormone has been used with limited success in small studies to counteract growth inhibition by high-dose steroids [138]. One study found an increase in bone turnover following 1 year of hGH therapy in glucocorticoid-dependent children, but there are no long-term studies of bone mineral accrual with this therapy [139].

E. Hyperprolactinemia Hyperprolactinemia is diagnosed typically in early adulthood [140] but has been reported as a cause of growth arrest [141], delayed puberty [142,143], secondary amenorrhea [143,144], and galactorrhea [145] in childhood or adolescence. In adult patients, deficits in cortical and trabecular bone mass result from inadequate bone accrual and/or accelerated bone loss [140,146]. In some, but not all studies, bone loss progresses over time. With drug-induced or spontaneous recovery of gonadal function, hyperprolactinemic patients show gains in bone mineral, but whether bone mass can be completely restored remains controversial [140,146,147]. The bone mineral status of hyperprolactinemic patients under age 18 has not been systematically examined. However, delayed puberty, secondary hypogonadism and GH deficiency associated with this disorder add to the risk of osteopenia [141,143]. The potential for recovery may be greater if growth and pubertal development resume with bromocriptine without or with hGH [141,143]. For this reason, it is appropriate to monitor bone mineral density in younger patients with hyperprolactinemia. Medical treatment with bromocryptine without or with human growth hormone has been shown to restore growth and pubertal development [141,143]; such therapy is likely to improve osteopenia as well. The reversibility of osteopenia in younger patients with successful treatment of prolactinomas warrants further study.

V. SYSTEMIC DISEASE Osteopenia has been identified in a growing number of chronic disorders of childhood. This increase likely reflects both the increased monitoring of bone health in childhood as well as the improvement in long-term survival.

A. Cystic Fibrosis The osteopenia seen in patients with cystic fibrosis (CF) exemplifies the deleterious effect of chronic multisystem

159 disease on skeletal health. CF, the most common recessive disorder in Caucasians, results in malnutrition, respiratory acidosis, growth failure, and gonadal insufficiency. With improvement in the life expectancy in CF, osteoporosis has emerged as a common cause of morbidity [148 – 154]. Deficits in bone mineral are common and profound and appear to increase with age [152,154]. Mean BMD Z scores in children with CF approximate 1.0 at the spine, 0.7 at the proximal femur, and 0.3 at whole body; 38% of patients under age 18 have BMD values more than 2 SD below expected for age [154,155]. In adults, mean BMD scores of 2.5 at the spine and femoral neck and 2.0 for whole body have been observed [155]; 34% of adults in the largest series had Z scores of 2.0 or less [151]. Since patients with cystic fibrosis are typically smaller than healthy controls, reliance upon BMC and BMD may be misleading. However, volumetric bone density measured by QCT [151] or estimated from DXA [149,152] is also significantly lower in cystic fibrosis subjects than in controls, indicating that the observed deficits are not simply an artifact of smaller bone size. The increased prevalence of kyphosis and pathologic fractures at spine, hip and ribs provide further evidence that bone strength is decreased in CF [148 – 153,155]. The abnormalities of bone metabolism causing osteopenia in this disorder have not been clearly defined. Bone biopsies from one adult patient with a history of fractures showed severe cortical and trabecular osteopenia, with several histological features not typical for either osteoporosis or osteomalacia [156]. Additional labeled bone biopsy studies would be valuable to determine the nature of the bone disease in CF. Both inadequate bone mineral accrual and increased bone loss appear to contribute to skeletal deficits in CF. Elevated concentrations of bone resorption markers and reduced values of bone formation markers have been observed in some patients, suggesting an imbalance favoring net loss of bone mineral [148,151 – 153,157]. In the largest longitudinal study of bone mineral to date, Bhudhikanok et al. observed bone loss in some CF patients and normal to increased rates of gain for age in others (Fig. 3). However, even for those showing catch-up in bone mineral accrual, the gains were not sufficient to restore bone mass to normal [155]. Mean Z scores declined during the 17-month study for subjects under age 18, indicating a worsening of osteopenia with age. Malnutrition, endocrine dysfunction, chronic illness, and reduced activity have been linked to reduced bone mineral in cystic fibrosis [148,152,154,155]. Patients with this disorder often have inadequate intake or absorption of calories, fat, and other nutrients to meet the increased metabolic demands of the disorder. Perhaps the most compelling evidence that nutritional factors are key to bone health in CF is the observation that well-nourished patients are not osteopenic [158]. Fat malabsorption interferes with vitamin D

160

LAURA K. BACHRACH

FIGURE 3

Change in lumbar spine BMD in females and males with cystic fibrosis. Subjects receiving regular glucocorticoid therapy are designated with filled symbols. The curve represents mean BMD for age and gender in healthy control subjects. Reprinted with permission from Bhudhikanok et al. [155].

absorption and serum concentrations of 25-hydroxyvitamin D may be low despite daily vitamin D supplements of 900 IU or more [151]. Serum concentrations of 25-hydroxyvitamin D fail to correlate with bone mineral status in most studies [150 – 154]. Elevated serum parathyroid hormone concentrations may prove to be a more sensitive indicator of vitamin D insufficiency [151]. Low sex steroid values associated with delayed puberty and secondary hypogonadism may also limit bone mineral accretion or accelerate loss [149,152]. Chronic glucocorticoid therapy is an additional risk factor for decreased bone mass, although the relationship between duration and steroid dose has been variable [148,152,155]. In many [151 – 154], but not all [149] studies, the severity of pulmonary disease has been a significant predictor of bone mineral status, reflecting the adverse effects of respiratory acidosis and chronic infection. Immobilization during pulmonary exacerbations may increase bone loss; physical activity has been shown to correlate with bone density in some studies [151,155] but not in others [149,153]. Cystic fibrosis transmembrane regulator (CFTR) genotype has not been shown to predict bone mineral status [151,155], but markers of bone turnover are higher in homozygotes for the delta F508 deletion [151]. Efforts to preserve and restore bone mineral in CF must address all risk factors. Nutrition should be optimized to foster weight gain and to ensure adequate intake of calories, protein, fat, and calcium. Vitamin D supplementation should be provided at a daily dose sufficient to normalize serum 25-hydroxyvitamin D, which may exceed 900 IU [148,151]. Glucocorticoid therapy should be reduced to the minimal dose effective to control respiratory symptoms; inhaled steroids may prove to be less deleterious to bone mineral than the orally administered formulations [129,159].

Testosterone therapy should be considered in hypogonadal men and in boys with significantly delayed puberty. Sex steroid replacement may also be appropriate for amenorrheic women if such therapy is tolerated without increasing airway reactivity and secretions and exacerbating pulmonary symptoms [160]. The use of anti-resorptive and anabolic agents in the treatment of osteopenia in CF remains experimental. Bisphosphonates or calcitonin may benefit those with increased bone loss associated with chronic glucocorticoid therapy or following lung transplantation [161]. Severe bone pain has been observed in adult patients following intravenous pamidronate, an adverse effect which may limit its use [162]. However, recurrence of pain appears to diminish with subsequent treatment in most patients. An effective anabolic agent would be the treatment of choice for other patients who are not losing bone but failing to gain adequate bone mineral. Growth hormone therapy has shown to increase weight gain and growth in short-term studies, but the effect of this hormone on bone mineral accrual has not been established [163]. The cost, daily injections, and risk of glucose intolerance associated with growth hormone must be weighed when considering the potential benefit of this anabolic agent. Controlled trials to evaluate the efficacy of nutrition, activity, and pharmacological interventions are urgently needed given the frequency and morbidity of osteopenia in CF.

B. Leukemia Long-term survival after acute lymphoblastic leukemia (ALL) and other childhood malignancies has become

161

CHAPTER 43 Osteoporosis in Childhood and Adolescence

common. Unfortunately, the quality of life in survivors may be marred by number of sequelae including thyroid and growth hormone deficiencies, hypogonadism, and osteopenia [164]. Musculoskeletal pain has been reported in 36% of children at the time of diagnosis of ALL and diffuse osteopenia and vertebral fractures are observed in 1 – 24% before therapy is initiated [165 – 167]. These deficits may result from infiltration of bone by leukemic cells or from tumor cell production of humoral factors such as parathyroid hormone-related peptide, prostaglandin E, or osteoblast inhibiting factor [165,167]. Circulating concentrations of 1,25-dihydroxyvitamin D are reduced and urinary calcium increased in ALL patients with osteopenia [167]. Bone biopsies in a small number of patients showed undermineralization of bone matrix [167]. Deficits in bone mass also develop following chemotherapy and radiation therapy for ALL [168 – 170]. Posttreatment osteopenia, bone pain and pathologic fractures occur generally in the spine and distal radius. Prednisone and methotrexate may cause acute changes in bone mineral metabolism and may contribute to bone loss [168]. Radiation therapy also appears to exacerbate trabecular bone loss; Gilsanz et al. observed osteopenia of the spine only in leukemic patients who had received cranial irradiation as well as chemotherapy [169]. Calciotropic hormones and suppressed osteocalcin concentrations normalize following completion of treatment [168]. The extent to which deficits in bone mineral can be restored with remission of ALL remains uknown. Peak bone mass was significantly reduced in 31 adults treated for childhood leukemia 7 – 29 years earlier; median BMD Z scores were 1.25 for vertebral trabecular BMD (QCT), 0.74 for spinal BMD (DXA), 1.35 for radius BMC, and 0.43 for femoral neck BMD [171]. Age at diagnosis, time since treatment, and growth hormone status were not predictive of bone mass. Further research is needed to determine if permanent deficits in bone mineral can be reduced or prevented in ALL and other childhood malignancies. Posttreatment deficiencies of thyroid hormone, growth hormone, and sex steroids should be replaced but this may not be sufficient to prevent osteopenia resulting from irradiation and/or chemotherapy [171].

C. Rheumatologic Disorders Osteopenia and osteoporosis are now recognized as an important cause of morbidity in rheumatoid arthritis (JRA), systemic lupus erythematosis (SLE), dermatomyositis and other chronic rheumatologic disorders [123,172 – 177]. Decreased trabecular and cortical BMD [172 – 174,176,177] has been observed in children with these disorders; in one study, 89% of those with active JRA had radius BMD more than 2 SD below expected [177]. The risk of osteopenia

increases with the duration [174] and severity [173,177] of JRA and with use of glucocorticoid therapy [123,172 – 174, 176]. Only one of these studies [123] adjusted BMD measurements for delayed skeletal and pubertal maturation or small bone size. Nonetheless, the occurrence of atraumatic fracture confirm that osteopenia can be a source of clinical morbidity. Reduced levels of bone formation and resorption markers have been observed in children with active JRA treated with steroids or nonsteroidal anti-inflammatory agents [175 – 177]. These data suggest that decreased bone turnover is a feature of JRA independent of drug therapy. The factors contributing to osteopenia in rheumatologic disorders include reduced physical mobility and therapy with glucocorticoids and methotrexate. As with other forms of chronic disease, delayed growth and puberty may account for some delays in bone mineral accrual. Increased production and exposure to inflammatory cytokines may also contribute to bone loss (see also Chapters 13 and 54). Treatment recommendations are similar to those outlined for other forms of chronic childhood disorders. Efforts to maximize mobility are of particular importance in this setting. Physical therapy that incorporates weight bearing has been shown to augment bone mineral accrual even in nonambulatory patients with cerebral palsy [178]. An open trial of alendronate therapy was shown to increase BMD in 43 osteopenic children with JRA, SLE, dermatomyositis, and other rheumatologic disorders; mean gains averaged 14.2  10% after 1 year as compared with pretreatment bone mineral accrual of only 1% annually [179]. Safety and efficacy of bisphosphonates must be explored further. The value of anabolic agents also warrants study because bone marker data indicate a low turnover state.

VI. IDIOPATHIC JUVENILE OSTEOPOROSIS Idiopathic juvenile osteoporosis (IJO) is a rare demineralization disorder of unknown etiology [180,181]. This condition typically presents in mid-childhood although the diagnosis has been made in patients as young as 1 year. Clinical findings include pathologic fractures involving long bones or vertebrae, bone pain, and difficulty walking. Osteopenia is more profound in the spine than at the radius [181]. Spontaneous resolution of symptoms and improvement in bone mass occurs at puberty in most patients, although disability and fractures may persist into adulthood [181]. The diagnosis of idiopathic juvenile osteoporosis can be made only after the exclusion of known causes of early osteopenia, such as osteogenesis imperfecta, leukemia, calcium or vitamin D deficiency, hyperparathyroidism, hyperthyroidism, or Cushing syndrome. No consistent biochemical abnormality has been associated with this disorder to serve as a

162 definitive diagnostic marker. Circulating calcium, phosphate, magnesium, PTH, alkaline phosphatase activity, calcitonin, and 25-hydroxyvitamin D concentrations are normal for age [181,182]. In some cases, reduced serum concentrations of 1,25-dihydroxyvitamin D have been observed [180,182], consistent with a partial defect in renal 1-alpha-hydroxylase activity. One patient was found to have elevated 1,25-dihydroxyvitamin D concentrations, thought to reflect calcitonin deficiency [183]. Baseline and stimulated values of osteocalcin in patients with IJO are similar to those seen in controls, suggesting that osteoblast function is normal [184]. Bone histomorphometry has revealed thin trabeculae or cortices in some and an increase in osteocytes in other patients [181]. Given the lack of specific clinical and laboratory findings, IJO may not represent a distinct entity but rather a heterogeneous group of osteopenic disorders, including variants of osteogenesis imperfecta not caused by a known defect in the type I collagen gene. The efficacy of therapies attempted in IJO is difficult to assess because of the difficulties in diagnosing the disorder, the small numbers treated, and the frequency of spontaneous recovery. Calcitriol was associated with reduced fracture rates and increased bone mass in three patients with low concentrations of 1,25-dihydroxyvitamin D, while one untreated patient showed no improvement [180]. Calcitonin therapy failed to improve the bone mineral status of the patient described above with elevated serum 1,25dihydroxyvitamin D [183]. In the absence of specific therapy, it appears reasonable to optimize vitamin D and calcium intake and to encourage weight bearing activity appropriate for the patient’s level of osteopenia. A controlled, randomized trial of alendronate therapy is being conducted at the National Institutes of Health. Fortunately, this poorly defined disorder resolves spontaneously as puberty progresses in the majority of patients [181].

VII. SUMMARY AND FUTURE DIRECTIONS Pediatric bone research has opened a Pandora’s box to reveal skeletal deficits in myriad chronic disorders. Not surprisingly, bone mineral accrual is compromised in the face of risk factors such as malnutrition, immobilization, glucocorticoid therapy, and other endocrinopathies. The list of disorders linked with early osteopenia is likely to expand and will encompass seemingly healthy youth. Children and adolescents screened because of forearm fractures have significantly lower bone density at all skeletal sites than those who have not broken a bone [185]. Further investigation of these youth may help to identify key risk factors for early bone fragility in children without chronic disease. To date, research on acquisitional osteopenia has been largely descriptive, cross-sectional, and limited to small

LAURA K. BACHRACH

cohorts of patients. In many cases, DXA results have not been corrected for all confounding factors such as bone size and skeletal or pubertal delay. Without longitudinal data, little is known about the natural history of childhood osteopenias. Do deficits result solely from inadequate accrual or is bone loss involved? How complete is recovery as risk factors are addressed and reversed? Despite recent progress, further studies are needed to develop effective treatment strategies for pediatric osteopenias. Larger cohorts must be studied longitudinally to define the natural history of childhood osteopenias and the reversibility of bone mineral deficits with general measures such improved nutrition, increased activity, and correction of endocrine deficits or excesses. There are indications that general measures do not suffice for complete restoration of peak bone mass, prompting the need for drug safety and efficacy studies. Controlled randomized trials of bisphosphonates or other anti-resorptive agents should be initiated for several high risk conditions, such as glucocorticoid-induced osteoporosis. Anabolic agents such as parathyroid hormone may prove to be more efficacious in reversing childhood osteopenias, but there is considerably less experience with these drugs even in adults. This research can best be accomplished in multicenter trials that will facilitate enrollment of sufficient subjects with less common, high-risk conditions.

References 1. S. L. Hui, C. W. Slemenda, and C. C. Johnston, The contribution of bone loss of post menopausal osteoporosis. Osteoporosis Int. 1, 30–34 (1990). 2. D. A. Bailey, R. A. Faulkner, and H. A. McKay, Growth, physical activity, and bone mineral acquisition. Exercise Sports Sci. Rev. 24, 233 – 266 (1994). 3. W. S. Pollitzer and J. J. B. Anderson, Ethnic and genetic differences in bone mass: A review with a hereditary vs environmental perspective. Am. J. Clin. Nutr. 50, 1244 – 1259 (1989). 4. J. A. Eisman, Genetics of osteoporosis. Endocr. Rev. 20, 788 – 804 (1999). 5. R. R. Recker, M. Davies, S. M. Hinders, R. P. Heaney, M. R. Stegman, and D. B. Kimmel, Bone gain in young adult women. JAMA 268, 2403 – 2408 (1992). 6. C. W. Slemenda, T. K. Reister, S. L. Hui, J. Z. Miller J. C. Christian, and C. C. Johnston, Influences on skeletal mineralization in children and adolescents: Evidence for varying effects of sexual maturation and physical activity. J. Pediatr. 125, 201 – 207 (1994). 7. R. P. Heaney, Nutritional factors in osteoporosis. Annu. Rev. Nutr. 13, 287 – 316 (1993). 8. D. A. Bailey, H. A. McKay, R. L. Mirwald, P. R. E. Crocker, and R. A. Faulkner, A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: The University of Saskatchewan Bone Mineral Accrual Study. J. Bone Miner. Res. 14, 1672 – 1679 (1999). 9. B. E. Bachrach and E. P. Smith, The role of sex steroids in bone growth and development: Evolving new concepts. The Endocrinologist 6, 362 – 368 (1996). 10. C. C. Johnston, Jr., J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation

CHAPTER 43 Osteoporosis in Childhood and Adolescence

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

27.

28.

and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). J.-Ph. Bonjour, A.-L. Carrie, S. Ferrari, H. Clavien, D. Slosman, G. Theintz, and R. Rizzoli, Calcium-enriched foods and bone mass growth in prepubertal girls: A randomized, double-bind, placebocontrolled trial. J. Clin. Invest. 99, 1287 – 1294 (1997). F. L. Morris, G. A. Naughton, J. L. Gibbs, J. S. Carlson, and J. D. Wark, Prospective 10-month exercise intervention in pre-menarcheal girls: Positive effects on bone and lean mass. J. Bone Miner. Res. 12, 1453 – 1462 (1997). M. Bradney, G. Pearce, G. Naughton, C. Sullivan, S. Bass, T. Beck, J. Carlson, and E. Seeman, Moderate exercise during growth in prepubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J. Bone Miner. Res. 13, 1814 – 1821 (1998). S. L. Hui, C. S. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). S. Bass, P. D. Delmas, G. Pearce, E. Hendrich, A. Tabensky, and E. Seeman, The differing tempo of growth in bone size, mass, and density in girls is region-specific. J. Clin. Invest. 104, 795 – 804 (1999). H. K. Genant, K. Engelke, T. Fuerst, C.-C. Gluer, S. Grampp, S. T. Harris, M. Jergas, T. Lang, Y. Lu, S. Majumdar, A. Mathur, M. Takada, Review: Non-invasive assessment of bone mineral and structure: State of the art. J. Bone Miner. Res. 11, 707 – 730 (1996). R. A. Faulkner, D. A. Bailey, D. T. Drinkwater, H. A. McKay, C. Arnold, and A. A. Wilkinson, Bone densitometry in Canadian children 8 – 17 years of age. Calcif. Tissue Int., 59, 344 – 351 (1996). L.K. Bachrach, T. Hastie, M.-C. Wang, B. Narasimhan, and R. Marcus, Bone mineral acquisition in healthy Asian, Hispanic, Black and Caucasian youth. A longitudinal study. J. Clin. Endocrinol. Metab. 84, 4702 – 4712 (1999). A. M. Boot, M. A. J. De Ridder, H. A. P. Pols, E. P. Krenning, and M. P. F. de Muinck Keizer-Schrama, Bone mineral density in children and adolescents: Relation to puberty, calcium intake, and physical activity. J. Clin. Endocrinol. Metab. 82, 57 – 62 (1997). P. W. Lu, C. T. Cowell, S. A. Lloyd-Jones, J. N. Briody, and R. Howman-Giles, Volumetric bone mineral density in normal subjects, aged 5 – 27 years. J. Clin. Endocrinol. Metab. 81, 1586 – 1590 (1996). V. Gilsanz, D. L. Skaggs, A. Kovanlikaya, J. Sayre, M. L. Loro, F. Kaufman, and S. G. Korenman, Differential effect of race on the axial and appendicular skeletons of children. J. Clin. Endocrinol. Metab. 83, 1420 – 1427 (1998). C. K. Lum, M.-C. Wang, E. Moore, D. M. Wilson, R. Marcus, and L. K. Bachrach, A comparison of calcaneus ultrasound and dual energy x-ray absorptiometry in healthy American youth. J. Clin. Densitom. 2, 419 – 427(1999). M. B. Leonard, K. J. Propert, B. S. Zemel, V. A. Stallings, and H. I. Feldman, Discrepancies in pediatric bone mineral density reference data: Potential for misdiagnosis of ostepenia. J. Pediatr. 135, 182 – 188 (1999). H. Kroger, A. Kotaniemi, P. Vainio, and E. Alhava, Bone densitometry of the spine and femur in children by dual-energy x-ray absorptiometry. Bone Mineral 17, 75 – 85 (1992). D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone Miner. Res. 7, 137 – 145 (1992). C. Molgaard, B. L. Thomsen, and K. F. Michaelsen, Influence of weight, age, and puberty on bone size and bone mineral content in healthy children and adolescents. Acta Paediatr. 87, 494 – 499 (1998). E. Schonau and F. Rauch, Markers of bone and collagen metabolism — Problems and perspectives in paediatrics. Horm. Res. 48 (Suppl. 5), 50 – 59 (1997).

163 29. S. Mora, P. Pitukcheewanont, F. R. Kaufman, J. C. Nelson, and V. Gilsanz, Biochemical markers of bone turnover and the volume and the density of bone in children at different stages of sexual development. J. Bone Miner. Res. 14, 1664 – 1671 (1999). 29a. F. H. Glorieux, R. Travers, A. Taylor, J. R. Bowen, F. Rauch, M. Norman, and A. M. Parfitt, Normative data for iliac bone histomorphometry in growing children. Bone 26, 103 – 109 (2000). 30. T. L. Holbrook, K. Grazier, J. L. Kelsey, and R. N. Stauffen, “The Frequency of Occurrence, Impact and Cost of Selected Musculoskeletal Conditions in the United States.” American Academy of Orthopedic Surgeons, Chicago, (1984). 31. K. A. Carmichael and D. H. Carmichael, Bone metabolism and osteopenia in eating disorders. Medicine 4, 254 – 267 (1995). 32. L. K. Bachrach, D. Guido, D. Katzman, I. F. Litt, and R. Marcus, Decreased bone density in adolescent girls with anorexia nervosa. Pediatrics 86, 440 – 447 (1990). 33. A. Klibanski, B. M. K. Biller, D. A. Schoenfeld, D. B. Herzog, and V. C. Saxe, The effects of estrogen administration on trabecular bone loss in young women with anorexia nervosa. J. Clin. Endocrinol. Metab. 80, 898 – 904 (1995). 34. S. A. Abrams, T. J Silber, N. V. Esteban, N. E. Vieira, J. E. Stuff, R. Meyers, M. Majd, and A. L. Yergey, Mineral balance and bone turnover in adolescents with anorexia nervosa. J. Pediatr. 123, 326 – 331 (1993). 35. B. M. K. Biller, V. Saxe, D. B. Herzog, D. I. Rosenthal, S. Holzman, and A. Klibanski, Mechanisms of osteoporosis in adult and adolescent women with anorexia nervosa. J. Clin. Endocrinol. Metab. 68, 548 – 554 (1989). 36. L. K. Bachrach, D. K. Katzman, I. F. Litt, D. Guido, and R. Marcus, Recovery from osteopenia in adolescent girls with anorexia nervosa. J. Clin. Endocrinol. Metab. 72, 602 – 606 (1991). 37. W. Herzog, H. Minne, C. Deter, G Leidig, D. Schellberg, C. Wuster, R. Gronwald, E. Sarembe, F. Kroger, and G. Bergmann, Outcome of bone mineral density in anorexia nervosa patients 11.7 years after first admission. J. Bone Miner. Res. 8, 597 – 605 (1993). 38. N. A. Rigotti, R. M. Neer, S. J. Skates, D. B. Herzog, and S. R. Nussbaum, The clinical course of osteoporosis in anorexia nervosa. A longitudinal study of cortical bone mass. JAMA 265, 1133 – 1138 (1991). 39. T. N. Hangartner and C. C. Johnston, Influence of fat on bone measurements with dual-energy absorptiometry. Bone Miner. 9, 71 – 81 (1990). 40. R. E. Kreipe, D. G. Hicks, R. N. Rosier, and J. E. Puzas, Preliminary findings on the effects of sex hormones on bone metabolism in anorexia nervosa. J. Adolesc. Health 14, 319 – 324 (1993). 41. E. Seeman, G. I. Szmukler, C. Formica, C. Tsalamandris, and R. Mestrovic, Osteoporosis in anorexia nervosa: The influence of peak bone density, bone loss, oral contraceptive use and exercise. J. Bone Miner. Res. 7, 1467 – 1474 (1992). 42. NIH Consensus Development Panel on Optimal Calcium Intake, Optimal calcium intake. JAMA 272, 1942 – 1948 (1994). 43. J. M. Joyce, D. L. Warren, L. L Humphries, A. J. Smith, and J. S. Coon, Osteoporosis in women with eating disorders: Comparison of physical parameters, exercise and menstrual status with SPA and DPA evaluation. J. Nucl. Med. 31, 325 – 331 (1990). 44. N. A. Rigotti, S. R. Nussbaum, D. B. Herzog, and R. M. Neer, Osteoporosis in women with anorexia nervosa. N. Engl. J. Med. 311, 1601 – 1606 (1984). 45. J. S. Hay, A. Hall, J. W. Delahunt, G. Harper, A. W. Mitchell, and C. Salmond, Investigation of osteopaenia in anorexia nervosa. Aust. NZ J. Psychiatry 23, 261 – 268 (1989). 46. C. M. White, A. C. Hergenroeder, and W. J. Klish, Bone mineral density in 15- to 21-year-old eumenorrheic and amenorrheic subjects. Am. J. Dis. Child 146, 31 – 35 (1992).

164 47. K. H. Myburgh, J. Hutchins, A. B. Fataar, S. F. Hough, T. D. Noakes, Low bone density is an etiologic factor for stress fractures in athletes. Ann. Intern. Med. 113, 754 – 759 (1990). 48. K. H. Myburgh, L. K. Bachrach, B. Lewis, K. Kent, and R. Marcus, Low bone mineral density at axial and appendicular sites in amenorrheic athletes. Med. Sci. Sports Exercise 25, 1197 – 1202 (1993). 49. C. M. Snow-Harter, Bone health and prevention of osteoporosis in active and athletic women. Clin. Sports Med. 13, 389 – 404 (1994). 50. N. T. Frusztajer, S. Dhuper, M. P. Warren, J. Brooks-Gunn, and R. P. Fox, The role of nutritional intake in the incidence of stress fractures in ballet dancers. Am. J. Clin. Nutr. 51, 779 – 783 (1990). 51. M. L. Rencken, C. H. Chestnut, and B. L. Drinkwater, Bone density at multiple skeletal sites in amenorrheic athletes. JAMA 276, 238 – 240 (1996). 52. N. Young, C. Formica, G. Szmukler, and E. Seeman, Bone density at weight-bearing and nonweight-bearing sites in ballet dancers: The effects of exercise, hypogonadism, and body weight. J. Clin. Endocrinol. Metab. 78, 449 – 454 (1994). 53. T. L. Robinson, C. Snow-Harter, D. R. Taaffe, D. Gillis, J. Show, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10, 26 – 35 (1995). 54. S. Bass, G. Pearce, M. Bradney, E. Hendrich, P. D. Delmas, A. Harding, and E. Seeman, Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepubertal and retired female gymnasts. J. Bone Miner. Res. 12, 500 – 507 (1998). 55. G. E. Theintz, H. Howald, U. Weiss, and P. C. Sizonenko, Evidence for a reduction in growth potential in adolescent female gymnasts. J. Pediatr. 12, 306 – 313 (1993). 56. S. Bass, M. Bradney, G. Pearce, E. Hendrich, S. Stuckey, S. K. Lo, and E. Seeman, Short stature and delayed puberty in gymnasts: Influence of selection bias on leg length and the duration of training on trunk length. J. Pediatr. 136, 149 – 155 (2000). 57. W. J. Gadpaille, C. F. Sanborn, and W. W. Wagner, Athletic amenorrhea, major affective disorders and eating disorders. Am. J. Psychiatry 114, 939 – 942 (1987). 58. M. P. Warren, Amenorrhea in endurance runners. J. Clin. Endocrinol. Metab. 75, 1393 – 1397 (1992). 59. B. A. Bullen, G. S. Skinar, I. Z. Beitins, G. Von Merging, B. A. Turnbull, and J. W. McArthur, Induction of menstrual disorders by strenous exercise in untrained women. N. Engl. J. Med. 312, 1349 – 1353 (1985). 60. A. C. Hergenroeder, Bone mineralization, hypothalamic amenorrhea, and sex steroid therapy in female adolescents and young adults. J. Pediatr. 126, 683 – 689 (1995). 61. B. A. Bullen, M. P. Warren, J. M. Stager, B. Ritchie-Flanagan, and D. Robertshaw, Reversibility of amenorrhea in athletes. N. Engl. J. Med. 310, 51 – 52 (1984). 62. D. C. Cumming, Exercise-associated amenorrhea, low bone density, and estrogen replacement therapy. Arch. Intern. Med. 156, 2193 – 2195 (1996). 63. J. H. Gibson, A. Mitchell, J. Reeve, and M. G. Harries, Treatment of reduced bone mineral density in athletic amenorrhea: A pilot study. Osteoporosis Int. 10, 284 – 289 (1999). 64. B. L. Drinkwater, K. Nilson, S. Ott, and C. H. Chestnut, III, Bone mineral density after resumption of menses in amenorrheic athletes. JAMA 256, 380 – 382(1986). 65. J. S. Finkelstein, R. M. Neer, B. M. K. Biller, J. D. Crawford, and A. Klibanski, Osteopenia in men with a history of delayed puberty. N. Engl. J. Med. 326, 600 – 604 (1992). 66. J. S. Finkelstein, A. Klibanski, and R. M. Neer, A longitudinal evaluation of bone mineral density in adult men with histories of delayed puberty. J. Clin. Endocrinol. Metab. 81, 1152 – 1155 (1996).

LAURA K. BACHRACH 67. B. Moore, J. Briody, C. T. Cowell, and E. Mobbs, Does maturational delay affect bone mineral density (BMD)? Horm. Res. (Suppl. 2) 48, 91 (1998) [Abstract]. 68. S. Bertelloni, G. I. Baroncelli, M. Ferdeghini, G. Perri, and G. Saggese, Normal volumetric bone mineral density and bone turnover in young men with histories of constitutional delay of puberty. J. Clin. Endocrinol. Metab. 83, 4280 – 4283 (1998). 69. S. Bertelloni, G. I. Baroncelli, R. Battini, G. Perri, and G. Saggese, Short-term effect of testosterone treatment on reduced bone density in boys with constitutional delay of puberty. J. Bone Miner. Res. 10, 1488 – 1495 (1995). 70. D. I. Rosenthal, W. Mayo-Smith, C. W. Hayes, J. S. Khurana, B. M. K. Biller, R. M. Neer, and A. Klibanski, Age and bone mass in premenopausal women. J. Bone Miner. Res. 4, 533 – 538 (1989). 71. A. C. Hergenroeder, E. O’Brian Smith, R. Shypallo, L. A. Jones, W. J. Klish, and D. Ellis, Bone mineral changes in young women with hypothalamic amenorrhea treated with 12 months. Am. J. Obstet. Gynecol. 176, 1017 – 1025 (1997). 72. E. K. Neely, L. K. Bachrach, R. L. Hintz, R. L. Habiby, C. W. Slemenda, L. Feezle, and O. H. Pescovitz, Bone mineral density during treatment of central precocious puberty. J. Pediatr. 127, 819 – 822 (1995). 73. A. M. Boot, S. M. P. F. De Muinck Keizer-Schrama, H. A. P. Pols, E. P. Krenning, and S. L. S. Drop, Bone mineral density and body composition before and during treatment with gonadotroin-releasing hormone agonist in children with central precocious and early puberty J. Clin. Endocrinol. Metab. 83, 370 – 373 (1998). 74. F. Antoniazzi, F. Bertoldo, S. Lauriola, S. Sirpresi, E. Gasperi, G. Zamboni, and L. Tato, Prevention of bone demineralization by calcium supplementation in precocious puberty during gonadotropin-releasing hormone agonist therapy. J. Clin. Endocrinol. Metab. 84, 1992 – 1999 (1999). 75. E. S. Surrey and H. L. Judd, Reduction of vasomotor symptoms and bone mineral denisty loss with combined norethindrone and longacting gonadotropin-releasing hormone agonist therapy of symptomatic endometriosis: A prospective randomized trial. J. Clin. Endocrinol. Metab. 75, 558 – 563 (1992). 76. M. C. Davies, B. Gulekli, and H. S. Jacobs, Osteoporosis in Turner’ syndrome and other forms of primary amenorrhoea. Clin. Endocrinol. 43, 741 – 746 (1995). 77. G. Saggese, S. Bertelloni, G. I. Baroncelli, R. Battini, and G. Franchi, Reduction of bone density: An effect of gonadotropin releasing hormone analogue treatment in central precocious puberty. Eur. J. Pediatr. 152, 717 – 720 (1993). 78. S. Bertelloni, G. I. Baroncelli, M. C. Sorrentino, G. Perri, and G. Saggese, Effect of central precocious puberty and gonadotropinreleasing hormone analogue treatment on peak bone mass and final height in females. Eur. J. Pediatr. 157, 363 – 367 (1998). 79. S. Heger, C. -J. Partsch, and W. G. Sippell, Long-term outcome after depot gonadotropin-releasing hormone agonist treatment of central precocious puberty: Final height, body proportions, body composition, bone mineral density, and reproductive function. J. Clin. Endocrinol. Metab. 84, 4583 – 4590 (1999). 80. P. Bergmann, J. Valsamis, J. Van Perborgh, J. De Schepper, and G. Van Vliet, Comparative study of the changes in insulin-like growth factor-I, procollagen-III N-terminal extension peptide, bone Gla-protein, and bone mineral content in children with Turner’s syndrome treated with recombinant growth hormone. J. Clin. Endocrinol. Metab. 71, 1461 – 1467 (1990). 81. J. L. Ross, L. M. Long, P. Feuillan, F. Cassorla, and G. B. Cutler, Normal bone density of the wrist and spine and increased wrist fractures in girls with Turner’s syndrome. J. Clin. Endocrinol. Metab. 73, 355 – 369 (1991). 82. L. K. Bachrach, Osteopenia in Turner girls. In “Turner Syndrome in a Life Span Perspective: Research and Clinical Aspects

165

CHAPTER 43 Osteoporosis in Childhood and Adolescence

83.

84.

85.

86.

87.

88.

89.

90.

91.

92. 93.

94.

95.

96.

97.

98.

99.

(K. Albertsson-Wikland and M. B. Ranke, eds.), pp. 233 – 240. Elsevier Science, The Netherlands. K. S. Sykes and E. K. Neely, Long-term outcome of height, bone density, and body composition in Turner syndrome. Curr. Opinion. Endo. Diabetes. 4, 100 – 107 (1997). E. K. Neely, R. Marcus, R. G. Rosenfeld, and L. K. Bachrach, Turner syndrome adolescents receiving growth hormone are not osteopenic. J. Clin. Endocrinol. Metab. 76, 861 – 866 (1993). S. Mora, G. Weber, M. P. Guarneri, G. Nizzoli, D. Pasoline, and G. Chiumello, Effect of estrogen replacement therapy on bone mineral content in girls with Turner syndrome. Obstet. Gynecol. 79, 747 – 751 (1992). R. T. Kirkland, T.-H Lin, A. D. LeBlanc, and J. L. Kirkland, Effects of hormonal therapy on bone mineral density in Turner syndrome. In “Turner Syndrome.” (R. G. Rosenfeld and M. M. Grumbach, eds.), pp. 319 – 325. Marcel Dekker, Inc., New York (1990). J. J. Stepan, J. Musilova, and V. Pacovsky, Bone demineralization, biochemical indices of bone remodeling, and estrogen replacement therapy in adults with Turner’s syndrome. J. Bone Miner. Res. 4, 193 – 198 (1989). R. W. Naeraa, K. Brixen, R. M. Hansen, C. Hasling, L. Mosekilde, J. -H Andresen, P. Charles, J. Nielsen, Skeletal size and bone mineral content in Turner’s syndrome: Relation to karyotype, estrogen treatment, physical fitness, and bone turnover. Calcif. Tissue. Int. 49, 77 – 83 (1991). M. C. Davies, B. Gulekli, and H. S. Jacobs, Osteoporosis in Turner’s syndrome and other forms of primary amenorrhoea. Clin. Endocrinol. 43, 71 – 746 (1995). K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C.-C Gluer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The Study of Osteoporotic Fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). G. Saggese, G. Federico, B. Silvano, and G. I. Baroncelli, Mineral metabolism in Turner’s syndrome: Evidence for impaired renal vitamin D metabolism and normal osteoblast function. J. Clin. Endocrinol. Metab. 75, 998 – 1001 (1992). L. G. Raisz, Local and systemic factors in the pathogenesis of osteoporosis. N. Engl. J. Med. 318, 818 – 828 (1988). S. E. Inzucchi and R. J. Robbins, Effects of growth hormone on human bone biology. J. Clin. Endocrinol. Metab. 79, 691 – 693 (1994). G. Saggese, B. I. Baroncelli, S. Bertelloni, L. Cinquanta, and G. Di Nero, Effects of long-term treatment with growth hormone on bone and mineral metabolism in children with growth hormone deficiency. J. Pediatr. 122, 37 – 45 (1993). G. D. Ogle, A. R. Rosenberg, D. Calligeros, and G. Kainer, Effects of growth hormone treatment for short stature on calcium homeostasis, bone mineralisation, and body composition. Horm. Res. 41, 16 – 20 (1994). A. M. Boot, M. A. M. J. Engels, G. J. M. Boerma, E. P. Krenning, and S. M. P. F. De Muinck Keizer-Schrama, Changes in bone mineral density, body composition, and lipid metabolism during growth hormone (GH) treatment in children with GH deficiency. J. Clin. Endocrinol. Metab. 82, 2423 – 2428 (1997). G. Zambone, F. Antoniazzi, G. Radetti, C. Musumeci, and L. Tato, Effects of two different regimens of recombinant human growth hormone therapy on the bone mineral density of patients with growth hormone deficiency. J. Pediatr. 119, 483 – 485 (1991). D. J. O’Halloran, A. Tsatsoulis, R. W. Whitehouse, S. J. Holmes, J. E. Adams, and S. M. Shalet, Increased bone density after recombinant human growth hormone (GH) therapy in adults with isolated GH deficiency. J. Clin. Endocrinol. Metab. 76, 1344 – 1348 (1993). L. K. Bachrach, R. Marcus, S. M. Ott, A. L. Rosenbloom, O. V. Vasconez, A. L. Martinez, R. G. Rosenfeld, and J. Guevara-Aguirre, Bone mineral, histomorphometry and body composition in adults

100.

101.

102.

103.

104.

105. 106.

107.

108.

109.

110.

111.

112.

113.

114. 115.

116.

117. 118.

119.

with growth hormone receptor deficiency. J. Bone Miner. Res. 13, 415 – 421 (1998). R. Lanes, P. Gunczler, and J. R. Weisinger, Decreased trabecular bone mineral density in children with idiopathic short stature: Normalization of bone density and increased bone turnover after 1 year of growth hormone treatment. J. Pediatr. 135, 177 – 181 (1999). J. -M. Kaufman, P. Taelman, A. Vermeulen, and M. Vandeweghe, Bone mineral status in growth hormone-deficient males with isolated and multiple pituitary deficiencies of childhood onset. J. Clin. Endocrinol. Metab. 74, 118 – 123 (1992). H. De Boer, G. J. Block, A. Van Lingen, G. J. J. Teule, P. Lips, and E. A. Van Der Veen, Consequences of childhood-onset growth hormone deficiency for adult bone mass. J. Bone Miner. Res. 9, 1319 – 1326 (1994). F. Albright and E. C. Reifenstein, “The Parathyroid Glands and Metabolic Bone Disease: Selected Studies.” Williams & Wilkins, Baltimore, (1948). M. E. Levin, V. C. Boisseau, and L. V. Avioli, Effects of diabetes mellitus on bone mass in juvenile and adult-onset diabetes. N. Engl. J. Med. 294, 241 – 245 (1976). R. Bouillon, Diabetic bone disease. Calcif. Tissue Int. 49, 155 – 160 (1991). B. Piepkorn, P. Kann, T. Forst, J. Andreas, A. Pfutzner, and J. Beyer, Bone mineral density and bone metabolism in diabetes mellitus. Horm. Metab. Res. 29, 584 – 591 (1997). J. V. Santiago, W. H. McAlister, S. K. Ratzan, Y. Bussman, M. W. Haymond, G. Shackelford, and V. V. Weldon, Decreased cortical thickness and osteopenia in children with diabetes mellitus. J. Clin. Endocrinol. Metab. 45, 845 – 848 (1977). A. L. Rosenbloom, D. C. Lezotte, F. T. Weber , J. Gudat, D. R. Heller, M. L. Weber, S. Klein, and B. B. Kennedy, Diminution of bone mass in childhood diabetes. Diabetes 26, 1052 – 1055 (1977). R. M. Shore, R. W. Chesney, R. B. Mazess, P. G. Rose, and G. J. Bargman, Osteopenia in juvenile diabetes. Calcif. Tissue Int. 33, 455 – 457 (1981). P. McNair, S. Madsbad, C. Christiansen, O. K. Faber, I. Transbol, and C. Binder, Osteopenia in insulin treated diabetes mellitus. Diabetologia 15, 87 – 90 (1978). T. F. Roe, S. Mora, G. Costin, F. Kaufman, M. E. Carlson, and V. Gilsanz, Vertebral bone density in insulin-dependent diabetic children. Metabolism 40, 967 – 971 (1991). S. W. Ponder, D. P. McCormick, D. Fawcett, A. D. Tran, G. W. Ogelsby, B. H. Brouhard, and L. B. Travis, Bone mineral density of the lumbar vertebrae in children and adolescents with insulin-dependent diabetes mellitus. J. Pediatr. 120, 541 – 545 (1992). S. L. Hui, S. Epstein, and C. C. Johnston, A prospective study of bone mass in patients with type I diabetes. J. Clin. Endocrinol. Metab. 60, 74 – 80 (1985). M. Leon, L. Larrodera, G. Lledo, and F. Hawkins, Study of bone loss in diabetes mellitus type I. Diabetes Res. Clin. Pract. 6, 237 – 242 (1989). M. P. Guarneri, G. Weber, P. Gallia, and G. Chiumello, Effect of insulin treatment on osteocalcin levels in diabetic children and adolescents. J. Endocrinol. Invest. 16, 505 – 509 (1993). G. Saggese, S. Bertelloni, G. I. Baroncelli, and P. Ghirri, Calcitonin secretion in children with insulin-dependent diabetes mellitus. Eur. J. Pediatr. 150, 464 – 467 (1991). L. Wartofsky, Osteoporosis and therapy with thyroid hormone. The Endocrinologist 1, 57 – 61 (1991). P. J. Meunier, G. G. S. Bianchi, C. M. Edouard, J. C. Bernard, P. Courpron, and G. E. Vignon, Bone manifestations of thyrotoxicosis. Orthoped. Clin. North Am. 3, 745 – 774 (1972). S. Mora, G. Weber, K. Marenzi, E. Signorini, R. Rovelli, M. C. Proverbio, and G. Chiumello, Longitudinal changes of bone density and bone resorption in hyperthyroid girls during treatment. J. Bone Miner. Res. 14, 1971 – 1977 (1999).

166 120. G. Radetti, C. Castellan, L. Tato, K. Platter, L. Gentili, and S. Adami, Bone mineral density in children and adolescent females treated with high doses of L-thyroxine. Horm. Res. 3, 127 – 131 (1993). 121. S. Heyerdahl, B. F. Kase, and S. O. Lie, Intellectual development in children with congenital hypothyroidism in relation to recommended thyroxine treatment. J. Pediatr. 118, 850 – 857 (1991). 122. C. M. Leinung and D. Zimmerman, Cushing’s disease in children. Endocrinol. Metab. Clin. N. Am. 23, 629 – 639 (1994). 123. A. Kotaniemi, A. Savolainen, H. Kautiainen, and H. Kroger, Estimation of central osteopenia in children with chronic polyarthritis treated with glucocorticoids. Pediatrics 91, 1127 – 1130 (1993). 124. S. Varanos, B. M. Ansell, and J. Reeve, Vertebral collapse in juvenile chronic arthritis: Its relationship with glucocorticoid therapy. Calcif. Tissue Int. 41, 75 – 78 (1987). 125. B. P. Lukert and L. G. Raisz, Glucocorticoid-induced osteoporosis: Pathogenesis and management. Ann. Intern. Med. 112, 352 – 364 (1990). 126. S. Crowley, P. C. Hindmarsh, D. R. Matthews, and C. G. D. Brook, Growth and the growth hormone axis in prepubertal children with asthma. J. Pediatr. 126, 297 – 303 (1995). 127. G. W. Leong, L. B. Mercaado-Asis, J. C. Reynolds, S. C. Hill, E. H. Oldfied, and G. P. Chrousos, The effect of Cushing’s disease on bone mineral density, body composition, growth, and puberty: A report of an identical adolescent twin pair. J. Clin. Endocrinol. Metab. 81, 1905 – 1911 (1996). 128. O. S. Gluck, W. A. Murphy, T. J. Hahn, and B. Hahn, Bone loss in adults receiving alternate day glucocorticoid therapy. A comparison with daily therapy. Arthritis Rheum. 24, 892 – 898 (1981). 129. C. Heuck, O. D. Wolthers, G. Kollerup, M. Hansen, and B. Teisner, Adverse effects of inhaled budesonide (800 g) on growth and collagen turnover in children with asthma: A double-blind comparison of once-daily versus twice-daily adminstration. J. Pediatr. 133, 608 – 612 (1998). 130. J. Loftus, R. Allen, R. Hesp, J. David, D. M. Reid, D. J. Wright, J. R. Green, J. Reeve, B. M. Ansell, and P. M. Woo, Randomized, doubleblind trial of deflazacort versus prednisone in juvenile chronic (or rheumatoid) arthritis: A relatively bone -sparing effect of deflazacort. Pediatrics 88, 428 – 436 (1991). 131. P. Sambrook, J. Birmingham, P. Kelly, S. Kempler, T. Nguyen, N. Pocock, and J. Eisman, Prevention of corticosteroid osteoporosis. A comparison of calcium, calcitriol, and calcitonin. N. Engl. J. Med. 328, 1747 – 1752 (1993). 132. J. D. Adachi, W. G. Bensen, J. Brown, D. Hanley, A. Hodsman, R. Josse, D. L. Kendler, B. Lentle, W. Olszynski, L.-G. Ste.-Marie, A. Tenenhouse, and A. A. Chines, Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N. Engl. J. Med. 337, 382 – 287 (1997). 133. K. G. Saag, R. Emkey, T. J. Schnitzer, J. P. Brown, F. Hawkins, S. Goemaere, G. Thamsborg, U. A. Liberman, P. D. Delmas, M.-P. Malice, M. Czachur, and A. G. Daifotis, Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N. Engl. J. Med. 339, 292 – 299 (1998). 134. T. Nishioka, H. Kurayama, T. Yasuda, J. Udagawa, C. Matsumura, and H. Niimi, Nasal administration of salmon calcitonin for prevention of glucocorticoid-induced osteoporosis in children with nephrosis. J. Pediatr. 118, 703 – 707 (1991). 135. F. H. Glorieux, N. J. Bishop, H. Plotkin, G. Chabot, G. Lanoue, and R. Travers, Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N. Engl. J. Med. 339, 947 – 952 (1998). 136. C. Brumsen, N. A. T. Hamdy, and S. E. Papapoulos, Long-term effects of bisphosphonates on the growing skeleton. Studies of young patients with severe osteoporosis. Medicine 76, 266 – 283 (1997). 137. J. Allgrove, Bisphosphonates. Arch. Dis. Child 76, 73 – 75 (1997).

LAURA K. BACHRACH 138. D. Allen and B. Goldberg, Stimulation of collagen synthesis and linear growth by growth hormone in glucocorticoid-treated children. Pediatrics 89, 416 – 421 (1992). 139. C. P. Sanchez, W. G. Goodman, D. Brandli, M. Goldenhersh, C. Murray, E. Carlton, T. Hahn, and I. B. Salusky, Skeletal response to recombinant human growth hormone (rhGH) in children treated with long-term corticosteroids. J. Bone Miner. Res. 10, 2 – 6 (1995). 140. S. L. Wardlaw and J. P. Bilezikian, Editorial: Hyperprolactinemia and osteopenia. J. Clin. Endocrinol. Metab. 75, 690 – 691 (1992). 141. S. E. Oberfield, M. Nino, L. Riddick, S. Pang, M. Nagel, A. Khandji, R. Kairam, and L. S. Levine, Combined bromocriptine and growth hormone (GH) treatment in GH-deficient children with macroprolactinoma in situ. J. Clin. Endocrinol. Metab. 75, 87 – 90 (1992). 142. M. L. Patton and P. D. Woolf, Hyperprolactinemia and delayed puberty: A report of three cases and their response to therapy. Pediatrics 71, 572 – 575 (1983). 143. D. Tyson, D. Reggiardo, C. Sklar, and R. David, Prolactin-secreting macroadenomas in adolescents. Am. J. Dis. Child 147, 1057 – 1061 (1993). 144. E. W. Ridgeway, S. J. Emans, E. G. Fisher, and J. F. Crigler, Prolactinomas causing primary and secondary amenorrhea in adolescents. J. Adolesc. Health Care 3, 148 (1982). 145. R. D. Rohn, Galactorrhea in the adolescent. Sex. Act. Teenagers 2, 21 – 33 (1988). 146. J. Schlechte, L. Walkner, and M. Kathol, A longitudinal analysis of premenopausal bone loss in healthy women and women with hyperprolactinemia. J. Clin. Endocrinol. Metab. 75, 698 – 703 (1992). 147. B. M. K. Biller, H. B. A. Baum, D. I. Rosenthal, V. C. Saxe, P. M. Charpie, and A. Klibanski, Progressive trabecular osteopenia in women with hyperprolactinemic amenorrhea. J. Clin. Endocrinol. Metab. 75, 692 – 697 (1992). 148. S. M. Ott and M. L. Aitken, Osteoporosis in patients with cystic fi brosis. Clin. Chest. Med. 19, 555 – 567 (1998). 149. L. K. Bachrach, C. W. Loutit, R. B. Moss, and R. Marcus, Osteoporosis in adults with cystic fibrosis. Am. J. Med. 96, 27 – 34 (1994). 150. R. M. Aris, J. B. Renner, A. D. Winders, H. E. Buell, D. B. Riggs, G. E. Lester, and D. A. Ontjes, Increased rate of fractures and severe kyphosis: Sequelae of living into adulthood with cystic fibrosis. Ann. Int. Med. 128, 186 – 193 (1998). 151. C. S. Haworth, P. L. Selby, A. K. Webb, M. E. Dodd, H. Musson, R. Niven, McL., G. Economou, A. W. Horrocks, A. J. Freemont, E. B. Mawer, and J. E. Adams, Low bone mineral density in adults with cystic fibrosis. Thorax 54, 961 – 967 (1999). 152. G. S. Bhudhikanok, J. Lim, R. Marcus, A. Harkins, R. B. Moss, and L. K. Bachrach, Correlates of osteopenia in patients with cystic fi brosis. Pediatrics 97, 103 – 111 (1995). 153. A. B. Grey, R. W. Ames, R. D. Matthews, and I. R. Reid, Bone mineral density and body composition in adult patients with cystic fibrosis. Thorax 48, 589 – 593 (1993). 154. R. C. Henderson and C. D. Madsen, Bone density in children and adolescents with cystic fibrosis. J. Pediatr. 128, 28 – 34 (1996). 155. G. S. Bhudhikanok, M.-C. Wang, R. Marcus, A. Harkins, R. B. Moss, and L. K. Bachrach, Bone acquisition and loss in children and adults with cystic fibrosis: A longitudinal study. J. Pediatr. 133, 18 – 27 (1998). 156. C. S. Haworth, A. J. Freemont, A. K. Webb, M. E. Dodd, P. L. Selby, E. B. Mawer, and J. E. Adams, Hip fracture and bone histomorpometry in a young adult with cystic fibrosis. Eur. Respir. J. 14, 478 – 479 (1999). 157. G. I. Baroncelli, F. De Luca, G. Magazzu, T. Arrigo, C. Sferlazzas, C. Catena, S. Bertellone, and G. Saggese, Bone demineralization in cystic fibrosis: Evidence of imbalance between bone formation and degradation. Pediatr. Res. 41, 397 – 403 (1997). 158. F. Salamoni, M. Roulet, F. Gudinchet, M. Pilet, D. Thiebaud, and P. Burckhardt, Bone mineral content in cystic fibrosis patients:

CHAPTER 43 Osteoporosis in Childhood and Adolescence

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

Correlation with fat-free mass. Arch. Dis. Child 74, 314 – 318 (1996). P. Konig, Hillman, C. Cervantes, C. Levin, C. Maloney, B. Douglass, L. Johnson, and S. Allen, Bone metabolism in children with asthma treated with inhaled beclomethasone dipropionate. J. Pediatr. 122, 219 – 226 (1993). E. F. Juniper, P. A. Kline, R. S. Roberts, F. E. Hargreave, and E. E. Daniel, Airway responsiveness to methacholine during the natural menstrual cycle and the effect of oral contraceptives. Am. Rev. Resp. Dis. 135, 1039 – 1042 (1987). R. M. Aris, G. Lester, J. Renner, A. Winders, I. P. Neuringer, and D. Ontjes, Pamidronate for osteoporosis after lung transplantation in cystic fibrosis patients. Am. J. Respir. Crit. Care Med. 159, A537 (1999). C. S. Haworth, P. L. Selby, A. K. Webb, E. B. Mawer, J. E. Adams, and T. J. Freemont, Severe bone pain after intravenous pamidronate in adult patients with cystic fibrosis. Lancet 352, 1753 – 1754 (1998). D. S. Hardin and J. P. Sy, Effects of growth hormone treatment in children with cystic fibrosis: The National Cooperative Growth Study experience. J. Pediatr. 131, S65 – 69 (1997). S. M. Shalet and D. J. O’Halloran, Growth and endocrine sequelae following the treatment of childhood cancer. The Endocrinologist 4, 44 – 55 (1994). G. M. Samuda, M. Y. Cheng, and C. Y. Yeung, Back pain and vertebral compression: An uncommon presentation of childhood acute lymphoblastic leukemia. J. Pediatr. Orthop. 7, 175 – 178 (1987). S. L. Cohn, E. R. Morgan, and L. E. Mallette, The spectrum of metabolic bone disease in lymphoblastic leukemia. Cancer 59, 346 – 350 (1987). J. M. Halton, S. A. Atkinson, L. Fraher, C. E. Webber, W. P. Cockshott, C. Tam, and R. D. Barr, Mineral homeostasis and bone mass at diagnosis in children with acute lymphoblastic leukemia. J. Pediatr. 126, 557 – 564 (1995). S. A. Atkinson, L. Fraher, C. M. Gundberg, M. Andrew, M. Pai, and R. D. Barr, Mineral homeostasis and bone mass in children treated for acute lymphoblastic leukemia. J. Pediatr. 114, 793 – 800 (1989). V. Gilsanz, M. E. Carlson, T. F. Rose, and J. A. Ortega, Osteoporosis after cranial irradiation for acute lymphoblastic leukemia. J. Pediatr. 117, 238 – 244 (1990). R. Vassilopoulou-Sellin and I. Ramirez, Severe osteopenia and vertebral compression fractures after complete remission in an adolescent with acute leukemia. Am. J. Hematol. 39, 142 – 143 (1992). B. M. Brennan, A. Rahim, J. A. Adams, O. B. Eden, and S. M. Shalet, Reduced bone mineral density in young adults following cure of acute lymphoblastic leukaemia in childhood. Br. J. Cancer 79, 1859 – 1863 (1999).

167 172. F. Falcini, S. Trapani, R. Civinini, A. Capone, M. Ermini, and G. Bartolozzi, The primary role of steroids on the osteoporosis in juvenile rheumatoid patients evaluated by dual energy x-ray absorptiometry. J. Endocrinol. Invest. 19, 165 – 169 (1996). 173. A. Cetin, R. Celiker, F. Dincer, and M. Ariyurck, Bone mineral density in children with juvenile chronic arthritis. Clin. Rheumatol. 17, 551 – 553 (1998). 174. R. M. Pereira, J. E. Corrente, W. H. Chahade, and N. H. Yoshinari, Evaluation by dual X-ray absorptiometry (DXA) of bone mineral density in children with juvenile chronic arthritis. Clin. Exp. Rheum. 15, 495 – 501 (1998). 175. F. Falcini, M. Ermini, and F. Bagnoli, Bone turnover is reduced in children with juvenile rheumatoid arthritis. J. Endocrinol. Invest. 21, 31 – 36 (1998). 176. R. Brik, Z. Keidar, D. Schapira, and O. Israel, Bone mineral density and turnover in children with juvenile chronic arthritis. J. Rheumatol. 25, 990 – 992 (1998). 177. A. M. Reed, M. Haugen, L. M. Pachman, and C. B. Langman, Repair of osteopenia in children with juvenile arthritis. J. Pediatr. 122, 693 – 196 (1993). 178. K. E. Chad, D. A. Bailey, H. A. McKay, G. A. Zello, and R. E. Snyder, The effect of a weight-bearing physical activity program on bone mineral content and estimated volumetric density in children with spastic cerebral palsy. J. Pediatr. 135, 115 – 117 (1999). 179. R. Cimaz, R. Falcini, M. Bardare, F. Zulian, L. Lepore, A. Buoncompagni, and M. L. Bianchi, Safety and efficacy of alendronate for the treatment of osteoporosis in pediatric rheumatic diseases. Arthritis Rheum. 42 (Suppl.), S400 (1999). 180. G. Saggese, S. Bertelloni, G. I. Baroncelli, G. Perri, and A. Calderazzi, Mineral metabolism and calcitriol therapy in idiopathic juvenile osteoporosis. Am. J. Dis. Child. 145, 457 – 462 (1991). 181. R. Smith, Idiopathic juvenile osteoporosis: Experience of twentyone patients. Br. J. Rheumatol. 34, 68 – 77 (1995). 182. H. K. Marder, R. C. Tsang, G. Hug, and A. C. Crawford, Calcitriol deficiency in idiopathic juvenile osteoporosis. Am. J. Dis. Child. 136, 914 – 917 (1982). 183. E. C. Jackson, F. Strife, R. C. Tsang, and H. K. Marder, Effect of calcitonin replacement therapy in idiopathic juvenile osteoporosis. Am. J. Dis. Child. 142, 1237 – 1239 (1988). 184. S. Bertelloni, G. I. Baroncelli, G. Di Nero, and G. Saggese, Idiopathic juvenile osteoporosis: Evidence of normal osteoblast function by 1,25-dihydroxyvitamin D3 stimulation test. Calcif. Tissue Int. 51, 21 – 23 (1992). 185. A. Goulding, R. Cannan, S. M. William, E. J. Gold, R. W. Taylor, and N. J. Lewis-Barned, Bone mineral density in girls with forearm fractures. J. Bone Miner. Res. 13, 143 – 148 (1998).

CHAPTER 44

Glucocorticoid-Induced Osteoporosis Basic Pathological Mechanisms, Clinical Features, and Management in the New Millennium GARY M. LEONG, JACQUELINE R. CENTER, N. KATHRYN HENDERSON, AND JOHN A. EISMAN Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia

I. Introduction II. The Scope of the Problem and Economic Considerations III. Effect of Glucocorticoids on Bone Density and Fracture Incidence IV. Glucocorticoid-Induced Muscle Weakness V. Glucocorticoids: Methods of Administration and Dosage Effects

VI. The Pathogenesis and Molecular Basis of Glucocorticoid Action on Bone Metabolism and Development VII. Glucocorticoids and Calcium Homeostasis VIII. Treatment Options IX. Treatment and Fracture Outcomes X. Management of Glucocorticoid-Induced Osteoporosis in Children References

I. INTRODUCTION

longer-term users (greater than 1 year) of glucocorticoids have osteoporosis and a large proportion of them will suffer from fractures [2 – 4]. The mechanisms of glucocorticoid-induced osteoporosis are complex and effective preventive regimens have been elusive. This has been at least partly due to the wide variety of conditions for which glucocorticoids are prescribed, many of which may affect bone density independently of the glucocorticoid treatment, as well as the wide dosage

The deleterious effects of glucocorticoids on bone have been recognized since Harvey Cushing first described endogenous glucocorticoid excess over 50 years ago [1]. However, now with the widespread use of glucocorticoids for systemic diseases affecting nearly every bodily system, exogenous glucocorticoid is an important cause of osteoporosis. It has been estimated that up to 50% of

OSTEOPOROSIS, SECOND EDITION VOLUME 2

169

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

170 range of glucocorticoids prescribed and varying lengths of treatment. Thus preventive therapy has been suboptimal and only infrequently prescribed. However, there have been exciting developments in the molecular mechanisms underlying the pathogenesis of glucocorticoid bone disease. This review will highlight major aspects of these advances as well as current perspectives on treatment options.

II. THE SCOPE OF THE PROBLEM AND ECONOMIC CONSIDERATIONS Glucocorticoids are frequently used for their antiinflammatory and immunosuppressive properties. Community survey data suggest that the prevalence of use of oral glucocorticoids is 0.5% rising to 1.7% in women over 55 [5]. The systemic use of glucocorticoids is associated with numerous side effects, one of the most incapacitating being osteoporosis. The incidence of glucocorticoid osteoporosis is approximately 50% in patients treated for more than 6 months [2], and it has been estimated that over 34% of patients on long-term glucocorticoids have had fractures [3,4]. Despite the high prevalence of this iatrogenic morbidity associated with long-term use of glucocorticoids, co-prescription of therapy for osteoporosis is low, ranging from 5.6 to 14% [5,6]. In addition to the morbidity associated with osteoporotic fracture, consideration must also be given to the substantial economic costs associated with the management of osteoporotic fracture. These costs are substantial and for osteoporosis in general have been estimated at $US50 – 100 million per annum per million of population [7,8]. In view of the relatively high fracture risk in patients receiving glucocorticoid therapy, prophylactic therapy is likely to be cost-effective.

III. EFFECT OF GLUCOCORTICOIDS ON BONE DENSITY AND FRACTURE INCIDENCE The adverse effect of glucocorticoid therapy on bone mineral density (BMD) is not linear with time. Bone loss and fracture risk seems to be related, at least to some extent, to the dose and duration of glucocorticoid exposure. The risk of fracture increases rapidly after the commencement of glucocorticoid therapy, but does decrease again after cessation of therapy [9]. Glucocorticoids contribute to increased fracture risk through a number of direct and indirect mechanisms that result in accelerated bone loss. The bone loss is biphasic, precipitous during the first 12 months and more gradual but continuous in subsequent years as measured by DXA or histomorphometry [10,11]. For example, iliac crest trabecular bone volume decreased an average of 27% in 16

GARY M. LEONG ET AL

patients following 5 – 7 months of prednisone therapy, while no further loss was found in 6 of these patients reassessed after a further 12 months of therapy. The rate of bone loss is similar in the lumbar spine and femoral neck when measured using DXA [12,13]. However, it has been suggested that bone loss is most rapid in trabecular bone [14,15], hence quantitative computer tomography or lateral spine DXA scans may be more sensitive to bone loss than more commonly used AP spine and proximal femoral measures [15]. In adults after discontinuation of glucocorticoid therapy, there is restoration of bone loss [14,16] and a reduction of excess fracture risk [9]. This is comparable to the recovery of bone after successful treatment of endogenous Cushing’s syndrome [17 – 19]. As glucocorticoid excess in children or adolescents may interfere with accrual of peak bone mass, it is unclear whether a similar recovery of bone occurs [20]. The early rapid bone loss suggests that optimum effects of prophylactic therapy are most likely to be achieved if a primary prevention approach is adopted, particularly if there is a high likelihood of longterm glucocorticoid therapy. There is some evidence to suggest that fracture susceptibility may be higher in glucocorticoid-induced osteoporosis than in involutional osteoporosis [21,22]. Glucocorticoiddependent asthmatic patients with vertebral fracture have higher BMD than patients with vertebral fracture secondary to idiopathic osteoporosis [21]. Moreover, a disturbed relationship between BMD and fracture has been observed in postmenopausal women receiving glucocorticoid therapy for rheumatoid arthritis. In one study, a six fold increase in the risk of vertebral fracture was associated with 1 SD or less decrease in lumbar spine (LS) BMD [22]. It has been proposed that alterations in bone quality independent of BMD may explain these observations [22]. More recently, the distribution of LS BMD in patients with fracture was found to be similar regardless of whether they had been receiving glucocorticoid therapy or not [23]. Regardless of whether the BMD threshold for fracture is altered in patients on glucocorticoid therapy, the risk of fracture is elevated in this group of patients, averaging about 30% in adults treated for 5 years or longer [24]. Fracture risk is greatest at predominantly trabecular sites such as the vertebral bodies and ribs [25] but the risk of hip fracture is also approximately doubled in glucocorticoid users [26]. In a large retrospective cohort study, the relative rate of vertebral, nonvertebral, and hip fractures was found to be significantly higher in oral glucocorticoid users compared with topical glucocorticoid users (2.2 [95% CI 2.31 to 2.92]; 1.33 [95% CI 1.29 to 1.38]; and 1.61 [95% CI 1.47 to 1.76], respectively) [9]. Furthermore, a significant dose response was found for all fracture types except forearm. For example, the rate of hip fracture increased by 77% with an oral glucocorticoid dose of 2.5 – 7.5 mg/day and by

171

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

127% with doses of 7.5 mg/day. The rapid increase in risk of fracture on commencement of glucocorticoid therapy and rapid decline on cessation suggests changes in addition to those measured by bone density [9,27].

IV. GLUCOCORTICOID-INDUCED MUSCLE WEAKNESS Significant muscle weakness is another recognized complication of long-term glucocorticoid therapy [28]. As such, the associated muscle weakness in patients with glucocorticoid-induced osteoporosis may predispose to falls, which particularly in the setting of reduced BMD, could increase the risk of fracture. Certainly, in the aging population, quadriceps muscle weakness is a risk factor for fracture independent of BMD [29]. The pathophysiology of glucocorticoid-induced muscle weakness is multifactorial and related to direct and indirect effects stimulating protein degradation and inhibiting protein synthesis [30,31]. Resistance training has been reported to attenuate the amount of glucocorticoid-induced muscle weakness, suggesting that the major deficit in function may be related to reduced capacity to generate muscle force rather than impaired muscle endurance [30,32]. The clinical presentation of glucocorticoid-induced muscle weakness is usually insidious, symmetrical, pain-free, and mostly in the proximal musculature. Initially the lower limbs are affected resulting in difficulty with tasks such as stair climbing, followed by proximal upper limb and finally peripheral muscle involvement [28]. Functional limitations may be profound, as has been shown in a group of heart transplant recipients whose mean age was 51 years, and were found to have knee extensor strength equivalent to that of untrained, sedentary 75-year-old subjects [33]. Although few human studies have been performed, there is evidence suggesting that glucocorticoid-induced muscle weakness can be prevented by resistance exercise training. In one controlled trial of patients receiving oral glucocorticoid therapy for conditions other than respiratory diseases, specific resistance training of the inspiratory muscles prevented the significant deterioration in inspiratory muscle strength and endurance that occurred in the control group [34]. Heart transplant recipients on a conservative program of resistance exercise had four fold greater improvements in muscle strength than matched controls with usual rehabilitation care [33]. The mechanisms by which exercise prevents glucocorticoid-induced muscle weakness are unclear. Proposed mechanisms include altered receptor function, elevated concentration of circulating androgens and inhibition of glutamine synthetase activity [30]. In view of the efficacy of exercise in the prevention and treatment of glucocorticoid-induced muscle weakness, a

program of strength and endurance training would be prudent for all patients receiving long-term glucocorticoid therapy, particularly where high doses are required for disease control. Recommendations for exercise need to be individualized depending on the patient’s underlying disease status and exercise should be monitored and progressed regularly [30].

V. GLUCOCORTICOIDS: METHODS OF ADMINISTRATION AND DOSAGE EFFECTS A. Oral Glucocorticoids There is evidence to suggest that glucocorticoid-induced bone loss may be dose-dependent. However, it is unclear whether low doses cause bone loss in all patients. The beneficial effects of glucocorticoids in controlling the disease process, which may itself be deleterious to the skeleton, must be weighed against their potential negative effects on bone metabolism. In one study of 84 patients with rheumatoid arthritis, bone density was compared between 44 patients treated with low dose prednisone and those not taking glucocorticoids [35]. The mean (SE) dose of users was 8.0  0.5 mg/day with a mean duration of use of 90  12 months. There were significant reductions in bone density in all patients whether or not they were taking prednisone. However, there was no significant difference between those on glucocorticoids and those on no treatment although the prednisone group had a lower bone density. These results suggest that low dose glucocorticoids do not increase the risk of osteoporosis in rheumatoid arthritis patients. In a cross-sectional study of 139 patients with rheumatoid arthritis the relationships between lateral LS BMD and different doses of glucocorticoids were examined [36]. While no difference in BMD of patients taking 4mg/day of prednisone was observed when compared with patients not taking prednisone, patients taking 5 mg/day had significantly lower BMD than controls. Control for potential covariates such as disease severity or disease duration did not obscure the relationship. Another longitudinal study of patients with rheumatoid arthritis reported an 8% decrease in spinal trabecular BMD in those treated with 10 mg/day prednisone measured at 20 weeks compared with no change in the placebo-treated group [14]. Twenty-four weeks after stopping the prednisone, BMD had recovered 5% in the treatment group. Interestingly, 20% of patients receiving glucocorticoid treatment in this study did not lose bone. Individual variation in the susceptibility to glucocorticoids may reflect genetic differences or variability

172

GARY M. LEONG ET AL

in the pharmacokinetics of glucocorticoids between individuals [37]. Clinically, identification of individuals at increased risk of glucocorticoids side effects would be of great value and further studies of these factors need to be undertaken. A dose of 7.5 mg/day of prednisone (or equivalent doses of other glucocorticoids) has been proposed as the threshold beyond which most people will lose significant amounts of bone [38,39]. The concept of a threshold dose is controversial, as physiological evidence suggests that even low doses of glucocorticoids are likely to have detrimental effects on the skeleton [40]. For example, inhaled glucocorticoids have been found to reduce markers of bone formation [41,42], and a single oral dose of 2.5 mg of prednisone suppresses the normal nocturnal rise in osteocalcin [43]. It appears that the beneficial effects of glucocorticoids in controlling the disease process, which may itself be deleterious to the skeleton, may outweight their potential negative effects on bone metabolism. There is evidence that disease control may be as important as the dose of glucocorticoids in determining the effects on the skeleton. In a longitudinal study of patients with early rheumatoid arthritis, bone loss was related to disease activity [44]. In this study, greater bone loss occurred in the group treated with 1 to 5 mg/day of oral glucocorticoids than in those on higher doses, probably because of less effective disease control in the low-dose group. In addition to minimizing disease related bone loss, good disease control will help to ensure maintenance of exercise tolerance reducing the risk of falling and thus potentially decreasing fracture risk [45]. Thus although there is no clear threshold effect below which glucocorticoids will not affect bone, a dose of less than 5 mg of prednisone would appear to be relatively safe.

of bone loss even high-dose inhaled glucocorticoids are preferable to oral glucocorticoid use.

Many mechanisms have been proposed for the bone loss associated with glucocorticoid excess. These include decreased bone formation; increased bone resorption; decreased gut calcium absorption and increased renal calcium excretion resulting in increased secretion of parathyroid hormone; and sex hormone deficiency. While some, albeit varying degrees of evidence supports all of these factors, it has become increasing clear that a major effect of glucocorticoids is on osteoblast function. The main histological features of glucocorticoid-induced osteoporosis are decreased rates of bone formation, decreased trabecular wall thickness and cancellous bone area and in situ evidence of bone cell death [48]. These features can all be explained by major effects on osteoblasts. Recent studies suggest that in addition to a decrease in osteoblast differentiation, a key factor may be induction of osteoblast and osteocyte apoptosis [49]. Clinically, there is a decrease in markers of bone formation, while markers of bone resorption in general have not been shown to increase [24]. Recent molecular findings surrounding glucocorticoid effects on osteoblast differentiation and function, and the interlinkage of osteoblast and osteoclast function underline the occurrence of glucocorticoid-induced osteoporosis.

B. Inhaled Glucocorticoids

A. Osteoblast and Osteoclast Development

Respiratory disease is one of the most common indications for glucocorticoid therapy, with inhaled therapy more commonly prescribed [5,6]. Inhaled beclomethasone dipropionate has been found to reduce markers of bone formation, whereas fluticasone propionate does not have a significant effect [41,42]. However, the effects of inhaled glucocorticoids on bone density are less clear as many of the studies are confounded by concomitant oral glucocorticoid therapy. In a longitudinal trial of low and high doses of fluticasone propionate and beclomethasone dipropionate there was no significant bone loss detectable using quantitative computer tomography [46]. Similarly, when newly diagnosed asthmatic women, whose only glucocorticoid treatment was inhaled beclomethosone dipropionate (1000 g/day) were compared with healthy control women of similar age, neither group had bone loss detectable by DXA during the 12-month follow-up period [47]. This evidence suggests that in terms

Osteoblasts (bone forming cells) develop from pluripotential mesenchymal stem cells, which can also develop into chondrocytes, myocytes, fibroblasts, adipocytes, and hempoietic cells [50] (see Chapter 2). By contrast, the precursors of osteoclasts are hemopoietic cells of the monocyte – macrophage lineage [51] (see Chapter 3). Commitment to the osteoblast lineage is initiated by bone morphogenetic proteins (BMPs), members of the transforming growth factor- (TGF-) superfamily (see Chapter 14). Bone morphogenetic proteins-2 (BMP-2) and BMP-4 and their receptors are required for formation of both osteoblasts and osteoclasts [52]. BMPs stimulate transcription of the gene encoding for Cbfa1 (core binding factor a1) (see Chapters 2 and 6), which is also known as osteoblastspecific transcription factor (Osf2). Cbfa1 induces osteoblast differentiation by activating osteoblast-specific genes to produce alkaline phosphatase, osteopontin, bone

VI. THE PATHOGENESIS AND MOLECULAR BASIS OF GLUCOCORTICOID ACTION ON BONE METABOLISM AND DEVELOPMENT

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

Sialoprotein, type 1 procollagen, and osteocalcin [53,54]. In a mouse gene knockout model, absence of Cbfa1 prevented osteoblast development and, because of the dependence of osteoclastogenesis on mesenchymal cell differentiation, also led to abnormal osteoclast development [55]. The molecular basis of the dependency of osteoclastogenesis on osteoblast precursors occurs through a membrane-bound receptor called RANK (receptor of NFB ligand) which belongs to the tumor necrosis factor (TNF)related cytokine gene superfamily [56,57]. This receptor is expressed on committed osteoclastic cells. The RANK ligand (RANKL) which is also known as osteoprotogerin ligand (OPG-L) or osteoclast differentiation factor (ODF) or TRANCE, is expressed by cells of osteoblastic lineage and required for osteoclastogenesis. Transgenic mice with an absence of RANKL develop osteopetrosis, while administration of RANKL to normal mice produces hypercalcemia, increased bone turnover and bone loss [55]. A third factor, a secreted glycoprotein called osteoprotegerin (OPG), binds to RANK L and by acting as a decoy inhibits RANK L– RANK interaction and hence osteoclastogenesis and bone resorption in vitro and in vivo [58]. The RANKL binds to its receptor on the surface of hemopoietic osteoclast progenitor cells and, with macrophage colony-stimulating factor (M-CSF), is sufficient to induce osteoclastogenesis. The promoter of the murine and human RANKL gene contains two functional Cbfa1 binding sites providing further evidence at the molecular level of the essential and inseparable relationship between the bone formation and resorption pathways [49]. Furthermore, BMP-2 and BMP-4 stimulate Cbfa1 expression, suggesting the presence of a BMP – Cbfa1 – RANK L molecular cascade involved in osteoblast/ osteoclast bone interactions on remodeling. Thus, any factor such as glucocorticoids which has a major effect on osteoblast formation will also have a profound effect on bone resorption.

B. The Glucocorticoid Receptor: Molecular Mechanisms of Action Glucocorticoids exert their genomic effects through binding to a ubiquitously expressed intracellular receptor, the glucocorticoid receptor (GR) [59]. There exist two receptor isoforms GR and GR (777 aa and 727 aa, respectively), derived from alternative splicing of the same gene on chromosome 5 [60]. Most glucocorticoid actions appear to be mediated via GR. In fact, GR does not bind ligand, and it has been proposed that it may act as a dominant negative regulator of GR activity. GR is a 94-kDa protein which, in the absence of hormone is predominantly present in the cytoplasm of the cell. There it is bound to a multiprotein complex which includes two molecules of heat shock chaperone protein 90

173 (hsp90) and one molecule each of hsp70, hsp56, and hsp26 [61]. Upon hormone binding, GR undergoes a conformational change which enables it to dissociate from the hsp complex. The receptor becomes hyperphosphorylated, allowing its nuclear translocation. Recent evidence also suggests that multiple cofactor protein complexes are critical in nuclear receptor gene transcription, some of which are involved in chromatin remodeling [62]. This model of multiple transcription factor complexes seems to be common to the mechanism of action of other nuclear hormone receptors. GR is a member of the steroid – thyroid – retinoic-acid receptor ligand-activated transcription gene superfamily [63]. Over the last decade great progress has been made in understanding the molecular mechanisms of action of this gene superfamily [62]. These receptors regulate the expression of target genes by binding directly or indirectly to cis-acting sequences. The GR binds as a homodimer to a consensus glucocorticoid response element (GRE) DNA sequence of inverted repeats. This is composed of two sixbase-pair half sites arranged as an imperfect palindrome with a three base-pair interval [64]. A number of bonerelated genes have been demonstrated to contain GREs, including the human osteocalcin gene promoter which contains a negative GRE [65], the interleukin 1 gene [66], and the bone sialoprotein (BSP) gene promoter [67,68]. In addition to the classical DNA-dependent mechanisms, DNA-independent GR-mediated transcription can occur through protein – protein interactions with DNAbound transcriptional activators, cofactors [69] and/or components of the basal transcription machinery. For example, the collagenase A gene lacks a GRE, but GR interacts via a DNA bound AP1 transcription factor complex consisting of a Jun – Jun homodimer [70]. Likewise the bone sialoprotein promoter has an overlapping GRE/AP-1 site [68]. All members of the nuclear hormone receptor family share a characteristic three-domain structure. The Nterminal domain is the least conserved and may vary greatly in length, containing sequences responsible for activation of target genes (activation-function-1 or AF-1 domain). The DNA-binding domain (DBD) is a highly conserved cysteine-rich region that forms two zincfinger structures through which the receptor interacts with DNA. This domain also participates in receptor dimerization and is involved in translocation to the nucleus. The C-terminal ligand-binding domain (LBD) is a less well conserved and hydrophobic region in all receptors. This domain possesses the sequences important for ligand recognition and ensures both specificity and selectivity of the physiological response. It also contains regions involved in heat shock protein binding, nuclear translocation, receptor dimerization, and ligand-dependent transactivation [63].

174

GARY M. LEONG ET AL

C. In Vivo and in Vitro Localization of the Glucocorticoid Receptor In rat osteoblastic cells cytosolic binding studies show the presence of specific glucocorticoid binding sites [71,72]. Specific binding in vitro has also been shown in human osteoblasts and osteosarcoma cells [73 – 76]. GR mRNA has been detected in human osteoblasts [73] and the human cell lines SaOS-2, MG-63, and TE85 [77 – 79]. The recent in vivo localization in human bone of GR protein confirm the wide ranging targets of glucocorticoids and imply some functional consequences on bone formation. Using immunohistochemistry and a polyclonal antibody against GR, Abu et al. [80] localized GR in neonatal, adolescent, and adult bone samples from various sites. GR was expressed predominantly in hypertrophic chondrocytes and chondrocytes within the cartilage, as well as in osteoblasts at sites of bone growth and modeling in neonatal bone and in osteocytes in adult bone. Within the bone marrow, GR was detected in mononuclear cells and in endothelial cells of blood vessels. Strikingly, GR expression was not detected in osteoclasts. Overall, these results suggest that any effects of glucocorticoids on bone resorption are mediated indirectly and that the primary cell targets of glucocorticoids include osteoblasts and hypertrophic chondrocytes in cartilage [80].

D. Effects of Glucocorticoids on Osteoblast Function and Survival 1. ANIMAL Rodent models of the effects of glucocorticoids on bone have been inconsistent, with both positive and negative effects on bone [81 –83]. In both rats and mice glucocorticoidmediated effects on osteoblast differentiation occur at very specific developmental time points, and are most potent on proliferating cells [84,85]. The effects of glucocorticoids in vitro appear to be species-dependent. Osteoprogenitor cells from rat and mice differ in their response to glucocorticoids with corticosterone stimulating cell growth and differentiation in rat calvaria cultures, but inhibiting these processes in mouse cultures [81,86]. Also in a murine osteoblast cell line (MC3T3E1), chronic dexamethasone exposure blocks osteoblast differentiation, in contrast to enhancing it in rat calvarial cultures [84]. By contrast to rat data, in mice, it has been reported recently [87] that glucocorticoids in vivo lead to inhibition of osteoblastogenesis and increase in osteoblast apoptosis. These results are more consistent with the human in vivo data. 2. HUMAN The average life span of an osteoblast is 3 months and osteoclast is 2 weeks. It has been estimated that the majority (50 – 70%) of osteoblasts and osteoclasts die by apoptosis

[48]. The remaining osteoblasts either become embedded in mineral matrix as osteocytes (90% of cells in bone) or become elongated lining cells covering newly formed bone surface. It has been clearly demonstrated in glucocorticoidinduced osteoporosis that there are decreased osteoblast numbers on bone surfaces [88]. Further evidence suggests that the mechanism for the decreased bone mass observed secondary to glucocorticoids is due to increased apoptosis or cell death of bone cells [48]. The major histologic features of glucocorticoid-induced osteoporosis are decreased bone formation rate, decreased trabecular wall thickness, and apoptosis of bone cells [48] (Fig. 1, see also color plates). In bone obtained from patients on chronic prednisone treatment, increased presence of apoptotic osteoblast cells has been observed compared to bone from normal controls [89]. Furthermore, the osteocyte-canaliculi-lining cell network was reported to be disrupted, suggesting interference with bone repair and hence increased bone fragility [48]. In vitro studies on human bone cells conflict with in vivo findings and indicate that, contrary to being detrimental, glucocorticoids are actually required for normal osteoblast differentiation and may prevent osteoblast apoptosis under certain conditions [90]. In human osteoblast cultures, dexamethasone increased alkaline phosphatase activity [73, 91 – 93] and was required for production of mineralizing bone “nodules” [94]. The effect of dexamethasone on osteocalcin concentration was more variable [73,91] and in primary osteoblast cultures these effects may depend on donor age [93]. Thus, glucocorticoids may be essential for normal osteoblast differentiation including mineralization but, at high pharmacological concentrations, may inhibit osteoblast function and lead to enhanced apoptosis.

E. Effects of Glucocorticoids on Osteoclast Function and Survival 1. ANIMAL In rodent cells, glucocorticoids have been shown to inhibit rat osteoclast function and decrease their survival [95]. However, no effects on osteoclast apoptosis or histomorphometric changes in bone resorption were observed in a recent study of rat long bones in which high-dose corticosterone (10 mg/day) significantly increased osteoblastic apoptosis and histomorphometric evidence of decreased bone formation [96]. Interestingly, there was no evidence of colocalization of GR with apoptosis induction, suggesting indirect modulation of the apoptotic effects of glucocorticoids. In mice, the principal effects of glucocorticoids are to stimulate bone resorption and osteoclast formation [97,98], an effect which appears to be mediated directly by the GR. Glucocorticoids may also modulate osteoclast responsiveness by

FIGURE 1 Glucocorticoid-induced osteoporosis. (A) Apoptotic osteoblasts (arrows) are identified by the brown staining and nuclear condensation. In situ detection of DNA fragmentation, a marker of apoptosis, was by the TUNEL reaction (transferase-mediated biotin–dUTP nick end-labeling). The section was counterstained with methyl green and viewed by Nomarski differential interference contrast microscopy, original magnification 400. (B) Chronic glucocorticoid therapy causes the accumulation of apoptotic osteocytes and lining cells. Original magnification 400. (Courtesy of Dr. Robert S. Weinstein, University of Arkansas.) (See also color plates.)

175

176

GARY M. LEONG ET AL

regulating expression of calcitonin receptor [99]. Thus, overall the animal models do not really clarify the human data as bidirectional effects on osteoblast and osteoclast function and activity have been observed. 2. HUMAN While it has been recognized for many years that glucocorticoids have major effects on decreasing bone formation, glucocorticoids effects on osteoclast function have been more controversial with conflicting findings. Glucocorticoids are reported to inhibit OPG and concurrently stimulate the expression of RANKL in human osteoblasts and primary bone marrow cells in vitro [100]. These changes would be expected to promote osteoclastogenesis. However long-term effects of glucocorticoids on osteoclast function are inhibitory rather than stimulatory. These stimulatory effects on osteoclast differentiation may possibly occur at an early and/or transient phase of glucocorticoid-induced osteoporosis [101]. These observations may explain the finding that osteoclast surface is not increased by glucocorticoids [102]. Moreover, glucocorticoids have been shown to inhibit rather than elevate markers of bone resorption [20,103] (Fig. 2).

F. Cellular Mechanisms by Which Glucocorticoids May Exert Their Effect on Bone Cells 1. GLUCOCORTICOID EFFECTS ON THE GROWTH HORMONE (GH) AND INSULIN-LIKE GROWTH FACTOR (IGF) AXIS As IGFs and glucocorticoids have opposite effects on bone formation, it is plausible that changes in the IGF axis would be involved in inhibitory actions of glucocorticoids in vitro and in vivo. In vitro IGF-I stimulates human osteoblast proliferation and partly reverses the glucocorticoidinduced inhibition of osteoblast proliferation [104]. Glucocorticoids inhibit GH release in normal men in response to GHRH [105], though serum IGF-I concentrations have been reported to be normal in subjects with glucocorticoidinduced osteoporosis [106,107]. IGF-I and IGF-II are weak mitogens that increase the replication of cells of the osteoblast lineage [108]. Skeletal cells express IGF-I and -II receptors (see Chapter 14). In FIGURE 2 (A) Overnight diurnal osteocalcin secretion is suppressed in Cushing’s syndrome. The normal diurnal pattern of secretion of osteocalcin is shown in a normal, healthy 15-year-old girl (open circles), in contrast to its complete suppression in her hypercortisolemic identical twin (closed circles) secondary to Cushing’s disease (preoperative 24-h urinary free cortisol was 1509 nmol, NR 55 – 281 nmol) [20,277]. (B). Changes in markers of bone turnover in the Cushing twin (closed circles) and her identical healthy, co-twin (open circles) at diagnosis and during 27 months follow-up after surgical cure of Cushing disease. At diagnosis in the Cushing twin

there was marked suppression of markers of bone formation (osteocalcin) and resorption (24-h pyridinium crosslinks). These significantly increased after surgical cure of the Cushing disease, suggesting a state of increased coupled bone turnover 20. The age and sex-matched normal range for osteocalcin, 24 hour urinary pyridinoline and deoxypyridinoline are shown within the hatched areas (Corning Nichols Institute, Vienna, VA). These data are adapted from Leong et al. with permission20.

177

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

osteoblasts, glucocorticoids decrease IGF-II synthesis by transcriptional mechanisms [109,110], while glucocorticoids have inconsistent effects on the level of IGF-I receptors [110 – 112]. Glucocorticoids may alter the binding of one or more of the six different serum IGF binding proteins (IGFBP-1- to -6) to modulate their function. All the IGFBPs, which are expressed on osteoblasts are considered important in storage and transport of IGFs locally [113,114]. Specifically, glucocorticoids increase IGFBP-1 levels [115], an effect which has been associated with glucocorticoid-induced fetal growth retardation [116]. In human bone marrow cells, dexamethasone increases IGFBP-2 levels in association with increases in IGF-II, but decreases IGFBP-3, -5, and -6 levels that could increase bioavailability of IGF-II [117]. Glucocorticoids also decreased expression of IGFBP-3, -4, and -5 in osteoblasts [118,114]. While IGFBP-4 inhibits, IGFBP-5 stimulates bone cell growth [119,120]. Cortisol increases the activity of collagenase 3, an enzyme known to degrade IGFBP-5 and therefore may modify the stability of the IGFBP-5 [121]. As IGFBP-5 stimulates bone cell growth and enhances the action of IGF-I, the reduced level of this binding protein in the bone microenvironment may be important in the inhibitory effects of glucocorticoids on bone formation. IGFBP-6 binds IGF-II with 20 – 100 times more affinity than with IGF-I, whereas other binding proteins have similar affinities for the two IGFs [121]. IGFBP-6 inhibits the effect of IGF-II on DNA and glycogen synthesis in osteoblasts, but has only minor effects on IGF-mediated actions [122]. In contrast to its effects in bone marrow cells, in osteoblasts cortisol increases IGFBP-6 transcription and synthesis causing a decrease in free IGF-II concentration [123]. In rat calvarial cells cortisol also causes a dose- and time-dependent increase in Mac25 gene expression and its associated gene product, the IGFBP-related-Peptide-1 (IGBP-r-P1). As IGFBP-r-P1 binds IGF-I and insulin, its increased expression may also contribute to the inhibitory effects of glucocorticoids on bone [124]. Cortisol may also increase collagenase activity and its effect on collagen breakdown [125]. Thus glucocorticoids could lead to decreases in free IGF-I and IGF-II in bone cells which may contribute to the inhibitory actions of glucocorticoids on bone formation. While IGF-I protects some other cell types from apoptosis, e.g., chondrocytes [126], it is not known whether it affects osteoblast apoptosis. Although such data are not available for osteoblasts, glucocorticoid effects on osteoblast survival may in part be mediated through IGF-I modulation of apoptosis. Recent evidence suggests that glucocorticoids at physiological concentration have synergistic effects with cyclic AMP-inducing hormones and transiently induce CCAATenhancer binding proteins (C/EBP and C/EBP ) expression in osteoblasts [127]. C/EBP directly binds the HS3D

FIGURE 3

Model for glucocorticoid action in the osteoblast and adipocyte differentiation pathways. Glucocorticoids at physiological levels promote C/EBP expression which upregulates IGF-1 and IGFBP-5 gene transcription and leads to bone promoting effects. These effects also occur secondary to PGE2 and intermittent PTH administration. By contrast, glucocorticoids at pharmacological levels cause suppression of IGF-1 transcriptional activity and increases C/EBP and PPAR-2 expression to promote adipocyte differentiation at the expense of osteoblast differentiation and function.

DNA element in the rat, human, and chicken IGF-I promoter, upregulating transcription of the IGF-I gene in osteoblasts [128]. Furthermore, glucocorticoids may modulate protein kinase A (PKA) and protein kinase C (PKC) effects on IGF-I expression. These effects may occur through glucocorticoid modulation of both PTH and PGE2 [129 – 131]. At pharmacological levels of glucocorticoids, the above pathways are inhibited, ultimately suppressing IGF-I expression. C/EPBs also play a major role in adipogenesis through upregulation of PPAR-2 expression [132]. In the presence of pharmacological levels of glucocorticoids, glucocorticoid-mediated adipogenesis occurs at the expense of osteoblastogenesis [49] (Fig. 3). 2. GLUCOCORTICOID EFFECTS ON TRANSFORMING GROWTH FACTOR- (TGF-) AND BONE MORPHOGENETIC PROTEINS (BMPS) Transforming growth factor- (TGF-) plays a major role in bone remodeling (see Chapter 14). It is anabolic for bone formation though its effects on bone resorption depend on the experimental model and the culture conditions [133]. TGF- binds to three cell surface receptors (TGF-)-1, -2, and -3, which are all expressed in osteoblast cells [134,135]. Recently the intracellular transducers for TGF- signaling have been identified as the Smad proteins [136,137]. Interestingly, the liganded GR interacts with Smad3 proteins to repress TGF- signaling, a finding consistent with earlier observations showing glucocorticoids decrease plasminogen activator activity in normal rodent osteoblast and UMR10601 cells [138] through upregulation of the plasminogen

178 activator inhibitor (PAI-1) mRNA and protein levels [139]. The PAI-1 gene has been shown to possess Smad3-binding sites through which multiple factors, including GR, can modulate its activity [140]. These observations suggest an ever increasing complexity of GR regulation on cell function and bone modeling/remodeling. The rat TGF- type 1 receptor-1 (TR-1) promoter contains binding sites for Cbfa1 which increase in parallel with osteoblast differentiation [141]. Glucocorticoids in part decrease TR-1 expression through downregulation of Cbfa1 expression [142]. In addition the TGF- type 3 receptor promoter possesses glucocorticoid sensitive sites which may play a similar role in bone matrix production [143]. These actions may have inhibitory effects on matrixproduction and contribute to the bone fragility observed in glucocorticoid-induced osteoporosis. Bone morphogenetic proteins (BMPs) induce cartilage and bone differentiation in vivo and promote osteoblast differentiation from calvarial and marrow stromal cell preparations [144] (see Chapters 5 and 14). In fetal rat calvarial cell culture system BMP-2, -4, and -6 had differential effects on early osteoblast differentiation in association with glucocorticoids or recombinant BMP [145]. BMP-6 was a more potent inducer of osteoblast differentiation than BMP-2 or -4. The effects of all three of these BMPs were potentiated up to 10-fold by cotreatment or pretreatment with the glucocorticoid triamcinolone [146]. Thus, glucocorticoids at physiological concentrations promote osteoblast differentiation from fetal calvarial cells, calvarial organ cultures, and bone marrow stromal cells [50]. 3. GLUCOCORTICOIDS EFFECTS ON ADIPOGENESIS Glucocorticoids increase the production of peroxisome proliferator-activated receptor (PPAR)-2 promoting adipogenesis at the expense of osteoblastogenesis [142,132]. The PPAR ligand TZD (5-(4-N-methyl-N(2-pyridyl)amino) ethoxy)benzyl-thiazolidione-2,4-dione) decreased alkaline phosphatase activity and expression of collagen type 1 mRNA levels in a murine osteoblast (MB 1.8) and human osteosarcoma cell line (SaOs-2/B10), but not a non-bonederived cell line (CV1) [147]. At the same time this PPAR ligand upregulated adipocyte fatty acid binding protein and enhanced the effect of dexamethasone to stimulate transcription of a glucocorticoid-inducible reporter gene (MMTV-luciferase). As the GR antagonist RU 486 was able to block the dexamethasone and TZD responses these effects appear to be mediated via endogenous GR. 4. GLUCOCORTICOIDS AND PROSTAGLANDINS Prostaglandins (PGs) produced by bone cells are potent stimulators of bone formation and resorption [148] (see Chapter 13). They also have inhibitory effects on fully differentiated osteoblasts and osteoclasts. This complex, multifunctional regulation may be mediated by different PG

GARY M. LEONG ET AL

receptors. Endogenous PGs are produced by osteoblasts in response to parathyroid hormone (PTH), interleukin-1 (IL-1), TGF-, and to mechanical strain, by induction of prostaglandin G/H synthase (PGHS)-2 (COX-2) which is inhibited by glucocorticoids both basally and at an induced level [149 – 151]. Transgenic PGHS-2 deficient mice exhibit reduced osteoclast formation in response to 1,25(OH)2D3 and PTH. This effect was reversed by exogenous PGE2, and may be mediated by reduced RANKL expression on osteoblasts [152]. In rodent organ calvarial cultures PGs have both stimulatory and inhibitory effects on bone formation depending on the dose and hormonal milieu [153,154]. Dose-dependent inhibition of osteoclast formation and lacunar resorption was seen when PGE2 was added to human peripheral blood mononuclear cell (PBMC) cultures in the presence or absence of dexamethasone [155]. Calcitriol, 1,25(OH)2D3 in the presence of RANKL and macrophage colony-stimulating factor (M-CSF), decreases osteoclast differentiation and function. Dexamethasone, however, reverses these 1,25(OH)2D3-dependent effects on osteoclast differentiation and function with a marked increase in resorption pit formation. Thus, 1,25(OH)2D3 and PGE2 not only influence osteoclast formation in the presence of bone stromal cells, but also act directly on circulating osteoclast precursors to influence osteoclast differentiation. Glucocorticoids may augment this process, as has been observed in mouse bone cell cultures treated with dexamethasone and PGE2, as well as dexamethasone with hPTH 1 – 34 [98]. IGFBP-5 synthesis is also upregulated by PGE2, PTH, and other agents that stimulate cAMP synthesis [149,156,157]. 5. GLUCOCORTICOIDS AND CYTOKINES Interleukin-1 (IL-1) and interleukin-6 (IL-6) are involved in postmenopuasal bone loss, in which bone resorption is increased and bone formation decreased (see Chapters 13 and 41). Glucocorticoids inhibit the expression and action of most cytokines [158]. The IL-6 gene contains a GRE through which GR may block cytokine binding and transcriptional effects [159]. In addition, GR binds to and inhibits the activity of proinflammatory transcription factors (e.g., AP-1 and NFkB via protein – protein interactions, sometimes in a DNA-independent manner [160,161]. The inhibition of interleukin-2 (IL-2) transcription is proposed to depend on such a mechanism [162]. For the IL-6 and IL-8 genes, binding of NFkB has been reported through GR association with the p65 subunit of NFkB [163,164]. Treatment of osteoblast cells with dexamethasone increases expression of interleukin-6 receptor (IL-6R) mRNA. Osteoblastic cells from transgenic mice constitutively expressing human IL-6R could support osteoclast development in the presence of human IL-6 in cocultures with normal spleen cells. In contrast, osteoclast progenitors in spleen cells from transgenic mice overexpressing human

179

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

IL-6R were not able to differentiate into osteoclasts in response to IL-6 in cocultures with normal osteoblastic cells. These results clearly indicate that the ability of IL-6 to induce osteoclast differentiation depends on signal transduction mediated by IL-6R expressed on osteoblastic cells but not on osteoclast progenitors. In this regard it is interesting that dexamethasone, but not sex steroids, inhibited constitutive and cytokine-stimulated IL-6 secretion from human osteoblasts [165]. 6. GLUCOCORTICOID EFFECTS ON NOVEL BONE-RELATED GENES The integrin proteins, a family of heterodimeric transmembrane glycoproteins, are the major cell surface receptors responsible for cell adhesion to the matrix. Human osteoblasts express a variety of integrins that bind matrix proteins such as osteopontin, bone sialoprotein, vitronectin, type 1 collagen, thrombospondin, and fibronectin [166] (see Chapter 4). As integrins control cell migration and adhesion of several bone matrix proteins through interaction with V3 and V5 integrins, they may be involved in glucocorticoid effects on osteoblast survival. Dexamethasone causes a time-dependent regulation of V3 and V5 expression in normal human osteoblast cells [166]. Short-term exposure to dexamethasone increased the abundance of the integrins on the surface and cell adhesion to osteopontin and vitronectin, whereas long-term exposure (8 days) decreased expression of both integrins and inhibited cell adhesion to matrix proteins. These studies also demonstrated transcriptional effects in MC3-T3-E-1 osteoblast-like cells of dexamethasone on the 3 and 5 promoter of these genes. The marked dexamethasone-mediated suppression of V3 and associated decrease in osteoblast – osteopontin interaction may lead to inhibition of osteoblast differentiation, and hence at least partly explain glucocorticoid-mediated effects on osteoblast cell survival and differentiation [147,167,169]. Other possible effects of glucocorticoids on other genes involved in cell adhesion, such as the cadherins, may also play a role in modulating glucocorticoids effects on osteoblast differentiation [170]. Glucocorticoids also have effects on the synthesis of extracellular matrix proteins in human bone marrow stromal osteoprogenitor cells and mature osteoblasts. Specifically, glucocorticoids modulate small proteoglycans, increasing decorin and decreasing biglycan gene expression [171], which may lead to changes in bone cell integrity (see Chapter 4). Glucocorticoids inhibit proliferation of many cell types, but the events leading from the activated GR to growth arrest are not understood. Glucocorticoids may modulate cell proliferation through direct effects on cell cycle regulatory proteins such as cyclin-dependent kinases (CDKs) CDK4 and CDK6, as well as their regulatory partner, cyclin D3 [172]. Glucocorticoids reduce expression of E2F-1 and c-Myc, transcription factors involved in the G1-to-S-phase

transition, and induce expression of the CDK inhibitors (CDIs) p27 and p21. Thus, GR regulation of the cell cycle machinery may involve transcriptional repression of G1 cyclins and CDKs as well as involving enhanced transcription of CDIs. Glucocorticoids in vitro also modulate expression of a number of other growth factor-related genes, including decreasing hepatocyte growth factor/ scatter factor in rat calvaria cells and in MC3T3 cells [173] and in human osteoblast cells [174].

VII. GLUCOCORTICOIDS AND CALCIUM HOMEOSTASIS A. Effect of Glucocorticoids on Parathyroid Hormone (PTH) Function and Activity Secondary hyperparathyroidism from decreased intestinal calcium absorption and hypercalciuria has been proposed for many years as one mechanism for the observed bone loss in glucocorticoid-induced osteoporosis. However, increased, unchanged, or decreased concentrations of PTH have been found in various studies of patients receiving exogenous glucocorticoids or in endogenous Cushing’s syndrome [103,175 – 178]. Differences in assay type with different hormone fragments may explain some of these conflicting results [175,179 – 181]. Using intact PTH or mid-region fragment PTH measurements, unchanged PTH concentrations have been reported [43,182 – 185]. Glucocorticoids may enhance the sensitivity of osteoblast responsiveness to PTH by increasing expression of PTH receptors [186]. Parathyroidectomy prevents the excessive bone resorption in rats associated with glucocorticoids, consistent with enhanced secretion of or sensitivity to PTH by rat osteoblasts [187]. However, while glucocorticoids acutely stimulate bone resorption in rat bone cultures, they inhibit bone resorption in long-term cultures [188]. Thus, as there is no consistent change in PTH following glucocorticoid administration, a role for altered PTH secretion and function is not established.

B. Effect of Glucocorticoids on Gut Calcium Absorption and Vitamin D Metabolites Malabsorption of calcium can occur within 2 weeks of glucocorticoid treatment though it is not a universal phenomenon. Normal gut calcium absorption has been reported in glucocorticoid-treated patients except in those with fractures [106,189 – 192]. Glucocorticoids induce the expression of calbindin-D28K, a protein involved in intestinal calcium transport [177,192,193]. Vitamin-D-dependent

180

GARY M. LEONG ET AL

calcium binding protein is reduced in chickens on glucocorticoids, but there are no comparable human data [194]. Defective vitamin D metabolism resulting in hyperparathyroidism has also been proposed as a mechanism for the bone loss seen with glucocorticoids. However, no change in vitamin D metabolites has been documented in patients with glucocorticoid-induced-osteoporosis [195]. Thus it is highly unlikely that changes in vitamin D metabolism play a major role in glucocorticoid-induced osteoporosis [196].

C. Effect of Glucocorticoids on Renal Calcium Handling Glucocorticoid excess promotes hypercalciuria with fasting urinary calcium excretion double that of control patients [197]. There is some evidence that tubular reabsorption of calcium is decreased [197]. However, this may only occur at high equivalent doses of glucocorticoids (25 mg prednisone/day) in which case the increase in urinary calcium may be explained by a greater filtered load secondary to increased serum calcium concentration [183]. Acutely, i.e., within 2 h of an intravenous dose of methylprednisone, serum 1,25(OH)2D3 concentrations rise and those of phosphorus decrease. These changes were associated with progressive but transient increases in circulating PTH. In this study, all biochemical variables had returned to normal by 3 weeks despite continuation of oral prednisone [198]. The mechanisms of these responses to calcium flux in the kidney are unclear, but may include direct effects of glucocorticoids on renal tubular calcium reabsorption. While some studies of patients on longer-term exposure to glucocorticoids or endogenous Cushing’s syndrome have found elevated PTH and 1,25(OH)2D3 concentration and reduced tubular reabsorption of phosphate [180,185,199], this finding has not been universal. This may relate to the underlying pathological states necessitating glucocorticoid therapy, the dose of glucocorticoid used, and the age of subjects.

D. Glucocorticoid Effects on Sex Steroid Hormone Regulation of Bone Metabolism High levels of circulating glucocorticoids may lead to changes in sex hormone production. The effects occur at various levels of the hypothalamic – pituitary – gonadal axis including suppression of lutenizing hormone (LH) and follicle-stimulating hormone (FSH) secretion from the pituitary [20,200,201]. In addition, glucocorticoids have been shown to modify the hypothalamic – pituitary – gonadal axis response [202]. They enhance androgen receptor expression in human osteoblast cells and, with estradiol and

1,25 (OH)2D3, enhance the mitogenic response to dihydrotestosterone [203]. This may relate to enhanced aromatase cytochrome p450 gene activity [204]. These results suggest that bone tissue can synthesize estrogen from adrenal androgens by a unique aromatase activity depending on the VDR level. Glucocorticoids have direct effects on gonads, inhibiting FSH-induced estrogen production by ovarian granulosa cells and testosterone production by Leydig cells with a decrease of nearly 50% in serum testosterone in men [205 – 208]. There is a direct correlation between BMD and plasma estradiol concentration in women treated with glucocorticoids [209], while women receiving estrogen – progesterone replacement therapy and men given medroxyprogesterone acetate while taking glucocorticoids were protected against bone loss [210,211]. Overall these effects lead to decreases in various circulating sex steroids which may potentially contribute to the bone loss in glucocorticoid-induced osteoporosis.

E. Summary Glucocorticoids have myriad effects on bone cells, which may in part, be mediated through effects on calcium homeostasis and the GH – IGF-I – sex-steroid axes. However, the recent molecular evidence suggests that a major effect of glucocorticoids on bone cells involves compromising osteoblast survival through enhancement of apoptosis. These effects are dependent on the concentrations of circulating hormone, as at physiological levels glucocorticoids play a major role in osteoblast differentiation. Osteoblast differentiation may in part be mediated by various growth factors, such as IGF-I/IGFBP-5, TGF-, and BMPs, as well as other factors, such as prostaglandins and cytokines. Future studies will continue to unravel the complex molecular regulation of bone cells by glucocorticoids, and may lead to the development of specific therapeutic targets aimed at preventing or treating glucocorticoid-induced osteoporosis.

VIII. TREATMENT OPTIONS There are several issues to consider in reviewing studies on the efficacy of therapies available for glucocorticoidinduced osteoporosis. These include primary versus secondary prevention, underlying disease states, and fracture prevention. A distinction needs to be made between primary and secondary prevention trials. The former commence at or soon after glucocorticoid therapy is begun as the greatest rate of bone loss due to glucocorticoids occurs within the first 6 – 12 months of use. Secondary prevention, or treatment trials are those commencing after the subject

181

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

TABLE 1

Approach to Management of Patients Receiving Glucocorticoids

1. Minimize systemic glucocorticoid use. If possible use inhaled or topical glucocorticoids. 2. Maintain a healthy lifestyle including performing adequate exercise. In long-term glucocorticoid recipients, initiate a strength and endurance training program to minimize glucocorticoid-induced muscle weakness to reduce falls risk. 3. Ensure good dietary calcium intake (1000 – 1500 mg/day in adults) and vitamin D sufficiency. If in doubt, supplement with vitamin D (400 – 1000 IU/day). 4. Treat sex hormone deficiency in both men and women. 5. Use 1-hydroxylated vitamin D metabolites/analogs if glucocorticoid use is likely to be long-term. 6. Use potent bisphosphonates for prevention, if glucocorticoid use is likely to be long-term, and for treatment if evidence of osteoporosis by BMD or fracture. 7. Monitor bone mineral density

has already taken glucocorticoids for a significant period of time and in whom bone loss has already occurred. This section will deal with primary and secondary prevention studies separately. Glucocorticoids are used to treat a wide variety of conditions. These include rheumatological diseases such as rheumatoid arthritis, respiratory diseases such as asthma, and gastroenterological diseases such as primary biliary cirrhosis; they are also beneficial in the prevention of transplant rejection. This leads to considerable heterogeneity in the data and studies may not be comparable as the disease itself may cause osteoporosis in its own right, particularly if it is not well controlled. In addition, studies vary in the degree of osteoporosis at baseline, length of time on glucocorticoids, and age of subjects. There are few data on fracture outcome, the important clinical measure, as most studies have been insufficiently powered to detect effects on fracture and have used bone mineral density as the outcome measure. Despite these difficulties, there is substantial evidence for the effectiveness of a number of therapies used in the prevention and treatment of glucocorticoid-induced osteoporosis. Thus in the light of this knowledge there is no excuse for ignoring this devastating complication of glucocorticoid use. An approach to the management of patients receiving glucocorticoids is summarized in Table 1.

A. Calcium There are no controlled trials of calcium alone versus placebo in the prevention or treatment of glucocorticoidinduced osteoporosis as most studies have used calcium in conjuction with other therapies. However, it has been used in the placebo arms of several studies [182,212 – 214].

From these studies it is clear that calcium alone does not prevent the bone loss that occurs on initiation of glucocorticoids and it does not increase bone density in subjects who are already osteoporotic. However, inadequate calcium intake in postmenopausal osteoporosis has been shown to result in bone loss. Thus subjects prescribed glucocorticoids are advised to have an adequate calcium intake of 1500 mg per day due to deleterious effects on gut calcium absorption and on renal calcium reabsorption.

B. Vitamin D Trials involving simple vitamin D (calciferol) and its active metabolites, calcitriol (1,25(OH)2D3) and alphacalcidiol (1(OH) D) and dihydrotachysterol, are often discussed as if these treatments are synonymous. This is potentially misleading as the latter involve pharmacological doses, whereas the former generally involves therapeutic dosing. In addition, vitamin D is metabolized to compounds other than its active form that may have as yet unknown effects on bone metabolism. To date there have been very few direct comparisons between simple vitamin D and its activated metabolites; thus this review will consider them separately. 1. PRIMARY PREVENTION A recent randomized placebo controlled prevention study of calciferol (50,000 IU/week) plus calcium (1000 mg/day) in a group of patients with mixed rheumatic diseases found no beneficial effect of treatment at 3 years, although lumbar bone density decreased to a lesser extent in the treatment arm at one year [215]. In contrast, calcium plus calcidiol (40 g/day) in transplant patients resulted in a lesser decline in lumbar bone density compared with historical controls. Unfortunately, this study was not placebo controlled [216]. 2. SECONDARY PREVENTION The results of treatment studies have also shown mixed results. The use of 50,000 IU of calciferol three times a week and 25-OH vitamin D (40 g/day) has been reported to prevent bone loss and produce small increases in BMD [181,177]. A recent 2-year study in rheumatoid arthritis subjects on low-dose glucocorticoids found an increase in BMD in both the lumbar spine and trochanter in those supplemented with calcium 1000 mg and calciferol 500 IU/day while the placebo group lost bone [217]. However, 1000 mg calcium plus a lower dose of vitamin D (250 IU/day) conferred no benefit after 1 year in a group of patients taking glucocorticoids for inflammatory bowel disease [218]. It is unclear whether these differences relate to the different patient groups and underlying disease states, or the glucocorticoid dosages, baseline vitamin D status, or dosages of vitamin D given.

182

GARY M. LEONG ET AL

C. Activated Vitamin D Metabolites 1. PRIMARY PREVENTION Calcitriol was demonstrated to be effective in preventing bone loss at the lumbar spine in a group of patients starting glucocorticoids for a mixture of conditions [182]. In this study, patients were randomized to receive calcium 1000 mg, calcium plus calcitriol (0.5 – 1.0 g/day), or calcium, calcitriol, and calcitonin (400 IU/day). The groups receiving calcitriol with or without calcitonin lost less bone at the lumbar spine (0.2 and 1.3%, respectively) than the group receiving calcium alone (4.3%) at 1 year. However, there was no protection at the femoral neck or distal radius. In the second year only the group receiving calcitriol and calcitonin were protected against further bone loss compared with the calcium-alone or calcitriol groups. However, glucocorticoid doses were greater in the calcitriol group. Alphacalcidiol plus calcium was also found to be effective in preventing bone loss over 1 year in a more recent study in a mixed group of patients just commencing glucocorticoids (bone loss of 1 – 5.7% in the calcium alone arm compared with 0.4% increase in the treatment arm) [219]. 2. SECONDARY PREVENTION There have been very few treatment or secondary prevention studies using the activated metabolites of vitamin D. One such study, showed no benefit of calcitriol but both control and treatment groups received simple vitamin D (400 IU/day) [214]. The studies were all too small to show differences in fracture as an endpoint but a recent metaanalysis concluded that there was a trend to fracture reduction in the treatment groups [220]. In conclusion, vitamin D or its activated metabolites used together with calcium supplementation results in less bone loss than calcium alone and appear to maintain bone when given to long-term glucocorticoid users. However, the results are variable and the different compounds and dosages used in the studies make it difficult to identify or propose the most effective regime. In the only small comparison trial between vitamin D and its activated metabolites in subjects with established glucocorticoid-induced osteoporosis, 1 g alphacalcidiol plus calcium was more effective than 1000 IU vitamin D plus calcium over 3 years [221]. Perhaps most importantly, none of the vitamin D compounds appear to be as effective as bisphosphonate therapy (see below), where evidence for benefit is more consistent [220,222]. These data underline the clear recommendation that vitamin D replacement should be given for any patient who is or may be vitamin D deficient. Calcium supplementation should be given for all patients but used cautiously in those receiving calcitriol due to the risk of hypercalcemia and hypercalciuria. Therapy is best begun on initiation of glucocorticoids and the choice of activated

vitamin D compounds versus bisphosphonates should be individualized depending on baseline bone density, age of patient, duration of glucocorticoid use, and response to treatment.

D. Bisphosphonates Bisphosphonates are potent inhibitors of bone resorption that have clearly been shown to reduce fracture risk in postmenopausal osteoporosis [223,224] (see Chapters 16 and 72). The efficacy of a number of bisphosphonates with differing antiresorptive potencies has been reported in various studies of glucocorticoid-induced osteoporosis. In general these studies have found a significant beneficial effect in terms of BMD, but as the changes in bone density in many of these studies have been followed for only 12 months it is possible they are related to the bone remodeling transient. Few studies have been sufficiently powered to evaluate effi cacy in fracture prevention. Among those studies where incidence of vertebral fracture has been monitored there is a consistent trend for patients receiving bisphosphonates to have a lower incidence of vertebral fracture [222,225 – 227], but larger studies of longer duration are required to confirm these trends. Interestingly, bisphosphonates, along with PTH, have been recently found to have antiapoptotic effects on osteoblasts and osteocytes [228], which may explain part of their effectiveness in glucocorticoid-induced osteoporosis. However, a recent study of the acute changes in bone markers in women on high doses of prednisone found that etidronate increased osteocalcin concentrations [229]. These authors proposed an effect of either PTH or 1,25(OH)2D3 on bone formation as a potential mechanism for this effect. Comprehensive studies are required to elucidate this possibility. 1. PRIMARY PREVENTION Primary prevention studies are few in comparison with secondary prevention/treatment studies but there are several evaluating the effectiveness of cyclical etidronate. In one 12-month randomized placebo-controlled study, patients were recruited within 100 days of commencing high-dose therapy with prednisone or equivalent [222]. In the etidronate group, lumbar spine and trochanteric BMD did not change significantly during the 12-month study period, whereas it declined in the placebo group. There was a significant between-group difference in the percentage change in BMD of 3.7% for the lumbar spine (P  0.02), 4.1% for the trochanter (P  0.02), and 1.9% for the femoral neck (P  0.63). The percentage change in BMD in all three of the etidronate treated subgroups (men, premenopausal women, and postmenopausal women) was favorable compared with that of the comparable control subgroups.

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

Early administration of intravenous pamidronate has also been found to effectively protect against glucocorticoid-induced osteoporosis [230]. In patients starting high dose prednisone, mainly for inflammatory rheumatic conditions, 30 mg IV pamidronate at 3-month intervals resulted in differences in BMD at 12 months of 8.9 and 7.5% at the spine and femoral neck, respectively, compared with calcium supplementation alone. The usefulness of alendronate in preventing glucocorticoid-related bone loss has been assessed in patients with sarcoidosis [231]. Forty-three previously untreated patients participated in the study. One group (n  13), whose disease was not severe enough to require glucocorticoid therapy acted as the control group. The remaining 30 patients treated with glucocorticoids were randomly divided to receive a placebo or 5 mg oral alendronate per day and were followed for 12 months. BMD of the ultradistal radius, measured with DPA, increased 0.8% in the group treated with alendronate compared with a bone loss of 4.5% in the patients on placebo and 0.6% in the patients not receiving glucocorticoids. In these patients there was a significant reduction in markers of bone formation independent of treatment with alendronate but the increase in bone resorption markers secondary to glucocorticoids was counteracted by alendronate. The efficacy of risedronate as primary prevention of glucocorticoid-induced osteoporosis has also been reported in a randomized, double-blind, placebo-controlled study in 224 men and women commencing long-term glucocorticoid treatment for diverse conditions [226]. All participants received 500 mg of elemental calcium daily plus 2.5 or 5 mg of risedronate or a placebo. Risedronate prevented bone loss so that at 12 months there was 3.8% difference in BMD between the 5 mg risedronate group and the control group. 2. SECONDARY PREVENTION There have been numerous studies of secondary prevention of glucocorticoid-induced osteoporosis using bisphosphonates. In a cohort study of 68 individuals with a variety of disorders necessitating glucocorticoid use for a minimum of 1 year, BMD of the lumbar spine increased in the patients treated with cyclical etidronate (3.8%) but decreased in the comparison group (1.8%) [232]. The difference between groups was significant (P 0.001) in the lumbar spine but there were no differences within or between groups in femoral neck BMD. In part these effects, as previously mentioned, may largely be due to the remodeling transient. In another cohort study with a 36-month follow-up period, changes in BMD in patients treated with etidronate therapy were compared with those in untreated controls [233]. There was an increase in lumbar spine BMD in the group treated with etidronate; however, most of this in-

183 crease occurred during the first 24 months of treatment. Interestingly, the control group lost bone during the first year but regained sufficient bone during the second and third years, such that there was no residual net change. There were no significant changes in BMD of the hip in either group. In a 2-year placebo-controlled study of 49 patients on long-term glucocorticoids for a variety of disorders, cyclical etidronate increased lumbar spine BMD significantly compared with no change in the placebo group [234], resulting in a between-group difference of 4.5% compared with baseline (P  0.007). There were no differences between groups at any of the hip sites or with bone markers. In patients on long-term glucocorticoid therapy for respiratory disease, collagen vascular disease, or ulcerative colitis, 12 months therapy with oral pamidronate (150 mg/day) plus calcium (1 g/day) produced a 19.6% increase in vertebral BMD measured using quantitative computed tomography compared with a decline of 8.8% in controls receiving calcium alone (P  0.005 for comparison between groups) [212]. Bone histomorphometry and biochemical indices of bone turnover were consistent with decreased bone turnover. Alendronate has also been found to be effective for the treatment of glucocorticoid-induced osteoporosis [225]. In a randomized placebo-controlled study of 477 men and women receiving long-term glucocorticoid therapy for various diseases, the effect of alendronate, 5 or 10 mg/day, on lumbar spine BMD was measured by DXA. At 48 weeks there were significant increases in lumbar spine and femoral neck BMD in patients receiving 5 or 10mg alendronate (P  0.01) compared with decreases in the placebo group. The efficacy of alendronate did not vary according to dose or duration of glucocorticoid therapy. In patients treated with high-dose oral glucorticoid therapy for an average of 5 years, risedronate 2.5 or 5 mg/day with calcium and vitamin D had significant effects on BMD at the lumbar spine, femoral neck, and trochanter [227]. These results were indicative of increased BMD with risedronate therapy, compared with no change in the placebo group. The 5-mg dose was more effective than 2.5 mg, with only the 5-mg group differing significantly from controls at 12 months. The absence of a reduction in BMD of the control group in this and some of the other secondary prevention studies may reflect the administration of vitamin D and calcium, as similar effects were noted in a study of calcitriol in the treatment of patients with heart or lung transplantation [235]. In summary, there is consensus among studies that bisphosphonates increase bone density at the spine with a smaller improvement seen at the hip. Meta-analysis suggests that the response to bisphosphonate therapy is

184

GARY M. LEONG ET AL

greater in primary than in secondary prevention trials. The major effect of bisphosphonates in primary prevention is to attenuate bone loss while in the treatment of established osteoporosis, bisphosphonates result in bone accrual [236].

E. Hormone Replacement Therapy 1. ESTROGEN There have been no primary prevention studies of hormone replacement therapy (HRT), although there have been a few studies of secondary prevention in postmenopausal women. In these studies, women receiving estrogen and calcium increased lumbar spine BMD (2 – 4%), while controls lost bone [211,237]. This benefit of HRT seems to be similar to data in postmenopausal women who are not on glucocorticoids. There are no data on the combined use of bisphosphonates and HRT in glucocorticoid-induced osteoporosis but in postmenopausal osteoporosis, alendronate plus HRT increased bone density more than HRT alone; thus one would suspect that this would also be the case in women receiving glucocorticoids. Hormone replacement therapy has not been used in premenopausal women and for those who are cycling normally, as there is no place for hormonal therapy in this situation. However, there is evidence that oligomenorrheic or amenorrheic premenopausal women have a lower bone mass than eugonadal women, leading to the proposition that the oral contraceptive pill or hormone replacement should be given to hypogonadal women on glucocorticoids. 2. TESTOSTERONE Like estrogen, testosterone therapy in men has only been used in secondary prevention trials based on the observation that testosterone levels are reduced in men on glucocorticoids [24]. In a crossover study of asthmatic men who had received high-dose glucocorticoids for a mean of 8 years, treatment with intramuscular testosterone resulted in a 5% increase in lumbar BMD, but no changes in femoral BMD after 1 year, compared with no change in the group not given testosterone. In addition, there were increases in lean mass and decreases in fat mass in the testosterone treated group [238].

F. Anabolic Steroids Nandronlone decanoate was studied in a randomized trial of women on long-term glucocorticoids for a variety of rheumatic conditions [239]. After 18 months of therapy, the group receiving the active treatment had a mean increase in forearm bone density of 5% compared with the control

groups who lost bone. However, virilizing side effects were described in half of the subjects, although they were considered acceptable by all subjects. Thus the role of anabolic glucocorticoids in treatment of osteoporosis is still controversial.

G. Calcitonin Calcitonin has been studied in several randomized controlled trials using both subcutaneous and intranasal preparations. For the primary prevention trials, a number of studies demonstrated less bone loss at the lumbar spine in the calcitonin than the placebo arm [240,241], although there was no demonstrated benefit at the hip [241], and in one study no beneficial effect of calcitonin over and above calcitriol [182]. Nor was there any additional benefit of calcitonin over 2 years in another study where all subjects were given calcium and 400 IU vitamin D [242]. In treatment studies, both nasal (100 and 200 IU/day) and subcutaneous (100 U twice daily) calcitonin resulted in an increase or stabilization of bone density (increases between 2 and 4% at the lumbar spine) compared with placebo groups where bone density declined (2.5 – 8%) [243 – 246]. Thus calcitonin appears to be beneficial, especially in secondary prevention, but patient acceptability is less than other treatments, particularly for the subcutaneous route, where side effects can be significant.

H. Sodium Fluoride Unlike the antiresorptive agents described above, fluoride is an anabolic agent affecting osteoblasts (see Chapters 74 and 75). It is therefore particularly attractive in glucocorticoid-induced osteoporosis as the major effect of glucocorticoids on suppressing osteoblast results in decreased bone formation. Low-dose monofluorophosphate and sodium fluoride has been shown to increase spinal bone density in subjects with established osteoporosis without a demonstrated increase in bone density at the hip [247 – 250]. However, there is no proven anti-fracture efficacy of fluoride. Indeed, fluoride has resulted in increased vertebral fractures in postmenopausal osteoporosis despite increases in bone density [251]. Thus fluoride should only be considered experimental in glucocorticoid-induced osteoporosis.

I. PTH Daily subcutaneous PTH is a novel therapy that has recently been shown to increase bone mass in established

185

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

glucocorticoid-induced osteoporosis [252] (see Chapter 77). In a 1-year randomized controlled study of daily PTH plus hormone replacement therapy versus hormone replacement therapy alone, the women taking PTH experienced an 11% increase in spinal bone density and a smaller increase (2%) at the hip. Two-year follow-up revealed maintenance of the spinal increase in bone density and a further increase in hip density to 4.7% [253]. Changes in bone remodeling were suggested by elevations of bone formation and resorption markers, with the former peaking earlier. Unlike the bisphosphonates that act via decreasing osteoclast resorption, PTH is an anabolic agent, stimulating bone formation. Indeed, continuous higher concentrations of PTH, as in hyperparathyroidism, are associated with widespread bone resorption. Recent evidence suggests that PTH prevents osteoblast and osteocyte apoptosis [49,254] which is a major pathological mechanism of glucocorticoid-induced bone loss. PTH is thus a potentially exciting development for the treatment of glucocorticoid-induced osteoporosis but as yet is still experimental and not available for regular use [255].

IX. TREATMENT AND FRACTURE OUTCOMES Although fracture incidence is relatively high in patients on long-term systemic glucocorticoid therapy, studies on fracture prevention have been relatively few. Studies that have addressed prevention of glucocorticoid-related bone loss have not been designed to address fracture outcomes. Within these limitations, bisphosphonates, including risedronate, alendronate, and etidronate have all been shown to reduce vertebral fractures in glucocorticoid-induced osteoporosis. Etidronate significantly reduced the incidence of vertebral fractures from 0.3 per person-year in the control group of glucocorticoids to 0.1 per person-year in the etidronate group [222]. Among postmenopausal women, there was an 85% reduction in the proportion with vertebral fracture during the study period in the etidronate group. In two recent studies, risedronate reduced fracture incidence in patients receiving glucocorticoids by as much as 70% [226,227]. In a study of alendronate there was no overall difference between groups in the incidence of vertebral fractures during the study period [225]. However, when postmenopausal women receiving gluccorticoids were considered alone, the alendronate group had significantly fewer fractures [225]. As previously mentioned, calcium and vitamin D have not been shown to be effective in reducing fracture rates in patients on long-term glucocorticoid therapy [215]. In addition, not all studies using calcium supplementation for their control groups have shown a protective effect [226]. No studies on the effects of sex hormone therapy or 1-vitamin D metabolites in glucocorticoid-induced

osteoporosis have been adequately powered to address changes in fracture incidence. None of the studies of calcitonin use in glucocorticoid-induced osteoporosis that examined fractures, demonstrated any significant decrease in fracture rates, but the sample sizes of these studies have been small. In summary, there is evidence of improvement in BMD with several agents in glucocorticoid-induced osteoporosis. Bisphosphonates have been the main effective agents studied for the prevention of fractures, though other agents may also be effective [255].

X. MANAGEMENT OF GLUCOCORTICOID-INDUCED OSTEOPOROSIS IN CHILDREN Very few prospective studies have been undertaken in children to determine the long-term effects on bone mass of chronic glucocorticoid administration [20]. However, as the large majority of peak bone mass is acquired during the first two decades of life [256] (see Chapter 25), glucocorticoid exposure during this critical time of bone mass accrual may impact negatively on peak bone mass [257]. While some studies in children with osteoporosis using bisphosphonates have now been reported [258], few, if any, such studies in children with glucocorticoid-induced osteoporosis have been conducted [259]. The mainstay of management in children with glucocorticoid-induced-osteoporosis involves minimization of glucocorticoid exposure. This can be achieved either through reduction of the total glucocorticoid dose, use of alternative daily dosage regimens, or replacement of systemic glucocorticoids with inhaled or topical glucocorticoids, which may lessen the adverse side effects on bone mass accrual and growth [260 – 264]. In addition, optimization of other factors are warranted, including adequate calcium and vitamin D intake as per current recommended intakes [265 – 267], exercise [268 – 270], and regular follow-up of bone mass by DXA [271,272]. Normative data for childhood BMD are now available from various populations [39,273 – 276].

References 1. H. Cushing, The basophil adenomas of the pituitary body ad their clinical manifestations (pituitary bsophilism). Bull Johns Hopkins Hosp. 50, 137 – 195 (1932). 2. I. R. Reid, Glucocorticoid-induced osteoporosis — Assessment and treatment. J. Clin. Densitom. 1, 65 – 73 (1998). 3. B. A. Michel, D. A. Bloch, and J. F. Fries, Predictors of fractures in early rheumatoid arthritis. J. Rheum. 18, 804 – 808 (1991). 4. A. D. Adinoff, and J. R. Hollister, Steroid-induced fractures and bone loss in patients with asthma. N. Engl. J. Med. 309, 265 – 268 (1983). 5. L. J. Walsh, C. A. Wong, M. Pringle, and A. E. Tattersfield, Use of oral corticosteroids in the community and the prevention of

186

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

23.

24. 25.

GARY M. LEONG ET AL secondary osteoporosis: A cross sectional study. Br. Med. J. 313, 344 – 346 (1996). I. D. Peat, S. Healy, D. M. Reid, and S. H. Ralston, Steroid induced osteoporosis: An opportunity for prevention? Ann. Rheum. Dis. 54, 66 – 68 (1995). A. Randell, P. N. Sambrook, T. V. Nguyen, et al., Direct clinical and welfare costs of osteoporotic fractures in elderly men and women. Osteoporosis Int. 5, 427 – 432 (1995). N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton, Medical Expenditures For the Treatment of Osteoporotic Fractures in the United States in 1995 - Report From the National Osteoporosis Foundation. Journal of Bone & Mineral Research 1997; 12: 24 – 35. T. P. Van Staa, H. G. M. Leufkens, L. Abenhaim, B. Zhang, and C. Cooper, Use of oral corticosteroids and risk of fractures. J. Bone Miner. Res. 15, 993 – 1000 (2000). N. K. Henderson, P. N. Sambrook, P. J. Kelly, et al., Bone mineral loss and recovery after cardiac transplantation [letter]. Lancet 346, 905 (1995). V. LoCascio, P. Ballanti, S. Milani, et al., A histomorphometric longterm longitudinal study of trabecular bone loss in glucocorticoidtreated patients: Prednisone versus deflazacort. Calci. Tissue Int. 62, 199 – 204 (1998). P. Sambrook, J. Birmingham, S. Kempler, et al., Corticosteroid effects on proximal femur bone loss. J. Bone Miner. Res. 5, 1211 – 1216 (1990). J. F. Marystone, E. L. Barrett-Connor, and D. J. Morton, Inhaled and oral corticosteroids: Their effects on bone mineral density in older adults. Am. J. Public Health 85, 1693 – 1695 (1995). R. F. Laan, P. L. van Riel, L. B. van de Putte, L. J. van Erning, M. A. van’t Hof, and J. A. Lemmens, Low-dose prednisone induces rapid reversible axial bone loss in patients with rheumatoid arthritis. A randomized, controlled study. Ann. Intern. Med. 119, 963 – 968 (1993). I. R. Reid, M. C. Evans, and J. Stapleton, Lateral spine densitometry is a more sensitive indicator of glucocorticoid-induced bone loss. J. Bone Miner. Res. 7, 1221 – 1225 (1992). G. Rizzato and L. Montemurro, Reversibility of exogenous corticosteroid-induced bone loss. Eur. Respir. J. 6, 116 – 119 (1993). N. A. Pocock, J. A. Eisman, C. R. Dunstan, R. A. Evans, D. H. Thomas, and N. L. Huq, Recovery from steroid-induced osteoporosis. Ann. Intern. Med. 107, 319 – 323 (1987). P. J. Manning, M. C. Evans, and I. R. Reid, Normal bone mineral density following cure of Cushing’s syndrome. Clin. Endocrinol. 36, 229 – 234 (1992). A. R. Hermus, A. G. Smals, L. M. Swinkels, et al. Bone mineral density and bone turnover before and after surgical cure of Cushing’s syndrome. J. Clin. Endocrinol. Metab. 80, 2859 – 2865 (1995). G. M. Leong, L. B. Mercado-Asis, J. C. Reynolds, S. C. Hill, E. H. Oldfield, and G. P. Chrousos. The effect of Cushing’s disease on bone mineral density, body composition, growth, and puberty: A report of an identical adolescent twin pair. J. Clin. Endocrinol. Metab. 81, 1905 – 1911 (1996). M. Luengo, C. Picado, L. Del Rio, N. Guanabens, J. M. Montserrat, and J. Setoain, Vertebral fractures in steroid dependent asthma and involutional osteoporosis: a comparative study. Thorax 46, 803 –806 (1991). N. F. Peel, D. J. Moore, N. A. Barrington, D. E. Bax, and R. Eastell, Risk of vertebral fracture and relationship to bone mineral density in steroid treated rheumatoid arthritis. Ann. Rheum. Dis. 54, 801 – 806 (1995). P. L. Selby, J. P. Halsey, K. R. H. Adams, et al. Corticosteroids do not alter the threshold for vertebral fracture. J. Bone Miner. Res. 15, 952 – 956 (2000). I. R. Reid, Glucocorticoid osteoporosis — Mechanisms and management. Eur. J. Endocrinol. 137, 209 – 217 (1997). A. Verstraeten and J. Dequeker, Vertebral and peripheral bone mineral content and fracture incidence in postmenopausal patients with

26. 27. 28. 29.

30.

31. 32.

33.

34. 35.

36.

37. 38.

39.

40.

41.

42.

43.

44.

45. 46.

rheumatoid arthritis: Effect of low dose corticosteroids. Ann. Rheum. Dis. 45, 852 – 857 (1986). C. Cooper, C. Coupland, and M. Mitchell, Rheumatoid arthritis, corticosteroid therapy and hip fracture. Ann. Rheum. Dis. 54, 49 – 52 (1995). S. C. Manolagas Corticosteroids and fractures: A close encounter of the third cell kind. J. Bone Miner. Res. 15, 1001 – 1005 (2000). A. Askari, P. J. Vignos, Jr., and R. W. Moskowitz, Steroid myopathy in connective tissue disease. Am. J. Med. 61, 485 – 492 (1976). T. Nguyen, P. Sambrook, P. Kelly, et al. Prediction of osteoporotic fractures by postural instability and bone density. Br. Med. J. 307, 1111 – 1115 (1993). T. K. LaPier, Glucocorticoid-induced muscle atrophy. The role of exercise in treatment and prevention. J. Cardiopulm. Rehab. 17, 76 – 84 (1997). W. D. McArdle, F. I. Katch, and V. L. Katch, “Exercise Physiology. Williams & Wilkins, Baltimore, 1996. M. T. Falduto, S. M. Czerwinski, and R. C. Hickson, Glucocorticoidinduced muscle atrophy prevention by exercise in fast-twitch fibres. J. Appl. Physiol. 63, 2048 – 2052 (1990). R. W. Braith, M. A. Welsch, R. M. Mills, Jr., J. W. Keller, and M. L. Pollock, Resistance exercise prevents glucocorticoid-induced myopathy in heart transplant recipients. Med. Sci. Sports Exercise 30, 483 – 489 (1998). P. Weiner, Y. Azgad, and M. Weiner, Inspiratory muscle training during treatment with corticosteroids in humans. Chest 107, 1041–1044 (1995). P. N. Sambrook, J. A. Eisman, M. G. Yeates, N. A. Pocock, S. Eberl, and G. D. Champion, Osteoporosis in rheumatoid arthritis: Safety of low dose corticosteroids. Ann. Rheum. Dis. 45, 950 – 953 (1986). L. M. Buckley, E. S. Leib, K. S. Cartularo, P. M. Vacek, and S. M. Cooper, Effects of low dose corticosteroids on the bone mineral density of patients with rheumatoid arthritis. J. Rheum. 22, 1055 – 1059 (1995). P. N. Sambrook and G. Jones, Corticosteroid osteoporosis. Br. J. Rheumatol. 34, 8 – 12 (1995). B. P. Lukert, Glucocorticoid-induced osteoporosis. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), pp. 292 – 296. Lippincott Williams & Wilkins, Philadelphia, 1999. R. Eastell, Management of corticosteroid-induced osteoporosis. UK Consensus Group Meeting on Osteoporosis. J. Intern. Med. 237, 439 – 447 (1995). N. E. Lane, B. Lukert, The science and therapy of glucocorticoidinduced bone loss. Endocrinol. Metab. Clin. North Am. 27, 465 – 483 (1998). E. M. Pouw, M. F. Prummel, H. Oosting, C. M. Roos, and E. Endert, Beclomethasone inhalation decreases serum osteocalcin concentrations. Br. Med. J. 302, 627 – 628 (1991). G. P. Bootsma, P. N. Dekhuijzen, J. Festen, P. G. Mulder, L. M. Swinkels, C. L. van Herwaarden, Fluticasone propionate does not influence bone metabolism in contrast to beclomethasone dipropionate. Am. J. Respir. Crit. Care Med. 153, 924 – 930 (1996). H. K. Nielsen, K. Thomsen, E. F. Eriksen, P. Charles, T. Storm, and L. Mosekilde, The effects of high-dose glucocorticoid administration on serum bone gamma carboxyglutamic acid-containing protein, serum alkaline phosphatase and vitamin D metabolites in normal subjects. Bone Miner. 4, 105 – 113 (1988). A. K. Gough, J. Lilley, S. Eyre, R. L. Holder, and P. Emery, Generalised bone loss in patients with early rheumatoid arthritis. Lancet 344, 23 – 27 (1994). A. Woodcock, Effects of inhaled corticosteroids on bone density and metabolism. J. Allergy Clin. Immunol. 101, S456 – S459 (1998). T. C. Medici, E. Grebski, M. Hacki, P. Ruegsegger, C. Maden, and J. Efthimiou, Effect of one year treatment with inhaled fluticasone propionate or beclomethasone dipropionate on bone density and bone metabolism: A randomised parallel group study in adult asthmatic subjects. Thorax 55, 375 – 82 (2000).

CHAPTER 44 Glucocorticoid-Induced Osteoporosis 47. J. Herrala, H. Puolijoki, O. Impivaara, K. Liippo, E. Tala, and M. M. Nieminen, Bone mineral density in asthmatic women on highdose inhaled beclomethasone dipropionate. Bone 15, 621 – 623 (1994). 48. R. S. Weinstein and S. C. Manolagas, Apoptosis and osteoporosis. Am. J. Med. 108, 153 – 164 (2000). 49. S. C. Manolagas and R. S. Weinstein, New developments in the pathogenesis and treatment of steroid-induced osteoporosis. J. Bone Miner. Res. 14, 1061 – 1066 (1999). 50. J. E. Aubin, Bone stem cells. J. Cell. Biochem. 73 – 82 (1998). 51. S. C. Manolagas and R. L. Jilka, Mechanisms of disease — Bone marrow, cytokines, and bone remodeling — Emerging insights into the pathophysiology of osteoporosis N. Engl. J. Med. 332, 305 – 311 (1995). [Review] 52. E. Abe, M. Yamamoto, Y. Taguchi, et al. Essential requirement of BMPs-2/4 for both osteoblast and osteoclast formation in murine bone marrow cultures from adult mice: Antagonism by noggin. J. Bone Miner. Res. 15, 663 – 673 (2000). 53. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1 — A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 54. P. Ducy, M. Starbuck, M. Priemel, et al. A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). 55. Y. Y. Kong, W. J. Boyle, and J. M. Penninger, Osteoprotegerin ligand: A common link between osteoclastogenesis, lymph node formation and lymphocyte development. Immunol. Cell Biol. 77, 188 – 193 (1999). 56. D. L. Lacey, E. Timms, H. L. Tan, et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 57. J. Li, I. Sarosi, X. Q. Yan, et al. RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Nat. Acad. Sci. USA 97, 1566 – 1571 (2000). 58. W. S. Simonet, D. L. Lacey, C. R. Dunstan, et al. Osteoprotegerin — A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 59. S. M. Hollenberg, C. Weinberger, E. S. Ong, et al. Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318, 635 – 641 (1985). 60. C. M. Bamberger, H. M. Schulte, and G. P. Chrousos, Molecular determinants of glucocorticoid receptor function and tissue sensitivity to glucocorticoids. Endocr. Rev. 17, 245 – 261 (1996). 61. E. R. Sanchez, J. L. Hu, S. Zhong, P. Shen, M. J. Greene, and P. R. Housley, Potentiation of glucocorticoid receptor-mediated gene expression by heat and chemical shock. Mol. Endocrinol. 8, 408 – 421 (1994). 62. C. K. Glass and M. G. Rosenfeld, The coregulator exchange in transcriptional functions of nuclear receptors. Genes Dev. 14, 121 – 141 (2000). 63. L. L. Issa, G. M. Leong, and J. A. Eisman, Molecular mechanism of vitamin D receptor action. Inflammation Res. 47, 451 – 475 (1998). 64. M. Beato, Gene regulation by steroid hormones. Cell 56, 335 – 344 (1989). 65. N. A. Morrison, J. Shine, J. C. Fragonas, V. Verkest, M. L. McMenemy , and J. A. Eisman, 1,25-dihydroxyvitamin D-responsive element and glucocorticoid repression in the osteocalcin gene. Science 246, 1158 – 1161 (1989). 66. G. Zhang, L. Zhang, and G. W. Duff, A negative regulatory region containing a glucocorticosteroid response element (nGRE) in the human interleukin-1beta gene. DNA Cell Biol. 16, 145 – 152 (1997). 67. J. Sodek, R. H. Kim, Y. Ogata, et al. Regulation of bone sialoprotein gene transcription by steroid hormones. Connect. Tissue Res. 32, 209 – 217 (1995).

187 68. M. Yamauchi, Y. Ogata, R. H. Kim, J. J. Li, L. P. Freedman, and J. Sodek. Ap-1 regulation of the rat bone sialoprotein gene transcription is mediated through a Tpa response element within a glucocorticoid response unit in the gene promoter. Matrix Biol. 15, 119 – 130 (1996). 69. A. B. Hittleman, D. Burakov, J. A. Iniguez-Lluhi, L. P. Freedman, and M. J. Garabedian. Differential regulation of glucocorticoid receptor transcriptional activation via AF-1-associated proteins. EMBO J. 18, 5380 – 5388 (1999). 70. M. I. Diamond, J. N. Miner, S. K. Yoshinaga, and K. R. Yamamoto. Transcription factor interactions: Selectors of positive or negative regulation from a single DNA element. Science 249, 1266 – 1272 (1990). 71. T. L. Chen, L. Aronow, and D. Feldman, Glucocorticoid receptors and inhibition of bone cell growth in primary culture. Endocrinology 100, 619 – 628 (1977). 72. S. C. Manolagas and D. C. Anderson, Detection of high-affinity glucocorticoid binding in rat bone. J. Endocrinol. 76, 379 – 380 (1978). 73. M. Subramaniam, D. Colvard, P. E. Keeting, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg. Glucocorticoid regulation of alkaline phosphatase, osteocalcin, and proto-oncogenes in normal human osteoblast-like cells. J. Cell. Biochem. 50, 411 – 424 (1992). 74. P. Liesegang, G. Romalo, M. Sudmann, L. Wolf, and H. U. Schweikert. Human osteoblast-like cells contain specific, saturable, high-affinity glucocorticoid, androgen, estrogen, and 1 alpha,25-dihydroxycholecalciferol receptors. J. Androl. 15, 194 – 199 (1994). 75. L. N. Song. Effects of retinoic acid and dexamethasone on proliferation, differentiation, and glucocorticoid receptor expression in cultured human osteosarcoma cells. Oncol. Res. 6, 111 – 118 (1994). 76. R. Bland, C. A. Worker, B. S. Noble, et al. Characterization of 11beta-hydroxysteroid dehydrogenase activity and corticosteroid receptor expression in human osteosarcoma cell lines. J. Endocrinol. 161, 455 – 464 (1999). 77. A. Mahonen and P. H. Maenpaa. Steroid hormone modulation of vitamin D receptor levels in human MG-63 osteosarcoma cells. Biochem. Biophys. Res. Commun. 205, 1179 – 1186 (1994). 78. S. Tanaka, M. Haji, R. Takayanagi, Y. Sugioka, and H. Nawata, 1,25-Dihydroxyvitamin D3 enhances the enzymatic activity and expression of the messenger ribonucleic acid for aromatase cytochrome P450 synergistically with dexamethasone depending on the vitamin D receptor level in cultured human osteoblasts. Endocrinology 137, 1860 – 1869 (1996). 79. J. Codon, C. Gosden, D. Gardener, et al. Expression of type 2 11beta-hydroxysteroid dehydrogenase and corticosteroid hormone receptors in early human fetal life. J. Clin. Endocrinol. Metab. 83, 4490 – 4497 (1998). 80. E. O. Abu, A. Horner, V. Kusec, J. T. Triffitt, and J. E. Compston, The localization of the functional glucocorticoid receptor alpha in human bone. J. Clin. Endocrinol. Metab. 85, 883 – 889 (2000). 81. C. G. Bellows, A. Ciaccia, and J. N. M. Heersche, Osteoprogenitor cells in cell populations derived from mouse and rat calvaria differ in their response to corticosterone, cortisol, and cortisone. Bone 23, 119 – 125 (1998). 82. J. L. Ferretti, S. O. Vazquez, C. J. Delgado, R. Capozza, and G. Cointry, Biphasic dose-response curves of cortisol effects on rat diaphyseal bone biomechanics. Calcif. Tissue Int. 50, 49 – 54 (1992). 83. M. Li, Y. Shen, B. P. Halloran, B. D. Baumann, K. Miller, and T. J. Wronski, Skeletal response to corticosteroid deficiency and excess in growing male rats. Bone 19, 81 – 88 (1996). 84. J. B. Lian, V. Shalhoub, F. Aslam, et al. Species-specific glucocorticoid and 1,25-dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: dexamethasone inhibits osteoblast differentiation and vitamin D down-regulates osteocalcin gene expression. Endocrinology 138, 2117 – 2127 (1997). 85. S. M. Pockwinse, J. L. Stein, J. B. Lian, and G. S. Stein, Developmental stage-specific cellular responses to vitamin D and glucocorticoids

188

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

GARY M. LEONG ET AL during differentiation of the osteoblast phenotype: Interrelationship of morphology and gene expression by in situ hybridization. Exp. Cell Res. 216, 244 – 260 (1995). S. Advani, D. LaFrancis, E. Bogdanovic, P. Taxel, L. G. Raisz, and B. E. Kream, Dexamethasone suppresses in vivo levels of bone collagen synthesis in neonatal mice. Bone 20, 41 – 46 (1997). R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids — Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). E. Canalis, Clinical review 83: Mechanisms of glucocorticoid action in bone: Implications to glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 81, 3441 – 3447 (1996). R. S. Weinstein, Apoptosis of osteocytes in glucocorticoid-induced osteonecrosis of the hip. J. Clin. Endocrinol. Metab. 85, 2907 – 2912 (2000). T. Nakashima, H. Sasaki, M. Tsuboi, et al. Inhibitory effect of glucocorticoid for osteoblast apoptosis induced by activated peripheral blood mononuclear cells. Endocrinology 139, 2032 – 2040 (1998). M. M. Wong, L. G. Rao, H. Ly, et al. Long-term effects of physiologic concentrations of dexamethasone on human bone-derived cells. J. Bone Miner. Res. 5, 803 – 813 (1990). K. Iba, H. Chiba, N. Sawada, S. Hirota, S. Ishii, and M. Mori, Glucocorticoids induce mineralization coupled with bone protein expression without influence on growth of a human osteoblastic cell line. Cell Struct. Funct. 20, 319 – 330 (1995). M. S. K. Sutherland, L. G. Rao, S. A. Muzaffar, et al. Age-dependent expression of osteoblastic phenotypic markers in normal human osteoblasts cultured long term in the presence of dexamethasone. Osteoporosis Int. 5, 335 – 343 (1995). J. Janssen, R. Bland, M. Hewison, et al.. Estradiol formation by human osteoblasts via multiple pathways: Relation with osteoblast function. J. Cell. Biochem. 75, 528 – 537 (1999). D. W. Dempster, B. S. Moonga, L. S. Stein, W. R. Horbert, and T. Antakly, Glucocorticoids inhibit bone resorption by isolated rat osteoclasts by enhancing apoptosis. J. Endocrinol. 154, 397 – 406 (1997). G. Silvestrini, P. Ballanti, F. R. Patacchioli, et al. Evaluation of apoptosis and the glucocorticoid receptor in the cartilage growth plate and metaphyseal bone cells of rats after high-dose treatment with corticosterone. Bone 26, 33 – 42 (2000). H. H. Conaway, D. Grigorie, and U. H. Lerner, Stimulation of neonatal mouse calvarial bone resorption by the glucocorticoids hydrocortisone and dexamethasone. J. Bone Miner. Res. 11, 1419 – 1429 (1996). H. Kaji, T. Sugimoto, M. Kanatani, K. Nishiyama, and K. Chihara, Dexamethasone stimulates osteoclast-like cell formation by directly acting on hemopoietic blasts cells and enhances osteoclast-like cell formation stimulated by parathyroid hormone and prostaglandin E2. J. Bone Miner. Res. 12, 734 – 741 (1997). S. Wada, N. Udagawa, T. Akatsu, N. Nagata, T. J. Martin, and D. M. Findlay, Regulation by calcitonin and glucocorticoids of calcitonin receptor gene expression in mouse osteoclasts. Endocrinology 138, 521 – 529 (1997). L. C. Hofbauer, F. Gori, B. L. Riggs, et al. Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140, 4382 – 4389 (1999). S. C. Manolagas, Editorial: Cell number versus cell vigor — What really matters to a regenerating skeleton? Endocrinology 140, 4377 – 4381 (1999). D. W. Dempster, Bone histomorphometry in glucocorticoid-induced osteoporosis. J. Bone Miner. Res. 4, 137 – 141 (1989).

103. G. Pearce, D. A. Tabensky, P. D. Delmas, H. W. Baker, and E. Seeman, Corticosteroid-induced bone loss in men. J. Clin. Endocrinol. Metab. 83, 801 – 806 (1998). 104. K. B. Jonsson, A. Frost, R. Larsson, S. Ljunghall, and O. Ljunggren, A new fluorometric assay for determination of osteoblastic proliferation — Effects of glucocorticoids and insulin-like growth factor-I. Calcif. Tissue Int. 60, 30 – 36 (1997). 105. S. Kaufmann, K. L. Jones, W. B. Wehrenberg, and F. L. Culler, Inhibition by prednisone of growth hormone (GH) response to GHreleasing hormone in normal men. J. Clin. Endocrinol. Metab. 67, 1258 – 1261 (1988). 106. B. P. Lukert and L. G. Raisz, Glucocorticoid-induced osteoporosis: Pathogenesis and management. Ann. Intern. Med. 112, 352 – 364 (1990). 107. M. Gourmelen, F. Girard, and M. Binoux. Serum somatomedin/ insulin-like growth factor (IGF) and IGF carrier levels in patients with Cushing’s syndrome or receiving glucocorticoid therapy. J. Clin. Endocrinol. Metab. 54, 885 – 892 (1982). 108. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122, 254 – 60 (1988). 109. T. L. McCarthy, M. Centrella, and E. Canalis, Cortisol inhibits the synthesis of insulin-like growth factor-I in skeletal cells. Endocrinology 126, 1569 – 1575 (1990). 110. S. Rydziel and E. Canalis, Cortisol represses insulin-like growth factor II receptor transcription in skeletal cell cultures. Endocrinology 136, 4254 – 4260 (1995). 111. A. Bennett, T. Chen, D. Feldman, R. L. Hintz, and R. G. Rosenfeld, Characterization of insulin-like growth factor I receptors on cultured rat bone cells: Regulation of receptor concentration by glucocorticoids. Endocrinology 115, 1577 – 1583 (1984). 112. A. M. Delany and E. Canalis, Transcriptional repression of insulinlike growth factor I by glucocorticoids in rat bone cells. Endocrinology 136, 4776 – 4781 (1995). 113. M. M. Rechler, Insulin-like growth factor binding proteins. Vitamins Horm. 47, 1 – 114 (1993). 114. R. Okazaki, B. L. Riggs, and C. A. Conover, Glucocorticoid regulation of insulin-like growth factor-binding protein expression in normal human osteoblast-like cells. Endocrinology 134, 126 – 132 (1994). 115. C. A. Conover, P. D. K. Lee, B. L. Riggs, and D. R. Powell, Insulinlike growth factor-binding protein-1 expression in cultured human bone cells — Regulation by insulin and glucocorticoid. Endocrinology 137, 3295 – 3301 (1996). 116. W. A. Price, A. D. Stiles, B. M. Moats-Staats, and A. J. D’Ercole, Gene expression of insulin-like growth factors (IGFs), the type 1 IGF receptor, and IGF-binding proteins in dexamethasone-induced fetal growth retardation. Endocrinology 130, 1424 – 1432 (1992). 117. S. L. Cheng, S. F. Zhang, S. Mohan, et al. Regulation of insulin-like growth factors I and II and their binding proteins in human bone marrow stromal cells by dexamethasone. J. Cell. Biochem 71, 449 – 458 (1998). 118. B. Gabbitas, J. M. Pash, A. M. Delany, and E. Canalis, Cortisol inhibits the synthesis of insulin-like growth factor-binding protein-5 in bone cell cultures by transcriptional mechanisms. J. Biol. Chem. 271, 9033 – 9038 (1996). 119. D. LaTour, S. Mohan, T. A. Linkhart, D. J. Baylink, and D. D. Strong. Inhibitory insulin-like growth factor-binding protein: Cloning, complete sequence, and physiological regulation. Mol. Endocrinol. 4, 1806 – 1814 (1990). 120. D. L. Andress and R. S. Birnbaum, Human osteoblast-derived insulin-like growth factor (IGF) binding protein-5 stimulates osteoblast mitogenesis and potentiates IGF action. J. Biol. Chem. 267, 22467 – 22472 (1992). 121. L. A. Bach, S. Hsieh, K. Sakano, H. Fujiwara, J. F. Perdue, and M. M. Rechler, Binding of mutants of human insulin-like growth

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134. 135.

136.

137. 138.

139.

factor II to insulin-like growth factor binding proteins 1-6. J. Biol. Chem. 268, 9246 – 9254 (1993). M. C. Kiefer, C. Schmid, M. Waldvogel, et al. Characterization of recombinant human insulin-like growth factor binding proteins 4, 5, and 6 produced in yeast. J. Biol. Chem. 267, 12692 – 12699 (1992). B. Gabbitas and E. Canalis, Cortisol enhances the transcription of insulin-like growth factor-binding protein-6 in cultured osteoblasts. Endocrinology 137, 1687 – 1692 (1996). R. C. Pereira, F. Blanquaert, and E. Canalis, Cortisol enhances the expression of mac25/insulin-like growth factor-binding protein-related protein-1 in cultured osteoblasts. Endocrinology 140, 228 – 232 (1999). E. Canalis, S. Rydziel, A. M. Delany, S. Varghese, and J. J. Jeffrey, Insulin-like growth factors inhibit interstitial collagenase synthesis in bone cell cultures. Endocrinology 136, 1348 – 1354 (1995). R. F. Loeser, and G. Shanker, Autocrine stimulation by insulin-like growth factor 1 and insulin-like growth factor 2 mediates chondrocyte survival in vitro. Arthritis Rheum. 43, 1552 – 1559 (2000). T. L. McCarthy, C. Ji, Y. Chen, K. Kim, and M. Centrella, Time- and dose-related interactions between glucocorticoid and cyclic adenosine 3,5-monophosphate on CCAAT/enhancer-binding proteindependent insulin-like growth factor I expression by osteoblasts. Endocrinology 141, 127 – 137 (2000). Y. Umayahara, J. Billiard, C. Ji, M. Centrella, T. L. McCarthy, and P. Rotwein, CCAAT/enhancer-binding protein delta is a critical regulator of insulin-like growth factor-I gene transcription in osteoblasts. J. Biol. Chem. 274, 10609 – 10617 (1999). T. L. McCarthy, M. Centrella, and E. Canalis, Cyclic AMP induces insulin-like growth factor I synthesis in osteoblast-enriched cultures. J. Biol. Chem. 265, 15353 – 15356 (1990). J. M. Pash, A. M. Delany, M. L. Adamo, C. T. Roberts, D. Jr., LeRoith, and E. Canalis, Regulation of insulin-like growth factor I transcription by prostaglandin E2 in osteoblast cells. Endocrinology 136, 33 – 38 (1995). D. P. Bichell, P. Rotwein, and T. L. McCarthy, Prostaglandin E2 rapidly stimulates insulin-like growth factor-I gene expression in primary rat osteoblast cultures: Evidence for transcriptional control. Endocrinology 133, 1020 – 1028 (1993). X. M. Shi, H. C. Blair, X. L. Yang, J. M. McDonald, and X. Cao, Tandem repeat of C/EBP binding sites mediates PPAR gamma 2 gene transcription in glucocorticoid-induced adipocyte differentiation. J. Cell. Biochem. 76, 518 – 527 (2000). M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). E. Canalis, J. Pash, and S. Varghese, Skeletal growth factors. Crit. Rev. Eukaryotic Gene Expression 3, 155 – 166 (1993). M. Centrella, T. L. McCarthy, and E. Canalis, Glucocorticoid regilation of transforming growth factor beta 1 activity and binding in osteoblast-enriched cultures from fetal rat bone. Mol. Cell. Biol. 11, 4490 – 4496 (1991). E. Piek, C. H. Heldin, and P. Ten Dijke, Specificity, deversity, and regulation in TGF-beta superfamily signaling. FASEB J. 13, 2105 – 2124 (1999). J. Massague and D. Wotton, Transcriptional control by the TGFbeta/Smad signaling system. EMBO J. 19, 1745 – 1754 (2000). J. A. Hamilton, S. Lingelbach, N. C. Partridge, and T. J. Martin, Regulation of plasminogen activator production by bone-resorbing hormones in normal and malignant osteoblasts. Endocrinology 116, 2186 – 2191 (1985). S. Fukumoto, E. H. Allan, R. Zeheb, T. D. Gelehrter, and T. J. Martin, glucocorticoid regulation of plasminogen activator inhibitor-1 messsenger ribonucleic acid and protein in normal and malignant rat osteoblasts. Endocrinology 130, 797 – 804 (1992).

189 140. S. L. Stroschein, W. Wang, and K. Luo, Cooperative binding of Smad proteins to two adjacent DNA elements in the plasminogen activator inhibitor-1 promoter mediates trnsforming growth factor betainduced smad-dependent transcriptional activation. J. Biol. Chem. 274, 9431 – 9441 (1999). 141. C. Banerjee, L. R. McCabe, J. Y. Choi, et al. Runt homology domain proteins in osteoblast differentiation — Am13/Cbfa 1 is a major component of a bone-specific complex. J. Cell. Biochem. 66, 1 – 8 (1997). 142. D. J. Chang, C. Ji, K. K. Kim, S. Casinghino, T. L. McCarthy, and M. Centrella, Reduction in transforming growth factor beta receptor I expression and transcription factor CBFa1 on bone cells by glucocorticoid. J. Biol. Chem. 273, 4892 – 4896 (1998). 143. C. Ji, Y. Chen, T. L. McCarthy, and M. Centrella, Cloning the promoter for transforming growth factor-beta type III receptor. Basal and conditional expression in fetal rat osteoblasts. J. Biol. Chem. 274, 30487 – 30494 (1999). 144. J. M. Schmitt, K. Hwang, S. R. Winn, and J. O. Hollinger, Bone morphogenetic proteins: An update on basic biology and clinical relevance. J. Orthop. Res. 17, 269 – 278 (1999). 145. S. D. Boden, K. McCuaig, G. Hair, et al. Differential effects and glucocorticoid potentiation of bone morphogenetic protein action during rat osteoblast differentiation in vitro. Endocrinology 137, 3401 – 3407 (1996). 146. S. D. Boden, G. Hair, L. Titus, et al. Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 147. T. E. Johnson, R. Vogel, S. J. Rutledge, G. Rodan, and A. Schmidt, Thiazolidinedione effects on glucocorticoid receptor-mediated gene transcription and differentiation in osteoblastic cells. Endocrinology 140, 3245 – 3254 (1999). 148. L. G. Raisz, Prostaglandins and bone: Physiology and pathophysiology. Osteoarthritis Cartilage 7, 419 – 421 (1999). 149. T. L. McCarthy, S. Casinghino, D. W. Mittanck, C. H. Ji, M. Centrella, and P. Rotwein, Promoter-dependent and -independent activation of insulin-like growth factor binding protein-5 gene expression by prostaglandin E2 in primary rat osteoblasts. J. Biol. Chem. 271, 6666 – 6671 (1996). 150. J. L. Masferrer, K. Seibert, B. Zweifel, and P. Needleman, Endogenous glucocorticoids regulate an inducible cyclooxygenase enzyme. Proc. Nat. Acad. Sci. USA 89, 3917 – 3921 (1992). 151. H. Kawaguchi, L. G. Raisz, O. S. Voznesensky, C. B. Alander, Y. Hakeda, and C. C. Pilbeam, Regulation of the two prostaglandin G/H synthases by parathyroid hormone, interleukin-1, cortisol, and prostaglandin E2 in cultured neonatal mouse calvariae. Endocrinology 135, 1157 – 1164 (1994). 152. Y. Okada, J. A. Lorenzo, A. M. Freeman, et al. Prostaglandin G/H synthase-2 is required for maximal formation of osteoclast-like cells in culture. J. Clin. Invest. 105, 823 – 832 (2000). 153. L. G. Raisz and P. M. Fall, Biphasic effects of prostaglandin E2 on bone formation in cultured fetal rat calvariae: Interaction with cortisol. Endocrinology 126, 1654 – 1659 (1990). 154. L. G. Raisz, P. M. Fall, B. Y. Gbbitas, T. L. McCarthy, B. E. Kream, and E. Canalis, Effects of prostaglandin E2 on bone formation in cultured fetal rat calvariae: Role of insulin-like growth factor-1. Endocrinology 133, 1504 – 1510 (1993). 155. I. Itonaga, A. Sabokbar, S. D. Neale, and N. A. Athanasou, 1,25Dihydroxyvitamin D-3 and prostaglandin E-2 act directly on circulating human osteoclast precursors. Biochem. Biophys. Res. Commun. 264, 590 – 595 (1999). 156. T. L. McCarthy, S. Casinghino, M. Centrella, and E. Canalis, Complex pattern of insulin-like growth factor binding protein expression in primary rat osteoblast enriched cultures: regulation by prostaglandin E2, growth hormone, and the insulin-like growth factors. J. Cell. Physiol. 160, 163 – 175 (1994).

190 157. C. A. Conover, L. K. Bale, J. T. Clarkson, and O. Torring, Regulation of insulin-like growth factor binding protein-5 messenger ribonucleic acid expression and protien availablity in rat osteoblast-like cells. Endocrinology 132, 2525 – 2530 (1993). 158. R. Brattsand and M. Linden, Cytokine modulation by glucocorticoids — Mechanisms and actions in cellular studies. Alimentary Pharmacol. Ther. 10, 81 – 90 (1996). 159. A. Ray, K. S. LaForge, and P. B. Sehgal, On the mechanism for efficient repression of the interleukin-6 promoter by glucocorticoids: Enhancer, TATA box, and RNA start (Inr motif) occlusion. Mol. Cell. Biol. 10, 5736 – 5746 (1990). 160. C. Jonat, H. J. Rahmsdorf, K. K. Park, et al. Antitumor promotion and antiinflammation: Down-modulation of AP-1 (Fos/Jun) activity by glucocorticoid hormone. Cell 62, 1189 – 1204 (1990). 161. A. Ray and K. E. Prefontaine, Physical association and functional antagonism between the p65 subunit of transcription factor NFkappa B and the glucocorticoid receptor. Proc. Natl. Acad. Sci. USA 91, 752 – 756 (1994). 162. F. Paliogianni, A. Raptis, S. S. Ahuja, S. M. Najjar, and D. T. Boumpas, Negative transcriptional regulation of human interleukin 2 (IL-2) gene by glucocorticoids through interference with nuclear transcription factors AP-1 and NF-AT. J. Clin. Invest. 91, 1481 – 1489 (1993). 163. A. Ray, D. H. Zhang, M. D. Siegel, and P. Ray, Regulation of interleukin-6 gene expression by steroids. Ann. NY Acad. Sci. 762, 79 – 87 (1995). 164. R. I. Scheinman, A. Gualberto, C. M. Jewell, J. A. Cidlowski, and A. S. Baldwin, Jr. Characterization of mechanisms involved in transrepression of NF-kappa B by activated glucocorticoid receptors. Mol. Cell. Biol. 15, 943 – 953 (1995). 165. T. Hierl, I. Borcsok, U. Sommer, R. Ziegler, and C. Kasperk, Regulation of interleukin-6 expression in human osteoblastic cells in vitro. Exp. Clin. Endocrinol. Diabetes 106, 324 – 333 (1998). 166. S. L. Cheng, C. F. Lai, A. Fausto, et al. Regulation of alpha v beta 3 and alpha v beta 5 integrins by dexamethasone in normal human osteoblastic cells. J. Cell. Biochem. 77, 265 – 276 (2000). 167. Y. K. Liu, T. Uemura, A. Nemoto, et al. Osteopontin involvement in integrin-mediated cell signaling and regulation of expression of alkaline phosphatase during early differentiation of UMR cells. FEBS Lett. 420, 112 – 116 (1997). 168. G. A. Gronowicz and M. B. McCarthy, Glucocorticoids inhibit the attachement of osteoblasts to bone extracellular matrix proteins and decrease beta(1)-integrin levels. Endocrinology 136, 598 – 608 (1995). 169. W. J. Doherty, M. E. Derome, M. B. McCarthy, and GA Gronowicz. The effect of glucocorticoids on osteoblast function — The effect of corticosterone on osteoblast expression of beta(1) integrins. J. Bone J. Surg. 77A, 396 – 404 (1995). 170. F. Lecanda, S. L. Cheng, C. S. Shin, et al. Differential regulation of cadherins by dexamethasone in human osteoblastic cells. J. Cell. Biochem. 77, 499 – 506 (2000). 171. F. Suarez, B. Rothhut, C. Comera, L. Touqui, F. R. Marie, and C. Silve, Expression of annexin-I, annexin-Ii, annexin-V, and annexin-Vi by rat osteoblasts in primary culture — stimulation of annexin-I expression by dexamethasone. J. Bone Miner. Res.8, 1201 – 1210 (1993). 172. I. Rogatsky, J. M. Trowbridge, and M. J. Garabedian, Glucocorticoid receptor mediated cell cycle arrest is achieved through distinct cellspecific transcriptional regulatory mechanisms. Mol. Cell. Biol. 17, 3181 – 3193 (1997). 173. F. Blanquaert, R. C. Pereira, and E. Canalis, Cortisol decrease hepatocyte growth factor/scatter factor expression and induces c-met transcripts in osteoblasts. Am. J. Physiol. Endocrinol. Metab. 278, E509 – E515 (2000). 174. S. Skrtic and C. Ohlsson, Cortisol decreases hepatocyte growth factor levels in human osteoblast-like cells. Calcif. Tissue Int. 66, 108 – 112 (2000).

GARY M. LEONG ET AL 175. R. F. Fucik, S. C. Kukerja, G. K. Hargis, E. N. Bowser, W. J. Henderson, and G. A. Williams, Effect of glucocorticoids on function of the parathyroid glands in man. J. Clin. Endocrinol. Metab. 40, 152 – 155 (1975). 176. E. Paz-Pacheco, G. E. Fuleihan, and M. S. LeBoff, Intact parathyroid hormone levels are not elevated in glucocorticoid-treated subjects. J. Bone Miner. Res. 10, 1713 – 1718 (1995). 177. T. J. Hann, L. R. Halstead, S. L. Teitelbaum, and B. H. Hahn, Altered mineral metabolism in glucocorticoid-induced osteopenia. Effect of 25-hydroxyvitamin D administration. J. Clin. Invest. 64, 655 – 665 (1979). 178. A. T. Hattersley, K. Meeran, J. Burrin, P. Hill, R. Shiner, H. K. Ibbertson, The effect of long- and short-term corticosteroids on plasma calcitonin and parathyroid hormone levels. Calcif. Tissue Int. 54, 198 – 202 (1994). 179. R. Marcus, Secondary forms of osteoporosis. In “Disorders of Bone and Mineral Metabolism” F. L. Coe, and N. J. Favus, eds. Raven Press, New York, 1992. 180. B. P. Lukert and J. S. Adams, Calcium and phosphorus homeostasis in man. Effect of corticosteroids. Arch. Int. Med. 136, 1249 – 1253 (1976). 181. T. J. Hahn and B. H. Hahn, Osteopenia in patients with rheumatic disease: Principles of diagnosis and therapy. Seminars in Arthritis & Rheumatism 6, 165 – 188 (1976). 182. P. Sambrook, J. Birmingham, P. Kelly, et al. Prevention of corticosteroid osteoporosis—A comparison of calcium, calcitriol and calcitonin. N. Engl. J. Med. 328, 1747–1752 (1993). 183. M. F. Prummel, W. M. Wiersinga, P. Lips, G. T. Sanders, and H. P. Sauerwein, The course of biochemical parameters of bone turnover during treatment with corticosteroids. J. Clin. Endocrinol. Metab. 72, 382 – 386 (1991). 184. B. H. Jennings, K. E. Andersson, and S. A. Johansson, The assessment of the systemic effects of inhaled glucocorticoids. The effects of inhaled Budesonide vs oral prednisolone on calcium metabolism. Eur. Clin. Pharmacol. 41, 11 – 16 (1991). 185. D. D. Bikle, B. Halloran, L. Fong, L. Steinbach, and J. Shellito, Elevated 1,25-Dihydroxyvitamin D levels in patients with chronic obstructive pulmonary disease treated with prednisone. J. Clin. Endocrinol. Metab. 76, 456 – 461 (1993). 186. P. Urena, A. Iidaklein, X. F. Kong, et al. Regulation of parathyroid hormone (Pth) Pth-related peptide receptor messenger ribonucleic acid by glucocorticoids and Pth in Ros 17/2.8 and OK cells. Endocrinology 134, 451 – 456 (1994). 187. T. L. Chen and D. Feldman, glucocorticoid receptors and actions in subpopulations of cultured rat bone cells. Mechanism of dexamethasone potention of parathyroid hormone-stimulated cycle AMP production. J. Clin. Invest. 63, 750 – 758 (1979). 188. G. Gronowicz, M. B. McCarthy, and L. G. Raisz, glucocorticoids stimulate resorption in fetal rat parietal bones in vitro. J. Bone Miner. Res. 5, 1223 – 1230 (1990). 189. B. E. Nordin, D. H. Marshall, R. M. Francis, and R. G. Crilly, The effects of sex steroid and corticosteroid hormones on bone. J. Steroid Biochem. 15, 171 – 174 (1981). 190. R. G. Klein, S. B. Arnaud, J. C. Gallagher, H. F. Deluca, and B. L. Riggs, Intestinal calcium absorption in exogenous hypercortisonism. Role of 25-Dihydroxyvitamin D and corticosteroid dose. J. Clin. Invest. 60, 253 – 259 (1977). 191. M. Luengo, C. Picado, C. Piera, et al. Inestinal calcium absorption and parathyroid hormone secretion in asthmatic patients on prolonged oral or inhaled steroid treatment. Eur. Respir. J. 4, 441 – 444 (1991). 192. H. A. Morris, A. G. Need, P. D. O’Loughlin, M. Horowitz, A. Bridges, and B. E. Nordin, Malabsorption of calcium in corticosteroid-induced osteoporosis. Calcif. Tissue Int. 46, 305 – 308 (1990). 193. R. A. Corradino, and C. S. Fullmer, Positive cotranscriptional regulation of intestinal calbindin-D28K gene expression by

CHAPTER 44 Glucocorticoid-Induced Osteoporosis

194.

195.

196. 197.

198.

199.

200.

201.

202.

203.

204. 205.

206.

207.

208. 209.

210.

211.

212.

213.

1,25-Dihydroxyvitamin D3 and glucocorticoids. Endocrinology 128, 944 – 950 (1991). M. Tohmon, M. Fukase, M. Kishihara, S. Kadowaki, and T. Fujita, Effect of glucocorticoid administration on intestinal, renal, and cerebellar calbindin-D28K in chicks. J. Bone Miner. Res. 3, 325 – 331 (1988). T. J. Hahn, L. R. Halstead, and D. T. Baran, Effects off short term glucocorticoid administration on intestinal calcium absorption and circulating vitamin D metabolite concentrations in man. J. Clin. Endocrinol. Metab. 52, 111 – 115 (1981). I. R. Reid, Glucocorticoid effects on bone [editorial]. J. Clin. Endocrinol. Metab. 83, 1860 – 1862 (1998). I. R. Reid and H. K. Ibbertson, Evidence for decreased tubular reabsorption of calcium in glucocorticoid-treated asthmatics. Horm. Res. 27, 200 – 204 (1987). F. Cosman, J. Nieves, J. Herbert, V. Shen, and R. Lindsay, High-dose glucocorticoids in multiple sclerosis patients exert direct effects on the kidney and skeleton. J. Bone Miner. Res. 9, 1097 – 1105 (1994). J. W. Findling, N. D. Adams, J. Lemann, R. W. Jr., Gray, C. J. Thomas, and J. B. Tyrrell, Vitamin D metabolites and parathyroid hormone in Cushing’s syndrome: Relationship to calcium and phosphorus homeostasis. J. Clin. Endocrinol. Metab. 54, 1039 – 1044 (1982). J. P. Luton, P. Thieblot, J. C. Valcke, J. A. Mahoudeau, and H. Bricaire, Reversible gonadotropin deficiency in male Cushing’s disease. J. Clin. Endocrinol. Metab. 45, 488 – 495 (1977). M. Sakakura, K. Takebe, and S. Nakagawa, Inhibition of luteinizing hormone secretion induced by synthetic LRH by long-term treatment with glucocorticoids in human subjects. J. Clin. Endocrinol. Metab. 40, 774 – 779 (1975). B. Lukert and B. E. Kream, Clinical and basic aspects of glucorticoid action in bone. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, G. A. Rodan, eds.) pp. 533 – 548. Acadmic Press, San Diego, 1996. C. Kasperk, A. Helmboldt, I. Borcsok, et al. Skeletal site-dependent expression of the androgen receptor in human osteoblastic cell populations. Calcif. Tissue Int. 61, 464 – 473 (1997). H. Tanaka and Y. Seino, Does growth hormone treatment prevent corticosteroid-induced osteoporosis? Bone 18, 493 – 494 (1996). P. Doerr and K. M. Pirke, Cortisol-induced suppression of plasma testosterone in normal adult males. J. Clin. Endocrinol. Metab. 43, 622 – 629 (1976). A. J. Hsueh and G. F. Erickson, Glucocorticoid inhibition of FSHinduced estrogen production in cultured rat granulosa cells. Steroids 32, 639 – 648 (1978). I. R. Reid, H. K. Ibbertson, J. T. France, and J. Pybus, Plasma testosterone concentrations in asthmatic men treated with glucocorticoids. Br. Med. J. 291, 574 (1985). I. R. Reid, A. G. Veale, J. T. France, Glucocorticoid osteoporosis. J. Asthma 31, 7 – 18 (1994). C. Montecucco, R. Caporali, P. Caprotti, M. Caprotti, and A. Notario, Sex hormones and bone metabolism in postmenopausal rheumatoid arthritis treated with two different glucocorticoids. J. Rheumatol. 19, 1895 – 1900 (1992). E. O. Grecu, A. Weinshelbaum, and R. Simmons, Effective therapy of glucocorticoid-induced osteoporosis with medroxyprogesterone acetate. Calcif. Tissue Int. 46, 294 – 299 (1990). B. P. Lukert, B. E. Johnson, and R. G. Robinson, Estrogen and progesterone replacement therapy reduces glucocorticoid-induced bone loss. J. Bone Miner. Res. 7, 1063 – 1069 (1992). I. Reid, C. Alexander, A. King, and H. Ibbertson, Prevention of steroid-induced osteoporosis with (3-amino-1-hydroxyprpylidene)-1, 1-bisphosphonate (APD). The Lancet 143 – 146 (1988). J. W. Bijlsma, J. A. Raymakers, C. Mosch, et al. Effect of oral calcium and vitamin D on glucocorticoid-induced osteopenia in patients with rheumatic diseases. Clin. Exp. Rheumatical 6, 113 – 119 (1988).

191 214. T. R. Dykman, K. M. Haralson, O. S. Gluck, et al. Effect of oral 1,25-dihydroxyvitamin D and calcium on glucocorticoid-induced osteopenia in patients with rheumatic disease. Arthritis Rheum. 27, 1336 – 1343 (1984). 215. J. D. Adachi, W. G. Bensen, F. Bianchi, et al. Vitamin D and calcium in the prevention of corticosteroid induced osteoporosis: A 3 year followup. J. Rheumatol. 23, 995 – 1000 (1996). 216. M. Talalaj, L. Gradowska, E. Marcinowska-Suchowierska, M. Durlik Z. Gaciong, and M. Lao, Efficiency of preventive treatment of glucocorticoidinduced osteoporosis with 25-hydroxyvitamin D3 and calcium in kidney transplant patients. Transplant. Proc. 28, 3485–3487 (1996). 217. L. M. Buckley, E. S. Leib, K. S. Cartularo, P. M. Vacek, and S. M. Cooper,. Calcium and vitamin D-3 supplemntation prevents bone loss in the spine secondary to low-dose corticosteroids in patients with rheumatoid arthritis — A randomized, double-blind, placebocontrolled trial. Ann. Intern. Med. 125, 961 (1996). 218. C. N. Bernstein, L. L. Seeger, P. A. Anton, et al. A randomized, placebo-controlled trial of calcium supplementation for decreased bone density in corticosteroid-using patients with inflammatory bowel disease: A pilot study. Alimentary Pharmacol. Ther. 10, 777 – 786 (1996). 219. J. Y. Reginster, D. Kuntz, W. Verdickt, et al. Prophylactic use of alfacalcidol in corticosteroid-induced osteoporosis. Osteoporosis Int. 9, 75 – 81 (1999). 220. S. Amin, M. P. LaValley, R. W. Simms, and D. T. Felson, The role of vitamin D in corticosteroid-induced osteoporosis: A meta-analytic approach. Arthritis Rheum. 42, 1740 – 1751 (1999). 221. J. D. Ringe, A. Coster, T. Meng, E. Schacht, and R. Umbach, Treatment of glucocorticoid-induced osteoporosis with alfacalcidol/calcium versus vitamin D/calcium. Calci. Tissue Int. 65, 337 – 340 (1999). 222. J. D. Adachi, W. G. Bensen, J. Brown, et al. Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N. Engl. J. Med. 337, 382 – 387 (1997). 223. I. R. Reid, Pharmacological management of osteoporosis in postmenopausal women: A comparative review. Drugs Aging 15, 349 – 363 (1999). 224. P. J. Meunier, P. D. Delmas, R. Eastell et al. Diagnosis and management of osteoporosis in postmenopausal women: Clinical guidelines. International Committee for Osteoporosis Clinical Guidelines. Clin. Ther. 21, 1025 – 1044 (1999). 225. K. G. Saag, R. Emkey, T. J. Schnitzer, et al. Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N. Engl. J. Med. 339, 292 – 299 (1998). 226. S. Cohen, R. M. Levy, M. Keller, et al. Risedronate therapy prevents corticosteroid-induced bone loss: A twelve-month, multicenter, randomized, double-blind, placebo-controlled, parallel-group study. Arthritis Rheum. 42, 2309 – 2318 (1999). 227. D. M. Reid, R. A. Hughes, R. Laan, et al. Efficacy and safety of daily risedronate in the treatment of corticosteroid-induced osteoporosis in men and women: A randomized trial. J. Bone Miner. Res. 15, 1006 – 1013 (2000). 228. L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Roberson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999). 229. A. Struijs, A. Smals, S. A. Witte de, W. H. L. Hackeng, and H. Mulder, Acute effects of etidronate on glucocorticoid-induced bone degradation. Rheumatology 39, 523 – 529 (2000). 230. Y. Boutsen, J. Jamart, W. Esselinckx, M. Stoffel, and J-P. Devogelaer, Primary prevention of glucocorticoid-induced osteoporosis with intermittent intravenous pamidronate: A randomized trial. Calcif. Tissue Int. 61, 266 – 271 (1997). 231. S. Gonnelli, P. Rottoli, C. Cepollaro, et al. Prevention of corticosteroid-induced osteoporosis with alendronate in sarcoid patients. Calci. Tissue Int. 61, 382 – 385 (1997).

192 232. J. Adachi, A. Cranney, C. H. Goldsmith, et al. Intermittent cyclic therapy with etidronate in the prevention of corticosteroid induced bone loss. J. Rheumatol. 21, 1922 – 1926 (1994). 233. R. J. Sebaldt, G. Ioannidis, J. D. Adachi, et al. 36 month intermittent cyclical etidronate treatment in patients with established corticosteroid induced osteoporosis. J. Rheumatol. 26, 1545 – 1549 (1999). 234. P. Pitt, F. Li, P. Todd, D. Webber, S. Pack, and C. A. Moniz, double blind placebo controlled study to determine the effects of intermittent cyclical etidronate on bone mineral density in patients on longterm oral corticosteroid treatment. Thorax 53, 351 – 356 (1998). 235. P. Sambrook, N. K. Henderson, A. Keogh, et al. Effect of calcitriol on bone loss after cardiac or lung transplantation. J. Bone Miner. Res. in press. 236. J. E. Homik, A. Cranney, B. Shea, et al. A metaanalysis on the use of bisphosphonates in corticosteroid induced osteoporosis. J. Rheumatol. 26, 1148 – 1157 (1999). 237. G. M. Hall, M. Daniels, D. V. Doyle, and T. D. Spector, Effect of hormone replacement therapy on bone mass in rheumatoid arthritis patients treated with and without steroids. Arthritis Rheumatism 37, 1499 – 1505 (1994). 238. I. R. Reid, D. J. Wattie, M. C. Evans, and J. P. Stapleton, Testosterone therapy in glucocorticoid-treated men. Arch. Intern. Med. 156, 1173 – 1179 (1996). 239. S. Adami, V. Fossaluzza, M. Rossini, et al. The prevention of corticosteroid-induced osteoporosis with nandrolone decanoate. Bone Miner. 15, 73 – 81 (1991). 240. L. Montemurro, G. Schiraldi, P. Fraioli, G. Tosi, A. Riboldi, and G. Rizzato, Prevention of corticosteroid-induced osteoporosis with salmon calcitonin in sarcoid patients. Calcif. Tissue Int. 49, 71 – 76 (1991). 241. J. D. Adachi, W. G. Bensen, M. J. Bell, et al. Salmon calcitonin nasal spray in the prevention of corticosteroid-induced osteoporosis. Br. J. Rheumatol. 36, 255 – 259 (1997). 242. J. H. Healey, S. A. Paget, P. Williams-Russo, et al. A randomized controlled trial of salmon calcitonin to prevent bone loss in corticosteroidtreated temporal arteritis and polymyalgia rheumatica. Calci. Tissue Int. 58, 73 – 80 (1996). 243. A. Kotaniemi, H. Piirainen, L. Paimela, et al. Is continuous intranasal salmon calcitonin effective in treating axial bone loss in patients with active rheumatoid arthritis receiving low dose glucocorticoid therapy? J. Rheumatol. 23, 1875 – 1879 (1996). 244. M. Luengo, C. Picado, L. Del Rio, N. Guanabens, J. M. Montserrat, and J. Setoain, Treatment of steroid-induced osteopenia with calcitonin in corticosteroid-dependent asthma. A one-year follow-up study. Am. Rev. Respiratory Dis. 142, 104 – 107 (1990). 245. J. D. Ringe and D. Welzel, Salmon calcitonin in the therapy of corticoidinduced osteoporosis. Eur. J. Clin. Pharmacol. 33, 35 – 39 (1987). 246. M. Luengo, F. Pons, M. J. Martinez, de Osaba, and C. Picado, Prevention of further bone mass loss by nasal calcitonin in patients on long term glucocorticoid therapy for asthma: A two year follow up study. Thorax 49, 1099 – 1102 (1994). 247. W. F. Lems, W. G. Jacobs, J. W. Bijlsma, et al. Effect of sodium fluoride on the prevention of corticosteroid-induced osteoporosis. Osteoporosis Int. 7, 575 – 582 (1997). 248. W. F. Lems, J. W. Jacobs, J. W. Bijlsma, et al. Is addition of sodium fluoride to cyclical etidronate beneficial in the treatment of corticosteroid induced osteoporosis? Ann. Rheum. Dis. 56, 357 – 363 (1997). 249. G. Guaydier-Souquieres, P. O. Kotzki, J. P. Sabatier, B. BasseCathalinat, and G. Loeb, In corticosteroid-treated respiratory diseases, monofluorophosphate increses lumbar bone density: A doublemasked randomized study. Osteoporosis Int. 6, 171 – 177 (1996). 250. R. Rizzoli, T. Chevalley, D. O. Slosman, and J. P. Bonjour, Sodium monofluorophosphate increases vertebral bone mineral density in patients with corticosteroid-induced osteoporosis. Osteoporosis Int. 5, 39 – 46 (1995).

GARY M. LEONG ET AL 251. M, Kleerekoper and D. B. Mendlovic, Sodium fluoride therapy of postmenopausal osteoporosis. Endocr. Rev. 14, 312 – 323 (1993). 252. N. E. Lane, S. Sanchez, G. W. Modin, H. K. Genant, E. Ini, and C. D. Arnaud, Parathyroid hormone treatment can reverse corticosteroid-induced osteoporosis. Results of a randomized controlled clinical trial. J. Clin. Invest. 102, 1627 – 1633 (1998). 253. N. E. Lane, S. Sanchez, G. W. Modin, H. K. Genant, E. Pierini, C. D. Arnaud, Bone mass continues to increase at the hip after parathyroid hormone treatment is discontinued in glucocorticoid-induced osteoporosis: Results of randomized controlled clinical trial. J. Bone Miner. Res. 15, 944 – 951 (2000). 254. R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). 255. E. Canalis, Novel treatments of osteoporosis. J. Clin. Invest. 106, 177 – 179 (2000). 256. L. K. Bachrach, Making an impact on pediatric bone health. J. Pediatr. 136, 137 – 139 (2000). [Editorial Comment]. 257. C. W. Slemenda, T. K. Reister, S. L. Hui, J. Z. Miller, J. C. Christian, and C. C. Johnston, Jr. Influences on skeletal mineralization in children and adolescents: Evidence for varying effects of sexual maturation and physical activity. J. Pediatr. 125, 201 – 207 (1994). 258. F. H. Glorieux, N. J. Bishop, H. Plotkin, G. Chabot, G. Lanoue, and R. Travers, Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N. Engl. J. Med. 339, 947 – 952 (1998). 259. T. Nishioka, H. Kurayama, T. Yasuda, J. Udagawa, C. Matsumura, and H. Niimi, Nasal administration of salmon calcitonin for prevention of glucocorticoid-induced osteoporosis in children with nephrosis. J. Pediatr. 118, 703 – 707 (1991). 260. R. Rao, R. K. Gregson, A. C. Jones, E. A. Miles, M. J. Campbell, and J. O. Warner, Systemic effects of inhaled corticosteroids on growth and bone turnover in childhood asthma: A comparison of fluticasone with beclomethasone. Eur. Respir. J. 13, 87 – 94 (1999). 261. S. Balsan, D. Steru, A. Bourdeau, R. Grimberg, and G. Lenoir, Effects of long-term maintenance therapy with a new glucocorticoid, deflazacort, on mineral metabolism and statural growth. Calcif. Tissue Int. 40, 303 – 309 (1987). 262. J. Loftus, R. Allen, R. Hesp, et al. Randomized, double-blind trial of deflazacort versus prednisone in juvenile chronic (or rheumatoid) arthritis: A relatively bone-sparing effect of deflazacort. Pediatrics 88, 428 – 436 (1991). 263. A. Markham and H. M. Bryson, Deflazacort. A review of its pharmacological properties and therapeutic efficacy. Drug 50, 317 – 333 (1995). 264. J. David, J. Loftus, R. Hesp, B. M. Anselt, J. Reeve, and P. M. Woo, Spinal and somatic growth in patients with juvenile chronic arthritis treated for up to 2 years with deflazacort. Clin. Exp. Rheumatol. 10, 621 – 624 (1992). 265. S. S. Baker, C. A. Flores, M. K. Georgieff, M. S. Jacobson, T. Jaksic, and N. F. Krebs, Calcium requirements of infants, children, and adolescents. Pediatrics 104, 1152 – 1157 (1999). 266. American Academy of Pediatrics Committee on Nutrition. Calcium requirements of Infants, children and adolescents. Pediatrics 104, 1152 – 1157 (1999). 267. C. C. Johnston, Jr., J. Z. Miller, C. W. Slemenda, et al. Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87 (1992). 268. M. Bradney, G. Pearce, G. Naughton, et al. Moderate exercise during growth in prepubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J. Bone Miner. Res. 13, 1814 – 1821 (1998). 269. C. W. Slemenda, J. Z. Miller, S. L. Hui, T. K. Reister, and C. C. Johnston, Jr. Role of physical activity in the development of skeletal mass in children. J. Bone Miner. Res. 6, 1227 – 1233 (1991).

CHAPTER 44 Glucocorticoid-Induced Osteoporosis 270. D. C. Welten, H. C. Kemper, G. B. Post, et al. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 9, 1089 – 1096 (1994). 271. E. J. Semeao, A. F. Jawad, B. S. Zemel, K. M. Neiswender, D. A. Piccoli, and V. A. Stallings, Bone mineral density in children and young adults with Crohn’s disease. Inflammatory Bowel Dis. 5, 161–166 (1999). 272. H. D. Allen, I. G. Thong, P. Clifton-Bligh, S. Holmes, L. Nery, and K. B. Wilson, Effects of high-dose inhaled corticosteroids on bone metabolism in prepubertal children with asthma. Pediatr. Pulmonol. 29, 188 – 193 (2000). 273. L. K. Bachrach, T. Hastie, M. C. Wang, B. Narasimhan, and R. Marcus, Bone mineral acquisition in healty Asian, Hispanic, Black, and Caucasian youth: A longitudinal study. J. Clin. Endocrinol. Metab. 84, 4702 – 4712 (1999).

193 274. C. Molgaard, B. L. Thomsen, and K. F. Michaelsen, Influence of weight, age and puberty on bone size and bone mineral content in healthy children and adolescents. Acta Paediatr. 87, 494 – 499 (1998). 275. G. Jones and T. Dwyer, Bone mass in prepubertal children: Gender differences and the role of physical activity and sunlight exposure. J. Clin. Endocrinol. Metab. 83, 4274 – 4279 (1998). 276. S. Bass, P. D. Delmas, G. Pearce, E. Hendrich, A. Tabensky, and E. Seeman, The differing tempo of growth in bone size, mass, and density in girls is region-specific. J. Clin. Invest. 104, 795 – 804 (1999). 277. C. M. Gundberg, M. E. Markowitz, M. Mizruchi, and J. F. Rosen, Osteocalcin in human serum: A circadian rhythm. J. Clin. Endocrinol. Metab. 60, 736 – 739 (1985).

CHAPTER 45

Familial Osteoporosis ALAN L. BURSHELL* AND STEVEN R. SMITH*,† *Department of Internal Medicine, Alton Ochsner Medical Foundation, Ochsner Clinic, New Orleans, Louisiana 70131; and † Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana 70808

I. Introduction II. Importance of and Obstacles to the Study of Familial Osteoporosis III. Genetic Analysis of Complex Traits IV. Animal Models V. Selection of Families

VI. Epidemiologic Data VII. Polymorphisms and Bone Mineral Density VIII. Collagen as a Candidate Gene for Familial Osteoporosis IX. Conclusions References

II. IMPORTANCE OF AND OBSTACLES TO THE STUDY OF FAMILIAL OSTEOPOROSIS

I. INTRODUCTION Osteoporosis is a generalized condition of bone whose hallmark is increased bone fragility. As described in other chapters of this volume, multiple factors, both environmental and genetic, contribute to the development of osteoporosis. With recent developments in molecular genetics and the eventual sequencing of the human genome, understanding of the hereditary basis of osteoporosis will accelerate. Environmental factors, well known to be important in the development of osteoporosis, are addressed elsewhere in this volume and are not discussed in this chapter. However, as our knowledge of the genes that control bone acquisition and turnover expand, it is anticipated that gene–environment as well as gene–gene interactions will become apparent. It is the purpose of this chapter to outline approaches to study the familial nature of osteoporosis, to review the clinical evidence to date on the hereditary nature of osteoporosis, and to propose a candidate gene that may predispose an individual to its development. It is hoped that this chapter will encourage clinicians to identify families with hereditary osteoporosis and to develop a repository of such families for further genetic testing.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

The identification of families with hereditary osteoporosis will facilitate the diagnosis, prevention, therapy, and understanding of osteoporosis. The positive feedback loop in Fig. 1 demonstrates this concept. The key aspect to enter the loop is the identification of individuals with familial osteoporosis, “familial” being defined as affecting several members of the same family. The implication is that a familial disorder involves a genetic abnormality as opposed to clustering of environmental factors within a family. It seems somewhat contradictory that bone mass is under hereditary control and yet the identification of families with hereditary osteoporosis is uncommon. This is in part due to difficulties in the early recognition of an osteoporosis phenotype. It is also likely that nongenetic factors may be of sufficient magnitude to overwhelm subtle genetic influences. Fractures are the most important clinical end point, but a difficult one for research studies since they occur late in life, depend on additional factors, such as falls, and require large populations followed over many years. An alternative

195

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

196

FIGURE 1 Identification of families with osteoporosis may lead to new preventive and therapeutic strategies.

approach is to measure the bone mineral density (BMD), and many studies demonstrate that BMD is an excellent predictor of fracture risk [1–3]. Therefore, this measurement could be used as a screening tool to identify family members at increased risk for fractures. The fundamental assumption of this approach is that peak bone mass is under hereditary control and is an important component of the final risk of fractures. However, there are potential limitations to the use of a single measurement of BMD to screen families. If the rate of bone loss is the major determinant of bone deficits in a family, single BMD measurements in young family members would be an ineffective screening strategy. If a given familial disorder preferentially affected a particular skeletal envelope, it would be important that the BMD measurement be taken at an appropriate skeletal site, e.g., forearm, lumbar spine, hip, or total body. Various sites begin to lose bone at different ages and rates of change also vary [4]. It is also necessary to establish a diagnostic criterion for affected members, i.e., a level of BMD that is below the mean for age-matched controls by some defined amount. Another problem with this approach is the cost of measurements. Obviously, an inexpensive screening test, such as blood pressure or serum cholesterol measurement, would be preferred. A careful history and physical examination of index cases may identify other aspects of the phenotype that permit diagnosis. The assumption is that there may be skeletal, anthropomorphic, or extraskeletal differences that are associated with the genes responsible for osteoporosis. For example, in multiple endocrine neoplasia type II B (MEN II B), there are mucosal neuromas and Marfanoid habitus in addition to the classic manifestations of medullary cancer and pheochromocytoma. In some families with osteogenesis imperfecta (OI), in addition to frequent fractures, presence of blue sclerae is frequently a differentiating factor. In other families with OI, the sclerae are white, but increased joint flexibility is a predominant manifestation, indicating multiple different genetic abnormalities in

BURSHELL AND SMITH

various families. We have been able to identify families with familial osteoporosis on the basis of the following criteria: increased fractures, short stature, increased joint flexibility, thin skin, scoliosis, and/or early kyphosis. Because of the overlap of these signs with mild osteogenesis imperfecta, the collagen gene is an excellent candidate for osteoporosis in these families. As described later, there are additional reasons to test this hypothesis, but it is likely that other matrix proteins may also be candidate genes. Type I collagen may be adversely affected by protein interactions or gene regulation of transcription or translation. The potential complexity is great, but by careful identification and selection of these families and coordination with molecular geneticists, this candidate gene and other genes can be simultaneously tested. The key, however, is the accurate identification of families with osteoporosis. The essence of the familial approach is exemplified by the fact that whether or not a specific collagen gene hypothesis is or is not correct for a given family, the family members may still be highly informative. Assuming that genetic abnormalities of collagen or its regulation are not discovered in our families with hereditary osteoporosis, these families can still be tested for other candidate genes such as the vitamin D receptor. If genetic abnormalities of collagen turn out to be only a rare cause of hereditary osteoporosis, they still may be of great importance. For example, mutations in the lowdensity lipoprotein (LDL) cholesterol receptor are known to cause severe hypercholesterolemia but only rarely cause the more common multifactorial hypercholesterolemia. However, identification of the LDL receptor and the discovery of its importance in regulating circulating LDL cholesterol markedly improved our understanding of the etiology of coronary artery disease and led to several drug therapies that are widely beneficial to individuals with hypercholesterolemia. Similarly, discovering genes important in some families with osteoporosis may lead to new therapies. Finally, the identification of a specific gene abnormality in a family should permit earlier diagnosis of the disease process. Ideally, this would lead to earlier intervention with preventive lifestyle changes, medications, or dietary therapies. There is also the potential for generating animal models based on the human genetic defect and the development of specific therapies via this route. Therefore, the identification of families with osteoporosis may facilitate the development of new prevention and therapeutic strategies.

III. GENETIC ANALYSIS OF COMPLEX TRAITS It is likely that osteoporosis, like most disease processes, does not follow simple Mendelian genetics but is a complex trait. There are many excellent reviews of genetic

197

CHAPTER 45 Familial Osteoporosis

approaches [5] and the purpose of this chapter is simply to raise the level of awareness of the power of genetics and the potential for collaboration with geneticists. Gene mapping permits identification of a gene location without knowledge of its structure. Complex traits are difficult to study because of incomplete penetrance, genetic heterogeneity, and polygenic inheritance. There may be multiple genes important for continuous variables like BMD, and fracture may be a discrete variable representing a threshold effect. There are four major genetic approaches to identifying the critical genes in complex traits: Linkage analysis Allele sharing Association studies Polygenic analysis of experimental crosses (animal models only). The techniques which are frequently employed in familial studies are linkage analysis and allele sharing methods. Again, there are basic reviews [6] concerning both techniques and they will not be detailed here. In general, linkage analysis involves testing a transmission model to explain the inheritance in a pedigree such that a specific chromosomal location is more likely to explain the data than an unlinked locus [6]. This is frequently expressed by the lod score, Z  log 10 likelihood ratio. The larger the family, the more informative the linkage analysis. This works particularly well for simple Mendelian traits. On the other hand, use of the allele sharing method is an attempt to demonstrate that in a region of the chromosome, inheritance is not consistent with random Mendelian segregation. The association technique simply examines affected and unaffected individuals in a population, e.g., individuals with ankylosing spondylitis are more likely to be HLA-B27 positive than other individuals in the population. Newer techniques, such as serial analysis of gene expression (SAGE), cDNA arrays, and proteomics, use the power of the Human Genome projects and EST/protein data bases to test the simple hypothesis that one or more genes will be differentially expressed between normal and diseased individuals. SAGE is a molecular technique that allows for the simultaneous semiquantitative detection of all mRNAs expressed within a single tissue or cell type [7]. Comparison of normal and affected individuals has the potential to detect differentially expressed genes. Although labor and cost intensive, the technique has been adapted to small tissue samples [8] and might be useful for comparing gene expression in bone tissues from individuals with familial osteoporosis to that in their normal family members. Several other techniques have similar potential to rapidly compare genes differentially expressed in a given tissue such as bone [9]. Proteomics, the identification of proteins that are differentially expressed between normal and diseased tissues, is another powerful tool [10]. This

technique uses two-dimensional gel electrophoresis to compare differentially expressed proteins between two tissues. The proteins are then isolated from the gel and the amino acid sequence determined by mass spectrometry. Although an expensive technique, the expanding database of identified proteins and standardization of protein separation techniques makes this approach feasible. Last, the cDNA microarray is able to compare thousands of mRNAs between tissues. Current technologies can examine approximately 50,000 expressed genes in a single experiment. If the Human Genome project is completed as predicted, the number of genes in a single cDNA microarray is likely to be equal to all genes in the genome. The cDNA array technique was recently used to identify a single gene mutation in an animal model [11] and has the potential to identify dysregulated or mutated genes in human bone tissues from individuals with familial osteoporosis. All of these new technologies are likely to be extensively used in animal models.

IV. ANIMAL MODELS Animal models can be useful tools to investigate the heritable nature of osteoporosis (see Chapter 37). Bone metabolism in rodents, for example, provides an excellent model system by which to understand the pathophysiology of human menopause and test new therapeutic approaches such as gene therapy delivery systems [12,13]. In addition, there are well-known sequence similarities and functional homologies in genes from multiple physiologic systems and species. Several recent examples illustrate the utility of animal models in the investigation of familial (heritable) polygenic diseases. Positional cloning of a mouse obesity (OB) gene by Zhang and colleagues in 1994 [14] exemplifies the use of an inbred strain of mouse to identify a potentially important gene in human disease. The OB mouse develops a phenotype of obesity and diabetes similar to that in humans. This approach utilized physical mapping to locate a region of mouse chromosome 6 that contained the gene. Subsequently, yeast artificial chromosomes (YACs) and P1 bacteriophage clones were screened to isolate and sequence the mouse OB gene. Based on the mouse sequence, the human gene was identified as well. This approach is time-and resource-intensive, but has the ability to identify genes that may be important in human disease. Many other approaches are available for identifying genes related to a given phenotype but are beyond the scope of this dicussion. Again, the key to each is the identification and quantification of the particular phenotype. Animal models of osteoporosis exist (OIM, SAM-P6, and fro/fro, for example [15 –17] and may be amenable to similar approaches.

198 Studies of heritability of bone phenotypes in animals may provide insight into the complexity of genetic control of bone mass and fracture risk. Several groups are using quantitative trait loci (QTL) to map high and low BMD by dual-energy X-ray absorptimetry (DXA) in recombinant inbred strains [17,18]. Beamer et al. [17] used pQCT to determine the heritability of bone mass and density in multiple strains of mice. Importantly, they concluded that cortical bone density at each of three different sites was regulated by a common set of genes. In contrast, femur length and density were controlled by a different set of genes, illustrating the complexity of the genetic control of bone mass and density. This study serves to illustrate that the genes that control one aspect of the osteoporotic phenotype (bone density) may be different from the genes that control another phenotype (bone size). The mouse and human genome projects are likely to be completed in the next few years and the mouse is likely to become the preferred rodent for genetic studies to identify new genes that are important for the development of osteoporosis. Finally, knock out of various genes in mice permits the identification of genes which may be important in osteoporosis and establish phenotypes that may occur in humans, i.e., estrogen receptor defects.

V. SELECTION OF FAMILIES The selection of families becomes the first and most crucial step in research of familial osteoporosis. As noted above, osteoporosis, similar to many other disease processes, is likely to reflect complex or multifactorial inheritance. A possible strategy for selection of families might include: (i) identification of index cases, (ii) examination of family members, and (iii) measurement of phenotypic end points. Although it is possible to separate these different entities, in practice they interact. Preferred index cases are those individuals with the most severe disease. Since men are less frequently affected by osteoporosis, a man with severe osteoporosis would actually be an ideal candidate for an index case. Using this approach, 38 men with fractures and osteoporosis were identified and 73 relatives studied [19]. The mean Z score in relatives was decreased by 1.28 SD at the lumbar spine and 1.03 SD at the femoral neck, suggesting that identifying men with osteoporosis is a reasonable strategy for identifying families with osteoporosis. In addition, individuals who develop symptoms and signs of osteoporosis at younger ages will also be good candidates. Whenever, possible, individuals who do not have obvious environmental risks are also more suited for genetic evaluation as they are more likely to have a genetic predisposition to disease. A positive family history is helpful since this again suggests a genetic abnormality. Finally, the larger the family and the more generations that are available, the better for genetic study.

BURSHELL AND SMITH

Examination of family members may indicate which members are and are not affected. Where it is unknown whether an individual is affected, he or she should not be assumed to be unaffected. One key factor will be the family history of fractures. On physical examination, anthropomorphic and extraskeletal phenotype should be carefully documented in the index case and all family members. From the perspective of laboratory evaluation, at the present time the best marker for osteoporosis is BMD. The relationship between BMD and heredity is complex. First, DXA BMD measures both bone size and density and thus genes identified by this technique may affect either the size, density, or both. Some genes may affect all sites similarly, but it is probable that other genes may affect the hip more than the lumbar spine, etc. Thus, multiple BMD sites should be studied. It is possible that, in specific families, other measurements may further differentiate the phenotype, such as biochemical markers of bone turnover, hip axis length, flexibility, and skin thickness. The genes responsible for endochondral and intramembranous bone formation as well as those for modeling and remodeling may be different. It is recommended to measure BMD in younger members of the family who are still growing as well as adults, look at footprints at birth, measure the width and length of the long bones, and measure the span, height, upper segment, and lower segment. It is important to appreciate that periosteal expansion is an intramembranous process, whereas longitudinal growth of long bones is endochondral. Eunuchoid proportions may reflect greater endochondral compared to intramembranous bone growth. The skeletal phenotype is likely to be composed of a vast array of genes which may differentially affect development, growth, remodeling, and specific bone sites. These genes may have important extraskeletal effects. In summary, an approach to identifying families could be simplified by studying individuals with more severe disease, performing history and physical examinations on index cases and family members, and, finally, measuring the BMD. Blood should be obtained for DNA and saved for future collaboration with geneticists using linkage analysis, allele sharing, or association techniques, as described above.

VI. EPIDEMIOLOGIC DATA A. Fracture Data Most of the epidemiologic information related to bone have been obtained from healthy populations using BMD as the primary end point. Fractures are the preferred end point but are seldom used because the events are relatively rare and require large populations at great expense. One important exception is the currently ongoing Study of Osteoporotic Fractures (SOF). Cummings et al. [20]

199

CHAPTER 45 Familial Osteoporosis

recently published their experience with 9516 white women above 65 years of age followed for hip fracture for 4.1 years. A daughter had a relative risk of 2.0 for fracturing her hip if her mother had suffered a hip fracture and this risk was increased if the mother fractured before age 80. The risk of wrist fracture was increased threefold if there was a paternal history of wrist fracture. This important study indicates that fractures are hereditary and site specific and that paternal history is of significance. A major problem in the study is that because of the age of the subjects, their parents are not alive and thus the fracture history cannot easily be authenticated. Further analysis of the SOF data including BMD suggested that the family history predicts fractures in a site-specific manner not dependent on BMD effects [21]. In another study, a family history of hip or wrist fractures was associated with an odds ratio of 2.02 for all fractures compared to a negative family history of fractures. If there was a family history of a wrist fracture then the odds ratio of a wrist fracture rose to 4.24 and these relationships remained after correcting for the BMD [22]. Finally, prevalent vertebral fractures identified by spine X-ray predict vertebral and hip fractures [23]. Thus, there is evidence that a history of fractures is associated with both an increased risk of site specific fractures and fractures in other locations.

B. BMD in Daughters of Osteoporotic Mothers As previously noted, the fact that fractures tend to occur at older ages makes the study of heredity much more difficult. Mother–daughter studies have been utilized to evaluate the vertical transmission of genes. In general, the mother is identified as having osteoporosis and BMD measurements of both mother and daughter are compared to those of a normal population. Seeman et al. utilized this approach in two studies. In the first study [24] a history of vertebral fractures in a mother was associated with a low BMC and BMD in the lumbar spine and femoral neck of the daughter. The authors estimated that approximately 80% of the variance was on the basis of heredity and concluded that low peak bone mass was at least one component in the risk for osteoporosis. In the second study [25], 31 women with hip fractures were identified and their 41 daughters were compared to a reference population. This study demonstrated that a low peak BMD of the hip was at least partially responsible for the risk of hip fractures. There appears to be specificity of bone deficits since the lumbar spine BMD was normal in the offspring of the women with hip fractures. Corroborating data from France was obtained by identifying osteoporotic women with a low BMD of the lumbar spine or femoral neck and comparing to matched women with normal BMD. The daughters of osteoporotic mothers had significantly lower Z scores BMD at the lumbar spine than control daughters [26].

C. Low Bone Mineral Density Associated with Fracture History in Relatives Armamento-Villareal et al. [27] and Soroko et al. [28] studied family history of hip fracture as a risk factor for a low BMD. Armamento-Villareal [27] found the major predictors of BMD in 63 premenopausal Caucasian women to be an estimate of lifetime estrogen exposure and heredity. Subjects with a lumbar spine BMD 1 SD below age-predicted means had a positive maternal fracture history significantly more often than those with normal BMD. In a study by Grainge et al., young English postmenopausal women with a family history of fractures had a lower femoral neck BMD than those women with a negative family history of fractures [29]. There was a relationship between fracture history in sisters and mothers and a low femoral neck BMD. Soroko et al. [28] studied 600 noninstitutionalized men and 877 women. They determined the family history of osteoporosis in parents and sisters and measured BMD of the lumbar spine and hip. In men, a positive family history was associated with a low hip BMD, whereas in women it was associated with a low spine BMD. The more family members with a positive family history, the lower the BMD of the individual. Paternal history contributed significantly to BMD of both the male and the female offspring, and the predictive value of the family history was estimated to be about 23%. These studies together indicate that family history predicts BMD in both premenopausal and postmenopausal women and elderly men.

D. Twin Studies in Normal Populations Twin studies have been utilized to separate contributions of genetic from environmental variables. The basic concept is that a monozygotic twin pair will have less variability than a dizygotic pair if heredity is an important factor. The greater the difference in variability of mono- and dizygotic twins, the greater the role of heredity. Multiple twin studies have used BMD as the end point in nonosteoporotic populations, but we could locate no twin studies in osteoporotic populations. Investigators at Indiana University have followed a group of male twins since 1973 and studied juvenile twins and adult women twins [30–32]. Their results indicated that radial bone mineral density was under hereditary control, that these effects were greatest at the time of maximal bone mass, and that the hereditary effects declined with advancing years. The Indiana group [33] studied 124 monozygotic and 47 dizygotic pairs of women ages 25 to 80 and measured the BMD of the lumbar spine, hip, and mid-radius. All heredity indices were in the 80 to 105% range, which appeared to be unrealistically high and suggested that twin studies may overemphasize the role of

200

BURSHELL AND SMITH

genetics. It was their interpretation that the high correlation for the monozygotic twins may reflect environmental factors which are more similar for monozygotic than for dizygotic twins. Multiple studies from Australia have also used twin methodology. Pocock et al. [34] found that the hereditary estimates were highest in the lumbar spine (92%), Ward’s triangle (85%), and femoral neck (73%). These authors suggested that a single gene or set of genes might be responsible for bone mass at multiple sites. However, only the serum osteocalcin level showed evidence for genetic control, with a heredity estimate of approximately 80%. Kelly et al. [35] studied biochemical markers in twins and concluded that osteocalcin was under genetic control and that since osteocalcin concentrations reflect the overall rate of bone turnover, bone formation was genetically regulated as well. Kelly et al. [36] also examined hereditary contributions to bone loss in 21 monozygotic and 19 dizygotic twin pairs, ages 24 to 75 years. BMD was measured at the lumbar spine and hip on two to four occasions over a period of 1 to 5.5 years. At both sites, the correlation of the monozygotes was significantly greater than that of the dizygotes, indicating that variations in the rate of bone loss were genetically determined. However, the rate of bone loss at the femoral neck showed no genetic influence. This study suggests that bone loss at some sites remains under some genetic control, in contrast to the previous study from Indiana, which found bone loss predominantly to be under environmental influence. Further studies are needed to resolve the influence of heredity on bone loss. On the basis of these reports, the following generalizations seem compatible with the findings. Heredity has a major influence upon BMD at multiple sites including the lumbar spine, hip, and radius. This effect diminishes with age, possibly secondary to a greater role for environmental factors.

change or that the environment plays a greater role with age. Futhermore, in that study bone loss began earlier at the hip than at the spine. Tylavsky et al. [40] measured BMC and BMD of the mid- and distal radius in 84 premenopausal mothers (average age, 44.2) and their daughters (average age, 18.6). Familial resemblance was significant at both sites but greater for the mid-radius, where the mother–daughter BMC correlation was r  0.47. Matkovic et al. [41] measured radiogrammetry of the second metacarpal bone to determine the resemblance of mothers, fathers, and daughters. In addition, they measured BMD of the distal and 1/5 region of the forearm and of the lumbar spine. The 14-year-old adolescents and their mothers (only one postmenopausal) and fathers showed average parent–daughter correlations of 0.60, 0.72 and 0.52 at the spine, distal radius, and proximal radius, respectively. In addition, the authors noted that bone size and mass were all under greater genetic control than was BMD. Ferrari et al. studied 138 premenopausal mothers (mean age 40) and their daughters (age 8.1) at 2-year intervals. At age 8, the daughters reached 59–78% of their mother’ areal BMD, but 75–105% of their mother’s bone mineral apparent density. During the next 2 years the size of the bones increased, but volumetric BMD did not change. Thus, the familial resemblance between mother’s and daughter’s BMD is present prior to puberty [42]. The mother–daughter studies again emphasize the importance of genetics. The genetic influence is easier to demonstrate in premenopausal mothers, suggesting that menopause and age are asociated with environmental effects that tend to dilute the genetic effects. The hereditary component influencing maximal BMD may be programmed before puberty and puberty may be a time of increase bone size rather than true BMD a period associated more with bone growth and increases in bone mineral content rather than large increases in volumetric BMD.

E. Mother–Daughter Studies in the Normal Population

F. Nuclear Families

Mother–daughter studies have also been used to estimate the normal hereditary component of bone mass. Lutz et al. [38] evaluated radial BMC and BMD in 26 mother–daughter pairs, deriving heredity estimates of 72% for BMC and 57% for BMD. The authors viewed the hereditary estimate as a resemblance measurement because there could also be a shared environment. Lutz et al. [39] performed another mother–daughter study in 37 pairs, but measured BMD in the spine and hip. Thirty of the mothers were pr menopausal. At all sites there were significant mother– daughter BMD correlations which were greater in the premenopausal group, suggesting that the genetic influence on bone loss may

Krall and Dawson-Hughes [43] studied 40 nuclear families composed of a premenopausal daughter, postmenopausal mother (average age 60), father (average age 63), and brother (average age 32) and measured BMD at five different sites. All measurements were adjusted for age, weight, and height and were expressed as Z scores. The mid-parent (i.e., averaged values of mother and father) Z score correlated significantly with the son’s BMD at the os calcis, femoral neck, and with total body calcium. The midparent Z score correlated with the daughter’s bone mineral density at the radius, os calcis, and total body calcium. There was no significant familial correlation for BMD at the lumbar spine. The mid-parent correlations were, in

201

CHAPTER 45 Familial Osteoporosis

general, higher than the mother–daughter correlations. The authors estimated that 46 to 62% of the variance was secondary to genetics. The reason for the poor correlation of the lumbar spine is unknown. The higher-correlations of the mid-parent BMD with their offspring supports the genetic importance of the father to the BMD.

G. Summary of Epidemiologic Data The twin, mother–daughter, nuclear family, and fracture data taken together permit the following conclusions: 1. Peak bone mass and BMD in the normal population are under hereditary control. 2. Hereditary factors for BMD derive from both parents. 3. Bone loss begins at different ages at different sites. Although it is more difficult to study (since it requires at least two points over time), bone loss may be affected by genetic factors which could be the same gene(s) or different gene(s) than those influencing peak bone mass. 4. Environmental effects exert a greater influence on bone mass in the elderly, whereas the effect of genetics is greatest in the young. 5. Heredity is an important factor in the development of osteoporosis. Heredity influences BMD as measured by DXA and BMD is well known to predict fractures. However, there are likely to be genes that are associated with fractures that are independent of bone size and density and therefore the fracture history of parents and other family members is important information.

VII. POLYMORPHISMS AND BONE MINERAL DENSITY There are multiple studies of multiple genetic polymorphisms in multiple populations claiming or disclaiming correlations with fractures and/or BMD. This is a complex area of study and is discussed in detail in other sections of this book (see Chapter 26).

following section will review similarities of phenotype between osteoporosis and osteogenesis imperfecta (a known disorder of type I collagen and discussed in Chapter 50) and propose a schema for the study of the role of type I collagen in familial osteoporosis (see Fig. 2). Emphasis is placed upon the phenotype of mild Type I OI since it may be clinically relevant to understanding osteoporosis. We hope that by sharing some of our clinical findings that it will encourage other physicans to more carefully study and examine their patients with familial osteoporosis. The intent is to review the candidate gene approach in a practical manner. Collagen is simply one of many potential candidate genes. For example, another extracellular matrix protein, biglycan is associated with osteoporosis as shown in the biglycan-deficient mouse [45].

A. Osteogenesis Imperfecta as a Model for Familial Osteoporosis OI is an autosomal dominant disorder of type I collagen which result in a phenotype of (i) increased skeletal fragility, (ii) thin skin, (iii) failure to attain appropriate peak bone mass, (iv) increased joint mobility, (v) blue sclerae, and (vi) dentinogenesis imperfecta. The severity of OI varies from fractures in utero and death to a milder form of the disease with fractures occuring only infrequently and sometimes difficult to differentiate from osteoporosis. Type I collagen is present throughout the body as a structural protein, but is most abundant in skin, ligaments, tendons, and bone, where it is the primary structural protein of the skeleton. Thin skin and joint hypermobility are important, but often unrecognized components of the OI syndrome. Similarly, osteoporotic patients are seldom evaluated for their joint flexibility and skin thickness. In our clinical experience, a subgroup of families with osteoporosis also have joint laxity. Joint

VIII. COLLAGEN AS A CANDIDATE GENE FOR FAMILIAL OSTEOPOROSIS Fuller Albright, in his 1941 article “Postmenopausal Osteoporosis” noted, “Their skin is noticeably thin, this condition suggesting that the atrophy is more widespread than just in the bone matrix” [44]. Based on our clinical observations and (the similarity) similarities in phenotype between osteoporosis and osteogenesis imperfecta, we propose that the type I collagen genes or other matrix proteins are excellent candidate genes for familial osteoporosis. The

FIGURE 2

The overlap of osteogenesis imperfecta and osteoporosis.

202 laxity decreases with age and thus joint laxity may be more obvious in younger family members.

B. Skin and Bone (Both Major Sites of Type I Collagen Synthesis) Are Coordinately Regulated in Many Diseases Type I collagen appears to be coordinately regulated in the skin and bone in many systemic illnesses. For example, glucocorticoid excess [46], rheumatoid arthritis [47], and anorexia nervosa [47] are associated with thin skin and osteoporosis. Acromegaly (growth hormone excess) produces the reverse situation, where cortical BMD is increased and the skin is thickened [48,49]. In acromegaly, trabecular bone loss commonly occurs but may be related to gonadal deficiency that frequently accompanies the pituitary tumor rather than an effect of growth hormone. Effective therapy of acromegaly [48] decreases BMD as well as skin thickness [50]. A majority of investigators who have studied this relationship have demonstrated associations between skin thickness and bone mineral [49]. Thinning of the skin occurs in postmenopausal women and can be prevented or reversed by estrogen replacement therapy [51]. This is strikingly similar to the prevention of postmenopausal osteoporosis by estrogen. Thus, the close association of skin and bone in disease and aging emphasizes the importance of collagen or other matrix proteins. On the basis of the physiology, collagen would be a reasonable candidate gene for further examination.

C. The Role of Collagen in Osteoblast Function Type I collagen is also important in the regulation of maturation and differentiation of cultured osteoblasts. Lynch et al. [52] demonstrated that osteoblastic cells grown on a type I collagen matrix increased expression of the late osteoblastic markers alkaline phosphatase and osteocalcin. This is presumed to occur through the interaction of collagen with the alpha2 beta1 integrin receptor which is expressed on transformed and normal human osteoblastic cells [53]. Francheschi and Iyers [54] demonstrated that ascorbate, a necessity for collagen synthesis in culture, was able to stimulate osteoblastic cell maturation as evidenced by induction of both alkaline phosphatase and osteocalcin. Specific inhibitors of collagen synthesis were able to block the maturation of osteoblastic cells. This suggests that binding of osteoblasts to collagen results in autocrine or paracrine signaling that is required for osteoblastic maturation. Harada and colleagues also examined the role of collagen in osteoblastic maturation. They demonstrated that ascorbate stimulation of osteoblastic collagen production and cell proliferation was blocked by collagen synthesis in-

BURSHELL AND SMITH

hibitors and by peptides that competitively block collagen binding to integrin receptors [55]. Similarly, Francheschi and colleagues [56] provided evidence for a model which proposes that ascorbic acid induces differentiation of osteoblasts by allowing the production and secretion of type I collagen into the extracellular matrix which then binds to cell surface integrins and stimulates osteoblast maturation. Matsumoto et al. [57] suggest that IGF-I production and action require collagen binding to an RGD receptor as well. In summary, in vitro evidence suggests that type I collagen is involved in the normal proliferation and maturation of cultured osteoblastic cells. This basic scientific information concerning collagen emphasizes its potential importance for bone and its potential genetic importance.

D. Genotypic Overlap between OI and Osteoporosis Most patients with OI are diagnosed at birth or in early childhood but milder cases may be difficult to differentiate from osteoporosis (see Fig. 2). Shapiro et al. [58] reported a case of familial osteopenia, hypermobile joints, short stature, blue sclerae, and scoliosis. Mutation detection and sequence analysis revealed a G–T point mutation at the first portion of the helical codon 43 causing a cysteine for glycine substitution in type I 1 collagen which was responsible for the phenotype. Analysis of the family confirmed linkage of the mutation to the phenotype. Spotila and colleagues [59] reported a case of phenotypic overlap between OI and osteoporosis with fractures of the axial skeleon, hearing loss, and osteopenia. DNA sequencing revealed a serine-for-glycine substitution at position 661 and evidence for posttranslational overmodification of type I 2 collagen. The authors screened 26 patients with low bone mineral density, most of whom had a family history of fracture or osteopenia but none of whom fulfilled criteria for OI. Point mutations in the 2 (I) and 1 (I) collagen genes were found in 3 patients. The remaining 23 patients had normal DNA sequences for both collagen genes. No family data were available to confirm linkage of these mutations with the phenotype but the mutations were not found in 81 normal individuals or in 37 additional osteopenic patients. The prevalence of mutations in the general population or in other defined populations such as familial osteoporosis or postmenopausal osteoporosis is unknown. To confirm the relationship between the mutation and a given phenotype, family or larger population studies are required. Once again, description of the phenotype of familial osteoporosis and identification of affected families becomes crucial. Thus, from a genetic perspective, there is growing evidence that in some families, abnormalities of collagen may present clinically as familial osteoporosis.

CHAPTER 45 Familial Osteoporosis

In support of the above hypothesis, Grant et al. [60] have identified a polymorphism in the first intron of type I 2 collagen. This G-to-T substitution was overrepresented in patients with a low BMD of the lumbar spine compared to normal controls.

E. Using Physical Examination to Increase Yield in Testing Collagen as a Candidate Gene in Osteoporosis We are studying a large Acadian kindred with Type I mild OI, because of our interest in the effect of Type I collagen on skeletal and extraskeletal tissues and to further explore the similarities and differences between Type I OI and osteoporosis. We have identified a single-base mutation within chromosome 17 in the Procollagen alpha I region in affected family members and have proven this to be a null cell mutation. Affected family members only have transcription of one of their two procollagen 1 genes and thus produce less type I collagen protein. Thus, the primary problem is a quantitative decrease in type I collagen rather than a qualitative defect in collagen. As expected children in the family have long-bone fractures and the frequency of fractures declines after puberty, which may be related to the increase in bone size. Vertebral and hip fractures tend to occur after menopause. All affected members of the family have blue sclera and adults may have premature hearing loss. The skin of the dorsum of the hand is thinner and the flexibility at the left second metacarpophalangeal joint greater in affected OI patients. The shoe width in the women is frequently AA and the width of long bones, i.e., radial width, is narrower than unaffected family members. In addition, the OI face has a characteristic appearance. Preliminary studies of the feet indicate a collapse of the hindfoot, i.e., a lower calcaneal declination angle and a calcaneal cone (narrowing of the posterior portion of the calcaneus on lateral radiographs). The calcaneous has an anterior endochondral ossification center and a posterior intramembranous center which may be related to the calcaneal cone. These clinical observations are compatible with OI Type I being primarily an intramembranous bone problem and affecting bone size, particularly long-bone width. The narrow long bones are more likely to fracture because of the bending forces. It is likely that these genetic changes begin in utero. In addition, there are multiple extraskeletal manifestations, including joint laxity, and the ligamentous changes may indirectly affect, the bone. Although some of these findings may be speculative they emphasize the potential importance of the phenotype. Our present thoughts are that a quantitative abnormality of type I collagen may cause developmental changes beginning in utero, preferentially affect intramembranous bone, and have extraskeletal manifestations. A subgroup of patients with

203 osteoporosis may have a similar phenotype. Thus, the small petite woman or man with osteoporosis may have a quantitative decrease in collagen or other matrix proteins contributing to their phenotype and related to the interplay of multiple genes. The phenotype may contribute to earlier diagnosis, a better understanding of pathophysiology, more specific therapies, and a better prediction of prognosis. It may also help with the development of a classification scheme for osteoporosis based on phenotype and genetics. Ehlers–Danlos Syndrome (EDS) is another disease in which the physical examination may help to identify patients at risk for osteoporosis. EDS is a group of inherited connective tissue disorders with skin fragility and joint hypermobility [61] and genetic and phenotypic similarities with both osteoporosis and OI. EDS VII is caused by abnormalities of type I collagen or an abnormality of the enzyme, N-terminal procollagen protease and patients with this disease have joint dislocations and may fracture [62]. Type IV EDS is caused by mutations of type III collagen which lead to aneurysmal ruptures. A mouse model shows that the type III collagen is necessary for type I fibrillogenesis [63]. Types I and II EDS have been associated with abnormalities of type V collagen in humans and mice [64]. Collagen V forms heterotypic fibrils with collagen I in many tissues and plays an important role in collagen I fibrillogenesis. Thus, EDS is associated with many mutant genes which may affect type I collagen and thus may have adverse effects on bone. These genes may be independently regulated, increasing the complexity of genes involved in the regulation of type I collagen. Thus, it is possible that type I collagen may be adversely effected by multiple genes and simply looking at the procollagen 1 and  2 genes may not be an adequate strategy. However, the phenotype is crucial to discovering and understanding the genetic relationships. Thus, our approach to type I collagen and families with osteoporosis is evolving. We continue to see patients with osteoporosis who also have other manifestations similar to OI Type I or EDS including thin skin, increased joint flexibility, and scoliosis. It is hoped that the family approach will yield either a higher incidence of collagen abnormalities or the potential for future genetic studies to identify other candidate genes which are known to effect collagen function. As discussed in section III, a simple approach to identifying collagen abnormalities is to use linkage analysis. The location of type I collagen is known to be on chromosome 17(pro a1) and chromosome 7(pro a2) and there are nucleotide repeats in close proximity to both genes. Nucleotide repeats are found throughout the human genome, are more variable than most alleles, are easily measured by polymerase chain reaction (PCR), and are highly informative. Other approaches, such as restriction fragment length polymorphism (RELP), are also convenient means to examine gene structure. Although, initially we hoped that abnormalities in the type I collagen genes would be frequent and

204 related to the development of osteoporosis, it seems more likely that multiple complex genes that control proteins that interact with type I collagen in the extracellular matrix will also require investigation. It is also possible that collagen both genetically and pathophysiologically does not play a major role in the development of osteoporosis. Even then the identification of large families with hereditary osteoporosis will still be helpful for the study of the genetics of osteoporosis. The critical aspect for the success of any of these approaches is to identify large families with hereditary osteoporosis and to carefully study the affected individuals.

IX. CONCLUSIONS Although not generally appreciated, the overwhelming epidemiologic evidence is that osteoporosis is primarily a genetic disease. The importance of genetics is greater in the young, with environmental factors becoming more predominant with age. Parallel to research in many other fields, the genetic aspects are now only beginning to be appreciated, but are likely to improve our understanding of the pathophysiology of osteoporosis as well as aid in developing new strategies of prevention and therapy. Similar to other diseases, the hereditary aspects of osteoporosis are likely to be due to multiple complex genes. The genetic approaches utilizing polymorphic markers to perform linkage analysis are readily available. We clinicians have the opportunity to identify key families with hereditary osteoporosis and colloborate with geneticists and molecular biologists to identify the genetic abnormalities. Since osteoporosis is frequently not apparent until advanced age, the phenotype may help identify and predict future fractures. At the present time, the BMD is the best predictor of future fractures and peak bone mass is under hereditary control, suggesting that measurement of the BMD may be able to separate affected and unaffected individuals. This assumption is not yet proven and may be true for some families and not others who may have osteoporosis more on the basis of bone loss than peak bone mass. A practical approach might be to identify index cases who have more severe disease defined by number of fractures, young age at fractures, male sex, and few environmental risk factors. A careful family history of fractures, physical examination for other extraskeletal findings, and BMD measurements should be performed on these index cases. The best location for BMD measurements is not obvious and multiple sites should be considered including the hip, total body calcium, lumbar spine, and radius. It is possible that specific locations are under different genetic controls. Once families with hereditary osteoporosis are identified, in collaboration with geneticists, the genetic abnormality may be discovered. One approach outlined in this chapter is the modified candidate gene approach. It is hoped that a national repository for specimens could then be established.

BURSHELL AND SMITH

References 1. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804–1809 (1988). 2. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Baseline measurement of bone mass predicts fracture in white women. Ann. Intern. Med. 111, 355–361 (1989). 3. P. D. Ross, R. D. Wasnich, and J. M. Vogel. Detection of prefracture spinal osteoporosis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1–11 (1988). 4. M. A. Hansen, K. Overgaard, and C. Christiansen, Spontaneous postmenopausal bone loss in different skeletal areas—Followed up for 15 years. J. Bone Miner. Res. 10, 205–210 (1995). 5. M. W. Thompson, R. R. McInnes, and H. F. Willard, “Genetics in Medicine,” 5th ed. Saunders, Philadelphia, 1991. 6. E. Lander and N. J. Schork, Genetic dissection of complex traits. Science 265, 2037–2048 (1994). 7. V. E. Velculescu, L. Zhang, B. Vogelstein, and K. W. Kinzler, Serial analysis of gene expression. Science 270, 484–487 (1995). 8. N. A. Datson, J. van der Perk-de Jong, M. P. van den Berg, E. R. de Kloet, and E. R. Vreugdenhil, MicroSAGE: A modified procedure for serial analysis of gene expression in limited amounts of tissue. Nucleic Acids Res. 27(5), 1300–1307 (1999). 9. D. G. Spinella, A. K. Bernardino, A. C. Redding, P. Koutz, Y. Wei, E. K. Pratt, K. K. Myers, G. Chappell, S. Gerken, and S. J. McConnell, Tandem arrayed ligation of expressed sequence tags (TALEST): A new method for generating global gene expression profiles. Nucleic Acids Res. 27(18), e22 (1999). 10. P. R. Jungblut, U. Zimny-Arndt, E. Zeindl-Eberhart, J. Stulik, K. Koupilova, K. P. Pleissner, A. Otto, E. C. Muller, W. SokolowskaKohler, G. Grabher, and G. Stoffler, Proteomics in human disease: Cancer, heart and infectious diseases. Electrophoresis 20(10), 2100–2110 (1999). [Review] 11. T. J. Aitman, A. M. Glazier, C. A. Wallace, L. D. Cooper, P. J. Norsworthy, F. N. Wahid, K. M. Al-Majali, P. M. Trembling, C. J. Mann, C. C. Shoulders, D. Graf, E. St. Lezin, T. W. Kurtz, V. Kren, M. Pravenec, A. Ibrahimi, N. A. Abumrad, L. W. Stanton, and J. Scott, Identification of Cd36 (Fat) as an insulin-resistance gene causing defective fatty acid and glucose metabolism in hypertensive rats. Nat. Genet. 21(1), 76–83 (1999). 12. M. Oyama, A. Tatlock, S. Fukuta, K. Kavalkovich, K. Nishimura, B. Johnstone, P. D. Robbins, C. H. Evans, and C. Niyibizi, Retrovirally transduced bone marrow stromal cells isolated from a mouse model of human osteogenesis imperfecta (oim) persist in bone and retain the ability to form cartilage and bone after extended passaging Gene Ther. 6(3), 321–329 (1999). 13. D. N. Kalu, The ovariectomized rat model of postmenopausal bone loss. Bone Miner. 15, 175–191 (1991). 14. Y. Zhang, R. Proenca, M. Maffel, M. Barone, L. Leopold, and J. M. Friedman, Positional cloning of the mouse obese gene and its human homologue. Nature 372, 425–432 (1994). 15. M. P. Muriel, J. Bonaventure, R. Stanescu, P. Maroteaux, J. L. Guenet, and V. Stanescu, Morphological and biochemical studies of a mouse mutant (fro/fro) with bone fragility. Bone 12, 241–248 (1991). 16. T. Tsuboyama, K. Takahashi, T. Yamomuro, M. Hosokawa, and T. Takeda, Cross-mating study on bone mass in the spontaneously osteoporotic mouse. Bone Miner. 23, 57–64 (1993). 17. W. G. Beamer, L. R. Donahue, C. J. Rosen, and D. J. Baylink, Genetic variability in adult bone density among inbred strains of mice. Bone 18(5), 397–403 (1996). 18. R. F. Klein, S. R. Mitchell, T. J. Phillips, J. K. Belknap, and Orwoll, Quantitative trait loci affecting peak bone mineral density in mice. J. Bone Miner. Res. 13(11), 1648–1656 (1998).

CHAPTER 45 Familial Osteoporosis 19. M. E. Cohen-Solal, C. Baudoin, M. Omouri, D. Kuntz, and M. C. De Vernejoul, Bone mass in middle-aged osteoporotic men and their relatives: familial effect. J. Bone Miner. Res. 13(12), 1909–1914 (1998). 20. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Engl. J. Med. 332, 767–773 (1995). 21. K. M. Fox, S. R. Cummings, K. Powell-Threets, and K. Stone, Family history and risk of osteoporotic fracture. Study of Osteoporotic Fractures Research Group. Osteoporos Int. 8(6), 557–562 (1998). 22. R. W. Keen, D. J. Hart, N. K. Arden, D. V. Doyle, and T. D. Spector, Family history of appendicular fracture and risk of osteoporosis: A population-based study. Osteoporosis Int. 10(2), 161–166 (1999). 23. D. M. Black, N. K. Arden, L. Palermo, J. Pearson, and S. R. Cummings, Prevalent vertebral deformities predict hip fractures and new vertebral deformities but not wrist fractures. Study of Osteoporotic Fractures Research Group. J. Bone Miner. Res. 14(5), 821–828 (1999). 24. E. Seeman, J. L. Hooper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554–558 (1989). 25. E. Seeman, C. Tsalamandris, C. Formica, J. L. Hopper, and J. McKay, Reduced femoral neck bone density in the daughters of women with hip fractures: The role of low peak bone density in the pathogenesis of osteoporosis. J. Bone Miner. Res. 9, 739–743 (1994). 26. N. Barthe, B. Basse-Cathalinat, P. J. Meunier, C. Ribot, X. Marchandise, J. P. Sabatier, P. Braillon, J. Thevenot, and B. Sutter, Measurement of bone mineral density in mother–daughter pairs for evaluating the family influence on bone mass acquisition: A GRIO survey. Osteoporosis Int. 8(4), 379–384. 27. R. Armamento-Villareal, D. T. Villareal, L. V. Avioli, and R. Civitelli, Estrogen status and heredity are major determinants of premenopausal bone mass. J. Clin. Invest. 90, 2464–2471 (1992). 28. S. B. Soroko, E. Barrett-Conner, S. L. Edelstein, and D. Kritz-Silverstein, Family history of osteoporosis and bone mineral density at the axial skeleton: The Rancho Bernardo study. J. Bone Miner. Res. 9, 761–769 (1994). 29. M. J. Grainge, C. A. Coupland, S. J. Cliffe, C. E. Chilvers, and D. J. Hosking, Association between a family history of fractures and bone mineral density in early postmenopausal women. Bone 24(5), 507–512 (1999). 30. D. M. Smith, W. E. Nance, K. W. Kang, J. C. Christian, and C. C. Johnston, Jr., Genetic factors in determining bone mass. J. Clin. Invest. 52, 2800–2808 (1973). 31. J. C. Christian, P. Yu, C. W. Slemenda, and C. C. Johnston, Jr., Heritability of bone mass: A longitudinal study in aging male twins. Am. J. Hum. Genet. 44, 429–433 (1989). 32. C. W. Slemenda, J. C. Christian, T. Reed, T. K. Reister, C. J. Williams, and C. C. Johnston, Jr., Long term bone loss in men: Effects of genetic and environmental factors. Ann. Intern. Med. 117, 286–291 (1992). 33. C. W. Slemenda, J. C. Christian, C. J. Williams, J. A. Norton, and C. C. Johnston, Jr., Genetic determinants of bone mass in adult women: A reevaluation of the twin model and the potential importance of gene interaction on heritability estimates. J. Bone Miner. Res. 6, 561–567 (1991). 34. N. A. Pocock, J. A. Eisman, J. L. Hopper, M. G. Yeates, P. N. Sambrook, and S. Eberi, Genetic determinants of bone mass in adults. A twin study. J. Clin. Invest. 80, 706–710 (1987). 35. P. J. Kelly, J. L. Hopper, G. T. Macaskill, N. A. Pocock, P. N. Sambrook, and J. A. Eisman, Genetic factors in bone turnover. J. Clin. Endocrinol. Metab. 72, 808–813 (1991). 36. P. J. Kelly, T. Nguyen, J. Hopper, N. Pocock, P. Sambrook, and J. Eisman, Changes in axial bone density with age: A twin study. J. Bone Min. Res. 8, 11–17 (1993). 37. J. Dequeker, J. Nijs, A. Verstraeten, P. Geusens, and G. Gevers, Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8, 207–209 (1987).

205 38. J. Lutz, Bone mineral, serum calcium, and dietary intakes of mother/daughter pairs. Am. J. Clin. Nutr. 44, 99–106 (1986). 39. J. Lutz and R. Tesar, Mother–daughter pairs: Spinal and femoral bone densities and dietary intakes. Am. J. Clin. Nutr. 52, 872–877 (1990). 40. F. A. Tylavsky, A. D. Bortz, R. L. Hancock, and J. J. B. Anderson, Familial resemblance of radical bone mass between premenopausal mothers and their college-age daughters. Calcif. Tissue Int. 45, 265–272 (1989). 41. V. Matkovic, D. Fontana, C. Tominac, P. Goel, and C. H. Chesnut, Factors that influence peak bone mass formation: A study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr. 52, 878–888 (1990). 42. S. Ferrari, R. Rizzoli, D. Slosman, and J. P. Bonjour, Familial resemblance for bone mineral mass is expressed before puberty. J. Clin. Endocrinol. Metab. 83(2), 358–361 (1998). 43. E. Krall and B. Dawson-Hughes, Heritable and life-style determinants of bone mineral density. J. Bone Miner. Res. 8, 1–9 (1993). 44. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis. JAMA 116, 2465–2474 (1941). 45. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nat Genet. 20(1), 78–82 (1998). 46. B. P. Lukert and L. G. Raisz, Glucocoriticoid-induced osteoporosis: Pathogenesis and management. Ann. Inter. Med. 112, 352–364 (1990). 47. R. W. E. Mellish, M. M. O’Sullivan, N. J. Garrahon, and J. E. Compston, Iliac crest trabecular bone mass and structure in patients with nonsteroid treated rheumatoid arthritis. Ann. Rheum. Dis. 46, 830–836 (1987). 48. R. H. Sheppard and H. E. Meema, Skin thickness in endocrine disease. Ann. Int. Med. 66, 531–539 (1967). 49. T. Diamond, L. Nery, and S. Posen, Spinal and peripheral bone mineral densities in acromegaly: The effects of excess growth hormone and hypogonadism. Ann. Int. Med. 111, 567–573 (1989). 50. J. K. Ferguson, R. A. Donald, T. S. Weston, and E. A. Espiner, Skin thickness in patients with acromegaly and Cushing’s syndrome and response to treatment. Clin. Endocrinol. 18, 347–353 (1983). 51. M. Brincat, Longterm effects of the menopause and sex hormones on skin thickness. Br. J. Obstet. Gynaecol. 92, 256–259 (1985). 52. M. P. Lynch, G. S. Stein, J. L. Stein, and J. B. Lian (Introduction by L. Barone). “The Role of Collagen Type I in the Development Expression of the Osteoblast Phenotype.” Fifteenth Annual Meeting of the American Society for Bone and Mineral Research, p. 506, Tampa, 1993. 53. R. Dunlay, S. L. Teitelbaum, K. Sago, P. Price, and F. P. Ross, “Differential Attachment to Collagen by Human Bone Cells and Osteosarcoma Cells Lines Correlates with Surface Expression of the Integrin.” Fifteenth Annual Meeting of the American Society for Bone and Mineral Research, p. 516, Tampa, 1993. 54. R. T. Franceschi and B. S. Iyer, The role of ascorbic acid in mesenchymal differentiation. J. Bone Miner. Res. 7, 235–245 (1992). 55. S. Harada, T. Matsumoto, and E. Ogata, Role of ascorbic acid in the regulation of proliferation in osteoblast-like MC3T3-E1 cells. J. Bone Miner. Res. 6, 903–908 (1991). 56. R. T. Franceschi, B. S. Iyer, and Y. Cui, Effects of ascorbic acid on collagen matrix formation and osteoblast differentiation in murine MC3T3-E1 Cells. J. Bone Miner. Res. 9, 843–854 (1994). 57. T. Matsumoto, S. Harada, H. Kawaguchi, and E. Ogata, Matrix proteins regulate the proliferation of osteoblastic cells by affecting the actions of IGF-I. Osteoporosis Int. 3(Suppl. 1), 117–120 (1993). 58. J. R. Shapiro, M. L. Stover, V. E. Burn, M. B. McKinstry, A. L. Burshell, S. D. Chipman, and D. W. Rowe, An osteopenic nonfracture syndrome with features of mild osteogenesis imperfecta associated with the substitution of a cysteine for glycine at triple helix position 43 in the pro alpha 1(I) chain of type I collagen. J. Clin. Invest. 89, 567–573 (1992).

206 59. L. D. Spotila, C. D. Constantinou, L. Sereda, A. Ganguly, B. L. Riggs, and D. J. Prockop, Mutation in a gene for type I procollagen (COL1A2) in a woman with postmenopausal osteoporosis: Evidence for phenotype and genotypic overlap with mild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423–5427 (1991). 60. S. F. A. Grant, D. M. Reid, and S. H. Ralston, “Osteoporotic Fracture and Reduced Bone Density Related to a Polymorphism in the Transcriptional Control Region of the Collagen Type I 1 Gene.” Seventeenth Annual Meeting of the American Society for Bone and Mineral Research, p. 165, Baltimore, 1995. 61. P. H. Byers, Ehlers–Danlos syndrome: Recent advances and current understanding of the clinical and genetic heterogeneity. J. Pediatr. 135(4), 494–499 (1999).

BURSHELL AND SMITH 62. P. H. Byers, M. Duvic, M. Atkinson, M. Robinow, L. T. Smith, S. M. Krane, M. T. Greally, M. Ludman, R. Matalon, S. Pauker, D. Quanbeck, and U. Schwarze, Ehlers–Danlos syndrome type VIIA and VIIB result from splice-junction mutations or genomic deletions that involve exon 6 in the COL1A1 and COL1A2 genes of type I collagen. J. Invest. Dermatol. 103(5 Suppl.), 47S–52S (1994). 63. X. Liu, H. Wu, M. Byrne, S. Krane, and R. Jaenisch, Type III collagen is crucial for collagen I fibrillogenesis and for normal cardiovascular development. Hum. Mutat. 9(4), 300–315 (1997). 64. A. De Paepe, L. Nuytinck, I. Hausser, I. Anton-Lamprecht, and J. M. Naeyaert, Mutations in the COL5A1 gene are causal in the Ehlers-Danlos syndromes I and II. Nat. Genet. 9(1), 31–36 (1995).

CHAPTER 46

Immobilization Osteopenia B. JENNY KIRATLI

I. II. III. IV.

Spinal Cord Injury Center, Veterans Affairs Palo Alto Health Care System, Palo Alto, California 94304

V. Intervention Attempts VI. Clinical Relevance References

Overview Clinical Conditions of Immobilization Evidence of Skeletal Response to Immobilization Revelent Sequelae in Spinal Cord Injury Patients

I. OVERVIEW

these different measurement techniques, information has been obtained on the cellular, tissue, and organ level; however, historically, the choice of technique has usually been limited by availability. Bone loss with immobilization has also been induced in a variety of animal models; experimental designs have included encasement of a limb or the entire body in plaster, suspension of the body or portion of the body, and surgical elimination of innervation or muscular attachment to a designated limb. Significant bone loss occurs in all animal models with reduction of mechanical loading, and data have been collected on structural and mechanical changes as well as cellular mechanisms which cannot be directly studied in humans. This chapter, however, will be restricted to a discussion of clinical evidence and human experimental studies of immobilization osteopenia. There is a large body of evidence from human and animals studies, utilizing a variety of measurement techniques, which indicates that bone loss follows immobilization, and reduction in mechanical loading is usually stated as the “cause.” The specific causal mechanism, however, has not been determined. Over 30 years ago (1968), Hattner and McMillan reviewed the evidence from human and animal studies available at that time and discussed three possible mechanisms: (i) reduction of stresses and strains secondary to weightbearing and muscle tension, (ii) neurally mediated influences, and (iii) vascular changes. Their conclusion was

Reduction in mechanical loading results in bone loss, often called “disuse osteoporosis.” Since the 1940s, this phenomenon has been described for a variety of pathological, traumatic, and experimental conditions. Perhaps a more correct term would be discuse (or immobilization) osteopenia because the observed bone loss is not always excessive (depending on the level of immobilization) and does not always result in fracture, although heightened fracture risk may be present. Further, fracture risk may be difficult to assess under some conditions of immobilization, where habitual loading patterns differ from “normal” and result in different bone strength requirements. That is, the femur of a spinal-cord-injured person who never stands or stands only to transfer may not be subjected to the same loading forces as the femur of an ambulatory person; does a fracture risk assessment based on ambulatory normative data apply to this individual? The objective of this chapter is to describe the clinical evidence for and consequences of bone loss (osteopenia) associated with immobilization. Various methods have been used to examine the skeletal response to immobilization, including assays of biochemical markers of bone turnover, collection of histologic and histomorphometric data, and measurements of bone mass by rediographic and densitometric techniques. By use of

OSTEOPOROSIS, SECOND EDITION VOLUME 2

207

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

208

B. JENNY KIRATLI

that “the most important determinants appear to be mechanical strain and compression of bone by interaction of muscles and gravity.” The same theoretical approach is followed here, but potential contributions of neural and vascular influences in causing or mediating the observed bone responses should not be overlooked. Furthermore, even if immobilization “directly” causes bone loss through elimination of normal (and necessary) mechanical stimuli, the mechanism of this action is not definedd and may certainly include hormonal, neurohumoral (direct neural effect and/or neural factors), and piezoelectric influences. Few data are available on the relative importance of these mechanisms or on the details of their potential effects.

II. CLINICAL CONDITIONS OF IMMOBILIZATION A continuum of human conditions of immobilization are presented below in which bone atrophy has been observed, ranging from complete motor paralysis through reversible immobilization. Viewed together, the results from these studies provide evidence on the contributions of weightbearing and muscular activity in regulation of human bone response. There is much concordance of observed response among the different immobilization conditions, despite different methodologic approaches.

A. Paralysis Complete paralysis due to traumatic spinal cord injury represents the most extreme case of immobilization, where neither muscle activity nor weightbearing are present in body regions below the level of neurologic lesion. It has long been known that bone loss occurs in spinal-cord-injured patients invilving affected body region(s) [1 – 3]; this osteopenia is accompained by related disturbances, such as incresed occurrence of renal calculi and ossification of periarticular soft tissues (heterotopic ossification). Both associated conditions presumably result from increased mobilization of calcium from the bones. Further, fracture risk is elevated with spinal cord injury and is higher in patients with complete lesions than those with incomplete lesions where residual muscle function may afford some degree of protection [4,5]. Much of this discussion will focus on evidence from studies of bone response following spinal cord injury as this is the most profound clinical occurrence of immobilization and is currently a fairly active area of research. It should be noted, however, that spinal cord injury does not always in complete cessation of motor function (paralysis), and with improved acute medical care, more idividuals survive actue spinal trauma with incomplete motor injuries and retain

some voluntary motor function. Furthermore, bone loss subsequent to spinal cord injury occurs only in affected (paralyzed) regions; nonaffected regions may show disparate responses related to localized mechanical loading.

B. Paretic Disorders 1. POLIOMYELITIS The next level of immobilization, related in some ways to spinal cord injury (and paralysis), can be found with poliomyelitis, a paretic (and sometimes paralytic) disease of the peripheral nervous system. Much of the early clinical awareness of bone loss due to immobilization came from studies of poliomyelitis [6]. Affected patients showed a wide spectrum of clinical findings, as many were initially paralyzed but later recovered differing amounts of function, and many of those affected were children. These individuals were often encoureged to participate in rigorous physical therapy regimens, and ambulation was accomplished with therapist assistance and orthotic aids. The majority of the published work describes metabolic responses only, as densitometry was not yet available, and little recent work has been conducted on postpolio survivors. 2. HEMIPLEGIA AND HEMIPARESIS A unique set of conditions under which to study bone response with reduced mobility is found with the hemiparetic (hemiplegic) occurrences of stroke or Brown-Sequard syndrome, cerebral palsy, and unilateral amputation. In hemiparetic conditions, the nonaffected limb can be considered a physiologic control for study of regional changes due to immobilization, thus eliminating the confounding factors of genetics and life style (nutrition, medication, cigarette or alcohol use, etc.) which clearly do not differ between the limbs. One critical assumption underlying assesment of bone response in hemiparetic patients is that of bilateral symmetry. In studies of bilaterality in nonathletes, however, differences in bone mineral or cross-sectional properties are usually observed between limbs, as high 7% (metacarpal) [7] and 4.6% (humerus) [8] in the upper extremity and 11% (proximal femur) in the lower limb [9,10], while one study found no difference in bone mass between femurs [11]. The observed differences are often attributed to handedness and footedness (although the latter is less easily determined) and may thus reflect a difference in habitual use and mechanical loading history. A decrease in asymmetry with age [8], which presumably accompanies diminished use and activity of the dominant limb, is suggested as further evidence that asymmetry in bone properties reflects asymmetrical mechanical usage patterns. Therefore, both immobilization and underlying dominance may contribute to observations of asymmetry in bone mass following a hemiparetic event.

209

CHAPTER 46 Immobilization Osteopenia

C. Temporary Immobilization A final category of human studies involves nonpermanent immobilization due to medical or therapeutic cause, voluntary bed rest, and spaceflight. Clinical investigations have determined that therapeutic recumbency and limb or body immobilization result in regional bone loss which is largely reversible with return to normal mobility. These studies are primarily observational and uncontrolled. Studies of astronauts before, during, and after spaceflight have provided evidence of the effects of hypogravity and hypodynamics on bone response. Research has also been conducted on human volunteers undergoing prolongedd bed rest, originally begun as groundbased simulations of spaceflight. The direct relevance of data from astronaut studies to clinical immobilization might be questioned because of the overwhelming condition of hypogravity, in addition to reduced activity, as well as other undetermined infulences of spaceflight. However, data obtained from bed rest studies provide many relevant parallels to clinical immobilization.

III. EVIDENCE OF SKELETAL RESPONSE TO IMMOBILIZATION There are generally three approaches to quantifying the skeletal response: biochemical measurement of markers of bone metabolism, histomorphometric assessment of local bone activity, and regional analysis of bone mass or mineral content. Results from these different techniques are related but different processes and different scales of effects are represented. Biochemical assessment can identify the earlist response as initial alterarion of bone metabolism occurs. However, while urine and serum assays represent a systemic level response, the changes in bone cell activity are usually quite localized; thus, measured (metabolic) responses are diluted and cannot be attributed to any specific body region(s). Histomorphometric assessment of transiliac biopsies provides evidence of local cellular activity, which is the next sequential response, but the results only directly applicable to the iliac region and extrapolations to other skeletal regions are limited. Direct measurement of bone mass or mineral content at specific regions provides that most important clinical level of assessment and represents the third sequential effect. Measureable evidence of bone remodeling may take many weeks and months, depending on the assessment technique and the magnitude of the response. In the studies discussed below, various measurement methods and combinations of methods are used in prospective and retrospective evaluation of bone response to immobilization. While not all methods are used in all studies, and seemingly discrepant observations may result in some instances from the various study designs, in general, the total

picture presented is consistent. It is also important to note that the scale of the effect differs and that this may influence the apparent timing of the response. That is, while metabolic activity may increase immediately, the magnitude of the change in skeletal tissue will determine when a detectable change in bone mass occurs.

A. Bone Loss with Paralysis and Paresis Acute diminution of mechanical loading due to paralysis or paresis is immediately accompanied by increased bone remodeling activity. The net result is elevated resorption and bone loss. Evidence of changes in metabolic and histologic activity and in bone mass are reviewed below. 1. METABOLIC EVIDENCE Much of the early research on bone loss following spinal cord injury focused on changes in metabolites in the serum and urine as indicators of bone resorption and formation activity. Urinary hydroxyproline and calcium reflect collagen turnover and demineralization, respectively. Serum calcium is also expected to reflect elevated bone resorption, and bone formation activity is indicated by serum concentration of alkaline phosphatase, an enzyme produced by osteoblasts and associated with osteoblastic activity. Longitudinal studies of blood and urine chemistery in recently paralyzed patients demonstrates immediate elevation of bone resorption as well as bone formation [12 – 14]. Urinary calcium is elevated immediately after injury and peaks at 8 – 10 weeks at approximately twice the normal level. After 15 weeks, calciuria then falls to a stable concentration, still above normal levels. Normal calcium excretion is observed after 7 months. Urinary hydroxyproline shows a pattern similar to that of urinary calcium. Hydroxyproline output is elevated immediately, peaks at 3 months, remaining high for as long as 8 months, and then declines steadily to a stable level within 1 year. Serum calcium in adult spinal-cord-injury patients is not significantly different from normal while hypercalcemia with immobilization is commonly found in children and adolescents [15 – 17] and is attributed to their being in a rapid growth phase [18,19]. Good to moderate correlation is reported between urinary levels of calcium and hydroxyproline in spinal-cordinjury patients [12,20]. The excretion of urinary glycosaminoglycans (a constituent of noncollagenous bone matrix) also increases after immobilization in spinal-cordinjury patients and then declines to a reduced steady state after 8 to 9 months [14,21]. This pattern is parallel to that of hydroxyproline excretion and confirms that bone resorption includes both organic and inorganic fractions of bone tissue. That is, demineralization does not occur separately from matrix degradation.

210 In contrast to the above, Claus-Walker et al. [22] reports that levels of calciuria and hydroxyprolinuria are uncorrelated in paralyzed patients (from onset to 84 weeks duration); elevation in hydroxyproline excretion is observed immediately after injury while hypercalciuria is not found until a few weeks later. The authors conclude that bone collagen resorption precedes demineralization. It is further suggested that immobilization causes increased collagen formation, in addition to elevated bone resorption, and that the hyperhydroxyprolinuria which continues after hypercalciuria has resolved might be the result of breakdown of newly formedd collagen, which is more soluble. Elevations in total serum alkaline phosphatase, an indicator of bone formation, occur gradually over 5 to 9 weeks postinjury. While Klein and associates [23] report no correlations between serum alkaline phosphatase and calciuria or hydroxyprolinuria, Bergman et al. [12] find a moderate correlation between levels of serum alkaline phosphatase and urinary hydroxyproline. Ohry et al. [24] also observe serum alkaline phosphatase to be elevated above control values but still within normal limits. It is suggested by Klein and co workers that the elevation in alkaline phosphatase represents an “abortive attempt to reactivate matrix in the face of massive loss of bone.” Regional remodeling activicty has been measured by calcium accretion in paraplegic patients, using radioactively labeled calcium [12]. Calcium accretion, indicating mineral deposition in new bone, peaks between 3 and 10 months, and a moderate correlation exists between calcium accreation and hydroxyprolinuria. Initially (during the first 2 months), turnover activity observed in the nonparalyzed skeleton above the level of the spinal cord lesion is greater than that in the lower skeleton, but no net bone mineral loss occurs in the upper body. After 2 months, bone remodeling is more active in the lower skeleton, and calcium balance is negative, reflecting a net loss. Similar results are reported by Chantraine and colleagues [20], who find bone accretion to be normal or elevated in paraplegic patients. Therefore, the net bone loss which occurs below the lesion must result from resorption accelerated above the also-increased level of formation. In summary, biochemical markers provided valuable information about bone tissue activity, expecially before the advent of reliable and accurate methods of measuring bone mass. The atrophy involves both the organic and the inorganic constituents of bone tissue, and total bone mass is reduced. Although data on the effects of paralysis on bone formation are less consistent than data on bone resorption, much of the evidence indicates that formation is also elevated. This is not surprising as the processes are normally coupled with bone formation stimulated by resorptive activity. The drawback of metabolic evidence is that no information is available regarding the specificity of tissue response, and because bone responds locally to changes in

B. JENNY KIRATLI

mechanical stimuli, estimates of regional response and change in specific bones is needed. 2. HISTOLOGIC EVIDENCE Bone remodeling has beeen examined in tibial and iliac bone biopsies from paraplegic patients [25]. The degree mineralization of the organic matrix of newly formed bone is determined by bone particle fractionation. Evidence of increased turnover, “increased proportion of little mineralized particles,” is found in both cortical and cancellous bone in the newly injured patients compared with normal controls. However, paraplegic patients measured 2 years after injury display fractionation profiles similar to normal. These data are consistent with metabolic evidence of increased remodeling rate with the onset of paralysis and resumption of a normal rate of remodeling with chronic paralysis. Histomorphometric data collected on iliac biopsies from spinal cord injury patients have been used to quantify cellular changes in bone with immobilization [13]. Trabecular bone volume (the percentage of bone occupied by trabeculae) decreases at the rate of 6% per month and is correlated with the duration of paralysis, up to 25 weeks after onset. After this, there is no further relation with duration of paralysis, and a new steady state is assumed to have been reached. The number of trabecular osteoclastic resorptive surfaces is elevated and remains high for the frist 16 weeks, gradually returning to normal by 40 weeks. These results are consistent with clinical evidence of hypercalciuria and hyperhydroxyprolinuria. In addition, a significant correlation is found between the number of resorptive surfaces and the levels of urinary hydroxyproline. Finally, osteoblastic appositional rate, assessed by double tetracycline labeling, is greatly reduced following injury. Thus, it is concluded that bone loss occurs due to reduced bone formation and a transitory increase in osteoclastic resorption, followed by a new steady state characterized by “lazy” bone, which “produces far less bone per unit time than it did before.” However, this result is inconsistent with the metabolic evidence of elevation in formation activity, presented above (increased calcium accretion and elevated serum alkaine phosphatase levels). 3. CHANGES IN BONE MASS Localized osteopenia due to paralysis following spinal cord injury or polio myelitis has been known for over four decades, originally documented by radiography. More recently, with the development and availability of densitometric techniques, it has also been possible to study the magnitude and time course of bone atrophy with immobilization. The importance of site specific response is evident from consideration of unilateral paralysis and paresis and disparate observations between paraplegia and quadriplegia. Table 1 summarizes results from studies of bone mass in patients with acute and chronic paralysis of paresis.

211

CHAPTER 46 Immobilization Osteopenia

TABLE 1 Medical circumstances

Summary of Studies Reporting Bone Response following Paralysis and Paresis Subjects (age range)

Assessment method

Abbreviated results

Ref.

Permanent bedfastness due to developmental disability

299 M/F (5 – 30)

Radiography and photometry of calcaneus

Compared with age-matched, BMC reduction of 21 – 42%. Greater loss with greater hypokinesis (low activity level).

26

SCI (para/ quadriplegia), chronic injury

36 M/F (20 – 55)

Densitometry of forearm; hand radiography

Loss of trabecular bone in distal forearm, no cortical effect. No correlation with duration of injury.

27

SCI (paraplegia), acute and chronic injury

66 M (10 – 63)

Densitometry of femoral shaft

Compared with able-bodied, reduced BMC of femoral shaft (27%). No correlation with duration of injury, nor with spasticity.

28

SCI (paraplegia), wheelchair athlete champions

2F (25,27)

Densitometry of forearm

Compared with able-bodied female nonathletes, increased BMD in radius (18%) and ulna (26%).

29

SCI (para/ quadriplegia), chronic injury

40 M/F (13 – 69)

Radiographic studies of acetabulum

Double cortical lines in acetabulum observedin most (37 patients).

30

SCI (para/ quadriplegia), chronic injury

26 M/F (20 – 46)

Densitometry of spine, femoral neck proximal tibia

Compared with able-bodied, reduced BMC of femoral neck (22%) and tibia (50%); increased spine (12%). Lower values in quadriplegic than paraplegic subject. No effects of spasticity nor daily leg brace use.

31

SCI (para/ quadriplegia), acute and chronic injury

45 M/F (14 – 60)

Densitometry of spine and proximal femur

Acute BMD reduction in proximal femur 1 – 2% per month; no change in spine. Chronic BMD reduction in femur (28%) compared with expected; slight, but not significant, increase in spine (4%).

32

SCI (para/ quadriplegia), acute injury

8 M/F (18 – 49)

Densitometry of spine, forearm, femoral neck and shaft, proximal tibia

Rapid BMC loss immediately after spinal cord-injury, new steady state in 2 years at 40 – 50% (femur) and 60 – 70% (tibia) of expected values. Slower loss in the femoral shaft and no evidence of cessation or reaching new steady state. No change in forearm or spine BMC.

33

SCI (para/ quadriplegia), acute and chronic injury

45 M (All 40)

Densitometry of whole body

Chronic spinal cord injury show reduction in BMC of arms (26%), pelvis (46%), legs (49%); no change in trunk; increase in head (3%).

34

SCI (para/ quadriplegia), with varying mobility

27 M (Mean  42)

Densitometry of tibia, bilateral measurements

Reduction in tibia BMC. Correlation with mobility index and asymmetrical motor sparing (in Brown-Sequard patients).

35

SCI (paraplegia), chronic injury

19 M (15 – 64)

Densitometry of forearm and tibia

Reduction in forearm BMC of distal diaphysis(5%), distal metaphysis (13%). Reduction in tibia BMC of distal diaphysis (26%), distal metaphysis (45%).

36

SCI, chronic injury

14 M (Mean  32)

Densitometry of spine and femoral neck

Reduction in femoral neck BMD (86% age-matched). No reduction in spine (101% age-matched).

37

SCI (para/ quadriplegia), acute and chronic injury

37 M/F (19 – 64)

Densitometry o f proximal and distal tibia

Expected reduction in proximal tibia trabecular subcortical, and cortical bone (52, 42, and 32%, respectively) after 2 years; similar results for distal site. Additional losses projected for 2 – 7 years (14, 13, 11%, respectively).

38

Stroke with unilateral weakness (hemiparesis), chronic injury

25 M/F (29 – 85)

Radiography of humerus, radius, metacarpal, bilateral measurements

Reduction in combined cortical thickness of humerus (25%), radius (19%), and metacarpal (24). Less cortical thinning with better motor function. No effect of spasticity.

39

Stroke with unilateralweakness (hemiparesis), acute and chronic injury

74 M/F (Mean  69)

Densitometry of forearm, bilateral measurements

Bone loss positively correlated with duration of stroke, negatively with motor function. Estimated loss rate for trabecular (1.3%) and cortical bone (1.5) per year for affected versus normal limb; no evidence of steady state. Negative association between trabecular but not cortical bone and spasticity.

40

(continues)

212

B. JENNY KIRATLI

TABLE 1 Medical circumstances

Subjects (age range)

Assessment method

(continued) Abbreviated results

Ref.

Stroke with unilateralweakness (hemiparesis), chronic injury

15 M/F (Mean  63)

Densitometry of whole body, bilateral measurements

Reduction in arm BMC (10%) and leg BMC (4%). No correlation with spasticity.

41

Stroke with unilateral weakness (hemiparesis), chronic injury

30 M (55 – 75)

Densitometry of whole body, bilateral measurements

Reduction in arm BMD (1 4 % ) , BMD (8%). Reduction in leg BMC (5%), BMD (3%). Duration of paralysis more correlated with arms than legs.

42

Note. SCI, spinal cord injury.

Most of the bone decline with spinal cord injury and paralysis occurs during the first year after injury [31 – 33]. Based on data from cross-sectional studies, spinal cord injury patients with “relatively long-standing” paralysis do not continue to lose bone; i.e., bone mass is not related to duration of injury in patients injured between 2 and 40 years earlier [27,28,32,43]. In contrast, Prince et al. [40] suggest that bone loss continues up to 15 years after immobilization by stroke. Perhaps this seemingly contrary result reflects a difference between complete and partial paralysis; there may be continued bone loss where residual muscle forces maintain at least partial stimulation of bone cell activity. Although there are some discrepancies among studies, it is generally found that bone loss occurs regionally (only in affected areas), is related to the completeness of paralysis, and is more evident in trabecular than cortical bone. There is substantial evidence of reduction in lower extermity bone density of paraplegic and quadriplegic patients [28,31 – 33] with reduction in bone mass of the tibia twice as great as loss in the femur. Bone loss begins immediately after injury with a mean rate of loss in the hip of greater than 2% per month for the first 4– 6 months and approximatley 1% per month for the latter part of the first year [32]. This significant rate of loss resolves during the third or fourth year, and there is little evidence of measureable loss afterward. While less loss occurs with incomplete paralysis related to the amount of residual muscle function, no influence on bone response can be found related to spasticity or use of standing braces. Lumbar spine bone mass is not different from expected control (ambulatory) values in both retrospective and prospective examinations of paralyzed individuals [32,33]. It is suggested that the mechanical loading of the lumbar spine in the upright sitting posture and during transfer activity is sufficient to protect against bone loss. In fact, there appears to be a slight increase in lumbar spine bone mass with chronic paralysis which might result from unusual

loading patterns (e.g., abnormal postural positions and excessive ligamentous loads) incurred with sustained wheelchair-sitting. In a study of total body bone response, significant decreases are noted in the upper extremity, pelvis,and throughout the lower extremity during the first 16 months postinjury, compared with controls, but no differences are observed in bone mass of the trunk or head [43]. The findings of total body and lower extremity bone loss and absent trunk response are consistent with results from longitudinal studies of acute spinal cord injury [31,32]. However, the finding of upper extremity bone loss in paraplegic patients is not consistent with results from several other reports of acute and chronic paralysis [29,33,44]. Upper extremity bone response has been studied in paralyzed and stroke patients for 25 years, and most of the evidence is in accordance with the expectation that immobility leads to bone atrophy. An early stay utilizing radiographic comparisons between affected and unaffected humeri, radii, and third metacarpals of hemiplegic stroke patients demonstrates significantly smaller cortical thicknesses on the affected side [39]. Further, cortical thinning is related to the degree of residual motor function in patients who recovered some motor ability, but the presence or absence of spasticity does not alter the response. In a study of spinal-cord-injured individuals, the metacarpal cortical indices (total thickness of cortical bone divided by the diameter) are found to be normal in paraplegics and decreased in quadriplegics, wheras femoral cortical indices are generally reduced in both [45]. However, a similar study finds no evidence of upper extermity cortical atrophy following spinal cord injury while significant trabecular loss is observed [27]. Further, although this finding is more marked in quadriplegic patients, it is also apparent in some of the paraplegic patients. While expected in the quadriplegic limb where function is greatly compromised, osteopenia is unexpected in the paraplegic upper extermity which is subjected to greater

213

CHAPTER 46 Immobilization Osteopenia

than normal loading during lifting maneuvers and wheelchair mobility; these loads might be expected to preserve or augment skeletal mass. Two recent studies show this same discrepancy. Garland et al. [34] finds reduced arm bone mass in both quadriplegic and paraplegic subjects, while Kiratli and colleagues [44] describe significantly greater bone mass in paraplegic subjects as well as quadriplegic subjects who habitually use manual wheelchairs but reduced bone mass in quadriplegics who are unable to use manual wheelchairs. There are no explanations for these inconsistent results (i.e., absence or presence of atrophy in the paraplegic upper extremity), except that some of the samples are small and might have involved nonrandom group differences in activity level, muscle strength, age, or some other confounding factor which affects arm bone mass. Densitometric measurments comparing the normal and affected arms of hemiplegic stroke patients have been used to differentiate the effects of immobilization, and residual activity on trabecular bone [40]. The decline in cortical bone depends primarily on the duration of immobilization, while decline in trabecular bone with duration of paralysis is influenced by muscle stresses imposed by spasticity and functional motor ability. The predominance of trabecular bone response over cortical has also been demonstrated by Iverson et al. [41] in a study somparing arms and legs of hemiparetic and hemiplegic patients by whole body bone densitometry. The decrese (difference between affected and unaffected, divided by the unaffected) in the mostly trabecular distal arm bone (13%) significantly exceeds that in the mostly cortical proximal site (8%). Further, there is a greater loss in the arms (10%)than in the legs (4%). To explain the latter result, the authors propose that a hospitalization-dependent in overall activity leads to a decline in bone mass of the unaffected leg as well; this might effectively reduce the observed “loss,” as determined by the ratio of affected to unaffected bone mass. These data also indicate the importance of local muscular strain in maintenance of normal bone of the upper limbs. The evidence of bone loss with complete or partial paralysis is relatively consistent, but there remains much to be studied. For instance, there is little information regarding the importance of previous physical condition or differences in response due to age gender (as the majority of acutely injured patients are young and male). With current advances in early medical attention, many newly injured patients sustain incomplete spinal cord injury and retain or regain substantial motor function. Little research has yet been conducted on this growing patient population. Finally, although it is little discussed, there is much individual variability in observed responses. Most studies hav focused on common trends in order to describe the phenomenon of disuse osteopenia, but there may also be essential information in case studies where unusual or uncommon responses are noted.

B. Bone Loss with Immobilization Due to Nonparalytic Medical Conditions A number of clinical conditions involve immobilization either for therapeutic or treatment purpose or beacuse of the condition itself. Examples of treatment-related immobilization include complete bed rest prescribed for patients with prolapsed disc or corrective surgery for scoliosis and cast or brace immobilization of fractures of joint injuries. In most cases, this type of immobilization is temporary, but subsequent activity may be altred, depending on the severity of the condition or fracture. Frozen shoulder syndrome result in severely reduced use of the affected arm, and amputation of a limb, part of limb, or finger result in permanently reduced use of the adjacent anatomy. Table 2 summarizes nonparalytic conditions which involve immobilization either because of the condition or because of the treatment in which bone loss has been studied. 1. METABOLIC EVIDENCE Elevation in serum ionized calcium activity and urinary calcium excretion has been noted in patients with femur and vertebral fractures immobilized up to 4 months [46]. Similar biochemical findings are reported in two studies of patients during therapeutic bed rest for vertebral disc disorders. Van der wiel [47] reports immediate increases in urinary creatinine ratios of calcium and hydroxyproline and incresed serum and phosphate concentrations, but decresed circulating 1,25-dihydroxyvitamin D. Krolner [48] finds increased serum calcium (1.3%) and phosphorus (5.9%) concentrations and increased urinary excretion of creatinine (17%) but not calcium. Thus, in the few studies of bone metabolism under conditions of medical immobilization, urinary and serum changes

TABLE 2

Summary of Nonparalytic Conditions with Associated Osteopenia

Condition

Region studied

Frozen shoulder syndrome

Bilateral comparisons of humeral head

Unilateral amputation

Bilateral comparisons of distal femur (above knee amputation) and metacarpal (digital amputation)

Ligamentous knee injury

Bilateral comparisons of proximal tibia and fibula

Prior tibial fracture

Bilateral comparisons of calcaneus, tibia, patella, femur, spine, and forearm

Vertebral disc prolapse

Lumbar spine

Scoliosis correction

Fourth lumbar vertebra

214

B. JENNY KIRATLI

are generally indicative of elevated bone resorption similar to that seen with spinal cord injury and poliomyelities. No information is available on bone formation indices.

10 and 50% in bilateral comparisons, and some authors have demonstrated an association between duration of condition and the amount of bone reduction while others have found the opposite [49,52,55].

2. CHANGES IN BONE MASS Under conditions which involve unilateral effects, marked osteopenia is usually observed in the immobilized limb compared with the unaffected one. Examples of long-term responses include 50% less bone mineral in the “frozen” shoulder with no evidence of reversal after recovery of function [49], 28% reduction in femoral neck bone mineral [50] and 14% larger medullary cavity with extensive intracortical porosity [51] following above knee amputation, and 13% less metacarpal cortical thickness following digital amputation [52]. Within 12 weeks, tibia bone mineral is reduced by 10% in nonoperated and 18% in operated cases of ligamentous knee injury with no evidence of restitution after 1 year [53]. Lumbar spine bone mineral loss occurs at 2% per week (individual rates range from 1.3 to 4.8%) following scoliosis correction treated by 3 – 6 weeks of therapeutic bed rest [54] and 0.9% per week in patients with prolapsed disc treated with up to 8 weeks of bed rest [48]. Reduced mineral in several bones of the injured extermity after tibial fracture is accompanied by 10 – 12% reduction in lumbar spine bone mineral, compared with controls, presumably related to the 16 – 27 weeks of therapeutic bed rest [55]. Although the study designs and measurement methodologies differ, these studies, considered together, demonstrate several trends. Bone loss is localized to the affected limb or bone and involves both cortical and cancellous bone. Rates of loss differ across these studies by site and measurement technique, but bone loss has been observed as great as 2% per week. Long-term bone loss ranges between

TABLE 3 Treatment

C. Bone Loss with Voluntary Immobilization In the past three decades, studies have been conducted in which healthy young male volunteers are placed under continuous bed rest for up to 9 months. A chronologic summary of published studies is provided in Table 3. The results from these studies are converted to weekly or monthly rates for clarity of comparisons. 1. METABOLIC EVIDENCE Net bone loss following immobilization is thought result from concurrent elevation of bone resorption and depression of bone formation. Studies of bone-specific markers and hormones as well as histomorphometric studies of bone cell activity provide both supporting and equivocal evidence for this explanation. Data indicating bone resorption are consistent across all studies; urinary excretion of calcium increases immediately after initiation of bed rest, reaches a plateau between 4 and 8 weeks, and remains elevated throughout the treatment; this hypercalciuria is accompained by elevated urinary hydroxyproline and phosphorus output [57,59]. Urinary excretion of indicators of collagen breakdown, tartrateresistant acid phosphatase and pyridinium cross-links (pyridinoline and deoxypyridinoline), increases immediately after initiation of bed rest [64], and decreased serum concentrations of parathyroid hormone and 1,25-dihydroxyvitamin D are also observed [63]. Histomorphometric data

Summary of Studies Reporting Bone Responses to Voluntary Bed Rest Duration

Subjects 14 M

Assessment method Biochemistry (urine)

Ref.

Horizontal bed rest

6 Weeks (1.5 months)

56

Horizontal bed rest

30 – 36 Weeks (7.5 – 9 months)

3M

Gamma ray transmission scanning; biochemistry

57

Horizontal bed rest

24 – 30 Weeks (6 – 7.5 months)

5M

Gamma ray transmission scanning; biochemistry

58

Horizontal bed rest

(5 – 36 Weeks) (multiple studies)

90 M

SPA of calcaneus; biochemistry

59

Horizontal bed rest

5 Weeks (1  months)

6M

DPA; biochemistry

60

Horizontal bed rest

17 Weeks (4  months)

6M

SPA and DPA of various sites; calcium balance

61

Bed rest with 5° head-down tilt

16 Weeks (4 months)

8M

Histomorphometry of trans-illiac bone biopsy

62

Bed rest with 6° head-down tilt

1 Week

8M

Histomorphometry of trans-illiac bone biopsy; biochemistry

63

Bed rest with 6° head-down tilt

1 Week

8M

Biochemistry

64

215

CHAPTER 46 Immobilization Osteopenia

obtained after 4 months of bed rest show a reduction in the number of iliac crest trabeculae with an indication of thickening of the remaining trabeculae (attributed to compensatory architectural adaptation). There is little evidence of other changes in bone mass nor any significant effect on resorption parameters [62]. The evidence regarding bone formation is less consistent, but has also been less studied; observations from the same study demonstrate increased serum osteoclacin concentrations, a specific marker of osteoblast activity, but no change in skeletal alkaline phosphatase activity [64]. Decreased bone formation and mineral apposition rates were seen on iliac creast bone biopsies [63]. Total body calcium loss estimated from balance studies is shown to be approximately 0.1% per week [60,61] and appears to continue throughout the period of immobilization up to 36 weeks [58]. A similar profile of loss is found when total body calcium is assessed by use of dual energy X-ray absorptiometry (DXA), and these result are strongly correlated with data from balance studies (r 0.88) [61]. This close association suggests that total body densitometry might be substituted for the more complex and expensive calcium balance studies in long-term studies of immobilization and/or spaceflight. One notable finding is the wide variability in individual responses; “higher losses than average in a few personnel will need to be anticipated.” As with paralysis-induced osteopenia, it may be instructive to better understand the reasons underlying tis individual variability in response. 2. CHANGES IN BONE MASS Neither systemic measures of bone metabolism nor total body data fully explain bone mineral responses. Characteristic differences in regional responses may reflect both the degree and the duration of unloading (with bed rest), as well as the habitual loading conditions. The greatest bone loss is observed in the calcaneus, with measured changes ranging from 0.6 to 1.8% per week [57,58,61]. Moderate losses are seen in the tibia, femur, and spine (0.13 – 0.27% per week), although, not surprisingly, no change is observed in the spine after only 1 week of bed rest [60]. No change is reported in the radius with many months of bed rest [58,61], and a slight increase is noted in head BMC with head-down tilt [61]. The potential long-term effect of continued immobilization can be predicate by projection to annual changes. Thus, the total yearly reduction in hip and spine would be slightly over 10%, the calcaneal loss might exceed 50% in 1 year, and total body mineral loss would approximate 8%. Examination of repeated measurements taken over 17 weeks indicates that bone loss continues to be linear over that period [61]. Evidence from paralyzed patients suggests that the rate of bone loss slows after 6 – 8 months; however, as there are no bed rest studies with duration longer than 9 months, no data available to support or refute a similar change in loss pattern.

The differential pattern of bone response through the body is thought to reflect different magnitudes of change in mechanical loading from normal mobility to bed rest. With pre-bed-rest activity and ambulation, the highest loads occur in the heel with relatively lesser loads through the leg and axial body. Thus, bed rest results in excessive loss from the calcaneus and moderate losses from the legs and spine, relatedd to the amount of reduction in habitual load. Morever, the imposition of bed rest may not greatly alter the loading patterns of the arms, especially as movement of the upper extremity does not appear to be particularly restricted; this is supported by an absence of observed change in arm bone mass. The increases in head bone mineral in studies with a head-down tilt are less easily explained with this approach, although the suggestion is offered that stresses in the head may be increased under the bed rest condition with sitting up to eat or read [61]. Another potential explanation is that this non-weight-bearing region may serve as an internal “sink” for calcium an dphosphorus released during recumbency. Incresed mineral mass has also been attributed to a fluid shift and altered vascular perfusion with the head-down orientation. Although similar patterns are apparent in comparisons of the effects of paralysis, therapeutic bed rest, and prolonged voluntary bed rest, the magnitude of effect is sonsistently greater wherever disuse is more extreme. For example, even in continuous voluntary bed rest, the individual retains normal muscle function and the ability to move at will; weight bearing is effectively eliminated while muscular function is only partially reduced (greatly dependent on the compliance of the volunteer). Thus the greatest effect is observed in the calcaneus, and relatively less bone atrophy is seen elsewhere in the body. Conversely, the spinal-cord-injured individual has no intact neural input to muscle or bone, thus eliminating the influences of neuromuscular function as well as direct neural input, and regional losses are much larger than those observed with voluntary bed rest. Finally, the patient who is placed at therapeutic bedrest for a vertebral disorder experiences great reduction in both muscle activity and weight bearing infulences in the vertebral region (although voluntary motor function remains), and rapid bone loss occurs locally in the spine.

IV. RELEVANT SEQUELAE IN SPINAL CORD INJURY PATIENTS A. Heterotopic Ossification Heterotopic ossification (HO), or ectopic bone formation, is a commonfinding associated with spinal cord injury, occurring most commonly around the hip and knee joints as well as in the distal femur and elbow and shoulder joints.

216

B. JENNY KIRATLI

FIGURE 1

Bilateral heterotopic ossification of proximal femur joints in a 61-year-old male, with complete quadriplegia

for 7 years.

Figure 1 demonstrates bilateral HO in an individual with chronic paraplegia. HO is also commonly found in patients with head injury and after total joint arthroplasty, but there appear to be differences in type and distribution of bone formation in the three patient groups [43]. The etiology is not well understood, but evidence suggests that local trauma and subsequent hemorrhage may contribute as well as immobility, and HO is often found as a sequel to deep vein thrombosis [65 – 67]. Early evidence of HO includes elevated serum alkaline phosphatase activity and local edema preceeding positive findings on Technetium-99 bone scan and radiography. HO has been reported to occur in 10 – 53% of patients with acute spinal cord injury [68,69]; the wide range arises from variable diagnostic methods. If untreated, HO can progress to ankylosis resulting in impaired (or absent) joint movement and significant reduction in the ability of an individual to function normally. Advanced HO is usually treated with high dose radiation and/or surgical resection, but there is a high prevalence of recurrence [43,70,71]. Disodium etidronate has been found by some to be effective in acute spinal cord injury for prevention or reduction in severity of HO, but the effect seems to require continued etidronate therapy, and symptoms recur with discontinuation of the drug [72,73]. A number of patients (nonresponders) show no benefit. Etidronate ad-

ministration is generally recommended for periods of only 3 months; long-term administration has been associated with defective mineralization and elevated risk of fracture in patients with paget’s disease [74 – 76]. Thus, the potentoal beneficial effects (against HO) must be weighed against the likelihood of future fractures. Newer bisphosphonates may prove to be more effective.

B. Pathological Fracture 1. INCIDENCE AND OUTCOMES Spinal-cord-injured individuals are now living longer and may, with good medical management, expect a normal life expectancy. However, with longer life spans, a commensurate increase in fracture occurrence may be expected, presumably due to a combination of bone atrophy from disuse and normal age-related bone loss, as well as increased activity levels and greater exposure to potential trauma. Long bone fractures occur in individuals with chronic spinal cord injury, often without any discernible trauma, such as during a transfer or during range-of-motion exercises. In the absence of pain sensation, spinal cord injury individuals are often unaware that a fracture has occurred, and thus do not seek medical attention until more extensive

217

CHAPTER 46 Immobilization Osteopenia

FIGURE 2

Subcapital fracture of the right proximal femur in a 43-year-old male with complete paraplegia at the T1011 level for 22 years. Note that the disarticulated femoral shaft has migrated superiorly. The patient was unaware that the fracture had occurred and no treatment was offered.

and serious symptoms (such as increased spasticity, fever, and autonomic dysreflexia) are present. Figures 2 – 4 depict common fractures occurring in spinal cord injured persons. The majority of these fractures occur in the lower extremity, with the highest prevalence in the distal femur; the term “paraplegic fracture” is widely used to describe supracondylar femoral fractures (the most common type) in these patients. Lower extremity fracture occurrence has been estimated to approximate 6% in chronic spinal cord injury [4,5,77 – 79]. Table 4 summarizes data from studies of lower extremity fractures in spinal cord injury patients. However, several of the studies do not distinguish between fractures which occur at the time of the initial injury and those which occur later; fractures which occur concurrent with a traumatic spinal cord injury can hardly be considered “pathologic.” Furthermore, although all of these studies report “incidence,” most actually present prevalence data. The specific location or morphologic type of these fractures is infrequently reported, although such date may be

essential for proper evaluation. For instance, complications, such as hermatoma and hypercoagulability, and subsequent healing depend on the amount of blood flow to a region; as cancellous bone has greater perfusion than cortical, certain types of fracture might involve greater or lesser blood loss. Therefore, while excessive bleeding might be a frequent complication, healing might also be accelerated where more cancellous bone is involved. In addition to medical considerations, there are notable economic and social consequences of sustaining a fracture. Spinal cord injury patients usually require at least [1– 2] weeks of in-hospital care, even without surgical treatment. Loss of income may result during hospitalization and/or due to restricted activity and function. Self-care may be greatly affected by the need to wear a brace or otherwise immobilize a limb such that additional attendant care is required — at additional cost. Finally, overall quality of life may be reduced by limitations in activities of daily living (due to balance problems, difficulty with transfers, etc.) and altered body image.

218 FIGURE 3

Spiral fracture of the distal femur in a 46-year-old male, with complete paraplegia at the T2-4 level for 20 years. This fracture occurred during daily range of motion activity and was treated acutely with a hinged brace. (a) Note the extreme cortical thinning. (b) Healing has begun 4 months later. Note exuberant callus formation. (c) Healed fracture 1 year later. Little evidence of change is noted on later films obtained 4 years postfracture.

219

CHAPTER 46 Immobilization Osteopenia

FIGURE 4

(a) Comminuted fracture of the distal femur in a 55-year-old female with incomplete paraplegia at the T11 – 12 level. Fracture occurred 10 years post-spinal-cord-injury and was acutely treated with a leg brace. (b) Little evidence of healing is apparent 6 months later. Note inferior placement of femoral shaft and poor alignment of fragments. This fracture presents as a nonunion and pseudoarthrosis 3 years after the fracture occurrence.

There is generally perception that long bonb fractures subsequent to spinal cord injury are becoming more common and that the management of these fractures will be increasingly important in rehabilitation medicine of the future. Yet, there are no estimates of current incidence nor clear prescriptions for optimal management. These fractures often occur without evidence of trauma and go undiagnosed (acutely) due to absent sensation and lack of awareness of the fracture event. If untreated, they may lead to complications including hemorrhage and infection, or, if healed without proper management, may result in deformity, loss of motion, and impaired function. 2. MANAGEMENT Treatment regimens are varied and based mostly on the experience of the treating physicians rather than consensus protocols. In general, conservative, nonoperastive treatments are followed, such as pillow splinting or external

bracing; casts should only be used if extremely well-padded and frequently removed for skin inspection [4,78,80]. Surgical interventions are usually contraindicated because of preexisiting osteopenia (unsuitable for internal fixation of orthopedic harware), skin fragility and excessive spasticity (risk of decubitus ulcers with casting), and elevated susceptibility to infection (contraindication for open reduction), as well as heightened risk of further complications including continuous drainage, edema, hematoma, secondary hemorrhage, and refracture. Healing is often reported as rapid involving as “exuberant callus,” and nonunion is relatively rare [77 – 79]. Assessment is made of the patient’s prefracture functional abilities, and treatment is directed to maximize return to equivalent function. However, one orthopedic surgeon with wide experience treating fractures in in spinal cord injury patients reports the contrary for almost all of these statements [81,82]. Specifically, he describes slowed healing

TABLE 4 Reference no. and year of study

Frx (Pts)/ total ptsa

Summary of Data from Previous Studies of Fractures in Spinal Cord Injury Patients

Frx

Femoral neck

Inter/ subtrochanteric

Femoral shaft

Supracondylar

Tibial plateau

Proximal shaft

Tibial midshaft

Distal shaft

Other leg

119 (81)/1363

9b

12

—

7

40

—

8

2

24

26

78, 1965b

45 (23)/264

17b

9

—

7

10

—

7

5

7

—

79, 1981

27 (18)/546

5

4

8

4

5

—

2

1

3

—

5, 1981

33 (23)/578

6

1

2

10

11

3

—

6

—

—

77, 1989

33 (25)/526

6

6

—

5

10

3

—

8

—

1

257 (170)

—

12.5%

3.9%

12.8%

29.6%

2.3%

6.6%

8.6%

13.2%

10.5%

4, 1962

220

Compiled fractures a b

The number of fractures (frx) and patients (pts) per total number of patients surveyed is presented. No distinction between fractures which occur concurrently with the initial injury (acute spinal cord injury) and those that occur later (chronic spinal cord injury).

221

CHAPTER 46 Immobilization Osteopenia

and common malunion and suggests the necessity for open reduction and internal fixation for successful management of fractures in these patients. Prolonged immobilization subsequent to the fracture may present additional risks, resulting in postural deformities and joint contractures as well as general reduction in cardiovascular fitness and possibly altered bone metabolism leading to accentuated osteopenia. Comlications can include malalignment and spontaneous arthrodesis of adjacent joints, while some patients exhibit a paucity of bony deposition and tend not to heal. Nontheless, healing is ususally adequate. Little is known about the reasons for there different responses, nor is it known how often fractures in these patients mights heal spontaneously in satisfactory positions. Finally, the quesion remains whether all fractures in the spinal cord injury patient require treatment. For instance, an untreated fracture which results in a fibrous union may produce motion (as a pseudoarthrosis) near an immobile ankylosed joint, enabling activities such as sitting which were not previously possible. Moreover, treatment of a subcapital femoral fracture mayneedlessly expose a patient to surgical complications. Untreated fractures of this type often resolve naturally in a noncomplicated, flail joint where the femoral neck completely resorbs, the head remains in the acetabulm, and the shaft is separate (effectively, a “Girdlestone Joint” ) with no impairment in function. However, treatment of other fracture types is necessary for maintenance of functional joints and to prevent deformities which may interfere with activities of daily living or lead to pressure sores. 3. CAUSES AND PREVENTION Characteristics which appear to be associated with elevated fracture risk include: completeness of injury (presence/absence of sensation and proprioception and motor control), flaccidity, and greater duation since injury. Although osteopenia is frequently referred to as a relevant risk factor, few data are available which document this association [83,84]. There is a general awareness among rehabilitation physicians, therapists,and nurses regarding heightened fracture risk with long-standing paralysis, and broad sanctions are common against use of standing frames or initiation of strenuous activity in persons with chronic spinal cord injury. However, there are few data to support such recommendations, and fractures often occur during prescribed activities such as range-of-motion exercises and transfers rather than during proscribed activities. Mechanical failure (fracture) results when local stresses exceed the ultimate strenght of bone in a given region, and the loads generated by daily activities are equally important as bone material quality (see also Chapter 19 ). Poor bone quality may increase risk and predispose spinal cord injury persons to

fractures, but in most cases, the activity is the proximate cause of the fracture (i.e., applied load greater than bone strenght). Thus prevention may rely on improved identification of “risky” behaviors and activities that frequently result in fractures. By targeting these activities and developing safer alternatives, it may be possible to reduce fracture occurrence.

V. INTERVENTION ATTEMPTS A. General Considerations Reversal or prevention of immobilization-related bone atrophy has been attempted by several means. These include simulation or resumption of weight bearing, stimulation of muscular contractions, and use of pharmaceutical agents. The former two approaches are clearly based on the assumtion that the predominant factor underlying bone loss (directly or indirectly) is the elimination of “normal” mechanical loading through standing, walking and mucular activity. Reintroduction of loads equivalent to those lost should therefore positively influence bone metabolism. Practical difficulties in these approaches include the inability of researchers and practitioners to duplicate physiologic loading patterns and the poor or limited compliance of immobilized patients. Usually in these studies, one motor action or set of actions is isolated and evaluated in a controlled manner (e.g., functional electrical stimulation of quadriceps and hamstrings without weight bearing and without other mucsular agonists and antagonists). This provides the advantage of examining single effects but carries the disadvantage of reduced effectiveness due to excluded effects. Often the intensity and/or magnitude of loads are well below physiologic levels. Further, in both clinical and experimental studies, the intervention is usually applied in defined bouts (15 to 60-min sessions) on a regular schedule (3– 5 times per week) for a specified period (3– 8 months); yet, although clearly defined for experimental purposes, these applications may well be inadequate in terms of both daily and long-term frequency. Pharmacologic intervention is unrelated to the cause of immobilization bone loss but rather is based on known bone metaboloc responses to immobilization (e.g., bisphosphonate used to inhibit bone resorption). The use of pharmacologic therapy does not attempt to “normalize” the situation but to correct it. One potential difficulty with a pharmacologic countermeasure is the unbalancing of metabolic homeostasis. Bone metabolic markers and hormones appear to return to normal levels and normal turnover is reestablished within 1 year of injury, with diminished bone mass, and within 2– 4 years, there appears to be little further bone loss.

222

B. JENNY KIRATLI

However, the long-term administration of an anti-resorptive agent (i.e., bisphosphonate) under conditions where bone resorption is not elevated will not restore bone mass and may lead to defective mineralization. Another undetermined parameter is the length of treatment. The most effective time period for pharmacologic intervention would seem to be the acute period of immobilization. Further, the long-term effectiveness of such an intervention is unclear. If it were possible to retard or prevent initial bone loss by drug therapy, would it be necessary to continue the therapy indefinitely, or would “normal” bone metabolism be restored with time even in the absense of “normal” activity and mobilization? Moreover, will bone loss occur (or recur) with cessation of treatment? Finally, patient compliance is an issue which must be considered. Many patients are unwilling to adhere to a regular pill or injection schedule for potential prevention of a nonobvious, non-life-threatening condition.

B. Weightbearing A variety of physical treatments have been proposed for reversal or prevention of bone loss in healthy volunteers undergoing voluntary immobilization. In a 1949 study, Whedon and coworkers [85] reported decreased hypercalciuria in subjects immobilized in body casts who underwent 8– 21 h per day of oscillation from horizontal to 20° foot-down tilt. Nearly two decades later, Issekutz et. al. [56] observed reduction in hypercalciuria in bed rest subjects who stood quietly for 3 h per day but no effects in those who performed supine cycling (4 h/day) or sat (up to 8 h per day). Similarly, no significant effects on calcium metabolism or calcaneal bone mass were detected in bed rest volunteers subjected to the following treatments: passive leg exercise with a pulley, static or intermittent longitudinal compression up to 80% body weight, lower body negative pressure, and impact loading to the foot (consisting of 20-and 36-lb. loads delivered 40 times per minute, for 6 and 8 h per day, respectively) [59,86]. In several other bed rest studies, however, reambulation led to return of normal calcium metabolism and restitution of calcaneal bone mass [57,58,61]. Early weight bearing and assisted ambulation are recommended in most rehabilitation setting following immobilization injuries. These activities have much therapeutic value on circulation, urinary function, spasticity, and body image, but few data support the role of weight bearing for prevention of bone loss. It is of interest that almost 50 years ago Abramson [1] made the following observations: 1. Bone atrophy occurring in paraplegia is of an osteoporotic nature. 2. There is close association if urinary calculi and soft-

tissue ossifications with the excessive calcium mobilization from the limbs. 3. If calcium could be held in the bone, urinary calculi, soft-tissue ossifications, and pathological fractures could be prevented. 4. Pressure produces bone matrix, thus allowing the deposition of calcim. 5. The only logical therapeutic agent which applies pressure intermittently is ambulation. Attempts to implement this thesis have involved therapeutic applications of compressive or vertical loading through crutch ambulation, underwater exercise, and rocking bed but have not yielded changes in calcium metabolism in individuals with paralytic poliomyelitis or spinal cord injury [6,15,87,88]. However, reduction in hypercalciuria and positive calcium balance have been demonstrated with assisted ambulation and tilt table treatments in acutelty injured patients [89,90]; greater effects are observed in those injured less than 3 months. One caveat to these results should be noted: hypercalciuria resolves naturally within months after injury, and it may be difficult to distinguish the effects of ambulation from the natural process. Claus-Walker and associates report reduction in hydroxyproline, but not calcium, excretion in quadriplegics who are less recumbent and more active in rehabilitation exercises; however, these activities are not enumerated or quantified in this paper [22]. Finally, passive loading through sitting, tilt-table treatments, and standing in a standing frame have shown no effects on calcium metabolism or bone mass in individuals with chronic spinal cord injury [15,91].

C. Muscular Activity Functional electrical stimulation [FES] is a method of eliciting active muscle contractions from paralyzed muscle. Reintroduction of muscular activity by this technique has been examined for effects on cardiopulmonary, endocrine, and musculoskeletal systems in spinal-cord-injury subjects. Much of that literature is outside the scope of this discusion, but there are clear indications for improved fitness and enhhanced muscle strengh and endurance with FES treatment. With evidence of reversal of muscle atrophy by this technique comes the supposition that bone may be similarly affected. Skeletal assessments have been made of lumbar spine, hip, and proximal and distal tibia by multiple densitometric techniques in individuals with acute and chronic spinal cord injury participating in regular FES programs [38,92 – 95]. Treatment is usually provided in three to five sessions per week for approximately 3 months, but one study lasted for 7 months and one for over 10 months. Most of these trials have involved participants injured longer than 1 year previously although several include subjects

223

CHAPTER 46 Immobilization Osteopenia

with acute injury. However, while muscle parameters may be consistently improved with FES, no study has been able to clearly document a change in bone mass. In one recent study, the authors suggest that although no increase in bone density was observed, they may have reduced the bone loss by their treatment [38]. This is an important and valid distiniction and should be considered in any intervention studies involving acutely injured patients. However, the study involved a mixture of patients with acute and chronic injuries, and the authors’ method of predicting “expected loss” was original but somewhat problematic. Cross-sectional data and regression analysis were used to project bone loss from 0.15 months through 22 years postinjury. Actual study duration was adjusted per subject so that the total number of sessions (36) were achieved, regardless of how many weeks this involved. Therefore, in order to “normalize” results to account for this time discrepancy, the authors projected all data to a yearly rate. Annualized results for each subject were then compared with his/her expected rate of loss. Incorrect prediction of “expected loss” and/or augmentation of predicted change with annualized data would certainly have influenced these results. A natural occurrence of residual muscular activity in the spinal cord injury individual is spasticity, or involuntary muscle contractions below the lesion (in uppermotor neuron injuries). These spasms can be quite intense and can occur frequtnly; muscle mass is often visibly larger in paralyzed persons who are spastic than in those who are flaccid. Based on these observations, spasticity is comonly assumed to prevent bone loss, but there is little evidence to support this conjecture [31,39], although flaccid patients predominate among spinal cord injury persons who have sustained long bone fracture [5].

D. Pharmacologic Treatments Attempts to reverse bone loss have been made by use of pharmacologic agents, including calcitonin, bisphosphonates, and oral phosphate and have involved both spinal cord injured patients and bed rest subjects. While variable effects on bone turnover markers have been shown, there is little clear evidence of reversal of bone loss with any of these agents. Calcitonin used alone had some efficacy in reducing hypercalcemia in acute spinal cord injury, but this effect appears to be transient, and better results are achieved when it is combined with etidronate or prednisone [96 – 99]. Calcitonin alone has no effect on reducing negative calcium or phosphorus balance in bed rest subjects, while administration of oral phosphate or calcium plus phosphorus has been found to prevent hypercalciuria and result in a less negative calcium balance [58,59,86]. Low-dose etidronate (5

mg/kg/day) used alone has little effect on urinary excretion of calcium or hydroxyproline or on negative calcium or phosphorus balance, but high-dose etidronate (20 mg/kg/day) appears to reduce hyperhydroxyprolinuria, cause positive shifts in mineral balances, and suppress elevated turnover in bed rest subjects [59,100]. While the effects of etidronate administered in spinal cord injury patients for prevention of heterotopic ossification have been reviewed above, there are no reports of etidronate used alone for prevention of bone metabolic changes or bone loss. However, two other bisphosphonates, clodronate and tiludronate, have been studied. Encouraging results were observed with administration of clodronate in acute spinal cord injury, including reduced urinary excretion of calcium and hydroxyproline, reduction in number of osteoclasts, and prevention of loss of tibial bone mineral [99]. A similar reduction in osteoclast number is found with use of tiludronate in spinal cord injury patients, but there is no evidence of effect on bone volume, osteoid parameters, or eroded surfaces. Furthermore, cancellous mineral apposition rate is normal with low-dose tiludronate but reduced in both high-dose and placebo groups [101].

E. Recovery and Reversibility A final consideration is whether bone loss can be reversed with restoration of ambulation and normal activity following temporary immobilization. In healthy individuals subjected to voluntary bed rest and patients undergoing therapeutic immobilization, reambulation is accompanied by return of normal mineral metabolism. No data are available concerning histomorphometric changes with recovery of normal loading. The data regarding changes in bone mass are somewhat inconsistent, but generally indicate that full restitution of normal bone mass may not occur [102]. Increases are observed in some, but not all, studies, and rates of change differ by site and individual. In most studies where bone gain is observed, the rate of increase is much slower than the prior loss rate. Some studies show that recovery may require many years [48,54,61,102] while other studies report no evidence of recovery in later years even after patients have returned to previous activity levels [49,53,54]. One explanation for the lack of observed recovery is that bone loss has resulted in elimination of trabeculae which severely limits available surfaces for new bone formation. Another consideration is that some individuals may not actually return to their previous activity levels and thus may not provide sufficient stimulus for bone accretion. In some cases, reversal of bone loss may not be feasible even if it is potentially possible; that is, patients may be unwilling to commit to the exercise regimes necessary to promote bone formation.

224

VI. CLINICAL RELEVANCE Reduction or elimination of mechanical loading has a clear impact on the skeleton. The effects are regional, related to the magnitude of unloading (i.e., partial paralysis has less of an effect than complete), and, at least in some circumstances, reversible. There is also overwhelming evidence of individual variability of responses, as is commonly the case with human conditions. Factors such as prior habitual activity patterns and initial absolute bone mass may determine the level of the response to acute immobilization. Would an individual who was more physically active or who has greater initial bone mass prior to immobilization lose bone more rapidly? This might be expected if skeletal maintenance is based on habitual mechanical usage (i.e., activity patterns). However, as investigational studies of disuse osteopenia are limited by available patient or subject populations, these factors have not yet been assessed. Other factors, such as genetic predisposition and lifestyle habits must also play a role in any observed bone response, but quantification of these is even more problematic. A more complete understanding of the contributions of hormonal, neural, and other causal influences will be necessary before effective treatment will be possible. Although it is beyond the scope of this chapter to discuss underlying mechanisms, nonetheless there is much evidence that unloading has a direct or indirect effect on bone metabolism, resulting in net bone loss. It is further clear that the main effect on skeletal homeostasis occurs immediately but is soon resolved at a new, lower steady state. Although rate and absolute change appear to be individually variable, there also appears to be a less variable later response with long-duration immobilization (paralysis). For example, the mean bone mass of the femoral neck following chronic spinal cord injury is approximately 65 – 75% of (ambulatory) normative values, and there is heightened fracture risk below this level. In fact, a general “fracture threshold” (for age-related osteoporosis) has been suggested to be in this same range. Relatively few spinal cord injury individuals have femoral neck bone mass greater than 75% of expected, and the variance around the mean is quite small; this observation is consistent across a number of studies. This suggests the possibility of a physiological threshold, maintained by basal hormonal activity and cellular processes, that is reached in the absence of mechanical stimuli. If so, identification of the regulatory controls might be applicable clinically for maintenance of normal bone mass and prevention of disuse osteopenia. Further, there are occasional spinal cord injury individuals with extremly low bone mass (sometimes unmeasureable as mineralized tissue by densitometry). Investigation of why these people do not maintain bone at this physiological level might also contribute informa-

B. JENNY KIRATLI

tion about mechanisms of bone loss. A recurrent pattern in the studies reviewed here is that of individual variability. Most of the published data are expressed as rates of change. However, if the magnitude of reduction in loading is important and if there is a lower threshold of bone mass which is achieved with elimination of mechanical stimuli, different rates of changes might be expected as each individual adjusts to specific conditions of disuse, relative to his or her habitual activity and previous bone mass. The clinical outcomes of immobilization osteopenia alone are generally not serious. However, the sequelae such as unhealed fracture, extensive renal calculi, and excessive heterotopic ossification can be devasting to health and/or quality of life. Reversal of later bone atrophy, once a new steady state has been reached, is probably not possible, and thus the most feasible approach will be to attempt early prevention. Furthermore, a more accurate estimation of “safe” versus “risky” bone mass levels for particular conditions may be useful for influencing therapeutic management of patients. That is, if individuals with extremely low bone mass are identified, they may be directed toward activities that produce lower stresses and therefore reduce fracture risk. Immobilization osteopenia is considered currently as a condition of “accelerated aging” associated with spinal cord injury; changes in skin integrity and endocrine function are other elements of this phenomenon. Conversely, knowledge of bone response with immobilization may be relevant in treatment or prevention of age-related osteoporosis, as a component of bone loss in many aging individuals may be attributable to increased inactivity. Evidence from immobilization studies may clarify the separate effects of different types of mechanical loading. For instance, weight bearing appears to exert a greater influence on bone response than muscular activity; this may be relevant for therapeutic planning in that expectations may differ for strength versus mobility training. Finally, elucidation of the mechanisms (cellular, electrical, hormonal, etc.) underlying immobilization osteopenia will certainly contribute to an enhanced understanding of the causes and consequences of other forms of osteoporosis.

References 1. A. Abramson, Bone disturbances in injuries to the spinal cord and cauda equina (paraplegia). J. Bone Jt. Surg. 30A, 982 – 987 (1948). 2. L. Guttmann, “Spinal Cord Injuries: Comprehensive Management and Research” Blackwell Scientific, England, 1976. 3. D. Yashon, “Spinal Injury.” Appleton – Century – Crofts, New York, 1978. 4. A. E. Comarr and R. H. Hutchinson, Extremity fractures of patients with spinal cord injuries. Am. J. Surg. 103, 732 – 739 (1962). 5. K. Ragnarsson and G. Sell, Lower extremity fractures after spinal cord injury: A retrospective study. Arch. Phys. Med. Rehabil. 62, 418 – 423 (1981).

CHAPTER 46 Immobilization Osteopenia 6. G. D. Whedon and E. Shorr, Metabolic studies in paralytic acute anterior poliomyelitis. II. Alterations in calcium and phosphorus metabolism. J. Clin. Invest. 36, 966 – 981 (1957). 7. C. Plato, J. Wood, and A. Norris, Bilateral asymmetry in bone measurement of the hand and lateral hand dominance. Am. J. Phys. Anthropol. 52, 27 – 31 (1980). 8. C. Ruff and H. Jones, Bilateral asymmetry in cortical bone of the humerus and tibia: Sex and age factors. Hum. Biol. 53, 69 – 86 (1981). 9. J. Balseiro, F. Fahey, H. Ziessman, and T. Le, Comparison of bone mineral density in both hips. Radiology 167, 151 – 153 (1988). 10. H. Lessig, M. Meltzer, and J. Siegel, The symmetry of hip bone mineral: A dual photon absorptiometry approach. J. Nucl. Med. 12, 811 – 812 (1988). 11. B. Nilsson and N. Westlin, Bone density in athletes. Clin. Orthop. 77, 179 – 182 (1971). 12. P. Bergmann, A. Heilporn, A. Schoutens, J. Paternot, and A. Tricot, Longitudinal study of calcium and bone metabolism in paraplegic patients. Paraplegia 15, 147 – 159 (1978). 13. P. Minaire, P. Meunier, C. Edouard, J. Bernard, P. Courpron, and J. Bourret, Quantitative histological data on disuse osteoporosis: Comparison with biological data. Calcif. Tissue Int. 17, 57 – 73 (1974). 14. G. Pilchonery, P. Minaire, J. J. Milan, and A. Revol, Urinary elimination of glycosaminoglycans during the immobilization osteoporosis of spinal cord injury patients. Clin. Orthop. 174, 230 – 235 (1983). 15. F. M. Maynard and K. Imai, Immobilization hypercalcemia in spinal cord injury. Arch. Phys. Med. Rehabil. 58, 16 – 24 (1977). 16. F. M. Maynard and K. Imai, Immobilization hypercalcemia following spinal cord injury. Arch. Phys. Med. Rehabil. 67, 41 – 44 (1986). 17. J. Tori and L. Hill, Hypercalcemia in children with spinal cord injury. Arch. Phys. Med. Rehabil. 59, 443 – 447 (1978). 18. A. Wolf, R. Chuinard, R. Riggins, R. Walter, and T. Depner, Immobilization hypercalcemia: A case report and review of the literature. Clin. Orthop. 118, 124 – 129 (1976). 19. F. Steinberg, S. Birge, and N. Cooke, Hypercalcemia in adolescent tetraplegic patients: Case report and review. Paraplegia 16, 60 – 67 (1978 – 1979). 20. A. Chantraine, Clinical investigation of bone metabolism in spinal cord lesions. Paraplegia 8, 253 – 259 (1971). 21. J. Singh, N. DiFerrante, M. Strashun, and F. Gyorkey, Urinary excretion of glycosaminoglycans in quadriplegia. Proc. Soc. Exp. Biol. Med. 156, 488 – 490 (1977). 22. J. Claus-Walker, W. A. Spencer, R. E. Carter, L. S. Halstead, R. H. Meier, and R. J. Campos, Bone metabolism in quadriplegia: Dissociation between calciuria and hydroxyprolinuria. Arch. Phys. Med. Rehabil. 56, 327 – 332 (1975). 23. L. Klein, “The Relationship between Urinary Hydroxyproline, Calcium and Serum Alkaline Phosphatase during Fracture Healing and Acute Disuse Osteoporosis.” The Fourth European Symposium on Calcified Tissues, Leiden, (1966). 24. A. Ohry, Y. Shemesh, R. Zak, and M. Herzberg, Zinc and osteoporosis in patients with spinal cord injury. Paraplegia 18, 174 – 180 (1980). 25. A. Chantraine, B. Nusgens, and C. M. Lapiere, Bone remodeling during the development of osteoporosis in paraplegia. Calcif. Tissue Int. 38, 323 – 327 (1986). 26. G. Vose and D. Keele, Hypokinesia of bedfastness and its relationship to x-ray determined skeletal density. Tex. Rep. Biol. Med. 28, 123 – 130 (1970). 27. H. J. Griffiths, B. Bushueff, and R. E. Zimmerman, Investigation of the loss of bone mineral in patients with spinal cord injury. Paraplegia 14, 207 – 212 (1976).

225 28. D. A. Hancock, G. W. Reed, P. J. Atkinson, J. B. Cook and P. H. Smith, Bone and soft tissue changes in paraplegic patients. Paraplegia 17, 267 – 271 (1980). 29. L. Lussier, J. Knight, G. Bell, T. Lohman, and A. Morris, Body composition comparison in two elite female wheelchair athletes. Paraplegia 21, 16 – 22 (1983). 30. R. Yagan, M. Radivoyevitch, and M. Khan, Double cortical line in the acetabular roof: A sign of disuse osteoporosis. Radiology 165, 171 – 175 (1987). 31. R. Biering-Sorensen and H. Bohr, Bone mineral content of the lumbar spine and lower extremities years after spinal cord lesion. Paraplegia 26, 293 – 301 (1988). 32. B. J. Kiratli, “Skeletal Adaptation to Disuse: Longitudinal and Cross-Sectional Study of the Response of the Femur and Spine to Immobilization (Paralysis).” Ph.D. thesis, University of Wisconsin — Madison, 1989. 33. F. Biering-Sorensen, H. Bohr, and O. Schaadt, Longitudinal study of bone mineral content in the lumbar spine, the forearm and the lower extremities after spinal cord injury. Eur. J. Clin. Invest. 20, 330 – 335 (1990). 34. D. Garland, C. Stewart, R. Adkins, S. Hu, C. Rosen, F. Liotta, and D. Weinstein, Osteoporosis after spinal cord injury. J. Orthop. Res. 10, 371 – 378 (1992). 35. R. Saltzstein, S. Hardin, and J. Hastings, Osteoporosis in spinal cord injury: Using an index of mobility and its relationship to bone density. J. Am. Paraplegia Soc. 15, 232 – 234 (1992). 36. V. Finsen, B. Indredavik, and K. Fougner, Bone mineral and hormone status in paraplegics. Paraplegia 30, 343 – 347 (1992). 37. W. Leslie and P. Nance, Dissociated hip and spine demineralization: A specific finding in spinal cord injury. Arch. Phys. Med. Rehabil. 74, 960 – 964 (1993). 38. T. Hangartner, M. Rodgers, R. Glaser, and P. Barre, Tibial bone density loss in spinal cord injured patients: Effects of FES exercise. J. Rehabil. Res. Dev. 31, 50 – 61 (1994). 39. N. Panin, W. J. Gorday, and B. J. Paul, Osteoporosis in hemiplegia. Stroke 2, 41 – 47 (1971). 40. R. L. Prince, R. I. Price, and S. Ho, Forearm bone loss in hemiplegia: A model for the study of immobilization osteoporosis. J. Bone Miner, Res. 3, 305 – 310 (1988). 41. E. Iversen, C. Hassager, and C. Christiansen, The effect of hemiplegia on bone mass and soft tissue body composition. Acta Neurol. Scand. 79, 155 – 159 (1989). 42. R. Hamdy, G. Krishnaswamy, V. Cancellaro, K. Whalen, and L. Harvill, Changes in bone mineral content and density after stroke. Am. J. Phys. Med. Rehabil. 4, 188 – 191 (1993). 43. D. Garland, A. Clinical perspective on common forms of acquired heterotopic ossification. Clin. Orthop. 263, 13 – 26 (1991). 44. B. Kiratli, A. Smith , J. Chen , and I. Perkash , Functional ability affects bone and muscle mass after spinal cord injury . J. Bone Miner. Res. 10 (Supp1.1), S361 (1995). [Abstract] 45. V. Nikolic, P. Vladovia, D. Sajko, B. Zimmermann, M. Hudec, A. Vladovic, J. Hancevic, and Z. Mijatovic, Bone mass and the safety factor of bone strength in lower extremities of patients with paraplegia. Calcif. Tissue Res. 22 (Suppl.), 303 – 306 (1977). 46. H. Heath, J. Earll, M. Schaaf, J. Piechocki, and T. Li, Serum ionized calcium during bed rest in fracture patients and normal men. Metabolism 21, 633 – 640 (1972). 47. H. van der Wiel, P. Lips, J. Nauta, J. Netelenbos, and G. Hazenberg, Biochemical parameters of bone turnover during ten days of bed rest and subsequent mobilization. Bone Miner. 13, 123 – 129 (1991). 48. B. Krolner and B. Toft, Vertebral bone loss: An unheeded side effect of therapeutic bed rest. Clin. Sci. 64, 537 – 540 (1983). 49. B. Lunberg and B. Nilsson, Osteopenia in the frozen shoulder. Clin. Orthop. 60, 187 – 191 (1968).

226 50. P. Rush, J. Wong, J. Kirsh, and M. Devlin, Osteopenia in patients with above knee amutation. Arch. Phys. Med. Rehabil. 75, 112 – 115 (1994). 51. D. P. Jenkins and T. H. Cochran, Osteoporosis: The dramatic effect of disuse of an extremity. Clin. Orthop. 64, 128 – 134 (1969). 52. T. Cundy and A. Grey, Mechanisms of cortical bone loss from the metacarpal following digital amputation. Calcif. Tissue Int. 55, 164 – 168 (1994). 53. S. Andersson and B. Nilsson, Changes in bone mineral content following ligamentous knee injuries. Med. Sci. Sports 11, 351 – 353 (1979). 54. T. Hansson, B. Roos, and A. Nachemson, Development of osteopenia in the fourth lumbar vertbra during prolonged bed rest after operation for scoliosis. Acta Orthop. Scand. 46, 621 – 630 (1975). 55. P. Kannus, M. Jarvinen, H. Sievanen, P. Oja, and I. Vuori, Osteoporosis in men with a history of tibial fracture. J. Bone Miner. Res. 9, 423 – 429 (1994). 56. B. Issekutz, J. Blizzard, N. Birkhead, and K. Rodah, Effect of prolonged bed rest on urinary calcium output. J. Appl. Physiol. 21, 1013 – 1020 (1966). 57. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers, and D. E. McMillan, Effect of prolonged bed rest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 58. S. B. Hulley, J. M. Vogel, and C. L. Donaldson, Effect of supplemental calcium and phosphorus on bone mineral changes in bed rest. J. Clin. Invest. 50, 2506 – 2518 (1971). 59. V. Schneider and J. McDonald, Skeletal calcium homeostasis and countermeasures to prevent disuse osteoporosis. Calcif. Tissue Int. 36 (Suppl.), S151 – S154 (1984). 60. A. Leblance, V. Schneider, J. Krebs, H. Evans, S. Jhingram, and P. Johnson, Spinal bone mineral after 5 weeks of bed rest. Calcif. Tissue Int. 41, 259—261 (1987). 61. A. LeBlanc, V. Schneider, H. Evans, D. Engelbretson, and J. Krebs, Bone mineral loss and recovery after 17 weeks of bed rest. J. Bone Miner. Res. 5, 843—850 (1990). 62. S. Palle, L. Vico, S. Bourrin, and C. Alexandre, Bone tissue response to four-month antiorthostatic bedrest: A bone histomorphometric study. Calcif. Tissue Int. 51, 189 – 194 (1992). 63. S. Arnaud, D. Sherrard, N. Maloney, R. Whalen, and P. Fung, Effects of 1-week head-down tilt bed rest on bone formation and the calcium endorcine system. Aviat. Space Environ. Med. 63, 14 – 20 (1992). 64. S. Lueken, S. Arnaud, A. Taylor, and D. Baylink, Changes in markers of bone formation and resorption in a bed rest model of weightlessness. J. Bone Miner. Res. 8, 1433 – 1438 (1993). 65. A. Chantraine and P. Minaire, Para-osteo-arthropathies. Scand. J. Rehabil. Med. 13, 31 – 37 (1981). 66. A. Perkash, G. Sullivan, L. Toth, L. Bradleigh, S. Linder, and I. Perkash, Persistent hypercoagulation associated with heterotopic ossification in patients with spinal cord injury long after injury has occured. Paraplegia 31, 653 – 659 (1993). 67. S. C. Colachis and D. M. Clinchot, The association between deep venous thrombosis and heterotopic ossification in patients with acute traumatic spinal cord injury. Paraplegia 31, 507 – 512 (1993). 68. R. Wittenberg, L. Peschke, and U. Botel, Heterotopic ossification after spinal cord injury. J. Bone Jt. Surg. Br. 74B, 215 – 218 (1992). 69. L. Jensen, E. Halar, J. Little, and M. Brooke, Neurogenic heterotopic ossification. Am. J. Phys. Med. 66, 351 – 362 (1988). 70. S. Stover, K. Niemann, and J. Tulloss, Experience with surgical resection of heterotopic bone in spinal cord injury patients. Clin. Orthop. 263, 71 – 77 (1991). 71. D. Ayers, V. Pellegrini, and C. Evarts, Prevention of heterotopic ossification in heigh-risk patients by radiation therapy. Clin. Orthop. 263, 87 – 93 (1991).

B. JENNY KIRATLI 72. K. Banovac, F. Gonzalez, N. Wade, and J. Bowker, Intravenous disodium etidronate therapy in spinal cord injury patients with heterotopic ossification. Paraplegia 31, 660 – 666 (1993). 73. S. Strover, K. Niemann, and J. I. Miller, Disodium etidronate in the preventation of postoperative recurrence of heterotopic ossification in spinal cord injury patients. J. Bone Jt. Surg. 58A, 683 – 688 (1976). 74. C. Mautalen, D. Gonzales, E. Blumenfeld, E. Araujo, and F. Schajowicz, Spontaneous fracture of uninvolved bones in patients with Paget’s disease during unduly prolonged treatment with disodium etidronate. Clin. Orthop. 207, 150 – 155 (1986). 75. C. Johnston, R. Altman, R. Canfield, G. Finerman, J. Taulbee, and M. Ebert, Review of fracture experience during treatment of Paget’ disease of bone with etidronate disodium. Clin. Orthop. 172, 186 – 194 (1983). 76. S. Krane, Etidronate disodium in the treatment of Pagent’s disease of bone. Ann. Int. Med. 96, 619 – 625 (1982). 77. R. Ingram, R. Suman, and P. Freeman, Lower limb fractures in the chronic spinal cord injured patient. Paraplegia 27, 133 – 139 (1989). 78. A. Freehafer and W. Mast, Lower extremity fractures in patients with spinal cord injury. J. Bone Jt. Surg. 47A, 683 – 694 (1965). 79. A. Freehafer, M. Coletta, and C. Becker, Lower extremity fractures in patients with spinal cord injury. Paraplegia 19, 367 – 372 (1981). 80. S. Eichenholtz, Management of long-bone fractures in paraplegic patients. J. Bone Jt. Surg. 45A, 299 – 310 (1963). 81. D. Garland, T. Saucedo, and T. Reiser, The management of tibial fractures in acute spinal cord injury patients. Clin. Orthop. 213, 237 – 239 (1986). 82. D. Garland, Clinical observations on fractures and heterotopic ossifi cation in the spinal cord and traumatic brain injured populations. Clin. Orthop. 233, 86 – 101 (1988). 83. D. Garland, Z. Maric, R. Adkins, and C. Stewart, Bone mineral density about the knee in spinal cord injured patients with pathologic fractures. Contemp. Orthop. 26, 375 – 379 (1993). 84. B. Kiratli, T. Nauenberg, and I. Perkash, Estimation of femoral strength from bone mineral and geometric data in spinal cord injured patients with and without lower extremity fractures. Trans. Orthop. Res. Soc. 19, 435 (1994). 85. G. Whedon, J. Deitrick, and E. Shorr, Modification of the effects of immobilization upon metabolic and physiologic functions of normal men by use of an oscillating bed. Am. J. Med. 6, 684 – 711 (1949). 86. D. A. Hantmann, J. M. Vogel, C. L. Donaldson, R. Friedman, R. S. Goldsmith, and S. B. Hulley, Attempts to prevent disuse osteoporosis by treatment with calcitonin, longitudinal compression and supplementary calcium and phosphate. J. Clin. Endocrinol. Metab. 36, 845 – 858 (1973). 87. F. Plum and M. F. Dunning, The effect of therapeutic mobilization on hypercalciuria following acute poliomyelitis. Arch. Intern. Med. 101, 528 – 536 (1958). 88. A. Abramson and E. Delagi, Influence of weight-bearing and muscle contraction on disuse osteoporosis. Arch. Phys. Med. Rehabil. 42, 147 – 151 (1961). 89. P. Kaplan, W. Roden, E. Gilbert, L. Richards, and J. Goldschmidt, Reduction of hypercalciuria in tetraplegia after weight-bearing and strengthening exercises. Paraplegia 19, 289 – 293 (1981). 90. P. E. Kaplan, B. Gandhavadi, L. Richards, and J. Goldschmidt, Calcium balance in paraplegic patients: Influence of injury and ambulation. Arch. Phys. Med. Rehabil. 59, 447 – 450 (1978). 91. C. Kunkel, A. Scremin, B. Eisenberg, J. Garcia, S. Roberts, and S. Marinez, Effects of “standing” on spasticity, contracture and osteoporosis in paralyzed males. Arch. Phys. Med. Rehabil. 74, 73 – 78 (1993).

CHAPTER 46 Immobilization Osteopenia 92. M. Rodgers, R. Glaser, S. Figoni, S. Hooker, B. Ezanwa, S. Collins, T. Mathews, A. Suryaprasad, and S. Gupta, Musculoskeletal responses of spinal cord injured individuals to functional neuromuscular stimulation-induced knee extension exercise training. J. Rehabil. Res. Dev. 28, 19 – 26 (1991). 93. P. Pacy, R. Heep, D. Halliday, D. Katz, G. Cameron, and J. Reeve, Muscle and bone in paraplegic pateints, and the effect of functional electric stimulation. Clin. Sci. 75, 481 – 487 (1988). 94. E. M. Leeds, J. Klose, W. Ganz, A. Serafini, and B. S. Green, Bone mineral density after bicycle ergometry training. Arch. Phys. Med. Rehabil. 71, 207 – 209 (1990). 95. K. Sloan, L. Bremmer, J. Byrne, R. Day, and E. Scull, Musculoskeletal effects of an electric stimulation induced cyclingprogramme in the spinal injured. Paraplegia 32, 407 – 415 (1994). 96. G. Merli, G. McElwain, A. Adler, J. Martin, J. Roberts, B. Schnall, and J. Ditunno, Immobilization hypercalcemia in acute spinal cord injury treated with etidronate. Arch. Intern. Med. 144, 1286 – 1288 (1984).

227 97. D. Carey and L. Raisz, Calcitonin therapy in prolonged immobilization hypercalcemia. Arch. Phys. Med. Rehabil. 66, 640 – 644 (1985). 98. J. Meythaler, J. Stephen, M. Tuel, and L. Cross, Successful treatment of immobilization hypercalcemia using calcitonin and etidronate. Arch. Phys. Med. Rehabil. 74, 316 – 319 (1993). 99. P. Minaire, J. Depassio, E. Berard, P. Meunier, C. Edouard, G. Pilchonery, and G. Goedert, Effects of clodronate on immobilization bone loss. Bone 8, S63 – S68 (1987). 100. D. Lockwood, J. Vogel, V. S. Schneider, and S. Hulley, Effect of the diphosphonate EHDP on bone mineral metabolism during prolonged bed rest. J. Clin. Endocrinol. Metab. 41, 533 – 541 (1975). 101. D. Chappard, P. Minaire, C. Privat, E. Berard, J. MendozaSarmiento, H. Tournebise, M. Basle, M. Audran, A. Rebel, C. picot, and C. Gaud, Effects of tiludronate on bone loss in paraplegic patients. J. Bone Miner. Res. 10, 112 – 128 (1995). 102. A. LeBlanc and V. Schneider, Can the adult skeleton recover lost bone. Exp. Gerontol. 26, 189 – 201 (1991).

CHAPTER 47

Thyroid Hormone and the Skeleton DANIEL BARAN

I. II. III. IV.

Departments of Medicine, Orthopedics, and Cell Biology, University of Massachusetts Medical Center, and Merck & Company Inc., Worcester, Massachusetts 01655

Introduction Thyroid Hormone and Skeletal Growth and Maturation Thyroid Hormone and Mineral Metabolism Thyroid Hormone and Skeletal Metabolism

V. Bone Mass and Fracture Risk in Thyroid Disease VI. Prevention of Thyroid Hormone-Induced Bone Loss References

I. INTRODUCTION

detrimental to bone mass, raising the risk of subsequent fracture.

Sufficient exposure to thyroid hormones is necessary for normal skeletal development. Growth and maturation of the skeleton are complex events that result from the interaction of nutritional, genetic, and hormonal factors. In both the developing and the adult skeleton, thyroid hormone is necessary for the recruitment and maturation of bone cells. Deficient thyroid hormone production in utero and in the neonate retards growth and delays skeletal maturation. The presence of thyroid hormone increases bone remodeling. Excess thyroid hormone alters the hormonal regulation of calcium metabolism and can contribute to bone loss. Mineral metabolism is controlled through the interaction of parathyroid hormone (PTH) and 1,25-dihydroxyvitamin D (1,25(OH)2D). These hormones, acting systemically, along with growth factors and cytokines produced and acting locally regulate the activity of bone cells, skeletal remodeling, and calcium metabolism. Thyroid hormone alters mineral metabolism by acting directly on bone cells to increase bone resorption with secondary changes in PTH, 1,25(OH)2D, and calcium. The changes in mineral metabolism induced by excess thyroid hormone are potentially

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. THYROID HORMONE AND SKELETAL GROWTH AND MATURATION The decrease in skeletal growth noted in thyroiddeficient animals may be due in part to decreased production of growth hormone (GH), insulin-like growth factor I (IGF-I), and poor nutrition [1]. GH and thyroid hormone appear to potentiate each other’s effect on skeletal growth in hypophysectomized animals [2– 4]. Thyroid hormone alone minimally stimulates growth in hypophysectomized animals [5,6]. The greatest improvement in growth is noted when both hormones are simultaneously administered [7– 9]. Thyroxine and GH potentiate the effect of each other on skeletal growth in rats [6,10,11] and humans [7,9], supporting the view of the synergistic actions of each of these hormones on skeletal growth. The effects of thyroid hormone on bone cells may be mediated in part by stimulation of IGF-I production. IGF-I

229

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

230 increases osteoblast number, enhances osteoblast differentiation, and increases bone remodeling. Circulating IGF-I levels are decreased in hypothyroidism [12 – 20]. Thyroid hormone exerts a direct effect on osteoblasts to increase production of IGF-I mRNA, IGF-I, and IGF binding protein II [21,22]. In addition, thyroid hormone may regulate IGF-I receptors since decreased IGF-I receptor numbers in the pituitary gland of hypothyroid animals normalize by thyroid replacement [23]. In hypothyroid patients treated with thyroxine, there is a positive correlation between free triiodothyronine and IGF-I concentration after treatment [18]. Nutritional state also affects IGF-I levels. In children [24 – 26] and adults [19,27], impaired nutritional state is the most likely explanation for the correlation between IGF-I and thyroid hormone levels in nonthyroidal illness. The combined data suggest that thyroid hormone, in combination with growth factors and nutritional state, is instrumental in normal skeletal development.

III. THYROID HORMONE AND MINERAL METABOLISM Abnormalities in serum calcium concentrations are observed in patients with hyperthyroidism. Mild hypercalcemia occurs in 20% of patients with thyrotoxicosis [28 – 30]. As many as 50% of patients with hyperthyroidism have elevations in ionized calcium activity, but the modest degree of hypercalcemia rarely causes symptoms. Serum PTH concentrations and bioactivity [31,32], serum 1,25(OH)2D [33,34], and intestinal calcium absorption [35,36] are decreased in thyrotoxic patients, suggesting that thyroid hormone-induced bone resorption explains the occurrence of hypercalcemia. Hypercalciuria is common in hyperthyroid patients even in the absence of hypercalcemia and normalizes after treatment [28]. The increased urinary calcium excretion appears to be secondary to suppression of PTH secretion, resulting in decreased renal tubular calcium reabsorption. Thyrotoxicosis also increases fecal calcium excretion. Changes in intestinal secretion, enteric circulation of bile, and intestinal transit time along with steatorrhea account for the increased fecal calcium loss. The net effect of diminished intestinal calcium absorption and increased fecal and urinary calcium excretion in thyrotoxic patients is negative calcium balance. Abnormalities in serum calcium are uncommon in patients with hypothyroidism. Patients with hypothyroidism have a blunted response to hypocalcemia, presumably due to decreased renal and bone sensitivity to PTH[37], since PTH secretion itself is increased in hypothyroidism [37,38]. The increased circulating PTH is responsible for increased 1,25(OH)2D levels [39], which in turn increase intestinal calcium absorption [40,41]. Calcium losses in urine and

DANIEL BARAN

feces are decreased in hypothyroidism. Decreased release of calcium from bone apparently increases the steady-state of concentration of PTH.

IV. THYROID HORMONE AND SKELETAL METABOLISM Thyroid hormone increases bone remodeling [42]. Although both osteoblast and osteoclast activity are increased by elevated levels of thyroid hormone, osteoclast activity predominates with a resultant loss of bone mass. Thyroid hormones stimulate osteoclastic bone resorption by an indirect effect mediated by osteoblasts, which possess thyroid hormone receptors, and whose presence is required for increased bone resorption to occur [43 – 47]. Bone-cell-specific triiodothyronine responses reflect differing patterns of receptor gene and osteblastic phenotype expression. Thyroid hormone directly stimulates osteoblast production of alkaline phosphatase [48,49], osteocalcin [22,50], and IGF-I [21]. Thyrotoxicosis is associated with increased serum levels of osteocalcin [51 – 53] and alkaline phosphatase [54]. Despite increased osteoblastic activity, the enchanced bone formation cannot compensate for thyroid hormone-induced increments in bone resorption. Increased bone resorption is detected by increased urinary excretion of hydroxyproline and collagen cross-links in thyrotoxic patients [42,55 – 57]. The levels of these markers of bone turnover correlate with circulating levels of thyroid hormones [58]. In contrast, serum alkaline phosphatase activity and osteocalcin concentration [59,61] are frequently decreased in hypothyroid patients, and urinary hydroxyproline excretion is decreased despite the elevated PTH concentration in these patients. In thyrotoxic bones, the surface area of unmineralized matrix (osteoid) is increased. In contrast to osteomalacia, mineralization rates are increased. The elevated bone turnover is characterized by an increase in the numbers of osteoclasts and prevalence of resorption sites and the ratio of resorptive to formative surfaces. In contrast to the normal bone remodeling cycle which lasts about 200 days, in hyperthyroid patients the cycle is shortened primarily due to a decrease in the length of the formation period, resulting in greater than normal inefficiency in replacing resorbed bone [42]. Cortical bone in hyperthyroidism is characterized by increased porosity [42]. Administration of thyroid hormone to rats increases gene transcripts for osteoclast and osteoblast markers in the appendicular, but not axial, skeleton [62]. Thyroid hormoneinduced increases in tartrate-resistant acid phosphatase mRNA (an osteoclast marker), alkaline phosphatase mRNA (an osteoblast marker), and histone H4 mRNA (a cell proliferation marker) in the femurs of treated rats correlate with decreases in bone mineral density.

231

CHAPTER 47 Thyroid Hormone and the Skeleton

V. BONE MASS AND FRACTURE RISK IN THYROID DISEASE Bone mass is reduced in patients with thyrotoxicosis [63 – 67]. The detrimental skeletal effects of thyroid hormone occur more frequently in female patients [63]. As a result of the decrease in bone density, individuals with a history of thyrotoxicosis have an increased risk of fracture [68] and sustain fractures at an earlier age than individuals who have never been thyrotoxic [69] (Fig. 1). The decreased bone density noted in thyrotoxic patients is reversible after effective treatment. Normalization of thyroid function results in significant increases in axial and appendicular bone density compared to pretreatment values [70 – 74]. The detrimental skeletal effects of thyroid hormone are not restricted to individuals with thyrotoxicosis, but may also occur in patients administered supraphysiologic doses of thyroid hormone to suppress thyroid-stimulating hormone(TSH) secretion in the treatment of differentiated thyroid carcinoma or nontoxic goiter. In patients with low bone mass, this therapy may aggravate fracture risk. Initial studies demonstrated that supraphysiologic doses of thyroid hormone in premenopausal women were associated with decreased forearm [75] and femoral neck [76] bone mineral density. These studies were confounded by the inclusion of women who had previously been thyrotoxic. Some studies of premenopausal women who had never been thyrotoxic confirmed that TSH-suppressive doses of thyroid hormone decreased axial and appendicular bone mass [77 – 82], but other studies have not confirmed such a decrease [83 – 91].

Similarly, in postmenopausal women, TSH-suppressive doses of thyroid hormone have been reported to decrease [77,79,84,85,88,92,93] or have no effect on [83,87,91,94,95] bone mineral density. In general, supraphysiologic doses of thyroxine do not appear to decrease bone mineral density in men [84,87,96]. A meta analysis of the reports in which bone mineral density was assessed in women receiving TSHsuppressive doses of thyroxine concluded that treatment did not appear to significantly reduce bone mass in premenopausal women, but did lead to a 1% increase in annual bone loss in postmenopausal women [97]. A subsequent meta analysis confirmed the detrimental effects of suppressive doses of thyroid hormone on bone mass in postmenopausal women [98]. Suppressive therapy did not appear to affect bone mineral density in premenopausal women. The authors cautioned against overzealous or irrelevant use of thyroid hormone. Although these meta analyses suggest that TSH-suppressive doses of thyroxine are not immediately detrimental to bone mass in premenopausal women, there is no reason to assume that premenopausal women will not be at increased risk with long-term treatment. Hypothyroidism decreases recruitment, maturation, and activity of bone cells, leading to decreased bone resorption and formation [42]. Despite these decreases, trabecular bone volume and bone mineral density are similar to that of euthyroid subjects. It is unclear whether replacement therapy with doses of thyroid hormone that maintain TSH within the physiologic range affects bone mass. Treatment of hypothyroid patients with replacement doses of thyroid hormone has been reported to decrease [99 – 101] and to have no effect [102 – 103] on bone mineral density. A meta analysis concluded that replacement therapy is associated with decreased bone mineral density at the spine and hip in premenopausal, but not postmenopausal, women [98]. At present it seems prudent to assess bone mineral density in hypothyroid patients in whom replacement therapy is initiated.

VI. PREVENTION OF THYROID HORMONE-INDUCED BONE LOSS

FIGURE 1 Age at which the first fracture occurred in 82 white postmenopausal women, 37 with a history of thyroid disease, 45 without a history of thyroid disease. The difference is depicted using survival techniques; Wilcoxon  6.95; P  0.008. Reproduced with permission from [69].

Treatment of thyrotoxic patients increases bone density compared to pretreatment values [70 – 74]. A more difficult situation is presented by the thyroid cancer patient who requires TSH-suppressive doses of thyroid hormone. Bone mass measurements at 1-to 2-year intervals will detect patients with accelerated rates of bone loss. In animal models of thyroid hormone-induced TSH suppression, bisphosphonates prevented the detrimental effects of thyroid on the skeleton [62,104,105] while calcitonin did not [106]. Bisphosphonates also prevent the increases in the biochemical markers of osteoblast and osteoclast activity that occur after thyroid hormone administration in humans [107]. In a cross-sectional study, estrogen appeared to negate thyroid

232 hormone-associated loss of bone density in postmenopausal women taking replacement doses of thyroid hormone [108] (Fig. 2). Women taking suppressive doses of thyroid hormone who were also taking estrogen had significantly higher bone mineral density values than women who were on suppressive doses of thyroid hormone alone and had values similar to those taking neither hormone. In contrast, bone density values did not differ by estrogen status in women taking replacement doses of thyroid hormone [108]. In a prospective study of patients with thyroid cancer receiving suppressive doses of thyroxine, pamidronate suppressed bone resorption and increased bone mineral density [91]. Although patients receiving the bisphosphonate experienced the expected increase in bone mineral density, patients in the control group receiving thyroxine alone did not show accelerated bone loss. In that study [91] only 10% of the subjects were postmenopausal women, perhaps explaining the absence of an effect of suppressive doses of thyroxine on bone mass.

FIGURE 2

Mean bone mineral densities (95% confidence intervals) by current thyroid hormone use and estrogen use adjusted for age, body mass index, smoking, and use of thiazide diuretics and oral corticosteroids in women, Rancho Bernardo, California, 1988 to 1991. Reproduced with permission from [108]. © 1994 American Medical Association.

DANIEL BARAN

Thus, current evidence suggests that women with a history of thyrotoxicosis or TSH-suppressive therapy should have assessment of bone mineral density [109]. Estrogen or bisphosphonate therapy should be considered for those individuals who demonstrate accelerated rates of bone loss or who already manifest decreased bone mass.

References 1. M. L. Mitchell, R. J. Hermos, M. Feingold, and A. C. Moses, The relationship of insulin-like growth factor-I to total thyroxine in normal and low birth weight infants. Pediatr. Res. 25, 336 – 338 (1989). 2. C. Dieguez, V. Jordan, P. Harris, S. Foord, M. D. Rodriguez-Amao, A. Gomez-Pan, and M. F. Scanlon, Growth hormone responses to growth hormone-releasing factor 1-29 in euthyroid, hypothyroid, and hyperthyroid rats. J. Endocrinol. 109, 53 – 56 (1986). 3. A. W. Root, D. Shulman, J. Root, and F. Diamond, The interrelationships of thyroid and growth hormones: Effect of growth hormone releasing hormone in hypo- and hyperthyroid male rats. Acta Endocrinol. Suppl. 279, 367 – 373 (1986). 4. H. Katakami, T. R. Downs, and L. A. Frohman, Decreased hypothalamic growth hormone-releasing hormone content and pituitary responsiveness in hypothyroidism. J. Clin. Invest. 77, 1704 – 1711 (1986). 5. R. O. Scow, Effect of growth hormone and thyroxine on growth and chemical composition of muscle, bone, and other tissues in the thyroidectomized – hypophysectomized rats. Am. J. Physiol. 196, 859 – 870 (1959). 6. K. G. Thorngren and L. I. Hansson, Effect of thyroxine and growth hormone on longitudinal bone growth in the hypophysectomized rat. Acta Endocrinol. 74, 24 – 40 (1973). 7. Y. Harada, Y. Okada, T. Hikita, T. Niomiya, and Y. Ishihara, Comparison of growth acceleration of pituitary dwarfs treated with anabolic steroid and thyroid hormone with normal growth spurt. Acta Endocrinol. 74, 237 – 249 (1973). 8. M. M. Martin and L. Wilkins, Pituitary dwarfism: Diagnosis and treatment. J. Clin. Endocrinol. Metab. 18, 679 – 687 (1958). 9. J. L. Van den Brande, J. J. Van Wyk, F. S. French, A. L. Strickland, and W. B. Radcliffe, Advancement of skeletal age of hypopituitary children treated with thyroid hormone plus cortisone. J. Padiatr. 82, 22 – 27 (1973). 10. K. G. Thorngren and L. I. Hansson, Effect of thyroxine and growth hormone on longitudinal bone growth in the hypophysectomized rat. Acta Endocrinol. 74, 24 – 40 (1973). 11. D. Levenson, G. M. Bialik, and Z. Ochberg, Differential effects of hypothyroidism on the cartilage and the osteogenic process in the mandibular condyle: Recovery by growth hormone and thyroxine. Endocrinology 135, 1504 – 1510 (1994). 12. P. J. Burstein, B. Draznin, C. J. Johnson, and D. S. Schalch, The effect of hypothyroidism on growth, serum growth hormone, the growth hormone-dependent somatomedin, insulin-like growth factor, and its carrier protein in rats. Endocrinology 104, 1107 – 1111 (1979). 13. H. Cavaliere, M. Knobel, and G. Mederios-Neto, Effect of thyroid hormone therapy on plasma insulin-like growth factor I levels in normal subjects, hypothyroid patients and endemic cretins. Horm. Res. 25, 132 – 139 (1987). 14. S. D. Chernausek and R. Turner, Attenuation of spontaneous, nocturnal growth hormone secretion in childern with hypothyroidism and its correlation with plasma insulin-like growth factor I concentrations. J. Pediatr. 114, 968 – 972 (1989). 15. B. Draznin, P. J. Burstein, U. E. Heinrich, C. B. Johnson, C. A. Emler, and D. S. Schalch, Insulin-like growth factor and its carrier

CHAPTER 47 Thyroid Hormone and the Skeleton

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28. 29. 30.

31.

32.

33.

protein in hypopituitary and hypothyroid children and adults. Clin. Endocrinol. 12, 137 – 142 (1980). N. Hogerbrugge-vd, N. Linden, H. Jansedn, W. C. Hulsmann, and J. C. Birkenhager, Relationship between insulin-like growth factor-I and low-density lipoprotein cholesterol levels in primary hypothyroidism in women. J. Endocrinol. 123, 341 – 345 (1989). S. G. Ren, S. Malozowski, C. Simoni, H. Garcia, M. CarusoNicoletti, G. B. Cutler, and F. Casorla, Doss – response relationship between thyroid hormone and growth velocity in cynomolgus monkeys. J. Clin. Endocrinol. Metab. 66, 1010 – 1013 (1987). R. Valcavi, C. Dieguez, M. Preece, A. Taylor, I. Portioli, and M. F. Scanlon, Effect of thyroxine replacement therapy on plasma insuline-like growth factor 1 levels and growth hormone responses to growth hormone releasing factor in hypothyroid patients. Clin. Endocrinol. 27, 85 – 90 (1988). M. Valimaki, Liewendahl, S. L. Karonen, T. Helenius, and A. M. Suikkari, Concentrations of somatomedin-C and triiodothyronine in patients with thyroid dysfunction and nonthyroidal illnesses. J. Enocrinol. Invest. 13, 155 – 159 (1990). K. Westermark, L. Alm, A. Skottner, and A. Karlsson, Growth factors and the thyroid: Effects of treatment for hyper- and hypothyroidism on serum IGF-I and urinary epidermal growth factor concentrations. Acta Endocrinol. 118, 415 – 421 (1988). C. Schmid, I. Schlapfer, E. futo, M. Waldvogel, J. Schwander, J. Zapf, and E. R. Froesch, Triiodothyronine (T3) stimulates insulin-like growth factor (IGF)-1 and IGF binding protein (IGFBP)-2 production by rat osteoblasts in vitro. Acta Endocrinol. 126, 467 – 473 (1992). M. Milne, M.-I. Kang, J. M. Quail, and D. T. Baran, Thyroid hormone excess increases insulin-like growth factor I transcripts in bone marrow cell cultures: Divergent effects on vertebral and femoral cell cultures. Endocrinology 139, 2527 – 2534 (1998). K. Matsuo, S. Yamashita, M. Niwa, M. Kurihara, S. Harakawa, M. Izumi, S. Nagataki, and Melmed, Thyroid hormone regulates rat pituitary insulin-like growth factor-I receptors. Endocrinology 126, 550 – 554 (1990). D. B. Grant, J. Hambley, D. Becker, and B. L. Pimstone, Reduced sulphation factor in undernourished childern. Arch. Dis. Child. 48, 596 – 600 (1973). R. L. Hintz, R. Suskind, K. Amatayakul, O. Thanangkul, and R. Olsen, Plasma somatonedin and growth hormone values in childern with protein-caloric malnutrition. J. Pediatr. 92, 153 – 156 (1978). A. R. Zucker, B. Chernow, A. I. Fields, W. Hung, and K. D. Burman, Thyroid function in critically ill childern. J. Pediatr. 107, 552 – 554 (1985). D. R. Clemmons, A. Klibanski, L. E. Underwood, J. W. McArthur, E. C. Ridgway, I. Z. Beitins, and J. J. Van Wyk, Reduction of plasma immunoreactive somatonedin C during tasting in humans. J. Clin. Endocrinol. Metab. 53, 1547 – 1250 (1981). J. D. Baxter and P. K. Bondy, Hypercalcemia of thyrotoxicosis. Ann. Intern. Med. 65, 429 – 442 (1966). A. E. Farnsworth and B. M. Dobyns, Hypercalcemia and thyrotoxicosis. Med. J. Aust. 2, 782 – 784 (1974). D. L. Gordon, S. Suvanich, V. Erviti, M. A. Schwartz, and C. J. Martinez, The serum calcium level and its significance in hyperparathyroidism: A prospective study. Am. J. Med. Sci. 268, 31– 41 (1974). M. D. Fallon, H. M. Perry III, M. Bergfeld, D. Droke, S. L. Teitelbaum, and L. V. Avioli, Exogenous hyperthyroidism with osteoporosis. Arch. Intern. Med. 143, 442 – 444 (1983). L. Mosekilde and M. S. Christensen, Decreased parathyroid function in hyperthyroidism: Interrelationship between serum parathyroid hormone, clacium-phosphorus metaboism and thyroid function. Acta Endocrinol. 84, 566 – 575 (1977). R. Bouillon, E. Muls, and P. DeMoor, Influence of thyroid function on the serum concentration of 1,25-dihydroxyvitamin D3. J. Clin. Endocrinol. Metab. 51, 793 – 797 (1980).

233 34. B. Jastrup, L. Mosekilde, f. Melsen, Bi. Lund, Bj. Lund, and O. H. Srensen, Serum levels of vitamin D metabolism and bone remodeling in hyperthyroidism. Metab. Clin. Exp. 31, 126 – 132 (1982). 35. B. Haldimann, E. M. Kaptein, F. R. singer, J. T. Nicoloff, and S. G. Massry, Intestinal calcium absorption in patients with hyperthyroidism. J. Clin. Endocrinol. Metab. 51, 995 – 997 (1980). 36. P. Singhelakis, C. C. Alveizaki, and D. G. Ikkos, Intestinal calcium absorption in hyperthyroidism. Metab. Clin. Exp. 23, 311 – 321 (1974). 37. J. H. Castro, S. M. Genuth, and L. Klein, Comparative response to parathyroid hormone in hyperthyroidism and hypothyroidism. Metab. Clin. Exp. 24, 839–848 (1975). 38. R. Bouillon, and P. De Moor, parathyroid function in patients with hyper and hypothyroidism. J. Clin. Endocrinol. Metab. 38, 999 – 1004 (1974). 39. R. Bouillon, E. Muls, and P. De Moor, Influence of thyroid function on the serum concentration of 1,25-dihydroxyvitamin D3. J. clin. endocrinol. Metab. 51, 793 – 797 (1980). 40. J. F. F. Lekkerkerker and H. Doorenbos, The influence of thyroid hormone on calcium absorption from the gut in relation to urinary calcium excretion. Acta Endocrinol. 73, 672 – 680 (1973). 41. J. F. F. Lekkerkerker, F. Van Woudenberg, H. Beekhuis, and H. Doorenbos, Enchancement of calcium absorption in hypothyroidism. Isr. J. Med. Sci. 7, 399 – 400 (1971). 42. L. Mosekilde, E. F. Eriksen, and P. Charles, Effects of thyroid hormone on bone and mineral metabolism. Endocrinol. Metab. Clin. North Am. 19, 35 – 63 (1990). 43. T. J. Allain, T. J. Chambers, A. M. Flanagan, and A. M. McGregor, Triiodothyronine stimulates rat osteoclastic bone resorption by an indirect effect. J. Endocrinol. 133, 327 – 331 (1992). 44. J. M. Britto, A. J. Fenton, W. R. Holloway, and G. C. Nicholson, Osteoblasts mediate thyroid hormone stimulation of osteoclastic bone resorption. Endocrinology 134, 169– 176 (1994). 45. G. R. Williams, R. Bland, and M. C. Sheppard, Characterization of thyroid hormone (T3) receptors in three osteosarcoma cell lines of distinct osteoblast phenotype: Interactions among T3, vitamin D3, and retinoid signaling. Endocrinology 135, 2375– 2385 (1994). 46. M. Milne, M.-I. Kang, G. Cardona, J. M. Quail, L. E. Braverman, W. W. Chin, and D. T. Baran, Expression of multiple thyroid hormone receptor isoforms in rat femoral and vertebral bone and in bone marrow osteogenic cultures. J. Cell. Biochem. 74, 684 – 693 (1999). 47. R. Gruber, K. Czerwenka, F. Wolf, G. M. Ho, M. Wilheim, and M. Paterlik, Expression of the vitamin D receptor, of estrogen and thyroid hormone receptor alpha-and beta-isoforms, and of the androgen receptor in cultures of native mouse bone marrow and of stromal/ osteoblastic cells. Bone 24, 465 – 473 (1999). 48. K. Kasono, K. Sato, D. C. Han, Y. Fujii, T. Tsushima, and K. Shizume, Stimulation of alkaline phosphatase activity by thyroid hormone in mouse osteoblast-like cells (MC3T3-E1): A possible mechanism of hyperalkaline phosphatasia in hyperthyroidism. Bone Miner. 4, 355 – 363 (1988). 49. K. Sato, D. C. Han, Y. Fujii, T. Tsushima, and K. Shizume, Thyroid hormone stimulates alkaline phosphatase activity in cultured rat osteoblastic cells (ROS 17/2.8) through 3,5,3’-triiodo-L-thyronine nuclear receptors. Endocrinology 120, 1873 – 1881 (1987). 50. R. Rizzoli, J. Poser, and U. Burgi, Nuclear thyroid hormone receptors in cultured bone cells. Metab. Clin. Exp. 35, 71 – 74 (1986). 51. J. Faber, H. Perrild, and J. S. Johansen, Bone gla protein and sex hormone-binding globulin in nontoxic goiterL Parameters for metabolic status at the tissue level. J. Clin. Endocrinol. Metab. 70, 49 – 55 (1990). 52. D. R. Garrel, P. D. Delmas, L. Malaval, and J. Tourniaire, Serum bone gla protein: A marker of bone turnover in hyperthyroidism. J. Clin. Endorinol. Metab. 62, 1052 – 1055 (1986).

234 53. B. P. Lukert, J. C. Higgins, and M. M. Stoskopf, Serum osteocalcin is increased in patients with hyperthyroidism and decreased in patients receiving glucocorticoids. J. Clin. Endocrinol. Metab. 62, 1056 – 1058 (1986). 54. D. S. Cooper, M. M. Kaplan, E. C. Ridgway, F. Maloof, and G. H. Daniels, Alkaline phosphatase isoenzyme patterns in hyperthyroid. Ann. Intern. Med. 90, 164 – 168 (1979). 55. R. D. Harvey, K. C. McHardy, I. W. Reid, F. Patterson, P. D. Bewsher, A. Duncan, and S. P. Robins, Measurement of bone collagen degradation in hyperthyroidism and during thyroxine replacement therapy using pyridinium cross-links as specific urinary makers. J. Clin. Endocrinol. Metab. 72, 1189 – 1194 (1991). 56. J. C. Krakerauer and M. Kleerekoper, Borderline low serum thyrotropin level is correlated with increased fasting urinary hydroxyproline excretion. Arch. Intern. Med. 152, 360 – 364 (1992). 57. P. Garnero, V. Vassy, A. Bertholin, J. P. Riou, and P. D. Delmas, Markers of bone turnover in hyperthyroidism and the effects of treatment. J. Clin. Endocrinol. Metab. 78, 955 – 959 (1994). 58. M. Okawa, K. Kushida, M. Takahashi, T. Ohishi, H. Hoshino, M. Suzuki, H. Ogihara, G. Ishigaki, and T. Inoue, Bone turnover and cortical bone mineral density in the distal radius in patients with hyperthyroidism being treated with antithyroid drugs for various periods of time. Clin. Endocrinol. 50, 171 – 176 (1999). 59. K. Brixen, H. K. Nielsen, E. F. Eriksen, P. Charles, and L. Mosekilde, Efficacy of wheat germ lectin-precipitated alkaline phosphatase in serum as an estimator of bone mineralization rate: Comparison to serum total alkaline phosphatase and serum bone gla-protein. Calcif. Tissue Int. 44, 93 – 98 (1989). 60. J. Cassar and S. Joseph, Alkaline phosphatase levels in thyroid disease. Clin. Chim. Acta 23, 33 – 37 (1969). 61. P. Charles, J. W. Poser, L. Mosekilde, and F. T. Jensen, Estimation of bone turnover evaluated by 47Ca-kinetics. Efficiency of serum bone gamm-carboxyglutamic acid-containing protein, serum alkaline phosphatase, and urinary hydroxyproline excretion. J. Clin. Invest 76, 2254 – 2258 (1985). 62. B. Ongphiphadhanakul, L. G. Jenis, L. E. Braverman, S. Alex, G. S. Stein, J. B. Lian, and D. T. Baran, Etidronate inhibits the thyroid hormone-induced bone loss in rats assessed by bone mineral density and messenger ribonuclei acid markers of osteoblast and osteoclast function. Endocinology 133, 2502 – 2207 (1993). 63. M. S. Lee, S. Y. Kim, M. C. Lee, B. Y. Cho, H. K. Lee, C. S. Koh, and K. K. Min, Negative correlation between the change in bone mineral density and serum osteocalcin in patients with hyperthyroidism. J. Clin. Endocrinol. Metab. 70, 766 – 770 (1990). 64. J. Linde and T. Friis, Osteoporosis in hyperthyroidism estimated by photon absorptiometry. Acta Endocrinol. 91, 437 – 448 (1979). 65. S. H. Toh, B. C. Claunch, and P. H. Brown, Effect of hyperthyroidism and its treatment on bone mineral content. Arch. Intern. Med. 145, 883 – 886 (1985). 66. B. Krolner, Vesterdal J. Jorgensen, and S. Pors Nielsen, Spinal bone mineral content in myxedema and thyrotoxicosis: Effects of thyroid hormone(s) and antithyroid treatment. Clin. Endocrinol. 18, 439 – 446 (1983). 67. C. Gomez-Acotto, A. M. Schoot, D. Hans, H. Niepumniszcze, C. A. Mautalen, and P. J. Meunier, Hyperthyroidism influences ultrasound bone measurements on the os calcis. Osteoporosis Int. 8, 455 – 459 (1998). 68. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrad, J. Cauley, D. Black, and T. M. Vogt, For the study of Osteoporotic Fractures Research Group, Risk factors of hip fracture in white women. N. Engl. J. Med. 332, 767 – 773 (1995). 69. B. L. Solomon, L. Wartofsky, and K. D. Burman, Prevalence of fractures in postmenopausal women with thyroid disease. Thyroid 3, 17 – 23 (1993).

DANIEL BARAN 70. C. J. Rosen and R. A. Alder, Longitudinal changes in lumbar bone densith among thyrotoxic patients after attainment of euthyroidism. J. Clin. endocrinol. 75, 1531 – 1534 (1992). 71. T. Diamond, J. Vine, R. Smart, and P. Butler, Thyrotoxic bone disease in women: A potentially reversible disorder. Ann. Intern. Med. 120, 8 – 11 (1994). 72. M. Wakasugi, R. Wakao, M. Tawata, N. Gan, M. Inoue, K. Koizumi, and T. Onaya, Changes in bone mineral density in patients with hyperthyroidism after attainment of euthyroidism by dual energy x-ray absorptiometry. Thyroid 4, 179 – 182 (1994). 73. A. H. Mudde, A. J. Houben, A. J. Nieuwenhuizen, and A. C. Kruseman, Bone metabolism during antithyroid drug treatment of endogenous subclinical hyperthyroidism. Clin. Endocrinol. 41, 421 –424 (1994). 74. N. Arata, N. Momotani, H. Maruyama, T. Saruta, Tsukatanik, A. Kubo, K. Ikemoto, and K. Ito, Bone mineral density after surgical treatment for Graves disease. Thyroid 7, 547 – 554 (1997). 75. D. S. Ross, R. M. Neer, E. C. Ridgeway, and G. H. Daniels, Subclinical hyperthyroidism and reduced bone density as a possible result of prolonged suppression of the pituitary—thyroid axis with L-thyroxine Am. J. Med. 82, 1167 – 1170 (1987). 76. L. T. Paul, J. Kerrigan, A. M. Kelly, L. E. Braverman, and D. T. Baran, Long term L-thyroxine therapy is associated with decreased hip bone density in premenopausal women. JAMA 259, 3137 – 3141 (1988). 77. T. Diamond, L. Nery, and I. Hales, A therapeutic dilemma: Suppressive doses of thyroxine significantly reduced boen mineral measurements in both premenopausal and postmenopausal women with thyroid carcinoma. J. Clin. Endocrinol. Metab. 72, 1184 – 1188 (1991). 78. G. Pioli, M. Pedrazzoni, E. Palummeri, M. Sianesi, R. Delfrate, P. P. Vescovi, M. Prisco, V. Olivetti, D. Costi, and M. Passeri, Longitudinal study of bone loss after thyroidectomy and suppressive threapy in premenopausal women. Acta Endocrinol. 126, 238 – 242 (1992). 79. S. Greenspan, N. M. Resnick, J. E. Block, A. L. Friedlander, and H. K. Genant, Skeletal integrity in premenopausal and postmenopausal women receiving long-term L-thyroxine therapy. Am. J. Med. 91, 5 – 14 (1991). 80. G. Radetti, C. Castellan, L. Tato, K. Platter, L. Gentili, and S. Adami, Bone mineral density in childern and adolescent females treated with high doses of L-thyroxine. Horm. Res. 39, 127– 131 (1993). 81. G. DeRosa, A. Testa, D. Giacomini, C. Carrozza, P. Astazi, and P. Caradonna, Prospective study of bone loss in pre- and postmenopausal women on L-thyroxine therapy for non-toxic goiter. Clin. Endocrinol. 47, 529 – 535 (1997). 82. E. Jodar, M. Begona-Lopez, L. Garcia, D. Rigopoulou, G. Martinez, and F. Hawkins, Bone changes in pre- and postmenopausal women with thyroid cancer on levothyroxine therapy: Evolution of axial and appendicular bone mass. Osteoporos. Int. 8, 311 – 316 (1998). 83. C. Ribot, F. Tremollieres, J. M. Pouilles, and J. P. Louvet, Bone mineral density and thyroid hormone therapy. Clin. Endocrinol. 33, 143 – 153 (1990). 84. J. J. Stephan and Z. Limanova, Biochemical assessment of bone loss in patients on long term thyroid hormone treatment. Bone Miner. 17, 377 – 388 (1992). 85. P. Taelman, J. M. Kaufman, X. Janssens, H. Vandecauter, and A. Vermeulen, Reduced forearm bone mineral content and biochemical evidence of increased bone turnover in women with euthyroid goiter treated with thyroid hormone. Clin. Endocrinol. 33, 107 – 117 (1990). 86. C. Marcocci, F. Golia, G. Bruno- Bossio, E. Vignoli, and A. Pinchera, Carefully monitored levothyroxine suppressive therapy is not associated with bone loss in premenopausal women. J. Clin. Endocrinol. Metab. 78, 818 – 823 (1994). 87. J. A. Franklyn, J. Betteridge, J. Daykin, R. Holder, G. D. Oates, J. V. Parle, J. Lilley, D. A. Heath, and M. C. Sheppard, Long-term thyroxine treatment and bone mineral density. Lancet 340, 9 – 13 (1992).

CHAPTER 47 Thyroid Hormone and the Skeleton 88. J. Lehmke, V. Bogner, D. Felsenberg, H. Peter, and H. Schleusener, Determination of bone mineral density by quantitative computed tomography and single photon absorptiometry in subclinical hyperthyroidism: A risk of early osteopenia in postmenopausal women. Clin. Endocrinol. 36, 511 – 517 (1992). 89. G. Goerres, R. Theiler, and J. Muller-Brand, Interfemur variation of bone mineral density in patients receiving high does thyroxine therapy. Calcif. Tissue Int. 63, 98 – 101 (1998). 90. J. Toivonen, R. Tahtela, K. Laitinen, J. Risteli, and M. J. Valimaki, Markers of bone turnover in patients with differentiated thyroid cancer with and following withdrawal of thyroxine suppressive therapy. Eur. J. Endocrinol. 138, 667 – 773 (1998). 91. H. N. Rosen, A. C. Moses, J. Garber, D. S. Ross, S. L. Lee, L. Ferguson, V. Chen, K. Lee, and S. L. Greenspan, Randomized trial of pamidronate in patients with thyroid cancer: Bone dinsity is not reduced by suppressive doses of thyroxine, but is increased by cyclic intravenous pamidronate. J. Clin. Endocrinol. Metab. 83, 2324–2330 (1998). 92. G. M. Stall, S. Harris, L. J. Sokoll, and B. Dawson-Hughes, Accelerated bone loss in hypothyroid patients overtreated with L-thyroxine. Ann. Intern. Med. 113, 265 – 269 (1990). 93. E. V. Adlin, A. H. Maurer. A. D. Marks, and B. J. Channick, Bone mineral dinsity in postmenopausal women treated with L-thyroxine. Am. J. Med. 90, 360 – 366 (1991). 94. A. N. Gam, G. F. Jensen, K. Hasselstrom, M. Olsen, K. Siersbae, and K. Nielsen, Effect of thyroxine therapy on bone metabolism in substituted hypothyroid patients with normal or suppressed levels of TSH. J. Endocrinol invest. 14, 451 – 456 (1991). 95. D. C. Bauer, M. C. Nevitt, and B. Ettinger, et al., Low thyrotropin levels are not associated with bone loss in older women: A prospective study. J. Clin. Endocrinol. Metab. 82, 2931 – 2936 (1997). 96. C. Marcocci, F. Golia, and E. Vignoli, et al., Skeletal intergrity in men chronically treated with suppressive doses of L-thyroxine. J. Bone Miner. Res. 12, 72 – 77 (1997). 97. J. Faber and A. M. Galloe, Changes in bone mass during prolonged subclinical hyperthyroidism due to L-thyroxine treatment: A metaanalysis. Eur. J. Endocrinol. 130, 350 – 356 (1994). 98. B. Uzzam, J. Campus, and M. Cucherat, et al., Effects on bone mass of long-term treatment with thyroid hormones: A meta analysis. J. Clin. Endocrinol. Metab. 81, 4278 – 4289 (1996).

235 99. B. Krolner, J. V. Jorgensen, and S. P. Nielsen, Spinal bone mineral content in myoedema and thyrotoxicosis. Effects of thyroid hormone(s) and antithyroid treatment. Clin. Endocrinol. 18, 439 – 446 (1983). 100. P. Lakatos. G. Tarjan. J. Foldes, I. Krasnai, and I. Hollow, The effect of thyroid hormone treatment on bone mineral content in patients with hypothyroidism and euthyroid benign adenoma. In “osteoporosis” (C. Christianse, J. S. Johansen, B. J. Riis, eds.), pp. 452 – 453. Norhaven A/S Viborg, Denmark, 1987. 101. N. Demeester-Mirkine, P. Bergmann, J-J Body, and J. Corvilain, Calcitonin and bone mass status in songenital hypothyroidism. Calcif. Tissue Int. 46, 222 – 226 (1990). 102. S. H. Toh and P. H. Brown, bone mineral content in hypothyroid male patients with hormone replacement: A 3-year study. J. Bone Miner. Res. 5, 463 – 467 (1990). 103. W. E. Duncan, A. Chang, B. Solomon, and L. Wortofsky, Influence of clinical characteristics and parameters associated with thyroid hormone therapy on the bone mineral density of woman treated with thyroid hormone. Thyroid. 4, 183 – 189 (1994). 104. M. Yamamoto, A. Markatos, J. J. Seedor, P. Masarachia, M. Gentile, G. A. Rodan, and R. Galeena, The effects of aminobisphosphonate alendronate on thyroid hormone-induced osteopenia in rats. Calcif. Tissue Int. 53, 278 – 282 (1993). 105. H. N. Rosen, E. K. Sullivan, V. L. Middlebrooks, H. J. Zeind, C. Gundberg, R. Dresner-Pollack, L. A. Maitland, J. M. Hock, A. C. Moses, and S. L. Greenspan, Parental pamidronate prevents thyroid hormone-induced bone loss in rats. J. Bone Miner. Res. 8, 1255–1261 (1993). 106. B. Ongphiphadhanakul, S. Alex, L. E. Braverman, and D. T. Baran, TSH-suppressive L-thyroxine therapy decreases bone density in the rat: Effect of hypogonadism and calcitonin. J. Bone Miner. Res. 7, 1227 – 1231 (1992). 107. H. N. Rosen, A. C. Moses, C. Gundberg, V. T. Kung, S. M. Seyedin, T. Chen, M. Holick, and S. L. Greesnpan, Therapy with parenteral pamidronate prevents thyroid hormone-induced bone turnover in human. J. Clin. Endocrinol Metab. 77, 664 – 669 (1993). 108. D. L. Schneider, E. L. Barrett-Connor, and D. J. Morton, Thyroid hormone use and bone mineral density in elderly women: Effects of estrogen. JAMA 271, 1245 – 1249 (1994). 109. S. L. Greenspan and F. S. Greenspan, The effect of thyroid hormone on skeletal integrity. An. Intern. Med. 130, 750 – 758 (1999).

CHAPTER 48

Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases DANIEL D. BIKLE

Department of Veterans Affairs, San Francisco Veterans Affairs Medical Center, San Francisco, California 94121

I. Physiologic Considerations II. Gastrointestinal Diseases III. Hepatic Diseases

IV. Bone Disease Complicating Therapy References

I. PHYSIOLOGIC CONSIDERATIONS

A. Calcium Absorption Intestinal calcium absorption occurs throughout the intestine, although the highest rates of absorption are found in the duodenum [1]. Calcium absorption occurs through both transcellular and paracellular pathways; net absorption is reduced by calcium secretion and endogenous calcium losses associated with the sloughing of cells into the lumen [2]. Vitamin D, through its active metabolite 1,25(OH)2D, controls primarily the transcellular pathway. At low calcium intakes, the transcellular pathway dominates and provides a highly efficient means of absorption. However, as calcium intake increases, nonsaturable but less efficient pathways come into play, and calcium absorption falls to approximately 10% of the amount ingested at calcium intakes above 500 mg/day [3,4]. With age the efficiency of calcium absorption and the ability of the intestine to adapt to decreased calcium intake fall.

The mineral constituents of bone come from the diet and must be absorbed from the ingested food in the intestine. Vitamin D, through its active metabolite 1,25dihydroxyvitamin D (1,25[OH]2D), regulates the intestinal absorption of the two major mineral constituents of bone, calcium and phosphate. Although synthesized in the skin under the influence of ultraviolet light, vitamin D is also an important dietary constituent especially in circumstances of reduced exposure to ultraviolet light. Therefore, the bone is dependent on the adequate supply of calcium, phosphate, and vitamin D from the diet, and abnormalities of the hepato-gastrointestinal tract which impair their absorption cause bone disease. Vitamin D is comprehensively discussed in Chapter 9.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

237

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

238 Most reports using radioisotope absorption techniques have documented a reduction of intestinal calcium absorption with age especially in osteoporotic patients [5–12], although this was not observed in a more recent study by Eastell et al. [13] using a stable isotope procedure which traced all meals. Balance studies by Heaney et al. [14] indicate that with age and menopause the amount of dietary calcium required to maintain a positive balance increases, again consistent with the concept that calcium absorption efficiency decreases with age. Surgical induction of menopause (ovariectomy) results in a fall in intestinal calcium absorption which can be prevented by estrogen replacement [15]. 1,25(OH)2D corrects the decrease in calcium absorption with age [11,12,16]. Postmenopausal women who respond to the precursor of 1,25(OH)2D, namely 25-hydroxyvitamin D (25OHD), with a rise in intestinal calcium transport also show an increase in 1,25(OH)2D concentrations following 25OHD administration, whereas the nonresponders do not [17]. Estrogen administration to postmenopausal women raises circulating 1,25(OH)2D [11,15,18] and increases intestinal calcium transport [11,15]. Although estrogen also increases the concentration of the vitamin D binding protein, this does not account for the entire increase in 1,25(OH)2D since both free and total 1,25(OH)2D levels are raised by estrogen administration [19,20]. The stimulation of calcium absorption by estrogen may not be limited to increased 1,25(OH)2D production since estrogen receptors have been described in intestinal epithelial cells which may respond directly to estrogen with increased calcium transport [21]. Thus, a fall in 1,25(OH)2D levels, perhaps secondary to a fall in 1,25(OH)2D production by the aging or estrogen deprived kidney, could account for the decrease in intestinal calcium absorption with age. On the other hand, the aging intestine may become more resistant to 1,25(OH)2D action with respect to calcium absorption [22], as is suggested by the age related fall in vitamin D receptors in the intestinal epithelium [23]. Other dietary constituents alter calcium absorption. Lactose increases calcium absorption in a number of animals including humans [24,25], and lactase deficiency has been associated with an increased risk for osteoporosis [26 – 28]. To the degree that lactase deficiency would reduce calcium intake, the loss of this enzyme could predispose to osteoporosis. However, it is not clear that lactase deficiency per se reduces the efficiency of calcium absorption [9,29 – 31]. Phosphate increases fecal loss of calcium in part by increased endogenous calcium secretion [32]. However, phosphate reduces urinary loss of calcium [32 – 34], so the net effect of phosphate on calcium balance is not obviously harmful [35]. Nevertheless, at least in short-term studies, increased dietary phosphate can result in increased PTH secretion [36 – 38] and decreased 1,25(OH)2D production [39,40], hormonal changes which may be detrimental to bone. A diet rich in fiber could reduce calcium absorption by chelating calcium and other cations [41,42] and so decrease calcium

DANIEL D. BIKLE

balance [43,44]. However, a high-fiber diet has not been correlated with the development of osteoporosis.

B. Vitamin D Absorption Although the skin has the capability to produce adequate amounts of vitamin D given enough sunlight of sufficient intensity, because of our indoor lifestyle, modesty with respect to amount of skin exposed, and fear of cancer leading to zealous use of sun screens, this biologic pathway does not always suffice. Thus, dietary intake and intestinal absorption of vitamin D become important. Vitamin D is absorbed in the jejunum and ileum [45,46] by a mechanism capable of absorbing approximately 75% of the vitamin D administered [47]. Vitamin D appears in both the portal system and lymphatics, indicating that both pathways are utilized [45,46], although the lymphatic route may be preferred in humans [48]. In lymph, approximately 50% of vitamin D is found in the chylomicron fraction [46]. Fatty acids reduce vitamin D absorption, but this can be reversed with the addition of bile acids [45,46]. 25OHD (calcifediol) is better absorbed than vitamin D [46,49] especially in the presence of steatorrhea [50,51]. Vitamin D metabolites also undergo an extensive enterohepatic circulation. Arnaud et al. [52] noted the appearance in the duodenal lumen of 33% of the label 24 h after an intravenous dose of radiolabeled 25OHD. Nearly all of the secreted vitamin D metabolites were reabsorbed. The appearance of label following the intravenous administration of the dihydroxylated metabolites 1,25(OH)2D [53] and 24,25(OH)2D [54] in bile is even faster than that following 25OHD administration. In contrast, the appearance of radiolabel in the bile following the administration of radiolabeled vitamin D is slower and less extensive [55]. Primary biliary cirrhosis further reduces the appearance of vitamin D metabolites in the bile [55,56]. This chapter will emphasize the bone disease resulting from vitamin D and calcium malabsorption that complicates disorders in the hepato-gastrointestinal tract. Many of these disorders result in both osteoporosis and osteomalacia. Figure 1 shows six points at which vitamin D and/or calcium absorption could be affected by such disorders. First, adequate intake of vitamin D and calcium is required especially in an individual who otherwise fails to synthesize sufficient quantities of vitamin D in the skin. Milk and other dairy products are a good source of both, if these products are supplemented with vitamin D. Second, vitamin D absorption requires an intact small intestine, pancreas, and liver to provide the milieu (lipase, bile acids) required for vitamin D absorption. Partial gastrectomy, chronic pancreatic insufficiency, intrinsic small bowel disease, disorders of the biliary tract, and surgical bypass procedures of the jejunum and ileum can all cause problems

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

239

FIGURE 1

Six steps in vitamin D and calcium absorption and handling that may be altered by hepato-gastrointestinal disorders and so lead to bone disease. (1) Decreased intake of vitamin D and calcium. (2) Decreased absorption of vitamin D secondary to disorders in biliary secretion, pancreatic enzymes, enterocyte function, or intestinal anatomy. (3) Abnormal production of 25OHD by the liver secondary to hepatic parenchymal disease. (4) Disruption in the enterohepatic circulation of vitamin D metabolites and conjugates secondary to disorders in biliary secretion. (5) Reduced delivery of vitamin D metabolites to target tissues secondary to decreased DBP and albumin synthesis. (6) Decreased response of the diseased intestine to 1,25(OH)2D with respect to Ca and P absorption.

here. Third, vitamin D that enters the body must be further metabolized to active metabolites. Diseases of the liver, where the first step in bioactivation takes place, or drugs such as phenytoin, which alter this first metabolic step, lead to deficiency of the active metabolites. Fourth, the vitamin D metabolites undergo an enterohepatic circulation, being secreted in bile in conjugated form with subsequent reabsorption in the small intestine. Disruption of this pathway may contribute to vitamin D deficiency in certain diseases of the liver and small intestine. Fifth, vitamin D and its metabolites are poorly soluble in water and must be transported in blood bound to proteins, vitamin D binding protein (DBP) and albumin, which are synthesized in the liver. Decreased synthesis of these proteins may impair the delivery of the vitamin D metabolites to the target tissues. Finally, the diseased or surgically altered intestine may fail

to respond normally to the active vitamin D metabolites with respect to calcium and phosphate absorption. Clearly one disease may impact adversely on bone by several mechanisms involving aberrations in vitamin D and calcium absorption, metabolism, or function, and the systemic effects of the disease or its treatment may aggravate the abnormalities in vitamin D and calcium absorption, metabolism, or function. For example, chronic illness may limit the ability of the patient to get outdoors into the sunlight, thus decreasing the epidermal production of vitamin D. If this patient were also intolerant of milk products or had a condition in which malabsorption of calcium and vitamin D were present the stage would be set for vitamin D deficiency. Glucocorticoid therapy is used for a number of conditions discussed in this chapter. Glucocorticoid therapy by itself leads to osteoporosis, and when used to treat a

240

DANIEL D. BIKLE

disease which has already jeopardized skeletal integrity because of an abnormality in calcium and vitamin D absorption, such therapy can be especially detrimental. Hypogonadism may accompany gastrointestinal, hepatic, and pancreatic diseases and potentiate their deleterious actions on bone. In this chapter, specific disease entities will be discussed individually even though different diseases may impact on bone by similar or identical mechanisms.

II. GASTROINTESTINAL DISEASES Gastrointestinal diseases lead to abnormalities in bone primarily through the malabsorption of vitamin D and calcium, although the presence of disease may itself lead to reduced intake of vitamin D and calcium or limited exposure to sunlight. Each disease discussed below has its own subtle variations on this prevailing theme (Table 1).

A. Postgastrectomy Bone Disease 1. INCIDENCE AND PREVALENCE In a large (9704 subjects) study of older women, gastrectomy correlated with an 8.2% decrease in bone density

TABLE 1 Prevalence

[57]. In another study comparing 342 postgastrectomy patients with 180 unoperated patients of similar age with peptic ulcer disease, Eddy [58] observed osteopenia of the spine in 24% of the postgastrectomy patients compared to 4% of the unoperated controls. Pseudofractures and fractures were found in 2.4 and 5.2% of gastrectomized patients, respectively, whereas none were seen in the controls. Bone pain or tenderness was observed in 26% of the gastrectomized patients compared to 4% of controls. Bone biopsies of 84 gastrectomized patients showed widened osteoid seams in 32% compared to none of the 9 controls biopsied. In a more recent series, Mellstrom et al. [59] reported spinal fractures in 19% of males with partial gastrectomies compared to 4% of age matched controls. Smokers were at particularly high risk. An even higher prevalence of spinal osteopenia (69% of females, 41% of males) was observed by Deller et al. [60,61] in a study of 100 unselected patients following partial gastrectomy; however, 41% of females and 13% males in the age-matched unoperated control group with peptic ulcer disease also had spinal osteopenia. Twelve of 20 patients selected for bone biopsy because of the severity of their bone disease had increased osteoid seam width (12 mm), and 17 of 20 had decreased trabecular bone volume. In contrast to the high prevalence of bone disease among gastrectomized patients in these

Bone Disease Associated with Gastrointestinal Disorders Clinical features

Pathogenesis

Treatment

1. Postgastrectomy Up to 70%, increase with age

Older adults, females males, bone pain common; p Ca, P, 25(OH)D, q alk. phosphatase, normal PTH, p urine Ca; osteopenia by x-ray; q osteoid, p TBV

p Ca, vit. D absorption, ?20 to p acid, duodenal bypass, increased motility, steatorrhea

Vitamin D or calcifediol, calcium

Children, younger adults, growth retardation, steatorrhea, response to gluten-free diet; p 25(OH)D, q alk. phosphatase, may be occult

p Ca, vit. D absorption 20 abnormal enterocyte function from gliadin toxicity

Gluten free diet

Younger adults, steatorrhea, frequent ileal resection, frequent glucocorticoid use, p 25(OH)D; Ca, P, alk. phosphatase often normal; osteopenia by x-ray; q osteoid, p TBV

p Ca, vit. D absorption 20 disruption of enterohepatic circulation and abnormal jejunal-ileal function; glucocorticoid use

Vitamin D or calcifediol, calcium, minimize steroids and bile acid binders

p Ca, Mg, albumin, p 25(OH)D, q alk. phosphatase, PTH; osteopenia uncommon; q osteoid, p bone formation

p Ca, vit. D absorption 20 bypass of distal small intestine

Vitamin D or calcifediol, calcium, restore normal anatomy

Associated cholestasis or alcohol abuse, steatorrhea

May be due to associated conditions more than to fat malabsorption per se

Vitamin D, enzyme replacement

2. Celiac disease Up to 80% if untreated

3. Crohn’s disease 30%

4. Jejuno-ileal bypass Up to 60%

5. Pancreatic insufficiency Uncommon unless accompanied by other dsorders

TBV, total bone volume; 20, secondary

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

studies, two British studies showed a lower prevalence. In a survey of 1228 patients following partial gastectomy, Morgan et al. [62,63] identified only 6 who had symptoms, biochemical features, and bone biopsy evidence of bone disease. However, these investigators were looking for osteomalacia, not osteoporosis. Tovey et al. [64,65] found evidence for osteomalacia (increased osteoid, decreased calcification front) in only 10 of 227 postgastrectomy patients followed over a 25-year period (only 23 patients were selected for biopsy, however). Most patients identified as having osteomalacia were females. In contrast, 22% of the males and up to 86% of the females (percentage increased with age) had radiologic evidence of osteopenia. Thus, osteopenia appears to be quite common in patients following partial gastrectomy especially as they age, although frank osteomalacia is much less frequently seen. Females in particular are predisposed to developing bone disease following gastrectomy. 2. CLINICAL FEATURES Peptic ulcer leading to gastrectomy is a problem primarily of middle-aged adults, and bone disease is not likely to develop until several years after the procedure. Therefore, the clinical presentation is often that of osteoporosis in the elderly. Distinguishing between the bone disease accompanying the aging process and that due to gastrectomy is not always obvious even with a bone biopsy unless the biopsy shows frank osteomalacia. In a recent study of 471 patients following operation for peptic ulcer disease, Melton et al. [65a] noted increased fracture risk but concluded that associated conditions including age, use of glucocorticoids, thyroid hormone replacement, and chronic anticoagulation could account for these findings. They found no relationship between the type of operation and fracture risk. Bone pain or tenderness is generally found in patients in whom osteomalacia is eventually diagnosed but is not a reliable indicator. Symptomatic patients often provide a history of at least modest fat malabsorption [60,62,66,67] and milk intolerance [68]. Routine laboratory assessment of patients following partial gastrectomy reveals a reduction in serum calcium and phosphate concentrations (generally within the normal range) and an increase in alkaline phosphatase activity and osteocalcin in 10 – 25% [59,61,62]. Urinary calcium excretion tends to be low and phosphate clearance increased [61]. PTH values may be elevated [68a] but are normal in most patients [59,69 – 71], although urinary cAMP levels may be increased [72], suggesting mild hyperparathyroidism. 25OHD levels tend to be reduced in most studies [59,70 – 75]. The significance of this is obscured by concomitant reductions in the transport proteins of the vitamin D metabolites, vitamin D binding protein and albumin [73,75,76], and no correlation between 25OHD level and bone disease has been established [64]. However, 24,25(OH)2D concentrations are even further reduced

241

[69,71], while 1,25(OH)2D levels are normal or slightly increased [68a,69 – 71]. This profile suggests mild hyperparathyroidism secondary to early vitamin D and calcium deficiency. As described above, radiologic assessment demonstrates the frequent association of osteopenia and spinal fractures with gastrectomy [58 – 60,65,65a]. Pseudofractures and fractures of the hip [77] are less common but occur more frequently in these patients than in aged matched controls [58,66,78]. The type of bone disease reported on bone biopsy varies widely from study to study, reflecting the often subtle nature of the findings and the lack of tetracycline labeling in earlier reports to determine the dynamic parameters associated with bone formation. Osteoporosis, as indicated by reduced trabecular bone volume, and osteomalacia, as indicated by increased osteoid volume and a reduced calcification front, frequently coexist [58,61,67,79]. A more recent report [69] using double-label tetracycline found normal mineral apposition rates, normal mineralization lag time, and slightly increased bone formation rates along with increased osteoid volume in 16 asymptomatic patients with partial gastrectomies. This suggests that most “osteomalacia” diagnosed in other studies not using tetracycline labeling may represent early vitamin D deficiency and/or secondary hyperparathyroidism. Some patients have increased marrow fibrosis and osteoclast numbers, clearly indicating that secondary hyperparathyroidism occurs at least occasionally [66]. 3. PATHOGENESIS Absorption of both vitamin D (or 25OHD) [80 – 82] and calcium [68,70,83,84] is reduced in postgastrectomy patients, especially those who have evidence of bone disease. Such patients tend to have mild degrees of fat malabsorption [58,60,80] in the absence of small bowel disease [58]. Milk intolerance contributes to the reduced oral intake of vitamin D and calcium in at least some patients [85]. Normal calcium absorption has been thought to require the acid environment of the stomach to solubilize calcium salts prior to their absorption in the small intestine. Thus, procedures which reduce acid output would reduce calcium absorption. However, this concept has been tested [86,87], and it appears that gastric acid is not required for normal calcium absorption. The duodenum plays an important role in the vitamin D-regulated absorption of calcium [1], so that with duodenal bypass as in the Billroth II procedure, calcium absorption is likely to be reduced. Fat malabsorption would be expected to reduce calcium absorption both directly by the formation of calcium complexes as well as indirectly by the accompanying malabsorption of vitamin D. Inadequate mixing of bile and pancreatic enzymes with the luminal contents, which could occur in duodenal isolation procedures such as Billroth-II operations, would be expected to decrease absorption of fat soluble vitamins such

242

DANIEL D. BIKLE

as vitamin D. However, evidence supporting a greater incidence of low 25OHD levels or bone disease in general in subjects with Billroth-II procedures than in those with other surgical procedures is not strong [65,65a,70]. 4. TREATMENT The osteomalacic component of the bone disease responds to vitamin D and calcium supplements [63], but the osteoporotic component may not [88,89]. Distinguishing between these two components without a bone biopsy is difficult, and unless a biopsy is obtained to exclude osteomalacia, a clinical trial with vitamin D and calcium is indicated especially if the serum 25OHD concentration is low. Since malabsorption of these substances variably accompanies gastrectomy, the amount of either agent required to correct the deficiency will vary and needs to be individually established. Serum 25OHD concentrations are good indicators of vitamin D absorption adequacy. If steatorrhea is a major problem, calcifediol (25OHD) rather than vitamin D may be preferable because it is better absorbed under these conditions [50,51]. Since renal function is normal in most patients, treatment with calcitriol (1,25(OH)2D) is seldom indicated.

B. Celiac Disease 1. INCIDENCE AND PREVALENCE The prevalence of bone disease among patients with celiac disease depends on the age at which the diagnosis was made and treatment with a gluten-free diet started. Most untreated adults have reduced bone mineral density at time of diagnosis [90–93c]. Similarly, 60% of children have bone growth retardation at the time of diagnosis [94]. Those who fail to respond to a gluten free diet with improved intestinal morphology continue to have decreased bone density in comparison to their successfully treated peers [95]. Treatment started after childhood, because of delayed recognition of the disease, may be less successful and lead to persistent osteopenia [92,96]. Lindh et al. [97] found that 11 of 92 individuals with osteoporosis had IgA antibodies to gliadin compared to 3% of age matched controls, suggesting that occult, untreated celiac disease contributes to the development of osteoporosis in a substantial portion of the population. Of six that underwent jejunal biopsy, villous atrophy was found in three. The bone disease associated with celiac disease can present as osteoporosis or osteomalacia or both [98–101]. A recent survey found osteomalacia in 3 and osteoporosis in 32 of 56 patients with celiac disease [102]. 2. CLINICAL FEATURES The finding of bone disease in patients with celiac disease is usually made in association with malabsorption, although steatorrhea may be occult [98] or absent [93a,99,

102a]. The upper small intestine is usually more affected than the ileum. Untreated patients tend to have reduced serum and urine calcium levels and elevated values for serum alkaline phosphatase, PTH, and urine hydroxyproline [92,93a,96,101,102a]. Net calcium absorption can be reduced, but this has been ascribed to increased endogenous fecal calcium losses [101]. Of nine untreated patients with celiac disease Dibble et al. [73] found two with low (5 ng/ml) 25OHD concentrations. A higher (3/7) prevalence of low 25OHD levels was found by Arnaud et al. [103] using a higher value of 25OHD as the lower limit of normal. Bone mineral density (BMD) measurements have been found to correlate with circulating 25OHD levels [93a]. Serum 1,25(OH)2D levels tend to be elevated [102a]. With gluten-free diet these biochemical abnormalities improve even if BMD does not [73,96,101]. Radiologic evidence for osteopenia is common in untreated individuals, but if treatment is started in childhood and successfully reverses villous atrophy, peak bone mass can be achieved and maintained [95]. Spinal and rib fractures occur, but pseudofractures are uncommon even in individuals with osteomalacia on bone biopsy [101]. Studies incorporating bone biopsies have found evidence for both osteomalacia and osteoporosis [101,102]. Seven of nine untreated patients studied by Melvin et al. [101] had increased osteoid, three had in addition decreased calcification front, and none had decreased bone volume on bone biopsy. In contrast, a more recent study found that osteoporosis was more common than osteomalacia [102]. 3. PATHOGENESIS Vitamin D and calcium absorption are abnormal in untreated patients with celiac disease [47,101] as part of their general disorder in enterocyte function. Part of the increased fecal losses of calcium may result from increased endogenous secretion through the deranged epithelium [101]. 4. TREATMENT The treatment of choice is a gluten-free diet. This will correct the disorder in calcium metabolism in most cases. Vitamin D (or 25OHD) and calcium supplementation should be reserved for the individual on a gluten-free diet who fails to normalize serum calcium, phosphorus, alkaline phosphatase, and 25OHD levels or urine calcium excretion [104].

C. Inflammatory Bowel Syndromes 1. INCIDENCE AND PREVALENCE Of the two major forms of inflammatory bowel disease, Crohn’s disease and ulcerative colitis, severe bone disease is most frequently associated with Crohn’s disease especially when treated with ileal resection and glucocorticoids [105]. Twenty-three subjects of an unselected series of 75 patients

243

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

with inflammatory bowel disease had BMD 2 SD below normal, demonstrated by single photon absorptiometry of the radius or quantitative computed tomography of the spine [105,105a]. Six of these subjects had spinal fractures. Eighteen had ileal resections. In a separate study, Bernstein et al. [105a] found that corticosteroid use correlated better with the bone loss than did disease diagnosis (Crohn’s disease vs ulcerative colitis). In an earlier retrospective study of 700 patients with inflammatory bowel disease, osteoporosis was noted in only 3% [106], but sensitive methods to make the diagnosis were not used in that study. Osteomalacia may also be present [107]. The clinical syndrome of bone pain, weakness, elevated alkaline phsophatase activity, and radiologic features were found in 5% of subjects with Crohn’s disease [107]. Six of 9 patients studied by Driscoll et al. [108] and 9 of 25 patients studied by Compston et al. [109] had increased osteoid on bone biopsy despite the fact that most of these patients had few clinical features of osteomalacia. Adolescents have a high likelihood of developing osteopenia and retarded bone growth [110]. 2. CLINICAL FEATURES The clinical features of bone disease in patients with Crohn’s disease are usually subtle. Most patients are young adults with a variety of gastrointestinal and extragastrointestinal concerns that obscure symptoms of bone disease. Ileal resection, malabsorption, and glucocorticoid treatment are common and relevant to the bone disease that ensures. Routine serum biochemical measurements are generally normal but calcium, phosphorus, and magnesium may be low and the alkaline phosphatase activity may be high [108,111]. The level of alkaline phosphatase activity may correlate negatively with the degree of osteopenia [112]. However, serum 25OHD concentrations are reduced in up to 65% of patients [73,105,108], especially those who have undergone ileal resection [73,105]. Osteopenia is commonly observed both in cortical [105,111] and cancellous [105] bone, but less than 10% of patients will have fractures or pseudofractures [105,107,108]. Bone biopsy is the only means of diagnosing osteomalacia in most of these patients. Biopsies frequently show reduced trabecular bone volume and increased osteoid [105,110,111]. However, one study [113] using double tetracycline labeling failed to show a reduction in bone formation, mineral apposition, or mineralization lag time in 30 unselected patients despite decreased trabecular bone volume, suggesting that osteomalacia is less common than osteoporosis. These patients also had normal vitamin D metabolites, inactive disease, and were not on glucocorticoid therapy, so these results may not be applicable to sicker patients. 3. PATHOGENESIS Patients with Crohn’s disease have multiple reasons for developing bone disease. Vitamin D [50] and calcium [114]

absorption are reduced. Vitamin D is absorbed primarily in the jejunum and ileum via a process expedited by bile salts and inhibited by fat [115]. Therefore, disease or resection of this portion of the intestine will result in reduced vitamin D absorption. Concurrent use of cholestyramine or the development of hepatobiliary complications will reduce the availablity of bile salts for vitamin D absorption. Vitamin D metabolites undergoing enterohepatic circulation [52] cannot be reabsorbed by a diseased or resected ileum. Calcium malabsorption reflects both the state of vitamin D insufficiency and the steatorrhea. Low dietary intake of nutrients including milk products often compounds the problem of absorption. Glucocorticoid therapy is frequently used during active disease and can contribute to the calcium malabsorption and bone loss [105,105a,108]. 4. TREATMENT Vitamin D in doses of 4000 to 12,000 units (100 – 300 mg) per day is generally adequate therapy for patients with low serum 25OHD levels, although the appropriate dose must be determined for each patient [111]. Calcifediol (25OHD) at doses starting with 50 mg three times per week may be used instead of vitamin D, if vitamin D therapy is not effective. Calcifediol is better absorbed than vitamin D in patients with steatorrhea [50,51]. Dietary counseling to ensure that adequate calcium is also being ingested should be performed. Serum 25OHD and urine calcium levels provide good markers of treatment. Vitamin D treatment will reduce the osteomalacic component of the bone disease [108], but available data are not sufficient to determine whether the osteoporotic component will improve.

D. Jejuno – Ileal Bypass 1. INCIDENCE AND PREVALENCE Fortunately, the popularity of this procedure as treatment for massive obesity has waned because of the large number of undesirable side effects, of which bone disease is one. Initially following the operation, nearly all patients undergo at least a transient change in calcium homeostatic mechanisms [116]. Recovery occurs such that little or no osteopenia or osteoporosis can be appreciated by routine radiologic procedures [116 – 119]. However, osteomalacia and/or reduced trabecular bone volume on bone biopsy have been found in up to 60% of unselected individuals evaluated several years after bypass surgery [120 – 124]. 2. CLINICAL FEATURES Few individuals present with bone pain or fractures, and radiologic evidence of osteopenia is generally not found [116,119]. However, reductions in serum calcium and magnesium concentrations with an increase in serum alkaline phosphatase activity are observed in most subjects within

244

DANIEL D. BIKLE

3 months of the operation and persist in approximately half the patients for years [116 – 119,125,126]. Serum phosphrous levels are generally normal. At least part of the fall in serum calcium can be attributed to the fall in albumin concentration. Serum 25OHD levels, which tend to be reduced in the obese patient prior to operation, fall even further in most individuals [116 – 119,126]. 1,25(OH)2D levels are often normal but may also be reduced [117,121,126]. PTH levels tend to be increased [117,121,122,126]. Reductions in bone density are generally not found, and fracture rates do not appear to be increased. Osteomalacia is the dominant lesion found on bone biopsy [117 – 120], even when more restrictive criteria incorporating the results from double tetracycline labeling are used [121 – 123]. Clinical features do not predict the existence of bone disease on bone biopsy [120]. 3. PATHOGENESIS The ileum and much of the jejunum are effectively cut off from the flow of nutrients by this procedure. This region of the intestine is responsible for the absorption of vitamin D and the metabolites of vitamin D secreted in the bile. Malabsorption of vitamin D [117], calcium [116,125], and magnesium [127] have all been demonstrated. Fatty infiltration of liver is often found, but the degree to which liver disease contributes to the bone disease is not clear. 4. TREATMENT Vitamin D and the analog of 1,25(OH)2D, 1OHD, have been used successfully to treat the bone disease following jejuno-ileal bypass [128,129]. However, since 1,25(OH)2D levels are usually normal in these patients, vitamin D or calcifediol (25OHD) should be tried first using doses which normalize serum 25OHD levels. Calcium supplementation to normalize urine calcium excretion is also indicated. Treatment failures may respond to antibiotic treatment of bacterial overgrowth in the bypassed segment. If these measures fail, reanastomosis of the bypassed segment may be required [130].

E. Pancreatic Insufficiency 1. INCIDENCE AND PREVALENCE Clinically significant bone disease in patients with isolated pancreatic insufficiency is unusual [131]. However, reduced bone density and a high rate of fractures have been found in studies of children and young adults with cystic fibrosis [132–135a]. These patients have multiple risk factors including poor nutrition, pancreatic insufficiency, reduced absorption of calcium and vitamin D, reduced physical activity, pulmonary disease, delayed and reduced production of sex steroids, use of corticosteroids, and increased circulating concentrations of osteoclast activating cytokines.

These studies demonstrate a progressive decrement in bone mass relative to age group as these individuals pass through puberty into young adulthood. Biliopancreatic bypass procedures for obesity are also accompanied by severe bone disease with evidence for osteomalacia on bone biopsy being found in 73% of 41 subjects 1–5 years after this procedure [136]. 2. CLINICAL FEATURES The clinical features of pancreatic insufficiency include diabetes mellitus and steatorrhea. Although diabetes mellitus could contribute to the reduction in bone mass, steatorrhea is the feature that should most affect vitamin D and calcium absorption. However, the link between steatorrhea and bone disease is not established for this condition. 25OHD and serum and urine calcium values are generally in the low or low normal range [133,134], although normal values have been found in some series [136]. The 25OHD concentrations are more likely to be reduced if the pancreatic disease is associated with cholestasis, small bowel, and/or liver disease [73]. The presence of bone disease in a patient with malabsorption thought to be secondary to pancreatic insuffi ciency should lead to a search for complicating features such as alcohol abuse, cholestasis, cirrhosis, or intrinsic small bowel involvement. Case reports have been published demonstrating ostemalacia in patients with cystic fibrosis each of whom had liver involvement [131,137]. Osteomalacia has been found in a high percentage of patients with biliopancreatic bypass [136]. Nevertheless, current data are insufficient to determine the extent to which the osteopenia observed radiologically in most cases of pancreatic insuffi ciency represents osteoporosis versus osteomalacia. 3. PATHOGENESIS The infrequency of serious bone disease in patients with pancreatic insufficiency and steatorrhea who do not also have other risk factors indicates that the steatorrhea resulting from decreased pancreatic enzyme secretion is not sufficient to cause major impairment of vitamin D and calcium absorption. However, the lack of pancreatic enzymes appears to be synergistic with disruption of bile secretion and/or intrinsic small bowel disease, producing bone disease when these other risk factors are present. 4. TREATMENT Patients with low 25OHD concentrations should be given sufficient amounts of vitamin D to restore their 25OHD levels to normal. The dose will vary from patient to patient. In patients with substantial amounts of steatorrhea, calcifediol (25OHD) may be required. The diet should be supplemented with enough calcium to raise urinary calcium excretion to above 150 mg/day (adults). These patients are likely to require pancreatic enzyme replacement and supplementation with other fat-soluble vitamins.

245

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

TABLE 2 Prevalence

Bone Disease Associated with Liver Disorders

Clinical features

Pathogenesis

Treatment

1. Primary biliary cirrhosis Up to 80%

10 females, bone pain, jaundice; p Ca, 25(OH)D, PTH; qq alk. phosphatase (liver); osteopenia by x-ray; Fxs uncommon, p TBV more common than q osteoid, generally low turnover osteoporosis

p Ca, P, vit. D absorption ? 20 p bile, q urinary losses of vit. D conjugates, 25 hydroxylation of vit. D intact

Osteomalacia responds to vit. D, osteoporosis does not

Patients often on glucocorticoids; bone disease usually asymptomatic; p 25(OH)D, osteopenia by x-ray; p TBV

Bone disease 20 to glucocorticoid use more than to liver disease

Ensure adequate nutrition, limit glucocorticoid dose

Back pain, fractures; p Ca, Mg, P, albumin, p 25(OH)D, q PTH; osteopenia, fractures by x-ray; p TBV, p bone formation ETOH abuse, increases with age

Poor diet, ETOH induced q urinary losses of Ca, Mg, possible direct toxic affects of ETOH on bone

Stop ETOH, q Ca, Mg, P in diet; vit. D if p 25(OH)D and osteomalacia on biopsy

2. Chronic active hepatitis 50%

3. Alcoholic cirrhosis Most alcoholics with 10 years

TBV, total bone volume; 20, secondary; ETOH, ethanol

III. HEPATIC DISEASES For this discussion on the relationship of liver disease to bone disease, four categories of liver disease will be considered: chronic cholestatic disease of which the most common is primary biliary cirrhosis, chronic active hepatitis, viral hepatitis, and alcoholic cirrhosis (Table 2). The major themes linking liver disease to bone are: (i) the ability of the liver to convert vitamin D to 25OHD; (ii) the role of the hepatically produced vitamin D transport proteins, albumin and DBP, in transport of the vitamin D metabolites to their target tissues; (iii) the degree to which the enterohepatic circulation of the vitamin D metabolites contributes to the maintenance of vitamin D metabolite levels; and (iv) the role of bile in promoting vitamin D and calcium absorption. Each type of liver disease has its own nuances which contribute to these major themes such as the use of bile acid binding resins in cholestatic diseases, the use of glucocorticoids and other immunosuppressives to treat chronic active hepatitis, and the direct skeletal toxicity of alcohol in subjects with alcoholic cirrhosis. The recent increase in liver transplantation procedures, the immunosuppression following which can accelerate bone loss at least initially, makes it important to prevent or treat these disorders at an early stage before substantial bone loss occurs.

A. Chronic Cholestatic Diseases 1. INCIDENCE AND PREVALENCE Of the various chronic cholestatic diseases, the bone disease in primary biliary cirrhosis (PBC) has been best studied. PBC is a disease primarily of middle-aged women, an age at which postmenopausal osteoporosis is

common and not readily distinguished from the osteoporosis of the liver disease. Both osteomalacia and osteoporosis occur in PBC, but estimates of the prevalence of these forms of bone disease vary widely. Some studies indicate that patients with PBC have primarily osteomalacia [138,139], others find mostly osteoporosis [140 – 143], others find a high prevalence of both osteomalacia and osteoporosis [144,145], and still others find little of either [146 – 149]. Although studies differ somewhat with respect to the severity of the disease in the study population or in the criteria used to diagnose the bone disease, these differences do not fully account for the differences in results. Newly diagnosed patients who have received no treatment and have mild cholestasis appear to have less bone disease than those with more severe liver disease of greater duration [73,147,150]. The likelihood of developing osteomalacia as well as or instead of osteoporosis may depend on the prevalence of osteomalacia in the population since the reports of osteomalacia in PBC tend to come from the United Kingdom and Scandanavian countries where osteomalacia is more likely to be found in the general population [151,152]. Since the more recent studies show the least amount of bone disease, it may be that with the heightened awareness of the potential for bone disease in patients with PBC more attention is being paid to nutritional factors which can prevent or forestall this complication. In contrast to PBC, biliary atresia is a disease of infants and children. Children with this condition, even if surgically corrected, have a high likelihood of developing rickets [153], which is readily treated with vitamin D. 2. CLINICAL FEATURES Patients with PBC are often asymptomatic, although bone pain is common in patients subsequently shown to

246 have osteoporosis or osteomalacia on bone biopsy [138,141]. Laboratory assessment tends to show normal or slightly reduced serum and urine calcium, low normal serum phosphorus, and normal serum magnesium levels. PTH concentrations may be low even in subjects with decreased circulating calcium and 25OHD [138,141,146,154, 155], although reports of elevated PTH can also be found [140,143, 148,156]. Alkaline phosphatase activity in serum is increased, but the source is the liver, not bone. Osteocalcin and urinary hydroxproline levels are normal [138,146, 156,157]. Serum 25OHD levels can be normal in asymptomatic patients but fall as the disease progresses [139,141, 158]. The 25OHD level is not a good predictor of bone disease, however [138,141].1,25(OH)2D concentrations are generally normal [154 – 157,159]. Subjects with PBC have an increased prevalence of fractures and decreased bone mineral density [141,146,150,160,161], although pseudofractures are rare. Children with biliary atresia often present with florid rickets [153,162]. Bone biopsy is required to make a definitive diagnosis of osteomalacia, a finding which was commonly described in the early reports from the United Kingdom and Scandanavia [138,151,152]. However, the most common lesion seen in more recent studies using double lable tetracycline is reduced trabecular bone volume with normal or low amounts of osteoid, reduced bone formation rates, and increased mineralization lag time — characteristics of low turnover osteoporosis [140 – 142,146,154]. High turnover osteoporosis has also been found in a subset of patients [157], which may account in part for the surprisingly rapid loss of bone seen by Matloff et al. [141] and Herlong et al. [142] during a 1-year follow-up period. 3. PATHOGENESIS Bone disease in PBC has several potential etiologies. Intestinal malabsorption of calcium [141,142,154,163], phosphate [164], and vitamin D [139,163,165] have all been demonstrated to occur. Vitamin D absorption is further impaired in patients treated with cholestyramine [166]. Although some reports [165,166] indicate that the hepatic hydroxylation of vitamin D to 25OHD is impaired, this does not appear to be a problem in most patients. 25OHD concentrations are readily increased with vitamin D therapy [139,167,168], although some patients have required 1,25(OH)2D to treat the osteomalacic component of the bone disease [169]. Disruption of the enterohepatic circulation of vitamin D metabolites with increased losses in the urine has been postulated to lead to vitamin D deficiency in PBC [52,53], but it has also been proposed that the lack of biliary secretion of 1,25(OH)2D accounts for the normal level of 1,25(OH)2D and decreased concentrations of PTH seen in many patients [56]. Finally, an abnormality in the bone forming cell itself, the osteoblast, has been postulated

DANIEL D. BIKLE

to account for the failure of adequate vitamin D and mineral levels to correct the reduction in bone formation seen in patients with PBC [140 – 142,146]. 4. TREATMENT When present, osteomalacia responds readily to vitamin D, calcifediol (25OHD), calcitriol (1,25[OH]2D), or 1OH vitamin D [145,169]. However, osteoporosis has not been successfully treated with vitamin D or its metabolites, and calcifediol may even be detrimental [141,142,169]. If the serum 25OHD concentration is reduced, vitamin D should be given to restore the level to normal. If vitamin D is ineffective, calcifediol should be tried. The use of vitamin D or calcifediol for patients with normal 25OHD values is not justified. The rationale for using calcitriol or 1OH vitamin D in the absence of renal failure is weak. Such therapy ignores the potential benefit of 24,25(OH)2D which is generally low in patients with low 25OHD concentrations and which is increased by vitamin D or calcifediol therapy. Supplementing this regimen with calcium has been shown to increase the effectiveness of vitamin D therapy [170].

B. Chronic Active Hepatitis 1. INCIDENCE AND PREVALENCE Patients with chronic active hepatitis may not have the same increased prevalence of fractures and decreased bone density as those with other forms of chronic liver disease unless they are treated with glucocorticoids [150,171]. Osteopenia of the distal radius or reduced trabecular bone volume on bone biopsy was found in 47% of patients with chronic active hepatitis treated with glucocorticoid hormones [172], but the prevalence of bone disease in the absence of such treatment is not established. 2. CLINICAL FEATURES The bone disease associated with chronic active hepatitis is usually asymptomatic. Many patients are treated with glucocorticoids which may account for much of the bone disease which manifests primarily as osteopenia or osteoporosis, although osteomalacia has been described [173]. Patients with chronic active hepatitis tend to have 25OHD concentrations below the normal range and comparable to those seen in patients with alcoholic cirrhosis or PBC [73,158]. The reduction in 25OHD is accompanied by a reduction in DBP levels [73], suggesting that the free 25OHD concentration may be normal in a number of patients whose total 25OHD values are low. No osteomalacia was observed by Stellon et al. [172] in bone biopsies from 36 patients with chronic active hepatitis, although reduced trabecular bone volume was frequently seen. In contrast Dibble et al. [114] reported the presence of osteomalacia in

247

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

the bone biopsies from two of seven patients with chronic active hepatitis.

demonstrated by increased level of bone markers and by bone histomorphometry [173b – 173d]. However, in these cases bone mass is increased, not decreased.

3. PATHOGENESIS The scarcity of data specific to the impact of this disease on bone mineral metabolism makes problematic the compilation of a pathogenetic mechanism for the bone disease in chronic active hepatitis. As for other liver diseases calcium and vitamin D deficiency secondary to malabsorption or impaired hepatic conversion of vitamin D to 25OHD may be implicated but appears not to play a major role. Glucocorticoid therapy is likely to be more important in the pathogenesis of the bone disease in these patients.

3. PATHOGENESIS The lack of data makes it difficult to formulate a well documented pathogenesis. However, the finding of increased bone turnover in this disease suggests that the chronic inflammation in the liver is accompanied by increased cytokines systemically that serve to stimulate bone cell activity. In general this leads to bone loss, but in those unusual cases with osteosclerosis this leads to bone gain. 4. TREATMENT

4. TREATMENT Limiting the use of glucocorticoid therapy, ensuring adequate nutrition, and encouraging sunlight exposure are recommended first steps. The role of vitamin D and calcium supplementation in patients requiring glucocorticoid therapy is not clear and cannot be recommended unless malabsorption of these substances is strongly suspected.

Treatment of the underlying hepatitis is likely to prove most efficacious. Bisphosphonate therapy may prove useful in reducing what appears to be increased bone resorption, but this has not yet been tested in clinical trials.

D. Alcoholic Cirrhosis 1. INCIDENCE AND PREVALENCE

C. Viral Hepatitis 1. INCIDENCE AND PREVALENCE Although hepatitis C is assuming almost epidemic proportions, there is surprisingly little study of the skeleton in this disorder. Anecdotally, we are seeing a number of young males with this disease in our osteoporosis clinic who have substantial reduction in their bone mass, but no large study has yet been reported to provide incidence and prevalence data. A rare syndrome in which osteosclerosis but not osteopenia accompanies hepatitis C has been reported. 2. CLINICAL FEATURES In one recent study [173a] 32 males mean age 58 with hepatitis B or C were evaluated. Cirrhosis on liver biopsy was documented in 25. No differences in biochemistries were found between the two forms of hepatitis. BMD of the lumbar spine and femoral neck correlated with disease severity (Child-Pugh score). 25OHD and PTH concentrations decreased with increasing severity of the liver disease while the marker of bone resorption, urine deoxypyridinoline cross links (D-pyr) increased. D-pyr correlated negatively with BMD. Insulin-like growth factor I (IGF-I) also decreased with increasing severity of the liver disease, and correlated with the fall in bone mineral density. These data suggest a high state of bone turnover leading to net bone loss. The reported cases of osteosclerosis associated with hepatitis C also demonstrated increased bone turnover

Alcohol-induced bone disease is not restricted to those individuals who develop cirrhosis. This subject will be covered in depth elsewhere (see Chapter 31). However, since alcohol abuse is a major cause of liver disease, and the cirrhosis contributes to the severity of the bone disease, this problem will also be reviewed here. Saville [174] was the first to call attention to the high prevalence of osteopenia in alcoholics in his study of bone biopsies from cadavers in the New York City morgue. Spinal osteopenia may be observed in 50% of ambulatory male alcoholics by routine Xray procedures [175], and fractures of ribs or vertebrae occur in nearly 30% of this population [176,177]. Caucasians may be more susceptible to alcohol-induced bone disease than Blacks [177a]. This prevalence of fractures is much higher than in other types of liver disease [176]. The likelihood of developing a fracture increases rapidly beyond age 45 [178]. Partial gastrectomy increases the likelihood of developing osteopenia and fractures [179,180]. Bone densitometry and bone biopsy have demonstrated osteopenia in most patients with a prolonged history of heavy alcohol abuse [181 – 183]. Osteoporosis is the disease usually found histologically [181 – 185], although osteomalacia does occur [186,187] and may be more likely in patients who have had a partial gastrectomy [188]. 2. CLINICAL FEATURES Alcoholism can be a subtle disease and may be undetected unless and even if the patient is specifically questioned about alcohol intake. The presentation is often

248

DANIEL D. BIKLE

that of idiopathic osteoporosis, discovered by chance on radiologic assessment for low back pain or pulmonary complaints. Aseptic necrosis of the hip is associated with alcoholism, but the incidence of this disease in alcoholics is low [187,189,190]. Serum concentrations of calcium, phosphorus, and magnesium tend to be low normal in ambulatory alcoholics [181,182,185]. However, following a binge or when other alcohol-related medical problems are serious enough to require hospitalization, serum levels of these minerals can be sufficiently reduced to cause neuromuscular disturbances and rhabdomyolysis [191]. Part of the reduction in serum calcium is accounted for by a reduction in serum albumin concentration. Serum PTH and urinary cAMP levels may be elevated or high normal in part because of the lowered calcium and magnesium levels [181,185,192], although acute administration of alcohol can lower the PTH level [193]. 25OHD levels are usually low [138,156,181,186,192] and correlate with the low albumin and DBP concentrations [194,195]. 1,25(OH)2D concentrations have been variably reported as low [186,196], normal [181], or high [192]. Low levels of 1,25(OH)2D are found in the alcoholics with the severest liver disease, and like 25OHD levels, correlate with serum albumin and DBP [196]. The free or unbound concentrations of the vitamin D metabolites are generally normal [195,196] (Table 3). It is now appreciated that much of the reduction in the total concentrations of the vitamin D metabolites is a direct result of the reduction in the circulating levels of the carrier proteins [76,197] (Fig. 2). The radiologic assessment of bone reveals osteopenia or osteoporosis. Cancellous bone is more affected than cortical bone. Fractures are common, often following minimal trauma, but pseudofractures are rare in this population. The bone biopsy usually reveals reduced trabecular bone vol-

TABLE 3 Total and Free Vitamin D Metabolite Levels in Subjects with Alcoholic Liver Disease Liver disease

Normal

Total (ng/ml)

10.9  9.5*

19.2  6.6

Free (pg/ml)

6.6  4.6

5.9  2.3 41.5  11.5

25(OH)D

1,25(OH)2D Total (pg/ml)

22.6  12.5*

Free (fg/ml)

209  91

174  46

DBP (g/ml)

188  105*

404  124

Albumin (g/dl)

2.8  0.7*

4.5  0.2

*Significantly lower than normal. Data taken from Bikle et al., J. Clin. Invest. 74, 1966 – 1971 (1984) and J. Clin. Invest. 78, 748 – 752 (1986).

ume with normal or decreased amounts of osteoid [178,181,185], although a few patients will have increased osteoid volume [186,187]. Marrow fibrosis is uncommon. Bone formation and active bone resorption are generally reduced [181 – 187], although in younger patients high rates of bone turnover may be observed [178] (Fig. 3). 3. PATHOGENESIS The original reports of low 25OHD levels in alcoholics led to suggestions that poor nutrition [198], decreased sunlight [199], vitamin D malabsorption [200], or defective hydroxylation of vitamin D to 25OHD [194,201] might be involved in alcoholic bone disease. Vitamin D deficiency, therefore, could account for the osteomalacia seen in some alcoholic patients. Hypophosphatemia due to poor intake, malabsorption, concomitant use of aluminum-containing antacids, or increased renal excretion [202] could enhance the mineralization defect. However, the infrequency of osteomalacia [181], the finding of normal free levels of the vitamin D metabolites [195], and the realization that the low total concentrations of the vitamin D metabolites reflect decreased hepatic production of DBP and albumin not decreased hepatic production of 25OHD [203 – 205] all indicate that for most individuals the bone disease is not one of vitamin D deficiency. Calcium deficiency from poor intake, malabsorption [206,207], or increased urinary excretion [207] could lead to osteoporosis especially if associated with secondary hyperparathyroidism [181,185,192]. Mild degrees of hypomagnesemia could aggravate this picture (very low magnesium levels cause hypoparathyroidism). However, evidence for hyperparathyroidism is seldom seen on bone biopsies; rather the picture is usually one of inactive bone at least in the older individual. Failure to explain the bone disease of alcoholics on the basis of changes in the calciotropic hormones has led to the hypothesis that the prime offender is alcohol or one of its metabolites such as acetaldehyde causing direct inhibition of bone cell activity [181,185,193]. 4. TREATMENT Cessation of alcohol consumption appears to arrest the progression of the bone disease, and may reverse it [208,209]. Vitamin D therapy should be considered if the 25OHD levels are lower than what would be expected for the reduction in albumin and DBP. Such therapy will reverse osteomalacia if present and may help to restore bone mass [185]. Vitamin D itself is effective in most subjects since malabsorption is usually not severe [204] and 25OHD production is usually intact [204,205]. If oral vitamin D therapy does not raise 25OHD levels, calcifediol (25OHD) can be used. Ensuring adequate nutrition including calcium, magnesium, and phosphate also is appropriate. However, the degree to which osteoporosis can be reversed with current therapeutic measures remains unclear.

249

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

FIGURE 2

Correlation of total 25OHD (A) and 1,25(OH)2D (C) levels to DBP; lack of correlation of free 25OHD (B) and 1,25(OH)2D (D) levels to DBP. Normal subjects (), subjects with liver disease (, ), and subjects on oral contraceptives (). In C and D, data from pregnant women () are also included. These data demonstrate the dependence of total 25OHD and 1,25(OH)2D concentrations on DBP levels which are reduced by liver disease. However, the free concentrations of 25OHD and 1,25(OH)2D are normal in most patients with liver disease. Reprinted with permission from the American Society for Clinical Investigation.

IV. BONE DISEASE COMPLICATING THERAPY In this section, two iatrogenic bone diseases arising as a result of efforts to treat disorders of the hepatogastrointestinal tract will be considered — that attending liver transplantation and that accompanying total parenteral nutrition (Table 4). In each case the treatment aggravates the preexisting bone disease accompanying the disorder for which

treatment is intended, and rapid onset of clinically significant bone disease occurs.

A. Liver Transplantation 1. INCIDENCE AND PREVALENCE As the survival of patients undergoing liver transplantation improves (currently approximately 75% 5-year

250

DANIEL D. BIKLE

developed new fractures. In the 68 non-PBC/PSC patients, bone loss was more gradual, and only 3 of these patients developed new fractures. Of the 12 patients who developed aseptic necrosis, 11 were in the PBC and PSC group. Most of the fractures occur in the first 6 months following transplantation [171,211]. The accelerated bone loss appears to be due to increased bone turnover as demonstrated histomorphometrically by Vedi et al. [211a]. 2. CLINICAL FEATURES

FIGURE 3

Correlation of active resorption surface (OcS/BS) and bone formation rate (BFR/BV) with age in alcoholics. The data for active resorption surface and bone formation rate are expressed a percentile of age-matched controls. The correlation of the active parameters of bone remodeling with age are negative and significant. Reprinted with permission from Williams and Wilkins.

survival) [210], the metabolic complications of this procedure become more important. Osteoporosis is one such complication (see Chapter 52). In a series of 146 patients surviving for at least 1 year, Porayko et al. [171] found accelerated bone loss in nearly all patients, but the degree of morbidity in the skeletal system depended on the underlying liver disease prior to transplantation. Patients with PBC and primary sclerosing cholangitis (PSC) had the highest prevalence of osteopenia and osteoporosis prior to surgery (54% of 78 subjects had spinal bone densities below the fracture threshold of 0.98 g/cm2, 5 of whom had fractures) compared to patients with chronic active hepatitis (n  44) and a miscellaneous group (n  24) (15% of whom had spinal bone densities below the fracture threshold, 1 of whom had a fracture). Following transplantation the patients with PBC and PSC showed accelerated bone loss (up to 30 times normal) for the first 3 months, and 29 (37%) of these patients TABLE 4 Prevalence

Many patients being considered for liver transplantation already have bone disease [211b]. Such patients are started on high doses of glucocorticoids (e.g., 200 mg/day prednisolone) and immunosuppressives such as cyclosporine and azathioprine. Following transplantation osteocalcin levels may increase [212,213] and vitamin D metabolite concentrations fall [213]. Bone mineral density falls rapidly [171,210,211]. Fractures and aseptic necrosis appear within months [171,211]. These fractures tend to occur primarily in the spine and ribs, although hip fractures are also observed, and a single patient may have several fractures in rapid succession [214]. Bone biopsy data show high turnover osteoporosis [211a]. 3. PATHOGENESIS The drugs used to prevent and treat rejection are almost certainly the cause of the rapid loss of bone. Patients with preexisting bone disease are particularly susceptible. Contributing factors include immobilization and wasting, which accompany any major surgical procedure. 4. TREATMENT Prior to transplantation all factors predisposing to bone disease should be corrected if possible, including calcium and vitamin D deficiencies. Newer immunosuppressive agents may permit lower doses of glucocorticoid hormones to be used, but it remains to be seen whether such changes will alter the rapidity and extent of bone loss following transplantation. Calcium and 1OHD did not prevent bone loss in one study [211]. Antiresorptive agents such as bisphosphonates may prove useful, but studies

Bone Disease Associated with Liver Transplantation and Total Parenteral Nutrition Clinical features

Pathogenesis

Treatment

1. Liver transplantation Nearly 100% in patients with cholestatic diseases; less in other liver diseases 2. TPN

Rapid onset of Fxs: ribs, spine; loss of BMD, 25(OH)D and 1,25(OH)2D may p

High dose glucocorticoids, immobilization

Correct vit. D and Ca deficiencies pre OP; reduce glucocorticoids if possible

Risk of osteomalacia decreasing with newer TPN solutions; osteopenia in 50%

Bone pain; q Ca, P, alk. phosphatase, normal 25(OH)D, p 1,25(OH)2D; p TBV, q osteoid, p bone formation

Aluminum contamination in casein hydrolysate and other components of TPN solution

Discontinue TPN, substitute purified amino acids for casein hydrolysate

TPN, total parenterol nutrition; TBV, total bone volume; pre OP, preoperatively

251

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

employing these agents for this condition have not been reported.

B. Total Parenteral Nutrition 1. INCIDENCE AND PREVALENCE At the time the relationship between total parenteral nutrition (TPN) and bone disease was first described [215,216], 20 – 30% of patients on long-term TPN complained of bone pain often occurring within 1 year of beginning treatment. Infants similarly treated presented with radiologic and biochemical evidence of rickets [217]. Most of the patients who were biopsied showed evidence of osteomalacia [215,216] regardless of the presence of bone pain. In these early studies, casein hydrolysate was used as the source of amino acids. Casein hydrolysate was subsequently shown to contain high concentrations of aluminum [218], a contaminant strongly implicated in the osteomalacia associated with hemodialysis and which was found in high concentrations in the bone of patients on long-term TPN [219]. Substituting purified amino acids for casein hydrolysate has markedly reduced the incidence of bone pain and prevalence of osteomalacia on bone biopsy, although osteopenia still occurs and may affect approximately 50% of patients on long-term therapy [220 – 222]. Although TPN appears to result in progressive loss of bone, patients requiring TPN often have preexisting bone disease [223]. 2. CLINICAL FEATURES When initially described, TPN-induced bone disease resulted in severe bone pain primarily affecting the lower extremities, lower back, and ribs. Some patients could not walk as a result of the pain. These symptoms resolved when TPN was discontinued. This clinical picture is seldom seen today with newer formulations of TPN solutions. In the initial studies [215,216,224,225], serum calcium, alkaline phosphatase, and phosphorus levels were elevated. At least part of the increased alkaline phosphatase was hepatic in origin as other liver function tests were abnormal [226]. Hypercalciuria exceeding the infused amount of calcium was observed. PTH and 1,25(OH)2D levels were low despite normal levels of 25OHD and adequate amounts of vitamin D in the TPN solution. Substituting purified amino acids for casein hydrolysate [220,221] resulted in lower serum calcium levels and normal serum phosphorus, PTH, 25OHD, and 1,25(OH)2D concentrations. Alkaline phosphatase activity continued to be elevated in these patients and osteopenia was still found radiologically [223]. In the original reports bone biopsies showed reduced trabecular bone volume, increased osteoid, and decreased mineralization characteristic of osteomalacia [215,216,218,219]. More recent reports of patients on TPN supplemented with purified amino acids rather than casein hydrolysate show

normal levels of osteoid and normal bone formation rates, although reduced trabecular bone volume is still seen [221,222]. 3. PATHOGENESIS In the original studies by Shike et al. [216,225], vitamin D itself was implicated in the genesis of the bone disease, although the mechanism for this was obscure. This explanation has given way to the hypothesis that aluminum is the likely culprit for many of the abnormalities. Aluminum contaminates not only casein hydrolysate, but also albumin, phosphate, and calcium solutions [227,228]. However, casein hydrolysate appears to be the major source of aluminum contamination, and replacing this with purified amino acids has resulted in a marked reduction in aluminum concentrations in the blood, urine, and bone of patients receiving TPN [218]. Changing from casein hydrolysate to purified amino acids has reduced the amount of clinically evident bone disease and altered the morphologic picture from osteomalacia to osteopenia. The reduction in aluminum has also corrected the low levels of PTH and 1,25(OH)2D and improved the hypercalciuria which characterized the original syndrome. The reasons for the persistence of the bone disease in patients receiving the newer formulations of TPN are not yet clear. At least some of the patients have bone disease before they begin TPN because of the underlying gastrointestinal disorder which leads them to require TPN. Furthermore, the amino acids in the TPN solution may induce hypercalciuria and subtle hyperparathyroidism if the infused amounts are high [229]. 4. TREATMENT Discontinuing TPN, when feasible, may correct the bone disease [215]. Adjusting the vitamin D, amino acid, and calcium concentration to achieve a positive calcium balance needs to be done in those patients who cannot discontinue TPN. Reducing the aluminum contamination of the solutions to the lowest possible level has proven to be of great importance. Other trace contaminants or deficiencies which impact on the skeleton may be found in the future.

References 1. D. D. Bikle, R. L. Morrissey, D. T. Zolock, and H. Rasmussen, The intestinal response to vitamin D. Rev. Physiol. Biochem. Pharmacol. 89, 63 – 143 (1981). 2. R. P. Heaney and T. G. Skillman, Secretion and excretion of calcium by the human gastointestinal tract. J. Lab. Clin. Med. 64, 29 – 41 (1964). 3. R. P. Heaney, Calcium in the prevention and treatment of osteoporosis. J. Intern. Med. 231, 169 – 180 (1992). 4. C. Hasling, P. Charles, F. Taagehoj Jensen, and L. Mosekilde, Calcium metabolism in postmenopausal osteoporosis: The influence of dietary calcium and net absorbed calcium. J. Bone Miner. Res. 5, 939 – 946 (1990).

252 5. J. R. Bullamore, J. C. Gallagher, R. Wilkinson, B. E. C. Nordin, and D. H. Marshall, Effect of age on calcium absorption. Lancet 11, 535 – 537 (1970). 6. R. G. Crilly, R. M. Francis, and B. E. C. Nordin, Steroid hormones, aging and bone. Clin. Endocrinol. Metab. 10, 115 – 139 (1981). 7. P. Ireland and J. S. Fordtran, Effect of dietary calcium on jejunal calcium absorption in humans studied by intestinal perfusion. J. Clin. Invest. 52, 2672 – 2681 (1973). 8. J. C. Gallagher, B. L. Riggs, J. Eisman, A. Hamstra, S. B. Arnaud, and H. F. DeLuca, Intestinal absorption and vitamin D metabolites in normal subjects and osteoporotic patients: Effect of age and dietary calcium. J. Clin. Invest. 64, 729 – 736 (1979). 9. R. P. Heaney, R. R. Recker, M. R. Stegman, and A. J. Moy, Calcium absorption in women: Relationships to calcium intake, estrogen status, and age. J. Bone Miner. Res. 4, 469 – 475 (1989). 10. A. Caniggia, C. Gennari, and V. Bianchi, Intestinal absorption of 45 Ca in senile osteoporosis. Acta Med. Scand. 173, 613 – 617 (1963). 11. J. C. Gallagher, B. L. Riggs, and H. F. DeLuca, Effect of estrogen on calcium absorption and serum vitamin D metabolites in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 51, 1359 – 1364 (1980). 12. B. L. Riggs and K. I. Nelson, Effect of long term treatment with calcitriol on calcium absorption and mineral metabolism in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 61, 457 – 461 (1985). 13. R. Eastell, A. L. Yergey, N. E. Vieira, S. L. Cedel, R. Kumar, and B. L. Riggs, Interrelationship among vitamin D metabolism, true calcium absorption, parathyroid function, and age in women: Evidence of an age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J. Bone Miner. Res. 6(2), 125 – 132 (1991). 14. R. P. Heaney, R. R. Recker, and P. D. Saville, Calcium balance and calcium requirements in middle-aged women. Am. J. Clin. Nutr. 30, 1603 – 1611 (1977). 15. C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D3 in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). 16. A. Caniggia, R. Nuti, F. Lore, G. Martini, V. Turchetti, and G. Righi, Long-term treatment with calcitriol in postmenopausal osteoporosis. Metabolism 4(1), 43 – 49 (1990). 17. J. E. Zerwekh, K. Sakhaee, K. Glass, and C. Y. C. Pak, Long term 25-hydroxyvitamin D3 therapy in postmenopausal osteoporosis: Demonstration of responsive and nonresponsive subgroups. J. Clin. Endocrinol. Metab. 56, 410 – 413 (1983). 18. J. L. Stock, J. A. Coderre, and L. E. Mallette, Effects of a short course of estrogen on mineral metabolism in postmenopausal women. J. Clin. Endocrinol. Metab. 61, 595 – 600 (1985). 19. D. D. Bikle, B. P. Halloran, S. T. Harris, and A. A. Portale, Progestin antagonism of estrogen stimulated 1,25-dihydroxyvitamin D levels. J. Clin. Endocrinol. Metab. 75, 519 – 523 (1992). 20. C. Cheema, B. F. Grant, and R. Marcus, Effects of estrogen on circulating “free” and total 1,25-dihydroxyvitamin D and on the parathyroid-vitamin D axis in postmenopausal women. J. Clin. Invest. 83, 537 – 542 (1989). 21. B. H. Arjmandi, M. A. Salih, D. C. Herbert, S. H. Sims, and D. N. Kalu, Evidence for estrogen receptor-linked calcium transport in the intestine. Bone Miner. 21, 63 – 74 (1993). 22. R. M. Francis, M. Peacock, G. A. Taylor, G. H. Storen, and B. E. C. Nordin Calcium malabsorption in elderly women with vertebral fractures. Evidence for resistance to the action of vitamin D metabolites on the bowel. Clin. Sci. 66, 103 – 107 (1984). 23. P. R. Ebeling, M. E. Sandgren, E. P. DiMagno, A. W. Lane, H. F. DeLuca, and B. L. Riggs, Evidence of an age-related decrease in intestinal responsiveness to vitamin D: Relationship between serum 1,25-dihydroxyvitamin D3 and intestinal vitamin D receptor concen-

DANIEL D. BIKLE

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35. 36.

37.

38.

39.

40.

41.

42.

43.

trations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). A. Kobayashi, S. Kawai, Y. Ohbe, and Y. Nagashima, Effects of dietary lactose and a lactose preparation on the intestinal absorption of calcium and magnesium in normal infants. Am. J. Clin. Nutr. 28, 681 – 683 (1975). E. E. Ziegler and S. J. Fomon, Lactose enhances mineral absorption in infancy. J. Pediatr. Gastroenterol. Nutr. 2, 288 – 294 (1983). S. J. Birge, H. T. Kentmann, P. Cuatrecasas, and G. D. Whedon, Osteoporosis, intestinal lactase deficiency and low dietary calcium intake. N. Engl. J. Med. 276, 445 – 448 (1967). A. D. Newcomer, S. F. Hodgson, D. B. McGill, and P. J. Thomas, Lactase deficiency: Prevalence in osteoporosis. Ann. Intern. Med. 89, 218 – 220 (1978). M. Horowitz, J. Wishart, L. Mundy, and B. E. C. Nordin, Lactose and calcium absorption in postmenopausal osteoprosis. Arch. Intern. Med. 147, 534 – 536 (1987). J. Kocian, I. Skala, and K. Bakos, Calcium absorption from milk and lactose-free milk in healthy subjects and patients with lactose intolerance. Digestion 9, 317 – 324 (1973). B. Cochet, A. Jung, M. Griessen, P. Bartholdi, P. Schaller, and A. Donath, Effects of lactose on intestinal calcium absorption in normal and lactase deficient subjects. Gastroentorology 48, 935 – 940 (1983). M. Griessen, B. Cochet, F. Infante, A. Jung, P. Bartholdi, A. Donath, E. Loizeau, and B. Courvoisier, Calcium absorption from milk in lactase-deficient subjects. Am. J. Clin. Nutr. 49, 377 – 384 (1989). R. P. Heaney and R. R. Recker, Effects of nitrogen, phosphorus, and caffeine on calcium balance in women. J. Lab. Clin. Med. 99, 46 – 55 (1982). H. Spencer, L. Kramer, D. Osis, and C. Norris, Effects of phosphorus on the absorption of calcium and on the calcium balance in man. J. Nutr. 108, 447 – 457 (1978). M. Hegsted, S. A. Schuette, M. B. Zemel, and H. M. Linkswiler, Urinary calcium and calcium balance in young men as affected by level of protein and phosphorus intake. J. Nutr. 111, 553 – 562 (1981). M. S. Calvo, Dietary phosphorus, calcium metabolism and bone. J. Nutr. 123, 1627 – 1633 (1993). E. Reiss, J. M. Canterbury, M. A. Bercovitz, and E. L. Kaplan, The role of phosphate in the secretion of parathyroid hormone in man. J. Clin. Invest. 49, 2146 – 2149 (1970). S. J. Silverberg, E. Shane, T. L. Clemens, D. W. Dempster, G. V. Segre, R. Lindsay, and J. P. Bilezikian, The effect of oral phosphate administration on major indices of skeletal metabolism in normal subjects. J. Bone Miner. Res. 1, 383 – 388 (1986). S. J. Silverberg, E. Shane, L. de la Cruz, G. V. Segre, T. L. Clemens, and J. P. Bilezikian, Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D3 formation in women with osteoporosis. N. Engl. J. Med. 320, 277 – 281 (1989). A. A. Portale, B. P. Halloran, and R. C. Morris, Jr., Physiologic regulation of the serum concentration of 1,25-dihydroxyvitamin D by phosphorus in normal men. J. Clin. Invest. 83, 1494 – 1499 (1989). A. A. Portale, B. P. Halloran, M. M. Murphy, and R. C. Morris, Jr., Oral intake of phosphorus can determine the serum concentration of 1,25-dihydroxyvitamin D by determining its production rate in humans. J. Clin. Invest. 77, 7 – 12 (1986). R. A. McCance and E. M. Widdowson, Mineral metabolism of healthy adults on white and brown bread dietaries. J. Physiol. 101, 44 – 85 (1942). J. G. Reinhold, B. Faradji, P. Abadi, and F. Ismail-Beigi, Decreased absorption of calcium, magnesium, zinc and phosphorus by humans due to fiber and phosphorus consumption as wheat bread. J. Nutr. 106, 493 – 503 (1976). H. H. Sandstead, L. M. Klevay, R. A. Jacob, J. M. Munoz, G. M. Logan, and S. L. Reck, Effects of dietary fiber and protein level on

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58. 59.

60.

61.

62.

63.

mineral element metabolism. In (G. E. Inglett and S. I. Falkehaged, eds.), pp. 147 – 156. Academic Press, New York, 1979. J. L. Kelsey, K. M. Behall, and E. S. Prather, Effect of fiber from fruits and vegetables on metabolic responses on human subjects. 11. Calcium, magnesium, iron and silicon balances. Am. J. Clin. Nutr. 32, 1876 – 1880 (1979). D. Hollander, K. S. Muralidhara, and A. Zimmerman, Vitamin D-3 intestinal absorption in vivo: Influence of fatty acids, bile salts, and perfusate pH on absorption. Gut 19, 267 – 272 (1978). M. D. Sitrin, K. L. Pollack, M. J. G. Bolt, and I. H. Rosenberg, Comparison of vitamin D and 25-hydroxyvitamin D absorption in the rat. Am. J. Physiol. 242, G326 – G332 (1982). G. R. Thompson, B. Lewis, and C. C. Booth, Absorption of vitamin D3-3H in control subjects and patients with intestinal malabsorption. J. Clin. Invest. 45, 94 (1966). R. Blomstrand and L. Forsgren, Intestinal absorption and esterification of vitamin D3-1,2-3H in man. Acta Chem. Scand. 21, 1662 (1967). D. Hollander, E. Rim, and D. Morgan, Intestinal absorption of 25hydroxyvitamin D3 in unanesthetized rat. Am. J. Physiol. 236, E441 – E445 (1979). M. Davies, E. B. Mawer, and E. L. Krawitt, Comparative absorption of vitamin D3 and 25-hydroxyvitamin D3 in intestinal disease. Gut 21, 287 – 292 (1980). E. L. Krawitt and B. F. Chastenay, 25-hydroxyvitamin D absorption test in patients with gastrointestinal disorders. Calcif. Tissue Int. 32, 183 – 187 (1980). S. B. Arnaud, R. S. Goldsmith, P. W. Lambert, and V. L. W. Go, 25-hydroxyvitamin D3: Evidence of an enterohepatic circulation in man. Proc. Soc. Exp. Biol. Med. 149, 570 – 572 (1975). R. H. Wiesner, R. Kumar, E. Seeman, and V. L. W. Go, Enterohepatic physiology of 1,25-dihydroxyvitamin D3 metabolites in normal man. J. Lab. Clin. Med. 96, 1094 – 1100 (1980). R. Kumar, S. Nagubandi, V. R. Mattox, and J. M. Londowski, Enterohepatic physiology of 1,25-dihydroxyvitamin D3. J. Clin. Invest. 65, 277 – 284 (1980). L. V. Avioli, S. W. Lee, J. E. McDonald, J. Lund, and H. F. DeLuca, Metabolism of vitamin D3-3H in human subjects: Distribution in blood, bile jfeces, and urine. J. Clin. Invest. 46, 983 – 992 (1967). R. Kumar, Hepatic and intestinal osteodystrophy and the hepatobiliary metabolism of vitamin D. Ann. Intern. Med. 98, 662 – 663 (1983). D. C. Bauer, W. S. Browner, J. A. Cauley, E. S. Orwoll, J. C. Scott, D. M. Black, J. L. Tao, and S. R. Cummings, Factors associated with appendicular bone mass in older women. Ann. Intern. Med. 118, 657 – 665 (1993). R. L. Eddy, Metabolic bone disease after gastrectomy. Am. J. Med. 50, 442 – 449 (1971). D. Mellstrom, C. Johansson, O. Johnell, G. Lindstedt, P. A. Lundberg, K. Obrant, I-M. Schoon, G. Toss, and B-O. Ytterberg, Osteoporosis, metabolic aberrations, and increased risk for vertebral fractures after partial gastrectomy. Calcif. Tissue Int. 53, 370 – 377 (1993). D. J. Deller, M. D. Begley, R. G. Edwards, and M. Addison, Metabolic effects of partial gastrectomy with special reference to calcium and folic acid. Gut 5, 218 – 225 (1964). D. J. Deller, R. G. Edwards, and M. Addison, Calcium metabolism and the bone after partial gastrectomy: The nature and cause of the bone disorder. Aust. Ann. Med. 12, 295 – 309 (1963). D. B. Morgan, C. R. Paterson, C. G. Woods, C. N. Pulvertaft, and P. Fourman, Search for osteomalacia in 1228 patients after gastrectomy and other operations on the stomach. Lancet 2, 1085 – 1088 (1965). D. B. Morgan, C. R. Paterson, C. G. Woods, C. N. Pulvertaft, and P. Fourman, Osteomalacia after gastrectomy: A response to very small doses of vitamin D. Lancet 2, 1089 – 1091 (1965).

253

64. F. I. Tovey, D. G. Karamanolis, J. Godfrey, and C. G. Clark, Post-gastrectomy nutrition: Methods of outpatient screening for early osteomalacia. Hum. Nutr. Clin. Nutr. 39C, 439 – 446 (1985). 65. F. I. Tovey, J. E. Godfrey, and M. R. Lewin, A gastrectomy population: 25 – 30 Years on. Postgrad. Med. J. 66, 450 – 456 (1990). 65a. L. J. Melton III, C. S. Crowson, S. Khosla, and W. M. OFallon, Fracture risk after surgery for peptic ulcer disease: A populationbased cohort study: Bone 25, 61 – 67 (1999). 66. D. B. Morgan, G. Hunt, and C. R. Paterson, The osteomalacia syndrome after stomach operations. Q. J. Med. 39, 395 – 410 (1970). 67. R. Garrick, A. W. Ireland, and S. Posen, Bone abnormalities after gastric surgery: A prospective histological study. Ann. Intern. Med. 75, 221 – 225 (1971). 68. D. H. Gregory and R. Van Uelft, Calcium absorption following gastric resection. Am. J. Gastroenterol. 57, 34 – 40 (1972). 68a. M. Davies, S. E. Heys, P. L. Selby, J. L. Berry, and E. B. Mawer, Increased catabolism of 25-hydroxyvitamin D in patients with partial gastrectomy and elevated 1,25-dihydroxyvitamin D levels. Implications for metabolic bone disease, J. Clin. Endocrinol Metab. 82, 209 – 212 (1997). 69. K. B. Klein, E. S. Orwoll, D. A. Lieberman, D. E. Meier, M. R. McClung, and A. M. Parfitt, Metabolic bone disease in asymptomatic men after partial gastrectomy with billroth. II. Anastomosis. Gastroentorology 92, 608 – 616 (1987). 70. L. Nilas and C. Christiansen, Influence of PTH and 1,25(OH)2D on calcium homeostasis and bone mineral content after gastric surgery. Calcif. Tissue Int. 37, 461 – 466 (1985). 71. K. Kozawa, M. Imawari, H. Shimazu, O. Kobori, T. Osuga, and Y. Morioka, Vitamin D status after total gastrectomy. Dig. Dis. Sci. 29, 411 – 416 (1984). 72. H. von Lilienfeld-Toal, K. G. Mackes, G. Kodrat, H. Ochs, and A. Sonnenberg, Plasma 25-hydroxyvitamin D and urinary cyclic AMP in German patients with subtotal gastrectomy. Dig. Dis. Sci. 22, 633 – 636 (1977). 73. J. B. Dibble, P. Sheridan, and M. S. Losowsky, A survey of vitamin D deficiency in gastrointestinal and liver disorders. Q. J. Med. 209, 119 – 134 (1984). 74. M. S. Schoen, J. Lindenbaum, M. S. Roginsky, and P. R. Holt, Significance of serum levels of 25-hydroxycholecalciferol in gastrointestinal disease. Dig. Dis. Sci. 23, 137 – 142 (1978). 75. M. Imawari, K. Kozawa, Y. Akanuma, S. Koizumi, H. Itakura, and K. Kosaka, Serum 25-hydroxyvitamin D and vitamin D-binding protein levels and mineral metabolism after partial and total gastrectomy. Gastroenterology 79, 255 – 258 (1980). 76. D. D. Bikle, E. Gee, B. Halloran, M. A. Kowalski, E. Ryzen, and J. G. Haddad, Assessment of the free fraction of 25-hydroxyvitamin D in serum: Its regulation by albumin and the vitamin D-binding protein. J. Clin. Endocrinol. Metab. 63, 954 – 959 (1986). 77. B. E. Nilsson, Conditions contributing to fracture of the femoral neck. Acta Chir. Scand. 136, 383 – 384 (1970). 78. J. Chalmers, W. D. G. Conacher, D. L. Gardner, and P. J. Scott, Osteomalacia, a common disease in elderly women. J. Bone J. Surg. 49B, 403 (1967). 79. J. E. Compston, S. Chadha, L. W. L. Horton, and A. L. Merrett, Postgastrectomy osteomalacia. Gut 21, A904 (1980). 80. J. M. Gertner, M. Lilburn, and M. Domenech, 25-Hydroxycholecalciferol absorption in steatorrhoea and postgastrectomy osteomalacia. Br. Med. J. 1, 1310 – 1312 (1977). 81. G. R. Thompson, B. Lewis, and C. C. Booth, Vitamin-D absorption after partial gastrectomy. Lancet 1, 457 – 458 (1966). 82. T. C. Stamp, Intestinal absorption of 25-hydroxycholecalciferol. Lancet 2, 121 – 123 (1974). 83. A. Caniggia, C. Gennari, and L. Cesari, Intestinal absorption of 45Ca and dynamics of 45Ca in gastrectomy osteoporosis. Acta Med. Scand. 176, 599 – 605 (1964).

254 84. V. Harvald, A. R. Krogsgaard, and P. Lous, Calcium deficiency following partial gastrectomy. Acta Med. Scand. 172, 497 – 503 (1962). 85. H. R. Ekbom, Calcium studies in partially gastrectomized patients, with special reference to the oral intake of calcium. Acta Med. Scand. 178, 193 – 201 (1965). 86. G. W. Bo-Linn, G. R. Davis, D. J. Buddrus, S. G. Morawski, C. Santa Ana, and J. S. Fordtran, An evaluation of the imrotance of gastric acid secretion in the absorption of dietary calcium. J. Clin. Invest. 73, 640 – 647 (1984). 87. R. R. Recker, Calcium absorption and achlorhydria. N. Engl. J. Med. 313, 70 – 73 (1985). 88. L. Tougaard, H. Rickers, P. Rodgro, E. H. Thaysen, M. S. Christensen, B. Lund, and O. H. Sorensen, Bone composition and vitamin D after Polya gastrectomy. Acta Med. Scand. 202, 47 – 50 (1977). 89. E. M. Alhava, S. Aukee, M. Juuti, P. Karjalainen, and K. Kettunen, The effect of vitamin D2 treatment on bone mineral after partial gastrectomy. Scand. J. Gastroenterol. 12, 497 – 499 (1977). 90. S. Mora, G. Weber, G. Barera, A. Bellini, D.Pasolini, C. Prinster, C. Bianchi, and G. Chiumello, Effect of gluten-free diet on bone mineral content in growing patients with celiac disease. Am. J. Clin. Nutr. 57, 224 – 228 (1993). 91. N. Molteni, M. P. Caraceni, M. T. Bardella, S. Ortolani, G. G. Gandolini, and P. Bianchi, Bone mineral density in adult celiac patients and the effect of gluten-free diet from childhood. Am. J. Gastroenterol. 85, 51 – 53 (1990). 92. M. P. Caraceni, N. Molteni, M. T. Bardella, S. Ortolani, A. Nogara, and P. A. Bianchi, Bone and mineral metabolism in adult celiac disease. Am. J. Gastroenterol. 83, 274 – 277 (1988). 93. K. E. W. Melvin, G. W. Hepner, P. Bordier, G. Neale, and G. F. Joplin, Calcium metabolism and bone pathology in adult coeliac disease. Q. J. Med. 39, 83 – 113 (1970). 93a. T. Kemppainen, H. Kroger, E. Janatuinen, I. Arnala, V. M. Kosma, P. Pikkarainen, R. Julkunen, J. Jurvelin, E. Alhava, and M. Uusitupa, Osteoporosis in adult patients with celiac disease. Bone 24, 249 – 255 (1999). 93b. D. Gonzalex, R. Mazure, C. Mautalen, H. Vazquez, and J. Bai, Body composition and bone mineral density in untreated and treated patients with celiac disease. Bone 16, 231 – 234 (1995). 93c. R. Mazure, H. Vazquez, D. Gonzalez, C. Mautalen, S. Pedreira, L. Boerr, and J. C. Bai. Bone mineral affection in asymtomatic adult patients with celiac disease. Am. J. Gastro. 89, 2130 – 2134 (1994). 94. G. U. Exner, M. Sacher, D. H. Shmerling, and A. Prader, Growth retardation and bone mineral status in children with coeliac disease recognized after the age of 3 years. Helv. Paediatr. Acta 33, 497 – 507 (1978). 95. T. Valdimarsson, G. Toss, I. Ross, O. Lofman, and M. Strom, Bone mineral density in coeliac disease. Scand. J. Gastroenterol. 29, 457 – 461 (1994). 96. B. Bode, C. Hassager, E. Gudmand-Hoyer, and C. Christiansen, Body composition and calcium metabolism in adult treated coelic disease. Gut 32, 1342 – 1345 (1991). 97. E. Lindh, S. Ljunghall, K. Larsson, and B. Lavo, Screening for antibodies against gliadin in patients with osteoporosis. J. Intern. Med. 231, 403 – 406 (1992). 98. J. L. Jurgens, D. A. Scholz, and E. E. Wolla ger, Severe osteomalacia associated with occult steatorrhea due to nontropical sprue: Report of five cases. Arch. Intern. Med. 98, 774 – 782 (1956). 99. A. J. Moss, C. Waterhouse, and R. Terry, Gluten sensitive enteropathy with osteomalacia but without steatorrhea. N. Engl. J. Med. 272, 825 – 830 (1965). 100. H. A. Salvesen and J. Boe, Osteomalacia in sprue. Acta Med. Scand. 1466, 290 – 299 (1953). 101. K. E. Melvin, G. W. Hepner, P. Bordier, G. Neale, and G. F. Joplin, Calcium metabolism and bone pathology in adult coeliac disease. Q. J. Med. 39, 83 – 113 (1970).

DANIEL D. BIKLE 102. E. Janatuinen, E. Alhava, I. Arnala, R. Julkunen, V. M. Kosma, H. Kroger, P. Pikkarainen, M. Uusitupa, and M. Voutilainen, Bone density and histomorphometry in adult coeliac diseae. Gastroenterology 102, 229 (1992) [abstract]. 102a. G. R. Corazza, A. Di Sario, L. Cecchetti, R. A. Jorizzo, M. Di Stefano, L. Minguzzi, G. Brusco, M. Bernardi, and G. Gasbarrini. Influence of pattern of clinical presentation and of gluten-free diet on bone mass and metabolism in adult coeliac disease. Bone 18, 525 – 530 (1996). 103. S. B. Arnaud, A. D. Newcomer, E. R. Dickson, C. D. Arnaud, and V. L. W. Go, Altered mineral metabolism in patients with gastrointestinal (GI) diseases. Gastroenterology 70, 860 (1976). 104. G. W. Hepner, J. Jowsey, C. Arnaud, S. Gordon, J. Black, M. Roginsky, H. F. Moo, and J. F. Young, Osteomalacia and celiac disease: Response to 25-hydroxyvitamin D. Am. J. Med. 65, 1015 – 1020 (1978). 105. J. E. Compston, D. Judd, E. O. Crawley, W. D. Evans, C. Evans, H. A. Church, E. M. Reid, and J. Rhodes, Osteoporosis in patients with inflammatory bowel disease. Gut 28, 410 – 415 (1987). 105a. C. N. Bernstein, L. L. Seeger, J. W. Sayre, P. A. Anton, L. Artinian, F. Shanahan. Decreased bone density in inflammatory bowel disease is related to corticosteroid use and not disease diagnosis. J. Bone Min. Res. 10, 250 – 256 (1995). 106. A. J. Greenstein, H. D. Janowitz, and D. B. Sachar, The extraintestinal complications of Crohn’s disease and ulcerative colitis: A study of 700 patients. Medicine 55, 401 – 412 (1976). 107. W. T. Cooke, Survey of results of treatment of Crohn’s disease. Clin. Gastroenterol. 1, 521 – 531 (1972). 108. R. H. J. Driscoll, S. C. Meredith, M. Sitrin, and I. H. Rosenberg, Vitamin D deficiency and bone disease in patients with Crohn’s disease. Gastroenterology 83, 1252 – 1258 (1982). 109. J. E. Compston, A. B. Ayers, L. W. Horton, J. R. Tighe, and B. Creamer, Osteomalacia after small-intestinal resection. Lancet 1, 9 – 12 (1978). 110. R. M. Issenman, S. A. Atkinson, C. Radoja, and L. Fraher, Longitudinal assessment of growth, mineral metabolism, and bone mass in pediatric Crohn’s disease. p 17, 401 – 406 (1993). 111. M. Sitrin, S. Meredith, and I. H. Rosenberg, Vitamin D deficiency and bone disease in gastrointestinal disorders. Arch. Intern. Med. 138, 886 – 888 (1978). 112. S. H. Scharla, H. W. Minne, U. G. Lempert, G. Leidig, M. Hauber, R. Raedsch, and R. Ziegler, Bone mineral density and calcium regulating hormones in patients with inflammatory bowel disease (Crohn’s disease and ulcerative colitis). Exp. Clin. Endocrinol. 102, 44 – 49 (1994). 113. I. Hessov, L. Mosekilde, F. Melsen, S. Fasth, L. Hulten, B. Lund, and O. H. Sorensen, Osteopenia with normal vitamin D metabolites after small-bowel resection for Crohn’s disease. Scand. J. Gastroenterol. 19, 691 – 696 (1984). 114. E. L. Krawitt, W. L. Beeken, and C. D. Janney, Calcium absorption in Crohn’s disease. Gastroenterology 71, 251 – 254 (1976). 115. L. V. Avioli, Absorption and metabolism of vitamin D3 in man. Am. J. Clin. Nutr. 22, 437 – 446 (1969). 116. J. H. Sellin, S. C. Meredith, S. Kelly, H. Schneir, and I. H. Rosenberg, Prospective evaluation of metabolic bone disease after jejunoileal bypass. Gastroenterology 87, 123 – 129 (1984). 117. S. L. Teitelbaum, J. D. Halverson, M. Bates, L. Wise, and J. G. Haddad, Abnormalities of circulating 25-OH vitamin D after jejunalileal bypass for obesity: Evidence of an adaptive response. Ann. Intern. Med. 86, 289 – 293 (1977). 118. H. Rickers, C. Christiansen, I. Balslev, H. Foltved, P. Rodbro, and M. S. Christensen, Vitamin D and bone mineral content after intestinal bypass operation for obesity. Gut 24, 67 – 72 (1983). 119. H. Rickers, C. Christiansen, I. Balslev, and P. Rodbro, Impairment of vitamin D metabolism and bone mineral content after intestinal

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

135. 135a.

136.

bypass for obesity. A longitudinal study. Scand. J. Gastroenterol. 19, 184 – 189 (1984). J. E. Compston, L. W. Horton, M. F. Laker, A. B. Ayers, J. S. Woodhead, H. J. Bull, J. C. Gazet, and T. R. Pilkington, Bone disease after jejuno– ileal bypass for obesity. Lancet 2, 1 – 4 (1978). L. Mosekilde, F. Melsen, I. Hessov, M. S. Christensen, B. J. Lund, B. I. Lund, and O. H. Sorensen, Low serum levels of 1.25-dihydroxyvitamin D and histomorphometric evidence of osteomalacia after jejunoileal bypass for obesity. Gut 21, 624 – 631 (1980). A. M. Parfitt, J. Podenphant, A. R. Villanueva, and B. Frame, Metabolic bone disease with and without osteomalacia after intestinal bypass surgery: A bone histomorphometric study. Bone 6, 211 – 220 (1985). F. Melsen and L. Mosekilde, Trabecular bone mineralizationlag time determined by tetracycline double labeling in normal and certain pathological conditions. Acta Pathol. Microbiol. Scand. A 88, 83 – 88 (1980). I. T. Gilmore, J. B. Eastwood, W. R. Ellis, V. A. Luck, S. Fenton, A. J. O’Grady, E. Harris, and I. M. Murray-Lyon, Bone disease after jejuno-ileal bypass for morbid obesity. Postgrad. Med. J. 59, 299 – 303 (1983). P. Dano and C. Christiansen, Calcium absorption and bone mineral contents following intestinal shunt operation in obesity. Scand. J. Gastroenterol. 9, 775 – 779 (1974). H. Hey, K. H. Stokholm, B. Lund, and O. H. Sorensen, Vitamin D deficiency in obsese patients and changes in circulating vitamin D metabolites following jejunoileal bypass. Int. J. Obes. 6, 473 – 479 (1982). S. A. J. Swenson and R. Oberst, Pre- and postoperative care of the patient with intestinal bypass for obesity. Am. J. Surg. 129, 225 – 228 (1975). P. Charles, L. Mosekilde, K. Sondergard, and F. T. Jensen, Treatment with high-dose oral vitamin D2 in patients with jejunoileal bypass for morbid obesity. Effects on calcium and magnesium metabolism, vitamin D metabolites, and faecal lag time. Scand. J. Gastroenterol. 19, 1031 – 1038 (1984). W. A. Franck, G. S. Hoffman, J. S. Davis, H. D. Alpern, and J. E. Olson, Osteomalacia and weakness complicating jejunoileal bypass. J. Rheumatol. 6, 51 – 56 (1979). H. Hey, B. Lund, and O. H. Sorensen, Delayed fracture healing following jejunoileal bypass surgery for obesity. Calcif. Tissue Int. 34, 13 – 15 (1982). H. Z. Friedman, C. B. Langman, and M. J. Favus, Vitamin D metabolism and osteomalacia in cystic fibrosis. Gastroenterology 88, 808 – 813 (1985). R. Assan, J. H. Alexandre, A. Tiengo, M. Marre, L. Costamailleres, and C. Lhomme, Survival and rehabilitation after total pancreatectomy. A follow-up of 36 patients. Diabetes Metab. 11, 303 – 309 (1985). T. J. Hahn, A. E. Squires, L. R. Halstead, and D. B. Strominger, Reduced serum 25-hydroxyvitamin D concentration and disordered mineral metabolism in patients with cystic fibrosis. J. Pediatr. 94, 38 – 42 (1979). R. J. Stead, S. Houlder, J. Agnew, M. Thomas, M. E. Hodson, J. C. Batten, and P. Dandona, Vitamin D and parathyroid hormone and bone mineralisation in adults with cystic fibrosis. Thorax 43, 190 – 194 (1988). L. K. Bachrach, C. W. Loutit, R. B. Moss, and R. Marcus, Osteopenia in adults with cystic fibrosis. Am. J. Med. 96, 27 – 34 (1994). R. M. Aris, J. B. Renner, A. D. Winders, H. E. Buell, D. B. Riggs, G. E. Lester, D. A. Ontjes. Increased rate of fractures and severe kyphosis: sequelae of living into adulthood with cystic fibrosis. Ann. Int. Med. 128, 186 – 193 (1998). J. E. Compston, S. Vedi, E. Gianetta, G. Watson, D. Civalleri, and N. Scopinaro, Bone histomorphometry and vitamin D status after

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

151. 152. 153.

154.

155.

255

biliopancreatic bypass for obesity. Gastroenterology 87, 350 – 356 (1984). J. Scott, E. Elias, P. J. Moult, S. Barnes, and M. R. Wills, Rickets in adult cystic fibrosis with myopathy, pancreatic insufficiency and proximal renal tubuler dysfunction. Am. J. Med. 63, 488 – 492 (1977). R. G. Long, E. Meinhard, R. T. Skinner, Z. Varghese, M. R. Wills, and S. Sherlock, Clinical, biochemical, and histological studies of osteomalacia, osteoporosis, and parathyroid function in chronic liver disease. Gut 19, 85 – 90 (1978). M. Davies, E. B. Mawer, H. J. Klass, G. A. Lumb, J. L. Berry, and T. W. Warnes, Vitamin D deficiency, osteomalacia, and primary biliary cirrhosis. Response to orally administered vitamin D3. Dig. Dis. Sci. 28, 145 – 153 (1983). A. J. Stellon, A. Webb, J. Compston, and B. Williams, Low bone turnover state in primary biliary cirrhosis. Hepatology 7, 137 – 142 (1987). D. S. Matloff, M. M. Kaplan, R. M. Neer, M. J. Goldberg, W. Bitman, and H. J. Wolfe, Osteoporosis in primary biliary cirrhosis: effects of 25-hydroxyvitamin D3 treatment. Gastroenterology 83, 97 – 102 (1982). H. F. Herlong, R. R. Recker, and W. C. Maddrey, Bone disease in primary biliary cirrhosis: Histologic features and response to 25hydroxyvitamin D. Gastroenterology 83, 103 – 108 (1982). M. J. Atkinson, I. Vido, E. Keck, and R. D. Hesch, Hepatic osteodystrophy in primary biliary cirrhosis: A possible defect in Kupffer cell mediated cleavage of parathyroid hormone. Clin. Endocrinol. 19, 21 – 28 (1983). J. M. Bengoa, M. D. Sitrin, S. Meredith, S. E. Kelly, N. Shah, A. L. Baker, and I. H. Rosenberg, Intestinal calcium absorption and vitamin D status in chronic cholestatic liver disease. Hepatology 4, 261 – 265 (1984). J. S. Reed, S. C. Meredith, B. A. Nemchausky, I. H. Rosenberg, and J. L. Boyer, Bone disease in primary biliary cirrhosis: Reversal of osteomalacia with oral 25-hydroxyvitamin D. Gastroenterology 78, 512 – 517 (1980). S. F. Hodgson, E. R. Dickson, H. W. Wahner, K. A. Johnson, K. G. Mann, and B. L. Riggs, Bone loss and reduced osteoblast function in primary biliary cirrhosis. Ann. Intern. Med. 103, 855 – 860 (1985). H. C. Mitchison, A. J. Malcolm, M. F. Bassendine, and O. F. W. James, Metabolic bone disease in primary biliary cirrhosis at presentation. Gastroenterology 94, 463 – 470 (1988). J. A. Cuthbert, C. Y. Pak, J. E. Zerwekh, K. D. Glass, and B. Combes, Bone disease in primary biliary cirrhosis: Increased bone resorption and turnover in the absence of osteoporosis or osteomalacia. Hepatology 4, 1 – 8 (1984). M. S. Shih and C. Anderson, Does “hepatic osteodystrophy” differ from peri- and postmenopausal osteoporosis? A histomorphometric study. Calcif. Tissue Int. 41, 187 – 191 (1987). T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, J. Roche, and S. Posen, Osteoporosis and skeletal fractures in chronic liver disease. Gut 31, 82 – 87 (1990). R. G. Long and D. S. Sherlock, Vitamin D in chronic liver disease. Prog. Liver Dis. 6, 539 – 555 (1979). J. G. Heaf, Hepatic osteodystrophy. Am. J. Gastroenterol. 20, 1035 – 1040 (1985). J. B. Griffin, C. D. Smith, and G. M. Bright, Bone disease in biliary atresia: A not uncommon complication. Curr. Surg. 42, 475 – 477 (1985). N. Guanabens, A. Pares, L. Marinoso, A. Brancos, C. Piera, S. Serrano, F. Rivera, and J. Rodes, Factors influencing the development of metabolic bone disease in primary biliary cirrhosis. Am. J. Gastroenterol. 85, 1356 – 1367 (1990). F. N. R. Van Berkum, R. Beukers, J. C. Birkenha ger, P. P. M. Kooij, S. W. Schalm, and H. A. P. Pols, Bone mass in women with primary

256 biliary cirrhosis: The relation with histological stage and use of glucocorticoids. Gastroenterology 99, 1134 – 1139 (1990). 156. V. Fonseca, O. Epstein, D. S. Gill, R. K. Menon, M. Thomas, N. McIntyre, and P. Dandona, Hyperparathyroidism and low serum osteocalcin despite vitamin D replacement in primary biliary cirrhosis. J. Clin. Endocrinol. Metab. 64, 873 – 877 (1987). 157. S. F. Hodgson, E. R. Dickson, R. Eastell, E. F. Eriksen, S. C. Bryant, and B. L. Riggs, Rates of cancellous bone remodeling and turnover in osteopenia associated with primary biliary cirrhosis. Bone 14, 819 – 827 (1993). 158. R. G. Long, M. R. Wills, R. K. Skinner, and S. Sherlock, Serum 25dihydroxyvitamin D in untreated parenchymal and cholestatic liver disease. Lancet 11, 650 – 652 (1976). 159. M. M. Kaplan, M. J. Goldberg, D. S. Matloff, R. M. Neer, and D. B. Goodman, Effect of 25-hydroxyvitamin D3 on vitamin D metabolites in primary biliary cirrhosis. Gastroenterology 81, 681 – 685 (1981). 160. A. J. Stellon, A. Davies, J. Compston, and R. Williams, Osteoporosis in chronic cholestatic liver disease. Q. J. Med. 57, 783 – 790 (1985). 161. A. Danielsson, R. Lorentzon, and S. E. Larsson, Intestinal absorption and 25-hydroxylation of vitamin D in patients with primary biliary cirrhosis. Scand. J. Gastroenterol. 17, 349 – 355 (1982). 162. M. E. Holda and J. R. Ryan, Hepatobiliary rickets. J. Pediatr. Orthop. 2, 285 – 287 (1982). 163. A. K. Kehayoglou, C. D. Holdsworth, J. E. Agnew, M. J. Whelton, and S. Sherlock, Bone disease and calcium absorption in primary biliary cirrhosis with special reference to vitamin-D therapy. Lancet 1, 715 – 718 (1968). 164. R. G. Long, Z. Varghese, M. R. Wills, and S. Sherlock, Plasma calcium and magnesium fractions in liver disease. Clin. Chim. Acta 84, 239 – 245 (1978). 165. M. D. Sitrin and J. M. Bengoa, Intestinal absorption of cholecalciferol and 25-hydroxycholecalciferol in chronic cholestatic liver disease. Am. J. Clin. Nutr. 46, 1011 – 1015 (1987). 166. W. G. Thompson and G. R. Thompson, Effect of cholestyramine on the absorption of vitamin D3 and calcium. Gut 10, 717 – 722 (1969). 167. R. K. Skinner, S. Sherlock, R. G. Long, and M. R. Willis, 25Hydroxylation of vitamin D in primary biliary cirrhosis. Lancet 1, 720 – 721 (1977). 168. E. B. Mawer, H. J. Klass, T. W. Warnes, and J. L. Berry, Metabolism of vitamin D in patients with primary biliary cirrhosis and alcoholic liver disease. Clin. Sci. 69, 561 – 570 (1985). 169. R. G. Long, Hepatic osteodystrophy: Outlook good but some problems unsolved. Gastroenterology 78, 644 – 647 (1980). 170. O. Epstein, Y. Kato, R. Dick, and S. Sherlock, Vitamin D, hydroxyapatite, and calcium gluconate in treatment of cortical bone thinning in postmenopausal women with primary biliary cirrhosis. Am. J. Clin. Nutr. 36, 426 – 430 (1982). 171. M. K. Porayko, R. H. Wiesner, J. E. Hay, R. A. Krom, E. R. Dickson, S. Beaver, and L. Schwerman, Bone disease in liver transplant recipients: incidence, timing, and risk factors. Transplant. Proc. 23, 1462 – 1465 (1991). 172. A. J. Stellon, A. Davies, J. Compston, and R. Williams, Bone loss in autoimmune chronic active hepatitis on maintenance corticosteroid therapy. Gastroenterology 89, 1078 – 1083 (1985). 173. J. B. Dibble, P. Sheridan, R. Hampshire, G. J. Hardy, and M. S. Losowsky, Osteomalacia, vitamin D deficiency and cholestasis in chronic liver disease. Q. J. Med. 201, 89 – 103 (1982). 173a. F. J. Gallego-Rojo, J. L. Gonzalez-Calvin, M. Munoz-Torres, J. I. Mundi, R. Fernandez-Perez, and D. Rodrigo-Moreno. Bone mineral density, serum insulin-like growth factor I, and bone turnover markers in viral cirrhosis. Hepatology 28, 695 – 699 (1998). 173b. M. P. Whyte, S. L. Teitelbaum, and W. R. Reinus. Doubling skeletal mass during adult life: The syndrome of diffuse osteosclerosis after intravenous drug abuse. J. Bone Miner. Res. 11, 554 – 558 (1996).

DANIEL D. BIKLE 173c. T. Diamond and B. Depcynski, Acquired osteosclerosis associated with intravenous drug abuse and hepatitis c infection. Bone 19, 679 – 683 (1996). 173d. A. A. K. Hassoun, T. B. Nippolot, R. D. Tiegs, and S. Khosla, Hepatitis C-associated osteosclerosis: An unusual syndrome of acquired osteosclerosis in adults. Am. J. Med. 103, 70 – 73 (1997). 174. P. D. Saville, Changes in bone mass with age and alcoholism. J. Bone Jt. Surg. 47, 492 – 499 (1965). 175. H. Spencer, N. Rubio, E. Rubio, M. Indreika, and A. Seitam, Chronic alcoholism: Frequently overlooked cause of osteoporosis in men. Am. J. Med. 80, 393 – 397 (1986). 176. D. R. M. Lindsell, A. G. Wilson, and J. D. Maxwel, Fractures on the chest radiograph in detection of alcoholic liver disease. Br. Med. J. 285, 597 – 599 (1982). 177. Y. Israel, H. Orrego, S. Holt, D. W. Macdonald, and H. E. Meemak, Identification of alcohol abuse: Thoracic fractures on routine chest X-rays as indicators of alcoholism. Alcohol. Clin. Exp. Res. 4, 420 – 422 (1980). 177a. C. V. Odvina, I. Safi, C. H. Wojtowicz, E. I. Barengolts, P. Lathon, A. Skapars, P. N. Desai, and S. C. Kukreja. Effect of heavy alcohol intake in the absence of liver disease on bone mass in black and white men. J. Clin. Endocrinol. Metab. 80, 2499 – 2503 (1995). 178. D. D. Bikle, A. Stesin, B. Halloran, L. Steinbach, and R. Recker, Alcohol-induced bone disease: Relationship to age and parathyroid hormone levels. Alcohol Clin. Exp. Res. 17, 690 – 695 (1993). 179. B. E. Nilsson and N. E. Westlin, Femur density in alcoholism and after gastrectomy. Calcif. Tissue. Res. 10, 167 – 170 (1972). 180. B. E. Nilsson and N. E. Westlin, Changes in bone mass in alcoholics. Clin. Orthop. 90, 229 – 232 (1973). 181. D. D. Bikle, H. K. Genant, C. Cann, R. R. Recker, B. P. Halloran, and G. J. Strewler, Bone disease in alcohol abuse. Ann. Intern. Med. 103, 42 – 48 (1985). 182. M. C. De Vernejoul, J. Bielakoff, M. Herve, J. Gueris, M. Hott, D. Modrowski, D. Kuntz, L. Miravet, and A. Ryckewaert, Evidence for defective osteoblastic function: A role for alcohol and tobacco consumption in osteoporosis in middle-aged men. Clin. Orthop. 179, 107 – 115 (1983). 183. J. A. Jorge-Hernandez, C. E. Gonzales-Reimers, A. TorresRamirez, F. Santolaria-Fernandez, C. Gonzales Garcia, J. N. Batista-Lopez, M. Pestana-Pestana, and L. Hernandez-Nieto, Bone changes in alcoholic liver cirrhosis: A histomorphometrical analysis of 52 cases. Dig. Dis. Sci. 33, 1089 – 1095 (1988). 184. C. M. Schnitzler and L. Solomon, Bone changes after alcohol abuse. South Afr. Med. J. 66, 730 – 734 (1984). 185. S. A. Mobarhan, R. M. Russell, R. Recker, D. B. Posner, F. L. Iber, and P. Miller, Metabolic bone disease in alcoholic cirrhosis: A comparison of the effect of vitamin D2, 25-hydroxyvitamin D, or supportive treatment. Hepatology 4, 266 – 273 (1984). 186. B. C. Lalor, M. W. France, D. Powell, P. H. Adams, and T. B. Counihan, Bone and mineral metabolism and chronic alcohol abuse. Q. J. Med. 59, 497 – 511 (1986). 187. M. E. Arlot, M. Bonjean, P. M. Chavassieux, P. J. Meunier, and L. France, Bone histology in adults with aseptic necrosis. J. Bone Jt. Surg. 65A, 1319 – 1327 (1983). 188. O. Johnell, B. E. Nilsson, and P. E. Wiklund, Bone morphometry in alcoholics. Clin. Orthop. 165, 253 – 258 (1982). 189. L. Solomon, Idiopathic necrosis of the femoral head: Pathogenesis and treatment. Can. J. Surg. 24, 573 – 578 (1981). 190. E. W. Gold and P. J. Cangemi, Incidence and pathogenesis of alcohol-induced osteonecrosis of the femoral head. Clin. Orthop. 143, 222 – 226 (1979). 191. T. O. Pitts and D. H. Van Thiel, Disorders of divalent ions and vitamin D metabolism in chronic alcoholism. Recent Dev. Alcohol 4, 357 – 377 (1986).

CHAPTER 48 Osteoporosis in Gastrointestinal, Pancreatic, and Hepatic Diseases 192. S. Feitelberg, S. Epstein, F. Ismail, and C. D. Amanda, Deranged bone mineral metabolism in chronic alcoholism. Metabolism 36, 322 – 326 (1987). 193. K. Laitinen, C. Lamber-Allardt, R. Tunninen, S-L. Karonen, R. Tahtela, R. Ylikahri, and M. Valimaki, Transient hypoparathyroidism during acute alcohol intoxication. N. Engl. J. Med. 324, 721 – 727 (1991). 194. M. S. Roginsky, I. Zanzi, and S. H. Cohn, Skeletal and lean body mass in alcoholics with and without cirrhosis. Calcif. Tissue. Res. 21, 386 – 391 (1976). 195. D. D. Bikle, B. P. Halloran, E. Gee, E. Ryan, and J. G. Haddad, Free 25-hydroxyvitamin D levels are normal in subjects with disease and reduced total 25-hydroxyvitamin D levels. J. Clin. Invest. 78, 748 – 752 (1986). 196. D. D. Bikle, E. Gee, B. Halloran, and J. G. Haddad, Free 1,25dihydroxyvitamin D levels in serum from normal pregnant subjects, and subjects with liver disease. J. Clin. Invest. 74, 1966 – 1971 (1984). 197. D. D. Bikle, P. K. Siiteri, E. Ryzen, and J. G. Haddad, Serum protein binding of 1,25-dihydroxyvitamin D: A reevaluation by direct measurement of free metabolite levels. J. Clin. Endocrinol. Metab. 61, 969 – 975 (1985). 198. C. M. Leevy, A. Thompson, and H. Baker, Vitamin D and liver injury. Am. J. Clin. Nutr. 23, 493 – 499 (1970). 199. R. T. Jung, M. Davie, J. O. Hunter, T. M. Chalmers, and D. E. M. Lawson, Abnormal vitamin D metabolism in cirrhosis. Gut 19, 290 – 293 (1978). 200. M. Meyer, S. Wechsler, and S. Shibolet, Malabsorption of vitamin D in patients and rats with liver cirrhosis, In “Vitamin D: Biochemical, Chemical and Clinical Aspects Related to Calcium Metabolism” (A. W. Norman, K. Schaeffer, J. W. Coburn, and H. F. DeLuca, Eds.), pp. 831 – 833. Walter de Gruyter, Berlin/New York, 1977. 201. G. W. Hepner, M. Roginsky, and H. F. Moo, Abnormal vitamin D metabolism in patients with cirrhosis. Am. J. Dig. Dis. 21, 527 –531 (1976). 202. A. J. Adler, S. Gudis, and G. M. Berlyne, Reduced renal phosphate threshold concentration in alcoholic cirrhosis. Miner. Electrolyte. Metab. 10, 63 – 66 (1984). 203. J. M. Barragry, R. G. Long, and M. W. France, Intestinal absorption of cholecalciferol in alcoholic liver disease and primary biliary cirrhosis. Gut 20, 559 – 564 (1979). 204. B. Lund, O. H. Sorensen, and M. Hilden, The hepatic conversion of vitamin D in alcoholics with varying degrees of liver affection. Acta Med. Scand. 202, 221 – 224 (1977). 205. D. B. Posner, R. M. Russell, S. Absood, T. B. Connor, D. Davies, L. Martin, J. B. Williams, A. H. Norris, and C. Merchant, Effective 25-hydroxylation of vitamin D2 in alcoholic cirrhosis. Gastroenterology 74, 866 – 870 (1978). 206. E. L. Krawitt, Ethanol inhibits intestinal calcium transport in rats. Nature 243, 88 – 89 (1973). 207. A. Pascual-Garcia, A. Donath, and B. Courvoisier, Plasma 25OHD, bone mass densitometry and Ca intestinal absorption deficiency in chronic alcoholism. In “Vitamin D: Biochemical, Chemical and Clinical Aspects Related to Calcium Metabolism” (A. W. Norman, K. Schaefer, J. W. Coburn, et al., Eds.), pp. 819 – 821. Walter de Gruyter, New York, 1977. 208. T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, and S. Posen, Ethanol reduces bone formation and may cause osteoporosis. Am. J. Med. 86, 282 – 288 (1989). 209. J. Lindholm, T. Steiniche, E. Rasmussen, G. Thamsborg, I. O. Nielsen, H. Brockstedt-Rasmussen, T. Storm, L. Hyldstrupt, and C. Schou, Bone disorder in men with chronic alcoholism: A reversible disease? J. Clin. Endocrinol. Metab. 73, 118 – 124 (1991). 210. E. B. Haagsma, I. J. Klompaker, R. Verwer, and M. J. Slooff, Longterm results after liver transplantation in adults. Scand. J. Gastroenterol. Suppl. 188, 38 – 43 (1991).

257

211. E. B. Haagsma, C. J. P. Thijn, J. G. Post, M. J. H. Slooff, and C. H. Gips, Bone disease after orthotopic liver transplantation. J. Hepatol. 6, 94 – 100 (1988). 211a. S. Vedi, S. Greer, S. J. Skingle, N. J. Garrahan, M. Ninkovic, G. A. Alexander, and J. E. Compston, Mechanism of bone loss after liver transplantation: a histomorphometric analysis. J. Bone Miner. Res. 14, 281 – 287 (1999). 211b. O. M. Crosbie, R. Freaney, M. J. McKenna, and J. E. Hegarty, Bone density, vitamin D status, and disordered bone remodeling in endstage chronic liver disease. Calcif. Tissue Intern. 64, 295 – 300 (1999). 212. F. G. Hawkins, M. Leon, M. B. Lopez, M. A. Valero, L. Larrodera, I. Garcia-Garcia, C. Loinaz, and E. Moreno Gonzalez, Bone loss and turnover in patients with liver transplantation. Hepatogastroenterology 41, 158 – 161 (1994). 213. M. Rabinovitz, J. Shapiro, J. Lian, G. D. Block, I. S. Merkel, and D. H. Van Thiel, Vitamin D and osteocalcin levels in liver transplant recipients. Is osteocalcin a reliable marker of bone turnover in such cases? J. Hepatol. 16, 50 – 55 (1992). 214. R. Eastell, E. R. Dickson, S. F. Hodgson, R. H. Wiesner, M. K. Porayko, H. W. Wahner, S. L. Cedel, B. L. Riggs, and R. A. Krom, Rates of vertebral bone loss before and after liver transplantation in women with primary biliary cirrhosis. Hepatology 14, 296 – 300 (1991). 215. G. L. Klein, C. M. Targoff, M. E. Ament, D. J. Sherrard, R. Bluestone, J. H. Young, A. W. Norman, and J. W. Coburn, Bone disease associated with total parenteral nutrition. Lancet 2, 1041 – 1044 (1980). 216. M. Shike, J. E. Harrison, W. C. Sturtridge, C. S. Tam, P. E. Bobechko, G. Jones, T. M. Murray, and K. N. Jeejeebhoy, Metabolic bone disease in patients receiving long-term total parenteral nutrition. Ann. Intern. Med. 92, 343 – 350 (1980). 217. G. L. Klein, R. A. Cannon, M. Diament, H. Kangarloo, M. E. Ament, A. W. Norman, and J. W. Coburn, Infantile vitamin D resistant rickets associated with total parenteral nutrition. Am. J. Dis. Child. 136, 74 – 76 (1982). 218. G. L. Klein, A. C. Alfrey, N. L. Miller, D. J. Sherrard, T. K. Hazlet, M. E. Ament, and J. W. Coburn, Aluminum loading during total parenteral nutrition. Am. J. Clin. Nutr. 35, 1425 – 1429 (1982). 219. S. M. Ott, N. A. Maloney, G. L. Klein, A. C. Alfrey, M. E. Ament, J. W. Coburn, and D. J. Sherrard Aluminum is associated with low bone formation in patients receiving chronic parenteral nutrition. Ann. Intern. Med. 98, 910 – 914 (1983). 220. J. H. Vargas, G. L. Klein, M. E. Ament, S. M. Ott, D. J. Sherrard, R. L. Horst, W. E. Berquist, A. C. Alfrey, E. Slatopolsky, and J. W. Coburn, Metabolic bone disease of total parenteral nutrition: Course after changing from casein to amino acids in parenteral solutions with reduced aluminum content. Am. J. Clin. Nutr. 48, 1070 – 1078 (1988). 221. E. W. Lipkin, S. M. Ott, and G. L. Klein, Heterogeneity of bone histology in parenteral nutrition patients. Am. J. Clin. Nutr. 46, 673 – 680 (1987). 222. M. Shike, M. E. Shils, A. Heller, N. Alcock, V. Vigorita, R. Brockman, M. F. Holick, J. Lane, and C. Flombaum, Bone disease in prolonged parenteral nutrition: Osteopenia without mineralization defect. Am. J. Clin. Nutr. 44, 89 – 98 (1986). 223. J. Foldes, B. Rimon, M. Muggia-Sullam, Z. Gimmon, I. Leichter, R. Steinberg, J. Menczel, and H. R. Freund, Progressive bone loss during long-term home total parenteral nutrition. J. Parenter. Enteral Nutr. 14, 139 – 142 (1990). 224. G. L. Klein, R. L. Horst, A. W. Norman, M. E. Ament, E. Slatopolsky, and J. W. Coburn, Reduced serum levels of 1 alpha,25dihydroxyvitamin D during long-term total parenteral nutrition. Ann. Intern. Med. 94, 638 – 643 (1981). 225. M. Shike, W. C. Sturtridge, C. S. Tam, J. E. Harrison, G. Jones, T. M. Murray, H. Husdan, J. Whitwell, D. R. Wilson, and K. N. Jeejeehoy, A

258 possible role of vitamin D in the genesis of parenteral-nutrition-induced metabolic bone disease. Ann. Intern. Med. 95, 560 – 568 (1981). 226. G. L. Klein and J. W. Coburn, Metabolic bone disease associated with total parenteral nutrition. Adv. Nutr. Res. 6, 67 – 92 (1984). 227. A. B. Sedman, G. L. Klein, R. J. Merritt, N. L. Miller, K. O. Weber, W. L. Gill, H. Anand, and A. C. Alfrey, Evidence of aluminum loading in infants receiving intravenous therapy. N. Engl. J. Med. 312, 1337 – 1343 (1985).

DANIEL D. BIKLE 228. D. S. Milliner, J. H. Shinaberger, P. Shuman, and J. W. Coburn, Inadvertent aluminum administration during plasma exchange due to aluminum contamination of albumin-replacement solutions. N. Engl. J. Med. 312, 165 – 167 (1985). 229. J. M. Bengoa, M. D. Sitrin, R. J. Wood, and I. H. Rosenberg, Amino acid-induced hypercalciuria in patients on total parenteral nutrition. Am. J. Clin. Nutr. 38, 264 – 269 (1983).

CHAPTER 49

Primary Hyperparathyroidism and Hyperparathyroid Bone Disease LORRAINE A. FITZPATRICK HUNTER HEATH III

I. II. III. IV.

Division of Endocrinology and Metabolism and Internal Medicine, Mayo Clinic and Foundation, Rochester, Minnesota 55905 U.S. Medical Division, Eli Lilly and Company, Indianapolis, Indiana 46285

Introduction The Actions of Parathyroid Hormone on Bone PTH Secretion in Primary Hyperparathyroidism Clinical Manifestations of Bone Disease in Primary Hyperparathyroidism

V. Effects of Therapy on Bone Mass in Primary Hyperparathyroidism VI. Summary and Conclusions References

I. INTRODUCTION

warranted to protect bone [5 – 7]. This chapter will review the current state of knowledge about 1°HPT and bone and highlight recent long-term data.

The effects of primary hyperparathyroidism (1°HPT) on bone are of great interest to clinicians, because they affect the therapeutic decision-making process for many patients each year. 1°HPT is a common disease of middle-aged and older adults and predominantly of women over the age of 50 [1], those most susceptible to postmenopausal osteoporosis. In the past, views of clinicians were diverse. Some authors claimed that avoidance or reversal of osteopenia necessitates surgical treatment of all pateints with 1°HPT [2,3]; others proposed that bone is “protected” in women having 1°HPT [4]; and still others maintained that therapy of the hyperparathyroidism is not

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. THE ACTIONS OF PARATHYROID HORMONE ON BONE A detailed treatise on the cellular and molecular actions of parathyroid hormone (PTH) on bone is beyond the scope of this chapter, but a brief review is necessary to set the stage for understanding clinical aspects of bone in 1°HPT. The reader is referred to recent surveys for details [8 – 10] (see also Chapters 7, 12, and 77).

259

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

260 Both animal studies and clinical observations document a “parathyroid paradox”: in some situations, PTH excess causes severe bone disease with focal destruction, fractures, and fibrous cysts (osteitis fibrosa cystica) [11,12]. In other circumstances, both endogenous and exogenous PTH may be anabolic to bone, even causing at one extreme diffuse osteosclerosis [13]. In other settings, 1°HPT is associated with diffuse osteopenia and osteoporotic fractures [14]. How can one explain such discrepant actions of one hormone? The answers appear to lie in fundamental concepts of PTH action on bone [8 – 10].

A. Mechanisms of PTH Action on Bone in Vitro In mammals, PTH administration increases bone resorption by osteoclasts within minutes to a few hours (see Chapter 3). Immunochemical and radiolocalization studies with PTH suggest that the primary cellular targets of the hormone in bone are mononuclear cells [15 – 17]. It remains possible that PTH binds to osteoclast precursors and stimulates their maturation into osteoclasts, but a more current view is that PTH acts indirectly to foster bone resorption. In contrast to the uncertainty about PTHs actions on osteoclasts and their precursors, an abundant literature documents effects of PTH on cells of the osteoblast lineage in virtually every species and system tested. Most osteoblastlike cell lines possess high-affinity receptors for PTH, manifest activation of adenylyl cyclase and phospholipase C by PTH, and are affected in morphology, growth, and differentiation by PTH. The bone-resorbing actions of PTH in vitro are dependent upon the presence of osteoblast-like cells [18]. In vitro, one can obtain widely varying actions of PTH on bone rudiments and bone cells, depending upon the species from which the cells came, whether the cells are in bone, in primary culture, or transformed, and how they are treated before and during exposure to PTH. The phenomenon of “downregulation” or desensitization to PTH is noteworthy: cells exposed to PTH for even relatively brief periods suffer reduced ability to respond to reexposure to the hormone [19 – 21]. This phenomenon may be related to differing effects of PTH in vivo when administered continuously vs periodically, as described below.

B. In Vivo Studies of PTH Effects on Bone Continuous infusion of synthetic PTH fragment (1 – 34) to greyhound dogs or rats in pharmacologic amounts increases bone turnover with resorption sometimes exceeding formation, leading to net loss of trabecular bone [22,23]. However, periodic administration of PTH (1 – 34) once or twice daily by parenteral routes is anabolic to bone

FITZPATRICK AND HEATH

[22 – 24] (see Chapter 77). While both bone resorption and formation are stimulated by any method of administering PTH, only intermittent dosing regimens increase formation over resorption with net gain of trabecular bone. Rats given PTH in this fashion may develop radiographically evident osteosclerosis [25]. In greyhound dogs, continuous administration of human PTH(1 – 34) (hPTH(1 – 34) increased bone turnover without significant change in trabecular bone volume, while once-daily injection increased trabecular bone volume. Most experiments in animals and humans have been done with synthetic PTH fragment (1 – 34), but the region of the PTH molecule distal to residue 34 may also have effects on bone turnover. Kaji et al. [26] demonstrated increased osteoclastic activity in vitro with administration of human carboxyl-terminal region PTH fragments. Taken together, these data suggest a profoundly important temporal effect of PTH on bone that must be understood before we can comprehend why hyperparathyroid patients might or might not develop osteopenia. PTH may induce osteoblasts to produce the bone-resorbing cytokine, interleukin-6 (IL-6). In patients with untreated 1°HPT, circulating serum concentrations of interleukin-6 were 16-fold higher than in normal controls. Circulating interleukin-6 receptor and tumor necrosis factor (TNF) were also significantly elevated. After surgical extirpation of the abnormal parathyroid glands, levels of these cytokines returned to normal. Concentrations of circulating IL-6 and TNF correlated with biochemical markers of bone turnover, and regression analysis suggested that Il-6 was an independent predictor of bone resorption. These data suggest that PTH may mediate IL-6-induced bone resorption in 1°HPT [27]. The anabolic effects of PTH(1 – 34) have been demonstrated in human beings and exploited in the treatment of osteoporosis (see Chapter 77). The anabolic actions of PTH on bone have been nicely summarized by Dempster et al. [28]. Four hundred units (about 25 mcg) of human PTH(1 – 34) injected subcutaneously daily increased spinal bone mineral content, in men with idiopathic osteoporosis [29] and in estrogen-treated postmenopausal women with osteoporosis [30], and prevented spinal bone mineral loss in 20 young hypogonadal women during treatment with GnRH analogs [31,32]. A 3-year randomized controlled trial evaluated the effects of human PTH(1 – 34) in postmenopausal women with osteoporosis who were taking hormone replacement therapy. The control group was taking hormone replacement therapy alone. The mean increase in vertebral bone mineral density at 3 years of therapy with PTH(1 – 34) was 13.0%, and was 2.7% at the hip. Increased bone mineral density was associated with a reduced incidence of vertebral fractures, which was statistically signifi cant when the fractures were evaluated as ≥15% reduction in vertebral height. The responses of bone markers were unexpected. Serum osteocalcin concentration, which

261

CHAPTER 49 Primary Hyperparathyroidism and Hyperparathyroid Bone Disease

reflects bone formation, increased by more than 55%. A bone resorption marker, urinary cross-linked N-telopeptide, decreased by 20%. Osteocalcin levels rose within 1 month, while the marker of bone resorption, N-telopeptide, did not rise until 6 months of treatment. This suggests an uncoupling of bone resorption and formation, as bone formation was stimulated prior to bone resorption [33]. A separate 2-year, randomized, double-blind, placebocontrolled trial evaluated the skeletal efficacy of daily subcutaneous hPTH(1 – 34) plus oral estrogen. Women enrolled were at least 5 years postmenopausal and had lumbar spine or hip bone mineral density (BMD) more than 2.5 SD below the mean for normal young women. Subjects had been receiving hormone replacement therapy for at least 1 year prior to entry. Patients received supplemental vitamin D and calcium for total intakes of 800 IU and 1500 mg, respectively. Sixty subjects completed the treatment protocol. At 1 year, bone mineral density of the lumbar spine had increased 20.6% and femoral neck BMD increased 4.0%. By 2 years, the increase in BMD at the lumbar spine was 29.2%, and 11.0% at the femoral neck. Thus, hPTH(1 – 34) given concomitantly with estrogen therapy produces large increases in BMD at the lumbar spine and femoral neck [34].

III. PTH SECRETION IN PRIMARY HYPERPARATHYROIDISM The secretion of PTH is not fully autonomous in most patients having 1°HPT. Calcium chelation with EDTA increases and calcium infusion decreases PTH secretion in most affected people [35 – 37]. Moreover, the secretion of PTH by parathyroid tumors is not continuous, but often appears to be episodic or pulsatile [38,39]. Because of the expense of two-site assays for PTH and technical constraints, the full minute-to-minute, 24 h per day secretory profile for PTH has not been determined in hyperparathyroid patients, but there is surely a complex cyclicity of hormone release in most individuals, encompassing both diurnal rhythm and pulsatility on a shorter time scale [38 – 44]. In contrast to the situation of once-daily administration of PTH, however, most hyperparathyroid patients have plasma PTH concentrations that are at all times above the normal reference range. Why then would the bone of hyperparathyroid patients not respond as in animals given PTH by constant infusion? The episodic or pulsatile secretion of PTH by parathyroid tumors may offer a partial answer. While constant infusion of PTH may blunt PTH’s effect on bone cells, it is not abolished. Periodic respite of bone from PTH action — even if only relative — may allow “recovery” of target cells. Alternatively, constant administration and discontinuous administration may have differential effects on cell signaling, leading to opposite effects on differentiated osteoblast function [45 – 47].

IV. CLINICAL MANIFESTATIONS OF BONE DISEASE IN PRIMARY HYPERPARATHYROIDISM A. Osteitis Fibrosa Cystica Osteitis fibrosa cystica generalisata, or OFC, is the least common form of hyperparathyroid bone disease, but the first recognized [11]. Patients may present with diffuse or focal bone pain, or pathologic fracture through an osteoclastic “brown tumor.” Radiographically, there are focal areas of osteolysis and bone expansion that may be mistaken for neoplastic metastasis (Figs. 1 and 2) [11,48]. Subperiosteal resorption of bone is present in almost all cases, most easily visualized on high-resolution radiographs of the hands (Fig. 1). OFC is a rare occurrence in patients with 1°HPT, but when it occurs is an absolute indication for surgical intervention. The complexity of the skeletal lesions associated with 1°HPT has been illustrated by dual-energy X-ray absorptiometry (DXA) studies. Two young women with 1°HPT and severe osteitis fibrosa cystica allowed postoperative follow-up of BMD by DXA. Both patients had preoperative T scores for femoral neck and lumbar spine below 4.0. Postoperatively, these patients had dramatic and sustained increases in BMD, reaching 555% above baseline. These patients demonstrate the remarkable capacity of the skeleton to mineralize after surgery for 1°HPT [49].

B. Diffuse Osteosclerosis Diffuse osteosclerosis is seldom seen in patients with 1°HPT [13], but is more common in patients with secondary HPT caused by chronic renal insufficiency. It is not clear why renal patients are so much more likely to have an anabolic response to PTH excess; there are several possibilities. Patients on hemodialysis have plasma concentrations of PTH that are often much higher than in 1°HPT, plasma of inorganic phosphorus concentrations are also high, and those of 1,25-dihydroxyvitamin D are generally low. These variations may affect the skeletal response to PTH.

C. Diffuse Osteopenia with or without Clinical Signs The advent of single photon absorptiometry (SPA) permitted precise measurement of forearm BMD, and the first such reports of osteopenia in hyperparathyroid individuals appeared in the 1970s. Osteopenia of varying degrees has been detected in hyperparathyroid patients by many different techniques [50 – 73]. Deficits in BMD have been

262

FITZPATRICK AND HEATH

FIGURE 1

Subperiosteal resorption of the hands in hyperparathyroidism. In patients with hyperparathyroidism, generalized demineralization with hazy and indistinct trabecular pattern is seen with a loss of a distinct cortical margin. Periosteal resorption, especially of the radial aspect of the second to fourth distal phalanges, is noted. These series of photographs indicate bone resorption in the phalanges and tufts of both hands with increasing severity (left to right) in three patients with primary hyperparathyroidism. Long bones may also show periosteal and endosteal resorption. Resorption of the lamina dura around the teeth results in a “floating teeth” picture. (Bottom) Magnified hand film of a patient with primary hyperparathyroidism. Note the subperiosteal resorption along the cortical shaft and at the fingertips.

CHAPTER 49 Primary Hyperparathyroidism and Hyperparathyroid Bone Disease

FIGURE 2

Brown tumors are focal areas of resorption producing cyst-like areas on radiograph. Fibroblastic tissue fills the defect and at times it is so exuberent that it bulges the bone and suggests a primary neoplasm. Brown tumors occur in any bone although the mandible, pelvis, and femur are frequently involved. Pathologic fractures can occur through brown tumors or through the abnormally weakened bones. In the first radiograph (top), a mixed sclerotic, lytic lesion in the right ilium and left supraacetabular region is presented in a patient with hyperparathyroidism. The second radiograph (bottom) shows a cyst-like area involving the shoulder which is also a brown tumor.

263

264

FITZPATRICK AND HEATH

described for the total skeleton, the radius and ulna, the lumbar spine, the proximal femur, the tibial cortex, the calcaneus, and the phalanges. A number of papers suggest that 1°HPT is most closely associated with deficits of cortical rather than trabecular bone [60,62,71], but others find roughly equivalent deficits of cortical and trabecular bone or a worse reduction in trabecular bone [61]. There have been many reports over 20 years. Generally, these reports document BMD decrements of about 10% or 1 SD from age-predided means. One isolated report found no differences of BMD between 41 postmenopausal women with 1°HPT and 43 controls [72]. From a counseling standpoint, advice regarding dietary calcium intake in 1°HPT has substantial clinical significance. In 1°HPT, dietary calcium intake does not significantly alter BMD measurements [73].

volume and normal or greater trabecular connectivity than normal subjects in spite of higher bone turnover. The authors suggested that 1°HPT may “protect” against the structural changes in bone histology seen in biopsies from patients with postmenopausal osteoporosis [86]. Studies on BMD and fracture rates, however, provide compelling data regarding the appropriate approach in a clinical setting (see below). In a recent histomorphometric analysis, iliac crest biopsies from patients with 1°HPT had higher values for trabecular wall width, adjusted apposition rate and active formation period than did samples from controls. Thus, the stimulatory effect of PTH on the duration of active bone formation may account, in part, for preservation of cancellous bone in postmenopausal females with 1°HPT [87].

D. Histology of Bone in Primary Hyperparathyroidism

E. Skeletal Fractures in Patients Having Primary Hyperparathyroidism

The appearance of bone in mild to moderate 1°HPT is known mostly from biopsies of the iliac crest and generalized to the rest of the skeleton. Available data are largely congruent, with findings of reduced cortical thickness, preserved or even increased trabecular or cancellous bone volume and connectivity, and evidence of increased bone turnover [73 – 84]. Signs of increased bone resorption consist of increased eroded surfaces and elevated osteoclast counts. Signs of increased bone formation include increased osteoid surface with normal mineralization (by tetracycline labeling), increased osteoblast numbers, and increased mineral apposition rate [73 – 84]. Cortical bone remodeling and its association with fracture risk has been of recent interest. Careful analyses of iliac crest bone biopsies have provided interesting, albeit confusing, information regarding bone architecture in 1°HPT. In biopsies from 39 patients with 1°HPT, cortical porosity was increased in 1°HPT due to high bone turnover with expansion of the remodeling space compared to age- and sexmatched controls. In 23 postmenopausal patients with 1°HPT, a significant increase in resorption depth was balanced by an increase in wall width. These changes in resorption indices may be responsible for the cortical thinning associated with 1°HPT. Postsurgical biopsies were performed in 9 subjects with striking changes in cortical bone metabolism. Activation frequency, porosity, and remodeling space decreased, indicating normalization of the increased bone turnover seen in the preoperative biopsies. These histomorphometric changes are consistent with the increase in BMD that occurs after surgical cure of 1°HPT [85]. In a comparison study of bone biopsies from patients with mild 1°HPT or postmenopausal osteoporosis vs normal subjects, several differences in histology were highlighted. Patients with 1°HPT had normal trabecular bone

The evidence concerning fracture risk in 1°HPT is not extensive. There are few prospective studies, and some of these suffer from a number of study design flaws that limit their generalizability or applicability to subgroups. The reductions of bone mineral density are, on average, modest, being usually 1 SD or less below the age- and sex-specific mean values. Most published studies were small, lacking in control subjects, focused on only one skeletal site or method, and did not permit stratification of patients by such relevant variables as sex, race/ethnic background, concurrent disease, or other risk factors for osteopenia. Several observational studies have assessed the prevalence of vertebral crush fractures in patients with 1°HPT. Results are mixed and difficult to interpret into clinically relevant recommendations. Several studies suggest a small, but significant increase in vertebral fracture risk [15,85,87] but no increases were found in other studies [2,7]. Fewer studies evaluated hip [88,90,71] or distal radius [89] fractures. Recently, Khosla and colleagues evaluated the incidence of age-related fractures in a population-based cohort of 407 cases of 1°HPT observed over 28 years (1965 – 1992) [88]. Standardized incidence ratios (SIR) were determined by comparing new fractures at each site to the number expected from age- and gender-specific fracture evidence rates for the general population. The study subjects had mild 1°HPT, with mean serum calcium of 10.9 mg/dl. Primary HPT was associated with an increased risk of vertebral fractures (SIR 3.2, 95% CI 2.5 – 4.0), distal forearm (SIR 2.2, 95% CI 1.6 – 2.9), rib (SIR 2.7, 95% CI 2.1 – 3.5), and pelvic fractures (SIR 2.1, 95% CI 1.1 – 3.5). Increasing age, female gender, and higher serum levels of calcium were associated with increased fracture risk by univariate analyses. By multivariate analysis, only age and female gender remained significant independent predictors of

CHAPTER 49 Primary Hyperparathyroidism and Hyperparathyroid Bone Disease

fracture risk [88]. With the current trend for conservative, nonsurgical management of 1°HPT, understanding the risks of no intervention in these patients is an important and relevant challenge. Clearly, most of the available data are retrospective or observational, and have limited power, but it is reasonable to conclude that mild-to-moderate 1°HPT does not cause very large increases of fracture risk. This observation seems paradoxical in view of the histological evidence for high bone turnover, cortical bone thinning, and osteopenia at multiple sites. However, the decrements of bone mass are at least partly due to increased remodeling space [73 – 84] and are on average no more than 1 SD below the normal mean. Note that some patients having 1°HPT lose substantial bone mass and that bone mineral density decreases of 1 SD are supposedly associated with increased fracture risk of about twofold, consistent with the findings of Khosla et al. [88]. Currently available data do not help in dealing with the subgroup of hyperparathyroid patients whose bone mass is unusually low; it is unknown whether fracture risk is especially high in such cases. Given the power of low BMD to predict increased fracture risk, and the increases in BMD that occur regularly after surgical cure of 1°HPT, we believe that low BMD in this disorder should be regarded seriously.

V. EFFECTS OF THERAPY ON BONE MASS IN PRIMARY HYPERPARATHYROIDISM While BMD is not strikingly low in most patients having 1°HPT, what of those in whom it is? Can the skeleton recover substantially if hyperparathyroidism is treated medically or surgically? There are extreme views, ranging from the belief in the “irreversibility” of such osteopenia to that of major regain, but most reports support the latter view. Reported effects of parathyroid adenomectomy on radius BMD vary considerably [89 – 91]. Other studies have utilized small numbers of subjects or older, less precise techniques to measure BMD [92 – 96]. Most studies found modest increases in BMD after treatment of 1°HPT, but some reported no postoperative change [90 – 97]. However, more recent studies by DXA have found dramatic postoperative increases in lumbar spine and hip BMD. The most striking recent data on skeletal recovery after treatment of 1°HPT came from Silverberg and colleagues in New York [98], who enrolled 121 patients in a long-term prospective observational study of 1°HPT. These patients have been followed, with or without surgery, for about 10 years. Sixty-one patients were selected for operation, based on published criteria for 1°HPT management [99]. Fortynine surgical patients were asymptomatic and 12 were symptomatic with nephrolithiasis. After surgery, BMD rose strikingly at the lumbar spine and femoral neck and in-

265

creases continued out to 10 years [98]. For the femoral neck, the mean increase over baseline was 12.7  3.8% (mean  SE), and for the lumbar spine 12.8  2.8%. For the radius, there was an increase of 4.0  1.5% at 3 years but no significant chage at year 10. There were no differences in BMD gain between symptomatic and asymptomatic surgical patients. Of the 60 patients who elected medical management, 52 were asymptomatic. In the latter untreated cohort, there were no changes in serum calcium, bone turnover markers, or BMD at any site over a 10-year period, suggesting that ongoing bone loss is iminimal in mild 1°HPT [71]. This observation was also true for a subset of 29 asymptomatic postmenopausal women who might be at higher risk of bone loss, due to estrogen deficiency. By contrast, another subset of 11 patients with the highest baseline serum calcium levels had a mean decrease in BMD of more than 10% over 10 years. Five patients experienced the onset of menopause, which was the only factor identified in logistic-regression analysis as associated with increased risk of BMD loss. Several other subjects had apparent progression of their 1°HPT during observation. Fourteen of 52 asymptomatic patients developed indications for parathyroidectomy, including marked (12 mg/dl) hypercalcemia (n  2), marked hypercalciuria (n  8), and low cortical (distal third of the radius) bone density (Z  2) (n  6). Our views about surgical intervention in 1°HPT based on low BMD have changed because of the prospective 10year study in New York [98]. Initially, it was thought in view of the relative rarity of severe osteopenia [100] and the fact that little bone is lost in untreated patients [71], that medical follow-up would be the best treatment for the asymptomatic 1°HPT patient. This hypothesis is supported by the stability of BMD and biochemical parameters in patients with 1°HPT, and suggested a skeletal anabolic action of endogenous parathyroid hormone. Marcus has discussed these data cogently [101]: “ . . . the substantial gains in BMD that follow parathyroidectomy amount to almost a full standard deviation and should be associated with a substantial improvement in long-term fracture risk. It is difficult to deny patients such [an] advantage by withholding treatment.” Taken together with retrospective studies of fracture rates [88], we believe that a strong argument can be made for surgical intervention in hyperparathyroid patients found to have low BMD or frank osteoporosis. Estrogen replacement therapy has been advocated as treatment for 1°HPT postmenopausal women to attenuate the effects of PTH on bone resorption [102], and favorable effects on BMD have been recently demonstrated. McDermott et al. found 11 to 20% reductions of bone mineral density from control means in 59 women with mild asymptomatic 1°HPT [69]. Osteopenia was absent in those who were taking estrogen replacement therapy. In a doubleblind, randomized, placebo-controlled trial, 42 postmenopausal women with mild 1°HPT were randomly

266

FITZPATRICK AND HEATH

TABLE 1.

Clinical Approach Recommended Based on Synthesis of the Current Literature

• Measurement of BMD at spine and hip is needed before committing to medical follow-up of mild to moderate 1°HPT • If BMD is not reduced more than 2 SD from the young normal mean, then the decision to operate or not may be based on nonskeletal factors • If BMD is more than 2 SD below the young normal mean, then strong consideration must be given to surgical therapy because of the compelling evidence for substantial, long-term gains of bone mass in treated patients • The worse the initial BMD, the greater the relative recovery of skeletal mass, and the greater the need for definitive therapy of the 1°HPT for bone health • If surgical therapy cannot be given, estrogen replacement therapy may mitigate the osteopenia in postmenopausal women • Yearly measurement of BMD and biochemical laboratory tests are warranted for the rare nonsurgical candidate with known osteopenia or osteoporosis Note. No firm assurance can be given to patients about reduction of fracture risk; it can only be reasoned from first principles and retrospective analysis to be likely.

assigned to receive placebo or conjugated equine estrogen 0.625 mg/day and medroxyprogesterone 5 mg/day. After 2 years, BMD in the treatment group increased in value ranging from 3.6 to 6.6% over baseline at all sites except at Wards triangle. No changes were noted in serum calcium or PTH concentrations; however, biochemical markers of bone turnover indicated reduced bone turnover in the treatment group [103]. The effects of progestins on bone in subjects with 1°HPT was evaluated in 15 women. Subjects treated with norethindrone had mean increases of forearm BMC of 1.9% per year over 2 years [104]. Thus, it seems likely that modest increases in BMD can be achieved with hormone replacement therapy in women with untreated 1°HPT. There are no data to indicate whether such effects would lower fracture risk.

VI. SUMMARY AND CONCLUSIONS Variable degrees of osteopenia are common in patients having 1°HPT and osteoporosis may be evident at the diagnosis of 1°HPT. The skeletal deficits occasionally are severe, but usually of undetermined relationship to the hyperparathyroidism. On average, the decrements of bone mass suggest only about a doubling of fracture risk, an increment not discernible in the small studies done to date. The few prospective studies of fracture risk in 1°HPT were not sufficiently powered to adequately address the issue. Osteopenia may be worst at primarily cortical sites, which would suggest a greater risk of appendicular than of spinal crush fractures. Regardless of site or severity of osteopenia, surgical therapy of 1°HPT causes substantially increased BMD at most sites, on the order of 10 to 12%. Increases of such magnitude are rarely seen in therapy of osteoporosis by any other means. Moreover, the increases are larger and may go on for longer periods than could be accounted for by simple filling in of remodeling space. One must reason that decrements of bone mass similar to those seen in 1°HPT increase fracture risk under other circumstances, and assure that restoration of BMD after parathyroid adenomectomy in hyperparathyroid patients should substantially reduce fracture risk.

Severe bone disease caused by 1°HPT is rare. As a group, hyperparathyroid patients have mildly to moderately reduced bone mineral density (about 10% or 1 SD) that may be worst for cortical bone, but which has been observed at all sites. Removal of parathyroid adenomas and restoration of normal parathyroid function causes substantial, lasting increases of BMD (averaging 10 to 12%). Gain of bone occurs at all sites, may go on for up to 10 years, and is greatest in patients having the greatest baseline decrements of BMD. While hyperparathyroid patients as a group appear not to lose bone at greater rates than normal, some individuals do lose bone at abnormal rates. Clearly, a hyperparathyroid patient whose bone mass is below average may become clinically osteoporotic sooner than would otherwise occur. While the foregoing might argue forcefully for early surgical intervention in 1°HPT to protect bone [105], studies of fracture incidence in such patients are few. Thus, there is a great need for a large-scale, prospective treatment trial in 1°HPT that would determine whether definitive treatment of hyperparathyroidism prevents fractures. Until such data are available, clinicians will have to rely on clinical judgment and the suggested clinical approach presented here (see Table 1). At a minimum, the potential for skeletal benefit must be considered in decisionmaking about definitive treatment of 1°HPT.

References 1. H. I. Heath, S. Hodgson, and M. Kennedy, Primary hyperparathyroidism: Incidence, morbidity, and economic impact in a community. N. Engl. J. Med. 302, 189 – 193 (1980). 2. G. Stenstrom and P. Heedman, Clinical findings in patients with hypercalcemia. A final investigation based on biochemical screening. Acta Med. Scand. 195, 473 – 477 (1974). 3. S. Silverberg, J. Bilezikian, G. Bone, G. Talpos, M. Horwitz, and A. Stewart, Therapeutic controversies in primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 84, 2275 – 2285 (1999). 4. M. Parisien, S. Silverberg, E. Shane, L. de la Cruz, R. Lindsay, J. Bilezikian, and D. Dempster, The histomorphometry of bone in primary hyperparathyroidism: Preservation of cancellous bone. J. Clin. Endocrinol. Metab. 70, 930 – 938 (1990).

CHAPTER 49 Primary Hyperparathyroidism and Hyperparathyroid Bone Disease 5. J. Soreide, J. van Heerden, C. Grant, C. Yau Lo, C. Schleck, and D. Ilstrup, Survival after surgical treatment for primary hyperparathyroidism. Surgery 122, 1117 – 1123 (1997). 6. R. Wilson, D. Rao, B. Ellis, M. Kleerekoper, and M. Parfitt, Mild asymptomatic primary hyperparathyroidism is not a risk factor for vertrebral fractures. Ann. Intern. Med. 109, 959 – 962 (1988). 7. W. Vant Hoff, F. Ballardie, and E. Bicknell, Primary hyperparathyroidism: The case for medical management. Br. Med. J. 287, 1605 – 1608 (1983). 8. L. Raisz, Osteoporosis: Current approaches and future prospects in diagnosis, pathogenesis, and management. J. Bone Miner. Metab. 17, 79 – 89 (1999). 9. R. Marcus, Endogenous and nutritional factors affecting bone. Bone 18, 11S – 13S (1996). 10. J. Heersche, C. Bellows, and J. Aubin, Cellular actions of parathyroid hormone on osteoblast and osteoclast differentiation. In J. Bilezikian, R. Marcus, and M. Levine, eds. “The Parathyroids, Basic and Clinical Concepts” pp. 83–92. Raven Press, New York, 1994. 11. F. Albright and E. J. Reifenstein, “The Parathyroid Glands and Metabolic Bone Disease. Selected Studies.” Williams and Wilkins, Baltimore, 1948. 12. H. Steinbach, G. Gordan, E. Eisenberg, J. Carne, S. Silverman, and L. Goldman, Primary hyperparathyroidism: A correlation of roentgen, clinical, and pathologic features. Am. J. Roentgenol. Radium Ther. Nucl. Med. 86, 329 – 343 (1961). 13. H. Genant, J. Baron, F. Straus, E. Paloyan, and J. Jowsey, Osteosclerosis in primary hyperparathyroidism. Am. J. Med. 59, 104 – 113 (1975). 14. R. Dauphine, B. Riggs, and D. Scholz, Back pain and vertebral crush fractures: An unemphasized mode of presentation for primary hyperparathyroidism. Ann. Intern. Med. 83, 365 – 367 (1975). 15. E. Filvaroff and R. Derynck, Bone remodelling: A signalling system for osteoclast regulation. Curr. Biol. 8, R679 – R682 (1998). 16. K. Fuller, J. Owens, and T. Chambers, Induction of osteoclast formation by parathyroid hormone depends on an action on stromal cells. J. Endocrinol. 158, 341 – 350 (1998). 17. M. Rouleau, J. Mitchell, and D. Goltzman, In vivo distribution of parathyroid hormone receptors in bone: Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 123, 187 – 191 (1988). 18. T. Suda, N. Takahashi, and T. Martin, Modulation of osteoclast differntiation. Endocr. Rev. 13, 66 – 80 (1992). 19. M. Herrmann-Erlee, J. van der Meer, and J. Hekkelman, In vitro studies of the adenosine 3,5-monophosphate (cAMP) response of embryonic rat calvaria to bovine parathyroid hormone-(1 – 84) [bPTH-(1 – 84)], bPTH-(1 – 34), and bPTH-(3 – 34) and the loss of cAMP responsiveness after prolonged incubation. Endocrinology 106, 2013 – 2018 (1980). 20. I. Yamamoto, C. Shigeno, J. J. Potts, and G. Segre, Characterization and agonist-induced down-regulation of parathyroid hormone receptors in clonal rat osteosarcoma cells. Endocrinology 122, 1208 – 1217 (1988). 21. J. Mitchell and D. Goltzman, Mechanisms of homologous and heterologous regulation of parathyroid hormone receptors in the rat osteosarcoma cell line UMR-106. Endocrinology 126, 2650 – 2660 (1990). 22. R. Podbesek, C. Edouard, P. Meunier, J. Parsons, J. Reeve, R. Stevenson, and J. Zanelli, Effects of two treatment regimes with synthetic human parathyroid hormone fragment on bone formation and the tissue balance of trabecular bone in greyhounds. Endocrinology 112, 1000 – 1006 (1983). 23. C. Tam, J. Heersche, T. Murray, and J. Parsons, Parathyroid hormone stimulates the bone apposition rate independently of its resorptive action: Differential effects of intermittent and continuous administration. Endocrinology 110, 506 – 512 (1982).

267

24. J. Hock and I. Gera, Effects of continuous and intermittent administration and inhibition of resorption on the anabolic response of bone to parathyroid hormone. J. Bone Miner. Res. 7, 65 – 72 (1992). 25. E. Hefti, U. Trechsel, J. Bonjour, H. Fleisch, and R. Schenk, Increase of whole-body calcium and skeletal mass in normal and osteoporotic adult rats treated with parathyroid hormone. Clin. Sci. 62, 389 – 396 (1982). 26. H. Kaji, T. Sugimoto, M. Kanatani, A. Miyauchi, T. Kimura, S. Sakakibara, M. Fukase, and K. Chihara, Carboxyl-terminal parathyroid hormone fragments stimulate osteoclast-like cell formation and osteoclastic activity. Endocrinology 134, 1897 – 1904 (1994). 27. A. Grey, M. Mitnick, S. Shapses, A. Ellison, C. Gundberg, and K. Insogna, Circulating levels of interleukin-6 and tumor necrosis factor-alpha are elevated in primary hyperparathyroidism and correlate with markers of bone resorption — A clinical research center study. J. Clin. Endocrinol. Metab. 81, 3450 – 3454 (1996). 28. D. Dempster, F. Cosman, M. Parisien, and V. Shen, Anabolic actions of parathyroid hormone on bone. Endocr. Rev. 14, 690 – 709 (1993). 29. D. Slovik, D. Rosenthal, S. Doppelt, J. Potts, Jr., M. Daly, J. Campbell, and R. Neer, Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1 – 34) and 1,25-dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 381 (1986). 30. R. Lindsay, J. Nieves, E. Henneman, V. Shen, and F. Cosman, Subcutaneous administration of the amino-terminal fragment of human parathyroid hormone (1 – 34): Kinetics and biochemical response in estrogenized osteoporotic subjects. J. Clin. Endocrinol. Metab. 77, 1535 – 1539 (1993). 31. J. Finkelstein, A. Klibanski, E. Schaefer, M. Hornstein, I. Schiff, and R. Neer, Parathyroid hormone for the prevention of bone loss induced by estrogen deficiency. N. Engl. J. Med. 331, 1618 – 1623 (1994). 32. J. Finkelstein, A. Klibanski, A. Arnold, T. Toth, M. Hornstein, and R. Neer, Prevention of estrogen deficiency-related bone loss with human parathyroid hormone-(1 – 34): A randomized controlled trial. JAMA 280, 1067 – 1073 (1998). 33. R. Lindsay, J. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, D. Dempster, and F. Cosman, Randomized controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550 – 555 (1997). 34. E. B. Roe, S. D. Sanchez, G. A. del Puerto, E. Pierini, P. Bacchetti, C. E. Cann, and C. D. Arnaud, Parathyroid hormone 1 – 34 (hPTH 1 – 34) and estrogen produce dramatic bone density increases in postmenopausal osteoporosis — Results from a placebo-controlled randomized trial. J. Bone Miner. Res. 14 (Suppl. 1), 1019 (1999). 35. T. Murray, M. Peacock, D. Powell, J. Monchik, and J. J. Potts, Nonautonomy of hormone secretion in primary hyperparathyroidism. Clin. Endocrinol. 1, 235 – 246 (1972). 36. H. I. Heath and D. Purnell, Urinary cyclic 3,5-adenosine monophosphate responses to exogenous and endogenous parathyroid hormone in familial benign hypercalcemia and primary hyperparathyroidism. J. Lab. Clin. Med. 96, 974 – 984 (1980). 37. S. Khosla, P. Ebeling, A. Firek, M. Burritt, P. Kao, and H. I. Heath, Calcium infusion suggests a “set-point” abnormality of parathyroid gland function in familial benign hypercalcemia and more complex disturbances in primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 76, 715 – 720 (1993). 38. H. Harms, K. Prank, U. Brosa, et al., Classification of dynamical diseases by new mathematical tools: Application of multi-dimensional phase space analyses to the pulsatile secretion of parathyroid hormone. Eur. J. Clin. Invest. 22, 371 – 377 (1992). 39. H. Harms, E. Schlinke, O. Neubauer, C. Kayser, P. Wüstermann, R. Horn, W. Külpmann, A. von zur Mühlen, and R. Hesch, Pulse amplitude and frequency modulation of parathyroid hormone in primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 78, 53 – 57 (1994).

268 40. M. Samuels, J. Veldhuis, C. Cawley, R. Urban, M. Luther, R. Bauer, and G. Mundy, Pulsatile secretion of parathyroid hormone in normal young subjects: Assessment by deconvolution analysis. J. Clin. Endocrinol. Metab. 76, 399 – 403 (1993). 41. K. Herfarth, H. Schmidt-Gayk, S. Graf, and A. Maier, Circadian rhythm and pulsatility of parathyroid hormone secretion in man. Clin. Endocrinol. 37, 511 – 519 (1992). 42. G. el-Hajj Fuleihan, E. Klerman, E. Brown, Y. Choe, E. Brown, and C. Czeisler, The parathyroid hormone circadian rhythm is truly endogenous — A general clinical research center study. J. Clin. Endocrinol. Metab. 82, 281 – 286 (1997). 43. F. Logue, W. Fraser, S. Gallacher, D. Cameron, D. O’Reilly, G. Beastall, U. Patel, and I. Boyle, The loss of circadian rhythm for intact parathyroid hormone and nephrogenous cyclic AMP in patients with primary hyperparathyroidism. Clin. Endocrinol. 32(4), 475 – 483 (1990). 44. T. Sinha, S. Miller, J. Feming, R. Khairi, J. Edmondson, C. Johnston, Jr., and N. Bell, Demonstration of a diurnal variation in serum parathyroid hormone in primary and secondary hyperparathyroidism. J. Clin. Endocrinol. Metab. 41, 1009 – 1013 (1975). 45. H. Donahue, M. Fryer, E. Eriksen, and H. I. Heath, Differential effects of parathyroid hormone and its analogues on cytosolic calcium ion and cAMP levels in cultured rat osteoblastlike cells. J. Biol. Chem. 263, 13522 – 13527 (1988). 46. J. Bidwell, M. Fryer, A. Firek, H. Donahue, and H. I. Heath, Desensitization of rat osteoblast-like cells (ROS17/2.8) to parathyroid hormone uncouples the adenosine 3,5 monophosphate and cytosolic ionized calcium response limbs. Endocrinology 128, 1021 – 1028 (1991). 47. J. Bidwell, W. Carter, M. Fryer, and H. I. Heath, Parathyroid hormone (PTH)-induced intracellular Ca2 signalling in naive and PTH-desensitized osteoblast-like cells (ROS 17/2.8): Pharmacological characterization and evidence for synchronous oscillation of intracellular Ca2. Endocrinology 129, 2993 – 3000 (1991). 48. J. Joyce, R. Idea, S. Grossman, R. Liss, and J. Lyons, Multiple brown tumors in unsuspected primary hyperparathyroidism mimicking metastatic disease on radiograph and bone scan. Clin. Nucl. Med. 19, 630 – 635 (1994). 49. C. Kulak, C. Bandeira, D. Voss, S. Sobieszczyk, S. Silverberg, F. Bandeira, and J. Bilezikian, Marked improvement in bone mass after parathyroidectomy in osteitis fibrosa cystica. J. Clin. Endocrinol. Metab. 83, 732 – 735 (1998). 50. N. Dalen and B. Hjern, Bone mineral content in patient with primary hyperparathyroidism without radiological evidence of skeletal changes. Acta Endcrinol. 75, 297 – 304 (1974). 51. C. Pak, R. Kaplan, C. Notz, A. Stewart, A. Bone, and R. Browne, Photon absorptiometric analysis of bone density in primary hyperparathyroidism. Lancet 2, 7 – 8 (1975). 52. L. Tougaard, C. Hau, P. Rodbro, and J. Ditzel, Bone mineralization and bone mineral content in primary hyperparathyroidism. Acta Endocrinol. 85, 314 – 319 (1977). 53. E. Seeman, H. Wahner, K. Offord, R. Kumar, W. Johnson, and B. Riggs, Differential effects of endocrine dysfunction on the axial and appedicular skeleton. J. Clin. Invest. 69, 1302 – 1309 (1982). 54. M. Fuss, C. Gillet, J. Simon, J. Vandewalle, A. Schoutens, and P. Bergmann, Bone mineral content in idiopathic renal stone disease and in primary hyperparathyroidism. Eur. Urol. 9, 32 (1983). 55. R. Possi-Mucelli, A. Kanter, H. Genant, C. Cann, B. Ettinger, and F. Kolb, Quantitative bone mineral analyses in primary hyperparathyroidism. J. Comput. Assisted Tomogr. 7, 555 (1983). 56. L. Jimenez, T. Spinks, A. Ranicar, and G. Joplin, Total body calcium mass in primary hyperparathyroidism and long-term changes. Calcif. Tissue Int. 36, 14 – 18 (1984). 57. J. Andresen, H. Nielsen, and K. Albertsen, Metacarpal bone measurements in renal transplant recipients, in corticosteroid-treated

FITZPATRICK AND HEATH

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

patients with polymyalgia rheumatica and in patients with primary hyperparathyroidism. Acta Med. Scand. 219, 99 – 104 (1986). B. Greenspan, A. Strashun, D. Bergman, M. DaCosta, M. DeLaney, and S. Goldsmith, Evaluation of bone mineral loss in hyperparathyroidism by single and dual photon absorptiometry. J. Nucl. Med. 27, 886 (1986). M. Richardson, R. Pozzi-Mucelli, A. Kanter, F. Kolb, B. Ettinger, and H. Genant, Bone mineral changes in primary hyperparathyroidism. Skeletal Radiol. 15, 85 – 95 (1986). J. Wishart, M. Horowitz, A. Need, and B. Nordin, Relationship between forearm and vertebral mineral density in postmenopausal women with primary hyperparathyroidism. Arch. Intern. Med. 150, 1329 – 1331 (1990). R. Hesp, M. Tellez, L. Davidson, A. Elton, and A. Reeve, Trabecular and cortical bone in the radii of women with parathyroid adenomata: A greater trabecular deficit, with a preliminary assessment of recovery after parathyroidectomy. Bone Miner. 2, 301 – 310 (1987). J. Adams, R. Whitehouse, P. Adams, and M. Davies, The effects of primary hyperparathyroidism on trabecular and cortical bone mass. Calcif. Tissue Res. 46, 144 (1990). S. Silverberg, E. Shane, T. Jacobs, E. Siris, F. Gartenberg, D. Seldin, T. Clemens, and J. Bilezikian, Nephrolithiasis and bone involvement in primary hyperparathyroidism. Am. J. Med. 89, 327 – 334 (1990). Y. Negoro, A. Tomita, T. Masahashi, M. Suzuki, M. Noguchi, and M. Nakanishi, Bone mass and biochemical parameters in metabolic bone diseases. Rinsho Byori 39, 934 – 939 (1991). J. Lopez and G. Gonzalez, Asymptomatic primary hyperparathyroidism: Evaluation of bone mass using dual-photon bone densitomtery. Rev. Med. Chile 120, 282 – 287 (1992). A. Bergenfelz, B. Lindergard, and B. Ahren, Biochemical variables associated with bone density in patients with primary hyperparathyroidism. Eur. J. Surg. 158, 473 – 476 (1992). J. Pfeilschifter, E. Siegrist, C. Wuster, E. Blind, and R. Ziegler, Serum levels of intact parathyroid hormone and alkaline phosphatase correlate with cortical and trabecular bone loss in primary hyperparathyroidism. Acta Endocrinol. 127, 319 – 323 (1992). J. Napal, Y. Cuerno, J. Olmos, J. Riancho, J. Amado, and J. GonzalezMacias, Changes in bone mass in hepatic cirrhosis, chronic obstructive pulmonary disease, insulin-dependent diabetes, and primary hyperparathyroidism. Med. Clin. Barcelona 100, 576 – 579 (1993). M. McDermott, J. Perloff, and G. Kidd, Effects of mild asymptomatic primary hyperparathyroidism on bone mass in women with and without estrogen replacement therapy. J. Bone Miner. Res. 9, 509 – 514 (1994). J. Casez, A. Troendle, K. Kippuner, and P. Jaeger, Bone mineral density at distal tibia using dual-energy x-ray absorptiometry in normal women and in patients with vertebral osteoporosis or primary hyperparathyroidism. J. Bone Miner. Res. 9, 1851 – 1857 (1994). S. Silverberg, F. Gartenberg, T. Jacobs, E. Shane, E. Siris, R. Staron, and J. Bilezikian, Longitudinal measurements of bone density and biochemical indices in untreated primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 80, 723 – 728 (1995). A. Grey, M. Evans, J. Stapleton, and I. Reid, Body weight and bone mineral density in postmenopausal women with primary hyperparathyroidism. Ann. Intern. Med. 121, 745 – 749 (1994). F. Becker, R. Eisenstein, T. Schwartz, and S. Economou, Needle bone biopsy in primary hyperparathyroidism. Arch. Intern. Med. 131, 650 – 656 (1973). C. Tam, A. Bayley, E. Cross, T. Murray, and J. Harrison, Increased bone apposition in primary hyperparathyroidism: Measurements based on short interval tetracycline labeling of bone. Metab. Clin. Exp. 31, 759 – 765 (1982). S. Charhon, C. Edouard, M. Arlot, and P. Meunier, Effects of parathyroid hormone on remodeling of iliac trabecular bone packets in patients with primary hyperparathyroidism. Clin. Orthop. 162, 255 – 263 (1982).

CHAPTER 49 Primary Hyperparathyroidism and Hyperparathyroid Bone Disease 76. P. Delmas, P. Meunier, E. Faysse, and E. Saubier, Bone histomorphometry and serum bone gla-protein in the diagnosis of primary hyperparathyroidism. World J. Surg. 10, 572 – 578 (1986). 77. E. Eriksen, L. Mosekilde, and F. Melsen, Trabecular bone remodeling and balance in primary hyperparathyroidism. Bone 7, 213 – 221 (1986). 78. G. Delling, Bone morphology in primary hyperparathyroidism — A qualitative and quantitative study of 391 cases. Bone Pathol. 5, 147 – 159 (1987). 79. M. Parisien, S. Silverberg, E. Shane, L. de la Cruz, R. Lindsay, J. Bilezikian, and D. Dempster, The histomorphometry of bone in primary hyperparathyroidism: Preservation of cancellous bone structure. J. Clin. Endocrinol. Metab. 70, 930 – 938 (1990). 80. P. Christiansen, T. Steiniche, L. Mosekilde, I. Hessov, and F. Melsen, Primary hyperparathyroidism: Changes in trabecular bone remodeling following surgical treatment — Evaluated by histomorphometric methods. Bone 11, 75 – 79 (1990). 81. M. Parisien, R. Mellish, S. Silverberg, E. Shane, R. Lindsay, J. Bilezikian, and D. Dempster, Maintenance of cancellous bone connectivity in primary hyperparathyroidism: Trabecular strut analysis. J. Bone Miner. Res. 7, 913 – 919 (1992). 82. P. Christiansen, T. Steiniche, A. Vesterby, L. Mosekilde, I. Hessov, and F. Melsen, Primary hyperparathyroidism: Iliac crest trabecular bone volume, structure, remodeling, and balance evaluated by histomorphometric methods. Bone 13, 41 – 49 (1992). 83. P. Christiansen, T. Steiniche, H. Brockstedt, L. Mosekilde, I. Hessov, and F. Melsen, Primary hyperparathyroidism: Iliac crest cortical thickness, structure, and remodeling evaluated by histomorphometric methods. Bone 14, 755 – 762 (1993). 84. L. van Doorn, P. Lips, J. Netelenbos, and W. Hackeng, Bone histomorphometry and serum concentrations of intact parathyroid hormone [PTH(1 – 84)] in patients with primary hyperparathyroidism. Bone Miner. 23, 233 – 242 (1993). 85. H. Brockstedt, P. Christiansen, L. Mosekilde, and F. Melsen, Reconstruction of cortical bone remodeling in untreated primary hyperparathyroidism and following surgery. Bone 16(1), 109 – 117 (1995). 86. M. Parisien, F. Cosman, R. Mellish, M. Schnitzer, J. Nieves, S. Silverberg, E. Shane, D. Kimmel, R. Recker, and J. Bilezikian, Bone structure in postmenopausal hyperparathyroid, osteoporotic, and normal women. J. Bone Miner. Res. 10, 1393 – 1399 (1995). 87. D. Dempster, M. Parisien, S. Silverberg, X. Liang, M. Schnitzer, V. Shen, E. Shane, D. Kimmel, R. Recker, R. Lindsay, and J. Bilezikian, On the mechanism of canellous bone preservation in postmenopausal women with mild primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 84, 1562 – 1566 (1999). 88. S. Khosla, L. I. Melton, R. Wermers, C. Crowson, W. OFallon, and B. Riggs, Primary hyperparathyroidism and the risk of fracture: A population-based study. J. Bone Miner. Res. 14, 1700 – 1707 (1999). 89. P. Martin, P. Bergmann, C. Gillet, M. Fuss, P. Kinnaert, J. Corvilain, and J. van Geertruyden, Partially reversible osteopenia after surgery for primary hyperparathyroidism. Arch. Intern. Med. 146, 689 – 691 (1986).

269

90. P. Martin, P. Bergmann, C. Gillet, M. Fuss, J. Corvilain, and J. van Geertrudyen, Long-term irreversibility of bone loss after surgery for primary hyperparathyroidism. Arch. Intern. Med. 150, 1495 – 1497 (1990). 91. C. Mautalen, H. Reyes, G. Ghiringhelli, and G. Fromm, Cortical bone mineral content in primary hyperparathyroidism. Changes after parathyroidectomy. Acta Endocrinol. 111, 494 – 497 (1986). 92. E. Scalzetti, E. Levinsohn, P. Numann, and D. Bassano, Vertebral mineralization after parathyroidectomy: Measurement by quantitative CT. Am. J. Roentgenol. 151, 533 – 535 (1988). 93. E. Alhava, P. Karjalainen, and M. Paakkonen, Bone mineral density and surgical treatment of primary hyperparathyroidism. Acta Chir. Scand. 154, 345 – 347 (1988). 94. R. Antonelli, D. Sudhaker Rao, R. Kleerekoper, D. Nelson, and A. Parfitt, Incomplete recovery of bone mineral deficit after parathyroidectomy (PTX) in primary hyperparathyroidism. Calcif. Tissue Res. 46, 144 (1990). 95. M. Block, G. I. Dailey, and D. Muchmore, Bone demineralization, a factor of increasing significance in the management of primary hyperparathyroidism. Surgery 106, 1063 – 1068 (1989). 96. J. Warner, P. Clifton-Bligh, S. Posen, A. McElduff, and L. T. R. Delbridge, Longitudinal changes in forearm bone mineral content in primary hyperparathyroidism. J. Bone Miner. Res. 6 (Suppl. 2), S91 – S95 (1991). 97. A. Bergenfelz, B. Lindergard, and B. Ahren, Prediction of changes in bone density after operation for primary hyperparathyroidism. Ann. Chir. Gynaecol. 82, 245 – 249 (1993). 98. S. Silverberg, E. Shane, T. Jacobs, E. Siris, and J. Bilezikian, A 10-year prospective study of primary hyperparathyroidism with or without parathyroid surgery. N. Engl. J. Med. 341, 1249 – 1255 (1999). 99. Anonymous, Diagnosis and management of asymptomatic primary hyperparathyroidism. National Insitutes of Health Consensus Development. Consensus Statement 8, 1 – 18 (1990). 100. S. Silverberg, E. Shane, L. de la Cruz, D. Dempster, T. Jacobs, E. Siris, M. Parisien, D. Seldin, and J. Bilezikian, Skeletal disease in primary hyperparathyroidism. J. Bone Miner. Res. 3, S91 (1988). 101. R. Marcus, Bones of contention: The problem of mild hyperparathyroidism. J. Clin. Endocrinol. Metab. 80, 720 – 722 (1995). 102. R. Marcus, P. Madvig, M. Crim, A. Pont, and J. Kosek, Conjugated estrogens in the treatment of postmenopausal women with hyperparathyroidism. Ann. Intern. Med. 100, 633 – 640 (1984). 103. A. Grey, J. Stapleton, M. Evans, M. Tatnell, and I. Reid, Effect of hormone replacement therapy on bone mineral density in postmenopausal women with mild primary hyperparathyroidism. A randomized, controlled trial. Ann. Intern. Med. 125, 360 – 368 (1996). 104. J. Wishart, M. Horowitz, A. Need, B. Chatterton, and B. Nordin, Treatment of postmenopausal hyperparathyroidism with norethindrone. Long-term effects on forearm mineral content. Arch. Intern. Med. 150, 1951 – 1953 (1990). 105. S. J. Silverberg, J. P. Bilezikian, H. G. Bone, G. B. Talpos, M. J. Horwitz, and A. F. Stewart, Therapeutic controversies in primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 84, 2275 – 2285 (1999).

CHAPTER 50

Osteogenesis Imperfecta and Other Defects of Bone Development as Occasional Causes of Adult Osteoporosis JAY R. SHAPIRO

I. II. III. IV.

Kennedy Krieger Institute, Baltimore, Maryland 21224; and Uniformed Services University, Bethesda, Maryland 20814

V. Osteoporosis in Inherited Hematologic Disorders VI. Osteoporosis in Inherited Hepatic Disease References

Scope of the Problem Osteogenesis Imperfecta as a Cause of Adult Osteoporosis Histology of Bone in Osteogenesis Imperfecta Osteoporosis in the Heritable Disorders of Connective Tissue

I. SCOPE OF THE PROBLEM

osteoporosis is being recognized with increasing frequency in young adult women and men, the clinician is faced with a differential diagnosis that may range from an inherited disorder such as mild osteogenesis imperfecta, to include acquired endocrine, gastrointestinal, and renal disorders. Representative of these are hyperparathyroidism, occult malabsorption, and idiopathic hypercalcuria. Osteoporosis associated with malignancy in a young adult is also a factor in differential diagnosis. However, a difficult diagnostic and therapeutic situation confronts the clinician attending a young male or premenopausal female when these disorders are eliminated and the diagnosis is primary idiopathic osteoporosis.

It is recognized that the age-specific incidence of osteoporosis appears to be increasing both in women and in men [1]. Although viewed as an issue primarily for postmenopausal women and the elderly, interest in the prevention of osteoporosis has focused attention on its recognition in younger populations and in men [2]. This is facilitated by the widespread availability of bone density measurements and the newer biochemical markers of bone turnover that together provide clinicians with valid estimates of skeletal function. Both modalities assist in earlier diagnosis and a better evaluation of the results of treatment. Because

OSTEOPOROSIS, SECOND EDITION VOLUME 2

271

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

272

JAY R. SHAPIRO

TABLE 1

Adult Osteoporosis: Heritable Defects in Skeletal Development

Heritable disorders of connective tissue Osteogenesis imperfecta Homocystinuria Marfan syndrome Ehlers – Danlos syndrome Copper deficiency syndromes Idiopathic juvenile osteoporosis Idiopathic osteoporosis in young adults (?) Hematologic disorders Thalassemia syndromes (thalassemia minor) Sickle cell anemia (SC disease, etc.) Pernicious anemia (cobalamin deficiency) Gastrointestinal disorders Hemochromatosis

A consideration of the etiology of osteoporosis in young adults between the ages of puberty and age 50 years invokes two concerns: (1) the genetic background of the individual and (2) the role of a heritable defect in connective tissue synthesis leading to diminished bone mass [3]. Both factors may underlie the failure to achieve peak young adult bone mass, now recognized as leading to osteoporosis in later life (see Chapters 25 and 43). New developments in the molecular biology of connective tissues, particularly the definition of mutations affecting the synthesis of types I, II, and III collagens, have focused research perspectives on the possibility of mutations affecting other skeletal matrix components besides the collagens. Thus, osteoporosis related to a heritable disorder may present at any age. This chapter is about the differential diagnosis of osteoporosis in adults, in particular where the clinical findings suggest osteogenesis imperfecta, idiopathic osteoporosis, or other genetic disorders that have clinical features in common with these syndromes (Table 1). Where possible, we will discuss the probable etiology of these disorders, differences in clinical presentation among the different syndromes, and appropriate therapies. The reader is forewarned that osteopenic young adults are commonly encountered in clinical practice and that the distinction among these various syndromes may be difficult. Here, the focus will be on current events in a field where advances in technology have quickened the pace of discovery.

II. OSTEOGENESIS IMPERFECTA AS A CAUSE OF ADULT OSTEOPOROSIS A. Introduction Osteogenesis imperfecta (OI), the brittle bone syndrome, was initially listed among the heritable disorders of

connective tissue by Victor McKusick in 1972 [4]. The efforts of many investigators and the application of state-ofthe-art methods in protein chemistry and molecular biology reveal the central role of mutations affecting the type I collagen genes COLA1 and COLA2 in the genesis of this disorder [5]. As our understanding of the genetic basis of OI has broadened, and as previously unsuspected cases are encountered, the disorder has moved from the status of a “rare” genetic syndrome to one that is more commonly encountered in the community. Although the more severe phenotypes are usually apparent at birth or in early childhood, mild OI may not be suspected until later in life when it becomes a diagnostic consideration in an osteopenic adult.

B. Definition OI is an inherited syndrome characterized by fragile bones (fragilitas ossium) and recurrent fractures that, in severe cases, lead to skeletal deformities. It is a heterogeneous disorder both in terms of inheritance and phenotypic expression [6]. Associated clinical signs of OI include short stature, blue sclerae, dentinogenesis imperfecta, adult onset hearing loss, scoliosis, and joint laxity. In milder cases, skeletal deformities may not occur in spite of multiple fractures and there may be near-normal height. More severely affected individuals may be born with multiple fractures and have skeletal deformity from birth onward.

C. Clinical Classification of Osteogenesis Imperfecta The classification of OI now in general use was devised by Sillence and coworkers in 1979 and modified in 1986 at the 7th International Congress of Human Genetics, Berlin (Table 2) [7]. The distinguishing characteristics in this schema include blue sclerae, the mode of inheritance (dominant, recessive, sporadic/new mutation), and the severity of the disorder as determined by the incidence of fractures and the degree of skeletal deformity. Subjects without dentinogenesis imperfecta in each clinical category are designated as “A” group. “B” refers to subjects with dentinogenesis imperfecta. However, in as many as 25% of OI subjects, accurate classification is hampered because phenotypic overlap complicates estimates of clinical severity and scleral color in an individual patient. To illustrate this problem, type I (mild) OI is uniformly associated with blue sclerae. Scleral color is usually white in adults with severe type III disease but may occasionally be blue in children and in some type III adults. Type IV OI is defined on the basis of sclerae that are blue at younger ages but white in adults [6]. However, some adults considered as type IV (moderately

273

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

TABLE 2 Type

Clinical Classification of Osteogenesis Imperfecta Features

Inheritance

I. Mild IA

IB II. Lethal

IIA

IIB IIC

III. Severe, progressive

Osseous fragility Short stature Blue sclerae Normal teeth Presenile hearing loss Above, plus dentinognensis imperfecta

Autosomal dominant (heterogeneous)

Perinatal severe malformation Intrauterine fractures Stillbirth or neonatal death Broad crumpled bones Beaded ribs Short Broad bones, no/or limited beading Slender limb bones Thin nonbeaded ribs

Autosomal dominant Gonadal/somatic mosaicism

Severe osseous fragility Marked limb shortening Severe scoliosis Growth retardation Neonatal fractures Severe deformity White sclerae Dentinogenesis imperfecta

Autosomal dominant or autosomal recessive

IV. Moderately severe IVA

IVB

Autosomal dominant Scoliosis Blue sclerae when young White sclerae as adult Moderate skeletal deformities Scoliosis Normal teeth Same as above plus dentinogenesis imperfecta

severe) also have blue sclerae. Furthermore, patients report day-to-day variation in scleral color. Another problem in classification arises when an infant with apparent type II (lethal) disease, survives the neonatal period and is labeled a type III case (severe progressive disease). For these reasons, the author recommends that one avoid strict assignment of a clinical type early in the course of OI because the long-term prognostic value is limited, particularly with respect to the incidence of fractures, the level of disability, and the future social achievement of the individual. Clinical definition of OI is also complicated by the fact that, at this time, there is no correlation between a clinical phenotype and a specific type I collagen mutation. It does appear however, that mutations localized to the N-terminal domain of the collagen helix are associated with milder phenotypes, while those at the C-terminus (from which intracellular collagen assembly is initiated) tend to be more severe.

D. Prevalence of OI Given that many subjects with mild disease remain undiagnosed, there are probably 50,000 affected individuals in the United States. Mild type I disease accounts for approximately 60% of patients. The estimate of occurrence for severe or lethal disease is about 3 – 4 cases/100,000 births. Less severe disease is estimated to occur in 4 – 5 cases/100,000 births [8,9]. Estimates based on the presence of fractures at birth range from 1.6/100,000 in Singapore [10] to 3.3/100,000 in France [11] and 15/100,000 in the United Kingdom [12]. The incidence of OI is also underestimated in the case of aborted lethal or severe type III cases where diagnosis at birth is not possible. Mild cases also go undiagnosed, many until later in life. The estimated incidence of sporadic, as opposed to familial, disease ranges from 19 to 34% [13,14]. Parental somatic or gonadal mosaicism may hinder understanding transmis-

274

JAY R. SHAPIRO

sion of the disorder (14a). In certain individuals, in the absence of bone density measurements to document osteopenia, blue sclerae may be the only overt sign of the disorder. Fractures may be rare or may never occur in some individuals. An additional point of diagnostic confusion is that blue sclerae occur in several of the heritable disorders of connective tissue (e.g., in some patients with Ehlers – Danlos syndrome) as well as in apparently normal individuals [4]. OI has been reported throughout the world with no selection for race or gender.

E. Mutations of Type I Collagen in Osteogenesis Imperfecta 1. COLLAGEN Type I collagen is the major structural protein of bone, skin, tendon, ligament, and dentin. Exciting advances have been made during the past decade in the detection of mutations affecting the two genes, COLIA1 and COLIA2, that encode type I procollagen in subjects with osteogenesis imperfecta. Polymorphisms identified in the promotor region of type I collagen genes have recently been associated with idiopathic and with postmenopausal osteoporosis in adults [15]. Progress in this field has been summarized in several recent reviews and the reader is directed to these for an overview of the molecular biology of collagen genes and its application to OI and the other heritable disorders of connective tissue [16,17] (see also Chapter 4). Current methods for mutation analysis involve a combination of denaturing gel electrophoresis combined with sequencing of gene segments identified by electrophoresis [18]. Direct sequencing of pro-alpha 1(I) and pro-alpha 2(I) genes is also employed for mutation analysis. Over 200 type I collagen mutations have been reported in studies using cultured OI fibroblasts and osteoblasts. These mutations have been identified in approximately 90% of the cases that have been studied. However, several reports have failed to define collagen mutations in some OI cases, even after detailed analysis of these genes [19]. Thus, the possibility exists that defects in the synthesis of other matrix components may also be responsible for the OI phenotype. The collagens are a family of proteins that share certain structural homologies but differ to the extent that they serve tissue-specific functions. Nineteen collagen types that contain at least one collagen triple helix, encompassing at least 28 separate genes have been identified [20,21]. These have been separated into several groups: Class I collagens are fibril-forming banded collagens (types I, II, III, V, XI) that may contain two or three collagen types in each fibril. Class II collagens comprise collagen types IX and XII, which adhere to the surface of banded (Class I) collagens. Class III collagens include molecules that form independent fiber systems,

such as basement membranes (type IV), beaded filaments (type VI), and anchoring fibrils (type VII), as well as type X collagen, which forms a network surrounding hypertrophic chondrocytes in cartilage. Class IV collagens contain several proteins with unknown fiber forms and with undefined functions. The “FACIT” or fibril-associated collagens with interrupted triple helices are types IX, XII, XIV, and XIX. All these collagens have short triple helical regions interrupted by short noncollagenous segments. The range of collagen functional specificity is illustrated by the large type I triple-helical collagen polymers that provide strength and elasticity to bone matrix and tendon, the short type VII fibrils that form anchoring fibrils to bind epithelial membranes to dermis, and the type IV and VIII collagens that form basement membranes and Decemet’ membrane [22]. It has been recently appreciated that, in addition to types IX and XII collagen, types III and V collagen are associated with the surface of type I collagen and that type XI is associated with type II collagen [23,24]. Illustrative of the potential complexity of the matrix environment for collagen is decorin, a ubiquitous small proteoglycan that is associated with the surface of types I, II, and VI collagen fibers [25,26]. Decorin may be involved in the inhibition of cell proliferation, perhaps mediated by its ability to bind transforming growth factor (TGF)beta [25]. The biochemistry and molecular biology of collagen, decorin, and other bone matrix constituents are discussed comprehensively in Chapter 4 by Robey and Boskey. Type I collagen is a heterotrimer composed of three polypeptide chains termed alpha [(I)] chains. Two pro- (I) and one pro-2(I) chains are coiled around each other in a triple-helical configuration of approximately 1000 amino acids. Assembly of procollagen -chains occurs in the rough endoplasmic reticulum, the chains assembling in the carboxy-to amino-terminal direction. N- and C-terminal extension peptides maintain solubility of the procollagen molecule during intracellular processing. The basic unit of the collagen (I) chain is the repeating triplet (gly-x-y) where about 20% of the x and y residues are proline and prolines in the y position are hydroxylated to hydroxyproline. Glycine residues facilitate the helical configuration due to their small size and repeating position in the triplet. Collagen biosynthesis involves a series of complex intracellular posttranslational modifications [28] (Fig.1). Hydroxylation of proline residues stabilizes the triple-helical configuration while the formation of lysine aldehyde groups (lysyl oxidase mediated) facilitates the formation of intramolecular cross-links. The collagenbinding chaperone protein HSP47 interacts with Gly-x-y repeats in the triple helical region and acts to stabilize the procollagen molecule in the endoplasmic reticulum [29]. Intracellular procollagen processing and its subsequent secretion into the extracellular space leads to cleavage of procollagen extension peptides by specific N- and C-terminal

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

275

proteases. Self-assembly and cross-linking of individual triple helical molecules into a large collagen polymer occurs in the extracellular space. Hydroxylation and glycosylation of lysine residues permits the formation of stabilizing cross-links between collagen fibrils. The cleaved C-terminal propeptide extensions are not further metabolized but circulate in plasma as markers of collagen biosynthesis. Measurement of these propeptides is used clinically to estimate collagen turnover [30] (see Chapter 60). 2. TYPE I COLLAGEN MUTATIONS AND OI

FIGURE 1 The intracellular and extracellular events involved in the formation of a collagen fibril. Note that assembly is from the C- to the N-terminus. In addition to transcriptional events, a large number of enzymatic steps are involved in post-translational processing of the procollagen molecule. As an example of how the collagen fibrils can form ordered arrays in the extracellular space, they are shown assembling into large collagen fibers. The covalent cross-links that stabilize the extracellular fibrils are not shown. Reprinted with permission from Garland Publishing.

The type I collagen genes, COLIA1 and COLIA2, are large genes encoding the repeating (gly-x-y) triplet and procollagen extension peptides. COLIA1 (chromosome 17q23.3 – q22) is 51 exons or 18 kb long. The COLIA2 gene (chromosome 7q21.3 – q22) is 52 exons and 40 kb in size. The exons encoding the triple helical portion of the protein are multiples of the basic 9-base-pair sequence (e.g., 54, 108 base pairs) that encodes the (gly-x-y) triplet. Over 200 type I collagen mutations have been reported to date in OI patients. The reader can access a complete list of type I (and type III) collagen mutations in the Database of Human Type I and Type III Collagen Mutations compiled by R. Dalgliesh [31] (http://www.le.ac.ul/genetics/ collagen) [31]. Collagen mutations have been classified by Cole [32] according to the locus of the mutation. These include: (i) null allelic mutations, (ii) helical mutations (glycine substitutions, splicing mutations, and helical deletions and insertions), (iii) carboxy-telopeptide and carboxy-propeptide mutations, and (iv) amino-telopeptide and aminopropeptide mutations. Mapping mutations in these large genes poses a technical obstacle. Several methods have been developed to define mutations in cultured fibroblasts and osteoblasts. On get electrophoresis of secreted pro-1(I) and pro-2(I) chains by cultured fibroblasts, one can recognize abnormal peptide mobility due either to posttranslational overmodification (excess glycosylation or hydroxylation) resulting from slowed intracellular processing in the presence of a mutation or an alpha chain doublet representing the product of a mutated allele [33]. Detection of collagen gene mutations has been facilitated by the use of the polymerase chain reaction (PCR) to amplify mutated segments of the gene [18]. PCR-based methods now in common use include analysis by single-strand conformational polymorphisms (SSCP) [34], electrophoresis of heteroduplex DNA [35], and chemical cleavage [36] or RNAse protection of mismatched sense – antisense cDNA or RNA hybrids [37]. Technology now permits direct sequencing of double-stranded PCR products but this is a complex and labor-intensive undertaking [38]. Any alteration in the normal amino acid composition of the type I procollagen chains will destabilize the triplehelical portion of the molecule. Although frequently

276 reflected as diminished thermal stability (i.e., melting temperature) of the procollagen heterotrimer, normal thermal stability has been observed in the presence of similar mutations, demonstrating that the effects of glycine substitutions on the thermal unfolding of type I collagen may be highly position-specific [39]. This was illustrated by an 1 gly598serine mutation that dramatically decreased thermal stability, whereas an 1 gly631-serine mutation had normal thermal stability [40,41]. The substitute amino acid also affects thermal stability in that glycine – arginine mutations have a greater effect on stability than do glycine – cysteine mutations. However, there is no relationship between thermal stability and phenotype. Two general classes of type I collagen mutation have been described in OI. In one, “the null allele,” a mutation affecting either the pro-1(I) or pro-2(I) alleles impairs gene transcription. mRNA stability, or intracellular processing of a mutated polypeptide chain so that only half of the expected amount of heterotrimer is secreted into the extracellular space. The secreted procollagen is normal in composition but deficient in quantity. This mechanism has been recognized in cases of mild, type I OI [42]. A second mechanism involves structurally abnormal pro-1(I) chains that are assembled into heterotrimers that, although defective, are secreted and incorporated into extracellular matrix. As a consequence of intracellular degradation of mutant collagen termed “protein suicide,” the net result of either mechanism is that less bone matrix is synthesized [43]. Where defective procollagen is incorporated into matrix, the bone will be both quantitatively and qualitatively defective. Either mechanism ultimately increases the susceptibility to fracture. A third possible type of mutation, one affecting the regulatory portions (promoter, enhancer regions) of the COLIA1 or COLIA2 genes, has not been reported. a. The Concept of Dominant/Negative Mutation Type I procollagen is a heterotrimer consisting of two identical pro1 chains and a structurally different pro-2 chain. A mutation affecting one 1(I) allele will alter the synthesis of 50% of those 1 chains with incorporation of either one or two mutated chains into 43 of the total number of procollagen molecules. Thus, the negative effect of the one dominant mutation is amplified. With the 2(I) chain, a mutated pro-2(I) would be incorporated into 50% of the type I molecules. This assumes equal production of the mutated and normal chains as well as equal access to procollagen assembly. b. Relating Genotype to Phenotype in OI It has proven difficult to formulate a cohesive theory to explain the relationship between specific collagen gene mutations and the resulting OI phenotypes. Because the molecule assembles from the C- to the N-terminal direction, C-terminal mutations tend to be clinically more severe than N-terminal mutations. However, this rule is breached by several exam-

JAY R. SHAPIRO

ples of mutations that have inconsistent effects on the resulting phenotype both within and among affected families [44]. One explanation gleaned from the effects of different mutation loci on the patterns of thermal unfolding of type I collagen suggests that specific domains (“cooperative melting domains”) of the procollagen chains constitute regions specifically susceptible to altering chain assembly or stability [45]. Similar mechanisms may explain the manner in which like mutations located at adjacent loci have dramatically different effects on the expressed phenotype. For example, deletion of exon 11 produces a phenotype like Ehlers – Danlos syndrome without significant bone disease, while deletion of exon 12 produces type IV OI. The basic unit of the triple-helical region is the repeating triplet gly-x-y. Point substitutions affecting the first two nucleotides of the GGN codon for glycine would be expected to produce first position glycine substitutions by eight amino acids: alanine, arginine, aspartic acid, cysteine, glutamic acid, serine, valine, and tryptophan [18]. Substitutions of glycine by cysteine are the most frequent. Glycine substitutions involving larger amino acids (arginine, alanine, aspartate, and serine) have been associated with greater disruption in procollagen assembly and, in general, more severe phenotypes. In addition to point substitutions, there occur deletions and insertions of various sizes from single bases to entire exons that affect the helical portion of the molecule. Mutations that alter consensus donor splice sites, leading to deletions or insertions of whole introns or exons, have been reported. In addition, such mutations induce the formation of premature stop codons that terminate translation, resulting in truncated  chains and triple-helix destabilization. Mutations that affect the coding region for the helical portion of the pro- chains are frequently associated with metabolic abnormalities including delayed or retarded secretion into the extracellular space, increased intracellular degradation of the mutant chain, and decreased thermal stability of the protein [5]. Dilatation of the endoplasmic reticulum of OI fibroblasts occurs due to retained intracellular protein. Secreted heterotrimeric collagen molecules containing mutated  chains will intefere with normal fibril formation [46]. One mutation, glycine 748-cysteine induced the formation of a kink at the site of mutation and altered a proteinase cleavage site a sizable distance away [47]. Although initial reports of mutations failed to disclose the same mutation in different families, this has now been documented. These mutations include an 2(I) gly859 – ser substitution in two unrelated subjects with OI type III [48], two OI type II individuals having the same gly 154 – arg substitution, and two with type II OI sharing an 1(I) gly1003 – ser substitution [49]. Wenstrup reported two families with OI type IV in which affected individuals were found to harbor an 2(I) gly646 – ser mutation [50]. Unrelated subjects with the same 1(I) gly352 – ser mutation have also been identified [51 – 53].

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

FIGURE 2 Type I OI. Osteopenia is the characteristic radiologic finding in type I OI. Bone density may vary from near normal to markedly deficient. Typically the cortices are thin and the medullary trabecular pattern is deficient. However, unlike more severe phenotypes, there is normal architecture in the epiphyseal zone and the growth plate remains intact.

F. Clinical Overview of OI Approximately 60% of recognized OI cases are type I, 15% are type II, 20% are type III, and 5% are classified as type IV (Table 2). 1. TYPE I OI Type I OI is the mildest and the most prevalent form of the disease (Fig. 2). However, even within this group there is considerable phenotypic heterogeneity. Transmission is autosomal dominant. The diagnosis may be missed in very mild cases even with several affected members in one family. For example, a 52-year-old woman considered to represent a case of postmenopausal osteoporosis was recently reported in whom an [2(I)] glycine 661 – serine collagen mutation consistent with OI was found [54]. However, attention to the history and physical findings revealed that her first of five fractures occurred at age 7 and that her 26-yearold son had also suffered fractures. The patient had blue sclerae and slight hearing loss. Thus, although the age of onset or number of fractures may be of little help in establishing the diagnosis of OI, the family history may be important in establishing the probable genetic basis of the disorder.

277

The cardinal manifestations of type I disease include a history of multiple fractures, usually dating from childhood, blue sclerae, hearing loss that is evident between ages 20 and 30, mild joint laxity, and short stature [52]. One-third of type IA patients are less than the 3rd percentile in height, an equal number are between the 3rd and the 50th percentile, and 10% may have normal height [55]. Subjects with type IB OI tend to be shorter. A characteristic triangular facies occurs in many patients. In spite of multiple fractures, and in contrast to more severe OI types, skeletal deformity may be mild or absent. Dentinogenesis imperfecta (type IB) occurs in between 5 and 15% of type I subjects [56]. The fracture rate is reported to be higher in subjects with dentinogenesis imperfecta. The expression of each of these traits, including the incidence of fractures, is highly variable among and within families. A partial explanation for significant variation in expression is somatic or gonadal mosaicism (vide infra). Individuals with type I disease usually have fractures in early childhood, but they may not experience a first fracture until their teens or later. Fracture incidence has a biphasic pattern, decreasing after puberty, and rising again in women and men with increasing age [57]. Scoliosis is also of a mild degree. Radiologic examination shows a well-proportioned outline of the appendicular skeleton with intact epiphyseal architecture. There are varying degrees of osteopenia that may approach normal bone density in certain patients [58]. Vertebral osteoporosis is usually present and vertebral compression fractures may occur by the 30s. However, there are mild type I OI cases, in whom glycine mutations in type I collagen  chanis have been found, who have not had fractures despite having radiologic osteoporosis, blue sclerae, joint laxity, short stature, and a dominant pattern of transmission [31]. Here, the differential diagnosis would include idiopathic osteoporosis in a young adult [59]. However, distinctions have been drawn between these syndromes. In the author’s experience, subjects with idiopathic osteoporosis have white sclerae, do not have dentinogenesis imperfecta, and tend to be taller than individuals with type I OI. In common with OI, these individuals have mild joint laxity and mild scoliosis. Examination of tetracycline-labeled bone biopsy specimens from individuals with type I disease has usually revealed low bone turnover in adults. However, elevated remodeling rates previously reported [60] for other OI types [61] may also occur in type I OI children [62]. a. Null Allelic Muations in Type I OI The majority of cultured fibroblast strains from subjects with type I OI secrete approximately one half the normal amounts of collagen, i.e., the product of one normal allele [42,63]. Also, the secreted protein is not abnormal by PAGE, suggesting that the product of the mutated allele has either not been

278

JAY R. SHAPIRO

transcribed or has been degraded intracellularly (e.g., null allele). The ratio of pro-1(I) to pro-1(III) collagen is also abnormal approximating I rather than the expected 3 [64]. The defect in type I collagen synthesis is not restricted to pro-1(I) since certain OI strains have been found to harbor mutations affecting pro-2(I) in subjects expressing the same clinical phenotype. Diminished pro-1(I) mRNA has also been reported in association with the null allele [42]. It is probable that several types of mutations may be expressed as a null allele. As indicated above, these may involve mutations affecting the promoter or enhancer regions, splicing defects, deletions or insertions that induce a premature stop codon, and finally, structural rearrangements that induce a nonfunctional allele [63]. It appears that may be associated with an increased nuclear to cytoplasmic ratio of pro-1(I) mRNA secondary to a failure of translocation of mutant mRNA out of the nucleus [64,65]. This was detected only in strains where a splicing mutation produced a product that was out of frame (frameshift mutations) and for unexplained reasons, was retained in the nucleus. mRNA bearing in-frame mutations are apparently transported to the cytosol, processed, and secreted into the extracellular matrix. This type of mutation is potentially associated with a more severe clinical phenotype [66]. 2. TYPE II OI (LETHAL PERINATAL) Lethal OI has been the subject of considerable investigation as to the relationship between the loci of COLIA1 and COLIA2 mutations and the devastating effects that these have on the formation of extracellular matrix in bone and other organs. Infants with lethal disease usually succumb to pulmonary insufficiency during the first 3 months of life. Although initially considered the result of recessive mutations, type II OI is now recognized as the result of dominant heterozygous disease. Gonadal mosaicism has also been associated with type II disease and accounts for instances of variable expressivity within affected families [67]. These infants are small for dates, with shortend, deformed extremities and deep-blue sclerae. Organ involvement is widespread, manifested as severe skeletal deformity with diffuse fracture involvement of the extremities, ribs, calvarium, and spine. Apgar scores are depressed in type II OI at birth. The major life-threatening complication is ventilatory insufficiency due either to mechanical factors secondary to multiple rib fractures or to primary pulmonary insufficiency. Secondary complications include traumatic brain hemorrhage, spinal cord injury, and avulsion of body parts during delivery. The radiologic picture of type II OI is distinctive (Fig. 3). The differential diagnosis includes severe infantile hypophosphatasia, thanatophoric dwarfism, asphyxiating thoracic dystrophy and achondroplasia. However, the

FIGURE 3 Type II OI. The hallmarks of this phenotype in the neonate are multiple fractures of the extremities associated with a characteristic “concertina” deformity of the lower extremities. Both a broad bone and a narrow bone appearance of the extremities may occur. The ribs have been fractured in utero and there is beading or callus seen at birth. In certain cases there is relatively little beading and the ribs appear narrow. latter differ in that widespread failure to ossify characterizes hypophosphatasia and alkaline phosphatase activity is low. In achondroplasia, the bones are short and tubular. Asphyxiating thoracic dystrophy is associated with a narrow or bell-shaped thorax and may mimic the pulmonary insufficiency of type II OI [68]. In OI, the cranial vault is markedly undermineralized, paper thin, and may show considerable molding. Wormian bones, multiple small islands of underossified cranium, are visible in the occipital and parietal regions [69]. The extremities appear foreshortened due to fractures and appear widened, crumpled (“concertina” appearance), and markedly demineralized. Both the upper and the lower extremities, clavicles, and ribs contain multiple fractures. The spine shows platyspondyly. The ribs are typically narrow, and sometimes exhibit a beaded appearance due to the presence of healing intrauterine fractures. Variations in the X-ray appearance have led to a subclassification of type II OI based on the appearance of the extremities and ribs [70].

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

In type IIA, the extremities are short and broad in appearance, the ribs are beaded, and there is platyspondyly. Rib changes are less marked in type IIB, whereas in type IIC, the limbs show slender bones and the ribs are thin with little beading. Infants with type II OI either die at birth or survive for periods of days to weeks. Occasionally infants will survive for several months depending on the available nutritional and ventilatory support. However, the integrity and maturity of the thoracopulmonary system usually determines the outcome, with infection always a risk. Pulmonary hypoplasia occurs in type II disease, although the prevalence of this lesion is unknown. In one case associated with a mutation affecting pro-1(I) collagen, arrest of bronchoalveolar development appeared to occur at the 10th week of gestation [71]. The histology of bone in type II OI demonstrates markedly defective cortical and trabecular bone formation (woven bone). The process of endochondral bone formation at the epiphysis is disorganized, leading to persistent islands of cartilage and undermineralized bone. Membranous bone formation is similarly deficient, resulting in marked calvarial thinning. a. Mutations Associated with Type II OI Mutations associated with the lethal form of OI frequently involve single base substitutions and deletions or insertions of various sizes affecting the C-terminal domains of the helical region or the C-propeptide of either the COLIA1 or the COLIA2 genes. However, as in phenotypes III and IV, splicing mutations, deletions, and insertions of central (gly256 – val]) [72] or N-terminal domains (gly97 – asp)] have also been described [5]. It is of interest that similar mutations have also been associated with milder type I disease [73]. Multiple single-base substitutions have been reported, the majority in the C-terminal portion of the helical region. Glycine to alanine [74], arginine [74], aspartate [75,76], serine [74,49], and valine substitutions have been particularly common in lethal OI. Lethal disease associated with a glycine to arginine substitution at position 298, a more Nterminal site has also been reported [78]. A COLIA2 gly343 – glutamine substitution was also reported in type II disease [79]. Large deletions [80 – 83] or insertions (exon duplication) affecting the helical region have also been associated with lethal disease [84 – 86]. Unlike the null allele, copolymerization of the mutated pro- chain with normal chains occurs, and the structurally abnormal heterotrimeric product is secreted into the extracellular matrix. Helical deletions in type II disease include deletion of COLIA1 exons 22 – 24 [80,81] and two examples of deletions of nine base pairs each involving the C-terminal portion (gly868-ala-pro and gly974-ala-pro) of the helical segment [82,83]. Mutations involving the COLIA2 gene also include gly to arg and asp, large deletions (exons 34 – 40) [87], and a splicing mutation that deleted exon 33 [88].

279

3. TYPE III OSTEOGENESIS IMPERFECTA (SEVERE, PROGRESSIVE OI) Infants born with severe OI present a characteristic appearance due to deformities of the limbs and the presence of multiple fractures. Birth weight and length are initially within normal range, although retarded growth appears within the first year of life. Fractures at birth involve the cranium, ribs, clavicles, and long bones. The cranium, although normal in circumference appears relatively large. With growth, an occipital overhang or “helmutshadel” deformity of the calvarium may develop. Molding of the cranium alters facial proportions so that a “sunset” appearance to the eyes may occur. Type III malocclusion is associated with mild prognathism. Sclerae are usually blue at birth but this decreases with age so that white sclerae are more typical of adults with type III OI. Moderate thoracic deformity with a pectus carinatum may be present but rib fractures are uncommon. Scoliosis may be mild initially but, with growth, approaches moderate to severe proportions (Fig. 4). Multiple vertebral fractures may be present at an early age. Vertebral fractures contribute to the progression of scoliosis (Fig. 5). The limbs are deformed by the pull of muscles and ligaments on the undermineralized bone. Individuals with type III disease have a characteristic high-pitched voice. There is a profound failure of somatic growth, many patients reaching only 3 – 4 feet in height. Deformities of the upper and lower extremities are present from birth and aggravated following recurrent fractures (Fig. 6). Because of skeletal deformities and severe osteopenia with a propensity to fracture, type III individuals tend to be wheelchair bound. Complications in the adult include the ever-present risk of traumatic fracture, a syndrome of chronic headaches (occipital cough headache) [56] related to basilar invagination, hearing loss, and progressive pulmonary insufficiency. A critical problem may occur if the medullary respiratory center is compromised by basilar invagination. The radiologic appearance may be of either the “broad bone” or “narrow bone” type, both representing a severe defect in skeletal modeling (Fig. 7). The epiphyses are poorly defined in these children, perhaps accounting for the limited skeletal growth. The epiphyses may contain irregular areas of poorly mineralized whorls of connective tissue. Kyphoscoliosis develops during childhood and may progress with increasing age, leading to restrictive pulmonary disease in the adult. Osteoporotic vertebral fractures increase the tendency toward spinal curvature. Following puberty, as is common with other OI types, the incidence of fractures declines markedly. a. Mutations in Severe Nonlethal Type III OI As in types I and II disease, a variety of mutations affecting various domains in the type I procollagen genes have been reported. These range from an N-terminal glycine 154 – arginine

280

FIGURE 4

Type III OI. This 25-year-old woman has been wheelchair bound due to thin, dysplastic lower extremities (see Fig. 7). Her current problems are related to: (i) a slowly progressive increase in scoliosis and (ii) basilar invagination as a consequence of relative softening of the base of the skull. Future complications may include restrictive pulmonary disease as well as neurologic symptoms secondary to medullary compression.

single-base substitution [46] to glycine, cystine mutations in the central helical region (gly cys 610, 526, and 415) [89,90]. Serine mutations were reported to pro-1(I) positions 1009 [72], 844, [91], 460, and 415 [77,78,92]. Wallis reported an exon 30 skip in type III OI [93]. An identical mutation involving has been reported in 2 members of unrelated families [94]. 4. TYPE IV OSTEOGENESIS IMPERFECTA This group, the least common of the OI phenotypes, is recognized as being clinically heterogeneous. It was initially categorized as having blue sclerae at a young age that faded to a white hue in adulthood [6]. However, individuals with this phenotype may retain blue sclerae as adults. This phenotype is inherited as an autosomal dominant trait. However, both the mild and severe extremes of this phenotype may be confused with type I or III OI. Clinically, these individuals have short stature and a tendency to cranial molding, and dentinogenesis imperfecta

JAY R. SHAPIRO

FIGURE 5 Type III OI. Severe osteoporosis and scoliosis. Note the thin markedly demineralized ribs. may affect approximately 25% of cases. Molding of the calvarium persists into adulthood. Basilar impression is reported to occur in 71% of type IVB OI patients [56]. In type IV, both vertebral and appendicular bone are more osteoporotic and dysplastic (Fig. 8). Scoliosis may be prominent. Pelvic deformity is common in these individuals. Joint laxity may disrupt the architecture of the ankle joint with a tendency to inversion, and dislocation of the knees may occur. Growth in height is intermediate between those with types I and III OI. In type IV OI there is more extensive skeletal deformity than in type I disease and the osteopenia of underlying bone is more severe. As a consequence, many individuals rely on either a cane or crutches for ambulation. a. Mutations in Type IV Disease Type IV is the least common of the OI phenotypes and for that reason, relatively few mutations have been reported. Both pro-1(I) and pro-2(I) chains have been affected. In general, these mutations occur near the central region of the helical chain consistent with a gradient effect on phenotype. The

281

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

FIGURE 7 Type III OI. The lower extremity demonstrates a narrow bone appearance due to faulty modeling. The epiphyseal zone is dysplastic: no growth plate is seen. The epiphysis contains whorls of partially calcified connective tissue termed “popcorn” calcifications.

FIGURE 6

Type III OI. Upper extremity deformity and nonunion of a

G. Inheritance Patterns and Osteogenesis Imperfecta

fracture.

initial report of a mutation in type IV OI involved that pro-2(I) chain, subsequently defined as an exon 12 skip secondary to a GT substitution affecting the consensus donor splice site [37]. Intron mutations leading to exon skip in type IV disease have been reported to involve pro1(I) exon 8 and pro-2(I) exons 12 and 21 [95,5]. The exon 21 skip was in a boy with short stature, osteoporosis, and dentinogenesis but no fractures [96]. A pro-1(I) gly352 – ser mutation affecting the helical region was also reported [92]. Interestingly, both pro-2(I) gly646 – cys and gly661 – ser mutations have been reported in type IV disease [50,54].

There are no good data relating the incidence of sporadic vs familial occurrence of OI. Estimates for the prevalence of sporadic disease vary from 19 to 25% [13,14]. As indicated below, the occurrence of mosaicism has made this question even more uncertain. For families in which dominant inheritance is based on a structural mutation (types I and IV and most cases of type III OI) the risk in successive pregnancies is 50%. For many years, approximately 25% of cases were thought to be recessively inherited. Recent data indicate that only 5 – 8% of OI cases can be attributed to recessive inheritance, i.e., the subsequent risk to parents of a child with sporadic lethal disease [85,97]. However, in the case of gonadal or somatic mosaicism, the involved parents appearing normal, the risk on successive pregnancies could be 50%. Recessive

282

JAY R. SHAPIRO

FIGURE 8 Type IV OI. This phenotype is of moderate severity. Note the extensive osteoporosis and deformity involving the pelvis and femurs. Cystic changes are present in the right femur. Not fully visible is the scoliosis that is common in type IV disease. inheritance plays a role in types II and III disease. Along with the recognition that most cases are due to structural mutations affecting the type I collagen genes came the realization that the majority of OI cases of each phenotype involve autosomal dominant transmission.

H. Prenatal Diagnosis of OI Prenatal diagnosis of OI is usually made by radiographic or ultrasonographic imaging. Current practice relies on detailed ultrasonography during the second trimester, a procedure mainly of value in type II, type III, or type IV disease; milder disease with less deformity may be missed on ultrasound. There are several reports of type III or II disease being diagnosed prior to the 20th week of gestation, even as early as the 12th week of gestation in a woman who previously had an affected child and whose husband was known to be mosaic for a COLIA2 gene mutation [99,100]. Diagnostic features include enlargement of the cranium, reduced echogenicity (low bone density), and deformities or shortening of the extremities as a consequence of intrauterine fractures. At-

testing to the variability in appearance, bowing with or without shortening, has been observed in the late second or third trimester with grossly normal mineralization [101]. Anencephaly detected by transvaginal sonography has been reported in a fetus with OI [102]. Chorionic villus biopsy with analysis of DNA for a type I collagen mutation is an invasive method that has successfully confirmed or eliminated the diagnosis of OI in several reported cases [103]. This is applicable during the 10th – 12th week of gestation in situations where a mutation has been previously detected in a family member. Type II OI was diagnosed prenatally by analysis of DNA obtained from chorionic villus biopsy at 12 weeks in a woman who previously had an affected child and whose husband was known to be mosaic for a mutation in COLIA2 gene [99].

I. Somatic and Gonadal Mosaicism in Osteogenesis Imperfecta This mode of inheritance of OI was first recognized when a man fathered two children with lethal OI with two partners [104 – 106]. Somatic mosaicism has also

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

283

been considered a mechanism for explaining the occasional marked variability in phenotypic expression seen within certain kindreds. Both somatic and germ line mosaicism has been demonstrated using sperm, skin fibroblasts, and blood. Mosaicism has been demonstrated using DNA derived from dermal fibroblasts, blood lymphocytes, and sperm in varying percentages for each tissue. An interesting observation about the mosaic parents has been their normal phenotype: in one instance, the only expression of OI was short stature despite the finding that 100% of dermal fibroblasts and 40% of sperm contained the mutations [104]. The frequency of mosaicism becomes a significant issue when genetic counseling is required because we do not know how commonly this occurs in lethal disease with apparently normal parents or how commonly mosaicism underlies milder forms of the disorder.

III. HISTOLOGY OF BONE IN OSTEOGENESIS IMPERFECTA The abnormalities in bone histology and histomorphometry generally parallel the severity of the OI phenotype. However, interpretation of bone histology is subject to qualification depending on the site of the biopsy, including proximity to an area of recent trauma, patient age, and the influence of medication that may alter bone turnover. Thus, to a varying extent, it is characterized by decreased trabecular volume and diminished cortical width, reflecting deficient matrix formation. There is mild increase in unmineralized osteoid. A striking finding in OI is the presence of increased numbers of osteocytes embedded in trabecular bone (Fig. 9). This has been confirmed by direct counting of osteocytes in type I subjects (J. R. Shapiro, unpublished data). Increased numbers of osteocytes have also been reported in mild OI [9,107]. Although plump osteoblasts are readily identified along trabecular margins, there is no increase in osteoclastic bone resorption. Electron microscopy of osteoblasts in OI has demonstrated dense material in the Golgi, glycogen deposits, and decreased alkaline phosphatase in the cell membrane [108]. The impact of a type I collagen mutation on endochondral bone formation can be seen in the growth plate in severe cases of OI (type II, severe type III). Cartilage columns appear to develop normally up to the point that endochondral bone formation occurs. In type II OI there is a failure of normal lamellar formation. Rather than the bony trabeculae normally present at the time of birth, there are disorganized islands of cartilagenous core surrounded by islands of poorly mineralized woven bone. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) and electron diffraction stud-

FIGURE 9

Hyperosteocytosis in a trabecular bone biopsy from a 40year-old woman with type IV OI.

ies of OI bone have disclosed abnormalities in hydroxyapatite crystal size as well as in the morphology of type I collagen fibers and the organization of lamellar plates in bone. SEM of type I OI bone did not differ significantly from normal in ultrastructure [109]. In severe OI the lamellar structure of bone is disconnected and separated by open spaces in regions. Type II OI bone presented a spongy appearance. TEM of OI type III bone displayed a matrix of loose fibrous mineral that was undermineralized with abnormally oriented small crystals, poorly organized in relation to collagen fibrils. However, even in severe phenotypes, normally oriented lamellar bone structure may still be preserved. Electron diffraction analysis of mineral crystals and direct measurement of crystal length from OI bone show that these are small in size and may be smaller in more severe phenotypes [102]. The orientation of crystals in collagen fibers may also be abnormal. X-ray scattering studies of crystal orientation in bones from animals with the OIM (osteogenesis imperfecta murine) mutation have also re-

284

JAY R. SHAPIRO

vealed faulty orientation of crystals in collagen fibrils [110]. Immunohistochemical studies of bone from type II OI subjects have demonstrated nests of cartilage with type II collagen and increased amounts of type III collagen in bone matrix. The persistence of type III collagen mimicking a fetal pattern was shown in both type III and type IV OI bone.

A. Bone Mineral Density in OI Normally, bone mass is determined by genetic, hormonal, and lifestyle factors. Most OI subjects have diminished bone density (osteoporosis) by X-ray, although, both in children and in adults, bone mass may occasionally appear radiologically normal [58]. Dual-energy X-ray absorptiometry measurements in OI have demonstrated both decreased bone mass and the ability of mineral mass to increase to a limited extent with age in some subjects. Both trabecular bone density (vertebral, wrist, ribs) and cortical bone density (humerus and femur) are decreased. In OI bone mass will decrease during pregnancy [53], after menopause, and after age 50 in men. The fact that fracture rates increase after the menopause and with increasing age in men indicates the need for aggressive treatment to limit the rate of bone loss [57].

B. Bone Histomorphometry in OI The issue of bone turnover in OI is important as it relates to selection of treatment. For example, calcitonin might be recommended where bone turnover is high and a bisphosphonate avoided in the presence of very low bone turnover. Ramser and Frost examined bone turnover in the rib of a woman with type I OI and determined that cortical bone turnover was increased threefold, while that in the periosteal layer was diminished [111]. This discrepancy between cortical and periosteal bone formation was unresolved but was proposed to contribute to diminished width of the ribs. Albright observed that the surface involved in new bone formation was increased as was resorption and the size of osteocytic lacunae [112]. The presence of osteocytic resorption in OI bone has not been confirmed. Bone turnover has been studied using double-tetracycline labeling prior to biopsy. St. Marie analyzed iliac crest bone biopsies following tetracycline labeling and observed that in types I and IV OI, trabecular bone volume was decreased, calcification rate was reduced, and apparent bone formation rate at the cellular level was decreased. Studies in children ages 6 – 15 years with mild OI demonstrated increased turnover rate with decreased osteoblastic activity. Glorieux et al. obtained iliac crest bone

biopsies from 44 OI children ages 2 – 14 years with nonlethal disease [113]. Common to all OI types were decreases in cancellous bone volume, cortical width, and trabecular thickness. In an expanded series of 70 children, ages 1.5 to 13.5 years, Rauch et al. found that decreased cancellous bone volume was due to a 41 to 57% reduction in trabecular number and a 15 to 27% decrease in trabecular thickness [114]. Surface-based parameters of bone remodeling were increased in all OI types. However, no defect in matrix mineralization was found. It is this increase in bone remodeling in OI children that may explain the effectiveness of antiresorptive therapy with bisphosphonates in children. Histomorphometric evidence of low bone turnover has also been observed in a cohort of type I OI adults [60]. The differences in bone turnover between these recent results and earlier studies noted above may be due to improved technology but more likely results from more consistent clinical classification.

C. Biochemical Markers of Bone Turnover in OI Biomarkers of bone formation, serum osteocalcin, serum procollagen type I C-terminal and N-terminal propeptides (PICP and PINP), and bone-specific alkaline phosphatase have been measured in OI as have markers of bone resorption, urinary excretion of deoxy-pyridinoline cross-links, and the collagen N-telopeptide cross-link [27]. These are potentially important because of histologic data suggesting that bone turnover was generally increased in OI [61]. Brenner et al. found PICP concentrations decreased in various forms of OI and more so in type I subjects [115]. Osteocalcin concentrations were increased in patients during the first decade, but in only 1 of 18 older patients. In a subsequent report, elevation in D-pyridinoline was reported, suggesting that increased resorption was a factor in the osteopenia of OI [116]. However, increased resorption as a contributing factor to osteopenia in OI is not supported by histomorphometric analysis of bone in children or adults and has been disputed by Minisola et al. [117] and Shapiro, who find markers of both formation and resorption commonly (but not uniformly) decreased in OI [118]. In a study of bone biomarkers in 78 OI patients. PICP and PINP were generally low and lower in patients with type I disease and a quantitative defect (diminished bone formation) than in more severely affected OI patients with a qualitative collagen I defect [119]. Biomarker results suggest that bone turnover is reduced in OI children and mildly affected adults, and bone resorption is increased in severely affected adults. This is an important consideration as it may influence the choice of therapy, e.g., the use of antiresorptive agents vs agents to increase bone formation of osteoblastic bone formation in subjects with varying rates of bone remodeling.

285

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

D. Ocular Features of OI Scleral color may be a distinguishing feature of certain OI types; however, it is important to note that scleral hue may vary during a subject’s lifetime and that among connective tissue disorders, blue sclerae are not unique to OI [120]. Blue sclerae have been described in subjects with the Ehlers – Danlos syndromes, including unusual cases of human dermatosparaxis [121,122], Marfan syndrome with contractures [123], and lethal hypophosphatasia [124]. Blue sclerae occur uniformly in subjects with type I disease. Sclerae may be blue or white in individuals with type III disease and are frequently blue at a young age, fading to white in those with type IV disease. Although the cause of blue scleral color remains unknown, it may be related to a matrix abnormality of the scleral coat since a positive relationship between the intensity of color and deformability of the globe has been demonstrated as well as decreased corneoscleral rigidity [125,126]. In terms of physical characteristics, the blue color is a product of reflectance rather than absorbance, again suggesting that it depends on altered matrix composition. Thin scleral coat as a cause of blue sclerae has not been a consistent finding, although Chan et al. reported that in lethal OI, both corneal fiber diameter and the diameter of scleral collagen fibers were reduced by 25 and 50%, respectively [127]. Arcus senilis (embryotoxon) is the second most frequent abnormality in OI following blue sclerae, being observed in 28% of affected individuals [4,8,9]. It may occur as an opacity or arcus at the periphery of the cornea at a young age and appear as annulus senilis in an older individual. The cause of the lesion is unknown: it is not related to abnormal lipid metabolism. Isolated instances of several other occular abnormalities have been reported in OI [9]. These include keratocornus, thin cornea, corneal rupture, and rare instances of subluxed lenses [128,129].

E. Dentinogenesis Imperfecta Two dental lesions have been recognized in OI, dentinogenesis imperfecta (DI) and multiple radiolucent bone cysts, which is a rare occurrence. Associated defects in the maxillofacial bone include condylar deformities with dislocation of the mandibular condyle, prognathic mandible (type III malocclusion), hypoplastic hemimandible, and depressed zygoma [130]. The most common oral manifestation of OI is dentinogenesis imperfecta (DI). There are two types of DI. That common to OI is dentinogenesis imperfecta type II [131]. Although dentinogenesis imperfecta occurs in approximately 5 – 15% of each OI type, it is more frequent in

type III OI and uncommon in type I OI. As a rule, DI tracks with bone disease so that individuals with DI in a family should be evaluated for bone disease. Affected teeth demonstrate a bulbous crown and increased coronal angle and may lack pulp space. Permanent teeth are less severely affected than deciduous teeth. Electron microscopy of dentin shows disorganization of dentinal tubules [132]. It is this defect in dentin that interferes with the adherence of enamel to dentin and leads to chipping and erosion of the tooth. Dentinogenesis imperfecta type II is an autosomal-dominant disorder of dentin formation which has been mapped to the 6.6 D4S2691 – D4S2692 interval at human chromosome 4q21. In the current investigation, the use of four short tandem repeat polymorphisms has allowed the critical region to be refined to an interval of less than 2 defined by recombination events in unrelated, affected individuals from two families, both of which show independent evidence for linkage to chromosome 4q21 [133]. O’Connell evaluated 40 children (age range 1 – 17.5 years) with types III and IV osteogenesis imperfecta. The incidence of dentinogenesis imperfecta was greater than 80% in the primary dentition. Class III dental malocclusion occurred in 70 to 80% of this osteogenesis imperfecta population. A delay in dental development was observed in 21% of patients type III osteogenesis imperfecta, whereas accelerated development was noted in 23% of the patients with type IV. In addition, ectopic eruption occurred [134]. Bone cysts of the jaw occur infrequently. A recent report describes a 23-year-old woman with OI and DI who developed multiple unilocular bilateral radiolucent cysts of the mandible 5 years after a condylar fracture.

F. Hearing Loss in OI Diminished audioacuity is a frequent manifestation of OI having been recognized by testing in approximately 30% of all OI groups. Socially impaired hearing function occurs in approximately 10% of patients. Hearing loss has been detected in children, although it most commonly occurs by young adulthood. In a series of type I patients, hearing loss greater than 30 dB was observed for 51% of subjects between the ages of 20 and 60 years [135]. Hearing loss threshold increased between the ages of 10 and 45 years at a rate of 1 to 1.7 dB/year. Multiple functional lesions have been described. These include conductive defects and mixed or sensorineural lesions. Conductive loss is due to traumatic defects in the stapes crura or to fibrosis at the stapes footplate. The high incidence of mixed defects and sensorineural loss implies involvement of the cochlea or cochlear nerve [136,137]. High-resolution CT and scintigraphy of the labyrinthine capsule has been performed in nine subjects with OI [138]. A severe decrease in bone density in the pericochlear

286

JAY R. SHAPIRO

region was found in subjects with mixed hearing loss although normal density was found in other affected individuals. Tympano-cochlear scintigraphy suggested increased bone metabolism in this region. A major question is whether stapedectomy is of longrange value in treating hearing loss in OI. Functional improvement occurs following stapedectomy [139]. Garretsen et al. found a gain in hearing in 85% of 58 ears after 3 months and in 68% of 40 ears followed for 9.6 years and no loss of their postoperative gain in hearing [140]. Where loss of hearing occurred, it was due to progression of the sensorineural defect either immediately after surgery or 1 year later.

G. Cardiac Lesions in OI Clinically significant cardiac lesions are infrequent in osteogenesis imperfecta, in contrast to other of the heritable disorders of connective tissue such as Marfan and Ehlers – Danlos syndromes. Right-sided heart failure may complicate progressive pulmonary insufficiency in type II OI, in type III disease in infants, and also in type III disease in adults. Mitral valve prolapse has been found on auscultation in 1 – 2% of OI subjects, and by ECHO cardiography in 10% as contrasted with a 6% incidence in the general population [141,142]. Approximatly 10% of OI subjects have dilatation of the aortic root but this is usually not associated with aortic regurgitation that has been infrequently observed [143]. Effective surgical correction of aortic valvular defects has been reported in OI subjects with severe disease [144]. The presence of a connective tissue dysplasia may complicate these procedures or healing postoperatively.

H. Neurologic Disorders in OI A variety of neurologic lesions have been reported in OI patients; the majority involve individuals with moderate to severe skeletal deformity. In this category are hydrocephalus, basilar invagination with brain-stem compression, and cortical atrophy [145]. Basilar invagination allows upward displacement of the upper cervical spine and clivus into the foramen magnum. Brain-stem compression may result in compression of the upper cord, with sensory neuropathy as well as impaired respiratory function. Syringomyelia has been reported in the presence of basilar impression [146,147]. A chronic headache syndrome, perhaps due to increased intracranial pressure occurs in individuals with type III OI where the base of the calvarium is deformed. Trigeminal neuralgia may accompany this syndrome [145]. A second category of neurologic complications occurs in individuals

with mild to moderate disease which are secondary to skeletal deformities. In this category are nerve entrapment syndromes following fracture healing and nerve root lesions secondary to scoliosis or vertebral collapse. Charnas et al. have reported on neurologic disease in 76 OI children, mean age 8 years, the majority with types III and IV disease [148]. Ten patients had macrocephaly, although head circumference was generally normal. Cerebral atrophy was identified in 17 individuals ranging from 7 to 17 years. Eight subjects, the majority with type III disease, had basilar invagination. Seizures occurred in five patients. Ten subjects had suffered skull fractures. Surgical decompression of the spinal cord may be required in severely affected subjects [149]. In the past, a distinction between OI and the osteoporosis – pseudoglioma syndrome has not been well defined. This syndrome includes serious visual impairment, dwarfing or short stature, and skeletal deformities. However, fibroblast collagen metabolism was normal in two patients, thus differentiating this syndrome from severe OI, which it may superficially resemble [150].

I. Pulmonary Disease in OI Pulmonary insufficiency is a major problem for two groups of OI subjects: (i) those neonates with lethal disease and (ii) adult type III patients with severe scoliosis. Adults with type IV disease have moderately severe scoliosis but are not at risk for pulmonary insufficiency. Restrictive thoracic disease may be complicated by the presence of pectus carinatum or pectus excavatum deformities. It is common for infants with lethal perinatal disease to succumb during the first few weeks of life from pulmonary insufficiency with superimposed pulmonary infection. In most cases, the reasons for this sequence of events is unknown. Neonatal pulmonary insufficiency may be secondary to the presence of intrauterine rib fractures (e.g., type IIA OI with beading of the ribs), giving rise to dyssynergic state of the thoracic musculature. Ventilatory support is required from shortly after birth. Pulmonary hypoplasia has been occasionally reported in postmortem examinations of infants with type II disease. A case of type II OI with pulmonary hypoplasia was studied to determine the potential effect of a mutation affecting the synthesis of pro-1(I) collagen on bronchoalveolar development. In this case it was determined that development of the bronchoalveolar tree had ceased at about the 10th week of gestation. The frequency of such lesions as a cause of lethal pulmonary disease is currently unknown [71]. Another cause of pulmonary insufficiency in severe type III OI is alveolar hypoventilation as a consequence of compression of the brain- stem secondary to basilar invagination [151].

287

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

The type III OI subject may have a restrictive pulmonary disorder secondary to severe scoliosis in association with decreased volume of the thoracic cage (Fig. 4). Deformity of the chest may increase as the scoliosis worsens. Since both tend to worsen with age, the patient, already wheelchair-bound, becomes dyspneic with little effort. It was initially observed that pulmonary insufficiency in severe OI correlated with the presence of kyphoscoliosis [152]. We have evaluated sleep patterns with reference to sleep apnea as well as pulmonary function in subjects with type III disease. The results indicate that despite the presence of type III malocclusion of the jaw and displacement of the tongue in certain patients, sleep apnea was found in only 2 of 6 type III subjects. Mild sleep apnea was noted in one type I OI subject age 70. The treatment of these individuals is complicated by their small thoracic volume. Tracheostomy may be required. Mechanical ventilatory support including the administration of oxygen at home is frequently required.

J. Scoliosis in OI The general incidence of spinal curvature approaches 70%. Contributing to the development of scoliosis are vertebral defects secondary to osteoporosis and laxity of the spinal ligaments. In general, the more severe types of OI are associated with greater deformity of the spine. A study of vertebral compression in 46 children with OI demonstrated that vertebral compression began shortly after birth in types III and IV OI and progressed to puberty. The pediatric OI spine was unusual due to the presence of posterior compression fractures, particularly at the L4 and L5 levels [153]. Additional deformities include platyspondly, thoracic kyphosis and increased lumbar lordosis. A survey of 102 patients revealed that curvature was mild (40°) in 50% moderate (60°) in 6%, and severe (80°) in 2% [154]. Deformities of the thoracic cage are almost always present in subjects whose scoliosis is greater than 40°. Thus, scoliosis tends to be mild in type I OI, more marked in type IV disease, but of major concern in type III disease since: (i) progression of the deformity may occur with age, leading to compromised pulmonary function, and (ii) surgical stabilization may be complicated by the osteoporotic quality of bone that limits effective placement of a prosthesis such as the Harrington rod. As in other forms of scoliosis, exaggeration of the deformity occurs after the age of 5 and may increase at the time of pubertal growth spurt. As a general rule, progression of scoliosis will occur in most patients. Bracing is of little value in this circumstance and it may further compromise pulmonary function.

K. Hyperplastic Callus Formation This is a very uncommon complication that appears during the healing phase of fractures, including areas that have previously fractured and healed without incident [155]. Any clinical type including type I OI may be affected. A painful tumor-like inflamed excessive deposition of callus forms from extracellular matrix. Callus tissues in OI have been studied without finding a significant deviation from the normal pattern of fracture healing aside from overmodification of types I and III collagen [156]. The course of hyperplastic callus is gradual resolution, usually over a period of several weeks to months. Recurrence has been observed in a few patients. Osteosarcoma is a consideration in the differential diagnosis of localized pain and swelling, but osteosarcoma is extremely rare in OI patients. Treatment should include administration of glucocorticoids to suppress matrix formation and the apparent inflammatory component of this reactive lesion. Low-dose radiation therapy has been reported to be successful in decreasing pain and swelling associated with hyperplastic callus formation [157].

L. Osteoblast Metabolism in OI The osteoblast developmental cascade proceeds from bone marrow osteoprogenitor cells to preosteoblasts and terminally differentiated osteoblasts that do not further divide. Osteoblasts give rise to bone lining cells and osteocytes. Cultured human osteoblastic cells may be grown as explants from minced trabecular bone and subjected to metabolic study for as long as 35 days in tissue culture. These cells do divide and are at the preosteoblastic stage [158]. An assessment of expressed osteoblastic markers by cultured cells from subjects with different OI types has revealed that the production of osteocalcin in response to 1,25(OH)2D3 stimulation was similar in OI types I, III, and IV vs controls [159]. However, both osteocalcin production and cAMP response to 1,25(OH)2D3 were decreased in cultured type II OI osteoblasts as contrasted with fetal controls. Several metabolic abnormalities related to extracellular matrix synthesis have been reported when cultured osteoblasts from OI subjects are compared to age-matched normals [160,161]. Although the growth curve of cultured human osteoblasts is slower than that of normal fibroblasts, OI osteoblasts (but not OI fibroblasts) have slower rates of proliferation when compared to age-matched normals [162,165,166]. The synthesis of type I collagen as measured by [3H]proline incorporation is decreased in cultured OI osteoblasts, as is the synthesis of the matrix associated proteins osteocalcin and osteonectin. The synthesis of the matrix proteoglycans decorin, biglycan, and the large chondroitin sulfate proteoglycan is also decreased [163]. However, the

288 synthesis of hyaluronan and bone sialoprotein by OI preosteoblasts appears to be increased, consistent with the increased amounts of bone sialoprotein and osteocalcin isolated from OI bone. Osteonectin content was also reduced in OI bone. It is assumed that these defects are secondary to the type I collagen mutation that in some manner alters the synthesis of other matrix proteins. However, the bone decorin levels were not altered in contrast to the observed decrease in decorin synthesis by OI osteoblasts in tissue culture. Collagen matrix deposition by cultured OI dermal fibroblasts from a spectrum of OI phenotypes was also decreased, indicating that diminished collagen secretion is not limited to subjects with the null allele phenotype [164]. Studies of collagen extracellular matrix deposition and turnover by fibroblasts from a case of lethal OI reveal the limitations on collagen secretion by these cells [165]. OI fibroblasts with a gly667 – arg mutation deposited onefourth the type I collagen that control fibroblasts did. However, the reduction in collagen matrix deposition was not due to less total collagen synthesis; rather, the incorporation of the mutant collagen into matrix was less efficient. Such mutant collagen as incorporated into matrix appeared more subject to proteolytic breakdown, suggesting faulty copolymerization of the mutant with the normal procollagen. Analysis of OI bone for collagen cross-link content has provided divergent results. Although decreased levels of lysine-based cross-links in skin have been reported [166], analysis of OI bone from subjects with types I, III, and IV OI for the mature collagen cross-links, hydroxylysylpyridinoline and lysylpyridinoline, disclosed similar concentrations in OI and normal subjects [167]. An analysis of compact bone from 30 patients ages 2 to 9 years with various OI phenotypes demonstrated low collagen and low total protein content per milligram of DNA or OI compared to that in age-matched controls [145]. This analysis also found increased content of type V but not type III collagen in OI compact bone. Type III collagen is not normally present in adult bone. However, increased types II and V collagen were found in collagen extracted from OI infants with lethal disease [168]. In contrast to type I subjects where overhydroxylation of collagen was not found, posttranslation overmodification was found in bone collagen from subjects with types II and III OI. Hydroxyapatitite crystal size in bone specimens from children and adolescents reveals reduced c-axis crystallinity of apatite in types III and IV OI specimens and reduced crystal size during childhood only in type I OI subjects [169,170]. It was postulated that reduced crystallinity was in some manner related to the defect in collagen synthesis because crystal size appeared reduced in more severely affected children. Reduced bone apatite

JAY R. SHAPIRO

crystal size had previously been demonstrated in two strains of cattle reported having a phenotype that included blue sclerae, marked joint laxity, and osteoporosis but in these animals type I collagen synthesis was normal. The mechanism of altered crystal size in OI thus remains uncertain.

M. Animal Models of Osteogenesis Imperfecta Several animals of different species have been described as having the osteogenesis imperfecta phenotype based on the presence of osteopenia with findings indicative of a connective tissue dysplasia. One example is the murine fragilitas ossium (fro/fro mouse) mutation developed as a result of treatment of male germ cells with tris(1-aziridinyl phosphine) sulfide [171]. Collagen metabolism was normal in this model. A second model, termed bovine osteogenesis imperfecta was reported in the offspring of unrelated Friesan cattle in Australia and Holstein cattle in Texas [172]. These animals had intrauterine or postnatal fractures, blue sclerae, and severe joint laxity. Although bone proteoglycan and sialoprotein content were depressed, type I collagen synthesis was normal [173]. Severe canine OI has been described in association with an 1(I) Gly208 –ACA mutation [173a]. The application of transgenic methodology has led to the development of several models of human OI based on mutations in the type I collagen molecule. While several of these are lethal in the homozygous state, the MOV 13 heterozygous mouse model has been suggested as a prototype for human osteoporosis. In this model, integration of a murine retrovirus in the first intron of COLIA1 blocks transcription simulating a null allele phenotype [174]. A transgenic model containing a 45-bp pro-1(I) deletion, first described in a lethal case of human OI, has proved viable [175]. This minigene was designed to generate shortened pro-1(I) chains to mimic the mutation, leading to intracellular “protein suicide.” In this model, 6% of the offspring were lethal, 33% had fractures, and 61% had no fractures at birth, suggesting mosaicism. This variable expression raises interesting questions about putative factors in transgenic models that may vary the expression of the mutant gene in different tissues [176]. A viable, naturally occurring murine model that duplicates the phenotype of moderately severe human OI was described by Chipman et al. [179]. This animal has a single base deletion in the C-terminal propeptide (exon 52) that prevents synthesis of pro-2,1(I) and leads to the accumulation of 1(I)3 homotrimer in all tissues. Interestingly, it is not lethal despite the location of the mutation. Osteoporosis, fractures, joint laxity, and diminished somatic growth are features common to human OI. The mutation, a single base deletion, mimics a 4-bp deletion in

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

the same exon (exon 52) in a child with type III OI [178,179].

N. Medical Treatment in OI The literature contains many references to the treatment of OI with a variety of hormonal and bone-active agents. Unfortunately, many of these involve single-patient reports, and most are uncontrolled. When reviewing treatment options for OI, it is important to recall that mutations in collagen synthesis underlie the osteoporosis in most cases; whether any treatment alters collagen production or stabilizes collagen in the extracellular matrix remains to be seen. Fluoride treatment of osteopenic syndromes has been under study for several years. Fluoride will increase bone mass by densitometric measurement and will increase bone formation as determined histologically. However, fluoride will lead to a mineralization lag and may induce synthesis of irregular bone matrix. At issue is the question of whether treatment with fluoride will lead to a decrease in fracture rate (see Chapters 74 and 75). Fluoride in a slowly absorbed preparation has recently been reported to decrease fracture rate in postmenopausal osteoporosis [180] but its use in severe OI has not proven effective [181]. The use of calcitonin in OI has provided variable results. Although no positive effects on the appearance of bone biopsy specimens has been reported [182] in type III OI, decreased fracture rates were reported in uncontrolled studies [183]. Recently, 10 patients treated with calcitonin by injection and nasal spray for periods of 28 to 76 months had apparent decreases in fracture rate [184]. As the authors indicate, spontaneous or age-related changes in fracture rate could not be discounted. Because of the frequent occurrence of short stature in OI children, and the positive effects of GH treatment to subjects with Turner’s syndrome, several laboratories [185,186] have initiated clinical trials of growth hormone therapy. In general, baseline provocative tests of GH and IGF-I have been normal in OI subjects although a blunted somatomedin generation test response to growth hormone was observed in 13 of 22 children [187]. In general, GH treatment of non-GH-deficient osteoporotic adults has given highly variable results, in part because the effects of GH on bone turnover are in the direction of an increase in both bone formation and bone resorption with a net increase in resorption. GH treatment of OI patients has been reported from several centers with variable results. GH was found to increase the rate of linear growth velocity, increase bone turnover and mineral content in lumbar trabecular bone in patients with collagen defects typical of mild/moderate OI [188]. In a large clinical

289

trial, a modest positive effect of GH in OI was seen only in type IV subjects (Marini) [189]. Bisphosphonates are widely used in the treatment of postmenopausal osteoporosis and have been effectively used in children with different skeletal disorders and in OI subjects [190]. Intravenous pamidronate (Novartis Pharmaceuticals), a “second generation” agent, has been administered to adult type I OI subjects [191]. In adults treated with pamidronate, 30 mg intravenously every 3 months, trabecular bone volume increased [192]. However, clearly positive effect in terms of bone mineral density and a decrease in fracture rate has been reported in children with type III and IV disease [193,194]. Other reports of bisphosphonates use involve single patients, again uncontrolled, in which beneficial effects on bone density were observed [195, 196,197]. Current investigational studies involve the administration of oral bisphosphonates, alendronate and residronate. Because of the reported increase in fracture rate with increasing age it is recommended that postmenopausal women be placed on conventional hormone replacement therapy with estrogen and progesterone. However, there are no data, either histology or bone density measurements to support a beneficial effect of hormone replacement in the OI subject.

O. Rehabilitation and Physical Therapy in OI Consistent and coordinated rehabilitative and physical therapy are critical to the successful development of children with OI and to the maintenance of effective daily activities for adults with this disorder. Functional independence is the ultimate goal. Binder et al. have stated the two central tenets in approaching rehabilitative care: (i) the variability associated with the OI phenotypes makes it difficult to predict which children are at risk for significant disability and (ii) one must fully evaluate a child’s functional abilities and potential for rehabilitation [198]. Children and adults with severe disease may be unable to sit independently, and many adults with type III disease are wheelchair bound, or as with type IV subjects, dependent on lower extremity braces, canes, or crutches for ambulation. Important issues involve the development and maintenance of muscle strength in the upper and lower extremities, the prevention of joint contractures at several joints including the shoulders, hands, hips, and feet, poor joint alignment, and disturbances of gait and low endurance in those who ambulate. To approach optimal rehabilitation and maintenance of function, children were classified as to the severity of disease and potential for rehabilitation and a specific program was outlined for each group [199]. Modalities included custom-molded seats to align hips, knees, and ankles to prevent stress in the spine, therapeutic water activities,

290

JAY R. SHAPIRO

soft tissue mobilization techniques to control joint contractures, and specific exercises to strengthen muscle groups, improve posture and balance, and increase endurance. Specific bracing techniques involving both joint support and lightweight lower extremity braces were employed to limit joint instability and permit early ambulation. As a result, improvements were found in head and trunk control, joint alignment, strength in the extremities, and the ability to ambulate.

P. Orthopedic Treatment in OI Orthopedic management of OI starts literally at the moment that the diagnosis is made. In the infant this may involve aligning fractures to minimize deformities and rodding bones to decrease the occurence of fractures and improve function [200]. Intramedullary rodding is, in the author’s opinion, indicated as early as required to decrease

the risk of fracture. Both nonexpanding and expanding rods are used, depending on the age of the child, the size of the bone, and the severity of the osteopenia. Expanding (Bailey – Dubow) rods can elongate with growth of the bone. However, these rods may be associated with bending, unscrewing of the T-shaped end from the sleeve of the rod, and breakage [201]. These rods may fail to elongate and the T-end of the rod may migrate within the bone. Protrusion through the bone may occur (Fig. 10). Overall, complications occur in about one-third of cases and these require reoperation in approximately 20% of cases after 5 years [202]. In the teenager or adult, orthopedic care is required to assess and surgically correct scoliosis, stablize lax joints with equalization of the length of extremities, and correct deformity by osteotomy. Protrusio acetabuli, a source of chronic pain and limited mobility has been reported in 29% of subjects with type III OI [204]. Joint replacement of the hip and knee has been performed with satisfactory results for five of six patients having the following complications of OI: osteoarthrosis of the hip and knee, severe deformity of the hip associated with posttraumatic arthritis, acetabular fracture, and nonunion of a subtrochaneric fracture [205]. Placement of Harrington rods and fusion for progressive scoliosis is a major orthopedic concern. The therapeutic result may be complicated by the quality of the bone into which a prosthesis is seated, the degree of scoliosis, and the age of the patient relative to growth rate.

IV. OSTEOPOROSIS IN THE HERITABLE DISORDERS OF CONNECTIVE TISSUE A. Homocystinuria as a Cause of Adult Osteoporosis

FIGURE 10

Expanding rods in femur and tibia in OI. Note that the distal end of the femur rod is protruding into the joint. Epiphyseal architecture remains intact in the presence of these rods.

Homocystinuria, one of the heritable disorders of connective tissue when fully expressed, is associated with mental retardation, ectopia lentis, marfanoid habitus, and thrombotic vascular disease that occurs at an early age. Premature osteoporosis in association with other skeletal alterations occurs in teenagers and may be associated with vertebral and appendicular fractures [4]. Additional skeletal findings include the development of scoliosis, increased length of long bones and growth arrest lines, bowing and fracture of long bones, arachnodactyly, enlarged carpal bones and pectus excavatum, and carinatum deformities of the sternum. Homocysteine is an intermediate in the transsulfuration pathway which converts methionine to cysteine and ultimately to sulfate [177]. The occurrence of homocystine in urine may result from seven different genetic abnormalities, the most common of which is cystathionine b-synthetase deficiency [178]. Cystathionine b-synthetase defi

291

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

ciency is inherited as an autosomal recessive trait [208]. These subjects accumulate abnormal amounts of homocysteine and methionine in plasma. Other inborn errors similar to that in vitamin B12 (cobalamin) deficiency affect the remethylation pathway of homocysteine back to methionine [209]. Serum methionine concentrations are elevated in cystathione b-synthasedeficiency but are low or normal when 5-methyltetrahydrofolate-dependent homocysteine methylation is diminished. Although significant progress has been made in understanding the relationship between elevated plasma homocysteine and premature atherosclerosis in this disorder, little has occured toward understanding the development of osteopenia at an early age in this syndrome. Osteoporosis is one of the more common and consistent manifestations of elevated serum levels in homocystinuric subjects. In a series of 26 subjects, 25 were found to be osteoporotic [210]. Approximately 50% of affected individuals will be osteoporotic by the start of their third decade. Vertebral bodies appear osteoporotic, flatter than normal, and elongated in the anteroposterior axis with a posteriorly placed biconcave deformity similar to that seen in hemolytic disorders [211]. A quandary related to the frequent appearance and early onset of osteoporosis is the relationship of elevated plasma homocysteine in the vascular compartment to defective collagen synthesis and, ultimately, to bone deficits. This question remains unresolved in spite of early studies suggesting an effect of homocysteine, in vitro, on collagen cross-linking, as demonstrated by increased skin collagen solubility and the failure of collagen to form a stable gel after heating to 37°C and cooling [212]. Homocysteine at concentrations of 104 to 105 M, levels found in patient’s sera, will not inhibit lysyl oxidase activity but will prevent formation of insoluble fibrils and bifunctional cross-links [207,208]. Several reports have documented an association between mild homocystemia and premature vascular disease. These subjects are heterozygotes for cystathione b-synthase deficiency [213]. It is not known whether these individuals are at risk for osteoporosis. The treatment of patients with cystathionone b-synthase deficiency includes administration of vitamin B6, low methionine diets, and recently, betaine. This agent will lower plasma homocysteine by increasing the methylation of homocysteine to methionine [208].

B. Marfan Syndrome The Marfan syndrome is an autosomal dominant disorder characterized by the presence of skeletal abnormalities, cardiovascular lesions, and ocular defects [214]. Research into the connective tissue dysplasia of Marfan syndrome

has defined mutations involving one of the two genes (FBNI gene, chromosome 15) coding for the 350-kDa glycoprotein fibrillin that forms microfibrils associated with elastin and the extracellular matrix [29,215]. One unanswered question was whether the connective tissue lesion in Marfan syndrome extended to bone. Osteopenia has now been reported by several groups [216]. Kohlmeier et al. measured bone density in 17 women with Marfan syndrome. Significant deficits in bone density were observed for the whole body and proximal femur and for the femoral neck when corrected for bone size. No relationship existed between bone density and scoliosis and there was no occurrence of nontraumatic fracture. Similarly, no loss of radius bone mineral was observed in 14 Marfan subjects [217]. The status of bone density in males with Marfan syndrome remains uncertain at this time. However, since fibrillin mutations have not been reported in all Marfan families, the possibility remains that mutations in other extracellular proteins contribute to the development of this syndrome. Three examples have been reported: Pulkkinen et al. reported deficient expression of the gene coding for the matrix proteoglycan decorin in a lethal form of Marfan syndrome [218]. In support of this observation, fibroblasts from a patient with neonatal Marfan syndrome were found to deposit markedly less fibrillin in extracellular matrix than that normally seen. Fibrillin mRNA and synthesis appeared normal. In addition, decorin mRNA and biosynthesis were decreased, as was decorin deposition in extracellular matrix [219]. Another neonatal case with the Marfan phenotype was reported to have a deficiency of both laminin and fibronectin in skin [220]. The results suggested that a defect in fibrillin caused a secondary defect in decorin biosynthesis. Normally, decorin is integrated with the type I collagen molecule. In addition, a type I collagen 2(I) arginine 618 – glutamine mutation was reported in a 30-year-old severely affected individual (progressive lumbar scoliosis and aortic root disease but no lens dislocation) with Marfan syndrome [221]. This mutation was also found in the proband’s father, who suffered from kyphosis and chronic lung disease but had neither arachnodactyly nor joint hypermobility.

C. Ehlers–Danlos Syndrome and Adult Osteoporosis Ehlers – Danlos syndrome (EDS) constitutes a heterogeneous group of connective tissue disorders that have in common joint and skin laxity and excess bruisability [222]. The individual phenotypes vary in the extent of joint laxity and skin fragility and the expression of other characteristics that clinically identify each variant. Additional manifestations of EDS include recurrent joint dislocations as well

292

JAY R. SHAPIRO

TABLE 3 Type Classical type

Ehlers – Danlos Syndromes Former classification

Inheritance

Gravis, EDS type I

AD

Mitis EDS type II

AD

Hypermobile type

Hypermobile EDS type III

AD

Kyphoscoliosis type

Ocular scoliotic (EDS type VI)

AD

Arthrochalasia multiplex congenita

AD

AR Arthochalasia type

(EDS type VIIA and VIIB) Dermatosparaxis type Human dermatosparaxis (EDS type VIIC) Other forms

AR

X-linked EDS (EDS type X) Periodontitis type (EDS type VIII) Fibronectin-deficient type (EDS type X)

?

Progeroid EDS

?

Unspecified types EDS, Ehlers–Danlos syndromes; AD, autosomal dominant; AR, autosomal recessive.

as fragile skin with characteristic “cigarette paper” scars, mitral and tricuspid valve prolapse, kyphoscoliosis and ocular fragility (type VI), and fragile vascular and cavity lining tissues with arterial and gastrointestinal rupture (type IV). Types I, II, and III are dominantly inherited and vary in terms of joint laxity and involvement of the skin. Several syndromes, type V (lysyl oxidase deficient), type VII (short stature, multiple joint dislocations, round facies), type VIII (peridontitis), and type IX (bladder diverticulae, occipital horn syndrome), have unique facies and body habitus that aid in diagnosis. The EDS types listed above had previously been defined based solely on clinical signs. However, recent advances in discovery of molecular lesions have raised uncertainty about the inclusion of certain phenotypes in the EDS syndrome: this will undoubtedly be resolved as studies progress. In 1997, a simplified classification was proposed dividing EDS into six major clinical types, including genetic defects where previously known [223] (see Table 3). This classification grouped EDS types I and II into the classical type, former type III EDS into the hypermobility type, EDS type IV into a vascular type, a kyphoscoliosis type includes EDS VI, and former types VIIA and VIIB are now grouped into an arthrochalasia type. Several poorly differentiated EDS types are grouped into other forms pending biochemical confirmation of their identity. These include human dermatosparaxis (EDS VIIC), X-linked EDS type V, EDS VIII associated with periodontitis, and EDS type X or fibronectin-deficient EDS associated with prominent bruising. Type IX EDS (occipital horn syndrome) and X-linked recessive disorder shares biochemical features with Menkes

disease and has been categorized as a disorder of copper metabolism [224]. Type V collagen, a member of the group of fibrillar collagens, is composed of three alpha polypeptide chains 1(V), 2(V) and, 3(V), which are products of the COL5A1, COL5A2, and COL5A3 genes located respectively on chromosomes 9q, 2q, and 19p. Recent evidence indicates that mutations involving the type V collagen is causally linked to EDS I and II [225]. Subsequent studies have shown both clinical and biochemical heterogeneity. Type IV EDS is the result of different mutations involving type III collagen [226]. The majority of mutations involve point substitutions of arginine, serine, valine, aspartic acid, or glutamic acid for glycine in the triple-helical domain. Small genomic deletions, multiple exon deletions, and exon skipping mutations have also been reported in EDS IV. In the original description of EDS VI, it was found that the hydroxylysine content of collagen in skin and tendons was less than 1 residue per 1000 total amino acids in contrast to 4 residues per 1000 in control tissue. The hydroxylysine content was somewhat less than normal in bone and was normal in cartilage. Subsequently, both low lysyl hydroxylase and normal enzyme activity has been reported in fibroblasts obtained from individuals with this form of EDS [229]. The group of disorders classified as EDS type VIIA and EDS VIIB are the result of mutations involving the Nterminal alpha chain pro-peptide cleavage site [227]. EDS VIIC, the homolog of dermatosparaxis in sheep and cattle, is the result of defects in the converting enzyme procollagen N-peptidase [228]. The previously categorized EDS IX is a rare X-linked condition characterized by skeletal dysplasia, characteristic occipital “horns” that appear during adolesence, diarrhea due to increased bowel motility, and obstructive uropathy due to bladder diverticulae that appear during the first decade. This condition is allelic with Menkes disease and along with Wilson’s disease is one of the three hereditary disorders of copper metabolism. Cells from patients with these disorders have elevated levels of intracellular copper due to defective copper transport. A defective copper transporting ATPase gene demonstrated in Menkes syndrome may underlie the mechanism of this disease [230]. Patients with EDS may show vertebral abnormalities including wedged vertebrae and spondylolisthesis. Few data exist on bone mineral density in the various EDS phenotypes. However, Coelho et al. recently assessed bone mineral density in four patients, ages 16 – 25 years, with EDS I [231]. Bone density at the lumbar spine was persistently 1 SD below average for age and gender. However, this difference was not present for the femoral neck. In the context of adult osteoporosis, diagnostic uncertainty involves those younger subjects with idiopathic osteoporosis

293

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

in whom joint laxity is of moderate degree and the question of mild EDS arises. Solution of this question may require genetic analysis for detection of specific collagen mutations.

D. Idiopathic Juvenile Osteoporosis Idiopathic juvenile osteoporosis (IJO) is included in this section because this is considered along with mild OI in the differential diagnosis of idiopathic osteopenia in teenagers and young adults. Idiopathic juvenile osteoporosis is an uncommon self-limited disorder of children and teenagers characterized by potentially reversible osteoporosis that usually appears in the prepubertal years. Norman found approximately 60 cases of IJO reported in the literature between 1939 and 1991 [232]. There is no gender selection. Although differentiation from type I OI may be difficult, IJO is not familial and not associated with blue sclerae, dentinogenesis imperfecta, or short stature. Certain patients may have a pectus carinatum chest deformity. The disorder usually has its onset 2 – 3 years before puberty, although the age at onset may vary from 3 to 16 years [233]. It usually runs its course over 2 – 4 years. Children complain of the gradual onset of pain in the back, knees, and ankles. Wedge compression fractures of the spine and lower extremities may occur, causing kyphoscoliosis, as may fractures of the knees and ankles. Although bone loss eventually ceases and remineralization proceeds, mild cases are left with short stature and mild kyphosis. In severe cases, IJO may lead to marked deformities of the extremities and pulmonary insufficiency due to kyphoscoliosis and collapse of their rib cage [234]. Serum biochemistries are normal in idiopathic juvenile osteoporosis. Calcium balance may be negative during the period of rapid bone loss. Urine calcium excretion has been reported to be normal or increased and urinary hydroxyproline excretion may be increased [235]. The question of bone turnover in IJO remains unsettled. Bone biopsies have revealed evidence of either increased bone resorption [236] or normal resorption and decreased bone formation [237]. Using double tetracycline labeling. Evans has reported normal endosteal bone formation by histomorphometry in a 12-year-old boy with severe disease [238]. It is likely that measured rates of bone turnover may vary depending on the site of the biopsy, age, pubertal status, and the stage of evolution of the disorder. Radiologic examination reveals generalized osteopenia associated with a decrease in height of vertebral bodies due to wedge-shaped fracture or misshapen vertebral bodies due to collapse of the end plates [239]. Long bones are osteopenic and may show osteoporosis of bone newly formed during the pubertal growth spurt (neoosseous osteoporosis).

Linear metaphyseal rarefaction is a clue to this disorder. It results from impaction type fractures that occur at the growing ends of weight bearing bones [9]. These fractures are typically seen at the distal tibia, adjacent to ankle joint and adjacent to knee and hip joint. Pocock et al. have reported that dermal fibroblasts from two brothers with IJO secreted a reduced amount of type I collagen, a defect similar to that reported in type I OI [240]. However, other fibroblast cell lines from similar patients secreted from 57 to 155% of expected amounts of type I collagen. In addition an 2(I) chain mutation (gly436 – arg) has been reported in siblings with juvenile osteoporosis [241]. A case report indicates that treatment with bisphosphonates may be beneficial in this condition [242].

E. Idiopathic (Adult) Osteoporosis (see Chapter 45) As the availability of bone mineral density measurements has increased, accurate determinations of bone mass are available for patients in whom abnormal bone loss is suspected by routine radiographs or a fracture after minimal trauma. This has led to the recognition of a cohort of young or middle-aged adults, both women and men, who show significant bone loss. Because clinical and laboratory evaluation does not disclose the cause of the bone loss, mild osteogenesis imperfecta becomes a diagnostic possibility. Unlike OI, these individuals do not have blue sclerae, dentinogenesis imperfecta, or hearing loss. They do not have short stature. However, the phenotype of idiopathic osteoporosis includes mild joint laxity and mild scoliosis. A positive family history of osteoporosis is found in both OI and individuals with idiopathic osteoporosis. Spotila et al. [243] analyzed COLA1 and COLA2 in 26 patient with this phenotype and reported three subjects with mutations that altered an encoded amino acid in either the pro1 or pro-2 chain. However, the role of these mutations in the genesis of osteopenia was considered uncertain [243]. The author has observed idiopathic osteoporosis in the young adult has been also associated with COL1A1 Sp1 transcription site polymorphisms but this is also of uncertain signficance. Decreased serum IGF-I levels have been observed in many men under 50 years with idiopathic osteoporosis [244]. However, the significance of this is uncertain because the GH axis appears normal in these subjects and because serum IGF-I reflects alterations in multiple metabolic processes. Effective treatment for idiopathic osteoporosis remains uncertain. Adequate clacium intake is advisable. Calcitonin may be useful where bone turnover is elevated. Elevated bone turnover has been observed in a Norwegian group of patients with idiopathic osteoporosis [245]. However, this is not likely to be a uniform finding because of the

294

JAY R. SHAPIRO

heterogeneity inherent in this cohort. Although bisphosphonate therapy may also be helpful in retarding bone loss preliminary results indicate that certain patients may not respond to these agents. The role of treatment with supplemental vitamin D is untested in this group of patients.

V. OSTEOPOROSIS IN INHERITED HEMATOLOGIC DISORDERS Hematologic disorders may be associated with osteoporosis, and this is probably a consequence of a high rate of bone turnover. It is assumed that local production of bone resorbing cytokines (IL-1, IL-4, IL-6, tumor necrosis factors, prostaglandins) play a role in facilitating resorption. Although diagnostic confusion with idiopathic osteoporosis may not exist in the presence of homozygous disease as with severe sickle cell anemia or thalassemia major, uncertainty as to the cause of adult osteoporosis may exist in the presence of mild genetically related hematologic disease such as thalassemia minor or mild pernicious anemia.

A. Adult Osteoporosis Associated with Thalassemic Disorders Thalassemia major, a hereditary disorder due to the inability to synthesize the b-chain of adult hemoglobin, is associated with severe anemia and a variety of skeletal abnormalities [206,246,247]. There are widened medullary spaces in the tubular bones of the hands and feet, and the calvarium is thickened with widened diploic spaces. There is diffuse osteoporosis with widening of medullary spaces and loss of cortical bone. Osteoporosis tends to be more severe in males. Skeletal maturation is delayed. These changes are in response to a hyperactive erythroid marrow. Patients with thalassemia major may fail to undergo spontaneous puberty so that hypogonadism may contribute to low bone mass [248]. By contrast, subjects with b-thalassemia trait have a mild, asymptomatic anemia. These individuals may present with osteoporosis in the absence of other metabolic cause and generate concern for the presence of associated OI. Greep et al. have reported a 53-year-old postmenopausal woman with thalassemia minor who had osteoporosis of the spine with bone mineral density values unexpectedly low, more than 3 standard deviations below normal [249]. Bone histomorphometric studies demonstrated a high remodeling rate. Eleven other subjects with thalassemia minor had a mean Z score for lumbar spine density of 0.78 and hip density of 0.54. There was no correlation between diminished bone mineral density and

the severity of the hemolytic anemia or hemoglobin A2 levels. There was a greater deficiency in axial than appendicular bone mass perhaps related to increased turnover of trabecular bone in the vertebrae. By contrast is the report of Kalef-Ezra et al. in which bone mineral density was assessed in 22 premenopausal women and 21 men with thalassemia minor. Both serum and urine markers of bone turnover, as well as bone mineral density measurements, were the same as for matched controls [250]. These divergent results may depend on the particular populations studied and their age and menopausal status. Pending additional studies, it is possible that thalassemia minor, as a relatively common genetic trait, may present as osteoporosis in a young adult.

B. Osteoporosis Associated with Pernicious Anemia As discussed above (homocystinuria and osteoporosis), vitamin B12 deficiency is associated with homocystinemia and osteoporosis. Congenital pernicious anemia, the result of a failure to secrete intrinsic factor, usually is recognized early in life but some patients are diagnosed in adolescence or adulthood. Pernicious anemia has recently been recognized as a cause of osteoporosis and fractures. In a series of postmenopausal women studied at the Mayo Clinic, pernicious anemia was found associated with reduced bone mineral density of the spine and with vertebral fractures [251]. A population-based study was conducted of all Rochester, Minnesota, residents diagnosed with pernicious anemia from 1950 to 1979. Subjects with pernicious anemia had a 1.9-fold increase in proximal femur fractures, a 1.8-fold increase in vertebral fractures, and 2.6-fold increase in distal forearm fractures. This increased fracture risk did not appear to be related to neurologic complications of the anemia. Melton et al. have reported 2-year follow-up data on a 68-year-old man with pernicious anemia and multiple osteoporotic compression fractures. This patient exhibited a “dramatic” increase in bone mineral density at the spine and hip (79% increase in bone density in the femoral neck region) to treatment with vitamin B12 and cyclic etidronate therapy [252]. This response to bisphosphonate treatment, a suppressor of osteoclastic resorption, points to increased bone turnover as a probable factor in this hematologic syndrome. A direct relationship between vitamin B12 and osteoblast function was proposed by Carmel et al., who reported decreased serum levels of osteocalcin and bone-specific alkaline phosphatase in vitamin B12-deficient subjects [253]. Vitamin B12 replacement restored the serum levels of these osteoblastic marker proteins to normal. Vitamin B12 has been shown to augment bone marrow osteoprogenitor cell production of alkaline phosphatase and to increase the proliferation of these cells

295

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

[254]. There does not appear to be a relationship between achlorhydria and calcium absorption that would adversely affect bone mass [255].

VI. OSTEOPOROSIS IN INHERITED HEPATIC DISEASE A. Hemochromatosis Hemochromatosis is characterized by the presence of iron overload in multiple tissues including the liver, pancreas, parathyroid glands, and testes. Osteoporosis is a well-recognized complication of hemochromatosis, having been identified in 15 – 66% of patients. Although it is frequently recognized late in the course of the disease, hemochromatosis may be diagnosed only following the occurrence of an osteoporotic fracture [256]. An unresolved question is whether the primary cause of the bone loss is due to the toxic effects of iron overload, hypogonadism, liver disease, or other factors. Twenty-two men with idiopathic hemochromatosis were evaluated by serum biochemistries, spine radiography, bone mineral densitometry, and bone histomorphometry [257]. Ten of 22 patients were osteoporotic with decreased trabecular bone volume: no patient had osteomalacia. Eugonadal subjects treated with venesection (phlebotomy) had higher osteoid and osteoblastic surfaces than did nonvenesected eugonadal men. Hypogonadal men with low serum free testosterone levels had the lowest bone mass but these individuals tended to have more severe liver dysfunction due to iron overload. However, the severity of hepatic iron deposition did not correlate with the degree of osteoporosis. Thus, hepatic disease, iron excess, and gonadal deficiency each impact bone mass in idiopathic hemochromatosis.

B. Hepatolenticular Degeneration (Wilson’s disease) Copper is an essential cofactor for lysyl oxidase which acts to form stable collagen cross-links. Skeletal and connective tissue disease are characteristic of both Menkes disease and the occipital horn syndrome which result from defective intestinal copper transport secondary to mutations involving P-type ATPase, leading to excessive intestinal cellular storage and low hepatic copper concentrations [258]. Skeletal osteoporosis occurs. The role of copper deficiency in the genesis of osteoporosis has been the subject of several studies [259]. Hepatolenticular degeneration is characterized by the presence of hypoceruloplasminemia, net hypocupremia, increased nonceruloplasmin serum copper, and increased urine copper excretion. Osteoporosis was recorded in 79%

of affected individuals [260]. Osteomalacia has also been reported in certain patients. Although reduced skeletal mass has been documented in Wilson’s disease for several years, the mechanism involved remains undefined [261,262]. However, dietary copper supplementation for 6 weeks failed to alter biochemical markers of bone in healthy adults [263].

References 1. J. Kanis and F. A. Pitt, Epidemiology of osteoporosis, Bone 13, S7 – S15 (1992). 2. E. Seeman, The dilemma of osteoporosis in men. Am. J. Med. 98, 76S – 88S (1995). 3. E. Krall and B. Dawson Hughes, Heritable and life – style determinants of bone mineral density. J.Bone Miner. Res. 8, 1 – 10 (1993). 4. V. A. McKusick, Heritable, “Heritable Disorders of Connective Tissue.” Mosby, St. Louis, 1972. 5. P. Byers, Brittle bones – fragile molecules: Disorders of collagen gene structure and expression. Trends’Genet. 6, 293 – 300 (1990). 6. D. Sillence, A. Senn, and D. Danks, Genetic heterogeneity in osteogenesis imperfecta. J. Med Genet. 16, 101 – 116 (1979). 7. P. Beighton, A De Paepe, B. Steinmann, P. Tsipovras, and R. J. Wenstrup, Ehlers – Danlos syndromes: Revised nosology, Villefranche 1997. Ehlers – Danlos National Foundation (USA) and Ehlers – Danlos Support Group (UK). Am. J. Med. Genet. 77(1), 31 – 37 (1998). 8. G. Smars, “Osteogenesis Imperfecta in Sweden: Clinical, Genetic Epidemiological and Socio-medical Aspects.” Stockholm Svenska Bokforlaget, 1961. 9. R. Smith, M. Francis, and G. Houghton, “The Brittle Bone Syndrome: Osteogenesis Imperfecta.” Butterworths, London, 1983. 10. K. Tan and E. Tock, Osteogenesis imperfecta congenita. Aust Paediatr. J. 7, 49 – 53 (1971). 11. J. Robert, J. Gremeau, A. Notter, and J. Guilhot, Hereditary bone fragility. Lyon Med. 219, 881 – 982 (1968). 12. S. Navani and B. Sarzin, Intrauterine osteogenesis imperfecta: Review of the literature and a report of the radiological and necropsy findings in two cases. Br. J. Radiol. 40, 449 – 452 (1967). 13. G. Schroder, Zeit. Mensch. Vererbunges konstitutionslehre 37, 632 – 676 (1964). 14. C. Herndon, Osteogenesis imperfecta: Some clinical and genetic considerations. Clin. Orthop. Relat. Res. 8, 132 – 141 (1956). 15. F. E. MeGuigan, D. M. Reid, and S. H. Ralston, Susceptibility to osteoporotic fracture is determined by allelic variation at the Spl site, rather than other polymorphic sites at the COLIAI locus. Osteoporosis Int. 11, 338 – 342 (2000). 16. D. Tilstra and P. Byers, Molecular basis of hereditary disorders of connective tissue. Annu. Rev. Med. 45, 149 – 163 (1994). 17. E. S. Kurland, C. J. Rosen, F. Cosman, D. McMahon, F. Chan, E. Lindsay, D. Dempster and J. P. Belezikian, Insulin-like growth factor-I in men with idiopathic osteoporosis, J. Clin. Endo. Metab. 82, 2799 – 2805 (1997). 18. J. Korkko, S. Annunen, T. Pihlajamaa, D. J. Prockop, and L. Ala – Kokko, Conformation sensitive gel electrophoresis for simple and accurate detection of mutations: Comparison with denaturing gradient gel electrophoresis and nucleotide sequencing. Proc. Natl. Acad. Sci. USA 95, 168 1– 1685 (1998). 19. G. Wallis, B. Sykes, P. H. Byers, C. G. Matthews, D. Viljoen, and P. Beioghton, Osteogenesis imperfecta type III: Mutations in the type I collagen structural genes, COLIA1 and COLIA2, are not necessarily responsible. J. Med. Genet. 30, 192 – 196 (1993).

296 20. S. A. Jiminez, Collagens. In “Principles of Molecular Rheumatology” (G. C. Tsokos, ed.), pp. 175 – 196. Humana Press, Totowa, NJ, 2000. 21. R. Burgeson and M. Nimni, Collagen types. Molecular structure and tissue distribution. Clin. Orthop. 282, 250 – 272 (1992). 22. M. van der Rest and R. Garrone, Collagen family of proteins. FASEB J. 5, 2814 – 2823 (1991). 23. D. Birk, J. Fitch, J. Babiarz, and T. Linsmayer, Collagen type I and type V are present in the same fibril in the avian corneal stroma. J. Cell. Biol. 106, 999 – 1008 (1988). 24. M. Van Der Rest and R. Mayne, Type IX collagen proteoglycan from cartilage is covalently linked to type II collagen. J. Bio1. Chem. 263, 1615 – 1618 (1988). 25. R. Iozzo and I. Cohen, Altered Proteoglycan gene expression and the tumor stroma. Experientia 49, 447 – 455 (1993). 26. K. Vogel, B. Paulsson, and D. Heinegard, Specific inhibition of type I and type II collagen fibrillogenesis by the small proteoglycan of tendon. Biochem. J. 223, 587 – 597 (1984). 27. Y. Yamaguchi, D. Mann, and E. Ruoslahti, Negative regulation of transforming growth factor-beta by the proteoglycan decorin. Nature 436, 281 – 284 (1990). 28. K. Kivirikko, Collagens and their abnormalities in a wide spectrum of diseases. Ann. Med. 25, 113 – 126 (1993). 29. Tasab, M. R. Batten and N. J. Bulleid, Hsp47: A molecular chaperone that interacts with and stabilizes correctly-folded procollagen. EMBO J. 19, 2204 – 2211 (2000). 30. P. Delmas, Biochemical markers of bone turnover. I. Theoretical considerations and clinical use in osteoporosis. Am. J. Med. 95, 11S – 16S (1993). 31. R. Dalgleish, The human collagen mutation database 1998. Nucl. Acids Res. 26, 253 – 255 (1998). 32. W. G. Cole, The Nicholas Andry Award, 1996. The molecular pathology of osteogenesis imperfecta Clin. Orthos. 343, 235 – 248, (1997). 33. S. D. Chipman, J. R. Shapiro, M. B. McKinstry, M. L. Stover, P. Branson, and D. W. Rowe, Expression of mutant alpha (I) – procollagen in osteoblast and fibroblast cultures from a proband with osteogenesis imperfect type IV. J. Bone Miner. Res. 7, 793 – 805 (1992). 34. K. Mackay, P. H. Byers, and R. Dalgleish, An RTPCR-SSCP screenig strategy for detection of mutations in the gene encoding the alpha I chain of type I collagen: Application to four patients with osteogenesis imperfecta. Hum. Mol. Genet. 2, 1155 – 1160 (1993). 35. M. White, M. Calvalho, D. Derse, S. O’Brien, and M. Dean, Detecting single base substitutions as heteroduplex polymorhpisms. Genomics 12, 301 – 306 (1992). 36. C. Genovese, A. Brufsky, J. Shapiro, and D. Rowe, Detection of mutations in human type I collagen mRNA in osteogenesis imperfecta by indirect RNase protection. J. Biol. Chem. 264, 9632 – 9637 (1989). 37. J. Marini, D. Grange, G. Gottesman, M. Lewis, and D. Koeplin, Osteogenesis imperfect type IV. Detection of a point mutation in one alpha I (I) collagen allele (COLIAI) by RNA/RNA hybrid analysis. J. Biol. Chem. 264, 11893 – 11900 (1989). 38. P. Bianco, M. Riminucci, A. Majolagbe, S. A. Kuznetsov, M. T. Collins, M. H. Mankani, A. Corsi, H. G. Bone, S. Wientroub, A. M. Spiegel, I. W. Fisher, and P. G. Robey, Mutations of the GNASI gene, stromal cell dysfunciton, and osteomalacic changes in nonMcCune-Albright fibrous dysplasia of bone. J. Bone Miner. Res. 15, 120 – 128 (2000). 39. H. Bachinger, N. Morris, and J. Davis, Thermal stability and folding of the collagen triple helix and the effects of mutations in osteogenesis imperfecta on the triple helix of type I collagen. Am. J. Med. Genet. 45, 152 – 162 (1993). 40. A. Sieron, A Fertala, L. Ala-kokko, and D. Prockop, Deletion of a large domain in recombinant human procollagen II does not after the thermal stability of the triple helix. J. Biol. Chem. 268, 21232 – 21237 (1993).

JAY R. SHAPIRO 41. A. Westerhausen, J. Kishi, and D. Prockop, Mutations that substitute serine for glycine alpha 1-598 and glycine alpha 1-631 in type I procollagen. The effects on thermal unfolding of the triple helix are positon-specific and demonstrate that the protein unfolds through a series of cooperative blocks. J. Biol. Chem. 265, 13995 – 14000 (1990). 42. D. Rowe, J. Shapiro, M. Porior, and S. Schlesinger, Diminished type I collagen synthesis and reduced alpha 1(I) collagen messenger RNA in cultured fibroblasts from patients with dominantly inherited (type I) osteogenesis imperfecta. J. Clin. Invest. 76, 604 – 611 (1985). 43. D. Prockop, Osteogenesis imperfecta: Phenotypic heterogeneity, protein suicide, short and long collagen Am. J. Hum. Genet. 36, 499 – 505 (1984.) 44. M. Wirtz et al., A cysteine for glycine substitution at position 175 in an alpha 1(I) chain of type I collagen produces a clinically hterogeneous form of osteogenesis imperfecta. Connect. Tissue Res. 29, 1 – 11 (1993). 45. S. Deak et al., The substitution of arginine for glycine 85 of the alpha 1(I) procollagen chain results in mild osteogenesis imperfecta. The mutation provides direct evidence fot three discrete domains of cooperative melting of intact type I collagen. J. Biol. Chem. 266, 21827 – 21832 (1991). 46. K. Kadler et al., A type I collagen with substitution of a cysteine for glycine-748 in the alpha 1(I) chain copolymerizes with normal type I collagen and can generate fractal-like structures. Biochemistry 30, 5081 – 5088 (1991). 47. B. Vogel et al., A substitution of cysteine for glycine 748 of the alpha 1 chain produces kink at this site in the procollagen I molecule and an altered N-proteinase cleavage site over 225 nm alway. J. Biol. Chem. 263, 19249 – 19255(1988). 48. N. Rose, K. MacKay, P. Byers, and R. Dalgleish, A Gly859Ser substitution in the triple helical domain of the alpha 2 chain of type I collagen resulting in osteogenesis imperfecta type III in two unrelated individuals. Hum. Mutat. 3, 391 – 394 (1994). 49. C. Pruchno et al., Osteogenesis imperfecta due to recurrent point mutations at CpG dinucleotides in the COLIAI gene of type I collagen. Hum. Genet. 87, 33 – 40 (1991). 50. R. Wenstrup et al., The effects of different cysteine for glycine substitutions within alpha 2(I) chains. Evidence of distinct structural domains within the type I collagen triple helix. J. Biol. Chem. 266, 2590 – 2594 (1991). 51. J. Marini, M. Lewis, Q. Wang, K. Chen, and B. Orrison, Serine for glycine substitutions in type I collagen in two cases of type IV osteogenesis imperfecta (OI). Additional evidence for a regional model of OI pathophysiology. J. Biol. Chem. 268, 2667 – 2673 (1993). 52. J. Bateman et al., Chemical cleavage method for the detection of RNA base changes: Experience in application to collagen mutations in osteogenesis imperfecta. Am. J. Med. Genet. 45, 233 – 240 (1993). 53. W. Cole, Collagen genes: Mutations affecting collagen structure and expression. Prog. Nucleic Acid Res. Mol. Biol. 47, 29 – 80 (1991). 54. L. Spotila et al., Mutation in a gene for type I procollagen (COLIA2) in a woman with postmenopausal osteoporosis: Evidence for phenotypic and genotypic overlap with mild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423 – 5427 (1991). 55. C. Paterson, S. McAllion, and R. Miller, Heterogeneity of osteogenesis imperfecta type I. J. Med. Genet. 20, 203 – 205 (1983). 56. D. Sillence, “Fifth International Conference on Osteogenesis Imperfecta.” Oxford, 1993. 57. C. Paterson, S. McAllion, and J. Stellman, Osteogenesis imperfecta after the menopause. N. Engl. J. Med. 310, 1694 – 1696 (1984). 58. C. Paterson and P. Mole, Oxford, Sept (1993). 59. B. Riggs and L. I. Melton, “Osteoporosis: Etiology, Diagnosis and Management”. Raven Press, New York, 1988.

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development 60. E. McCarthy, K. Earnst, K. Rossiter, and J. Shapiro, Bone histomorphometry in adults with type IA osteogenesis imperfecta. Clin. Orthop. Pediatr. Res. 173, 204 – 207 (1997). 61. R. Baron, J. Gertner, R. Lang, and A. Vignery, Increase bone turnover with decreased bone formation by osteoblasts in childern with osteogenesis imperfecta tarda. Pediatr. Res. 17, 204 – 207 (1983). 62. F. Rauch, R. Travers, A. M. Parfitt and F. H. Glorieux, Static and dynamic bone histomorphometry in childern with osteogenesis imperfecta Bone 26, 581 – 589 (2000). 63. M. C. Willing, S. P. Deschenes, R. L. Slayton, and E. J. Roberts, Premature chain termination in a unifying mechanism for COLIAI nulkl alleles in osteogenesis imperfecta type I cell strains. Am. J. Hum. Genet. 59, 799 – 809 (1996). 64. B. Sykes, M. Francis, D. Phil, and R. Smith, Altered relation of two collagen types in osteogenesis imperfect. N. Engl. J. Med. 296, 1200 – 1203 (1977). 65. D. A. Redford-Badwal, M. L. Stover, M. Valli, M. B. McKinstry, and D. W. Rowe, Nuclear retention of COLIAI messenger RNA identifies null alleles causing mild osteogenesis imperfecta. J. Clin. Invest. 97, 1035 – 1040 (1996). 66. C. Genovese and D. Rowe, Analysis of cytoplasmic and nuclear messenger RNA in fibroblasts from patients with type I osteogenesis imperfecta. Methods Enzymol. 145, 223 – 235 (1987). 67. M. Edwards, R. Wenstrup, P. Byers, and D. Cohn, Recurrence of lethal osteogenesis imperfecta due to parental mosaicism for mutation in the COLIA2 gene of type I collagen. The mosaic parent exhibits phenotypic features of a mild form of the disease. Hum. Mutat. 1, 47 – 54 (1992). 68. S. Rabushka, L. Love, and H. Kadison, Isolated thoracie dysostosis Radiology 106, 161 – 175 (1973). 69. B. Cremin, H. Goodman, J. Spranger, and P. Beighton, Wormian bones in osteogenesis imperfecta and other disorders. Skeletal Radiol. 8, 35 – 38 (1982). 70. D. Sillence, K. Barlow, A. Garber, J. Hall, and D. Rimoin, Osteogenesis imperfecta type II delineation of the phenotype with reference to genetic heterogeneity. Am. J. Med. Genet. 17, 407 – 423 (1984). 71. J. Shapiro, V. E. Burn, S. D. Chipman, J. B. Jacobs, B. Schloo, L. Reid, N. Larsen, F. Louis, Pulmonary hypoplasia and osteogenesis imperfecta type II with defective synthesis of alpha I(1) procollagen. Bone 10, 165 – 171 (1989). 72. E. Patterson, E. Smiley, and J. Bonadio, RNA sequence analysis of a perinatal lethal osteogenesis imperfecta mutation. J. Biol. Chem. 264, 10083 – 10087 (1989). 73. M. Willing, D. Cohn, and P. Byers, Frameshift mutation near the 3’ end of the COLIA1 gene of type I collagen predicts an elongated Pro alpha 1(I) chain and results in osteogenesis imperfecta type I. J. Clin. Invest. 85, 282 – 290 (1990). 74. S. Lamande, H. Dahl, W. Cole, and J. Bateman, Characterization of point mutations in the collagen COLIA1 and COLIA2 genes causing lethal perinatal osteogenesis imperfecta. J. Biol. Chem. 264, 15809 – 15812 (1989). 75. D. Cohn, B. Starman, B. Blumberg, and P. Byers, Recurrence of lethal osteogenesis imperfecta due to parental mosaicism for dominant mutation in a human type I collagen gene (COLIA1). Am. J. Hum. Genet. 46, 591 – 601 (1990). 76. D. Cohn, G. Wallis, M. Edwards, B. Starman, and P. Byers, “International Conference on Osteogenesis Imperfecta,” Pavia, 1990. 77. D. Cohn, G. Wallis, and X. Zhang, and P. Byers, Matrix 10, 236 (1990). 78. P. Byers, G. Wallis, M. Willing, Osteogenesis Imperfecta: Translation of mutation to phenotype. J. Med. Genet. 28, 433 – 442 (1991). 79. K. MacKay, A. DePaepe, B. Steinman, and R. Dalgleish, “Fifth International Conference on Osteogenesis Imperfecta,” Oxford, 1993.

297

80. M. Chu, V. Gargiulo, C. Williams, and F. Ramirez, Multiexon deletion in an Osteogenesis Imperfecta variant with increased type III collagen mRNA. J. Biol. Chem. 260, 691 – 694 (1985). 81. G. Barsh, C. Roush, J. Bonadio, P. Byers, and R. Gelinas, Intronmediated recombination may cause a deletion in an alpha 1 type I collagen chain in a lethal from of Osteogenesis Imperfecta Proc. Natl. Acad. Sci. USA 82, 2870 – 2874 (1985). 82. J. Hawkins, A. Superti-Furga, B. Steinman, and R. Dalgleish, A 9-base pair deletion in COLIAI in a lethal variant of osteogenesis mperfecta J. Biol. Chem. 266, 22370 – 22374 (1991). 83. G. Wallis, B. Starman, S. Chessler, M. Willing, and P. Byers, J. Clin. Invest. 84, 1206 – 1214 (1990). 84. L. Cohen-Solal, L. Zylberberg, A. Sangalli, M. Lira, and M. Mottes, Substitution of an aspartic acid for glycine 700 in the alpha 2(I) chain of typeI collagen in a recurrent lethal type II osteogenesis Imperfecta dramatically affects the mineralization of bone. J. Biol. Chem. 269(20), 14751 – 14758 (1994). 85. P. Byers, P. Tsipouras, J. Bonadio, B. Starman, and R. Schwartz, Perinatal lethal osteogenesis Imperfecta (OI type II): A biochemically heterogeneous disorder usually due to new mutations in the genes for type I collagen. Am. J. Hum. Genet. 42, 237 – 248 (1988). 86. D. Cohn, C. Pruchno, X. Zhang, and P. Byers, Am. J. Hum. Genet. 47, 110 – 115 (1990a). 87. M. Willing et al., Heterozygosity for large deletion in the alpha 2(I) collagen gene has a dramatic effect on type I collagen secretion and produces preinatal lethal osteogenesis Imperfecta. J. Biol. Chem. 263, 8398 – 8404 (1988). 88. A. Ganguly et al., Heterozygous mutation in the G+5 position of intron 33 of the pro-alpha 2(I) gene (COL1A2) that causes aberrant RNA splicing and lethal osteogenesis Imperfecta. Use of carbodiimide methods that decrease the extent of DNA sequencing necessary to define an unusual mutation. J. Biol. Chem, 266, 12035 – 12040 (1991). 89. B. Starman et al., Osteogenesis Imperfecta. The position of substitution for glycine by cysteine in the triple helical domain of the pro alpha 1(I) chains of type I collagen determines the clinical phenotype. J. Clin. Invest. 84, 1206 – 1214 (1989). 90. A. Nicholls, J. Oliver, D. Renour, M. Keston, and F. Pope, Subsititution of cysteine for glycine at residue 415 of one allele of the alpha 1(I) chain of type I procollagen in type III/IV osteogenesis Imperfecta. J. Med. Genet. 28, 757 – 764 (1991). 91. M. Pack, C. Constantinou, K. Kalia, K. Nielson, and D. Prockop, Substitution of serine for alpha 1(I)-glycine 844 in a severe variant of osteogenesis imperfecta minimally destabilizes the triple helix of type I procollagen. The effects of glycine substitutions on thermal stability are either position of amino acid specific. J. Biol. Chem. 264, 19694 – 19699 (1989). 92. J. Batman et al., “IV International Conference on Osteogenesis Imperfecta” Pavia, 1990. 93. G. Wallis, B. Starman, and P. Byers, Clinical heterogeniety of OI explained by molecular heterogeneity and somatic mosaicism. Am. J. Hum. Genet. 45, A228 (1989). 94. A. M. Lund, E. Astrom, S. Soderhall, and M. Schwartz, and F. Skovby, Osteogenesis Imperfecta: Mosaicism and refinement of the genotype-phenotype map in OI type III. Mutations in brief no. 242. Online. Hum. Mutat. 13(6), 503 (1999). 95. J. Batman, M. Hannagan, S. Lamande, I. Moeller, D. Chan, and W. Cole, “IV International Conference on Osteogenesis Imperfecta” pavia, Italy 1990. [Abstract 2] 96. A. Superti-Furga, M. Raghunath, F. Pistone, C. Romano, and B. Steinmann, “IV International Conference on Osteogenesis Imperfecta” Pavia, 1990. 97. F. Thompson, I. Young, C. Hall, and M. Pembrey, Recurrence risks and prognosis in severe sporadic Osteogenesis Imperfecta. J. Med. Genet. 24, 390 – 405 (1987).

298 98. E. H. Glorieux, F. Rauch, H. Plotkin, L. Ward, R. Travers, P. Roughley, L. Lalic, D. F. Glorieux, F. Fassier, N. Bishop, Type V osteogenesis imperfecta: A new form of brittle bone disease. J. Bone Miner. Res. 15, 1650 – 1650, (2000). 99. M. DiMaio et al., First-trimester prenatal diagnosis of osteogenesis imperfecta type II by DNA analysis and sonography. Prenatal Diagn. 13, 589 – 596 (1993). 100. R. Sharony, C. Browne, R. Lachman, and D. Rimoin, Prenatal diagnosis of the skeletal dysplasias. Am. J. Obstet. Gynoecol. 169, 668 – 675 (1993). 101. D. Bulas et al., Variable prenatal appearance of osteogenesis imperfecta J. Ultrasound Med. 13, 419 – 426 (1994). 102. M. Bronshtein and Z. Weiner, Anencephaly in a fetus with osteogenesis imperfecta: Early diagnosis by transvaginal sonography. Prenatal Diagn. 12, 831 – 834 (1992). 103. M. Raghunath, B. Steinmann, C. Delozier-Blanchet, P. Extermann, and A. Superti-Furga, Prenatal diagnosis of collagen disorders by direct biochemical analysis of chorionic villus biopsies. Pediatr. Res. 36, 441 – 448 (1994). 104. M. J. Edwards, R. J. Wenstrup, P. H. Byers and D. H. Cohn, Recurrence of lethal osteogenesis imperfecta due parental mosaicism for a mutation in the COL1A2 gene of type I collagen. The mosaic parer exhibits phenotypic features of a mild form of the disease. Hum. Mutat. 1, 47 – 54 (1992). 105. M. Mottes et al., Paternal mosaicism for a COL1A1 dominant mutation (alpha 1 Ser – 415) causes recurrent osteogenesis imperfecta Hum. Mutat. 2, 196 – 204 (1993). 106. C. Constantinou-Deltas, R. Ladda, and D. Prockop, Somatic cell mosaicism: Another source of phenotypic heterogeneity in nuclearfamilies with osteogenesis imperfecta. Am. J. Med. Genet. 45, 246 – 251 (1993). 107. K. Falvo and P. Bullough, Osteogenesis imperfecta: A histometric analysis. J. Bone Jt. Surg. 56(A), 783 – 793 (1973). 108. S. Doty and R. Matthews, Electron microscopic and histochemical investigation of osteogenesis imperfecta tarda. Clin. Orthop. Relat. Res. 80, 191 – 201 (1971). 109. W. Traub, T. Arad, U. Vetter, and S. Weiner, Ultrastructural studies of bones from patients with osteogenesis imperfecta. Matrix Biol. 14, 337 – 345 (1994). 110. P. Fratzl, O. Paris, K. Klaushofer, and W. Landis, J. Bone. Miner. Res. 9 (Suppl. 12), A160 (1994). 111. J. Ramser and H. Frost, The study of a rib biopsy from a patient with ostegenesis imperfecta. A method using in vivo tetracycline labeling. Acta. Orthop. Scand. 37, 229 – 240 (1996). 112. J. Albright and J. A. Grunt, Studies of patients with osteogenesis imperfecta. J. Bone Jt. Surg. 53(A), 1415 – 1425 (1971). 113. F. Glorieux, R. Travers, G. Chabot, and G. Lanoue, J. Bone Miner. Res. 9, A504 (1994). 114. F. Rauch, R. Travers, A. M. Parfitt, F. H. Glorieux Static and dyanamic bone histomorphometry in childern with osteogenesis imperfecta. Bone 26, 581 – 589 (2000). 115. R. Brenner, U. Vetter, J. Ittner, and W. Teller, Bone resorption assessed by immunoassay of urinary cross-linked collagen peptides in patients with osteogenesis imperfecta. Acta Pediatr. 82, 764 – 767 (1993). 116. R. Brenner, U. Vetter, A. Bollen, M. Morike, and D. Eyre, Bone resorption assessed by immunoassay of urinary cross-linked collagen peptides in patients with oteogenesis imperfecta. J. Bone Miner. Res. 9, 993 – 997 (1994). 117. S. Minisola, R. Rosse, and, E. Romagnoli, Bone resorption assessed by immunoassay of urinary cross-linked collagen peptides in patients with osteogenesis imperfecta. J. Bone Miner. Res. 10, 335 – 336 (1995). 118. J. R. Shapiro, Comment. J. Bone Miner. Res. 10, 338 – 339 (1995).

JAY R. SHAPIRO 119. A. M. Lund, M. Hanser, G. Kollerup, A. Juul, B. Teisner, and F. Skovby, Collagen-derived markers of bone metabolism in osteogenesis impertecta. Acta Paediatr. 87, 1131 – 1137 (1998). 120. D. Siuence, Natural history of blue sclerae in osteogenesis imperfecta. Am. J. Med. Genet. 45, 183 – 186 (1993). 121. P. Beighton, “The Ehlers – Danlos Syndrome.” Heineman, London, 1970. 122. E. Petty, et al., Dermatosparaxis in children. A case report and review of the newly recognized phenotype Arch. Dermatol. 129, 1310 – 1315 (1993). 123. F. Meire, J. Van Egmond, and M. Hanssens, Congenital Marfan syndrome with contractures. A clinicopathological report. Bull. Soc. Belge Ophthlmol. 245, 91 – 97. 124. M. Shohat, D. Rimoin, H. Gruber, and R. Lachman, Perinatal lethal hypophosphatasia: Clinical, radiologic and morphologic findings. Pediatr. Radiol. 21, 421 – 427 (1991). 125. M. Kaiser Kupfer, L. McClain, and J. Shapiro, Low ocular rigidity in patients with osteogenesis imperfecta. Invest. Opthalmol. Vis. Sci. 20, 807 – 812 (1993). 126. M. Vrba and K. Cada, [Results of measurement of the corneoscleral rigidity in otosclerosis and Van der Hoeve-de Kleyn syndrome]. Czech. Otolaryngol. 20, 79 – 81 (1971). 127. C. Chan, W. Green, Z. de la Cruz, and A. Hillis, Ocular findings in osteogenesis imperfecta congenita. Arch. Ophthalmol. 100, 1458 – 1463 (1982). 128. R. Ellis, Proc. R. Soc. Med. 24, 421 – 427 (1931). 129. P. Remigio and H. Grinvalsky, Osteogenesis impferfecta congenita. Association with conspicuous extraskeletal connective tissue dysplasia. Am. J. Dis. Childs. 119, 524 – 528 (1970). 130. L. Bergstrom, Osteogenesis imperfecta: Otologic and axillofacial aspects. Laryngoscope 87, 1 – 50 (1977). 131. H. Ranta, P. Lukinmaa, and J. Waltimo, Heritable dentin defects: Nosology, pathology, and treatment. Am. J. Med. Genet. 45, 193 – 200 (1993). 132. L. Levin, J. Brady, and M. Meinick, Scanning electron microscopy of teeth in dominant osteogenesis imperfecta: Support for genetic heterogeneity. Am. J. Med. Genet. 5, 189 – 199 (1980). 133. H. M. Aplin, K. L. Hirst, and M. J. Dixon, Refinement of the dentinogenesis imperfecta type II locus to an interval of less than 2 Centimorgans at chromosome 4q21 and the creation of a yeast artificial chromosome contig of the critical region. J. Dent. Res. 78, 1270 – 1276 (1999). 134. A. C. O’Connel and J. C. Marini, Evaluation of oral problems in an osteogenesis imperfecta population. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 87, 189 – 196 (1999). 135. A. J. Garretsen, C. W. Cremers, and P. L. Huygen, Hearing loss (in nonoperated ears) in relation to age in osteogenesis imperfecta type I. Ann. Otol. Rhinol. Laryngol. 106, 575 – 582 (1997). 136. E. Reidner, S. Levin, and M. Holliday, Hearing patterns in dominant osteogenesis imperfecta. Arch. Otolaryngol. 106, 737 – 740 (1980). 137. J. Shapiro, A. Pikus, G. Weiss, and D. W. Rowe, Hearing and middle ear function in osteogenesis imperfecta. J. Am. Med. Assoc. 247, 2120 – 2126 (1982). 138. U. Ross, R. Laszing, H. B. N. Ornemann, and C. Ulrich, Osteogenesis imperfecta: Clinical symptoms and update findings in computed tomography and tympano-cochlear scintigraphy. Acta. Otolaryngol. 113, 620 – 624 (1997). 139. R. Dieler, J. Muller, and J. Helms, Stapes surgery in osteogenesis imperfecta patients Eur. Arch. Ororhinolaryngol. 254, 120 – 127 (1997). 140. T. Garretsen and C. Cremers, Ear surgery in osteogenesis imperfecta. Clinical finding and short-term and long-term results. Arch. Otolaryngol. Head Neck Surg. 116, 317 – 323 (1990).

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development 141. N. White, C. Winearls, and R. Smith, Cardiovascular abnormalities in osteogenesis imperfecta. Am. Heart J. 106, 1416 – 1420 (1983). 142. J. Hortop, P. Tsipouras, and J. Hanley, Cardiovascular involvement in osteogenesis imperfecta. Circulation 73, 54 – 61 (1986). 143. R. Heppner, H. Babitt. J. Bianchine, and J. Warbasse, Cardiovascular involvement in osteogenesis imperfecta. Am. J. Cardiol. 31, (1973). 144. R. Zegdi, N. D’ Attellis, P. Fornes, J. F. Fuzellier, J. P. Carteaux, J. N. Fabiani, and A. Carpentier, Aortic valve surgery in osteogenesis imperfecta: Report of two cases and review of the literature. J. Heart Valve Dis. 7, 510 – 514 (1998). 145. M. Hayes, G. Parker, J. Ell, and D. Sillence, Basilar Impression complicating osteogenesis imperfecta type IV. The clinical and neurolradiological finding in four cases. J. Neurol. Neurosurg. Psychiatry 66, 357 – 362 (1999). 146. M. Kurimoto, S. Ohara, and A. Takaku, Basilar impression in osteogenesis imperfecta trada. Case report. J. Neurosurg. 74, (1991). 147. L. Hurwitz and R. McSweney, Brain 83, 138 – 149 (1960). 148. L. Charnas and J. Marini, Communicating hydrocephalus, basilar invagination, and other neurologic features in osteogenesis imperfecta. Neurology 43, 2603 – 2608 (1993). 149. R. S. Bhangoo, and H. A. Crockard, Transmaxillary anterior decompression in patients with severe basilar invagination. Clin. Orthop. 359, 115 – 125 (1999). 150. A. De Paepe, J. Leroy, L. Nuytinck, F. Meire, and J. Capoen, Osteoporosis – pseudoglioma syndrome. Am. J. Med. Genet. 45, 30 – 37 (1993). 151. T. Wang, G. Yang, and A. Alba, Chronic ventilator use in osteogenesis imperfecta congenita with basilar impression: A case report Arch. Phys. Med. Rehabil. 75, 699 – 702 (1994). 152. K. Falvo, D. Klain, A. Krauss, L. Root, and P. Auld, Pulmonary function studies in osteogenesis imperfecta. Am. Rev. Respir. Dis. 108, 1258 (1973). 153. L. Tosci, L. Gerber, and J. Marini, “Fifth International Conference on Osteogenesis Imperfecta.” Oxford, 1993. 154. A. Karbowski, M. Schwitalle, and A. Eckhardt, Scoliosis in patients with osteogenesis imperfecta: A federal nation-wide cross-sectional study. Z. Orthop. Ihre Grezgeh. 137, 219 – 222 (1999). 155. H. Lehmann, A. Nerlich, R. Brenner, M. Bodo, and P. Muller, Hyperplastic callus formation in osteogenesis imperfecta. Eur. J. Pediatr. Surg. 2, 281 – 284 (1993). 156. R. Brenner et al., Biochemical analysis of callus tissue in osteogenesis imperfecta type IV. Evidence for transient overmodification in collagen types I and III. Clin. S. Invest. 84, 915 – 921 (1989). 157. O. Micke, W. Wagner, R. Potter, F. Prott, and A. Krbowski, [The radiotherapy of hyperplastic heterotopic ossification in osteogenesis imperfecta. 2 case reports]. Strahlentherpie Onkol. 170, 347 – 351 (1994). 158. N. Fedarko, U. Vetter, S. Weinstein, and P. Gehron Robey, Age-related changes in hyaluronan, proteoglycan, collagen, and osteonectin synthesis by human bone cells. J. Cell. Physiol. 151, 215 – 227 (1992b). 159. M. Morike et al., In vitro expression of osteoblastic markers in cells isolated from normal fetal and postnatal human bone and from bone of patients with osteogenesis imperfecta. J. Cell. Physiol. 1587, 439 – 444 (1993). 160. N. Fedarko, M. Moerike, R. Brenner, P. Gehron Robey, and U. Vetter, Extracellular matrix formation by osteoblasts from patients with osteogenesis imperfecta J. Bone Miner. Res. 7, 921 – 930 (1994a). 161. U. Vetter et al., Osteogenesis imperfecta: Changes in nocollagenous proteins in bone. J. Bone Miner. Res. 6, 501–505 (1991a). 162. N. S. Fedarko, P. D’Avis, C. R. Frazier, M. J. Burrill, V. Fergusson, M. Tayback, P. D. Sponseller, J. R. Shapiro, Cell proliferation of human fibroblasts and osteoblasts in osteogenesis imperfecta: Influence of age. J. Bone. Miner. Res. 10, 338 – 339 (1995).

299

163. N. S. Fedarko, P. D. Sponseller, and J. R. Shapiro, Long-term extracellular matrix metabolism by cultured human osteogenesis imperfecta osteoblasts. J. Bone Miner. Res. 11, 800 – 805 (1996). 164. G. Cetta et al., Deposition of mutant type I collagen in the exracellular matrix of cultured dermal fibroblasts in osteogenesis imperfecta. Connect. Tissue Res. 29, 41 – 49 (1993). 165. J. Bateman and S. Golub, Deposition and selective degradation of structurally-abnormal type I collagen in a collagen matrix produced by osteogenesis imperfecta fibroblasts in vitro. Matrix Biol. 14, 251 – 262 (1994). 166. K. Fujii, T. Kajawara, H. Kurosu, and M. Tanzer, Osteogenesis imperfecta: Altered content of type III collagen and proportion of the crosslinks in skin. FEBS Lett. 82, 251 – 254 (1993). 167. U. Vetter, M. Weis, M. Morike, E. Eanes, and D. Eyre, Collagen crosslinks and mineral crystallinity in bone of patients with osteogenesis imperfecta. J. Bone Miner. Res. 8, 133 – 137 (1993). 168. J. Bateman, D. Cahn, T. Mascara, J. Rogers, and W. Code, Collagen defects in lethal perinatal osteogenesis imperfecta. Biochem. J. 217, 103 – 115 (1986). 169. L. Fisher, E. Eanes, L. Denholm, B. Heywood, and J. Termine, Two bovine models of osteogenesis imperfecta exhibit decreased apatite crystal size. Calcif. Tissue Int. 40, 282 – 285 (1987). 170. U. Vetter, E. Eanes, J. Kopp, J. Termine, and P. Gehron Robey, Changes in apatite crystal size in bones of patients with osteogenesis imperfecta. Calcif. Tissue Int. 49, 248 – 250 (1991). 171. M. Muriel et al., Morphological and biochemical studies of a mouse mutant (fro/fro) with bone fragility. Aud. Bone. 12, (1983). 172. L. Denholm and W. Cole, Heritable bone fragility, joint laxity and dysplastic dentin in Friesian calves: A bovine syndrome of osteogenesis imperfecta. Aust. Vet. J. 60, 9 – 17 (1983). 173. J. Termine et al., Osteonectin, bone proteoglycan, and phosphophoryn defects in a form of bovine osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 81, 2213 – 2217 (1984). 173a.B. Campbell, J. A. Wootton, J. N. MacLeod, R. R. Minor, Sequence of normal canine COL1A1 cDNA and identification of a heterozygous alpha (1) collagen GLY208 – ALA mutation in a severe case of canine osteogenesis imperfecta. Arch. Biochem. Biophys. 384, 37 – 46 (2000). 174. J. Bonadio et al., Transgenic mouse model of the mild dominant from of osteogenesis imperfecta. Proc. Natl. Acad. Sci. (1990). 175. J. Khillan, A. Olsen, S. Kontussari, B. Sokolov, and D. Prockop, Transgenic mice that express a mini-gene version of the human gene for type I procollagen (COLIA1) develop a phenotype resembling a lethal form of osteogenesis imperfecta. J. Biol. Chem. 266, 23373 – 23379 (1991). 176. R. Pereira, J. Killan, H. Helminen, E. Hume, and D. Prockop, Transgenic mice expressing a partially deleted gene for type I procollagen (COLIA1). A breeding line with a phenotype of spontaneous fractures and decreased bone collagen and mineral. J. Clin. Invest. 91, 709 – 716 (1993). 177. S. Chipman, H. Sweet, D. McBride, W. Davisson, S. Marks, A. Shuldiner, R. Wenstrup, and D. Rowe, Defective pro alpha 2(I) collagen synthesis in a recessive mutation in mice: A model of human osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993). 178. S. Deak, A. Nicholls, F. Pope, and D. Prockop, The molecular defect in a nonlethal variant of osteogenesis imperfecta. Synthesis of pro-alpha 2(I) chains which are not incorporated into trimers of type I procollagen. J. Biol. Chem. 258, (1983). 179. A. Nicholls, F. Pope, and H. Schloon, Biochemical heterogeneity of osteogenesis imperfecta: New variant. Lancet 1, 1193 (1979). 180. C. Pak et al., Slow-release sodium fluoride in the management of postmenopausal osteoporosis. A randomized controlled trial. Ann. Int. Med. 120, 625 – 632 (1994).

300 181. M. Whyte, personal communication, 2000. 182. G. August, J. Shapiro, and W. Hung, Calcitonin therapy of childern with osteogenesis imperfecta. J. Pediatr. 91, 1001 – 1005 (1977). 183. S. Castells et al., Therapy of osteogenesis imperfecta with synthetic salmon calcitonin. J. Pediatr. 95, 807 – 811 (1979). 184. Y. Nishi et al., Effect of long-term calcitonin therapy by injection and nasal spray on the incidence of fractures in osteogenesis imperfecta. J. Pediatr. 121, 477 – 480 (1992). 185. J. Marini, B. S., G. Heavner, S. Rose, and G. P. Chrousos, Evaluation of growth hormone axis and responsiveness to growth stimulation of short childern with osteogenesis imperfecta. Am. J. Med. Genet. 45, 261– 264 (1993). 186. K. Morley, M. Wilson, S. Likitmaskul, J. Ault, C. Cowell, H. Giles, and D. Sillence, The effect of somatotropin on bone density in osteogenesis imperfecta. 187. J. Marini, S. Bordenick, G. Heavner, S. Rose R. Hintz, R. Rosenfeld, and G.P. Chrousos, The growth hormone and somatomedin axis in short childern with osteogenesis imperfecta. J. Clin. Endocrinol. Metab. 76, 251 – 256 (1993b). 188. F. Antoniazzi, F. Bertoldo, M. Mottes, M. Valli, S. Sirpresi, G. Zamboni, R. Valentini, and L. Tato, Growth hormone treatment in osteogenesis imperfecta with quantitative defect of type I collagen synthesis. J. Pediatr. 129, 432 – 439 (1996). 189. N. E. Vieira, R. E. Goans, G. H. weiss, E. Hopkins, J. C. Marini, and Yergey, A. L. Calcim kinetics in children with osteogenesis imperfecta type III and IV: Pre- and post-growth hormone therapy. Calcif. Tissue Int. 67(2), 97 – 100 (2000). 190. C. Brumsen, N. A. Hamdy, and S. E. Papapoulos, Long-term effects of bisphosphonates of the growing skeleton. Studies of young patients with severe osteoporosis. Medicine (Baltimore). 76(4), 266 – 283 (1997). 191. F. H. Glorieux, Bisphosphonate therapy for severe osteogenesis imperfecta. J. Pediatr. Endocrinol. Metab. 13 (Suppl 2), 989 – 992 (2000). 192. J. R. Shapiro, E. F. McCarthy, K. Rossiter, K. Ernest, R. Gelman, N. Fedarko, J. Parry, and M. Bober, The effect of intravenous pamidronate on bone mineral density, bone histomorphometry and parameters of bone turnover in adults with type IA osteogenesis imperfecta (submitted). 193. F. H. Glorieux, N. J. Bishop, H. Plotkin, G. Chabot, G. Lanoue, and R. Travers, Cyclic administration of pamidronate in childern with severe osteogenesis imperfecta. N. Engl. J. Med. 339, 947 – 952 (1998). 194. H. Plotkin, F. Rauch, N. J. Bishop, K. Montpetit, J. Ruck-Gibis, R. Travers, and F. H. Glorieux, Pamidronate treatent of severe osteogenesis imperfecta in children under 3 years of age. J. Clin. Endocrinol. Metab. 85, 1846 – 1850 (2000). 195. B. Bembi, A. Parma, M. Bottega, S. Ceschel, M. Zanatta, C. Martini, and G. Ciana, Intravenous pamidronate treatment in osteogenesis imperfecta. J. Pediatr. 131(4), 622 – 625 (1997). 196. I. Fujiwara, E. Ogawa, Y. Igarashi, M. Ohba, and A. Asanuma, Intravenous pamidronate treatment in osteogenesis imperfecta. Eur. J. Pediatr. 157, 261 – 262 (1998). 197. E. Astrom and S. Soderhall, Beneficial effect of bisphosphonate during five years of treatment of severe osteogenesis imperfecta. Acta Paediatr. 87, 64 – 68 (1998). 198. H. Binder et al., Comprehensive rehabilitation of the child with osteogenesis imperfecta. Am. J. Med. Genet. 45, 265 – 269 (1993). 199. H. Binder, A. Conway, and L. Gerber, Rehabilitation approaches to childern with osteogenesis imperfecta: A ten-year experience. Arch. Phys. Med. Rehab. (1993). 200. W. Cole, Early Surgical management of severe forms of osteogenesis imperfecta. Am. J. Med. Genet. 45, 270 – 274 (1993).

JAY R. SHAPIRO 201. P. Brunelli and P. Novati, Complications of elongating intramedullary rodding in osteogenesis imperfecta. Am. J. Med. Genet. 45, 275 (1993). 202. S. J. Lyhmann, J. J. Sheridan, A. M. Capelli, and P. L. Schoenecker. Management of lower-extremity deformities in osteogenesis imperfecta with extensible intramedullary rod technique: A 20-year experience. J. Pediatr. Orthop. 18, 88 – 94 (1998). 203. Deleted at proof. 204. J. King and W. Bobechko, J. Bone Jt. Surg. 53(B), 72 – 85 (1971). 205. P. Papagelopoulos and B. Morrey, Hip and knee replacement in osteogenesis imperfecta. J. Bone Jt. Surg. 75(A), 572 – 5807 (1993). 206. F. Skovby, Inborn errors of metabolism causing homocysteinemia and related vascular involvement. Haemostasis 19 (Suppl. 1), 4 – 9 (1989). 207. S. Mudd, F. Skovby, and H. Levy, The natural history of homocystinuria due to cystathionine beta-synthase deficiency. Am. J. Hum. Genet. 37, 1 – 31 (1985). 208. S. H. Mudd, H. L. Levy, and F. Skovby, Disorders of transsulfuration. In “The Metabolic Basis of Inherited Disease” (C. Scriver, A. Beaudet, W. Sly, and D. Valle, eds.), Vol. 6, pp. 1279 – 1328, McGraw-Hill, New York, 1989. 209. S. Mudd, Homocystinuria and homocysteine: Selected aspect. In “Heritable Disorders of Amino Acid Metabolism” (W. Nyhan, Ed.), p. 429. Wiley, New York, 1974. 210. C. Morreels, B. Fletcher, R. Weilbaecher, and F. Dorst, Radiology 90, 1150 – 1158 (1953). 211. D. Brenton, J. Clin. Invest. 52, 488 (1977). 212. A. Kang and R. Trelstad, A collagen defect in homocystinuria. J. Clin. Invest. 52, 2571 – 2578 (1973). 213. D. Wilcken and N. Dudman, Mechanisms of thrombogenesis and accelerated atherogenesis in homocysteinaemia. Haemostasis 19 (Suppl. 1), 14 – 23 (1989). 214. R. Pyeritz and V. McKusick, The Marfan syndrome: Diagnosis and management. N. Engl. J. Med. 300, 772 – 777 (1979). 215. D. Milewicz, R. Pyritz, E. Crawford, and P. Byers, Marfan syndrome: Defective synthesis, secretion, and extracellular matrix formation of fibrillin by cultured dermal fibroblasts. J. Clin. Invest. 89, 79–86 (1992). 216. L. Kholmeier, C. Gasner, and R. Marcus, Bone mineral status of women with Marfan syndrome. Am. J. Med. 95, 568 – 572 (1993). 217. J. Gray et al., Osteoporosis and the Marfan syndrome. Postgrad. Med. J. 69, 373 – 375 (1993). 218. L. Pulkkinen et al., Deficient expression of the gene coding for decorin in a lethal form of Marfan syndrome. J. Biol. Chem. 265, 17780-17785 (1990). 219. M. Raghunath, A. Superti-Furga, M. Godfrey, and B. Steinman, Decreased extracellular deposition of fibrillin and decorin in neonatal Marfan syndrome fibroblasts. Hum. Genet. 90, 511 – 515 (1993). 220. D. Bonneau et al., Recurrent ctb (7) (q31.3) and possible laminin involvement in a neonatal cutis laxa with a Marfan phenotype. Hum. Genet. 87, 317 – 319 (1991). 221. C. Phillips, A. Shrago-Howe, S. Pinell, and R. Wenstrup, A substitution at a non-glycine position in the triple-helical domain of proalpha2(I) collagen chains present in an individual with a variant of the Marfan syndrome. J. Clin. Invest. 86, 1723 – 1728 (1990). 222. P. Byers, Ehlers – Danlos syndrome: Recent advances and current understanding of the clinical and genetic heterogeneity. J. Invest. Dermatol. 103, 47S – 52S (1994). 223. P. Beighton, A. De Paepe, B. Steinmann, P. Tsipouras, R. J. Wenstrup, Ehlers – Danlos syndromes: Revised nosology, Villefranche, 1997. Ehlers – Danlos National Foundation (USA) and Ehlers-Danlos Support Group (UK). Am. J. Med. Genet. 77(1), 31–37 (1998). 224. H. Kuivaniemi, L. Peltonen, K. I. Kivirikko Type IX Ehlers-Danlos syndrome and Menkes syndrome: The decrease in lysyl oxidase

CHAPTER 50 Osteogenesis Imperfecta and Other Defects of Bone Development

225.

226.

227.

228.

229.

230.

231. 232.

233.

234. 235. 236.

237. 238. 239. 240. 241.

242.

243.

244.

activity is associated with a corresponding deficiency in the enzyme protein. Am. J. Hum. Genet. 37, 798 – 808 (1985). A. De Paepe, L. Nuytinck, I. Haysser, I. Anton-Lamporecht, and J. M. Naeyaert, Mutations in the COL5A1 gene are causal in the Ehlers – Danlos syndromes I and II. Am. J. Hum. Genet. 50, 547 – 514 (1997). L. Smith, U. Schwartz, J. Goldstein, and P. H. Byers, Mutations in the COL3A1 gene result in the Ehlers – Danlos syndrome type IV and alterations in the size and distribution of the major collagen fibrils of the dermis. J. Invest. Dermatol. 108; 241 – 247 (1997). C. Giunta, A. Superti-Furga, S. Spranger, W. G. Cole, and B. Steinmann, Ehlers – Danlos syndrome type VII: Clinical features and molecular defects. J. Bone. J. Surg. (Am) 81, 225-238 (1999). B. V. Nusgens, T. Hermanns-Le, A. DePaepe, L. Nuytinck, G. E. Pierard, and C. M. Lapiere, Evidence for a relationship between Ehlers-Danlos syndrome type VIIIC in humans and bovine dermatosparaxix Nat. Genet. 1, 212 – (1992). H. N. Yeowell and L. C. Walker, Mutations in the lysyl hydroxylase 1 gene that result in enzyme deficiency and the clinical phenotype of ehlers-danlos syndrome type VI. Mol. Genet. Metab 71, 212 – 224 (2000). Kuivaniemi, L. Peltonen, and K. Kivirikko, Type IX Ehlers – Danlos syndrome and Menkes syndrome: The decrease in lysyloxidase activity is associated with a corresponding deficiency in the enzyme protein. Am. J. Hum. Genet. 53, 798 – 804 (1985). P. Coelho, R. Santos, and J. Melo Gomes, Osteoporosis and Ehlers – Danlos syndrome. Ann. Rheum. Dis. 212 – 213 (1994). M. Norman, In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. Favus ed.), pp. 245 – 248. Raven, New York, 1993. G. Exner, A. Prader, U. Elasser, and M. Anliker, Idiopathic osteoporosis in a three-year-old girl. Follow-up over a period of 6 years by computed tomography bone densitometry (CT). Helv. Pediatr. Acta. 39, 517 – 528 (1984). C. Dent and M. Friedman, Pregnancy and idiopathic osteoporosis. Q. J. Med. 34, 177 – 210 (1965). J. Jowsey and K. Johnson, Juvenile osteoporosis: Bone findings in seven patients. J. Pediatr. 81, 511 – 517 (1972). M. Cloutier, A. Hayles, B. Riggs, J. Jowsey, and W. Bickel, Juvenile osteoporisis: Report of a case including a description of some metabolic and microradiographic studies. Pediatrics 40, 649 – 655 (1967). R. Smith, J. Bone Jt. Surg. 62(B), 417 – 427 (1980). R. Evans, Bone metabolism in idiopathic juvenile osteoporosis: A case report. C. Dunstan, and E. Hills, Calcif. Tissue Int. 35, 5 – 8 (1983). E. Jones and R. Hensinger, Spinal deformity in idiopathic juvenile osteoporosis. Spine 6, 1 – 4 (1981b). A. Pocock, M. Francis, and R. Smith, “Fifth International Conference on Osteogenesis Imperfecta,” Oxford, 1993. P. A. Dawson, T. E. Kelly, and J. C. Marini, Extension of phenotype associated with structural mutation in type I collagen: Siblings with juvenile osteopofosis have an alpha2(I) Gly436 – Arg substitution. J. Bone Miner. Res. 14, 449 – 455, 1999. K. Hoekman, S. E. Papapouhos, A. B. C. Peters, and O. L. Bijvoet, Characteristics and bisphosphonate treatment of a patient with juvenile osteoporosis. J. Clin. Endo. Metab. 61, 952 – 956 (1985). L. D. Spotila, J. Caminis, M. Devoto, K. Shimoya, L. Sereda, J. Ott, M. P. Whyte, A. Tenenhouse, D. J. Prockop and Osteopenia in 37 members of seven families: Analysis based on a model of dominant inheritance. Mol. Med. 3, 313 – 324 (p382). E. S. Kurland, F. K. Chan, C. J. Rosen, J. P. Bilezikian, Normal growth hormone secretory reserve in men with idiopathic osteporo-

245.

246.

247.

248.

249. 250. 251. 252.

253.

254.

255. 256. 257.

258.

259.

260.

261.

262.

263.

301

sis and reduced circulating levels of insulin-like growth factor-I J. Clin. Endo. Metab. 83, 2576-2579 (1998). J. Halse, K. Nordal, A. Attramadal, and E. Dahl, [Idiopathic osteoporosis in middle-aged men — A new disease]? Tidskr. Norske Laegeforen. 114, 439 – 442 (1994). B. Wonke, C. Jensen, J. J. Hanslip, E. Prescott, M. Lalloz, M. Layton, S. Erten, S. Tuck, J. E. Agnew, J. E. Raja, K. Davies, A. Hoffbrand, Genetic and acquired predisposing factors and treatment of osteoporosis in thalassemia major. J. Pediatr. Endocrinol. Metab. 11,, 795 – 801 (1998). L. Rioja, R. Girot, M. Garabedian, and G. Cournot-Witmer, Bone disease in children with homozygous beta-thalassemia. Bone Miner. 8, 69 – 86 (1990). M. L. Anapliotou, I. T. Kastanias, P. Psara, E. A. Evangelou, M. Liparaki, and P. Dimitriou. The contribution of hypogonadism to the development of osteoporosis in thalassemia major: New therapeutic approaches. Clin. Endocrinol. 42, 279 – 287 (1995) N. Greep, A. L. Andersen, and J. Gallagher, Thalassemia minor: A risk factor for osteoporosis. Bone Miner. 16, 63 – 72 (1992). J. Kalef-Ezra et al., Bone minerals in beta-thalassemia minor. Bone 16, 651 – 655 (1995). J. Goerss et al., Risk of fractures in patients with pernicious anemia. J. Bone Miner. Res. 7, 573 – 579 (1992). M. Melton and M. Kochman, Reversal of severe osteoporosis with vitamin B12 and etidronate therapy in a patient with pernicious anemia. Metabolism 43, 468 – 469 (1994). R. Carmel, K.-H. Lau, D. Baylink, S. Sexena, and S. Singer, Cobalamin and osteoblast-specific proteins. N. Engl. J. Med. 319, 70 – 75 (1988). G. S. Kim, C. H. Kim, J. Y. Park, K. U. Lee, and C. S. Park, Effects of vitamin B12 on cell proliferation and cellular alkaline phosphatase activity in human bone marrow stromal osteoprogenitor cells and UMR 106 osteoblastic cells. Metabolism 45, 1443 – 1446 (1996). R. Recker, Calcium absorption and achlorhydria. N. Engl. J. Med. 313, 70 – 73 (1985). K. Eyres et al., Osteoporotic fractures: An unusual presentation of haemochromatosis. Bone 13, 431 – 433 (1992). T. Diamond, D. Stiel, and S. Posen, Osteoporosis in hemochromatosis: Iron excess, gonadal deficiency, or other factors? Ann. Intern. Med. 110, 430 – 436 (1989). S. G. Kaler, Metabolic and molecular bases of Menkes disease and occipital horn syndrome. Pediatr. Dev. Pathol. 1, 85 – 98 (1998). C. D. Yee, K. S. Kubena, M. Walker, T. H. Champeny, and H. W. Sampson, The relationship of nutritional copper to the development of postmenopausal osteoporosis in rats Biol. Trace Elem. Res. 48, 1 – 11 (1995). E. R. Barbosa, M. Scaff, L. R. Comerlatti, and H. M. Canelas, Hepatolenticular degeneration: Critical evaluation of the diagnostic criteria in 95 cases. Arq Neuropsiquiatr. 43, 234 – 242 (1995). Y .Z. Xie, X. Z. Zhang, X. H. Xu, Z. X. Zhang, and Y. K. Feng, Radiologic study of 42 cases of Wilson disease. Skeletal Radiol 13, 114 – 119 (1985). G. Stavrakakis, M. Spengos, and S. Scarpalezos. Skeletal mass conversions in hepatolenticular degeneration. Neuroradiolgy 10, 169 – 172 (1975). A. Baker, E. Turley, M. P. Bonham, J. M. O’Connor, J. J. Strain, A. Flynn, and K. D. Cashman, No effect of copper supplementation on biochemical markers of bone metabolism in healthy adults 1999. Br. J. Nutr. 82, 283 – 290.

CHAPTER 51

Osteoporosis Secondary to Illnesses and Medications ADINA SCHNEIDER ELIZABETH SHANE

Mount Sinai Hospital, New York, New York 10029 College of Physicians and Surgeons of Columbia University, New York, New York 10032

I. Introduction II. Hematologic Disorders III. Metabolic Disorders

IV. Medications References

I. INTRODUCTION

II. HEMATOLOGIC DISORDERS A. Multiple Myeloma

Primary osteoporosis is a condition characterized by increased skeletal fragility that develops in association with normal processes of menopause and advancing age. Secondary osteoporosis may be caused by specific clinical disorders, including a variety of endocrinopathies and genetic diseases, that cause low bone mineral density, either by interfering with attainment of peak bone mass or by increasing rates of involutional bone loss. In addition, many drugs are associated with alterations in bone remodeling that may lead to loss of bone mineral (Table 1). Bone loss resulting from these diseases and drugs is superimposed upon the primary processes that cause osteoporosis, exacerbating normal bone loss in affected individuals. In this chapter, certain disorders and drugs that are frequently associated with osteoporosis will be considered. Those addressed here will be indicated in boldface type in Table 1, whereas the appropriate chapter in which they are considered will accompany those covered elsewhere in this book.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Bony destruction is a common clinical manifestation of multiple myeloma [1]. Osteolytic lesions and pathologic fractures, the most frequent abnormalities observed on plain radiographs, are often accompanied by skeletal demineralization. However, generalized osteopenia alone may be the presenting feature of multiple myeloma, thus making this disease part of the differential diagnosis of patients with postmenopausal or other forms of osteoporosis. Moreover, since the pathologic fractures commonly involve the vertebral bodies, multiple myeloma must also be considered in the evaluation of patients with acute vertebral crush fractures. Myeloma bone disease develops in response to local production of cytokines by abnormal plasma cells that stimulate osteoclast-mediated bone resorption [2]. These cytokines, previously referred to as osteoclast-activating factors (OAFs) [3], are now known to include tumor necrosis factor , lymphotoxin, and interleukins 1 and 6 [4 – 7].

303

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

304

SCHNEIDER AND SHANE

TABLE 1

Secondary Causes of Osteoporosis

Genetic disorders

Chapter 50

TABLE 1 Hemophilia

Ehlers – Danlos

Multiple myeloma

Glycogen storage diseases

Leukemias and lymphomas

Gaucher’s disease

Systemic mastocytosis

Hemochromatosis

Rheumatologic diseases

Homocystinuria

Chapter 54

Ankylosing spondylitis

Hypophosphatasia

Rheumatoid arthritis

Marfan’s syndrome

Nutritional deficiencies

Menkes steely hair syndrome

Calcium

Osteogenesis imperfecta

Magnesium

Porphyria

Chapter 27

Vitamin D

Riley – Day syndrome Hypogonadal states

(continued )

Drugs Chapter 29

Anticoagulants (heparin and warfarin)

Androgen insensitivity

Anticonvulsants

Anorexia nervosa/bulemia

Cyclosporines and tacrolimus

Athletic amenorrhea

Cytotoxic drugs

Hyperprolactinemia

Glucocorticoids (and ACTH)

Chapter 44

Panhypopituitarism

Gonadotropin-releasing hormone agonists

Chapter 29

Premature menopause

Lithium

Turner’s and Kleinfelter’s syndrome

Methotrexate

Endocrine disorders

Thyroxine

Acromegaly

Chapter 47

Miscellaneous

Adrenal insufficiency Cushing’s syndrome

Chapter 52

Alcoholism Chapter 44

Diabetes mellitus

Chapter 31

Amyloidosis Chronic metabolic acidosis

Hyperparathyroidism (1° and 2°)

Chapter 49

Congestive heart failure

Chapter 52

Thyroid disease

Chapter 47

Cystic fibrosis

Chapter 43

Chapter 48

Emphysema

Gastrointestinal diseases Gastrectomy

End stage renal disease

Inflammatory bowel disease

Idiopathic hypercalciuria

Malabsorption

Idiopathic scoliosis

Celiac disease

Immobilization

Primary biliary cirrhosis

Multiple sclerosis

Hematologic disorders Sickle cell disease Thalassemia

They act at all phases of osteoclast development and differentiation to stimulate bone resorption [8 – 10]. In vitro, these cytokines also have been shown to inhibit bone formation [7]. Their effect to uncouple resorption from formation causes osteolytic lesions and progressive demineralization. Histomorphometric analysis of bone biopsies from patients with multiple myeloma has demonstrated increased numbers of osteoclasts and decreased numbers of osteoblasts in areas adjacent to plasma cells [11]. There is also evidence that osteoclasts produce interleukin 6 (IL-6) which stimulates growth of myeloma cells, thereby leading

Organ transplantation

Chapter 46 Chapter 52

Parenteral nutrition Sarcoidosis

to a vicious cycle of myeloma cell growth and increased bone resorption [12]. Other factors that may contribute to the development of osteopenia in patients with multiple myeloma include prednisone therapy which may further inhibit osteoblast function and may also exacerbate the hypogonadism that frequently accompanies severe systemic illness, radiation, and chemotherapy. Immobilization, secondary to the bone pain of multiple myeloma, may also accelerate bone loss. The predominant symptoms of myeloma bone disease are bone pain and pathologic fractures. The bone pain,

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

305

which may be intractable, most frequently involves the lower back and ribs, is exacerbated by movement, and is the presenting complaint in more than 80% of patients [1,13,14]. Pathologic fractures commonly involve bones with a high marrow content, such as the vertebral bodies, sternum, and ribs. In addition, hypercalcemia, related in part to renal insufficiency that impairs urinary calcium excretion, may develop in 20 to 40% of patients during the course of their illness [3,10]. When present, hypercalcemia is associated with low concentrations of 1,25-dihydroxyvitamin D [15] and parathyroid hormone (PTH) as measured by assays that detect only the intact PTH molecule [16]. Classic laboratory features of myeloma bone disease helpful in distinguishing it from other forms of osteoporosis include anemia, elevation of the erythrocyte sedimentation rate, abnormal serum protein electrophoresis, and Bence Jones proteinuria [1]. However, since approximately 10% of myelomas are nonsecretory, these features may be absent. In such cases, definitive diagnosis requires bone marrow biopsy. There are also subtle biochemical hallmarks of osteoporosis due to multiple myeloma that reflect the distinctive pathogenesis of the bony destruction. Increased osteoclastmediated bone resorption is associated with elevated resorption markers such as urinary pyridinium cross-link and N-telopeptide excretion [17], and serum immunoreactive bone sialoprotein [18]. In contrast, serum alkaline phosphatase activity and osteocalcin concentrations, markers of osteoblast activity, may be normal or suppressed [19]. A variety of radiographic techniques may be useful in the evaluation of patients with osteoporosis due to multiple myeloma. Plain radiographs may reveal multiple circumscribed lytic lesions distributed throughout the calvarium, pelvis, proximal long bones, clavicles, scapulae, sternum, and spine. Generalized osteopenia may accompany these typical osteolytic lesions, but may also occur in their absence in approximately 10% of patients [1,13]. Bone scans, which depend upon osteoblast function, are usually negative, despite extensive bony involvement. Vertebral collapse due to osteoporosis or metastatic disease may be distinguished from that due to multiple myeloma by the presence of a paraspinal mass as well as the fact that the intervertebral disc and vertebral pedicles are rarely involved in the former [1,13,20]. Quantitative computerized tomography (QCT) and magnetic resonance imaging (MRI) may have greater sensitivity than plain radiographs in the evaluation of focal lytic lesions [21,22]. MRI findings such as diffuse involvement, posterior epidural mass, and lack of cortical destruction are suggestive of hematologic malignancies rather than metastases [23]. In addition, MRI scanning of the spine may reveal marrow inhomogeneity of the involved vertebral bodies and apparent bright signal from the intervertebral discs when compared to scans from patients with vertebral collapse due to osteoporosis (Fig. 1). On a cautionary note,

FIGURE 1 Magnetic resonance imaging (MRI) of the spine. (A) A patient with a compression fracture due to postmenopausal osteoporosis. The T1 weighted spin echo sagittal view of the thoracolumbar spine demonstrates an anterior wedge compression fracture of L1. The marrow of the vertebral bodies is homogeneous and shows no evidence of replacement by tumor. The heterogenous fat distribution in the marrow of the lower thoracic vertebrae is the result of discogenic degeneration (spondylosis deformans). The intervertebral discs appear darker than the marrow, which is normal. (B) A patient with multiple compression deformities of the thoracolumbar spine due to multiple myeloma. The T1 weighted spin echo sagittal view of the spine demonstrates a heterogeneous appearance of the marrow of the vertebral bodies that is due to replacement by tumor. The marrow signal is darker than normal, giving the intervertebral discs a brighter appearance than the vertebral bodies, which is abnormal. however, a recent study of 224 vertebral compression fractures in 27 patients with multiple myeloma found that most pathologic fractures could not be distinguished from benign osteoporotic fractures by MRI [24]. MRI may also be useful in predicting risk of vertebral fractures in patients with multiple myeloma. A recent study identified four MRI patterns of bone marrow involvement: A, normal marrow; B, fewer than 10 focal lesions; C, more than 10 focal lesions; D, diffuse infiltration. Patients with patterns A and B had significantly longer fracture-free survival than patients with patterns C and D [25]. Measurements of bone mineral density (BMD) in patients with multiple myeloma, while not useful diagnostically, have confirmed the subjective impression of osteopenia observed on plain radiographs. In one study, the mean lumbar spine Z score (standard deviation from age-predicted mean) in a group of 10 patients with severe multiple myeloma was –2.69 and only 2 of the 10 patients had Z scores above 0 [26]. A much larger study of 168 patients also revealed significant reductions in lumbar and femoral

306 neck Z scores, while Z scores at the radial diaphysis were normal [27]. Therapy of osteoporosis in patients with multiple myeloma has focused upon relief of bone pain, prevention of bone loss or progressive lytic bone disease, and, to a lesser extent, restoration of bone mass and recalcification of osteolytic lesions. When discomfort is local and related to a particularly painful lesion, radiotherapy is often effective in reducing pain, although radiographic healing is unlikely. Chemotherapy with agents such as melphalan, aimed at reducing the tumor burden, may also be effective in reducing bone pain that involves the low back and ribs [28]. Moreover, in patients who received aggressive combination chemotherapy, total body irradiation and autologous marrow transplantation for severe myeloma, lumbar spine bone mass rose by 7.7% during the 12 months after treatment [26]. Glucocorticoids, a mainstay of therapy for multiple myeloma, have unpredictable effects on bone mass in patients with this disease. Their antitumor effect might be expected to improve myeloma bone disease while their well-known effects to decrease osteoblast function, enhance osteoclastic bone resorption, and decrease gonadal function, could be deleterious. Recently, bisphosphonates, have been shown to be of great value in the prevention and therapy of myeloma bone disease. Studies conducted with an early bisphosphonate, clodronate (dichloromethylene diphosphonate), suggested that this drug both inhibited bone resorption and reduced bone pain [29,30]. In contrast, a recent placebo-controlled trial of clodronate did not reveal a significant decrease in bone pain or fracture rates [31]. Pamidronate (3-amino-1hydroxypropylidene-1, 1-bisphosphonate) has also been reported to decrease bone pain in patients with multiple myeloma [32 – 34]. One of the most important advances in the care of patients with multiple myeloma has come from the work of Berenson and colleagues. In a randomized, double-blind trial, they determined that monthly infusions of pamidronate not only decreased bone pain but also reduced skeletal events in patients with advanced multiple myeloma [35,36]. Moreover, recent studies suggest that bisphosphonates may actually improve survival and decrease tumor burden in multiple myeloma. This is currently an area of active investigation (37). Newer and more potent oral and intravenous bisphosphonates, such as ibandronate and zoledronate, are under investigation and may provide alternatives in the future [38 – 40]. Calcitonin, another osteoclast inhibitor, has been shown to decrease biochemical indices of bone resorption in studies of short duration [41], but is probably relatively ineffective in comparison to bisphosphonates. Gallium nitrate blocks bone resorption due to IL-1, tumor necrosis factor, and parathyroid hormone in vitro [42,43]. This drug also stimulates type 1 collagen synthesis, increases the deposition of calcium into bone, and reduces the biochemical

SCHNEIDER AND SHANE

markers of bone resorption in patients with bone metastases. Subcutaneous administration of this drug to a small number of patients with multiple myeloma was associated with an increase in total body calcium and a decrease in vertebral fracture incidence and bone pain [44]. Although this drug shows promise in the management of myeloma bone disease [45], more and larger studies are needed.

B. Lymphoproliferative Disorders The lymphoproliferative disorders include a heterogeneous group of neoplasms in which there is malignant transformation of hematopoietic cells. Those that primarily originate in the bone marrow are classified as leukemias, while those that originate in extramedullary sites such as the lymph nodes are classified as lymphomas. Since leukemic cells proliferate in the bone marrow, they interfere with the processes of hematopoiesis and immunity. Bone pain and sternal tenderness occur in approximately half of patients with acute leukemia although osteolytic lesions are less frequent [46]. Whether because they occupy marrow space or because they produce factors that affect normal skeletal metabolism, the acute leukemias have also been associated with diffuse osteoporosis and fractures [47,48]. In contrast, skeletal involvement by lymphoma tends to be focal rather than diffuse [49]. While generalized osteoporosis as a presenting symptom of lymphoma has been documented [50], it is more often reported in association with leukemias, particularly acute leukemias of childhood and adolescence, than with either chronic leukemias or lymphomas. Skeletal abnormalities reported in children with acute lymphoblastic leukemia include both osteolytic and osteosclerotic lesions, generalized osteoporosis, vertebral compression fractures, as well as transverse radiolucent bands at the metaphyseal ends of long bones and periosteal new bone formation [51 – 58]. Children and adolescents may present with fractures at the time of diagnosis or may develop them after therapy is initiated [57]. The early work of Cohn et al. [55] suggested that the prevalence of symptomatic fractures and hypercalcemia at presentation in children with acute lymphoblastic leukemia (ALL) may be quite low (8 of 720 patients or approximately 1% in a retrospective study). However, a more recent prospective study of 40 children with ALL revealed that 10% of these children had radiologic evidence of osteopenia at presentation [59]. Hypercalciuria and decreased serum osteocalcin and 1,25-dihydroxyvitamin D concentrations were also present at diagnosis. The prevalence of osteopenia increased with time after diagnosis, in that 64% and 76% had evidence of osteopenia at 12 and 24 months, respectively [59 – 61]. Moreover, 39% of the study population developed fractures over the 2-year period of observation [60]. Another recent

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

307

prospective study of children with cancer, which included 27 patients with leukemia or lymphoma, reported similar findings. At diagnosis bone mineral density was normal, while markers of bone metabolism revealed evidence of increased destruction and decreased formation. Significant decreases in volumetric BMD were seen after 1 year of follow-up [62,63]. The prevalence of low bone mass is also high in cross-sectional studies of long-term survivors of ALL. One study reported that mean bone density measured 2.3 standard deviations below normal in 16 patients after therapy for leukemia; 11 of the patients had a history of bone pain, limp, or fracture [64]. Another study of 33 survivors of ALL found a significant decrease in BMD compared with 31 healthy sibling controls and 20 survivors of other malignancies [65]. The cause of the bone destruction may be related to the disease itself or to its therapy. The mechanism is still not completely understood but appears to be related to uncoupling of bone formation and resorption. In those patients with fractures at presentation, the osteoporosis may be due to the occupation of marrow by malignant cells or to boneresorbing cytokines produced by transformed lymphocytes. In those individuals who develop osteoporotic fractures after therapy, the use of chemotherapeutic agents and cranial radiation may directly and indirectly contribute to skeletal destruction. Virtually all regimens designed to achieve initial remission of acute leukemia include highdose glucocorticoids, which predispose to development of symptomatic vertebral fractures. In addition, most maintenance chemotherapeutic regimens include methotrexate, a drug for which there is considerable evidence for adverse skeletal effects (vide infra) [66]. Central nervous system (CNS) irradiation, prescribed in high doses as prophylaxis for leukemic meningitis, may constitute an independent risk factor for low bone mass in children treated for acute lymphoblastic leukemia [67]. The mechanism for the association between CNS irradiation and low bone density may be related to impaired growth hormone secretion [68]. A recent study of survivors of childhood ALL, observed a significant reduction in height that appeared to be related to decreased GH secretion, likely secondary to cranial irradiation [69]. Another study of 95 survivors of ALL, evaluated 11 years after diagnosis, found decreased bone mass secondary to reduced height for age, bone area for height, and bone mineral content for bone area. The reduced bone size was associated with cranial irradiation [70]. Growth is generally retarded during active therapy for acute leukemia [46]. Although catch-up growth generally occurs after remission and these children often achieve normal weight and height, there is certainly the potential for an adverse effect of these treatment regimens upon the achievement of peak bone mass. Another potential cause of low bone density is the hypogonadism commonly induced by radiation and chemotherapy.

Although recovery of gonadal hormone secretion may occur, any prolonged period of gonadal hormone deficiency, particularly during childhood and adolescence, may contribute to the loss of bone mass. It should be noted, however, that the aforementioned study of ALL survivors revealed preservation of gonadal function [69]. On a more hopeful note, a case of almost complete resolution of multiple vertebral compression deformities and resolution of dorsal kyphosis has been reported in a 9-year-old child treated successfully for acute leukemia [57]. There are few or no data available regarding the pathophysiology and the precise nature of the remodeling alterations that underlie the skeletal abnormalities observed in the acute leukemias. Similarly, published reports of prevention or therapy of bone loss and fractures in patients with leukemia are currently lacking. However, there is now ample evidence to support measurement of bone mineral density at diagnosis, during treatment, and after remission. The general principles relevant to the management of any individual with an unavoidable skeletal insult should apply in this clinical situation, with certain exceptions. Calcium and vitamin D supplements, commonly prescribed in patients with other secondary forms of osteoporosis, must be used with caution, if at all, in these individuals because of the potential for hypercalciuria and hypercalcemia. This precaution is of particular importance in patients with both Hodgkins and non-Hodgkins lymphoma, in whom unregulated production of 1,25-dihydroxyvitamin D by the malignant cells may be associated with enhanced gastrointestinal calcium absorption [71,72,73]. Rehabilitation therapy should be prescribed to maintain physical fitness and regular weight bearing exercise encouraged to prevent the deleterious effects of immobilization on the skeleton. Replacement of gonadal steroids should be considered in hypogonadal adults and adolescents. Drugs that inhibit bone resorption may be of potential efficacy in the therapy of patients with leukemias and lymphomas, as they are in multiple myeloma. There are currently very few data on the use of bisphosphonates in children or adults with lymphoproliferative disorders. However, an uncontrolled study of 12 children with idiopathic osteoporosis revealed that bisphosphonate use over a period of 2 to 8 years was safe, did not decrease linear growth, and was associated with normal lamellar structure on histomorphometric evaluation [74]. Clearly, well-designed clinical investigations of the biochemical and histomorphometric alterations that accompany leukemia- and lymphoma-associated osteoporosis are needed to permit the design of rational regimens for its prevention and therapy. Controlled trials of bisphosphonates in such children are urgently needed to prevent skeletal complications such as the 39% fracture rate reported in a cohort of children with ALL [60].

308

SCHNEIDER AND SHANE

C. Systemic Mastocytosis Systemic mastocytosis is an unusual condition characterized by abnormal proliferation of mast cells that infiltrate the skin, bone marrow, spleen, liver, and lymph nodes [72]. Mast cells are connective tissue cells, located most commonly at mucosal and cutaneous surfaces and in the bone marrow. They typically contain cytoplasmic granules that stain dark blue or violet with toluidine blue and pink or red with basic blue dye [75]. The cytoplasmic granules contain several potent chemical mediators including histamine, heparin, cytokines, serotonin, and bradykinin. Upon degranulation, prostaglandins of the D series are also released [76]. The clinical presentation of systemic mastocytosis is variable. This disease may masquerade as many different disorders [72]. The signs and symptoms of the disease are related to both the infiltration of various organs by mast cells and the release of the stored pharmacologic mediators upon degranulation which may cause fever, pruritis, flushing, diarrhea, weight loss, episodic hypotension and syncope, and peptic ulceration. The cutaneous form of the disease, which is caused by accumulation of mast cells in the skin, is the most common manifestation. The majority of patients (99%) present with the typical small, reddishbrown pigmented macules of urticaria pigmentosa. Urticaria, due to release of histamine, develop upon stroking of the skin. Approximately, 70% of patients with systemic mastocytosis have radiographically detectable abnormalities of bone. In the majority of patients the axial skeleton is predominantly affected. Poorly demarcated sclerotic and lucent lesions are the characteristic radiologic lesions and involvement may range from focal to generalized [77]. In a review of 58 patients with systemic mastocytosis, both diffuse sclerosis (10%) and generalized osteopenia (28%) were observed [72]. Recently, several cases have been reported in which osteopenia, with or without painful vertebral compression fractures, has been the sole manifestation of systemic mastocytosis, underscoring the importance of considering this disease in the differential diagnosis of osteoporosis [78 – 80]. It should be emphasized that the majority of these patients were men, rather than women, who are more commonly affected with osteoporosis. However, systemic mastocytosis may be underdiagnosed in women who are commonly assumed to have osteoporosis due to menopausal or age-related bone loss. The pathogenesis of the bony lesions of systemic mastocytosis is uncertain. However, mast cells secrete a number of bioactive substances that are known to affect bone metabolism [81]. In particular, heparin is associated with increased bone resorption in vitro and with the development of severe bone loss and fractures in patients on chronic long-term therapy [82,83]. An additional product of mast cell degranulation, prostaglandin D2, not only

weakly stimulates bone resorption but also causes the release of prostaglandin E2, a more potent bone resorbing agent [84]. However, mast cells also release chemical mediators such as histamine that stimulate formation of fi brous tissue and low concentrations of prostaglandin E2 may also increase bone formation [84]. Elevated excretion of a histamine metabolite has been shown to be associated with increased bone density, indicating that increased histamine release by mast cells may protect against osteoporosis [78]. Moreover, IL-6, which also stimulates bone resorption and fibrosis, is among the cytokines produced by mast cells [85]. Thus, it is postulated that the variable patterns of bony abnormalities observed in systemic mastocytosis may be due to complex interactions among these, as well as other, mast cell products. The diagnosis of systemic mastocytosis in patients who present with osteoporosis in the absence of typical skin lesions can present difficulties. A thorough history and complete physical examination are essential since a patient with only a single small lesion of urticaria pigmentosa has been reported [86]. The routine biochemical evaluation generally undertaken in the investigation of patients with osteoporosis may be unrevealing. However, several patients have been reported with mild elevations of serum osteocalcin concentration [80,86], increased alkaline phosphatase activity [87], and elevated urinary hydroxyproline excretion [88]. Measurement of plasma and urinary histamine may be helpful if increased; however, normal values do not exclude the diagnosis [72]. Bone mineral density (BMD) measurements may be normal [89], increased, or low [78,80,86,89]. Likewise, the radionuclide bone scan may be normal, may demonstrate unifocal or multifocal lesions, or may appear diffusely abnormal in affected patients [90]. A case report of a patient with known systemic mastocytosis who developed low back pain revealed normal plain radiographs of the spine, but an unusual mosaic pattern on MRI [91]. The diagnosis of systemic mastocytosis may be made on histologic grounds by either skin or bone marrow biopsy since the pathologic changes have been well characterized [72]. On a cautionary note, however, degranulation of mast cells secondary to the trauma of the biopsy procedure or to the decalcification process used during specimen preparation may prevent recognition of the singular staining characteristics of this cell and make recognition more difficult [92]. Therefore, iliac crest bone biopsy may be superior to bone marrow aspiration or biopsy because of the larger sample and the lack of a decalcification step in specimen processing [79,80,86 – 89]. Examination of undecalcified sections may reveal round, oval, or spindle-shaped cells containing numerous granules with typical staining characteristics of mast cells. The infiltration may be massive and diffuse or more focal and granulomatous in nature. Granulomas containing contain eosinophils, lymphocytes, and plasma cells are found

309

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

nine patients, who, despite increased static parameters of bone formation, had normal mineralizing surface and elevated mineral apposition rate, resulting in a normal bone formation rate [87]. Therapy of systemic mastocytosis is generally aimed at controlling symptoms. Antihistamines, H2 receptor blockers, and cromolyn sodium may be efficacious [72]. There are few published data on the therapy of osteoporosis due to systemic mastocytosis. Good responses to mast cell stabilizers such as ketotifen [96] and cromolyn sodium [97] have been reported in single case format. The histomorphometric data suggest that antiresorptive therapy with bisphosphonates or calcitonin may have greater efficacy than strategies designed to stimulate bone formation. In this regard, both dichloromethylene diphosphonate [98] and pamidronate [99,100] have been associated with clinical improvement in individual cases.

III. METABOLIC DISORDERS A. Diabetes

FIGURE 2

Mean number of mast cells per mm2 in bone from osteoporotic (O) and normal (N) women. Compared to normal women, women with postmenopausal osteoporosis have more marrow mast cells, but comparable numbers of cortical – endosteal mast cells. Reprinted with permission from Ref. 68.

adjacent to bony trabeculae or marrow vessels. Other marrow elements are usually present in normal numbers. Peritrabecular marrow fibrosis may also be present. Also of interest is a report of excessive marrow mast cells in women with postmenopausal osteoporosis [93 – 95] (Fig. 2). The relationship between these observations and the spectrum of systemic mastocytosis is unclear. Alterations in parameters of bone remodeling have also been characterized [79,80,86 – 89]. In general, it can be said that the histomorphometric picture is consistent with elevated bone remodeling. Cancellous bone volume may be normal, elevated, or decreased. Static indices of bone formation, osteoid volume and surface, and osteoblastic surface are consistently elevated and osteoid thickness is normal. Osteoclast surface and number are also increased. Dynamic indices of bone formation have been reported in two studies. De Gennes et al. described seven osteoporotic patients with systemic mastocytosis, all of whom manifested increased bone formation rate, mineralizing surface, and mineral apposition rate [86]. However, the quantity of bone deposited during each remodeling cycle as measured by the wall thickness was normal. Chines et al. reported

The association between diabetes mellitus and osteoporosis has received considerable attention. However, despite the large number of publications that have addressed this problem, a great deal of controversy remains regarding the prevalence of disturbed mineral metabolism, low bone mass, and fracture in diabetic patients. 1. BONE MINERAL DENSITY IN TYPE I DIABETES Low BMD has been found in the majority of investigations of patients with insulin-dependent (type I) diabetes mellitus [101 – 113]. The majority of these studies have utilized radiographic techniques such as radiogrammetry or single photon absorptiometry that focus primarily upon the appendicular skeleton [114]. In general, the degree of appendicular osteopenia has been modest, with bone density averaging 8% (range, 1 to 13%) or 0.82 standard deviations (range, 0.25 to 1.24) below that of controls [114]. Relatively few studies have examined axial BMD in insulin-dependent diabetics. In most, lumbar spine or proximal femur BMD is lower than in controls. In one group of 31 patients, mean age 41 years, with long-standing (18 years on average) diabetes, 38% had lumbar spine BMD measurements more than 10% below that of controls [115]. In another group of 41 adults with insulin-dependent diabetes, lumbar spine BMD did not differ between subjects and controls, but, femoral neck BMD was 10% and BMD of the distal lower limb was 12% below normal [116]. In a group of 48 subjects ranging in age from 5 to 20, in whom vertebral BMD was measured by QCT, trabecular BMD did not differ significantly from controls but cortical density was 3.5% lower [117]. In contrast, lumbar spine bone

310 density measured by dual-energy X-ray absorptiometry (DXA) was 5.4% higher in a group of 20 premenopausal diabetic women (mean age 32) than in a control population, and femoral neck BMD was normal [118]. This study, which also investigated certain biochemical indices of mineral metabolism, found significantly higher levels of serum alkaline phosphatase activity and urinary hydroxyproline excretion in the diabetic patients, suggesting that bone turnover was increased. There is considerable disagreement in the literature as to the effect of gender and duration of diabetes upon the prevalence of osteopenia in insulin-dependent diabetics. Several investigations have found gender-related differences in severity of bone loss. However, there is disagreement upon whether the degree of osteopenia is more severe in women [102,109,113] or in men [103,110]. Studies evaluating the relationship between duration of diabetes and bone mineral density have also been conflicting. Diabetic children often, but not always [117], demonstrate retarded bone age. Most studies conducted at or close to the time of presentation have found osteopenia [101,102,106,109,112]. However, a study of children with diabetes of less than 5 years duration revealed normal lumbar spine areal BMD values [119]. Other studies have suggested that BMD declines with increasing duration of diabetes [105,113]. These conflicting results are probably related to the predominance of cross-sectional studies, differences in study design, patient selection, relatively small sample size, and the use of different techniques for measurement of BMD. Relatively few studies have applied dual-energy X-ray absorptiometry and other state-of-the-art technologies for measuring bone mass. Few longitudinal studies of bone mass in insulin-dependent diabetic patients exist. One small study of seven patients studied twice, 11 years apart with photon absorptiometry of the forearm, revealed a clinically insignificant decline of less than 1% [120]. However, problems related to the reliability of data obtained with photon absorptiometry at such distant time points and the small sample size preclude generalizations based upon this study. 2. BONE MINERAL DENSITY IN TYPE II DIABETES Studies of patients with type II diabetes mellitus have yielded even more conflicting results; low, normal, and increased BMD measurements have been reported [101,121 – 132]. Again, vast differences in study design, bone mass measurement technology, and patient selection, particularly with regard to body weight and duration of disease, may account for some of these discrepant results. An Italian study of 110 type II diabetic patients reported that BMD was reduced at both the lumbar spine and the femoral neck compared to controls [126]. In contrast, a more recent study compared BMD in type I and II diabetics with controls. All subjects in this study had developed diabetes after

SCHNEIDER AND SHANE

age 30 and all were using insulin. Lower BMD values were found in type I than in type II diabetics and controls. Type II diabetic subjects did not differ from controls even after adjustment for BMI and age [133]. This study suggests that bone mass is low in type I and normal in type II diabetes. Insulin use and timing of disease onset (usually before achievement of peak bone mass in type I and after in type II) do not account for BMD differences between type I and type II patients. 3. FRACTURE Whether diabetics are at increased risk for most types of fractures is also controversial. It is clear that stress fractures of the tarsal and metatarsal bones are common in diabetic patients. In a study of 9704 women, 65 years of age or older, insulin-dependent diabetes was associated with an increased risk (RR 2.9) of foot fracture [134]. Such distal limb fractures may be of neuropathic or ischemic etiology, rather than due to generalized reduction of bone mass. Epidemiologic studies suggest that fractures of the femoral neck are also more frequent in diabetics [135 – 140], though none of these studies gave information on type or duration of diabetes. One pooled estimate suggests that there is approximately a twofold increased risk of fracture in diabetes [114]. Although a retrospective case – control study from the Mayo Clinic failed to show any increase in the incidence of fracture in either type I or type II diabetics [141], a more recent study reported that the relative risk of hip fracture is 6.9 and 1.8 in women with type I and type II diabetes, respectively [142]. Patients who undergo renal transplantation for diabetic nephropathy have a significantly higher fracture rate after transplantation than those transplanted for other causes [143] and fracture rates are extremely high (49%) in kidney – pancreas transplant recipients [144]. 4. PATHOGENESIS OF BONE LOSS IN DIABETICS The pathogenesis of bone loss associated with insulindependent diabetes mellitus is not well understood. Some studies have found a relationship between the severity of osteopenia and metabolic control of the diabetes [114,118], while others have not [113]. Poor diabetic control could have a negative impact upon bone mass by influencing biochemical indices of bone turnover. Diabetic patients with hyperglycemia, glucosuria, and higher hemoglobin A1C levels have been shown to have lower BMD, hypercalciuria, and biochemical evidence of secondary hyperparathyroidism [126]. These observations suggest that poorly controlled diabetes is associated with excessive renal calcium losses due to osmotic diuresis, resulting in compensatory increases in PTH secretion and loss of bone mineral. Improved control of diabetes has been associated with improvement in the biochemical parameters of bone turnover as well as in bone mass [126].

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

311

Recent studies are lending support to the notion that deficiencies of insulin itself and insulin-like growth factor I (IGF-I) may be important in the pathogenesis of low bone mass in diabetics. Insulin acts as a skeletal growth factor, stimulating amino acid uptake and collagen synthesis by bone cells [145,146]. A study of 5931 elderly patients (The Rotterdam Study) that related BMD values to insulin and glucose levels 2 h after an oral glucose load, found a direct correlation between higher glucose and insulin levels and higher BMD [147]. Thus the normal or elevated insulin levels found in type II diabetics may be anabolic for bone and protect against bone loss, while the lower levels seen in type I diabetes may predispose to bone loss by depressing bone formation. Insulin-like growth factor (IGF-I) is also an important regulator of bone formation. The actions of IGF-I are modulated by inhibitory (IGFBP-1 and -4) and stimulatory (IGFBP-3 and -5) binding proteins. A cross-sectional study of 52 type I and II diabetics found significantly lower levels of IGF-I and IGFBP-3 and higher levels of IGFBP-4 in type I diabetics than either type II diabetics or healthy controls [148]. Type I diabetics had significantly lower BMD of the hip and spine than type II subjects and IGFBP-1 was negatively correlated with BMD in the type I patients. In addition, serum proinsulin correlated positively with BMD in type II diabetics [148]. The incidence of insulin-dependent diabetes mellitus peaks in the second decade of life (around puberty), a time that corresponds to rapid skeletal growth and to the achievement of peak bone mass. A detrimental effect of insulin deficiency (or IGF-I deficiency) on bone formation at this critical juncture might be expected to interfere with attainment of normal adult bone mass, thus predisposing the insulin-dependent diabetic to osteopenia. In contrast, the peak incidence of non-insulin-dependent diabetes occurs after the fourth decade of life and after peak bone mass has been attained, making low BMD less likely. However, it should be noted that Touminen et al. controlled for onset of type I diabetes after adolescence and still demonstrated lower BMD in the type I patients [133]. Ninety percent of non-insulin-dependent diabetics are obese, which also may protect against osteoporosis. The different natural history and pathogenesis of insulin-dependent and non-insulin-dependent diabetes may account for the observed differences in bone mass. The heterogeneity of non-insulin-dependent diabetics with respect to age of onset, insulin levels, and obesity may account for the conflicting results of different studies of bone density in these patients.

[149]. Summarizing the results of many investigations, he has reported that urinary excretion of calcium, phosphate, and magnesium are increased, plasma concentrations of magnesium and ionized calcium are decreased, and those of PTH and 1,25-dihydroxyvitamin D are, for the most part, normal to low. It has been postulated that a relative hypoparathyroid state exists secondary to magnesium deficiency [112]. Again, however, there are discrepancies in the literature. Shao et al. documented elevated PTH concentrations in 11 patients with insulin-dependent and 19 patients with non-insulin-dependent diabetes mellitus [150], while Pietschmann found no difference between diabetic patients and controls [151,152]. Gregorio et al. also found evidence of hypercalciuria and increased circulating parathyroid hormone, but only in poorly controlled, non-insulin-dependent diabetic patients [126]. A study of 46 young type I diabetics revealed lower levels of 1,25-dihydroxyvitamin D and osteocalcin in those with persistent microalbuminuria than those without microalbuminuria [153]. Most histomorphometric studies have demonstrated low bone formation in diabetes mellitus [151,154,155] and these histological findings have been supported by reports of low osteocalcin levels in diabetic patients [41,152].

IN

5. MINERAL METABOLISM AND BONE TURNOVER DIABETES

Studies of biochemical indices of mineral metabolism in insulin-dependent diabetes have been reviewed by McNair

6. VITAMIN D Vitamin D metabolism may play a role in the pathogenesis of diabetes. The vitamin D system may be important in the regulation of insulin secretion. Calcitriol receptors have been found on pancreatic islet cells and vitamin D deficiency appears to be associated with impaired insulin secretion [156]. There are also clinical studies linking vitamin D deficiency and diabetes. A large population-based case – control study revealed a decreased risk of developing type I diabetes in patients who had received vitamin D supplementation [157]. A prospective study of 142 elderly Dutchmen found an inverse association between serum 25OH vitamin D and glucose concentrations following oral glucose tolerance testing [158]. This is an area that warrants further investigation. 7. GENETICS Bone mass has been shown to be largely determined by genetic factors. The vitamin D receptor is one of several genes linked to bone mineral density. Certain vitamin D receptor polymorphisms have also been shown to be associated with susceptibility to diabetes in some ethnic groups. A study of 93 Asian Indian families suggested that a polymorphism of the vitamin D receptor gene may modify susceptibility to type I diabetes in subjects living in South India [159]. In addition, genetic variations in the vitamin D binding protein have been shown to be associated with differences in oral glucose tolerance in nondiabetic Pima Indians [160]. The gene for type I collagen (COL1A1) is another likely regulator of bone mass. A study of 52 diabetics

312

SCHNEIDER AND SHANE

found an association between lower BMD at the femoral neck and the COL1A1 ‘s’ genotype [161]. 8. TREATMENT The general approach to the diabetic patient with low bone mineral density should encourage excellent glycemic control, adequate calcium and vitamin D, weight bearing exercise, and the avoidance of other potential risk factors for osteoporosis. If the patient is fracturing or if BMD is very low or decreasing, pharmacologic intervention should be considered. Antiresorptive agents have not been tested specifically in diabetic patients. If the etiology is low peak bone mass and bone formation is low, such drugs may theoretically be less effective. However, if renal function is normal, a trial of a bisphosphonate would be a reasonable approach.

B. Hemochromatosis Hemochromatosis, a disease associated with increased iron stores, may be classified into at least two types [162]. Familial or genetic hemochromatosis is an autosomal recessive disease associated with histocompatibility antigens HLA A3, B7, and B14. This disease is characterized by abnormal regulation of intestinal iron absorption such that excessive iron is absorbed. In the acquired form of hemochromatosis, iron overload arises as a consequence of another disease or its treatment, usually an iron loading anemia such as thalassemia or sideroblastic anemia or hypoproliferative anemias requiring multiple blood transfusions. The human body can excrete only limited amounts of iron (men, 1 mg/day; premenopausal women, 1.5 mg/day). Thus, long-standing, inappropriately high intestinal iron absorption or exposure to other sources of iron eventually leads to excessive iron stores. In both genetic and acquired forms of hemochromatosis, massive iron deposits accumulate in parenchymal tissues. Intracellular iron accumulation causes cellular dysfunction, necrosis, and fibrosis, most commonly involving the liver and pancreas, heart, and skin. In addition, there may be associated endocrinopathies. Hemochromatosis is 5 to 10 times more common in men than in women. The clinical features of hemochromatosis include hepatomegaly, which may be cirrhotic or precirrhotic, excessive skin pigmentation due to increased melanin and iron in the dermis and epidermis, diabetes mellitis, and congestive heart failure. Hypogonadotrophic hypogonadism due to iron deposition in the pituitary occurs commonly, causing decreased libido and testicular atrophy. Adrenal insufficiency, hypothyroidism, and hypoparathyroidism have also been described but are much less common. The association between hemochromatosis and osteoporosis was first described by Delbarre in 1960 [163]; since

that time, a few additional reports have appeared in the literature [164,165]. There are few data available regarding the prevalence of low bone density in hemochromatosis. In one small study, 10 of 22 men (45%) with genetic hemochromatosis had low BMD at either the distal one-third site of the forearm (by single-photon absorptiometry) or the lumbar spine (by single-energy QCT) and 4 men had vertebral fractures [164]. Another study reported low forearm BMD in 5 of 6 men with genetic hemochromatosis [166]. Czink et al. measured bone density at the radius, lumbar spine, and femoral neck in 5 patients with genetic hemochromatosis and 5 with transfusion-related hemochromatosis; bone density was decreased at one or more sites in the majority (9 of 10) of the patients [167]. These data suggest that osteoporosis may commonly accompany hemochromatosis. The pathogenesis of the osteoporosis is uncertain and is likely to be multifactorial. Hypogonadism and diabetes mellitus, both very common in hemochromatosis, are frequently associated with osteoporosis. While no study of pateints with hemochromatosis has controlled for diabetes, Diamond et al. clearly showed that hypogonadal men with this disease have lower forearm bone mass and trabecular bone volume than eugonadal men [164]. Another factor that could contribute to bone loss is associated liver disease. Severe hepatic fibrosis may be associated with decreased 25hydroxylase activity which in turn could lead to lower body stores of vitamin D, impaired intestinal calcium absorption, and defective bone mineralization. However, while low 25hydroxyvitamin D concentrations have been reported in some studies, histological osteomalacia has not been reported [164,166]. Yet another potential etiology for low bone mass in hemochromatosis is multiple hormonal defi ciency as has been recently reported in all of 17 children with thalassemia [168]. Delayed puberty, growth hormone deficiency, and primary hypothyroidism, the most common defects detected, may all interfere with attainment of peak bone mass and have lasting effects upon skeletal integrity. Only two studies have evaluated alterations in bone remodeling in hemochromatosis. Diamond et al. documented low cancellous bone volume and trabecular thickness and increased resorption surface and osteoclast number; however, these abnormalities were confined to those men who were also hypogonadal [164]. Urinary hydroxyproline excretion, a marker of bone resorption, was also significantly higher in the hypogonadal men. Although no patient had osteomalacia, indices of bone formation (osteoblast surface, mineralizing surface, adjusted apposition rate, bone formation rate, and mineralization lag time) were suggestive of impaired osteoblast function, particularly in hypogonadal and eugonadal untreated patients. These data were essentially in agreement with those reported by Conte et al. unfortunately, however, the latter authors did not give any information regarding the presence of hypogonadism in

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

313

their patients [166]. Thus, it would appear that the basic remodeling defect in hemochromatosis is an imbalance between increased bone resorption and impaired synthesis of osteoid. Phlebotomy is the cornerstone of therapy of genetic hemochromatosis while chelation of excess iron stores with deferrioxamine is the primary treatment modality in the acquired form of this disease [162]. With adequate treatment, most of the signs and symptoms of this disease will improve, with the unfortunate exception of the testicular atrophy. Thus, it is essential to evaluate and treat patients for hypogonadism. A recent study reported the effect of 24 months of therapy upon forearm and spinal bone mass in hypo- and eugonadal men with hemochromatosis [169]. Both groups of men were treated with regular phlebotomy and the hypogonadal men also received testosterone injections. Only testosterone therapy was associated with an increase in bone mass at both sites. Eugonadal men treated with phlebotomy alone sustained a decline in spinal bone mass while forearm bone mass remained stable.

Interestingly, there was no difference in BMD between active and controlled acromegalics [175]. A study of 23 female patients with acromegaly revealed increased spine BMD in menstruating patients but not in amenorrheic patients [176]. These data suggest that low bone density is particularly common in acromegalic patients who are also hypogonadal. The effects of GH on bone metabolism have been recently investigated. In studies of fetal rat bone in organ culture, GH stimulated IGF-I production, osteoblast progenitor proliferation, and bone formation [177]. IGF-I also stimulates synthesis of bone collagenous and noncollagenous proteins, growth of cartilage, and proliferation of chondrocytes and osteoblasts [178,179]. These experimental data would suggest that excess GH and IGF-I would increase bone metabolic activity. In this regard, two recent studies are of interest. Ezzat et al. measured serum and urine markers of bone and mineral metabolism in 27 untreated acromegalic patients [174]. While mean concentrations of serum calcium, intact parathyroid hormone, calcitonin, 25hydroxyvitamin D, and calcitriol were normal, 22% of the patients had elevated urinary calcium excretion. Biochemical markers of bone resorption were elevated and correlated with GH and IGF-I values. Serum osteocalcin concentrations were elevated in 50% of subjects, confirming the results of two previous studies [180,181]. Osteocalcin and the carboxyterminal peptide of type 1 procollagen (PICP), both considered to be markers of bone formation, were increased in 14 patients with untreated acromegaly [182]; there was a significant reduction in both during therapy with the longacting somatostatin analog, octreotide. Legovini et al. also observed a decrease in most serum markers of bone turnover in acromegalic patients treated with octreotide [183]. However, a mild increase in PTH concentration was also observed. Thus, there is biochemical evidence for both increased bone resorption and formation in adults with untreated acromegaly. Significantly higher concentrations of IGF-I, IGF-II, and IGFBP-5 have been found in the cortical bone of acromegalic patients than controls, a finding that is likely secondary to chronic excess of GH/IGF-I [184]. It remains unclear, however, whether the increased bone turnover is a primary effect of GH/IGF-I excess or whether hypogonadism also plays a role. Histomorphometric studies of patients with acromegaly are few [171,185 – 187]. Several investigators have described elevated cancellous bone volume and increased trabecular plate thickness in active acromegaly. A recent histomorphometric evaluation of the vertebral body of a 44-year-old hypogonadal woman with acromegaly and clinically evident osteoporosis and fractures revealed low cancellous bone volume compared to that of a control population [186]. There are no published data that address the effect of therapy for acromegaly upon bone mass. Replacement of

C. Acromegaly The effects on the adult skeleton of chronic, long-standing growth hormone (GH) excess are poorly understood. It is clear that during childhood, GH secretion stimulates linear bone growth via its effect upon insulin-like growth factor-I (IGF-I). However, the reported effects of GH excess due to acromegaly upon skeletal mass and bone metabolism in adults have been conflicting. Some of the inconsistencies in the literature can be accounted for by the common association of acromegaly and hypogonadism. In any case, an increased propensity to fracture has not been reported in acromegalic patients. While acromegaly is often included in lists of endocrinopathies associated with osteoporosis [170], some authors have reported normal or increased bone mass in this disorder. Studies performed in the 1970s, using techniques that measure predominently cortical bone, revealed increased forearm bone density [171,172]. More recent studies suggest that axial bone mass is normal in eugonadal and reduced in hypogonadal patients with acromegaly. For example, lumbar spine BMD was found to be normal in 7 acromegalic patients who had normal gonadal function [172]. Diamond et al. measured lumbar spine BMD by dual-photon absorptiometry in 12 hypogonadal and 7 eugonadal acromegalic patients and reported either reduced or normal values [173]. Ezzat et al. used QCT to measure vertebral BMD in 14 Caucasian patients with acromegaly [174]. One of 7 eugonadal patients had elevated BMD (Z score, 1.2). Low BMD was found in all of the 7 hypogonadal subjects, with Z scores ranging from -1 to -3. Lesse et al. also found lower BMD in hypogonadal acromegalics.

314

SCHNEIDER AND SHANE

gonadal steroids is likely to help and they should be prescribed when they are lacking and if the clinical situation permits. Theoretically, other agents that inhibit osteoclastic bone resorption also may be of use.

IV. MEDICATIONS A. Anticoagulants Heparin-induced osteoporosis was described in 1964 by Griffith and Silverglade [188]. These authors initially reported that 6 of 10 patients who had received between 15,000 and 30,000 units of heparin daily for 6 months or longer developed spontaneous vertebral or rib fractures. The following year, these investigators reported a larger series of 117 patients; osteoporosis developed only in those patients who had received more than 15,000 units daily and not in those who had received 10,000 units or less [189]. Also in 1965, Jaffe and Willis reported the occurrence of spontaneous rib and vertebral fractures in a 41-year-old man treated with 20,000 units of heparin daily for 6 months [190]. These early observations were strongly suggestive of a link between heparin therapy and osteoporosis. Today, long-term heparin therapy is limited primarily to patients with end-stage congestive heart failure who are awaiting transplantation and pregnant women with recurrent thromboembolic disease or prosthetic valves. Warfarin is contraindicated during pregnancy because of the teratogenic effect on the fetus during the first trimester and the potential for increased fetal wastage due to hemorrhage when given close to term [191]. In recent years the use of lowmolecular-weight heparin has grown considerably. Data on the long-term effects of low-molecular-weight heparin on bone are limited. 1. HEPARIN The subsequent literature on heparin-induced osteoporosis has been limited to a few case reports and several small series [192 – 197]. As previously noted by Griffith [189], the majority of fractures have occurred in patients exposed to a minimum dose of 15,000 units for 3 or more months [198]. The incidence of symtomatic vertebral fractures in heparin-treated patients is uncertain. However, a recent series of 184 women observed that symptomatic vertebral fractures developed in 2.2%, a rate consistent with the rate of 3% reported in another study [199,200]. However, one patient in that series fractured whose exposure was less than 2 months. In contrast, Rupp et al. have reported a much higher fracture rate of 24% in heparin-treated women [201]. Given the potential for fractures in heparin-treated patients, there has been interest in investigating the effect of heparin therapy on bone mass. The type of bone most

affected, the amount of bone loss, and the reversibility of any deficit have all been addressed. In a cross-sectional study of 61 premenopausal women exposed to high dose, long-term heparin therapy, there was no significant difference in mean bone density of the spine and radius compared to controls matched for age, parity, and time since last delivery. However, significantly more cases than controls had low bone mass at the lumbar spine and radius [202]. In 1982, Rupp et al. reported a series of 25 patients treated for recurrent thromboembolism with subcutaneous heparin for 105 weeks [201]. The drug was administered by continuous infusion and the dose was adjusted to achieve a plasma level of 0.2 units/ml. Although six patients (24%) developed vertebral fractures, forearm BMD measured by SPA did not change. Since that report, several more prospective studies have been published, all in pregnant women. In 1983, DeSwiet et al. reported a decrease in metacarpal and phalangeal cortical area in 20 women treated with 20,000 units of heparin daily for 6 to 32 weeks, a deficit that remained apparent 6 months after therapy was discontinued [203]. More recently, Barbour and colleagues used DXA to study changes in BMD at the proximal femur in 14 pregnant women who required heparin prophylaxis [204]. The doses of heparin were adjusted to achieve a level between 0.05 and 0.2 units/ml and ranged between 12,000 and 21,000 units daily. The heparin-treated group was compared to 14 normal pregnant women, matched for age and race. Proximal femur BMD was measured at the start of heparin therapy. Proximal femur and vertebral BMD were measured when heparin therapy was discontinued immediately postpartum and 6 months later. Between baseline and immediately postpartum, proximal femur bone mass decreased significantly by 5% in the patients, while there was no change in controls. Thirty-six percent of the patients had more than a 10% decline in bone mass. By 6 months postpartum, bone density remained significantly below baseline in the treated patients, although there was a trend toward recovery. Vertebral BMD did not change between the immediate and 6-month-postpartum measurement. These authors found no relationship between heparin dose and change in bone density. Dahlman et al. also prospectively evaluated bone density in 39 pregnant women at two forearm sites using SPA [205]. Measurements of the distal onethird (predominantly cortical bone) and the ultradistal (55% cancellous bone) sites were made at institution of heparin therapy and immediately and 7 weeks postpartum. The patients, 10 of whom had undergone previous treatment with heparin, did not differ at the outset from a matched group of normal pregnant women with respect to bone mass. The mean dose of heparin was 17,300 units daily. In the heparin-treated women, there was a significant 4.9% decline in BMD at the ultradistal site while the control group manifested a nonsignificant trend toward a 2.3% decline. Neither group of women demonstrated a significant change in bone

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

315

mass at the predominantly cortical site. One woman taking heparin had spinal crush fractures. There was a trend toward recovery of bone mass by the 7-week postpartum measurement in the patients, despite the fact that all women breast-fed and continued to take heparin at a somewhat reduced dose of 15,000 units daily. The reasons for the discrepancies among these various studies with respect to the type of bone affected (cortical versus cancellous), the reversibility of the bone loss, and the relationship of the bone loss to heparin dose are not immediately apparent. What seems clear, however, is that incidence of fracture is significant and that the clinical presentation is consistent with predominant involvement of sites rich in cancellous bone, such as the vertebral bodies. The balance of the evidence seems to favor the notion that there is at least some recovery of the lost bone. Moreover, it does not appear to be necessary to discourage breast feeding in women who have been treated with heparin during their pregnancies [205]. Little information is available in human subjects on the biochemical changes that accompany heparin induced bone loss or fractures. There is some experimental evidence in the Japanese quail for an inhibitory effect of heparin on renal 1  hydroxylase activity [206]. Mutoh et al. documented progressive bone loss accompanied by normal total and ionized serum calcium, elevated parathyroid hormone, and decreased 1,25-dihydroxyvitamin D concentrations in heparin-treated rats compared to those seen in controls [207]. The authors therefore suggested that perturbations in the parathyroid – vitamin D axis were involved in the pathogenesis of heparin-induced osteopenia. In this regard, Aarskog and colleagues demonstrate significantly lower plasma levels of 1,25-dihydroxyvitamin D in a group of 10 pregnant women treated with 15,000 units of heparin daily than in a group of 22 normal pregnant women [193,208]. Taken together, these data are consistent with the hypothesis that heparin therapy causes increased bone resorption indirectly by virtue of an effect to decrease renal 1,25-dihydroxyvitamin D production and increase parathyroid hormone secretion. However, these observations must be viewed in the context of both in vitro and in vivo studies that have shown that heparin directly stimulates bone resorption [209 – 211] and reduces bone formation [212,213]. It is likely that the skeletal effects of heparin are multiple and involve both stimulation of bone resorption and inhibition of bone formation, whether due to direct actions on bone cells through the osteoclastic 53 integrin receptor or indirect, hormonally mediated effects. Clearly, more studies are needed to clarify the mechanism(s) by which heparin administration induces demineralization. Even less information is available regarding the effect of heparin on bone histomorphometry in humans. There have been only two case reports describing histomorphometric analysis of transiliac crest bone biopsies after therapy with

heparin [196,214]. Both reports were in pregnant women who had suffered vertebral compression fractures during the course of their treatment. In the first report published by Meunier and colleagues in 1982 [196], a biopsy was obtained 1 month after delivery and revealed severely decreased cancellous bone volume and normal cortical parameters. Resorption surfaces and osteoclast number were markedly elevated while there was clear evidence of osteoblastic depression. Bone formation, as measured by tetracycline uptake, was normal in cortical bone, whereas it was extremely low in cancellous bone. In a second case reported by Zimran et al. in 1986 [214], the biopsy was obtained 6 weeks postpartum. In contrast to the first patient, cancellous bone volume was reasonably normal, parameters of bone resorption were decreased, osteoid surfaces were increased, and the bone formation rate was normal. The reasons for the discrepancies between these two patients are uncertain. Both biopsies were compared to normative data from nonpregnant women, so it remains uncertain whether the results reflected recovery from pregnancy. Therapy of heparin-induced osteoporosis is uncertain. No studies have been published. As in other forms of osteoporosis, prevention is likely to be more effective than therapy of established fractures. The lowest possible dose of heparin, preferably less than 15,000 units daily, should be used and adequate calcium and vitamin D intake should be ensured. A symptomatic clinical response to substitution of warfarin for heparin has been reported in patients with fractures [189]. However, in the context of the currently most common indication for prolonged heparin therapy today, namely the treatment or prophylaxis of thromboembolic episodes in pregnancy, this may not be possible. In this regard, some authors recommend using heparin for the first trimester and then changing to warfarin until the middle of the third trimester when heparin should be reinstituted until term [215,216]. This strategy may be particularly helpful for women with systemic lupus erythematosis and high levels of antiphospholipid antibodies, who require glucocorticoids as well as heparin. Therapy with antiresorptive drugs, such as bisphosphonates or calcitonin, may represent a reasonable approach in patients who are not pregnant. However, the decision to institute any pharmacologic treatment in a pregnant woman requires due consideration of the risk to benefit ratio. In this regard, there have been no trials of either calcitonin or bisphosphonates in pregnant women. The potential for adverse effects of bisphosphonates upon skeletal development in the fetus is considerable and their use cannot be recommended in this situation. 2. LOW-MOLECULAR-WEIGHT HEPARIN Low-molecular-weight heparin (LMWH) has become more widely used in recent years. However, evidence that it is less toxic to the skeleton is conflicting [217] and its longterm effects on bone are unclear. A histomorphometric

316

SCHNEIDER AND SHANE

comparison of the effects of heparin and low-molecularweight heparin in cancellous bone of rats revealed less osteopenia in the LMWH-treated rats. While both heparin and LMWH decreased osteoblast surface, only heparin increased osteoclast surface [218]. There are very few human data available on the effect of LMWH on bone. Of concern is a report of a 29-year-old woman who developed osteoporosis and vertebral fractures after 3 months of treatment with LMWH [219]. 3. WARFARIN There are no reported cases of osteoporosis complicating warfarin therapy. However, osteocalcin, a major noncollagenous vitamin K-dependent protein secreted by osteoblasts, is inhibited by warfarin [220 – 222]. Moreover, warfarintreated lambs have lower circulating osteocalcin concentration and higher fraction of noncarboxylated osteocalcin in serum and developed osteopenia after 3 months of therapy. These changes were associated with decreased cancellous bone area, eroded surface, and bone formation rate consistent with both decreased bone resorption and formation [223]. In addition, warfarin exposure during fetal development may produce skeletal abnormalities in humans [224]. However, self-reported warfarin use was not associated with lower bone density or higher fracture rates in more than 6000 postmenopausal women who participated in the Study of Osteoporotic Fractures [225]. Thus, although a possible effect of warfarin therapy on bone remodeling in humans cannot be excluded, there appears to be no increase in fracture rates among the most vulnerable of populations, namely elderly postmenopausal women.

B. Methotrexate Methotrexate is a member of the antimetabolite class of chemotherapeutic agents. An analog of folic acid that acts by competitive inhibition of folic acid reductase, methotrexate interferes with conversion of dihydrofolate to tetrahydrofolate and thus with normal cellular replication [226]. Methotrexate is used to treat a wide variety of childhood and adult malignancies including leukemias, lymphomas, osteogenic sarcoma, choriocarcinoma, and carcinomas of the breast, bladder, and pharynx [227]. It has been used for many years to treat severe psoriasis [227] and more recently to treat autoimmune disorders such as rheumatoid arthritis, polymyositis, systemic lupus erythematosus, polyarteritis nodosa, and Wegeners granulomatosis [227]. Initial reports linking osteoporosis to methotrexate therapy involved children treated for acute leukemia [66,228 – 230]. Similar reports have appeared in connection with children treated for osteogenic sarcoma [231] and lymphoma [229]. The syndrome consists of severe lower ex-

tremity bone pain, osteoporosis, and fractures [227,232] that primarily involve the lower extremities and may be multiple. Symptoms develop between 6 and 15 months after beginning methotrexate and are more frequently associated with long-term high-dose therapy. Radiographic features (Fig. 3) include thick dense provisional zones of calcification with growth arrest lines similar to the changes associated with scurvy [66,228,233]. In addition, nonunion or delayed union of fracture is common despite adequate immobilization; after methotrexate is discontinued, fractures usually heal [230]. Other authors have also noted clinical improvement after cessation of methotrexate [66,228]. The reported incidence of fractures in leukemic children treated with methotrexate has ranged from 12 to 45% [230,234]. However, subclinical disease may be much more common, since Stanisavljevic and Babcock reported low bone density in 20 of 37 children (54%). Similarly, Gnudi et al. measured forearm bone density in 59 patients treated for osteosarcoma with long-term high-dose (7500 mg/m2 or low-dose (750 mg/m2) methotrexate [235]. When compared to a control population, BMD was reduced at the distal forearm where cancellous bone predominates but not at the proximal forearm that consists of cortical bone. Moreover, the effect of methotrexate on forearm bone mass appeared to be limited to those patients exposed to the higher dose. Similarly, a study of 32 children with juvenile rheumatoid arthritis did not find an association between BMD and chronic low dose methotrexate use [236]. In contrast, Preston et al. reported 2 patients who developed stress fractures after receiving low weekly doses of methotrexate for treatment of psoriasis and inflammatory arthritis [237] and Singwe et al. described a patient on methotrexate for scleroderma who developed four stress fractures in a 13-month period (238). The pathogenesis of methotrexate-induced bone disease is uncertain. Methotrexate therapy does not impair growth in children with leukemia [239,240]. Moderat increases in urinary and fecal calcium loss were reported in eight patients on methotrexate for treatment of cancer [241]. However, detailed biochemical investigations of mineral metabolism, particularly utilizing currently available markers of bone turnover, are lacking. Histomorphic studies of methotrexate osteopathy are also few. Friedlander et al. studied the effects of short-term (5 days) high-dose methotrexate administration on histomorphometry of tail vertebrae in the rat [242]. They observed a 25% reduction in cancellous bone volume and a 60% decrease in bone formation in treated animals compared to controls. Similarly, May et al. found decrease in biochemical and histomorphometric parameters of bone formation and increased bone resorption in methotrexatetreated rats [243]. Preston et al. performed bone biopsies in two adults treated with methotrexate and observed changes consistent with osteoblast inhibition [237]. The mechanism

317

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

FIGURE 3 Methotrexate osteopathy. An 18-month-old boy with acute lymphocytic leukemia was treated with a variety of drugs, including methotrexate. Approximately 2 years later, these radiographs were obtained because the patient had pain and weakness in the legs. (A) Findings include osteopenia, periostitis, and fractures both in the diaphysis and in the metaphysis of the left femur (arrows). A growth recovery line is seen. (B) Observe osteopenia, growth recovery lines, and a fracture of the metaphysis of the right tibia. Although these fractures healed well, additional ones occurred subsequently. Reprinted with permission from Ref. 229. by which methotrexate impairs bone formation is unclear. Potential theories include imparied protein synthesis by osteoblasts. Radiographic changes resembling those of scurvy have raised the possibility that interference with the metabolism or actions of vitamin C, an essential cofactor for hydroxylation of proline and collagen synthesis, may be involved [232]. Specific therapy for this problem is lacking. In general, the lowest possible dose of methotrexate should be used for the shortest possible time. Calcium supplementation may be advisable in view of the potential for increased losses. Withdrawal of the drug is generally associated with symptomatic improvement within several weeks.

C. Anticonvulsant Drugs Metabolic bone disease may complicate the management of patients receiving several of the commonly used anticonvulsant drugs [191,233,244,245]. Diphenylhydantoin, phenobarbital, carbamazepine, and sodium valproate have all been implicated. Anticonvulsant bone disease occurs more frequently with long-term, high-dose, multidrug regimens (Table 2). Anticonvulsant-induced bone disease [233,244] may present as overt osteomalacia or rickets, with fractures, proximal muscle weakness, frank hypocal-

cemia, and hypophosphatemia. In the more common modern presentation [191,245], patients may present with an asymptomatic decrease in bone mineral density [246 – 249]. However, one recent study found a higher fracture rate among epilepsy patients than controls (RR 2.0); 34% of these fractures were related to seizures [250]. The majority of anticonvulsant drugs are thought to affect bone and mineral metabolism indirectly by causing abnormal metabolism of vitamin D. Diphenylhydantoin, phenobarbital, and carbamazepine stimulate hepatic mixed function oxidase activity [245,251] and therapy with these TABLE 2

Risk Factors for Anticonvulsant Bone Disease

1.

High-dose, multiple drug regimens

2.

Long-term therapy

3.

Low vitamin D intake

4.

Limited sunlight exposure

5.

Chronically ill, elderly or institutionalized patients

6.

Reduced physical activity levels

7.

Adjuvant therapy to induce chronic metabolic acidosis (acetazolamide or ketogenic diets)

8.

Concomitant therapy with drugs that induce hepatic enzymes (rifampin, glutethimide) Note. Adapted with permission from T. J. Hahn [245].

318 drugs is associated with increased degradation of steroid hormones. In contrast, sodium valproate does not affect vitamin D metabolism directly, but does cause renal toxicity which could indirectly affect bone and mineral metabolism. The effects of anticonvulsant drugs on serum 25-hydroxyvitamin D concentrations are unclear. Hahn et al. have shown that anticonvulsant drug administration is associated with increases in the disappearance of vitamin D and 25hydroxyvitamin D from the circulation and the appearance of inactive polar metabolites of vitamin D in the bile and urine [247]. In older studies conducted in the United States, serum concentration of 25-OHD were lower in long-standing epilepsy [251 – 253]. Similarly, a Finnish study found serum levels of 25-OHD to be lower in patients taking various combinations of diphenylhydantoin, phenobarbital, and carbamazepine [254]. However, two other studies found serum 25-OHD levels in patients on diphenylhydantoin to be no different from normal controls [255,256]. These discrepancies may be related to geographic differences in the study populations, since the studies with low 25-hydroxyvitamin D values were conducted in more northern latitudes where there is less sunlight exposure [245,254]. Thus, inadequate vitamin D stores may develop when taking these drugs, particularly when patients are from northern climates or institutionalized, and access to ultraviolet light may be diminished [191]. In addition to the actions of anticonvulsant drugs upon hepatic vitamin D metabolism, direct effects upon cellular metabolism have been described. Diphenylhydantoin, which interferes with cation transport in many tissues, directly inhibits intestinal calcium transport [257], and together with phenobarbital, has been shown to inhibit vitamin D-mediated calcium absorption [258]. Furthermore, Hahn and Halstead demonstrated that higher concentrations of vitamin D metabolites are required to maintain normal intestinal calcium absorption in patients taking anticonvulsant drugs [246]. In addition, both phenobarbital and diphenylhydantoin inhibit PTH-mediated bone resorption in vitro which could cause PTH resistance and secondary increases in PTH secretion [259,260]. Diphenylphydantoin, which appears to exert this effect by inhibition of hormonesensitive adenylate cyclase activity, is the more potent of the two drugs. Diphenylhydantoin also directly inhibits collagen synthesis and lysozomal enzyme release by bone explants [261] and cytochrome P-450 aromatase activity, which may lead to decreased circulating estrogen. A 17year-old epilpetic boy with severely reduced bone mass, who had been treated with multiple anticonvulsants, was found to have a serum estrogen concentration below 10 pg/ml. After 10 months of treatment with estrogen, he had a significant increase in bone density and bone age [262]. Two excellent histomorphometric studies by Melsen and Mosekilde [263] and Weinstein et al. [256] have contributed

SCHNEIDER AND SHANE

TABLE 3

Bone Histomorphometry in Patients Receiving Anticonvulsant Drugs

Static parameters

Patient (n  20)

Normal range

20.4  1.6

23.5  1.8

Static parameters Trabecular bone volume (%) Osteoid volume (%)

5.2  1.1*

2.1  0.7

Osteoid surface (%)

22.8  2.8†

11.0  2.8

Osteoid seam width (m)

11.7  0.9

9.9  1.5

Osteoblastic osteoid surface (%)

4.3  0.8

4.1  1.3

Osteoclast resorptive surface (%)

0.51  0.14

0.51  0.13

Calcification front (%)

85.2  4.3

80.3  4.2

Mineralization rate (m/day)

0.80  0.04‡

0.65  0.02

Mineralization lag time (days)

18.5  2.2*

23.5  2.5

Dynamic parameters

*P  0.05. † P  0.01. ‡ P  0.005. Note. Data are presented  SEM. Adapted with permission from Weinstein et al. [256].

in a major way to our current understanding of the skeletal effects of anticonvulsant drugs. Melsen and Mosekilde evaluated a group of 20 adults all treated with diphenylhydantoin for at least 10 years, some of whom were also taking other anticonvulsant drugs [263]. Compared to a group of normal controls, these patients had significantly reduced serum and urinary calcium levels and elevated serum total alkaline phosphatase activity. Analysis of transiliac crest bone biopsies revealed normal cancellous bone volume and increases in osteoid volume, active osteoid surface, mineralizing surface, and resorption surface. Although osteoid seams were slightly thicker than a control population, the mineral apposition rate was normal. Weinstein et al. [256] evaluated bone biopsies in 20 patients who had taken anticonvulsant drugs for 18 years (Table 3). They observed decreased serum ionized calcium activity and increased PTH concentrations and low cortical bone mineral density. Although cancellous bone volume was normal, cortical bone was thin and porous. Osteoid volume and surface were increased, osteoid width was normal, and the mineralization rate was normal or increased. Their findings were essentially in agreement with those of Melsen and Mosekilde [263], thus establishing anticonvulsant bone disease as a disorder of high remodeling rather than abnormal mineralization. Low 25-hydroxyvitamin D concentrations could impair intestinal calcium absorption and lead to secondary hyperparathyroidism, as well as the increased bone remodeling observed by Weinstein [256] and Mosekilde [263]. However, routine serum measurements of indices of bone and mineral metabolism are frequently unremarkable in patients taking anticonvulsant drugs. Biochemical evaluation may reveal

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

319

normal or minimally depressed serum calcium and phosphorus concentrations and serum total alkaline phosphatase activity may be normal or only slightly elevated. However, more sensitive biochemical techniques may reveal abnormalities in many patients in whom routine studies are normal [191,245]. Serum 25-hydroxyvitamin D concentrations are frequently reduced in anticonvulsant-treated patients and there may be mild elevations of PTH [246,254,256]. Serum 1,25-dihydroxyvitamin D concentrations may be normal, high, or low. Markers of bone formation such as osteocalcin [254,264,265], serum total [264,266], and bone-specific alkaline phosphatase [254,265] and procollagen carboxyterminal peptide [265] have been reported to be elevated in children and adults taking anticonvulsant drugs. With respect to markers of resorption, Ohishi et al. [267] documented that urinary hydroxyproline, pyridinoline, and deoxypyridinoline excretion were increased in 15 premenopausal women taking anticonvulsant drugs compared with 211 healthy age- and sex-matched controls. Elevated urinary excretion of hydroxyproline [264,266] and crosslinked carboxyterminal peptide of type I collagen [254] have also been reported in patients on antiepileptic drugs. Thus, the available data suggest that biochemical markers of bone turnover are elevated in patients on anticonvulsant drugs, consistent with the histomorphometric studies that have established this disorder as a state of increased bone remodeling [256,263]. Data on bone mineral density in children and adults taking anticonvulsant drugs are also conflicting. In a cross-sectional study of 226 free-living patients with epilepsy, radial bone mass was significantly reduced and, in 18% of the patients, was more than 2 SD below young normal controls [248]. Similarly, Hahn et al. found radial BMD to be 10% below normal in 22 children on anticonvulsant drugs [247]. A recent study of 44 epileptics, on medication for more than 5 years, found significantly decreased BMD (by DXA) at the hip and lumbar spine [268]. Valimaki et al. measured spine and hip bone density in 38 men and women between the ages of 20 and 49 who were taking diphenylhydantoin, carbamazepine, or both [254]. BMD was abnormal only in women and only at the hip. Sheth and colleagues evaluated both axial and appendicular bone mass by DXA in 26 children who had been taking sodium valproate or carbamezepine for more than 18 months [269]. While BMD was normal in patients on carbamezepine, those receiving sodium valproate had reduced BMD at both the lumbar spine and radius. In summary, bone turnover is frequently elevated in patients taking anticonvulsant drugs and bone mass is often but not always reduced. Risk factors for the development of anticonvulsant bone disease (Table 2) include high dose, multiple drug regimens, long duration of therapy, institutionalization, vitamin D deficiency due to either inadequate dietary intake or reduced sunlight exposure, physical inactivity, use

of ketogenic diets or acetazolamide to induce chronic metabolic acidosis, and concomitant therapy with other drugs that induce hepatic enzymes [245]. Anticonvulsant bone disease should be sought in individuals with a history of chronic long-term therapy and one or more of these risk factors. The routine biochemical evaluation may reveal hypocalcemia, hypophosphatemia, and increased alkaline phosphatase activity or may be normal. Urinary calcium excretion and serum 25-hydroxyvitamin D concentrations may be reduced and there may be mild elevations of serum PTH. Both serum osteocalcin concentrations and urinary pyridinium crosslink excretion may be increased. However, any or all of these biochemical variables may be normal and decreased bone mineral density may be the only manifestation. Alternatively, bone density may be normal in the face of clear-cut biochemical abnormalities. In difficult cases, bone biopsy may be helpful in establishing the diagnosis. Management of patients on chronic anticonvulsant drugs should include routine prophylaxis with a daily vitamin supplement that contains at least 400 IU of vitamin D. This is particularly important in the case of elderly or institutionalized patients. Although Barden et al. have demonstrated that this approach will prevent bone loss in institutionalized adults [270], Collins et al. have shown that higher doses (400 to 4000 IU) may be necessary to normalize serum levels in both institutionalized and noninstitutionalized patients [271]. The common side effects of vitamin D therapy — hypercalciuria and hypercalcemia — are unlikely to occur at the relatively small doses (usually 2000 IU) that have been shown to be effective. Treatment of established anticonvulsant bone disease by therapy with 2000 – 4000 IU of vitamin D daily has been demonstrated to result in improved calcium absorption, decreases in parathyroid hormone and urinary hydroxyproline excretion, and increases in bone mineral content [270]. All of these approaches should prove relatively cost-effective, particularly when contrasted with medical costs associated with fractures.

Acknowledgments The author thanks Dr. R. B. Staron for providing the radiographs for Fig. 1 and for helpful discussions regarding the use of magnetic resonance imaging for the differential diagnosis of vertebral compression fractures.

References 1. R. A. Kyle, Multiple myeloma: Review of 869 cases. Mayo Clin. Proc. 50, 29 – 40 (1975). 2. G. R. Mundy, Myeloma bone disease. Eur. J. Cancer 34, 246 – 251 (1998).

320 3. G. R. Mundy, L. G. Raisz, R. A. Cooper, G. P. Schechter, and S. E. Salmon, Evidence for the secretion of an osteoclast stimulating factor in myeloma. N. Engl. J. Med. 291, 1041 (1974). 4. P. Stashenko, F. E. Dewhirst, W. J. Peros, R. L. Kent, and J. A. Ago, Synergistic interactions between interkeukin 1, tumor necrosis factor and lymphotoxin in bone resorption. J. Immunol. 138, 1464 – 1468 (1987). 5. I. R. Garrett, B. G. M. Durie, G. E. Nedwin, A. Gillespie, T. Bringman, M. Sabatini, D. R. Bertolini, and G. R. Mundy, Production of lymphotoxin, a bone-resorbing cytokine, by cultured human myeloma cells. N. Engl. J. Med. 317, 526 – 532 (1987). 6. D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumor necrosis factors. Nature 319, 516 – 518 (1986). 7. P. Stashenko, F. E. Dewhirst, M. L. Rooney, L. A. Desjardins, and J. D. Heeley, Interleukin-1B is potent inhibitor of bone formation in vitro. J. Bone Miner. Res. 2, 559 – 565 (1987). 8. B. M. Thomson, J. Saklatvala, and T. J. Chambers, Osteoblasts mediate interleukin-1 stimulation of bone resorption by rat osteoclasts. J. Exp. Med. 164, 104 – 112 (1986). 9. B. M. Thompson, G. R. Mundy, and T. J. Chambers, Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J. Immunol. 138, 775 – 779 (1987). 10. G. R. Mundy, Y. Yoneda, and T. A. Guise, Hypercalcemia in hematologic malignancies and in solid tumors associated with extensive bone destruction. In “Primer on the Metabolic Bone Disease and Disorders of Mineral Metabolism (M. J. Favus, ed.), pp. 208 – 212 Raven Press, New York, 1999. 11. A. Valentin-Opran, S. A. Charhon, P. J. Meunier, C. M. Edouard, and M. E. Arlot, Quantitative histology of myeloma-induced bone changes. Br. J. Haematol. 52, 601 – 610 (1982). 12. J. Faust, P. Hunt, S. Scully, and V. Shalhoub, Multiple myeloma cells and cells of the human osteoclast lineage share morphological and cell surface markers. J. Cell Biochem. 71, 559 – 68 (1998). 13. S. B. Kapadia, Multiple myeloma: A clinicopathologic study of 62 consecutively autopsied cases. Medicine 59, 380 – 392 (1980). 14. I. Snapper, “Bone Diseases in Medical Practice.” Grune and Stratton, New York, 1957. 15. P. J. Lawson-Matthew, J. C. Clayton, D. F. Guilland-Cumming, et al., Defective production of 1,25(OH)2D3 in myeloma In “Vitamin D: A Chemical, Biochemical and Clinical Update” (A. W. Norman, K. Schaefer, H. G. Grigoleit, and D. V. Herrath, Eds.), pp. 891 – 892 (Walter de Gruyter, Berlin, 1985). 16. S. R. Nussbaum, R. J. Zahradnik, J. R. Lavigne, G. L. Brennan, L. Y. Nozawa-Ung, L. Y. Kim, H. T. Keutmann, C. A. Wang, J. T. Potts, and G. V. Segre, Highly sensitive two-site immunoradiometric assay of parathyrin, and its clinical utility in evaluating patients with hypercalcemia. Clin. Chem. 33, 1364 – 1367 (1987). 17. M. Pecherstorfer, I. Zimmer-Rth, T. Schilling, H. W. Woitge, H. Schmidt, G. Baumgartner, D. Thiebaud, H. Ludwig, and M. J. Seibel, The diagnostic value of urinary pyridiminum cross-links of collagen, serum total alkaline phosphatase, and urinary calcium excretion in neoplastic bone disease. J. Clin. Endocrinol. Metab. 80, 97 – 103 (1995). 18. H. W. Woitge, M. Pecherstorfer, Y. Li, A. V. Keck, E. Horn, R. Ziegler, and M. J. Seibel, Novel serum markers of bone resorption: Clinical assessment and comparison with established urinary indices. J. Bone Miner. Res. 14, 792 – 801 (1999). 19. K. Carlson, S. Ljunghall, B. Simonsson, and B. Smedmyr, Serum osteocalcin concentrations in patients with multiple myeloma — Correlation with disease stage and survival. J. Intern. Med. 231, 133 – 137 (1992). 20. H. G. Jacobson, M. H. Poppel, J. H. Shapiro, and S. Grossberger, The vertebral pedicle sign: A roentgen finding to differentiate

SCHNEIDER AND SHANE

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

metastatic carcinoma from multiple myeloma. Am. J. Roentgenol. 80, 817 – 821 (1958). A. Soloman, R. Rahamani, U. Seligsohn, and F. Ben-Artzi, Multiple myeloma: Early involvement assessed by computerized tomography. Skeletal. Radiol. 22, 254 – 261 (1984). L. Ludwig, D. Tscholakoff, A. Neuhold, F. Fruhwald, S. Rasoul, and E. Fritz, Magnetic resonance imaging of the spine in multiple myeloma. Lancet 2, 364 – 368 (1987). H. J. Kim, K. N. Ryu, W. S. Choi, B. K. Choi, J. M. Choi, and Y. Yoon, Spinal involvement of hematopoietic malignancies and metastasis: differentiation using MR imaging. Clin. Imaging 23, 125 – 133 (1999). F. E. Lecouvet, B. C. Vande Berg, B. E. Maldague, L. Michaux, E. Laterre, J. L. Michaux, A. Ferrant, and J. Malghem, Vertebral compression fractures in multiple myeloma. I. Distribution and appearance at MR imaging. Radiology 204, 195 – 1999 (1997). [see Comments] F. E. Lecouvet, J. Malghem, L. Michaux, J. L. Michaux, F. Lehmann, B. E. Maldague, J. Jamart, A. Ferrant, and B. C. Vande Berg, Vertebral compression fractures in multiple myeloma. Part II. Assessment of fracture risk with MR imaging of spinal bone marrow. Radiology 204, 201 – 205 (1997). [see Comments] X. Marietti, P. Khalifa, P. Ravaud, J. Frija, M. Laval-Jeantet, C. Chastang, J. C. Brouet, and J. P. Fermand, Bone densitometry in patients with multiple myeloma. Am. J. Med. 93, 595 – 598 (1992). M. V. Dhodapkar, R. Weinstein, G. Tricot, S. Jagannath, A. M. Parfitt, S. C. Manolagas, and B. Barlogie, Biologic and therapeutic determinants of bone mineral density in multiple myeloma. Leuk. Lymphoma 32, 121 – 127 (1998). L. H. Rodriguez, J. B. Finkelstein, C. C. Shullenberger, and R. Alexanian, Bone healing in multiple myeloma with melphalan chemotherapy. Ann. Intern. Med. 76, 551 – 556 (1972). E. S. Siris, W. H. Sherman, D. C. Baquiran, J. P. Schlatterer, E. F. Osserman, and R. E. Canfield, Effects of dichloromethylene diphosphonate on skeletal mobilization of calcium in multiple myeloma. N. Engl. J. Med. 302, 310 – 315 (1980). P. H. Delmas, S. Charhon, M. C. Chapuy, E. Vignon, D. Briancon, C. Edouard, and P. J. Meunier, Long-term effects of dichloromethylene diphosphonate (C12MDP) on skeletal lesions in multiple myeloma. Bone Dis. Relat. Res. 4, 163 – 168 (1982). R. Lahtinen, M. Laakso, I. Palva, P. Virkkunen, and I. Elomaa, Randomised, placebo-controlled multicentre trial of clodronate in multiple myeloma. Finnish Leukaemia Group [published erratum appears in Lancet 1992 Dec 5;340(8832): 1420]. Lancet 340, 1049 – 1052 (1992). [see Comments] F. J. M. v. Breukelen, O. L. M. Bijvoet, and A. T. v. Oosterom, Inhibition of osteolytic bone lesions by (3-amino-1hydroxypropylidene)-1, 1-bisphosphonate (A.P.D.). Lancet 1, 803 – 805 (1979). H. P. Sleeboom, A. T. v. Oosterom, O. L. M. Bijvoet, J. H. Gleed, and J. L. H. ORiordan, Comparison of intravenous (3-amino-1hydroxypropylidene)-1, 1-bisphosphonate and volume repletion in tumor-induced hypercalcemia. Lancet 1, 239 – 243 (1983). D. Thiebaud, S. Leyvraz, V. v. Fliedner, P. C. L. Perey, S. Thiebaud, and P. Burckhardt, Treatment of bone metastases from breast cancer and myeloma with pamidronate. Eur. J. Cancer 27, 37 – 41 (1991). J. R. Berenson, A. Lichtenstein, L. Porter, M. A. Dimopoulos, R. Bordoni, S. George, A. Lipton, A. Keller, O. Ballester, M. J. Kovacs et al., Efficacy of pamidronate in reducing skeletal events in patients with advanced multiple myeloma. Myeloma Aredia Study Group N. Engl. J. Med. 334, 488 – 493 (1996). [see Comments] J. R. Berenson, A. Lichtenstein, L. Porter, M. A. Dimopoulos, R. Bordoni, S. George, A. Lipton, A. Keller, O. Ballester, M. Kovacs, et al., Long-term pamidronate treatment of advanced multiple myeloma patients reduces skeletal events. Myeloma Aredia Study Group. J. Clin. Oncol. 16, 593 – 602 (1998). [see Comments]

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

321

37. C. M. Shipman, M. J. Rogers, J. F. Apperley, R. Graham, G. Russell, and P. I. Croucher, Anti-tumour activity of bisphosphonates in human myeloma cells. Leuk. Lymphoma 32, 129 – 138 (1998). 38. R. E. Coleman, O. P. Purohit, C. Black, J. J. Vinholes, K. Schlosser, H. Huss, K. J. Quinn, and J. Kanis, Double-blind, randomised, placebo-controlled, dose-finding study of oral ibandronate in patients with metastatic bone disease. Ann. Oncol. 10, 311 – 316 (1999). 39. J. J. Body, Clinical research update: Zoledronate. Cancer 80, 1699 – 1701 (1997). 40. F. Fulfaro, A. Casuccio, C. Ticozzi, and C. Ripamonti, The role of bisphosphonates in the treatment of painful metastatic bone disease: A review of phase III trials. Pain. 78, 157 – 169 (1998). 41. H. Rico, E. R. Hernandez, J. Diaz-Mediavilla, A. Alvarez, R. Martinez, and D. Espinos, Treatment of multiple myeloma with nasal spray calcitonin: A histomorphometric and biochemical study. Bone Miner. 8, 231 – 237 (1990). 42. R. P. Warrell, R. S. Bockman, C. J. Coonley, M. Isaacs, and H. Staszewski, Gallium nitrate inhibits calcium resorption from bone and is effective treatment for calcium-related hypercalcemia. J. Clin. Invest. 73, 1487 – 1490 (1984). 43. R. S. Bockman, M. A. Repo, R. P. Warrell, R. Israel, and J. Gabrilove, Gallium nitrate inhibits bone resorption induced by recombinant human tumor necrosis factor (TNF). Proc. Am. Assoc. Cancer Res. 28, 449 (1987). [Abstract] 44. R. P. Warrell, D. F. Lovett, A. Dilmarian, R. Schneider, and R. T. Heelan, Low-dose gallium nitrate for prevention of osteolysis in myeloma: Results of a pilot randomized study. J. Clin. Oncol. 11, 2443 – 2450 (1993). 45. R. P. Warrell, Jr., Gallium nitrate for the treatment of bone metastases. Cancer 80, 1680 – 1685 (1997). 46. R. Champlin, and D. W. Golde, The leukemias. In “Harrison’s Principles of Internal Medicine” (J. D. Wilson, eds.), pp. 1552 – 1561 McGraw Hill, New York, 1991. 47. A. J. Newman and D. K. Melhorn, Vertebral compression in childhood leukemia. Am. J. Dis. Child. 125, 863 – 865 (1973). 48. G. M. Samuda, M. Y. Cheng, and C. Y. Yeung, Back pain and vertebral compression: An uncommon presentation of childhood acute lymphoblastic leukemia. J. Pediatr. Orthop. 7, 175 – 178 (1987). 49. M. J. McKenna, Miscellaneous causes of osteoporosis. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), pp. 258 – 262 Raven Press, New York, 1993. 50. J. A. Child and I. E. Smith, Lymphoma presenting as “idiopathic” juvenile osteoporosis. Br. Med. J. 1, 720 – 721 (1975). 51. R. J. A. Aur, H. W. Westbrook, and W. Riggs, Childhood acute lymphoblastic leukemia: Initial radiologic bone involvement and prognosis. Am. J. Dis. Child. 124, 653 – 654 (1972). 52. R. G. Hughes and H. E. M. Kay, Major bone lesions in acute lymphoblastic leukemia. Med. Pediatr. Oncol. 10, 67 – 70 (1982). 53. G. Masera, V. Carnelli, M. Ferrari, M. Recchia, and F. Bellini, Prognostic significance of radiological bone involvement in childhood acute lymphocytic leukemia. Arch. Dis. Child. 52, 530 – 533 (1977). 54. J. Blatt, S. L. Martini, and L. Perchansky, Characteristics of acute lymphoblastic leukemia in children with osteopenia and vertebral compression fractures. J. Pediatr. 105, 280 – 282 (1984). 55. L. S. Cohn, E. R. Morgan, and L. E. Mallette, The spectrum of metabolic bone disease in lymphoblastic leukemia. Cancer 59, 346 – 350 (1987). 56. L. A. D. Castro, J. P. Kuhn, A. I. Freeman, L. F. Sinks, and E. R. Mindell, Complete remodeling of the vertebrae in a child successfully treated for acute lymphoblastic leukemia. Cancer 40, 398 – 401 (1977). 57. P. D. D’Angelo, V. Conter, G. DiChiara, C. Rizzari, A. Memeo, and P. Barigozzi, Severe osteoporosis and multiple vertebral collapses

in a child during treatment for B-ALL. Acta Haematol. 89, 38 – 42 (1993). P. C. Ribeiro, C. H. Pui, and M. J. Schell, Vertebral compression fractures as a presenting feature of acute lymphoblastic leukemia in children. Cancer 589 – 592 (1988). J. M. Halton, S. A. Atkinson, L. Fraher, C. E. Webber, W. P. Cockshott, C. Tam, and R. D. Barr, Mineral homeostasis and bone mass at diagnosis in children with acute lymphoblastic leukemia. J. Pediatr. 126, 557 – 564 (1995). J. M. Halton, S. A. Atkinson, L. Fraher, C. Webber, G. J. Gill, S. Dawson, and R. D. Barr, Altered mineral metabolism and bone mass in children during treatment for acute lymphoblastic leukemia. J. Bone Miner. Res. 11, 1774 – 1783 (1996). S. A. Atkinson, J. M. Halton, C. Bradley, B. Wu, and R. D. Barr, Bone and mineral abnormalities in childhood acute lymphoblastic leukemia: Influence of disease, drugs and nutrition. Int. J. Cancer Suppl. 11, 35 – 39 (1998). P. Arikoski, J. Komulainen, P. Riikonen, M. Parviainen, J. S. Jurvelin, R. Voutilainen, and H. Kroger, Impaired development of bone mineral density during chemotherapy: A prospective analysis of 46 children newly diagnosed with cancer. J. Bone Miner. Res. 14, 2002 – 2009 (1999). [see Comments] P. Arikoski, J. Komulainen, P. Riikonen, J. S. Jurvelin, R. Voutilainen, and H. Kroger, Reduced bone density at completion of chemotherapy for a malignancy. Arch. Dis. Child 80, 143 – 148 (1999). S. A. Atkinson, L. Fraher, C. M. Gundberg, M. Andrew, M. Pai, and R. D. Barr, Mineral homeostasis and bone mass in children treated for acute lymphoblastic leukemia. J. Pediatr. 114, 794 – 799 (1989). P. Arikoski, J. Komulainen, R. Voutilainen, P. Riikonen, M. Parviainen, P. Tapanainen, M. Knip, and H. Kroger, Reduced bone mineral density in long-term survivors of childhood acute lymphoblastic leukemia. J. Pediatr. Hematol. Oncol. 20, 234 – 240 (1998). A. H. Regab, R. S. Frech, and T. J. Vietti, Osteoporotic fractures secondary to methotrexate therapy of acute leukemia in remission. Cancer 25, 580 – 585 (1970). V. Gilsanz, M. E. Carlson, T. F. Rowe, and J. A. Ortega, Osteoporosis after cranial radiation for acute lymphoblastic leukemia. J. Pediatr. 117, 238 – 244 (1990). J. J. Hoorweg-Nijman, G. Kardos, J. C. Roos, H. J. van Dijk, C. Netelenbos, C. Popp-Snijders, C. M. de Ridder, and H. A. Delemarre-van de Waal, Bone mineral density and markers of bone turnover in young adult survivors of childhood lymphoblastic leukaemia. Clin. Endocrinol. (Oxford) 50, 237 – 44 (1999). N. H. Birkebaek, S. Fisker, N. Clausen, V. Tuovinen, S. Sindet-Pedersen, and J. S. Christiansen, Growth and endocrinological disorders up to 21 years after treatment for acute lymphoblastic leukemia in childhood. Med. Pediatr. Oncol. 30, 351 – 356 (1998). K. Nysom, K. Holm, K. F. Michaelsen, H. Hertz, J. Muller, and C. Molgaard, Bone mass after treatment for acute lymphoblastic leukemia in childhood. J. Clin. Oncol. 16, 3752 – 3760 (1998). J. F. Seymour, R. F. Gagel, F. B. Hagemeister, M. A. Dimopoulos, and F. Cabanillas, Calcitriol production in hypercalcemic and normocalcemic patients with non-Hodgkin lymphoma. Ann. Int. Med. 121, 633 – 640 (1994). W. D. Travis, C. Y. Li, E. J. Bergstralh, L. T. Yam, and R. G. Swee, Systemic mast cell disease. Analysis of 58 cases and literature review. Medicine 67, 345 – 368 (1988). R. H. Lands, Non-Hodgkin’s lymphoma, hypercalcemia, and elevated calcitriol levels: A report and review. J. Tenn. Med. Assoc. 88, 310 – 311 (1995). C. Brumsen, N. A. Hamdy, and S. E. Papapoulos, Long-term effects of bisphosphonates on the growing skeleton. Studies of young patients with severe osteoporosis. Medicine (Baltimore) 76, 266 – 283 (1997).

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

322 75. M. J. McKenna, D. Lundin, A. R. Villaneuva, and A. M. Parfitt, Staining for mast cells in undecalcified bone sections. J. Histotechnol. 13, 49 – 51 (1990). 76. M. J. McKenna and B. Frame, The mast cell and bone. Clin. Orthop. 200, 226 – 233 (1985). 77. D. Resnick and G. Niwayama, “Diagnosis of Bone and Joint Disorders, 2nd ed.,” pp. 1993 – 1994. Saunders, Philadelphia, 1988. 78. C. Johansson, G. Roupe, G. Lindstedt, and D. Mellstrom, Bone density, bone markers and bone radiological features in mastocytosis. Age Ageing 25, 1 – 7 (1996). 79. M. D. Fallon, M. P. Whyte, and S. L. Teitelbaum, Systemic mastocytosis associated with generalized osteopenia: Histopathological characterization of the skeletal lesions using undecalcified bone from two patients. Hum. Pathol. 12, 813 – 820 (1981). 80. C. Lidor, B. Frisch, D. Gazit, R. Gepstein, T. Hallel, and Y. A. Mekori, Osteoporosis as the sole presentation of bone marrow mastocytosis. J. Bone Miner. Res. 5, 871 – 876 (1990). 81. L. G. Raisz, Local and systemic factors in the pathogenesis of osteoporosis. N. Engl. J. Med. 318, 818 – 828 (1988). 82. L. V. Avioli, Heparin-induced osteopenia: An appraisal. Adv. Exp. Med. Biol. 52, 375 – 387 (1975). 83. C. Nelson-Piercy, Heparin-induced osteoporosis. Scand. J. Rheumatol. Suppl. 107, 68 – 71 (1998). 84. L. G. Raisz and T. J. Martin, Prostaglandins in bone and mineral metabolism. In “Bone and Mineral Research” (W. A. Peck, ed.), pp. 386 – 310. Elsevier, Amsterdam, 1983. 85. J. Grabbe, P. Welker, A. Moller, E. Dippel, L. K. Ashman, and B. M. Czarnetzki, Comparative cytokine release from human monocytes, monocyte-derived immature mast cells, and a human mast cell line (HMC-). J. Invest. Dermatol. 103, 504 – 508 (1994). 86. C. deGennes, D. Kuntz, and M. C. deVernejoul, Bone mastocytosis. A report of nine cases with a bone histomorphometric study. Clin. Orthop. Relat. Res. 279, 281 – 291 (1992). 87. A. Chines, R. Pacifici, L. V. Avioli, S. L. Teitelbaum, and P. E. Korenblatt, Systemic mastcytosis presenting as osteoporosis: A clinical and histomorphometric study. J. Clin. Endocrinol. Metab. 72, 140 – 144 (1991). 88. P. Shoenaers, L. S. DeClerck, U. Timmermans, and W. J. Stevens, Systemic mastocytosis, an unusual cause of osteoporosis. Clin. Rheumatol. 6, 458 – 462 (1987). 89. J. A. Harvey, H. C. Anderson, D. Borek, D. Morris, and B. P. Lukert, Osteoporosis associated with mastocytosis confined to bone: Report of two cases. Bone 10, 237 – 241 (1989). 90. E. R. Arrington, B. Eisenberg, M. F. Hartshorne, S. Vela, and R. I. Dorin, Nuclear medicine imaging of systemic mastocytosis. J. Nucl. Med. 30, 2046 – 2048 (1989). 91. K. Haney, W. Russell, F. A. Raila, A. C. Brower, and R. B. Harrison, MRI characteristics of systemic mastocytosis of the lumbosacral spine. Skel. Radiol. 25, 171 – 173 (1996). 92. T. A. Webb, C. Y. Li, and L. T. Yam, Systemic mast cell disease: A clinical and haematopathological study of 26 cases. Cancer 49, 927 – 938 (1982). 93. B. Frame and R. K. Nixon, Bone marrow mast cells in osteoporosis and aging. N. Engl. J. Med. 279, 626 – 630 (1968). 94. M. D. Fallon, M. P. Whyte, B. Craig, and S. L. Teitelbaum, Mastcell proliferation in postmenopausal osteoporosis. Calcif. Tissue Int. 35, 29 – 31 (1983). 95. M. J. McKenna, Histomorphometric study of mast cells in normal bone, osteoporosis and mastocytosis using a new stain. Calcif. Tissue Int. 55, 257 – 259 (1994). 96. L. Graves, D. J. Stechschulte, D. C. Morris, and B. P. Lukert, Inhibition of mediator release in systemic mastocytosis is associated with reversal of bone changes. J. Bone Miner. Res. 5, 1113 – 1119 (1990).

SCHNEIDER AND SHANE 97. R. R. Alexander, Disodium cromoglycate in the treatment of systemic mastocytosis involving only bone. Acta Haematol. 74, 108 – 110 (1985). 98. T. Cundy, M. H. C. Beneton, A. J. Darby, W. J. Marshall, and J. A. Kanis, Osteopaenia in systemic mastocytosis: Natural history and responses to treatment with inhibitors of bone resorption. Bone 8, 149 – 155 (1987). 99. A. Marshall, R. T. Kavanagh, and A. J. Crisp, The effect of pamidronate on lumbar spine bone density and pain in osteoporosis secondary to systemic mastocytosis. Br. J. Rheumatol. 36, 393 – 396 (1997). 100. R. A. Watts, D. G. I. Scott, and A. J. Crisp, Mastocytosis and osteoporosis. Br. J. Rheumatol. 31, 715 – 720 (1992). 101. M. E. Levin, V. C. Boisseau, and L. V. Avioli, Effects of diabetes mellitus on bone mass in juvenile and adult-onset diabetes. N. Engl. J. Med. 294, 241 – 245 (1976). 102. A. L. Rosenbloom, D. C. Lezotte, F. T. Weber, J. Gudat, D. R. Heller, M. L. Weber, S. Klein, and B. B. Kennedy, Diminution of bone mass in childhood diabetes. Diabetes 26, 1052 – 1055 (1977). 103. J. V. Santiago, W. H. McAlister, S. K. Ratzan, Y. Bussman, M. V. Haymond, G. Shakelford, and V. V. Weldon, Decreased cortical thickness and osteopenia in children with diabetes mellitus. J. Clin. Endocrinol. Metab. 45, 845 – 848 (1977). 104. C. Thoger, K. K. Ibsen, J. S. Christiansen, et al., Diabetes mellitus, skeletal age and bone mineral content in children. Acta Endocrinol. Suppl. 58, 219 (1978). 105. P. McNair, S. Madsbad, C. Christiansen, O. K. Faber, and G. Binder, Osteopenia in insulin-treated diabetes mellitus. Its relation to age at onset, sex and duration of disease. Diabetologia 15, 87 – 90 (1978). 106. R. M. Shore, R. W. Chesney, R. B. Mazess, P. G. Rose, and G. J. Bargman, Osteopenia in juvenile diabetes. Calcif. Tissue Int. 33, 455 – 457 (1981). 107. T. E. Frazer, N. H. White, S. Hough, J. V. Santiago, B. R. McGee, G. Bryce, J. Mallon, and L. V. Avioli, Alterations in circulating vitamin D metabolites in the young insulin-dependent diabetic. J. Clin. Endocrinol. Metab. 53, 1154 – 1159 (1981). 108. P. S. Wiske, S. M. Wentworth, J. A. Norton, S. Epstein, and C. C. Johnston, Evaluation of bone mass and growth in young diabetics. Metabolism 31, 848 – 854 (1982). 109. S. L. Hui, S. Epstein, and C. C. Johnston, A prospective study of bone mass in patients with type 1 diabetes. J. Clin. Endocrinol. Metab. 60, 74 – 80 (1985). 110. F. S. Hough, Alterations of bone and mineral metabolism in diabetes mellitus. Part II. South Afr. Med. 72, 120 – 126 (1987). 111. M. Leon, L. Larrodera, G. Lledo, and F. Hawkins, Study of bone loss in diabetes mellitus type 1. Diabetes Res. Clin. Pract. 6, 237 – 242 (1989). 112. G. Saggese, S. Bertelloni, G. I. Baroncelli, G. Federico, L. Calisti, and C. Fusaro, Bone demineralization and impaired mineral metabolism in insulin-dependent diabetes mellitus. Helv. Paediatr. Acta 43, 405 – 414 (1988). 113. G. Weber, L. Beccaria, M. de’Angelis, S. Mora, L. Galli, M. A. Cazzuffi, F. Turba, F. Frisone, M. P. Guarneri, and G. Chiumello, Bone mass in young patients with type 1 diabetes. Bone Miner. 8, 23 – 30 (1990). 114. R. Bouillon, Diabetic bone disease. Calcif. Tissue Int. 49, 155 – 160 (1991). 115. J. Auverx, J. Dequeker, R. Bouillon, P. Geusens, J. Nijs, Mineral metabolism and bone mass at peripheral and axial skeleton in diabetes mellitus. Diabetes 37, 8 – 12 (1988). 116. T. Forst, A. Pfutzner, P. Kann, B. Schehler, R. Lobmann, H. Schafer, J. Andreas, A. Bockisch, and J. Beyer, Peripheral osteopenia in adult patients with insulin-dependent diabetes mellitus. Diabetes Med. 12, 874 – 879 (1995).

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

323

117. T. F. Roe, S. Mora, G. Costin, F. Kaufman, M. E. Carlson, and V. Gilsanz, Vertebral bone density in insulin-dependent diabetic children. Metabolism 40, 967 – 971 (1991). 118. S. J. Gallacher, J. A. K. Fenner, B. M. Fisher, J. D. Quin, W. D. Fraser, F. C. Logue, R. A. Cowan, I. T. Boyle, and A. C. MacCuish, An evaluation of bone density and turnover in premenopausal women with type 1 diabetes mellitus. Diabetic Med. 10, 129 – 133 (1993). 119. J. De Schepper, J. Smitz, S. Rosseneu, P. Bollen, and O. Louis, Lumbar spine bone mineral density in diabetic children with recent onset. Horm. Res. 50, 193 – 196 (1998). 120. B. Mathiassen, S. Nielsen, J. Ditzel, and P. Rodbro, Long-term bone loss in insulin-dependent diabetes mellitus. J. Int. Med. 227, 325 – 327 (1990). 121. R. S. Weinstock, R. S. Goland, E. Shane, T. L. Clemens, R. Lindsay, and J. P. Bilezikian, Bone mineral density in women with type II diabetes mellitus. J. Bone Miner. Res. 4, 97 – 101 (1989). 122. H. E. Meema and S. Meema, The relationship of diabetes mellitus and body weight to osteoporosis in elderly females. J. Can. Med. Assoc. 96, 132 – 139 (1967). 123. H. Ishida, Y. Seino, S. Matsukura, M. Ikeda, M. Yawata, G. Yamashita, S. Ishizuka, and H. Imura, Diabetic osteopenia and circulating levels of vitamin D metabolites in type 2 (non-insulin-dependent) diabetes. Metabolism 34, 797 – 801 (1985). 124. G. Isaia, L. Bodrato, V. Carlevatto, M. Mussetta, G. Salamano, and G. M. Molinatti, Osteoporosis in type II diabetes. Acta Diabetol. 24, 305 – 310 (1987). 125. C. C. Johnston, S. L. Hui, and C. Longcope, Bone mass and sex steroid concentrations in postmenopausal Caucasian diabetics. Metabolism 34, 544 – 550 (1985). 126. F. Gregorio, S. Cristallini, F. Santeusanio, P. Fillipponi, and P. Fumelli, Osteopenia associated with non-insulin-dependent diabetes mellitus: What are the causes? Diabetes Res. Clin. Pract. 23, 43 – 54 (1994). 127. Y. Seino, H. Ishida, and H. Himura, Diabetic osteopenia in central Japan. Diabetes Metab. 11, 216 – 219 (1985). 128. M. Buysschaert, F. Cauwe, J. Jamart, C. Brichant, P. DeCoster, A. Magnan, and J. Donckier, Proximal femur density in type 1 and 2 diabetic patients. Diabetes Metab. 18, 32 – 37 (1992). 129. C. Montecucco, F. Baldi, R. Caporali, et al., Serum osteocalcin (bone gla-protein) and bone mineral content in non-insulin dependent diabetes. Diabetes Nutr. Metab. 4, 311 – 316 (1990). 130. E. Barrett-Connor and T. L. Holbrooke, Sex differences in osteoporosis in older adults with non-insulin-dependent diabetes mellitus. JAMA 268, 3333 – 3337 (1992). 131. A. Giacca, A. Fassina, F. Caviezel, A. G. Cattaneo, G. Caldirola, and G. Pozza, Bone mineral density in diabetes mellitus. Bone 9, 29 – 36 (1988). 132. M. Wakasugy, R. Wakao, M. Tawata, N. Gan, K. Koizumi, and T. Onaya, Bone mineral density measured by dual energy X-ray absorptiometry in patients with non-insulin dependent diabetes mellitus. Bone 14, 29 – 33 (1993). 133. J. T. Tuominen, O. Impivaara, P. Puukka, and T. Ronnemaa, Bone mineral density in patients with type 1 and type 2 diabetes. Diabetes Care 22, 1196 – 1200 (1999). 134. D. G. Seeley, J. Kelsey, M. Jergas, and M. C. Nevitt, Predictors of ankle and foot fractures in older women. The Study of Osteoporotic Fractures Research Group. J. Bone Miner. Res. 11, 1347 – 1355 (1996). 135. P. A. Alffram, An epidemiologic study of cervical and trochanteric fractures of the femur in an urban population. Acta Orthop. Scand. Suppl. 65, 75 – 77 (1964). 136. J. Menczel, M. Makin, and G. Robin, Prevalence of diabetes mellitus in Jerusalem; its association with osteoporosis. Isr. J. Med. Sci. 918 – 919 (1972).

137. T. A. Hutchinson, S. M. Polansky, and A. R. Feinstein, Postmenopausal oestrogens protect against fractures of hip and radius: A case – control study. Lancet 705 – 708 (1979). 138. J. C. Gallagher, L. J. Melton, and B. L. Riggs, Examination of prevalence rates of possible risk factors in population with a fracture of the proximal femur. Clin. Orthop. 153, 158 – 165 (1980). 139. S. Weintroub, D. Eisenberg, R. Tardiman, S. L. Weissman, and R. Salama, Is diabetic osteoporosis due to microangiopathy? Lancet 2, (1980). 140. A. L. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 141. H. Heath, L. J. Melton, and C. Chu-Pin, Diabetes mellitus and risk of skeletal fracture. N. Engl. J. Med. 303, 567 – 570 (1980). 142. L. Forsen, H. E. Meyer, K. Midthjell, and T. H. Edna, Diabetes mellitus and the incidence of hip fracture: Results from the Nord – Trondelag Health Survey. Diabetologia 42, 920 – 925 (1999). 143. U. Nisbeth, E. Lindh, S. Ljunghall, U. Backman, and B. Fellstrom, Increased fracture rate in diabetes mellitus and females after renal transplantation. Transplantation 67, 1218 – 1222 (1999). 144. M. Y. Chiu, S. M. Sprague, D. S. Bruce, E. S. Woodle, J. R. Thistlethwaite, Jr., and M. A. Josephson, Analysis of fracture prevalence in kidney – pancreas allograft recipients. J. Am. Soc. Nephrol. 9, 677 – 683 (1998). 145. P. J. Grant, M. H. Stickland, A. F. Bristow, R. F. Clarke, and J. K. Wales, Insulin increases somatomedin C (IGF-1) concentrations in Type 1 diabetic patients. Diabetic Med. 6, 245 – 248 (1989). 146. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor-1 has independent effects on bone matrix and cell replication. Endocrinology 122, 254 – 260 (1988). 147. R. P. Stolk, P. L. Van Daele, H. A. Pols, H. Burger, A Hofman, J. C. Birkenhager, S. W. Lamberts, and D. E. Grobbee, Hyperinsulinemia and bone mineral density in an elderly population: The Rotterdam Study [published erratum appears in Bone 1996 Nov;19(5):566]. Bone 18, 545 – 549 (1996). 148. P. M. Jehle, D. R. Jehle, S. Mohan, and B. O. Bohm, Serum levels of insulin-like growth factor system components and relationship to bone metabolism in Type 1 and Type 2 diabetes mellitus patients. J. Endocrinol. 159, 297 – 306 (1998). 149. P. McNair, Bone mineral metabolism in human type 1 (insulin-dependent) diabetes mellitus. Dan. Med. Bull. 35, 109 – 121 (1988). 150. A.-H. Shao, F. G. Wang, Y.-F. Hu, and L.-M. Zhang, Calcium metabolism and osteopathy in diabetes mellitus. Contrib. Nephrol. 90, 212 – 216 (1991). 151. M. Klein, K. Wu, and H. M. Frost, The numbers of bone resorption and formation foci in rib in diabetes mellitus. Henry Ford Hosp. Med. Bull. 12, 527 – 536 (1964). 152. P. Pietschmann, G. Schernthaner, and W. Woloszczuk, Serum osteocalcin levels in diabetes mellitus: Analysis of the type of diabetes and microvascular complications. Diabetologia 31, 892 – 895 (1988). 153. A. Verrotti, F. Basciani, F. Carle, G. Morgese, and F. Chiarelli, Calcium metabolism in adolescents and young adults with type 1 diabetes mellitus without persistant microalbuminuria. J. Endocrinol. Invest. 22, 198 – 202 (1999). 154. T. M. Melchior, O. H. Sorensen, G. Thamsborg, R. Sykuluski, and T. Storm, In “Tissue-Specific Metabolic Alterations in Diabetes” (F. Belfiore, G. M. Molinatti, and G. M. Reaven, Eds.), pp. 40 – 53. Karger, Basel, 1990. 155. J. Krakauer, M. J. McKenna, D. S. Rao, F. W. Whitehouse, and A. M. Parfitt, Low bone turnover in diabetes mellitus accounts for preservation of bone mineral density. Diabetologia 35, A192 (1992). 156. D. Mauricio, T. Mandrup-Poulsen, and J. Nerup, Vitamin D analogues in insulin-dependent diabetes mellitus and other autoimmune diseases: A therapeutic perspective. Diabetes Metab. Rev. 12, 57 – 68 (1996).

324 157. Vitamin D supplement in early childhood and risk for Type I (insulin-dependent) diabetes mellitus. The EURODIAB Substudy 2 Study Group. Diabetologia 42, 51 – 54 (1999). 158. K. C. Baynes, B. J. Boucher, E. J. Feskens, and D. Kromhout, Vitamin D, glucose tolerance and insulinaemia in elderly men [published erratum appears in Diabetologia 1997 Jul;40(7):870]. Diabetologia 40, 344 – 347 (1997). 159. M. F. McDermott, A. Ramachandran, B. W. Ogunkolade, E. Aganna, D. Curtis, B. J. Boucher, C. Snehalatha, and G. A. Hitman, Allelic variation in the vitamin D receptor influences susceptibility to IDDM in Indian Asians. Diabetologia 40, 971 – 5 (1997). 160. L. J. Baier, A. M. Dobberfuhl, R. E. Pratley, R. L. Hanson, and C. Bogardus, Variations in the vitamin D-binding protein (Gc locus) are associated with oral glucose tolerance in nondiabetic Pima Indians. J. Clin. Endocrinol. Metab. 83, 2993 – 2996 (1998). 161. G. Hampson, C. Evans, R. J. Petitt, W. D. Evans, S. J. Woodhead, J. R. Peters, and S. H. Ralston, Bone mineral density, collagen type 1 alpha 1 genotypes and bone turnover in premenopausal women with diabetes mellitus. Diabetologia 41, 1314 – 20 (1998). 162. M. M. Kaplan, Primary biliary cirrhosis, Wilson’s disease hemochromatosis, and other metabolic and fibrotic liver diseases. In “Internal Medicine” (J. H. Stein, ed.), pp. 622 – 624. (Mosby, St. Louis, MO, 1994). 163. F. Delbarre, L’osteoporose des hemochromatoses. Semin. Hop. Paris 36, 3279 – 3295 (1960). 164. T. Diamond, D. Stiel, and S. Posen, Osteoporosis in hemochromatosis: Iron excess, gonadal deficiency, or other factors? Ann. Intern. Med. 110, 430 – 436 (1989). 165. Y. Pawlotsky, Y. Lacien, G. Roudier, Y. Hany, J. Y. Louboutin, B. Ferrand, and M. Bourel, Histomorphometrie osseuse et manifestations osteo-articulaires de l’hemochromatose idiopathique. Rev. Rhum. Mal. Osteoartic. 46, 91 – 99 (1979). 166. D. Conte, M. P. Caraceni, J. Duriez, C. Mandelli, E. Corghi, M. Cesana, S. Ortolani, and P. A. Bianchi, Bone involvement in primary hemochromatosis and alcoholic cirrhosis. Am. J. Gastroenterol. 84, 1231 – 1234 (1989). 167. E. Czink, C. Horvath, A. A. Malek, and G. Siklosi, Is dehydroepiandrosterone sulfate a possible pathogenetic factor in osteopenia of patients with iron overload? Haematologia 24, 47 – 56 (1991). 168. K. E. Oerter, G. A. Kamp, P. J. Munson, A. W. Nienhuis, F. G. Cassorla, and P. K. Manasco, Multiple hormone deficiencies in children with hemochromatosis. J. Clin. Endocrinol. Metab. 76, 357 – 361 (1993). 169. T. Diamond, D. Stiel, and S. Posen, Effects of testosterone and venesection on spinal and peripheral bone mineral in six hypogonadal men with hemochromatosis. J. Bone Miner. Res. 6, 39 – 43 (1991). 170. F. Albright and E. C. Reifenstein, “The Parathyroid Glands and Metabolic Bone Disease: Selected Studies.” Williams and Wilkins, Baltimore, 1948. 171. L. Riggs, R. V. Randall, H. W. Wahner, J. Jowsey, P. J. Kelly, and M. Singh, The nature of metabolic bone disorder in acromegaly. J. Clin. Endocrinol. Metab. 34, 911 – 918 (1972). 172. E. Seeman, H. W. Wahner, K. P. Offord, R. Kumar, W. J. Johnson, and B. L. Riggs, Differential effects of endocrine dysfunction on the axial and the appendicular skeleton. J. Clin. Invest. 69, 1302 – 1309 (1982). 173. T. Diamond, L. Nery, and S. Posen, Spinal and peripheral bone mineral densities in acromegaly: The effects of excess growth hormone and hypogonadism. Ann. Intern. Med. 111, 567 – 573 (1989). 174. S. Ezzat, S. Melmed, D. Endres, D. Eyre, and F. R. Singer, Biochemical assessment of bone formation and resorption in acromegaly. J. Clin. Endocrinol. Metab. 76, 1452 – 1457 (1993). 175. G. P. Lesse, W. D. Fraser, R. Farquharson, L. Hipkin, and J. P. Vora, Gonadal status is an important determinant of bone density in acromegaly. Clin. Endocrinol. (Oxford) 48, 59 – 65 (1998).

SCHNEIDER AND SHANE 176. A. Scillitani, I. Chiodini, V. Carnevale, G. M. Giannatempo, V. Frusciante, M. Villella, M. Pileri, G. Guglielmi, A. Di Giorgio, S. Modoni, et al., Skeletal involvement in female acromegalic subjects: The effects of growth hormone excess in amenorrheal and menstruating patients. J. Bone Miner. Res. 12, 1729 – 1736 (1997). 177. H. Stracke, A. Schulz, D. Moeller, S. Rossol, and H. Schatz, Effect of growth hormone on osteoblasts and demonstration of somatomedin-C/IGF-I in bone organ culture. Acta Endocrinol. 107, 16 – 24 (1984). 178. E. Canalis, Effect of insulin-like growth factor I on DNA and protein synthesis in cultured rat calvaria. J. Clin. Invest. 66, 709 – 719 (1980). 179. E. Canalis, T. L. McCarthy, and M. Centrella, The role of growth factors in skeletal remodeling. Endocrinol. Metab. Clin. North Am. 18, 903 – 918 (1989). 180. J. S. Johansen, S. A. Pedersen, J. O. L. Jorgensen, B. J. Riis, C. Christiansen, J. S. Christiansen, and N. D. Skakkeback, Effects of growth hormone (GH) on plasma bone Gla protein in GH-defi cient adults. J. Clin. Endocrinol. Metab. 70, 916 – 919 (1990). 181. C. d. l. Piedra, E. Carbo, J. Larranaga, N. Castro, C. Horcajada, A. Rapado, J. L. Herrera-Pombo, and E. Carbo, Correlation among plasma osteocalcin, growth hormone, and somatomedin C in acromegaly. Calcif. Tissue Int. 43, 44 – 45 (1988). 182. M. Terzolo, A. Piovesan, G. Osella, A. Pia, G. Reimondo, C. Pozzi, C. Raucci, M. Torta, P. Paccotti, and A. Angeli, Serum levels of bone Gla protein (osteocalcin, BGP) and carboxyterminal propeptide of type 1 procollagen (PICP) in acromegaly: Effects of longterm octreotide treatment. Calcif. Tissue Int. 52, 188 – 191 (1993). 183. P. Legovini, E. De Menis, F. Breda, D. Billeci, A. Carteri, P. Pavan, and N. Conte, Long-term effects of octreotide on markers of bone metabolism in acromegaly: Evidence of increased serum parathormone concentrations. J. Endocrinol. Invest. 20, 434 – 438 (1997). 184. T. Ueland, J. Bollerslev, T. B. Hansen, E. N. Ebbesen, L. Mosekilde, K. Brixen, A. Flyvbjerg, and O. Djoseland, Increased cortical bone content of insulin-like growth factors in acromegalic patients. J. Clin. Endocrinol. Metab. 84, 123 – 127 (1999). 185. J. Halse, F. Melsen, and L. Mosekilde, Iliac crest bone mass and remodelling in acromegaly. Acta Endocrinol. 97, 18 – 22 (1981). 186. J. Diebold, B. Batge, H. Stein, G. Muller-Esch, P. K. Muller, and U. Lohrs, Osteoporosis in longstanding acromegaly: Characteristic changes of vertebral trabecular architecture and bone matrix composition. Virchow Archiv. Pathol. Anat. 419, 209 – 215 (1991). 187. G. R. Delling and A. Schulz, Bone cells and remodeling surfaces in acromegaly. Calcif. Tissue Res. (Suppl.) 22, 255 – 259 (1977). 188. G. C. Griffith and A. Silverglade, Symposium on heparin. Am. J. Cardiol. 14, 1 – 2 (1964). 189. G. C. Griffith, G. Nichols, J. D. Asher, and B. Flanagan, Heparin osteoporosis. JAMA 193, 91 – 95 (1965). 190. M. D. Jaffe and P. W. Willis, Multiple fractures associated with long-term heparin therapy. JAMA 193, 152 – 154 (1965). 191. R. Marcus, Secondary forms of osteoporosis. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 902 – 903. Raven Press, New York, 1992. 192. J. P. Sackler and L. Liu, Heparin-induced osteoporosis. Br. J. Radiol. 46, 548 – 550 (1973). 193. D. Aarskog and L. Aksnes, and V. Lehmann, Low 1,25-dihydroxyvitamin D in heparin-induced osteopenia. Lancet 2, 650 – 651 (1980). 194. P. H. Wise, and A. J. Hall, Heparin-induced osteopenia in pregnancy. Br. Med. J. 281, 110 – 111 (1980). 195. J. W. Squires and L. W. Pinch, Heparin-induced spinal fractures. J. Am. Med. Assoc. 241, 2417 – 2418 (1979). 196. M. Megard, M. Cuche, A. Grapeloux, C. Bojoly, and P. J. Meunier, Osteoporose de l’heparinotherapie. Analyse histomorphometrique de la biopsie osseuse. Une observation. Nous Press Med. 11, 261 – 264 (1982).

CHAPTER 51 Osteoporosis Secondary to Illnesses and Medications

325

197. M. Lequesne, J. Lamotte, and N. Foures, Osteoporose del’heparinotherapie, osteoporose de la mastocytose. In “Expression of Scientifique Francaise” (S. d. Seze, A. Ryckewaert, M. F. Kahn, and C. Guerin, eds.), pp. 105 – 115. (Francaise, Paris, 1982.) 198. T. Bardin and M. Lequesne, The osteoporosis of heparinotherapy and systemic mastocytosis. Clin. Rheumatol. 8, 119 – 123 (1989). 199. T. Dahlman, N. Lindvall, M. Hellgren, Osteopenia in pregnancy during long-term heparin treatment: A radiological study post partum. Br. J. Obstet. Gynaecol. 97, 221 – 228 (1990). 200. T. C. Dahlman, Osteoporotic fractures and the recurrence of thromboembolism during pregnancy and the puerperium in 184 women undergoing thromboprophylaxis with heparin. Am. J. Obstet. Gynecol. 168, 1265 – 1270 (1993). 201. W. M. Rupp, H. B. McCarthy, T. D. Rohde, P. J. Blackshear, F. J. Goldenberg, and H. Buchwald, Risk of osteoporosis in patients treated with long-term intravenous heparin therapy. Curr. Surg. 39, 419 – 422 (1982). 202. J. S. Ginsberg, C. T. Kowalchuk, J. Hirsch, P. Brill-Edwards, R. Burrows, G. Coates, and C. Webber, Heparin effect on bone density. Thromb. Haemostasis 64, 286 – 289 (1990). 203. M. DeSwiet, P. W. Dorrington, J. Fidler, A. Horsman, D. Katz, E. Letsky, M. Peacock, and P. H. Wise, Prolonged heparin therapy in pregnancy causes bone demineralization. Br. J. Obstet. Gynecol. 90, 1129 – 1134 (1983). 204. L. A. Barbour, S. D. Kick, J. F. Steiner, M. E. LoVerde, L. N. Heddleston, J. L. Lear, A. E. Baron, and P. L. Barton, A prospective study of heparin-induced osteoporosis in pregnancy using bone densitometry. Am. J. Obstet. Gynecol. 170, 862 – 869 (1994). 205. T. C. Dahlman, H. E. Sjoberg, and H. Ringertz, Bone mineral density during long-term prophylaxis with heparin in pregnancy. Am. J. Obstet. Gynecol. 170, 1315 – 1320 (1994). 206. W. J. F. v. d. Vijgh, A. B. J. Kruik, M. J. M. Jongen, P. Lips, and J. C. Netelenbos, Inhibition of 25(OH)D la-hydroxylase in kidney homogenates of Japanese quail by heparin. Netelenbos Calcif. Tissue Int. 35, 192 (1983). [Abstract] 207. S. Mutoh, N. Takeshita, T. Yoshino, and I. Yamaguchi, Characterization of heparin-induced osteopenia in rats. Endocrinology 133, 2743 – 2748 (1993). 208. D. Aarskog, L. Aksnes, T. Markestad, M. Ulstein, and N. Sagen, Heparin-induced inhibition of 1,25-dihydroxyvitamin D formation. Am. J. Obstet. Gynecol. 148, 1141 – 1142 (1984). 209. P. Goldhaber, Heparin enhancement factors stimulating bone resorption in tissue culture. Science 147, 407 – 408 (1965). 210. J. D. Asher and G. Nichols, Heparin stimulation of bone collagenese activity. Fed. Proc. 24, 211 (1965). 211. S. Sakamoto and M. Sakamoto, Bone collagenase, osteoblasts and cell-mediated bone resorption. Bone Miner. Res. 4, 49 – 102 (1986). 212. S. G. Shaughnessy, J. Hirsh, M. Bhandari, J. M. Muir, E. Young, and J. I. Weitz, A histomorphometric evaluation of heparin-induced bone loss after discontinuation of heparin treatment in rats. Blood 93, 1231 – 1236 (1999). 213. R. C. Thompson, Heparin osteoporosis: An experimental model using rats. J. Bone Jt. Surg. 55A, 606 – 612 (1973). 214. A. Zimran, S. Shilo, D. Fisher, and I. Bab, Histomorphometric evaluation of reversible heparin-induced osteoporosis in pregnancy. Arch. Intern. Med. 146, 386 – 388 (1986). 215. J. S. Ginsberg and J. Hirsh, Optimum use of anticoagulants in pregnancy. Drugs 36, 505 – 512 (1988). 216. A. Many, R. Pauzner, H. Carp, P. Langevitz, and U. Martinowitz, Treatment of patients with antiphospholipid antibodies during pregnancy. Am. J. Reprod. Immunol. 28, 216 – 218 (1992). 217. T. Matzsch, D. Bergvuist, U. Hedner, and B. Nilsson, Effects of low molecular weight heparin and unfragmented heparin on induction of osteoporosis in rats. Thromb. Haemostasis 63, 505 – 509 (1990).

218. J. M. Muir, J. Hirsh, J. I. Weitz, M. Andrew, E. Young, and S. G. Shaughnessay, A histomorphometric comparison of the effects of heparin and low-molecular-weight heparin on cancellous bone in rats. Blood 89, 3236 – 3242 (1997). 219. M. Sivakumaran, K. Ghosh, Y. Zaidi, and R. M. Hutchinson, Osteoporosis and vertebral collapse following low-dose, low molecular weight heparin therapy iin a young patient. Clin. Lab. Haematol. 18, 55 – 57 (1996). 220. P. A. Price and M. K. Williamson, Effects of warfarn on bone. Studies on the vitamin K-dependent protein of rat bone. J. Biol. Chem. 256, 12754 – 12759 (1981). 221. K. J. Obrant, S. M. Kakonen, J. Astermark, H. Lilja, T. Lovgren, K. Akesson, and K. Pettersson, The production of carboxylated to total or intact osteocalcin in serum discriminates warfarin-treated patients from control subjects. J. Bone Miner. Res. 14, 555 – 560 (1999). 222. P. V. Hauschka and M. L. Reid, Vitamin K dependence of a calcium binding protein containing gamma-carboxyglutamic acid in chicken bone. J. Biol. Chem. 253, 9063 – 9068 (1978). 223. P. Pastoureau, P. Vergnaud, P. J. Meunier, and P. D. Delmas, Osteopenia and bone-remodeling abnormalities in warfarin-treated lambs. J. Bone Miner. Res. 8, 1417 – 1426 (1993). 224. J. Hall, R. M. Pauli, and K. M. Wilson, Maternal and fetal sequelae of anticoagulation during pregnancy. Am. J. Med. 68, 122 – 140 (1980). 225. S. A. Jamal, W. S. Browner, D. C. Bauer, and S. R. Cummings, Warfarin use and risk for osteoporosis in elderly women. Ann. Intern. Med. 128, 829 – 832 (1998). 226. W. C. Werkheiser, Specific binding of 4-amino folic acid analogs by folic acid reductase. J. Biol. Chem. 236, 888 – 893 (1961). 227. D. J. Mazanec and J. M. Grisanti, Drug-induced osteoporosis. Cleveland Clin. J. Med. 56, 297 – 303 (1989). 228. S. O’Regan, D. K. Melhorn, and A. J. Newman, Methotrexate-induced bone pain in childhood leukemia. Am. J. Dis. Child. 126, 489 – 490 (1973). 229. A. M. Schwartz and J. C. Leonidas, Methotrexate osteopathy. Skeletal Radiol. 11, 13 – 16 (1984). 230. S. Stanisavljevic and A. L. Babcock, Fractures in children treated with methotrexate for leukemia. Clin. Orthop. Relat. Res. 125, 139 – 144 (1977). 231. J. Scalone and J. R. Bowen, Methotrexate osteopathy: Case report. Contemp. Orthop. 17, 46 – 49 (1988). 232. G. Jones and P. N. Sambrook, Drug-induced disorders of bone metabolism. Incidence, management and avoidance. Drug Safety 10, 480 – 489 (1994). 233. R. Kruse, Osteopathien bei antiepiliptscher langzeit-Therapie. Monatsschr. Kinderheilkd. 116, 378 – 380 (1968). 234. M. Nesbit, W. Krivit, R. Heyn, and H. Sharp, Acute and chronic effects of methotrexate on hepatic, pulmonary, and skeletal systems. Cancer 37, 1048 – 1054 (1976). 235. S. Gnudi, L. Butturini, C. Ripamonti, M. Avella, and G. Baci, The effects of methotrexate on bone. A densitometric study conducted on 59 patients with methotrexate administered at different doses. Ital. J. Orthop. Traumatol. 14, 227 – 231 (1988). 236. M. L. Bianchi, R. Cimaz, E. Galbiati, F. Corona, R. Cherubini, and M. Bardare, Bone mass change during methotrexate treatment in patients with juvenile rheumatoid arthritis. Osteoporosis Int. 10, 20 – 25 (1999). 237. S. J. Preston, T. Diamond, A. Scott, and M. R. Laurent, Methotrexate osteopathy in rheumatic disease. Ann. Rheum. Dis. 52, 582 – 585 (1993). 238. M. Singwe, L. Le Gars, A. Karneff, A. Prier, and G. Kaplan, Multiple stress fractures in a scleroderma patient on methotrexate therapy. Rev. Rhum. Engl. Ed. 65, 508 – 510 (1998). 239. L. B. Thomas, C. E. Forkner, I. E. Frei, J. B. E. Besse, and J. R. Stabenau, The skeletal lesions of acute leukemia. Cancer 14, 608 – 621 (1961).

326 240. H. A. Waisman and R. A. Harvey, Radiological evidence of growth in children with acute leukemia treated with folic acid antagonists. Radiology 62, 61 – 64 (1954). 241. H. B. Nevinny, M. J. Krent, and E. W. Moore, Metabolic studies on the effects of methotrexate. Metabolism 14, 135 – 139 (1965). 242. G. E. Friedlander, R. B. Tross, A. C. Doganis, J. M. Kirkwood, and R. Baron, Effects of chemotherapeutic agents on bone. 1. Shortterm methotrexate and doxorubicin (Adriamycin) treatment in a rat model. J. Bone Jt. Surg. 66, 602 – 607 (1984). 243. K. P. May, S. G. West, M. T. McDermott, and W. E. Huffer, The effect of low-dose methotrexate on bone metabolism and histomorphometry in rats. Arthritis Rheum. 201 – 206 (1994). 244. F. Schmid, Osteopathien bei antiepileptischer Dauerbehandlung. Fortscher Med. 85, 381 (1967). 245. T. J. Hahn, Steroid and drug-induced osteopenia. In Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism (M. J. Favus, ed.), pp. 252 – 255. Raven Press, New York, 1993. 246. T. J. Hahn and L. R. Halstead, Anticonvulsant drug-induced osteomalacia: Alterations in mineral metabolism and response to vitamin D3 administration. Calcif. Tissue Int. 27, 13 – 18 (1979). 247. T. J. Hahn, B. A. Hendin, C. R. Scharp, V. C. Biosseau, and J. J. G. Haddad, Serum 25-hydroxycalciferol levels and bone mass in children on chronic anticonvulsant therapy. N. Engl. J. Med. 292, 550 – 554 (1975). 248. C. Christiansen, P. Rodbro, and P. Lund, Incidence of anticonvulsant osteomalacia and effect of vitamin D: Controlled therapeutic trial. Br. Med. J. 4, 695 – 701 (1973). 249. C. Christiansen, P. Rodbro, and C. T. Nielsen, Iatrogenic osteomalacia in epileptic children: A controlled therapeutic trial. Acta Paediatr. Scand. 64, 219 – 224 (1975). 250. P. Vestergaard, S. Tigaran, L. Rejnmark, C. Tigaran, M. Dam, L. Mosekilde, Fracture risk is increased in epilepsy. Acta Neurol. Scand. 99, 269 – 275 (1999). 251. T. J. Hahn, B. A. Hendin, C. R. Scharp, and J. G. Haddad, Effect of chronic anticonvulsant therapy on serum 25-hydroxycholecalciferol levels in adults. N. Engl. J. Med. 287, 900 – 904 (1972). 252. T. J. Hahn, S. J. Birge, C. R. Scharp, and L. V. Avioli, Phenobarbital induced alterations in vitamin D metabolism. J. Clin. Invest. 51, 741 – 748 (1972). 253. T. C. B. Stamp, J. M. Round, D. J. F. Rowe, and J. J. G. Haddad, Plasma levels and therapeutic effect of 25-hydroxycholecalciferol in epileptic patients taking anticonvulsant drugs. Br. Med. J. 4, 9 – 12 (1972). 254. M. J. Valimaki, M. Tiihonen, K. Laitinen, R. Tahtela, M. Karkkainen, C. Lamberg-Allardt, P. Makela, and R. Tunninen, Bone mineral density measured by dual-energy X-ray absorptiometry and novel markers of bone formation and resorption in patients on antiepileptic drugs. J. Bone Miner. Res. 9, 631 – 637 (1994). 255. J. D. Wark, R. G. Larkins, D. Perry-Keene, C. T. Peter, D. L. Ross, and J. G. Sloman, Chronic diphenylhydantoin therapy does not reduce plasma 25-hydroxy-vitamin D. Clin. Endocrinol. 11, 267 – 274 (1979). 256. R. S. Weinstein, G. F. Bryce, L. J. Sappington, D. W. King, and B. B. Gallagher, Decreased serum ionized calcium and normal vitamin D metabolite levels with anticonvulsant drug treatment. J. Clin. Endocrinol. Metab. 58, 1003 – 1009 (1984). 257. H. C. Koch, D. Kraft, and D. v. Herrah, Influence of diphenylhydantoin and phenobarbital on intestinal calcium transport in the rat. Epilepsia 13, 829 – 841 (1972).

SCHNEIDER AND SHANE 258. H. C. Harrison and H. E. Harrison, Inhibition of vitamin D-stimulated active transport of calcium of rat intestine by diphenylhydantoin phenobarbital treatment. Proc. Soc. Exp. Biol. Med. 153, 220 (1976). 259. T. J. Hahn, C. R. Scharp, C. A. Richardson, L. R. Halstead, A. J. Kahn, and S. L. Teitelbaum, Interaction of diphenylhydantoin (phenytoin) and phenobarbital with hormonal mediation of fetal rat bone resorption in vitro. J. Clin. Invest. 62, 406 – 414 (1978). 260. M. V. Jenkins, M. Harris, and M. R. Willis, The effect of phenytoin on parathyroid extract and 25-hydroxycholecalciferol-inducedbone resorption: Adenosine 3,5-cyclic monophosphate production. Calcif. Red. 16, 163 – 167 (1974). 261. J. W. Dietrich and R. Duffield, Effects of diphenylhydantoin on synthesis of collagen and noncollagen protein in tissue culture. Endocrinology 106, 606 – 610 (1980). 262. Y. Yamano, S. Sakane, J. Takamatsu, and N. Ohsawa, Estrogen supplementation for bone dematuration in young epileptic man treated with anticonvulsant therapy: A case report. Endocr. J. 46, 301 – 7 (1999). 263. L. Mosekilde and F. Melsen, Dynamic differences in trabecular bone remodeling between patients after jejuno-ileal bypass for obesity and epileptic patients receiving anticonvulsant therapy. Metab. Bone Dis. Relat. Res. 2, 77 – 82 (1980). 264. N. Takeshita, Y. Seino, H. Ishida, H. Tanaka, C. Tsutsumi, K. Ogata, K. Kiyohara, H. Kato, M. Nozawa, Y. Akiyama et al., Increased circulating levels of a-carboxyglutamic acid-containing protein and decreased bone mass in children on anticonvulsant therapy. Cacif. Tissue Int. 44, 80 – 85 (1989). 265. K. H. Lau, O. Nakade, B. Barr, A. K. Taylor, K. Houchin, and D. J. Baylink, Phenytoin increases markers of osteogenesis for the human species in vitro and in vivo. J. Clin. Endocrinol. Metab. 80, 2347 – 2535 (1995). 266. L. Tjellesen, L. Hummer, C. Christiansen, and P. Rodbro, Different metabolism of vitamin D2/D3 in epileptic patients treated with phenobarbitone/phenytoin. Bone 7, 337 – 42 (1986). 267. T. Ohishi, K. Kushida, M. Takahashi, K. Kawana, K. Yagi, K. Kwakami, K. Horiuchi, and T. Inoue, Urinary bone resorption markers in patients with metabolic bone disorders. Bone 15, 15 – 20 (1994). 268. F. Kubota, A. Kifune, N. Shibata, T. Akata, K. Takeuchi, S. Takahashi, M. Ohsawa, and F. Takama, Bone mineral density of epileptic patients on long-term antiepileptic drug therapy: A quantitative digital radiography study. Epilepsy Res. 33, 93 – 97 (1999). 269. R. D. Sheth, C. A. Wesolowski, J. C. Jacob, S. Penney, G. R. Hobbs, J. E. Riggs, and J. B. Bodensteiner, Effect of carbamazepine and valproate on bone mineral density. J. Pediatr. 127, 256 – 262 (1995). [see Comments] 270. H. S. Barden, R. B. Mazess, P. G. Rose, and W. McAweeney, Bone mineral status measured by direct photon absorptiometry in institutionalized adults receiving long-term anticonvulsant therapy and multivitamin supplementation. Calcif. Tissue Int. 31, 117 – 121 (1980). 271. N. Collins, J. Maher, M. Cole, M. Baker, and N. Collaghan, A prospective study to evaluate the dose of vitamin D required to correct low 25-hydroxyvitamin D levels, calcium and alkaline phosphatase in patients at risk of developing antiepileptic drug-induced osteomalacia. Q. J. Med. 78, 113 – 122 (1991).

CHAPTER 52

Transplantation Osteoporosis SOL EPSTEIN ELIZABETH SHANE

Roche Laboratories, Nutley, New Jersey 07110 College of Physicians and Surgeons of Columbia University, New York, New York 10032

I. Immunosuppression and Osteoporosis II. Immunosuppressive Agents III. Clinical Impact of Transplantation on Bone

IV. Evaluation of Candidates for Transplantation V. Management of Transplantation Osteoporosis References

I. IMMUNOSUPPRESSION AND OSTEOPOROSIS

H-ras and c-myc [2]. Recently it has been shown that cyclosporine also increases expression of transforming growth factor (TGF-), a fact which may be of particular relevance to its actions on bone [2]. However, in recent animal studies, administration of TGF- does not alter cyclosporine-induced bone loss [3]. Immunologically, cyclosporine inhibits T-cell proliferation and generation of antigen-specific cytotoxic T lymphocytes. A newer compound, tacrolimus (FK506) [4], is a macrolide that binds to a similar class of binding protein (FK binding protein) as cyclosporine. FK506 acts via the same mechanism as cyclosporine and, as will be discussed below, has similar effects on bone [5]. In many transplantation programs, FK506 is the drug of choice as the initial immunosuppressant or as salvage agent after liver transplantation. In addition to the calcineurin-phosphatase inhibitors, two other drugs are commonly used to prevent allograft rejection. Azathioprine, a derivative of mercaptopurine, is a purine antagonist and antiproliferative agent. Mycophenolate mofetil, recently approved by the Food and Drug Administration (FDA), replaces azathioprine in some centers.

The bone disease associated with immunosuppression formerly was restricted to that resulting from glucocorticoid therapy. However, the advent of organ transplantation has required newer forms of immunosuppression specifically designed to prevent organ rejection. The discovery of a class of drugs called calcineurin inhibitors constituted the biggest advance in the field of immunosuppression and organ transplantation in the past two decades. These drugs have played a prominent role in the prevention of organ rejection and prolongation of life in transplant recipients. The first of these compounds, cyclosporine, is a small fungal cyclic peptide. Its activity depends upon the formation of a heterodimer consisting of cyclosporine and its cycloplasmic receptor, cyclophilin. This cyclosporine – cyclophilin heterodimer then binds to calcineurin [1,2]. The cyclosporine – cyclophilin – calcineurin complex functions as a phosphatase inhibitor that inhibits genes expressing interleukin (IL)-2, receptors for IL-2, and the protooncogenes

OSTEOPOROSIS, SECOND EDITION VOLUME 2

327

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

328

EPSTEIN AND SHANE

The majority of transplantation centers use glucocorticoids, cyclosporine or FK506, and azathioprine or mycophenolate mofetil [2]. The effects of these immunosuppressives on the skeleton are the main focus of this chapter. Depending upon the organ transplanted, the experience and preference of the team of physicians, the country, and the severity and frequency of rejection episodes, these drugs are used in differing dosages and for variable lengths of time. Such variations in drug regimens may account in large part for differences in survival rates and incidence of drug related complications including osteoporosis and fracture. In general, most experienced physicians endeavor to use the lowest doses possible that will prevent rejection and minimize complications. In this regard, one approach to minimizing side effects is to use glucocorticoids for a very short period immediately after transplantation and then switch to monotherapy using either cyclosporine or tacrolimus (FK506). Another approach is to add to the regimen newer drugs that have been shown experimentally to prolong graft survival. This may permit the use of smaller doses of glucocorticoids and calcineurin phosphatase inhibitors. These newer drugs include sirolimus (Rapamycin) and mycophenolate mofetil, both of which have recently been approved by the FDA, or mizoribine and brequinar sodium (Table 1) [2].

II. IMMUNOSUPPRESSIVE AGENTS A. Glucocorticoids A detailed consideration of glucocorticoid-induced osteoporosis is found in Chapter 44. However, since glucocorticoids constitute an integral component of most posttransplantation regimens, a brief description of their effects on bone and mineral metabolism in the setting of organ transplantation is given here. The immunosuppressive properties of corticosteroids (glucocorticoids) are related to their inhibition of the expression of a variety of cytokines including IL-1, IL-2, IL-6, interferon , and tumor necrosis factor (TNF-) [2]. This inhibition presumably occurs via binding of the glucocorticoid – glucocorticoid receptor protein complex to a glucocorticoid-responsive element in the regulatory region of the target genes. Glucocorticoids also inhibit the IGF regulatory system, including IGF-I expression in osteoblasts, an effect that likely contributes to their inhibitory effect on bone formation [6]. In addition, glucocorticoids inhibit T-cell proliferation [2]. It is of great interest that cytokines IL-1, IL-6, TNF-, and interferon , that are suppressed by glucocorticoids, have been found to stimulate bone resorption [7]. These observations suggest that glucocorticoids, which inhibit these cytokines, must cause bone loss and fractures via other mechanisms. These mechanisms

TABLE 1

Immunosuppressive Agents

Established

Action on bone

Glucocorticoids

Inhibit formation. Promote resorption.

Calcineurin-phosphatase inhibitors

Promote formation and resorption.

Cyclosporine and analogs Tacrolimus (FK506) Sirolimus (Rapamycin)

No effect short term (rat)

Azathioprine

No effect short term (rat)

Methotrexate

No effect short term but decreases bone volume long term in the rat

Newer Agents Mycophenolate mofetil

No effect (rat)

Mizoribine

?

Deoxyspergualin

?

Brequinar Sodium

?

Liflunomide

?

Azaspirane

?

are considered to include: (i) inhibition of gastrointestinal calcium absorption and stimulation of renal calcium loss, both of which predispose to negative calcium balance and secondary hyperparathyroidism [8]; (ii) suppression of the hypothalamic – pituitary – gonadal axis which decreases gonadal steroidogenesis; (iii) direct suppression of osteoblast recruitment and osteoblast function, including inhibition of osteocalcin synthesis; (iv) decreased transcription and synthesis of skeletal growth factors such as IGF-I, TGF-, and fibronectin; (v) enhanced synthesis of collagenase in osteoblasts; and (vi) induction of apoptosis of osteoblasts and osteocytes both in vitro and in vivo [9].

B. Cyclosporines In recent years it has become clear that other immunosuppressants have effects on bone and mineral metabolism that could contribute to bone loss after organ transplantation. Cyclosporine A and its analogs are the most commonly used agents to prevent rejection, whether used alone or in conjunction with other drugs [2]. Cyclosporine, when administered to rats in doses comparable to or in excess of those used to prevent allograft rejection, causes rapid and severe cancellous bone loss [10]. The loss of bone volume is accompanied by a decrease in limb ash weight and has been observed in both male and female rats between 3 and 9 months of age [10,11]. In this model, cyclosporine has also

329

CHAPTER 52 Transplantation Osteoporosis

been shown to exaggerate bone loss in oophorectomized rats [12]. The bone loss proceeds in the absence of significant renal impairment. Elevations of serum osteocalcin and 1,25(OH)2D concentrations consistently accompany cyclosporine-induced bone loss in the rat [11,12]. The elevated 1,25(OH)2D concentrations are due to increased renal production and are not accompanied by changes in the serum levels of calcium or PTH [13]. The observation that cyclosporine does not cause bone loss in the athymic nude Tcell-deficient rat suggests that the effect of this drug on the skeleton is mediated by T lymphocytes [14]. In support of this hypothesis is the observation that cyclosporine H, which does not inhibit T-cell function or have immunosuppressant properties, does not cause bone loss in the rat [15]. These experimental findings have been confirmed repeatedly with the exception of a study by Orcel et al. in which cyclosporine did not cause vertebral bone loss in weanling rats [16]. However, this particular model may not be an ideal one in which to investigate this issue, as it is characterized by accelerated and predominant bone formation. Further clues to the mechanisms by which cyclosporine affects bone at the cellular level have been provided by recent ex vivo studies utilizing extracted bone and bone marrow mRNA. These investigations demonstrated that osteocalcin, IL-1, and IL-6 gene expression may be increased by exposure to cyclosporine [17]. Moreover, recent animal studies have revealed that the presence of PTH modifies cyclosporine-induced bone loss, since parathyroidectomized rats lose less bone than intact animals when exposed to cyclosporine [18]. Endothelin, which has been implicated both in the nephrotoxicity and hypertension induced by cyclosporine, does not appear to influence the bone disease [19]. Recently, the calcineurin gene and its isoforms have been identified in osteoclasts and extracted whole rat bone. However, the calcineurin gene is not altered by cyclosporine action in bone [20]. These studies emphasize the dominant role of the T lymphocyte in cyclosporine-induced bone loss, a concept recently confirmed by Zahner and colleagues [21].

TABLE 2

C. Azathioprine This anti-proliferative agent is frequently used in combination with cyclosporine and/or glucocorticoids after organ transplantation. Azathioprine does not appear to influence bone volume in the rat model over the short term [22]. However, azathioprine was associated with a decline in serum osteocalcin values, which may represent an inhibitory effect on osteoblast function at a particular stage of development [22]. The long-term consequences of azathioprine administration in this model are unknown. However, no adverse effects of azathioprine administration alone on bone mass have been reported in human subjects.

D. Tacrolimus (FK506) FK506 is a macrolide that blocks T-cell activation in a manner similar to that of cyclosporine [3]. FK506 has been shown to cause bone loss in the rat model similar to that which occurs with cyclosporine (Table 2). The bone loss is accompanied by the same biochemical and histomorphometric alterations. Therefore, the incidence of bone loss and fractures in organ transplant recipients managed with FK506 may not differ significantly from those managed with cyclosporine. FK506 is being used predominantly for liver transplantation. Liver transplantation, particularly in primary biliary cirrhosis patients managed with cyclosporine, produces a very high rate of fracture [23]. It is unclear whether FK506 will confer any benefit over cyclosporine with regard to fracture incidence.

E. Sirolimus (Rapamycin) Rapamycin is a macrocyclic lactone which is structurally similar to FK506 and binds to the same binding protein. Nevertheless, the mechanism by which rapamycin

Histomorphometry of Proximal Tibial Metaphyseal Cancellous Bone after Rats Were Sacrificed on Day 28

Osteoid perimeter (mm) Mineral apposition rate (m / day) Percentage eroded perimeter Longitudinal growth rate (m / day)

Control

Vehicle (1 ml / kg)

Cyclosporine (15 mg / kg)

FK506 (5 mg / kg)

13.17  1.17

18.56  2.05

24.93  2.64**

25.97  3.02**

1.31  0.12

1.07  0.07

1.70  0.15*

1.54  0.11

9.85  1.76*

14.20  2.53**

9.55  1.30

34.79  7.78

22.9  3.62***

5.44  0.59

6.39  0.40

34.57  8.70

32.63  3.55

31.97  5.79

Rapamycin (2.5 mg / kg) 23.33  4.83* 1.81  0.12**

Note. The measured area was between 1 and 4 mm distal to the growth plate – metaphyseal junction. All data represent mean values  SE (n  8 or 9). *P  0.05 compared to control (group A). **P  0.01 compared to control (group A). ***P  0.001 compared to control (group A).

330

EPSTEIN AND SHANE

induces immunosuppression is distinct from both FK506 and cyclosporine. Studies in the rat model demonstrate that rapamycin has no adverse effect on bone (Table 2) [24]. However, in large doses, rapamycin causes hypogonadism in male rats that resulted in severe osteopenia with fractures [24]. This effect appears to be independent of alterations in the calmodulin – calcineurin phosphatase pathway. When combined with low-dose cyclosporine, rapamycin also has been shown to be bone sparing in rat studies (25). Thus newer combinations of immunosuppressive agents provide hope for achieving adequate immunosuppression while protecting the skeleton.

F. Methotrexate This folate antagonist is used in high doses to treat malignancies and more recently in low dose pulses to treat rheumatoid and other inflammatory arthritides. Short-term, 2-week therapy with methotrexate in the rat, at doses equivalent to those used clinically for arthritis, produced no change in trabecular bone volume. However, serum osteocalcin concentrations did decrease. Prolonged 16-week administration in the rat decreased both serum alkaline phosphatase activity and osteocalcin levels. Bone histomorphometry revealed decreased formation, increased resorption and diminished bone mass [26]. Prospective controlled clinical studies of the effects of methotrexate on bone have not been reported. Cross-sectional studies have produced contradictory results, with some studies demonstrating bone loss and others revealing no change [27,28]. Obviously, such studies can be influenced by the underlying disease for which methotrexate is prescribed.

G. Mycophenolate Mofetil and Other Drugs Mycophenolate displays no deleterious effect on bone in the rat model and should pose no problem when administered as an antirejection therapy [29]. There is little or no information available on the effects of other immunosuppressant agents such as mizoribine, deoxyspergualin, brequinar sodium, liflunomide, and azaspirane on bone (Table 1).

III. CLINICAL IMPACT OF TRANSPLANTATION ON BONE The majority of candidates for organ transplantation have risk factors that predispose toward osteopenia. These factors include general debilitation, loss of mobility, poor nutrition, cachexia and exposure to certain drugs. In addition, many women are postmenopausal and both premenopausal women and men with chronic illness may have

gonadal dysfunction. When the disease is present during childhood or adolescence, as is the case with cystic fibrosis, there may be interference with the attainment of peak bone mass. After transplantation, episodes of rejection, usually reversed by large doses of glucocorticoids and cyclosporine, compound the bone loss. Consideration of particular issues related to transplantation of specific organs follows.

A. Kidney Transplantation Patients who undergo renal transplantation have severe chronic renal insufficiency or end-stage renal disease (ESRD) and most have been dialyzed for varying intervals before transplantation. Preexisting bone disease is almost universal in this population. Renal osteodystrophy is a general term that encompasses all the bone histological alterations that may occur in uremic patients [30]. In a given individual, there may be evidence of hyperparathyroidism with or without osteitis fibrosa, osteomalacia, low turnover, or a dynamic bone disease due to aluminum accumulation (or other as yet poorly understood factors), osteosclerosis particularly of the vertebrae, and -macroglobulin amyloidosis. Many patients will have “mixed” renal osteodystrophy, a combination of one or more of the aforementioned lesions. Other factors that may affect the skeletal integrity of patients with ESRD include type I diabetes, hypogonadism secondary to uremia, and diseases such as systemic lupus erythematosus. Several drugs used routinely in the management of patients with renal disease, such as loop diuretics and aluminum containing phosphate binders, can also affect bone and mineral metabolism. In addition, some patients who are candidates for transplantation may have had previous exposure to glucocorticoids or cyclosporine as therapy for immune complex nephritis or other diseases and thus may already have sustained significant bone loss prior to transplantation. After renal transplantation, several investigations have documented a decline in bone mass [31,32] and increased fracture rate [33]. While the greatest insult to the skeleton is related to glucocorticoid and cyclosporine exposure, persistent hyperparathyroidism likely contributes to the declining bone mass. The rate of bone loss is greatest during the first 6 months after transplantation and at sites where cancellous bone predominates, such as the lumbar spine [31,32]. Indeed, some reports suggest an increase in bone mass at the radius, a site that consists predominantly of cortical bone [32]. The rate of lumbar spine bone loss varies between 6 and 18% per year, but tends to be somewhat lower than that observed for other transplanted organs such as the liver. This lower rate of bone loss may be because lower doses of glucocorticoids and cyclosporine are used for immunosuppression after kidney transplantation than after transplantation of

331

CHAPTER 52 Transplantation Osteoporosis

other solid organs [31,32]. Moreover, rejection is more easily diagnosed and therefore detected earlier than with other organs and lower doses of immunosuppressive drugs are more effective in reversing rejection when it is diagnosed at an earlier stage. It is accepted that the adverse skeletal effects of glucocorticoids and cyclosporine experimentally and clinically are dependent upon both dose and duration of exposure [34 – 36]. Julian et al. reported a decrease in lumbar spine BMD of 6.8  5.6% at 6 months and 8.8  7.0% at 18 months after transplantation [31]. Moreover by 18 months, bone density was below the “fracture threshold” in 10 of 17 patients. There appears to be a gender difference in the site at which bone is lost [32,37]. Men have been shown to lose more bone at the proximal femur than women in the first few months after transplantation. In contrast, radial bone density increased in men at 6 months posttransplantation but not in women. Bone biopsies performed prior to transplantation revealed typical changes of hyperparathyroidism. However, by 6 months after transplantation, the histomorphometric picture was more typical of a glucocorticoid effect, demonstrating osteoblast dysfunction and decreased mineral apposition [31]. Unfortunately, bone biopsies were not performed at 18 months when glucocorticoid doses were lower. Thus, there were no data either to exclude or to incriminate cyclosporine as a contributing factor to the bone loss. However, there is some evidence in the literature to support a role for cyclosporine in the pathogenesis of the high turnover state often apparent in renal transplant recipients by 1 year after renal transplantation [38 – 40]. The increased frequency of cyclosporine use as monotherapy in renal transplant patients should clarify the relative role of glucocorticoids and cyclosporine in the pathogenesis of posttransplantation bone loss. Recently 45 renal transplant recipients were evaluated with quantitative histomorphometry 120  70 months after transplantation. Those treated with cyclosporine monotherapy demonstrated significantly lower mineral apposition rates than those treated with azathioprine and prednisone [41]. In general, vertebral fractures are less common after kidney transplantation than after other types of organ transplantation. However, appendicular fractures are extremely common, particularly in patients transplanted for diabetic nephropathy, in whom the incidence of fracture has been reported as high as 45% [33]. While the reasons for this are not clear, microvascular disease and neuropathy (lack of proprioception and pain sensation) may affect fracture incidence by increasing the risk of falls. In addition, bone mass may be below normal in patients with type 1 diabetes mellitus (see Chapter 51) [42] even before renal transplantation, thus placing such patients at higher risk for fracture after transplantation. Avascular necrosis occurs commonly after renal transplantation [43 – 49]. The incidence in children is 6% [43, 44] and in adults is 8%. The hip is the most commonly

affected site. While the association of avascular necrosis with glucocorticoids is well established, cyclosporine has also been incriminated in producing avascular necrosis and bone pain of the hip and other weight bearing bones, such as the knees [50]. The known vasospastic or vasoconstrictive properties of cyclosporine may contribute to the development of avascular necrosis.

B. Cardiac Transplantation Osteoporosis and fractures constitute a major cause of morbidity after cardiac transplantation [51 – 56]. In early cross-sectional studies, the prevalence rate of vertebral fractures in cardiac transplant recipients ranged between 18 and 50% and moderate to severe bone loss was present in a substantial proportion of subjects at both lumbar spine and the femoral neck [52,57]. Risk factors that may predispose patients with end-stage cardiac failure to bone loss even before transplantation include exposure to tobacco, alcohol and loop diuretics, physical inactivity, hypogonadism, and anorexia which may contribute to dietary calcium deficiency. Hepatic congestion and prerenal azotemia may also affect mineral metabolism. Although the average bone mineral density of patients awaiting cardiac transplantation may not differ significantly from normal, it has been observed that approximately 8 to 10% fulfill World Health Organization criteria for osteoporosis and 40 to 50% have osteopenia or low bone mass [58]. Prospective longitudinal studies have documented rates of bone loss ranging from 2.5 to 20%, predominantly during the first year after transplantation [51,55,59 – 61]. Biochemical changes after cardiac transplantation include sustained increases in serum creatinine and decreases in 1,25-dihydroxyvitamin D concentrations [59]. On average, serum testosterone concentrations decrease in men with recovery by the sixth posttransplant month [56,59]. Serum osteocalcin falls precipitously and there is a sharp increase in markers of bone resorption (hydroxyproline and pyridinium crosslink excretion) during the first three months with return to baseline levels by the sixth month [56,59]. This biochemical pattern coincides with the period of most rapid bone loss and highest fracture incidence and suggests that the early posttransplant period is associated with uncoupling of formation from resorption. It is of interest that at least two studies of subjects treated with high doses of glucocorticoids alone confirm the decrease in serum osteocalcin but found no increase in markers of bone resorption [60,61]. This suggests that the pathogenesis of early bone loss after transplantation may be related both to the well-known inhibitory effects of glucocorticoids on bone formation, and to an effect of cyclosporine A or some other agent to increase bone resorption. There is also evidence for a high bone turnover state later in the

332

EPSTEIN AND SHANE

posttransplant course perhaps due to cyclosporine, characterized by elevations in both serum osteocalcin and urinary excretion of resorption markers [62,63]. In a recent crosssectional study, Eastell and colleagues [63] evaluated 50 men, ranging from 0.5 to 47 months after cardiac transplantation. They concluded that bone turnover is increased after cardiac transplantation, and that this increase is due in part to secondary hyperparathyroidism related to renal impairment. Thus biochemical changes later in the posttransplant course may be mediated, at least in part, by cyclosporine Ainduced renal insufficiency. The pattern of bone loss after cardiac transplantation is similar to that observed after renal [31] or liver transplantation [23,64 – 66]. Despite the predilection for glucocorticoids to affect the cancellous bone of the vertebrae to a greater extent than other sites, Shane et al. have reported that there is as much or more bone loss at the hip. Moreover, while bone loss at the lumbar spine slows or stops after the first 6 months, femoral neck bone loss continues during the second half of the first year after transplantation [59]. Moreover, these investigators in a recent longitudinal study demonstrated that 36% of patients (54% of the women and 29% of the men) suffered one or more fractures in the first year despite daily supplementation with calcium (1000 mg) and vitamin D (400 IU) [53]. Although the majority of fractures affected the vertebral bodies, two patients suffered multiple rib fractures and two had fractures of the femoral neck. In women, low pretransplant femoral neck bone density predicted the risk of vertebral fracture after transplantation. In men, however, it was the rate of bone loss after fracture rather than the pretransplant bone density that was associated with fracture risk. Also of note was the observation that patients with normal bone mass also fracture frequently [59]. A European study of 159 cardiac transplant recipients reported similar findings (54). This study underscores the need for a complete bone evaluation and bone mass measurements prior to, or immediately after, transplantation, as well as aggressive intervention to prevent bone loss and fractures in all patients regardless of age, sex, or pretransplant bone density. There are very few longitudinal data available on the pattern of bone loss during the second year after transplantation. However, data from Shane and colleagues suggests that the rate of bone loss slows or stops in the majority of patients, with some recovery at the lumbar spine noted during the third year of observation [59]. Bone loss also slows at the hip after the first year; however, in contrast to the spine, there has been no significant recovery by the fourth posttransplant year [59].

C. Liver Transplantation Patients with liver failure also have risk factors that may predispose to fracture after transplantation. In addition, the

doses of immunosuppressive drugs used are much larger than those commonly employed after renal transplantation. Moreover, since the liver plays a major role in cyclosporine metabolism, hepatic dysfunction may also influence its serum concentrations, possibly predisposing to cyclosporine toxicity. These factors may account for the observation that liver transplant recipients have higher rates of bone loss than cardiac and renal transplant recipients during the first year after transplantation [23,64 – 66]. However, the type of liver disease may also be an important risk factor for osteoporosis. In one study, 13 of 20 women with primary biliary cirrhosis, a disease characteristically associated with low turnover osteoporosis, suffered atraumatic fractures of vertebra, ribs, hips, and long bones during the first year after transplantation [23]. Patients with alcoholic cirrhosis may also have very low bone mass prior to transplantation, perhaps due to hypogonadotrophic hypogonadism and excess iron deposits in the skeleton. Eastell et al. reported that despite the high incidence of fractures in liver transplant recipients, bone mass recovers and bone histology normalizes with increasing survival time after transplantation [23]. This, however, has not been a uniform finding and other studies have found continued losses rather than recovery [66]. Depending upon the bone density at the time of transplantation, these patients may always be at risk for fractures as survival rates and duration increase. As is the case with renal and cardiac transplantation, the independent role of glucocorticoids and calcineurin phosphatase inhibitors in the pathogenesis of bone disease in liver transplant patients is difficult to assess since single drug therapy is uncommon. The mechanism of bone loss after liver transplantation has been studied by bone biopsy in 21 patients, who underwent tetracycline labeling and transiliac crest bone biopsy prior to and 3 months after transplantation. Before transplantation, a low turnover state was observed, with decreased wall width and erosion depth. Postoperative biopsies showed high turnover with increased formation rates and activation frequency and a trend toward increased indices of resorption [68]. In an earlier study, these investigators documented a significant increase in parathyroid hormone concentrations after liver transplantation [69]. While increased PTH could account for these histomorphometric findings, similar effects are observed in animals treated with calcineurin inhibitors without a rise in PTH concentrations.

D. Lung Transplantation Since the last edition of this book, there have been several reports of skeletal complications in lung transplantation candidates and recipients. As with any chronic illness severe enough to require transplantation for survival, patients who undergo lung transplantation have predisposing

333

CHAPTER 52 Transplantation Osteoporosis

factors for low bone mass prior to transplantation. In particular, tobacco exposure, chronic hypoxemia, immobility, glucocorticoid use, and certain underlying diseases such as cystic fibrosis are common in candidates for lung transplantation. The prevalence of osteoporosis (bone mineral density 2 standard deviations below aged-matched controls) varies between 55% and 66% at the spine and up to 78% for the hip [70,71]. The prevalence of vertebral fracture in candidates before transplantation is also high, varying between 25 and 29% [70, 71]. Adults with cystic fibrosis are probably at even greater risk for bone disease than other patients who undergo lung transplantation. A recent study of 70 adults with cystic fibrosis observed virtually all to have either osteoporosis or low bone mass. Reported fracture rates were 2-fold greater than the general population, while rib and vertebral fractures were 10-fold and 100-fold greater, respectively. The strongest predictors of BMD were body mass index, cumulative prednisone dose and pubertal age [72]. The incidence of new fractures after lung transplantation is also very high [73,74], even in patients who receive antiresorptive therapy. It has recently been reported that 11 (10 women) of 30 lung transplant recipients (37%), all of whom had received daily calcium (1000 mg) and vitamin D (800 IU) and antiresorptive therapy (injectable calcitonin, cyclic etidronate, pamidronate, or alendronate) sustained a total of 54 new fractures during the first year after transplantation [74]. The average time to first fracture was 4.5 months. The most common fracture sites were ribs, vertebrae, pelvis, and sacrum [74,75]. Moreover, despite antiresorptive therapy, 50% had a significant decrease in BMD. Biochemical markers of bone turnover were significantly higher in those who lost bone and in those who fractured. Pediatric lung transplantation is becoming more frequent and both osteoporosis and reduced growth velocity can be expected in these children [76]. Similar to other transplanted organs, a high bone turnover state with elevated osteocalcin concentrations has been reported after lung transplantation [77].

E. Bone Marrow Transplantation Bone marrow transplantation is performed with increasing frequency and for expanding indications. Low BMD has also been reported in patients after bone marrow transplantation [78], probably related to both pre- and posttransplant factors. In preparation for transplantation, patients receive myeloablative therapy (alkylating agents and/or total body irradiation) and commonly develop profound and frequently permanent hypogonadism, which almost certainly contributes to bone loss. Two recent studies have documented low BMD in hypogonadal women after bone marrrow transplantation [79] and that hormone replacement

therapy is associated with significant increases in BMD [80]. A study of 9 adults undergoing high-dose glucocorticoid and CSA therapy for graft-versus-host disease (GVHD) observed significant bone loss in most patients [81]. A more recent study documented that low bone mass antedates bone marrow transplantation, particularly in subjects with prior glucocorticoid exposure, and that postransplant bone loss is particularly severe in patients who undergo allogeneic bone marrow transplantation, probably because of their increased propensity for GVHD [82]. In these patients, the rate of bone loss was 11.7% at the femoral neck and 3.9% at the spine [82]. Biochemical data suggest that bone loss after marrow transplantation is a high turnover state with increases in resorption markers and alkaline phosphatase activity [82 – 84].

IV. EVALUATION OF CANDIDATES FOR TRANSPLANTATION There are now abundant data documenting the high prevalence of bone disease in candidates for all types of transplantation. Therefore a complete skeletal evaluation is indicated. This evaluation should occur prior to transplantation so that potentially treatable abnormalities of bone and mineral metabolism may be addressed and the skeletal condition of the patient optimized before transplantation. This evaluation should include a full history with particular emphasis upon risk factors for osteoporosis such as family history, medical conditions (thyrotoxicosis, renal disease, rheumatological, and intestinal disorders), poor lifestyle choices (physical inactivity, dietary calcium and vitamin D deficiency, excessive caffeine and alcohol intake, tobacco use), and exposure to drugs (diphenylhydantoin, lithium, loop diuretics, glucocorticoids, prolonged and large doses of heparin, thyroid hormone). Additional risk factors important in women include premature menopause, postmenopausal status, a history of anorexia nervosa or prolonged episodes of amenorrhea. In men, it is important to exclude hypogonadism. A physical examination should focus upon diseases that predispose to osteoporosis such as hypogonadism, thyrotoxicosis and Cushing’s syndrome. Risk factors for falling (sight, hearing, balance, and muscle strength) should also be assessed. Bone density of the spine and hip and plain radiographs of the thoracic and lumbar spine are the most important tests to obtain prior to transplantation. The biochemical evaluation should include a chemistry panel, thyroid function tests, intact parathyroid hormone (PTH) and vitamin D metabolites, and total and free testosterone, follicle-stimulating hormone (FSH), and lutenizing hormone (LH) concentrations in men. Markers of bone formation (serum osteocalcin and bone-specific alkaline phosphatase) and resorption (urinary deoxypyridinoline or N-telopeptide

334

EPSTEIN AND SHANE

TABLE 3 Evaluation of the Candidate for Organ Transplantation • History and physical examination with attention to risk factors for osteoporosis • Bone densitometry by dual-energy X-ray absorptiometry • Thoracic and lumbar spine radiographs • Serum calcium, parathyroid hormone, 25-hydroxyvitamin D, thyroid function tests • In men, serum total and/or free testosterone, FSH, and LH • Urine for calcium and markers of bone resorption (Optional)

excretion) can also be measured (Table 3) [85]. In the interest of cost control, it could be argued that the battery of biochemical tests may be unnecessary if the bone density measurement is normal and calcium and vitamin D supplementation are planned. However, if the pretransplant bone density is low, the biochemical evaluation can alert the physician to the etiology of low bone mass and guide appropriate therapy. After transplantation, serum and urine indices of mineral metabolism are less crucial. However, monitoring may be useful in detecting conditions that may contribute to bone loss (vitamin D deficiency or renal insufficiency with secondary hyperparathyroidism, hypercalciuria, or hypercalcemia related to pharmacologic doses of vitamin D). Measurement of bone density, on the other hand, remains important and should be performed at 6-month intervals for the first 2 years and annually thereafter. Bone biopsy may be necessary after renal transplantation since many experts remain reluctant to use bisphosphonates in patients with adynamic bone disease, despite lack of any data either to support or exclude the use of these drugs in this setting. Although transiliac crest bone biopsy remains a research tool, more histomorphometric studies would be very helpful in confirming theories of the pathogenesis of transplantation osteoporosis.

TABLE 4

V. MANAGEMENT OF TRANSPLANTATION OSTEOPOROSIS The general principles for the treatment of transplantation osteoporosis are similar to those for any type of osteoporosis. Therapy may be initiated during the waiting period before transplantation or in the initial 6 to 12 months after transplantation. In addition, the long-term transplant recipient with established osteoporosis and/or fractures should not be neglected. It must be emphasized that prevention of bone loss that accompanies transplantation is probably more effective in reducing morbidity than treatment of established osteoporosis in the transplant recipient. During the waiting period before transplantation, rehabilitation therapy should be prescribed as tolerated to maximize conditioning and physical fitness. In general, calcium supplementation should be prescribed at doses of 1 – 2 g per day, depending upon age, gender, menopausal status, and dietary intake. Either calcium citrate or calcium carbonate is acceptable; however, the carbonate form should be taken with food to enhance absorption and may cause constipation. All patients should receive the Recommended Daily Allowance of vitamin D (400 – 800 IU daily). Hormone replacement therapy should be considered in all postmenopausal women, as well as in premenopausal amenorrheic women where there are no contraindications to such therapy. Hypogonadal men should also be offered testosterone replacement. Generally accepted guidelines for gonadal hormone replacement should apply to these patients. After transplantation, pharmacologic strategies should be instituted immediately to prevent bone loss and fractures. It must be emphasized that very few controlled prospective studies have been reported that support the use of specific therapies (Table 4). The recommendations described in this chapter are based upon such data as exist, as well as experience in similar clinical situations and supportive experimental evidence.

Recommendations for Management of Organ Transplant Recipients

• Encourage transplant physicians to use the lowest dose of glucocorticoids possible and to consider alternate therapies for rejection (e.g., OKT3). • Obtain bone mineral density routinely in patients accepted for transplantation and to refer for evaluation/therapy all patients with low bone mass (T score between 1.0 and 2.5) or osteoporosis (T score  2.5). • Ensure calcium intake of 1500 mg daily both before and after transplantation. • Ensure a vitamin D intake of 400 – 1000 IU, or as needed to maintain serum 25-OHD concentrations in the upper half of the normal range. • Encourage participation in a physical rehabilitation program both before and after transplantation. • Replace gonadal steroids in hypogonadal women and men. • Begin antiresorptive therapy, preferably a bisphosphonate, before transplantation in patients with antecedent osteoporosis or low bone mass. • Begin antiresorptive therapy, preferably a bisphosphonate, immediately or as soon as possible after transplantation in patients with normal or low bone mass and continue for at least the first posttransplant year. • Measure BMD at 6-month intervals for the first 2 years after transplantation.

335

CHAPTER 52 Transplantation Osteoporosis

A. Vitamin D and Analogs Administration of vitamin D or its analogs is usually recommended after transplantation. Sambrook et al. [86] found calcitriol to be efficacious in patients with glucocorticoid-induced osteoporosis. Although the role of vitamin D and its analogs in transplantation osteoporosis remains unclear, these investigators recently reported, in abstract form, that calcitriol (0.5 – 0.75 g/day) prevented spine and hip bone loss during the first 6 months after heart or lung transplantation and was as effective as cyclic etidronate (87). Whether vitamin D should be prescribed as the parent compound in doses of 50,000 IU once a week, as calcidiol 25 g daily [88], or as calcitriol 0.25 to 0.75 g daily, or even at all is not known. Hypercalcemia and hypercalciuria are the major side effects of therapy with these agents. Either may develop suddenly and at any time during the course of treatment. Thus, frequent urinary and serum monitoring may be required. If hypercalcemia occurs, it must be recognized and reversed promptly because of the adverse effects on renal function and the life-threatening potential of a severely elevated serum calcium concentration. Supplemental calcium and any vitamin D preparations must be discontinued until calcium values normalize. Although one may be tempted to permanently discontinue pharmacologic doses of vitamin D or its metabolites in view of the necessary serial monitoring and potential dangers, it seems reasonable to recommence therapy at a lower dose. The exact mechanism by which vitamin D and its analogs may influence posttransplantation bone loss is uncertain. They may overcome glucocorticoid-induced decreases in intestinal calcium absorption, reduce the potential for secondary hyperparathyroidism, promote differentiation of osteoblast precursors into mature cells (see Chapter 2), or influence the immune system and potentiate the immunosuppressive action of cyclosporine [89,90]. The last of these potential mechanisms is of particular interest because cyclosporine increases both production and serum concentrations of 1,25(OH)2D in the rat and mouse [13]. However, in humans, no evidence has been found yet for a similar effect and, in fact, serum concentrations of 1,25(OH)2D have been

TABLE 5 Potential Therapies for Transplantation Osteoporosis • Calcium • Vitamin D and analogs • Estrogen • Testosterone • Calcitonin • Bisphosphonates • Fluoride

shown to fall after cardiac transplantation [59]. Vitamin D and its analogs may also promote mineralization as calcitriol interacts with the vitamin D-responsive element on the osteocalcin gene [91]. An increase in osteocalcin synthesis might improve recruitment of bone cells or the incorporation of calcium into the bone matrix. In the rat, 1,25(OH)2D administration alleviated the osteopenia produced by cyclosporine; however, hypercalcemia was an accompaniment [92]. In summary, given the requirement for serial monitoring and the narrow therapeutic window with regard to hypercalcemia and hypercalciuria, we regard pharmacologic doses of vitamin D and its analogs as adjunctive rather than primary therapy for the prevention and treatment of transplantation osteoporosis.

B. Estrogens In postmenopausal women or premenopausal women with amenorrhea or irregular menses, estrogen replacement should be recommended provided that there are no contraindications. The dose is the same as that used for prevention or therapy of postmenopausal osteoporosis and may be given either orally or transdermally. In women with an intact uterus, progesterone must be prescribed in addition to prevent endometrial cancer. Continuous rather than cyclical therapy is preferred after transplantation, as estrogen enhances hepatic metabolism of cyclosporine (and presumably FK506) and theoretically may compromise immunosuppression. Whether this occurs in patients is not known. For patients who cannot take estrogens, tamoxifen [93] or newer selective estrogen receptor modulators (SERMs) such as raloxifene [94] may well be a suitable alternative. Although no trials of these drugs in organ transplant recipients have been published, raloxifene has been shown to reduce cyclosporine-induced bone loss in the rat model [95]. The hyperlipidemia produced by glucocorticoids, cyclosporine, and its analogs may also be ameliorated by estrogen, tamoxifen, and raloxifene, although no systematic studies have been conducted that address this particular issue. Few trials have evaluated the efficacy of estrogen in preventing bone loss and fractures after transplantation. Therefore, the recommendation to prescribe estrogen for this purpose is based in part upon the observation that 17estradiol prevents cyclosporine-induced bone loss in the oophorectomized rat [96]. In addition, a wealth of clinical data supports the protective effect of estrogens on the skeleton (see Chapter 69) and one cross-sectional study suggests that estrogen and progesterone therapy are associated with higher bone density in women on glucocorticoids [8]. The exact mechanism by which estrogen protects the skeleton is unknown. However, estrogen is asociated with inhibition of bone resorbing cytokines, such as IL-1, IL-6, and IL-11

336

EPSTEIN AND SHANE

(see Chapters 13, 36, and 41). Recent work implicates IL-1 [17] as a potential candidate in the pathogenesis of high turnover bone loss in the cyclosporine-treated rat, making it conceivable that estrogen may modify bone loss in this model by inhibiting this cytokine. In bone marrow transplantation, in which hypogonadism is a predominant feature, 12 months of hormone replacement therapy (HRT) was associated with significantly increased bone density in a small number of women, without adversely affecting liver enzymes [80]. There may also be an advantage to combining HRT with a bisphosphonate, particularly in light of a recent study [97] in which the addition of alendronate in a group of postmenopausal women on stable doses of HRT caused a significant increase in BMD.

C. Testosterone Men with serious chronic illnesses, such as hepatic, renal or cardiac failure, are commonly hypogonadal. In addition, high doses of both glucocorticoids [8] and cyclosporine (greater than 15 mg/kg body weight) suppress the hypothalamic – pituitary – gonadal axis and produce hypogonadism [98, 99]. Testosterone deficiency, both before and after transplantation, may contribute to the risk of bone loss and fractures [55, 56,59]. However, given the complexity of the clinical situation after transplantation, it is not possible to ascertain whether testosterone deficiency is an independent risk factor for fracture or bone loss. In any event, since androgen deficiency is known to cause low bone mass in men, it is not unreasonable to prescribe testosterone in men who are truly hypogonadal. Testosterone therapy is not without risks. Of special concern is the potential for induction or exacerbation of hyperlipidemia in patients already prone to accelerated atherosclerosis from hypertension, diabetes, glucocorticoid, and cyclosporine drug therapy [1]. In addition, prostatic hypertrophy and liver abnormalities are side effects of androgen therapy. However, after successful transplantation serum testosterone levels frequently normalize so that androgen therapy may only be required as a temporary measure [55,56,59]. Thus, it can be argued that temporary administration of testosterone after transplantation constitutes physiologic hormone replacement rather than pharmacological therapy. The risks of prostatic hyperplasia and liver abnormalities may be minimized by administering testosterone transdermally, rather than by injection. If testosterone therapy is recommended, patients must be cautioned about potential risks and benefits. Monitoring should include monthly measurement of serum lipids and hepatic enzymes and regular prostate examinations. Whether testosterone replacement with calcium supplements and vitamin D is sufficient to prevent transplantation osteoporosis is not known. However, it would be more prudent to include additional antiresorptive therapy to protect the skeleton.

D. Calcitonin Calcitonin has long been used to treat Paget’s disease of bone, a disease characterized by focal areas of high bone turnover. In the therapy of osteoporosis, calcitonin has been shown to increase bone density in patients with high turnover osteoporosis (see Chapter 73). Moreover, this drug may have an analgesic effect upon acutely painful fractures and chronic pain due to multiple vertebral fractures. Despite many years of experience, the optimum dose, route of administration, and efficacy of continuous versus intermittent dosing remain unclear. Both injectable and inhaled calcitonin has been used successfully to treat glucocorticoidinduced bone loss in humans [100]. While its use in transplant recipients has not been established by controlled trials, experimental work in the rat model has demonstrated that cyclosporine-induced bone loss can be prevented by calcitonin [101]. Based on the results from clinical studies of patients with postmenopausal and glucocorticoid-induced osteoporosis and experimental rat studies, it is reasonable to use calcitonin to protect against transplantation osteoporosis. However, literature on the use of calcitonin in preventing bone loss and fractures after transplantation is not consistent. The usual practice is to prescribe synthetic salmon calcitonin, 100 units daily by subcutaneous injection or intranasal calcitonin (100 – 200 IU) as soon as immunosuppressive therapy is begun. This intranasal dose varies according to the organ transplanted [102,103]. Valero et al. administered either injectable calcitonin or etidronate to liver transplant recipients and found that lumbar spine bone density increased by 6 – 8% with no difference in effi cacy between the drugs [104]. In contrast, other investigators have not found calcitonin to be particularly effective [103,105].

E. Bisphosphonates Bisphosphonates, which act by inhibiting osteoclastic bone resorption, have been used successfully to prevent and treat glucocorticoid-induced bone loss. Published studies include first-, second-, and third-generation bisphosphonates such as Etidronate, Pamidronate, Tiludronate and, most recently, Alendronate [106 – 108]. Alendronate, approved by the FDA in 1995 for the treatment of postmenopausal osteoporosis (see Chapter 72) has also recently been approved for glucocorticoid-induced osteoporosis. Because transplantation osteoporosis can be considered one form of glucocorticoidinduced osteoporosis and as cyclosporine and tacrolimus-induced bone loss are characterized experimentally by increases in both formation and resorption, bisphosphonates may prove to be highly successful in prevention of transplantation osteoporosis. Some [109 – 111], although not all [112], studies suggest that bisphosphonates can prevent bone loss

337

CHAPTER 52 Transplantation Osteoporosis

and fractures after transplantation. A single intravenous dose of Pamidronate (60 mg), given during the first 2 weeks weeks after transplantation and followed by cyclical Editronate for the remainder of the first year, prevented lumbar spine and femoral neck bone loss and significantly reduced fracture incidence compared to patients who received only calcium and 400 IU of vitamin D [110]. Similar experience has been reported with intravenous Pamidronate in kidney [109] and even more dramatically in liver [111] transplant recipients where fractures were prevented in the group treated with Pamidronate. Clinical trials are currently underway with Alendronate in prevention of osteoporosis after cardiac, liver, and renal transplantation and our own clinical experience suggests that this drug is effective in this setting [113]. At present, these drugs constitute the most promising approach to the prevention of this often crippling form of osteoporosis. As with other forms of therapy, many issues remain to be resolved, such as whether continuous or intermittent (cyclical) therapy should be used, at what level of renal impairment these drugs should be avoided, whether impaired mineralization will develop after long-term therapy, or whether these agents will interfere with fracture healing. Based on preliminary data with the newer bisphosphonates [114], the latter concerns may not prove to be an issue.

F. Fluoride Despite the concerns related to fluoride therapy in the treatment of osteoporosis (see Chapter 74), this drug continues to be used extensively outside the United States. European investigators have demonstrated that administration of disodium monofluorophosphate to patients with glucocorticoid-induced osteoporosis resulted in a 63% increase in trabecular bone mass [115]. Meys et al. [88] used disodium monofluorophosphate (26.4 mg elemental fluoride) together with 1 g of elemental calcium and 25 mg (1000 IU) of calcidiol in cardiac transplant patients and compared these patients to another group who received the same dose of calcium and calcidiol without the fluoride compound. After 12 and 24 months of therapy, there was no decline in lumbar spine bone mass in those treated with calcium and calcidiol. In contrast, the group that received fluoride demonstrated an increase in lumbar bone density of 12.5% after 12 months and 29.5% after 24 months, respectively. Side effects were observed in 10 of 57 patients (gastric intolerance in 5 patients and lower limb pain in 5 patients). No hip, long-bone, or vertebral fractures were seen in the second year of treatment in either group. This study, although interesting, was open and uncontrolled, and no bone biopsy data were reported. Thus, the quality of the new bone was unknown. While encouraging, controlled trials are still necessary to assess the true value of this particular fluoride preparation in transplantation osteoporosis.

G. Newer Therapeutic Options Discussion of new advances in immune therapy which will prevent organ rejection but spare bone are beyond the scope of this chapter. At the present time, using the lowest possible doses of glucocorticoid and calcineurin phosphatase inhibitors offers the best option. Currently, the most exciting areas of investigation involve agents that stimulate bone formation (growth hormone, growth hormone releasing peptide, PTH or its analogs, the IGF family, including the IGF binding proteins, prostaglandins, particularly of the E series [116], and the TGF- superfamily, including bone morphogenic protein). Such drugs are of theoretical value, particularly in the setting of glucocorticoid therapy where inhibition of bone formation is a major contributor to the bone loss. In this regard the recent study by Lane and colleagues in which subcutaneous PTH markedly increased lumbar spine bone density in women with glucocorticoidinduced osteoporosis was most encouraging [117]. Newer analogs of vitamin D that promote calcium absorption and stimulate bone formation without hypercalcemia [90] may also be valuable additions to the therapeutic armamentarium. Drugs that inhibit bone resorption, such as newer potent bisphosphonates that can be given intravenously on an intermittent basis may also prove effective. Prospective, controlled clinical trials are sorely needed, not only to evaluate existing regimens, but also to study these newer therapies.

References 1. B. D. Kahan, Cyclosporine. N. Engl. J. Med. 321, 1725 – 1738 (1989). 2. M. Suthanthiran and T. B. Strom, Renal transplantation. N. Engl. J. Med. 331, 365 – 376 (1994). 3. G. Goodman I, Dissanayake A, Bowman, S. Pun, Y. F. Ma, W. S. Jee, and S. Epstein, “Interaction of Transforming Growth Factor Beta and Cyclosporine on Bone Mineral Metabolism in the Rat.” Presented at the Endocrine Society, Abstract 423, 1997. 4. U. S. Multicenter FK506 Liver Study Troup, A comparison of tacrolimus (FK506) and cyclosporine for immunosuppression in liver transplantation. N. Engl. J. Med. 331, 1110 – 1115 (1994). 5. M. Cvetkovic, G. N. Mann, D. F. Romero, X. Liang, F. Ma, W. S. S. Jee, and S. Epstein, The deleterious effects of long term cyclosporin A, cyclosporin G and FK506 on bone mineral metabolism in vivo. Transplantation 57, 1231 – 1237 (1994). 6. R. Okazaki, B. L. Riggs, and C. A. Conover, Glucocorticoid regulation of insulin-like growth factor binding protein expression in normal human osteoblast-like cells. Endocrinology 134, 126 – 132 (1994). 7. W. Y. Almawi, M. L. Lipman, A. C. Stevens, B. Zanker, E. T. Hado, and T. B. Strom, Abrogation of glucocorticoid mediated inhibition of T cell proliferation by the synergistic action of IL1 and IL6 and IFN. J. Immunol. 146, 3523 – 3527 (1991). 8. B. P. Lukert and L. G. Raisz, Glucocorticoid induced osteoporosis: Pathogenesis and management. Ann. Intern. Med. 112, 352 – 364 (1990).

338 9. R. Weinstein, R. Jilka, M. Parfitt, and S. Manolagas, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). 10. C. Movsowitz, S. Epstein, M. Fallon, F. Ismail, and S. Thomas, Cyclosporin A in vivo produces severe osteopenia in the rat: Effect of dose and duration of administration. Endocrinology 123, 2571 – 2577 (1988). 11. I. Katz, M. Li, I. Joffe, B. Stein, T. Jacobs, X. G. Liang, H. Z. Ke, W. Jee, and S. Epstein, Influence of age on cyclosporine A-induced alterations in bone mineral metabolism in the rat in vivo. J. Bone Miner. Res. 9, 59 – 67 (1994). 12. C. Movsowitz, S. Epstein, F. Ismail, M. Fallon, and S. Thomas, Cyclosporin A in the oophorectomized rat: Unexpected severe bone resorption. J. Bone Miner. Res. 4, 393 – 398 (1989). 13. B. Stein, P. Halloran, T. Reinhardt, G. W. Engstrom, C. W. Bales, M. K. Drezner, K. L. Currie, M. Takizawa, J. S. Adams, and S. Epstein, Cyclosporin A increases synthesis of 1,25 dihydroxyvitamin D3 in the rat and mouse. Endocrinology 128, 1369 – 1373 (1991). 14. F. Buchinsky, Y. Ma, G. Mann, B. Rucinski, H. Bryer, D. Romero, W. Jee, and S. Epstein, T lymphocytes play a critical role in the development of cyclosporine induced osteopenia. Endocrinology 2278 – 2285 (1996). 15. B. Rucinski, C. C. Liu, and S. Epstein, Utilization of cyclosporine H to elucidate the possible mechanisms of cyclosporine A induced osteopenia in the rat. Metabolism 43, 1114 – 1118 (1994). 16. P. Orcel, J. Bielakoff, D. Modrowski, L. Miravet, and M. C. de Vernejoul, Cyclosporin A induces in vivo inhibition of resorption and stimulation of formation in rat bone. J. Bone Miner. Res. 4, 387 – 391 (1989). 17. I. Marshall, J. A. Isserow, F. J. Buchinsky, B. V. Paynton, and S. Epstein, “Expression of Interleukin 1 and Interleukin 6 in Bone from Normal and Cyclosporin Treated Rats.” XII International Conference on Calcium Regulating Hormones, February, Melbourne, Australia, 1995. 18. G. Goodman, I. Dissanayake, A. Bowman, H. Zhou, Y. Ma, W.S. Jee, and S. Epstein, The effect of the interaction of parathyroid hormone and cyclosporineA on bone mineral metabolism in the rat. J. Bone Miner. Res. 12(Suppl. 1), 336 (1997). 19. B. Sodam, E. Awumey, and S. Epstein, The role of endothelin in cyclosporine induced bone loss in the rat. Calcif. Tissue Int. (1999). 20. E. Awumey, B. Moonga, B. Sodam, A. Koval, O. Adebanjo, M. Kumegawa, M. Zaide, and S. Epstein, Molecular and functional evidence for calcineurin alpha and beta isoforms in the osteoclasts. Novel insights into the mode of action of cyclosporine A. Biochem. Biophys. Res. Commun. 254, 148 – 252 (1999). 21. M. Zahner, F. Teraz, and R. Pacifici, T cells mediate a stimulatory effect of cyclosporine A on human osteoclastogenesis while immature osteoclast precursors are directly regulated by glucocortcoids. J. Bone Miner. Res. 12(Suppl. 1), 198 (1997). 22. H. P. Bryer, J. A. Isserow, E. C. Armstrong, G. N. Mann, B. Rucinski, J. F. Buchinsky, D. F. Romero, and S. Epstein, Azathioprine alone is bone sparing and does not alter cyclosporine A induced osteopenia in the rat. J. Bone Miner. Res. 10, 132 – 138 (1995). 23. R. Eastell, R. E. Dickson, S. F. Hodgson, R. H. Wiesner, M. K. Porayko, H. W. Wahner, S. L. Cedel, B. L. Riggs, and R. A. F. Krom, Rates of vertebral bone loss before and after liver transplantation in women with primary biliary cirrhosis. Hepatology 14, 296–300 (1991). 24. D. Romero, F. Buchinsky, B. Rucinski, M. Cvetkovic, H. Bryer, L. Ma, J. Webb, and S. Epstein, Rapamycin: A bone sparing immunosuppressant? J. Bone Miner. Res. 10, 1556 – 1565 (1995). 25. G. Goodman, I. Dissanayake, B. Sodam, E. Gorodetsky, J. Lu, Y. Ma, W. Jee, and S. Epstein, The effect of low dose combined immunosuppressants, cyclosporine A and Sirolimus on bone metabolism and the endocrine system. J. Bone Miner. Res. 12(Suppl. 1), 402 (1997).

EPSTEIN AND SHANE 26. K. P. May, G. W. Sterling, M. T. McDermott, and W. E. Huffer, The effect of low dose methotrexate on bone metabolism and histomorphometry in rats. Arthritis. Rheum. 27, 201 – 206 (1994). 27. S. Preston, T. Diamond , A. Scott, and M. Laurent, Methotrexate osteopathy in rheumatic disease. Ann. Rheum. Dis. 52, 582 – 585 (1993). 28. L. Buckley, E. Leib, K. Cartularo, P. Vacek, and S. Cooper, Effects of low dose methotrexate on bone mineral density of patients with rheumatoid arthritis. J. Rheumatol. 24, 1489 – 1494 (1997). 29. I. R. Dissanayake, G. R. Goodman, A. R. Bowman, Y. F. Ma, S. Pun, W. S. Jee, and S. Epstein, Mycophenolate mofetil: A promising new immunosuppressant that does not cause bone loss in the rat. Transplantation 65, 275 – 278 (1998). 30. W. G. Goodman, J. W. Coburn, E. Slatopolsky, and I. Salusky, Renal osteodystrophy in adults and children. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism (M. J. Favus, ed.), fourth ed., pp. 347 – 363. Lippincott-Raven, Philadelphia, 1999. 31. B. A. Julian, D. A. Laskow, J. Dubovsky, E. V. Dubovsky, J. J. Curtis, and L. D. Quarrles, Rapid loss of vertebral bone density after renal transplantation. N. Engl. J. Med. 325, 544 – 550 (1991). 32. F. F. Horber, J. P. Casez, U. Steiger, A. Czerniack, A. Montandon, and P. H. Jaeger, Changes in bone mass early after kidney transplantation. J. Bone Miner. Res. 9, 1 – 9 (1994). 33. U. Nisbeth, E. Lindh, S. Ljunghall, U. Backman, and B. Fellstrom, Fracture frequency after kidney transplantation. Transplant Proc. 26, 1764 (1994). 34. D. R. Mitchell and K.W. Lyles, Glucocorticoid-induced osteoporosis: Mechanisms for bone loss. Evaluation of strategies for prevention. J. Gerontol. 45, 153 – 158 (1990). 35. M. Schlosberg, C. Movsowitz, S. Epstein, F. Ismail, M.D. Fallon, and S. Thomas, The effects of cyclosporin A administration and its withdrawal on bone mineral metabolism in the rat. Endocrinology 124, 2179 – 2184 (1989). 36. V. LoCascio, E. Bonnuci, B. Imbimbo, et al., Bone loss after glucocorticoid therapy. Calcif. Tissue Int. 36, 435 – 438 (1984). 37. M. K. Almond, J. T. C. Kwan, K. Evans, and J. Cunningham, Loss of regional bone mineral density in the first 12 months following renal transplantation. Nephron 66, 52 – 57 (1994). 38. B. Bourbigot, M. C. Moal, and J. Cleder, Bone histology in renal transplant patients receiving cyclosporin. Lancet 1, 1048 – 1049 (1988). 39. J. Aubia, J. Masramon, and S. Serrano, Bone histology in renal patients receiving cyclosporin. Lancet 1, 1048 (1988). 40. J. M. Wilmink, J. Bras, S. Surachno, J. L. A. M. Heyst, and J. M. vd Horst, Bone repair in cyclosporine treated renal transplant patients. Transplant. Proc. 21, 1492 – 1494 (1989). 41. A. Cueto-Manzano, S. Konel, A.J. Hutchinson, et al., Bone loss in long term renal transplantation. Histopathology and densitometry analysis. Kidney Int. 55, 2021 – 2029 (1999). 42. E. Shane, Osteoporosis secondary to illness and medications. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.) 2nd ed., pp. 303–326. Academic Press, San Diego, 2001. 43. K. Nishiyama and A. Okinaga, Osteonecrosis after renal transplantation in children. Clin. Orthop. Relat. Res. 295, 168 – 171 (1993). 44. W. A. Briggs, C. L. Hampers, P. J. Merrell, E. B. Hayer, R. E. Wilson, A. G. Birtch, and J. E. Murray, Aseptic necrosis in the femur after renal transplantation. Ann. Surg. 175, 282 – 288 (1972). 45. E. Elmstedt and T. Szahn, Skeletal complications following renal transplantation. Acta Orthop. Scand. 52, 279 – 286 (1981). 46. J. P. Fryer, D. K. Granger, J. R. Leventhal, K. Gillingham, J. S. Najarian, and A. J. Matas, Steroid-related complications in the cyclosporine era. Clin. Transplant. 8, 224 – 229 (1994). 47. O. Ternoven, D. M. Mueller, E. L. Matteson, J. A. Velosa, W. W. Ginsburg, and R. L. Ehman, Clinically occult avascular necrosis of the hip: Prevalence in an asymptomatic population at risk. Radiology 182, 845 – 847 (1992).

339

CHAPTER 52 Transplantation Osteoporosis 48. J. M. LeParc, T. Andre, O. Helenon, J. Benoit, J. B. Paolaggi, and H. Kreis, Osteonecrosis of the hip in renal transplant recipients. Changes in functional status and magnetic resonance imaging findings over three years in three hundred five patients. Rev. Rheum. Engl. Ed. 63, 413 – 420 (1996). 49. I. P. Naiker, S. Govender, S. Naicker, S. Dawood, A. A. Haffejee, and Y. K. Seedat, Avascular necrosis of bone following renal transplantation. S. Afr. Med. J. 83, 646 – 649 (1993). 50. V. J. Gauthier and L. M. Barbosa, Bone pain in transplant recipients responsive to calcium channel blockers. Ann. Intern. Med. 121, 863 – 865 (1994). 51. J. S. Muchmore, D. K. C. Cooper, Y. Ye, V. J. Schlegel, and N. Zudhi, Loss of vertebral bone density in heart transplant patients. Transplant. Proc. 23, 1184 – 1185 (1991). 52. E. Shane, M. D. C. Rivas, S. J. Silverberg, T. S. Kim, R. B. Staron, and J. P. Bilezikian, Osteoporosis after cardiac transplantation. Am. J. Med. 94, 257 – 264 (1993). 53. E. Shane, M. Rivas, R. B. Staron, S. J. Silverberg, M. Seibel, J. Kuiper, D. Mancini, V. Addesso, R. E. Michler, and P. Factor-Litvak, Fracture after cardiac transplantation: A prospective longitudinal study. J. Clin. Endocrinol. Metab. 81, 1740 – 1746 (1996). 54. G. Leidig-Bruckner, S. Edwerbein, D. Czeczatka, P. Dodidou, T. Schilling, M. Pritsch, C. Klose, G. Otto, L. Theilmann, R. Zimmerman, R. Lange, and R. Ziegler, Incidence of osteoporotic fractures after liver and heart transplantation. J. Bone Miner. Res. 12 (Suppl. 1), 145 (1997). 55. P. N. Sambrook, P. J. Kelly, A. Keogh, P. Macdonald, P. Spratt, J. Freund, and J. A. Eisman, Bone loss after cardiac transplantation: A prospective study. J. Heart Lung Transplant. 13, 116 – 121 (1994). 56. P. N. Sambrook, P. J. Kelly, D. Fontana, T. Nguyen, A. Keogh, D. Macdonald, P. Spratt, J. Freund, and J. A. Eisman, Mechanisms of rapid bone loss following cardiac transplantation. Osteoporosis Int. 4, 273 – 276 (1994). 57. M. Rivas and E. Shane, Osteoporosis after organ transplantation. Am. J. Med. 104, 459 – 469 (1998). 58. E. Shane, D. Mancini, K. Aaronson, S. J. Silverberg, M. J. Seibel, V. Addesso, and D. J. McMahon, Bone mass, vitamin D deficiency and hyperparathyroidism in congestive heart failure. Am. J. Med. 103, 197 – 207 (1997). 59. E. Shane, M. Rivas, D. J. McMahon, R. B. Staron, S. J. Silverberg, M. J. Seibel, D. Mancini, R. Michler, K. Aaronson, V. Addesso, and S. H. Lo, Bone loss and turnover after cardiac transplantation. J. Clin. Endocrinol. Metab. 72, 382 – 386 (1997). 60. M. F. Prummel, W. M. Wiersinga, P. Lips, G. T. P. Sanders, and H. P. Sauerwein, The course of biochemical parameters of bone turnover during treatment with corticosteroids. J. Clin. Endocrinol. Metab. 72, 382 – 386 (1991). 61. W. F. Lems, M. I. Gerrits, J. W. G. Jacobs, R. M. van Vugt, H. J. M. van Rijn, and J. W. J. Bijlsma, Changes in (markers of) bone metabolism during high dose corticosteroid pulse treatment in patients with rheumatoid arthritis. Ann. Rheum. Dis. 55, 288 – 293 (1996). 62. D. Thiebaud, M. Krieg, D. Gillard-Berguer, A. Jaquet, J. Goy, and P. Burckhardt, Cyclosporine induces high turnover and may contribute to bone loss after heart transplantation. Eur. J. Clin. Invest. 26, 549 – 555 (1996). 63. C. Guo, A. Johnson, T. Locke, and R. Eastell, Mechanism of bone loss after cardiac transplantation. Bone 22, 267 – 271 (1998). 64. J. A. McDonald, C. R. Dunstan, P. Dilworth, K. Sherbon, A. G. Ross Sheil, R. A. Evans, and G. W. McCaughan, Bone loss after liver transplantation. Hepatology 14, 613 – 619 (1991). 65. E. Meys, E. Fontanges, N. Fourcade, A. Thomasson, M. Pouyet, and P. Delmas, Bone loss after orthotopic liver transplantation. Am. J. Med. 97, 445 – 450 (1994). 66. F. G. Hawkins, M. Leon, M. B. Lopez, M. A. Valero, L. Larrodera, I. Garcia-Garcia, C. Loinaz, and E. M. Gonzales, Bone loss and

67.

68.

69.

70.

71.

72.

73.

74.

75.

76. 77.

78.

79.

80.

81.

82.

83.

84.

turnover in patients with liver transplantation. Hepato-Gastroenterology 41, 158 – 161 (1994). M. B. Lopez, I. Gonzales Pinto, F. Hawkins, M. A. Valero, M. Leon, C. Loinaz, I. Garcia, and R. Gomez, Effect of liver transplantation and immunosuppressive treatment on bone mineral density. Transplant. Proc. 24, 3044 – 3046 (1992). J. Vedi, S. Greer, S. Skingle, N. Garrahan, M. Ninkovic, G. Alexander, and J. Compston, Mechanism of bone loss after liver transplantation: A histomorphometric analysis. J. Bone Miner. Res. 14, 281 – 287 (1999). J. Compston, S. Greer, S. Skingle, D. Stirling, C. Price, P. Friend, and G. Alexander, Early increase in plasma parathyroid hormone level following liver transplantation. J. Hepatol. 25, 715 – 718 (1996). R. Aris, I. Neuringer, M. Weiner, T. Egan, and D. Ontjes, Severe osteoporosis before and after lung transplantation. Chest 109, 1176 – 1183 (1996). E. Shane, S. J. Silverberg, D. Donovan, A. Papadopoulos, R. B. Staron, V. Addesso, B. Jorgensen, C. McGregor, and L. Shulman, Osteoporosis in lung transplantation candidates with end stage pulmonary disease. Am. J. Med. 101, 262 – 269 (1996). R. M. Aris, J. B. Renner, A. D. Winders, H. E. Buell, D. B. Riggs, G. E. Lester, and D. A. Ontjes, Increased rate of fractures and severe kyphosis: Sequelae of living into adulthood with cystic fibrosis. Ann. Intern. Med. 128, 186 – 193 (1998). S. L. Ferrari, L. P. Nicod, J. Hamacher, A. Spiliopoulos, Slosman, T. Rochat, J-P. Bonjour, and R. Rizzoli, Osteoporosis in patients undergoing lung transplantation. Eur. Respir. J. 9, 2378 – 2382 (1997). E. Shane, A. Papadopoulos, R. B. Staron, V. Addesso, D. Donovan, C. McGregor, and L. L. Schulman, Bone loss and fracture after lung transplantation. Transplantation 68, 220 – 227 (1999). L. Shulman, V. Addesso, R. Staron, C. McGregor, and E. Shane, Insufficiency fractures of the sacrum. A cause of low back pain after lung transplantation. J. Heart Lung Transplant. 16, 1081 – 1085 (1997). P. Stillwell and G. Mallory, Pediatric lung transplantation. Clin. Chest Med. 18, 405 – 414 (1997). M. Aringer, H. Kiener, M. Koeller, O. Artemiou, A. Zuckerman, G. Wieselthaler, W. Klepetko, G. Seidl, F. Kainberger, P. Barnecker, J. Smolen, and P. Pietschmann, High turnover bone disease following lung transplantation. Bone 23, 485 – 488 (1998). P. J. Kelly, K. Atkinson, R. L. Ward, P. N. Sambrook, J. C. Biggs, and J. A. Eisman, Reduced bone mineral density in men and women with allogeneic bone marrow. Transplantation 50, 881 – 883 (1990). S. Castaneda, L. Carmona, I. Carjaval, B. Arranz, A Diaz, and A. Garcia-Vadillo, Reduction of bone mass in women after bone marrow transplantation. Calcif. Tissue Int. 60, 343 – 347 (1997). C. Castelo Branco, M. Rovira, F. Pons , M. Duran, J. Sierra, A. Vives, J. Balasch, A. Fortuny, and J. Vanrell, The effect of hormone replacement therapy on bone mass in patients with ovarian failure due to bone marrow transplantation. Maturitas 23, 307 – 312 (1996). J. M. Stern, C. H. Chesnut III, B. Bruemmer, et al., Bone density loss during treatment of chronic GVHD. Bone Marrow Transplant. 17, 395 – 400 (1996). P. Ebeling, D. Thomas, B. Erbas, L. Hopper, J. Szer, and A. Grigg, Mechanism of bone loss following allogeneic and autologous hematopoeitic stem cell transplantation. J. Bone Miner. Res. 14, 342 – 350 (1999). W. Withold, H. Wolf, S. Kollbach, A. Heyll, W. Schneider, and H. Reinauer, Monitoring of bone metabolism after bone marrow transplantation by measuring two different markers of bone turnover. Eur. J. Clin. Chem. Clin. Biochem. 34, 193 – 179 (1996). K. Carlson, B. Simonsson, and S. Ljunghall, Acute effects of high dose chemotherapy followed by bone marrow transplantation on serum markers of bone metabolism. Calcif. Tissue Int. 5, 408 – 411 (1994).

340 85. T. W. Jacobs and S. Epstein, Current status of serum derived bone proteins. In “The Biological Mechanisms of Tooth Movement and Craniofacial Adaptation” (Z. Davidovitch, Ed.), pp. 149 – 158. Ohio State University College of Dentistry Columbus, 1992. 86. P. N. Sambrook, J. Birmingham, P. Kelly, S. Kempler, T. Nguyen, N. Pocock, and J. A. Eisman, Prevention of corticosteroid osteoporosis: A comparison of calcium, calcitriol and calcitonin. N. Engl. J. Med. 328, 1747 – 1752 (1993). 87. P. Sambrook, G. Marshall, K. Henderson, et al., Effect of calcitriol in the prevention of bone loss after cardiac or lung transplantation. J. Bone Miner. Res. 12(Suppl. 1), S400 (1997). 88. E. Meys, F. Terreaux-Duvert, T. Beaume-Six, G. Dureau, and P. J. Meunier, Effects of calcium, calcidiol, and monofluorophosphate on lumbar bone mass and parathyroid function in patients after cardiac transplantation. Osteoporosis Int. 3, 329 – 332 (1993). 89. J. M. Lemire, Immunomodulatory role of 1,25 Dihydroxyvitamin D3. J. Cell Biochem. 49, 26 – 31 (1992). 90. J. M. Lemire, D. C. Archer, and G. S. Reddy, Dihydroxy-24-oxo-16ene-vitamin D3, a renal metabolite of the vitamin D analog 1,25-dihydroxy-16ene-vitamin D3, exerts immunosuppressive activity equal to its parent without causing hypercalcemia in vivo. Endocrinology 135, 2818 – 2821 (1994). 91. J. Lian, C. Stewart, E. Puchacz, S. Mackowiak, V. Shalhoub, D. Collart, G. Zambetti, and G. Stein, Structure of the rat osteocalcin gene and regulation of vitamin D dependent expression. Proc. Natl. Acad. Sci. USA 86, 1143 – 1147 (1989). 92. S. Epstein, M. Schlosberg, M. Fallon, S. Thomas, C. Movsowitz, and F. Ismail. 1,25 Dihydroxyvitamin D3 modifies cyclosporine induced bone loss. Calcif. Tissue Int. 47, 152 – 157 (1990). 93. R. Love, R. Mazess, H. Barden, S. Epstein, P. Newcomb, V. Jordan, P. Carbone, and D. DeMets, Effect of tamoxifen and bone mineral density in postmenopausal women with breast cancer. N. Engl. J. Med. 326, 852 – 856 (1992). 94. B. Ettinger, D. M. Black, B. H. Mitlak, R. K. Knickerbocker, T. Nickelsen, H. K. Genant, C. Christiansen, P. D. Delmas, J. R. Zanchetta, J. Stakkestad, C. C. Gluer, K. Krueger, F. J. Cohen, S. Eckert, K. E. Ensrud, L. V. Avioli, P. Lips, and S. R. Cummings, Reduction of vertebral fracture risk in postmenopausal women treated with raloxifene: Results from a three-year randomized clinical trial. JAMA 282, 637 – 645 (1999). 95. A. Bowman, D. Sass, I. Marshall, H. Liang, Y. F. Ma, W. S. Jee, and S. Epstein, Raloxifene analog (Ly 117018-HCL) ameliorates cyclosporin A induced osteopenia. J. Bone Miner. Res. 22, 1191–1198 (1996). 96. I. Joffe, I. Katz, T. Jacobs, B. Stein, M. Talcizawa, C. Liu, J. Berlin, and S. Epstein, 17 Beta estradiol prevents osteopenia in the oophorectomized rat treated with cyclosporin A. Endocrinology 130, 1578 – 1586 (1992). 97. R. Lindsay, F. Cosman, R. A. Lobo, B. W. Walsh, S. T. Harris, J. E. Reagan, C. L. Liss, M. E. Melton, and C. A. Byrnes, Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84, 3076 – 3081 (1999). 98. T. Nieszporek, W. Grzeszczak, F. Kokot, E. Zukowska-Szczechowska, S. Kusmierski, and A. Szokodny, Influence of the type of immunosuppressive therapy on secretion of somatotropin and function of the pituitary – adrenal and pituitary – gonadal axis in patients with kidney transplant. Nephron 53, 65 – 69 (1989). 99. G. Ramirez, J. Navarte, P. A. Bittle, C. Ayeo-Chastain, and S. E. Dean, Cyclosporine induced alterations in the hypothalamic hypophyseal gonadal axis in transplant patients. Nephron 58, 27 – 32 (1991). 100. L. Montemurro, G. Schiraldi, P. Fraioli, and G. Rizzato, Prevention of corticosteroid induced osteoporosis with salmon calcitonin. Calcif. Tissue Int. 49, 71 – 76 (1991).

EPSTEIN AND SHANE 101. B. Stein, M. Takizawa, I. Katz, J. Berlin, I. Joffe, M. Fallon, and S. Epstein, Salmon calcitonin prevents cyclosporin A induced high turnover bone loss. Endocrinology 129, 92 – 98 (1991). 102. W. Grotz, A. L. Rump, H. Niessen, A. Schmidt-Gayt, G. Reichelt, G. Kirste, Olchewski, and P. Schollmeyer, Treatment of osteopenia and osteoporosis after kidney transplantation. Transplantation 66, 1004 – 1008 (1998). 103. I. Garcia-Delgado, S. Prieto, L. Gil Fragnas, E. Robles, T. Rufilanchas, and F. Hawkins, Calcitonin, editronate and calcidiol treatment in bone loss after cardiac transplantation. Calcif. Tissue Int. 60, 155 – 159 (1997). 104. M. Valero, C. Loinaz, L. Larrodera, M. Leon, E. Morena, and F. Hawkins. Calcitonin and bisphosphonate treatment in bone loss after liver transplantation. Calcif. Tissue Int. 57, 15 – 19 (1995). 105. M. Rodino and E. Shane, Osteoporosis after organ transplantation. Am. J. Med. 104, 459 – 469 (1998). 106. J. D. Adachi, W. G. Bensen, J. Brown, D. Hanley, A. Hodsman, R. Josse, D. L. Kendler, B. Lentle, W. Olszynski, L.-G. Ste-Marie, A. Tenehouse, and A. A. Chines, Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N. Engl. J. Med. 337, 382 – 387 (1997). 107. I. R. Reid, B. A. Schooler, and A. W. Stewart, Prevention of glucocorticoid-induced osteoporosis. J. Bone Miner. Res. 5, 619 – 623 (1990). 108. K. Saag, R. Emkey, T. Schnitzer, J. Brown, F. Hawkins, S. Goemare, G Thamsburg, U. Liberman, P. Delmas, M. Malice, M. Czachus, and A. Daifotis, Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N. Engl. J. Med. 339, 292 – 299 (1998). 109. S. Fan, M. K. Almond, E. Ball, K. Evans, and J. Cunningham, Pamidronate therapy as prevention of bone loss following renal transplantation. Kidney Int. 57, 684 – 690 (2000). 110. E. Shane, M. Rodino, D. J. McMahon, V. Addesso, R. B. Staron, M. J. Seibel, D. Mancini, R. E. Michler, and S. H Lo, Prevention of bone loss after cardiac transplantation with antiresorptive therapy: A pilot study. J. Heart Lung Transplant. 17, 1089 – 1096 (1998). 111. H. Reeve, R. Francis, D. Manas, M. Hudson, and C. Day, Intravenous bisphosphonate prevents symtomatic osteoporotic vertebral collapse in patients after liver tranplantation. Liver Transplant Surg. 4, 404 – 409 (1998). 112. S. C. Riemens, A. Oostdijk, J. van Doormaal, et al., Bone loss after liver transplantation is not prevented by cyclical etidronate, calcium and alpha calcidiol. Osteoporosis Int. 6, 213 – 218 (1996). 113. E. Shane, M. A. Rodino, S. J. Thys-Jacobs, V. Addesso, S. J. Silverberg, R. B. Staron, M. J. Seibel, and D. Mancini, Comparison of two antiresorptive regimens on rates of bone loss after cardiac transplantation. J. Bone Miner. Res. 12 (Suppl. 1), S181 (1997). 114. E. E. Opas, J. H. Seedor, H. Klein, D. Frankenfield, H. Quartuccio, C. Fioravanti, J. Clair, E. Brown, W. C. Hayes, and G. A. Rodan, The effects of a 2 year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry and bone strength in ovariectomized nonhuman primates. J. Clin. Invest. 92, 2577 – 2586 (1993). 115. P. J. Meunier, D. Brancon, P. Chavassieux, C. Edouard, G. Boivin, T. Conrozier, C. Macelli, P. Pastoureau, P. D. Delmas, and J. P. Casez, Treatment with fluoride. In “Osteoporosis” (C. Christiansen, J. S. Johansen, and B. J. Riis, Eds.), pp. 824 – 828. Osteopress, Copenhagen, 1987. 116. N. E. Lane, S. Sanchez, G. W. Modin, H. K. Genant, E. Pierini, and C. D. Arnaud, Parathyroid hormone can reverse corticosteroid-induced osteoporosis. J. Clin. Invest.. 102, 1627 – 1633 (1998). 117. I. A. Katz, W. S. S. Jee, I. Joffe, B. Stein, M. Takizawa, T. W. Jacobs, R. Setterberg, B. Y. Lin, L. Y. Tang, H. Z. Ke, Q. Q. Zeng, J. A. Berlin, and S. Epstein, Prostaglandin E2 alleviates cyclosporin A-induced bone loss in the rat. J. Bone Miner. Res. 4, 1191 – 1200 (1992).

CHAPTER 53

Osteoporosis Associated with Pregnancy LYNN KOHLMEIER ROBERT MARCUS

I. II. III. IV.

Spokane Osteoporosis Center, Endocrine Associates of Spokane, Spokane, Washington 99204 Veterans Affairs Medical Center, Palo Alto, California 94304; and Department of Medicine, Stanford University School of Medicine, Stanford, California 94305

V. Osteoporosis Associated with Magnesium Sulfate Therapy during Pregnancy VI. Paradoxical Bone Mineralization in Twin-to-Twin Transfusion Syndrome References

Introduction Calcium Homeostasis during Pregnancy and Lactation Osteoporosis Associated with Pregnancy Osteoporosis Associated with Heparin Therapy during Pregnancy

II. CALCIUM HOMEOSTASIS DURING PREGNANCY AND LACTATION

I. INTRODUCTION Normal adaptive responses during pregnancy and lactation allow for adequate delivery of mineral to the fetus or infant, while at the same time protecting the maternal skeleton. Current evidence suggests that bone is usually not lost during pregnancy. However, bone can transiently decrease with lactation and can be lost if lactation is prolonged. For additional discussion of the skeletal effects of normal pregnancy and lactation see Chapter 29. In this chapter we review the physiological adaptations in calcium homeostasis of pregnancy and lactation, rare forms of osteoporosis associated with pregnancy such as postpregnancy spinal osteoporosis and transient osteoporosis of the hip, and the skeletal consequences of heparin and magnesium salts when used during or after pregnancy.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

The calcium demands of pregnancy and lactation are great [1 – 3], and under rare conditions, such as postpregnancy spinal osteoporosis and transient osteoporosis of the hip in pregnancy, failure of the normal adaptations in maternal calcium metabolism may have placed the maternal skeleton in jeopardy. Pregnancy has been called a “physiologic absorptive hypercalciuric state” [4], in which intestinal calcium absorption normally doubles. This increase may reflect an elevation in circulating 1,25-dihydroxyvitamin D (calcitriol) [2] or possibly result from a direct intestinal effect of estrogen to stimulate calcium transport [5]. Increases in calcitriol concentrations, from 15 – 60 pg/ml in the nonpregnant state to 80 – 120 pg/ml [4,6], usually occur late, yet may begin as early as the first

341

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

342

KOHLMEIER AND MARCUS

trimester [7,8]. Calcitriol is produced in the placenta as well as in the kidney in pregnancy [9,10]. Though the circulating vitamin D binding protein increases, “free” calcitriol concentrations are also elevated [7], as are circulating albumin-adjusted calcium or ionized calcium concentrations [11] and urinary calcium excretion [12]. The average fasting albumin-corrected calcium concentrations in pregnant women are 9.56 mg/dl, compared to 9.2 mg/dl in nonpregnant women [13]. The calcium balance in pregnant women is positive. No consistent relationship has been found between intestinal calcium absorption and either dietary calcium intake or any index of calcitriol “bioavailability” in pregnant women [14]. In contrast, although maternal calcitriol production increases with lactation, it does not achieve the same concentrations seen during pregnancy, and there is no associated increase in intestinal calcium absorption [14,15]. Instead, calcium is conserved by a renal mecharism. Urinary calcium excretion decreases with lactation and may remain low for as long as 6 months postweaning [13,14].

A. A Possible Role for PTHrP in Normal Pregnancy For many years, pregnancy has been considered a “hyperparathyroid state,” yet we now know that this is incorrect insofar as it refers specifically to circulating concentrations of parathyroid hormone (PTH). Using specific immunoradiometric assays for intact hormone, circulating PTH remains stable and within the normal range throughout pregnancy [16,17]. A state of functional hyperparathyroidism may still occur however, reflecting the increased actions of PTH-related protein (PTHrP). PTHrP concentrations increase during pregnancy, particularly during the thrid trimester [18,19], and concentrations of both PTH and PTHrP are increased during lactation [12]. Sources of circulating PTHrP in pregnancy likely include the placenta and/or breast [18,20]. Hypercalcemia and hypercalciuria can be associated with lactation and extremely high PTHrP concentrations [20]. PTHrP shows powerful sequence homology to PTH in its first 13 amino terminal amino acids, and was first recognized because of its association with humoral hypercalcemia of malignancy (HHM) [22 – 24]. Binding of PTHrP to type I PTH/PTHrP receptors in bone and kidney leads to physiologic actions that mimic the bone-resorbing, phosphaturic, and hypocalciuric effects of PTH (see Chapter 7). [25]. PTHrP appears to be involved in maternal-fetal calcium transfer, parturition, embryogenesis, fetal growth and differentiation, and milk production [26-30]. Hypersecretion of calcitonin during pregnancy and/or lactation has been proposed to buffer the actions of PTH and PTHrP on maternal bone [31].

B. Changes in Bone Mass during Pregnancy and Lactation Under normal conditions, the greater efficiency of intestinal calcium absorption during pregnancy meets the increased calcium demand. Despite increased maternal bone resorption in the third trimester when fetal calcium needs are greatest, most studies find no change in bone mineral density (BMD) with pregnancy [32,33] and some actually report overcompensation, with increased bone mass [34, 35]. Although controversy persists, lactation has generally been associated with bone loss [1,34 – 38]. Animal studies indicate that lactation is accompanied by reduced bone mass and increased bone remodeling, attended by an expanded remodeling space and replacement of mature bone by osteoid and low mineral density bone [39]. In women, 6 months of lactation has been associated with as much as a 7% loss in maternal bone mass. If lactation ceases before 9 months, there appears to be full recovery of bone mass by 18 months postpartum [37].

III. OSTEOPOROSIS ASSOCIATED WITH PREGNANCY Postpregnancy spinal osteoporosis (PPSO) and transient osteoporosis of the hip (OHP) in pregnancy (Table 1) are rare conditions that involve different skeletal sites and have different times of onset. Nonetheless, it is entirely possible that these conditions represent the same disease.

A. Postpregnancy Spinal Osteoporosis PPSO appears characteristically within 3 months after delivery of the first child, yet 40% of affected women experience symptoms in their last trimester [40]. Patients typically have lumbar spine bone mineral densities (BMDs) 50 – 75% of age-predicted values [21]. Proximal femur BMD is often low as well, but appendicular BMD is usually normal. Symptoms include back pain and loss of height due to vertebral compression fracture. At least 80 cases have been reported [21,40 – 47], but the initial presentation followed delivery in less than half. PPSO is usually self-limited and tends not to recur with subsequent pregnancies. In 10 women affected during their first pregnancy, there were only 4 recurrences in 14 subsequent pregnancies [48]. Follow-up BMD measurements increase toward normal over time [49], but persistent BMD deficits are often seen [42]. Mechanisms underlying PPSO remain obscure. Several etiologies have been postulated, including excessive exposure to PTHrP, cytokines, and/or reduced osteoblast

343

CHAPTER 53 Osteoporosis Associated with Pregnancy

function [21,44,48]. PPSO has also been attributable to a low vertebral (peak) bone mass before pregnancy [46], but some patients are described as having normal prepregnancy bone mass [47]. Nonetheless, women with a low vertebral BMD might be at higher risk of PPSO, particularly associated with breast-feeding. In one illustrative case report [21], a 31-year-old woman was found postpartum to have multiple thoracic vertebral compression fractures, hypercalcemia, and an elevated PTHrP concentration. Although she was breast-feeding her infant, the elevation in PTHrP concentration was greater than that expected for lactating women. The patient was hypercalciuric and had a high urinary hydroxyproline excretion and a low serum PTH concentration. Calcitriol concentrations and urinary phosphate excretion were in the low-normal range. Lumbar and proximal femoral BMDs were 53 – 74%of age-predicted values, but BMD of the extremities was  90% of normal. Weaning at 4 months postpartum led to normalization of the serum calcium and PTH concentrations and a 50% reduction in urinary hydroxyproline excretion. Unfortunately, her BMD dropped another 10% during this time, and a 3-month trial of bisphosphonate was instituted. PTHrP concentrations remained high 4 months after weaning, but decreased to normal values by 16 months. At that time, BMD had increased, but was still below age-predicted values. In another case report [50], a 23-year-old woman developed severe back pain and a 5-cm height loss several weeks postpartum. She had four vertebral compression fractures, and a lumbar spine BMD Z score of 9.0 by quantitative computed tomography (QCT). Diagnostic evaluation was unremarkable. She was treated with Calcitonin, calcium and vitamin D, and fluoride. After 6 months of treatment, her spine BMD had increased 50%. At 18 months postpartum, her BMD had increased further, yet still was significantly low for age. It is difficult to know what portion of her improvement reflects a response to therapy and what portion represents the natural history of this condition. Although no skeletal problems have been associated with short courses of corticosteroids in pregnant women, it has been suggested that such women are at a greater risk for pregnancy-associated osteoporosis or for osteoporosis later in life [51]. Though lactation alone is associated with a generalized increase in bone turnover [13,38], patients with PPSO may demonstrate selective loss of trabecular bone. It is suggested that an increase in the bone resorption-inhibitory carboxyterminal PTHrP [107 – 139] peptide [52] may explain normocalcemia and 50% drop in urinary hydroxyproline, despite persistent high PTHrP concentrations in these patients [21]. It is possible that some PTHrP fragment that normally helps to protect bone during pregnancy is deficient in PPSO due to abnormal PTHrP processing.

An increase in the secretion of interleukin-1 (IL-1) may contribute to PPSO. In one reported patient, I1-1 production from cultured monocytes was 80-fold increased 6 months postpartum, returning to normal 9 months later [43]. Most transiliac bone biopsies in women with PPSO were obtained several years after the onset of fracture, making interpretation difficult. Several investigators reported “normal” bone turnover [53,54], where others found increased bone remodeling [43]. One case of osteomalacia was reported in a young woman with ankylosing spondylitis and PPSO [55]. Decreased bone formation with increased bone resorption has also been reported in several patients [42]. Despite an incomplete understanding of mechanisms behind this disorder, therapeutic interventions which have appeared to accelerate recovery include cessation of lactation and anti-resorptive therapy [21,42,54,56].

B. Transient Osteoporosis of the Hip in Pregnancy OHP has been termed “transient osteoporosis or `algodystrophy’ of the hip” [57 – 60] as well as “idiopathic osteoporosis in pregnancy” [61 – 63]. These disorders appear to be the same disease, which can be distinguished from PPSO by several characteristics (Table 1). As the name suggests, osteoporosis primarily, but not exclusively, affects the hip [64 – 67] and may be unilateral [66]. The onset of symptoms, periarticular and groin pain and limited range of hip motion, usually occurs in the third trimester with no history of trauma or antecedent illness. Hip fractures are frequent, and the period of maximal risk is near term. The first three cases of OHP were described in 1959 [68]. Since then, over 90 cases have been reported, though several authors consider the condition to be more common than current literature suggests [63,67,69]. The differential diagnosis for OHP includes osteonecrosis, infection, or inflammatory joint disease, primary and metastatic carcinoma, multiple myeloma, regional migratory osteoporosis, villonodular synovitis and stress fracture of the femoral head. In a separate but possibly related condition, men in their fourth or fifth decade can also be affected by transient osteoporosis of the hip, and over 400 cases have been reported [59]. Radiographic evidence of localized osteopenia in OHP includes an indistinct appearance of subchondral cortical bone [59,70 – 72] (Fig. 1). By contrast, in other osteoporotic disorders, preservation and even accentuation of the subchondral cortex is the rule. In OHP, magnetic resonance imaging (MRI) shows increased joint fluid in the affected hip and diffuse signal abnormalities in the marrow of the

344

KOHLMEIER AND MARCUS

TABLE 1

Comparison of Postpregnancy Spinal Osteoporosis and Transient Osteoporosis of the Hip in Pregnancy

Clinical features

Spinal osteoporosis

Transient osteoporosis of the hip

Usual onset

Within 3 months of delivery of first child

Third trimester

Symptoms

Back pain, height loss

Periarticular and groin pain, restricted hip motion

Prognosis

Usually self-limited, spontaneous recovery of bone mass

Usually self-limited, spontaneous recovery of bone mass, but hip pain may persist

Recurrence in later pregnancies

No

Rare

Proposed etiologies

Cytokines, PTHrP

Low 1,25(OH)2D, local ischemia, impaired venous flow, bone marrow edema, fibrinolysis, neurologic abnormalities

femoral head. T1-weighted signal are decreased and T2-weighted images show an increased signal. Ultrasound, though nonspecific, has been used to help diagnose OHP by detecting an effusion of the hip joint [73]. Radionuclide bone imaging with 99m-technetium diphosphonate may help to distinguish OHP from other conditions, such as osteonecrosis, infection, or inflammatory arthritis. Interpretation of the scan depends on when in the course of disease the image was taken [59]. Within a few days of symptom onset, patients with OHP show increased activity in the femoral head, while in early osteonecrosis, uptake is decreased. In osteonecrosis, the pattern of uptake usually lacks the intensity and regularity of the epiphyseal localization which is seen in OHP [74]. Though there can also be increased uptake in early infectious or inflammatory arthritis, the activity in these diseases is more characteristic of synovial inflammation than of bony uptake. In addition, cartilage loss and bony erosions are features of infection and rheumatoid arthritis, and not of OHP. The diagnosis of OHP is based primarily on clinical findings and, when performed, imaging studies. Synovial fluid and biopsies show only nonspecific inflammatory changes [59,75]. Bone biopsies, which are rarely done, have shown osteoporosis [59] and, sometimes, areas of mild inflammation or no specific abnormalities. Plasma concentrations of calcitriol were unexpectedly low in some patients [53,54]. Except for slight elevations in urinary hydroxyproline excretion, serum alkaline phosphatase activity and erythrocyte sedimentation rate (ESR), and low calcitriol concentrations in some patients, laboratory results in OHP are normal. One reported case of OHP

was a 29-year-old woman, 7 months pregnant, with a sacral insufficiency fracture [76]. She had been treated with heparin for 4 months because of a prior miscarriage. Her 25-hydroxyvitamin D concentration was very low (4 ng/ml, normal  10 – 35). PTH values were normal. Elevations were noted in her alkaline phosphatase activity and ESR as is characteristic for OHP. With vitamin D therapy and postpartum bed rest her pain resolved. Her femoral neck BMD and spine T scores after delivery were  2.02 and  1.21, respectively. The degree to which heparin contributed to this woman’s bone fragililty is not clear (see below). Etiologic explanations for OHP include: local ischemia of the hip, superficial thrombophlebitis, impairment of venous blood flow [77] bone marrow edema, abnormal fibrinolysis [78], and neurologic abnormalities [79] reminiscent of reflex sympathetic dystrophy. A review of 29 women with “idiopathic osteoporosis associated with pregnancy” found a significantly higher prevalence of previous fractures in the mothers of affected women than in a control population [63], suggesting a genetic predisposition to OHP. OHP is usually self-limited, and its course is not appreciably altered by intervention of any kind. Spontaneous recovery commonly occurs within 2 to 9 months postpartum [58], yet hip pain has persisted for an additional 6 months or longer. Rapid symptomatic improvement after termination of pregnancy has also been observed [80]. With clinical improvement, MRI and radiographic abnormalities all regress [66,81]. Both open and closed reduction with internal fixation after delivery have been reported in patients with hip fracture.

345

CHAPTER 53 Osteoporosis Associated with Pregnancy

note that bisphosphonates are not approved for use during pregnancy, and the long-term risks to the fetus are unknown.

C. Hyperthyroidism during Pregnancy

FIGURE 1 Radiograph of the hip obtained after delivery in a 34-yearold primigravida woman with osteoporosis of the hip in pregnancy. There is severe osteopenia with an indistinct appearance of subchondral cortical bone at the femoral head cortex. The cartilage spaces are normal. The patient noted progressive bilateral hip pain beginning in week 26 of gestation and stayed at complete bed rest from week 36 of gestation because of pain. She entered labor 2 weeks later, but severe pain prevented vaginal delivery. Caesarean section was performed. Six weeks later, she walked with assistance, and at 10 weeks postpartum, 5 months following symptom onset, radiographs were normal and she had resumed her normal activities. Reprinted with permission from Ref. 52.

Traditional management of OHP has been bed rest or limited mobility with protected weight bearing. Intensive range of motion exercises and progressive ambulation may be needed to prevent contractures. Nonsteroidal antiinflammatory medication and potent analgesics are often necessary. Therapy with sodium fluoride and calcium was described in three patients, of whom two showed an increase in BMD and one a stabilization [56]. These reported patients had multiple vertebral fractures, however, which are not commonly seen in OHP. Elective caesarian section has been performed in cases of bilateral hip pain preventing vaginal delivery [66], yet in most reported cases no adverse effects of vaginal delivery were noted. In another case report of OHP, a 36-year-old Japanese woman underwent an emergent cesarean section at 35 weeks gestation with twins [82]. She had severe bilateral hip pain and could not ambulate. Postpartum her symptoms did not improve, however, and radiographs revealed severe osteopenia limited to the hip and knee without fracture. Alendronate was initiated with functional recovery and symptom resolution. The authors appropriately

Thyrotoxicosis during pregnancy is not uncommon. Thyroid hormone directly stimulates bone turnover and, in amounts sufficient to suppress endogenous thyroid function, has been associated with bone loss [83 – 85] (see Chapter 47). Hyperthyroidism increases intestinal motility, decreases production of 1,25(OH)2 vitamin D, and, consequently, decreased intestinal calcium absorption [86]. Thus, the hyperthyroid state may counteract normal adaptive responses to pregnancy. A 35-year-old woman followed by us exhibited six vertebral compression fractures 6 months after a fullterm pregnancy which had been complicated by thyrotoxicosis. BMD Z score at the fourth lumbar vertebral body was  4.5. Antithyroid medication was initiated at 30 weeks gestation, although her symptoms were first noted 4 months earlier. Whether the bone deficit and fractures were due to the combined effects of pregnancy and sustained hyperthyroidism or were simply a severe example of PPSO cannot be known with certainty, but this patient serves as a reminder of the normal physiologic adaptations of pregnancy and the possibility that their disruption could lead to substantial bone loss.

IV. OSTEOPOROSIS ASSOCIATED WITH HEPARIN THERAPY DURING PREGNANCY Women with a prior history of thromboembolism have a 5 – 12% recurrence rate during pregnancy, and because the use of warfarin is associated with both embryopathy and devastating hemorrhagic events it is recommended that such women receive prophylactic antithrombotic therapy with heparin [87]. Patients requiring heparin during pregnancy are often treated for relatively prolonged periods. The incidence of heparin-induced osteoporosis can be as high as 30% [88], and that of symptomatic vertebral fractures as high as 2% [89 – 91]. One of the first reports of bone loss associated with heparin therapy in pregnancy described reduced phalangeal – cortical area ratio in women who received at least 25 weeks of heparin (20,000 IU daily) compared to the ratio in pregnant women who received the same daily dose for less than 7 weeks [92]. Another retrospective study found 17% of women exposed to heparin during pregnancy had osteopenia on postpartum spine or hip radiographs [93]. Half of the 70 women in this study were reevaluated 6 to 12 months postpartum, and bone mass appeared to recover in most cases.

346 Because radiographic assessment of bone mass is very insensitive, these initial studies are hard to interpret. Proximal femur BMD measured immediately postpartum revealed bone deficits of at least 10% from baseline in 4 of 14 patients (36%), whereas lumbar spine BMD did not change. Heparin dose did not influence BMD in that study. Bone mass remained lower than baseline at 6 months postpartum and lower than the BMD of control women without previous heparin exposure [88]. A prospective investigation of calcium homeostasis in 36 heparin-treated pregnant women revealed dose-dependence, where an average daily dose of 24,500 IU resulted in significantly higher total and ionized calcium and calcitionin concentrations and lower urinary calcium concentrations [94]. Possible mechanisms leading to reduced mineralization and elevated circulating calcium include increased osteoclastic activity, a decrease in bone collagen synthesis [95], and/or an increase in collagenase activity [96]. A role for secondary hyperparathyroidism [97], heparin-like cofactors of PTH [98], or PTHrP [94] remains inconclusive. Patient management generally includes using the lowest possible dose of heparin, even though some reports fail to establish dose-dependence, and supplemental calcium and/or calcitriol, which remain of unproven efficacy [94]. Patients in the studies previously discussed have almost all received a prenatal vitamin/mineral supplement containing vitamin D and at least 1000 mg of calcium. In a recent study, skeletal protection was reported for a hydroxyapatite supplement [99]. Postpartum antiresorptive therapy has been used in some patients, but to uncertain effect. Despite these measures, fractures are not completely avoided with either low-dose or short-term heparin prophylaxis [91]. Current data suggest a lower risk of heparin-induced osteoporosis with low-molecular-weight heparin (LMWH) which does not cross the placental barrier and has a longer half-life and increased bioavailability [100,101]. Other studies have not shown skeletal benefits of LMWH in comparison to standard unfractionated heparin [102]. A large series from 1990 to 1996 followed 61 women treated with LMWH during pregnancy. Though no women were symptomatic, 32% had BMD Z scores at least 1.0 or lower, and none completely normalized their BMD postpartum [103].

V. OSTEOPOROSIS ASSOCIATED WITH MAGNESIUM SULFATE THERAPY DURING PREGNANCY Tocolytic therapy with magnesium sulfate (MgSO4) is known to affect calcium homeostasis and result in hypocalcemia and hypercalciuria [104,105]. The effects of short-term MgSO4 therapy in women with preterm labor- or pregnancy-

KOHLMEIER AND MARCUS

induced hypertension are reversible once therapy is discontinued [104]. Long-term MgSO4 therapy is commonly instituted in tertiary care centers to delay preterm labor, and has been reported to be associated with infantile osteopenia [106] and congenital rickets [107]. One study in 20 women receiving 963 – 1405 of intravenous MgSO4 over 14 – 26 days revealed hypocalcemia and hypercalciuria [108]. Urinary excretion of magnesium and copper was increased as were serum magnesium, phosphorus, and parathyroid hormone concentrations. Further evaluation is needed to clarify the effects of long-term MgSO4 therapy on bone turnover and BMD. One impediment to such clarification is the frequent use of polypharmacy in these patients. For example, a 35-yearold woman with PPSO suffered bilateral calcaneal fractures after long-term magnesium sulfate tocolysis. However, she had also been at prolonged bed rest and received heparin and corticosteroids before delivery [109].

VI. PARADOXICAL BONE MINERALIZATION IN TWIN-TO-TWIN TRANSFUSION SYNDROME Twin-to-twin transfusion syndrome occurs in 7% of twins and is due to asymmetric blood flow. Twin paradoxical bone mineralization has been reported in a mother with preeclampsia [110], where one infant was relatively large, polycythemic, and osteopenic, and the other was small, anemic, and osteopetrotic. There was no evidence of subperiosteal resorption or fractures in either twin. Bone mineral content of the distal forearm was 0.028 g/cm in the larger infant and 0.074 g/cm in the smaller infant (normal 0.041  0.006 g/cm (mean  SD)). By 1 week of age, the osteopenic infant had laboratory evidence of hyperparathyroidism, with high calcium (3.2 mmol/liter), low phosphorus (0.75 mmol/liter), and increased alkaline phosphatase activity (668 U/liter). The osteopetrotic infant had a normal calcium (2.24 mmol/liter) concentration, a slightly high phosphorus concentration (2.0 mmol/liter), and normal alkaline phosphatase activity (315 U/liter). By 3 months of age all measureable variables had normalized in both infants. The authors speculate that alterations in macrophage derived osteoclastic activity may contribute to this disorder. As cell populations normalize with time, so does the bone mineral content and other parameters of bone metabolism and hematologic homeostasis [110].

References 1. P. Atkinson and R. West, Loss of skeletal calcium in lactating women. J. Obstet. Gynaecol. Br. Commonw. 77, 555 – 560 (1970). 2. R. Heaney and T. Skillman, Calcium metabolism in normal human pregnancy. J. Clin. Endocr. Metab. 33, 661 – 670 (1971). 3. J. Repke, Calcium homeostatis in pregnancy. Clin. Obstet. Gynecol. 37, 59 – 65 (1994).

CHAPTER 53 Osteoporosis Associated with Pregnancy 4. J. Gertner, D. Coustan, A. Kliger, L. Mallette, N. Ravin, and A. Broadus, Pregnancy as a state of physiologic absorptive hypercalciuria. Am. J. Med. 81, 451 – 456 (1986). 5. B. Arjmandi, B. Hollis, and D. Kalu, In vivo effect of 17b-estradiol on intestinal calcium absorption in rats. Bone Miner. 26, 181 – 189 (1994). 6. E. Seely, R. Wood, and E. Brown, Lower serum ionized calcium and abnormal calciotropic hormone concentrations in preeclampsia. J. Clin. Endocrinol. Metab. 74, 1436 – 1440 (1992). 7. R. Bouillon, F. VanAssche, H. VanBaelen, W. Heyns, and P. DeMoor, Influence of the vitamin D-binding protein on the serum concentration of 1,25-dihydroxyvitamin D. Significance of the free 1,25-dihydroxyvitamin D conentration. J. Clin. Invest. 67, 589 – 596 (1981). 8. D. Bikle, E. Gee, B. Halloran, and J. Haddad, Free 1,25dihydroxyvitamin D concentrations in serum from normal subjects, pregnant subjects, and subjects with liver disease. J. Clin. Invest. 74, 1966 – 1971 (1984). 9. G. Reddy, A. Norman, D. Willis, D. Goltzman, H. Guyda, S. Soloman, D. Philips, J. Bishop, and E. Mayer, Regulation of vitamin D metabolism in normal pregnancy. J. Clin. Endocrinol. Metab. 56, 363 – 370 (1983). 10. Y. Weisman, A. Harell, and S. Edelstein, 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 in vitro synthesis by human decidua and placenta. Nature 281, 317 (1979). 11. R. Payne, A. Little, and R. Evans, Albumin-adjusted calcium concentrations in serum increase during normal pregnancy. Clin. Chem. 36, 142 – 144 (1990). 12. G. Kent, R. Price, D. Gutteridge, M. Smith, J. Allen, C. Bhagat, M. Barnes, C. Hickling, R. Retallack, S. Wilson, R. Revlin, C. Davies, and A. St. John, Human lactation forearm trabecular bone loss, increased bone turnover and renal conservation of calcium and inorganic phosphate with recovery of bone mass following weaning. J. Bone Miner. Res. 5, 361 – 369 (1990). 13. G. Kent, R. Price, D. Gutteridge, J. Allen, S. Blakeman, C. Bhagat, A. St. John, M. Barnes, M. Smith, and D. Evans, Acute effects, of an oral calcium load in pregnancy and lactation: Findings on renal calcium concentration and biochemical indices of bone turnover. Miner. Electrolyte Metab. 17, 1 – 17 (1991). 14. G. Kent, R. Price, D. Gutteridge, K. Rossman, M. Smith, J. Allen, C. Hickling, and S. Blakeman, The efficiency of intestinal calcium absorption is increased in late pregnancy but not in established lactation. Calcif. Tissue Int. 48, 293 – 295 (1991). 15. B. Specker, N. Vieira, O. O’Brien, M. Ho, J. Heubi, S. Abrams, and A. Yerqey, Calcium kinetics in lactating women with low and high calcium intakes. Am. J. Clin. Nutr. 59, 593 – 599 (1994). 16. O. Davis, D. Hawkins, L. Rubin, J. Posillico, E. Brown, and I. Schiff, Serum parathyroid hormone (PTH) in pregnant women determined by an immunoradiometric assay for intact PTH. J. Clin. Endocrinol. Metab. 67, 850 – 852 (1988). 17. G. Saggese, G. Baroncelli, S. Bertelloni, and C. Cipolloni, Intact parathyroid hormone concentrations during pregnancy, in healthy term neonates and in hypocalcemic preterm infants. Acta Paediatr. Scand. 80, 36 – 41 (1991). 18. S. Bertelloni, G. Baroncelli, A. Pelletti, R. Battini, and G. Saggese, Parathyroid hormone-related protein in healthy pregnant women. Calcif. Tissue Int. 54, 195 – 197 (1994). 19. Y. Hiroto, T. Anai, and I. Miyakawa, PTHrP levels in maternal and cord blood Am. J. Obstet. Gynecol. 177, 702 – 770 (1997). 20. S. Khosla, J. van Heerden, H. Gharib, I. Jackson, J. Danks, and T. Martin, Parathyroid hormone-related protein and hypercalcemia secondary to massive mammary hyperplasia. N. Engl. J. Med. 322, 1157 (1990). 21. I. Reid, D. Wattie, M. Evans, and A. Budayr, Post-pregnancy osteoporosis associated with hypercalcemia. Clin. Endocrinol. 37, 298 – 303 (1992).

347 22. W. Burtis, T. Brady, J. Orloff, J. Ersback, R. J. Warrell, B. Olson, T. Wu, M. Mitnick, A. Broadus, and A. Stewart, Immunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcemia of cancer. N. Engl. J. Med. 322, 1106 – 1112 (1990). 23. J. Moseley, M. Kubota, H. Diefenbach-Jagger, R. Wettenhall, B. Kemp, L. Suda, C. Rodda, P. Ebeling, P. Hudson, J. Zajac, and T. Martin, Parathyroid hormone-related protein purified from a human lung cancer line. Proc. Natl. Acad. Sci. USA 84, 5048 – 5052 (1987). 24. A. Budayr, R. Nissenson, R. Klein, K. Pun, O. Clark, D. Diep, C. Arnaud, and G. Strewler, Increased serum concentrations of a parathyroid hormone-like protein in malignancy-associated hypercalcemia. Ann. Int. Med. 111, 807 – 812 (1989). 25. H. Kaji, T. Sugimoto, M. Fukase, and K. Chihara, Role of dual signal transduction systems in the stimulation of bone resorption by parathyroid hormone-related peptide. The direct involvement of cAMP-dependent protein kinase. Horm. Metab. Res. 25, 421 – 424 (1993). 26. R. MacIsaac, J. Heath, C. Rodda, J. Moseley, A. Care, J. Martin, and I. Caple, Role of the fetal parathyroid glands and parathyroid hormone-related protein in the regulation of placental transport of calcium, magnesium and inorganic phosphate. Reprod. Fertil. Dev. 3, 447 – 457 (1991). 27. J. Ferguson, J. Gorman, D. Bruns, E. Weir, W. Burtis, T. Martin and M. Bruns, Abundant expression of parathyroid hormone-related protein in human amnion and its association with labor. Proc. Natl. Acad. Sci. USA 89, 8384 – 8388 (1992). 28. J. Moseley, J. Hayman, J. Danks, D. Alcorn, V. Grill, J. Southby, and M. Horton, Immunochemical detection of parathyroidhormone-related protein in human fetal epithelia. J. Clin. Endocrinol. Metab. 73, 478 – 484 (1991). 29. P. Burton, C. Moniz, P. Quirke, A. Malik, T. Bui, H. Juppner, G. Segre, and D. Knight, Parathyroid hormone-related peptide: Expression in fetal and neonatal development. J. Pathol. 167, 291 – 296 (1992). 30. E. Bucht, M. Carlqvist, B. Hedlund, K. Bremme, and O. Torring, Parathyroid hormone-related peptide in human milk measured by a midmolecule radioimmunoassay. Metab. Clin. Exp. 41, 11 – 16 (1992). 31. M. Whitehead, G. Lane, O. Young, S. Campbell, G. Abeyasekera, C. Hillyard, I. MacIntyre, K. Phang, and J. Stevenson, Inter relations of calcium regulating hormones during pregnancy. Br. Med. J. 283, 10 – 12 (1981). 32. C. Christiansen, P. Rodbro, and B. Heinild, Unchanged total body calcium in normal human pregnancy. Acta Obstet. Gynecol. Scand. 55, 141 – 143 (1976). 33. G. Kent, R. Price, D. Gutteridge, J. Allen, K. Rosman, M. Smith, C. Bhagat, S. Wilson, and R. Retallack, Effect of pregnancy and lactation on maternal bone mass and calcium metabolism. Osteoporosis (Suppl. 1), S44 – S47 (1993). 34. M. Sowers, M. Crutchfield, M. Jannausch, S. Updike, and G. Corton, A prospective evaluation of bone mineral change in pregnancy. Ostet. Gynecol. 77, 841 – 845, (1991). 35. B. Drinkwater and C. Chesnut, Bone density changes during pregnancy and lactation in active women: A longitudinal study. Bone Miner. 14, 153 – 160 (1991). 36. G. Chan, M. McMurry, K. Westover, K. Englebert-Fenton, and M. Thomas, Effects of increased dietary calcium intake upon the calcium and bone mineral status of lactating adolescent and adult women. Am. J. Clin. Nutr. 46, 319 – 323 (1987). 37. M. Sowers, G. Corton, B. Shapiro, M. Jannausch, M. Crutchfield, M. Smith, J. Randolph, and B. Hollis, Changes in bone density with lactation. JAMA 269, 3130 – 3135 (1993). 38. M. F. Sowers, Pregnancy and lactation as risk factors for subsequent bone loss and osteoporosis. J. Bone Miner. Res. 11, 1052 – 1059 (1996).

348 39. M. Miller, T. Omura, and S. C. Miller, Increased cancellous bone remodeling during lactation in beagles. Bone 10, 279 – 285 (1989). 40. W. Khovidhunkit and S. Epstein, Osteoporosis in pregnancy. Osteo Int. 6, 345 – 354 (1996). 41. H. Chung, S. Lim, M. Lee, and K. Huh, Pregnancy associated osteoporosis. Yonsei Med. J. 29, 286 – 294 (1988). 42. N. Yamamoto, H. Takahashi, T. Tanizawa, T. Kawashima, and N. Endo, Bone mineral density and bone histomorphometric assessments of postpregnancy osteoporosis. Calcif. Tissue Int. 54, 20 – 25 (1994). 43. J. Blanch, R. Pacifici, and A. Chines, Pregnancy-induced osteoporosis: Report of two cases with long term bone density follow-up. Br. J. Rheumatol. 33, 269 – 272 (1994). 44. T. Anai, T. Tomiyasu, K. Arima, and I. Miyakawa, Pregnancy-Associated Osteoporosis with elevated levels of circulating PTHrP: A report of 2 cases. J. Obstet. Gyn. Res. 25, 63 – 67 (1999). 45. B. E. C. Nordin and A. Roper, Post Pregnancy osteoporosis A syndrome? Lancet 1, 431 – 434 (1955). 46. G. Khastgir, J. W. W. Studd, H. King, H. Abdalla, J. Jones, G. Carter, and J. Alaghband-Zadeh, Changes in bone density and biochemical markers of bone turnover in pregnancy-associated osteoporosis. Br. J. Obstet. Gynecol. 103, 716 – 718 (1996). 47. R. Smith and A. J. Phillips, Osteoporosis during pregnancy and its management. Scand. J. Rheum. 27, 66 – 67 (1998). 48. R. Smith, N. Athanasou, S. J. Ostlere, and S. E. Vipond, Pregnancy associated osteoporosis. J. Med. 88, 865 – 878 (1995). 49. A. J. Phillips, S. J. Ostlere, and R. Smith, Pregnancy associated osteoporosis: follow up bone mineral density. J. Bone Miner. Res. 13, 521 (1998). 50. Y. Liel, D. Atar, and N. Ohana, Pregnancy-associates osteoporosis: Preliminary densitometric evidence of extremely rapid recovery of bone mineral density. S. Med. J. 91, 33 – 35 (1998). 51. O. Ogueh, G. Khastgir, J. W. W. Studd, J. Jones, J. Alaghband-Zadeh, and M. R. Johnson, Antenatal corticosteroid therapy and risk of osteoporosis. Br. J. Obstet. Gyn. 105, 551 – 555 (1998). 52. A. Fenton, T. Martin, and G. Nicholson, Long-term culture of disaggregated rat osteoclasts: Inhibition of bone resorption and reduction of osteoclast-like cell number by calcitonin and PTHrP [107 – 139]. J. Cell Physiol. 155, 1 – 7 (1993). 53. R. Smith, J. Stevenson, C. Winearls, C. Woods, and B. Wordsworth, Osteoporosis of pregnancy. Lancet 1, 1178 – 1180 (1985). 54. H. Gruber, D. Gutteridge, and D. Baylink, Osteoporosis associated with pregnancy and lactation: Bone biopsies and skeletal features in three patients. Metab. Bone Dis. Relat. Res. 5, 159 – 165 (1984). 55. M. LeBoff and L. Kohlmeier, Postpartum Presentation of Osteoporosis, Osteomalacia and Vertebral Fractures in a Young Woman with Ankylosing Spondylitis: Case Report and Review of the Literature. J. Bone Miner. Res. (1996). 56. F. Moya, P. Persi, N. Guanabens, A. Monegal, J. Mitja, M. Centellas, and J. Munoz-Gomez, Osteoporosis associated with pregnancy. Description of 3 cases. Med. Clin. Barcelona 100, 743–745 (1993). 57. J. Loughead, Z. Mughal, F. Mimouni, R. Tsang, and A. Oestreich, Spectrum and natural history of congenital hyperparathyroidism secondary to maternal hypercalcemia. Am. J. Perinatol. 7, 350 – 355 (1990). 58. J. Beaulieu, C. Razzano, and R. Levine, Transient osteoporosis of the hip in pregnancy. Clin. Orthop. Rel. Res. 115, 165 – 168 (1976). 59. K. Bramlett, J. Killian, R. Nasca, and W. Daniel, Transient osteoporosis. Clin. Orthop. Rel. Res. 222, 197 – 202 (1987). 60. J. Brodell, J. Burns, and K. Heiple, Transient osteoporosis of the hip of pregnancy: Two cases complicated by pathologic fracture. J. Bone Jt. Surg. 71, 1252 – 1257 (1989). 61. C. Dent and M. Friedman, Pregnancy and idiopathic osteoporosis. Q. J. Med. 34, 341 – 357 (1965).

KOHLMEIER AND MARCUS 62. M. De Marini, Idiopathic osteoporosis in pregnancy: Review of the literature on the problem and presentation of personal case. Minerva Ginecol. 17, 960 – 968 (1965). 63. F. Dunne, B. Walters, T. Marshall, and D. Heath, Pregnancy associated osteoporosis. Clin. Endocrinol. 39, 487 – 490 (1993). 64. M. Rosenbaum and A. Nagler, Idiopathic transient osteoporosis of the hip. Harefuah 109, 143 – 144 (1985). 65. R. Kumar and R. House, Idiopathic transient osteoporosis of the hip in pregnancy. Tex. Med. 79, 33 – 34 (1983). 66. L. Shifrin, N. Reis, H. Zinman, and M. Besser, Idiopathic transient osteoporosis of the hip. Jt. Surg. 69B, 769 – 774 (1987). 67. G. Goldman, M. Friedman, and J. Ovadia, Idiopathic transient osteoporosis of the hip in pregnancy. Int. J. Gynecol. Obstet. 46, 317 – 320 (1994). 68. P. Curtiss and W. Kincaid, Transient demineralization of thehip in pregnancy: A report of three cases. J. Bone Jt. Surg. 41, 1327 (1959). 69. S. Siva and V. Raoch., Transient osteoporosis of the hip in pregnancy. Aust. NZ J. Obstet. Gynecol. 37, 261 – 266 (1997). 70. R. Rosen, Transitory demineralization of the femoral head. Radiology 94, 509 (1970). 71. F. Valenzuela, H. Aris, and S. Jacobelli, Transient osteoporosis of the hip. J. Rheum. 4, 59 (1977). 72. D. Schapira, Transient osteoporosis of the hip. Sem. Arthritis Rheum. 22, 98 – 105 (1992). 73. J. M. Koski and R. Mullykangas-Luosujarvi., Transient osteoporosis of the hip with joint effusion detected by ultrasonography. Clin. Rheum. 16, 404 – 408 (1997). 74. A. Gaucher, J. Colomb, A. Naoun, G. Faure, and P. Netter, The diagnostic value of 99m Tc-diphosphonate. Bone imaging in transient osteoporosis of the hip. J. Rheum. 6, 574 (1979). 75. M. Lequesne, M. Kerbaell, M. Bensasson, C. Perez, R. Dreisen, and A. Forest, Partial transient osteoporosis. Rev. Rheum. Mal. Osteoarticulaires 46, 111 – 121 (1979). 76. V. Breuil, O. Brocq, L. Euller-Ziegler, and A. Grimaud., Insufficient fracture of the sacrum revealing a pregnancy associated osteoporosis. First case reported. Ann. Rheum. Dis. 56, 278 – 279 (1997). 77. Y. Ben-David, J. Bornstein, Y. Sorokin, E. Morad, and H. Adramovici, Transient osteoporosis of the hip during pregnancy. A case report. J. Reprod. Med. 36, 672 – 674 (1991). 78. P. Van Veldhuizen, J. Neff, M. Murphey, D. Bodensteiner, and B. Skikne, Decreased fibrinolytic potential in patients with idiopathic avascular necrosis and transient osteoporosis of the hip. Am. J. Hematol. 44, 243 – 248 (1993). 79. I. Klier, J. Zoldan, Z. Yosipovitch, and N. Gadoth, Transient regional and migratory osteoporosis. A possible neural mechanism. Isr. J. Med. Sci. 24, 201 – 206 (1988). 80. M. Chigira, H. Watanabe, and E. Udagawa, Transient osteoporosis of the hip in the first trimester of pregnancy. A case report and review of Japanese literature. Arch. Orthop. Trauma Surg. 107, 178 – 180 (1988). 81. Y. Takatori, T. Kokubo, S. Ninomiya, T. Nakamura, I. Okutsu, and M. Kamogawa, Transient osteoporosis of the hip: Magnetic resonance imaging. Clin. Orthop. Rel. Res. 271, 190 – 194 (1991). 82. A. Samdani, E. Lachmann, and W. Nagler. Transient osteoporosis of the hip during pregnancy: A case report. Am. J. Phys. Med. Rehab. 77, 153 – 156, (1998). 83. G. Mundy and L. Raisz, Thyrotoxicosis and calcium metabolism. Miner. Electrolyte Metab. 2, 285 – 292 (1979). 84. B. Krolner, J. Jorgensen, and S. Nielsen, Spinal bone mineral content in myxoedema and thyrotoxicosis: Effects of thyroid hormone(s) and antithyroid treatment. Clin. Endocrinol. 18, 4439 – 4446 (1983). 85. S. Toh, B. Klaunch, and P. Brown, Effect of hyperthyroidism and its treatment on bone mineral content. Arch. Int. Med. 145, 883 – 886 (1985).

349

CHAPTER 53 Osteoporosis Associated with Pregnancy 86. B. Jastrup, L. Mosekilde, F. Melson, Bi. Lund, Bj. Lund, and O. Sorensen, Serum concentrations of vitamin D metabolites and bone remodeling in hyperthyroidism. Metabolism 31, 126 – 132 (1982). 87. J. Ginsberg and J. Hirsch, Use of antithrombotic agents during pregnancy. Chest 102, 385 – 390 (1992). 88. L. Barbour, S. Kick, J. Steiner, M. LoVerde, L. Heddleston, J. Lear, A. Baron, and P. Barton, A prospective study of heparin-induced osteoporosis in pregnancy using bone desitometry. Am. J. Obstet. Gynecol. 170, 862 – 869 (1994). 89. J. Niebyl, Drug therapy during pregnancy. Cur. Opin. Obstet. Gynecol. 3, 24 – 27 (1991). 90. G. Wunderer and G. Muller, Osteoporosis and pre- and postpartum heparin therapy (German). Geburt Frauen 50, 61 – 63 (1990). 91. T. Dahlam, Osteoporotic fractures and the recurrence of thromboembolism during pregnancy and the puerperium in 184 women undergoing thromboprophylaxis with heparin. Am J. Obstet. Gynecol. 168, 1265 – 1270 (1993). 92. M. deSwiet, P. Ward, J. Fidler, A. Horsman, D. Katz, E. Letsky, M. Peacock, and P. H. Wise, Prolonged heparin therapy in pregnancy causes bone demineralization. Br. J. Obstet. Gynecol. 90, 1129 – 1134 (1983). 93. T. Dahlman, N. Lindvall, and M. Hellgren, Osteopenia in pregnancy during long term heparin treatment. Br. J. Obstet. Gynecol. 97, 221 – 228 (1990). 94. T. Dahlman, H. Sjoberg, M. Hellgren, and E. Bucht, Calcium homeostasis in pregnancy during long-term heparin treatment. Br. J. Obstet. Gynecol. 99, 412 – 416 (1992). 95. R. Thompson, Heparin osteoporosis. An experimental model using rats. J. Bone Jt. Surg. 55A, 606 – 612 (1973). 96. B. Starcher, C. Hill, and J. Madaras, Effect of zinc deficiency on bone collagenase and collagen turnover. J. Nutr. 110, 2095 – 2102 (1980). 97. L. Avioli, Heparin-induced osteopenia: An appraisal. Adv. Exp. Med. Biol. 52, 375 – 387 (1975). 98. P. Goldhaber, Heparin enhancement of factors stimulating bone resorption in tissue culture. Science 147, 407 – 408 (1965). 99. J. Ringe and A. Keller, Risk of osteoporosis in long-term heparin therapy of thromboembolic diseases in pregnancy: Attempted pre-

100.

101. 102.

103.

104.

105.

106.

107.

108.

109.

110.

vention with ossein-hydroxyapatite. Geburtshilfe Frauenheilk. 52, 426 – 429 (1992). E. Melissari, C. Parker, N. Wilson, G. Monte, C. Kanthou, K. Pemberton, K. Nicolaides, J. Barrett, and V. Kakkar, Use of low molecular weight heparin in pregnancy. Thromb. Haemost. 68, 652 – 656 (1992). C. Nelson-Piercy, Heparin-induced ostoeporosis. Scand. J. Rheum. 27, 68 – 71 (1998). J. Shefras and R. G. Farquharson, Bone density studies in pregnant women receiving heparin. Eur. J. Obstet. Gynecol. Reprod. Biol. 65, 171 – 174 (1996). C. Nelson-Piercy, E. A. Letsky, and M. de Sweit, Low-molecularweight heparin for obstetric thromboprophylaxis: Experience of 69 pregnancies in 61 high risk women. Am. J. Obstet. Gynecol. 176, 1062 – 1066 (1997). D. Cruikshank, R. Pitkin, E. Donnelly, and W. Reynolds, Urinary magnesium, calcium, and phosphorus excretion during magnesium sulfate infusion. Obstet. Gynecol. 58, 430 – 434 (1981). I. Cholt, S. Steinberg, P. Tropper, H. Fox, G. Segre, and J. Bilezikian, The influence of hypermagnesemia on serum calcium and parathyroid hormone concentrations in human subjects. N. Engl. J. Med. 310, 1221 – 1225 (1984). W. Cumming and V. Thomas, Hypermagnesemia: A cause of abnormal metaphyses in the neonate. Am. J. Roentgenol. 152, 1071 – 1072 (1989). C. Lamm, K. Norton, R. Murphy, I. Wilkins, and J. Rabinowitz, Congenital rickets associated with magnesium sulfate infusion for tocolysis. J. Pediatr. 113, 1078 – 1082 (1988). L. Smith, P. Burns, and R. Schanler, Calcium homeostasis in pregnant women receiving long-term magnesium sulfate therapy for preterm labor. Am. J. Obstet. Gynecol. 167, 45 – 51 (1992). A. L. Levav, L. Chan, and R. J. Wapner, Long-term, magnesium sulfate tocolysis and maternal osteoporosis in a triplet pregnancy: A case report. Am. J. Perinatal 15, 43 (1998). N. Bishop, F. King, P. Ward, J. Rennie, and A. Dixon, Paradoxical bone mineralization in the twin to twin transfusion syndrome. Arch. Dis. Child. 65, 705 – 706 (1990).

CHAPTER 54

Osteoporosis Associated with Rheumatologic Disorders STEVEN R. GOLDRING

Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, and New England Baptist Bone and Joint Institute, Harvard Institutes of Medicine, Boston, Massachusetts 02215

I. Introduction II. Rheumatoid Arthritis III. Juvenile Rheumatoid Arthritis

IV. Seronegative Spondyloarthropathies V. Systemic Lupus Erythematosus References

I. INTRODUCTION

ticular tissues. These entities include rheumatoid arthritis, juvenile rheumatoid arthritis, the seronegative spondyloarthropathies, and systemic lupus erythematosus. It is important to note, however, that many of the related immunemediated rheumatic conditions such as systemic vasculitis may also be accompanied by adverse effects on skeletal remodeling and increased risk for osteoporosis, in part, related to the use of high-dose glucocorticoids and other therapies that adversely affect the skeleton.

The systemic and focal joint inflammation that characterizes many of the rheumatologic disorders is frequently accompanied by adverse effects on the skeleton. Much of the attention has focused on the focal bone resorption in articular and periarticular bone associated with disorders such as rheumatoid arthritis, the prototypical inflammatory joint disease. However, it is clear that many of the inflammatory rheumatic disorders have a marked effect on systemic bone remodeling and numerous studies have documented that osteoporosis and increased risk of fracture account for a substantial component of the morbidity associated with these conditions. In part, these adverse skeletal effects may be related to the therapies used to treat these diseases, as well as poor nutritional intake and other factors such as reduced physical activity that frequently occur in this population. This review will focus on the inflammatory rheumatologic disorders that target the articular and periar-

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. RHEUMATOID ARTHRITIS Rheumatoid arthritis (RA) represents an excellent model for gaining insights into the effects of local as well as systemic consequences of inflammatory processes on skeletal tissue remodeling. RA is a relatively common systemic inflammatory disorder affecting between 1 and 2% of the adult population throughout the world. The etiology of RA

351

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

352

STEVEN R. GOLDRING

FIGURE 1

Schematic depiction of the articular surface of a diarthrodial joint with rheumatoid pannus invading the articular cartilage and bone.

is unknown, but there is considerable evidence that both genetic and environmental factors are involved in its pathogenesis [1 – 3]. The hallmark of RA is the development of a chronic inflammatory process that targets the synovial lining of diarthrodial joints. The earliest changes involve the proliferation of the synovial lining cells which consist of a heterogeneous population of macrophage-like cells (A cells) and synovial fibroblasts (B cells). This process is accompa-

FIGURE 2

nied by endothelial cell proliferation with neovascularization and perivascular infiltration with lymphocytes, plasma cells, and activated macrophages consistent with local activation of an intense local immune reaction [1,4]. As the disease progresses, the inflamed synovial tissue migrates over the articular surface, forming the so-called pannus (Fig. 1). At the interface between the pannus and hyaline cartilage there is focal destruction of the cartilage matrix. At sites where the inflamed synovium comes into contact with the bone there is evidence of activation of osteoclastic bone resorption [1,5 – 8]. This process leads to focal destruction of the articular bone, producing the characteristic bone erosions that can be visualized radiographically (Fig. 2). In addition to these focal bone changes, there is also evidence of a generalized loss of periarticular bone in joints affected by the inflammatory synovial process. This type of bone loss also appears to be mediated by increased osteoclastic activity, although the sites of bone loss are frequently not in direct continuity with the pannus or inflamed synovium. This suggests that other mechanisms, including reduced joint mobility and immobilization, that accompany the joint inflammation may be involved in the enhanced bone resorption.

Radiograph of the hands demonstrating the characteristic radiological features of rheumatoid arthritis. Periarticular osteopenia is present in the proximal interphalangeal and metacarpal phalangeal joints and in the wrists. An arrow identifies a characteristic marginal joint erosion. There is joint space narrowing in several joints reflecting the destruction of the articular cartilage by the invading pannus.

CHAPTER 54 Osteoporosis Associated with Rheumatologic Disorders

Although the focal joint pathology accounts for a major component of the disease morbidity, RA is clearly a systemic illness that can produce adverse effects on extraarticular organs and is frequently accompanied by generalized systemic features of inflammation. In this respect the entire skeleton may be affected by this inflammatory process, resulting in progressive loss of bone mass leading to an increased risk for fracture. Thus, although the presence of focal bone erosions and juxtaarticular osteopenia have been regarded as the principal radiographic features of RA, generalized osteoporosis represents an additional skeletal manifestation of this disorder.

A. Focal Subchondral and Marginal Bone Erosions As suggested in the Introduction, the presence of articular bone erosions has been considered the radiographic hallmark of RA. The detection of these bone changes has been shown to have considerable clinical utility in diagnosis and monitoring of patients with joint inflammation. Studies employing radiographic analysis of progressive changes in articular and periarticular bone have helped to establish that, in general, patients with more extensive bone erosions exhibit more severe disease manifestations and demonstrate poorer clinical outcomes [9,10]. Additional interest in focal bone changes in RA has been generated by findings from recent clinical trials that have shown that certain treatment regimens can retard or even prevent the development and progression of focal bone loss [11 – 15]. These data indicate the utility of assessment of bone changes as an endpoint for evaluating treatment response and efficacy. Analysis of results from radiographic surveys of patients with RA have generated some confusion regarding the relationship between synovitis and focal bone loss. This is related to the failure in some instances to demonstrate a direct correlation between clinically evident synovitis and bone loss [16,17]. This has raised questions regarding the precise histopathological mechanisms responsible for the pathogenesis of the articular bone erosions. In part, the limitations in the sensitivity of conventional radiography may have contributed to this apparent discrepancy. The introduction of magnetic resonance imaging (MRI), which represents an excellent tool for detecting synovitis, has helped to establish that focal articular bone changes can be detected in a majority of patients beginning very early in the course of the disease and that the bone changes in RA are directly associated with the presence of synovitis [18 – 20]. Much of our present understanding of the pathogenesis of focal bone erosions in RA has come from careful histopathological examination of joint tissues from patients with RA. The early work of Bromley and Woolley has been particularly informative in providing insights into the cellular

353 mechanisms responsible for the pathogenesis of focal bone erosions [21 – 23]. They noted the presence of multinucleated cells with phenotypic features of osteoclasts in resorption lacunae at the pannus – bone interface. These observations have been supported by the work of Leisen et al. [24]. Using electron microscopy, they demonstrated resorption bays typical of osteoclastic activity in areas of pannus invasion into calcified cartilage and subchondral bone in metacarpal heads taken from patients with RA. In our own studies, we have used in situ hybridization techniques to demonstrate that the multinucleated (and some mononuclear) cells in resorption lacunae at the bone – pannus junction express the full repertoire of phenotypic markers of fully differentiated osteoclasts, including the expression of tartrate resistant acid phosphatase, cathepsin K, and calcitonin receptor mRNA [6]. The osteoclast-like cells were also examined for the expression of mRNA for the parathyroid hormone (PTH) receptor. No PTH receptor message was detected in these osteoclast-like cells, consistent with the observations of others who speculate that PTH does not act directly on osteoclasts [25 – 27]. Of interest, cells within the bone marrow and some cells lining the bone surface immediately adjacent to resorption bays expressed mRNA for the PTH receptor. These cells exhibited morphological features of osteoblasts and thus could represent a target for PTH action. An additional possibility is that they could be targets for parathyroid hormone-related protein (PTHrP), the humoral factor associated with the hypercalcemia of malignancy [28]. PTHrP has been detected in RA synovial fluids and in synovial tissues [29,30]. The PTHrP produced within the RA synovium could act on the PTH receptor-positive cells on the bone surface stimulating them to release products that enhance osteoclast-mediated resorption. The origin of the osteoclast-like cells in the RA lesions remains speculative. There is evidence that they may be derived from mononuclear cell precursors present within the inflamed synovium. Interaction with the bone surface, as well as the effects of cytokines produced locally within the RA synovium, may combine to induce these cells to differentiate into osteoclasts [6,21,23]. The mononuclear osteoclast precursors may also be derived directly from the circulation. In addition, in regions of subchondral or juxtaarticular trabecular bone, the bone marrow may also provide a source of osteoclast precursors. We have frequently observed infiltration of the bone marrow with inflammatory cells in regions adjacent to the subchondral bone of inflamed joints [6,7]. At these sites the bone marrow may be replaced with a loose network of fibrous connective tissue which is not in direct continuity with the invading pannus. These regions may correspond to the zones of marrow edema adjacent to synovial inflammation described by McGonagle et al. using MRI [19]. At these sites, cells expressing phenotypic features of osteoclast precursors, including TRAP and cathepsin K activity, are frequently present immediately adjacent to CTR-positive

354 osteoclast-like cells in resorption bays. These inflammatory marrow changes could be produced by invasion of the marrow with inflammatory cells that have gained access to the marrow space from sites of pannus erosion through the cortical bone. Cytokines produced by these cells could then act in a paracrine fashion to induce osteoclast formation and focal bone resorption. Additional analyses of the cell types at the bone – pannus junction and characterization of cells isolated from rheumatoid synovium lend further support to the concept that cells with phenotypic features and functional activities of authentic osteoclasts are responsible for at least a component of the focal bone resorption that characterizes the RA synovial lesion [31,32]. The demonstration of cells with phenotypic features of osteoclasts in resorption bays at the bone – pannus junction and within the marrow space does not exclude the possibility that other cell types could participate in the focal bone resorption. For example Hummel et al. have shown that synovial fibroblasts express mRNA for cathepsin K and have suggested that these cells may contribute directly to focal bone resorption [33]. Direct examination of macrophages and other cells types for bone-resorbing activity indicates that they do have the capacity to resorb bone [34 – 36]. However, their resorptive activity is very limited compared to authentic osteoclasts, and it is likely that osteoclasts mediate the major component of focal bone resorption in RA. Insights into the unique capacity of the rheumatoid pannus to induce osteoclast formation and osteoclast-mediated bone resorption has come from analysis of RA synovium for the presence of cytokines and other products implicated in the regulation of osteoclast differentiation and activity. These studies have shown that RA synovial tissues produce abundant quantities of IL-1  and , IL-6, IL-11, M-CSF, TNF-, and PTHrP [29,30,37 – 41]. Results from clinical trials that have targeted IL-1 and TNF- have provided the most compelling evidence that these cytokines play a critical role in the pathogenesis of focal bone erosions [13,42]. As described above, therapies that interfere with the activity of these cytokines have been shown to retard or prevent the progression of focal bone loss [11 – 15]. Further evidence that these cytokines play a role in the pathogenesis of focal bone resorption associated with inflammatory joint diseases is provided from animal models of arthritis using gene transfer or transgenic mouse models. These studies demonstrate that overexpression of TNF- or IL-1 in normal joints results in pannus formation that leads to focal bone and cartilage destruction [43, 44]. Of interest, IL-1 or TNF blockade ameliorate the arthritis in the collagen-induced arthritis model, but only the IL-1 blockade prevents the development of bone erosions, suggesting that it is possible to differentiate between the effects of these cytokines on synovial inflammation and osteoclast-mediated bone resorption [45,46].

STEVEN R. GOLDRING

RA synovial tissue is a source of another recently described factor that regulates osteoclast differentiation and activity [47 – 50]. This factor has been variously identified as osteoclast differentiation factor (ODF), RANK ligand (RANKL), and osteoprotegerin ligand (OPGL) [51 – 55]. It is a member of the TNF-ligand family of cytokines and studies demonstrate that many of the factors that enhance osteoclast formation or activity mediate their effects on bone via upregulating the expression of this cytokine [54, 55] (see Chapters 3 and 13). ODF/RANKL was originally cloned and characterized as a product of activated T cells and was designated as tumor necrosis factor-related activation-induced cytokine (TRANCE). Of interest, T cells within the RA synovium are a primary source of ODF/RANKL, as are synovial fibroblasts [47 – 50]. Direct evidence that ODF/RANKL plays an important role in the pathogenesis of focal bone erosions in inflammatory arthritis is provided by the recent studies of Kong et al. [47], who showed, using a rat model of adjuvant arthritis, that treatment with osteoprotegerin (OPG), the decoy receptor for ODF/RANKL, almost completely blocked the development of cortical and trabecular bone loss. Of interest, there was minimal effect of this treatment on joint inflammation or pannus, providing direct evidence that bone resorption can be inhibited even in the absence of an effect on the synovial inflammation. Although the optimal therapy for RA should suppress or prevent the progression of the inflammatory synovitis, therapies that reduce the damage to bone or articular cartilage, even in the absence of effects on the synovitis, have a potentially important role in the treatment strategy. Agents that specifically target osteoclastic bone resorption have been evaluated in several clinical studies in RA patients and in animal models of arthritis with the goal of blocking or reducing the progression of focal bone erosions. As will be discussed in the section on generalized osteoporosis, two studies have assessed the efficacy of bisphosphonates on the progression of focal bone loss in RA patients. Although treatments resulted in an increase in bone density and evidence of reduced urinary hydroxyproline excretion, there was no apparent alteration in the progression of the focal bone erosions [56,57]. In contrast, several studies in animal models of arthritis have been shown to retard or even prevent focal bone loss [58,59]. Although the validity of these models as surrogates for RA can be challenged, they provide a convincing argument for additional trials with antiresorptive therapies in RA.

B. Periarticular Bone Loss Periarticular osteopenia characteristically occurs in joints affected by active synovitis. The juxtaarticular bone loss usually precedes the appearance of focal bone erosions.

CHAPTER 54 Osteoporosis Associated with Rheumatologic Disorders

Histomorphometric examination of bone from sites of periarticular osteopenia demonstrate increased bone remodeling with a relative increase in resorption over formation [60]. The presence of increased osteoclastic resorption surfaces indicates that the juxtaarticular bone loss is mediated via osteoclasts. Since this resorption process occurs at sites removed from the direct pannus invasion or sites of inflammation, the responsible pathophysiological mechanism appears to differ from that associated with focal articular bone erosions. Elaboration of proinflammatory mediators within the marrow, hypervascularity of the synovium and surrounding tissue, and the effects of joint immobility, have been implicated as possible mechanisms for explaining this radiographic and histologic finding [1,37,40,61 – 63].

C. Generalized Bone Loss Several studies have provided evidence for generalized bone loss in the axial spine and appendicular skeleton distant from synovial-lined joints in patients with RA [64 – 72]. Identification of the independent role of the systemic inflammatory process on bone remodeling has been difficult because of the presence of multiple confounding factors, including reduced mobility, variation in the level of disease activity and duration, and the concomitant use of glucocorticoids and immunosuppressive therapies. In addition, many of the epidemiological studies have been flawed by design problems such as small numbers of subjects and inconsistent methods of measurement. Despite these limitations there does appear to be an increase in the risk for fracture among patients with RA [67,73 – 77]. Several studies have suggested that in patients with RA there is a relationship between systemic osteoporosis and disease activity [64,66,71,78 – 80]. For example, in a longitudinal study, Sambrook et al. [64] found that joint count and C reactive protein (CRP), correlated with trabecular bone loss. In the study by Hooyman et al., regression analysis suggested that, in addition to duration of disease and disability level, age, and glucocorticoid exposure were the major risk factors for fractures [75]. Similar findings were obtained by Michel et al., who analyzed data from five Arthritis, Rheumatism and Aging Medical Information System centers [76]. Using multivariate analysis, they identified an association between fracture risk and the years taking prednisone, disability, age, lack of physical activity, female sex, disease duration, impaired grip strength, and low body mass. More recently, bone mineral density (BMD) was assessed in 394 RA patients from the Oslo County Rheumatoid Arthritis Registry [81]. The authors found a twofold increase in osteoporosis (defined as a T score of equal to or greater than 2.5 standard deviations below the mean) in female patients at all ages examined [20 – 70 years]. A linear regression model was used to determine

355 individual predictors of BMD. In the final model, older age, low body weight, current use of corticosteroids, and lower functional status (assessed by modified HAQ score) were significant predictors of reduced bone mass. Although epidemiological studies indicate that glucocorticoid use in patients with RA is a risk factor for systemic osteoporosis, the effects of glucocorticoids on bone mass changes have been difficult to determine. In part, this is related to the capacity of low doses of glucocorticoids to improve symptoms and functional capacity in patients with RA which would be expected to have a beneficial effect on bone density, especially in the spine. In support of this concept, in cross-sectional and longitudinal studies by Sambrook, patients with RA had lower levels of BMD than did normal controls, but no difference was noted in the progression of bone loss in RA patients treated with or without low-dose glucocorticoids [65,82]. Lane et al. examined the association of steroid use with BMD in a community-based sample of ambulatory Caucasian women 65 and over with or without RA [83]. They found that women with RA who were current users of glucocorticoids had the lowest bone density at both the appendicular sites and at the hip. Women who had never used steroids also had a reduced bone density compared to the control population. They concluded that women with RA have lower appendicular and axial bone mass that is not attributable to the use of steroids. The lower bone density in the women on steroids could be accounted for by their lower functional status. Other studies, however, have suggested that even low dose glucocorticoids may have a detrimental effect on bone density. For example, in a cross-sectional analysis employing quantitative computed tomography, Laan et al. [84] demonstrated that low doses of glucocorticoids (mean dose, 6.8 mg prednisone/day) were associated with reduced bone density compared to non-steroid-treated patients. In a case – control analysis of 112 RA patients by Saag et al., long-term low-dose glucocorticoid was a significant predictor of adverse events, including fractures, gastrointestinal events, and infections [85]. There are conflicting results with respect to the mechanisms responsible for the reduced bone mass in patients with RA. In a recent longitudinal study, Gough et al. analyzed biochemical markers of bone turnover in a series of 232 patients with RA [86]. Over a 2-year period they detected a significantly greater rate of bone loss in the RA patients compared to the controls (greater than 3% at the spine and 5% at the hip). Results indicated that there was an increase in the rate of bone resorption in patients with RA based on elevations in urinary pyridinoline and deoxypyridinoline excretion. Of interest, the levels of these markers of bone resorption were highly correlated with Creactive protein (CRP) levels. The increase in bone resorption was not accompanied by changes in bone formation markers as assessed by serum alkaline phosphatase activity

356 and procollagen I carboxyterminal propeptide concentrations. Similar observations have been made in other studies which reported increases in urinary markers of bone resorption, particularly in patients with active disease who lost bone quickly, as determined by changes in BMD [79, 87 – 90]. The findings from these studies must be interpreted with caution, since in some instances the patients were receiving corticosteroids. In contrast to the results from studies in which bone markers have been used to assess bone remodeling, histomorphometric analysis of bone biopsies from patients with RA indicate that the cellular basis for the reduced bone mass is related to a decrease in bone formation rather than an increase in bone resorption [91 – 93]. These discrepancies could reflect differences in the stages of the disease in which bone remodeling was being evaluated as well as the confounding effects of corticosteroid treatment and disease activity in the patient populations.

III. JUVENILE RHEUMATOID ARTHRITIS Juvenile rheumatoid arthritis (JRA) is a systemic inflammatory joint disorder characterized by chronic synovitis affecting diarthrodial joints. Similar to the findings in patients with RA, children with JRA show evidence of multiple distinct patterns of bone loss, including focal marginal erosions, juxtaarticular osteopenia, and generalized osteoporosis. In some instances the peripheral joint inflammation is accompanied by enthesopathy and sacroiliitis. The disease may develop at any age during childhood and tends to affect girls more frequently than boys, although the sex ratios vary in the different subsets of the disease. Three major patterns or subsets of JRA have been described: pauciarticular, polyarticular, and systemic. It is not clear whether these subsets represent different disease entities that share in common their ability to produce joint inflammation or whether they are manifestations of a varying response to common pathogenic factors. There is little information concerning the histopathologic events associated with focal bone erosions and juxtaarticular osteopenia in children with JRA. It is likely, however, that the pathological processes responsible for the bone and cartilage destruction are similar to those that have been described in adults with RA. As in RA, generalized loss of bone mass is a common feature of all of the forms of JRA. The risk factors are similar to those associated with systemic osteoporosis in RA, including the effects of medications, reduced level of physical activity, dietary deficiencies and the adverse systemic effects of inflammatory mediators and cytokines [94]. The importance of disease activity on bone mass changes in children with JRA is demonstrated by the studies of Reed and coworkers [95], who

STEVEN R. GOLDRING

evaluated radial BMD over a 3-year period in children with JRA. Improvement in the disease activity was associated with an increase in BMD, although levels remained below the normal values. A unique aspect of the skeletal pathology in JRA is the effect of the inflammatory process on skeletal growth. Because the joint disease affects children during the period of skeletal acquisition there is often dramatic linear growth retardation in addition to the adverse effects on bone remodeling [96,97]. The suppression of bone formation in children, especially during the pubertal growth period, may have a major adverse effect on the achievement of optimal peak bone mass. The failure to achieve optimal peak bone mass predisposes these individuals to an increased fracture risk in adulthood [98,99]. Studies by Hillman et al. [100], who analyzed biochemical markers of bone remodeling and total body calcium in 44 children with active polyarticular or pauciarticular JRA, revealed a low bone formation rate with an overall reduction in bone remodeling. Similar results have been reported by Pepmueller et al. [101], who measured BMD and markers of bone remodeling in 41 children with JRA. Bone density was reduced at all sites. Low levels of osteocalcin and bone alkaline phosphatase were consistent with reduced bone formation. They noted that laboratory markers of disease activity were highly correlated with decreases in markers of bone formation but not with those of bone resorption. In general, the findings were similar in children with pauciarticular and polyarticular JRA, although the reduction in bone mass was greatest in the children with the polyarticular form of arthritis. Hopp et al. [102] studied spinal bone density in 20 children with active JRA and found reduced values in postpubertal girls compared with healthy controls. They noted that adolescents with active disease may be particularly vulnerable to the impact of the inflammatory process, in part because of the rapid skeletal acquisition that is normally associated with this stage of development. Of interest, bone mass in prepubertal girls did not differ from control at any of the skeletal sites examined. Henderson et al. [103] studied total body bone mineral content in noncorticosteroid-treated postpubertal females with JRA. They found that 30% of the children with mild to moderate JRA had low bone mass. Of interest, using stepwise linear regression they found that the predictor variable that significantly contributed to total-body bone mineral content was lean body mass.

IV. SERONEGATIVE SPONDYLOARTHROPATHIES The seronegative spondyloarthropathies represent a heterogeneous group of inflammatory disorders that include ankylosing spondylitis, reactive arthritis, Reiter’s syndrome,

CHAPTER 54 Osteoporosis Associated with Rheumatologic Disorders

spondylitis and arthritis associated with psoriasis or inflammatory bowel disease, and juvenile-onset spondyloarthropathy. Although these disorders may produce inflammation of peripheral joints, inflammation of the entheses (sites of tendinous or ligamentous attachment to bone), especially in the axial spine, represents the pathological hallmark of the spondyloarthropathies. Synovial hyperplasia, lymphoid infiltration, and pannus formation is frequently observed in affected joints. However, the inflammation is usually restricted to a limited number of joints and the pattern of distribution is typically asymmetrical, affecting distal as well as proximal joints. Insights into the topograhical localization of the inflammatory joint pathology has been provided by the recent introduction of MRI imaging techniques that utilize fat-supression sequences that are capable of delineating sites of bone and connective tissue inflammation [104,105]. With these techniques it has been possible to confirm that the inflammatory process that accounts for the initial joint pathology frequently begins at the enthesis. These changes are visualized as focal soft tissue edema that is maximal at regions adjacent to the entheseal insertions in peripheral joints. Similar changes have been observed in association with spondylitis suggesting a common pathophysiological process between spinal disease and peripheral joint inflammation [106]. There have been relatively few studies of the histopathology associated with the entheseal and synovial inflammation in the spondyloarthropathies. In addition to the differential localization of the inflammation to include the enthesis, the synovial inflammatory process in the spondyloarthropathies, unlike the joint inflammation in RA, may be accompanied by evidence of increased bone formation. Braun et al. used computer-assisted tomography to obtain biopsies of the sacroiliac joint in a series of patients with ankylosing spondylitis [107]. Immunohistologic examination of the tissue revealed dense infiltrates of T lymphocytes (CD4 and CD8) and macrophages (CD 14) in the synovial lining accompanied by localized nodules containing active foci of endochondral ossification. In situ hybridization demonstrated an abundant message for TNF- in the inflammatory cells. Of interest, abundant TGF-2 mRNA was expressed in cells at sites of new bone formation. These authors suggested that local production of bone growth factors such as TGF- by the inflammatory cells within the synovium could be responsible for the new bone formation. This process could account for bony ankylosis of the sacroiliac joints that is characteristic of ankylosing spondylitis and other forms of seronegative spondyloarthropathy. A similar process could contribute to the formation of syndesmophytes at the margins of the adjoining vertebral bodies with resultant ankylosis of the spine. Further evidence supporting a role for bone growth factors in the pathogenesis of bony ankylosis in the seronegative spondyloarhtropathies is provided by the

357 characterization of transgenic animals overexpressing bone morphogenic factor-6 (BMP-6) [108,109]. These animals develop psoriatic skin lesions and an osteoarthropathy similar to the joint and spine pathology of psoriasis. Despite the tendency of patients with spondylitis to develop bony ankylosis of the spine, there is evidence of vertebral osteopenia and an increased incidence of fractures (Fig. 3) [110 – 113]. Spencer has suggested that the decreased vertebral bone density is related to the effects of loss of spinal mobility [111]. However, this hypothesis has been challenged by Will et al. [113], who evaluated a series of patients with early ankylosing spondylitis using dualphoton absorptiometry and observed a significant reduction in bone mineral density in the lumbosacral spine and hip early in disease before bony ankylosis and spinal

FIGURE 3 Radiograph of the lumbar spine from a patient with ankylosing spondylitis. Note the presence of diffuse osteopenia. There are small erosions at the margins of L1/L2 with reactive bone formation.

358 immobility developed. They speculated that the reduced bone mass was related to the adverse effects of inflammation on bone remodeling. The reduced spinal bone mass in patients with spondylitis is associated with an increased risk for fracture. Ralston et al. [112] evaluated prospectively a group of 111 patients with ankylosing spondylitis. Fifteen patients developed radiographic evidence of vertebral compression fractures. These patients tended to have a greater degree of spinal deformity and less spinal mobility than individuals without fractures. He concluded that vertebral compression fractures secondary to spinal osteoporosis were a common but frequently unrecognized complication of ankylosing spondylitis and that they contributed to the pathogenesis of spinal deformity and back pain in this population. Of interest, bone mineral of the appendicular skeleton was normal, suggesting that osteoporosis in these patients is primarily localized to the axial spine. Similar observations have been reported by Devogelaer et al. [114], who observed that males with ankylosing spondylitis tended to develop significant bone loss in the vertebral bodies but exhibited bone density comparable to the controls in the appendicular skeleton.

V. SYSTEMIC LUPUS ERYTHEMATOSUS Systemic lupus erythematosus (SLE) is a systemic autoimmune disorder characterized by the production of a diverse array of so-called autoantibodies that are directed at a spectrum of self molecules found in the nucleus, cytoplasm, and on the surface of target cells. The pathologic changes in lupus occur throughout the body and are manifest by inflammation of blood vessels that is at least in part mediated by immune complex deposition. This inflammatory process can affect any organ system, but arthritis and arthralgias are the most common presenting manifestation of SLE. The joint inflammation may affect any joint, but the most frequently affected sites are the small joints of the hands, wrists, and knees. Unlike RA, joint inflammation in SLE does not produce focal bone erosions, although joint deformities may occur due to the development of joint capsule and ligamentous laxity that has been attributed to the effects of persistent inflammation in these soft tissues. This pattern of nonerosive arthritis accompanied by joint deformities has been referred to Jaccoud’s arthritis [115,116]. Van Vugt evaluated a series of 176 patients with SLE in an attempt to define the different patterns of Jaccoud’s arthritis. Three of the patients had an erosive form of arthritis indistinguishable from RA. These patients, however, were atypical and may represent a subset of patients with RA with overlap features of SLE. Eight patients had more typical features of Jaccoud’s arthropathy, characterized by severe deformity of the hands

STEVEN R. GOLDRING

(ulnar deviation, swan neck deformties, and Z-deformity of the thumb) and feet with multiple nonerosive subluxations. The patients experienced mild aching but exhibited little or no evidence of synovitis. In most of the patients the onset of the arthritis preceded the diagnosis of lupus. A striking relationship was observed between Jaccoud’s arthritis and fetal loss, arterial and venous thrombosis, and the presence of antiphospholipid antibodies. Jaccoud’s arthritis has also been described as a rare complication of rheumatic fever in which it is manifest by painless deforming arthritis affecting the hands and feet [117, 118]. All of the described cases have shown obvious rheumatic heart disease, although some patients may not recall distinct features of rheumatic fever. Distinctive socalled “hook” erosions may occur on the radial palmar aspect of the metacarpal heads. These can be easily distinguished from the marginal erosions of RA by their location and absence of other radiographic features of RA [119]. Similar patterns of nondestructive deforming arthritis have been described in sarcoidosis, Parkinson’s disease, so the specificity of Jaccoud’s arthropathy for SLE or rheumatic fever is not clear. Nevertheless, recognition of this form of arthropathy in SLE is of considerable clinical importance, particularly in light of its reported association with antiphospholipid syndrome [116]. Several cross-sectional studies have documented the increased incidence of bone loss and vertebral compression fractures in women with SLE [120 – 122]. As in patients with RA and other systemic inflammatory disorders, it is difficult to dissociate the effects of disease activity from other confounding variables that affect bone remodeling, including nutritional factors, physical activity, menopausal status, and the effects of therapies, especially corticosteroids. A recent analysis of fracture frequency in women with SLE indicates nearly a five-fold increase in the fracture occurrence in women with lupus compared to an age-matched population in the United States. These data were derived from the analysis of a cohort of 702 living women from the University of Pittsburgh Medical Center Lupus Registry. The women were followed for a total of 5951 person-years and fractures and associated risk factors for osteoporosis were ascertained by self-report and verified in a subset of patients. In this study, older age at lupus diagnosis and longer use of corticosteroids were associated with increased fracture risk. Although this study has several design limitations, it does provide evidence that the reduced bone mass in SLE patients is accompanied by an increased risk for fracture.

Acknowledgments I thank Daniel I. Rosenthal, MD Professor of Radiology, Massachusetts General Hospital, Boston, Massachusetts, for providing the radiographs for Figs. 2 and 3.

CHAPTER 54 Osteoporosis Associated with Rheumatologic Disorders

References 1. E. D. J. Harris, Rheumatoid arthritis. Pathophysiology and implications for therapy. N. Engl. J. Med. 322, 1277 – 1289 (1990). 2. C. M. Weyand, T. G. McCarthy, and J. J. Goronzy, Correlation between disease phenotype and genetic heterogeneity in rheumatoid arthritis. J. Clin. Invest. 95, 2120 – 2126 (1995). 3. D. A. Fox, The role of T cells in the immunopathogenesis of rheumatoid arthritis. Arthritis Rheum. 40, 598 – 609 (1997). 4. W. P. Arend, The pathophysiology and treatment of rheumatoid arthritis. Arthritis Rheum. 40, 595 – 597 (1997). 5. S. M. Krane, W. Conca, M. L. Stephenson, E. P. Amento, and M. B. Goldring, Mechanisms of matrix degradation in rheumatoid arthritis. Ann. NY Acad. Sci. 580, 340 – 354 (1990). 6. E. M. Gravallese, Y. Harada, J. T. Wang, A. H. Gorn, T. S. Thornhill, and S. R. Goldring, Identification of cell types responsible for bone resorption in rheumatoid arthritis and juvenile rheumatoid arthritis. Am. J. Pathol. 152, 943 – 951 (1998). 7. S. R. Goldring and E. M. Gravallese, Mechanisms of bone loss in inflammatory arthritis: Diagnosis and therapeutic implications. Arthritis Res. 2, 33 – 37 (2000). 8. S. R. Goldring and R. P. Polisson, Bone disease in rheumatological disorders, In “Metabolic Bone Disease” (L. Avioli and S. M. Krane, eds.), pp. 621 – 635. Academic Press, San Diego, (1998). 9. D. van Zeben, J. M. W. Hazes, A. H. Zwinderman, J. P. Vandenbroucke, and F. C. Breedveld, Factors predicting outcome of rheumatoid arthritis: Results of a followup study. J. Rheumatol. 20, 1288 – 1296 (1993). 10. J. T. Sharp, F. Wolfe, D. M. Mitchell, and D. A. Bloch, The progression of erosion and joint space narrowing scores in rheumatoid arthritis during the first twenty-five years of disease. Arthritis Rheum. 34, 660 – 668 (1991). 11. L. W. Moreland, M. H. Schiff, S. W. Baumgartner, E. A. Tindall, R. M. Fleischmann, K. J. Bulpitt, A. L. Weaver, E. C. Keystone, D. E. Furst, P. J. Mease, E. M. Ruderman, D. A. Horwitz, D. G. Arkfeld, L. Garrison, D. J. Burge, C. M. Blosch, M. L. M. Lange, N. D. McDonnell, and M. E. Weinblatt, Etanercept therapy in rheumatoid arthritis: A randomized, controlled trial. Ann. Intern. Med. 130, 478 – 86 (1999). 12. R. Maini, E. W. St Clair, F. C. Breedveld, D. Furst, J. Kalden, M. Weisman, J. Smolen, P. Emery, G. Harriman, M. Feldmann, and P. Lipsky, Infliximab (chimeric anti-tumour necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: A randomised phase III trial. ATTRACT Study Group. Lancet 354, 1932 – 1939 (1999). 13. B. Bresnihan, J. M. Alvaro-Gracia M. Cobby, M. Doherty, Z. Domljan, P. Emery, G. Nuki, K. Pavelka, R. Rau, B. Rozman, I. Watt, B. Williams, R. Aitchison, D. McCabe, and P. Musikic, Treatment of rheumatoid arthritis with recombinant human interleukin-1 receptor antagonist. Arthritis Rheum. 41, 2196 – 2204 (1998). 14. V. Strand, S. Cohen, M. Schiff, A. Weaver, R. Fleischmann, G. Cannon, R. Fox, L. Moreland, N. Olsen, D. Furst, J. Caldwell, J. Kaine, J. Sharp, F. Hurley, and I. Loew-Friedrich, Treatment of active rheumatoid arthritis with leflunomide compared with placebo and methotrexate. Arch. Int. Med. 159, 2542 – 2550 (1999). 15. R. Rau and G. Herborn, Healing phenomena of erosive changes in rheumatoid arthritis patients undergoing disease-modifying antirheumatic drug therapy. Arthritis Rheum. 39, 162 – 168 (1996). 16. D. Mulherin, O. Fitzgerald, and B. Bresnihan, Clinical improvement and radiological deterioration in rheumatoid arthritis: Evidence that the pathogenesis of synovial inflammation and articular erosion may differ. Br. J. Rheumatol. 35, 1263 – 1268 (1996). 17. J. R. Kirwan, The relationship between synovitis and erosions in rheumatoid arthritis. Br. J. Rheumatol. 36, 225 – 228 (1997).

359 18. D. Foley-Nolan J. P. Stack, M. Ryan, U. Redmond, C. Barry, J. Ennis, and R. J. Coughlan, Magnetic resonance imaging in the assessment of rheumatoid arthritis — A comparison with plain film radiographs. Br. J. Rheumatol. 30, 101 – 106 (1991). 19. D. McGonagle, P. G. Conaghan, P. O’Connor, W. Gibbon, M. Green, R. Wakefield, J. Ridgway, and P. Emery, The relationship between synovitis and bone changes in early untreated rheumatoid arthritis: A controlled magnetic resonance imaging study. Arthritis Rheum. 42, 1706 – 1711 (1999). 20. F. M. McQueen, N. Stewart, J. Crabbe, E. Robinson, S. Yeoman, P. L. J. Tan, and L. McLean, Magnetic resonance imaging of the wrist in early rheumatoid arthritis reveals a high prevalence of erosions at four months after symptom onset. Ann. Rheum. Dis. 57, 350 – 356 (1998). 21. M. Bromley and D. E. Woolley, Chondroclasts and osteoclasts at subchondral sites of erosion in the rheumatoid joint. Arthritis Rheum. 27, 968 – 975 (1984). 22. M. Bromley, W. D. Fisher, and D. E. Woolley, Mast cells at sites of cartilage erosion in the rheumatoid joint. Ann. Rheum. Dis. 43, 76 – 79 (1984). 23. M. Bromley and D. E. Woolley, Histopathology of the rheumatoid lesion: Identification of cell types at sites of cartilage erosion. Arthritis Rheum. 27, 857 – 863 (1984). 24. J. C. C. Leisen, H. Duncan, J. M. Riddle, and W. C. Pitchford, The erosive front: A topographic study of the junction between the pannus and the subchondral plate in the macerated rheumatoid metacarpal head. J. Rheumatol. 15, 17 – 22 (1988). 25. G. A. Rodan and T. J. Martin, Role of osteoblasts in hormonal control of bone resorption — A hypothesis. Calcif. Tissue Int. 33, 349 – 351 (1981). 26. T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endocr. Rev. 13, 66 – 80 (1992). 27. L. G. Raisz, Local and systemic factors in the pathogenesis of osteoporosis. N. Engl. J. Med. 318, 818 – 828 (1988). 28. G. J. Strewler, The physiology of parathyroid hormone-related protein. N. Engl. J. Med. 342, 177 – 185 (2000). 29. K. Okano, T. Tsukazaki, A. Ohtsuru, H. Namba, M. Osaki, K. Iwasaki, and S. Yamashita, Parathyroid hormone-related peptide in synovial fluid and disease activity of rheumatoid arthritis. Br. J. Rheumatol. 35, 1056 – 1062 (1996). 30. J. L. Funk, L. A. Cordaro, H. Wei, J. B. Benjamin, and D. E. Yocum, Synovium as a source of increased amino-terminal parathyroid hormone-related protein expression in rheumatoid arthritis: A possible role for locally produced parathyroid hormone-related protein in the pathogenesis of rheumatoid arthritis. J. Clin. Invest. 101, 1362 – 1371 (1998). 31. B. A. Ashton, I. K. Ashton, M. J. Marshall, and R. C. Butler, Localization of vitronectin receptor immunoreactivity and tartrate resistant acid phosphatase activity in synovium from patients with inflammatory and degenerative arthritis. Ann. Rheum. Dis. 52, 133 – 137 (1993). 32. J. S. Chang, J. M. Quinn, A. Demaziere, C. J. Bulstrode, M. J. Francis, R. B. Duthie, and N. A. Athanasou, Bone resorption by cells isolated from rheumatoid synovium. Ann. Rheum. Dis. 51, 1223 – 1229 (1992). 33. K. M. Hummel, P. K. Petrow, J. K. Franz, U. Muller-Ladner, W. K. Aicher, R. E. Gay, D. Bromme, and S. Gay, Cysteine proteinase cathepsin K mRN. is expressed in synovium of patients with rheumatoid arthritis and is detected at sites of synovial bone destruction. J. Rheumatol. 25, 1887 – 1894 (1998). 34. G. Hattersley and T. J. Chambers, Calcitonin receptors as markers for osteoclastic differentiation: Correlation between generation of boneresorptive cells and cells that express calcitonin receptors in mouse bone marrow cultures. Endocrinology 125, 1606 – 1612 (1989). 35. G. Hattersley and T. J. Chambers, Generation of osteoclastic function in mouse bone marrow cultures: Multinuclearity and tartrate-

360

36. 37.

38.

39. 40.

41. 42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

STEVEN R. GOLDRING resistant acid phosphatase are unreliable markers for osteoclastic differentiation. Endocrinology 124, 1689 – 1696 (1989). T. J. Chambers, B. M. Thomson, and K. Fuller, Resorption of bone by isolated rabbit osteoclasts. J. Cell Sci. 66, 383 (1984). C. Q. Chu, M. Field, S. Allard, E. Abney, M. Feldmann, and R. N. Maini, Detection of cytokines at the cartilage/pannus junction in patients with rheumatoid arthritis: Implications for the role of cytokines in cartilage destruction and repair. Br. J. Rheumatol. 31, 653 – 661 (1992). M. Feldmann, F. Brennan, E. Paleolog, P. Taylor, and R. N. Maini, Anti-tumor necrosis factor alpha therapy of rheumatoid arthritis. Mechanism of action. Eur. Cytokine Network 8, 297 – 300 (1997). M. Feldmann, F. M. Brennan, and R. N. Maini, Role of cytokines in rheumatoid arthritis. Annu. Rev. Immunol. 14, 397 – 440 (1996). B. W. Deleuran, C. Q. Chu, M. Field, F. M. Brennan, P. Katsikis, M Feldmann, and R. N. Maini, Localization of interleukin-1 alpha, type 1 interleukin-1 receptor and interleukin-1 receptor antagonist in the synovial membrane and cartilage/pannus junction in rheumatoid arthritis. Br. J. Rheumatol. 31, 801 – 809 (1992). E. Romas and T. J. Martin, Cytokines in the pathogenesis of osteoporosis. Osteoporosis Int. 7, S47 – S53 (1997). B. Finck, R. Martin, R. Fleischmann, L. Moreland, M. Schiff, and J. Bathon, A phase III trial of etanercept vs methotrexate (MTX) in early rheumatoid arthritis (Enbrel ERA trial) Arthritis Rheum. 42, S117 (1999) [abstract]. D. M. Butler, A. M. Malfait, L. J. Mason, P. J. Warden, G. Kollias, R. N. Maini, M. Feldmann, and F. M. Brennan, DBA/1 mice expressing the human TNF-alpha transgene develop a severe, erosive arthritis: Characterization of the cytokine cascade and cellular composition. J. Immunol. 159, 2867 – 2876 (1997). J. Keffer, L. Probert, H. Cazlaris, S. Georgopoulos, E. Kaslaris, D. Kioussis, and G. Kollias, Transgenic mice expressing human tumour necrosis factor: A predictive genetic model of arthritis. Embo J. 10, 4025 – 4031 (1991). L. A. B. Joosten, M. M. A. Helsen, F. A. J. van de Loo, and W. B. van den Berg, Anticytokine treatment of established type II collageninduced arthritis in DBA/1 mice: A comparative study using antiTNF alpha, anti- IL-1 alpha/beta, and IL-1Ra. Arthritis Rheum. 39, 797 – 809 (1996). L. A. B. Joosten, M. M. A. Helsen, T. Saxne, D. Heinegard, L. B. van de Putte, and W. B. van den Berg, Synergistic protection against cartilage destruction by low dose prednisolone and interleukin-10 in established murine collagen arthritis. Inflammation Res. 48, 48 – 55 (1999). Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). N. J. Horwood, V. Kartsogiannis, J. M. W. Quinn, E. Romas, T. J. Martin, and M. T. Gillespie, Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144 – 150 (1999). E. M. Gravallese, C. Manning, A. Tsay, A. Naito, C. Pan, E. Amento, and S. R. Goldring, Synovial tissue in rheumatoid arthritis is a source of osteoclast differentiation factor. Arthritis Rheum. 43, 250 – 258 (2000). H. Takayanagi, H. Iizuka, T. Juji, T. Nakagawa, A. Yamamoto, T. Miyazaki, Y. Koshihara, H. Oda, K. Nakamura, and S. Tanaka, Involvement of receptor activator of nuclear factor kappa-B ligand/osteoclast differentiation factor in osteoclastogenesis from synoviocytes in rheumatoid arthritis. Arthritis Rheum. 43, 259 – 269 (2000). D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu,

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63. 64.

65.

66.

67.

68.

J. Sullivan N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. I. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL Proc. Natl. Acad. Sci. USA. 95, 3597 – 3602 (1998). H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, M. Goto, S. I. Mochizuki, E. Tsuda, T. Morinaga, N. Udagawa, N. Takahashi, T. Suda, and K. Higashio, A novel molecular mechanism modulating osteoclast differentiation and function. Bone 25, 109 – 113 (1999). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). F. Eggelmeijer, S. E. Papapoulos, H. C. van Paassen, B. A. C. Dijkmans, R. Valkema, M. L. Westedt, J-O. Landman, E. K. J. Pauwels, and F. C. Breedveld, Increased bone mass with pamidronate treatment in rheumatoid arthritis: Results of a three-year randomized double-blind trial. Arthritis Rheum. 39, 396 – 402 (1996). S. H. Ralston, L. Hacking, L. Willocks, F. Bruce, and D. A. Pitkeathly, Clinical, biochemical, and radiographic effects of aminohydroxypropylidene bisphosphonate treatment in rheumatoid arthritis. Ann. Rheum. Dis. 48, 396 (1989). T. Osterman, K. Kippo, L. Lauren, I. Pasanen, R. Hannuniemi, and R. Sellman, A comparison of clodronate and indomethacin in treatment of adjuvant arthritis. Inflammation Res. 46, 79 – 85 (1997). L. Flora, Comparative antiinflammatory and bone protective effects of two diphosphonates in adjuvant arthritis. Arthritis Rheum. 22, 340 – 346 (1979). S. Shimizu, S. Shiozawa, K. Shiozawa, S. Imura, and T. Fujita, Quantitative histologic studies on the pathogenesis of periarticular osteoporosis in rheumatoid arthritis. Arthritis Rheum. 28, 25 – 31 (1985). G. S. Firestein, J. M. Alcaro-Garcia, and R. Maki, Quantitative analysis of cytokine gene expression in rheumatoid arthritis. J. Immunol. 144, 3347 – 3353 (1990). C. Q. Chu, M. Field, M. Feldmann, and R. N. Maini, Localization of tumor necrosis factor alpha in synovial tissues and at the cartilage – pannus junction in patients with rheumatoid arthritis. Arthritis Rheum. 34, 1125 – 1132 (1991). R. B. Mazess and G. D. Whedon, Immobilization and bone. Calcif. Tissue Int. 32, 265 (1983). P. Sambrook, B. Ansel, S. Foster, J. M. Gumpel, R. Hesp, and J. Reeve, Bone turnover in early rheumatoid arthritis; longitudinal bone density studies. Ann. Rheum. Dis. 44, 580 – 584 (1985). P. N. Sambrook, J. A. Eisman, M. G. Yeates, N. A. Pocock, S. Eberl, and G. D. Champion, Osteoporosis in rheumatoid arthritis: Safety of low dose corticosteroids. Ann. Rheum. Dis. 45, 950 – 953 (1986). P. N. Sambrook, A. Eisman, G. Champion, M. G. Yeates, N. A. Pocock, and S. Eberl, Determinants of axial bone loss in rheumatoid arthritis. Arthritis Rheum. 30, 721 – 728 (1987). N. F. Peel, R. Eastell, and R. G. G. Russell, Osteoporosis in rheumatoid arthritis — the laboratory perspective. Br. J. Rheum. 30, 84 – 85 (1991). A. D. Woolf, Osteoporosis in rheumatoid arthritis — The clinical viewpoint. Br. J. Rheum. 30, 82 – 84 (1991).

CHAPTER 54 Osteoporosis Associated with Rheumatologic Disorders 69. J. E. Compston, E. O. Crawley, C. Evans, and M.M. O’Sullivan, Spinal trabecular bone mineral content in patients with non-steroid treated rheumatoid arthritis. Ann. Rheum. Dis. 47, 660 – 664 (1988). 70. A. A. Deodhar and A. D. Woolf, Bone mass measurement and bone metabolism in rheumatoid arthritis: A review. Br. J. Rheumatol. 35, 309 – 322 (1996). 71. A. K. Gough, J. Lilley, S. Eyre, R. Holdin, and P. Emery, Generalized bone loss in patients with rheumatoid arthritis. Lancet 344, 23 – 27 (1994). 72. P. N. Sambrook, T. D. Spector, E. Seeman, N. Bellamy, R. R. Buchanan, D. L. Duffy, N. G. Martin, R. Prince, E. Owen, A. J. Silman, and J. A. Eisman, Osteoporosis in rheumatoid arthritis: A monozygotic co-twin control study. Arthritis Rheum. 38, 806 – 809 (1995). 73. A. M. Beat, D. A. Bloch, and J. F. Fries, Predictors of fractures in early rheumatoid arthritis. J. Rheumatol. 18, 804 – 808 (1991). 74. C. Cooper, P. L. Cook, C. Osmond, L. Fisher, and M. I. D. Cawley, Osteoarthritis of the hip and osteoporosis of the proximal femur. Ann. Rheum. Dis. 50, 540 – 542 (1991). 75. J. R. Hooyman, L. J. Melton, A. M. Nelson, W. M. O’Fallon, and B. L. Riggs, Fractures after rheumatoid arthritis: A population based study. Arthritis Rheum. 27, 1353 – 1361 (1984). 76. B. A. Michel, D. A. Bloch, F. Wolfe, and J. F. Fries, Fractures in rheumatoid arthritis: An evaluation of associated risk factors. J. Rheumatol. 20, 1666 – 1669 (1993). 77. T. D. Spector, G. M. Hall, E. V. McCloskey, and J. A. Kanis, Risk of vertebral fracture in women with rheumatoid arthritis. Br. Med. J. 306, 558 (1993). 78. R. F. Laan, P. L. van-Riel, and L. B. van-de-Putte, Bone mass in patients with rheumatoid arthritis. Ann. Rheum. Dis. 51, 826 – 832 (1992). 79. A. K. Gough, N. F. Peel, R. Eastell, R. L. Holder, J. Lilley, and P. Emery, Excretion of pyridinium crosslinks correlates with disease activity and appendicular bone loss in early rheumatoid arthritis. Ann. Rheum. Dis. 53, 14 – 17 (1994). 80. O. S. Als, A. Gotfredsen, B. J. Riis, and C. Christiansen, Are disease duration and degree of functional impairment determinants of bone loss in rheumatoid arthritis? Ann. Rhem. Dis. 44, 406 – 411 (1985). 81. G. Haugeberg, T. Uhlig, J. A. Falch, J. I. Halse, and T. K. Kvien, Bone mineral density and frequency of osteoporosis in female patients with rheumatoid arthritis: Results from 394 patients in the Oslo County Rheumatoid Arthritis Register. Arthritis Rheum. 43, 522 – 530 (2000). 82. P. N. Sambrook, M. L. Cohen, J. A. Eisman, N. A. Pocock, G. D. Champion, and M. G. Yeates, Effects of low-dose corticosteroid on bone mass in rheumatoid arthritis: A longitudinal study. Ann. Rheum. Dis. 48, 535 – 535 (1989). 83. N. E. Lane, A. R. Pressman, V. L. Star, S. R. Cummings, and M. C. Nevitt, Rheumatoid arthritis and bone mineral density in elderly women. The Study of Osteoporotic Fracture Research Group. J. Bone Miner. Res. 10, 257 – 263 (1995). 84. R. Laan, P van Riel, L van Erning, J. A. Lemmens, S. H. Ruijs, and L. B. van-de-Putte, Vertebral osteoporosis in rheumatoid arthritis patients: Effects of low-dose prednisone therapy. Br. J. Rheumatol. 31, 91 – 96 (1992). 85. K. Saag, K. Rochelle, J. Caldwell, R. Brasington, L. Burmeister, B. Zimmerman, J. Kohler, and D. Furst, Low dose longterm corticosteroid therapy in rheumatoid arthritis: An analysis of serious adverse events. Am. J. Med. 96, 115 – 123 (1994). 86. A. Gough, P. Sambrook, J. Devlin, A. Huissoon, C. Njeh, S. Robbins, T. Nguyen, and P. Emery, Osteoclastic activation is the principal mechanism leading to secondary osteoporosis in rheumatoid arthritis. J. Rheumatol. 7, 1282 – 1289 (1998). 87. D. Black, M. Marabani, R. D. Sturrock, and S. P. Robins, Urinary excretion of the hydroxypyridinium cross links of collagen in patients with rheumatoid arthritis. Ann. Rheum. Dis. 48, 641 – 644 (1989).

361 88. S. P. Robins, P. Stewart, C. Astbury, and H. A. Bird, Measurement of the cross linking compound, pyridinoline, in urine as an index of collagen degradation in joint disease. Ann. Rheum. Dis. 45, 969 – 973 (1986). 89. M. J. Seibel, A. Duncan, and S. P. Robins, Urinary hydroxy-pyridinium crosslinks provide indices of cartilage and bone involvement in arthritic diseases. J. Rheumatol. 16, 964 – 970 (1989). 90. G. M. Hall, T. D. Spector, and P. D. Delmas, Markers of bone metabolism in postmenopausal women with rheumatoid arthritis. Effects of corticosteroids and hormone replacement therapy. Arthritis Rheum. 38, 902 – 906. 91. H. Kroger, I. Arnala, and E. M. Alhava. Bone remodeling in osteoporosis associated with rheumatoid arthritis. Calcif. Tissue Int. 49, S90 (1991). 92. R. W. E. Mellish, M. M. O’Sullivan, N. J. Garrahan, and J. E. Compston, Iliac crest trabecular bone mass and structure in patients with non-steroid treated rheumatoid arthritis. Ann. Rheum. Dis. 46, 830 – 836 (1987). 93. J. E. Compston, S. Vedi, P. I. Croucher, N. J. Garrahan, and M. M. O’Sullivan, Bone turnover in non-steroid treated rheumatoid arthritis. Ann. Rheum. Dis. 53, 163 – 166 (1994). 94. J. T. Cassidy, C. B. Langman, S. Allen, and L. S. Hillman, Bone mineral metabolism in children with juvenile rheumatoid arthritis. Pediatr. Clin. North Am. 42, 1017 – 1033 (1995). 95. A. Reed, M. Haugen, L. Pachman, and C. Langman, Repair of osteopenia in children with juvenile rheumatoid arthritis. J. Pediatr. 122, 693 – 696 (1993). 96. B. M. Ansell and E. G. L. Bywaters, Growth in Still’s Disease. Ann. Rheum. Dis. 15, 295 – 318 (1956). 97. B. H. Bernstein, D. Stobie, B. H. Singsen, K. Koster-King, H. K. Kornreich, and V. Hanson, Growth retardation in juvenile rheumatoid arthritis (JRA). Arthritis Rheum. 20, 212 – 216 (1977). 98. B. W. D. Badley and B. M. Ansell, Fracture’s in Still’s disease. Ann. Rheum. Dis. 19, 135 – 138 (1960). 99. S. Varonos, B. M. Ansell, and J. Reeve, Vertebral collapse in juvenile chronic arthritis: Its relationship with glucocorticoid therapy. Calcif. Tissue Int. 41, 75 – 78 (1987). 100. L. Hillman, J. T. Cassidy, L. Johnson, D. Lee, and S. Allen, Vitamin D metabolism and bone mineralization in children with juvenile rheumatoid arthritis. J. Pediatr. 124, 910 – 916 (1994). 101. P. H. Pepmueller, J. T. Cassidy, S. H. Allen, and L. S. Hilman, Bone mineralization and bone mineral metabolism in children with juvenile rheumatoid arthritis. Arthritis Rheum. 39, 746 – 757 (1996). 102. R. Hopp, J. A. Degan, J. C. Gallagher, and J. T. Cassidy, Estimation of bone mineral density in children with juvenile rheumatoid arthritis. J. Rheumatol. 18, 1235 – 1239 (1991). 103. C. J. Henderson, B. L. Specker, R. I. Sierra, B. N. Campaigne, and D. J. Lovell, Total-body bone mineral content in non-corticosteroid treated postpubertal females with juvenile rheumatoid arthritis. Arthritis Rheum. 43, 531 – 540 (2000). 104. D. McGonagle, P. G. Conaghan, and P. Emery, Psoriatic arthritis. A unified concept twenty years on. Arthritis Rheum. 42, 1080 – 1086 (1999). 105. D. McGonagle, W. Gibbon, P. O’Connor, C. Pease, and P. Emery, Characteristic magnetic resonance imaging entheseal changes of knee synovitis in spondyloarthropathy. Arthritis Rheum. 41, 694 – 700 (1998). 106. V. Jevtic, B. Rozman, M. Kos-Golja, and I. Watt, MR imaging in seronegative spondyloarthritis. Radiologe 8, 624 – 631 (1996). 107. J. Braun, M. Bollow, L. Neure, E. Seipelt, F. Seyrekbasan, H. Herbst, U. Eggens, A. Distler, and J. Sieper, Use of immunohistologic and in situ hybridization techniques in the examination of sacroiliac joint biopsy specimens from patients with ankylosing spondylitis. Arthritis Rheum. 4, 499 – 505 (1995). 108. E. Marker-Hermann, H.-G. Fassbender, S. Kessler, H. Thomssen, Meyer zum Buschenfelde K-H, and M. Blessing, Transgenic mice

362

109.

110.

111.

112.

113.

114.

STEVEN R. GOLDRING with an epidermal overexpression of bone morphogenic protein-6 (BMP-6) present with psoriatic skin lesions and osteoarthropathy Arthritis Rheum. 40 (Suppl. 9), S261 (1997) [Abstract]. M. Blessing, P. Schirmacher, and S. Kaiser, Overexpression of bone morphogenic protein-6 (BMP-6) in the epidermis of transgenic mice: Inhibition or stimulation of proliferation depending on the pattern of transgene expression and formation of psoriatic lesions. J. Cell Biol. 135, 227 – 239 (1996). V. L. Szejnfeld, M. C. Monier-Faugere, B. J. Bognar, M. B. Ferraz, and H. H. Malluche, Systemic osteopenia and mineralization defect in patients with ankylosing spondylitis. J. Rheumatol. 24, 683 – 688 (1997). D. G. Spencer, W. M. Park, H. M. Dick, S. N. Papazoglou, and W. W. Buchanan, Radiologic manifestations in 200 patients with ankylosing spondylitis: Correlations with clinical features and HLAB27. J. Rheumatol. 6, 305 – 315 (1979). S. H. Ralston, G. D. Urquhart, M. Brzeski, and R. D. Sturrock, Prevalence of vertebral compression fractures due to osteoporosis in ankylosing spondylitis. Br. Med. J. 300, 563 – 565 (1990). R. Will, A. Bhalla, R. Palmer, F. Ring, and A. Calin, Osteoporosis in early ankylosing spondylitis; a primary pathological event? Lancet 23, 1483 – 1485 (1989). J.-P. Devogelaer, B. Maldague, J. Malghem, and C. N. de Deuxchaisnes, Appendicular and vertebral bone mass in ankylosing spondylitis: A comparison of plain radiographs with single- and dual-

115.

116.

117.

118. 119. 120.

121.

122.

photon absorptiometry and with quantitative computed tomography. Arthritis Rheum. 35, 1062 – 1067 (1992). J. G. Paredes, M. A. Lazaro, G. Citera, S. Da Representacao, and J. A. Maldonado Cocco, Jaccoud’s arthropathy of the hands an overlap syndrome. Clin. Rheumatol. 16, 65 – 69 (1997). R. M. van Vugt, R. H. Derksen, L. Kater, and J. W. Bijlsma, Deforming arthropathy or lupus and rhupus hands in systemic lupus erythematosus. Ann. Rheum. Dis. 57, 540 – 544 (1998). E. G. L. Bywaters, Relation between heart and joint disease including “rheumatic heart disease” and chronic post-rheumatic arthritis (type Jaccoud). Br. Heart Jr. 12, 101 – 107 (1950). T. Ignaczak, L. R. espinoza, O. S. Kantor, and C. K. Osterland, Jaccoud arthritis. Arch. Intern. Med. 135, 577 – 579 (1975). S. P. Pastershank, and D. Resnick, “Hook” erosions in Jaccoud’ arthropathy. J. Can. Assoc. Radiol. 31, 174 – 175 (1980). F. Formiga, I. Moga, J. M. Nolla, M. Pac, F. Mitjavila, and D. RoigEscofet, Loss of bone mineral density in premenopausal women with systemic lupus erythematosus. Ann. Rheum. Dis. 54, 274 – 276 (1995). A. A. Kalla, a. B. Fataar, S. J. Jessop, and L. Bewerunge, Loss of trabecular bone mineral density in systemic lupus erythematosus. Arthritis Rheum. 36, 1726 – 1734 (1993). Y. Kipen, R. Buchbinder, A. Forbes, B. Strauss, G. Littlejohn, and E. Morand, Prevalence of reduced bone mineral density in systemic lupus erythematosus and the role of steroids. J. Rheumatol. 24, 1922 – 1929 (1997).

CHAPTER 55

Oral Bone Loss and Systemic Osteopenia MARJORIE K. JEFFCOAT, MICHAEL S. REDDY, AND ARTHUR A. DECARLO Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294

I. II. III. IV. V.

VI.

Introduction Diagnosis Risk Factors for Intraoral Bone Loss Patterns of Progression of Oral Bone Loss Residual Ridge Resorption

VII. VIII.

I. INTRODUCTION

Periodontal diseases are initiated by a bacterial infection but are modified by host response factors, including factors that modulate bone remodeling. These host factors are discussed below. Periodontal diseases may be separated into two broad categories. Gingivitis is an inflammation of the gingival tissues without loss of supporting soft tissue or bone, while periodontitis is characterized by loss of bone and soft tissue attachment (Fig. 2). Periodontitis has been further subdivided into specific diseases based primarily on the clinical syndromes. Table 1 summarizes the forms of periodontitis. These classifications of periodontal diseases are continually evolving in light of our increasing knowledge of the pathogenesis of the disease process [2]. Many systemic diseases compromise the host’s ability to combat infection with resultant loss of oral bone. Diseases such as Papillon–Lefevre syndrome, Down’s syndrome, HIV infection, neutropenias, Chediak– Higashi syndrome and diabetes mellitus are associated with alveolar bone loss. Bone loss in the oral cavity may also occur as a result of caries invading the tooth pulp. Dental caries, which are

Bone loss in the oral cavity occurs due to many causes which may be grouped into diseases of primarily bacterial etiology, diseases of multifactorial etiology, and oral bone loss associated with systemic disease. Two diseases that are characterized by oral bone loss associated with bacterial plaque colonization are periodontitis and bone loss secondary to pulpal infection and necrosis. Periodontitis results in bone resorption and soft tissue destruction affecting the supporting tissues surrounding the roots of the teeth (Fig. 1). If periodontitis is left untreated bone is resorbed along the root surface, which can lead to abscesses and tooth loss. The most recent national survey of Oral Health of United States Adults [1] has shown that 94% of women above the age of 65 years have evidence of attachment loss. Thus, this disease constitutes a health problem of increasing concern. The American public pays approximately $1.5 billion every year for periodontal therapy, even though only a fraction of those who need treatment actually receive it.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Etiologic and Host Response Factors in Oral Bone Resorption Oral Bone Loss and Systemic Osteopenia: Are They Related? Treatment

363

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

364

JEFFCOAT, REDDY, AND DECARLO

FIGURE 2

Radiograph of a mandibular first molar exhibiting severe alveolar bone loss (see arrow).

FIGURE 1

Schematic representation of the anatomy of the tooth and alveolar bone. Alveolar bone loss due to periodontitis frequently results in a periodontal pocket between the attached gingiva and the tooth root. Note the presence of bacterial plaque on the tooth root within the pocket.

frequently the result of a Streptococcus mutans infection, may advance into the dental pulp with eventual pulpal infection and necrosis. This pathology frequently results in periapical bone loss that is visible radiographically and may compromise the buccal or lingual alveolar bone as well as the trabecular bone. Bone loss associated with endodontic infection is over 90% reversible with treatment of the carious lesion and placement of root canal fillings and will not be covered in detail here.

TABLE 1 Age of onset

Residual ridge resorption (RRR) refers to the resorption of the alveolar bone which may occur after tooth extraction (Fig. 3). The rate of resorption is highly variable and in some patients does not stop with the residual alveolar bone at the level of the tooth root apices [3,4]. The consequence of this pathologic resorption is an inability to stabilize a denture, resulting in loose prostheses. Severe cases may progress to exposure of the mandibular nerve, with pain and/or inability to wear a denture at all. Residual ridge resorption appears to be multifactorial in etiology. This chapter will focus on two major oral diseases, periodontitis and residual ridge resorption, and will review our current understanding of the etiology, prevalence, and treatment of these diseases.

Periodontitis: Clinical Classifications Bacteria

Host defense

Clinical appearance

Prepubertal

After eruption of teeth and before puberty

Pathogenic plaque

Abnormal

Localized and generalized forms

Juvenile

Circumpubertal

Preponderance of Actinobacillus Actinomycetemcomitans

Abnormal leukocyte chemotaxis and bacteriocidal activity

Rapidly progressive

20 Years or later

Bacteroides and other pathogenic plaque bacteria

Some abnormalities demonstrated

Adult

Usually over 35 years

Pathogenic plaque bacteria

Not demonstrated

Refractory

Any

Localized and generalized forms, less gingival inflammation than expected on severity, familial Rapid loss of bone familial distribution not prominent Loss of attachment and bone, usually gingival inflammation Heterogeneous group of patients unresponsive to any treatment rendered

365

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

FIGURE 3

(A) Panoramic radiograph of an edentulous mandible with relatively normal residual ridge height. (B) Panoramic radiograph of a mandible with severe residual ridge resorption.

II. DIAGNOSIS The diagnosis of loss of bone and soft tissue attachment due to periodontitis is accomplished with radiography and periodontal probing. The latter technique uses an instrument approximately 0.5 mm in diameter to measure the location where resistance is met by the periodontal probe at the base of the pocket. The periodontal pocket is the space between the tooth and the unattached gingival tissue. In Fig. 1, the bacterial plaque is shown in the periodontal pocket. The term “probing pocket depth” refers to the distance from the gingival margin to pocket base or soft tissue attachment, while the term “probing” or “clinical attachment level” refers to the distance from the cemento – enamel junction of the tooth to the pocket base. On a population basis, increased probing measurements correlate with degree of bone loss. In clinical practice, interpretation of pairs of radiographs is the most common method for identifying sites of existing

bone loss and progression of alveolar bone loss over time. Measurements taken from the radiographs, using either a grid or a Schei Ruler, which expresses bone loss as a percentage of root length [5 – 7], are simple and readily available, but not widely used in clinical practice. The ability of the clinician to detect small osseous changes over time is limited by variations in geometry, contrast and brightness of films taken at different examinations, and the superimposition of unchanging structures, such as tooth roots over areas losing bone. Digital subtraction radiography is an image processing technique that facilitates visualization of osseous changes too small to be seen by eye [8 – 16]. Two standardized radiographs are aligned, corrected for variations in contrast, brightness, and planar geometry, and components of the films which are identical are canceled, leaving areas of change readily visible against a neutral background. To enhance visualization, the region of osseous change may be colored [17,18]

366

JEFFCOAT, REDDY, AND DECARLO

367

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

FIGURE 5

Cross-sectional slices from computed tomography of the posterior mandible. Note the knife-edged ridge. The mental foramen is readily visualized in slice 38.

and superimposed on the original radiograph. Subtraction radiography has been shown to be over 90% accurate, sensitive, and specific in detecting small osseous lesions [10,19]. More recent quantitative methods exploit the gray level information in the subtraction image to calculate the mass and density of bony change (Fig. 4). Chips of known mass were used in validity studies of the subtraction radiography method. Calculated bone mass or density estimates correlated with actual chip mass with r2 greater than 92% [16,19]. To date, the use of this methodology is limited to clinical trials. Other aspects of image processing may prove valuable in the detection of osteopenia from oral radiographs. Southard and Southard [20] reported significant differences in image features derived from radiographs of alveolar bone when 70-year-old women were compared to 20-yearFIGURE 4

Example of the use of digital subtraction radiography to enhance visualization and measure changes in alveolar bone. The stippled area shows the region of bone loss detected by computer; changes in bone height, mass, and density are estimated. Images are for visualization only. Measurements must be made by software.

old women. Ruttimann [21] has recently reported that the fractal dimension of peridental alveolar bone may be indicative of generalized osteoporosis. Residual ridge resorption is diagnosed using clinical inspection, palpation, and radiographs. Both clinical inspection and palpation are limited in accuracy because the thickness of the gingiva overlying the residual ridge is indeterminate. Panoramic radiographs provide a two-dimensional tomogram through the arc of the mandible or maxilla (Fig. 3). This radiographic view is frequently used in clinical practice, but the information on the size of the residual ridge is limited to bone height; no information concerning ridge width is available. In order to assess the width of the residual ridge, tomographic methods are used [22 – 28] (Fig. 5). Computed tomography offers the advantages of providing simultaneous assessment of the ridge throughout the entire jaw and the ability to display the information in many forms including cross-sectional slices, frontal and lateral views, and three-dimensional views (Fig. 6). Furthermore, the computed tomography algorithms remove structures not in the plane of interest, resulting in an image with less blur and facilitating

368

JEFFCOAT, REDDY, AND DECARLO

FIGURE 6

Three-dimensional CT of an edentulous mandible. Note the knife-edged occlusal surface indicative of early residual ridge resorption.

interpretation by the clinician. This flexibility comes at a cost in terms of both radiation exposure to the patient and monetary cost of the service. When assessment of residual ridge anatomy can be limited to specific regions of interest within the jaws, motionbased tomography is frequently used to obtain cross-sectional slices. These tomograms can be made using in-office equipment at a relatively low cost and radiation burden. Current research is focused on developing algorithms to reduce the blur inherent to motion-based tomographic images.

III. RISK FACTORS FOR INTRAORAL BONE LOSS Risk factors in periodontal disease or residual ridge resorption have been determined primarily through epidemiologic or observational research (Table 2). Such research aids in the identification of high risk patients for progressive bone loss, thereby allowing clinicians to prescribe appropriate preventive care or treatment.

TABLE 2

A. General Studies indicate that one of the most important risk factors for progressive alveolar bone resorption in periodontitis is the presence of previous destruction. Grbic and Lamster [29] studied 65 subjects previously diagnosed with adult periodontitis in multiple tooth sites. Logistic modeling revealed that the primary risk factors for progressive loss of attachment around the teeth were age and a history of attachment loss. Subjects with severe bone loss ( 5 mm of mean bone loss) had 20.7 times the relative risk of progressive periodontitis over that of subjects with a mild initial loss of bone support. The age of the patient is another significant risk factor in the progression of periodontitis. Although the prevalence of bone loss increases with age in the population (Fig. 7), it should also be noted that not all subjects develop oral bone loss as they age and therefore periodontitis is not part of the aging process, but is a disease entity. A study of the U.S. employed population [30] showed that approximately 50% of individuals 18 to 19 years of age had one or more tooth sites

Risk Factors in Oral Bone Loss

Infection with bacterial plaque Other bacterial infections History of alveolar bone loss Cigarette smoking Age Endocrine dysfunction Diabetes Mellitus Corticosteroids Immunosuppression Neutropenias

FIGURE 7

Prevalence of attachment loss with age. Note that the percentage of employed adults with attachment loss increases with age. Reprinted with permission from the Survey of Oral Health of United States Adults 1987.

369

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

with attachment loss. This confirmed previous studies indicating that the adult form of periodontal destruction begins before the age of 20 [30 – 32]. In the national survey of employed adults the prevalence of bone loss at one or more sites by the age of 65 was over 90%. The natural history of periodontal progression in subjects, 15 to 45 years of age, who received no care and practiced little oral hygiene was studied in Sri Lanka [33]. In this longterm longitudinal study, subjects demonstrated on average an increase in attachment loss with increasing age at a rate of 0.1 – 0.3 mm per year. The subjects could be divided into three distinct groups exhibiting rapid progression, moderate progression, and no progression of attachment loss. These subjects had no adequate method for controlling bacterial plaque and the rate of progression or oral bone loss in developed countries would be expected to occur at a slower rate.

B. Cigarette Smoking Cigarette smoking as a risk factor for osteoporosis is discussed in Chapter 31 and is also a risk factor for periodontitis. Epidemiological studies have linked smoking with the resorption of alveolar bone between the teeth [34 – 37]. When subjects with excellent oral hygiene were examined, smokers exhibited significantly more bone loss around their teeth than matched control subjects. Although the epidemiological data indicate a clear effect of smoking on oral bone loss the pathogenesis is not fully understood. The potential mechanisms of destruction with smoking may involve immunosuppression, impaired soft tissue cell function, and impaired bone cell function [38,39]. The skeletal effects of cigarette smoking may be more pronounced in the oral cavity than elsewhere due to both systemic and local effects of smoking. The pattern of oral bone loss in smokers tends to differ from that in nonsmokers. Smokers exhibit more severe bone loss in the anterior region, while periodontitis patients generally show more severe loss in the posterior regions of the jaw where plaque control is more difficult to achieve.

C. Genetic Disorders Many of the early-onset forms of periodontitis that attack children and adolescents show a substantial familial predisposition. The prepubertal form of periodontitis is responsible for a devastating rate of bone loss in children. Generally this rare form of periodontitis is difficult to manage and leads to loss of teeth before adolescence. Prepubertal periodontitis has been associated with profound defects in monocyte and neutrophil adherence resulting from an inherited glycoprotein deficiency [40]. Precocious intraoral bone loss has also been associated with other hereditary diseases that involve phagocytic cell deficiencies, connec-

FIGURE 8

Pedigree of a family with juvenile periodontitis. Both parents and all children are affected.

tive tissue disorders, and enzyme defects [41] such as neutropenias, the Chediak – Higashi, Elhers – Danlos, and Papillon – Lefevre syndromes, trisomy 21, and hypophosphatasia. The association of severe bone loss with these various chromosomal disorders demonstrates the importance of the host genetic profile in oral bone loss. The onset of juvenile periodontitis occurs around puberty. Evidence from family studies seems to indicate that the prevalence is higher in females than males. An X-linked dominant gene model with reduced penetrance has been proposed [42]. The lack of clear father-to-son transmission of juvenile periodontitis has also been cited as evidence of an X-linked mode of inheritance. However, it should be noted that female adolescent patients tend to seek dental care at a much higher frequency than male patients, which may result in ascertainment basis. The early-onset forms of periodontitis tend to cluster within families (Fig. 8). From these data an autosomal dominant model of inheritance for juvenile periodontitis has also been postulated [43]. Some of the most convincing evidence for a genetic risk of oral bone loss in adults comes from twin studies. Michalowicz and coworkers [44 – 46] studied 110 pairs of adult monozygotic and dizygotic twins, including 96 twin pairs reared together and 14 pairs of monozygotic twins reared apart. The intraoral radiographic findings from this twin population demonstrated a significant genetic association in the proportion of alveolar bone loss observed between monozygotic twins. A brief discussion of the pathogenesis of bone loss is described below. Evidence continues to point to cytokines and matrix metaloproteinases as risk factors in oral bone loss. 1. CYTOKINES AS RISK FACTORS Cytokines were first identified in the gingival crevicular fluid in the 1980s. Since that time the recognition of various interleukins and cytokines that have been implicated in the tissue destruction associated with periodontitis and other chronic inflammatory diseases has rapidly expanded. The expression of IL-1 and IL-1 have been implicated in the pathogenesis of periodontitis [47]. Histologically the levels of IL-1 and connective tissue destruction in untreated adult periodontitis have been found to be highly correlated. In addition to IL-1 other pro-inflammatory cytokines such as tumor necrosis factor-, IL-2, IL-6, and IL-8 are believed to have a potential regulatory role in the inflammation

370

JEFFCOAT, REDDY, AND DECARLO

and subsequent tissue destruction in periodontitis [48 – 50]. A risk-indicator relationship between IL-1 and PGE2 and progressive periodontal destruction has been reported [51]. In now classic studies, PGE2 has been shown to be a risk factor for progressive periodontitis [52,55]. 2. METALLOPROTEINASES AS RISK FACTORS Metalloproteinases have been implicated in many pathologic conditions because of their ability to cleave extracellular matrix proteins. The balance of matrix metalloproteinases and the tissue inhibitors of metalloproteinases appears to play a major role in the connective tissue destruction associated with periodontitis, arthritis, atherosclerosis, pulmonary emphysema, and osteoporosis [54 – 57]. Nonsteroidal anti-inflammatory drugs and non-antibiotic tetracyclines have been used as adjuncts to manage excess metalloproteinase activity in arthritis and periodontitis [58]. 3. GENETIC POLYMORPHISMS AS RISK FACTORS The prevalence of severe periodontitis is approximately 15%. The susceptibility of some individuals to severe forms of periodontitis may at least be in part under genetic control. These studies indicated that a significant amount of the oral bone loss and attachment loss observed could be attributed to genetic variation. Recent data suggested that cytokine regulation may be related to specific genotypes [59]. These data indicate that a specific genotype of the polymorphic IL-1 gene cluster may be associated with periodontitis severity in nonsmokers.

D. Endocrine Disorders Alveolar bone loss is also associated with hormonal dysfunction. The most common endocrine disease, diabetes mellitus, has long-term ocular, renal, and vascular complications that are well established. In addition, these patients

FIGURE 9

may exhibit relatively rapid bone loss around teeth due to an aggressive form of periodontitis. The pathophysiology of this rapidly progressive bone loss is not well understood but a number of predisposing factors have been identified. Along with other peripheral vascular problems, changes in the periodontal vasculature may exacerbate bone loss [69]. Impaired neutrophil functions including chemotaxis [61], phagocytosis, intracellular killing, and adherence [62] have been reported. The crevicular fluid adjacent to teeth of diabetics has increased glucose content [63], which may serve as a source of nutrients for bacteria. The crevicular fluid also has an increased collagenase activity [51], possibly facilitating destruction of the alveolar bone matrix. This combination of factors seems to underlie the fulminating periodontitis in children and adults with uncontrolled diabetes which results in widespread destruction of alveolar bone (Fig. 9). Alveolar bone loss in diabetes is not limited to uncontrolled insulin-dependent diabetics. The Pima Indians of the southwestern United States represent a unique population in that approximately 50% of the adults over the age of 35 have non-insulin-dependent diabetes mellitus. In a largescale epidemiological study of 3219 Pima Indians, signifi cant loss of interproximal bone was observed both clinically and radiographically even after adjusting for demographics and amount of bacterial plaque [65]. This bone loss was strongly correlated with diabetic status. The odds ratio for diabetic subjects to exhibit bone loss was 3.43:1. Thus, even non-insulin-dependent diabetics are at risk for oral bone loss that cannot be explained on the basis of age, gender, oral hygiene, or general oral health. Endocrine dysfunctions resulting from long-term highdose corticosteroids have been implicated in oral bone loss associated with periodontitis. In subjects with acute nephrotic syndrome the mineral content of the mandible and the forearm were studied [66]. Steroid use has been hypothesized to be an etiologic factor for intraoral bone loss

Panoramic radiograph showing severe alveolar bone loss in a patient with diabetes.

371

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

because of the immunosuppressive effects of the steroids and the osteoporotic side effects. In these subjects, the degree of osteopenia in the mandible was comparable to the loss in other cortical bone sites examined. An “old wives tale” once indicated that a tooth is lost for each child that is born to a mother. Although this may not be factual it cannot be overlooked that pregnancy is associated with an increase in periodontal inflammation and to some extent with alveolar bone loss. Progesterone may influence the biosynthesis of prostaglandins in supporting tissues of the teeth leading to extreme gingival inflammation and interproximal bone loss [67]. Oral contraceptives may produce similar responses in some women. A two-to threefold higher incidence of osteitis has been reported after third molar extractions in women on oral contraceptives [68]. However, a clear effect on oral bone loss has not been established in women on long-term oral contraceptives. Hormonal changes associated with puberty are also implicated in the initiation of progressive bone loss in localized juvenile periodontitis [69]. This early-onset form of periodontitis is not clinically evident until puberty even though bacterial species highly associated with the disease may be present earlier. The exact interaction of sex hormone variations and oral bone loss is not clearly delineated; however, changes in hormone levels at puberty are considered a risk factor.

E. Immunosuppression Immunocompromised patients are at significant risk for rapid oral bone loss due to either disease or chemotherapy. The bone loss observed in such patients is severe largely because of the host’s inability to immunologically control the bacterial challenge that results in periodontitis. One of the early clinical signs of patients with leukemia may be severe gingival inflammation and destruction of the supporting bone of the dentition. Neutropenias also demonstrate characteristic gingival inflammation and bone loss. In patients with cyclic neutropenia [70] or agranulocytosis [71] bone loss may be so severe as to render the child edentulous before the permanent teeth erupt. Papillon – Lefevre syndrome [72], histiocytosis X [73], and trisomy 21 [74] all produce marked destruction of the alveolar bone as a manifestation of the illness. Rapid destruction of the supporting bone of the teeth often complicates HIV infection [75]. The extensive bone resorption that occurs as a result of intraoral infection that could be easily managed in a nonimmunocompromised patient has become a major quality of life issue for affected individuals. The management of bone loss of HIV-associated periodontitis has become increasingly challenging due to the necessity of avoiding the use of broad spectrum antibiotics. HIV periodontitis is best managed by early detection and rigorous preventive treatment, including strict oral hygiene.

IV. PATTERNS OF PROGRESSION OF ORAL BONE LOSS To determine the rate, pattern, and natural history of periodontal bone loss, researchers have assessed subjects with repeated measurements. Before these studies it was commonly assumed that bacterial plaque accumulation universally lead to gingival inflammation with subsequent unrelenting, albeit slowly progressive, bone resorption. The impression that alveolar bone loss was continuous over time and irreversible was developed by observing crosssectional populations over long periods of time. Papapanou and co-workers [76] studied over 200 subjects with full mouth radiographic surveys taken 10 years apart. Measurements of bone loss around each tooth revealed that the mean annual rate of loss of alveolar bone height varied by age. Subjects between the ages of 25 and 65 exhibited between 0.07 and 0.14 mm, whereas subjects over 70 had significantly higher rates of bone loss (0.28 mm). The data from this long-term study with only two major examination periods helped confirm the impression that bone loss was contiguous and slowly progressive. However, the data indicate that rates of progression varied widely among teeth and subjects. These findings are further supported by a similar 6-year study in elderly Chinese subjects where the individual range of bone loss varied dramatically from 0.00 to 0.53 mm/year [77] and the previously cited study by Löe and co-workers in Sri Lankan tea workers [33] which also reported wide variability in the rate of periodontal destruction. In a classic study, Goodson and co-workers [78] challenged the commonly held belief that oral bone loss proceeds in a gradual fashion (Fig. 10). In a series of studies they examined the individual tooth site for progressive bone resorption [79 – 82]. Twenty-two untreated subjects with existing periodontal pockets due to bone loss were studied for 1 year. In 15 subjects, tooth sites became significantly deeper, while other tooth sites appeared to gain attachment on the tooth. The results of this study indicated that destruction related to periodontal disease was a

FIGURE 10

Proposed models for bone loss in periodontitis. (Left) Continuous model. (Middle) Burst model. (Right) Burst with healing (or remission) model.

372 dynamic condition that exhibited exacerbations and remissions. The observed pattern of progression and regression has since been known as the “burst model” for periodontal disease progression. These classic studies utilized conventional clinical attachment level probing to detect sites exhibiting more than 2 mm of progressive attachment loss. Only 5% of tooth sites exhibited progressive attachment loss. A more recent study [83] which utilized a more sensitive electronic probe to measure attachment loss revealed that 29% of the tooth sites studied, in the population of adults previously diagnosed with periodontitis, showed progression over a 6-month period. Modeling of the data for the sites that lost bone over time showed that 76% of tooth sites lost attachment consistent with linear downhill patterns, 12% of tooth sites showed exacerbations and remissions, and 12% showed evidence of bursts of disease activity. Thus, the natural history and progression pattern of intraoral bone loss are not completely understood at this time. Current studies are turning increasingly to innovative sampling strategies and modeling techniques.

V. RESIDUAL RIDGE RESORPTION Residual ridge resorption occurs following tooth extraction. In the most severe cases, the denture may impinge on the exposed mandibular nerve, resulting in pain or total inability to tolerate the prostheses. There is no universally accepted definition of the amount of bone loss that must occur after extraction in order to confirm a diagnosis of residual ridge resorption. This lack of a working definition has likely hampered studies of the prevalence of the disease and clear statistics on its prevalence in the population are not available. Most studies aimed at defining etiologic factors and rate of residual ridge resorption have utilized small groups of patients [4]. These studies have shown that residual ridge resorption is chronic, progressive, and cumulative. Its rate is highly variable from patient to patient. Some patients may lose as much as 4.5 mm of ridge height per year in the anterior mandible in the first 2 years after extraction, while others lose as little as 0.75 mm per year [84]. It has been hypothesized that residual ridge resorption is a multifactorial disease with anatomic, metabolic, and biomechanical predisposing factors. It has been tempting to assume that biomechanical factors such as denture fit and occlusion are a major cause. Unfortunately, the few clinical trials dedicated to testing the hypothesis do not bear this out, and an effect of load bearing area on residual ridge resorption has not been conclusively demonstrated [85,86]. No significant difference in ridge height could be demonstrated between patients who wore their denture continuously and those who wore dentures only during the day [87]. Also, residual ridge resorption may occur in the ab-

JEFFCOAT, REDDY, AND DECARLO

sence of dentures [88]. In addition, Brehm [89] found no significant effect of denture occlusion on ridge height. Furthermore, the rate of residual ridge resorption slows over time, with the maximum rate occurring in the first 2 years after extraction [84,90]. It would appear that while studies do not definitively rule out a role for denture fit, the rate of residual ridge resorption must be related to more than anatomic factors. Diet, especially calcium intake, has long been thought to play a role in predisposing to residual ridge resorption. However, few clinical trials address this important issue. In a short-term study, Wical and co-workers [91] found that calcium and vitamin D supplementation reduced the rate of residual ridge resorption by 36% compared to patients receiving a placebo in the first year after extraction. Additional controlled studies are needed to elucidate further the etiology of residual ridge resorption. Oral bone loss is no longer considered a natural and universal consequence of aging. Clearly, bone loss in the oral cavity is multifactorial and may not occur at all. However, a number of researchers have demonstrated that many subjects undergo bone resorption at different rates and this has led to the study of various risk factors for intraoral bone loss. With an understanding of what puts the patient at risk for bone destruction, management of intraoral bone loss as a disease process can be greatly improved.

VI. ETIOLOGIC AND HOST RESPONSE FACTORS IN ORAL BONE RESORPTION Vertebrate bone is a dynamic mineralized connective tissue that remodels constantly throughout life, and the trabecular and cortical bone surrounding the teeth (alveolar bone) behave in a similar fashion. Bone remodeling occurs as two distinct and coupled processes: degradation or loss of bone (resorption) followed by formation (apposition) which occur on bone surfaces as discrete in foci bone remodeling units [92 – 96] (see Chapter 1). If the overall rate of bone resorption exceeds the rate of apposition, net loss of bone occurs over time. This concept of balance between the two processes of bone turnover applies to a very localized site around a single tooth as well as to the entire mandible or skeleton. Bone remodeling is comprehensively discussed in Chapters 1 and 39. While the osteoblast is considered to be responsible for bone formation, current evidence suggests that osteoblasts also have a role in bone resorption along with multinucleated osteoclasts [44,97 – 103]. Recent investigations suggest that bone resorption initiates with the induction of the osteoblast to remove a thin, nonmineralized surface collagenous osteoid layer, thereby prompting osteoclastic activity [97,104]. A

373

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

signal, possibly from these osteoblasts, induces preosteoclastic cells to migrate and differentiate. The bone lining cells are then replaced by preosteoclasts that subsequently fuse to form the multinucleated osteoclasts which then dissolve the mineral and degrade the bone matrix. Following this resorption phase, osteoblasts line the resorption unit and synthesize new bone matrix to fill the space [93,103]. Degradation of bone ultimately requires the degradation of type I collagen by active collagenase or by another active proteinase. In periodontal disease such proteinases may be released by bacteria or by the host cells and diffuse through the tissues to the bone surface, or they may be delivered directly to the remodeling surface by host cells (osteoblasts and/or osteoclasts) that reside on the bone. It has not been determined which mechanisms function in vivo so this review will consider the major sources of bone-degrading proteinases. We will also consider the etiology and control of bone remodeling as they relate to periodontal disease and osteoporosis. Periodontal diseases are characterized by inflammation around teeth and by loss of their ligamentous and bony support. In periodontitis, bone height is lost from the alveolar crest along the root surface of the tooth. This pattern of bone loss results in loss of both bone height and crestal and subcrestal bone density (Fig. 1). This clinical condition may be distinguished from loss of alveolar bone density which can occur independent of loss of alveolar bone height along the root surface. In most of the periodontal diseases, bacteria and their by-products are considered the primary etiology for the loss of osseous support around the teeth. The number of bacterial species that can infect the periodontal pocket has been estimated to be as high as 300 [105 – 107]. Loss of crestal bone height in periodontal disease is associated with (i) an increase in total bacterial load around the teeth and (ii) an increase in the numbers and proportion of gram-negative species [7,108 – 111]. Though no one bacterial species is thought to be causative for periodontal disease in adults or children, a small number of gram-negative anaerobic bacteria have been strongly associated with alveolar crestal bone loss (Table 3) [105,111 – 113]. Each of these suspected periodontopathogens elaborates powerful proteolytic enzymes, volatile and noxious shortchain organic acids, and lipopolysaccharide on their surface and into the surrounding milieu. Whole bacteria are rarely found beneath the epithelial lining of the pocket [114]. When considering the virulence of these suspected periodontopathogens and the pathogenesis of periodontal disease we must consider:(i) direct degradation of the surrounding connective tissues and bone by the release of bacterial collagenases and other bacterial proteinases and (ii) indirect stimulation of host degradative processes — a reaction to bacterial antigens and toxins — by immune cells, stromal cells, or the epithelial cells that line the periodontal pocket.

TABLE 3

Partial Listing of Putative Periodontopathic Bacteria: Adult Periodontitis P. gingivallis A. actinomycetemcomitans P. intermedius F. nucleatum F. forsythus C. rectus E. corrodens Selenomonas sp. Eubacterium sp. Spirochetes

A. Direct Bacterial Virulence in Periodontal Disease The bacteria associated with periodontal disease activity release potent proteinases capable of dissolving bone matrix [105,106,115 – 118]. Diffusion of various substances from the gingival crevice into the tissues has been demonstrated [119 – 121] and bacterial antigens have been localized to underlying tissues in histologic analysis of diseased periodontal tissues [123]. Up to 6 mg bacteria (wet wt) may be found around a periodontally infected tooth which may release significant proteolytic activity. These bacterial proteinases, which serve the nutritional needs of the microorganism, could theoretically diffuse beyond the epithelial lining of the periodontal pocket and into the underlying stromal connective tissues to the alveolar bone [121,122] (Fig. 11). Direct dissolution of the alveolar bone by bacterial collagenases or other bacterial proteinases, however, would require the unlikely avoidance of the ubiquitous serum and interstitial proteinase inhibitors such as 1-antitrypsin and 2-macroglobulin, which exist at concentrations of approximately 30 and 3 mM, respectively.

B. Indirect Bacterial Virulence in Periodontal Disease 1. INDIRECT EFFECTS OF BACTERIAL PROTEINASES Proteinases released by the suspected periodontopathogens are also capable of inducing a matrix-degrading phenotype in fibroblasts and epithelial cells by upregulation of collagenase and other matrix-metalloproteinases (MMPs) [124 – 126]. In vitro, these relatively quiescent cells acquire the ability to dissolve type I collagen substrates when treated with bacterial proteinases [124,127]. Stimulation of collagen degradation in these cells stems from two mechanisms: (i) the ability of the proteinases to upregulate the mRNA and secretory protein levels of the MMPs and (ii) the

374

JEFFCOAT, REDDY, AND DECARLO

FIGURE 11

Schematic representation of cellular events leading to bone loss.

ability of the bacterial proteinases to activate the MMPs which are all secreted as latent enzymes [111]. In histologic sections of periodontal tissues, collagenase-positive cells have been identified in association with inflammation [128, 129]. Bacterial proteinases could also inactivate serum proteinase inhibitors, hydrolyze immunoglobulins [130 – 132], and activate other mediators of inflammation such as complement [133] and the kininogens [134,135].

bone resorption and inflammation [138,140 – 144]. LPS from periodontopathogens has also been shown to induce collagenase secretion from macrophages [145,146] and from osteoblasts [147] but does not directly induce collagenase secretion from fibroblasts or keratinocytes. LPS does, however, appear to stimulate the release of collagenase-inducing interleukins from fibroblasts [148].

2. INDIRECT EFFECTS OF NOXIOUS BACTERIAL BY-PRODUCTS

THE IMMUNE

The predominantly gram-negative flora generate volatile short-chain fatty acids during metabolism such as butyric and proprionic acid. Many of the periodontopathogens also release the odorous and cytotoxic molecule hydrogen sulfide [136]. Cell cytotoxicity and increased epithelial permeability caused by these bacterial metabolites is suspected to contribute to the inflammatory response and the tissue destruction of periodontal disease [137,138]. Lipopolysaccharide (LPS) of gram-negative bacteria associated with human periodontal disease is a powerful stimulant of bone resorption in vitro [138,139]. Pathogen-associated LPS molecules can stimulate cells of the macrophage/monocyte lineage, endothelial cells, and epithelial cells to release cytokines and other mediators of

3. EFFECTS OF ANTIGENIC CHALLENGE OF SYSTEM

Bacteria infecting the periodontal pocket induce an antibody response which can be measured in the serum as well as the site of challenge [149 – 156]. These antibodies associated with chronic periodontitis are predominantly of the IgG and IgA isotypes and are specific for the infecting periodontopathogens. Gingival tissue around chronically infected teeth accumulate much higher numbers of antigenspecific B-cells, plasma cells, and T-cells in the surrounding tissues [150,157 – 159]. Bacterial antigens in periodontal disease include fimbrae, LPS, and other cell-wall components and can be mitogenic, eliciting a polyclonal response in addition to the antigen-specific response. Both the classical and alternate complement cascades are initiated by these antigens; antigen/antibody complexes and the complement fragments C3a and C5a which are chemotactic and

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

vasoactive can be found in the tissues and crevicular fluid of infected sites in the mouth [160,161]. Polymorphonuclear lymphocytes (PMNs) migrate quickly to regions of infection and remain at higher levels than at healthy sites throughout the period of infection due to continual antigenic challenge and possibly to the presence of interleukin9 (IL-9) [162]. PMNs may then directly contribute to interstitial and bone matrix degradation through stimulated degranulation and release of lysozomal collagenase and cathepsins. Macrophages gradually populate the already inflamed periodontal tissues and are thought to contribute significantly to the inflammatory response. Stimulated PMNs and macrophages may release IL-1 [163,164], products of arachadonic acid metabolism (prostaglandin E2, PGE2) [165], and reactive oxygen intermediates which have been associated with tissue damage [166,167]. The cathepsins released by PMNs will also activate collagenase and most other MMPs [168,169]. Antigenic challenge in periodontal disease creates not only a humoral but a cell-mediated response as well [170 – 173]. Both types involve a complex series of antigen presentation, cell-to-cell contacts, and cytokine-mediated communication among the macrophage/monocyte cells, Tcells, and B-cells (Fig. 11). During inflammation, whether in the periodontium, near accumulated bacterial plaque, or in the synovium of an arthritic joint, the levels of certain “pro-inflammatory” cytokines (IL-1, tumor necrosis factor [TNF]- and -, and IL-6, IL-8, and granulocyte – monocyte colony stimulating factor [GM-CSF]) have been shown to be increased compared to levels in healthy sites [47,155,174 – 176]. Each of these cytokines can affect bone remodeling and have been shown to initiate a strong local inflammatory response with osseous resorption accompanied by hypercalcemia [177]. These proinflammatory cytokines stimulate osteoclast formation and osteoclast activity in vitro and in vivo [98,177 – 179] while stimulating the expression of matrix-degrading collagenase from osteoblasts [180 – 182] and from neighboring epithelial cells and fibroblasts [183,184]. They can also downregulate bone matrix protein synthesis and inhibit bone formation in vitro as well as in vivo [185 – 188]. IL-1 stimulation of bone resorption is thought to occur through the generation of active phospholipase A2 leading to arachadonic acid metabolites including PGE2 and in stimulating the protein kinase C pathway, leading to MMP production by multiple cells including osteoblasts [181, 182]. TNF-, which has been demonstrated to be induced in cells from diseased gingival tissues [140,171], can upregulate IL-1 synthesis [189]. The reason that bone resorption appears to be very sensitive to both IL-1 and TNF- [185,190 – 192] may be explained by the need for IL-1-induced osteoblastic MMP secretion early in the degradation process discussed above. In inflamed periodontal tissues, IL-1 and - are present at concentrations of 109 –108M

375 [140,165,47,193], which would be sufficient to stimulate MMP expression in cultured fibroblasts [168,169]. The T-cells, which modulate the cellular and humoral responses are predominantly antigen-specific in chronically inflamed periodontal tissues and are increased in number with a relative increase in the suppressor CD8-type T-cell over the helper CD-4 type T-cell [194,195]. The CD8 suppressor cell population in periodontally diseased tissues has been shown to suppress specific antibody production via a macrophage-mediated pathway [195]. The majority of the T-cells in inflamed periodontal tissues have the Th1 cytokine profile which is IL-2, gamma-interferon (INF-), and TNF positive and IL-5 negative (Table 4). Only a small percentage of T-cells in periodontal patients are of the Th2 subtype, or IL-5, IL-4, IL-6, and IL-10 positive [196]. Recent evidence indicates that the Th2-derived IL-4 and IL-10 may suppress many of the actions associated with the Th1 lymphocytes [197 – 201]. Also, Th2-derived IL-4 induces apoptosis, potentially reducing the number of infiltrating macrophages [202,203], and upregulates the production of the IL-1 antagonist, inhibiting inflammatory pathways [200, 204]. In chronic inflammatory lesions including periodontal disease, evidence suggests that an initial lack of IL-4 early in the infection may contribute to this increase in Th1 lymphocytes and decrease of Th2 lymphocytes and to the inability of the inflammatory response to resolve. IL-4 is, indeed, deficient in chronic periodontal lesions and rheumatoid synovium [155,196,205,206]. Adoptive transfer of Th2-type T-cells in rodents increased antibody levels and decreased periodontal disease activity [207]. It is, therefore, possible that Th2-type lymphocytes could be protective in periodontal disease by helping antibody production while the Th1- and CD8-type cells may be destructive via prominent production of INF- and the potential for stimulation of macrophage IL-1 secretion with subsequent bonedestructive activity [194,207 – 209]. As a result of the inflammatory response in periodontal disease, prostaglandins are released from cell membranes and can be found in the surrounding tissues and fluids [163,210 – 212]. PGE2 is a membrane metabolite that can be released from most viable cells in small quantities. Levels are 3-fold higher than normal in gingival crevicular fluids from chronically inflamed gingiva and 18-fold higher in regions of the periodontium that are considered to be undergoing an overall net bone resorption [212]. PGE2 levels correlate well with the extent of bone loss around teeth [213]. Hyperresponsive secretion of PGE2 by monocytes and macrophages is considered characteristic in periodontitis patients and rheumatoid arthritis patients. PGE2 in vitro stimulates osteoclastic activity at high levels and at low levels may synergistically stimulate the osteoclast-inducing effects of IL-1 and other proinflammatory cytokines [209,214]. Also, the effects of IL-1 on bone degradation are thought to be mediated through an

376

JEFFCOAT, REDDY, AND DECARLO

TABLE 4 Cytokines Implicated in Periodontal Destruction Cytokine IL-1

Description Induces PGE2 from fibroblasts PGE2-dependent MMP induction in osteoblasts and fibroblasts Elevated in inflamed periodontal tissues IL-1 induces IL-6 Increases osteoclast activity Synergistic with bone resorbing effects of PTH

TNF-

Upregulates IL-1 and IL-6 synthesis; Elevated in inflamed periodontal tissues Elevated in rheumatoid arthritis synovium and sera

-INF

Activates macrophages to secrete IL-1 and increases B-cell antibody

alveolar bone destruction in periodontal disease [216,217, 218]. In summary, the bone loss in periodontal disease appears to occur as part of an inflammatory response to bacterial accumulation on teeth. The nonkeratinized nature of the epithelium that lines the periodontal pocket and the proximity of alveolar bone to the site of infection allow both cellular and humoral immune responses to develop. Inability of the lesion to heal is due to the continued accumulation or reinfection of bacteria, and possibly to the development of a chronic inflammatory response dominated by Th1/CD8 lymphocyte subtypes. Bone loss in the periodontium occurs from the osseous surface near the site of inflammation, presumably due to a gradient of proinflammatory and bone resorbing cytokines emanating from nearby leukocytes, stromal cells, and epithelial cells and in response to the bacteria within the periodontal pocket.

Elevated in inflamed periodontal tissues IL-4

Deficient in inflamed periodontal disease tissues Suppresses Th1 cytokine production Induces macrophage apoptosis Upregulates IL-1 antagonist

IL-5 IL-6

Increases division of activated B-cells Converts activated B-cells to plasma-type cells Plays a role in osteoclast development and recruitment Elevated in rheumatoid arthritis synovium and sera May mediate bone-resorbing activity of IL-1 High levels found in inflamed periodontal tissues

PGE2

Elevated in inflamed periodontal tissues and tissue levels correlate well with periodontal bone loss Stimulates osteoclast activity and synergistically stimulates the osteoclast-inducing effects of IL-1 and other pro-inflammatory cytokines Possibly mediates IL-1 effects Not involved in bone formation

IL-2

VII. ORAL BONE LOSS AND SYSTEMIC OSTEOPENIA: ARE THEY RELATED?

Could contribute to B-cell proliferation and differentiation but is conspicuously absent from inflamed periodontal tissues and other areas of chronic inflammation

IL-8

Powerful chemoattractant and activator of PMNs

GM-CSF

Inhibits collagen synthesis in osteoblasts Inhibits differentiation of pre-osteoclasts to osteoclasts Increases osteoclastic activity

arachadonic acid- and PGE2-dependent pathway [181, 182]. Further, extracellular components of some suspected periodontopathogens appear to inhibit osteoblast activity through a PG-mediated pathway [215]. Bone formation may be prostaglandin independent [208]. The recent success of prostaglandin inhibitors to reduce the rate of alveolar bone loss in adults with periodontitis compared to rates of bone loss in placebo-treated controls implicates prostaglandin-mediated pathways in the pathogenesis of

It has long been hypothesized that oral bone loss may be related to systemic conditions that predispose to osteoporosis. In fact, several of the same risk factors are present [219 – 221]. Older patients who smoke are at higher risk for both osteoporosis and oral bone loss. In a now classic study by Kribbs et al. [222], a strong correlation between dental bone mass and total bone mass was observed in women. In another study, mandibular bone mass was more closely related to skeletal bone mass than to age [223]. There is also evidence that loss of teeth may be related to overall bone mass. Severely osteoporotic women are three times as likely as controls to be edentulous [224]. Krall and co-workers [225] recently reported a significant association between the bone mineral density of the spine and radius and the number of remaining teeth, controlling for smoking, years postmenopause, education, and body mass index. Clearly, once the teeth are lost, the patients are at further risk for residual ridge resorption. It is, therefore, tempting to hypothesize a relationship between generalized osteopenia and residual ridge resorption. Existing data on the subject are limited. Habets [226] demonstrated osteopenia in iliac crest biopsies from 74 patients with severe mandibular residual ridge resorption. Studies have shown that women comprise a larger percentage of patients with residual ridge resorption than men [227] and it has been reported that the majority of patients referred for specialist prosthodontic treatment are women [228]. On the other hand, there is also evidence that bone mineral density is not related to progressive alveolar bone

377

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia

loss. Elders et al. [229] reported that lumbar bone mineral density measurements were not significantly different in edentulous and dentulous women, nor was a significant relationship observed between systemic bone mass and alveolar bone height. These authors conclude that systemic bone mass is not an important factor in the pathogenesis of periodontitis. Klemetti and co-workers [230] reported that the rate of crestal alveolar bone loss does not correlate strongly with existing trabecular or cortical density. Weyant et al. Reported that osteopenia did not correlate with soft tissue attachment loss around the teeth [231]. It has recently been reported that there is a correlation between oral bone mineral density and systemic bone mineral density [232]. In an Ancillary study to Women’s Health, a cross-sectional analysis was made of data obtained from the first 158 subjects. One goal was to determine whether or not image analysis of intraoral radiographs could be used to demonstrate a relationship of basal bone mineral density of the mandible to that of the hip, determined by dual-energy X-ray absorptiometry (DXA). Comprehensive medical histories and examinations were linked with results of oral examinations and quantitative digital intraoral radiography. General linear models of basal bone mineral density (of the mandible), hip bone mineral density, mid-root density, age, race, hormone replacement therapy, and calcium supplements were created. Figure 12 plots the hip bone mineral density as measured by DXA versus the basal bone mineral density measured from the intraoral radiographs. Significant correlations were found between mandibular basal bone mineral density and hip bone mineral density (r  0.74, P.001).

VIII. TREATMENT A. Periodontitis As the discussion on pathophysiologic mechanisms in periodontitis implies, treatment of periodontitis could involve either control of the pathogenic plaque bacteria or the destructive host response. Current periodontal treatment is focused primarily on controlling plaque bacteria. Patient’s self-administered plaque control and scaling and root planing performed by the dental professional reduces the mass of bacterial plaque [233]. Root planing smooths the root surface and removes decalcified cementum and bacterial components such as endotoxins. Periodontal surgical procedures provide access to the roots for effective root planing or to the bone for pocket reduction procedures [234]. Regenerative procedures using the principles of guided tissues regeneration or bone grafting materials may be used to aid the restoration of lost periodontal ligament and bone in selected cases. Nonspecific and specific control of plaque bacteria using antiseptics and antibiotics are also available. The mass of plaque bacteria may be controlled with topical antisepetic agents, such as chlorhexidine [233,235]. These nonspecific agents are effective for the reduction of plaque and gingivitis but their effect on periodontitis and alveolar bone loss has not been determined. Antibiotics provide more specific control of plaque bacteria [235 – 242]. To date antibiotics, especially tetracycline, have been most successful in reducing or eradicating the Actinobacillus actinomycetemcomitans of juvenile periodontitis. Although many systemic antibiotics (including tetracyclines, penicillins, metronidazole, clindamycin, and

FIGURE 12 Relationship between bone mineral density at hip (measured by DXA) and mandibular basal bone density (measured by digital radiography). Note the correlation between hip and basal bone mineral density.

378

JEFFCOAT, REDDY, AND DECARLO

erythromycins) have been tested in adult periodontitis most studies have addressed the effects on soft tissues and the results are mixed presumably due to the different etiologic bacteria involved in adult periodontitis. Antiseptics and antibiotics may be delivered locally to the periodontal pocket in both resorbable and nonresorbable vehicles [243 – 245]. Agents including chlorhexidine, tetracyclines, and metronidazole have significant effects on the reduction of pocket depths, but their effects on alveolar bone loss have not been widely evaluated. Recent studies are beginning to address agents that control the host side of the pathogenesis of periodontitis. As previously described, there is a considerable body of evidence pointing to prostaglandins and pro-inflammatory cytokines such as IL-1 as risk factors in progressive bone loss in periodontitis. [52,55]. Based on these findings, studies in beagles with periodontitis showed that decreases in inflammation could decrease the rate of progression of bone loss. Positive clinical trial results in humans have been reported for a range of nonsteroidal anti-inflammatory drugs including flurbiprofen, naproxen, and meclomen [216]. Metalloproteins have also been implicated in the pathogenesis of periodontitis. Tetracyclines have been shown to inhibit collagenase activity independent of their anti-bacterial effects [246]. Low-dose systemic doxycycline is now available for use as an adjunct to scaling and root planing for the treatment of periodontitis. It would seem logical that the treatment of osteoporosis and the progressive bone loss associated with periodontitis could utilize some of the same pharmacologic agents. This class of drugs, including bisphosphonates, may prove useful in this regard. It is important to distinguish between topical agents such as toothpastes that may include diphosphonates. These dentifrices are intended to control dental calculus (tartar) on the surface of tooth enamel but were not designed to affect the bone. Preliminary studies in dogs and monkeys have indicated that systemically administered bisphosphonates have a bone sparing effect in periodontitis [247,248]. A pilot study in patients with periodontitis has shown that the risk of progressive loss of bone and bone density is halved in patient taking alendronate when compared to placebo treated patients [249].

B. Residual Ridge Resorption The best treatment of residual ridge resorption is prevention by avoiding loss of teeth. Most treatment of residual ridge resorption involves improving patient comfort by increasing denture retention either by relining the denture or increasing ridge height through soft tissue and bone grafts. Placement of endosseous implants also provides improved denture retention.

IX. SUMMARY Oral bone loss is most often the result of a complex interplay between the pathogenic plaque colonizing the periodontal pocket and protective and destructive host responses. Progressive oral bone loss, termed residual ridge resorption, occurs in the absence of teeth. Treatment of bone loss focuses on maintaining teeth by removing the bacterial etiologic factors and enhancing the host response to permit the regeneration of lost bone wherever possible. Future advances in diagnostic and measurement techniques may facilitate a greater understanding of the etiology, prevention, and treatment of oral bone loss.

References 1. Oral Health of United States Adults, “The National Survey of Oral Health in the U.S. of Employed Adults and Seniors: 1985 – 1986. National Institute of Dental Research, 1987. 2. R. R. Ranney, Classification of periodontal disease. Periodontology 2000 2, 13 – 25 (1993). 3. D. A. Atwood, The reduction of residual ridges. A major oral disease entity. J. Prosthet. Dent. 26, 266 – 279 (1971). 4. D. A. Atwood, “The Problem of Reduction of Residual Ridges. Saunders, Philadelphia, PA, 1979. 5. H. Bjorn, A. Halling, and H. Thyberg, Radiographic assessment of marginal bone loss. Odontol. Rev. 20, 165 (1967). 6. S. Lavstedt, A methodological-roentgenologic investigation on alveolar bone loss. Acta Odont. Scand. (Suppl.), 33, 67 (1973). 7. O. Schei, J. Waerhaug, A. Lovdal, and A. Arno, Alveolar bone loss as related to oral hygiene and age. J. Periodontol. 30, 7 – 16 (1959). 8. R. L. Webber, U. E. Ruttiman, and H. G. Grondahl, X-ray image subtraction as a basis for assessment of periodontal changes. J. Periodontol. Res. 17, 509 – 511 (1982). 9. R. L. Webber, Oral imaging as a diagnostic tool for assessing osseous changes. J. Bone Miner. Res. 8, S543 – S548 (1993). 10. E. Hausmann, L. Christersson, R. Dunford, U. Wikesjo, J. Phyo, and R. J. Genco, Usefulness of subtraction radiography in the evaluation of periodontal therapy. J. Periodontol. 56, 4 – 7 (1985). 11. E. Hausmann, A contemporary perspective on techniques for the clinical assessment of alvolar bone. J. Periodontol. 61, 149 – 156 (1990). 12. U. Bragger, L. Pasquali, H. Rylander, D. Carners, and K. S. Kornman, Computer assisted densitometric image analysis in periodontal radiography. A methodological study. J. Clin. Periodontol. 15, 29 – 39 (1988). 13. U. Bragger, W. Burgin, N. Lang, and D. Buser, Digital subtraction radiography for the assessment of changes in peri-implant bone density Int. J. Oral Maxillofac. Radiol. 6, 160 – 166 (1991). 14. M. Jeffcoat, Radiographic methods for the detection of progressive alveolar bone loss. J. Periodontol. 63(Suppl.), 367 – 372 (1992). 15. M. K. Jeffcoat, M. S. Reddy, R. van den Berg, and E. Martens, Quantitative digital subtraction radiography for the assessment of peri-implant bone change. Clin. Oral Implants Res. 3, 22 – 27 (1992). 16. M. K. Jeffcoat and M. S. Reddy, Digital subtraction radiography for longitudinal assessment of peri-implant bone change: Method and validation. Adv. Dent. Res. 7, 196 – 201 (1993). 17. U. Braegger and L. Pasquali, Color conversion of alveolar bone density changes in digital subtraction images. J. Clin. Periodontol. 16, 209 – 214 (1989).

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia 18. M. S. Reddy, J. M. Bruch, M. K. Jeffcoat, and R. C. Williams, Contrast enhancement as an aid to interpretation in digital subtraction radiography. Oral. Surg. Oral Med. Oral Pathol. 71, 763 – 769 (1991). 19. M. K. Jeffcoat, M. S. Reddy, I. Magnusson, B. Johnson, M. P. Meredith, P. F. Cavanaugh, Jr., R. W. Gerlach, Efficacy of quantitative digital subtraction radiography using radiographs exposed in a multicenter trial. J. Periodontol Res. 31, 157 – 160 (1996). 20. K. A. Southard and T. E. Southard, Comparison of digitized radiographic alveolar feature between 20 and 70 year old women. Oral Surg. Oral Med. Oral Pathol. 74, 111 – 117 (1992). 21. U. E. Ruttimann, R. L. Webber, and R. B. Halzelrig, Fractal dimension from radiographs of peridental alveolar bone. A possible diagnostic indicator of osteoporosis. Oral Surg. Oral Med. Oral Pathol. 74, 98 – 110 (1992). 22. D. T. Allen, M. E. Gher, G. Quitero, and A. C. Richardson, Interpretation of linear and computed tomograms in the assessment of implant recipient sites. J. Periodontol. 64, 1243 – 1249 (1993). 23. J. A. Cliffe, Use of multiplanar computed tomography to aid in endosseous implant placement through hydroxyapatite overlay grafts. J. Oral Implantol. 16, 47 – 51 (1990). 24. D. K. Kassebaun, P. V. Nummikoski, R. G. Triplett, and R. P. Langlais, Cross sectional radiography for implant site assessment. Oral Surg. Oral Med. Oral Pathol. 70, 674 – 678 (1990). 25. R. A. Kraut, Utilization of 3D/Dental software for precise implant site selection: Clinical reports. Implant. Dent. 1, 134 – 139 (1992). 26. Y. Lamoral, M. Quirynen, P. Peene, F. Vanneste, S. F. Lemahieu, A. L. Baert, and D. van Steenberghe, Computed tomography in the preoperative planning of oral endo-osseous implant surgery. Fortschr. Geb. Rontgenstr. Nuklearmed. 153, 505 – 509 (1990). 27. S. L. Rothman, N. Chaftex, M. L. Rhodes, and M. S. Schwartz, CT in the preoperative assessment of the mandible and maxilla for endosseous implant surgery. Radiology 168, 171 – 175 (1988). 28. M. S. Schwartz, S. L. Rothman, N. Chafetz, and M. Rhodes, Computed tomography in dental implantation surgery. Dent. Clini North Am. 33, 555 – 597 (1989). 29. J. T. Grbic and I. B. Lamster, Risk indicators of future clinical attachment loss in adult periodontitis: Tooth and site variables. J. Periodontal Res. 63, 262 – 269 (1992). 30. A. J. Miller, “The National Survey of Oral Health in U.S. Employed Adults and Seniors: 1985 – 1986,” NIH Publication 87 – 2868. National Institutes of Health, Bethesda, MD, 1988. 31. C. D. Marshall-Day and K. L. Shourie, A roetgenographic survey of periodontal disease in India. J. Am. Dent. Assoc. 39, 572 (1949). 32. C. D. Marshall-Day, R. G. Stephens, and L. F. Quigley, Periodontal disease: Prevalence and incidence. J. Periodontol. 26, 185 – 203 (1955). 33. H. Loe, A. Anerud, H. Boysen, and E. Morrison, Natural history of periodontal disease in man. Rapid, moderate and no loss of attachment in Sri Lankan laborers 14 to 46 years of age. J. Clin. Periodontol. 13, 431 – 440 (1986). 34. A. Bolin, S. Lavstedt, L. Frithiof, and C. O. Henrikson, Proximal alveolar bone loss in a longitudinal radiographic investigation. IV. Smoking and other factors influencing the progress in a material of individuals with at least 20 remaining teeth. Acta Odontol. Scand. 44, 263 – 269 (1989). 35. J. Bergstrom and S. Eliasson, Cigarette smoking and alveolar bone in subjects with high standard of oral hygiene. J. Clin. Periodontol. 14, 466 – 469 (1987). 36. J. Bergstrom and S. Eliasson, Noxious effect of cigarette smoking on periodontal health. J. Periodontal Res. 22, 513 – 517 (1987). 37. J. Bergstrom, S. Eliasson, and H. Preber, Cigarette smoking and periodontal bone loss. J. Periodontol. 62, 242 – 246 (1991). 38. J. Haber, J. Wattles, M. Crowley, R. Mandell, K. Joshipura, and R. L. Kent, Evidence for cigarette smoking as a major risk factor for periodontitis. J. Periodontol. 64, 16 – 23 (1993).

379 39. J. Haber, Smoking is a major risk factor for periodontitis. Curr. Opinion Periodontol. 12 – 18 (1994). 40. T. C. Waldrop, D. C. Anderson, W. W. Hallmon, F. C. Schmalstieg, and R. L. Jacobs, Periodontal manifestations of the heritable Mac-1, LFA-1, deficiency syndrome. J. Periodontol. 58, 400 – 416 (1987). 41. T. C. Hart, Genetic considerations of risk in human periodontal disease. Curr. Opinion Periodontol. 3 – 11 (1994). 42. M. Melnick, E. Shields, and E. Bixler, Periodontosis: A phenotypic and genetic analysis. Oral Surg. 42, 32 – 41 (1976). 43. J. M. DiRienzo and J. Slots, Genetic approach to the study of epidemiology and pathogenesis of Actinobacillus actinomycemcomitans in localized juvenile periodontitis. Arch. Oral Biol. 35 (Suppl.), 79 – 84 (1990). 44. L. C. Macapanpan and J. P. Weinman, The influence of injury to the periodontal membrane on the spread of gingival inflammation. J. Dent. Res. 33, 263 – 272 (1954). 45. B. S. Michalowicz, D. Aeppli, J. G. Virag, D. G. Klump, J. E. Hinrichs, N. L. Segal, T. J. Bouchard, Jr., and B. L. Pihlstrom, Periodontal findings in adult twins. J. Periodontol. 62, 293 – 299 (1991). 46. B. S. Michalowicz, D. P. Aeppli, R. Kuba, J. E. Bereuter, J. P. Conry, N. L. Segal, T. J. Bouchard, Jr., and B. L. Pihlstrom, A twin study of genetic variation in proportional radiographic alveolar bone height. J. Dent. Res. 70, 1431 – 1435 (1991). 47. M. P. Massada, R. Persson, J. S. Kenney, S. W. Lee, R. C. Page, and A. C. Alisson, Measurement of interleukin-1 and 1 in gingival crevicular fluid: Implication for the pathogenesis of periodontal disease. J. Periodontal Res. 25, 156 – 163 (1990). 48. D. R. Bertolini, G. E. Nedwin, T. S. Bringmani, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumor necrosis factors. Nature 319, 561 – 518 (1996). 49. E. F. Rossomando, J. E. Kennedy, and J. Hadmimichael, Tumor necrosis factor alpha in gingival crevicular fluid as a possible indicator of periodontal disease in humans. Arch. Oral. Biol. 35, 431 – 434 (1990). 50. R. A. Reinhardt, M. P. Masada, W. B. Kaldahl, L. M. Dubois, K. S. Kornman, J. I. Choi, K. L. Kalkwarf, and A. C. Allison, Gingival fluid IL-1 and IL-6 levels in refractory periodontitis. J. Clin. Periodontol 20, 225 – 231 (1993). 51. P. F. Cavanaugh Jr., M. P. Meredith, W. Buchanan, M. J. Doyle, M. S. Reddy, and M. K. Jeffcoat, Coordinate production of PGE2 and IL-1 in the gingival crevicular fluid of adults with periodontitis: Its relationship to alveolar bone loss and disruption by twice daily treatment with ketorolac tromethamine oral rinse. J. Periodontal Res. 33, 75 – 82 (1998). 52. S. Offenbacher, B. M. Odle, and T. E. Van Dyke, The use of crevicular fluid prostaglandin E2 levels as a predictor of periodontal attachment loss. J. Periodontal Res. 21, 101 – 112 (1986). 53. S. Offenbacher, P. A. Heasman, and J. G. Collins, Modulation of host PGE2 secretion as a determinant of periodontal disease expression. J. Periodontol. 64, 432 – 444 (1993). 54. J. Woesnner Jr., The family of matrix metalloproteinases. Ann. NY Acad. Sci. 732, 11 – 21 (1994). 55. J. Woesnner Jr., Matrix metalloproteinases and their inhibitors in connective tissue remodeling. FASEB J. 5, 2145 – 2154 (1991). 56. H. Birkedal-Hansen, W. Moore, M. Bodden, L. J. Windsor, B. Birkedal-Hansen, A. DeCarlo, and J. A. Engler, Matrix metalloproteinases: A review. Crit. Rev. Oral. Biol. Med. 4, 197 – 250 (1993). 57. H. Birkedal-Hansen, Role of matrix metalloproteinases in human periodontal diseases. J. Periodontol 64, 474 – 484 (1993). 58. L. Golub, M. Wolff, S. Roberts, H-M. Lee, M. Leung, and G. Payonk, Treating periodontal diseases by blocking tissue-destructive enzymes. J. Am. Dent. Assoc. 125, 163 – 169 (1994). 59. K. S. Kornman, A. Crane, H.-Y. Wang, F. S. diGiovine, M. G. Newman, F. W. Pirk, T. G. Wilson Jr., F. L. Higginbottom, and G. W. Duff.

380

60.

61. 62. 63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73. 74.

75.

76.

77.

78.

79.

JEFFCOAT, REDDY, AND DECARLO The interleukin-1 genotype as a severity factor in adult periodontal disease. J. Clin. Periodontol. 24, 72 – 77 (1997). T. G. Frantzis, C. M. Reeve, and A. L. Brown, The ultrastructure of capillary basement membranes in the attached gingiva of diabetic and nondiabetic patients with periodontal disease. J. Periodontol. 42, 406 – 411 (1971). R. Clark, “Disorders of Granulocyte Chemotaxis.” Raven Press, New York, 1978. A. Saadoun, Diabetes and periodontal disease: A review and update. J. West. Soc. Periodontol. 23, 116 – 139 (1980). A. Ficara, M. Leven, M. Grower, and G. Fraker, A comparison of the glucose and protein content of gingival fluid from diabetics and nondiabetics. J. Periodontal Res. 10, 171 – 175 (1975). N. Rammamurthy and L. Golub, Diabetes increases collagenase activity in extracts of rat gingiva and skin. J. Periodontal Res. 18, 23 – 30 (1983). L. J. Emrich, M. Shlossman, and R. J. Genco, Periodontal disease in non-insulin-dependent diabetes mellitus. J. Periodontol. 62, 123 – 130 (1991). N. Von Wowern, B. Klausen, and K. Olgaard, Steriod-induced mandibular bone loss in relation to marginal periodontal changes. J. Clin. Periodontol. 19, 182 – 186 (1992). T. M. A. El-Attar, G. D. Roth, and A. Hugoson, Comparative metabolism of 4-C progesterone in normal and chronically inflamed human gingival tissue. J. Periodontal Res. 8, 79 – 83 (1973). J. B. Sweet and D. P. Butler, Increased incidence of postoperative localized osteitis in mandibular third molar surgery associated with patients using oral contraceptives. Am. J. Obstet Gynecol. 127, 518 (1977). K. S. Kornman, “Age, Supragingival Plaque, and Steroid Hormones as Ecological Determinants of the Subgingival Flora.” American Society for Microbiology, Washington, DC, 1982. J. F. Prichard, D. M. Ferguson, J. Windmiller, and W. C. Hart, Prepubertal periodontitis affecting the deciduous and permanent dentition in a patient with cyclic neutropenia. J. Periodontol. 55, 114 – 122 (1984). Y. Zubery, O. Moses, and A. Kozlovsky, Agranulocytosis: Periodontal manifestations and treatment of the acute phase: A case report. Clin. Prev. Dent. 13, 5 – 8 (1991). T. E. Van Dyke, M. A. Taubman, J. L. Ebersole, A. D. Haffajee, S. S. Socransky, D. J. Smith, and R. J. Genco, The Papillon-Lefevre syndrome: Neutrophil dysfunction with severe periodontal disease. Clin. Immunol. Immunopathol. 31, 419 – 429 (1984). A. Artzi, M. Gorsky, and M. Raviv, Periodontal manifestations of adult onset of histiocytosis X. J. Periodontol. 60, 57 – 60 (1989). L. Saxen, S. Aula, and T. Westermark, Periodontal disease associated with Down’s syndrome: An orthopantomographic evaluation. J. Periodontol. 48, 337 (1977). J. R. Winkler and P. B. Robertson, Periodontal disease associated with HIV infection. Oral Surg. Oral Med. Oral Pathol. 173, 145 – 150 (1992). P. N. Papapanou and J. L. Wennstrom, A 10-year retrospective study of periodontal disease progression. Clinical characteristics of subjects with pronounced and minimal disease development. J. Clin. Periodontol. 17, 78 – 84 (1990). V. Baelum, L. Wen-Min, G. Dahlen, O. Fejerskov, and C. Xia, Sixyear progression of destructive periodontal disease in 2 subgroups of elderly Chinese. J. Periodontol. 64, 891 – 899 (1993). J. M. Goodson, A. C. R. Tanner, A. D. Haffajee, G. C. Sornberger, and S. S. Socransky, Patterns of progression and regression of advanced periodontal disease. J. Clin. Periodontol. 9, 472 – 481 (1982). S. S. Socransky, A. C. R. Tanner, A. D. Haffajee, J. D. Hillman, and J. M. Goodson, “Present Status of Studies on the Microbial Etiology of Periodontal Diseases.” American Society for Microbiology, Washington, DC, 1982.

80. S. S. Socransky and A. D. Haffajee, The bacterial etiology of destructive periodontal disease: Current concepts. J. Periodontol. 63, 322 – 331 (1992). 81. A. D. Haffajee, S. S. Socransky, and J. M. Goodson, Clinical parameters as predictors of destructive periodontal disease activity. J. Clin. Periodontol. 10, 257 – 265 (1983). 82. A. D. Haffajee, S. S. Socransky, J. L. Ebersole, and D. J. Smith, Clinical microbiological, and immunological features associated with the treatment of active periodontosis lesions. J. Clin. Periodontol. 11, 600 – 618 (1984). 83. M. K. Jeffcoat and M. S. Reddy, Progression of probing attachment loss in adult periodontitis. J. Periodontol. 62, 185 – 189 (1991). 84. G. E. Carlsson and G. Persson, Morphologic changes of the mandible after extraction and wearing of dentures. Odontol. Rev. 18, 27 – 54 (1967). 85. H. Devlin and M. W. Ferguson, Alveolar ridge resorption and mandibular atrophy. A review of the role and local and systemic factors. Br. Dent. J. 170, 101 – 104 (1991). 86. J. C. Baxter, Relationship of osteoporosis to excessive residual ridge resorption. J. Prosthet. Dent. 46, 123 – 125 (1981). 87. B. Bergman, G. E. Carlsson, and S. Ericson, Effect of differences in habitual use of complete dentures on underlying tissues. Scand. J. Dent. Res. 79, 449 – 460 (1971). 88. R. L. Cambell, A comparative study of the resorption of the alveolar ridges in denture wearers and non-denture wearers. J. Am. Dent. Assoc. 60, 143 – 153 (1960). 89. T. W. Brehm and B. J. Abadhi, Patient response to variations in complete denture technique. J. Prosthet. Dent. 44, 491 – 494 (1980). 90. P. R. Likeman and D. M. Watt, Morphologic changes in the maxillary denture bearing area. A follow-up 14 – 17 years after tooth extraction. Br. Dent. J. 136, 500 – 503 (1974). 91. K. E. Wical and P. Brussee, Effects of calcium and vitamin D supplements on alveolar ridge resorption in immediate denture patients. J. Prosthet. Dent. 41, 4 – 11 (1979). 92. A. M. Parfitt, Quantum concept of bone remodeling and turnover: Implication for the pathogenesis of osteoporosis. Calcif. Tissue Int. 28, 1 – 5 (1979). 93. A. M. Parfitt, The cellular basis of bone remodeling: The quantum concept reexamined in light of recent advance in the cell biology of bone. Calcif. Tissue Int. 36, S37 — S45 (1984). 94. C. D. Rubin, Southwestern Internal Medicine Conference: Age-related osteoporosis. Am. J. Med. Sci. 301, 281 – 298 (1991). 95. R. Baron, A. Vignery, and M. Horowitz, “Lymphocytes, Macrophages and the Regulation of Bone Remodeling.” Elsevier, New York, 1984. 96. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr. Rev. 7, 379 – 408 (1986). 97. G. Vaes, J.-M. Delaisse, and Y. Eeckhout, “Relative Roles of Collagenase and Lysosomal Cysteine-Proteinases in Bone Resorption. Gustav Fischer, Stuttgart, 1992. 98. Y. Ishimi, C. Miyaura, C. Jin, H. Akatsu, T. Abe, Y. Nakamura, A. Yamaguchi, S. Yoshiki, T. Matsuda, T. Airano, T. Koshimoto, and T. Suda, IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297 – 3306 (1990). 99. P. M. J. McSheehy and T. J. Chambers, Osteoblast-like cells in the presence of parathyroid hormone release soluble factor that stimulates osteoclastic bone resorption. Endocrinology 119, 1654 – 1662 (1986). 100. P. M. J. McSheehy and T. J. Chambers, Osteoblastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology 118, 824 – 830 (1986). 101. P. M. J. McSheehy and T. J. Chambers, 1,25-Dihydroxyvitamin D stimulates rat osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption. J. Clin. Invest. 80, 425 – 429 (1987).

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia 102. B. M. Thompson, G. R. Mundy, and T. J. Chambers, Tumor necrosis factors  and b induce osteoblastic cells to stimulate osteoclastic bone resorption. J. Immunol. 138, 775 – 779 (1987). 103. M. J. Oursler, J. P. Landers, B. L. Riggs, and T. C. Spelsberg, Oestrogen effects on osteoblasts and osteoclasts. Ann. Med. 25, 361 – 371 (1993). 104. T. J. Chambers and K. Fuller, Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic contact. J. Cell Sci. 76, 155 – 164 (1991). 105. W. E. C. Moore, L. V. Holdeman, R. M. Smibert, D. E. Hash, J. A. Burmeister, and R. R. Ranney, Bacteriology of severe periodontitis in young adult humans. Infect. Immunol. 38, 1137 – 1148 (1982). 106. W. E. C. Moore, Microbiology of periodontal disease. J. Periodontal Res. 22, 335 – 341 (1987). 107. L. V. H. Moore, W. E. C. Moore, E. P. Cata, R. M. Smibert, J. A. Burmeister, and A. M. Best, Bacteriology of human gingivitis. J. Dent. Res. 66, 989 – 995 (1987). 108. A. Lovdal, O. Schei, J. Waerhaug, and A. Arno, Tooth mobility and alveolar bone resorption as a function of occlusal stress and oral hygiene and age. Acta Odontol. Scand. 17, 61 – 77 (1959). 109. H. Loe, E. Theilade, and S. B. Jensen, Experimental gingivitis in man. J. Periodontol. 36, 177 – 187 (1965). 110. J. Slots, H. S. Reynolds, and R. J. Genco, Actinobacillus actinomycetemcomitans in human periodontal disease: A crosssectional microbiologic investigation. Infect. Immunol. 29, 1013 – 1020 (1980). 111. A. C. R. Tanner, C. Haffer, G. T. Bratthall, R. A. Visconti, and S. S. Socransky, A study of the bacteria associated with advancing periodontitis in man. J. Clin. Periodontol. 6, 278 – 307 (1979). 112. J. M. Albandar, I. Olsen, and P. Gjermo, Associations between six DNA probe-detected periodontal bacteria and alveolar bone loss and other clinical signs of periodontitis. Acta Odontol. Scand. 48, 415 – 423 (1990). 113. A. C. R. Tanner, S. S. Socransky, and J. M. Goodson, Microbiota of periodontal pockets losing crestal alveolar bone. J. Periodontal Res. 19, 279 – 291 (1984). 114. H. Liakoni, P. Barber, and H. N. Newman, Bacterial penetration of pocket soft tissues in chronic adult and juvenile periodontitis cases. An ultrastructural study. J. Periodontol. 14, 22 – 28 (1987). 115. J. L. Dzink, S. S. Socransky, and A. D. Haffajee, The predominant cultivable microbiota of active and inactive lesions of destructive periodontal diseases. J. Clin. Periodontol. 15, 316 – 323 (1988). 116. A. D. Haffajee, S. S. Socransky, J. L. Dzink, M. A. Taubman, J. L. Ebersole, and D. J. Smith, Clinical, microbiological and immunological features of subjects with destructive periodontal diseases. 15, 240 – 246 (1988). 117. W. J. Loesche, The identification of bacteria associated with periodontal disease and dental caries by enzymatic methods. Oral Microbiol. Immunol. 1, 65 – 70 (1986). 118. J. Slots and R. J. Genco, Microbial pathogenicity. J. Dent. 63, 412 – 421 (1984). 119. R. L. Jensen and L. E. A. Folke, The passage of exogenous tritiated thymidine into gingival tissues. J. Periodontol. 45, 786 – 791 (1974). 120. J. Schwartz, F. L. Stinson, and R. B. Parker, The passage of tritiated bacterial endotoxin across intact gingival crevicular epithelium. J. Periodontol. 43, 270 – 276 (1972). 121. T. Takata, H. Nikai, N. Ijuhin, and H. Ito, Penetration and uptake of colloidal gold-labeled concanavalin A in the junctional epithelium of the rat. J. Periodontol. 59, 823 – 829 (1988). 122. T. Abe, Y. Hara, and M. Aono, Penetration, clearance and retention of antigen en route from the gingival sulcus to the draining lymph node of rats. J. Periodontal Res. 26, 429 – 439 (1991). 123. M. Tsutsui, N. Utsumi, and K. Tsubakimoto, Cellular components of staphylococci and streptococci in inflamed human gingiva. J. Dent. Res. 47, 663 (1968).

381 124. A. A. DeCarlo, “Matrix Metalloproteinase Activation and Induction in Keratinocytes by a Purified Porphyromonas Gingivalis Thiol-Proteinase”, Chap. 1. Univ. of Alabama at Birmingham, 1994. 125. E. M. Unemori and Z. Werb, Collagenase expression and endogenous activation in rabbit synovial fibroblasts stimulated by the calcium ionophore A23187. J. Biol. Chem. 263, 16252 – 16259 (1988). 126. Z. Werb and J. Aggeler, Proteases induce secretion of collagenase and plasminogen activator by fibroblasts. Proc. Natl. Acad. Sci. USA 75, 1839 – 1843 (1978). 127. H. Birkedal-Hansen, B. R. Wells, H. L. Lin, P. W. Caufield, and R. E. Taylor, Activation of keratinocyte-mediated collagen (Type I) breakdown by suspected human periodontopathogen. Evidence of a novel mechanism of connective tissue breakdown. J. Periodontal Res. 19, 645 – 650 (1984). 128. D. E. Woolley and R. M. Davies, Immunolocalization of collagenase in periodontal disease. J. Periodontal Res. 16, 292 – 297 (1981). 129. H. Takeyama, B. Birkedal-Hansen, W. G. I. Moore, M. K. Bodden, and H. Birkedal-Hansen, Immunolocalization of fibroblasts-type collagenase, Mr 72K gelatinase and TIMP-1 in inflamed human gingiva. J. Dent. Res. 72, 437 (1993). [Abstract] 130. J. Carlsson, B. F. Herrmann, J. F. Hofling, and G. K. Sundqvist, Degradation of the human proteinase inhibitors alpha-1antitrypsin and alpha2-macroglobulin by Bacteroides gingivalis. 43, 644 – 648 (1983). 131. B. F. Herrmann, J. Carlsson, and G. K. Sundqvist, Inactivation of guinea-pig serum proteinase inhibitors by Bacteroides gingivalis. Scand. J. Dent. Res. 93, 153 – 157 (1985). 132. M. Sato, M. Otsuka, R. Maehara, J. Endo, and R. Nakamura, Degradation of human secretory immunoglobulin A by protease isolated from the anaerobic periodontopathogenic bacterium Bacteroides gingivalis. Arch. Oral Biol. 32, 235 – 238 (1987). 133. J. A. Wingrove, R. G. DiScipio, Z. Chen, J. Potempa, J. Travis, and T. E. Hugli, Activation of complement components C3 and C5 by a cysteine proteinase (gingipain-1) from Porphyromonas (Bacteroides) gingivalis. J. Biol. Chem. 267, 18902 – 18907 (1992). 134. D. Hinode, A. Nagata, S. Ichimiya, H. Hayashi, M. Morioka, and R. Nakamura, Generation of plasma kinin by three types of protease isolated from Porphyromonas gingivalis 381. Arch. Oral Biol. 37, 859 – 861 (1992). 135. T. Imamura, R. N. Pike, J. Potempa, and J. Travis, Pathogenesis of periodontitis: A major arginine-specific cysteine proteinase from Porphyromona gingivalis induces vascular permeability enhancement through activation of the kallikrein/kinin pathway. J. Clin. Invest. 94, 361 – 367 (1994). 136. A. Horowitz and L. E. A. Folke, Hydrogen sulfide in the periodontal environment. J. Periodontol. 44, 390 – 395 (1973). 137. R. E. Singer and B. A. Buckner, Butyrate and propionate: Important components of toxic dental plaque extracts. Infect. Immunol. 32, 458 – 463 (1981). 138. S. Holt and T. E. Bramanti, Factors in virulence expression and their role in periodontal disease pathogenesis. Crit. Rev. Oral Biol. Med. 2, 177 – 281 (1991). 139. Y. Iino and R. M. Hopps, The bone resorbing activities in tissue culture of lipopolysaccharides from the bacteria Actinobacillus actinomycetemcomitans, Bacteroides gingivalis, and Capnocytophaga ochracea isolated from human mouths. Arch. Oral Biol. 29, 59 – 63 (1984). 140. C. G. McFarlane, J. J. Reynolds, and M. C. Meikle, The release of interleukin-1b, tumor necrosis factor- and interferon- by cultured peripheral blood mononuclear cells from patients with periodontitis. J. Periodontal Res. 25, 207 – 214 (1990). 141. R. A. Lindemann, J. S. Economou, and H. Rothermel, Production of interleukin-1 and tumor necrosis factor by human peripheral monocytes activated by periodontal bacteria and extracted lipopolysaccharides. J. Dent. Res. 67, 1131 – 1135 (1988).

382 142. S. W. Garrison and F. C. Nichols, LPS-elicited secretory responses in monocytes: Altered release of PGE2 but not IL-1b in patients with adult periodontitis. J. Periodont. Res. 24, 88 – 95 (1989). 143. J. Pugin, C. C. Schurer-Maly, D. Leturcq, A. Moriarty, R. J. Ulevitch, and P. S. Tobias, Lipopolysaccharide (LPS) activation of human endothelial and epithelial cells is mediated by LPS binding protein and soluble CD14. Proc. Natl. Acad. Sci. USA 90, 2744 – 2748 (1993). 144. C. R. H. Raetz, Biochemistry of endotoxins. Annu. Rev. Biochem. 59, 129 – 170 (1990). 145. L. Wahl, S. M. Wahl, S. E. Mergenhagen, and G. E. Martin, Collagenase production by endotoxin-activated macrophages. Proc. Natl. Acad. Sci. USA 71, 3598 – 3601 (1974). 146. J. Cury, E. J. Campbell, C. J. Lazarus, R. J. Albin, and H. G. Welgus, Selective upregulation of human alveolar macrophage collagenase production by lipopolysaccharide and comparison to collagenase production by fibroblasts. J. Immunol. 141, 4306 – 4312 (1988). 147. H. J. Sismey-Durrant, S. J. Atkinson, R. M. Hopps, and J. K. Heath, The effect of lipopolysaccharide from Bacteroides gingivalis and muramyl dipeptide on osteoblast collagenase release. Calcif. Tissue Int. 44, 361 – 363 (1989). 148. H. Takada, J. Mihara, I. Morisaki, and S. Hamada, Induction of interleukin-1 and -6 in human gingival fibroblast cultures stimulated with Bacteroides lipopolysaccharides. Infect. Immunol. 59, 295 – 301 (1991). 149. H. I. Baranowska, R. M. Palmer, and R. F. Wilson, A comparison of antibody levels to Bacteroides gingivalis in serum and crevicular fluid from patients with untreated periodontitis. Oral Microbiol. Immunol. 4, 173 – 175 (1989). 150. J. R. McGhee, K. W. Beagley, J. H. Eldridge, K. Fujihashi, C. Lue, A. Moldoveanu, J. Mestecky, J. Radl, and H. Kiyono, Interleukin cascade for the regulation of IgA synthesis and immune responses. Protides Biol. Fluids 36, 183 – 191 (1989). 151. K. Monefeldt and T. Tollefsen, Serum IgG antibodies reactive with lipoteichoic acid in adult patients with periodontitis. J. Clin. Periodontol. 16, 519 – 524 (1989). 152. P. A. Murray, D. A. Burstein, and J. R. Winkler, Antibodies to Bacteroides gingivalis in patients with treated and untreated periodontal disease. J. Periodontol. 60, 96 – 103 (1989). 153. K. Schenck, S. R. Porter, T. Tollefsen, J. R. Johansen, and C. Scully, Serum levels of antibodies against Actinobacillus actinomycetemcomitans in various forms of human periodontitis. Acta Odontol. Scand. 47, 271 – 277 (1989). 154. D. W. Turner, G. J. Dai, B. Merrell, and P. J. Robinson, Serum and gingival tissue antibody levels to oral microbial antigens in human chronic adult periodontitis. Microbios 60, 133 – 140 (1989). 155. K. Fujihashi, J. R. McGhee, M. Yamamoto, K. W. Beagley, and H. Kiyono, “Cytokine Networks and Immunoglobulin Synthesis in Inflamed Gingival Tissues.” American Society for Microbiology Press, Washington, DC, 1994. 156. T. Ogawa, Y. Kono, M. L. McGhee, J. R. McGhee, J. E. Roberts, S. Hamada, and H. Kiyono, Porphyromonas gingivalis-specific serum IgG and IgA antibodies originate from immunoglobulin-secreting cells in inflamed gingiva. Clin. Exp. Immunol. 83, 237 – 244 (1991). 157. B. Afar, D. Engel, and E. A. Clark, Activated lymphocyte subsets in adult periodontitis. J. Periodontal Res. 27, 126 – 133 (1992). 158. T. Ogawa, M. L. McGhee, Z. Moldoveanu, S. Hamada, J. Mestecky, J. R. McGhee, and H. Kiyono, Bacteroides-specific IgG and IgA subclass antibody-secreting cells isolated from chronically inflamed gingival tissues. Clin. Exp. Immunol. 76, 103 – 110 (1989). 159. J. Tew, D. Engel, and D. Mangan, Polyclonal B-cell activation in periodontitis. J. Periodontal Res. 24, 225 – 241 (1989). 160. H. A. Schenkein and R. J. Genco, Gingival fluid and serum in periodontal diseases. II. Evidence for cleavage of complement components C3, C3 proactivator (Factor B) and C4 in gingival fluid. J. Periodontol. 48, 778 – 784 (1977).

JEFFCOAT, REDDY, AND DECARLO 161. H. A. Schenkein, The role of complement in peridontal diseases. Crit. Rev. Oral Biol. Med. 2, 65 – 81 (1991). 162. M. Bickel, The role of interleukin-8 in inflammation and mechanisms of regulation. J. Periodontol. 64, 456 – 460 (1993). 163. N. McCartney-Francis, D. Mizel, H. Wong, L. M. Wahl, and S. M. Wahl, TGF-b regulates production of growth factors and TGF-b by human peripheral blood monocytes. Growth Factors 4, 27 – 35 (1990). 164. P. T. Marucha, R. A. Zeff, and D. L. Kreutzer, Regulation of IL-1b gene expression in human peripheral blood PMN. J. Periodontal Res. 26, 264 – 267 (1991). 165. S. Offenbacher, P. A. Heasman, and J. G. Collins, Modulation of host PGE2 secretion as a determinant of periodontal disease expression. J. Periodontol. 64, 432 – 444 (1993). 166. M. Zembala and G. L. Asherson, “Human Monocytes.” Academic Press, San Diego, 1989. 167. U. Skaleric, J. B. Allen, P. D. Smith, S. E. Mergenhagen, and S. M. Wahl, Inhibitors of reactive oxygen intermediates suppress bacterial cell wall-induced arthritis. J. Immunol. 147, 2559 – 2564 (1991). 168. H. Birkedal-Hansen, Role of matrix metalloproteinases in human periodontal diseases. J. Periodontol. 64, 474 – 484 (1993). 169. H. Birkedal-Hansen, W. G. I. Moore, M. K. Bodden, L. J. Windsor, B. Birkedal-Hansen, A. DeCarlo, and J. A. Engler, Matrix metalloproteinases: A review. Crit. Rev. Oral Biol. Med. 4, 197 – 250 (1993). 170. E. Gemmell and G. J. Seymour, Interleukin 1, interleukin 6 and transforming growth factor-b production by human gingival mononuclear cells following stimulation with Porphyromonas gingivalis and Fusobacterium nucleatum. J. Periodontal Res. 28, 122 – 129 (1993). 171. Y. Matsuki, T. Yamamoto, and K. Hara, Detection of inflammatory cytokine messenger RNA (mRNA)-expressing cells in human inflamed gingiva by combined in situ hybridization and immunohistochemistry. Immunology 76, 42 – 47 (1992). 172. R. A. Reinhardt, M. P. Masada, W. B. Kaldahl, L. M. DuBois, K. S. Kornman, J.-I. Choi, K. L. Kalkwarf, and A. C. Allison, Gingival fluid IL-1 and IL-6 levels in refractory periodontitis. J. Clin. Periodontol. 20, 225 – 231 (1993). 173. Y. Takeuchi, H. Yoshie, and K. Hara, Expression of interleukin-2 receptor and HLA-DR on lymphocyte subsets of gingival crevicular fluid in patients with periodontitis. J. Periodontal Res. 26, 502 – 510 (1991). 174. P. Stashenko, J. J. Jandinski, P. Fujiyoshi, J. Rynar, and S. S. Socransky, Tissue levels of bone resorptive cytokines in periodontal disease. J. Periodontol. 62, 504 – 509 (1991). 175. S. M. Wahl, “Cellular and Molecular Interaction in the Induction of Inflammation in Rheumatic Diseases.” Dekker, New York, 1991. 176. E. F. Rossomando and L. B. White, A novel method for the detection of TNF-alpha in gingival crevicular fluid. J. Periodontol. 64, 445 – 449 (1993). 177. G. R. Mundy, Cytokines and local factors which affect osteoclast function. Int. J. Cell Cloning 10, 215 – 222 (1992). 178. M. Kawano, I. Yamamoto, K. Iwato, H. Tanaka, H. Akoku, O. Tanabe, H. Ishikawa, M. Noduyoshi, Y. Ohuto, Y. Husi, and A. Kuramoto, Interleukin-1b rather than lymphotoxin as the major bone resorbing activity in human multiple myeloma. Blood 73, 1646 – 1649 (1989). 179. J. A. Lorenzo, S. L. Sousa, J. M. Fonseca, J. M. Hock, and E. S. Medlock, Colony stimulating factors regulate the development of multinucleated osteoclasts from recently replicated cells in vitro. J. Clin. Invest. 80, 160 – 164 (1987). 180. K. Heath, S. J. Atkinson, M. C. Meikle, and J. J. Reynolds, Mouse osteoblasts synthesize collagenase in response tobone resorbing agents. Biochem. Biophys. Acta 802, 151, 163 (1984).

CHAPTER 55 Oral Bone Loss and Systemic Osteopenia 181. S. B. Mizel, J. M. Dayer, S. M. Krane, and J. Mergenhagen, Stimulation of rheumatoid synovial cell collagenase and prostaglandin production by partially purified lymphocyte-activating factor (interleukin 1). Proc. Natl. Acad. Sci. USA 78, 2474 – 2477 (1981). 182. S. C. Gilman, J. Chang, P. R. Zeigler, J. Uhl, and E. Mochan, Interleukin-1 activates phopholipase A2 in human synovial cells. Arthritis Rheum. 31, 126 – 130 (1988). 183. A. Ito, Y. Itoh, Y. Sasaguri, M. Morimatsu, and Y. Mori, Effects of interleukin-6 on the metabolism of connective tissue components in rheumatoid synovial fibroblasts. Arthritis Rheum. 35, 1197 – 1201 (1992). 184. G. J. Lyons, H.-Y. Lin, T. Salo, H. Larjava, A. DeCarlo, B. BirkedalHansen, and H. Birkedal-Hansen, “Expression of Collagen-Cleaving Matrix Metalloproteinases by Keratinocytes: Effect of Growth Factors and Cytokines and of Microbial Mediators.” Quintessence, Tokyo, 1991. 185. D. R. Bertonlini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumor necrosis factors. Nature 319, 516 – 518 (1986). 186. M. S. Nanes, J. Rubin, G. Titus, G. Hendy, and B. D. Catherwood, Tumor necrosis factor- inhibits 1,25-dihydroxyvitamin D3stimulated bone Gla protein synthesis in rat osteosarcoma cells (ROS 17/2.8) by a pretranslational mechanism. Endocrinology 128, 2577 – 2584 (1991). 187. P. Stashenko, M. S. Obernesser, and F. E. Dewhirst, Effect of immune cytokines on bone. Immunol. Invest. 18, 239 – 249 (1989). 188. Y. P. Li and P. Stashenko, Pro-inflammatory cytokines tumor necrosis factor- and IL-6, but not IL-1, down-regulate the osteocalcin gene promoter. J. Immunol. 148, 788 – 794 (1992). 189. C. A. Brenner, R. R. Adler, D. A. Rappolee, R. A. Pedersen, and Z. Werb, Genes for extracellular-matrix-degrading metalloproteinases and their inhibitor, TIMP, are expressed during early mammalian development. Genes Dev. 3, 848 – 859 (1989). 190. M. Gowen and G. R. Mundy, Actions of recombinant interleukin1, interleukin 2, and Interferon- on bone resorption in vitro. J. Immunol. 136, 2478 – 2482 (1986). 191. J. Saklatvala, Tumour necrosis factor  stimulates resorption and inhibits synthesis of proteoglycan in cartilage. Nature 322, 547 – 549 (1986). 192. J. K. Heath, J. Saklatvala, M. C. Meikle, S. J. Atkinson, and J. J. Reynolds, Pig interleukin-1 (catabolin) is a potent stimulator of bone resorption in vitro. Calcif. Tissue Int. 37, 95 – 97 (1985). 193. J. Honig, C. Rordorf-Adam, D. Siegmund, W. Wiedemann, and R. Erard, Increased interleukin-1 beta (IL-1b) concentration in gingival tissue from periodontitis patients. J. Periodontal Res. 24, 362 – 367 (1989). 194. E. D. Stoufi, M. A. Taubman, J. L. Ebersole, D. J. Smith, and P. P. Stashenko, Phenotypic analyses of mononuclear cells recovered from healthy and diseased human periodontal tissues. J. Clin. Immunol. 7, 235 – 245 (1987). 195. M. A. Taubman, J. W. Eastcott, H. Shimauchi, O. Takeichi, and D. J. Smith, “Modulatory Role of T Lymphocytes in Periodontal Inflammation.” American Society for Microbiology Press, Washington, DC, 1994. 196. K. Fujihashi, J. Yoshiharu, K. W. Beagley, M. Yamamoto, J. R. McGhee, J. Mestecky, and H. Kiyono, Cytokines and periodontal disease: Immunopathological role of interleukins in chronic inflamed gingival tissues. J. Periodontol. 64, 400 – 406 (1993). 197. H. L. Wong, M. T. Lotze, L. M. Wahl, and S. M. Wahl, Administration of recombinant IL-4 to humans regulates gene expression, phenotype and function in circulating monocytes. J. Immunol. 148, 2118 – 2125 (1992). 198. D. F. Fiorentino, M. W. Bond, and T. R. Mosmann, Two types of mouse helper cells. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones. J. Exp. Med. 170, 2081 – 2095 (1989).

383 199. M. Howard, A. O’Garra, H. Ishida, R. DeWaal Malefyt, and J. De Vries, Biological properties of interleukin 10. J. Clin. Immunol. 12, 239 – 247 (1992). 200. H. L. Wong, M. T. Lotze, L. M. Wahl, and S. M. Wahl,Interleukin-4 differentially regulates monocyte IL-1 family gene expression and synthesis in vitro and in vivo. J. Exp. Med. 177, 775 – 781 (1993). 201. P. Moissec, J. Briolay, J. Dechaniet, J. Wijdenes, H. MartinezValdez, and J. Banchereau, Inhibition of the production of proinflammatory cytokines and immunoglobulins by interleukin-4 in an ex vivo model of rheumatoid synovitis. Arthritis Rheumatol. 35, 874 (1992). 202. D. F. Mangan, S. E. Mergenhagen, and S. M. Wahl, Apoptosis in human monocytes: Possible role in chronic inflammatory diseases. J. Periodontol. 64, 461 – 466 (1993). 203. D. F. Mangan, B. Robertson, and S. M. Wahl, IL-4 enhances programmed cell death (apoptosis) in stimulated human monocytes. J. Immunol. 148, 1812 – 1816 (1992). 204. M. J. Fenton and J. A. Burns, IL-4 reciprocally regulates IL-1 and IL-4 antagonist expression in human monocytes. J. Immunol. 149, 925 – 931 (1992). 205. W. A. Falkler, Jr., R. Lai, J. W. Vincent, L. Dober, C. Spiegel, and S. Hayduk, The Elisa system for measuring antibody reactive with Fusobacterium nucleatum in the sera of patients with chronic periodontitis J. Periodontol. 53, 762 – 766 (1982). 206. K. Fujihashi, K. W. Beagley, Y. Kono, W. A. Aicher M. Yamamoto, S. DiFabio, J. Xu-Amano, J. R. McGhee, and H. Kiyono, Gingival mononuclear cells from chronic inflammatory periodontal tissues produce interleukin (IL)-5 and IL-6 but not IL-2 and IL-4. Am. J. Pathol. 142, 1239 – 1250 (1993). 207. K. Yamashita, J. W. Eastcott, M. A. Taubman, D. J. Smith, and D. S. Cox, Effect of adoptive transfer of cloned Actinobacillus actinomycetemcomitans-specific T helper cells on periodontal disease. Infect. Immunol. 59, 1529 – 1534 (1991). 208. P. Stashenko, L. Nguyen, and Y.-P. Li, “Mechanisms of Regulation of Bone Formation by Proinflammatory Cytokines.” American Society for Microbiology Press, Washington, DC, 1994. 209. C. A. Dinarello, Interleukin-1 and its biologically related cytokines. Adv. Immunol. 44, 153 – 205 (1989). 210. P. A. Heasman, J. G. Collins, and S. Offenbacher, Changes in crevicular fluid levels of IL-1b, LTB4, PGE2, TxB2 and TNF in experimental gingivitis in humans. J. Periodontal Res. 28, 241 – 247 (1993). 211. S. Offenbacher, A. Soskolne, and J. G. Collins, “Markers of Periodontal Disease Susceptibility and Activity Present within the Gingival Crevicular Fluid.” Cambridge Univ. Press, Cambridge, 1991. 212. S. Offenbacher, B. M. Odle, and T. E. Van Dyke, The use of crevicular fluid prostaglandin E2 levels as a predictor of periodontal attachment loss. J. Periodontal Res. 21, 101 – 112 (1986). 213. S. Offenbacher, J. G. Collins, and R. R. Arnold, New clinical diagnostic strategies based on pathogenesis of disease. J. Periodontal Res. 28, 523 – 535 (1993). 214. B. F. Boyce, T. B. Aufemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interleukin-1 on bone turnover in normal mice. Endocrinology 123, 1142 – 1150 (1989). 215. S. Meghji, B. Henderson, S. Nair, and M. Wilson, Inhibition of bone DNA and collagen production by surface-associated material from bacteria implicated in the pathology of periodontal disease. J. Periodontol. 63, 736 – 742 (1992). 216. M.K. Jeffcoat, Periodontitis induced alveolar bone loss and its treatment. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. Favus, Ed), pp. 461 – 465. Lippincott, Williams and Willis, New York, 2000. 217. M. S. Reddy, K. G. Palcanis, M. L. Barnett, S. Haigh, C. H. Charles, and M. K. Jeffcoat, Efficacy of Meclofenamate Sodium in the treatment of rapidly progressive periodontitis. J. Clin. Periodontol. 20, 635 – 640 (1993).

384 218. R. C. Williams, M. K. Jeffcoat, T. H. Howell, A. Rolla, D. Stubbs, K. W. Teoh, M. S. Reddy, and P. Goldhaber, Altering the progression of human alveolar bone loss with the non-steroidal anti-inflammatory drug flurbiprofen. J. Periodontol 60, 485 – 490 (1989). 219. M. K. Jeffcoat and C. Chesnut, Systemic osteoporosis and oral bone loss J. Am. Dent. Assoc. 124, 49 – 56 (1993). 220. M. K. Jeffcoat, Bone loss in the oral cavity. J. Bone Miner. Res. 8, S467 – S474 (1993). 221. M. K. Jeffcoat, Osteoporosis a possible modifying factor in oral bone loss. Ann. Periodontol. 3, 312 – 321, (1998). 222. P. J. Kribbs, D. E. Smith, and C. H. Chesnut, Oral findings in osteoporosis. II. Relationship between residual ridge and alveolar bone resorption and generalize skeletal osteopenia. J. Prosthet. Dent. 50, 719 – 724 (1983). 223. P. J. Kribbs, C. H. Chesnut, S. M. Ott, and R. F. Kilcyne, Relationships between mandibular and skeletal bone in a population of normal women. J. Prosthet. Dent. 63, 86 – 89 (1990). 224. H. W. Daniel, Postmenopausal tooth loss. Arch. Intern. Med. 143, 1678 (1983). 225. E. A. Krall, B. Dawson-Hughes, A. Papas, and R. I. Garcia, Tooth loss and skeletal bone density in healty postmenopausal women. Osteoporosis Int. 4, 104 – 109 (1994). 226. L. Habets, J. Bras, and J. P. R. van Merkstyn, Mandibular atrophy and metabolic bone loss. Histomorphometry of iliac crest biopsies. J. Oral Maxillofac. Surg. 17, 325 – 329 (1988). 227. L. F. Ortman, E. Hausmann, and R. G. Dunford, Skeletal osteopenia and residual ridge resorption. J. Prosthet. Dent. 61, 321 – 325 (1989). 228. R. Yemm, Analysis of patients referred over a period of five years to a teaching hospital consultant service in dental prosthetics. Br. Dent. J. 159, 304 – 306 (1985). 229. P. L. Elders, L. L. Habets, J. C. Netelenbos, L. W. van der Linden, and P. F. van der Stelt, The relationship between periodontitis and systemic bone mass in women between 46 – 55 years of age. J. Clin. Periodontal. 19, 492 – 496 (1992). 230. E. Klemetti, P. Vainio, V. Lassila, and E. Alhava, Trabecular bone mineral density of mandible and alveolar height in postmenopausal women. Scand. J. Dent. Res. 101, 166 – 170 (1993). 231. R. J. Weyant, M. E. Pearlstein, A. P. Churak, K. Forrest, P. Famili, and J. Cauley, The association between osteopenia and periodontal attachment loss in older women. J. Periodontol. 70, 982 – 991 (1999). 232. M. K. Jeffcoat, C. E. Lewis, M. S. Reddy, C. Y. Wang, and M. Redford, Postmenopausal bone loss and its relationship to oral bone loss. Periodontology 2000 23, 94 – 102, (2000). 233. “Consensus Report”, Section II. World Workshop in Clinical Periodontics, The American Academy of Periodontology, Princeton, NJ, 1989. 234. “Consensus Report,” Section IV. World Workshop in Clinical Periodontics, American Academy of Periodontology, Princeton, NJ, 1989. 235. C. H. Drisko. Non-surgical pocket therapy. Pharmacotherapeutics 1, 491 – 566 (1996).

JEFFCOAT, REDDY, AND DECARLO 236. S. Aitken, P. Birek, G. V. Kulkarni, W. L. Lee, and C. A. McCulloch, Serial doxycycline and metronidazole in prevention of recurrent periodontitis in high risk patients. J. Periodontol. 63, 87 – 92 (1992). 237. W. Al-Joburi, T. C. Quee, C. Lautar, I. Iugovaz, J. Bourgouin, F. Delorme, and E. C. S. Chan, Effects of adjunctive treatment of periodontitis with tetracycline and spiramycin. J. Periodontol. 60, 533 – 539 (1989). 238. S. S. Ciancio, Chemotherapeutic agents and periodontal therapy. Their impact on clinical practice. J. Periodontol. 57, 108 – 111 (1986). 239. J. Gordon, C. Walker, C. Hovliaris, and S. Socransky, Efficacy of clindamycin hydrochloride in refractory periodontitis: 24-Month results. J. Periodontol. 61, 686 – 691 (1990). 240. W. J. Loesche, E. Schmidt, B. A. Smith, E. C. Morrison, R. Caffesse, and P. P. Hujoel, Effects of metronidazole on periodontal treatment needs. J. Periodontol. 62, 247 – 257 (1991). 241. J. Slots and T. E. Rams, Antibiotics in periodontal therapy: Advantages and disadvantages. J. Clin. Periodontol. 17, 479 – 493 (1990). 242. P.-O. Soder, L. Frithiof, S. Wikner, F. Wouters, P.-E. Engstrom, B. Rubin, U. Nedlich, and B. Soder, The effect of systemic metronidazole after non-surgical treatment in moderate and advanced periodontitis in young adults. J. Periodontol. 61, 281 – 288 (1990). 243. K. Kornman, Controlled release local delivery antimicrobials in periodontics: Prospects for the future. J. Periodontol. 64, 782 – 791 (1993). 244. M. K. Jeffcoat, et al. Adjunctive use of a subgingival controlled-release chlorhexidine chip (PerioChip™) reduces probing depth and improves attachment level compared with scaling and root planing alone. J. Periodontol. 69, 989 – 997 (1998). 245. J. M. Goodson, M. A. Cugini, R. L. Kent, G. A. Armitage, C. M. Cobb, D. Fine, M. E. Fritz, E. Green, M. J. Imoberdorf, W. J. Killoy, C. Mendieta, R. Niederman, S. Offenbacher, E. J. Taggert, and M. Tonetti, Multicenter evaluation of tetracycline fiber therapy. II. Clinical response. J. Periodont. Res. 26, 371 – 379 (1991). 246. L. M. Golub, S. Ciancio, N. S. Ramamurthy, M. Leung, and T. F. McNamara, Low-dose doxycycline therapy: Effect on gingival and crevicular fluid collagenase activity in humans. J. Periodont. Res. 25, 321 – 330 (1990). 247. M. S. Reddy, T. W. I. Weatherford, C. A. Smith, B. D. West, M. K. Jeffcoat, and T. M. Jacks, Alendronate treatment of naturally occurring periodontitis in beagles. J. Periodontal. 66, 211 – 217 (1995). 248. M. A. Brunsvold, E. S. Chaves, K. S. Kornman, T. B. Aufdemorte, and R. Wood, Effects of a biphosphonate on experimental periodontitis in monkeys. J. Periodontol. 63, 825 – 830 (1992). 249. M. K. Jeffcoat and M. S. Reddy, Alveolar bone loss and osteoporosis: evidence for a common mode of therapy using the bisphosphonate alendronate. In “Biological Mechanisms of Tooth Movement and Craniofacial Adaptation” (Z. Davidovitch and L. A. Norton, eds.), pp. 365 – 373. Harvard Society for the Advancement of Orthodontics, Boston, MA, 1996.

CHAPTER 56

Localized Osteoporosis D. J. SCHURMAN

Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305

W. J. MALONEY R. L. SMITH

I. II. III. IV. V. VI. VII. VIII. IX.

Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis, Missouri 63110 Orthopedic Research Laboratory, Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305; and Veterans Affairs Medical Center, Palo Alto, California 94304

Introduction Mechanical Stress and Normal Bone Architecture Localized Osteoporosis: Generalizations Localized Osteoporosis and Fracture Internal Fixation Devices Inflammatory Disease-Associated Localized Osteoporosis Reflex Sympathetic Dystrophy Transient Osteoporosis of the Hip Total Hip Replacement

X. Prosthetic Design XI. Immobilization Osteoporosis: Animal Studies of Localized Bone Loss XII. Aging XIII. Hormones and Drugs XIV. Local Cellular Mechanisms XV. Summary References

I. INTRODUCTION

mechanical loading experience at a specific bone or bone site [1]. The particular loading history of a chain of bones, a single bone, or a portion of a bone is a critical factor in the alteration or maintenance of bone density. Mechanical stress applied to bone is not itself a biologic process but a trigger or effector that stimulates biologic processes to allow bone to respond or adapt to applied loads. Since the time of the Wolff’s law postulate [2] it has been accepted that mechanical stress is a principal factor in determining changes in local bone geometry (see Chapter 17). These changes include shape, size, and density. A bone that is not mechanically stressed loses its density and becomes osteoporotic. Such an event might occur, for example, in an extremity that is painful or immobilized. The painful or immobilized bone will become less dense and rapidly so. With

The organization and general structure of bone and joint formation are probably the result of embryonic processes dependent on developmental genes (see Chapter 5). Vertebrate homeotic genes are similar to the Drosophila hedgehog proteins that provide position information in the developing embryo as well as in the patterning of cells. The anatomic patterning of specific bones and their position in the skeleton are likely to be explained in the coming years. Once the specific bones and joints are organized according to a predetermined genetic and embryonic plan, different processes and modulators direct specific bone growth and architectural change of individual bones. Hypertrophy, atrophy, and density are influenced by muscular activity and

OSTEOPOROSIS, SECOND EDITION VOLUME 2

385

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

386 the reinstitution of normal function and reapplication of normal intermittent stress loading, such bones will recover their density. Clinical observations of this sort have been the stimulus and foundation for developing phenomenologic theory that can be tested, for example, by computer modeling [3 – 5]. The theory used is responsible for the computer simulation results. Verification of the theory is confirmed by the extent to which it directs changes in the computer-simulated bone to mimic known anatomic results that occur under equivalent in vivo circumstances. In the past few decades finite element analysis (FEM) has been used to great advantage in developing and validating theory so it can then be utilized to predict experimental results (see also Chapter 17). The term “localized osteoporosis” is used here in a straightforward fashion. Localized osteoporosis in the context used indicates osteoporosis affecting either part of one bone or several bones in a chain. Osteoporosis is also used interchangeably with decreased bone density or osteopenia if there is little likelihood of osteomalacia. Localized osteoporosis could be explained in terms of many different methodologies or disciplines. Aspects of bone loss could deal with histology or bone histomorphometry. It could be explained in terms of changes in blood flow or metabolic changes such as the rate of bone accretion or loss. Osteoblasts, osteoclasts, growth factors, cytokines, morphogens, and active peptides all participate in the local osteoporotic process. Most occurrences of localized osteoporosis are related to a well-characterized clinical event. Many of these clinical situations involve a decrease in mechanical stress to the affected bone. This observation was defined heuristically in the 19th century by Wolff, who predicted that increased stress to a bone would lead to hypertrophy and decreased stress would lead to its atrophy. Atrophy or hypertrophy of a bone finally and always depends on biologic transformations. The method by which mechanical stress alters bone mass, therefore, must take place through a putative mechanoreceptor(s). While the mechanoreceptor has not yet been isolated, its existence is self-evident. A dead bone subjected to any sort of stress never hypertrophies or atrophies. If intermittent stresses are applied to a region of the musculoskeletal system, the bones in that region will hypertrophy and likewise if stress is reduced the bones will atrophy. Skeletal ossification to a very important degree depends on maintenance of cyclic loading of the skeletal system (see Chapter 17). The dependency of bone maintenance on a recurring mechanical load is illustrated by Urist’s well-known experiment in which an injection of bone morphogenic protein extract into a mouse muscle thigh led to extensive ossification of this muscle [6]. Less appreciated is the fact that because this ossified muscle is not functional and was not subject to recurrent mechanical load, with time the bone in the muscle disappeared.

SCHURMAN, MALONEY, AND SMITH

II. MECHANICAL STRESS AND NORMAL BONE ARCHITECTURE The structure of a particular bone in the musculoskeletal system is directly related to its function. Few examples of the biologic relationship between structure and function are as clearly visualized as in a radiograph of a long bone. The direction, density, and thickness of the cortical and trabecular bone mirror the direction and magnitude of local bone stress. At the ends of bone, stress caused by axial forces is much greater than forces caused by bending and torsion. A structure of uniform density most efficiently receives an axial loading. Axial stresses at a joint are multidirectional, unpredictable, and concentrated at different sites from one moment to the next. Enlargement of the bone ends protects the structure by maintaining the force per unit area in a safe range. This explains rather simply why the ends of long bones are knobby, examples being knees and knuckles. An enlarged diameter at the bone end provides muscles with large moment arms, reducing forces generated at the joints. The knobby area, i.e., the epiphysis, also contains the most uniformly dense trabeculae. These trabeculae handle stress that the bone receives from a wide range of directions transferred from the opposite joint surface. Stress travels through the metaphysis and diaphysis with greater uniformity of direction compared to the epiphysis. Forces are transmitted across the diaphysis efficiently in a smaller, lighter tubular structure with its hollowed chambers filled by marrow and its outer shell concentrating the forces in its dense thick cortical bone. The appearance of trabecular and cortical anatomy on a long bone radiograph depict the average direction and stress applied to the bone. The size and orientation of a trabeculum reflects the magnitude and direction of the typical stress encountered at the site. Proximal load bearing trabeculae stream into the cortical bone as its presence emerges. In the proximal epiphysis, at first, there is no visual cortex on X-ray. The cortex begins at the confluence of the more proximal trabecular endings. Trabecular bone continues to flow into the cortical bone with a progressive increase in the thickness of the cortical bone moving from epiphysis through the metaphysis as the cortex reaches its greatest uniform thickness until it reaches the metaphysis at the other end of the long bone, where the anatomy reverses itself. Relating structure to function improves understanding. Early in the embryonic stage, bone is composed of undifferentiated mesenchymal cells that develop into anlagen. By the time joints appear as cavitations, these discontinuous anlagen are fully cartilaginous and the basic shapes and positions of the bones as well as the muscles, tendons, ligaments, and their attachments are established. This process is fully evident by 8 weeks’ gestation in the human.

CHAPTER 56 Localized Osteoporosis

387

The process by which the cartilage anlagen are replaced by bone and the sequence of bone differentiation of this anlagen is unusually well predicted by Carter et al. [7]. This theory hypothesizes that intermittent strain energy accelerates cartilage degeneration and leads to ossification. Compressive hydrostatic stresses inhibit cartilage degeneration and therefore preserve it. A typical type of strain energy would be that of shear. Hyaline cartilage has the capacity for proliferation, degeneration, and ossification. Hyaline cartilage degeneration is accelerated by intermittent strain energy, whereas it is inhibited by cyclic hydrostatic pressure. For bone, the apparent density and orientation are regulated by the transfer of mechanical energy. Theories governing the relation of applied load to bone can be powerful predictors of local changes in bone morphology and density [8,9]. An example of this bone forecasting ability is illustrative. Using the general principals of how stress prophesizes bone formation and regulation found in Carter et al. [10], the specific details of femoral bone formation can be elucidated. Finite element modeling of femoral anlagen, composed initially only of dense cartilaginous cells, can be taken through its embryologic development to mature bone. Starting with an appropriate femoral shape, simply loading this solid cartilaginous object results in iterative changes that stepwise reproduce the normal embryologic developmental sequence (Fig. 1). To do this, a combination of loads applied to the femur as would ordinarily occur are required. FIGURE 2 The result after loading a finite element model (FEM) of a completely cartilaginous solid anlage. The strain energy density levels depicted indicate the development of mid-diaphyseal bone formation confined to the periphery, leaving an intramedullary cavity.

FIGURE 1 Human histology of a femur anlage at 8 weeks with progressive ossification at 11 and 35 weeks of gestation. The numbers in millimeters reflect the length of the different bones.

The first change occurs at the center of the diaphysis where the stress results in a replacement of the cartilaginous material with a mid-diaphyseal collar of bone, leaving a hollow intramedullary shell (Fig. 2). By recopying this model in its predicted new structure and again loading the FEM anlage a second iteration is accomplished. A progression in the formation of the diaphysis toward either bone end of the anlage takes place just as one sees in normal embryologic development of bone (Fig. 3). The third iteration establishes a bony metaphysis (Fig. 4) but the fourth iteration does not predict further bone development in the outer shell of the anlage. Instead, bone is predicted to form in the center of the anlage head (Fig. 5). The model thus predicts the site of the secondary ossific nucleus, which normally forms in an infant at 3 months of age. The site of the proximal femoral growth plate is predicted also. The next iteration of the model predicts further ossification of the femoral head, leaving less dense sites where vessels are known to penetrate the femoral head. This step also defines the articular cartilage surface, which is the persistence of the cartilaginous anlage at the surface. The cartilage is preserved because the stress at this site is predominantly hydrostatic.

388

SCHURMAN, MALONEY, AND SMITH

injury, or disease. Trabecular bone loss occurs measurably within days and can become important in just 1 week. Cortical bone loss is much more restricted. A decrease in cortical bone density of significant magnitude takes months to a couple of years to stabilize even in response to a permanent change in loading. For example, with an immobilized limb, cortical bone mass will change little even after a couple of months. In the case of total joint replacement or plate fixation to a bone for fracture fixation, cortical remodeling generally takes 2 years to develop a complete picture of changes on X-ray.

IV. LOCALIZED OSTEOPOROSIS AND FRACTURE Following a bone fracture two paradoxical events occur simultaneously. The body part at the site of the fracture is not used, muscle function markedly decreases, and, therefore, the intermittent stress normally experienced by that bone is significantly reduced. This would

FIGURE 3

After the second FEM iteration, the ossification of the anlage progresses toward the metaphysis.

The FEM simulation correctly predicts thickened articular cartilage at the center of the convex surface of the femoral head. Articular cartilage found toward the periphery of the femoral head becomes progressively thinner. Bone remodeling theories carried forward by Carter, Beaupré, Huskies [10 – 12], and others have been predictive of numerous specific clinical situations which result in bone changes in a highly specific way. Localized osteoporosis of a specific region of a bone or a whole bone or set of bones is most commonly seen under circumstances of “stress shielding.” The most common conditions under which this occurs are summarized in Table 1.

III. LOCALIZED OSTEOPOROSIS: GENERALIZATIONS When localized osteoporosis occurs it does so more quickly at the site of trabecular bone than cortical bone. This will be a recurring theme in all specific conditions of localized bone loss whether due to immobilization,

FIGURE 4 Following the third FEM iteration, ossification advances into the metaphysis and epiphyseal region, depicting bone formation starting to take place in sites other than the outer cortical margins.

389

CHAPTER 56 Localized Osteoporosis

(Fig. 6). However, the fracture repair process unleashes a cascade of bone-stimulating proteins and events resulting in a very localized and powerful bone formation at the fracture site (see also Chapter 12). Callus formation and bone maintenance are regulated differently. This independence may appear to be counterintuitive, but consider that even severely porotic paralyzed limbs, when fractured, have no special problem healing. Displaced fractures commonly heal with a large callus of woven bone exceeding in quantity the amount of original bone at the site. The stimulus for this enlarged callus to persist is withdrawn once the callus bridges the two fractured bone ends. Often much of the original callus is redundant. The return of normal function to the fractured bone with the completion of healing callus returns normal intermittent stress loading to the bone. The transmission of this stress across the fracture site takes place in a direct path that leaves much of the originally formed callus relatively unstressed. With time the unloaded portion of the callus resorbs. Restoration of normal bone density begins after callus bridges the fracture gap and osteoporotic bone begins a gradual return to its prefracture density. After a few years the healed bone will look similar to the prefractured bone.

V. INTERNAL FIXATION DEVICES Figure 5

After the fourth FEM iteration, the secondary ossific nucleus in the femoral head and the site of the proximal femoral physeal plate are indicated.

predict general bone loss and localized osteoporosis to that bone or set of bones. In fact, this does occur. Trabecular bone loss precedes cortical loss and proceeds at a faster pace. Nevertheless, formation of callus and bone evolves rapidly at the fracture site as a healing response. These paradoxical responses result from two different control processes. The fracture leads to a loss of normal stimulus to bone maintenance with resultant atrophy and localized osteoporosis TABLE 1 Mechanical Causes of Localized Osteoporosis 1. Bone fracture 2. Immobilization and paralysis 3. Painful clinical diseases a. Reflex sympathetic dystrophy b. Infection c. Inflammation, e.g., rheumatoid arthritis 4. Total joint replacement 5. Internal fixation devices, e.g., plates and screws

Internal fixation devices used at the time of surgery to hold broken bones together such as screws, plates, and intramedullary nails become part of the composite structure of the bones into which they are placed. The fixation devices reconstitute bone anatomy allowing applied loads to be passed through the bone while maintaining stability. The fixation devices stabilize and allow the fractured bone ends to be held in reasonable proximity to one another, thus making bone repair easier. Typically from the time fracture fixation devices are inserted they participate in sharing some of the mechanical load normally transmitted through the intact bone. If the fracture is reconstituted with good alignment and stabilized by the adjunct fixation device, the applied forces will be satisfactorily transmitted from one side of the fractured bone to the other. The internal fixation device transmits stress extraosseously, diverting intermittent stress from the bone [13,14]. The fractured bone is therefore stress shielded. The stress shielding leads to atrophy and osteoporosis of the healing bone. The use of intramedullary nail fixation for long bone fractures emphasizes stabilization of alignment with less load sharing than when the same fracture is fixed by a plate. Intramedullary rods allow the bone to carry its normal axial load and less stiffly hold the bone.

390

SCHURMAN, MALONEY, AND SMITH

FIGURE 6

A segmental fracture of the tibia with marked osteoporosis (left). (Middle) Callus formation at the fracture sites despite marked osteoporosis of bone. (Right) Less than ideal intramedullary fixation in a tibial fracture with early osteoporosis. Reprinted with permission from Watson-Jones Fractures and Joint Injuries, (J. N. Wilson, Ed.), pp. 455. Churchill Livingstone, Edinburgh, 1976.

Since there is less load sharing by the fixation device there is less tendency for either trabecular or cortical atrophy. Plate fixation, on the other hand, typically participates in intimate load sharing because of its stiff fixation, resulting in weaker, less dense, and more osteoporotic bone. Unnecessarily large or stiff fixation devices increase the amount of stress shielding to the fractured bone. The greater the load sharing by the fixation device, the more atrophy there is of the intact bone at the site of the fixation device. Even when the fracture site heals, bones that share loads with the fixation devices will not return to normal strength. When fracture fixation devices are removed the bone next to these devices, being weak, is predisposed to fracture at much lower applied forces than normal. Screw holes are stress risers that concentrate applied stress. Removing the screws and leaving a screw hole predisposes the weak bone to fracture.

Stiff fixation devices steal more load from the bone and lead to a greater degree of “localized osteoporosis” at the site of fixation. Therefore, there is a premium on using stabilizing devices with decreased stiffness to maintain the strength of the fractured bone. Unfortunately, to achieve adequate device strength usually means a larger sized, and therefore stiffer, device. If one uses a stabilization device that is not strong enough, the penalty for the device breaking before the fracture heals is severe. There are many ways to fix a bone that is fractured. When a long bone is broken the clinician may have a choice between an intramedullary rod, a single plate, or a double plate. The stability of the fixation increases correspondingly but the resultant atrophy of the adjacent bone also increases. The proper treatment of a particular fracture, therefore, is highly dependent on physician judgment.

391

CHAPTER 56 Localized Osteoporosis

VI. INFLAMMATORY DISEASEASSOCIATED LOCALIZED OSTEOPOROSIS Localized osteoporosis occurs adjacent to sites of infectious or inflammatory arthritis, regardless of type. The first X-ray sign of rheumatoid arthritis is localized osteoporosis near the inflamed joint. This precedes all other roentgenographic abnormalities including the juxtaarticular erosion of the radial side of the first metacarpal, another early and diagnostic sign of rheumatoid arthritis. The first roentgenographic abnormality of an infected joint is also localized osteoporosis. The clinical course of a pyogenic joint includes progressive roentgenographic changes secondary to the rapid progressive destruction of joints. These changes include articular erosions, loss of articular cartilage, gross destruction of bone ends, and frank osteomyelitis. With more indolent organisms, such as tuberculosis, osteopenia on X-ray can precede more destructive events by many months. Inflammatory arthritis, such as rheumatoid or psoriatic arthritis, may proceed with time to very destructive events such as major bone loss or joint arthrodesis. The rate of bone destruction from inflammatory arthritis is typically slower to materialize than the progressive, destructive events of a tuberculosis infection. The localized osteoporosis of inflammatory or infectious arthritis is for the most part due to the same cascades of biologic mediators and cellular and biochemical processes. The bacteria causing sepsis may contribute specific mediators that initiate osteoporosis such as a staphylococcus factor [15]. Inflammatory events bring to the local site prostaglandin E2 and all the necessary osteoclast activation triggers such as interleukin-1, tumor necrosis factor, and cytokines. In addition, recruitment of osteoclasts by granulocyte-macrophage colony-stimulating factor and interleukin-6 is enhanced by the inflammatory process. Infection or inflammatory arthritis is painful and site specific. Pain limits local function. Less use brings less applied loads to bone, leading to local disuse osteoporosis.

VII. REFLEX SYMPATHETIC DYSTROPHY Reflex sympathetic dystrophy (RSD) most frequently represents a local response to traumatic injury that is characterized by multiple clinical features, of which prolonged, severe, and intense pain is always prominent. Osteoporosis limited to the painful region of the affected limb is associated with vasomotor instability, swelling of soft tissue, tenderness, and diminished motor function of the involved part. Patchy osteoporosis is the primary roentgenologic sign of the condition. The appearance of osteoporosis evolves from the initial spotty or patchy changes to a more diffuse

involvement by 3 months. At 6 to 9 months, the bony cortices are thinned and the osteoporosis is more homogeneous [16]. The most common explanation for RSD is that the initial traumatic injury leads to vasomotor reflex spasm and then progressively to a condition of persistent vasodilatation. Bone resorption is associated with these progressive changes [17]. Magnetic resonance imaging (MRI) was performed on 17 consecutive patients with early signs of RSD. The patients’ clinical courses were followed to establish diagnosis. The MR findings were completely normal in 10 and in the other 7 patients the findings were nonspecific [18]. Blockade of the sympathetic nervous system is the most important form of therapy and also the most likely to bring complete relief, though not all patients respond to treatment. The rapidity and progressive nature of the osteoporosis has been tied to the vasodilatation associated with this sympathetic abnormality. Historically, RSD was called by a number of different names. The first characterization was described during the Civil War as Causalgia [19], a name adopted to indicate that the condition was associated with a burning pain. After the introduction of X-rays, Sudeck [20] described the same condition in civilians. The presence of osteoporosis in his finding enlarged the description and added an air of objectivity. In the mid-20th century distinctions were made in diagnosis according to the inciting injury using the terms major and minor causalgia. Shoulder – hand syndrome was used to group patients where the condition of RSD existed in a particular anatomic location associated with minor injuries. The extent of osteoporosis is directly correlated to clinical symptomatology. If the condition improves, the osteoporosis improves. So long as the symptoms remain so does the osteoporosis. The presumption is that the osteoporosis is a result and not a cause of the other problems associated with RSD.

VIII. TRANSIENT OSTEOPOROSIS OF THE HIP This affliction is relatively rare, but many independent groups of patients have been reported. Transient osteoporosis of the hip (TOH) occurs in pregnant woman in the third trimester and middle-aged men (see also Chapter 53). The course is benign with full recovery after some months. The clinical syndrome includes the onset of significant pain in the hip, limitation of motion, and limping. The plain radiographs demonstrate osteopenia. Bone scans with technetium diphosphonate are positive. Rest of the hip and pain medication result in recovery over a period of several months. There are no known sequelae to TOH. There are,

392 however, rare complications including compression fracture of the femoral neck. One feature readily distinguishes TOH from osteonecrosis of the hip. Osteoporosis involves the head and the neck of the femur in TOH, but is limited to the femoral head in patients with osteonecrosis. Groups of patients followed throughout their disease course have been studied with MR and bone biopsy. In these situations one study concluded that 8 of 10 patients had some bone necrosis similar to patients with osteonecrosis, but with no evidence of “osteoporosis” [21]. A similar study indicated no osteonecrosis but reported instances of fat necrosis. Other findings included an increase in bone resorption and reactive bone formation in biopsy specimens [22]. Both studies demonstrated bone marrow edema with complete resolution by 6 to 8 months from the time of onset. Focal osteopenia is radiographically evident within 8 weeks following onset of pain. Along with all other signs and symptoms, the osteoporosis resolves completely. There is no joint space narrowing or subsequent arthritis in any of the reported patients. A related benign and self-limiting condition known as transient regional osteoporosis involves other joints and is less common. One form of this condition labeled migrating transient regional osteoporosis travels to different large joints only to resolve eventually and completely. Thus far, all conditions of transient osteoporosis are idiopathic, cultures and full laboratory analyses having failed to indicate specific etiologies. The condition of transient osteoporosis is occasionally listed with reflex sympathetic dystrophy, but their course, treatment, and outcomes are so different that these conditions may have no causal relationship.

IX. TOTAL HIP REPLACEMENT The skeletal changes that occur in the proximal femur following total hip replacement surgery are an excellent example of disuse or mechanically induced osteoporosis that can occur in an otherwise normal skeleton. It is widely accepted that mechanical loading environment influences the balance between bone formation and resorption [23 – 26]. The stresses and strains that occur within a given bone depend on the shape and internal organization of that bone and the external loads applied to it. In the normal hip joint, load is transmitted from the femoral head through the femoral neck to the cortical bone of the proximal femur. Following total hip arthroplasty, load transmission to the proximal femur is markedly altered. Load normally carried by the bone alone is now shared by the implant and bone. As a result of load sharing, the stresses and strains in the proximal femur are reduced. Laboratory experiments using strain gauges designed to evaluate the mechanical alter-

SCHURMAN, MALONEY, AND SMITH

ations in the proximal femur after hip replacement have demonstrated that implantation of both cemented and cementless femoral components results in a marked reduction of strain in the proximal medial femoral cortex (called the calcar) [27 – 29]. On average, the strain measured in the calcar region after insertion of a cemented femoral component is only 15% of that measured in the intact femur prior to implantation. Similar results have been reported after insertion of extensively porous coated implants. There exists a gradient of strain reduction from proximal aspect of the femur to the level of the stem tip. The closer to the tip of the implant, the less the magnitude of the strain reduction. These findings confirm predictions made by finite element studies [30 – 32]. The reduction in stress and strain drives the adaptive remodeling process and results in bone resorption. This is evident on clinical radiographs that demonstrate periprosthetic osteopenia most easily seen in the calcar region (Figs. 7a and 7b). This phenomenon has been widely referred as “stress shielding” and is somewhat of a misnomer. As a result of load sharing, adaptive bone remodeling theory predicts that resorptive remodeling would continue following implantation of a femoral component until there was a normalization of cortical strain patterns. However, analyses of autopsy specimens have not confirmed this prediction. In a study of cadavaric femurs in patients who had previously undergone cemented femoral replacement, the strain patterns along the medial femoral cortex were analyzed [28]. Immediately following implantation of a cemented femoral component, the strain measured in the calcar region decreases by 90%. With the adaptive remodeling that occurs in vivo, the strain in the calcar region increases to approximately 40% of preimplant values. This process occurs relatively rapidly in the first 1 to 2 years after surgery. Although a new “strain equilibrium” appears to be reached, the strain patterns do not return to normal by 17 years after surgery. In a similar autopsy study, analysis of the strain patterns around extensively porous-coated boneingrown stems up to 7.5 years after surgery failed to show that cortical strains increased significantly despite marked in vivo remodeling [33]. This information is difficult to interpret with simple theories. It is also interesting to note that despite the fact that the calcar strains did not increase significantly with mechanically induced, resorptive remodeling, the bone in the calcar region did not completely disappear. In examining the bony remodeling that occurs clinically after total hip arthroplasty, it is important to understand the normal age-related changes that occur in the proximal femur [34 – 37]. During aging, the diameter of the femoral canal enlarges particularly after age 45, especially in females. These changes are greater in the diaphysis than in the metaphysis. In a radiographic study, Smith and Walker [36] reported that the femoral diaphyseal diameter

393

CHAPTER 56 Localized Osteoporosis

FIGURE 7 (a) Early postoperative radiography following a cementless total hip arthroplasty. (b) Radiograph 8 years following surgery. Note the marked osteopenia and thinning of the medial and lateral femoral cortex. Also note the hypertrophy of bone distally reflecting the region of maximal load transfer from stem to bone.

increased an average of 4.4 mm in women between the ages of 45 and 90. Trotter and Peterson [37] performed direct measurements using archeological specimens and compared them to radiographic measurements. The changes that they noted were similar to those reported by Smith and Walker, but on direct measurement the magnitudes of the changes were less. In addition to canal widening, cortical bone tends to become thinner and more porotic with advancing age. A variety of techniques have been used to evaluate femoral bone remodeling after hip replacement surgery. These include clinical radiographs [38 – 41] dual-energy X-ray absorptiometry [42 – 45], and direct examination of autopsy specimens [27,28,46 – 49]. Based on radiographic studies, several authors suggested that skeletal remodeling after total hip arthroplasty (THA) may be associated with aseptic loosening of the implants. Hoffman et al. [39] compared 30 patients with aseptic loosening of a cemented femoral component to a group of matched controls. In those patients with aseptic loosening, the femoral canal expanded at a rate four times that of the control group and they suggested that medullary enlargement

might play a role in failure. Morsher and Ittenson [40] found that canal expansion occurs at an accelerated rate for the first 2 years after THA and also thought that this may play a role in the loosening process. In another radiographic study, Comadoll et al. [38] examined the radiographs on 26 cemented THAs an average of 10.4 years after surgery. They noted a significant decrease in the cortical thickness and an increase in canal diameter for both men and women and concluded that this may cause a separation between cement and bone over time. In contrast, Poss et al. [41] measured a mean rate of cortical expansion of 0.33 mm/year and average loss of cortical bone of 0.15 mm/year at a mean of 11.5 years after cemented THA, which is similar to what is observed in normal aging. They were unable to identify any consistent patterns when comparing the implanted femur to the contralateral intact femur in a small subgroup of patients. Autopsy studies have not supported the hypothesis that mechanically induced osteoporosis and canal widening associated with normal age-related bone remodeling play a role in loss of implant fixation [27,46,48,49].

394

SCHURMAN, MALONEY, AND SMITH

FIGURE 8 Cross-sectional radiographs through the diaphysis from specimens retrieved at autopsy 6 years following a unilateral cemented total hip replacement. The specimen on the left is from the femur that had a femoral replacement 6 years prior to death. On the right, the intact femur was implanted with a cemented femoral component in the laboratory for comparison. Note the decrease in cortical bone area in the in vivo remodeled specimen on the left as well as the preservation of bone at the cement – bone interface (inner cortex) and the beginnings of the formation of a second medullary canal between the inner and the outer cortexes. Although valuable information has been obtained from radiographic reviews (Figs. 8 and 9), there are several inherent difficulties in using clinical X-rays to assess architectural change in the proximal femur [50]. Radiographs are not sensitive enough to detect subtle changes in cortical density. It has been estimated that there has to be at least a 30% loss in bone mineral content to be reliably detected on clinical X-ray. Minor rotation of the femur leads to significant differences in the radiographic measurement of canal diameter and cortical thickness. Finally, once the loosening process begins, associated

bone resorption could lead to an overestimation of remodeling changes attributed to adaptive remodeling.

X. PROSTHETIC DESIGN Variables related to prosthetic design such as stem material and stiffness, extent of porous coating, and the presence of a collar have been proposed as being important in terms of the degree of disuse osteoporosis that occurs following hip replacement surgery [31,33,42,48,49,51 – 59]. The two

FIGURE 9 Cross-sectional radiographs through the diaphysis from specimens retrieved at autopsy 6 years following a unilateral noncemented total hip replacement. The specimen on the left is from the femur that had a femoral replacement 6 years prior to death. On the right, the intact femur was implanted with an identical porous-coated femoral component in the laboratory for comparison. Note the marked loss of cortical bone in the remodeled femur on the left compared to the intact femur on the right.

395

CHAPTER 56 Localized Osteoporosis

most commonly used implant materials in femoral components are chrome-cobalt and titanium alloy. The Young’s modulus of elasticity of titanium alloy is approximately half that of chrome – cobalt. Theoretical models such as finite element analysis predict improved load-sharing with titanium [54,58,60]. Animal studies have also been performed to compare the results of different materials on bone remodeling [52,53,59]. These studies suggest that titanium stems result in better bone preservation when compared to similarly designed stems made of cobalt – chrome. In one canine study, bilateral THAs were performed in eight animals [52]. On one side, a cobalt – chrome alloy implant was used. A hollowed titanium alloy stem, three to five times less stiff than the cobalt – chrome stem, was implanted in the contralateral femur. On average, the femora with the hollowed stems retained 30% more bone compared to the femora with the stiffer stems. The influence of stem size, which is directly proportional to stem stiffness, and the extent of porous coating was examined qualitatively by Engh and Bobyn who reviewed 411 cases of primary cementless THA at 2 years following surgery [55]. Pronounced resorption was defined as radiographic evidence of bone loss occurring at 5 of 16 sites on the anterior – posterior and lateral radiographs. Overall, pronounced resorption was noted in 18% of the cases. As a group, stem sizes  13.5 mm in diameter showed five times the incidence of pronounced resorption. Extensively (two-thirds or fully) porous coated stems demonstrated a two- to fourfold increase in pronounced resorption compared to stems with porous coating limited to the proximal one-third of the implant. The effect of the extent of porous coating has also been studied using a canine model. Turner et al. noted that the most severe loss of cortical bone occurred with extensively, circumferentially porous-coated stems [59]. They concluded that restriction of porous coating would limit the severity of adaptive remodeling. Some controversy exists as to whether or not a collar on a femoral component is capable of transferring load to the calcar region of the femur. Most collars are flat and are designed to sit on the cut surface of the femoral neck after planing the femoral neck with a calcar rasp. In vitro studies support the concept that it is possible to transmit at least some load through the collar to the medial femoral cortex [29]. In contrast, a canine study evaluated the results of porous-coated femoral components with those of a porous-coated collarless component inserted without cement [61]. At 12 weeks, the collarless components were associated with significantly less cortical porosity. In addition, there was no difference in axial bone strains acutely or after 12 weeks in vivo regardless of whether or not a collar was present. Fisher et al. have designed a conical collar intended to load the proximal femur more effectively [61a].

XI. IMMOBILIZATION OSTEOPOROSIS: ANIMAL STUDIES OF LOCALIZED BONE LOSS Much of our understanding of localized bone loss is derived from studies of animal models following short- or long-term periods of immobilization [62 – 66]. These studies addressed questions pertaining to the rate and extent of bone loss, changes in bony architecture, and potential for bone replacement through reparative processes. Consequences of limb immobilization have been evaluated concurrently with differences in age [63,64,67,68], hormonal status [62], and dietary factors [69,66]. In addition, modulation of localized bone loss by pharmacologic intervention has been tested by treatment with selected compounds during or after immobilization [65,67,70 – 73]. In general, animal models provide a dynamic system for assessment of bone loss using analytic techniques applicable to human disease. In the predominant experimental approach, a single hind limb is immobilized, leaving the contralateral limb to support unrestricted movement of the animal. In this model, the immobilized limb experiences reduced loading, whereas increased loads may occur on the nonimmobilized limb. Techniques for assessing bone loss have included chemical procedures such as determination of wet weight, dry weight, ash weight, and calcium content [74]. Nondestructive methods have included the use of single photon absorptiometry to quantify bone mineral density [64] and histomorphometric analysis of bone remodeling using bone-specific stains [74]. Biochemical methods have included analysis of bone collagen content, quantification of bone degradation products in the serum or urine [75], uptake of radioactive precursors into bone [76], and analysis of gene expression of bone proteins [77,78]. Results obtained by restricting hind limb loading by immobilization show that changes in bone metabolism in the affected limb occur rapidly and are concentrated within trabecular bone. For example, in adult and growing rats, onset of bone loss following immobilization by either cast or neurectomy takes place within 2 to 3 weeks of limb restriction [79,74]. In growing animals, immobilization-induced loss of bone weight occurs mainly from mineral losses as quantified by changes in wet weight, ash weight, and calcium content. Bone mineral deposition also decreases with immobilization as evidenced by incorporation of 45Ca. Trabecular bone accounts for most of the deficiency, whereas bone volume is not reduced. At 3 weeks, bone ash weights decrease approximately 12% for tibia and femur with immobilization. Bone mineral was partially restored during remobilization following either 1 week or 3 weeks of cast immobilization or neurectomy. Animal experiments have also been used to determine to what extent recovery of bone loss can occur with exercise

396 following periods of immobilization. In studies with retired breeder rats, bone mass and muscle did recover following a period of 6 weeks of cage activity following a period of immobilization of 6 weeks, but not to levels present prior to the period of bone loss [68]. In growing animals, bone mass recovered after removal of the cast following cast immobilization, and exercise but running did not statistically increase bone mass beyond that observed in animals allowed to move freely in their cages [76]. Induction of localized bone loss by immobilization also provides important insight into the mechanisms influencing both bone and cartilage metabolism within the joint. In a rabbit model, short-term immobilization (3 weeks) was used to analyze processes underlying osteoporosis and osteoarthrosis [80]. This study showed that joint surfaces of the immobilized limb were characterized by prominent subchondral vascular eruptions on the lateral tibial plateau and the lateral femoral condyle. The vascular eruptions were accompanied by decreased metaphyseal bone density of 27 and 18% in immobilization limbs of postadolescent and mature rabbits, respectively. The decrease in bone density was associated with increased calcein green fluorescence (1.9-fold) in the metaphyseal trabeculae of the immobilized limb. When bony remodeling was accelerated, cartilage glycosaminoglycan and hydroxyproline were unchanged although uptake of both sulfate and thymidine was increased. Thus, the changes in bone metabolism appeared to occur rapidly following decreased loading of the rabbit hind limb and preceded erosive cartilage degradation. Animal models provide experience and knowledge invaluable to understanding human disease. The ability to manipulate both time of onset and duration of conditions leading to skeletal remodeling replicate similar experiences observed clinically.

XII. AGING A decrease in mineral deposition also depended on the age of the animals. In 2.5-month-old female rats, immobilization of the right hind limb resulted in rapid loss of primary spongiosa at 2 weeks and secondary spongiosa bone at 8 weeks in the distal femoral metaphysis [63]. The negative bone balance could be attributed to a transient increase in bone resorption, decreased bone formation, and a decrease in longitudinal bone growth. By histomorphometric techniques, the secondary spongiosa showed that the eroded perimeter was transiently increased 55 to 82% and the labeled perimeter was decreased approximately 20%. By 20 weeks, the immobilization-induced bone loss equilibrated to a new steady-state level of less bone but with a normal (age-related control) bone turnover rate. Loss of bone following immobilization also occurred in older animals. In 9-month-old female rats, immobilization

SCHURMAN, MALONEY, AND SMITH

of the right hind limb for varying periods resulted in decreased muscle mass and cancellous bone loss in the unloaded limb by 2 weeks [81]. Effects on bone then stabilized at a 50% loss after 18 weeks and were characterized by a decrease in trabecular number and an increase in trabecular separation. In another study with older rats, 6 weeks of immobilization of the right hind limb produced decreased cancellous bone mass and decreased trabecular number in the proximal tibial metaphysis [68]. In the proximal metaphysis, formation was decreased and resorption was higher. Release of the hind limb followed by a 6-week period for recovery increased bone mass but not to the starting levels. Muscle weight also decreased and was not completely restored during the recovery period. Longer time periods may be required for full restoration of bone and muscle.

XIII. HORMONES AND DRUGS The impact of hormonal changes on bone loss following immobilization has been analyzed using animal models as well. The rat immobilization model coupled with ovariectomy provides a functional system for evaluation of reactive compounds targeted to the prevention of osteoporotic bone loss. Such an analysis showed that a 1,25-dihydroxyvitamin D analog could restore bone mineral density as determined by dry weight, ash weight, and ash content [82]. The effect of the analog was to suppress the elevated bone turnover, primarily by decreasing resorption. Similar studies have been used to assess the impact of adding prostaglandins to counteract bone loss [67,70]. Administration of PGE2 at 3 or 6 mg/kg/day for 8 weeks completely restored cancellous bone mass and reestablished bone structure in a rat model of hind limb immobilization. A second 8 weeks of treatment maintained bone mass and architecture. Treatment with PGE2 also increased bone mass in the overloaded limb and continued treatment maintained the increased bone mass at the elevated levels. Also, PGE2 was shown to restore cancellous bone mass and architecture in the proximal tibial metaphysis following continuous immobilization in female rats [83]. The model used right hind limb immobilization for underloading and the contralateral limb was overloaded during ambulation. After 4 weeks of immobilization, cancellous bone loss was decreased and trabecular thickness, number, and node density also decreased until a new steady-state level was established. Other animal models of immobilization have yielded insight into whether pharmacological intervention may prevent loss or reestablish bone mass. Although PGE2 improved bone mass in the example described above, in the setting of inflammatory processes this same agent may underlie an opposite effect. In a dog model, immobilization

397

CHAPTER 56 Localized Osteoporosis

using unilateral hind limb fiberglass cast-fixation resulted in a 50% decrease in bone mass after 4 weeks as determined by single-photon absorptiometry. In vitro release of PGE2 from the calcaneus, tibial cortical bone, tibial cancellous bone, and the ileum was increased twofold relative to that of the controls. Treatment of casted animals with aspirin reduced bone release of PGE2 in vitro by 65% and showed a 13% sparing effect on the decreased bone mass [65]. In another study testing the effects of nonsteroidal anti-inflammatory drug treatment, S-Ketoprofen at a dose of 2.5 mg/kg/day prevented the induced decrease in bone formation and the increase in bone resorption in a tenotomy model [73]. The use of tenotomy to induce immobilization resulted in a 54% decrease in cancellous bone mass in proximal tibial metaphyses of weanling rats. Agerelated bone gain was also inhibited in cortical bone sites using this model. Immobilization of the rat hind limb by casting or neurectomy to reduce bone mass was used to assess effects of hormonal modulation of bone loss. Administration of calcitonin (15 IU/kg given once daily) reduced the extent to which bone ash weight decreased with immobilization in the case of neurectomy but not in the case of casting [79].

XIV. LOCAL CELLULAR MECHANISMS Sustained changes in bone stability ultimately derive from the activity of osteoblasts laying down bone and osteoclasts removing mineral. Recent evidence shows that the relative contributions of osteoblasts and osteoclasts on bone homeostasis depend on two primary effectors, RANK ligand (RANK-L) and macrophage colony-stimulating factor [84]. Increased numbers of osteoclasts result from an upregulation in RANK-L production by osteoblast lineage cells resulting in an increase in loss of bone (see also Chapters 3, 12, and 13). The RANK-L acts by binding the osteoclast differentiation and activation receptor (receptor activator of NFB (RANK) on osteoclast lineage cells [84]. The RANK-L receptor is present on osteoclast precursors bound to membrane so that cell/cell contact is required for activity. Interaction between the RANK-L and its receptor results in osteoclast maturation, increased bone resorption, and decreased osteoclast apoptosis in the presence of macrophage colony-stimulating factor. These latter events shift the balance of bone metabolism so that loss exceeds deposition. Induction of bone loss by RANK-L is countered in part by the presence of a decoy receptor, osteoprotegerin (OPG) [85]. OPG is secreted by osteoblast lineage cells and acts to decrease osteoclastogenesis. In culture, addition of OPG reduces osteoclast differentiation, whereas the addition of RANK-L increases appearance of osteoclasts [86]. OPG

expression declines with age as determined using bone marrow cells in culture [87]. This may be of significance to age related osteoporosis. The release of factors from osteoblastic lineage cells represents a local control mechanism by which bone cells modulate the mineralized matrix under conditions requiring cell-to-cell contact. However, exogenous factors such as proinflammatory cytokines can overcome local control and force bone resorption [88]. There is at least one suggestion that the same can be said for promoting local bone formation [89].

XV. SUMMARY Common threads tie our current understanding of localized osteoporotic conditions together. First, the bone loss under conditions as divergent as paralysis, fracture, reflex sympathetic dystrophy, and rheumatoid arthritis are bound together by a decrease in intermittent loading of the bone. Second, the bone loss in these diverse conditions is reversible. If the fracture heals, the rheumatoid joint goes into remission, the paralysis is eliminated, or a ganglionic blocker abolishes the RSD, the local osteoporosis resolves. The challenge for the future is to understand the control processes at the molecular and genetic level and to reveal the mechanoreceptor that transduces strain energy into biologic events.

References 1. G. S. Beaupre, T. E. Orr, and D. R. Carter, An approach for timedependent bone modeling and remodeling — Theoretical development. J. Orthop. Res. 8(5), 651 – 661 (1990). 2. J. Wolff, “Das gesetz der transformation der knochen.” Ausg. Berlin, Hirschwald, Germany, 1892. 3. D. P. Fyhrie and D. R. Carter, A unifying principle relating stress to trabecular bone morphology. J. Orthop. Res. 4(3), 304 – 317 (1986). 4. D. R. Carter, Mechanical loading history and skeletal biology. J. Biomech. 20(11 – 12), 1095 – 1109 (1987). 5. R. Huiskes and S. J. Hollister, From structure to process, from organ to cell: Recent developments of FE-analysis in orthopedic biomechanics. J. Biomech. Eng. 115(4B), 520 – 527 (1993). 6. M. R. Urist, R. J. DeLange, and G. A. M. Finerman, Bone cell differentiation and growth factors. Science 220, 680 – 686 (1983). 7. D. R. Carter, T. E. Orr, D. P. Fyhrie, and D. J. Schurman, Influences of mechanical stress on prenatal and postnatal skeletal development. Clin. Orthop. Relat. Res. 219, 237 – 250 (1987). 8. M. E. Levenston, G. S. Beaupre, C. R. Jacobs, and D. R. Carter, The role of loading memory in bone adaptation simulations. Bone 15(2), 177 – 186 (1994). 9. M. G. Mullender, R. Huiskes, and H. Weinans, A physiolog-ical approach to the simulation of bone remodeling as a self-organizational control process. J. Biomech. 27(11), 1389 – 1394 (1994). 10. R. Huiskes, Stress shielding and bone resorption in THA: Clinical versus computer-simulation studies. Acta Orthop. Belg. 59 (Suppl. 1), 118 – 129 (1993).

398 11. T. E. Orr, G. S. Beaupre, D. R. Carter, and D. J. Schurman, Computer predictions of bone remodeling around porous-coated implants. J. Arthroplasty 5(3), 191 – 200 (1990). 12. M. C. van der Meulen, G. S. Beaupre, and D. R. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14(4), 635 – 642 (1993). 13. S. L. Woo, K. S. Lothringer, W. H. Akeson, R. D. Coutts, Y. K. Woo, B. R. Simmon, and M. A. Gomez, Less rigid internal fixation plates: Historical perspectives and new concepts. J. Orthop. Res. 1(4), 431 – 449 (1984). 14. G. S. Beaupre, D. R. Carter, T. E. Orr, and J. Csongradi, Stresses in plated long-bones: The role of screw tightness and interface slipping [published erratum appears in J. Orthop. Res. 6(3), 466 (1988)]. J. Orthop. Res. 6(1), 39 – 50 (1988). 15. R. J. Williams III, R. L. Smith, and D. J. Schurman, Septic arthritis: Staphylococcal induction of chondrocyte proteolytic activity. Arthritis Rheum. 233, 533 – 541 (1990). 16. L. L. Lankford and J. E. Thompson, Reflex sympathetic dystrophy, upper and lower extremity: Diagnosis and management. In “Instructional Course Lectures, AAOS,” Vol. 26, pp. 163 – 178. Mosby, St. Louis, MO, 1977. P. D. Inhofe and C. A. Garcia-Moral, “Reflex Sympathetic Dystrophy: “A Review of the Literature and Long Term Outcome Study Orthopaedic Review,” pp. 655 – 661. August, 1993. 17. L. S. Osher, G. J. Young, A. A. Edwards, P. Usewicz, J. E. Lichniak, Reflex sympathetic dystrophy syndrome. J. Mam. Podiatr. Med. SOASS 83(5), 276 – 283 (1993). 18. E. Koch, H. O. Hofer, G. Sialer, B. Marincek, and G. K. von Schulthess, Failure of MR imaging to detect reflex sympathetic dystrophy of the extremities. Am. J. Reontg. 56(1), 113 – 115 (1991). 19. S. W. Mitchell, G. R. Morehouse, and W. W. Keen, “Gunshot Wounds and Other Injuries of Nerves.” Lippincott, Philadelphia, PA, 1864. 20. P. Sudeck, Uber die acute entzundliche Knochenatrophie. Arch. Klin. Chir. 62, 47 – 156 (1990). 21. S. Hoffman, A. Engel, A. Neuhold, K. Leder, J. Kramer, and H. Plenk, Jr, Bone marrow edema syndrome in transient osteoporosis of the hip. An MRI controlled study of treatment by core decompression. J. Bone Jt. Surg. 75(B), 210 – 216 (1993). 22. H. Potter, M. Morn, R. Schneider, M. Bansal, C. Sherman, and J. Markisz, Magnetic resonance imaging in diagnosis of treatment of osteoporosis of the hip. Clin. Orthop. 280, 223 – 229 (1992). 23. A. E. Churches, C. R. Howlett, K. J. Waldron, and G. W. Ward, The response of living bone to controlled time varying loading: Method and preliminary results. J. Biomech. 12, 35 – 45 (1979). 24. H. H. Jones, J. D. Priest, W. C. Hayes, C. C. Tichenor, and D. A. Nagel, Humeral hypertrophy in response to exercise. J. Bone Jt. Surg. 59A, 204 – 208 (1977). 25. A. J. Tonino, C. L. Davidson, P. J. Klopper, and L. A. Linclau, Protection from stress in bone and it’s effects. J. Bone Jt. Surg. 58B, 107 – 112 (1976). 26. AL-Y. Woo, S. S. Kuei, D. Amiel, M. Gomez, W. C. Hayes, F. C. White, and W. H. Akeson, The effect of prolonged physical training on the properties of lone bone: A study of Wolff’s law. J. Bone Jt. Surg. 63A, 780 – 787 (1981). 27. C. A. Engh, D. O. O’Connor, M. Jasty, T. F. McGovern, J. D. Bobyn, and W. H. Harris, Quantification of implant micromotion, strain shielding, and bone resorption with porous-coated anatomic medullary locking femoral prostheses. Clin. Orthop. 285, 13 – 29, 1992. 28. W. J. Maloney, M. Jasty, D. W. Burke, D. O. O’Connor, E. B. Zalenski, C. A. Bragdon, and W. H. Harris, Biochemical and histological evaluation of cemented total hip replacements: A study of autopsy retrieved femurs after in vivo cycling. Clin. Orthop. 249, 129 (1989). 29. I. Oh, and W. H. Harris, Proximal strain distribution in the loaded femur. J. Bone Jt. Surg. 60A, 75 – 85 (1978).

SCHURMAN, MALONEY, AND SMITH 30. T. P. Harrigan and W. H. Harris, A finite element study of the effect of diametral interference gaps on the contact areas and pressures in uncemented cylindrical femoral total hip components. J. Biomech. 24, 87 (1991). 31. R. Huskes, The various stress patterns of press-fit, ingrown, and cemented femoral stems. Clin. Orthop. 261, 27 (1990). 32. R. Huskes, H. Weinans, B. V. Rietbergen, D. R. Sumner, T. M. Turner, and J. O. Galante, “Validation of Strain Adaptive Bone Remodeling Analysis to Predict Bone Morphology around Non-cemented THA.” AAOS 37th Annual Meeting, Anaheim, California, March 4 – 7, 1991. 33. C. A. Engh, J. D. Bobyn, and A. H. Glassman, Porous-coated hip replacement. The factors governing bone ingrowth, stress shielding, and clinical results. J. Bone Jt. Surg. 69B, 45 – 55 (1987). 34. M. F. Erikson, Aging changes in the medullary cavity of the proximal femur in American Blacks and Whites. Am. J. Phys. Anthropol. 51, 563 (1979). 35. C. B. Ruff and W. C. Hayes, Sex differences in age-related remodelling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). 36. R. W. Smith and R. J. Walker, Femoral expansion in aging women: Implication for osteoporosis and fracture. Science 145, 156 (1964). 37. M. Trotter and R. R. Peterson, Transverse diameter of the femur on roentgenograms and on bones. Clin. Orthop. 52, 233 (1967). 38. J. L. Comadoll, R. E. Sherman, R. B. Gustillo, and J. E. Bechtold, Radiographic changes in bone dimensions in asymptomatic cemented total hip arthroplasties. J. Bone Jt. Surg. 70A, 433 – 438 (1988). 39. A. A. Hoffman, R. B. Wyatt, E. P. France, G. T. Bigler, A. U. Daniels, and W. F. Hess, Endosteal bone loss after total hip arthroplasty. Clin. Orthop. 245, 138 – 144 (1989). 40. E. Morscher and F. Ittenson, Knochen-remodelling und lockerung des protheseenschaftes bei huftporthesen. Symposium des Argeitskreises fur Ostelologie Davos 12/13, 4 1984. Acta Med. Aust. 32, 16 (1984). 41. R. Poss, P. Staehlin, and M. Larson, Femoral expansion in total hip arthroplasty. J. Arthroplasty 2, 259 (1987). 42. C. A. Engh, T. F. McGovern, J. D. Bobyn, and W. H. Harris, A quantitative evaluation of periprosthetic bone-remodeling after cementless total hip arthroplasty. J. Bone Jt. Surg. 74A, 1009 – 1020 (1992). 43. S. S. Hughes, J. P. Furia, P. Smith, and V. D. Pelligrini, Atrophy of the proximal part of the femur after total hip arthroplasty without cement. J. Bone Jt. Surg. 77A, 231 – 239 (1995). 44. J. B. Kiratli, J. P. Heiner, A. A. McBeath, and M. A. Wilson, Determination of bone mineral density by dual x-ray absorptiometry in patients with uncemented total hip arthroplasty. J. Orthop. Res. 10, 836 – 844 (1992). 45. T. F. McGovern, C. A. Engh, K. Zettl-Schaffer, and J. P. Hooten, Cortical bone density of the proximal femur following cementless total hip arthroplasty. Clin. Orthop. 306, 145 – 154 (1994). 46. M. Jasty, W. J. Maloney, C. A. Bragdon, T. Haire, and W. H. Harris, Histomorphological studies of the long term skeletal response to well-fixed cemented femoral components. J. Bone Jt. Surg. 72A, 1220 (1990). 47. L. M. Kwong, M. Jasty, R. D. Mulroy, W. J. Maloney, C. A. Bragdon, and W. H. Harris, The histology of the radiolucent line. J. Bone Jt. Surg. 74B, 67 (1992). 48. W. J. Maloney, C. Bragdon, M. Jasty, K. Lee, and J. Lowenstein, “An In-Depth Study of Bone Remodeling following Cemented Femoral Replacement.” AAOS 41st Annual Meeting, Orlando, FL, February 16 – 21, 1995. 49. W. J. Maloney, C. Bragdon, C. Sychterz, T. McGovern, M. Jasty, C. A. Engh, and W. H. Harris, “Adaptive Bone Remodeling with WellFixed Cemented and Cementless Femoral Components.” Transactions of the Society for Biomaterials, 21st Annual Meeting, p. 70, 1995.

CHAPTER 56 Localized Osteoporosis 50. J. D. West, M. B. Mayor, and J. P. Collier, Potential errors inherent in quantitative densitometric analysis of orthopaedic radiographs. J. Bone Jt. Surg. 69A, 58 – 64 (1987). 51. W. H. Harris, E. A. Haywood, J. Lavorgna, and D. L. Hamblen, Spatial and temporal variations in cortical bone formation in dogs. J. Bone Jt. Surg. 50A, 1118 – 1128 (1968). 52. J. D. Bobyn, A. H. Glassman, H. Goto, J. J. Krygier, J. E. Miller, and C. E. Brooks, The effect of stem stiffness on femoral bone resorption after canine porous-coated total hip arthroplasty. Clin. Orthop. 261, 196 – 213 (1990). 53. J. D. Bobyn, E. S. Mortimer, A. H. Glassman, C. A. Engh, J. E. Miller, and C. E. Brooks, Producing and avoid stress shielding. Laboratory and clinical observations of noncemented total hip arthroplasty. Clin. Orthop. 274, 79 – 96 (1992). 54. P. D. Diegel, A. U. Daniels, and H. K. Dunn, Initial effect of collarless stem stiffness on femoral bone strain. J. Arthroplasty 4, 173 – 178 (1989). 55. C. A. Engh and J. D. Bobyn, The influence of stem size and extent of porous coating on femoral bone resorption after primary cementless hip arthroplasty. Clin. Orthop. 231, 7 – 28 (1988). 56. W. H. Harris, Is it advantageous to strengthen the cement-metal interface and use a collar for cemented femoral components of total hip replacements. Clin. Orthop. 285, 67 – 72 (1992). 57. J. L. Lewis, M. J. Askew, R. L. Wixon, G. M. Kramer, and R. R. Tarr, The influence of prosthetic stem stiffness and of a calcar collar on stresses in the proximal end of the femur with a cemented femoral component. J. Bone Jt. Surg. 66A, 280 – 286 (1984). 58. G. L. Maistrelli, V. Fornaiser, A. Binnington, K. McKenzie, V. Sessa, and I. Harrington, Effect of stem modulus in a total hip replacement. J. Bone Jt. Surg. 73B, 43 – 46 (1991). 59. T. M. Turner, D. R. Sumner, R. M. Urban, D. P. River, and J. Galante, A comparative study of porous coatings in a weight-bearing total hip-arthroplasty model. J. Bone Jt. Surg. 68A, 1396 – 1409 (1986). 60. R. R. Tarr, Comparison of lading behavior of femoral stemsof titanium-6AL-4V and cobalt chromium alloy: A three-dimensional finite element analysis. In “Titanium Alloys in Surgical Implants” (H. Luckey, ed.), pp. 88 – 101. American Society for Testing and Materials, Philadelphia, 1983. 61. S. L. Cheng, J. R. Davey, R. D. Inman, A. G. Binnington, and T. J. Smith, The effect of the medial collar in total hip arthroplasty with porous-coated components inserted without cement. J. Bone Jt. Surg. 77A, 118 – 123 (1995). 61a.K. J. Fisher, D. R. Carter, and W. J. Maloney, In vitro study of initial stability of a conical collared femoral component. J. Arthroplasty (Suppl.), 389 (1992). 62. C. M. Bagi, M. Mecham, J. Weiss and S. C. Miller, Comparative morphometric changes in rat cortical bone following ovariectomy and/or immobilization. Bone 14, 877 – 883 (1993). 63. M. M. Chen, W. S. Jee, H. Z. Ke, B. Y. Lin, Q. N. Li, and X. J. Li, Adaptation of cancellous bone to aging and immobilization in growing rats. Anat. Rec. 234, 317 – 334 (1992). 64. X. J. Li and W. S. Jee, Adaptation of diaphyseal structure to aging and decreased mechanical loading in the adult rat: A densitometric and histomorphometric study. Anat. Rec. 229, 291 – 297 (1991). 65. D. J. Waters, D. D. Caywood, G. J. Trachte, R. T. Turner, and S. F. Hodgson, Immobilization increases bone prostaglandin E. Effect of acetylsalicylic acid on disuse osteoporosis studied in dogs. Acta Orthop. Scand. 62, 238 – 243 (1991). 66. M. Yamaguchi, N. Shimokawa, and T. Hoshi, Oral administration of calcium hydroxide stimulates bone metabolism in the femoral diaphysis of rats with skeletal unloading. Chem. Pharm. Bull. 39, 3365 – 3367 (1991).

399 67. W. S. Jee, T. Akamine, H. Z. Ke, X. J. Li, L. Y. Tang, and Q. Q. Zeng, Prostaglandin E2 prevents disuse-induced cortical bone loss. Bone 13, 153 – 159 (1992). 68. H. Maeda, D. B. Kimmel, D. M. Raab, and N. E. Lane, Musculoskeletal recovery following hind limb immobilization in adult female rats. Bone 14, 153 – 159 (1993). 69. M. Weinreb, G. A. Rodan, and D. D. Thompson, Immobilization-related bone loss in the rat is increased by calcium deficiency. Calcif. Tissue Int. 48, 93 – 100 (1991). 70. M. Li, W. S. Jee, H. Z. Ke, X. G. Liang, B. Y. Lin, Y. F. Ma, and R. B. Setterberg, Prostaglandin E2 restores cancellous bone to immobilized limb and adds bone to overloaded limb in right hindlimb immobilization rats. Bone 14, 283 – 288, (1993). 71. T. Shimizu, N. Ishoguro, and T. Miura, The effect of calcitonin on osteoporosis of the rat hind limb induced by denervation and isograft transplantation. J. Reconstr. Microsurg. 8, 41 – 45 (1992). 72. D. D. Thompson, J. G. Seedor, M. Weinreb, S. Rosini, and G. A. Rodan, Aminohydroxybutane bisphosphonate inhibits bone loss due to immobilization in rats. J. Bone Miner. Res. 5, 279 – 286 (1990). 73. Q. Q. Zeng, W. S. Jee, H. Z. Ke, and W. J. Wechter, S-ketoprofen inhibits tenotomy-induced bone loss and dynamics in weanling rats. Bone Miner. 21, 203 – 218 (1993). 74. J. Tuukkanen, B. Wallmark, P. Jalovaara, T. Takala, S. Sjogren, and K. Vaananen, Changes induced in growing rat bone by immobilization and remobilization. Bone 12, 113 – 118 (1991). 75. J. K. Yeh, C. C. Liu, and J. F. Aloia, Effects of exercise and immobilization on bone formation and resorption in young rats. Am. J. Physiol. 264(2 Pt. 1), E182 — E189 (1993). 76. J. Tuukkanen, Z. Peng, and H. K. Vaananen, The effect of training on the recovery from immobilization-induced bone loss in rats. Acta Physiol. Scand. 145, 407 – 411 (1992). 77. G. K. Wakley, J. S. Portwood, and R. T. Turner, Disuse osteopenia is accompanied by down regulation of gene expression for bone proteins in growing rats. Am. J. Physiol. 263(6 Pt. 1), E1029 — E1034 (1992). 78. M. Wong, J. Germiller, J. Bonadio, and S. A. Goldstein, Neuromuscular atrophy alters collagen gene expression, pattern formation, and mechanical integrity of the chick embryo long bone. Prog. Clin. Biol. Res. 383B, 587 – 597 (1993). 79. J. Tuukkanen, P. Jalovaara, and K. Vaananen, Calcitonin treatment of immobilization osteoporosis in rats. Acta Physiol. Scand. 141, 119 – 124 (1991). 80. R. L. Smith, K. D. Thomas, D. J. Schurman, D. R. Carter, M. Wong, and M. C. van der Meulen, Rabbit knee immobilization: Bone remodeling precedes cartilage degradation. J. Orthop. Res. 10, 88 – 95 (1992). 81. X. J. Li, W. S. Jee, S. Y. Chow, and D. M. Woodbury, Adaptation of cancellous bone to aging and immobilization in the rat: A single photon absorptiometry and histomorphometry study. Anat. Rec. 227, 12 – 24 (1990). 82. H. Okumura, T. Yamamuro, S. Higuchi, M. Harada, T. Takamura, S. Otomo, H. Aihara, N. Ikekawa, and T. Kobayashi, 26,27Hexafluoro1,25-dihydroxyvitamin D3 (F6-1,25(OH)2D3) prevents osteoporosis induced by immobilization combined with ovariectomy in the rat. Bone Miner. 9, 101 – 109 (1990). 83. T. Akamine, W. S. Jee, H. Z. Ke, X. J. Li, and B. Y. Lin, Prostaglandin E2 prevents bone loss and adds extra bone to immobilized distal femoral metaphysis in female rats. Bone 13, 11 – 22 (1992). 84. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 85. L. C. Hofbauer, F. Gori, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Stimulation of osteoprotegerin ligand and

400 inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140, 4382 – 4389 (1999). 86. I. Itonaga, A. Sabokbar, D. W. Murray, and N. A. Athanasou, Effect of osteoprotegerin and osteoprotegerin ligand on osteoclast formation by arthroplasty membrane derived macrophages. Ann. Rheum. Dis. 59, 26 – 31 (2000). 87. H. A. Makhluf, S. M. Mueller, S. Mizuno, and J. Glowacki, Age-related decline in osteoprotegerin expression by human bone marrow cells cultured in three-dimensional collagen sponges. Biochem. Biophys. Res. Commun. 24, 669 – 672 (2000).

SCHURMAN, MALONEY, AND SMITH 88. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, T. Suda, Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J. Exp. Med. 17, 275 – 286 (2000). 89. M. Sakata, H. Shiba, H. Komatsuzawa, T. Fujita, K. Ohta, M. Sugai, H. Suginaka, and H. Kurihara, Expression of osteoprotegerin (osteoclastogenesis inhibitory factor) in cultures of human dental mesenchymal cells and epithelial cells. J. Bone. Miner. Res. 14, 1486 – 1492 (1999).

CHAPTER 57

Evaluation of the Patient with Osteoporosis or at Risk for Osteoporosis MICHAEL KLEEREKOPER

Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201

IV. The Evaluation of the Patient with Low BMD V. Summary and Conclusions References

I. Introduction II. The Decision to Measure BMD III. What BMD to Measure

I. INTRODUCTION

II. THE DECISION TO MEASURE BMD

In the past decade the definition of osteoporosis has changed from a disease of fractures to one of fracture risk with the fractures now being the complication of the disease. Accordingly the clinical and diagnostic evaluation of the patient with osteoporosis has changed substantially. This chapter will focus on patients who should be evaluated by measurement of bone density (or ultrasonometry) and what further evaluation is appropriate in patients who have been identified to have abnormal bone density (or ultrasonometry). Elsewhere in this text the several diagnostic modalities for evaluating bone strength non-invasively by radiographic and ultrasound techniques are discussed in detail (see Chapter 58 and 59). For ease of writing and reading, ultrasound techniques will be assumed to be synonymous with a peripheral measurement of bone mineral density (BMD) by radiographic techniques.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Until BMD measurements become so inexpensive that the performance of this test does not generate a bill for medical services (as is now the case for measurement of blood pressure), clinicians need to develop guidelines for the most appropriate application of BMD testing in individual patients. A number of such guidelines have been developed by appropriate scientific and regulatory bodies but none of these is directly applicable to most individual practices [1,2]. As a rule a BMD study should be obtained when the information to be gathered will influence individual patient management decisions. Such an approach works best in circumstances where specific intervention is unlikely to be recommended, no matter the outcome of the study.

403

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

404

MICHAEL KLEEREKOPER

TABLE 1 A. Who should not have a BMD study? •

Healthy children and adolescents

•

Healthy premenopausal women

•

Healthy men 65 years old

B. Who should have a BMD study? •

Premenopausal adult women with clinical evidence of hypoestrogenemia •

Anorexia/bulimia

•

Athletic amenorrhea

•

Prolactinoma

•

Therapy with •

GnRH agonists

•

Depo-Provera

•

All women at menopause who do not elect to begin ERT/HRT for non-skeletal reasons

•

All women over age 65

•

All patients with malabsorption

•

All patients with inflammatory bowel disease

•

All patients about to begin chronic (3 months) corticosteroid therapy

•

All patients already on chronic corticosteroid therapy

•

All men with hypogonadism

•

All patients with an unexplained fragility fracture

•

All patients with primary hyperparathyroidism

•

All patients anticipating organ transplantation

have a negative number for T score, 16% will have T score 1.0, and 2.5% will have T score 2.0. In none of these persons will it be possible to ascertain at the first BMD measurement whether they had ever achieved a higher peak bone mass and consequently whether they have ever sustained any bone loss. True, as pointed out in a recent NIH Consensus Development Conference on osteoporosis [5], there is reason to believe that such individuals are likely to be at increased risk for fracture as they begin to lose bone. However, until that happens there is no available intervention that will increase peak bone mass once attained and therefore no intervention (other than diet and lifestyle modification aimed at preventing bone loss) that should be instituted on the basis of the low BMD. In clinical practice this issue arises when a younger woman becomes concerned about the possibility of osteoporosis because an older female relative has recently sustained an osteoporotic fracture or been diagnosed with osteoporosis on the basis of a BMD study. It might be argued that aside from the cost of the study, BMD testing is safe and harmless. This is true if no emotional distress results from finding a low BMD and is compounded if such a finding results in the patient being placed on specific osteoporosis therapy. Younger women concerned about limited dietary calcium intake should be advised to increase their intake without the need for a BMD test.

C. Who might benefit from a BMD study? •

Patients with recurrent stress fractures

•

Patients with nephrolithiasis

•

Patients on chronic therapy with •

Anticonvulsants

•

Anticoagulants

•

Thyroxine

•

Patients with chronic alcoholism

•

All postmenopausal women irrespective of their initial decision about ERT/HRT

•

All men over age 65

A. Who Should Not Have a BMD Study? Premenopausal adult women [3,4] and men under age 65 (an arbitrary age) in whom there is no clinical suspicion of potential for bone loss will not benefit from a BMD study (see Table 1). For these persons it is generally sufficient to offer advice about diet, exercise, and avoidance of tobacco and alcohol without the need for BMD testing. That is not to say that some of these individuals will not have low BMD (osteopenia with T score 1.0) or even low enough to deserve the diagnostic label osteoporosis (T score 2.5). Since BMD is normally distributed about a bell-shaped curve, by definition 50% of normal people will

B. Who Should Have a BMD Study? In contrast to the above discussion there are several circumstances in which a BMD study is most appropriate in premenopausal women and men under age 65. Some are immediately obvious such as the patient who is initiating or already receiving chronic corticosteroid therapy [6]. Except in those circumstances such as an acute allergic reaction, in which it is assumed that steroid therapy will be very brief it is probably wise to err on the conservative side and measure BMD whenever the anticipated duration of steroid therapy is uncertain. In this event BMD testing can be considered inexpensive and without harm and provides important baseline information to consider when the possibility of prolonging or repeating steroid therapy exists. Likewise, persons with known intestinal malabsorption as a result of disease or surgery may benefit from a BMD study. Less clear is the approach to the patient with calcium oxalate nephrolithiasis in whom renal leak hypercalciuria may have long-term adverse effects on bone mass [7]. In general such a patient can be well managed without the BMD study but knowledge of BMD may conceivably act as a stimulus to adhere to chronic therapy such as a thiazide diuretic even though the patient is asymptomatic between bouts of acute renal colic or may never experience more than one stone episode.

405

CHAPTER 57 Evaluation of the Patient

Any clinical circumstance where there is a question about the integrity of the hypothalamic – pituitary – gonadal axis should be considered appropriate for a BMD study. In women this includes exercise-associated amenorrhea [8], anorexia [9,10], bulimia [11], prolactinoma [12], gonadotrophin-releasing hormone (GnRH) therapy for endometriosis [13,14], or irregular menses, particularly if associated with infertility. The one exception to this latter comment are women with polycystic ovarian syndrome where the hyperandrogenemia has been reported to be associated with a higher than normal BMD [15,16]. There are conflicting data on an association between oral contraceptive use and BMD [17,18] but there is increasing concern about the use of depot medroxyprogesterone acetate as a means of contraception, with most studies suggesting an adverse effect on bone health [19,20], particularly in teens. Men under age 65 with erectile dysfunction should be evaluated for hypogonadism and if present should have BMD tested, similarly with men being evaluated for infertility or other clinical evidence of hypogonadism. In the United States the decision about BMD testing in women age 65 or older has been made by the Federal Government by passage of the Bone Mass Measurement Act (BMMA) in 1997 [21]. Under this act, Medicare regulations mandate that BMD testing should be reimbursed in all estrogen-deficient women age 65 or older who are at risk for osteoporosis. Since that includes all women in this age range the decision-making is easy. Measure BMD! No such regulations yet exist for men.

C. Unresolved Issues It is unclear whether BMD should be measured in the person who sustains stress fractures of the lower extremities, particularly metatarsal stress fractures [22,23]. There is evidence that some of these patients will have low BMD but the effect of specific osteoporosis therapy has not been reported. If the stress fractures are clearly exercise related in someone who participates in frequent intense exercise, it seems prudent to recommend a change in the exercise program no matter what the BMD. However it is also conceivable that the stress fracture with low BMD might be the presentation of a previously unrecognized metabolic bone disease so that a BMD study would not be inappropriate in all persons with stress fractures. The most difficult decision about whether to order a BMD study occurs in women between menopause and age 65. Under the principle that the test should not be ordered if the result will not alter therapeutic considerations there is no apparent need to measure BMD in a woman who elects, for nonskeletal as well as skeletal reasons, to begin estrogen therapy at the menopause. However, the data on the very limited long-term compliance with postmenopausal

estrogen therapy [24,25] would suggest that a simple BMD study is appropriate even for these women. Furthermore, there is published evidence that knowledge of BMD leads to improved long-term compliance [26,27]. BMD testing is certainly indicated in women who do not immediately elect estrogen therapy at the menopause. All postmenopausal women lose bone, albeit at varying rates and with consequences dependent on the initial (at menopause) BMD and the subsequent rate of loss. Specific therapy to prevent postmenopausal bone loss (including estrogen in those women who choose not to start it for nonskeletal reasons) should not be prescribed without prior BMD measurement. Each of the current therapies are safe in all but a very small percentage of patients, so safe that some are being considered for over-the-counter sale without prescription. Should this occur it would clearly change the recommendation to measure BMD prior to their use but this would not necessarily be sound medical practice.

III. WHAT BMD TO MEASURE As with the decision about the patient in whom to obtain a BMD measurement, common sense applies to the decision about what to measure. All BMD and ultrasound devices approved in the United States by the Food and Drug Administration (FDA) have been documented to predict fracture risk. By that standard any measurement by any technology at any site will suffice to predict fracture risk in an individual patient. A problem arises when the BMD test is linked to reimbursement for the study since many insurers will reimburse only if a diagnosis of osteoporosis (T score 2.5) is established by the test. Several studies have demonstrated that the case detection rate with a single site BMD measurement is not very good (50 – 60%) in women between menopause and ages 65 – 70 [28]. In women over age 65 – 70 there is general agreement that hip BMD measurement, as the best predictor of hip fracture, is the most appropriate study to perform [1]. Outside of this group it would be appropriate to measure any and all sites for which the technology is readily available. The more sites that are measured, the more likely it is than one or more will detect osteoporosis. Conversely, the more sites that are measured and normal, the more confidence one can have that the patient has normal BMD.

A. Peripheral BMD All peripheral BMD instruments are smaller, less expensive, and more portable than central BMD. This makes these instruments far more suitable for the primary care physician’s office and useful for initial intervention decision-making. Peripheral BMD is apparently as good as

406

MICHAEL KLEEREKOPER

central BMD for overall fracture risk assessment in postmenopausal women under age 70, even if the value is not low enough to be diagnostic for osteoporosis. If, on the basis of a peripheral BMD study, a clinician is satisfied that the patient does not need specific osteoporosis therapy it would be appropriate to simply repeat the study in 3 – 5 years. If the clinician decides, on the basis of a peripheral BMD study, that specific intervention is appropriate no further BMD testing is really indicated. The regulations surrounding the BMMA do permit reimbursement for a central BMD if the initial intervention decision is made on the basis of a peripheral BMD but that is not always a necessary additional procedure to perform. From time to time, particularly with clinicians new to BMD measurements, a clear cut intervention decision cannot be made on the basis of the history, physical examination, and peripheral BMD. In that circumstance it is most appropriate to consider a central BMD study. The need for central BMD after an initial peripheral BMD will become less as the clinician becomes more familiar with the technology and its limitations.

TABLE 2 Laboratory studies indicated in all patients with low BMD •

Complete blood count

•

Biochemical profile (calcium, phosphate, phosphatase, creatinine, AST/ALT, protein)

•

TSH

•

24-Hour urine creatinine, calcium, sodium

•

Testosterone (men only)

•

Additional studies based on history, physical examination, and initial laboratory results

IV. THE EVALUATION OF THE PATIENT WITH LOW BMD A patient’s initial BMD measurement cannot distinguish a low BMD based on a failure of peak bone mass acquisition from loss of bone from a normal peak BMD, or a

Serum protein immunoelectrophoresis

•

Urine free cortisol

•

25-Hydroxyvitamin D

•

PTH

•

Anti-gliadin and anti-myosin antibodies

•

Urine histamine

Laboratory studies suggested prior to therapy for osteoporosis •

B. Central BMD At present central BMD using DXA is performed on instruments capable of measuring multiple skeletal sites, most often the spine, the proximal femur, and the forearm. It is advisable to measure all three sites in all patients unless precluded by technical problems (previous forearm fracture, hip prosthesis, etc.). This improves the ability to diagnose osteoporosis and predict fracture risk. As patients become older and the likelihood of degenerative changes in the spine increases, the value of this measurement site should be discounted. It is probably best not to rely only on spine BMD in most patients since all available instruments capable of measuring spine BMD can also measure the forearm and hip. Many BMD centers do not routinely measure forearm BMD. This is probably not a sound policy since both hyperthyroidism and hyperparathyroidism appear to have greatest effect on forearm BMD and might be missed if this study is not done. This is not to say that these two diseases should be diagnosed on the basis of an isolated low forearm BMD alone but such a finding on a BMD study should raise the level of suspicion about their presence.

•

Serum or urine bone resorption markers •

Pyridinoline

•

Deoxypyridinoline

•

N-telopeptide of collagen cross-links

•

C-telopeptide of collagen cross-links

•

Serum formation markers

•

Bone-specific alkaline phosphatase

•

Osteocalcin

combination of these two. The diagnostic approach must begin with an adequate history and physical examination (see Table 1). The key elements of the history begin with a family history of either osteoporosis or fragility fractures [29]. If a woman’s mother had either of these two conditions it is likely that she has inherited a low peak BMD. This does not minimize her own risk for developing osteoporosis or sustaining a fracture but does reduce the likelihood of further diagnostic studies uncovering a secondary cause for bone loss. A lifelong low dietary calcium intake has been implicated in a low peak BMD but not in accelerated bone loss in otherwise healthy persons. Studies have carefully sought an association between nulliparity and low BMD but this depends on whether the nulliparity was by choice or resulted from hormonal changes, leading to infertility. The association between oral contraceptive use and low BMD remains unresolved. Immobilization or restricted physical activity, particularly during the adolescent growth spurt, may contribute to deficits in peak BMD. Male hypogonadism is not often sought in a routine clinical history but is essential to seek out in men with low BMD. Unfortunately there are few diagnostic clues in the history with the most specific being a decrease in libido which many men deny even when directly questioned. If a specific clinical circumstance as described above was the

CHAPTER 57 Evaluation of the Patient

indication for the BMD test a low value will not be a surprise and not require further diagnostic evaluation. The physical examination will be normal in most people with low BMD without fractures. Occasionally surprises are detected and include the central obesity and dermal atrophy of Cushing’s syndrome, the rash and hepatomegaly of systemic mastocytosis, the peripheral manifestations of hyperthyroidism, or the blue sclerae and joint laxity of osteogenesis imperfecta or other inborn errors of collagen metabolism. Male hypogonadism may be difficult to detect on physical examination but a testicular examination should be done on all men with low BMD. Even when subsequently shown to be hypogonadal the testicular examination is often normal. One subtle clue is a hyperrugosity to the face with many fine wrinkles on the face of a man with limited facial hair. A few routine diagnostic studies are appropriate in all persons with otherwise unexplained or unexpected low BMD. Most of these will already be available to the treating physician and include a complete blood count (CBC), a biochemical profile, and measurement of TSH. The CBC will provide clues to the presence of malnutrition (macroor microcytic anemia) or hematologic malignancy which can present with low BMD, albeit rarely. It is much more common to find a low BMD in such patients than for the low BMD or fragility fracture to be the presenting complaint. On the biochemical profile it is important to check hepatic, renal, and parathyroid status as well as to consider malnutrition (low albumin) or myeloma (hyperglobulinemia). Assessment of renal and hepatic function is straightforward. Assessment of parathyroid function requires a little more attention. Hypercalcemia is obvious in primary hyperparathyroidism but secondary hyperparathyroidism not due to renal failure may be more subtle. An increasingly common cause, particularly in the unwell elderly or shut-ins, is vitamin D insufficiency or deficiency, for which hypocalcemia is a very late manifestation. Early in the course of the secondary hyperparathyroidism of vitamin D deficiency there will be mild hypophosphatemia and an increased total serum alkaline phosphatase activity. Many studies have addressed the issue of hyperthyroidism as a cause of accelerated bone loss and low BMD but most have found only mild effects if any. Nonetheless, thyroid disease is very prevalent in women and many postmenopausal women are taking thyroxine replacement for indications long since forgotten. As with any patient on thyroxine replacement the goal should be to maintain a euthyroid state (other than in suppressive therapy for thyroid cancer) and this is an appropriate indication to confirm that the patient on replacement therapy is euthyroid. Only infrequently will mild hyperthyroidism be detected by such screening in women not on replacement therapy. While serum estrogen concentrations are rarely indicated in postmenopausal women, even those with low BMD, all men with low BMD

407 should have measurement of serum testosterone. There is considerable debate as to the best measurement to assess gonadal function in men. Most authorities agree that bioavailable testosterone is the most specific test of male hypogonadism but this test is not performed well in many laboratories. Serum free testosterone is seemingly preferred over total serum testosterone measurement but this too is not often done well. Individual clinicians should check with the laboratory about the available options and their performance characteristics. Most patients do not appreciate having to provide a 24-h urine specimen but this may provide valuable information in the patient with unexplained or unexpected low BMD. The specimen should be analyzed for creatinine (as a measure of the completeness of the collection) as well as for calcium and sodium. Low urine calcium (50 mg/24 h) should raise the level of suspicion about inadequate calcium or vitamin D intake or absorption. If there is hypocalciuria, dietary calcium intake should be increased and the study repeated in 4 – 6 weeks. If the response is inadequate, 25-hydroxyvitamin D should be measured. A high urine calcium (4 mg/kg body weight/24 h) may simply reflect a high sodium diet (hence the need to measure urine sodium [30]) but this is nonetheless a cause of progressive negative calcium balance at the expense of the skeleton. If both the urine calcium and the urine sodium are high the patient should be advised to eat a lower sodium diet for 4 – 6 weeks before repeating the study. If the urine calcium, but not the urine sodium, is increased and the patient is taking a calcium supplement the study should be repeated after 4 – 6 weeks without the supplement. Isolated hypercalciuria without adequate explanation should be treated with a thiazide diuretic and in some patients this might be the only required intervention. It has been suggested that the likelihood of detecting a secondary cause for osteoporosis increases if the BMD is markedly lower than anticipated or if the patient has already sustained an unexplained or unexpected fragility fracture. The best clue to an unexpectedly low BMD is when the value is more than 2 standard deviations below the mean for persons of the same age, sex, and ethnicity (Z score 2.0). While this seems quite logical, there are no studies that fully support this suggestion and the prevalence of secondary osteoporosis is unknown in patients with either normal or low Z scores. Despite the limited firm evidence it would seem wise to proceed to additional studies in all patients where the clinician is uncomfortable, for whatever reason, with a diagnosis of primary osteoporosis. Studies to be considered include serum immunoelectrophoresis, parathyroid hormone, 25-hydroxyvitamin D, urine histamine, and urine free cortisol (or an overnight dexamethasone suppression test). Biochemical markers of bone remodeling, discussed in detail elsewhere in this textbook, deserve special mention in the context of the evaluation of the patient with low

408 BMD. There is increasing evidence that patients with elevated levels of bone remodeling markers in the untreated state can be expected to have significantly greater rates of bone loss if left untreated than those with normal pretreatment marker levels [31]. This is of limited importance if, on the basis of the history (including family history), physical examination, and BMD, a decision to offer specific osteoporosis therapy has been made. Likewise if a decision not to offer specific therapy has been made on the basis of the initial evaluation. In all other patients knowledge of the biochemical marker value should assist in the therapeutic decision making. It has been vigorously argued that the variability in the markers is so great that they cannot be relied on for individual patient care. This is probably an overstatement. The variability in the markers is no greater than with many other esoteric laboratory procedures in common use in clinical medicine. When the markers are used in the manner suggested above (i.e., when a therapeutic decision is not made definitively on the basis of known information about the patient) it is difficult to make a wrong intervention decision based on the marker value. As clinicians use these markers of bone remodeling more regularly they will gain confidence in their more precise role in their practices. All osteoporosis therapies that are available in the United States at the time of this writing are inhibitors of bone resorption. The controlled clinical trials on which approval was based all document a dose-dependent reduction in markers that differs significantly from pretreatment values earlier than a significant increase in BMD can be demonstrated. It appears logical to monitor anti-resorptive therapy with a marker of resorption. The positive predictive value for an early decrease in markers to herald a later increase in BMD is 90% (32). An early on-treatment decrease in markers should provide the clinician (and patient) with confidence that the follow-up BMD value will be improved or unchanged from the pretreatment BMD. An early on-treatment increase in resorption markers should prompt the clinician to determine that the patient is taking therapy regularly as prescribed. On this basis a case can be made for obtaining a pretreatment marker value in all patients, but this is not yet a widely held opinion by most authorities.

V. SUMMARY AND CONCLUSIONS The new definition of osteoporosis as a disease of fracture risk and the increased availability of BMD measurement devices has changed the approach to the patient with or who is considered at risk for osteoporosis. The disease or disease risk should be detected in the asymptomatic state by BMD measurement just as heart attack risk is assessed by serum lipid measurements or stroke risk assessed by blood pressure measurement in asymptomatic patients. BMD predicts fracture risk as well as blood pressure predicts risk of

MICHAEL KLEEREKOPER

stroke and better than total serum cholesterol predicts risk of acute myocardial infarction. All patients who sustain an unexplained fragility fracture should be considered to have osteoporosis and have BMD measured. This chapter has described the clinical situations where a BMD study is and is not appropriate on the basis of the history and physical examination and where it is not clearly indicated but might be beneficial. Once a low BMD has been documented a number of laboratory studies should be performed to make certain that a correctable secondary cause for osteoporosis does not exist. The potential roles of biochemical markers of bone remodeling in the evaluation of the patient with low BMD are also discussed. Specific details of the BMD test and of the biochemical markers are presented elsewhere in this text and the reader is encouraged to critically review those chapters.

References 1. Physician’s Guide to Prevention and Treatment of Osteoporosis. http://www.nof.org/physguide/entry_form.htm. 2. “AACE Clinical Practice Guidelines for Prevention and Treatment of Postmenopausal Osteoporosis.” http://www.aace.com/clinguideindex.htm. 3. C. A. Moreira Kulak, D. H. Schussheim, D. J. McMahon, E. Kurland, S. J. Silverberg, E. S. Siris, J. P. Bilezikian, and E. Shane, Osteoporosis and low bone mass in premenopausal and perimenopausal women. Endocr. Pract. 6, 296 – 304 (2000). 4. A. A. Licata, Does she or doesn’t she . . . have osteoporosis? The use and abuse of bone densitometry. Endocr. Pract. 6(4), 336 – 337 (2000). 5. “Osteoporosis Prevention, Diagnosis and Treatment.” http://consensus.nih.gov. 6. “Recommendations for the Prevention and Treatment of Glucocorticoid-Induced Osteoporosis.” http://www.rheumatology.org/research/guidelines/osteo/osteo.html. 7. J. S. Adams, C. F. Song, and V. Kantorovich, Rapid recovery of bone mass in hypercalciuric, osteoporotic men treated with hydrochlorothiazide. Ann. Int. Med. 130(8), 658 – 660 (1999). 8. J. H. Gibson, A. Mitchell, J. Reeve, and M. G. Harries, Treatment of reduced bone mineral density in athletic amenorrhea: A pilot study. Osteoporosis Int. 10(4), 284 – 289 (1999). 9. D. K. Baker, R. Roberts, and T. Towell, Factors predictive of bone mineral density in eating-disordered women: A longitudinal study. Int. J. Eating Disord. 27(1); 29–35 (2000). 10. L. A. Soyka, S. Grinspoon, L. L. Levitsky, D. B. Herzog, and A. Klibanski, The effects of anorexia nervosa on bone metabolism in female adolescents. J. Clin. Endo. Metab. 84(12), 4489 – 4496 (1999). 11. J. Sundgot-Borgen, R. Bahr, J. A. Falch, and L. S. Schneider, Normal bone mass in bulimic women. J. Clin. Endocr. Metab. 83(9), 3144 – 3149 (1998). 12. J. S. Sanfilippo, Implications of not treating hyperprolactinemia. J. Reprod. Med. 44(12S), 1111 – 1115 (1999). 13. N. Zamberlan, R. Castello, D. Gatti, M. Rossini, V. Braga, E. Fracassi, and S. Adami, Intermittent Etidronate partially prevents bone loss in hirsute hyperandrogenic women treated with GnRH agonist. Osteoporosis Int. 7(2), 133 – 137 (1997). 14. R. Revilla, M. Revilla, L. F. Villa, J. Cortes, I. Arribas, and H. Rico, Changes in body composition in women treated with gonadotropinreleasing hormone agonists. Maturitas 31(1), 63 – 68 (1998).

CHAPTER 57 Evaluation of the Patient 15. S. Adami, N. Zamberlan, R. Castello, F. Tosi, D. Gatti, and P. Moghetti, Effect of hyperandrogenism and menstrual cycle abnormalities on bone mass and bone turnover in young women. Clin. Endocr. 48(2), 169 – 173 (1998). 16. G. Lupoli, C. Di Carlo, V. Nuzzo, G. Vitale, D. Russo, S. Palomba, and C. Nappi, Gonadotropin-releasing hormone agonists administration in polycystic ovary syndrome. Effects on bone mass. J. Endocr. Invest. 20(8), 493 – 496 (1997). 17. J. A. Pasco, M. A. Kotowicz, M. J. Henry, S. Panahi, E. Seeman, and G. C. Nicholson, Oral Contraceptives and bone mineral density: A population-based study. Am. J. Obstet. Gynecl. 182(2), 265 – 269 (2000). 18. B. A. Cromer, Effects of hormonal contraceptives on bone mineral density. Drug Safety 20(3), 213 – 222 (1999). 19. O. S. Tang, G. Tang, P. Yip, B. Li, and S. Fan, Long-term depotmedroxprogesterone acetate and bone mineral density. Contraception 59(1), 25 – 29 (1999). 20. D. Scholes, A. Z. Lacroix, S. M. Ott, L. E. Ichikawa, and W. E. Barlow, Bone mineral density in women using depot medroxprogesterone acetate for contraception. Obstet. Gynecl. 93(2), 233 – 238 (1999). 21. http://www.medicare.gov/health/osteoporosisdetails.asp. 22. T. D. Lauder, S. Dixit, L. E. Pezzin, M. V. Williams, C. S. Campbell, and G. D. Davis, The relation between stress fractures and bone mineral density: Evidence from active-duty Army women. Arch. Phys. Med. Rehab. 81(1), 73 – 79 (2000). 23. A. D. Cline, G. R. Jansen, and C. L. Melby, Stress fractures in female army recruits: Implications of bone density, calcium intake, and exercise. J. Am. Coll. Nutr. 17(2), 128 – 135 (1998). 24. B. Ettinger, A. Pressman, and P. Silver, Effect of age on reasons for initiation and discontinuation of hormone replacement therapy. Menopause 6(4), 282 – 289 (1999).

409 25. B. Ettinger, A. Pressman, and C. Bradley, Comparison of continuation of postmenopausal hormone replacement therapy: transdermal versus oral estrogen. Menopause 5(3), 152 – 156 (1998). 26. C. Castelo-Branco, F. Figueras, A. Sanjuan, J. J. Vicente, M. J. Martinez de Osaba, F. Pons, J. Balasch, and J. A. Vanrell, Long-term compliance with estrogen replacement therapy in surgical postmenopausal women: Benefits to bone and analysis of factors associated with discontinuation. Menopause 6(4), 307 – 311 (1999). 27. R. P. Cole, S. Palushock, and A. Haboubi, Osteoporosis management: Physicians’ recommendations and womens’ compliance following osteoporosis testing. Women Health 29(1), 101 – 115 (1999). 28. D. A. Nelson, R. Molloy, and M. Kleerekoper, Prevalence of osteoporosis in women referred for bone density testing. J. Clin. Densitom. 1(1), 5 – 11 (1998). 29. R. W. Keen, D. J. Hart, N. K. Arden, D. V. Doyle, and T. D. Spector, Family history of appendicular fracture and risk of osteoporosis: a population-based study. Osteoporosis Int. 10(2), 161 – 116 (1999). 30. M. Cirillo, C. Ciacci, M. Laurenzi, M. Mellone, G. Mazzacca, and N. G. De Santo, Salt intake urinary sodium, and hypercalciuria. Miner. Electrolyte Metab. 23(3 – 6), 265 – 268 (1997). 31. P. Garnero, E. Sornay-Rendu, F. Duboeuf, and P. D. Delmas, Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: The OFELY study. J. Bone Miner. Res. 14(9), 1614 – 1621 (1999). 32. P. Ravn, D. Hosking, D. Thompson, G. Cizza, R. D. Wasnich, M. McClung, A. J. Yates, N. H. Bjarnason, and C. Christiansen, Monitoring of alendronate treatment and prediction of effect on bone mass by biochemical markers in the early postmenopausal intervention cohort study. J. Clin. Endocr. Metab. 84(7), 2363 – 2368 (1999).

CHAPTER 58

Imaging of Osteoporosis MICHAEL JERGAS AND HARRY K. GENANT Department of Radiology, University of California, San Francisco, San Francisco, California 94143

I. Introduction II. Principal Radiographic Findings in Osteopenia and Osteoporosis

III. Diseases Characterized by Generalized Osteopenia IV. Regional Osteoporosis References

I. INTRODUCTION

II. PRINCIPAL RADIOGRAPHIC FINDINGS IN OSTEOPENIA AND OSTEOPOROSIS

The term osteoporosis is widely used clinically to mean generalized loss of bone, or osteopenia, accompanied by relatively atraumatic fractures of the spine, wrist, hips, or ribs. Because of uncertainties of specific radiologic interpretation, the term osteopenia (“poverty of bone”) has been used as a generic designation for radiographic signs of decreased bone density. Radiographic findings suggestive of osteopenia and osteoporosis are frequently encountered in daily medical practice and can result from a wide spectrum of diseases ranging from highly prevalent causes such as postmenopausal and involutional osteoporosis to very rare endocrinologic and hereditary or acquired disorders (Table 1). Histologically, the result in each of these disorders is a deficient amount of osseous tissue, although different pathogenic mechanisms may be involved. Conventional radiography is widely available and, alone or in conjunction with other imaging techniques, it is widely used for the detection of complications of osteopenia, for the differential diagnosis of osteopenia, or for follow-up examinations in specific clinical settings. Bone scintigraphy, computed tomography and magnetic resonance imaging are additional diagnostic methods that are applied almost routinely to aid in the differential diagnosis of osteoporosis and its sequelae.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

The absorption of X-rays by a tissue depends on the quality of the X-ray beam, the character of the atoms composing the tissue, the physical density of the tissue, and the thickness of the penetrated structure. The amount of X-ray absorption defines the density of X-ray shadow that a tissue casts on the film. Because the absorption increases with the third power of the atomic number, and because calcium has a high atomic number, it is primarily the amount of calcium that affects the X-ray absorption of bone. The amount of calcium per unit mineralized bone volume in osteoporosis remains constant at about 35% [1,2]. Therefore, a decrease in the mineralized bone volume results in a decrease of the total bone calcium and consequently a decreased absorption of the X-ray beam. On the radiograph this phenomenon is referred to as increased radiolucency. At the same time as bone mass is lost, changes in the bone structure occur, and these can be observed radiographically. Bone is composed of two compartments: cortical bone and trabecular bone. The structural changes seen in cortical bone represent bone resorption at different sites (e.g., the inner and outer surfaces of the cortex, or within the cortex in the Haversian and Volkmann channels). These

411

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

412

JERGAS AND GENANT

TABLE 1 Disorders Associated with Radiographic Osteoporosis and Osteopenia Primary osteoporosis 1. Involutional osteoporosis (postmenopausal and senile) 2. Juvenile osteoporosis Secondary osteoporosis A. Endocrine 1. Adrenal cortex (Cushing’s disease) 2. Gonadal disorders (hypogonadism) 3. Pituitary (hypopituitarism) 4. Pancreas (diabetes) 5. Thyroid (hyperthyroidism) 6. Parathyroid (hyperparathyroidism) B. Marrow replacement and expansion 1. Myeloma 2. Leukemia 3. Metastatic disease 4. Gaucher’s disease 5. Anemias (sickle cell disease, thalassemia) C. Drugs and substances 1. Corticosteroids 2. Heparin 3. Anticonvulsants 4. Immunosuppressants 5. Alcohol (in combination with malnutrition) D. Chronic disease 1. Chronic renal disease

FIGURE 1

Patterns of bone resorption. Subperiosteal bone resorption characterizes hyperparathyroidism; endosteal resorption is prominent in senile osteoporosis. Intracortical and trabecular resorption are features of postmenopausal osteoporosis.

2. Hepatic insufficiency 3. Gastrointestinal malabsorption 4. Chronic inflammatory polyarthropathies 5. Chronic immobilization E. Deficiency states 1. Vitamin D 2. Vitamin C (scurvy) 3. Calcium 4. Malnutrition F. Inborn errors of metabolism 1. Osteogenesis imperfecta 2. Homocystinuria

three sites (endosteal, intracortical, and periosteal) may react differently to distinct metabolic stimuli, and careful investigation of the cortices may be of value in the differential diagnosis of metabolic disease affecting the skeleton (Fig. 1). Cortical bone remodeling typically occurs in the endosteal “envelope,” and the interpretation of subtle changes in this layer may be difficult at times. With increasing age, there is a widening of the marrow canal due to an imbalance of endosteal bone formation and resorption that leads

to a “trabecularization” of the inner surface of the cortex. Endosteal scalloping due to resorption of the inner bone surface can be seen in high bone turnover states such as reflex sympathetic dystrophy. Intracortical bone resorption may cause longitudinal striation or tunneling, predominantly in the subendosteal zone. These changes are seen in various high turnover metabolic diseases affecting the bone such as hyperparathyroidism, osteomalacia, renal osteodystrophy, and acute osteoporoses from disuse or the reflex sympathetic dystrophy syndrome but also postmenopausal osteoporosis (Fig. 2). Intracortical tunneling is a hallmark of rapid bone turnover. It is usually not apparent in disease states with relatively low bone turnover such as senile osteoporosis. Accelerated endosteal and intracortical resorption with intracortical tunneling and indistinct border of the inner cortical surface, is best depicted with high-resolution radiographic techniques. Intracortical tunneling must be distinguished from nutritional foraminae, which are isolated and present with an oblique orientation. Intracortical resorption is also a sign of bone viability and is not seen in necrotic or allograft bone.

413

CHAPTER 58 Imaging of Osteoporosis

FIGURE 2

Bone loss after oophorectomy. Accelerated intracortical resorption in the proximal phalanges at baseline (A) and 2 years following oophorectomy (B).

Subperiosteal bone resorption is associated with an irregular definition of the outer bone surface. This finding is pronounced in diseases with a high bone turnover, principally primary and secondary hyperparathyroidism. However, rarely it may also be present in other diseases. Cortical thinning with expansion of the medullary cavity occurs as endosteal bone resorption exceeds periosteal bone apposition in most adults. In the late stages of osteoporosis, the cortices appear paper thin with the endosteal surface usually being smooth (Fig. 3). The trabecular bone responds faster to metabolic changes than does cortical bone [3]. Trabecular bone changes are most prominent in the axial skeleton and in the ends of the long and tubular bones of the appendicular skeleton (juxtaarticular), e.g., proximal femur, distal radius. These are sites with a relatively great proportion of trabecular bone. Loss of trabecular bone (in cases of low rates of loss) occurs in a predictable pattern. Non-weight-bearing trabeculae are resorbed first. This leads to a relative prominence of the weight-bearing trabeculae. The remaining trabeculae may even become thicker, which may result in a distinct radiographic trabecular pattern. For example, early

changes of osteopenia in the lumbar spine typically include a rarefication of the horizontal trabeculae accompanied by a relative accentuation of the vertical trabeculae. This may lead to an appearance of vertical striation of the bone. With decreasing density of the trabecular bone the cortical rim of the vertebrae are more accentuated, and the vertebrae may have a “picture-frame” appearance. In addition to the changes in the trabecular bone, thinning of the cortical bone occurs. Changes of the bone structure at distinct skeletal sites are assessed for the differential diagnosis of various skeletal conditions. For the evaluation of very subtle changes, such as different forms of bone resorption, highresolution radiographic techniques with optical or geometric magnification may be required [4 – 6]. The anatomic distribution of the osteopenia or osteoporosis depends on the underlying cause. Osteopenia can be generalized affecting the whole skeleton, or regional, affecting only a part of the skeleton, usually in the appendicular skeleton. Typical examples of generalized osteopenias are involutional and postmenopausal osteoporosis and osteoporosis caused by endocrine disorders such as hyperparathyroidism, hyperthyroidism, osteomalacia, and

414

JERGAS AND GENANT

TABLE 2 Factors Influencing the Radiographic Appearance of Bones and Soft Tissue Radiation source

Exposure time Film-focus distance Anode characteristics Voltage Beam filtration

Object

Thickness of bone Bone mineral content Soft tissue composition Scattering

Film and screen

Film granularity Emulsion of film Film speed Screen properties

Film processing

Developing time Temperature of developer Type of developer Type of fixer Type of processing (automated vs manual)

After Heuck and Schmidt [7].

FIGURE 3 Cortical thinning in senile osteoporosis, primarily the result of endosteal bone resorption. hypogonadism. Regional forms of osteoporosis result from factors affecting only parts of the appendicular skeleton such as disuse, reflex sympathetic syndrome, and transient osteoporosis of large joints. The distribution of osteopenia may vary considerably between different diseases and may be suggestive of a specific diagnosis. Focal osteopenia primarily reflects the underlying cause such as inflammation, fracture, or tumor. Thus, it seems that a number of characteristic features by conventional radiography make the diagnosis of osteopenia or osteoporosis possible. However, the detection of osteopenia by conventional radiography is inaccurate since it is influenced by many technical factors such as radiographic exposure factors, film development, soft tissue thickness of the patient, etc. (Table 2) [7]. It has been estimated that as much as 20 to 40% of bone mass must be lost before a decrease in bone density can be seen in lateral radiographs of the thoracic and lumbar spine [8]. Finally, the diagnosis of osteopenia from conventional radiographs is also dependent on the experience of the reader and his/her subjective interpretation [9]. In summary, a radiograph may reflect the amount of bone mass, histology and gross morphology of the skeletal

part examined. The principal findings of osteopenia are increased radiolucency, changes in bone microstructure, e.g., rarefication of trabeculae, thinning of the cortices, eventually resulting in changes of the gross bone morphology, i.e., changes in the shape of the bone and fractures. Further characteristics of osteopenic and osteoporotic disease conditions and specific techniques for their radiological assessment are described in greater detail below.

III. DISEASES CHARACTERIZED BY GENERALIZED OSTEOPENIA A. Involutional Osteoporosis Involutional osteoporosis is the most common generalized skeletal disease. It has been classified as a type I or postmenopausal osteoporosis and a type II or senile osteoporosis [10,11]. Gallagher added a third type meaning secondary osteoporosis (Table 3) [12]. Even though the importance of estrogen deficiency for postmenopausal osteoporosis has been established, the distinction between the first two types of osteoporosis is not generally accepted. Distinctions between postmenopausal and senile osteoporosis may sometimes be arbitrary, and the assignment of fracture sites to the different types of osteoporosis is uncertain. Postmenopausal osteoporosis is believed to represent that process occurring in a subset of postmenopausal women, typically between the ages 50 and 65

415

CHAPTER 58 Imaging of Osteoporosis

TABLE 3 Type:

Classification of Osteoporosisa

I Postmenopausal

II Senile

III Secondary

Age (years)

55 – 70

75 – 90

Any age

Years past menopause

5 – 15

25 – 40

—

Sex ratio (Female:Male)

20:1

2:1

1:1

Fracture site

Spine

Hip, spine, pelvis, humerus

Spine, hip, peripheral skeleton

Trabecular







Cortical







Menopause







Age







Bone loss

Contributing factor

a After Albright [10], Riggs and Melton [11], and Gallagher [12]; adapted from Gallagher [144].

years. There is accelerated trabecular bone resorption related to estrogen deficiency, and the fracture pattern in this group of women primarily involves the spine and the wrist. In senile osteoporosis, there is a proportionate loss of cortical and trabecular bone. The characteristic fractures of senile osteoporosis include fractures of the hip, the proximal humerus, the tibia, and the pelvis in elderly women and men, usually 75 years or older. Major factors in the etiology of senile osteoporosis include the age-related decrease in bone formation, diminished adrenal function, reduced intestinal calcium absorption, and secondary hyperparathyroidism. The radiographic appearance of the skeleton in involutional osteoporosis may include all of the aforementioned characteristics for generalized osteoporosis. The high prevalence of involutional osteoporosis with its typical radiographic manifestations has led to numerous attempts to diagnose and quantify osteoporosis based on its radiographic characteristics.

B. Osteopenia and Osteoporosis of the Axial Skeleton

due to an accentuation of the cortical outline, and increased biconcavity of the vertebral endplates (Fig. 4). The biconcavity of the vertebrae results from protrusion of the intervertebral disk into the weakened vertebral body. A classification of these characteristics can be found with the Saville index (Table 4) [13]. This index, however, has never gained widespread acceptance, being prone to great subjectivity and experience of the reader. Doyle and colleagues found that neither of the aforementioned signs of osteopenia reflect the bone mineral status of an individual reliably and cannot be used for follow-up of osteopenic patients [14]. Thus, bone density measurements using dedicated densitometric methods have widely replaced the subjective analysis of bone density from conventional radiographs. Densitometric results may suggest osteopenia even if the bone loss is not detectable on a spine radiograph. Nevertheless, the aforementioned radiographic signs of osteoporosis have been found to be significantly related to measured bone density, and normal bone densitometry measurements may sometimes have to be considered false if the radiograph displays characteristic changes of osteopenia [15,16].

C. Vertebral Fractures and Their Diagnosis The radiographic manifestation of osteopenia of the axial skeleton includes increased radiolucency of the vertebrae. The vertebral body’s radiographic density may assume the density of the intervertebral disk space. Further findings include vertical striation of the vertebrae due to reinforcement of vertical trabeculae in the osteopenic vertebra, framed appearance of the vertebrae (picture framing or empty box)

Vertebral fractures are the hallmarks of osteoporosis, and even though one may argue that osteopenia per se may not be diagnosed reliably from spinal radiographs, spinal radiography continues to be a substantial aid in diagnosing and following vertebral fractures (Fig. 5) [17]. Changes in the gross morphology of the vertebral body have a wide

416

JERGAS AND GENANT

TABLE 4 Grade

Osteopenia Score for Vertebrae Radiographic appearance of vertebra

0

Normal bone density

1

Minimal loss of density; endplates begin to stand out giving a stenciled effect

2

Vertical striation is more obvious; endplates are thinner

3

More severe loss of bone density than grade 2; endplates becoming less visible

4

Ghost-like vertebral bodies; density is no greater than soft tissue; no trabecular pattern is visible After Saville [13].

inherent to a radiologist’s reading [21 – 23]. Especially with respect to its specificity there is some controversy on the usefulness of a purely morphometric analysis of vertebral deformities [24,25]. Drawing on the strength of each of the approaches, both a quantitative approach as well as a standardized visual approach may be applied in combination to reliably diagnose vertebral fractures in clinical drug trials [26,27]. More recently, dual X-ray absorptiometry scanners have been employed for acquiring lateral images of the thoracolumbar spine, and a visual or morphometric analysis of the vertebrae has been used to classify vertebral deformities according to the aforementioned criteria (Fig. 6) [28 – 30]. While morphometric X-ray absorptiometry (MXA) may be helpful in the serial assessment of vertebral deformities its diagnostic validity is still under investigation [31 – 35]. FIGURE 4 Overall loss of bone density results in increased radiolucency of the vertebrae with relative accentuation of the cortical rim (“picture framing”). range of appearances from increased concavity of the endplates to a complete destruction of the vertebral anatomy in vertebral crush fractures. In clinical practice conventional radiographs of the thoracolumbar region in lateral projection are analyzed qualitatively by radiologists or experienced clinicians to identify vertebral deformities or fractures. For an experienced radiologist, this assessment generally is uncomplicated, and it can be aided by additional radiographic projections such as anteroposterior and oblique views, or by complimentary examinations such as bone scintigraphy, computed tomography, and even magnetic resonance imaging [18 – 20]. Being the most frequent site of fracture in early postmenopausal women vertebral fractures have become the most important endpoints in epidemiological studies and clinical drug trials. In these settings conventional radiography is usually the only method applied to assess vertebral fractures. Several quantitative morphometric methods that rely entirely on measurements of vertebral heights for vertebral fracture diagnosis have been proposed to reduce the subjectivity that is

D. Osteopenia and Osteoporosis at Other Skeletal Sites The axial skeleton is not the only site where characteristic changes of osteopenia and osteoporosis can be depicted radiographically. Changes in the trabecular and cortical bone can also be seen in the appendicular skeleton. It is first apparent at the ends of long and tubular bones due to the predominance of cancellous bone in these regions. Endosteal resorption has a prominent role particularly in senile osteoporosis. The net result of this chronic process is widening of the medullary canal and thinning of the cortices. In late stages of senile osteoporosis, the cortices are paper-thin and the endosteal surfaces are smooth. In rapidly evolving postmenopausal osteoporosis accelerated endosteal and intracortical bone resorption may be seen and can be directly assessed by high-resolution radiographic techniques. Methods for quantitating the changes at the peripheral skeleton have been proposed and also clinically applied (e.g., Singh-index, radiogrammetry) [36 – 41]. Conventional radiography is the basis for a number of recent studies exploring new aspects of assessing bone structure using so-

417

CHAPTER 58 Imaging of Osteoporosis

FIGURE 5 Advanced involutional osteoporosis with multiple fractures including end plate, wedge, and compression fractures in the thoracic spine (A) and biconcave vertebrae with end plate fractures in the lumbar spine (B). phisticated image analysis procedures such as fractal analysis or fast Fourier transforms [42 – 48]. These techniques have also been applied to the study of bone structure using high-resolution images acquired with magnetic resonance imaging or computed tomography [49 – 62]. This scientific field is quite new, and the application of these techniques in a clinical setting awaits validation (see also Chapter 35).

E. Differential Diagnosis of Reduced Bone Mass Aside from senile and postmenopausal states there are various other conditions that may be accompanied by generalized osteoporosis. While most of the previously mentioned radiographic characteristics are shared by a variety of conditions, there may be some apparent differences in the appearance of osteoporosis as compared to involutional osteoporosis.

F. Endocrine Disorders Associated with Osteoporosis Increased serum concentrations of the parathyroid hormone in hyperparathyroidism may result from autonomous hypersecretion by a parathyroid adenoma or diffuse hyperplasia of the parathyroid glands (primary hyperparathyroidism). A long-sustained hypocalcemic stimulus may result in hyperplasia of all parathyroid glands and secondary hyperparathyroidism. The cause of hypocalcemia usually is chronic renal failure or rarely malabsorption states. Patients with long-standing hyperparathyroidism may develop autonomous function and hypercalcemia (tertiary hyperparathyroidism). While it is the increase in serum parathyroid hormone and calcium concentrations that establishes the diagnosis, radiographs document the severity and the course of the disease. Hyperparathyroidism leads to both increased bone resorption

418

JERGAS AND GENANT

FIGURE 6 Morphometric X-ray absorptiometry uses lateral scans of the thoracolumbar spine acquired on dual X-ray absorptiometry scanners. A morphometry analysis of the vertebral heights is used to identify vertebral deformities based on specific criteria. and bone formation. Changes induced by hyperparathyroidism may affect all bone surfaces resulting in subperiosteal, intracortical, endosteal, subchondral, subepiphyseal, subligamentous and subtendinous, and trabecular bone resorption [63 – 65] (see also Chapter 49). Subperiosteal bone resorption is the most characteristic radiographic feature of hyperparathyroidism [66]. It is especially prominent in the hand, wrist, and foot but may also be seen other sites. Radiographically, the outer margin of the bone becomes indistinct. Scalloping and spiculations of the cortex may occur in later stages (Fig. 7). Undermineralization of the tela ossea leads to the distinctive radiographic appearance of acro-osteolyses [67,68]. Intracortical resorption results in longitudinally oriented linear striations within the cortex, and endosteal bone resorption leads to scalloping of the inner cortex, cortical thinning, and widening of the medullary canal [69]. Subchondral bone resorption frequently also affects the joints of the axial skeleton, causing undermineralization of the tela ossea. For example, it may mimic widening of the sacroiliac joint space, leading to “pseudo-widening” of the

joint [70]. The osseous surface may collapse, and thus may simulate subchondral lesions of inflammatory disease. Osteopenia occurs frequently in hyperparathyroidism and may be observed throughout the skeleton. Other radiographic signs of hyperparathyroidism include focal bone lesions (“brown tumors”), cartilage calcification and also bone sclerosis (Figs. 8 and 9) [71]. Increased amounts of trabecular bone leading to bone sclerosis may occur especially in patients with renal osteodystrophy and secondary hyperparathyroidism [72,73]. Increased bone density may occur especially in the axial skeleton, sometimes leading to deposition of bone in subchondral areas of the vertebral body, resulting in an appearance of radiodense bands across the superior and inferior border and normal or decreased density of the center (rugger-jersey spine) [74]. While osteoporosis is defined by a reduction of regularly mineralized osteoid, findings in osteomalacia include an abnormally high amount of nonmineralized osteoid and a reduction in mineralized bone volume. Thus, radiographic abnormalities in osteomalacia include osteopenia (reduction of mineralized bone), coarsened, indistinct tra-

419

CHAPTER 58 Imaging of Osteoporosis

FIGURE 8

FIGURE 7 Primary hyperparathyroidism, subperiosteal and intracortical bone resorption. Radiograph of the middle phalanx shows irregular outline and spiculations at the radial aspect of the cortex (white arrows) resulting from subperiosteal bone resorption. Cortical tunneling indicating intracortical bone resorption is additionally present (black arrows).

beculae and unsharp delineation of cortical bone (excessive apposition of non-mineralized osteoid), deformities, insufficiency fractures and true fractures (bone softening and weakening) [75]. The deformations include bowing and bending of the long bones, and biconcave deformities of the vertebrae (Fig. 10) [76]. Pseudofractures, or Looser’s zones (focal accumulations of osteoid in compact bone at right angles of the long axis), are diagnostic of osteomalacia and are often bilateral and symmetrical. More than 50 different diseases may cause osteomalacia, of which chronic renal insufficiency and dietary deficiency are the most common [77,78]. One has to say though that modern patient management has resulted in typical radiographic features of osteomalacia being present

Brown tumor in secondary hyperparathyroidism. Radiograph of the lower leg demonstrates two lytic bone lesions in the tibia and in the fibula (long arrow). The lesions do not show sclerotic borders. The adjacent cortex is thinned and remodeled (short arrows), indicating the expansile nature of the lesion.

in only a minority of these patients [79]. A decrease of vitamin D and reduced responsiveness in chronic renal insufficiency leads to osteomalacia and rickets. The additional secondary hyperparathyroidism leads to a superimposition of radiographic changes of both osteomalacia and secondary hyperparathyroidism [80]. This radiographic appearance is termed renal osteodystrophy. A common finding in secondary hyperparathyroidism associated with renal osteodystrophy is the osteosclerosis resulting in typical appearance of the vertebral bodies as seen in the rugger-jersey spine (Fig. 11) [77]. Several other radiographic abnormalities may be frequently seen in renal osteodystrophy including amyloid deposits, destructive spondylarthropathy, juxtaarticular bone resorption, and avascular necrosis, soft tissue calcification and arteriosclerosis [81,82]. Hyperthyroidism is a high-turnover disease, and it is associated with an increase in both bone resorption and

420

JERGAS AND GENANT

FIGURE 9 Primary hyperparathyroidism and calcium pyrophosphate dihydrate deposition disease (CPPD). This radiograph of the left hand demonstrates findings of primary hyperparathyroidism such as increased radiolucency, cortical thinning, cortical tunneling and subperiosteal resorption. Additionally, calcifications of the triangular fibrocartilage (black arrow) and periarticular soft tissues (white arrows) are seen representing deposition of calcium pyrophosphate dihydrate crystals. bone formation [83]. Since bone resorption exceeds bone formation, rapid bone loss may occur and result in generalized osteoporosis with the largest effect on cortical bone [84 – 86]. This effect is especially pronounced in patients with thyrotoxicosis, or with a history of thyrotoxicosis [85,87]. TSH-suppressive doses of thyroid hormone have been reported to decrease, or have no effect on bone density [88,89]. Radiological findings of hyperthyroidism-induced osteoporosis are those that are commonly seen in involutional or senile osteoporosis including generalized osteopenia and cortical thinning and tunneling. The fractures associated with this condition affect the spine, the hip, and the distal radius [90,91] (see Chapter 47).

G. Medication-Induced Osteoporosis Hypercortisolism is probably the most common cause of medication induced generalized osteoporosis while the

endogenous form of hypercortisolism, Cushing’s disease, is relatively rare [92 – 95]. That is why this form of osteoporosis is listed in this section on medication-induced osteoporosis. Decreased bone formation and increased bone resorption have been observed in hypercortisolism. This has been attributed to inhibition of osteoblast formation, either direct stimulation of osteoclast activity or increased secretion of parathyroid hormone. The typical radiographic appearance of steroid-induced osteoporosis comprises generalized osteoporosis, at predominantly trabecular sites, with decreased bone density and fractures of the axial but also of the appendicular skeleton. A characteristic finding in steroid-induced osteoporosis is the marginal condensation of the vertebral bodies resulting from exuberant callus formation (Fig. 12). Osteonecrosis is another complication of hypercortisolism, most frequently involving the femoral head, and to a lesser extent the humeral head and the femoral condyles [96,97].

CHAPTER 58 Imaging of Osteoporosis

FIGURE 10 Osteomalacia. Lateral radiograph of the lumbar spine demonstrating increased radiolucency, picture framing, and vertebral deformity (fish vertebrae). In contrast to postmenopausal or senile osteoporosis, the trabecule have a fuzzy and indistinct apperance. Generalized osteoporosis has been observed in patients receiving high-dose heparin therapy [98 – 100]. The radiological features of heparin-induced osteoporosis include generalized osteopenia and vertebral compression fractures (Fig. 13) [101]. The mechanism of heparin-induced osteoporosis is not completely clear, and there may be a prolonged effect on bone even after cessation of therapy [102,103].

H. Other Causes of Generalized Osteoporosis Other causes of generalized osteoporosis include malnutrition, chronic alcoholism (if associated with malnutrition), smoking and caffeine intake, Marfan syndrome, and rather uncommonly pregnancy [104 – 108]. Marrow abnormalities associated with osteoporosis are anemias (sickle cell anemia, thalassemia), plasma cell myeloma, leukemia, Gaucher’s disease, and glycogen storage disease (Fig. 14) [109,110]. This list is far from complete but it includes some of the major causes of osteoporosis. Additional imaging techniques such as computed tomog-

421

FIGURE 11 Osteomalacia in chronic renal failure. In addition to the fuzzy and coarsened trabeculae there is subtle sclerosis of the vertebral end plates, resulting in transverse striation of the vertebral bodies (ruggerjersey spine).

raphy, magnetic resonance tomography, and bone scintigraphy as well as clinical information may be helpful in differential diagnosis of the various conditions associated with osteoporosis [111 – 115]. There are some conditions of the juvenile skeleton that result in generalized osteoporosis. Rickets is characterized by inadequate mineralization of the bone matrix, and some of its radiographic appearance may resemble that of osteomalacia [116]. Widening of the growth plates, cupping of the metaphysis, and decreased density and irregularities of the metaphyseal margins may be present [117]. Epiphyseal ossification centers may show delayed ossification and unsharp borders [118]. Overgrowth of the hyaline cartilage may lead to prominence of costochondral junctions of the ribs (rachitic rosary). The child’s age at the onset of the disease determines the pattern of bone deformity, with bowing of the long bone being more pronounced in infancy and early childhood and vertebral deformities and scoliosis in older children [119]. Further deformities that may be observed in rickets include pseudofractures, basilar invagination, and triradiate configuration of the pelvis (Fig. 15).

422

JERGAS AND GENANT

tients with type III disease have a significantly decreased bone density presenting with generalized osteopenia, thinned cortices, fractures of long bones and ribs, exuberant callus formation, and bone deformation (Fig. 16) [123]. The degree of osteopenia is highly variable, however, and at the mildest end of the spectrum some patients do not have any radiographic signs of osteopenia [124].

IV. REGIONAL OSTEOPOROSIS

FIGURE 12 Steroid-induced osteoporosis, marginal condensation. Lateral radiograph of the lumbar spine demonstrates increased radiolucency, wedge deformity of the first lumbar vertebra and sclerosis of the vertebral bodies adjacent to end plates (arrows). These “marginal condensations” result from exuberant callus formation in response to microfractures. Idiopathic juvenile osteoporosis is a self-limited disease of childhood with recovery occurring as puberty progresses [120]. A typical feature of this condition is the increased vulnerability of the metaphyses, often resulting in metaphyseal injuries of the vertebrae, knees, and ankles. Idiopathic juvenile osteoporosis must be distinguished from osteogenesis imperfecta, another disease often presenting with radiographic signs of generalized osteoporosis [121]. The pathogenesis of osteogenesis imperfecta involves quantitative or qualitative abnormalities of type I collagen. Osteogenesis imperfecta is divided into four major types, and the degree of osteoporosis in osteogenesis imperfecta depends strongly on the type of disease [122]. The clinical features of each type usually correspond to the type of mutation. The abnormal maturation of collagen seen in this disorder results in a primary defect in bone matrix. This, combined with a defective mineralization, results in overall loss of bone density involving both the axial and the peripheral skeleton. Pa-

Osteoporosis may also be confined to only a region of the body. This type of osteoporosis is called regional osteoporosis, and it is commonly caused by some disorder of the appendicular skeleton. Osteoporosis due to immobilization or disuse characteristically occurs in the immobilized regions of patients with fractures, motor paralysis due to central nervous system disease or trauma, and bone and joint inflammation [125]. Chronic and acute disease may vary in their radiographic appearance somewhat, showing diffuse osteopenia, linear radiolucent bands, speckled radiolucent areas, and cortical bone resorption. Reflex sympathetic dystrophy, sometimes also termed Sudeck’s atrophy or algodystrophy, has the radiographic appearance of a high bone turnover process. It most often occurs in patients with trauma, such as Colles’ fracture, but also in patients with any neurally related musculoskeletal, neurologic, or vascular condition such as hemiplegia or myocardial infarction [126 – 129]. This condition is probably related to overactivity of the sympathetic nervous system with increased blood flow and increased venous oxygen saturation in the affected extremity [130,131]. Its radiographic appearance includes soft tissue swelling as well as regional osteoporosis, showing with band-like, patchy, or periarticular osteoporosis (Fig. 17). Additional radiographic features include intracortical tunneling, endosteal bone resorption with initial excavation and scalloping of the endosteal surface and subsequent remodeling and widening of the medullary canal, as well as subchondral and juxtaarticular erosions (Fig. 17) [132]. Especially in the early stages of reflex sympathetic dystrophy, bone scintigraphy may be helpful to establish the diagnosis [133 – 135]. Transient regional osteoporosis includes conditions that have in common the development of self-limited pain and radiographic osteopenia affecting one or several major weight-bearing joints, most commonly the hip. Transient osteoporosis typically occurs in middleaged men and in women at the hip in the third trimester of pregnancy [136,137]. At the onset of clinical symptoms, there may be normal radiographic findings, and

423

CHAPTER 58 Imaging of Osteoporosis

FIGURE 13 Heparin-induced osteoporosis. Lateral radiograph of the lumbar spine (A) shows markedly increased radiolucency and picture framing of the vertebral bodies. Also note a moderate wedge deformity of L2. A vena cava filter is placed anterior to L3. The follow-up radiograph (B) shows severe progression of vertebral fractures, mainly affecting L1, L2, and L3.

within several weeks, patients develop variable osteopenia of the hip, sometimes involving the acetabulum. Some patients later develop similar changes in the opposite hip or in other joints, in which case the term regional migratory osteoporosis may be used [138]. No specific therapy is required, since patients recover. The cause of transient regional osteoporosis is not known, and it appears that it may be related to reflex sympathetic dystrophy. In some patients with clinically similar or identical manifestations, magnetic resonance imaging presents with transient regional bone marrow edema

[139,140]. Since not all patients with identical clinical symptoms and transient bone marrow edema develop regional osteoporosis, the sensitivity as to the detection of regional osteoporosis has to be questioned, as well as the interrelationship between transient regional osteoporosis and transient bone marrow edema. There also seems to be a relationship of transient bone marrow edema to ischemic necrosis of bone, and there is a need to define criteria for allowing differentiation of transient bone marrow edema and the edema pattern associated with osteonecrosis [141 – 143].

424

JERGAS AND GENANT

FIGURE 14 Multiple myeloma. Lateral radiograph of the lumbar spine (A) demonstrates severe diffuse osteopenia and moderate to severe vertebral fractures of L1 to L3. There are no focal areas of destruction. The radiograph of the forearm (B) demonstrates numerous, well demarcated, lytic defects in the humerus, the radius, and the ulna.

CHAPTER 58 Imaging of Osteoporosis

FIGURE 15

Rickets. Anteroposterior radiograph of the legs (A) and lateral radiograph of the left tibia and fibula (B) demonstrate markedly increased radiolucency, thinning and tunneling of the cortex, and severe bowing deformities of long bones. The growth plates are widened and the metaphyses are cupped and irregularly delineated (short arrows). In the proximal diaphysis of the left tibia a transverse lucency of the posteromedial cortex is seen (long arrow) which is surrounded by sclerotic bone (pseudofracture).

425

426

JERGAS AND GENANT

FIGURE 16

Osteogenesis imperfecta tarda. Lateral radiograph of the tibia and fibula demonstrates increased radiolucency, cortical thinning, and severe bowing deformities. Increased density in the distal diaphysis of the tibia indicates healing of a fracture (arrow).

CHAPTER 58 Imaging of Osteoporosis

FIGURE 17

Reflex sympathetic dystrophy. Radiograph of the left foot (A) demonstrates inhomogenously increased radiolucency (best seen around joints, white arrows), and subcortical radiolucent bands (black arrows). Computed tomography of the same patient (B) demonstrates subcortical radiolucent bands (arrows) and numerous spheroid lucent lesions in the cortical and trabecular bone.

427

428

JERGAS AND GENANT

References 1. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis. Its clinical features. JAMA 116(22), 2465 – 2474 (1941). 2. R. Z. LeGeros, Biological and synthetic apatites. In “Hydroxyapatite and Related Materials” (P. W. Brown, and B. Constantz, eds.), pp. 3 – 28. CRC Press, Boca Raton, (1994). 3. H. M. Frost, Dynamics of bone remodelling. In “Bone biodynamics.” (H. M. Frost, ed.), pp. 315 – 334. Little Brown, Boston (1964). 4. H. K. Genant, K. Doi, J. C. Mall, and E. A. Sickles, Direct radiographic magnification for skeletal radiology. Radiology 123(1), 47 – 55 (1977). 5. H. K. Genant, K. Doi, and J. C. Mall, Comparison of non-screen techniques (medical vs. industrial film) for fine-detail skeletal radiography. Invest Radiol. 11(5), 486 – 500 (1976). 6. T. M. Link, E. J. Rummeny, H. Lenzen, I. Reuter, N. Roos and P. E. Peters, Artificial bone erosions: Detection with magnification radiography versus conventional high resolution radiography. Radiology 192, 861 – 864 (1994). 7. F. Heuck and E. Schmidt, Die quantitative Bestimmung des Mineralgehaltes des Knochens aus dem Röntgenbild. Fortschr. Röntgenstr. 93, 523 – 554 (1960). 8. E. Lachmann and M. Whelan, The roentgen diagnosis of osteoporosis and its limitations. Radiology 26, 165 – 177 (1936). 9. M. Jergas, M. Uffmann, H. Escher, J. Schaffstein, E. Nitzschke, and O. Köster, Visuelle Beurteilung konventioneller Röntgenaufnahmen und duale Röntgenabsorptiometrie in der Diagnostik der Osteoporose. Z. Orthop. Grenzgeb. 132(2), 91 – 98 (1994). 10. F. Albright, Osteoporosis. Ann. Intern. Med. 27, 861 – 882 (1947). 11. B. L. Riggs and L. J. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1983). 12. J. C. Gallagher, The pathogenesis of osteoporosis. Bone Miner. 9, 215 – 227 (1990). 13. P. D. Saville, A quantitative approach to simple radiographic diagnosis of osteoporosis: Its application to the osteoporosis of rheumatoid arthritis. Arthritis Rheum. 10, 416 – 422 (1967). 14. F. H. Doyle, D. H. Gutteridge, G. F. Joplin, and R. Fraser, An assessment of radiological criteria used in the study of spinal osteoporosis. Br. J. Radiol. 40, 241 – 250 (1967). 15. M. Jergas, M. Uffmann, H. Escher, C. C. Glüer, K. C. Young, S. Grampp, et al., Interobserver variation in the detection of osteopenia by radiography and comparison with dual X-ray absorptiometry (DXA) of the lumbar spine. Skeletal Radiol. 23(3), 195 – 199 (1994). 16. A. I. H. Ahmed, D. Ilic, G. M. Blake, J. M. Rymer and I. Fogelman, Review of 3530 referrals for bone density measurements of spine and femur: Evidence that radiographic osteopenia predicts low bone mass. Radiology 207, 619 – 624 (1998). 17. H. K. Genant, C. Y. Wu, C. van Kuijk and M, Nevitt, Vertebral fracture assessment using a semi-quantitative technique. J. Bone Miner. Res. 8(9), 1137 – 1148 (1993). 18. P. C. McAfee, H. A. Yuan, B. E. Fredrickson, and J. P. Lubicky, The value of computed tomography in thoracolumbar fractures. An analysis of one hundred consecutive cases and a new classification. J. Bone Jt. Surg. Am. 65(4), 461 – 473 (1983). 19. S. E. Campbell, C. D. Phillips, E. Dubovsky, W. S. Cail, and R. A. Omary, The value of CT in determining potential instability of simple wedge-compression fractures of the lumbar spine. Am. J. Neuroradiol. 16(7), 1385 – 1392 (1995). 20. R. T. Ballock, R. Mackersie, J. J. Abitbol, V. Cervilla, D. Resnick, and S. R. Garfin, Can burst fractures be predicted from plain radiographs? J. Bone Jt. Surg. Br 74(1), 147 – 150 (1992). 21. E. V. McCloskey, T. D. Spector, K. S. Eyres, E. D. Fern, N. O’Rourke, S. Vasikaran, et al., The assessment of vertebral deformity: A method for use in population studies and clinical trials. Osteoporosis Int. 3(3), 138 – 147 (1993).

22. H. W. Minne, G. Leidig, C. Wüster, L. Siromachkostov, G. Baldauf, R. Bickel, et al., A newly developed spine deformity index (SDI) to quantitate vertebral crush fractures in patients with osteoporosis. Bone Miner. 3, 335 – 349 (1988). 23. M. Jergas and R. San Valentin, Techniques for the assessment of vertebral dimensions in quantitative morphometry. In “Vertebral Fracture in Osteoporosis” (H. K. Genant, M. Jergas, and C. van Kuijk, eds.), pp. 163 – 88. San Francisco, Radiology Research and Education Foundation, (1995). 24. D. Black, L. Palermo, M. C. Nevitt, H. K. Genant. R. Epstein, R. San Valentin, et al., Comparison of methods for defining prevalent vertebral deformities: The study of osteoporotic fractures. J. Bone Miner. Res. 10(6), 890 – 902 (1995). 25. C. Y. Wu, J. Li, M. Jergas, and H. K. Genant, Comparison of semiquantitative and quantitative techniques for the assessment of prevalent and incident vertebral fractures. Osteoporosis Int. 5(5), 354 – 370 (1995). 26. S. R. Cummings, L. J. Melton III, and D. Felsenberg, National Osteoporosis Foundation Working Group on Vertebral Fracture. Report: Assessing vertebral fractures. J. Bone Miner. Res. 10(4), 518 – 523 (1995). 27. H. K. Genant, M. Jergas, L. Palermo, M. Nevitt, R. San Valentin, D. Black, et al., Comparison of semiquantitative visual and quantitative morphometric assessment of prevalent and incident vertebral fractures in osteoporosis. J. Bone Miner. Res. 11(7), 984 – 996 (1996). 28. M. Jergas, T. F. Lang, and T. Fuerst, Morphometric X-ray absorptiometry. In “Vertebral Fracture in Osteoporosis” (H. K. Genant, M. Jergas, C. van Kuijk, eds.), pp. 331 – 348. Radiology Research and Education Foundation, San Francisco (1995). 29. D. Felsenberg, W. Gowin, E. Diessel, S. Armbrust, and J. Mews, Recent developments in DXA. Quality of new DXA/MXA-devices for densitometry and morphometry. Eur. J. Radiol. 20(3), 179 – 184 (1995). 30. J. A. Rea, P. Steiger, G. M. Blake, E. Potts, I. G. Smith, and I. Fogelman, Morphometric X-ray absorptiometry: Reference data for vertebral dimensions. J. Bone Miner. Res. 13(3), 464 – 474 (1998). 31. C. Chappard, S. Kolta, J. Fechtenbaum, M. Dougados, and C. Roux, Clinical evaluation of spine morphometric X-ray absorptiometry. Br. J. Rheumatol. 37(5), 496 – 501 (1998). 32. L. Ferrar and R. Eastell, Identification of vertebral deformities in men: Comparison of morphometric radiography and morphometric X-ray absorptiometry. Osteoporosis Int. 10(2), 167 – 174 (1999). 33. L. Ferrar, G. Jiang, N. A. Barrington, and R. Eastell, Identification of vertebral deformities in women: comparison of radiological assessment and quantitative morphometry using morphometric radiography and morphometric X-ray absorptiometry. J. Bone Miner. Res. 15(3), 575 – 585 (2000). 34. J. A. Rea, M. B. Chen, J. Li, E. Potts, B. Fan, G. M. Blake, et al. Morphometric X-ray absorptiometry and morphometric radiography of the spine: A comparison of analysis precision in normal and osteoporotic subjects. Osteoporosis Int. 9(6), 536 – 544 (1999). 35. J. A. Rea, M. B. Chen, J. Li, G. M. Blake, P. Steiger, H. K. Genant, et al., Morphometric X-ray absorptiometry and morphometric radiography of the spine: A comparison of prevalent vertebral deformity identification. J. Bone Miner. Res. 15(3), 564 – 574 (2000). 36. Y. M. Singh, A. R. Nagrath, and P. S. Maini, Changes in trabecular pattern of the upper end of the femur as an index of osteoporosis. J. Bone Jt. Surg. 52-A(3), 457 – 467 (1970). 37. Y. M. Singh, B. L. Riggs, J. W. Beabout, and J. Jowsey, Femoral trabecular-pattern index for evaluation of spinal osteoporosis. Ann. Intern. Med. 77, 63 – 67 (1972). 38. N. L. Jhamaria, K. B. Lal, M. Udawat, P. Banerji, and S. G. Kabra, The trabecular pattern of the calcaneum as an index of osteoporosis. J. Bone Jt. Surg. 65-B(2), 195 – 198 (1983).

CHAPTER 58 Imaging of Osteoporosis 39. E. Barnett and B. E. C. Nordin, The radiological diagnosis of osteoporosis: A new approach. Clin. Radiol. 11, 166 – 174 (1960). 40. E. Barnett and B. E. C. Nordin, Radiological assessment of bone density. I. The clinical and radiological problem of thin bones. Br. J. Radiol. 34, 683 – 692 (1961). 41. H. E. Meema and S. Meema, Radiogrammetry. In “Non-invasive Measurements of Bone Mass” (S. H. Cohn, ed.), pp. 5 – 50. CRC Press, Boca Raton (1981). 42. U. E. Ruttimann, R. L. Webber, and J. B. Hazelrig, Fractal dimension from radiographs of peridental alveolar bone. A possible diagnostic indicator of osteoporosis. Oral Surg. Oral Med. Oral Pathol. 74(1), 98 – 110 (1992). 43. C. L. Benhamou, E. Lespessailles, G. Jacquet, R. Harba, R. Jennane, T. Loussot, et al., Fractal organization of trabecular bone images on calcaneus radiographs. J. Bone Miner. Res. 9(12), 1909 – 1918 (1994). 44. W. G. M. Geraets, P. F. Van der Stelt, and P. J. M. Elders, The radiographic trabecular bone pattern during menopause. Bone 14, 859 – 864 (1993). 45. W. Geraets, P. Van der Stelt, P. Lips, P. Elders, F. Van Ginkel, and E. Burger, Orientation of the trabecular pattern of the distal radius around the menopause. J. Biomech. 30(4), 363 – 370 (1997). 46. W. Geraets, P. Van der Stelt, P. Lips, Ginkel, and F. Van, The radiographic trabecular pattern of hips in patients with hip fractures and in elderly control subjects. Bone 22(2), 165 – 173 (1998). 47. T. Link, S. Majumdar, W. Konermann, N. Meier, J. Lin, D. Newitt, et al., Texture analysis of direct magnification radiographs of vertebral specimens: Correlation with bone mineral density and biomechanical properties. Acad. Radiol. 4(3), 167 – 176 (1997). 48. E. Lespessailles, J. P. Roux, C. L. Benhamou, M. E. Arlot, E. Eynard, R. Harba, et al., Fractal analysis of bone texture on os calcis radiographs compared with trabecular microarchitecture analyzed by histomorphometry. Calcif. Tissue Int. 63(2), 121 – 125 (1998). 49. S. Majumdar, D. C. Newitt, M. Jergas, A. A. Gies, E. C. Chiu, D. Osman, et al., Evaluation of technical factors affecting the quantification of trabecular bone structure using magnetic resonance imaging. Bone 17, 417 – 130 (1995). 50. S. Majumdar, D. Newitt, A. Mathur, D. Osman, A. Gies, E. Chiu, et al., Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with x-ray tmographic microscopy and biomechanics. Osteoporosis Int. 6(5), 376 – 385 (1996). 51. S. Majumdar, H. K. Genant, S. Grampp, D. C. Newitt, V-H. Truong, J. C. Lin, et al., Correlation of trabecular bone structure with age, bone mineral density, and osteoporotic status: In vivo studies in the distal radius using high resolution magnetic resonance imaging. J. Bone Miner. Res. 12(1), 111 – 118 (1997). 52. T. M. Link, S. Majumdar, J. C. Lin, D. Newitt, P. Augat, X. Ouyang, et al., A comparative study of trabecular bone properties in the spine and femur using high resolution MRI and CT. J. Bone Miner. Res. 13(1), 122 – 132 (1998). 53. T. M. Link, S. Majumdar, J. C. Lin, P. Augat, R. G. Gould, D. Newitt, et al., Assessment of trabecular structure using high resolution CT images and texture analysis. J. Comput. Assist. Tomogr. 22(1), 15 – 24 (1998). 54. S. Majumdar, M. Kothari, P. Augat, D. C. Newitt, T. M. Link, J. C. Lin, et al., High-resolution magnetic resonance imaging: Threedimensional trabecular bone architecture and biomechanical properties. Bone 22(5), 445 – 454 (1998). 55. T. M. Link, S. Majumdar, P. Augat, J. C. Lin, D. Newitt, Y. Lu, et al., In vivo high resolution MRI of the calcaneus: Differences in trabecular structure in osteoporosis patients. J. Bone Miner. Res. 13(7), 1175 – 1182 (1998). 56. X. Ouyang, S. Majumdar, T. M. Link, Y. Lu, P. Augat, J. Lin, et al., Morphometric texture analysis of spinal trabecular bone structure assessed using orthogonal radiographic projections. Med. Phys. 25(10), 2037 – 2045 (1998).

429 57. T. M. Link, S. Majumdar, P. Augat, J. C. Lin, D. Newitt, N. E. Lane, et al., Proximal femur: Assessment for osteoporosis with T2* decay characteristics at MR imaging. Radiology 209(2), 531 – 536 (1998). 58. J. Millard, P. Augat, T. M. Link, M. Kothari, D. C. Newitt, H. K. Genant, et al., Power spectral analysis of vertebral trabecular bone structure from radiographs: Orientation dependence and correlation with bone mineral density and mechanical properties. Calcif. Tissue Int. 63(6), 482 – 489 (1998). 59. T. M. Link, J. C. Lin, D. Newitt, N. Meier, S. Waldt, and S. Majumdar, [Computer-assisted structure analysis of trabecular bone in the diagnosis of osteoporosis]. Radiologe 38(10), 853 – 859 (1998). 60. S. Majumdar, T. M. Link, P. Augat, J. C. Lin, D. Newitt, N. E. Lane, et al., Trabecular bone architecture in the distal radius using magnetic resonance imaging in subjects with fractures of the proximal femur. Magnetic Resonance Science Center and Osteoporosis and Arthritis Research Group. Osteoporosis Int. 10(3), 231 – 239 (1999). 61. A. Laib and P. Ruegsegger, Comparison of structure extraction methods for in vivo trabecular bone measurements. Comput. Med. Imaging Graph. 23(2), 69 – 74 (1999). 62. B. Cortet, P. Dubois, N. Boutry, P. Bourel, A. Cotten, and X. Marchandise, Image analysis of the distal radius trabecular network using computed tomography. Osteoporosis Int. 9(5), 410 – 419 (1999). 63. H. K. Genant, L. L. Heck, L. H. Lanzl, K. Rossmann, J. Vander Horst, and E. Paloyan, Primary hyperparathyroidism. A comprehensive study of clinical, biochemical and radiographic manifestations. Radiology 109, 513 – 519 (1973). 64. H. K. Genant, J. Vander Horst, L. H. Lanzl, J. C. Mall, and K. Doi, Skeletal demineralization in primary hyperparathyroidism. In “Proceedings of International Conference on Bone Mineral Measurement” (R. B. Mazess ed.), p. 177. Washington, DC, National Institute of Arthritis, Metabolism and Digestive Diseases (1974). 65. M. L. Richardson, R. S. Pozzi-Mucelli, A. S. Kanter, F. O. Kolb, B. Ettinger and H. K. Genant, Bone mineral changes in primary hyperparathyroidism. Skeletal Radiol. 15, 85 – 95 (1986). 66. J. D. Camp and H. C. Ochsner, The osseous changes in hyperparathyroidism associated with parathyroid tumor: A roentgenologic study. Radiology 17, 63 (1931). 67. D. Resnick and G. Niwayama, Parathyroid disorders and renal osteodystrophy. In “Diagnosis of Bone and Joint Disorders” (D. Resnick, ed.), 3rd ed., pp. 2012 – 2075. Saunders, Philadelphia, (1995). 68. F. H. W. Heuck, Endokrine, metabolische and medikamentös induzierte Knochen- und Gelenkerkrankungen. In “Radiologie der Knochen- und Gelenkerkrankungen” (A. Heuck, ed.), pp. 51 – 150. Georg Thieme Verlag, Stuttgart (1998). 69. H. E. Meema and S. Meema, Microradioscopic and morphometric findings in the hand bones with densitometric findings in the proximal radius in thyrotoxicosis and in renal osteodystrophy. Invest. Radiol. 7, 88 (1972). 70. C. W. Hayes and W. F. Conway, Hyperparathyroidism. Radiol. Clin. North Am. 29(1), 85 – 96 (1991). 71. H. L. Steinbach, G. S. Gordan, E. Eisenberg, J. T. Carne, S. Silverman, and L. Goldman, Primary hyperoarthyroidism: A correlation of roentgen, clinical, and pathologic features. Am J. Roentgenol. Radium Ther. Nucl. Med. 86, 239 – 243 (1961). 72. T. Crawford, C. E. Dent, and P. Lucas, Osteosclerosis associated with chronic renal failure. Lancet 2, 981 (1954). 73. J. G. Davis, Osseous radiographic findings of chronic renal insufficiency. Radiology 60, 406 (1953). 74. D. Resnick, The “rugger jersey” vertebral body. Arthritis Rheum. 24(9), 1191 – 1194 (1981). 75. A. J. Reginato, G. F. Falasca, R. Pappu, B. McKnight, and A. Agha, Musculoskeletal manifestations of osteomalacia: Report of 26 cases and literature review. Semin. Arthritis Rheum. 28(5), 287 – 304 (1999).

430 76. R. Kienböck, Osteomalazie, Osteoporose, Osteopsathyrose, porotische Kyphose. Fortschr. Röntgenstr. 61, 159 (1940). 77. M. J. Pitt, Rickets and osteomalacia are still around. Radiol. Clin. North Am. 29(1), 97 – 118 (1991). 78. F. Kainberger, O. Traindl, M. Baldt, T. Helbich, M. Breitenseher, G. Seidl, et al., renale Osteodytrophie: Spektrum der Röntgensymptomatik bei modernen Formen der Nierentransplantation und Dauerdialysetherapie. Fortschr. Röntgenstr. 157, 501 – 505 (1992). 79. J. E. Adams, Renal bone disease: radiological investigation. Kidney Int. Suppl. 73, S38 – S41 (1999). 80. M. Sundaram, Renal osteodystrophy. Skeletal Radiol. 18, 415 – 426 (1989). 81. J. S. Kriegshauser, R. G. Swee, J. T. McCarthy, and M. F. Hauser, Aluminum toxicity in patients undergoing dialysis: Radiographic findings and prediction of bone biobsy results. Radiology 164, 399 – 403 (1987). 82. M. D. Murphey, D. J. Sartoris, J. L. Quale, M. N. Pathria, and N. L. Martin, Musculoskeletal manifestations of chronic renal insufficiency. Radiographics 13(2), 357 – 379 (1993). 83. L. Mosekilde, E. F. Eriksen, and P. Charles, Effects of thyroid hormones on bone and mineral metabolism. Endocrinol. Metab. Clin. North Am. 19, 35 – 63 (1990). 84. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr. Rev. 7, 379 – 408 (1986). 85. S. H. Toh, B. C. Claunch, and P. H. Brown, Effect of hyperthyroidism and its treatment on bone mineral content. Arch. Intern. Med. 145, 883 – 886 (1985). 86. S. L. Greenspan and F. S. Greenspan, The effect of thyroid hormone on skeletal integrity. Ann. Intern. Med. 130(9), 750 – 758 (1999). 87. J. Linde and T. Friis, Osteoporosis in hyperthyroidism estimated by photon absroptiometry. Acta Endocrinol. 91, 437 – 448 (1979). 88. B. Krlner, J. V. Jrgensen, and S. P. Nielsen, Spinal bone mineral content in myxoedema and thyrotoxicosis. Effects of thyroid hormone(s) and antithyroid treatment. Clin. Endocrinol. 18, 439 – 446 (1983). 89. V. Nuzzo, G. Lupoli, A. Esposito Del Puente, E. Rampone, A. Carpinelli, and A. E. Del Puente et al., Bone mineral density in premenopausal women receiving levothyroxine suppressive therapy. Gynecol. Endocrinol. 12(5), 333 – 337 (1998). 90. F. S. Chew, Radiologic manifestations in the musculoskeletal system of miscellaneous endocrine disorders. Radiol. Clin. North Am. 29, 135 – 147 (1991). 91. B. L. Solomon, L. Wartofsky, and K. D. Burman, Prevalence of fractures in postmenopausal women with thyroid disease. Thyroid 3, 17 – 23 (1993). 92. R. F. Laan, W. C. Buijs, L. J. van Erning, J. A. Lemmens, F. H. Corstens, S. H. Ruijs, et al., Differential effects of glucocorticoids on cortical appendicular and cortical vertebral bone mineral content. Calcif. Tissue Int. 52(1), 5 – 9 (1993). 93. D. W. Brandli, G. Golde, M. Greenwald, and S. L. Silverman, Glucocorticoid-induced osteoporosis: a cross-sectional study. Steroids 56(10), 518 – 523 (1991). 94. J. K. Saito, J. W. Davis, R. D. Wasnich, and P. D. Ross, Users of low-dose glucocorticoids have increased bone loss rates: A longitudinal study. Calcif. Tissue Int. 57(2), 115 – 119 (1995). 95. J. D. Adachi, W. G. Bensen, and A. B. Hodsman, Corticosteroid-induced osteoporosis. Semin. Arthritis Rheum. 22(6), 375 – 384 (1993). 96. W. G. Heimann and R. H. Freiberger, Avascular necrosis of the femoral and humeral heads after heigh-dosage corticosteroid therapy. N. Engl. J. Med. 263, 672 – 674 (1969). 97. S. J. Hurel and P. Kendall-Taylor, Avascular necrosis secondary to postoperative steroid therapy. Br. J. Neurosurg. 11(4), 356 – 358 (1997).

JERGAS AND GENANT 98. G. C. Griffith, G. Nichols, J. D. Ashey, and B. Flannagan, Heparin osteoporosis. JAMA 193(2), 85 – 88 (1965). 99. W. M. Rupp, H. B. McCarthy, T. D. Rohde, P. J. Blackshear, F. J. Goldenberg, and H. Buchwald, Risk of osteoporosis in patients treated with long-term intravenous heparin therapy. Curr. Surg. 39, 419 – 22 (1982). 100. C. Nelson-Piercy, Heparin-induced osteoporosis. Scand. J. Rheumatol. Suppl. 107, 68 – 71 (1998). 101. J. P. Sackler and L. Liu, Heparin-induced osteoporosis. Br. J. Radiol. 46, 548 – 540 (1973). 102. J. M. Walenga and R. L. Bick, Heparin-induced thrombocytopenia, paradoxical thromboembolism, and other side-effects of heparin therapy. Med. Clin. North Am. 82(3), 635 – 658 (1998). 103. S. G. Shaughnessy, J. Hirsh, M. Bhandari, J. M. Muir, E. Young, and J. I. Weitz, A histomorphometric evaluation of heparin-induced bone loss after discontinuation of heparin treatment in rats. Blood 93(4), 1231 – 1236 (1999). 104. E. Seeman, G. I. Szmukler, C. Formica, C. Tsalamandris, and R. Mestrovic, Osteoporosis in anorexia nervosa: The influence of peak bone density, bone loss, oral contraceptive use, and exercise. J. Bone Miner. Res. 7(12), 1467 – 1474 (1992). 105. L. Kohlmeier, C. Gasner, and R. Marcus, Bone mineral status of women with Marfan syndrome. Am. J. Med. 95, 568 – 572 (1993). 106. R. Smith, J. C. Stevenson, C. G. Winearls, C. G. Woode, and B. P. Wordsworth, Osteoporosis of pregnancy. Lancet, 1178 – 1180 (1985). 107. J. L. Hopper and E. Seeman, The bone density of twins discordant for tobacco use. N. Engl. J. Med. 330, 387 – 392 (1994). 108. A. Diez, J. Puig, S. Serrano, M. -L. Marinoso, J. Bosch, and J. Marrugat, Alcohol-induced bone disease in the absence of severe chronic liver damage. J. Bone. Miner. Res. 9, 825 – 831 (1994). 109. D. Resnick, Hemoglobinopathies and other anemias. In “Diagnosis of Bone and Joint Disorders” (D. Resnick, ed.), 3rd ed., pp. 2107 – 2146. Saunders, Philadelphia, 1995. 110. D. Resnick, Plasma cell dyscrasias and dysgammaglobulinemias. In “Diagnosis of Bone and Joint Disorders” (D. Resnick, ed.), 3rd ed., pp. 2147 – 2189. Saunders, Philadelphia (1995). 111. A. Stäbler, A. Baur, R. Bartl, R. Munker, R. Lamerz, and M. F. Reiser, Contrast enhancement and quantitative signal analysis in MR imaging of multiple myeloma: assessment of focal and diffuse growth patterns in marrow correlated with biopsies and survival rates. Am. J. Roentgenol. 167 (4), 1029 – 1036 (1996). 112. A. Baur, A. Stabler, M. Steinborn, P. Schnarkowski, C. Pistitsch, R. Lamerz, et al., Magnetresonanztomographie beim Plasmozytom: Wertigkeit verschiedener Sequenzen bei diffuser und lokaler Infiltrationsform. RöFo Fortschr Röntgenstr. 168(4), 323 – 329 (1998). 113. L. A. Moulopoulos and M. A. Dimopoulos, Magnetic resonance imaging of the bone marrow in hematologic malignancies. Blood 90(6), 2127 – 2147 (1997). 114. F. Lecouvet, J. Malghem, L. Michaux, J. Michaux, F. Lehmann, B. Maldague, et al., Vertebral compression fractures in multiple myeloma. Assessment of fracture risk with MR imaging of spinal bone marrow. Radiology 204(1), 201 – 205 (1997). 115. F. Lecouvet, B. Van de Berg, B. Maldague, L. Michaux, E. Laterre, J. Michaux, et al., Vertebral compression fractures in multiple myeloma. Distribution and appearance at MR imaging. Radiology 204(1), 195 – 199 (1997). 116. W. M. Molpus, R. S. Pritchard, C. W. Walker, and R. L. Fitzrandolph, The radiographic spectrum of renal osteodystrophy. Am. Fam. Physician 43(1), 151 – 158 (1991). 117. M. J. Pitt, Rickets and osteomalacia. In (D. Resnick, ed.), “Diagnosis of Bone and Joint Disorders” 3rd ed., pp. 1885 – 1922. Saunders, Philadelphia (1995). 118. H. L. Steinbach, F. O. Kolb, and R. Gilfillan, A mechanism of the production of pseudofractures in osteomalacia (Milkman’s syndrome). Radiology 62, 388 (1954).

CHAPTER 58 Imaging of Osteoporosis 119. A. E. Rosenberg, The pathology of metabolic bone disease. Radiol. Clin. North. Am. 29, 19 – 35 (1991). 120. R. Smith, Idiopathic juvenile osteoporosis: experience of twenty-one patients. Br. J. Rheumatol. 34(1), 68 – 77 (1995). 121. M. E. Norman, Juvenile osteoporosis. In “Primer on the Metabolic Diseases and Disorders of Mineral Metabolism,” (M. J. Favus, ed.), 3 ed., pp. 275 – 278. Lippincott-Raven, Philadelphia (1996). 122. C. M. Minch and R. W. Kruse, Osteogenesis imperfecta: A review of basic science and diagnosis. Orthopedics 21(5), 558 – 567 (1998). 123. D. A. Hanscom, R. B. Winter, L. Lutter, J. E. Lonstein, B. A. Bloom, and D. S. Bradford, Osteogenesis imperfecta. Radiographic classification, natural history, and treatment of spinal deformities. J. Bone. Jt. Surg. Am. 74(4), 598 – 616 (1992). 124. L. E. Zionts, J. P. Nash, R. Rude, T. Ross, and N. S. Stott, Bone mineral density in children with mild osteogenesis imperfecta. J. Bone. Jt. Surg. Br. 77(1), 143 – 147 (1995). 125. B. J. Kiratli, Immobilization osteopenia. In “Osteoporosis” (R. Marcus, D. Feldman and J. Kelsey, ed.), Academic Press, San Diego, pp. 833 – 853 (1996). 126. P. Sudbeck, Über die akute (reflectorische) Knochenatrophie nach entzündungen und Verletzungen an den Extremitäten und ihre klinischen Erscheinungen. RöFo 5, 277 (1901). 127. D. S. Miller and G. DeTakats, Post-traumatic dystrophy of the extremities: Sudeck’s atrophy. Surg. Gynecol. Obstet. 75, 558 (1942). 128. W. J. Oyen, I. E. Arntz, R. M. Claessens, J. W. Van der Meer, F. H. Corstens, and R. J. Goris, Reflex sympathetic dystrophy of the hand: An excessive inflammatory response? Pain 55(2), 151 – 157 (1993). 129. P. P. Sarangi, A. J. Ward, E. J. Smith, G. E. Staddon, and R. M. Atkins, Algodystrophy and osteoporosis after tibial fractures. J. Bone. Jt. Surg. Br. 75(3), 450 – 452 (1993). 130. H. Gellman, M. A. Keenan, L. Stone, S. E. Hardy, R. L. Waters, and C. Stewart, Reflex sympathetic dystrophy in brain-injured patients. Pain 51(3), 307 – 311 (1992). 131. R. J. Schwartzman and T. L. McLellan, Reflex sympathetic dystrophy: A review. Arch. Neurol. 44, 555 – 561 (1987). 132. D. Resnick and G. Niwayama, In “Diagnosis of Bone and Joint Disorders” 3rd ed., (D. Resnick, ed.), pp. 1783 – 1853. Saunders, Philadelphia (1995).

431 133. M. Todorovic Tirnanic, V. Obradovic, R. Han, B. Goldner, D. Stankovic, D. Sekulic, et al., Diagnostic approach to reflex sympathetic dystrophy after fracture: radiography or bone scintigraphy? Eur. J. Nucl. Med. 22(10), 1187 – 1193 (1995). 134. H. Steinert and K. Hahn, Wertigkeit der Drei Phasen Skelettszintigraphie zur Fruhdiagnostik des Morbus Sudeck. Röfo Fortschr. Geb. Röntgenstr. Neuen. Bildgeb. Verfahr. 164(4), 318 – 323 (1996). 135. T. Leitha, A. Staudenherz, M. Korpan, and V. Fialka, Pattern recognition in five-phase bone scintigraphy: Diagnostic patterns of reflex sympathetic dystrophy in adults. Eur. J. Nucl. Med. 23(3), 256 – 262 (1996). 136. R. A. Rosen, Transitory demineralization of the femoral head. Radiology 94, 509 – 512 (1970). 137. G. G. Hunder and P. J Kelly, Roentgenologic transient osteoporosis of the hip. Ann. Intern. Med. 68, 539 – 552 (1968). 138. R. C. Gupta, M. M. Popovtzer, W. E. Huffer, and C. J. Smyth, Regional migratory osteoporosis. Arthritis Rheum. 21, 363 – 368 (1973). 139. C. W. Hayes, W. F. Conway, and W. W. Daniel, MR imaging of bone marrow edema pattern: Transient osteoporosis, transient bone marrow edema syndrome, or osteonecrosis. Radiographics 13(5), 1001 – 1011 (1993). 140. S. Boos, G. Sigmund, P. Huhle, and I. Nurbakhsch, Magnetresonanztomographie der sogenannten transitorischen Osteoporose. Primärdiagnostik und Verlaufskontrolle nach Therapie. Röfo Fortschr. Geb. Rontgenstr. Neuen. Bildgeb. Verfahr. 158(3), 201 – 206 (1993). 141. E. Trepman, T. V. King, Transient osteoporosis of the hip misdiagnosed as osteonecrosis on magnetic resonance imaging. Orthop. Rev. 21(9), 1089 – 1091, 1094 – 1098 (1992). 142. P. K. Froberg, E. M. Braunstein, and K. A. Buckwalter, Osteonecrosis, transient osteoporosis, and transient bone marrow edema: Current concepts. Radiol. Clin. North. Am. 34(2), 273 – 291 (1996). 143. J. J. Guerra and M. E. Steinberg, Distinguishing transient osteoporosis from avascular necrosis of the hip. J. Bone. Jt. Surg. Am. 77(4), 616 – 624 (1995). 144. J. C. Gallagher, Pathophysiology of osteoporosis. Sem. Nephrol. 12(2), 109 – 115 (1992).

CHAPTER 59

Clinical Use of Bone Densitometry KENNETH G. FAULKNER

I. II. III. IV.

GE Medical Systems, Madison, Wisconsin 53717

Densitometry Techniques Basic Principles of Densitometry Performing Densitometry Measurements with DXA Monitoring the DXA Scanner

V. Uses of Bone Densitometry VI. Conclusions References

method, radiogrammetry, is a simple technique by which the thickness of the bones in the hand (or some other site, such as the humerus or radius) is determined by direct measurement from the X-ray [3]. Cortical width is measured using a ruler and a magnifier, and the measurements are converted to a bone mass score. The primary advantage of these methods is equipment cost, as all medical institutions have standard X-ray units, and the aluminum wedge needed for calibration with RA can be obtained for little or no cost. However, there are potential disadvantages. The total costs associated with RA can be high, when one includes the radiologist’s fee for the hand X-rays, the centralized analysis fee, and the cost of shipping. Recent improvements include systems that permit local analysis of hand and forearm radiographs, eliminating the need for shipment to a central analysis facility (Figs. 1 and 2). Results are obtained using custom analysis software for evaluating the digitized images to produce BMD values. Despite these improvements, X-rays are still primarily qualitative images and are not specifically intended for measuring bone density. This has led to the development of devices specifically designed to quantitate bone density at the hand (Fig. 1) and other skeletal sites using X-ray and ultrasound technology.

I. DENSITOMETRY TECHNIQUES A. Radiographic Techniques Before the development of bone densitometers, bone density was estimated from conventional X-rays by comparing the brightness of the skeleton to the surrounding tissues. Dense bone appears relatively white on a standard Xray, while demineralized bone has an appearance closer to soft tissue. However, it has been suggested that bone mineral losses of at least 30% are required before they may be visually detected on a conventional X-ray [1]. Because of the insensitivity of X-ray images to bone density changes, several techniques have been developed to improve the accuracy and precision of conventional radiographs for bone mass assessment. Many of these techniques were based on measurements of the hand, because of the minimal soft tissue and easy access for measurement. Radiographic absorptiometry (RA, Fig. 1) is a technique by which the gray levels of a conventional X-ray are calibrated to an aluminum step wedge placed on the film [2]. Two films are obtained at different X-ray settings, and the films are sent to a central laboratory for analysis (Fig. 2). A second

OSTEOPOROSIS, SECOND EDITION VOLUME 2

433

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

434

KENNETH G. FAULKNER

FIGURE 1 Densitometry devices for measuring the hand. These include (clockwise from top left), radiographic absorptiometry, a portable dual X-ray system, and an ultrasound device designed to measure the fingers.

B. A Definition of Terms Before proceeding, it is useful to define some of the terms and units used in the field of bone densitometry. We speak of measuring bone density, but in reality the “true” density of the bone is never determined. The precise definition of bone density is the mass of bone per unit volume — exclusive of marrow and other nonbone tissue. In the field of densitometry, the term “bone mineral density” (abbreviated BMD) is related to the mass of bone tissue, including both bone and marrow components. Furthermore, most densitometric techniques are projectional, providing a two-dimensional image (or shadow) of the bone being measured. Therefore, the

BMD derived from projection techniques is the mass of bone tissue mass per unit area, not per unit volume. What is actually measured is the apparent bone mineral density, defined by the bone mineral content contained in the area scanned, usually expressed in g/cm2. The results are “apparent” in that the measurement is a combined value of bone, marrow, and other tissues. A measurement of the “true” density would require an isolated sample of a pure bone in three dimensions, excluding any marrow components. Even spinal QCT (quantitative computed tomography discussed below), which is a volumetric measure of vertebral trabecular bone (usually expressed in mg/cm3), is a measure of apparent density as it includes the marrow space of the vertebral body. Those who work in the field

435

CHAPTER 59 Clinical Use of Bone Densitometry

FIGURE 2

Examples of hand and finger bone density measurements.

have become accustomed to the term BMD representing the apparent area density in some cases, and volume apparent density for QCT. In this chapter, the term BMD will be used to represent the area bone density, in units of g/cm2, unless stated otherwise.

C. Single-Energy Densitometry Because of the problems and inaccuracies of using uncalibrated radiographs for measuring bone mass, researchers at the University of Wisconsin developed the first dedicated bone densitometer for measuring the forearm in the 1960s [4]. This device passed a beam of radiation through the forearm and determined the difference between the incoming (or incident) radiation and the outgoing (or transmitted) radiation (called the attenuation). The higher the bone mineral content, the greater the attenuation. The BMD was calculated by dividing the bone content (related to the attenuation of the radiation) by the bone area. This

technique is called single-photon absorptiometry, or SPA. In the late 1960s, SPA became the first commercially available technique for the noninvasive measurement of bone mineral density. Tabletop units were sold by several manufacturers for measuring the forearm, and a floor standing unit was developed to measure the bone density of both the forearm and the heel. With the introduction of these devices, physicians were now able to precisely measure bone density at a very low radiation dose, so that serial measurements in an individual were possible without concern for excess radiation exposure. For the first time, it was possible to monitor the changes in bone density that might occur as the result of aging or treatment. However, SPA was not without its limitations. First, SPA required a radioactive isotope as a radiation source. This was both expensive and inconvenient, as well as potentially causing errors in the measurements when sources were replaced. Second, SPA is limited to measuring peripheral bones such as the heel and forearm, as the measurement site must be immersed in water. Placing the measurement

436

KENNETH G. FAULKNER

FIGURE 3

Full table dual X-ray absorptiometry (DXA) systems.

site in water cancels the effect of the overlying soft tissues, so that only the differential attenuation by the bone is measured. This approach is reasonable for the measurement of the peripheral skeleton, but it is not practical to immerse the entire body in water to obtain measurements of the spine or the hip. The use of an isotope source was only a temporary disadvantage for SPA. With the development of small X-ray tubes, isotope-based devices were quickly replaced with Xray-based systems. These devices were called single X-ray absorptiometry (SXA) units, with the “X” representing the switch from an isotope based photon source to an X-ray tube source.

D. Dual-Energy Densitometry As stated above, the primary limitation of single-energy densitometry is that it is unable to directly measure the spine and hip. The challenge was to devise a method that eliminated the need for a water bath so that any skeletal site could be measured. Researchers found that if a dual-

energy radiation source was used, the influence of soft tissue could be eliminated without the need for a water bath to equalize soft tissue attenuation. As with single-energy densitometry, a radioactive source was originally used, but with an isotope which emits photons of two energy peaks. This technique was called dual-photon absorptiometry (DPA), with the word “photon” reflecting the use of the isotope source. Like with SPA, It did not take long for DPA manufacturers to replace the decaying isotope source of DPA with a highly stable dual-energy X-ray tube. The result was a system that, compared to DPA, had increased precision, significantly reduced scan times, and did not require periodic source replacement (Fig. 3). All previous DPA manufacturers have now switched to producing dual X-ray absorptiometry (DXA) scanners. Several abbreviations have been suggested (QDR, DEXA, DPX); however, DXA is preferred as it does not refer specifically to any single brand of equipment. The basic operation of a DXA system is the same as that for SXA, only on a larger scale. The radiation source is collimated to a pencil beam and aimed at a radiation detector

437

CHAPTER 59 Clinical Use of Bone Densitometry

FIGURE 4

X-ray-based bone densitometers for measuring the peripheral skeleton

placed directly opposite the site to be measured. The patient is placed on a table in the path of the radiation beam. The source/detector assembly is then scanned back and forth across the measurement site. The attenuation of the radiation beam is determined and is related to the bone mineral content [5]. The bone area of the scanned region is determined by a computer, and the BMD is calculated as the ratio of the bone content to the measured area. DXA systems can measure the BMD of the lumbar spine, proximal femur, forearm, and, in many cases, the total body. More advanced DXA scanners use a fan beam geometry which increases scan speed and reduces acquisition time. Recent advancements in DXA have provided improved image quality, allowing better visualization of the scan region as well as the ability to detect vertebral deformities. DXA technology has also been adapted for use in small, lower cost densitometers (Fig. 4). Simplified DXA systems

targeted for clinical use are available to measure just the spine and hip in an attempt to reduce equipment costs. Compact, portable peripheral DXA (pDXA) systems are now available for measuring the peripheral skeleton as well. These devices have essentially replaced single X-ray densitometers for measuring the forearm, heel, and hand.

E. Quantitative Computed Tomography Before the advent of DXA, several researchers used computed tomography (CT) scanners to obtain bone density measurements [6 – 8]. This technique was called quantitative CT (QCT) to differentiate it from imaging CT. QCT is the only noninvasive three-dimensional bone mass measurement technique available. With QCT, the result is a volumetric density (in mg/cm3) as opposed to the area

438

KENNETH G. FAULKNER

While most QCT studies are limited to the lumbar spine, specialized QCT systems (called peripheral QCT, or pQCT) have been introduced for measuring the forearm. This technique offers the advantages of measuring the volumetric density of the forearm, as well as providing measures of trabecular, cortical, and integral (trabecular plus cortical) bone. However, these scanners are limited to the forearm and can cost as much as full table DXA devices capable of density measurements at multiple skeletal sites.

F. Quantitative Ultrasound

FIGURE 5

Quantitative computed tomography (QCT) image of the lumbar spine. Note the calibration standard under they subject used for the determination of bone density.

density (in g/cm2) from other techniques. Initially, QCT was performed without any special equipment (other than the CT system) by measuring the average CT number of the vertebral body [6]. However, more advanced procedures were developed to increase the accuracy and precision of the measurement [7,8]. QCT is used clinically to measure the bone density of the spine, though methods have been developed for measuring the hip. It has the advantage of measuring the central bone of the vertebral body, which is a more sensitive site for detecting bone mineral changes than most other skeletal sites [9]. QCT can be performed on most commercial CT systems with the addition of a bone mineral standard for calibration of the CT measurement (Fig. 5). Several different types of calibration systems are commercially available from CT manufacturers and third-party vendors. In the standard QCT protocol, three to four lumbar vertebral bodies are measured using a single 8 to 10-mm slice through the center of each vertebra [10,11]. The calibration standard must also be measured, either at the same time or immediately after the patient. Low-dose settings are used on the CT scanner to reduce the radiation exposure well below a standard CT examination. From the CT images, the average attenuation of the vertebral body bone is determined as well as the attenuation of the calibration standard. Using the known density of each of the standards and the measured CT values of the bone mineral standard, the vertebral CT value is converted to a physical density.

Ultrasound has been used for many years to investigate the mechanical properties of various engineering materials. It offers the theoretical advantage of measuring material properties other than density. Recently several commercial ultrasound devices have been introduced for investigating bone status, primarily of the heel (Fig. 6). This technique is termed quantitative ultrasound (QUS) to distinguish it from the more common imaging ultrasound devices. QUS offers the advantages of small size, relatively quick and simple measurements, and no need for radiation. However, QUS measurements are much easier to perform at skeletal sites with minimal soft tissue covering. The majority of QUS devices measure the peripheral skeleton, including the heel, shin, forearm, and fingers. Several different QUS devices have been shown to predict hip fracture, independent of X-ray-based bone density measurements [12,13]. This has fueled the interest in QUS as a measure of bone quality as well as density. QUS has seen widespread use around the world and has recently been approved for clinical use in the United States. While the future of QUS appears promising, there are still some questions that remain to be answered. For example, researchers are still not certain exactly which mechanical and/or structural parameters of the bone are being measured with QUS. It has been speculated that QUS may be related to trabecular size, trabecular spacing, and parameters of bone mineralization, such as crystal size and orientation. However, it does appear that the majority of a QUS measurement is determined by bone density [14]. It remains to be determined how QUS can be used to monitor skeletal response to different therapies. Yet the compact size and non-radiation-based qualities of QUS make it an attractive choice for populationbased screening programs.

G. Comparison of Techniques From the above discussion, it is clear that each bone mass measurement technique has different advantages and

439

CHAPTER 59 Clinical Use of Bone Densitometry

FIGURE 6

Quantitative ultrasound densitometers used for measuring the heel.

H. Radiation Dose

disadvantages. There are differences in the skeletal sites that can be measured, clinical utility, radiation dose, availability, cost, and ease of use. Table 1 summarizes these factors based on the available research data and clinical experience.

TABLE 1 Method

Radiation dose is a common question among patients having a bone density measurement. The effective doses for

Comparison of Densitometry Techniques

Utility

Versatility

Ease

Availability

Cost

Dose

RA













SXA





 







pDXA





 







DXA

 

 









QCT











/

pQCT













QUS





 





 

, excellent; , good; , poor. Note: Techniques include radiographic absorptiometry (RA), single X-ray absorptiometry (SXA), peripheral dual X-ray absorptiometry (pDXA), full table dual X-ray absorptiometry (DXA), quantitative computed tomography (QCT), and quantitative ultrasonography (QUS).

440

KENNETH G. FAULKNER

TABLE 2 Effective Radiation Dose (in Microsieverts, Sv), for Various Densitometry Techniques Compared to Other Common Radiation Sources Radiation source

Effective dose (Sv)

SXA

1

DXA

1–5

QCT

60

Lateral spine film

700

Natural Background (per day)

5–8

8 to 10 Hour airplane flight

60

Note. Values are shown for single X-ray absorptiometry (SXA), dual X-ray absorptiometry (DXA), and quantitative computed tomography (QCT). Adapted from Ref. 15.

various bone density measurements are listed in Table 2, along with the effective doses from some common radiation sources for comparison [15]. The effective doses for bone densitometry are only a small fraction of the yearly natural background effective dose. Notice that even a short airplane flight is responsible for an effective dose equivalent to several bone density examinations.

coefficient is known, this equation can be solved for the amount of a single material scanned with such a system: m  k log (I0/I). In this equation, k is an experimentally determined constant related to the attenuation coefficient. Unfortunately, the body is not composed of a single material but consists of bone, muscles, and other tissues and organs. To get an accurate measure of bone density, it is necessary to remove the influence of the overlying tissues and determine the attenuation of the bone alone. For singleenergy densitometry, this is done by surrounding the measurement site in water, as water has attenuation characteristics similar to muscle. The depth of water is set to a known thickness across the X-ray path and is kept constant for all measurements. In this way, the attenuation component due to soft tissue and water is assumed to be constant for all individuals, and any attenuation differences are attributed to variations in bone density.

B. Dual Energy Densitometry

To understand the physics of densitometry, one must first start with the basic equation describing the attenuation of X-rays by a single material:

As mentioned previously, single-energy densitometry measurements are limited to sites which can be immersed in water to compensate for variable tissue thickness. Though it is conceivable to create a single-energy system for measuring the spine or hip, the practicalities argue against such a device. In dual-energy densitometry, the soft-tissue problem is solved in a different way [5]. Mathematically, the problem is one with two unknowns, the bone thickness and soft-tissue thickness. To determine a unique solution for a system with two unknown variables, it is necessary to have two independent equations. By using two X-ray energies, two equations can be derived by scanning the measurement site twice, once at each energy:

I  I0e( m),

IL  I0L [exp  (bL mb  tL mt)]

II. BASIC PRINCIPLES OF DENSITOMETRY A. Single-Energy Densitometry

IH  I0H [exp  (bH mb  tH mt)]

where I0

 incident radiation intensity

I

 transmitted radiation intensity



 mass attenuation coefficient for the attenuating material (cm2/g)

m

 area density of the attenuating material (g/cm2).

The mass attenuation coefficient is a physical property that describes how much a given material attenuates an X-ray beam. It depends on the type of material as well as the energy of the X-ray beam. If the linear attenuation

The subscripts b and t refer to the tissue and bone, respectively, while the superscripts L and H are for the low and high energies, respectively. In practice, the measurement site is not scanned twice, rather each point is either scanned simultaneously with both energies or by rapidly switching between X-ray energies at each measurement point. These two equations can be solved together to compute for the area bone mineral density at each point in the X-ray beam: mb 

(Ls /s ) log (I H / I H0 )  log (I L / I L0 ) . Lb  b (Ls / s )

This equation assumes that we know the attenuation coefficients for bone and soft tissue at both high and low

CHAPTER 59 Clinical Use of Bone Densitometry

X-ray energies. For bone, the attenuation coefficient is relatively constant from individual to individual. However, in the soft tissue, the attenuation coefficient varies greatly, mostly due to differences in body fat content and distribution. To account for these variations in the amount and distribution of fat, it is necessary to define the R value: R  (sL/sH). R can be determined by measuring the attenuation of the X-ray beam in soft-tissue at both high and low energies. It is related to the percentage of fat in the soft tissue and is also used for body composition studies. Once R is determined, and the values for the attenuation coefficient for bone at the high and low energies are identified, the amount of bone at each measurement point can be calculated. A computer evaluates the bone area based on the attenuation differences between the bone and soft tissue, and the BMD is determined as the total bone mass divided by the bone area. This can be done for the entire scan region or for regions of interest defined within the scan. This approach assumes that the X-ray tube, detectors and associated hardware will be stable and consistent over time. In the real world, this is never the case. To compensate for day-to-day fluctuations and drifts in the DXA system, the manufacturers have included calibration procedures to ensure stable performance. For most DXA systems, this involves measuring a set of stable calibration standards each day before acquiring patient data. Other systems have internal calibration systems which monitor machine performance continuously.

C. Dual X-ray Sources and Detectors Not just any two X-ray energies will work for DXA. It is important to choose two energies that maximize the differences in attenuation for bone and soft tissue. The first dual-energy systems used a gadolinium isotope source (153Gd) which has energy peaks at 44 and 100 keV. X-ray-based dual-energy systems are designed to mimic the 153Gd source using either an energy switching or filtered spectrum X-ray tube. Switched energy systems have the advantage of a simplified detector system, which is not required to distinguish high- and low-energy X-rays. By synchronizing the data collection and the switching X-ray tube, alternating high- and low-energy data points can be obtained and stored using a single, non-energy-discriminating detector. However, rapid voltage switching can induce Xray instabilities, requiring constant monitoring of the Xray beam. On voltage-switching systems, this is accomplished using a continuous internal calibration system which monitors the system stability at each measurement point.

441 Other DXA systems use filtered X-ray sources and energy-discriminating detector systems to simultaneously acquire high- and low-energy data. By choosing an appropriate type and thickness of filter, it is possible to use the K-edge phenomenon to split the spectrum into two peaks. Two filter materials are typically used; cerium (K-edge at 38 keV) and samarium (K-edge at 47 keV). When using a filtered X-ray source, the X-ray detector system must be capable of distinguishing the two different X-ray energies. This can be done using either stacked detectors (a low-energy detector directly below a high-energy detector) or an energy-discriminating detector (which separates the detected X-rays based on the amount of energy deposited in the detector). For filtered systems, highly stable X-ray tubes can be used, such that constant beam monitoring is not necessary. Daily checks of system stability using an external calibration standard are usually considered sufficient. For a typical DXA system, the X-ray source is mounted in the table below the patient. The detector is placed above the table with a C-arm attached to the X-ray tube assembly. With single-beam systems, the tube and detector move together across the measurement area in a serpentine fashion. Newer DXA systems are designed with fan-beam X-ray sources and an array of X-ray detectors which acquire data one line at a time as opposed to one point at a time. The result is faster scan speeds than the single-beam devices. Fan-beam geometry creates some special challenges for DXA that do not exist in the pencil-beam configuration. The most noticeable is the magnification error associated with any fan-beam imaging system. Distances are magnified as the object is moved closer to the radiation source. For the measurement of BMD, the effect is negligible, as both projected area and bone content are influenced to the same degree. However, the area and bone content measurements are estimates only and cannot be considered accurate. Fan-beam systems are also susceptible to parallax errors caused by placing the object off center in the radiation beam. This results in a variable path length through the measured object which can influence the measured BMD. In practice, these potential fan-beam errors are not of any clinical significance. As mentioned previously, the influence on BMD is minimal, resulting in a slight, if any, increase in precision error compared to pencil beam systems [16]. On the other hand, fan-beam systems offer several advantages. By acquiring entire lines of data as opposed to a single point at one time, measurement times decrease with the fan-beam geometry. A spine fan-beam measurement typically takes less than 1 min, while pencil-beam measurements using early systems requires 4 to 6 min. Newer model pencil-beam systems have reduced this time to 2 min, or less. Faster computers have reduced analysis time for all DXA devices. However, the time needed for patient preparation, positioning and other tasks remains similar

442

KENNETH G. FAULKNER

FIGURE 7

Examples of bone density measurements of the spine using a full table DXA system.

for fan-beam and pencil-beam systems. Many fan-beam systems are equipped with rotating C-arms which permit lateral scanning of the spine for density measurements as well as assessment of vertebral fractures.

III. PERFORMING DENSITOMETRY MEASUREMENTS WITH DXA A. PA Spine The most common scan performed with DXA is of the spine. PA indicates that the scan is done in the posterior – anterior projection, passing from the back of the body to the front. Often this exam is called an AP spine measurement, though due to the configuration of DXA scanners with the X-ray tube underneath the patient, the measurement is properly referred to as a PA spine scan. The PA spine measurement is limited to the lumbar vertebrae, as measurements in the thoracic spine are complicated by air in the lungs (which alters the soft tissue baseline) and the presence of the ribs and sternum overlying the scan field. The scan normally includes L1 – L4, though some clinics report only L2 – L4. It is recommended to scan the entire region from the middle of T12 (defined as the lowest vertebrae with ribs attached) through the middle of L5 (normally the last vertebral body directly above the sacrum). In this way, either L1 – L4, L2 – L4, or any other combination can be analyzed. In addition, the presence of ribs on T12 and the visualization of the iliac crest at L5 provide the anatomical landmarks necessary to accurately identify the vertebral levels. Routinely scanning L1 to L4

also has the advantage of sampling a larger volume of bone, and provides four vertebral bodies for measurement in case one (or maybe more) are unanalyzable due to technical reasons or deformities. During the measurement, look for any motion artifacts in the scan caused by sudden movements by the patient (due to coughing, tremors, etc.) These will appear as discontinuities in the scan, particularly noticeable at the bone edges. Large motions will require that the patient be rescanned, as the shift in position can cause errors in the measured BMD. In a properly acquired spine image, the spine should be centered in the scan field and properly aligned. Be alert for the appearance of artifacts in the scan field. Removable artifacts should have been eliminated before starting the scan; if artifacts are seen, they should be removed from the scan field and the measurement repeated. In most cases, the analysis of a PA spine scan is uncomplicated (Fig. 7). The entire L1 to L4 region should centered in the scan field. The ribs of T12 should be visible in the soft tissue lateral to L1, and the iliac crest should be seen at the lateral borders of the scan at or below the L4 level. A portion of the relatively broad and flat L5 should be seen directly below the dense H-shaped L4. L1 is identified as the first vertebra on the superior end of the scan that does not have ribs. Sometimes, the presence of ribs can be hard to detect — particularly the 12th ribs can be hard to see. Adjusting the viewing parameters of the scan (the contrast and gray level) will help this problem. Occasionally, a patient will have either an extra lumbar vertebra, or appear to be missing L5. When the vertebral level identification is uncertain, it is best to look for the characteristic H-shape of L4 as the benchmark for labeling

443

CHAPTER 59 Clinical Use of Bone Densitometry

the other vertebrae. This is different than the approach used by many radiographers for labeling vertebrae on standard X-rays. However, identifying L4 and labeling the vertebrae from the bottom up will produce the most consistent density result and avoid erroneously including T12, which is normally of much lower density than the other vertebral bodies [17]. Presence of vertebral fractures or deformities, degenerative disease, or severe curvature will alter the density of the vertebrae, such that the results will be of limited utility [18,19]. This is particularly true after age 65 to 70 when the presence of degenerative disease is common. A spine with marked degenerative change may be an indication for a lateral spine scan in clinics so equipped. However, in many cases, the lateral measurement may be no more useful than the AP scan, particularly in cases of severe scoliosis. In this case, a hip measurement may be the best alternative. When reviewing a PA spine measurement, look for large differences in vertebral height, area, BMC, and/or BMD between the different levels [17]. These differences are indicative of fracture or degenerative change which can influence the results. Exclude any vertebral bodies from the analysis that are fractured or deformed. If there is a question of whether or not a particular vertebral level should be excluded from the analysis, the scan can be analyzed both ways before making a the final judgment. Once properly placed, the location of the intervertebral markers should be recorded for future reference. Note if any of the markers were angled to accommodate curvature in the spine. In particular, the overall height of the L1 to L4 region should be recorded and duplicated for any future measurements of the same patient. Some DXA systems provide this capability semiautomatically by copying a region of interest from a previous scan for use on the current analysis. If the comparison feature is available, the intervertebral marker locations for each scan will automatically be stored. The analysis program will simply copy the intervertebral markers from a previous scan of the same patient and overlay them on the current scan. It may be necessary to adjust position of the region of interest on the new scan, but the intervertebral spacing should be preserved. If it appears that the height of either a single vertebral body or the entire L1 to L4 region has changed over time, this is indicative of a change in positioning or possibly an incident vertebral deformity. It is important to confirm visually that the bone edges are correctly identified by the computer. Manually altering the bone edges is rarely necessary and should only be done in circumstances where they are obviously wrong. Small changes should be avoided, as they will have little (if any) influence on the BMD and will be difficult to reproduce. However, if there are any “holes” contained within the bone edges, these should be filled. Some DXA systems perform this function automatically.

B. Proximal Femur The hip DXA scan includes the proximal end of the femur and portion of the pelvis, but only the bone in the proximal femur is evaluated. The proximal femur is the most difficult place to perform DXA measurements, as small changes in femoral rotation can cause large changes in BMD [20]. All DXA manufacturers supply special positioning aids for performing proximal femur scans. This positioning aid is normally a foot block or other device designed to position the measured leg at 15° to 30° inward rotation. By inwardly rotating the femur, the femoral neck is aligned parallel to the scanner table and perpendicular to the X-ray beam. In this way, the full femoral neck can be seen on the DXA image. Normally, the left and right femoral BMD will be similar within a patient [21,22]. Thus either hip can be measured. By routinely choosing either the left (or right) hip for all patients, the chance of scanning a different hip at follow-up is minimized. In cases of a previous fracture or suspected disease which affects only one hip, be sure to note the condition and scan the opposite hip. Likewise, if degenerative changes are seen or suspected in one hip, the other hip should be scanned as well. If both hips contain orthopedic hardware in the neck region, then scanning the hip is not recommended. Carefully observe the image as it appears on the screen. The lesser trochanter should be only slightly visible. This is an indication that the femur is rotated properly. If a prominent lesser trochanter is seen, the femur may not be completely rotated. Note that for some patients, arthritis or other conditions may preclude rotation of the femur to the desired angle. In this case, note the condition in the patient’s records to justify why the rotation angle was less than optimal. Look for motion artifacts, which appear as sharp discontinuities in the bone edges. While minor movements will not influence the BMD, large movements will necessitate a rescan. When evaluating the proximal femur scan, look for proper acquisition technique, with the femoral shaft aligned vertically and the hip rotated so that the lesser trochanter is minimally (if at all) visible (Fig. 8). Look for artifacts in the scan from clothing or surgical procedures which may alter the BMD. These should be noted and excluded from the analysis. All commercial DXA systems evaluate at least three different regions of interest at the proximal femur. These include the femoral neck, the trochanteric region, and Ward’s region. In addition, many systems include an intertrochanteric region and a “total” region (defined as the area-weighted average of the femoral neck, trochanteric, and intertrochanteric regions). Unlike the spine where the analysis region is consistently defined, the definitions for the proximal femur regions differ from manufacturer to

444

KENNETH G. FAULKNER

FIGURE 8

Examples of bone density measurements at the proximal femur using DXA.

manufacturer. It is important to carefully follow the manufacturer’s instructions for placement of these regions so that comparisons with the normative data will be valid. Because of differences in region definitions, comparisons of femoral BMD values (either BMD, standardized BMD, or T scores) across manufacturers cannot be considered as accurate as comparisons at the spine.

C. Forearm The forearm is a common site of osteoporotic fracture and, unlike the spine and the hip, the radial shaft is a good indicator of cortical bone density. It is also readily accessible and not subject to hidden artifacts and soft tissue fluctuations like the spine and hip. Thus precise measurements of the forearm BMD are relatively easier to achieve than those of other skeletal sites. Unlike the proximal femur, there is a tendency for the dominant arm to have greater bone density than the nondominant arm. This occurs in many individuals, but is particularly apparent in those who routinely use one arm more than the other (such as professional tennis players) [23]. Normally, the nondominant arm (i.e., the left arm on a right-handed person) is scanned, as it can be expected to have the lower density of the two extremities. However, if the patient has sustained a forearm fracture, the unfractured arm should be measured. If both forearms have been fractured, consider not performing the forearm scan, or scan the arm with the least recent fracture. When reviewing the forearm scan (Fig. 9), the arm should be straight and centered in the scan field. Look carefully for motion artifacts, which appear as discontinuities in the bone edges. Small side-to-side movements will

not significantly affect the measurement, but large shifts in the arm position will have a significant impact on the results. A strap or other object on the elbow during acquisition (placed outside of the scan field) will assist the patient to remain still. Forearm scans with excessive motion should be repeated unless the motion is unavoidable (such as from tremors). Artifacts such as jewelry are not normally a problem at the forearm, but previous fractures should be noted, particularly if the analysis will include any portion of the healed bone.

D. Total Body Total body scans are an excellent measure of cortical bone (the skeleton is 80% cortical) and are very precise because of the large sample of bone measured. Sometimes, total body scans are performed to evaluate body composition parameters, such as the lean body mass, and percent body fat, in addition to the bone content and density. Body composition measurements require special software, and possibly an additional calibration standard, that must be obtained directly from the DXA manufacturer. When performing a total body scan, it is necessity to clothe the patient in a hospital gown or surgical scrubs. Otherwise, belts, zippers, buttons, jewelry, objects in pockets, etc. will invariably be overlooked and appear in the scan. Some artifacts cannot be removed (such as an orthopedic implant), and must simply be noted in the patients records. Often, rings cannot (or will not) be removed by the patient; these should also be recorded in the patient’s records and be worn again at any follow-up examination.

445

CHAPTER 59 Clinical Use of Bone Densitometry

FIGURE 9

Examples of forearm bone density measurements using DXA.

For the total body scan, the subject must be aligned with the scanner axis and centered in the image (Fig. 10). The arms should be at the sides, slightly separated from the trunk. If possible, the hands should be palm down on

FIGURE 10 composition.

the table and separated from the thighs. For large or heavy patients, it may not be possible to place the hands flat and still keep them in the scan field. In these cases, the hands can be placed on end (i.e., rotated 90°) so that they can be

DXA measurements of the total body, used for measuring bone density and body

446

KENNETH G. FAULKNER

FIGURE 11 Lateral DXA images of the spine used for bone density assessment (left) and vertebral morphometry (right).

contained in the scan. The legs and feet should be placed together, toes pointing upward, and loosely secured with tape or straps to prevent movement. Any tape or straps should be radiolucent, such that they do not influence the measured BMD or body composition measurements. Be on the watch for motion artifacts, as the total body measurement generally takes more time than most other scans. When analyzing a total body scan, it is first necessary to verify that the entire body is contained inside the scan field. If any portion of the body is outside the scan, this should be noted. Check for artifacts and implants (such as joint replacements, pacemakers, breast implants) in the scan. Also look for metal jewelry, dense clothing, or other external artifacts. When performing a body composition analysis, it may be necessary to perform some additional analysis steps (such as locating a soft tissue phantom scanned with the patient).

E. Special Scans The scans described above represent the majority of the measurements performed using DXA. However, modern DXA systems are able to perform additional studies, including lateral spine BMD measurements, evaluation of vertebral fractures, assessment of bone around prosthetic implants, experimental scans of other skeletal sites, measurements in children and infants, small animal studies, and measurements of excised bone specimens. These special scans are not encountered in a general practice clinic, while in research centers they are more common. 1. LATERAL SPINE Lateral spine DXA was designed to overcome some of the disadvantages of PA spine measurements. By measuring the lumbar spine in the lateral projection (across the body), the vertebral body can be isolated and evaluated without

447

CHAPTER 59 Clinical Use of Bone Densitometry

the influence of the dense posterior elements (Fig. 11) In a PA spine scan, much of the measured BMD comes from the posterior elements, which are not the primary site for weight-bearing or fractures. In addition, aortic calcifications, which would overlay a PA scan, can be excluded in the lateral view (though the BMD error due to aortic calcifications is typically negligible). However, vertebral body osteophytes or bone spurs, as well as spinal scoliosis and fractures, can still influence lateral BMD measurements. First attempts at lateral scanning were performed with standard DXA systems with the patient lying on the side (lateral decubitus position). The knees were flexed, the arms placed in front of the face, and various pillows and supports were required to bring the spine parallel to the scanner table and keep the patient comfortable. However, even with cushions and positioning aids, this was uncomfortable for most people, as well as being very difficult to reproduce. Modern DXA systems have been designed with a rotating C-arm and fan-beam geometry which allow the patient to remain supine for the lateral spine scan. The patient lies normally on the table, with the arms placed over the head, out of the scan field. The C-arm is rotated such that the X-ray tube and detectors are placed opposite each other on the sides of the patient. Lateral scans are limited to the L2 – L4 region. L1 is obscured by ribs in the lateral projection and BMD measurements cannot be accurately made. In the decubitus position, often only L3 can be evaluated. For most individuals, the 12th rib extends to the level of L2 in the lateral projection, overlying the vertebral body. Also, the iliac crest often protrudes at least partially over the vertebral body of L4. Rarely is a complete analysis of L2 – L4 possible with the patient scanned on their side [24]. Precision of the decubitus lateral scans is also poor compared to other DXA measurements, because of the difficulties in positioning and defining a consistent soft tissue baseline. For these reasons, lateral DXA scans are usually performed in the supine position on systems with a rotating gantry. 2. VERTEBRAL MORPHOMETRY With rotating C-arm fan-beam systems, it is possible to scan the entire thoracic and lumbar spine for the evaluation of vertebral heights [25]. The patient is positioned similar to a supine lateral spine scan, flat on the table, with the arms over the head. With the C-arm rotated to the lateral position, the scanner acquires an image from the upper thoracic to the lower lumbar spine. This image is stored in the computer, and special software is used to measure the heights of each vertebral body. Differences in vertebral height between the posterior, mid, and anterior vertebral body dimensions indicate that a vertebra might be fractured. This technique offers several potentially attractive features. Evaluation of spine fractures can be performed without a conventional lateral spine X-ray and at the same time as the

BMD measurement. Also, the scanning geometry of the DXA system, with the X-ray beam and detector parallel to the disc spaces at all vertebral levels, is better for measuring vertebral heights. Conventional spine films suffer from distortions at the film edges, where the X-ray beam strikes the vertebral body at an angle. Finally, the DXA vertebral morphometry scan is easily analyzed on the system computer using software specifically designed for evaluating vertebral fractures. The computer (with the help of the technologist) places three points on the posterior, mid, and anterior margins of the superior and inferior endplates of each vertebra (Fig. 11). The vertebral heights are then calculated and compared to each other as well as to the expected normal dimensions. Changes in the anterior vertebral height compared to the posterior height are an indication of a wedge type fracture, while a decreased mid vertebral height suggests an endplate deformity. A decrease in all three heights compared to adjacent vertebrae or established norms suggests a crush type fracture. Despite the apparent advantages, the future of vertebral morphometry remains unclear. Skeletal radiologists have criticized the technique for being insensitive and inaccurate for detecting vertebral fractures. They point out that a few measurement points cannot accurately describe the complex shape of the vertebral body. Additionally, the placement of measurement points on the endplates can be very difficult and often ambiguous. A DXA morphometry scan is also of a lower resolution than a conventional X-ray, and might fail to identify other potential problems or diseases that would be apparent on a spine film. 3. OTHER USES OF DENSITOMETRY DEVICES One of the primary advantages of DXA is that it can measure any skeletal site. Various researchers have taken advantage of this fact to measure (or at least try to measure) virtually every bone in the body. DXA measurements have literally been attempted from the head (the skull and jaw) to the foot (the heel). They have also been used for measuring animals for research using special scan protocols and software. For patients who have had hip replacement surgery, most DXA manufacturers have special software available to measure the bone density around the metal implant. In this way, bone losses around the implant can be detected. It is hoped this might provide an early indicator of implant failure. Geometric measurements of the hip have been developed and incorporated into the DXA scanning software. The hip axis length (defined as the distance along the femoral neck axis from the base of the greater trochanter to the inner pelvic brim) has been identified as an independent indicator of hip fracture risk [26]. This measurement can be obtained from a standard DXA scan of the proximal femur, though on some systems an increased scan field may be required (Fig. 12). Optional software is available from most manufacturers to automatically determine the hip axis length.

448

KENNETH G. FAULKNER

FIGURE 12

Use of a proximal femur DXA image to measure hip axis length, defined as the distance along the femoral neck axis from the inner pelvic brim (A) to the base of the lesser trochanter (A).

More complex measures of femoral geometry have been developed, such as the cross-sectional moment of inertia (CSMI). The CSMI is a measure of how the bone is distributed in the femoral neck. This measurement can be obtained from a standard hip scan, but special software is needed to analyze the image [27]. The clinical use of CSMI has yet to be determined. It may be necessary to measure the bone density of children as part of a clinical evaluation. DXA scans of children are difficult for several reasons. Most children will not stay still for the time required, so that motion artifacts are common. The size and density of the bones in children are often a challenge for the DXA software to analyze. When evaluating changes in BMD of children, the measurement is influenced by growth of the bones and the child, so that comparisons over time are complicated. Because of limited experience with DXA in children, the amount of normative data is small.

F. Follow-up Scans When acquiring follow-up scans on patients, changes and variations in technologist performance will adversely effect results. These variations increase the precision error,

due to either inconsistencies in patient positioning, differences in analysis, or both. The primary way to minimize precision errors is to adopt and maintain a consistent scan procedure. If there were any deviations from the standard procedure at baseline, these should be carefully noted. Refer to a printout of the baseline scan before starting the follow-up measurement and duplicate the positioning as closely as possible. Check that the same scan mode and parameters are used at follow-up as were used on the baseline examination. For the hip and the forearm, the same side should be scanned as at baseline. If available, the comparison feature of the DXA analysis software may be helpful when analyzing follow-up measurements. However, analysis is rarely as simple as copying previous regions of interest over to the new image. Usually the regions will need to be adjusted in order to achieve a compatible analysis with the baseline. It is not unusual for a baseline analysis to be incompatible with a follow-up measurement due to differences in the scan field or other acquisition parameters. In this case, it may be necessary to reanalyze a baseline measurement in order to have the regions of interest be compatible with the follow-up analysis. If a baseline measurement is reanalyzed, all subsequent measurements should be reanalyzed as well using the newly defined baseline regions of interest. Changes to the baseline analysis should be made with caution. If significant alterations in region placement are required to match the follow-up measurement, this is indicative of inconsistent acquisition. In this case, rescanning may be the best option in order to provide better compatibility with baseline. Note that this may require a nonstandard acquisition of the follow-up measurement if the baseline was done poorly. If a baseline measurement is reanalyzed for any reason, retain a copy of the original analysis in the patient records, with a notation that the scan was reanalyzed.

G. Archiving Data The final step in analysis is to preserve the data. At least one printout should be generated for the patient records, with additional printouts as needed for the physician or other staff. In addition, all DXA systems provide a means of archiving the data to floppy or optical disk. This electronic archival is important for retrieving the scan in the future.

IV. MONITORING THE DXA SCANNER When performing DXA, the quality of the data is defined by both the precision and the accuracy of the measurements. Precision is defined as the reproducibil-

CHAPTER 59 Clinical Use of Bone Densitometry

449

FIGURE 13

(A) Quality control plot from a DXA system showing stable performance. Each of the points represents a single measurement of the quality control (spine) phantom. Note that all measurements are within 1.5% of the baseline mean. (B) Quality control plot from a DXA system that experienced a significant calibration shift. In mid-1996, a series of measurements were more than 1.5% below the mean, indicating a change in calibration. Upon servicing, the system was restored to proper calibration and remained stable.

ity, determined by scanning the same object several times and determining the coefficient of variation (CV). The CV is the standard deviation divided by the mean of the repeated measurements. Accuracy, on the other hand, is the difference between a measurement and “truth.” It is usually defined as the percentage difference between a measurement and a defined standard. Typically, problems with measurement precision can be attributed to inconsistencies in operator performance, while accuracy errors are due to machine malfunction. Both precision and accuracy errors must be minimized for the DXA measurement to be clinically useful.

A. Minimizing Instrument Errors To obtain an accurate BMD measurement, it is important that the DXA system is properly calibrated. At the time of manufacture, each DXA scanner is calibrated at the factory to a set of known bone density standards. Over time the system can be expected to drift due to X-ray tube aging, environmental changes, and other factors. To compensate for any potential drifts, the DXA manufacturers have provided ways of monitoring and, if necessary, correcting for any drifts in scanner performance. For many DXA systems, a daily calibration measurement is performed using standards supplied with the system. Each morning, the technologist is required to scan the calibration standard according to the manufacturer’s instructions. If a calibration shift is detected during the calibration measurement, the DXA system can self-adjust to bring the scanner back into the original calibration. However, if the detected shift is large, an error message is

displayed. In addition to the calibration check, most systems perform routine system diagnostics to verify that the detectors, X-ray system, and other subsystems are functioning properly. For details of the scanner’s daily calibration procedure, refer to the operators manual. For some DXA systems, the daily calibration measurement is replaced by an internal reference system that constantly monitors the system performance. The performance of the internal reference system is completely transparent to the technologist and requires no operator interaction. Despite daily or even constant calibration checks, drifts in scanner performance can still occur that are not compensated for by the calibration system (Fig. 13). As an independent check of scanner stability, each DXA manufacturer provides a quality control phantom to track machine performance over time. This phantom differs from the calibration standard previously described, as the results of scanning the phantom are not used by the DXA scanner to adjust the machine calibration if a drift has occurred. Instead, the periodic measurements of this quality control phantom provide an independent monitor of scanner stability. The different manufacturers have each developed their own quality control phantoms. They are typically fashioned to represent the lumbar spine, though some are shaped more like the spine than others. Several of these phantoms have “vertebrae” made of simulated bone material encased in plastic designed to mimic the soft tissues. Others are simply a machined aluminum bar which is submersed in water for scanning. It is recommend to measure these phantoms every day that the machine will be used. The spine phantom should be measured on the DXA system using the same settings as used for patient measurements. Always place the phantom in the same location near the

450

KENNETH G. FAULKNER

middle of the scanner table, on top of the table pad. It is acceptable to measure the phantom without a table pad, though it is best to measure the phantom in the same configuration as used for patient measurements. Start the scan at the same point in the phantom. The most important point to remember is consistency when measuring the spine phantom, as any changes in procedure may appear as an erroneous change in the phantom BMD. The phantom must be measured and analyzed using the spine analysis protocol normally used. Region of interest dimensions and intervertebral markers must be the same for all phantom measurements. BMD results should be recorded and plotted for review (some scanners have utilities to do this automatically). The first time the spine phantom is measured, it is important to establish a baseline BMD for the system by measuring the phantom 10 times on the same day. Subsequent BMD results from the daily phantom measurements are then compared to this established baseline value. Quality control rules specifically tailored for DXA phantom measurements have been established for this purpose. A list of suggested rules is given below [28]: 1. 1.5% Rule: One measurement more than 1.5% from the established baseline value. 2. 1.0% Rule: Two consecutive measurements more than 1.0% above or below the established baseline value. 3. 0.5% Rule: Four consecutive measurements more than 0.5% above or below the established baseline value. 4. Mean  10 rule: Ten consecutive measurements either above or below the established baseline value. The quality-control rules described above were designed for use in multicenter pharmaceutical studies with bone densitometry. In a clincal setting, system performance can be evaluated by plotting the daily spine phantom BMD as a percentage difference from the established baseline. To convert the daily phantom BMD to a percentage difference from baseline, use the following formula: %difference 

BMDph  BL  100%, BL

where BMDph  the BMD from the daily spine phantom measurement BL  average BMD of the 10 baseline measurements of the same phantom. Keep track of the date scanned, the BMD and percentage difference for each day the spine phantom is measured. If any single measurement is more than 1.5% from the baseline, repeat the phantom measurement. If the second measurement is more than 1.5% from baseline, the equipment service representative should be contacted for a more detailed system evaluation. If an adjustment to the system calibration is needed,

make sure that the service representative brings the measured spine phantom bone mineral density back to the established baseline value (within 1%). Also, be sure to check, and if necessary reestablish, the baseline by performing 10 measurements of the spine phantom after any service or relocation of the DXA scanner.

B. Calibration Differences At this point, it is important to mention the differences in calibration that exist between the densitometry systems from different manufacturers. At present, there is no consistent definition of “truth” when measuring BMD. For most quantitative measurements (such as time or length), there exists a carefully defined standard by which all other measurements are compared. In the field of densitometry, each manufacturer has independently defined “truth” based on several different standards. In some cases, samples of ashed bone, animal tissue, and animal fat were used as the basis for calibration. In other cases, chemical solutions of bone, tissue, and fat equivalent materials were used. While the different methods are approximately the same, they are not exactly the same. As a result, if the same skeletal site is measured on the same person, the reported BMD from two different systems will disagree. These differences can be as large as 10 to 15% at the lumbar spine [28]. In clinical practice, this calibration difference between manufacturers is not a problem, as long the same scanner is used for all measurements. However, if a switch is made to another type of DXA scanner, the BMD values will not agree with those previously measured. The DXA manufacturers are currently attempting to standardize their instruments based on measurements of the European Spine Phantom (ESP). The ESP is a semi-anthropomorphic spine phantom composed of calcium hydroxyapatite (the major constituent of bone mineral) in tissue-equivalent plastic. It contains three simulated vertebrae of low, medium, and high density. Results from the ESP have been compared to the PA spine BMD values from a group of female volunteers showing excellent agreement [29]. Independent researchers have used the results to compute a standardized BMD (sBMD) that is designed to correct for the calibration differences at the PA (L2 – L4) spine and total hip between the three major distributors of DXA systems, Hologic, Lunar, and Norland [30,31]: sBMD (L2 – L4 spine, mg/cm2)  1075.5 (Hologic BMD)  952.2 (Lunar BMD)  1076.1 (Norland BMD) sBMD (total hip, mg/cm2)  1008 (Hologic BMD)  0.006  979 (Lunar BMD)  0.031  1012 (Norland BMD)  0.026.

CHAPTER 59 Clinical Use of Bone Densitometry

The sBMD is defined in units of mg/cm2 (1000 mg  1 g), rather than g/cm2, to distinguish it from the manufacturer’s default BMD reading. Using these equations, it is possible convert the manufacturers L2 – L4 spine values so that they are approximately the same (within about 2 – 5%) for the three types of DXA scanners. However, the approximate nature of this correction must be stressed. Individuals can still show significant variations in sBMD between instruments.

C. Changing Equipment Any changes in instrumentation should be avoided. However, for several reasons, a change in equipment at some point in time is virtually guaranteed to occur. Upgrades to new technology will (hopefully) produce superior measurements in less time. For the benefit of the patients, clinics should be prepared to take advantage of these advances and upgrades. These changes should occur only after taking the appropriate steps to determine if the BMD results will be adversely affected. In addition, if the densitometry system is being used in any research trials, it will be necessary to clear any potential equipment changes with the study sponsor before installation. Software upgrades are the most common equipment change. Manufacturers routinely provide software updates, to both enhance the features of the system and correct errors. Anytime new software is received, it is important to receive verification from the manufacturer that the upgrade will not adversely affect the densitometry results. Before installing the software, a series of at least five spine phantom measurements should be done. The mean of these measurements should be within 1% of the established baseline. After installing the software upgrade, the phantom should be remeasured, and the BMD, BMC, and area results should be within 1% of the baseline value. Less commonly, changes in equipment will occur. Most often, these concern scanner maintenance, such as replacement of a detector system or X-ray source. These changes should be done by the service technician in such a way that the original calibration of the system is maintained. If possible, the calibration should be verified with at least five phantom measurements both before and after the service is performed. Unfortunately, a densitometry system cannot be expected to last forever and will someday need to be replaced. Therefore, a time will come when all active densitometry clinics will have to face the problems associated installing a new instrument. The following procedure is recommended when upgrading to a new system from the same manufacturer. At the time of the installation, the calibration of the new system should be matched to the existing system as closely as possible by the service representative. This

451 should be verified by the DXA technician using their spine phantom. The same spine phantom as supplied with the original scanner should be used for this verification and for all subsequent quality assurance measurements on both systems. If an additional spine phantom is supplied with the new system, it should not be used, as it can be expected to differ slightly from the original spine phantom. Verification consists of phantom measurements on the old system followed by 10 phantom measurements on the new system using the same or comparable scan mode. Measurements should be obtained sequentially without repositioning of the phantom. To pass the verification, the following criteria must be met: a. The coefficient of variation (CV) for each group of 10 measurements should be 0.5% or less. b. The difference between the previous baseline phantom BMD and the new BMD should be less than 1.0% After successful completion of the phantom verification as described above, a group of 10 patients should receive a complete set of measurements on both machines. Ideally the average difference between the systems should be less than 1% for the PA spine BMD. Small differences in BMD may be detected at the other scan locations, but they should be no more than 2%. When successfully completed, the two systems can be expected to be accurately cross-calibrated within the experimental error of the devices.

V. USES OF BONE DENSITOMETRY Each of the currently available bone density techniques offer different advantages and disadvantages for clinical use (Table 1). Currently, no single technique ideally addresses all the clinical requirements [9]. All bone density techniques have some clinical utility for assessing fracture risk [32]. However, some devices offer advantages in terms of versatility (i.e., the number of skeletal sites which can be measured), ability to monitor response, cost, availability and ease of use. Radiation exposure is very small for all techniques (in most cases, less than a chest X-ray or mammogram) and, in the case of quantitative ultrasound, nonexistent. So the question often is asked, which technique is the best to use? In clinical practice, the answer to this question depends on the purpose of the bone density measurement. Different indications for bone mass evaluation will necessitate the measurement of different skeletal sites, which will in turn dictate which technique should be ideally used. For example, assessment of fracture risk and monitoring of bone change provide very different challenges for bone density testing (Table 3). It is essential to first establish the purpose of the bone density test before the appropriate skeletal site and technique can be chosen. While there are many reasons for

452

KENNETH G. FAULKNER

TABLE 3 Technical Requirements for the Diagnosis of Fracture Risk and Monitoring Skeletal Response Diagnosis • Accuracy more important than precision • Valid reference range • Inclusion of other risk factors Age Prevalent fractures Family history

meanings in the field of statistics. However, in the field of densitometry, the Z score and T score are constantly mentioned, and it is important to be familiar with the difference between the two. Be aware that different DXA systems may have different names for these parameters. The age-matched or Z score is calculated as the difference between the patient’s bone density (BMD) and the normal BMD for those of the same age, sex, and race (agematched normal, or AMN), divided by the standard deviation of the normal population (SD). This is calculated by the DXA system using the following equation:

Other risk factors Monitoring • Precision more important than accuracy • Responsiveness of skeletal site measured • Appropriate follow-up time

measuring bone density, the typical clinical uses fall into three general categories: 1. confirmation of low bone mass by comparison with normative data 2. assessment of hip, spine, or overall fracture risk 3. monitoring skeletal change due to aging or in response to therapy

A. Comparisons with Normative Data For BMD measurements to be clinically useful, they need to be expressed in comparison to established normative data. All BMD manufacturers provide normative databases for this purpose. These databases are derived from measurements of large groups of both men and women of different ages and races. Comparisons are expressed both as the percentage of age-matched and young normal values, as well as standard deviation scores, that is, the number of standard deviations from the expected normal values. Percentage scores are determined as either the agematched normal BMD (AMN) or the young normal BMD (YN) using the following equations. Typically, the densitometer analysis software will calculate these values:  AMN  BMDYN   100% BMD  YN percent of age matched  1    100% YN

percent of age matched  1 

The age-matched standard deviation score is commonly referred to as the “Z score”, while the young normal standard deviation score has been labeled the “T score.” Each of these names is not a great choice, as they both have specific

Z score 

BMD  AMN . SD

The young normal or T score is defined in a similar fashion except the BMD difference is expressed in terms of the young normal (YN) bone density: T score 

BMD-YN . SD

For the diagnosis of osteoporosis, the World Health Organization has defined the following criteria for the assessment of osteoporosis based on the T score [33,34]: Normal: A BMD not more than 1 standard deviation below young normal (T score  1). Low bone mass (osteopenia): A BMD between 1 and 2.5 standard deviations below young normal (T score   1 and   2.5). Osteoporosis: A BMD 2.5 or more standard deviations below young normal (T score  2.5). Severe osteoporosis: A BMD 2.5 or more standard deviations below young normal (T score  2.5) and the presence of one or more fragility fractures. The WHO definition, although intended for use in defining populations and not for the diagnosis of osteoporosis in individual subjects, has nonetheless become commonly used for diagnosis in clinical practice. Recently, several researchers have pointed out the shortcomings of using T scores and the WHO criteria for individual diagnosis [35,36]. For example, it is known that the different skeletal sites demonstrate varied changes in T score with age. Figure 14 shows the mean age-related decline in T score for several different skeletal sites. The central skeleton, particularly the spine, shows the largest T score decline with age, while the hip and heel T score values decline less dramatically. Differences in the definition of young normal at the various skeletal sites contribute to this discrepancy. Variations in the young normal mean and standard deviation also result in significant differences in T scores obtained from different brands of instruments [36]. Currently, the various DXA manufacturers are attempting to standardize their normative data to eliminate this problem.

453

CHAPTER 59 Clinical Use of Bone Densitometry

In reality, osteoporosis is a multifactorial disease incorporating many different risk factors, of which bone density is a primary component. Age, family history, menopausal status (for women), concomitant therapies, and potential secondary causes of osteoporosis must all be carefully considered before therapeutic decisions can be made. In 1998, the National Osteoporosis Foundation (NOF) in the United States, in collaboration with 10 other professional societies, created a set of guidelines for the use and interpretation of BMD measurements [37]. They recommend BMD measurements for the following individuals:

• All postmenopausal women under the age of 65 who • • • •

have one or more additional risk factors for osteoporosis (besides menopause) All women age 65 and older regardless of additional risk factors Postmenopausal women who present with fractures Women considering therapy for osteoporosis, if BMD testing would facilitate the decision Women who have been on hormone replacement therapy for prolonged periods

Based on the bone density result, the NOF has recommended treatment for the following individuals:

• Women with BMD T scores below  2.0 in the ab-

ability for spine, heel, forearm, and hand measurements to assess risk in the elderly [40]. However, it is possible that degenerative change may influence the spinal measurements in the elderly, influencing the fracture relationship. It may be that direct spinal measurements in younger populations are a more sensitive measure of vertebral fracture risk — though this remains to be determined. To assess overall risk, spine, hip, and forearm BMD measurements appear to be relatively comparable in their predictive ability [41]. It is generally agreed that to predict fracture at a given skeletal site, a bone density measurement at that site will provide the best estimate of risk. When site-specific measurements are not available, other skeletal sites can be used to provide adequate fracture risk estimates. While the hip may have some advantage for predicting hip fracture, it is not necessary to measure the hip to predict hip fractures, just as it is not required to measure the forearm to predict forearm fractures or the spine to predict vertebral fractures. Indeed, to assess overall fracture risk, measurement of any skeletal site is acceptable. However, it is important to remember that these results were obtained primarily in older female populations, and that different recommendations may be appropriate for males and younger female populations.

sence of osteoporosis risk factors

• Women with BMD T scores below  1.5 in other risk factors are present. The NOF acknowledges that some patients (such as those over the age of 70 with multiple risk factors) are at sufficient risk for osteoporosis to warrant treatment without BMD testing.

B. Assessment of Fracture Risk Several different studies have shown a strong relationship between bone density at virtually any skeletal site and the subsequent risk for fractures of the spine, hip, and forearm [32]. The strength of this relationship depends on the type of fracture and the skeletal site measured. In most cases, the relationship between bone density and fracture is as strong or much stronger than the analogous relationship between lipid levels and coronary heart disease [33]. Almost all of the available data relating bone density to fracture risk has been obtained from elderly, primarily Caucasian, female populations. For determining hip fracture risk, direct measurements of the hip are the most sensitive predictors of hip fracture [38]. Heel measurements have also shown excellent utility for predicting hip fracture, with both SXA [39] and QUS [12,13]. For the prediction of spine fractures, data have shown a relatively comparable

C. Monitoring Skeletal Changes Changes in bone mass occur both as a consequence of aging and as a primary or secondary response to disease and/or therapy. When choosing the appropriate site for measuring bone mass changes, it is important to select a skeletal site that demonstrates the largest change, as this will maximize the potential for detection (Table 3). For assessing age-related bone loss, the metabolically active bone of the spine is often the most responsive skeletal site, particularly for women around the time of menopause [42]. With continued aging, the spine continues to be a sensitive monitor of bone loss, up until the age of 65 or so, when degenerative disease can mask age-related changes. In the elderly population, lateral spine measurements may show some advantage for evaluating age-related change by excluding the dense posterior elements [35]. Response to therapy has been studied by many different researchers. For the earliest detection of response to antiresorptive therapy, spinal measurements are preferred [43]. However, diseases or treatments known to preferentially influence cortical bone may show little effect at the spine. Hyperparathyroidism, calcium malabsorption syndromes, and renal calcium leaks will cause significant declines in cortical bone density (such as at the midshaft forearm), while spine density may remain relatively normal.

454

KENNETH G. FAULKNER

FIGURE 14 Age-related decline in mean Caucasian female T scores for different BMD technologies based on manufacturer reference ranges [36]. The DXA normative data for the PA spine (L1–L4), total hip, lateral spine (L2–L4), and forearm (1/3 region) were obtained from the Hologic (Bedford, MA) QDR-4500 densitometer. Heel normative data was taken from the estimated BMD for the Hologic Sahara ultrasound unit. Spinal QCT are those used by the Image Analysis (Lexington, KY) reference system.

D. The Use of Central and Peripheral Densitometry Based on the available studies, central densitometry (i.e., of the spine and/or hip) provides certain advantages in many clinical situations. However, it is also clear that peripheral measurements have specific utility when used appropriately in conjunction with central measurements. To understand the relative roles of peripheral and central densitometry in evaluating osteoporosis, it is helpful to consider the normal age-related course of bone density in women. From the time of peak bone mass, bone loss occurs at all skeletal sites, though somewhat faster at the spine and less so at the hip and peripheral sites. At the time of menopause, spinal bone loss is accelerated because of the relatively high metabolic activity of the vertebral trabecular bone. Within a few years of menopause, bone loss at other skeletal sites also increases, but typically only after significant changes have already occurred in the vertebral body. Around age 70, relative bone mass at all skeletal sites begins to coincide. Measurement of bone loss at the spine often begins to be influenced by degenerative change, masking the age-related changes at the spine which continue to be observed at other skeletal sites. From this model of the age course of bone mass, the relative utility of peripheral and central densitometry can be discerned [44]. In the premenopausal and early postmenopausal years (up until roughly age 65), spinal measurements typically provide the most accurate measure of skeletal state and response to aging and/or therapy. Hip

and heel measurements, as well as forearm and hand measurements, also have utility, with the caution that a normal value at the peripheral skeleton may not necessarily mean that the central skeleton is normal as well, particularly under the age of 65. In younger postmenopausal women with multiple risk factors, a normal peripheral measurement should be confirmed by a central measurement if possible [45]. In elderly women (65 and older), the potential for a false negative at the peripheral skeleton is reduced, as the relative bone mass at the different skeletal sites begins to more closely coincide. Thus for evaluating fracture risk, most skeletal sites can be expected to provide comparable information. However, the spine can be falsely elevated in many individuals in this age range, such that the PA spine measurement should be interpreted with care. If significant osteophytes, sclerosis, and/or scoliosis is suspected in the spine, either lateral spine, spinal QCT, or alternative skeletal sites should be evaluated to avoid underestimation of fracture risk.

E. Evaluating Changes in Bone Density When monitoring changes in bone density, it is crucial to confirm that any observed changes are real and not due to alterations in patient positioning, machine performance, or analysis technique. A detailed discussion of the technical factors that will influence measurement precision cannot be included here but must be considered for the proper evaluation of skeletal change. It is, however, noteworthy to reiterate that most densitometry devices (despite the fact that they are called “densitometers”) do not measure

CHAPTER 59 Clinical Use of Bone Densitometry

bone density, but rather they measure bone content and region area (or volume in the case of QCT). When reviewing the technical adequacy of a sequence of density measurements from an individual, it is wise to consider the bone content and area parameters for consistency. In particular, the area measurements should be relatively consistent across serial measurements, indicating consistent region of interest placement. If a true change in bone density has occurred, it should be reflected primarily in the bone content value. The area will also change slightly, as the computer edge detection routines identify either more or less bone depending on the direction of the change. However, the BMC and area should move in the same direction, with the majority of the difference occurring in the BMC. Basic statistical textbooks all contain an abundance of information regarding the evaluation of the significance of a difference between a group of measurements. This discussion considers the case of determining whether two bone density measurements of the same skeletal site have shown a significant interval change. In order to evaluate the significance of an interval change in bone density, it is essential to first determine the average variability, or precision error, for the measurement. Precision error is normally expressed as the coefficient of variation for a group of repeated measurements. The various bone densitometer manufacturers will often quote precision errors for their equipment on the order of 1% or better, determined by repeated measurements of a simulated spine. In clinical practice, the true precision error should be evaluated not in simulated bones but in a group of subjects representing the population normally seen by the bone density clinic. For most situations, the majority of subjects will be postmenopausal women referred for an osteoporosis evaluation. By taking a group of 14 individuals and measuring them each at least three times with interim repositioning, the precision error for that skeletal site can be estimated [46]. If multiple technologists are used to perform the bone density tests, then each technologist should perform at least one of the repeated measurements in order to include this variability factor in the experiment. To compute the estimated precision error, calculate standard deviation for each set of repeat measurements (a statistical calculator or spreadsheet program will perform this function) and then take the root mean square average across all subjects in the experiment [47]. The root mean square average standard deviation will be in units of the bone density test (usually g/cm2). To express this result as a coefficient of variation (CV), divide the standard deviation by the mean bone density of the group and express the result as a percent: CV 

SD  100%. mean BMD

455 The primary disadvantage to using the CV as a measure of precision rests in its dependence on the mean bone density of the precision group. Thus the CV will increase with decreasing bone density, not because the variability increases, but because the mean bone density goes down. If precision is expressed as the average standard deviation, the value remains essentially constant across a wide range of bone density. Once precision error has been determined, it is possible to evaluate the statistical significance of any observed change between two bone density measurements. If the precision error is small, then there is greater confidence that any observed change is real and not due to random chance. However, for the same observed change, a large precision error will diminish the confidence in the result (Fig. 15). The significance of a difference between values is expressed in statistical terms as percentage confidence. If

FIGURE 15 Illustration of percentage confidence. Shown are two sets of paired measurements, each differing by the same amount, but with different precision errors. The distribution around each measurement is defined by the standard deviation, which is related to the precision error. The percentage overlap in the two distributions is defined as the P value, while the area where the curves do not overlap is the percentage confidence. In the top example, the distributions share 37% of the area under the curves. This corresponds to a P value of 0.37 and a 63% confidence that the difference is statistically significant. In the lower example with better precision, the distribution overlap is only 4%, representing a P value of 0.04 and a 96% confidence in the difference.

456

KENNETH G. FAULKNER

FIGURE 16

(A) Baseline (left) and follow-up (right) proximal femur measurements performed 2 years apart in the same subject. Note the consistent positioning of the patient betweeen the two measurements, evidenced by the proper rotation of the femur, and the consistent projection of the lesser trochanter. Scan analysis is also consistent between baseline and follow-up, with some minor variation noted in the placement of the femoral neck region of interest. (B) Baseline (left) and follow-up (right) density measurements of the proximal femur showing significant variations in analysis. These measurements were obtained more than 3 years apart and show consistent patient positioning. However, the Ward’s region (the angled square) is both a different size and in a different location at follow-up. The femoral neck region at follow-up includes a portion of the greater trochanter, which is contrary to the manufacturer’s analysis procedures. In addition, the femoral neck region of interest in the follow-up scan includes a portion of sclerotic bone near the superior margin that was not present at baseline. This subject showed a 25% increase in femoral neck density between measurements, while the trochanteric region declined by 5% during the same time. The sclerotic changes were determined to be significantly influencing the measured density in the femoral neck, invalidating the results for this region.

457

CHAPTER 59 Clinical Use of Bone Densitometry

TABLE 4 Percent Confidence That Two BMD Measurements Are Different Based on Precision Error Precision (g/cm2) 0.020 0.030 0.040

BMD Change (g/cm2)

0.010

0.010

52

28

19

14

11

0.020

84

52

36

28

22

0.030

97

71

52

40

33

0.040

100

84

65

52

43

0.050

100

92

76

62

52

0.060

100

97

84

71

60

0.070

100

99

90

78

68

0.080

100

100

94

84

74

0.090

100

100

97

89

80

0.100

100

100

98

92

84

0.050

there is 95% confidence that two measurements are different, this suggests that there is only a 5% (5 of 100) chance that the difference is the result of random chance. The probability that a difference is the result of random chance is the P value. Thus a P value of 0.05 equates to a 95% confidence that the observed difference is real. Note that one of the benefits of statistics is that there is never 100% confidence in any result — or, in other words, statistics means never having to say you are certain. It is possible to assess the minimum significant difference between two bone density measurements given the precision error (s):

reading down the column with the appropriate precision error for the skeletal site measured. It is important to note that these confidence limits assume that the measurements are technically correct. If the measurements are inaccurate (due to calibration shift, equipment changes, or artifacts) or technically flawed (due to positioning and/or analysis errors), the statistical significance for any difference will be meaningless (Fig. 16).

VI. CONCLUSIONS Bone densitometry is a clinically accepted technique for assessing fracture risk and evaluating skeletal change. However, its proper clinical use requires an understanding of the available techniques, their appropriate application, and their potential sources of measurement error. Recent clinical guidelines recommend that all women over the age of 65, and all postmenopausal women with risk factors, should have their bone density assessed. With the advent of smaller, portable devices, bone density measurements are now widely available. In particular, ultrasound techniques, which do not use radiation, have particular promise for widespread screening applications. Peripheral densitometry alone cannot adequately address all clinical questions, particularly the question of monitoring subtle changes in bone density. For this purpose, central densitometers are still preferred. Yet for any bone density measurement to be clinically useful, it must be performed with careful attention to detail, particularly with regard to instrument calibration, patient positioning, measurement analysis, and interpretation.

at 99% confidence, 3.6s at 95% confidence, 2.8s at 90% confidence, 2.3s at 80% confidence, 1.8s. Thus for a center with a precision error of 0.01 g/cm2 at the spine, a change of 0.03 g/cm2 would be considered significant with 95 to 99% confidence. Note that it is preferable in this calculation to use the actual bone density units for both the precision and the observed change. If the CV and percent change are used, the problem of increasing CV with decreasing bone density must be considered. However, the same formulae will hold true for percentages if these are preferred. Thus for a center with 2% CV at the spine, a change of 6% would be significant at 95 to 99% confidence. To simplify the calculation of confidence, Table 4 provides a listing the percentage confidence that two BMD measurements are different based on the precision error and observed change (both in g/cm2) for a range of values. The percentage confidence for a given change is found by

References 1. D. Resnick, Osteoporosis: Radiographic – pathologic correlation. In “Osteoporosis Update 1987” (H. K. Genant, ed.), pp. 31 – 39. Radiology Research and Education Foundation, University of California, San Francisco (1987). 2. F. Cosman, B. Herrington, S. Himmelstein, et al., Radiographic absorptiometry: A simple method for determination of bone mass. Osteoporosis Int. 2, 34 – 38 (1991). 3. S. M. Garn, The, earlier gain and later loss of cortical bone. In “Nutritional Perspective” (S. M. Garn, ed.). C. C. Thomas, Springfield, IL, (1970). 4. J. R. Cameron and G. Sorenson, Measurements of bone mineral in vivo: An improved method. Science 142, 230 – 232 (1963). 5. R. H. Nord, Technical considerations in DPA. In “Osteoporosis Update 1987” (H. K. Genant, Ed.), pp. 203 – 212. Radiology Research and Education Foundation, University of California, San Francisco, (1987). 6. P. Ruegsegger, U. Elsasser, M. Anliker, H. Gnehn, H. Kind, and A. Prader, Quantification of bone mineralisation using computed tomography. Radiology 121, 93 – 97 (1976). 7. C. E. Cann and H. K. Genant, Precise measurement of vertebral mineral content using computed tomography. J. Comput. Assist. Tomogr. 4, 493 – 500 (1980).

458 8. H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gorday, Quantitative computed tomography of vertebral spongiosa: A sensitive method for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). 9. K. G. Faulkner, Bone densitometry: Choosing the proper skeletal site to measure. J. Clin. Densitom. 1(3), 279 – 285 (1998). 10. P. Steiger, J. Block, S. Steiger, A. F. Heuck, A. Friedlander, B. Ettinger, S. T. Harris, C. C. Glüer, and H. K. Genant, Spinal bone mineral density measured with quantitative CT: Effect of region of interest, vertebral level, and technique. Radiology 175, 537 – 543 (1990). 11. K. G. Faulkner, C. C. Glüer, S. Grampp, and H. K. Genant, Cross-calibration of liquid and solid QCT calibration standards: Corrections to the UCSF normative data. Osteoporosis Int. 3, 36 – 42 (1993). 12. D. Bauer, C. Glüer, J. Cauley, et al., Bone ultrasound predicts fractures strongly and independently of densitometry in older women: A prospective study. Arch. Intern. Med. 157, 629 – 634 (1997). 13. D. Hans, P. Dargent-Molina, A. Schott, J. Sebert, C. Cormier, P. Kotski, P. Delmas, J. Pouilles, G. Breart, and P. Meunier, Ultrasonic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 348, 511 – 514 (1996). 14. C. Wu, C. Glüer, Y. Lu, T. Fuerst, D. Hans, and H. K. Genant, Ultrasound characterization of bone demineralization. Calcif. Tissue Int. 62, 133 – 139 (1998). 15. W. A. Kalender, Efective dose values in bone mineral measurements by photon absorptiometry and computed tomography. Osteoporosis Int. 2, 82 – 87 (1992). 16. K. G. Faulkner, C. C. Glüer, M. Estilo, and H. K. Genant, Cross-calibration of DXA equipment: Upgrading from a Hologic QDR 1000/w to a QDR 2000. Calcif. Tissue Int. 52, 79 – 84 (1993). 17. N. F. A. Peel, A. Johnson, N. A. Barrington T. W. D. Smith, and R. Eastell, Impact of anomalous vertebral segmentation on measurements of bone mineral density. J. Bone Miner. Res. 8, 719 – 723 (1993). 18. E. S. Orwoll, S. K. Oviatt, and T. Mann, The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J. Clin. Endocrinol. Metab. 70, 1202 – 1207 (1990). 19. I. R. Reid, M. C. Evans, R. Ames, and D. J. Wattie, The influence of osteophytes and aortic calcification on spinal mineral density in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 1372– 1374 (1991). 20. J. C. H. Goh, S. L. Low, and K. Bose, Effect of femoral rotation on bone mineral density measurements with dual-energy x-ray absorptiometry. Calcif. Tissue Int. 57, 340 – 343 (1995). 21. K. G. Faulkner, H. K. Genant, and M. McClung, Bilateral comparison of femoral bone density and hip axis length from single and fan beam DXA scans. Calcif. Tissue Int. 56, 26 – 31 (1995). 22. S. L. Bonnick, D. L. Nichols, C. F. Sanborn, S. G. Payne, S. M. Moen, and C. J. Heiss, Right and left proximal femur analyses; Is there a need to do both? Calcif. Tissue Int. 57, 340 – 343 (1996). 23. P. Kannus, H. Haapasalo, H. Sievanen, P. Oja, and I. Vuori, The sitespecific effects of long-term unilateral activity on bone mineral density and content. Bone 15, 279 – 284 (1994). 24. R. C. Rupich, M. G. Griffin, R. Pacifici, L. V. Avioli, and N. Susman, Lateral dual-energy radiography: Artifact error from rib and pelvic bone. J. Bone Miner. Res. 7, 97 – 101 (1992). 25. P. Steiger, S. R. Cummings, H. K. Genant, and H. Weiss, Morphometric x-ray absorptiometry of the spine: Correlation in vivo with morphometric radiography. Osteoporosis Int. 4, 238 – 244 (1994). 26. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 27. T. J. Beck, C. B. Ruff, K. E. Warden, W. W. Scott, and G. U. Rao, Predicting femoral neck strength from bone mineral data: A structural approach. Invest. Radiol. 25, 6 – 18 (1990).

KENNETH G. FAULKNER 28. K. G. Faulkner and M. R. McClung, Quality control of DXA instruments in multicenter trials. Osteoporosis Int. 5, 218 – 227 (1995). 29. H. K. Genant, S. Grampp, C. C. Glüer, K. G. Faulkner, M. Jergas, K. Engelke, S. Hagiwara, and C. van Kuijk, Universal standardization of dual x-ray absorptiometry: Patient and phantom cross-calibration results. J. Bone Miner. Res. 9, 1503 – 1514 (1994). 30. P. Steiger, Standardization of spine BMD measurements. J. Bone Miner. Res. 10, 1602 – 1603 (1995). 31. J. Hanson, Standarization of femur BMD. J. Bone Miner. Res. 12, 1316 – 1317 (1997). 32. D. Marshall, O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312(7041), 1254 – 1259 (1996). 33. The WHO Study Group, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis.” World Health Organization, Geneva (1994). 34. J. A. Kanis, Assessment of fracture risk and is application to screening for postmenopausal osteoporosis: Synopsis of a WHO report. Osteoporosis Int. 4, 368 – 381 (1994). 35. S. L. Greenspan, L. Maitland-Ramsey, and E. Myers, Classification of osteoporosis in the elderly is dependent on site-specific analysis. Calcif. Tissue Int. 58, 409 – 414 (1996). 36. K. G. Faulkner, E. von Stetten, and P. Miller, Discordance in patient classification using T-scores. J. Clin. Densitom. 2(3), 343 – 350 (1999). 37. National Osteoporosis Foundation, “Physician’s Guide to Prevention and Treatment of Osteoporosis.” National Osteoporosis Foundation, Washington, DC, (1998). 38. S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, K. Ensrud, H. K. Genant, L. Palermo, J. Scott, and T. M. Vogt, Bone density at various sites for prediction of hip fractures. Lancet 341, 72 – 75 (1993). 39. S. R. Cummings, D. M. Black, M. C. Nevitt, W. S. Browner, J. A. Cauley, H. K. Genant, S. R. Mascioli, J. C. Scott, D. G. Seeley, P. Steiger, and T. Vogt, Appendicular bone density and age predict hip fracture in women. JAMA 263(5), 665 – 668 (1990). 40. P. Ross, C. Huang, J. Davis, K. Imose, J. Yates, J. Vogel, et al., Predicting vertebral deformity using bone densitometry at various skeletal sites and calcaneus ultrasound. Bone 16, 325 – 332 (1995). 41. L. Melton, E. Atkinson, W. O’Fallon, H. Wahner, and B. Riggs, Long-term fracture prediction by bone mineral assessed at different sites. J. Bone Miner. Res. 8, 1227 – 1233 (1993). 42. S. Harris and B. Dawson-Hughes, Rates of change in bone mineral density of the spine, heel, femoral neck, and radius in healthy postmenopausal women. Bone Miner. 17, 87 – 95 (1992). 43. K. G. Faulkner, M. R. McClung, P. Ravn, D. J. Hoskin, R. D. Wasnich, M. Daley, and A. J. Yates, Monitoring skeletal response to therapy in early postmenopausal women: Which bone to measure? J. Bone Miner. Res. 11(Suppl. 1), S96 (1996). 44. D. T. Baran, K. G. Faulkner, H. K. Genant, P. D. Miller, and R. Pacifici, Diagnosis and management of osteoporosis: Guidelines for the utilization of bone densitometry. Calcif. Tissue Int. 61, 433 – 440 (1997). 45. P. D. Miller, S. L. Bonnick, C. C. Johnston, M. Kleerekoper, R. L. Lindsay, L. Sherwood, and E. S. Siris, The Challenges of peripheral bone density testing: Which patients need additional central density skeletal measurements? J. Clin. Densitom. 1(3), 211 – 217 (1998). 46. C. C. Glüer, G. Blake, Y. Lu, B. A. Blunt, and H. K. Genant, Accurate assessment of precision errors: How to measure the reproducibility errors of bone densitometry techniques. Osteoporosis Int. 5, 262–270 (1995). 47. S. L. Bonnick, “Bone Densitometry in Clinical Practice.” Humana Press, Totowa, (1998).

CHAPTER 60

Biochemical Markers of Bone Turnover in Osteoporosis PATRICK GARNERO*,† AND PIERRE D. DELMAS*,‡ *

INSERM Research Unit 403, Hôpital E. Herriot, and †Synarc, and ‡ Claude Bernard University, Lyon 69003, France

I. Introduction II. Biochemical Markers of Bone Formation III. Biochemical Markers of Bone Resorption

I. INTRODUCTION Osteoporosis is a disease characterized by a low bone mass and by architectural deterioration of bone tissue that are both related to abnormalities of bone turnover. Bone turnover is characterized by two opposite activities, the formation of new bone by osteoblasts and the degradation (resorption) of old bone by osteoclasts. Both are tightly coupled in time and space in a sequence of events which define the same remodeling unit. Bone mass depends on the balance between resorption and formation within a remodeling unit and on the number of remodeling units which are activated within a given period of time in a defined area of bone. Osteoporosis is characterized by both an imbalance between resorption and formation within the remodeling unit, and by an increased activation frequency, the latter being responsible for the significant increase of bone turnover after the menopause. The rate of formation or degradation of bone matrix can be assessed either by measuring an enzymatic activity of the bone forming or resorbing cells — such as alkaline and

OSTEOPOROSIS, SECOND EDITION VOLUME 2

IV. Clinical Uses of Bone Markers in Osteoporosis References

acid phosphatase — or by measuring bone matrix components released into the circulation during formation or resorption (Table 1). These have been separated into markers of formation and resorption, but it should be kept in mind that in disease states where both events are coupled and change in the same direction, any marker will reflect the overall rate of bone turnover. Bone markers cannot discriminate between turnover changes in a specific skeletal envelope, i.e., trabecular versus cortical, but reflect whole body net changes. In osteoporosis, bone turnover markers have been suggested to predict the rate of postmenopausal bone loss, to predict the occurrence of osteoporotic fractures, and to monitor the efficacy of treatment, especially anti-resorptive therapy (hormone replacement therapy, bisphosphonates, and calcitonin). It has also been suggested that measurement of bone turnover before treatment might be useful to select the type of therapy (anti-resorptive or bone stimulating agent) and to predict the amplitude of the response to estrogen and bisphosphonate treatment, but there is only little solid evidence for these two latter concepts. In this paper, we will review the recent development in bone marker

459

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

460

GARNERO AND DELMAS

TABLE 1

Biochemical Markers for Bone Remodeling

Formation Serum Osteocalcin (bone Gla-protein) Total and bone alkaline phosphatase Collagen type I C-terminal and N-terminal propeptides (PICP and PINP) Resorption Plasma/serum Tartrate-resistant acid phosphatase (TRACP) Free pyridinoline and deoxypyridinoline C-terminal cross-linking telopeptide of type I generated by MMPs (CTX-MMP) N-terminal (S-NTX) and C-terminal (S-CTX) crosslinking telopeptide of type I collagen Bone sialoprotein (BSP) Urine Free pyridinoline and deoxypyridinoline N-terminal (U-NTX) and C-terminal (U-CTX) crosslinking telopeptide of type I collagen Calcium Hydroxyproline Galactosylhydroxylysine Note. The markers with the best performance characteristics in osteoporosis are in bold type.

technology and then discuss the use of these markers for the management of osteoporosis.

II. BIOCHEMICAL MARKERS OF BONE FORMATION

the bone compartments. Cortical bone has about twofold higher activity of B1 compared to B2, whereas trabecular bone has twofold higher activity of B2 compared with B1, suggesting that measurements of specific bone ALP activity may provide information of bone metabolism within specific bone envelopes [2]. Serum total alkaline phosphatase (total ALP) activity is the most commonly used marker of bone formation but for the above reasons it lacks sensitivity and specificity. Nevertheless, several studies have shown that its activity increases with aging in adults, especially in women after menopause [3]. In patients with vertebral osteoporosis, values are either normal or slightly elevated and correlate poorly with bone formation determined by iliac crest histomorphometry [4,5]. Also, a moderate increase in serum alkaline phosphatase activity is ambiguous as it may reflect a mineralization defect in elderly patients or the effect of one of the numerous medications which have been shown to increase production of the hepatic isoenzyme. In an attempt to improve the specificity and sensitivity of serum alkaline phosphatase measurement, techniques have been developed to differentiate the bone and the liver isoenzymes which differ only by posttranslational modifications as they are coded by a single gene. These techniques rely on the use of differentially effective activators and inhibitors (heat, phenylalanine, and urea), separation by electrophoresis, and indirect separation by liver-specific antibodies [6 – 8]. In general, these assays have slightly enhanced the sensitivity of this marker, but most of them are indirect and/or technically cumbersome. A real improvement has been achieved by using a monoclonal antibody that preferentially recognizes the bone isoenzyme. These direct immunoassays have been shown to exhibit a low cross-reactivity with the circulating liver isoenzyme (15 to 20%) and to be more sensitive than the total ALP activity to detect the increase in bone turnover following menopause [9,10].

A. Serum Alkaline Phosphatase The skeletal alkaline phosphatase (ALP) is an enzyme localized in the membrane of osteoblasts that is released into the circulation by an unclear mechanism. Among the several tissues containing alkaline phosphatase, the liver and bone isoenzymes are the major contributors to the serum level. The intestinal isoenzyme can also account for part of the circulating levels in some nonfasting patients, and placental alkaline phosphatase circulates during pregnancy. In serum, bone alkaline phosphatase (bone ALP) exists in three different isoforms including B/I (70% of bone and 30% of intestinal ALP), B1, and B2, the major forms being B1 and B2 which can be differentiated by high-performance liquid chromatography (HPLC) [1]. It has recently been shown that trabecular bone had higher total (B1  B2) ALP activity than cortical bone. In addition, the distribution of the two isoforms seems also different between

B. Serum Osteocalcin or Bone Gla Protein Osteocalcin (OC), also called bone gla-protein, is a small noncollagenous protein that is specific for bone tissue and dentin. Its precise function remains unknown [11] although recent data in osteocalcin-deficient mice suggest that OC could limit in vivo bone formation [12]. OC is predominantly synthesized by osteoblasts and is incorporated into the extracellular bone matrix, but a fraction of newly synthesized osteocalcin is released into the circulation where it can be measured by radioimmunoassay [13 –16]. OC mRNA has been detected in bone marrow megakaryocytes and peripheral blood platelets but the protein itself was undetectable in human platelets [17], suggesting that platelet osteocalcin is unlikely to contribute significantly to either serum or plasma levels. Circulating OC has a short

461

CHAPTER 60 Biochemical Markers of Bone Turnover

half-life and is rapidly cleared by the kidney [14,18]. Serum OC concentrations correlate with skeletal growth at the time of puberty, and are increased in a variety of conditions characterized by increased bone turnover, such as primary and secondary hyperparathyroidism, hyperthyroidism, Paget’s disease, and acromegaly. Conversely, they are decreased in hypothyroidism, hypoparathyroidism, in glucocorticoid-treated patients, and in some patients with multiple myeloma and malignant hypercalcemia [reviewed in 19]. Comparisons of serum OC values with iliac crest histomorphometry and calcium kinetic data have shown that under most conditions, osteocalcin is a valid marker of bone turnover when resorption and formation are coupled and is a specific marker of bone formation whenever formation and resorption are uncoupled [20 – 24]. Using a battery of monoclonal antibodies directed against various epitopes of the human osteocalcin molecule, we found that the intact molecule represents about one-third of the immunoreactivity in the adult serum (or plasma). One-third is represented by several small fragments and another third by a large N-terminal – mid-molecule fragment (N-mid fragment) [25]. This large fragment (about 43 amino acids compared to 49 for the intact molecule) is generated in vitro by osteoblastic cells in culture and circulates in vivo. After few hours at room temperature, a significant fraction of plasma intact osteocalcin is rapidly converted into the large N-mid fragment, resulting in a significant loss of immunoreactivity with most polyclonal antibodies because they recognize the C-terminal end of the molecule . This pattern could account for some of the discrepancies reported in the literature [26], and for the surprisingly very wide scatter of individual values observed in several studies. From a practical point of view, measuring both the intact molecule and the N-mid fragment with an assay using appropriate antibodies results in a more robust and sensitive assay. In addition, such an assay reduces by 50% the long-term precision error when osteocalcin measurement is repeated over months in a single patient.

serum bone ALP and OC concentration in osteoporotic patients treated with fluoride, serum PICP levels decreased under therapy [29]. Conversely, the decrease of serum PICP in osteoporotic patients treated with estrogen is consistent with the decrease of serum bone ALP and OC [30]. In contrast to serum PICP, which is a single protein with a size of the authentic propeptide [31], immunoreactive PINP circulates as two major components: the intact form corresponding to the authentic in vivo cleaved propeptide and a low-molecular-weight form [32]. Recently it has been shown that these two circulating forms represent intact alpha chains of type I collagen in trimeric and monomeric forms respectively [33]. The trimeric structure is unstable at 37°C and can be transformed to the stable monomeric forms. The first developed assays for serum PINP recognize both trimeric and monomeric forms [34] and, as serum PICP, have also been disappointing in osteoporosis [35, 36]. In contrast, the assay recognizing specifically the intact trimeric form of PINP [32] has been shown to be more sensitive than PICP, and of similar performance to OC and bone ALP for detecting the increase in bone turnover following menopause [37] and for monitoring the response to antiresorptive therapy [37 – 40]. The reasons for the different sensitivity of PICP and PINP are not clear, but could be related to the contribution of tissues other than bone to PICP [34] and/or to their metabolism. Both propeptides are not cleared by the kidney because they are either too large (PICP) or have an elongated shape (PINP). The propeptides are actively taken up and metabolized by the endothelia cells of the liver, via the mannose receptor for PICP [41] and the scavenger receptor for the PINP [42]. These two clearance systems are independent of each other and also seem to be regulated separately by hormones such as thyroid hormones and IGF-I [43].

III. BIOCHEMICAL MARKERS OF BONE RESORPTION

C. Procollagen Type I Propeptides

A. Fasting Urinary Calcium, Hydroxyproline, and Hydroxylysine Glycosides

During the extracellular processing of type I collagen, there is cleavage of the aminoterminal (PINP) and carboxyterminal (PICP) extension peptides prior to fibril formation. These peptides circulate in blood where they might represent useful markers of bone formation since collagen is by far the most abundant organic component of bone matrix. Serum PICP concentrations correlate weakly with histological bone formation in patients with vertebral osteoporosis (r  0.36 – 0.50) [27]. The menopause induces a significant but marginal ( 20%) increase in serum PICP concentration which is not correlated with the subsequent rate of bone loss measured by densitometry [28]. In contrast to significant increases in both

Fasting urinary calcium measured on a morning sample and corrected for creatinine excretion is certainly the cheapest assay of bone resorption. It is useful in detecting a marked increase of bone resorption but lacks sensitivity. Fasting urinary calcium reflects the amount of calcium released during resorption, but also the renal handling of calcium that is influenced by calcium regulating hormones and by estrogens. Hydroxyproline (Hyp) is found mainly in collagen and represents about 13% of the amino acid content of the molecule [44]. Hyp is derived from proline by a posttranslational hydroxylation occurring within the peptide chain. Because free Hyp released during degradation of

462 collagen cannot be reutilized in collagen synthesis, most of the endogenous Hyp present in biologic fluids is derived from the degradation of various forms of collagen [45]. Since half of human collagen resides in bone, where its turnover is probably faster than in soft tissues, excretion of Hyp in urine is regarded as a marker of bone resorption. Actually, the C1q fraction of complement contains significant amounts of Hyp and could account for up to 40% of urinary Hyp [46]. About 90% of the Hyp released by the breakdown of collagen in the tissues, and especially during bone resorption, is degraded to the free amino acid that circulates in plasma, is filtered, and is almost entirely reabsorbed by the kidney. It is eventually completely oxidized in the liver and is degraded to carbon dioxide and urea [47,48]. About 10% of the Hyp released by the breakdown of collagen circulates in a peptide-bound form, and these peptides are filtered and excreted in urine without any further metabolism. Thus, the urinary total Hyp represents only about 10% of total collagen catabolism. Colorimetric assay of Hyp is usually performed on a hydrolyzed urine sample and therefore reflects the total excretion of the amino acid. As a consequence of its tissue origin and metabolism pattern, urinary hydroxyproline is poorly correlated with bone resorption assessed by calcium kinetics or bone histomorphometry [19]. Hydroxylysine is another amino acid unique to collagen and proteins containing collagen-like sequences. Like hydroxyproline, hydroxylysine is not reutilized for collagen biosynthesis, and although it is much less abundant than hydroxyproline, it is a potential marker of collagen degradation. Hydroxylysine is present in part as galactosylhydroxylysine and in part as glucosyl-galactosyl-hydroxylysine. The relative proportion and total content of galactosyl hydroxylysine and glucosyl-galactosyl hydroxylysine varies in bone and soft tissues, with a higher content of galactosyl hydroylysine in bone suggesting that its urinary excretion might be a more sensitive marker of bone resorption than urinary hydroxyproline. Urinary galactosyl hydroxylysine increases with aging [49]. Recently a serum assay for free galactosyl hydroxylysine has been reported [50]. Using this serum assay, increased concentrations were found in patients with Paget’s disease and these decreased after treatment with the bisphosphonate etidronate with a magnitude similar to urinary galactosyl hydroxylysine. Finally an increase of urinary excretion of galactosyl hydroxylysine has recently been reported in women with a history of fracture compared to controls despite similar levels of bone resorption between the two groups as assessed by the urinary excretion of total deoxypyridinoline [51]. These data suggest that increased excretion of galactosyl hydroxylysine may reflect overglycosylation of hydroxylysine in

GARNERO AND DELMAS

bone matrix that may be associated with changes in bone collagen quality and increased susceptibility to fracture, an intriguing hypothesis that needs to be confirmed by prospective studies.

B. Plasma Tartrate-Resistant Acid Phosphatase Acid phosphastase is a lysosomal enzyme that is present primarily in bone, prostate, platelets, erythrocytes, and spleen. These different isoenzymes can be separated by electrophoretic methods, which lack sensitivity and specificity. Bone acid phosphatase is resistant to L()tartrate, whereas the prostatic isoenzyme is inhibited [52]. Acid phosphatase circulates in blood and shows higher activity in serum than in plasma because of the release of platelet phosphatase activity during the clotting process. In normal plasma, tartrate-resistant acid phosphatase (TRACP) corresponds to plasma isoenzyme 5, which originates partly from bone, as osteoclasts contain TRACP that is released into the circulation [53]. Isoenzyme 5 is represented by two subforms, 5a and 5b, and recent data indicate that TRACP 5b is more specific for osteoclasts [54]. Recently antibodies have been developed against TRACP isolated from spleens of patients with hairy leukemia [55], human cordon [56], or more recently recombinant human TRACP or bone TRACP [57,58] to develop immunoassays. Plasma TRACP activity is increased in a variety of metabolic bone disorders with increased bone turnover [59] and is elevated after oophorectomy [60] and in vertebral osteoporosis [61], but it is not clear whether this marker is actually more sensitive than urinary hydroxyproline [60]. When measured by immunoassay, serum TRACP has been shown to increase by about 40% after the menopause [58], but no extensive comparison with other resorption markers has yet been reported. Part of the instability of TRACP in plasma is related to the fact that this enzyme forms complexes with 2 macroglobulin which affects not only TRACP activity, but also immunological stability [62]. This complex can be only partly disrupted by acidification and or calcium chelators such as EDTA [57,62]. The lack of specificity of plasma TRACP activity for the osteoclast, its instability in frozen samples, and the presence of enzyme inhibitors in serum, are potential drawbacks which will limit the development of clinically useful enzymatic TRACP assays in osteoporosis.

C. Collagen Pyridinium Crosslinks and Associated Type I Collagen Peptides Pyridinoline (PYD) and deoxypyridinoline (DPD), also called respectively hydroxylysylpyridinoline (HP) and ly-

CHAPTER 60 Biochemical Markers of Bone Turnover

463

FIGURE 1

Type I collagen breakdown products as markers of bone resorption: Type I collagen molecules in bone matrix are linked by pyridinoline crosslinks (pyridinoline or deoxypyridinoline) in the region of N- and C-telopeptides. Pyridinoline (PYD) differs from deoxypyridinoline (DPD) by the presence of an hydroxyl residue shown in italic. During osteoclastic bone resorption, pyridinoline crosslinks are released into the circulation mainly as peptide-bound crosslinks, i.e., attached to fragments of C-terminal (CTX) or N-terminal (NTX) telopeptides. Part of peptide bound crosslinks are further degraded in the kidney in free crosslinks. Immunoassays detecting specifically free PYD, free DPD, CTX, and NTX in serum or urine are available.

sylpyridinoline (LP), are two nonreducible pyridinium crosslinks present in the mature form of collagen. This posttranslational covalent cross-linking generated from lysine and hydroxylysine residues is unique to collagen and elastin molecules. It creates interchain bonds which stabilize the molecule within the extracellular matrix. The highest concentration of PYD (expressed in mol/mol of collagen) is found in articular cartilage, while DPD is present in minute amounts in this tissue [63]. PYD and DPD are present in tendon and aorta but absent from the skin, an abundant source of type I collagen [63]. As both crosslinks result from a posttranslational modification of collagen molecules, they cannot be reused during collagen synthesis. PYD and DPD are released into the circulation after resorption of bone matrix and then excreted in urine where they are found as free and peptide-bound forms. It has been shown that the proportion of free crosslinks is about twofold lower in serum than in urine (16 – 20% vs 40%) and that the renal clearance is about fourfold higher for free than for peptide-bound crosslinks [64], suggesting a cleavage of peptide-bound crosslinks to free forms in the kidney (Fig. 1). If the conversion of peptide-bound to free crosslinks is saturable, this could explain part of the inverse relationship between increased bone turnover and decreased free crosslink fraction in urine [65]. The total amount of pyridinoline crosslinks can be measured by fluo-

rimetry after reversed-phase HPLC of a cellulose-bound extract of hydrolyzed urine [66,67]. In patients with vertebral osteoporosis, the urinary crosslink levels, especially of DPD, are correlated with bone turnover measured by calcium kinetics [68] and bone histomorphometry [69], in contrast to the poor correlations obtained with urinary Hyp. Immunoassays for the PYD crosslinks and related telopeptide fragments which advantageously substitute for the HPLC measurement of the total excretion are now available and represent currently the best indices of bone resorption. These comprise measurement of urinary free PYD (fPYD) and DPD (fDPD) [70,71] and of related peptides in urine and more recently in serum. These peptides include the C-terminal cross-linking telopeptide of type I collagen generated by matrix metalloproteases (CTX-MMP) [72], the C-terminal cross-linking telopeptide of type I collagen in serum (CTX-MMP) [73] and in urine (U-CTX) [74] and the N-terminal cross-linking telopeptide of type I collagen in serum (S-NTX) [75] and in urine (U-NTX) [76] (Fig. 1 and Table 2). The first peptide assay was a RIA for the S-CTX-MMP in serum [72]. The antibody was raised against a crosslink-containing collagen peptide (MW 8.5 kDa) isolated by trypsin digestion of human bone collagen. The antigenic determinant requires a trivalent crosslink, including two phenylalanine-rich domains of the telopeptide region of the alpha-1 chain of type I collagen.

464

GARNERO AND DELMAS

TABLE 2 Correlation between Baseline Levels of Biochemical Markers of Bone Turnover and Forearm Rate of Bone Loss in Postmenopausal Women: Effect of Measurement Errors Observed r value

r Value corrected for Error on bone loss

Error on bone marker

Error on rate of bone loss and bone marker

OC

 0.53

 0.72

 0.68

 0.91

PINP

 0.40

 0.54

 0.48

 0.66

U-NTX

 0.39

 0.53

 0.56

 0.76

U-CTX

 0.40

 0.55

 0.59

 0.81

S-CTX

 0.47

 0.64

 0.58

 0.79

Note. This study included 51 untreated women within 5 years of menopause from the OFELY prospective study [121]. The rate of bone loss was calculated in each woman by the slope of the regression obtained from four mid-radius BMD measurements by DXA performed at baseline, year 2, year 3, and year 4. The table shows the correlation coefficients (r) between levels of bone markers measured at baseline and the rate of bone loss over 4 years before and after correction for error in the estimation on individual rate of bone loss, error in individual marker levels, and both. The error on the rate of bone loss was estimated from the mean of the standard errors of the estimate for individual regression analyses of BMD against time. The error on the bone turnover marker levels was estimated from published intraindividual coefficients of variation of each marker. Correction of the observed correlation coefficients by errors on bone loss and bone marker estimations were based on the formula proposed by Hassager et al. (1991) (C. Hassager, S. B. Jensen, A. Gotfredsen, and C. Christiansen. The impact of measurement errors on the diagnostic value of bone mass measurements: Theoretical consideration. Osteoporosis Int. 1, 250–256 (1991)).

The tissue specificity and the clinical significance of CTXMMP levels are, however, unclear. In particular, CTXMMP concentrations increase after treatment with anabolic steroids which are believed to decrease bone resorption and to stimulate collagen synthesis [77]. Nevertheless, and in contrast to its poor sensitivity in osteoporosis [78], CTX-

MMP appears to be a useful bone resorption marker in patients with bone metastases from prostate cancer [79,80]. The discrepancy in the behavior of this marker in these two clinical situations is unclear. One hypothesis could be related to differences in the pattern of type I collagen degradation between osteoporosis and bone metastases because

FIGURE 2 Schematic representation of the different type I collagen telopeptide epitopes used as markers of bone resorption and sites of cleavage by cathepsin K (cat K) on type I collagen. The NTX epitope and CTX epitopes in the N and C telopeptide regions respectively are efficiently generated by cat K — the main enzyme responsible for type I collagen degradation in physiological conditions — but not by matrix metalloproteases (MMP) including MMP 1 and MMP 9 [83] which have been proposed to participate in bone resorption in physiological conditions, but also in metastatic processes. In contrast, CTX-MMP epitope is destroyed by the action of cat K but not by MMPs [81,83].

465

CHAPTER 60 Biochemical Markers of Bone Turnover

of changes in the relative contribution of the various collagenolotytic enzymes, namely the cysteine proteases and the matrix metalloproteinases (MMP). It has recently been shown that the epitope of CTX-MMP is destroyed by cathepsin K activity [81], an osteoclasticspecific cysteine protease which is likely to the key enzyme responsible for bone collagen degradation in osteoporosis [82,83], whereas it is not altered by MMPs thought to play an important role in collagen degradation associated with cancer (Fig. 2). The degradation of CTX-MMP epitope by cathepsin K probably explains the failure of this marker to increase in postmenopausal women and its paradoxical increase in patients with pycnodosystosis — a disease characterized by a mutation in the cathepsin K gene — contrasting with the expected decrease of CTX and NTX under this condition [84]. This is an intriguing hypothesis which requires further investigation and which may give insights into the pathophysiology of bone metastases. We recently used a new serum bone resorption assay based on two monoclonal antibodies directed against an eight-amino-acid sequence from the C-telopeptide of type I collagen (S-CTX). The epitope recognized by this assay is not destroyed by cathepsin K, but actually generated by this enzyme [83] such as the epitope recognized by the NTX assay (Fig. 2). We found that S-CTX increased by threefold compared to controls [85] with a sensitivity for detecting bone metastases which compared very well with that of urinary measurements of CTX (U-CTX) (Fig. 3). Urinary and serum markers of bone resorption have significant circadian rhythms [86 – 98]. Urinary excretion of pyridinoline and deoxypyridinoline peaks between

02:00 and 08:00 and reaches a nadir between 14:00 and 23:00. In healthy premenopausal women the magnitude of the rhythm may be as much as 100% of the 24-h mean with a decrease of 25 – 35% between 08:00 and 11:00 [93]. Urinary crosslinked telopeptides have similar rhythms as total crosslink excretion [87,88,92]. Serum CTX peaks between 01:30 and 04:30, with concentrations that are more than twice those at the nadir between 11:00 and 15:00 [90,98]. The amplitude of the rhythm of CTXMMP and S-NTX is only of 15 – 20% of the 24-h mean [89,96], suggesting that indeed the different type I collagen peptides may have different bone specificity and/or reflect different aspects of bone resorption. Circadian rhythms of markers of bone resorption may be attenuated by several factors Evening calcium supplementation may diminish the rhythmicity of markers of bone resorption, but this effect may depend whether it is given over a period of weeks or as a single dose [87,94]. Bisphosphonate therapy reduces the amplitude of the circadian rhythm of urinary NTX, although the pattern is maintained [95,96]. In two recent studies, fasting has been shown to result in a diminution in the amplitude of the circadian rhythm of urinary and serum CTX [97,98]. This effect remains to be investigated for the other resorption markers including the total urinary excretion of pyridinoline crosslinks, but, if confirmed, could give new insights in the pathophysiology and treatment of osteoporosis. In summary, although the amplitude of the circadian rhythm differs according to the markers, its effect is substantial for most of them and indicate the importance of standardizing the time of sampling.

FIGURE 3 Increased levels of bone formation (osteocalcin, bone ALP, and PICP) and bone resorption (urinary  CTX, urinary  CTX, and serum CTX) markers in patients with prostate cancer and bone metastases. The bars represent the median increase in the levels of markers in 39 patients with bone metastases from prostate cancer compared to levels of 355 healthy age-matched men. Reprinted with permission from Garnero et al. [85]. * P  0.0001 vs controls.

466

GARNERO AND DELMAS

FIGURE 4

Racemization and isomerization of type I collagen C-telopeptides. An attack by a peptide backbone nitrogen on the side chain carbonyl group of an adjacent aspartyl residue can result in the formation of a succinimide ring (I : II). The succinimide ring is prone to hydrolysis and racemization producing peptides and -aspartyl peptides in both the D and L configurations. Racemization is thought to proceed primarily through the succinimide pathway (II : V), but other pathways as direct proton abstraction (I 4 IV) and III 4 VI) may also contribute to the formation of D-aspartyl. Throughout the figure, the peptide backbone is shown as a bold line. The four types of C-telopeptides are present in bone matrix, the native form (L) and three age-related forms: an isomerized (L), a racemized (D), and an isomerized/racemized (D) form. With increasing age of type I collagen molecules, the proportion of  isomerized and D racemized form within bone matrix increases. Degradation products of these four CTX forms of type I collagen can be measured in urine independently by immunoassays using specific conformational monoclonal antibodies. Reprinted with permission from Cloos and Fledelius [100].

Posttransalation modifications of type I collagen other than pyridiroline crosslinks occur in bone matrix and some of them may be of clinical relevance for the investigation of metabolic bone diseases. Probably one of the most interesting of these modifications is the racemization and  isomerization of the Asp – Gly sequence in the 1209AHDGGR1214 sequence (CTX) of the C-telopeptide of type I collagen [99,100] generating four isomers: the native form (L) and three age- related forms: an isomerized (L), a racemized (D), and an isomerized/racemized (D) form (Fig. 4). The relative rate at which the age-related CTX-isoforms are generated in bone matrix has been shown to be L D D [100]. Thus, L-CTX reflects the resorption of newly synthesized bone, whereas L-, D-, and D-CTX reflect the degradation of aged bone, old bone and very old bone, respectively. Histological studies have shown a decreased degree of isomerization/racemization within the woven pagetic bone reflected by increased urinary L/L and L/D ratios [101, 102]. In other bone diseases also characterized by an increased bone turnover but without bone ma-

trix defect, such as primary hyperparathyroidism, the  CTX /-CTX ratio remains normal [101]. These results suggest that the newly synthesized collagen fibers found in the woven pagetic bone are characterized by a marked decrease in the degree of isomerization. The urinary ratio -CTX/ CTX returned to the normal range in most patients after several months of treatment with bisphosphonate, probably reflecting the progressive replacement of woven bone by a lamellar bone with a higher and normal degree of  isomerization of type I collagen, as previously documented by histology [102] (Fig. 5). Because patients with bone metastases from prostate cancer are also characterized by a marked increase of bone turnover in localized areas of the skeleton with the replacement of lamellar by woven bone [103] one could expect a higher increase of L-CTX compared to L-CTX and D-CTX leading to increased L/L and L/D ratio in these patients. In a recent study involving 39 patients with prostate cancer and bone metastases [85], although the L/L ratio was normal, the ratio between the urinary degradation products of the newly formed (L) and

467

CHAPTER 60 Biochemical Markers of Bone Turnover

Effect of bisphosphonate treatment on the urinary ratio of nonisomerized to -isomerized C-telopeptide breakdown products (-CTX/-CTX) in Paget’s disease. Twenty-eight patients with active Paget’s disease of bone were studied before and 2 months after a single injection of the bisphosphonate zoledronate (200 or 400 g) or placebo. Before treatment most of patients (93%) are characterized by elevated urinary -CTX/-CTX ratio with a mean 2.7-fold increase compared to controls (2.1 vs 0.8). Two months after treatment, the -CTX/-CTX was decreased by 50% in zoledronate-treated patients and returned within the normal range (gray area, mean 2 SD) in 65% of patients. No significant change was observed in the placebo-treated group. Reprinted with permission from Garnero et al. [102].

FIGURE 5

the oldest (D) forms of collagen was increased by twofold compared to controls. These data suggest that indeed the pattern of type I collagen isomerization/racemization is altered in sclerotic metastases and that the urinary L/D ratio may represent a useful index to assess abnormalities of bone matrix in that situation. Clearly these findings open new perspectives for the clinical use of bone markers, not only to measure quantitative changes of bone turnover, but also to assess changes of bone quality.

D. Bone Sialoprotein (BSP) BSP is a phosphorylated glycoprotein with an apparent MW of 70 – 80 kDa and accounts for 5 – 10% of the noncollageneous matrix of bone [104]. The protein has been shown to be a major synthetic product of active osteoblasts and odontoblasts, but was also found in osteoclast-like and malignant cell lines [105 – 107]. Compared to other noncollagenous proteins, the tissue distribution of BSP is quite restricted and mRNA has been detected in dentin and in bone with higher BSP concentration in the osteocartilagenous interfaces [108] that are involved early in joint diseases such as osteoarthritis and rheumatoid arthritis. However, BSP

immunoreactivity has also been detected in platelets [109]. BSP contains an Arg – Gly – Asp (RGD) integrin recognition sequence and binds preferentially to the 2 chain of collagen [110]. It nucleates hydroxyapatite crystal formation in vitro [111] and appears to enhance osteoclast mediated bone resorption [112], suggesting that this protein may play an important role in the local regulation of bone remodeling and in the organization of the extracellular matrix of mineralized tissues. Several immunoassays based on polyclonal antibodies raised against bovine or human BSP have been developed [109,113,114]. However, little is known about the exact nature of the respective epitopes recognized by these antibodies. Whether the assays recognized the intact newly synthesized molecule or fragments released during resorption of bone matrix and/or by proteolytic degradation of the protein in the circulation remain to be established. Increased serum BSP concentrations have been reported in patients with different metabolic bone diseases including Paget’s disease, primary hyperparathyroidism, patients with bone metastases and women with postmenopausal osteoporosis [115,116]. Based on these clinical data and the rapid decrease of serum BSP levels following intravenous bisphosphonate treatment, it is likely that serum BSP reflects processes mainly related

468

GARNERO AND DELMAS

to bone resorption [115]. Clearly the clinical utility of serum BSP in osteoporosis compared to other bone turnover markers require further investigation.

IV. CLINICAL USES OF BONE MARKERS IN OSTEOPOROSIS A. Bone Markers and Rate of Bone Loss Several cross-sectional studies indicate that bone turnover increases rapidly after the menopause. In contrast with what was originally thought, it has been shown that this increase in both bone formation and bone resorption is sustained long thereafter, up to 40 years [117]. BMD measured at various skeletal sites correlated negatively with bone turnover assessed by various markers in postmenopausal women. We have shown that the correlation between bone markers and BMD strengthens greatly with advancing age, so that in women more than 30 years after the menopause bone turnover accounts for 40 to 50% of the variance of bone mineral density of the whole skeleton [117]. These cross-sectional data suggest that a sustained increase of bone turnover in postmenopausal women induces a faster bone loss and therefore an increased risk of osteoporosis. Longitudinal studies, which are required to confirm that hypothesis, suffer from methodological issues. Indeed, when the rate of bone loss is assessed by annual measurement of BMD at the spine, hip, or radius over 2 to 4 years, the amount of bone loss is in the same order of the magnitude of the precision error of repeated measurements in a single individual, i.e., from 3 to 4%. This technical limitation impairs a valid assessment of the relationship between bone turnover and the subsequent rate of bone loss in individual postmenopausal women, and probably explains the conflicting results that have been published. When the precision error on the estimation of the rate of bone loss is reduced by performing nine BMD measurements over 24 months at a highly precise skeletal site such as the radius, the ability of baseline bone markers to predict rate of loss is markedly improved with correlation coefficients increasing from about 0.2 to 0.4 when rate of loss is assessed from yearly bone mass measurements to 0.7 – 0.8 [118]. Ultimately, the proof of evidence will come from long-term studies. At the present time, two such studies show that the assessment of bone turnover was correlated with the spontaneous rate of forearm and calcaneal bone loss over the next 12 – 13 years [119,120]. The first shows that, despite identical bone mass at baseline, women who were classified as fast losers at the time of menopause based on conventional bone markers including total alkaline phosphatase and urinary hydroxyproline had lost 50% more bone 12 years later than those diagnosed as slow losers (total bone

loss 26.6% versus 16.6%, P  0.001) [119]. A more recent study [120] performed over 13 years in older women (mean age at baseline, 62 years), where calcaneal BMD loss was assessed by eight measurements, showed that 1 standard deviation increase in new specific bone markers such as osteocalcin, bone specific alkaline phosphatase and free pyridinoline crosslinks was associated with a twofold increased risk of rapid bone loss defined as the upper tertile of rate of loss . However, one limitation of that study is that bone markers were measured at the end of the follow-up period, although bone turnover is likely to be stable in postmenopausal women. In a recent study of 305 untreated postmenopausal women [121], we found that baseline values of serum OC, serum PINP, serum and urinary CTX and urinary NTX were negatively correlated with the rate of bone loss at the forearm over the next 4 years with correlation coefficients in the order of 0.4 – 0.5. However, when correlation coefficients were corrected for the precision error in individual rate of bone loss and bone marker levels, values increased and reached 0.7 to 0.9 (Table 2). These data strongly suggest that reducing the short- and long-term variability of bone markers is also likely to improve their ability to predict the rate of bone mass change. Nevertheless, it appears at present that it would be difficult to predict absolute rate of bone loss in a individual woman using a measurement of bone markers. However, measurement of bone markers may be useful for identifying women at higher risk of accelerated bone loss in the following years. Interestingly, we found that women whose marker values exceeded mean values for premenopausal women by more than 2 SD lost bone two to six times faster than those whose marker values were normal (Fig. 6).

B. Bone Marker and Fracture Risk With the emergence of effective — but rather expensive — treatments, it is essential to detect those women at higher risk of fracture. Several prospective studies have shown that a SD decrease of BMD measured by dual X-ray absorptiometry (DXA) or heel ultrasound is associated with a two- to four-fold increase in the relative fracture risk at all sites. In this context the question arises to what extent bone markers could add to bone mass measurement in order to improve the assessment of fracture risk. Several retrospective studies have compared bone marker levels in patients with osteoporotic fractures and in controls. These retrospective studies suggest that in patients with fracture bone formation may be decreased, although this finding has been consistently found only for serum osteocalcin [122,123]. For bone resorption, studies using the most specific markers, i.e., urinary PYD, suggest that hip and other fractures cases are associated with increased bone resorption [124]. However, when biochemical markers are

469

CHAPTER 60 Biochemical Markers of Bone Turnover

FIGURE 6 Rate of bone loss in postmenopausal women with high and low bone turnover. Bone loss at the forearm was evaluated by measuring BMD by DXA on four occasions over a 4 - year period in 305 postmenopausal women (mean age, 64 years) as described in the legend of Table 2. For each bone marker each woman was placed in a low and high bone turnover using as a cutoff the mean value 2 SDs of 134 premenopausal women. The figure shows the percentage of bone loss from baseline during the 4 - year study period in low and high bone turnover groups. Women whose baseline marker values indicated high bone remodeling lost bone two to six times faster over the next 4 years than those whose marker values were normal. The serum concentration of the C-terminal cross-linking telopeptide of type I collagen (S-CTX) was measured by ELISA (Serum Crosslaps, Osteometer Biotech) and by automatic analyser (Elecsys, Roche Diagnostics). Reprinted with permission from Garnero et al. [121]. measured within the few hours after hip fracture one cannot exclude that part of these changes of bone turnover may be related to acute changes of body fluid and of hormonal levels related to the trauma. Thus, it seems difficult from retrospective studies to determine if differences in bone turnover levels are related to the underlying rate of bone turnover leading to fracture or to changes of bone turnover occurring after the fracture. Relating baseline bone turnover values to the subsequent risk of osteoporotic fractures is the valid methodology for assessing their clinical utility. Prospective studies relating concentrations of bone formation markers to risk of fracture have yielded conflicting results. Indeed, either a decrease, no difference, or an increase [125 – 128] of bone formation markers has been reported to be associated with increased fracture risk. The difference between studies may reflect the type of fracture, the population studied, or the duration of follow-up. Thus, whether bone formation marker levels are related to fracture risk remains unclear. In contrast, data on the relationship between bone resorption markers and fracture risk are very consistent. Riis et al. [129] reported that women within 3 years of menopause women classified as “fast bone losers” had a twofold higher risk of vertebral and peripheral fractures during a 15-year follow-up than women classified as “normal” or “slow” losers. Interestingly, BMD and rate of bone loss predispose to fractures to the same extent, with an odds ratio of about 2. Women with both a low BMD and a fast

rate of bone loss after the menopause had a higher risk for subsequent fractures than women with only one of the two risk factors. Concordant results have been obtained in three prospective studies (EPIDOS, Rotterdam, and OFELY), indicating that increased levels of bone resorption markers are associated with increased risk of vertebral, hip, and other nonvertebral fractures over follow-up periods ranging from 1.8 to 5 years [125 – 128,130]. This predictive value is consistently in the order of a twofold increase in the risk of fracture for levels above the upper limit of the premenopausal range. Both increased values of serum CTX and urinary CTX [127,130] and of free DPD [125 – 128,130] have been shown to be associated with a higher risk of hip, vertebral, and other nonvertebral fractures. Increased bone resorption is associated with increased risk of fracture only for values that exceed a threshold, suggesting that bone resorption becomes deleterious for bone strength only when it exceeds the normal physiological range. As bone resorption rate predicts fracture independently of BMD, these data suggest that increased bone resorption can lead to increased skeletal fragility by two factors. First, a prolonged increase in bone turnover will lead after several years to a lower BMD, which is a major determinant of reduced bone strength. Second, increased bone resorption above the upper limit of the normal range may induce microarchitectural deterioration of bone tissue such as perforation of trabeculae, a major component of bone strength.

470

GARNERO AND DELMAS

FIGURE 7 Combination of the assessment of bone mineral density (BMD) at the hip and of bone resorption rate to predict hip fracture risk in elderly women followed prospectively for 2 years: The Epidos study. Low BMD was defined according to the WHO guidelines, i.e., by a value lower than 2.5 SD below the young adult mean (T score 2.5). High bone resorption was defined by urinary CTX or Free DPD values higher than the upper limit (mean  2 SD) of the premenopausal range. Reprinted with permission from Garnero et al. [126]. OC contains three residues of -carboxyglutamic acid (GLA), a vitamin K-dependent amino acid. It was postulated that impaired -carboxylation of osteocalcin could be an index of both vitamin D and vitamin K deficiency in elderly populations. In two prospective studies performed in a cohort of elderly institutionalized women followed for 3 years [131,132] and in a population of healthy elderly women (EPIDOS study) [133], concentrations of serum undercarboxylated OC (ucOC) over the premenopausal range was associated with a two- to threefold increase in the risk of hip fracture. Like markers of bone resorption, the prediction was still significant after adjusting for hip BMD. Because increased levels of bone resorption markers and of ucOC have been shown to predict the risk of fracture independently of BMD, the combination of these two diagnostic tests could be useful in improving the identification of women at high risk for fracture. Using the EPIDOS database, it was shown that combining a bone resorption marker (or ucOC) and hip BMD measurement can detect women at very high risk of fracture. Indeed, women with both low hip BMD (according to the WHO definition of osteoporosis) and high bone resorption had a four- to fivefold higher risk than the general population (Fig. 7). This has been confirmed for vertebral and nonvertebral and nonhip fractures [128]. In addition, by using such combination the specificity of hip fracture prediction is increased without a loss of sensitivity [134]. However, because we are dealing with continuous risk factors, there is a large overlap between the distributions of these risk factors (markers and BMD) for the populations of women who will fracture and for the population of women who will not. Therefore, the increase in sensitivity is achieved at the expense of speci-

ficity, and conversely. Moreover, it seems difficult given this overlap and the quite low incidence of fracture, to obtain a very good trade-off between sensitivity and specificity. In that situation it may be more interesting to calculate the posterior probability of fracture, using the likelihood ratio for a positive test. Using this method, Ross et al. [128] showed that the predictive value of bone ALP and calcaneus BMD were similar. Similarly, we found in the OFELY study that for women with levels of S-CTX above the upper limit of the premenopausal range, the probability of fracture was 25% over a 5-year period, i.e., twofold higher than the average probability for all women of this population. Those women with both high CTX and low BMD had 55% probability of fracture compared to 13% for all women (Table 3). Clearly the clinical use of bone markers in the assessment of fracture risk of individual postmenopausal women can be already envisaged, probably to identify high-risk patients with other important risk factors including low BMD, personal and maternal history of fracture, and low body weight, although additional studies will be required to define the optimal strategy. Thresholds of increased bone resorption remains to be established for each type of fracture and the predictive value of bone markers in men has not been yet investigated.

C. Bone Markers for Monitoring Treatment of Osteoporosis with Antiresorptive Therapy As for most chronic diseases, monitoring the efficacy of treatment of osteoporosis is a challenge. The goal of treatment is to reduce the occurrence of fragility fractures, but

471

CHAPTER 60 Biochemical Markers of Bone Turnover

TABLE 3 Combination of Bone Mineral Density (BMD) and Bone Turnover Markers to Predict the Risk of Fractures in Postmenopausal Women Odds ratio (95% CI)

All women

Likelihood ratio

Probability of fracture (for 5 years) (%)

—

—

12.6

Low femoral neck BMD (T score  2.5)

2.8 (1.4 – 5.6)

2.8

39

High S-CTX (T score  2) Low BMD high S-CTX

2.1 (1.2 – 3.8)

1.7

25

3.8 (1.9 – 7.3)

3.7

54

Note. The OFELY study. Four hundred and thirty-five healthy untreated postmenopausal women (mean age, 64 years; from 50 to 89 years) were followed prospectively for an average of 5 years. During this follow-up period, 58 incident fractures (21 vertebral, 37 peripheral fractures) occurred in 55 women. The table shows the odds ratio (adjusted for age, prevalent fractures and physical activity), the likelihood ratio of fracture [sensitivity/(1specificity)], and the 5-year probability of fractures associated with low BMD, high bone resorption assessed by serum C-terminal cross-linking telopeptide of type I collagen (S-CTX) and the combination of both. The T scores of BMD and S-CTX were calculated from the mean and standard deviation of 134 premenopausal women from the same cohort.

their incidence is low and the absence of events during the first year(s) of therapy does not imply necessarily that treatment is effective. Measurement of BMD by DXA is a surrogate marker of treatment efficacy that has been widely used in clinical trials. Its use in the monitoring of treatment efficacy in individual patients, however, has not been validated. Given a short-term precision error of 1 to 1.5% for BMD measurement at the spine and hip, the individual change must be greater than 3 to 5% to be seen as significant. With bisphosphonates such as alendronate, repeating BMD 2 years after initiating therapy will allow one to detect if a patient is responding to therapy, i.e., showing a significant increase in BMD, at least at the lumbar spine,

FIGURE 8

which is the most responsive site. With treatments such as raloxifene or nasal calcitonin that induce much smaller increases in BMD, DXA is not appropriate for monitoring therapy and with any treatment, DXA does not allow identification of all responders within the first year of therapy. Failure to respond may be due to noncompliance —probably the most important single factor — poor intestinal absorption (i.e., bisphosphonates), other factors contributing to bone loss, or other unidentified factors. Monitoring may improve compliance, although this needs to be proven for osteoporosis treatment. Several randomized placebo-controlled studies found that anti-resorptive therapy was associated with a prompt

Effect of different doses of transdermal estradiol on markers of bone turnover. In this study, 201 postmenopausal women ages 40 to 60 years with a time since menopause shorter than 6 years were given either a placebo or a new matrix delivery transdermal 17  estradiol in a dosage of 25, 50, or 75 g twice a week for 28 days. The decrease of bone formation markers (osteocalcin and bone ALP) is delayed compared to that of bone resorption (urinary CTX), reflecting the coupling mechanism. This delayed decreased of bone formation is amplified by the transdermal route of administration of estradiol. Reprinted with permission from Cooper et al. [135].

472

GARNERO AND DELMAS

FIGURE 9

Relationships between early changes in bone resorption (urinary NTX) and changes in spine and hip BMD assessed by DXA at 2 years in early postmenopausal women treated with the bisphosphonate alendronate (5 mg day). This study included 387 early postmenopausal women from 45 to 59 year of age. The graphs show the percentage change from baseline at 24 months of lumbar spine and total hip by tertiles of percent changes from baseline at 6 months in urinary NTX in women treated with alendronate 5 mg/day. Reprinted with permission from Ravn et al. [143].

decrease in bone resorption markers that were seen as early as 2 weeks with a plateau reached within 3 to 6 months. The decrease of bone formation markers was delayed — reflecting the physiological coupling of formation to resorption — and a plateau was usually achieved within 6 to 12 months (Fig. 8) [135]. It has been suggested that baseline bone turnover is a determinant of BMD response, i.e., that patients with high-turnover osteoporosis show a higher increase in BMD than patients with low-turnover osteoporosis. Patients with high turnover have a significantly higher increase in spinal BMD in response to injectable or nasal calcitonin than those with low turnover [136]. A similar trend has been observed in patients treated with HRT and with alendronate, but there is a large overlap in the BMD response between the two groups, so that baseline bone turnover does not appear to be a useful parameter for predicting the individual response to therapy [137]. In contrast, the decrease of bone turnover markers under antiresorptive therapy, usually expressed in percentage of the initial value, is strongly correlated to the increase in BMD. Several studies of HRT [138 – 141], except one [142] have shown in the past 10 years that the short-term (3 to 6 months) decrease of bone turnover markers is significantly correlated with the long-term (1 to 2 years) increase in BMD at the spine, radius, and hip BMD (Fig. 9). A marked decrease of markers is associated with a positive subsequent BMD response while nonresponders show little or no changes of bone markers, suggesting that bone markers, especially new sensitive and specific ones, can be used

to monitor HRT. Similarly, studies with alendronate suggest that the magnitude of the short-term decrease of bone turnover is correlated with the magnitude of the increase of BMD [78,143 – 145], especially when placebo-treated patients are included in the analysis. Most of these studies, however, have not addressed the clinical use of these markers, i.e., how should they be used in the monitoring of the individual patient. For the clinician, the primary concern is the identification of nonresponders, i.e., of patients who will fail to demonstrate a significant increase of BMD after 2 years of treatment. A BMD response has been defined either as a positive BMD change or as a positive change greater than the precision error in a single individual, also called the least significant change. Several methods have been suggested to identify responders/non responders according to the bone marker response to therapy. One approach is to consider the least significant change of a bone marker (based on the short-term or long-term within subject variability in untreated women), regardless of the BMD response [140]. Another approach is to search for the minimum marker change associated with a positive BMD response as previously defined. The optimal threshold of bone marker change can be defined using receiver operating characteristics analysis, or by using logistic regression models [141,143,144]. The percentage change and/or the absolute value of the marker under treatment can be used [143], and cut-off values can be obtained with a prespecified sensitivity or specificity [141,145]. Retrospective

473

CHAPTER 60 Biochemical Markers of Bone Turnover

TABLE 4

Early Changes in Bone Remodeling Markers to Predict the Efficacy of Estrogen Replacement Therapy with 90% Specificity in Individual Patients Marker

Cutoff value for the bone marker decrease after 3 or 6 months (%)

Sensitivitya(%)

Likelihood of a positive responseb(%)

Serum osteocalcin (6 mth)

 21

51

89

Serum bone ALP (6 mth)

 20

49

89

Serum CTX (3 mth)

 33

68

87

Urinary CTX (3 mth)

 45

60

88

Note. In this study, 569 postmenopausal women aged 40 to 60 years with a time since menopause shorter than 6 years were given either a placebo or a transdermal estrogen in a dosage of 25, 50, or 75 g twice a week for 28 days (continuous treatment) or 50, 75, or 100 g twice a week for 25 days per cycle (cyclic therapy). Bone mineral density (BMD) at the spine was measured at baseline and after 2 years using dual-energy X-ray absorptiometry (DXA). Women with a BMD increase versus baseline greater than 2.26% (i.e., twice the short-term coefficient of variation for DXA) were classified as treatment responders and women with a BMD decrease versus baseline of more than 2.26% as nonresponders. The table shows the sensitivity and the likelihood of a positive response obtained using a 3-month to 6-month bone marker decrease cutoff associated with 90% specificity. Reprinted with permission from Delmas et al. [141]. a Proportion of women whose bone marker value decreased at 3 to 6 months into therapy was equal to or greater than the cutoff, among the women with a greater than 2.26% BMD increase two years into therapy. b Proportion of women with a greater than 2.26% BMD increase 2 years into therapy, among the women whose bone marker value decreased 3 to 6 months into therapy was equal to or greater than the cutoff.

analyses of several clinical trials using HRT or alendronate suggest that, for a given marker of resorption or formation, a cut-off value under treatment can be defined that provide adequate predictive value of the subsequent 2-year BMD response in a single patient (Table 4). These cut-off values should be tested in other cohorts using the same therapeutic regimens in order to strengthen their clinical utility. The value of BMD changes to predict the risk of fracture under treatment is debated, especially because some treatments –– such as raloxifene –– can induce a 30 to 50% reduction in vertebral fracture rate despite a small 2 to 3% increase of BMD at all skeletal sites. Thus, BMD changes may not be an adequate surrogate endpoint for analyzing the ability of bone markers to predict fracture risk. Unfortunately, there have been few attempts to correlate bone marker changes with fracture risk. In a retrospective analysis of a small placebo-controlled trial of HRT, Riggs [146] suggested that changes in bone turnover (assessed by histomorphometry) predict change in vertebral fracture risk as well as change in BMD in osteoporotic women. More recently, it was found that the short-term changes of serum osteocalcin under raloxifene were associated with the subsequent risk of vertebral fractures in a large subgroup of osteoporotic women enrolled in the MORE study, while changes in BMD were not predictive [147]. Clearly, such analyses should be performed in ongoing and recently completed large clinical trials performed in postmenopausal women with osteoporosis treated with bisphosphonates, HRT, or SERMs. In summary, changes of new markers of formation/ resorption under HRT, bisphosphonates and raloxifene have

been adequately documented in many clinical trials. The fact that they decrease rapidly and reach a drug- and dosedependant plateau within a few months suggests that they could be used to predict the longer term response to therapy. Statistical models have been developed recently, indicating that the percentage decrease of some bone markers after 3 to 6 months of HRT or alendronate can be used to predict the 2-year response in BMD with adequate sensitivity and specificity. These studies provide cut-off values to predict responders/nonresponders to therapy that should be tested in other cohorts. Importantly, the same approach should be applied to large trials with incident fracture as an endpoint.

References 1. P. Magnusson, O. Lofman, and L. Larsson, Determination of alkaline phosphatase isoenzymes in serum by high performance liquid chromatography with post-column reaction detection. J. Chromatogr. 576, 79 – 86 (1992). 2. P. Magnusson, L. Larsson, M. Magnussson, M. W. J. Davie, and C. A. Sharp, Isoforms of bone alkaline phosphatase: Characterization and origin in human trabecular and cortical bone. J. Bone Miner. Res. 14, 1926 – 1933 (1999). 3. R. G. Crilly, M. M. Jones, A. Horsman, et al., Rise in plasma alkaline phosphatase at the menopause. Clin. Sci. 53, 341 – 342 (1980). 4. J. P. Brown, P. D. Delmas, M. Arlot, et al., Active bone turnover of the cortico-endosteal envelope in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 64, 954 – 959 (1987). 5. J. Podenphant, J. S. Johansen, K. Thomsen, et al., Bone turnover in spinal osteoporosis. J. Bone Miner. Res. 2, 497 – 503 (1987). 6. D. W. Moss, Alkaline phosphatase isoenzymes. Clin. Chem. 28, 2007 – 2016 (1982).

474 7. J. R. Farley, C. J. Chestnut, and D. J. Baylink, Improved method for quantitative determination in serum alkaline phosphatase of skeletal origin. Clin. Chem. 27, 2002 – 2007 (1981). 8. R. J. Duda, J. F. O’Brien, J. A. Katzmann, et al., Concurrent assays of circulating bone gla-protein and bone alkaline phosphatase: Effects of sex, age, and metabolic bone disease. J. Clin. Endocrinol. Metab. 66, 951 – 957 (1988). 9. P. Garnero and P. D. Delmas, Assessment of the serum levels of bone alkaline phosphatase with a new immunoradiometric assay in patients with metabolic bone disease. J. Clin. Endocrinol. Metab. 77(4), 1046 – 1053 (1993). 10. B. Gomez, S. Ardakanis, J. Ju et al., Monoclonal antibody assay for measuring bone-specific alkaline phosphatase in serum. Clin. Chem. 41, 1560 – 1566 (1995). 11. P. A. Price, Vitamin K-dependent bone proteins. In “Calcium Regulation and Bone Metabolism: Basic and Clinical Aspects” (D. V. Cohn, T. J. Martin, and P. J. Meunier, eds.), Vol. 9, pp. 419 – 426. Elsevier Science, Amsterdam, 1987. 12. P. Ducy, C. Desbois, B. Boyce, et al. Increased bone formation in osteocalcin-deficient mice. Nature 382 – 452 (1996). 13. P. A. Price, J. G. Parthemore, and L. J. Deftos, New biochemical marker for bone metabolism. J. Clin. Invest. 66, 878 – 883 (1980). 14. P. A. Price, M. K. Williamson, and J. W. Lothringer, Origin of vitamin K-dependent bone protein found in plasma and its clearance by kidney and bone. J. Biol. Chem. 256, 12760 – 12766 (1981). 15. J. B. Lian and C. M. Gundberg, Osteocalcin. Biochemical considerations and clinical applications. Clin. Orthop. Rel. Res. 226, 267 –291 (1988). 16. P. D. Delmas, D. Stenner, H. W. Wahner, et al., Serum bone gla-protein increases with aging in normal women: Implications for the mechanism of age-related bone loss. J. Clin. Invest. 71, 1316 – 1321 (1983). 17. M. A. Thiede, S. L. Smock, D. N. Petersen, W. A. Grasser, D. D. Thompson, and S. K. Nishimoto, Presence of messenger of ribonucleic acid encoding osteocalcin, a marker of bone turnover, in bone marrow megakaryocytes and peripheral blood platelets. Endocrinology 135, 929 – 937 (1994). 18. P. D. Delmas, D. M. Wilson, K. G. Mann, et al., Effect of renal function on plasma levels of bone gla-protein. J. Clin. Endocrinol. Metab. 57, 1028 – 1030 (1983). 19. P. D. Delmas, Biochemical markers of bone turnover for the clinical assessment of metabolic disease. Endocrinol. Metab. Clin. North Am. 19(1), 1 – 18 (1990). 20. J. P. Brown, P. D. Delmas, L. Malaval, et al., Serum bone Glaprotein: A specific marker for bone formation in postmenopausal osteoporosis. Lancet 1, 1091 – 1093 (1984). 21. P. Charles, J. W. Poser, L. Mosekilde, and F. T. Jensen, Estimation of bone turnover evaluated by 47 calcium kinetics. Efficiency of serum bone gamma-carboxyglutamic acid containing protein, serum alkaline phosphatase and urinary hydroxyproline excretion. J. Clin. Invest. 76, 2254 – 2258 (1985). 22. P. D. Delmas, L. Malaval, M. E. Arlot, and P. J. Meunier, Serum bone gla-protein compared to bone histomorphometry in endocrine diseases. Bone 6, 329 – 341 (1985). 23. P. D. Delmas, B. Demiaux, L. Malaval, et al., Serum bone gla-protein (osteocalcin) in primary hyperparathyroidism and in malignant hypercalcemia. Comparison with bone histomorphometry. J. Clin. Invest. 77, 985 – 991 (1986). 24. R. Bataille, P. D. Delmas, and J. Sany, Serum bone gla-protein in multiple myeloma. Cancer 59, 329 – 334 (1987). 25. P. Garnero, M. Grimaux, P. Seguin, and P. D. Delmas, Characterization of immunoreactive forms of human osteocalcin generated in vivo and in vitro. J. Bone Miner. Res. 9(2), 255 – 264 (1994). 26. P. D. Delmas, C. Christiansen, K. G. Mann, and P. A. Price, Bone gla-protein (osteocalcin) assay standardization report. J. Bone Miner. Res. 1, 5 – 11 (1990).

GARNERO AND DELMAS 27. A. M. Parfitt, L. S. Simon, A. R. Villanueva, et al., Procollagen Type I carboxy-terminal extension peptide in serum as a marker of collagen biosynthesis in bone. Correlation with iliac bone formation rates and comparison with total alkaline phosphatase. J. Bone Miner. Res. 2, 427 – 436 (1987). 28. C. Hassager, G. Fabbri-Mabelli, and C. Christiansen, The effect of the menopause and hormone replacement therapy on serum carboxyterminal propeptide of type I collagen. Osteoporosis Int. 3, 50 – 52 (1993). 29. P. R. Ebeling, J. M. Peterson, and B. L. Riggs, Utility of Type I Procollagen propeptide assays for assessing abnormalities in metabolic bone diseases. J. Bone Miner. Res. 7, 1243 – 1250 (1992). 30. C. Hasling, E. F. Eriksen, J. Melkko, L. Ristelli, P. Charles, L. Mosekilde, and J. Ristelli, Effects of a combined estrogengestagen regimen on serum levels of the carboxy-terminal propeptide of human type I procollagen in osteoporosis. J. Bone Miner. Res. 6, 1295 – 1300 (1991). 31. J. Melkko, S. Niemi, L. Risteli, and J. Risteli, Radioimmunoassay of the carboxyterminal propeptide of human type I procollagen. Clin. Chem. 40, 1328 – 1332 (1990). 32. J. Melkko, S. Kauppila, S. Niemi, L. Risteli, K. Haukipuro, A. Jukkola, and J. Risteli, Immunoassay for intact amino-terminal propetide of human type I procollagen. Clin. Chem. 42, 947 – 954 (1996). 33. J. Brandt, T. Kroch, C. H. Jensen, J. K. Frederiksen, and B. Teisner, Thermal instability of the trimeric structure of the N-terminal propeptide of human procollagen type I in relation to assay technology. Clin. Chem. 45, 47 – 53 (1999). 34. J. Risteli and L. Risteli, Products of bone collagen metabolism. In “Dynamics of Bone and Cartilage Metabolism” (M. J. Seibel, S. P. robins, and J. P. Bilezikian, eds.), pp. 275 – 287. Academic Press, San Diego, 1999. 35. P. R. Ebeling, J. M. Peterson, and B. L. Riggs, Utility of type I procollagen propeptide assays for assessing abnormalities in metabolic bone diseases. J. Bone Miner. Res. 7, 1243 – 1250 (1992). 36. S. G. Linkhart, T. A. Linkhart, A. K. Taylor, J. E. Wergedal, P. Bettica, and D. J. Baylink, Synthetic peptide-based immunoassay for amino-terminal propeptide of type I procollagen: Application for evaluation of bone formation. Clin. Chem. 39, 2254 – 2258 (1993). 37. P. Garnero, P. Vergnaud, and P. D. Delmas, Amino terminal propeptide of type I collagen (PINP) is a more sensitive marker of bone turnover than C-Terminal propeptide in osteoporosis. J. Bone Miner. Res. 12(s1), s497 (1997). 38. C. D. Cabera, M. Henriquez, M. L. Traba, E. A. Villafane, and C. de la Piedra, Biochemical markers of bone formation in the study of postmenopausal osteoporosis. Osteoporosis Int. 8, 147 – 151 (1998). 39. R. Hannon, A. Blumsohn, K. Naylor, and R. Eastell, Response of biochemical markers of bone turnover to hormone replacement therapy: impact of biological variability. J. Bone Miner. Res. 13, 1124 – 1133 (1998). 40. P. Garnero, Y. Tsouderos, I. Marton, C. Pelissier, C. Varin, and P. D. Delmas, Effects of intranasal 17  estradiol on bone turnover and serum insulin like growth factor I in postmenopausal women. J. Clin. Endocrinol Metab. 84, 2390 – 2397 (1999). 41. B. Smedsrod, J. Melkko, L. Ristelli, and J. Ristelli, Circulating Cterminal propeptide of type I procollagen is cleared mainly via the mannose receptor in liver endothelial cells. Biochem. J. 271, 345 – 350 (1990). 42. J. Melko, T. Hellevick, L. Ristelli, J. Ristelli, and B. Smedsrod, Clerance of NH2 terminal propetide of types I and III procollagen is a physiological function of the scavenger receptor in liver endothelial cells. J. Exp. Med. 179, 405 – 412 (1994). 43. J. Toivonen, R. Tahtela, K. laitinen, J. Risteli and M. J. Valimaki, Markers of bone turnover in patients with differentiated thyroid

CHAPTER 60 Biochemical Markers of Bone Turnover

44.

45. 46.

47.

48.

49.

50.

51.

52. 53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

cancer on and off thyroxine suppressive therapy. Eur. J. Endocrinol. 138, 667 – 673 (1994). O. J. Prockop and K. I. Kivirikko, Hydroxyproline and the metabolism of collagen. In “Treatise on Collagen” (B. S. Gould, ed.), pp. 215 – 246. Academic Press, New York, 1968. O. J. Prockop, K. I. Kivirikko, K. Tuderman, et al., The biosynthesis of collagen ant its disorders. N. Engl. J. Med. 301, 13 – 23 (1979). S. M. Krane, F. G. Kantrowitz, M. Byrne, S. R. Pinnel, and F. R. Singer, Urinary excretion of hydroxylysine and its glycosides as an index of collagen degradation. J. Clin. Invest. 59, 819 – 827 (1977). K. I. Kivirikko, Excretion of urinary hydroxyproline peptide in the assessment of bone collagen deposition and resorption. In “Clinical Disorders of Bone and Mineral Metabolism” (B. Frame and J. T. Potts, eds.), pp. 105 – 107. Excerpta Medica, Amsterdam, 1983. M. Lowry, D. E. Hall, and J. J. Brosnan, Hydroxyproline metabolism by the rat kidney: Distribution of renal enzymes of hydroxyproline catabolism and renal conversion of hydroxyproline to glycine and serine. Metabolism 39, 955 (1985). L. Moro, R. S. P. Mucelli, C. Gazzarrini, et al., Urinary b-1-galactosyl-O-hydroxylysine (GH) as a marker of collagen turnover of bone. Calcif. Tissue Int. 42, 87 – 90 (1988). A. W. Al-Dehaimi, A. Blumsohn, and R. Eastell, Serum galactosyl hydroxylysine as a biochemical marker of bone resorption. Clin. Chem. 45, 676 – 681 (1999). V. Lo Cascio, F. Bertoldo, G. Gambaro, E. Gasperi, F. Furlan, F. Colapietro, C. Lo Cascio, and M. Campagnola, Urinary excretion of galactosyl-hydroxylysine in postmenopausal osteoporotic women: a potential marker of bone fragility. J. Bone Miner. Res. 14, 1420 –1424 (1999). C. Y. Li, R. A. Chuda, W. K. W. Lam, et al., Acid phosphatase in human plasma. J. Lab. Clin. Med. 82, 446 – 460 (1973). C. Minkin, Bone acid phosphatase: Tartrate-resistant acid phosphatase as a marker of osteoclast function. Calcif. Tissue Int. 34, 285 – 290 (1982). A. R. Hayman, A. J. Bune, Bradley, J. Rashbass, and T. M. Cox, Osteoclastic tartrate-resistant acid phosphatase (Acp 5). Its localization to dendritic cells and diverse murine tissues. Histochem. Cytochem. 48, 219 – 228 (2000). M. E. Kraenzlin, K. H. Lau, L. Liang, et al., Development of an immunoassay for human serum osteoclastic tartrate-resistant acid phosphatase. J. Clin. Endocrinol. Metab. 71, 442 – 451 (1990). C. Cheung, N. Panesar, C. Haines, J. Masarei, and R. Swaminathan, Immunoassay of a tartrate-resistant acid phosphatase in serum. Clin. Chem. 41, 679 – 686 (1995). J. M. Hallen, T. A. Hentunen, M. Karp, S -M. Kakonen, K. Pettersson, and H. K. Vaananen, Characterization of serum tartrate-resistant acid phosphatase and development of a direct two-site immunoassay. J. Bone Miner. Res. 13, 683 – 687 (1988). J. M. Halleen, M. Karp, S. Viloma, P. Laaksonen, J. Hellman, S. M. Kakonen, J. J. Stepan, S. Holmes, H. K. Vaananen, and K. Pettersson, Two-site immunoassays for osteoclastic tartrate-resistant acid phosphatase based on characterization of six monoclonal antibodies. Bone Miner. Res. 14, 464 – 469 (1999). J. J. Stepan, E. Silinkova-Malkova, T. Havrenek, et al., Relationship of plasma tartrate-resistant acid phosphatase to the bone isoenzyme of serum alkaline phosphatase in hyperparathyroidism. Clin. Chim. Acta 133, 189 – 200 (1983). J. J. Stepan, J. Pospichal, J. Presl, et al., Bone loss and biochemical indices of bone remodeling in surgically induced postmenopausal women. Bone 8, 279 – 284 (1987). C. Piedra, R. Torres, A. Rapado, et al., Serum tartrate resistant acid phosphatase and bone mineral content in postmenopausal osteoporosis. Calcif. Tissue Int. 45, 58 – 60 (1989). C. S. Brehme, S. Roman, J. Shaffer, and R. Wolfert, Tartrate resistant acid phosphatse forms complexes with a 2 macroglobulin in serum. J. Bone Miner. Res. 14, 311 – 318 (1999).

475 63. D. R. Eyre, T. J. Koob, and K. P. Van Ness, Quantitation of hydroxypyridinium crosslinks in collagen by high-performance liquid chromatography. Anal. Biochem. 137, 380 – 388 (1984). 64. A. Colwell and R. Eastell, Renal clearance of free and conjugated pyridinium crosslinks of collagen. J. Bone Miner. Res. 11, 1976 – 1980 (1996). 65. P. Garnero, E. Gineyts, P. Arbault, C. Christiansen, and P. D. Delmas, Different effects of bisphosphonate and estrogen therapy on free and peptide-bound crosslinks excretion. J. Bone Miner. Res. 10, 641 – 649 (1995). 66. D. R. Eyre, T. J. Koob, and K. P. Van Ness, Quantitation of hydroxypyridinium crosslinks in collagen by high-performance liquid chromatography. Anal. Biochem. 137, 380 – 388 (1984). 67. D. Black, A. Duncan, and S. P. Robins, Quantitative analysis of the pyridinium crosslinks of collagen in urine using ion-paired reversedphase high-performance liquid chromatography. Anal. Biochem. 169, 197 – 203 (1988). 68. R. Eastell, A. Colwell, L. Hampton, and J. Reeve, Biochemical markers of bone resorption compared with estimates of bone resorption from radiotracer kinetic studies in osteoporosis. J. Bone Miner. Res. 12, 59 – 65 (1997). 69. P. D. Delmas, A. Schlemmer, E. Gineyts, B. Riis, and C. Christiansen, Urinary excretion of pyridinoline crosslinks correlates with bone turnover measured on iliac crest biopsy in patients with vertebral osteoporosis. J. Bone Miner. Res. 6, 639 – 644 (1991). 70. S. Seyedin, R. Zuk, V. Kung, Y. Daniloff, and K. Shepard, An immunoassay to urinary pyridinoline: the new marker of bone resorption. J. Bone Miner. Res. 8, 635 – 642 (1993). 71. S. P. Robins, H. Woitge, R. Hesley, J. Ju, S. Seyedin, and M. J. Seibel, Direct, enzyme-linked immunoassay for urinary deoxypyridinolin as a specific marker for measuring bone resorption. J. Bone Miner. Res. 9, 1643 – 1649 (1994). 72. J. Ristelli, I. Elomaa, S. Niemi, A. Novamo, and L. Ristelli, Radioimmunoassay for the pyridinoline cross-linked carboxy-terminal telopeptide of type I collagen: A new serum marker of bone collagen degradation. 39, 635 – 640 (1993). 73. C. Rosenquist, C. Fledelius, S. Christgau, B. J. Pedersen, M. Bonde, P. Qvist, and C. Christiansen, Serum Crosslaps one step ELISA. First application of monoclonal antibodies for measurement in serum of bone-related degradation products from C-terminal telopeptides of type I collagen. Clin. Chem. 44, 2281 – 2289 (1998). 74. B. Bonde, P. Qvist, C. Fledelius, B. J. Riss, and C. Christiansen, Immunoassay for quantifying type I collagen degradation products in urine evaluated. Clin. Chem. 40, 2022 – 2025 (1994). 75. J. D. Clemens, M. V. Herrick, F. R. Singer, and D. R. Eyre, Evidence that serum NTX (collagen I N-telopeptides) can act as an immunochemical marker of bone resorption. Clin. Chem. 43, 2058 – 2063 (1997). 76. D. A. Hanson, M. A. E. Weiss, A. M. Bollen, S. L. Maslan, F. R. Singer, and D. R. Eyre, A specific immunoassay for monitoring human bone resorption: Quantitation of type I collagen crosslinked N-telopeptides in urine. J. Bone Miner. Res. 7, 1251 – 1258 (1992). 77. C. Hassager, L. T. Jensen, J. Podenphant, K. Thomsen, and C. Christiansen, The carboxy-terminal pyridinoline cross-linked telopeptide of type I collagen in serum as a marker of bone resorption: The effect of Nandrolone decanoate and hormone replacement therapy. Calcif. Tissue Int. 54, 30 – 33 (1994). 78. P. Garnero, W. J. Shih, E. Gineyts, D. Karpf, and P. D. Delmas, Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate. J. Clin. Endocrinol. Metab. 79, 1693 – 1700 (1994). 79. H. Maeda, M. Koizumi, K. Yoshimura, T. Yamauchi, T. Kawai, and E. Ogata, Correlation between bone metabolic markers and bone scan in prostatic cancer. J. Urol. 157, 539 – 543 (1997).

476 80. K-I Yoshida, S. Sumi, K. Arai, F. Koga, H. Umeda, Y. Hosoya, M. Honda, M. Yano, H. Moriguchi, and S. Kitahara, Serum concentration of type I collagen metabolites as a quantitative marker of bone metastases in patients with prostate cancer. Cancer 80, 1760 – 1767 (1997). 81. M - L Sassi, H. Eriksen, L. Risteli, S. Niemi, J. Mansell, M. Gowen, and J. Ristelli, Immunochemical characterization of assay for carboxyterminal telopeptide of human type I collagen: Loss of antigenicity by treatment with cathepsin K. Bone 26, 367 – 373 (2000). 82. P. Saftig, E. Hunziker, O. Wehmeyer, S. Jones, A. Boyde, W. Rommerskirch, J. D. Moritz, and K. von Figura, Impaired osteoclastic bone resorption leads to osteopetrosis in cathepsin K deficient mice. Proc. Natl. Acad. Sci. USA 95, 13453 – 13458. 83. P. Garnero, O. Borel, I. Byrjalsen, M. Ferraras, F. Drake, N. T. Foged, P. D. Delmas, and J - M Delaissé, The collagenolytic activity of cathepsin K is unique amongst mammalian proteinases. J. Biol. Chem. 273, 32347 – 32352 (1998). 84. Y. Nishi, L. Atley, D. R. Eyre, J. G. Edelson, A. Superti-Furga, T. Yasuda, R. J. Desnick, and B. D. Gelb, Determination of bone markers in Pycnodysostosis: Effects of cathepsin K deficiency on bone matrix degradation. J. Bone Miner. Res. 1902 – 1908 (1999). 85. P. Garnero, J. Buchs, J. Zekri, R. E. Coleman, and P. D. Delmas, Bone turnover markers for the management of patients with bone metastases from prostate cancer. Br. J. Cancer 82, 858 – 864 (2000). 86. R. Eastell, M. S. Calvo, M. F. Burritt, K. P. Offord, R. G. Russell, and B. L. Riggs, Abnormalities in circadian patterns of bone resorption and renal calcium conservation in type I osteoporosis. J. Clin. Endocrinol. Metab. 74, 487 – 494 (1992). 87. A. Blumsohn, K. Herrington, and R. A. Hannon, P. Shao, D. R. Eyre, and R. Eastell, The effect of calcium supplementation on the circadian rhythm of bone resorption. J. Clin. Endocrinol. Metab. 79, 730 – 735 (1994). 88. S. L. Greenspan, R. Dresner-Pollak, R. A. Parker, D. London, and L. Ferguson, Diurnal variation of bone mineral turnover in elderly men and women. Calcif. Tissue Int. 60, 419 – 423 (1997). 89. C. Hassager, J. Risteli, L. Risteli, S. B. Jensen, and C. Christiansen, Diurnal variation in serum markers of type I collagen synthesis and degradation in healthy premenopausal women. J. Bone Miner. Res. 7, 1307 – 1311 (1992). 90. M. Wichers, E. Schmidt, F. Bidlingmaier, and D. Klingmuller, Diurnal rhythm of CrossLaps in human serum. Clin. Chem. 45, 1858 – 1860 (1999). 91. A. Schlemmer, C. Hassager, S. B. Jensen, and C. Christiansen, Marked diurnal variation in urinary excretion of pyridinium crosslinks in premenopausal women. J. Clin. Endocrinol. Metab. 74, 476 – 480 (1992). 92. H. Aosahima, K. Kushida, M. Takahashi, T. Ohishi, H. Hoshino, M. Suzuki et al., Circadian variation of urinary type I collagen crosslinked C-telopeptide and free and peptide-bound forms of pyridinium crosslinks. Bone 22, 73 – 78 (1998). 93. A. Schlemmer, C. Hassager, B. J. Pedersen, and C. Christiansen, Posture, age, menopause, and osteopenia do not influence the circadian variation in the urinary excretion of pyridinium crosslinks. J. Bone Miner. Res. 9, 1883 – 1888 (1994). 94. S. Sairanen, R. Tahtela, K. Laitinen, S. L. Karonen, and M. J. Valimaki, Nocturnal rise in markers of bone resorption is not abolished by bedtime calcium or calcitonin. Calcif. Tissue Int. 55, 349 – 352 (1994). 95. S. Sairanen, R. Tahtela, K. Laitinen, E. Loyttyniemi, and M. J. Valimaki, Effects of short-term treatment with clodronate on parameters of bone metabolism and their diurnal variation. Calcif. Tissue Int. 60, 160 – 163 (1997). 96. B. J. Gertz, J. D. Clemens, S. D. Holland, W. Yuan, and S. Greenspan, Application of a new serum assay for type I collagen cross-linked N- telopeptides: Assessment of diurnal changes in bone turnover with and without alendronate treatment. Calcif. Tissue Int. 63, 102 – 106 (1998).

GARNERO AND DELMAS 97. A. Schlemmer and C. Hassager, Acute fasting diminishes the circadian rhythm of biochemical markers of bone resorption. Eur. J. Endocrinol. 140, 332 – 337 (1999). 98. S. Christgau, O. B. Jensen, N. H. Bjarnason, C. Rosenquist, P. Alexandersen, D. B. Henriksen, and C. Christiansen, Serum CrossLaps provides a rapid assessmsnt of therapy respornse compared to BMD measurements. Bone (In press). 99. C. Fledelius, A. H. Johnsen, P. A. C. Cloos, M. Bonde, and P. Qvist, Characterization of urinary degradation products derived from Type I collagen. J. Biol. Chem. 272, 9755 – 9763 (1997). 100. P. A. C Cloos and C. Fledelius, Collagen fragments in urine derived from bone resorption are highly racemized and isomerized: a biological clock of protein aging with clinical potential. Biochem. J. 345, 473 – 480 (2000). 101. P. Garnero, C. Fledelius, E. Gineyts, C - M Serre, E. Vignot, and P. D. Delmas, Decreased  isomerisation of C-telopeptides of type I collagen in Paget’s disease of bone. J. Bone Miner. Res. 12, 1407 – 1415 (1997). 102. P. Garnero, E. Gineyts, A. V. Shaffer, S. Seaman, and P. D. Delmas, Measurement of urinary excretion of nonisomerized and -isomerized forms of type I collagen breakdown products to monitor the effects of the bisphosphonate zoledronate in Paget’s disease. Arthritis Rheum. 41, 354 – 360 (1998). 103. N. W. Clarke, J. McClure, and N. J. R. George, Morphometric evidence for bone resorption and replacement in prostate cancer. Br. J. Urol. 68, 74 – 80 (1991). 104. L. W. Fisher, S. W. Whitson, L. W. Avioli, and J. D. Termine, Matrix sialoprotein of developing bone. J. Biol. Chem. 258, 12723 – 12727 (1983). 105. J. Chen, H. S. Shapiro, J. L. Wrana, S. Reimers, and J. Heersche, and J. Sodek, Localization of bone sialoprotein (BSP) expression to sites of mineral tissue formation in fetal rat tissue by in situ hybridization. Matrix 11, 133 – 143 (1991). 106. H. S. Shapiro, J. Chen, J. L. Wrana, Q. Zhang, M. Blum, and J. Sodek, Characterization of porcine bone sialoprotein: primary structure and cellular expression. Matrix 13, 431 – 444 (1993). 107. P. Bianco, L. W. Fisher, M. F. Young, J. D. Termine, and P. G. Robey, Expression of bone sialoprotein (BSP) in developing human tissues. Calcif. Tissue Int. 49, 421 – 426 (1991). 108. E. Debri, F. P. Reinholt, D. Heinegard, S. Mengarelliholm, Norgard, and O. Svensson, Bone sialoprotein and osteopontin distribution at the osteocartilagineous interface. Clin. Orthop. 330, 251 – 260 (1996). 109. C. Chenu and P. D. Delmas, Platelets contribute to circulating levels of bone sialoprotein in humans. J. Bone Miner. Res. 7, 47 – 54 (1992). 110. R. Fujisawa, Y. Nodasaka, and Y. Kuboki, Further characterization of interaction between bone sialoprotein (BSP) and collagen. Calcif. Tissue Int. 56, 140 – 144 (1995). 111. G. K. Hunter and H. A. Goldberg, Modulation of crystal formation by bone phosphoproteins: Role of glutamic acid-rich sequences in the nucleation of hydroxyapatite by bone sialoprotein. Biochem. J. 302, 175 – 179 (1994). 112. C Raynal, P. D. Delmas, and C. Chenu, Bone sialoprotein stimulates in vitro bone resorption. Endocrinology 137, 2347 – 2354 (1996). 113. T. Saxne, L. Zunino, and D. Heinegard, Increased release of bone sialoprotein into synovial fluid reflects tissue destruction in rheumatoid arthritis. Arthritis Rhem. 1, 82 – 90 (1995). 114. M. Karmatschek, H. W. Woitge, F. P. Armbruster, R. Ziegler, and M. J. Seibel, Improved purification of human bone sialoprotein and development of a homologous radioimmunoassay. Clin. Chem. 43, 2076 – 2082 (1997). 115. M. J. Seibel, H. W. Woitge, M. Pecherstorfer et al., Serum immunoreactive bone sialoprotein as a new marker of bone turnover in metabolic and malignant bone disease. J. Clin. Endocrinol. Metab. 81, 3289 – 3294 (1996).

477

CHAPTER 60 Biochemical Markers of Bone Turnover 116. H. W. Woitge, M. Pecherstorfer, Y. Li, A - V. Keck, E. Horn, R. Ziegler, and M. Seibel, Novel Markers of bone resorption and comparison with established urinary indices. J. Bone Miner. Res. 14, 792 – 801 (1999). 117. P. Garnero, E. Sornay-Rendu, M. C. Chapuy, and P. D. Delmas, Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J. Bone Miner. Res. 11, 337 – 349 (1996). 118. D. Uebelhart, A. Schlemmer, J. Johansen, E. Gineyts, C. Christiansen, and P. D. Delmas, Effect of menopause and hormone replacement therapy on the urinary excretion of pyridinium crosslinks. J. Clin. Endocrinol. Metab. 72, 367 – 373 (1991). 119. M. A. Hansen, O. Kirsten, B. J. Riss, and C. Christiansen, Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 Years study. Br. Med. J. 303, 961 – 964 (1991). 120. P. D. Ross and W. Knowlton, Rapid bone loss is associated with increased levels of biochemical markers. J. Bone Miner. Res. 13, 297 – 302 (1998). 121. P. Garnero, E. Sornay-rendu, F. Duboeuf, and P. D. Delmas, Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: The Ofely Study. J. Bone Miner. Res. 14, 1614 – 1621 (1999). 122. K. Akesson, P. Vergnaud, E. Gineyts, P. D. Delmas, and K. Obrant, Impairment of bone turnover in elderly women with hip fracture. Calcif. Tissue Int. 53, 162 – 169 (1993). 123. K. Akesson, P. Vergnaud, P. D. Delmas, and K. J. Obrant, Serum osteocalcin during fracture healing in elderly women with hip fracture. Bone 16, 427 – 430 (1995). 124. L. J. Melton III, S. Khosla, E. J. Atkinson, W. M. O’Fallon, and B. L. Riggs, Relationship of bone turnover to bone density and fractures. J. Bone Miner. Res. 12, 1083 – 1091 (1997). 125. P. L. A. van Daele, M. J. Seibel, H. Burger, A. Hofman, D. E. Grobbee, J. P. T. M. van Leeuwen, J. C. Birkenhager, and H. A. P. Pols, Case-control analysis of bone resorption markers, disability, and hip fracture risk: The Rotterdam study. Br. Med. J. 312, 482 – 483 (1996). 126. P. Garnero, E. Hausher, M. C. Chapuy, et al., Markers of bone resorption predict hip fracture in elderly women: The Epidos prospective study. J. Bone Miner. Res. 11, 1531 – 1538 (1996). 127. P. Garnero, E. Sornay-Rendu, B. Claustrat, and P. D. Delmas, Biochemical markers of bone turnover, endogenous hormones and the risk of fractures in postmenopausal women: The OFELY study. J. Bone Miner. Res. 15, 1526 – 1536 (2000). 128. P. D. Ross, B. C. Kress, R. E. Parson, R. D. Wasnich, K. A. Armour, and I. A. Mizrahi, Serum Bone alkaline phosphatase and calcaneus bone density predict fractures: A prospective study. Osteoporosis Int. 11, 76 – 82 (2000). 129. S. B. J. Riis, A. M. Hansen, K. Jensen, K. Overgaard, and C. Christiansen, Low bone mass and fast rate of bone loss at menopause-equal risk factors for future fracture. A 15 year follow-up study. Bone 19, 9 – 12 (1996). 130. R. D. Chapurlat, P. Garnero, G. Brèart, P. J. Meunier, and P. D. Delmas, Afternoon sampled serum type I collagen breakdown product (serum CTX) predicts hip fracture in elderly women: Results of the Epidos study. Bone 27, 283 – 286 (2000). 131. P. Szulc, M. C. Chapuy, P. J. Meunier, and P. D. Delmas, Serum undercarboxylated osteocalcin is a marker of the risk of hip fracture in elderly women. J. Clin. Invest. 91, 1769 – 1774 (1993). 132. P. Szulc, M. C. Chapuy, P. J. Meunier, and P. D. Delmas, Serum undercarboxylated osteocalcin is a marker of the risk of hip fracture: a three year follow-up study. Bone 5, 487 – 488 (1996). 133. P. Vergnaud, P. Garnero, P. J. Meunier et al., Undercarboxylated osteocalcin measured with a specific immunoassay predicts hip frac-

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

ture in elderly women: The EPIDOS study. J. Clin. Endocrinol. Metab. 82, 719 – 724 (1997). P. Garnero, P. Dargent-Molina, D. Hans, A. M. Schott, G. Bréart, P. J. Meunier, and P. D. Delmas, Do markers of bone resorption add to bone mineral density and ultrasonographic heel measurement for the prediction of hip fracture in elderly women? The EPIDOS prospective study. Osteoporos. Int. 8, 563 – 569 (1998). C. Cooper, J. A. Stakkestad, S. Radowicki, P. Hardy, C. Pilate, M. P. Dain, and P. D. Delmas, Matrix delivery transdermal 17  estradiol for the prevention of bone loss in postmenopausal women. Osteoporosis Int. 9, 358 – 366 (1999). R. Civitelli, S. Gonelli, F. Sacchei, et al., Bone turnover in postmenopausal osteoporosis. Effect of calcitonin treatment. J. Clin. Invest. 82, 1268 – 1274 (1988). S. L. Greenspan, R. H. Parker, L. Fergusson, H. N. Rosen, L. Maitland-Ramsey, and D. B. Karpf, Early changes in biochemical markers of bone turnover predict the long term response to alendronate therapy in representative elderly women: a randomized clinical trial. J. Bone Miner. Res. 13, 1431 – 1438 (1998). J. S. Johansen, B. J. Riis, P. D. Delmas, et al., Plasma BGP: An indicator of spontaneous bone loss and effect of estrogen treatment in postmenopausal women. Eur. J. Clin. Invest. 18, 191 – 195 (1988). C. J. Rosen, C. H. Chesnut III, and N. J. S. Mallinak, The predictive value of biochemical markers of bone turnover for bone mineral density in early postmenopausal women treated with hormone replacement or calcium supplementation. J. Clin. Endocrinol. Metab. 82, 1904 – 1910 (1997). R. Hannon, A. Blumsohn, K. Naylor, and K. Eastell, Response of biochemical markers of bone turnover to hormone replacement therapy: Impact of biological variability. J. Bone Miner. Res. 13, 1124 – 1133 (1998). P. D. Delmas, P. Hardy, P. Garnero, and M. P. Dain, Monitoring individual response to hormone replacement therapy with bone markers. Bone 26, 553 – 560 (2000). R. Marcus, L. Holloway, B. Wells, G. Greendale, M. K. James, C. Wasilauskas, and J. Kelaghan, The relationship of biochemical markers of bone turnover to bone density changes in postmenopausal women: Results from the postmenopausal estrogen/progestin interventions (PEPI) trial. J. Bone Miner. Res. 14, 1583 – 1595 (1999). P. Ravn, B. Clemmesen, and C. Christiansen, Biochemical markers can predict the response in bone mass during alendronate treatment in early postmenopausal women. Bone 24, 237 – 244 (1999). P. Ravn, D. Hosking, G. C. Thompson, G. Cizza, R. D. Wasnich, M. Mcclung, A. J. Yates, N. H. Bjarnasson, and C. Christiansen, Monitoring of alendronate treatment and prediction of effect on bone mass by biochemical markers in early postmenopausal intervention cohort of study. J. Clin. Endocrinol. Metab. 84, 2363 – 2368 (1999). P. Garnero, C. Darte, and P. D. Delmas, A model to monitor the Efficacy of alendronate treatment in women with osteoporosis using a biochemical markers of bone turnover. Bone 24, 603 – 609 (1999). B. L. Riggs, L. J. Melton III, and W. M. O’Fallon, Drug therapy for vertebral fractures in osteoporosis: Evidence that decreases in bone turnover and increases in bone mass both determine antifracture efficacy. Bone 18(3 Suppl), 197S-201S (1996). N. H. Bjarnason, C. Christiansen, S. Sarkar, B. Mitlak, R. Knickerbocker, P. Delmas, and S. Cummings for the MORE Study Group, Six-Month changes in biochemical markers predict 3-year response in vertebral fracture rate in postmenopausal, osteoporotic women: Results from the MORE study. J. Bone Miner. Res. 14(S1), pS157 (1999).

CHAPTER 61

Osteoporosis and Its Nonskeletal Consequences Their Impact on Treatment Decisions DEBORAH T. GOLD KENNETH W. LYLES KATHY M. SHIPP MARC K. DREZNER

Departments of Psychiatry and Behavioral Sciences, Sociology, Psychology, and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Departments of Medicine and Aging Center, Sarah W. Stedman Center for Nutritional Studies, GRECC, Veterans Affairs Medical Center, Duke University Medical Center, Durham, North Carolina 27710 Departments of Physical Therapy and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Department of Medicine, University of Wisconsin Medical School, Madison, Wisconsin 53792

I. Introduction II. Primary Nonskeletal Outcomes of Osteoporosis III. Other Nonskeletal Consequences of Osteoporosis

I. INTRODUCTION Osteoporosis, the most prevalent metabolic bone disease, results in low bone mass and deterioration of the microarchitecture of bone [1]. The most frequent skeletal consequences of this disease are low-energy fractures of the hip, wrist, and vertebrae. The goal of much osteoporosis research in the last two decades has been to eliminate the characteristic disequilibrium of bone remodeling that leads

OSTEOPOROSIS, SECOND EDITION VOLUME 2

IV. Conclusions References

to bone density loss. Such research has led to the development of five FDA-approved anti-resorptive pharmacological agents of differing effectiveness for the prevention and/or treatment of this disease. However, underutilization of diagnostic and treatment options means that large numbers of people still develop the familiar physical sequelae of osteoporosis [2,3]. Comprehensive care for osteoporosis must include both anti-resorptive therapy as well as management strategies for

479

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

480

DEBORAH T. GOLD ET AL

its psychosocial and functional challenges [4]. Unfortunately, health care providers can spend little time with their patients in managed care environments. Thus, other ways in which to manage the psychosocial and functional problems of osteoporosis must be found. In this chapter, we identify the signs and symptoms of osteoporosis that have psychological, social, and functional consequences and suggest potential treatment strategies which can be utilized in a comprehensive care plan for this disease.

II. PRIMARY NONSKELETAL OUTCOMES OF OSTEOPOROSIS Low bone density and deterioration of the microarchitecture of bone lead directly to skeletal and, with advanced disease, anatomic changes that are the hallmarks of osteoporosis. In this section, we identify those physical outcomes of fractures that result in psychological and social harm to patients with osteoporosis.

A. Pain Of all outcomes of osteoporosis, the pain that results from multiple fractures has perhaps the most overwhelming impact on the people with this disease [5 – 7]. Indeed, osteoporotic pain reveals itself in unique ways. For some individuals, pain is not a component of osteoporosis. Some people have silent fractures that are often discovered secondary to diagnosis or treatment of other problems. Others may experience the acute pain that accompanies a fracture. The etiology of this pain is straightforward, and it will diminish over time. Because it will remit, coping with acute pain is relatively easy. Third, chronic pain can be the result of multiple fractures. This is typically associated with multiple vertebral fractures and often results in a complex psychosocial reaction including anxiety, depression, interpersonal relationship problems, and reduced quality of life [8]. Because analgesics are only marginally effective, there is virtually no pain relief, and chronicity becomes particularly bothersome from a social and psychological perspective [4]. The quest for relief of this chronic pain can overtake the life of an affected patient.

B. Physical Changes Elderly people most frequently display the overt physical changes caused by vertebral fractures and osteoporosis. Kyphosis, most frequently caused by vertebral compression fractures, influences the configuration of the entire body [9 – 11]. This spinal deformity is a constant reminder of osteoporosis, and changes in spinal configuration cause other

physiological manifestations to occur. In severe cases with progressive vertebral fractures, downward angulation of the ribs eventually leads to the seventh rib resting on the iliac crest as the gap between the ribs and ilium narrows. As a result, the abdomen protrudes with distention, constipation, and eructation as frequent symptoms. Additionally, neck and head positions over the body change so that upright posture becomes impossible. In the most severely affected patients, pseudospondylosthetic abnormalities with a posterior pelvic tilt, hamstring contracture, and hip joint flexion are characteristic. These conspicuous alterations of physical shape affect the self-image of people with osteoporosis and may negatively influence both interpersonal relations and self-esteem. Osteoporotic deformity also has profound effects on mobility. Changes in body configuration and posture often result in impaired gait and create difficulty with basic tasks such as bed mobility and rising from sitting to standing [12,13]. Gait velocity slows, and endurance becomes limited. Patients with severe deformity become unstable and frequently require assistive devices for ambulation. The mechanisms behind these effects have not yet been identified. However, we do know that the changed shape of the axial skeleton reduces lung capacity [14] and that overall deconditioning is a consequence of reduced overall activity level. Difficulty with basic mobility patterns such as changing from sitting to standing or moving in bed may result from altered alignment of body segments (which requires more effort for all mobility), reduced range of motion in the spine [12] and extremities, or pain. Enhanced efficiency, ease of movement, and reduced pain during mobility can be achieved through intervention, which addresses range of motion restrictions and teaches the patient how to align body parts as close to normal as possible and then maintain this alignment as movement tasks are accomplished. Cook et al. [15] report that fear of fracture and fear of falling were associated with osteoporosis. Ross et al. [16] also found that height loss is a strong predictor of fear of falling. These two fears often result in restriction of overall activity level, which reduces cardiovascular endurance and promotes muscular weakness and bone loss. Patients may develop fear of activity (and therefore fear of fractures) on their own if no guidelines for appropriate exercise and movement are provided. People with osteoporosis then search the Internet, hoping to find more specific information but are often misled there as well. If we instruct patients how to limit amounts of loads lifted, how to avoid impact forces, and how to avoid flexion and rotation of the spine with exercise and daily activities [17], and encourage them to continue all other activity, including appropriate exercise, this iatrogenic fear of activity diminishes. Trunk strength is also decreased in people with osteoporosis compared to age-matched individuals without osteoporosis [12,

481

CHAPTER 61 Osteoporosis and Its Nonskeletal Consequences

18]. If hyperkyphosis also is present, erect posture will be severely compromised. This results because greater strength of the erector spinae muscles is required to maintain erect alignment when the head, shoulders, and upper back are positioned anteriorly as occurs with hyperkyphosis. An exercise program designed to strengthen the trunk extensors is therefore strongly indicated. In fact, a study by Malmros et al. showed positive effects of a 10-week exercise program on pain and psychosocial function [19]. Balance capability may also be affected by the multiple physical impairments described above (deformity, range of motion restrictions, and reduced strength). We found that functional reach [20], a physical performance test that serves as a marker of balance, was reduced in women with two or more vertebral fractures when they were compared with age-matched nonfractured women [12]. In another study on functional reach, Duncan et al. found that limitations in this measure have also been associated with increased risk of falls [20]. Appropriate interventions geared toward reducing risk of falls by eliminating environmental hazards as well as addressing range of motion and strength deficits are needed for patients with established osteoporosis. Further, trochanteric hip pads may prevent hip fractures from occurring as a result of the falls that cannot be avoided [21,22].

C. Nutritional Limitations Calcium deficiency is a major risk factor for the evolution of osteoporosis and, as a result, health care professionals recommend that people at risk of osteoporosis meet their individual needs for calcium daily. Although the task of adding calcium to a diet may seem simple, unfortunately many individuals are unable to reach this goal. Lack of compliance occurs for complex reasons including lactose intolerance and fear of high-calorie foods. Despite the availability of calcium-rich dairy foods in the United States, many adults — especially women — receive less than an estimated 60% of their daily calcium intake from dairy products [23]. Consequently, increased consumption of calcium-rich low- or nonfat dairy foods offers a convenient and acceptable means of improving calcium intake via food sources that allow for a varied and balanced diet. Unfortunately, many people with or at risk of osteoporosis find the simple increase in dairy product intake intolerable. Lactose-intolerance is especially common in older adults but is substantial in adults of any age. Indeed, up to 15% of adult Caucasian and as high as 80 – 90% of the adult African-American and Asian populations exhibit lactase deficiency [24]. Moreover, although this defect is generally hereditary, acquired deficiency is often seen secondary to innocuous disorders such as viral and bacterial infections of the gastrointestinal tract and giardiasis. When

adults who have these disorders increase their dairy product intake, the result is often painful abdominal cramps, bloating and distention, or diarrhea. The intensity of these symptoms often precludes the desired increase in calcium consumption. However, research suggests that lactose intolerance should not be seen as an impediment to obtaining 1500 mg of calcium a day [25]. Other health care professionals may be reluctant to increase dairy product consumption in people who have other age-related health problems, such as hypercholesterolemia. In these cases, low- or nonfat dairy products should be incorporated into daily meal plans. An even greater challenge to calcium intake occurs for those who try to increase dietary calcium intake through increasing consumption of calcium-rich vegetables. Broccoli and certain greens (e.g., turnip, beet, and mustard) all have reasonable calcium content and availability; however, they all may substantially increase flatulence. Further, the amount of calcium provided in typical servings of these foods may make their use impossible. However, digestive enzyme supplements are available at a low cost to those who are intolerant. Complications such as these may lead physicians to adopt less efficient modes of nutritional therapy. This is unfortunate, since lifetime compliance with pharmaceutical therapy may be more viable if a change in dietary behavior can also be achieved. Patients should be taught about both sources of calcium: dietary and supplemental. Because calcium supplements are not regulated in the United States, consumers may not be certain of their calcium content, and an inordinate number are likely to discontinue supplement therapy because the associated constipation is intolerable. Interest is growing in an area of nutrition that is particularly relevant for people with osteoporosis: dietary supplements. Phytoestrogens, including ipriflavone, are purported to have a weak antiresorptive effect on the skeleton. Additionally, trace minerals such as boron, manganese, magnesium, and zinc, may also contribute to skeletal health. Research in this area has been going on for over two decades, yet it is only recently that a broader interest has been evidenced. This interest has become so much a part of mainstream scientific research that the National Institutes of Health have established, within the Office of the Director, an Office of Dietary Supplements.

III. OTHER NONSKELETAL CONSEQUENCES OF OSTEOPOROSIS In this section, we discuss some of the secondary consequences of osteoporosis. These may not be evident on physical examination or may be excluded from the medical history because patients themselves are reluctant to discuss or complain about them. We classify these secondary

482

DEBORAH T. GOLD ET AL

outcomes into three conceptual areas: psychological, functional, and social.

A. Psychological Consequences The relationship between mind and body is undeniably strong, and there is overwhelming empirical evidence that positive attitudes can overcome serious disease. We also know that such disease can result in major psychosocial consequences as well as physical problems. There are three common psychological reactions to osteoporosis and its related nontraumatic fractures. The first of these to appear is typically anxiety. In fact, the diagnosis of osteoporosis often causes individuals great apprehension about their future with this disease. Not only does the diagnosis of osteoporosis generate concern, but so does the fear of future fractures [26]. Anxiety does appear to diminish as time passes but can be profound at the discovery of osteoporosis [5]. Another psychological consequence of osteoporosis is the presence of depressive symptoms. Depression is the most prevalent mental health problem of older adults and it can also be a major problem for people with osteoporosis [4,28 – 29]. Depressive symptoms can be both physical and psychological and include sleep disorders, uncontrolled appetite, and lack of vigor or physical energy. Psychologically, patients report apathy, dejection, and uselessness. Social role loss may also contribute to depressive symptoms in osteoporotic patients. It appears that depressive symptoms are associated with low bone mass as shown by Michelson et al. [30] in their study comparing the bone density of matched groups of premenopausal women, one with and one without histories of depression. At five different sites, those subjects with histories of depression had significantly lower bone mass than did those without depressive histories. Another study from Portugal [31] looked at postmenopausal women and determined their bone densities. Those who had experienced depressive episodes had lower bone densities than did their nondepressed counterparts. If the presence of depressive symptoms in people with osteoporosis is substantial, is there also a high prevalence of major depressive disorders? In other words, do osteoporosis patients experience the number and duration of symptoms to qualify for a DSM-IV diagnosis of major depression? Although one might logically conclude that depressive symptoms lead to diagnoses of depression, we have no empirical evidence to support such a conclusion. In addition to anxiety and depression, a third psychological consequence can result from osteoporosis: loss of selfesteem [8]. This disease leads to deformity, disablement, and pain, all of which help rob patients of their self-esteem. In addition, severe osteoporosis can lead to almost total dependence on others. In the United States, those who are not

productive are marginalized as is easily seen with our attitudes toward older adults. We also marginalize those who, for health reasons, cannot participate in the usual activities of adult life. If such marginalization were not to occur, the self-esteem of those with osteoporosis would not be threatened.

B. Functional Consequences In addition to the psychological costs of osteoporosis, functional loss is also a major consequence. Such loss is not typically detected in a routine physical examination. Thus, specific questions about function must be included in any patient – provider interaction. This is especially true when primary care providers are managing this disease. Those patients with established bone loss who are seen by metabolic bone specialists have reported the following activities as problematic: bending, rising from a chair [31], walking and dressing [31 – 32], carrying [15,31], standing [15,32], fixing hair, getting into and out of bed, cooking, washing, bathing, and using the toilet [32]. In addition, those with severe bone loss experience impairment in the functional tasks they can perform [12]. In the study by Ryan and colleagues [32], there was a significant correlation between impaired performance of activities and both the number of collapsed vertebrae and a summed score of the severity of collapse of all vertebrae. Bathing and dressing were also significantly correlated with degree of kyphosis as measured from radiographs. Sixty percent of the subjects had disturbed sleep, and 42% reported difficulty in finding suitable clothes to wear [32]. Another study conducted by Ettinger and others as part of the Study of Osteoporotic Fractures [33] found that a significantly higher proportion of subjects with severe vertebral deformity (deformity graded  4 SD from the population mean) had great difficulty doing or could not do one or more of the following activities because of back pain: bending, dressing, lifting, reaching, getting out of an automobile, and standing for 2 h. As a result, the researchers concluded that substantial pain and disability occurred only in the presence of severe vertebral deformity It is clear that, once an individual has sustained a vertebral fracture, the risk of subsequent fracture is high, and the probability of progressive worsening of deformity is considerable [16]. In an earlier study, Ettinger and colleagues had found that 8.4% of subjects with moderate to severe deformity required some help at home [34]. In our own research, we found that women with two or more vertebral compression fractures needed significantly greater assistance with daily activities than did age-matched women without vertebral fractures. The fracture group also reported significantly more pain and difficulty with activities than did the nonfracture group [12].

483

CHAPTER 61 Osteoporosis and Its Nonskeletal Consequences

C. Social Consequences As noted above, there are several unique ways in which osteoporosis can intrude on an individual’s life. However, the impact of osteoporosis is not limited to within-person issues; there can be relationship problems that result from the disease as well. These social consequences of osteoporosis rarely surface immediately, and some individuals with mild or no symptoms never face these social problems. Unfortunately, someone with a hip fracture or multiple vertebral fractures is extremely vulnerable to limitations in social activities, with social isolation a frequent result [5, 35]. One social problem caused by osteoporosis is the social isolation that can result from an inability to engage in predictable interpersonal interactions. Other than a support group, people with severe osteoporosis often find it difficult to participate in social events because of their functional impairment and fear of future fractures. Chronic pain can also limit social interactions and preclude social reciprocity. Further, osteoporosis can lead to a loss of social roles. A social role is a position in society with clearly defined responsibilities such as a friend, a grandparent, or an accountant. All people have multiple social roles, usually in the family or job arena. As we age, we may naturally give up certain roles (such as employee or boss) and be forced to accept new roles which may be undesirable (e.g., widow). Many of these losses occur with aging, including the loss of “worker” roles and others which require physical competence. However, most older adults keep the majority of the social roles they have held throughout life (e.g., volunteer, bridge player). For the person with moderate to severe osteoporosis, social roles become serious challenges [8,36]. Certain roles such as parent or grandparent are easy to maintain on the surface. However, role-related tasks may become impossible. For example, a grandparent with multiple vertebral fractures cannot lift or carry a grandchild. For the person healing from a recent hip fracture, many family activities are simply impossible. Regardless of the nature of the social role and its seeming desirability, the inability to fulfill it leads to feelings of inadequacy and failure. Thus, serious psychological issues can emerge directly or secondarily from the social problems caused by osteoporosis. The final social variable upon which osteoporosis has an impact is giving and receiving social support. Social support is the provision of instrumental and emotional assistance to any individual in need [35,36]. Early in adulthood, instrumental social support is rarely needed, and individuals establish reciprocal emotional support interactions (e.g., exchanging tips on child rearing). This is an easy exchange when partners are healthy and when need is approximately equal. As people age, however, the need for social support increases, and the opportunity for reciprocity diminishes

[37]. Add the disability associated with osteoporosis to aging, and support exchanges become more difficult. Although family and friends can provide short-term, essential support in acute crises (e.g., after a hip or vertebral fracture), chronic support needs are less easily met. The person with osteoporosis becomes dependent and loses the ability to provide reciprocal support, therefore creating imbalance in the short-term relationship. Although life-long family relationships may not depend on direct reciprocity, most other social relationships do. Without some reciprocity, relationships typically end.

IV. CONCLUSIONS Treatment for osteoporosis in the past typically centered on the relationship between physician and patient. However, given the increasing demands of the managed care environment on physician time, it is impractical to assume that psychological, nutritional, and social needs can be addressed adequately in the context of the medical exam. Rather than ignore these challenging consequences of osteoporosis, we must find ways of providing these services within the current and future medical context. As we have discussed elsewhere [38], the multidisciplinary team approach to the management of osteoporosis allows team numbers to address specific problems in their areas of expertise. Osteoporosis is a complex disease with impact in multiple arenas of life, and no single practitioner has either the time or the training to manage all of its consequences. Patients with osteoporosis have had their functional and psychosocial problems ignored, in large part because physicians were unaware of the broad impact of this disease. Now, as we learn more about the ways in which osteoporosis affects quality of life, we need to identify and treat the psychosocial, nutritional, and functional limitations caused by this disease. Without such treatment, patients with osteoporosis are far more likely to experience impaired quality of life.

References 1. J. C. Stevenson and M. I. Whitehead, Postmenopausal osteoporosis. Br. Med. J. 285, 585 – 588 (1982). 2. M. Kleerekoper and L. V. Alvioli, Evaluation and treatment of postmenopausal osteoporosis. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), 2nd ed., pp. 223 – 229. Raven Press, New York, 1993. 3. D. J. Togerson and P. Dolan, Prescribing by general practitioners after an osteoporotic fracture. Ann. Rheum. Dis. 57, 378 – 379 (1998). 4. D. T. Gold, C. W. Bales, K. W. Lyles, and M. K. Drezner, Treatment of osteoporosis: The psychological impact of a medical education program on older patients. J. Am. Geriatr. Soc. 37, 417 – 422 (1989). 5. D. T. Gold, Chronic musculoskeletal pain: Older women and their coping strategies. J. Women Aging 6, 43 – 58 (1994).

484 6. M. C. Nevitt, B. Ettinger, D. M. Black, K. Stone, S. A. Jamal, K. Ensrud, M. Segal, H. K. Genant, and S. R. Cummings, The association of radiographically detected vertebral fractures with back pain and function: A prospective study. Ann. Intern. Med. 128, 793 – 800 (1998). 7. A. A. Ismail, C. Cooper, D. Felsenberg, J. Varlow, J. A. Kanis, A. J. Silman, and T. W. O’Neill, Number and type of vertebral deformities: Epidemiological characteristics and relation to back pain and height loss. Osteoporos. Int. 9, 206 – 213 (1999). 8. D. T. Gold, K. Stegmaier, C. W. Bales, K. W. Lyles, R. E. Westlund, and M. K. Drezner, Psychosocial functioning and osteoporosis in late life: Results of a multidisciplinary intervention. J. Women’s Health 2, 149 – 155 (1993). 9. U. Patel, S. Skingle, G. A. Campbell, A. J. Crisp, and I. T. Boyle, Clinical profile of acute vertebral compression fractures in osteoporosis. Br. J. Rheum. 30, 418 – 421 (1991). 10. A. C. Scane, R. M. Francis, A. M. Sutcliffe, M. J. D. Francis, D. J. Rawlings, and C. L. Chapple, Case-control study of the pathogenesis and sequelae of symptomatic vertebral fractures in men. Osteoporos. Int. 9, 91 – 97 (1999). 11. G. Leidig-Bruckner, H. W. Minne, C. Schlaich, et al., Clinical grading of spinal osteoporosis: quality of life components and spinal deformity in women with chronic low back pain and women with vertebral osteoporosis. J. Bone Miner. Res. 12, 663 – 675 (1997). 12. K. W. Lyles, D. T. Gold, K. M. Shipp, and C. F. Pieper, Osteoporotic vertebral fractures: Their association with impaired functional status. Am. J. Med. 94, 595 – 601 (1993). 13. C. Schlaich, H. W. Minne, T. Bruckner, et al., Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteoporosis Int. 8, 261 – 267 (1998). 14. J. A. Leech, C. Dulberg, S. Kellie, L. Pattee, and J. Gay, Relationship of lung function to severity of osteoporosis in women. Am. Rev. Respir. Dis. 141, 68 – 71 (1990). 15. D. J. Cook, G. H. Guyatt, J. D. Adachi, J. Clifton, L. E. Griffith, R. S. Epstein, and E. F. Juniper, Quality of life issues in women with vertebral fractures due to osteoporosis. Arthritis Rheum. 36, 750 – 756 (1993). 16. P. D. Ross, B. Ettinger, J. W. Davis, L. J. Melton, and R. D. Wasnich, Evaluation of adverse health outcomes associated with vertebral fractures. Osteoporosis Int. 1, 13 – 40 (1991). 17. M. Sinaki and B. A. Mikkelson, Postmenopausal spinal osteoporosis: Flexion versus extension exercises. Arch. Phys. Med. Rehab. 65, 593 – 596 (1984). 18. M. Sinaki, S. Khosla, P. J. Limburg, J. W. Rogers, and P. A. Murtaugh, Muscle strength in osteoporotic versus normal women. Osteoporosis Int. 3, 8 – 12 (1993). 19. B. Malmros, L. Mortensen, M. B. Jensen, and P. Charles, Positive effects of physiotherapy on chronic pain and performance in osteoporosis. Osteoporosis. Int. 8, 215 – 221 (1998). 20. P. W. Duncan, S. Studenski, J. Chandler, and B. Prescott, Functional reach: Predictive validity in a sample of elderly male veterans. J. Gerontol. 47, M93 – M98 (1992). 21. P. Lips, Epidemiology and predictors of fractures associated with osteoporosis. Amer. J. Med. 103, 3S – 8S (1997).

DEBORAH T. GOLD ET AL 22. S. N. Robinovitch, W. C. Hayes, and T. A. McMahon, Distribution of contact force during impact to the hip. Ann. Biomed. Engineer. 25, 499 – 508 (1997). 23. F. Bronner, Calcium and osteoporosis. Am. J. Clin. Nutr. 60, 831 – 836 (1994). 24. T. Sahi, Genetics and epidemiology of adult-type hypolactasia. Scand. J. Gastroenterol. 202, 7 – 20 (1994). 25. F. L. Suarez, J. Adshead, J. K. Furne, and M. D. Levitt, Lactose maldigestion is not an impediment to the intake of 1500 mg calcium daily as dairy products. Am. J. Clin. Nutr. 68, 1118 – 1122 (1998). 26. L. F. Berkman, L. Leo-Summers, and R. I. Horwitz, Emotional support and survival after myocardial infarction. Ann. Intern. Med. 117, 1003 – 1009 (1992). 27. D. T. Gold, S. D. Smith, C. W. Bales, K. W. Lyles, R. E. Westlund, and M. K. Drezner, Osteoporosis in late life: Does health locus of control effect psychosocial adaptation? J. Am. Geriatr. Soc. 29, 670 – 675 (1991). 28. V. L. Rose, Consensus statement update on depression in late life is issued by the NIH. Am. Fam. Phys. 57, 2013 (1998). 29. J. L. Fessler, Depression among the elderly. Wisc. Med. J. 95, 695 – 696 (1996). 30. D. Michelson, C. Stratakis, L. Hill, J. Reynolds, E. Galliven, G. Chrousos, and P. Gold, Bone mineral density in women with depression. N. Engl. J. Med. 335, 1176 – 1181 (1996). 31. R. Coelho, C. Silva, A. Maia, J. Parata, and H. Barros, Bone mineral density and depression: A community study in women. J. Psychosomat. Res. 46, 29 – 35 (1999). 32. G. Leidig, H. W. Minnie, P. Sauer, C. Wuster, J. Wuster, M. Lojen, F. Raue, and R. Ziegler, A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. 8, 217 – 229 (1990). 33. P. J. Ryan, G. Blake, R. Herd, and I Fogelman, A clinical profile of back pain and disability in patients with spinal osteoporosis. Bone 15, 27 – 30 (1994). 34. B. Ettinger, D. M. Black, M. C. Nevitt, A. C. Rundle, J. A. Cauley, S. R. Cummings, H. K. Genant, and Study of Osteoporotic Fractures Group, Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). 35. B. Ettinger, J. E. Block, R. Smith, S. R. Cummings, S. T. Harris, and H. K. Genant, An examination of the association between vertebral deformities, physical disabilities, and psychosocial problems. Maturitas 10, 283 – 296 (1988). 36. K. A. Roberto, Stress and adaptation patterns of older osteoporotic women. Women Health 14, 105 – 119 (1989). 37. K. A. Roberto, Women with osteoporosis: The role of the family and service community. Gerontology 28, 224 – 228 (1988). 38. N. Krause, Life stress, social support, and self-esteem in an elderly population. Psychol. Aging 2, 349 – 355 (1987). 39. D. T. Gold and M. K. Drezner, Quality of life and osteoporosis: Impact on delivery of care. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs, and L. J. Melton III, eds.), 2nd ed., pp. 475 – 486. Raven Press, New York, 1995.

CHAPTER 62

An Orthopedic Perspective of Osteoporosis MICHAEL H. HEGGENESS AND KENNETH B. MATHIS Department of Orthopaedic Surgery, Center for Spinal Disorders, Baylor College of Medicine, Houston, Texas 77030

I. Introduction II. Biomechanics III. Discussion of Specific Injuries

IV. Prevention of Fracture References

I. INTRODUCTION

II. BIOMECHANICS

The clinical significance of osteoporosis is overwhelmingly related to fracture events in affected patients. Management of these fractures comprises a large portion of the efforts of nearly all practitioners of orthopedic surgery. The goal of treatment of any osseous injury is a rapid return to normal function. This almost always includes an interval of immobilization of the injured bone by externally applied casts or braces or internally stabilizing it by operatively placed internal fixation devices. A disappointing number of fractures occur within or adjacent to articular surfaces. Accurate reduction of intraarticular fracture fragments and preservation of motion in the joint are important treatment objectives. Displacement of nonarticular fractures can also have important functional significance; translational, angular, and rotational deformities may also add complexity to the problems of a fracture management program.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Fractures of bone are complex events, and the process by which a bone is fractured depends on both extrinsic and intrinsic factors. Extrinsic factors are the direction, magnitude, and duration of the force acting on the bone as well as the rate at which the bone is loaded. Intrinsic properties include a bone’s geometry, energy-absorbing capacity, modulus of elasticity, and density. For example, the more rapidly a bone is loaded, the more energy it absorbs prior to fracture. Therefore, fractures which result from slow loading events are usually simple, with minimal comminution or fragmentation. In contrast, fractures produced by high loading rates are associated with sudden release of large amounts of energy, causing complex fracture patterns with multiple fragments. The complex shape of the bones of the human body, their variable densities, and the extremely complex loading patterns make the study of osseous fractures a challenging field. The density of the bone in question is only

485

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

486

HEGGENESS AND MATHIS

one of many features that affect not only the manner in which these fractures occur, but their treatment as well. Many fractures are most appropriately managed by nonoperative methods, such as external immobilization with a cast or brace. These treatments are usually not strongly affected by bone mineral density. In contrast, many other fractures, such as those involving an articular surface which must be very accurately reconstructed, are best managed by operatively placed internal fixation devices. The strength and efficacy of internal fixation is very strongly affected by bone quality. Great care, and occasionally drastically altered surgical technique is needed in the operative treatment of severely osteoporotic patients.

III. DISCUSSION OF SPECIFIC INJURIES A. Fractures of the Upper Extremity 1. COLLES’ FRACTURE Colles’ fracture of the distal radius is the most common fracture of the upper extremity. There are two peaks of incidence that occur, a pediatric group and a geriatric group, with females predominating in the elderly [1]. A dorsally displaced fracture of the distal radius commonly results from a fall onto the outstretched hand. In the majority of cases, this injury can be managed by closed reduction and immobilization in a cast. It is important that the original length of the radius be restored, as well as the normal, slightly palmar angulation to its distal articular surface. Should either objective be incompletely realized by closed reduction, either open reduction and internal fixation or, more commonly, application of an external fixation device may be indicated. Such devices, consisting of transfixing pins in the metacarpals and the radial shaft, may allow maintenance of this reduction (Fig. 1). While the results of these treatments are generally good, mild to moderate residual pain, stiffness of the wrist and fingers, osteoarthritic changes, causalgia, and diminished function are described in up to 31% of patients by Cooney [2]. 2. PROXIMAL HUMERUS FRACTURES Fractures of the proximal humerus at the junction of the humeral head and shaft are most commonly seen in elderly osteoporotic individuals. These fractures respond well to simple immobilization for 3 to 4 weeks, although rehabilitation of the muscles and restoration of joint motion may require weeks or months. Four major fragments are usually seen in proximal humerus fractures: the shaft, the head, and the greater and lesser tuberosities.

FIGURE 1 Lateral and PA views of a Colles fracture of the wrist. This fracture has been stabilized by the use of an external fixation device which allows control of the fracture fragments with ongoing distraction forces. Fractures are considered to be displaced if any of the four major segments is displaced over 1 cm or angulated more than 45°. More aggressive management is indicated when this occurs, usually with some type of internal fixation [3,5]. Unfortunately, many such fractures separate both the greater and the lesser tuberosities of the humerus, leaving the most proximal fragment (containing the articular surface) with essentially no significant muscular attachment or blood supply. These fractures are extremely difficult to treat since the poor bone quality of the humeral head makes internal fixation with pins or screws likely to fail. Even if internal fixation succeeds and the bone heals, there is still the risk of avascular necrosis of the head due to the disruption of blood supply in the head fragment at the time of injury. For these reasons, such badly comminuted fractures of the proximal humerus are frequently managed by primary hemiarthroplasty (replacement of the humeral head with a metal prosthesis)[4]. This procedure allows nearly immediate resumption of motion and in general provides functional results superior to those obtained with internal fixation (Fig. 2).

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

487

FIGURE 2 A comminuted fracture (A) of the proximal humerus managed by primary replacement (B) of the humeral head as fractures such as these put the patient at very high risk of avascular necrosis of the humeral head.

B. Fractures of the Lower Extremity 1. FRACTURES OF THE HIP The number of hip fractures is increasing worldwide due to the increasing number of elderly in the population. In the United States alone, approximately 275,000 hip fractures occur annually at a cost to the health care system of over $7.3 billion. The mortality risks associated with this diagnosis have been reported in past studies to be as high as 50% at 1 year. However, recent studies with more modern surgical management report a 1-year mortality of approximately 20%. Mortality risks become much higher if the patient has an altered mental status, is institutionalized, has severe medical problems, or is very old [6]. The primary goal in treatment of hip fractures is early mobilization. The only possible exception to this would be a nonambulatory patient who is either demented and or has little or no pain related to the fracture. In almost all other cases, operative intervention is preferred. Past experience in orthopedics has proven that nonoperative treatment results in very high rates of morbidity, mortality, nonunion, and malunion of these fractures. An additional factor to consider is that early operation, preferably within the first 24 h, decreases morbidity and mortality [7]. 2. CLASSIFICATION Hip fractures are classified by location into three major types: femoral neck, intertrochanteric, and subtrochanteric

fractures. The treatment of each type of fracture differs significantly. a. Femoral Neck Fractures Fractures of the femoral neck, sometimes referred to as subcapital fractures, lie entirely within the capsule of the hip joint (Fig. 3). These fractures rarely involve significant comminution, and, since the fracture hematoma is contained within the joint capsule do not generally result in major blood loss. Femoral neck fractures, particularly when displaced, do, however, have a reduced healing potential compared to other fractures of the hip. In addition, fractures in this location often compromise the blood supply to the femoral head, so that a high risk of avascular necrosis is associated with this particular injury. b. Intertrochanteric Fractures Fractures of the proximal femur at the level of the greater and lesser trochanter are frequently seen (Fig. 4). These fractures are frequently subjected to considerable comminution, and frank displacement of the greater and lesser trochanters, with their muscular attachments, frequently occurs. Since this fracture is extraarticular, substantial loss of blood into the proximal thigh can occur. The healing potential of these fractures is high and with adequate stabilization or immobilization, nonunions are rare. c. Subtrochanteric Fractures Fractures of the femur distal to the lesser trochanter are less common than the femoral neck and intertrochanteric fractures. Like the

488

HEGGENESS AND MATHIS

FIGURE 3 A displaced fracture of the femoral neck (A). These fractures occur within the joint capsule and when displaced put the femoral head at high risk of avascular necrosis. Nonunions are also common. For this reason displaced fractures are most frequently managed by primary hemiarthroplasty as shown in (B). Nondisplaced femoral neck fractures can be successfully managed by pins as shown (C).

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

489

FIGURE 4

(A) A fracture of the hip at the level of the lesser trochanter. These fractures are most commonly managed by a screw and plate device such as that shown in (B). These devices provide sufficient stability that immediate weight-bearing is possible. Nonunions are very rare even in extremely osteoporotic individuals.

intertrochanteric fracture, these injuries often occur with significant comminution. Operative treatment is complicated further by the need for extensive surgical exposure. Subsequently, intraoperative blood loss is often significant. Often these fractures require the use of more complex internal fixation devices which may also increase perioperative morbidity. 3. TREATMENT Treatment of hip fractures depends on many variables. Of primary concern is the anatomic location of the fracture. This is important as it may indicate if the blood supply to the femoral head has been compromised. The arterial supply to the femoral head is supplied by three routes: (i) ascending cervical arteries on the surface of the femoral neck, (ii) the artery of the ligamentum teres (which provides minimal blood supply to the head), (iii) intraosseous cervical vessels. Consequently, a displaced femoral neck fracture, in contrast to an intertrochanteric fracture, can disrupt both the surface vessels as well as

the intraosseous vessels, whereas the intertrochanteric fracture (which is below the level of the ascending cervical arteries) would not likely affect blood supply to the femoral head. Therefore, in displaced femoral neck fractures in the elderly, prosthetic replacement is indicated, as opposed to internal fixation. By contrast, internal fixation is almost always preferred for intertrochanteric fractures. Prosthetic replacement offers distinct advantages as primary treatment of femoral neck fractures. Most importantly, it allows immediate mobilization and full weight bearing on the affected limb. This is important in elderly debilitated patients who cannot maintain the required 6 weeks of non-weight-bearing for internal fixation because of their poor upper body strength and balance and/or dementia. Prosthetic replacement also diminishes the need for subsequent operations in the femoral neck fracture patient. Patients who have a poor outcome from fracture repair, due to either nonhealing related to poor bone quality, with subsequent loss of fixation, or to

490 aseptic necrosis from the initial vascular disruption, face the risk of further surgery. In one series, the incidence of reoperation for these complications after internal fixation was approximately 30%, while after prosthetic replacement the risk of reoperation was considerably less, and there were fewer complications, less pain, and better functional results at 2-year follow-up [13]. Fixation of prosthetic devices has evolved over the past century into three main categories. The first involves what is termed an interference or “press” fit. This type of fixation relies on obtaining stable fixation by the fit of the prosthesis at the time of surgery. The second type uses a cold-curing acrylic polymer cement (polymethylmethacrylate) to achieve fixation of the prosthesis to host bone. This type of fixation allows immediate weight bearing even for severely osteopenic bone, and is indicated for femoral component fixation in most middle aged and elderly patients. However, concerns over longterm deterioration of the mechanical properties of bone cement led to the most recent fixation strategy that employs porous metals to achieve biologic fixation. Initial prosthetic stability is obtained by a “press fit” and then the fixation is augmented by bone ingrowth into the porous interstices on the prosthesis. This type of fixation is usually indicated in younger active patients with otherwise healthy bones. It is hoped that this strategy will provide greater long-term success in joint arthroplasty (see Chapter 51 for further discussion of porous implants). Internal fixation devices of the hip are generally divided into three categories. (i) Cannulated pins or screws: These are primarily used for femoral neck fractures. Usually, three such screws are used in the femoral head. (ii) Sliding hip nails or screws: These devices rely on a single large flanged nail or screw that can slide within a barrel attached to the femur via a plate. This construct allows both femoral neck and intertrochanteric fractures to collapse in a controlled fashion that allows compression at the fracture site and enhances stability for fracture healing while also preventing the screw from penetrating out of the femoral head. (iii) Intramedullary devices: These implants rely on an intramedullary rod used with transfixion screws that pass through the rod into the femoral head and shaft. Theoretically, this device will bear a reduced bending moment due to its medial position inside the femur, and may have fewer mechanical failures than will nail-plate devices. Intramedullary devices are used primarily for unstable intertrochanteric and subtrochanteric fractures, however, as insertion requires increased operative dissection and blood loss. Systemic disease can influence selection of treatment for hip fracture. For example, in patients with Paget’s disease, hip fractures are generally best treated by prosthetic replacement unless the fracture is absolutely nondisplaced. Care is taken to inspect the acetabulum at the time of surgery. An acetabular replacement which would convert

HEGGENESS AND MATHIS

the procedure to a total hip replacement can be placed if significant degeneration of the acetabular surface is seen. A nondisplaced fracture in a patient with severe arthritic changes may be better treated by total hip replacement than by fracture repair as a return to the operating room at a later date for conversion to a total hip replacement may otherwise be difficult to avoid. Those few fractures that cannot be reduced or have failed internal fixation should be considered for insertion of a prosthesis. However, some relative contraindications to the use of a prosthesis also exist. These include young active patients and patients with either preexisting joint infection or with recurrent bacteremia. A “total joint” prosthesis is at higher risk for hematogenous bacterial seeding than is a hemiarthroplasty. 4. TIBIAL PLATEAU FRACTURES Fractures of the tibial plateau occur often in elderly patients and nearly always involve a primary injury force of axial loading. This is frequently also combined with angular and rotational forces. An osteoporotic individual can acquire a badly comminuted fracture of the proximal tibia through as simple an act as stepping off a curb. These fractures almost invariably involve the articular surface, which is usually displaced and impacted down into the softer cancellous bone of the proximal tibia. Restoration of an adequate joint surface is an important treatment consideration, as is preservation of the complex ligamentous stabilizing structures of the knee. Since the tibial condylar surface slopes posteroinferiorly 10° to 15°, accurate assessment of the plateau fracture requires that the X-ray beam be angled 10° to 15° inferiorly. Alternatively, tomography may be required if plain radiographs are inconclusive. In addition, since fracture of the weight bearing surface of the tibial plateau can lead to instability, stress radiographs are commonly obtained. Treatment of these fractures is directed at maintaining knee motion and reasonable alignment, and avoiding instability. Fractures with less than 5 mm of condylar depression do not cause instability in stress testing, and where less than 10 mm of central articular depression occurs, may be treated nonoperatively. This is best accomplished with early motion and a cast brace. Early reports showed that prolonged immobilization of these fractures allowed healing, but at the cost of dense intraarticular adhesions and poor knee motion. Patients with greater displacement of the fracture are best treated by open reduction and internal joint alignment with rigid internal fixation of the bone. As this injury frequently results in crushing and loss of bone stock in the proximal tibia, the void left in the tibial metaphysis when the the articular surface is elevated requires a graft both to fill the void and also to buttress the joint fragments. The usual source of graft is the iliac crest, although cadaveric bone can also be used. Ceramics are

491

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

FIGURE 5

(A) A severe fracture of the distal tibia. This fracture severely disrupts the ankle joint, with extension of the fracture into the shaft of the tibia. (B) A combination of internal and external fixation to reduce and stabilize this complex injury.

under investigation as a bone graft substitute for these metaphyseal fractures. 5. ANKLE FRACTURES Fractures of the ankle are common in all age groups, and the elderly osteoporotic patient is no exception. While these fractures sometimes involve the weight bearing surface of the tibia (the plafond), they more frequently involve displaced fractures of the medial or lateral malleolus. A variety of torsional and translational forces can cause these injuries, which almost always result in loss of the integrity of the precise anatomy of the ankle joint. According to Ramsey, even a talar shift of 1 mm in the joint can result in a 42% reduction in tibiotalar contact and can lead to higher point contact stresses and posttraumatic arthritis [14]. Because of this fact, most of these injuries are managed operatively, with open reduction and internal fixation (Fig. 5). The distal tibia, or medial malleolus, is usually best stabilized by one or two transfixing screws. The distal fibula, or

lateral malleolus, is more commonly treated by a screw plate device. Precise reconstruction of the normal joint anatomy is important to minimize late osteoarthritic change. This joint has a poor tolerance for even minor incongruencies in its surface. 6. METATARSAL FRACTURES The most common fracture of the metatarsal bones is an avulsion fracture of the base of the fifth metatarsal, an avulsion of the tuberosity caused by contraction of the peroneus brevis muscle. This is frequently a consequence of stepping in a hole, off a step, or twisting the foot. The avulsion fragments rarely displace enough to cause difficulty and generally heal well with symptomatic treatment alone. In contrast, a fracture of the shaft at the proximal fifth metatarsal just a few millimeters distal to those noted above, often requires more aggressive treatment, as nonunion and delayed union are frequent. On rare occasion, internal fixation of this fracture is indicated.

492

HEGGENESS AND MATHIS

FIGURE 6 Schematic axial views. (A) An unfractured vertebra. (B) A compression fracture. Bony injury to the anterior portion of the vertebral body results in loss of height; however, the bony borders of the spinal canal are uninjured. (C) Schematic of a burst fracture. Fracture of the vertebral body with involvement of the borders of the spinal canal. This injury is often associated with retropulsion of bony fragments into the canal, which can cause neurologic deficit.

C. Fractures of the Axial Skeleton 1. CERVICAL SPINE Osteoporotic fractures of the cervical spine are very rare indeed. All fractures of the cervical spine require aggressive work-up and imaging by either CT or MRI. A detailed medical history and neurologic examination are also mandatory. Management of elderly osteoporotic patients with cervical spine fractures is essentially the same as that recommended for younger individuals. One exception may be the minimally displaced fracture of the dens (odontoid), which in some elderly patients may lead to minimal morbidity. A high suspicion for metastatic disease should be maintained in any patient who relates a history of minimal trauma leading to a cervical spine fracture. 2. FRACTURES OF THE THORACIC AND LUMBAR SPINE Vertebral fractures associated with osteoporosis are commonly seen, although prevalence and incidence rates

are extremely difficult to estimate as discussed elsewhere in this volume. Precise diagnostic criteria are difficult to establish. Minor degrees of end plate collapse are often quite difficult to appreciate. In addition, projectional artifact and the low mineral density of the bones themselves contribute to the difficulty [23]. Classic epidemiologic work of Urist, Saville, Riggs, Melton, and others has demonstrated a wide variation in the presentation and progression of vertebral collapse in osteoporosis [15 – 23], from which some general conclusions can be formed. In general, the earliest fracture events occur in the upper thoracic spine. Interestingly, many of these fractures are asymptomatic. Progressive collapse of multiple vertebrae in this area can lead to a signifi cant upper thoracic kyphosis, often unfortunately referred to as a “dowager’s hump.” Adding to the diagnostic challenge patients not infrequently present with acute back pain without initial radiographic evidence of fracture. Frequently, images taken days or weeks later will document vertebral collapse. This phenomenon illustrates an important point about osteo-

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

porotic compression fractures: the fractured vertebra frequently demonstrates insidious progressive collapse over weeks or months to a degree not seen in younger patients. a. Classification An orthopedic surgeon or neurosurgeon, when evaluating axial load will classify these injuries initially on the basis of the integrity of the posterior cortex of the vertebral body. Fractures which involve the vertebral end plate and anterior cortex, and spare the posterior cortex, are called compression fractures (Fig. 6B). Those in which the posterior cortex of the body is also involved are referred to as burst injuries (Fig. 6C). Burst fractures in all cases are potentially much more morbid because of the potential for nerve compression. In addition, when the posterior cortex (or “middle column”) of the vertebral body is injured, a much greater potential for collapse, angulation, and progression of deformity exists. The majority of osteoporotic vertebral fractures result from failure under axial compression [28,29]. These injuries are commonly referred to as compression fractures although they are known quite frequently to involve the posterior cortex of the vertebral body. Many such fractures are therefore technically true burst fractures. As very few of these are imaged by CT or MRI, the true incidence of middle column injury is not known. Classic anatomic studies by Schmorl, Jaffee, and others have shown, however, that this is a very frequent occurrence [30,31]. Most attempts to classify these injuries have been based on plain radiographic criteria, and the classification system of Eastell et al. [32] has proven useful. Vertebral fractures are often referred to by their gross morphology as “bioconcave” or “codfish” fractures, “wedge” fractures, or “crush” fractures. The crush fracture represents a gross failure of anterior and posterior cortex of the vertebral body. It is not known what percentage of biconcave or wedge fractures also have some posterior cortical involvement. Indeed, as mentioned above, it is frequently observed that what appears to be an isolated fracture may undergo progressive collapse to a wedge and subsequently a crush fracture appearance over days and weeks of observation. On rare occasion, this sequence of events can lead to devastating late neurologic dysfunction [33 – 38]. b. Management Management of patients with osteoporosis-related compression fracture includes investigation of other possible causes of pathologic fractures and, in the vast majority of cases, nonoperative care. Appropriate evaluation and therapy of such patients is fully discussed in other chapters (see Chapters 57 and 61 respectively) and will be briefly given here. The patient’s history should include specific reference to osteoporosis risk factors, such as history of smoking or excessive alcohol intake, and detailed surgical and medical history. The possibility of multiple myeloma must be kept specifically in mind. A history of

493 weight loss may be particularly suggestive of malignancy. Radiographs should be carefully examined for fracture morphology. A history of previous fracture can be very useful. A laboratory assessment should be routinely performed to include complete blood count, sedimentation rate, serum protein electrophoresis, urinalysis, and thyroid function tests. In an elderly caucasian woman without evidence of other contributing history, a tentative diagnosis of osteoporosis may be entertained if this workup is negative. Men or young women withosteoporosis may require additional workup and endocrinologic consultation. A careful neurologic examination, particularly of the lower extremities, is mandatory. A general physical examination to include breast exam and palpation of the thyroid is encouraged. The presence of objective neurologic dysfunction presents a strong indication for CT or magnetic resonance imaging. Education of the patient should be part of the initial phase of management. It is very useful for the patient to understand the diagnosis and its implications. The possibility of subsequent fractures should be discussed, although it is important to stress that vertebral fractures do heal successfully in the overwhelming majority of cases and that spontaneous resolution of pain may be expected in 2 to 10 weeks, regardless of treatment. Patients are counseled to seek prompt reevaluation should neurological signs or symptoms develop. Pain management is a critical concern in these patients. When pain is inadequately addressed, many patients become limited to bed rest, which places them at risk for venous thrombosis and a worsening of their osteoporosis on the basis of inactivity. The exact effect of prolonged bed rest on the mineral density of osteoporotic patients has not been studied. Bed rest studies on younger patients, however, indicate that bone loss of up to 1% per week can be expected [39]. On this basis, the authors strongly discouraged bed rest as treatment and defined patient mobilization as a critical aspect of care. A short course of oral narcotic analgesics is often indicated to allow patients a reasonable level of activity. The use of braces for acute fracture is controversial. It is certainly true that many elderly patients, despite their pain, are unable to tolerate a brace. Attempts to brace fractures in the upper thoracic spine are particularly difficult and too rarely successful for pain management. On the other hand, simple braces, such as canvas corsets, can be extremely useful for lumbar fractures and often afford a dramatic level of pain relief. Bracing of thoracolumbar and mid-thoracic fractures is more difficult as a lumbosacral corset often does not provide adequate support to this region. Many patients with such thoracolumbar fractures find significant relief with the use of custom molded soft foam braces. There is also ongoing controversy about brace use for these problems because of the theoretical possibility that stress shielding of the spine may occur with exacerbation of

FIGURE 7

This series of radiographs demonstrates the progressive insidious collapse often seen in vertebral fractures. (A) A lateral radiograph of an 82-year-old male patient, 3 weeks after the onset of acute back pain. An image obtained 3 weeks later is shown in (B). Note the continued loss of height demonstrated in this L-2 fracture. This patient had complaints of leg pain and weakness and the presence of bony fragments within the canal is demonstrated on the CT images shown in (C). This patient was managed nonoperatively with a brace and analgesics and did make a complete recovery.

494

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

495

FIGURE 8 Another example of the insidious collapse sometimes seen with vertebral fractures in osteoporotic patients. (A) A compression fracture following a minor fall in a 78-year-old female. Fracture of L-2 is noted. (B) Lateral radiograph of the lumbar spine of this patient obtained 7 weeks later. Note the profound collapse of L-2 and the fracture of the adjacent L-3 body. At this time this patient was complaining of severe leg pain with associated profound motor deficits. osteoporosis. It is the author’s experience that osteoporotic patients will use the brace only so long as it is useful for severe pain management, and the benefits of keeping the patient ambulatory may outweigh potential risks of stress shielding. Unfortunately, no firm data exist on which to form a firm conclusion on this issue. A minority of patients will experience such severe pain and physical limitation from fracture that hospitalization is required for supportive care and parenteral pain medication. Mobilization of these patients, even when hospitalized, is encouraged. Parenteral calcitonin use is gaining in the acute management of vertebral compression fractures. For unknown reasons, many patients who have sustained vertebral compression fractures obtain dramatic analgesia from the use of calcitonin. While the basis of this phenomenon may lie in the documented CNS receptors for this hormone, its precise mode of action in analgesia remains unknown (see Chapter 73 by Civitelli). Parenteral doses of calcitonin of approximately 100 IU per day are extremely effective in pain relief for some patients. Hypersensitivity reactions have been de-

scribed, however, and many patients do experience transient gastrointestinal symptoms of nausea and vomiting during the initial days of therapy. Because of this, smaller doses are usually given initially (5 – 20 IU) and the dosage is slowly increased into the therapeutic range over 3 to 5 days. Symptomatic treatment of nausea is often helpful during this interval. The authors are aware that a nasal spray form of calcitonin can be useful for analgesic purposes. We find empirically that the subcutaneous administration is more effective. A controlled clinical study on this problem would be most welcome. Calcitonin treatment and external bracing can usually be discontinued within 4 to 10 weeks of the fracture event. Some patients unfortunately experience the relentless occurrence of multiple vertebral fractures through their sixth, seventh, and eighth decades. Dramatic kyphotic deformity and severe postural impairments often result. Chronic back pain with associated degenerative disease and kyphosis can be an extremely frustrating problem.

496

HEGGENESS AND MATHIS

FIGURE 8 (continued) An MRI image (C) documents disruption of the borders of the vertebral canal with bone fragments impinging the neural elements. This patient was treated surgically by an anterior approach to the lumbar spine, debridement of the fractured bone with decompression of the neural elements, bone grafting, and anterior instrumentation. A postoperative image demonstrating the Kostuik Harrington rods is shown in (D). Progression of kyphosis will usually stop when the lower ribs begin to impinge on the iliac wings. Unfortunately, this is frequently associated with local pain due to irritation of soft tissues and costal nerves in this area. In rare cases, severe intractable pain may be managed by costal nerve blocks. Spinal osteotomy, rib resection, and multiple level spinal fusions are strongly contraindicated. c. Spine Fracture with Neurologic Deficit The literature suggests that while the incidence of vertebral fracture in the aging population is high, neurologic dysfunction results from these fractures only in extremely rare cases. Reports of such cases have appeared with much greater frequency in recent years, however, and it is likely that this phenomenon is more common than has been previously appreciated [33 – 35,37,38]. All reports describe common features in the clinical presentation of these injuries. In nearly every case, fractures occurred either spontaneously or after minor trauma, and the initial presentation involved a complaint of back pain

only. The patient subsequently experienced progressive, insidious collapse of the fractured vertebra and acquired radicular pain and neurological deficit weeks or months after the index fracture event (Fig. 7). It is the authors’ observation that radicular pain always precedes the development of motor deficits. Shikata [36], Keneda [37], and others have advocated aggressive surgical decompression and operative stabilization of these injuries. These authors have found that conservative management, consisting of aggressive bracing, analgesics and physical therapy can also give excellent results in patients with relatively minor deficits. In patients with major neurological deficits and dramatic motor dysfunction, there is general agreement that operative management, while difficult, is usually indicated. These patients are frequently elderly and often have associated medical problems associated with smoking, alcohol, or other complicating illness. Surgical techniques for dealing with these injuries must be individualized. In general, operative intervention consists of an anterior approach to the spine, corpectomy, and reconstruction.

497

CHAPTER 62 An Orthopedic Prespective of Osteoporosis

FIGURE 9 An alternative to anterior surgery and instrumentation is a posterior pedicle screw device. This image indicates the postoperative appearance of the patient who has undergone posterior instrumentation and stabilization for this fracture of L-3. We favor anterior instrumentation, although simultaneous or staged posterior stabilization may also be appropriate, depending on surgeon preference and the individual clinical situation. Iliac crest autograft struts as well as allograft and ceramic spacers may all have a role in the surgical treatment of this injury. Instrumentation of the anterior spine in severely osteopenic patients requires meticulous technique and the creation of a construct with load sharing between unfractured posterior elements, the bone graft itself, and the instrumentation system. The authors personally favor the Kostuik – Harrington device [40] (Fig. 8) for instrumentation of these fractures as the screw thread design allows excellent purchase in the vertebral bodies. Extended biomechanical studies have not addressed the question of the relative advantage of different screw designs. However, large cancellous threads would seem to offer some advantage in this situation. Bicortical technique is also advised as this does add substantially to the biomechanical stability of such screws [41]. Biomechanical studies of posterior instrumentation techniques have suggested that hook constructs may offer some advantage over pedicle screw constructs in osteoporotic individuals [42], although specific techniques such as rigid cross-linking and hook screw constructs leave many options for the surgeon to individualize treatment (Fig. 9). The authors’ experience suggests that this clinical phenomenon of neurological deficit of osteoporosis related spine fracture is probably a good deal more common than is generally appreciated and the insidious presentation of the neurological deficits frequently leads to a missed or delayed

diagnosis. Increased awareness of this injury may lead to more prompt and accurate diagnosis and, it is to be hoped, more appropriate treatment. Late follow-up of conservatively treated patients who have resolved their neurological deficit reveals progressive resorption and remodeling of the offending bone fragments within the canal, although the resorption occurs much more slowly than the clinically observed neurological recovery. d. Fractures of the Sacrum The patient with severe osteoporosis is also at risk for spontaneous fracture of the sacrum. While numerous fracture patterns have been documented, the most common, “sacral insufficiency fractures,” involve bilateral fractures of the sacral alae, which in these authors’ experience are usually not visualized on plain roentgenography. A CT image will reveal the characteristic vertical fracture lines. These fractures are almost always managed conservatively and excellent results are usually attained with symptomatic treatment alone. On rare occasion, dramatic displacement of these fracures can occur.

IV. PREVENTION OF FRACTURE Surgical management of a patient with an acute fracture, where decreased mineral density is suspected, should include an assessment of fracture risk and appropriate treatment and counseling for the patient’s overall osteoporotic condition as well as his or her recent fracture.

498

HEGGENESS AND MATHIS

Immediately following fracture, the hospital environment is an excellent place to initiate counseling with regard to diet, calcium and vitamin D supplementation, and potential antiresorptive therapies. A bone mineral density examination should be considered. The fracture event provides an opportunity to initiate ongoing treatment for bone fragility, which should continue long after hospital discharge. Some patients are already taking appropriate treatment for osteoporosis at the time that they sustain a fracture. With few exceptions they should be encouraged to remain on medications throughout the treatment phase of their injury. The treatment of spinal osteoporosis and fracture can be very frustrating for the physician and patient alike. It is likely that the best solution for this problem may result not from its cure but from its prevention. Lifestyle modification, including increased exercise throughout life, postmenopausal estrogen supplementation, and adequate calcium and vitamin D intake together, may significantly decrease the magnitude of this problem in the future. To be effective, however, this approach must rely on identification, education, and treatment of high-risk individuals in adolescence or in early and middle adult years. For patients with established osteoporosis, acute treatment of fractures significantly decreases suffering. Home safety and fall prevention, however, may assist the patients in avoiding these problems entirely. Activities and situations where falls are likely should be approached with great caution. An obstacle-strewn living environment (with exposed lamp cords and throw rugs), icy sidewalks, and dimly lit stairways represent preventable causes of fracture. A discussion of these issues should be part of the treatment of any patient with osteoporosis (see Chapter 32).

References 1. P. A. Alffram and G. C. H. Bauer, Epidemiology of fractures of the forearm: A biomechanical investigation of bone strength. J. Bone Jt. Surg. 44A, 105 – 114 (1962). 2. W. P. Cooney III, J. H. Dobyns, and R. L. Linscheid, Complications of Colles’ fractures. J. Bone Jt. Surg. 62A, 613 – 618 (1980). 3. R. J. Hawkins, R. H. Bell, and K. Gurr, The three part fracture of the proximal part of the humerus: Operative treatment. J. Bone Jt. Surg. 68A, 1410 – 1414 (1986). 4. Cofield, R. Comminuted fractures of the proximal humerus. Clin. Orthop. 230, 49 – 57 (1988). 5. E. L. Flatow, F. Cuomo, M. G. Maday, S. R. Miller, S. J. McIlveen, and L. U. Bigliani, Open reduction and internal fixation of two-part displaced fractures of the greater tuberosity of the proximal humerus. J. Bone Jt. Surg. 73A, 1213 – 1218 (1991). 6. G. K. Ions and J. Stevens, Prediction of survival in patients with femoral neck fractures. J. Bone Jt. Surg. 69B, 384 – 387 (1987). 7. S. Sexton and J. Lehner, Factors affecting hip fracture mortality. J. Orthop. Trauma 1, 298 – 305 (1988). 8. T. J. Wilton, D. J. Hosking, E. Pawley, A. Stevens, and L. Harvey, Osteomalacia and femoral neck fractures in the Elderly patient. J. Bone Jt. Surg. 69B, 388 – 390 (1987).

9. M. Lender, M. Makin, G. Robin, R. Steinberg, and J. Menczel, Osteoporosis and fractures of the neck of the femur: Some epidemiologic considerations. Isr. J. Med. Sci. 12, 596 – 600 (1976). 10. G. P. Vose and R. M. Lockwood, Femoral neck fracturing: Its relationship to radiographic bone density. J. Gerontol, 20, 300 – 305 (1965). 11. M. Makin, Osteoporosis and proximal femoral fractures in the female elderly of jersusalem. Clin. Orthop. 218, 19 – 23 (1987). 12. I. Sernbo and O. Johnell, Changes in bone mass and fracture type in patients with hip fractures. Clin. Orthop. 238, 139 – 147 (1989). 13. C. Zetterberg, S. Elmersson, and G. B. Andersson, Reoperations of hip fractures. Acta Orthop. Scand. 56, 8 – 11 (1985). 14. P. Ramsey and W. Hamilton, Changes in tibiotalar area of contact caused by lateral talar shift. J. Bone. Jt. Surg. 58A, 356 (1976). 15. M. R. Urist, M. S. Gurvey, and D. O. Fareed, Long term observations on aged women with pathologic osteoporosis In “osteoporosis,” pp. 3 – 37. Grune & Stratton, New York, 1970. 16. P. D. Saville, Observations on 80 women with osteoporotic spine fractures In “Osteoporosis”, pp. 38 – 46. Grune & Stratton, New York, 1970. 17. J. A. Kanis and F. A. Pitt, Epidemiology of osteoporosis. Bone 13, S7 – S15 (1992). 18. A. P. Iskrant and R. W. Smith, Osteoporosis in Women 45 Years and over related to subsequent fractures. Public Health Rep. 84, 33 – 38 (1969). 19. L. J. Melton III, S. H. Kan, M. A. Frye, H. W. Wagner, W. M. N. O’Fallon, and B. L. Riggs, Epidemiology of vertebral fractures during 30 years. Calcif. Tissue Int. 42, 293 – 296 (1988). 20. U. Bengner, O. Johnell, and I. Redlund-Johnell, Changes in incidence and prevalence of vertebral fractures during 30 years. Calcif. Tissue Int. 42, 293 – 296 (1988). 21. L. V. Aviolo, Significance of osteoporosis: A growing international health problem. Calcif. Tissue Int. 49, 55 – 57 (1991). 22. G. Leidig, H. W. Minne, P. Sauer, C. Wuster, J. Wuster, M. Logen, F. Raue and R. Ziegler, A Study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. Density 8, 217 – 229 (1990). 23. C. Cooper, E. Atkinson, M. O’Falcon, and L. J. Melton III, Incidence of clinically diagnosed vertebral fractures: ‘A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 220 (1990). 24. B. L. White, W. D. Fisher, and C. A. Laurin, Rate of mortality for elderly patients after fracture of the hip in the 1980’s. J. Bone Jt. Surg. 60A, 930 – 934 (1978). 25. C. W. Miller, Survival and ambulation following hip fracture. J. Bone Jt. Surg. 60A, 930 – 934 (1978). 26. M. Kleerekoper and D. A. Nelson, Vertebral fracture or vertebral deformity? Calcif. Tissue Int. 50, 5 – 6 (1992). 27. J. M. Lane, C. N. Cornell, and J. H. Healey, Orthopaedic consequences of Osteoporosis In “Osteoporosis: Etiology, Diagnosis and Management,” pp. 433 – 455. Raven Press, New York, 1988. 28. R. Denis, The three column spine and its significance in the classifi cation of acute thoracolumbar spinal injuries. Spine 8, 8176 – 831 (1983). 29. F W. Holdsworth, Fractures, dislocations and fracture dislocations of the spine. J. Bone Jt. Surg. 52A, 1534 – 1551 (1970). 30. G. Schmorl and H. Junghans, “The Human Spine in Health and Disease.” Grune & Stratton, New York, 1971. 31. H. J. Jaffe, “Metabolic Degenerative and Inflammatory Diseases of Bone and Joints.” Lea & Febiger, Philadelphia, PA, 1972. 32. R. Eastell, S. L. Cedel, H. W. Wahner, B. L. Riggs, and L. J. Melton III, Classification of vertebral fractures. J. Bone Miner. Res. 6, 2076 – 215 (1991).

CHAPTER 62 An Orthopedic Prespective of Osteoporosis 33. R. A. Arciero, K. Y. K. Leung, and J. H. Pierce, Spontaneous unstable burst fracture of the thoracolumbar spine in osteoporosis: A report of two cases. Spine 14, 114 – 117 (1989). 34. C. Salomon, D. Chopin, and M. Benoist, Spinal cord compression: An exceptional complication of spinal osteoporosis. Spine 13, 222 – 224 (1988). 35. S. B. Tan, J. A. Kozak, and M. E. Mawad, The limitations of magentic resonance imaging in the diagnostic of pathologic vertebral fractures. Spine 16, 919 – 923 (1991). 36. J. Shikata, T. Yamamuro, H. Iida, K. Shimizu, amd J. Yoshikawa, Surgical treatment of paraplegia resulting from vertebral fractures in senile osteoporosis. Spine 15, 485 – 489 (1990). 37. K. Kaneda, S. Asano, T. Hashimoto, S. Satoh, and M. Fujiya, The treatment of osteoporotic – posttrumatic vertebral collapse using the Kaneda device and a bioactive ceramic vertebral prosthesis. Spine 17(S), 295 – 303 (1992).

499 38. M. H. Heggeness, Spine fracture with neurological deficit in osteoporosis. Osteoporosis Int. 3, 215 – 221 (1993). 39. A. D. LeBlanc, V. S. Schneider, H. J. Evans, D. A. Engelbretson, and J. M. Krebs, Bone mineral loss and recovery after 17 weeks of reduction. J. Bone Miner. Res. 5, 843 – 850 (1990). 40. J. P. Kostuik, Anterior fixation for burst fractures of the thoracic and lumbar spine with or without neurological involvement. Spine 13, 286 – 293 (1988). 41. S. Breeze, J. Alexander, P. S. Noble, and M. H. Heggeness, “A Biomechanical Study of Thoracolumbar Screw Fixation”. Presented North American Spine Society, Minneapolis, MN, 1994. 42. J. D. Coe, K. E. Warden, M. A. Herzig, and P. C. McAfee, Influence of bone mineral density on the fixation of thoracolumbar implants. Spine 15, 902 – 907 (1988).

501

CHAPTER 63 Clinical Use of Bone Biopsy

CHAPTER 63

Clinical Use of Bone Biopsy P. CHAVASSIEUX, M. ARLOT, AND P. J. MEUNIER INSERM Unité 403, Faculté RTH Laennec, 69372 Lyon, France

I. Introduction II. Methodology III. Indications for Bone Biopsy

IV. Conclusions References

I. INTRODUCTION

thereby allowing the possibility to make a diagnosis of mineralization defect from a dynamic analysis. Clinical use of bone biopsy and bone histomorphometry now serves three main needs in osteoporotic patients: the diagnosis of the bone disease in order to expressly exclude osteomalacia, systemic mast-cell disease, or any other marrow abnormality; the evaluation of the mechanisms underlying osteoporosis at the tissue or basic multicellular unit (BMU) level in order to identify low or high turnover; and the evaluation of the effects of therapeutic agents on bone remodeling and quality. The latter includes skeletal microarchitecture, the texture of bone matrix, and the possible presence of mineralization defects.

Bone biopsy is an important tool for diagnosis and research in osteoporotic patients, on the condition that it is processed without prior decalcification and analyzed using well-standardized histomorphometric methods. Bone histomorphometry consists of measuring parameters reflecting bone structure and turnover and is usually performed on transiliac bone biopsies. Today, precise noninvasive measurements of bone mineral content or bone mineral density accurately quantify trabecular and cortical bone from different sites (e.g., spine, femur, radius, whole body) and biochemical markers of bone turnover provide precise information on whole body bone remodeling. However, histomorphometry remains the only method that allows the study of bone at the tissue or cell level to enable measurements at intermediary levels of organization of bone, i.e., the osteon in cortical bone and the trabecular basic structural unit in spongy bone. With the use of the tetracycline double labeling procedure it also provides dynamic parameters [1]. This process has permitted the introduction of a time dimension into the quantitative analysis,

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. METHODOLOGY Bone histomorphometry requires high-quality undecalcified sections of bone. Strict methodologic conditions must be observed in order to obtain a complete and unbroken bone sample. Only adequate bone histologic procedures allow the measurement of reliable parameters.

501

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

502

CHAVASSIEUX, ARLOT, AND MEUNIER

FIGURE 1

Section of undecalcified transiliac biopsy in a case of trabecular osteoporosis with a marked atrophy of spongy bone. Stain, solochrome cyanin R. Original magnification, 10.

A. Bone Biopsy 1. CLINICAL PROCEDURE Rib biopsy, initially used by Epker and Frost [2], has now been abandoned in human adults because it necessitates orthopedic surgery and provides little spongy bone. Iliac biopsy is universally used because it does not require an extensive surgical protocol, provides large areas of spongy bone, and is safe. The bone biopsy may be taken either by a vertical approach as described by Sacker and Nordin [3] or by a horizontal transiliac method as described by Bordier et al. [4]. This latter method gives a full-thickness specimen with cortex at both ends which is a representative volume of the iliac bone (Fig. 1). The preferred site is situated 2 cm behind the anterosuperior iliac spine and 2 cm below the summit of the crest. The procedure requires local anesthesia in patients having received a sedative premedication. Local anesthetic should be applied to the skin and underlying soft tissue as well as the outer and inner periosteum, and a 1- to 2-cm-long skin and muscle incision should be performed. A trephine with an internal diameter of 7.5 mm should be used to give an adequate core for histomorphometry. Jowsey [5] showed that a trephine having a 5-mm inner diameter was responsible for a large sampling variation. Faugere and Malluche [6] compared Jamshidi needle biopsies with larger cores and concluded that larger biopsies are necessary for adequate evaluation of bone turnover and mineralization rate. The trephine teeth must be perfectly sharpened to avoid compression artifacts and fracture of the specimen that may markedly compromise the histomorphometry. Furthermore, the risk of having a poor sample is doubled when the operator is inexperienced [7].

In a multicenter survey on the side effects of transiliac bone biopsy, an overall incidence of complications of 0.52% was reported [7]. They were hematomas (0.24%), pain (0.11%), femoral neuropathy (0.09%), and skin infection (0.07%). Double tetracycline labeling involves the administration of two short courses of tetracycline which is deposited along the calcification front in bone as two distinct lines visualized on bone sections under fluorescent light [1] (Fig. 2). A typical labeling procedure is 10 mg/kg/day demethylchlortetracycline or tetracycline hydrochloride orally for 2 days, 12 days off, 4 days on. The bone biopsy is taken 4 – 6 days later. This technique provides very useful information on bone mineralization. 2. PREPARATION

OF

UNDECALCIFIED SECTIONS

Bone samples must be processed without prior decalcification. Fixation has to be immediate and may be accomplished in 70% ethanol, in absolute methanol, or in 10% phosphate-buffered formalin (pH 7). However, alcoholic fixatives are recommended to preserve the tetracycline labels. After dehydration in graded concentrations of alcohol, bone specimens are embedded in plastic monomers which must infiltrate and not only surround the sample. Embedding material must be as hard as the bone tissue to avoid vibration and fractures during sectioning. Methyl methacrylate is the most convenient plastic. Nonconsecutive serial sections, 5 to 20 m thick, are cut using special microtomes equipped with tungsten carbide-edged knives (e.g., Reichert Jung Polycut) or with diamond or glass knives. Before staining, some laboratories dissolve the plastic. This technique has some advantage for the

503

CHAPTER 63 Clinical Use of Bone Biopsy

FIGURE 2

Tetracycline double label. The osteoblasts (OB) lining the osteoid seam are visible. Unstained undecalcified bone section under fluorescent light. Original magnification,  250.

observation of cellular detail but it usually creates artifacts like folds and alters the skeletal architecture. The ideal stain should permit the distinction between calcified matrix and osteoid and allow easy identification of osteoblasts and osteoclasts from hematopoietic cells. Solochrome cyanin R, in 1% aqueous solution, is a reliable stain, giving an unequivocal identification of osteoid, calcified matrix, and bone cells. It is also convenient for the measurement of bone volume by using automatic image analyzers because it clearly differentiates trabeculae from the marrow spaces [8]. Goldner’s trichrome gives a good discrimination of osteoclasts and of resorption cavities, but underestimates the amount of osteoid tissue [8,9]. Toluidine blue, which is widely used, clearly stains osteoblasts. It has been reported also to identify the calcification front as a granular metachromatic dark line at the osteoid – calcified bone interface [10]. However, the reproducibility of this method is poor and analysis of tetracycline-labeled surfaces is preferred.

B. Bone Histomorphometry 1. MEASURING METHODS These methods are based on Delesse’s principle, which permits deduction of three-dimensional numerical information from two-dimensional images [11]. Measurements of histomorphometric parameters may be performed with either the manual point-counting method or image-analyzing computers. Because of the intersection

variation, several nonconsecutive sections must be measured. The point-counting method consists of projecting an integrating eyepiece over the slide. The percentage of hits overlying the structure or the number of intersections between the lines and the bone perimeter gives an accurate estimation of volume or surface of the analyzed structure [12]. However, this method is tedious and time-consuming and is almost abandoned. More advanced techniques use semiautomatic or automatic image analyzers. Semiautomatic systems are composed of a microscope equipped with a drawing tube, a digitizing tablet with a cursor. The image of the luminous cursor is projected through the drawing tube onto the microscope field. The observer traces the structure to be measured by moving the cursor on the digitizing tablet. The computer integrates the x and y coordinates of each point of the tablet and gives the results according to the previously selected program package [13,14]. Automatic methods consist of a TV camera that projects the field of the microscope on a screen. In the earliest systems, the image was analyzed by a computer according to its shading of gray. The observer had only to determine the shade of grey of the structure [15]. These analyzers were used only for the measurement of bone volume. More recently, measurement of remodeling surfaces could be performed with an automatic color analyzer (Visiolab, Biocom, France). This system is equipped with specific software for bone histomorphometry. A 3 CCD color camera captures the image of the analyzed structure to obtain a digital image that is recorded in the computer. Then, the image is displayed on the high-definition video monitor [16].

504 2. HISTOMORPHOMETRIC PARAMETERS In 1987, a committee from the American Society for Bone and Mineral Research proposed a standardized nomenclature of histomorphometric parameters [17]. The abbreviations of parameters listed here are expressed in three dimensions. a. Parameters of Bone Structure PARAMETERS EXPRESSING THE AMOUNT OF BONE (SEE ALSO CHAPTER 35) Cancellous bone volume (Cn-BV/TV, %) is the percentage of spongy bone tissue including mineralized bone and osteoid. The mineralized bone volume (Cn-Md.V/TV) is expressed as the percentage of mineralized bone tissue, i.e., bone volume-osteoid volume. Total bone volume (BV/Tt.CV, %) represents the amount of the total bone including cortices and spongy bone in percentage of the core volume. It is an estimate of bone mass and may be compared with the bone mineral density measured by noninvasive methods [18]. Cortical width (Ct.Wi, mm) is the mean thickness of cortices and is expressed in millimeters. Wall thickness (W.Th, mm) represents the amount of bone formed in any forming unit at the intermediary level of organization of bone and is the width of a complete cancellous bone. PARAMETERS REFLECTING THE MICROARCHITECTURE OF SPONGY BONE (SEE ALSO CHAPTER 35) Besides the bone density, the microarchitecture of spongy bone is an essential determinant of the bone quality [19]. Parameters reflecting the three-dimensional trabecular network have been recently defined. Trabecular thickness (Tb.Th, mm), trabecular separation (Tb.Sp, mm), and trabecular number (Tb.N, /mm) represent the width of trabeculae, the distance between trabeculae, and the density of trabeculae, respectively. They are calculated from the cancellous bone area and volume and the length of the bone marrow interface according to Parfitt’s formulae [17]. Study of trabecular connectivity consists of the count of nodes (N.Nd), i.e., branch points, and termini (N.Tm), i.e., free-ends, and the measurement of the distance between nodes (Nd.Nd), between termini (Tm.Tm), or between nodes and termini (Nd.Tm). These measurements are performed after skeletonization of the trabecular network by either a manual [20] or an automatic [21] method. The ratio of nodes to termini (N.Nd/N.Tm) is an index of spatial connectivity [22]. Star volume in spongy bone represents a three-dimensional estimation of the mean size of the medullary space and should reflect its connectivity [23]. It is calculated for individual points in the marrow and represents the total volume which can be reached from each point in a straight line

CHAVASSIEUX, ARLOT, AND MEUNIER

without interruption by trabeculae or subcortical perimeter. Star volume is not routinely measured. Euler number is a term based on the number of holes and structural components that are linked. It is a direct measurement of the connectivity. Derived from this method, the ConnEulor may evaluate three-dimensional values from two-dimensional measurements performed on sections 10 to 40 m apart and named dissector [24]. Fractal geometry describes a point according to its spatial position and the complexity of its structure [25,26]. It may be correlated with the trabecular number and separation. b. Bone Formation Parameters STATIC PARAMETERS Osteoid volume (OV/BV, %) represents the fraction of cancellous bone tissue which is not calcified. Osteoid surface (OS/BS, %) is the percentage of total cancellous surface covered with osteoid. Osteoid thickness (O.Th, m) is the average width of osteoid seams. DYNAMIC PARAMETERS Mineral apposition rate (MAR, m/day) is determined by measuring the mean distance between the midpoints of two tetracycline labels and expresses the rate of progression of the mineralization front [1]. Mineralizing surfaces represent the extent of double (dLS/BS, %) or single (sLS/BS, %) or total (LS/BS, %) tetracycline-labeled surfaces, expressed as percentage of total cancellous bone surfaces. DERIVED PARAMETERS Bone formation rate (Cn-BFR/BS, m3/m2/day) represents the amount of mineralized bone made per unit of cancellous surface per day. It is calculated as the product of MAR and mineralizing surfaces (LS or dLS or 21 sLS  dLS). Adjusted apposition rate (Aj.AR, m/day), calculated as BFR/OS, represents the amount of mineralized bone made per day per unit of osteoid-covered surface. Formation period (FP, days) is the mean time required to rebuild a new bone structural unit and is given by W.Th/Aj.AR. Mineralization lag time (Mlt, days) is the mean interval between the deposition of osteoid and its subsequent mineralization. It is calculated by O.Th/Aj.Ar. c. Bone Resorption Parameters Eroded surface (ES/BS, %) is the percentage of cancellous bone surface eroded, i.e., where osteoclastic resorption is continuing or where resorption has ceased but where the osteoblasts have not yet started to refill the Howship’s lacunae. Osteoclast number is expressed either per millimeter of cancellous surface (N.Oc/BS) or per mm2 of bone section.

505

CHAPTER 63 Clinical Use of Bone Biopsy

Erosion depth (E.De, m) can be derived either from the number of eroded lamellae in each resorption cavity multiplied by the lamellar thickness [27] or after the rebuilding of this cavity on image analyzer [16,28]. d. Activation Frequency Activation frequency represents the probability that a new cycle of remodeling will be initiated at any point on the bone surface. It is calculated as Cn-BFR/W.Th and expressed per day.

Measurement of biochemical markers of bone turnover may reflect whole body bone formation and resorption activities [37], but histomorphometry is the only tool for assessing bone turnover at the trabecular level and for differentiating the cell and tissue levels of remodeling activity.

A. Utility of Bone Biopsy for the Diagnosis of Osteoporosis and Other Metabolic Bone Diseases 1. OSTEOPOROSIS

3. REPRODUCIBILITY OF BONE HISTOMORPHOMETRY a. Inter- and Intraobserver Variations Whatever the method used, inter- and intraobserver variations are low for all parameters [29] except for cortical width. Its measurement is subject to a large interobserver variation because the limit between cortical and cancellous bone is difficult to identify [30]. b. Intermethod Variation A highly significant correlation between manual and computerized methods has been reported [29,31,32]. Computerized methods are generally less time-consuming for all parameters. The automatic method is preferable for Cn-BV/TV and parameters of bone microarchitecture because it is the fastest one. c. Intersample Variation The iliac crest being not perfectly isotropic, the architectural organization of trabeculae explains the variability between sites within the ilium [33]. Repeat biopsies are often performed to evaluate therapeutic effects in individuals suffering from metabolic bone disease. Thus, this variation should be taken into account to affirm that observed changes are due to the treatment. In osteoporotic patients, the confidence interval for Cn-BV/TV is 29% in 1 individual, 10% for a group of 10 patients, and 7% for a group of 20 subjects [34]. Thus the value of CnBV/TV on a second biopsy must be out of that range to conclude that the second biopsy significantly differs, in an individual or in a group. This difference is lower in osteoporotic patients than in controls [31] and is increased if the two biopsies are more than 1.5 cm apart [35,36].

III. INDICATIONS FOR BONE BIOPSY Bone histomorphometry is the only procedure which allows assessment of bone turnover and bone cell activity. Ten years ago, it was used as a routine technique for the diagnosis of osteoporosis and other metabolic bone diseases, but bone mass is now measured by noninvasive methods, such as dualenergy X-ray absorptiometry. The latter do not, however, provide information on bone quality or microarchitecture.

Anatomically, osteoporosis is characterized by a decrease in bone volume and microarchitectural alterations which cause fragility fractures. The amount of bone may be assessed by photon absorptiometry, but bone biopsy provides data about the mechanism of bone loss and changes in bone turnover [38]. Age-related changes in bone are characterized by trabecular thinning associated with a loss of connectivity of trabeculae [39 – 42]. This decrease in trabecular thickness is consistent with the reduction in the amount of bone formed in each remodeling unit. The spacing between bone structures increases and the risk of trabecular perforation is augmented when the number or the activity of osteoclasts is increased [43]. This is the case after menopause when the activation frequency is markedly increased by estrogen deprivation. Histological heterogeneity in bone turnover has been reported in osteoporotic patients [44]. Bone loss results from an unbalanced coupling between formation and resorption, but some patients have increased resorption alone, some have increased resorption and formation, and some have reduced bone formation [44 – 47]. The distinction of low, high, or normal turnover is interesting for the selection of the best therapeutic agent to be used. Patients with high turnover should receive drugs that reduce the differentiation of osteoclasts or the duration of osteoclast activity without depressing osteoblast activity (e.g., calcitonin or bisphosphonates). When osteoporosis results from a defect of bone formation, treatment must increase the number and/or the activity of osteoblasts, as fluoride salts. Differences in bone turnover can be assessed by specific and sensitive biochemical markers of bone remodeling (e.g., serum alkaline phosphatase or osteocalcin for bone formation, urinary pyridinoline or deoxypyridinoline for bone resorption) [37]; therefore, the clinical utility of a bone biopsy for a patient with a recent fragility fracture is questionable. At present, the indication should be restricted to patients for whom history, signs, X-rays, or biochemical profile suggest the possibility of osteomalacia, mast cell bone disease, nonsecreting myeloma, sarcoidosis, or other rare conditions, or to patients with osteoporosis that is particularly serious or that is nonresponsive to usual therapies.

506

CHAVASSIEUX, ARLOT, AND MEUNIER

FIGURE 3

Undecalcified bone section in a case of osteomalacia characterized by (A) abnormal thick osteoid seams (Ost) (stain, solochrome cyanin R, original magnification, 125). (B) Only single tetracycline labels (arrows) are present, indicating a decreased mineralization rate (unstained section, original magnification, 250).

2. OSTEOMALACIA This disorder is characterized by the presence of abnormally thick osteoid seams, decreased mineralization rate, and prolonged mineralization lag time (Fig. 3). Bone biopsy is not required for the diagnosis of osteomalacia in patients with typical biochemical and radiological signs, but it is very useful in borderline cases. The use of these dynamic criteria distinguishes osteomalacia from other conditions in which there may be increased osteoid surfaces with a normal mineralization rate. Osteoid accumulation in such cases results from an increased number of apposition sites due to an enhanced activation frequency of new basic multicellular units, as in primary or secondary hyperparathyroidism and thyrotoxicosis. Bone biopsy is particularly indicated when osteomalacia occurs as a complication of other disorders, such as complete lack of sunlight exposure, previous gastric surgery, coeliac disease, Crohn’s disease, or long-term parenteral nutrition [48]. Osteomalacia can also complicate chronic renal failure after dialysis and is due to the disturbances of vitamin D and calcium metabolism or to the effects of aluminum toxicity. Aluminum deposits at the calcification front inhibit mineralization [49]. However, in hemodialysis patients with aplastic bone disease, aluminum induces a mineralization defect with osteoid seams of normal thickness [50]. Aluminum is identified on histological bone sections stained with aurine tricarboxylic acid. It appears as a deep pink to red line on mineralized bone surfaces.

3. SYSTEMIC MAST CELL DISEASE This disorder is associated with either osteoporosis or osteosclerosis in the presence or absence of skin lesions. A study performed in 21 male patients has shown unbalanced coupling between resorption and formation in favor of resorption without any mineralization defect [51]. The diagnosis is confirmed by the count of mast cells. Only the observation of histological sections stained for the identification of mast cells (May – Grünwald – Giemsa, acridine orange, or toluidine blue at pH 2.6) allows this diagnosis (Fig. 4).

B. Bone Histomorphometry to Assess Bone Quality in Osteoporosis The amount of bone is not the only determinant of mechanical strength of bone. The quality of bone, i.e., the texture of bone matrix (lamellar or woven bone), the microarchitecture of the trabecular network, and the presence of mineralization defects are also major factors. Bone histomorphometry is the only tool to assess these parameters of bone quality. Recently, bone biopsy has been used to evaluate the degree of mineralization of bone structural units by microradiography. In adult bone, the mean degree of mineralization depends on the rate of turnover. A preliminary study performed in mature female baboons demonstrated a strong

507

CHAPTER 63 Clinical Use of Bone Biopsy

FIGURE 4 Undecalcified bone section in a case of mast cell disease showing the presence of numerous mast cells (arrows). OC, osteoclast. Stain: toluidine blue, pH 2.6.  400.

hypomineralization of bone from the ovariectomized baboon due to the increased remodeling activity in response to hormonal defect. In contrast, the degree of mineralization from ovariectomized baboons treated for 2 years with alendronate was similar to that measured in a control animal [52]. In postmenopausal women treated for 2 and 3 years with 10 mg/day of alendronate, the mean degree of mineralization of bone was increased when compared to the placebo group [53]. These observations suggest that a substantial part of the increase in BMD induced by alendronate may be explained by the increase in the mean degree of mineralization and not by an increase in bone mass reflected by cancellous bone volume. The latter parameter was not significantly increased in osteoporotic patients after 2 or 3 years of alendronate [54]. The results indicate that the increased degree of matrix mineralization was due to an alendronate-related reduction in bone turnover followed by a prolonged secondary mineralization period, the net effect of which was an increase in bone strength and a reduction in fracture rate.

C. Bone Histomorphometry for Evaluating the Effects of Treatments Bone biopsy is not always required to assess treatment efficacy, since measurement of bone mass by noninvasive methods is generally sufficient. However, when efficacy of treatment is unknown, bone histomorphometry is the most

valuable tool for explaining the nonresponse to the drug. Furthermore, in osteoporotic patients treated with fluoride salts, bone biopsy allows the measurement of bone fluoride content [55] which reflects the cumulative dose of fluoride ion taken up in bone and patient compliance with therapy.

D. Bone Histomorphometry as a Research Tool Bone biopsies are required to evaluate the effects on bone remodeling of any new therapeutic agent in both human trials and experimental studies. Such assessment of bone quality is now required by the Food and Drug Administration and other agencies for any long-term clinical trials. The effects on bone quality and remodeling have been assessed in 231 postmenopausal women with osteoporosis treated with oral alendronate (5, 10, or 20 mg/day) after 2 and 3 years of treatment [54]. Mineral apposition rate was unaffected by treatment. Osteoid thickness, volume and surface significantly decreased. Whatever the dose, mineralizing surface and activation frequency markedly decreased at each time point (92% after 2 years and  96% after 3 years for 10 mg daily dose). These findings have also shown that alendronate had no adverse effect on bone structure or mineralization and that alendronate dose-dependently and markedly decreased the rate of bone turnover. As mentioned before, the observed increases in BMD reported in alendronate-treated patients could result from an augmentation of the mean degree of mineralization of bone resulting from the low bone turnover. Similar

508

CHAVASSIEUX, ARLOT, AND MEUNIER

effects of alendronate on bone turnover have also been observed in patients with corticosteroid-induced osteoporosis after 1 year of treatment [56].

IV. CONCLUSIONS Bone biopsy was first developed to evaluate the amount of bone and bone turnover in osteoporotic patients. This information has now been supplanted by dual-energy X-ray absorptiometry and measurements of biochemical markers. The main diagnosic utility for biopsy is now to exclude osteomalacia and mastocytosis. However, it is the only method to assess the intermediary level of organization of bone and particularly to evaluate static and dynamic aspects of osteoblast function. Biopsy is an essential tool for assessing the pathophysiology of bone loss, the mechanisms of action of any treatment, and the safety of new therapeutic agents evaluated by assessment of the quality of bone after long-term treatment.

References 1. H. M. Frost, Tetracycline based histological analysis of bone remodeling. Calcif. Tissue Res. 3, 211 – 237 (1969). 2. B. N. Epker and H. M. Frost, A histological study of remodeling at the periosteal, haversian canal. Cortical endosteal and trabecular endosteal surfaces in human ribs. Anat. Rec. 152, 129 – 135 (1965). 3. L. S. Sacker and B. E. C. Nordin, A sample bone biopsy needle. Lancet 1, 347 (1954). 4. P. Bordier, H. Matrajt, L. Miravet, and D. Hioco, Mesure histologique de la masse et de la resorption des travées osseuses. Pathol. Biol. 12, 1238 – 1243 (1964). 5. J. Jowsey, “The Bone Biopsy.” Plenum Medical Book, New York/ London, 1977. 6. M. C. Faugere and H. H. Malluche, Comparison of different bone biopsy techniques for qualitative and quantitative diagnosis of metabolic bone diseases. J. Bone Jt. Surg. 65a, 1314 – 1319 (1983). 7. H. Duncan, S. D. Rao, and A. M. Parfitt, Complications of bone biopsies. In “Bone Histomorphometry 1980” (W. S. S. Jee and A. M. Parfitt, eds.), pp. 483 – 486. Armour Montagu, Paris, 1981. 8. P. J. Meunier, C. Edouard, P. Courpron, and F. Toussaint, Morphometric analysis of osteoid in iliac trabecular bone. In “Vitamin D and Problems Related to Uremic Bone Disease” (A. W. Norman, K. Schaefer, H. G. Grigoleit, D. V. Herrath, and E. Ritz, eds.), pp. 149 – 155. Walter de Gruyter, Berlin/New York, 1975. 9. P. Lips, “Metabolic Causes and Prevention of Femoral Neck Fractures,” Ph.D. thesis. Free University of Amsterdam, Rodopi, Amsterdam, 1982. 10. P. J. Bordier and S. Tun Chot, Quantitative histology of metabolic bone diseases. Clin. Endocrinol. Metab. 1, 197 – 215 (1972). 11. M. Delesse, Procédé mécanique pour determiner la composition des roches. C. R. Acad. Sci. 25, 544 – 552 (1847). 12. R. K. Schenk, W. A. Merz, and J. A. Muller, A quantitative histological study of bone resorption in human cancellous bone. Acta Anat. 74, 11 – 53 (1969). 13. H. H. Malluche, D. Sherman, W. Meyer, and S. G. Massry, A new semi-automatic method for quantitative static and dynamic bone histology. Calcif. Tissue Res. 34, 439 – 448 (1982).

14. R. C. Manaka and H. H. Malluche, A program package of the quantitative analysis of histologic structure and remodeling dynamics of bone. Comp. Prog. Biomed. 13, 191 – 201 (1981). 15. P. J. Meunier, Use of an image analysing computer for bone morphometry. In “Clinical Aspects of Metabolic Bone Disease” (A. W. Norman, K. Schaefer, H. G. Grigoleit, D. V. Herrath, and E. Ritz, eds.), pp. 148 – 151. Excerpta Medica, Amsterdam, 1973. 16. J. P. Roux, M. E. Arlot, E. Gineyts, P. J. Meunier, and P. D. Delmas, Automatic — interactive measurement of resorption cavities in transiliac bone biopsies and correlation with deoxypyridinoline. Bone 17, 153 – 156 (1995). 17. A. M. Parfitt, M. K. Drezner, F. H. Glorieux, J. A. Kanis, H. Malluche, P. J. Meunier, S. M. Ott, and R. R. Recker, Bone histomorphometry: Standardization of nomenclature, symbols and units. J. Bone Miner. Res. 6, 595 – 610 (1987). 18. P. D. Delmas, E. Fontanges, F. Duboeuf, G. Boivin, P. Chavassieux, and P. J. Meunier, Comparison of bone mass measured by histomorphometry on iliac biopsy and by dual photon absorptiometry of the lumbar spine. Bone 9, 209 – 213 (1988). 19. A. M. Parfitt, Implications of architecture for the pathogenesis and prevention of vertebral fracture. Bone 13, S41 – S47 (1992). 20. R. W. E. Mellish, M. W. Ferguson-Pell, G. V. B. Cochran, R. Lindsay, and D. W. Dempster, A new manual method for assessing two-dimension cancellous bone structure: Comparison between iliac crest and lumbar vertebra. J. Bone Miner. Res. 6, 689 – 696 (1991). 21. N. J. Garranhan, R. W. E. Mellish, S. Vedi, and J. E. Compston, A new method for the two dimensional analysis of bone structure in human iliac crest biopsies. J. Microsc. 142, 341 – 349 (1986). 22. J. E. Compston, R. Mellish, P. Groucher, and R. Newcombe, Structural mechanisms of trabecular bone loss in man. Bone Miner. 6, 339 – 350 (1989). 23. A. Vesterby, H. J. G. Gundersen, and F. Melsen, Star volume of marrow space and trabeculae of the first lumbar vertebra: Sampling effi ciency and biological variation. Bone 10, 7 – 13 (1989). 24. H. J. G. Gundersen, R. W. Boyce, J. R. Nyengaard, and A. Odgaard, The Conneulor: Unbiased estimation of connectivity using physical disectors under projection. Bone 14, 217 – 222 (1993). 25. E. Lespessailles, C. L. Benhamou, D. Tourliere, T. Loussot, G. Jacquet, R. Harba, and R. Jennane, Fractal evaluation of trabecular bone microarchitecture of calcaneus: Study of reproductibility. J. Bone Miner. Res. 8 (Suppl. 1), S264 (1993). 26. R. S. Weinstein, S. Majumdar, and H. K. Genant, Fractal geometry applied to the architecture of cancellous bone biopsy specimens. Bone 13, A38 (1992). 27. E. F. Eriksen, L. Mosekilde, and F. Melsen, Trabecular bone resorption depth decreases with age: Differences between normal males and females. Bone 6, 141 – 146 (1985). 28. N. J. Garrahan, P. I. Croucher, and J. E. Compston, A computerized technique for the quantitative assessment of resorption cavities in trabecular bone. Bone 11, 241 – 245 (1990). 29. P. M. Chavassieux, M. A. Arlot, and P. J. Meunier, Intermethod variation in bone histomorphometry: Comparison between manual and computerized methods applied to iliac bone biopsies. Bone 6, 221 – 229 (1985). 30. J. E. Compston, S. Vedi, and A. J. Stellon, Inter-observer and intraobserver variation in bone histomorphometry. Calcif. Tissue Int. 38, 67 – 70 (1986). 31. M. C. de Vernejoul, D. Kuntz, L. Miravet, D. Goutallier, and A. Ryckewaert, Bone histomorphometry reproductibility in normal patients. Calcif. Tissue Res. 33, 369 – 374 (1981). 32. H. H. Malluche, D. Sherman, R. Manaka, and S. G. Massry, Comparison between different histomorphometric methods. In “Bone Histomorphometry 1980” (W. S. S. Jee and A. M. Parfitt, eds.), pp. 449 – 451, Armour Montagu, Paris, 1981. 33. W. J. Whitehouse, The quantitative morphology of anisotropic trabecular bone. J. Microsc. 101, 153 – 168 (1974).

509

CHAPTER 63 Clinical Use of Bone Biopsy 34. P. M. Chavassieux, M. A. Arlot, and P. J. Meunier, Intersample variation in bone histomorphometry: Comparison between parameter values measured on two contiguous transiliac bone biopsies. Calcif. Tissue Int. 37, 345 – 350 (1985). 35. C. Bergot, A. M. Laval-Jeantet, and M. Laval-Jeantet, Etude critique des variations de la mesure de la masse osseuse par biopsie iliaque. Rev. Rhum. 45, 317 – 324 (1978). 36. F. Melsen, B. Melsen, and L. Mosekilde, An evaluation of the quantitative parameters applied in bone histology. Acta Pathol. Microbiol. Scand. A 86, 63 – 69 (1978). 37. P. D. Delmas, Biochemical markers of bone turnover in osteoporosis. In “Osteoporosis: Etiology, Diagnosis and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 297 – 316. Raven Press, New York, 1988. 38. P. J. Meunier, Assessment of bone turnover by histomorphometry in osteoporosis. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 317 – 332. Raven Press, New York, 1988. 39. J. E. Aaron, N. B. Makins, and K. Sagreiya, The microanatomy of trabecular bone loss in normal aging men and women. Clin. Orthop. Rel. Res. 215, 260 – 271 (1987). 40. R. W. E. Mellish, N. J. Garrahan, and J. E. Compston, Age-related changes in trabecular width and spacing in human iliac crest biopsies. Bone Miner. 6, 331 – 338 (1989). 41. A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationships between surface volume and thickness of iliac trabecular bone in aging and in osteoporosis. J. Clin. Invest. 72, 1396 – 1409 (1983). 42. R. S. Weinstein and M. S. Huston, Decreased trabecular width and increased trabecular spacing contribute to bone loss with aging. Bone 8, 137 – 142 (1987). 43. P. Chavassieux and P. J. Meunier, Mecanismes de la perte osseuse dans les osteoporoses. Rev. Prat. 45, 1083 – 1088 (1995). 44. M. P. Whyte, M. A. Bergfeld, W. A. Murphy, L. V. Avioli, and S. L. Teitelbaum, Postmenopausal osteoporosis: A heterogeneous disorder as assessed by histomorphometric analysis of iliac crest bone from untreated patients. Am. J. Med. 72, 193 – 201 (1982). 45. M. E. Arlot, P. D. Delmas, D. Chappard, and P. J. Meunier, Trabecular and endocortical bone remodeling in postmenopausal osteoporosis: Comparison with normal postmenopausal women. Osteoporosis Int. 1, 41 – 49 (1990). 46. P. J. Meunier, S. Sellami, D. Briancon, and C. Edouard, Histological heterogeneity of apparently idiopathic osteoporosis. In “Osteoporosis:

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Recent Advances in Pathogenesis and Treatment” (H. F. de Luca, H. Frost, W. Jee, C. Johnston, and A. M. Parfitt, eds.), pp. 293 – 301. University Park Press, Baltimore, 1981. S. L. Teitelbaum, M. A. Bergfeld, L. V. Avioli, and M. P. Whyte, Failure of routine biochemical studies to predict the histological heterogeneity of untreated postmenopausal osteoporosis. In “Osteoporosis: Recent Advances in Pathogenesis and Treatment” (H. F. de Luca, H. Frost, W. Jee, C. Johnston, and A. M. Parfitt, eds.), pp. 303 – 309. University Park Press, Baltimore, 1981. S. M. Ott, N. M. Maloney, and G. L. Klein, Aluminum is associated with low bone formation in patients receiving chronic parenteral nutrition. Ann. Int. Med. 98, 910 – 914 (1983). A. B. Hodsman, D. J. Sherrard, A. C. Alfrey, S. Ott, A. S. Brickman, N. L. Miller, N. A. Maloney, and J. W. Coburn, Bone aluminum and histomorphometric features of renal osteodystrophy. J. Clin. Endocrinol. Metab. 54, 539 – 546 (1982). M. Parisien, S. A. Charhon, M. Arlot, E. Mainetti, P. Chavassieux, M. C. Chapuy, and P. J. Meunier, Evidence for a toxic effect of aluminium on osteoblasts: A histomorphometric study in hemodialysis patients with aplastic bone disease. J. Bone Miner. Res. 3, 259 – 267 (1988). E. Meys, J. P. Roux, A. Buffet, C. Edouard, and P. J. Meunier, Bone involvement in systemic mast cell disease. J. Bone Miner. Res. 7(Suppl. 1), S298 (1992). P. J. Meunier and G. Boivin, Bone mineral density reflects bone mass but also the degree of mineralization of bone: Therapeutic implications. Bone 21, 373 – 377 (1997). G.Boivin, K. Klaushofer, P. Roschger, S. Rinnerthaler, P. Fratzl, P. Chavassieux, A. C. Santora, A. J. Yates, and P. J. Meunier, Alendronate increases the mean degree of mineralization of bone and the uniformity of mineralization of bone in osteoporotic women. Bone 23, 5 (Suppl.), F282 (1998). P. Chavassieux, M. A. Arlot, C. Reda, L. Wei, A. J. Yates, and P. J. Meunier, Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. J. Clin. Invest. 100, 1475 – 1480 (1997). P. J. Meunier and G. Boivin, Fluoride salts for vertebral osteoporosis: The benefit-to-risk ratio depends on the cumulative dose reaching bone. Osteoporosis Int. 3 (Suppl. 1), 211 – 214 (1993). P. J. Meunier and P. Chavassieux, Bone histomorphometry in untreated and treated corticosteroid-induced osteoporosis. Osteoporosis. Int. 9 (Suppl.1) S3 (1999).

CHAPTER 64

Design Considerations for Clinical Investigations of Osteoporosis ROBERT P. HEANEY

I. II. III. IV.

Creighton University, Omaha, Nebraska 68178

V. Design Alternatives VI. Design Issues of Special Relevance to Investigation of Bone VII. Burden of Proof and the Null Hypothesis References

Background: Inference from Phenomena Subject to Variation Types of Investigational Designs Principal Types of Bias with Various Designs When Controlled Trials and Observational Studies Disagree

I. BACKGROUND: INFERENCE FROM PHENOMENA SUBJECT TO VARIATION

Thus, when we perform an investigation and find, for example, a difference between treatment groups, we must decide whether that difference represents an effect of the drug we have investigated or is, instead, simply an expression of the ubiquitous underlying differences between patients — something we model under the name “chance.” This inherent, biological variability would create less of a problem with true breakthrough developments, which tend to be easily detectable. However, these are rare in medicine and certainly so in the clinical investigation of osteoporosis. As noted elsewhere in this volume, osteoporosis is a decidedly multifactorial disorder. Among other things, this means that etiologic studies of unselected individuals will find only small effects for any given putative cause, since at a population level there will be many causes operating with varying degrees of intensity in different individuals. This means also that preventive measures, inescapably

It is a commonplace of both clinical and investigative experience that things vary. No two individuals have quite the same bone size or density. Nor does bone mineral content respond to an intervention equally in every individual treated with a given agent. Moreover, the patients enrolled in an investigation of a new treatment for osteoporosis may differ, not only from one another, but from the larger group of patients the drug is intended to treat. Finally, those patients selected to receive a particular treatment (e.g., hormone replacement therapy) may be different from those not selected. (Indeed, one would hope the treatment choices by physicians and patients had a rational basis and were not decided by flips of a coin.)

OSTEOPOROSIS, SECOND EDITION VOLUME 2

513

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

514 operating on only one or two causal pathways, will also produce relatively small effects. Thus the differences found in investigations will often be of a magnitude not much larger than inherent biological variability. The purpose of formal investigation is to help us find the signal (if any) embedded in all that noise. Investigative noise is of two general kinds: random biological (and measurement) variability, and bias. The laws of probability help us deal with the first of these — the random variability component — by allowing us to state the likelihood that the difference we observed would have been produced solely by chance. In other words, if there were actually no difference between treatments, and if we randomly assigned subjects to the various treatments and performed the same investigation over again many times, the laws of probability tell us how often we would have observed, simply from the workings of chance, a difference as large as the one we actually found. If the only sources of variability in our investigation were random chance and the agent being tested, then a finding that chance would rarely produce a result as large as the one we found allows us to infer that the investigative agent probably is responsible. But often there are other factors at work. For example, if we assign certain patients to the active treatment because we judge that they are the ones more likely to respond, and others to the placebo because we judge that they are less likely to respond, then the difference we may find, while unlikely to have been produced by chance, will reflect prior differences between subjects rather than, or in addition to, true treatment effects. Indeed, it is actually rather silly to do any sort of statistical test in such a situation. Our goal in designing an investigation is to reduce, insofar as possible, the many sources of variability to just two, random chance and the effects (if any) of the active agent. However, in this case, we have deliberately let another source of variability enter into the picture, namely our judgement about treatment assignment. Whenever we apply the laws of random chance to situations in which, under the null hypothesis, other factors are operative, we are guilty of bias. Bias takes many forms, to be described below: treatment bias (as in the foregoing example), volunteer bias, medical center bias, ascertainment bias, admission rate bias, reclassification bias, and many, many more. In general, the purpose of investigative design is to eliminate, or reduce to the extent possible, all bias in our investigation. This brief overview is nothing more than elementary design theory. It is important to understand it, however, since it provides the basis for all investigative design and analysis. Further, as we look at the unique investigative challenges presented by osteoporosis, we need to understand what the various features of the available designs are intended to accomplish. All chronic diseases, and osteoporosis in particular, create problems which can vitiate

ROBERT P. HEANY

even the best designs. In trying to avoid one bias we can inadvertently create another which may, actually, be worse. In what follows I will examine the basic design options, point out some of the biases they are prone to, call attention to osteoporosis-specific problems, and suggest tentative compromises where the theoretically optimal design presents problems.

II. TYPES OF INVESTIGATIONAL DESIGNS Investigations can be classified into two large groups: those which seek to describe or characterize a population, and those which seek to test an hypothesis, termed by Mainland exploratory and explanatory, respectively [1]. Exploratory, or descriptive, investigation asks such questions as “What is the distribution of spine BMD values in 65-year-old Caucasian women?” “How common are vertebral compression deformities in Hispanic women over 60 years of age?” Often these are one-variable questions, as in the preceeding examples. (Or, if they contain more than one variable, no particular relationship is postulated among those variables). Explanatory, or analytic, investigation, by contrast, tests hypotheses; it looks for differences between groups or seeks causal relationships. It asks such questions as “Does this treatment slow age-related bone loss?” or “Does obesity protect against fracture?” Always at least two variables are involved, and a relationship (causal, associative, difference) is postulated between them. Explanatory, or analytic, investigations are divided into two main design types: true experiments, in which the investigator controls the assignment of the independent variable to the subjects being studied (the sampling units), and observational studies, in which no such control exists. Observational studies are further subdivided into case – control and cohort studies, distinguished by whether the subjects are sorted for analysis by outcomes (case – control) or by exposure to the independent variable (cohort). Cohort studies, in turn, are subdivided further on the basis of whether the exposure to the independent variable occurs concurrently with the investigation or prior to it. The first is termed a concurrent cohort study; the second, nonconcurrent. (These design types, as well as others, are discussed in more detail in Chapter 20). A summary of the principal design types, together with their associated problems, is presented in Table 1. This scheme is not the only way to classify investigations, but it 1 Classification schemes for investigations abound, and can be a source of confusion for the novice. Kelsey and Sowers (see Chapter 20) use the terms “Descriptive” and “Analytic” where Mainland uses “Exploratory and “Explanatory”.

CHAPTER 64 Design Considerations for Clinical Investigations

TABLE 1 Investigational Types and Some of the Biases/Problems to Which They Are Prone Investigational type

Potential problem/Bias

Exploratory/descriptive

Nonrandom sample of population of interest, usually because of selection or medical center bias

Case – control study

Admission rate bias

Concurrent cohort study

Admission rate bias; Ascertainment bias; Unequal interference (placebo effect); Lost sampling units

Nonconcurrent cohort study

Admission rate bias; Lost sampling units

Experiment (RCT)

Inadequate randomization; Unequal interference (placebo effect); Lost sampling units; Compression of response range.

is a particularly useful one for our purposes because each type has a different profile of biases. The section which follows develops these problems in modest detail. Chapter 20 and standard works on investigational design will provide additional information and examples [1 – 5]. Before proceeding, it is useful to recognize that these investigational types commonly arise out of the exigency of the situation; that is, they are not all equally available to an investigator attempting to answer a research question. While it is possible for an investigator to use a true experimental design to test the hypothesis that a particular treatment slows age-related bone loss, that is, by assigning the treatment to some subjects and not to others, that option would not be open for the hypothesis concerning obesity and fracture. An investigator cannot assign some people to be fat and others to be lean, and then follow both groups for several years. Instead, the researcher must either determine the prevalence of obesity in fracture patients and compare with obesity prevalence in a suitable nonfracture control group (a case – control study), or follow for several years groups of subjects who, altogether apart from input by the investigatior, happen to be fat and lean and than count fractures in each (a cohort study).

III. PRINCIPAL TYPES OF BIAS WITH VARIOUS DESIGNS A. Exploratory/Descriptive Studies The principal problem besetting much exploratory/descriptive investigation is sampling bias (or selection bias), that is, the tendency for the study itself (including such features as its organizational locus and its leaders) to influence who gets selected for study, i.e., who gets counted or measured. The result is that the subjects (and the findings) are not representative of the larger population about which we

515 presumably wanted to draw conclusions. Selection bias arises the moment we draw inferences from a nonrandom sample in a way that is valid only for samples drawn randomly from the population. It is important to bear in mind that, unless we study an entire population, the purpose of most exploratory investigation is not actually to describe those whom we study, but rather those whom we cannot ourselves measure or count. The sample exists simply to give us a window on a certain population. A biased sample distorts the view through that window. Sampling bias takes two principal forms in the investigation of disorders such as osteoporosis, volunteer bias and medical center bias. Volunteer bias reflects the fact that patients and controls who volunteer to be studied often differ in important ways from the larger population we seek to understand and characterize. Reasons might include concern about, or interest in, health issues or the specific disease being studied. An example is the finding by BarrettConnor [6] that women who take estrogen after menopause exhibit more health-promoting behaviors than women generally. However, often such explanations will not be obvious, even though operative. In any event the impact of these selection factors can never be accurately estimated. An example of an unexplained effect of volunteer bias is seen in the study of Stegman et al. [7], who set out to characterize the distribution of values for sonic velocity through bone in an aging population. The investigators had taken a community-based, stratified random sample in which to make the measurements. The associated, enhanced level of community awareness led unselected people to ask if they could have their bones measured too. The investigators complied, but segregated the data derived from these volunteers. They found that the volunteers had lower sonic velocity values than the true random sample. The difference was highly significant, statistically, but more importantly, it was large as well, and would have produced a serious distortion in the estimate of the population distribution had the volunteers been used to estimate the population value. Medical center bias is the distortion in disease prevalence, or expression of disease characteristics, which results because medical centers are at the top of a multilevel referral pyramid that shunts the more difficult or serious cases upward toward specialists in the diseases concerned. Our disease descriptions in textbook chapters are generally written by these same specialists. The irony is that they are usually the only ones with a sufficient patient experience to conduct the studies; yet precisely that experience is distorted by virtue of where they work and how patient referrals operate. Such distortions can be of many kinds, but usually there will be an overestimate of disease severity and an underestimate of disease prevalence (both occurring because the mild or asymptomatic cases will less commonly be seen in the referral centers, or will not be seen at all).

516 Examples include: (i) the finding by Cooper et al., when they estimated crush fracture prevalence from a populationbased sample, that only about 35% of cases had been sufficiently symptomatic to come to medical attention [8]; and (ii) the notorious overestimate of thyroid cancer prevalence by Lahey et al. [9], which led to an explosion of unnecessary and inappropriate thyroidectomies  40 years ago. These latter authors found that 10.2% of solitary thyroid nodules removed at surgery in their clinic were malignant, a figure judged sufficiently higher than the operative complication rate to justify an aggressive search for and removal of, solitary nodules. The authors ignored the fact that the Lahey Clinic was at the apex of a referral chain which, at every step in the evaluation process, would have concentrated cancers in the group referred upward. We now know that medical center bias in this instance exaggerated thyroid cancer prevalence in solitary nodules by more than a factor of 10. This example illustrates a common feature of all bias. Mainland describes bias as “mislabeling” [1]. That is what happened here. The Lahey study (basically not a planned study, but a review of nearly 2000 Clinic records) was appropriate enough so far as it went, and if it had been used only to evaluate the continued surgical approach to patients referred to the Clinic with solitary nodules, would have been unexceptionable. However, when the authors advocated a search for nodules among asymptomatic individuals, the authors can be said to have labeled their sample as “typical of all patients with single nodules.” Because this label was incorrect, their sample was biased. The only way to avoid selection bias is to take a random sample from the population we wish to characterize, as Stegman et al. did for sonic velocity (or, equivalently, to obtain a near total sample, as Cooper et al. did for crush fracture prevalence). But even here, common sense is needed. What population do we really want to know about? Public health or policy issues will commonly require population-based samples, but treatment-related studies often do not. For example, cost effectiveness studies of a menopausal bone density measurement, or of postmenopausal estrogen hormone replacement therapy (HRT), often start with the entire population concerned and then apply population-level estimates of response to the information or compliance with the therapy, and calculate the cost for each fracture averted. While seemingly reasonable, the result may have no applicability to any individual group of patients, who may have quite different motivators and compliance rates. One could argue also that the whole exercise is a little silly to begin with, since no apparatus exists in this country for saturation treatment of a population in the first place. So the assumptions behind the calculation do not apply to actual practice. By contrast, Lafferty and Fiske presented results of a meticulously performed, prospective study of HRT in an upper middle-class female population

ROBERT P. HEANY

with a strong personal relationship with their physician [10]. While the results may well not be generalizable to other cultural, ethnic, or socioeconomic groups, this study contains more applicable information about patients of this type than true random, population-based samples. (In the final analysis, physicians do not treat random samples of the total population.)

B. Case – Control and Cohort Studies A major source of bias in case – control and cohort studies involves the selection of a suitable contrast group (the so-called “control group”). Contrast groups represent the two (or more) values of the relevant variables: their members either were or were not exposed to the causative or therapeutic agent (the cohort design), or their members either did or did not develop the disease or other outcome concerned (the case – control design). Both study designs are commonly used to investigate causal factors, since the true experimental design is usually precluded in investigation of disease causality. If the two contrast groups are not derived from the general population, various factors may influence subject admission into study, giving rise to what has been termed the “admission rate bias”, or “Berkson’ bias” (after its discoverer). Briefly, factors related either to the putative cause, or to its presumed outcome, can influence who gets into the study, thereby distorting any differences which may be present — either obliterating real differences or creating spurious ones. The admission rate bias involves a double dose of selection bias, but with the filtering or selection being different for each of the contrast groups. This unequal selection is what distorts underlying similarities or differences between the groups. An example is a study by Kochersberger et al. [11], designed to test whether asymptomatic hyperparathyroidism in postmenopausal women carried any perceptible morbidity. The investigators chose to look at low trauma fractures in a group of patients with a diagnosis of hyperparathyroidism on the presumption that, if hyperparathyroidism weakened bone, fractures would be more frequent. More frequent than what? More frequent than in a similar group of women without hyperparathyroidism. The challenge, as always, was to find such a group. It is a notoriously difficult thing to do, fraught with many pitfalls and traps. In this instance the authors chose women of the same age and ethnic background admitted to their medical center for cholecystectomy. They obtained fracture history in both groups and they found significantly more fractures in the hyperparathyroid women, thereby seemingly supporting their hypothesis. How might the admission rate bias have operated here? Three plausible mechanisms come immediately to mind. Women with gallbladder disease tend to weigh more than

517

CHAPTER 64 Design Considerations for Clinical Investigations

patients without, and weight is positively associated with bone mass and negatively with fractures. Also, gallbladder disease is a recognized complication of ERT, and hence a group of postmenopausal cholecystectomy patients will contain a disproportionate number of women receiving estrogen, which also protects bone mass. Thus, for two reasons, the selected nonhyperparathyroid contrast group would be expected to have fewer fractures than the sought for group of “normal controls.” The authors, in fact, recognized these associations and attempted to control for obesity and ERT use with statistical methods. However, they were unable to deal with a third possibility, namely, that the fractures themselves may have led to blood tests which in turn uncovered asymptomatic hyperparathyroidism, a problem referred to as ascertainment bias. In other words, the hyperparathyroid contrast group itself may well have had a higher prevalence of fractures than hyperparathyroid patients generally. Thus the selection factors that brought these two groups of women into the study exaggerated (or created entirely) the observed difference in fracture rates. I select this example because it allows us to see concretely how the admission rate bias operates. But the bias would be there nonetheless, even if we were not able to discern the causal connections that, as we can see in this case, plausibly distort prevalence of fractures in both of our contrast groups. Whenever subjects in the contrast groups are not randomly selected from the general population, the admission rate bias operates, and it would not have mattered whether the investigators chose patients undergoing cholecystectomy, or refractions in the eye clinic, or simply visitors in the waiting room. (Recall the quite unexplained, low sonic velocities in Stegman’s volunteers [7].) We will deal with the problems of lost sampling units and of unequal interference in greater depth in the next section. However, it is useful to note here an important difference among the various design types in regard to these potential problems. In both the case – control and the nonconcurrent cohort designs, the investigative procedures follow the exposure of the subjects to the causative or treatment agent; hence the investigation cannot interfere with (or alter) the treatment effect. In the experiment and concurrent cohort designs, however, the study and the treatment proceed together, and the two may interact in unexpected ways. Therefore, unless a double-blind is used (not possible with the concurrent cohort design), the placebo effect is likely to be greater with the treated group than with the controls. (This is what is meant by unequal interference.) In brief, in any investigation in which the independent variable cannot be randomly assigned to the individuals being studied, there is always a possibility that one or more factors other than the one we are studying will have

produced the difference between groups that we might have found. The laws of probability (and their application in statistical tests) can help us determine the likelihood that random chance is behind the observed difference. However, those tests cannot help us with bias. That is why neither the case – control nor the cohort study design permits strong causal inference.

C. Experiments: The Randomized Controlled Trial By contrast, the randomized controlled trial (RCT) is called a strong design because it often permits a clear test of the effect of a causative or therapeutic agent. That follows from the fact that, when an RCT is appropriately planned and executed, only chance determines the allocation of treatments to the individuals being studied. We can estimate from probability theory how much difference chance alone might have produced, and if the difference we find is greater, then we can infer that the variable being tested is probably responsible. For this reason the RCT now constitutes the minimum acceptable evidence for approval of a new treatment agent by the Food and Drug Administration of the United States [12], as well as by the cognate agencies of many other governments. What is little appreciated is the fact that proper design and execution of an RCT often impose requirements which may be impossible to meet. As a consequence, imperfectly executed experiments (which is what we are left with) may actually be less strong than designs (such as a nonconcurrent cohort study) that are inherently weaker. Also, even when perfectly executed, RCTs may mitigate against finding real effects of importance in quite unanticipated ways — especially in disorders such as osteoporosis. It is important to examine these issues in some detail. Also the participants of these trials are often so atypical that the generalizability of the results is questionable. Proper execution of an RCT involves four features [1]: (i) use of contrast groups (new drug, untreated controls, and possibly one or more comparison treatments); (ii) randomization through every step of the investigation (from treatment assignment to reading slides or X-rays); (iii) equalization of interference by means of a double-blind; and (iv) zero loss of subjects. The first two, though often imperfectly applied, present no conceptual problem and will not be considered further. The last two are a widely ignored source of serious difficulty. 1. EQUALIZATION OF INTERFERENCE Experienced clinical investigators recognize that investigation itself has a powerful effect, both on human performance and on the course of clinical illness. This effect goes by the name “placebo effect.” It can be thought of as the

518 enabling of innate, but latent healing powers. It not only makes patients feel better; it produces appreciable substantial objective improvement. The placebo effect is generally considered to be substantially greater in the investigative context than in routine health care delivery [13 – 15]. It does not require actual administration of a placebo; instead, the very process of looking — of making observations during a trial, with all of the attendant procedures and precautions — invokes the effect and alters what we are trying to measure. It is a kind of clinical investigative analog of Heisenberg’s uncertainty principle. The double-blind does not eliminate this problem; rather it ensures that all of the contrast groups will receive an approximately equal dose of the placebo effect.2 2. CONSTRICTION OF THE RESPONSE RANGE A little appreciated feature of both the RCT and the concurrent cohort study is that the placebo effect may narrow the response range available to the active agent, even a highly potent one. This is a little appreciated weakness of the RCT, built into its very structure. The response range may be further constricted by cotherapy — which will usually be an integral part of any treatment regimen for osteoporosis. (Double-blind, placebo-controlled trials in patients with the crush fracture syndrome will always ensure, in both groups, adequate nutrition, especially calcium and vitamin D, some measure of physical therapy, and finally the socialization with other sufferers which is usually a part of studies of this sort. All of these cotherapies can be expected to produce benefit.) The combination of cotherapy and the placebo effect can so severely compress the response range that it may be very difficult to find a real, additional benefit from the treatment, even if present. A concrete example will help make this point. Assume that a treatment agent is capable of increasing bone mass in the spine to fully normal levels. Assume, also, that it is being studied in a double-blind, placebo-controlled RCT design, in a group of individuals with vertebral osteoporosis who, on entry, have a minimum of two compression fractures, plus low values for BMD. Assume also a basal fracture rate for this group, untreated, of 400 new fracture events per thousand patient years. One might presume that the response window extends from that 400 per thousand patient year figure all the way down to zero, but that would not be realistic. Even fully normal individuals do not have a zero fracture rate. Furthermore, prevalent vertebral deformities create abnormal weight-bearing forces which, even in a spine with normal BMD, would be expected to produce an increased fracture rate. Similarly, the loss of trabecular 2 One suspects that some investigators are not aware of the fact that all treatment groups get the placebo effect, that the group receiving the active agent gets a response that is a composite of the placebo effect and the effect (if any) of the test drug.

ROBERT P. HEANY

connectivity which is widely believed to be a factor in vertebral osteoporosis, would not be expected to be corrected by increasing the density of the bony elements remaining in the vertebral bodies at the start of study. So, even restoring density to normal would probably not restore full normal strength. Finally, the treatment agency does not produce an instantaneous effect: it takes time to build the spine back up to a normal density value. Thus, although no one can specify what the exact figure might be, it is clear that, in a 3 to 4-year study, the minimum achievable fracture rate will be substantially above zero. For purposes of this example, assume that it is 200 new fractures per thousand patient years. The available range is thus from 400 down to 200. However, some of this range will be taken up by cotherapy and placebo effects, which will operate in both groups. Assume, once again purely for purposes of illustration, that placebo and cotherapy by themselves reduce fracture events by 100 per thousand patient years, and that the combination with the active agent reduced cases by a further 50. (It could not account for all of the remaining 100, since in the early years of the study, the effect of the agent on BMD would not yet have been fully expressed. See The Long Response Time of Bone, below). These relationships are shown graphically in Fig. 1. Thus, we are left with a situation in which the placebotreated control group, under the conditions of the study, has

FIGURE 1 (Top) Outcome of a hypothetical randomized controlled trial (RCT). (Bottom) Plausible underlying reality reflected in the results of the RCT, showing first the untreated (and unmeasured) background fracture rate and then the reductions therein produced by, first, cotherapy and the placebo effect, and then the active agent. These effects are shown both as they might be observed in an RCT and as might occur in routine clinical practice. Note that the control group fracture rate in the top panel is less than the untreated rate in the bottom panel, inasmuch as it reflects the influence of cotherapy and the placebo effect. Reprinted with permission from Robert P. Heaney.

CHAPTER 64 Design Considerations for Clinical Investigations

a fracture rate of 300 (not 400), and the treated group 250. The investigation may well be very significantly underpowered to find a difference this small. Even if statistically significant, the apparent reduction in fracture (from 300 to 250) per 1000 patient years) might seem biologically small. However, it would be a serious mistake to conclude anything about effect size in such a situation. In a routine clinical situation, with less of a placebo effect, there would likely have been a much wider response range available to the active drug, and the effect of the agent might well prove to be substantially greater. This is pure speculation, of course, but the point is that the placebo effect inevitably narrows the response range and thereby makes it more difficult to find real treatment effects. It is necessary to stress here that, while the effects attributed to cotherapy and placebo in this case are intended mainly to be illustrative, they are not implausible. In general, biomedical scientists greatly underestimate the degree of objective improvement which the placebo effect can produce [13 – 15]. The irony of this situation is that, while a positive finding in such a study clearly establishes that an agent has a greater effect than a placebo, negative studies may mean little, and the usual power calculations are not useful in helping us discern why. Furthermore, whether an agent would produce its putative benefits outside the investigative context, when the placebo effect is not fully engaged, can never be determined by using a randomized controlled trial. Nor can an RCT tell us whether the agent is better than no treatment at all in the usual therapeutic context. Thus, without doubting their inferential power when properly executed, RCTs can sometimes seem to be hothouse flowers, with little connection to what grows in the garden of clinical practice [16].

519 cohort design the entry is nonrandom, while in the RCT it is exit that is nonrandom, but the inferential effect is qualitatively the same.) In fact, a concurrent cohort design (which is what such a degenerate RCT becomes) may well not have been the preferable option, since it is burdened with all of the problems of unequal interference and nonrandom losses, just discussed. A nonconcurrent cohort study might well have been preferable. However, subjects drop out of cohort studies as well, and so it is necessary to examine the impact of such losses on the conclusions we can draw from any such investigation. Losses are both common and frequently large in a disorder such as osteoporosis, both because studies are performed in elderly, infirm individuals who may not be willing or able to maintain the needed commitment, and because the studies must be of long duration. Contemporary investigations usually recognize the likelihood of such losses and attempt to recruit enough subjects so that, at the conclusion of the investigation, sufficient power will remain to discern effects that would be clinically or physiologically interesting. Unfortunately, this stratagem is almost never adequate. Subject losses erode the randomization because they are seldom random (and virtually never demonstrably so). The result, even for small losses, is a substantial widening of the confidence interval for estimates of the frequency of the outcome measure in both of the contrast groups, and an even greater loss of ability to discern a difference between them. This broadening effect is illustrated in Fig. 2. Design experts, in handling the results from such investigations, will commonly insist that lost sampling units be counted both ways, i.e., as treatment failures or as

3. ZERO LOSS OF SUBJECTS The final criterion for proper execution of an RCT, zero loss of subjects, seems both unreasonable and impossible to achieve. It is important, therefore, to understand the reason for this criterion and what failure to observe it does to the power of an RCT. Recall that the inferential power of an RCT is based on the fact that only random chance determines which subjects get allocated to which treatment. Therefore, the laws of probability can be used to determine whether differences between groups are larger than chance alone might have produced. However, experienced investigators recognize that subjects drop out for reasons often related to either the treatment or the disease. Even if we start a study by assigning patients to treatments randomly, if they drop out of study nonrandomly, then what we have left are nonrandom treatment groups, i.e., the experiment has been reduced to the equivalent of a concurrent cohort study. It makes no sense to assert that the experiment is a stronger design than a cohort study, if, in execution, the two become the same thing. (The biases operate differently, since in the

FIGURE 2 Graphical representation of the effect of sample losses on the uncertainty range of an estimate of population prevalence. For a random sample of 90 subjects, 36 of whom exhibited fractures (40% prevalence in the sample), the uncertainty range for population prevalence is 29.8 – 50.9%. But if the actual sample size was 100, and 10 units were lost, then the actual sample composition is uncertain, and the corresponding uncertainty range for the estimate of population prevalence is much broader, i.e., from 26. 6 – 56.3%. Reprinted with permission from Robert P. Heaney.

520 treatment successes, i.e., they developed the outcome measure or they did not. If doing this alters the conclusion, then the study must be considered inconclusive. While it is easy to accept this point in principle, it almost always comes as a surprise when we calculate how large an effect even small losses can have. Assume an investigation of a new treatment agent for osteoporosis. We anticipate that 40% of the untreated subjects will develop a fracture over the period of observation and we hope the new agent will produce at least a 50% reduction in this fracture rate (i.e., to 20% or below). Standard power calculations show that, for this kind of an expected difference, two samples of 100 each will produce a power of 0.84 at the usual  value of 0.05. Next, assume that this is a relatively brief investigation and that we anticipate no more than a 10% loss of sampling units, that is, that we will have 90 subjects in each of the contrast groups at the completion of the investigation. (For simplicity, Fig. 2 depicts this situation for only one of the two contrast groups.) We recognize that this loss will reduce power, and the usual sort of calculation readily shows that, for 90 subjects in each of two groups, the power is 0.79, only slightly below the original power estimate and still a value that would generally be considered acceptable. What is not so often recognized, however, is that the revised power calculation assumes that we started with 180 subjects, randomly allocated into two groups of 90 each, which is not what happened. If, in this hypothetical example, the observed fracture rate in the untreated group was 36 of 90 (i.e., the expected value of 40% — see Fig. 2), and if those 90 had been randomly obtained from the population of all such patients, then the exact 95% confidence interval for the actual incidence rate in the population is 29.8 – 50.9%.3 However, the original sample was 100, not 90. We know for certain that there were 36 fractures in those 90 subjects who remained in study, but we have no information on the 10 who were lost. Perhaps all of them suffered fractures; perhaps none of them did. This additional uncertainty concerning the composition of our sample expands the range of uncertainty for our estimate of the population value at both ends, so that it now becomes 26.6 – 56.3%; or a range half again as large as would have been the case had we started and finished with 90 sampling units. It turns out that an uncertainty range of this size is exactly what would have been produced with original sample sizes of only 43 each, not 90 or 100. As a consequence, for the expected prevalences in this sample, we have only the power that would be produced by samples of 43 each, which can be calculated to be 0.42. In other words, the nonrandom loss of only 10% of our sampling units has reduced our effective power by half.

3

Derived from standard tables of the binomial distribution.

ROBERT P. HEANY

In studies of disorders such as osteoporosis, sample losses of only 10% would generally be considered unusually good. An even more probable 20% loss would reduce effective power to only 0.22 in the foregoing example, and a 30% loss to 0.11. It is sobering to note that, in this latter case, such a power value means that we have only one chance in nine of finding a 50% reduction in fracture rate if it were really there, even though we still have 70% of the original subjects in each group. This is, of course, a worst case scenario. The best case would occur if the sampling unit losses are truly random, or if the true (but undetected) outcome measure prevalence in our lost sampling units was the same as in those subjects who remained in the study. Then the power is equivalently what we would have gotten from two samples of 90 each (or 80, or 70). However, as already noted, it is never safe to assume that subject losses are unrelated to the treatment being tested or the outcomes sought [1,7]. Thus the actual impact of subject losses lies at some indeterminate point between these best- and worst-case scenarios, and we can never know how bad the damage from subject losses may be. Two strategies are generally employed when, at the time of analysis, one confronts the loss of sampling units. The first is a test for differences between the subjects who were lost and those who remained, in terms of certain features they exhibited on entry, features like age, comorbidity, disease severity, and so forth, which might plausibly be considered to have, in their own right, some effect on the outcome measure. Although laudable in intent, this stratagem usually fails on several counts. First, there is almost never enough power to find differences that may be important. Investigations are not designed to test for important differences in subjects who drop out of study. In fact, it is hard to imagine how one might even go about designing such a study. A further problem with this approach is that we can test only those potential factors that we are able to think of, or for which we have obtained measurement values. An inescapable feature of our doing research is that we often work in areas in which we do not know all of the questions, and thus we may not have made the relevant measurements. Predictably, therefore, we generally find that the subjects who dropped out of study seem similar to those who remain, and we lapse into a false sense of security that, indeed, the two groups were not different after all. The null hypothesis being tested in this context (see also below), that the group that dropped out was not different from the group that remained, contains a bias toward preserving the appearance of validity for the investigation. A safer course, more in keeping with traditional scientific skepticism, would be to assume, until evidence of substantial similarity can be produced, that the lost sampling units were different. After all, they did drop out, while the others remained.

521

CHAPTER 64 Design Considerations for Clinical Investigations

A second analytic stratagem is to use what is termed the “intention to treat” criterion in the analysis of data. In this approach, all of the original sampling units assigned to the two (or more) treatment groups are counted, whether or not they complied with the assigned treatment, and often whether or not we have outcome measures for them. When we use the outcome measure in the treated noncompliers, we reduce the effect size in the treatment group. When we do not know the outcome (because subjects were lost), we have little choice but to treat them all the same; i.e., they developed, or did not develop, the outcome of interest (fractures in this case). Both approaches minimize the difference in the outcome measure between groups, if any, and further reduce the power of the investigation. “Intention to treat,” in effect, creates a bias toward the null hypothesis. While this approach may be inferentially preferable to comparing only those subjects who remain, still it works against the core goal of all investigational design, which is to avoid biases wherever possible, not to build them into the investigational strategy. “Intention to treat” is sometimes defended as being representative of what may happen in actual practice, since real patients often fail to adhere to prescribed regimens, just as do subjects in an RCT. However, this defense relates to a secondary question. The primary question in these trials is “Is the agent effective?” We have to find that out first before we deal with compliance issues. Nevertheless, “intention to treat” is generally considered a necessary conservative hedge, given stopping of assigned therapy or loss of sampling units. It is important to stress, however, that “intention to treat” in no way compensates for subject loss, nor does it help us answer the question that stands behind the original investigation. It simply reduces the chances that, if we do find a significantly positive effect of treatment, it will have been produced by biases introduced by nonrandom losses of sampling units. In other words, it reduces the Type I error risk in such investigations. In effect, the statistician-designer, insisting upon intention to treat, seems to emphasize keeping us from falling into the trap of finding an effect that is not real. By contrast, the clinical investigator is concerned with avoiding the trap of missing an effect that might be beneficial. There is always this tension in any investigation. Losses of sampling units and our attempts to deal with them bring that tension onto center stage. That is why design experts insist on zero loss of sampling units. Peto et al. [5] simply assert, “One excellent policy is to accept no withdrawals under any circumstances.” Since this advice is nearly impossible to follow (and would be considered unethical today), it is generally ignored, and along with doing so, we lose sight of the reason why it is necessary. What can investigators do under such circumstances? First, in designing an investigation, appropriate attention must be given to the stratagems necessary to retain sam-

pling units and to track those who drop out in order to obtain the outcome measure in lost units whenever necessary. Subjects should be followed as they move about the country; they should be visited in their new locations; their medical records should be perused in detail; their attending physicians interviewed; etc. Trials can also be designed with prerandomization, “run-in” periods, during which subject compliance and reliability can be assessed, and subjects can themselves decide whether the project is something they are willing to commit to. However, all groups should receive placebo or an agent deemed likely to provoke noncompliance during the run-in, since the randomization cannot be done until it is clear who will likely remain in study. Failing either the implementation or the success of such strategies, we should recognize that otherwise welldesigned investigations, in which there may well be substantial losses of sampling units, will predictably give uncertain results. We should also recognize that negative trials mean far less than properly designed and executed positive ones, since the intention to treat analytic approach, as noted above, biases the investigation toward the null hypothesis. Unfortunately, negative studies have a disproportionately dampening effect on the field. An inconclusive or negative study can actually do considerable harm, since it makes it more difficult to take a second or third look at what might, in fact, be a promising therapeutic advance.

IV. WHEN CONTROLLED TRIALS AND OBSERVATIONAL STUDIES DISAGREE The focus of this chapter is on design considerations, not on interpretation of study results. Nevertheless, the two issues cannot be cleanly separated, and interpretation of results from one design often leads directly to design of other investigations. Hence it will be useful to look briefly at discordances of results derived from different investigational designs. As any field of clinical investigation develops, the usual pattern is that observational studies uncover associations that suggest hypothetical causal connections. Often these are then formally tested in an RCT. If the hypothesis is supported in this strong design, then the observed association is confirmed and the connection deemed causal. If the RCT is negative, the casual connection is considered unsupported even though, as has already been noted, negative findings in RCTs do not definitively exclude a useful and perhaps even clinically important effect of the agency being tested. But what about the case of a positive RCT and observational studies that are mixed — some negative and some positive? (It is hard to conceive of a situation in which all

522 the observational studies are negative and one or more RCTs are positive, since with no evidence of a connection there would have been scant reason for entering into the expense and work of an RCT.) The principle still applies: negative observational studies cannot trump a positive RCT. Of course, single studies of any sort may be statistical flukes, but when multiple RCTs are positive, the causal connection must be considered established. Generally policy decisions are made in accord with this principle. A good example is the recent recommendation concerning folate intake in women of reproductive age. Several of the large observational studies had failed to find a connection between folate intake and neural tube defects, while the RCTs clearly showed protection. Public policy was promptly changed in the direction indicated by the RCTs. Reaching this conclusion does not require understanding of the reason for the discrepancy between results produced by the two study types. Design specialists will simply note that, in addition to the factor being explicitly tested, observational studies are always beset by uncontrolled (and often unknown) factors that may influence the outcome variable. Presumably such factors got in the way of finding a real association with respect to folate intake in this case. However, investigators interested more in the biological question than in the methods used to answer it will still seek to explain the discrepancy. In the case of hypotheses involving nutrient intakes, much of the explanation must lie in weak ability to quantify exposure to the nutrient in question. This is as true for calcium and vitamin D in the osteoporosis field as it was for folate. In all such cases, intake is estimated from responses to questions asked subjects about what they ate. This approach is notoriously inaccurate (see “Estimating Nutrient Intakes,” below). To return to the topic of this section, there is only limited reason for performing observational studies testing an hypothesis that has already been established using stronger designs. Further testing of the hypothesis itself does not fall within those limits. Observational studies may, however, play a useful secondary role when their findings are concordant with those of an RCT. They can greatly extend the generalizability of the conclusions of the RCT, inasmuch as they show that the relationship can be found outside the “hot house” environment of the controlled trial. Thus, in early stages of knowledge in a field, observational studies lead to RCTs, and in later stages they complement RCTs when positive. But they do not refute RCTs when negative.

V. DESIGN ALTERNATIVES The nonconcurrent cohort study can sometimes offer a useful alternative, particularly when, using an RCT, the problem is compression of the response range. Cohort de-

ROBERT P. HEANY

signs, like RCTs, sort the sampling units into contrast groups by exposure to treatment. Nonconcurrent cohort studies have an advantage over RCTs in that there is much reduced opportunity for investigative interference, i.e., for the placebo effect. Unlike experiments, however, the exposure is never randomly assigned, and this creates a design weakness. However, by carefully enumerating the various biases which may lead to selection of one course of treatment over another in advance of the analysis, it may be possible to create a set of criteria for admission of subjects into the analyzed treatment groups which can be applied retroactively to select individuals for study. This means that not all the untreated patients will be included in the control group, and not all the treated patients will be in the active agent group. This approach is not foolproof, by any means, but as we have just seen, neither is the RCT. Also, often the information one needs to make treated and untreated groups similar may simply not be available. Other nonrandom treatment assignment methods have been proposed [17,18], although none enjoys the acceptance of the RCT. One that seems particularly attractive, although virtually confined to social sciences research, is the regression discontinuity design [19]. With this scheme, instead of random assignment to treatment, assignment is totally determined by clearly specified a priori criteria usually based on a pretreatment value of the variable the treatment is designed to change. For example, instead of randomly assigning a group of women to receive calcium or placebo in a study to see if extra calcium will augment bone gain or slow bone loss, the regression discontinuity approach would confine the treatment to those deemed most likely to benefit, i.e., those with the lowest starting values for BMC/BMD. The contrast, untreated (or placebo) group consists of those with higher values for BMC/BMD. The null hypothesis is that the treatment makes no difference (regardless of whom you give it to), and results of the trial are analyzed simply by plotting the terminal measure against the baseline measure, as in Fig. 3. Under the null hypothesis all the data points should fall along the same regression line. If, in fact, treatment does make a difference, then there will be an evident discontinuity in the regression relationship, which can be detected and tested in a number of ways. Figure 3 depicts only one of the possible outcome patterns of an experiment employing the regression discontinuity design. Figure 4 shows the principal outcomes in schematic form and illustrates the analytic flexibility of this design. In each panel, the vertical axis represents the posttreatment value of BMC/BMD, and the horizontal axis the corresponding pretreatment value. Figure 4A, with no effect, reflects the null hypothesis. However, even here, the slope of the line contains useful information. That slope will be unity in a no-change situation, greater than 1.0 during growth, and less than 1.0 during age-related bone loss.

523

CHAPTER 64 Design Considerations for Clinical Investigations

FIGURE 3

A plot of terminal bone mass values against entry values for a simulated trial using the regression discontinuity design. 200 subjects spanning the typical range of values for TBBM/Ht were allocated to treatment if, on entry, they had values below 1500 g/m. Otherwise they received placebo. Note the discontinuity in the regression relationship, which appears at precisely the cut-off point between those treated and those not treated. Reprinted with permission from Robert P. Heaney.

Figure 4B reflects what is termed the “main effect,” which is the outcome depicted in Fig. 3. Actually, since the discontinuity is approximately the same at all values within the treated range of BMC/BMD values, it shows both that

FIGURE 4

the treatment produced an effect and that the effect was not, after all, related to the starting value. By contrast Fig. 4C and 4D depict what is termed an “interaction effect.” In both cases the effect is greater at low starting values of BMC/BMD than at values closer to the treatment cut-off point, i.e., the treatment and the starting value “interact.” In Fig. 4C the treatment cut-off point has been fortuitously chosen to lie at precisely the BMC/BMD value above which treatment would no longer make a difference. In this case there is no discontinuity between the lines, but their slopes are different. More likely is the outcome shown Fig. 4D, in which there is both a slope difference and a discontinuity; i.e., the selected treatment cut-off point and the observed no-effect point for the population do not coincide. While the same kinds of outcomes and analyses would be available in a typical RCT, employing standard multivariate methods in the analysis, the approach used by the regression discontinuity method to conceptualize the design and to plot the results provides a satisfying visual representation of what is involved, which would not be so readily accessible to the clinical investigator with standard multivariate methods. The inferential power of the regression discontinuity design can be fully as great as that of the typical RCT. A difficulty may arise from unequal interference, since the

Principal patterns encountered in analyzing data produced under the regression discontinuity design. See text for details. Reprinted with permission from Robert P. Heaney.

524

ROBERT P. HEANY

placebo effect will be greater in any group explicitly selected for treatment. A double-blind is possible in this situation, but it requires concealing the baseline BMC/BMD data (in this example) from the investigators until the study is completed. The regression discontinuity design is intuitively appealing, particularly in testing remedies for various kinds of deficiencies (from nutritional to educational), since, unlike drug response, one predicts in these situations that there will be a gradient of responsiveness which depends upon the pre-treatment value of the principal outcome measure. The design also offers an ethical advantage over the usual RCT because it seems to ensure that, if the treatment is effective, those most in need will be certain to benefit. Thus, it can be attractive both to investigators and to institutional ethics committees.

VI. DESIGN ISSUES OF SPECIAL RELEVANCE TO INVESTIGATION OF BONE A. The Long Response Time of Bone We have already seen that one of the reasons that the entire response range would not be available to even a potent treatment agent is that the full treatment effect would not have been produced during the course of even a 3 to 4-year study, and certainly not in the early phases thereof. It is curious that, on the one hand, experienced investigators take this slow response of bone for granted, and, on the other, we continue to conduct our investigations of bone-active agents as if we were treating a rapidly responding disorder such as diabetes or hypertension. In hypertension, for example, one can produce an immediate, major reduction in myocardial work and in vessel wall stress by lowering blood pressure or by use of agents such as betablockers. In osteoporosis, however, where one of the problems we are attempting to remedy is that we have too little bone, even the most potent bone-forming agent is able to change bone mass by only a few percentage points per year. Thus, in our fracture studies, we end up looking for an effect before we have produced much change.4 At very least, therefore, one may reasonably exclude first year fractures in any such study, both because the active agent will not yet have had much chance to affect bone mass, and because fractures, by definition, reflect preexisting fragility.

4

On the other hand, if fragility is partly due to level of remodeling activity, then early data may make very good sense, since bone-active agents act very promptly on remodeling.

Counting fractures that occur soon after starting treatment includes the impact of pretreatment conditions as well as that of the treatment. More logical would be to treat for four or five years, or until one has achieved the targeted increase in bone mass, and only then begin the measurements which constitute the formal study. This would, of course, be extremely expensive and difficult to do, which may explain why, instead, we do what is easier, i.e., test the effects of a density-increasing agent before it has had a chance to produce much change. This way of looking at the problem suggests an intermediate stratagem for studies with a bone mass-related fracture end point, and that is to use change in BMC/BMD produced by the treatment, rather than the fact of treatment itself, as the independent variable in our analysis. That makes sense because the research question, when we stop to think about it, can usually be phrased as follows: “If I were to increase bone mass with agent X, would I reduce risk of fracture?” For example, the question is not: “Does fluoride reduce spine fractures?,” but “Does the increase in bone density in the spine produced by fluoride reduce spine fractures?” Since not all patients respond to various therapies with an increase in bone mass, and because those who do respond do so to varying degrees, it may often make sense to plan the investigations at the outset so that bone mass change, rather than treatment per se, will be used as the basis for sorting patients into contrast groups. This can be done retrospectively as well and is a stratagem applicable to both experimental and cohort designs. Farley et al. [20] have employed this approach to advantage in the analysis of fluoride effects.

B. Filling up the Remodeling Space Many, perhaps most, bone-active agents alter the component processes of bone remodeling. Indeed, we use them precisely because of their presumed or demonstrated capacity to alter bone formation or bone resorption. Because of the ways these processes are coupled, the effect of such alterations will often produce transient changes in measurable bone mass which can be misinterpreted, and which therefore create special problems for design of trials of such agents. A brief review of remodeling biology will be necessary to understand the remodeling transient and how it operates (see also Chapters 12 and 15). 1. REMODELING BIOLOGY It has been recognized ever since bone remodeling could be measured that bone resorption and bone formation were coupled processes; i.e., as one increased, the other usually increased more or less to match, and vice versa. Even in the outspoken bone-wasting or bone-gaining disorders, in which imbalance between resorption and formation is re-

525

CHAPTER 64 Design Considerations for Clinical Investigations

sponsible for the gain or loss, the two processes tend to vary together, and the difference between them is usually no more than a small fraction of the absolute value of either. This coupling is now known to have two biologic bases, one local and the other systemic. Locally, remodeling is a surface process, working inward into the bony substance from a vascular, periosteal, or trabecular-endosteal surface. These bone surfaces are normally covered by lining cells which, in effect, insulate the bony material from the bulk of the body fluids. The first step in remodeling, termed activation, involves retraction of the cells of this lining membrane at a particular locus, exposing the bony material underneath to the extracellular fluid and to the humors and cells circulating therein. Certain proteins in bone are chemotactic for osteoclasts, and thus osteoclast precursors congregate at the exposed site and begin the resorption process. After a certain volume of bone is removed, the osteoclastic process ceases and osteoblasts are recruited to the site, possibly because some of the proteins released from bone during the resorption process function locally in a paracrine fashion, to stimulate and localize osteoblast activity. Typically the resorptive phase at any single site lasts for a few days to at most a few weeks; osteoblast deposition of new bone takes 3 – 4 months, by which time the new matrix is about 70% mineralized; then passive mineralization continues at a slow pace for another 4 – 5 months. Coupling occurs locally both because each step triggers the next, and because, in a trivial sense, at least for intracortical remodeling, there cannot be formation until an excavation has made room for it. Coupling also occurs systemically, through the parathyroid axis. Since the blood is supersaturated with respect to hydroxyapatite, newly mineralizing bone depletes the blood flowing past it of up to half its calcium, and thereby creates a condition of hypocalcemia to which the parathyroid glands respond. Parathyroid hormone (PTH), in turn, is a principal determinant of the bone activation threshold, and it thereby effectively stimulates osteoclastic resorption. In so doing, it provides the calcium needed to mineralize forming bone elsewhere in the skeleton. Thus, spatially, current formation pulls current resorption at remote loci in the skeleton; while, temporally, current resorption pushes formation at the same locus by setting the stage for later formation. About 8 – 12% of the skeleton is remodeled each year. Because bone mass measurement methods detect the mineral component of bone, which is the first to be removed at resorbing sites and the last to be restored at forming sites, there will always be less bone measurable in the body than is potentially there. The volume of bone temporarily out of service in this way is termed “the remodeling space.” In a healthy adult skeleton it amounts to about 2% of total skeletal volume, and probably about 6% of the volume of bones with a high specific surface, such as vertebrae [21].

FIGURE 5 Time course for bone mass in the spine in a healthy adult at peak bone mass, subjected to 1 year of treatment with an agent that produces a 50% suppression of remodeling activation. The transient lasts 40 weeks under typical conditions, and no further change in mass occurs even though treatment is continued. When treatment is withdrawn, the reverse transient occurs. The first change represents a shrinkage of the remodeling space, and the second and expansion. Reprinted with permission from Robert P. Heaney.

2. THE BONE REMODELING TRANSIENT The foregoing discussion of coupling and its bases is pertinent here since the temporal and spatial separation of the component processes mean that interference with one or the other step in remodeling will produce a temporary uncoupling. When a bone-active agent alters the activation threshold, or interferes with osteoclastic work efficiency, the immediate effect is to resorb less bone per unit time. However, since formation at sites already resorbed and/or part way through their forming phase continues according to its own schedule, there will be more bone being formed than resorbed, and measurable bone mass will increase. This effect lasts only for a time equal to the length of the remodeling cycle. Any reduction in activation frequency will reduce the size of the remodeling space, even if the agent responsible has no effect on the steady-state balance between formation and resorption. Thus, a 50% reduction in activation in a fully normal, young skeleton, will, over a period of 7 – 9 months result in a one-time increase in spine BMD of about 3% (Fig. 5). While that gain will be retained so long as treatment is continued, no further change will occur after the first remodeling cycle. Then if treatment is withdrawn, the remodeling space expands again, and the apparent gain in BMD is lost. That is why the term “transient” is used for this phenomenon. The size of the transient, that is, the size of the one-time gain in measurable bone produced by reducing the remodeling space, will depend upon baseline bone mass, baseline activation rate, and the length of the remodeling period, as well as upon the dose or potency of the administered activation suppressive agent.

526

ROBERT P. HEANY

There are several implications for the design of clinical trials in this remodeling biology. In theory, all activation suppressors will produce a positive remodeling transient. Bisphosphonates, calcitonin, estrogen, and calcium all do so. However, it is less clear which ones, if any, will alter steady-state remodeling balance. Hence the study designer must first be clear about what he or she is actually looking for. In patients with osteoporosis the remodeling space is commonly larger than normal. One reason is trivial, but often ignored: even an absolutely normal remodeling space will constitute a higher percentage of a reduced skeletal mass. Since the results of trials using bone mass (or density) as the endpoint typically express results as percent change from baseline, it follows that an identical constriction of the remodeling space will be expressed as a larger percentage change in a depleted skeleton than in a normal one. A second reason is that activation is often absolutely elevated in osteoporosis. A third is that the remodeling period is also frequently prolonged. In brief, the remodeling space in osteoporosis is nearly always relatively large and often absolutely so, as well. Many studies in the past have ignored the transient and have become entangled in the type of analytical problem illustrated in Fig. 6. The figure shows the true bone mass curve (first without and then with the usual measurement

noise) which could be produced solely by constriction of the remodeling space in a patient with osteoporosis. As can be seen, there is gain during the first year of treatment, and then slow loss thereafter, just as there was before starting treatment. Because of measurement precision errors, the curve will never be as sharply defined as the figure shows, and there will be an analytic tendency to compute a linear slope through the data, or to compare bone mass between groups at the study endpoints. Either way, the percentage gain is greatest at 1 year and then seems to become smaller every year thereafter. Clearly this is the inevitable result of averaging a positive change of fixed duration with a smaller negative change, the duration of which varies with the length of observation. It should be stressed that an agent that did nothing more than reduce the remodeling space could still be very benefi cial, particularly in patients with substantial bone loss or high remodeling rates (or both). An activation suppressor can, in a few months’ time, produce a 10 – 20% increase in usable bone mass in such patients. However, it is important also to recognize that this one time transient gain is not the same thing as a steady-state effect. Figure 6 makes clear that studies intended to test for the presence or magnitude of the remodeling transient should typically be of only about 1 year’s duration. Conversely, studies looking for steady-state effects cannot begin until at

FIGURE 6 Three-year time course for true bone mass/density in patients with osteoporosis treated with an agent which suppresses remodeling activation by 50% but has no effect on remodeling balance. (Top left) The underlying reality; (bottom left) what might have been observed given a measurement precision of  1.5%. The effect, if expressed as percentage change per year (right), declines with study duration. Reprinted with permission from Robert P. Heaney.

527

CHAPTER 64 Design Considerations for Clinical Investigations

FIGURE 7 Time course for true bone mass/density for two agents that suppress remodeling activation, but only one of which also favorably alters remodeling balance. Reprinted with permission from Robert P. Heaney. least a full year of treatment has passed. As Fig. 6 also makes clear, it will not be possible to tell whether an agent alters steady-state remodeling balance until after the transient has played itself out. That means that studies intended to detect effects on the underlying pathophysiology of the disease will not be able to use data derived during the transient phase (approximately the first year of treatment). In brief, the measurements on which the steady-state effect will be based should not begin until at least 12 months after starting therapy. Moreover, since steady-state effects will predictably be smaller than the transient, the study will have to be designed with sufficient power to find the small differences likely to be present. These points are illustrated in Figs. 7 and 8, which show the true bone mass curves (again, without measurement

FIGURE 8 Simulation of the time course for true bone mass/density in a controlled trial in which the combination of investigational agent and cotherapy suppresses remodeling activation by 50%, and the cotherapy alone by 20%. The interval from A to B reflects the time period over which data relative to the efficacy of the agent should be accumulated. Reprinted with permission from Robert P. Heaney.

noise) in patients with osteoporosis who are losing bone prior to treatment at a rate of 1% per year. Figure 7 contrasts the changes that would be produced by two agents, each of which suppresses activation by 50%. One agent fails to alter the underlying remodeling balance, and so, after the transient has been fully expressed, bone loss continues, although at half the rate that existed prior to treatment (simply because remodeling has been reduced by half). The other agent, by contrast, produces a favorable change, in that formation is greater than resorption, and bone is now being gained at a rate of 0.5% per year. The difference between them, a change in BMD of 1% per year, would clearly be therapeutically beneficial and hence worth detecting. As Fig. 7 makes clear, the initial bone mass curves are nearly identical for the two agents. This is as would be expected, since the two have the same effect on activation. The figure shows that there is a small difference at 1 year, which would probably not be detectable in any practicable study. From that time on, slow bone loss occurs with one agent and slow bone gain with the other. The challenge is to design an investigation that will detect the difference in these slopes. Power calculations would typically have to be made on such a basis (see Timing of Bone Mass Measurement, below). The important thing to recognize is that it is not change from baseline that is being evaluated, since most of that difference is a reflection of the transient, but change from a new, posttransient starting point. Exactly the same challenge exists when one is comparing an active drug with a placebo. Figure 8 illustrates that situation. Here the bone mass difference between the groups is larger at all points after starting treatment, and one might be tempted to think that it would, therefore, be easier to detect. (Note that the figure contains a small transient for the placebo group; this is because calcium and vitamin D cotherapy, which would almost certainly be given to both groups in any such experiment, will, by suppressing tonic PTH secretion, also produce a typical — although small — remodeling transient.) It is the difference in rates of change of bone mass that we are testing, not the differences in bone mass between the groups, and here the situation is exactly the same as in Fig. 7. For any feasible study duration, most of the actual difference between groups would still be accounted for by the larger transient produced by the active agent. Since transient effects are not the object of the investigation, it is the posttransient difference in slopes between the two groups that we must look for.

C. Timing of Bone Mass Measurement It is the usual practice, in studies in which bone mass is a primary or secondary outcome measure, to spread measurements evenly over the course of the study, typically every 6

528 months. In a 4 year study, that means nine measurement points, and if three skeletal sites are measured in each woman, 27 measurements in all. For every 100 women enrolled, that work will cost between $0.5 and $1.0 million — not a trivial expenditure. It is useful, therefore, to examine our objectives in making these measurements and to see whether this way of timing them best meets those objectives. Reasons for multiple, spaced measurements would include catching nonlinear trends in the data, assessing longterm measurement stability, and improving the chances of having terminal data for early dropouts from study. Only the study designers can say how important each of these may be. For most studies, probably only the last reason (salvaging as much data as possible from dropouts) will be important.5 Note that improving the precision of the slope or trend estimate was not one of the reasons listed for multiple, spaced measurements. Or, at least, if that is an objective of the design, then evenly spaced measurements are not a good way of achieving it. The closer measurements are to the center of a study time span, the less they contribute to the slope estimate [22]. (At dead center, they contribute nothing at all). The best way to improve slope estimates is not by spreading measurements out but by doubling up measurements at beginning and end. For example, in a 2 year study, double measurements at beginning and end (four measurements in all) yield a distribution of slopes with a variance only half as large as would be produced by five measurements spaced every 6 months over the same 2year period. So, for 20% less work (and money) this timing strategy produces a substantial improvement in investigative power. The remaining consideration, salvaging data for those who leave study early, presents trade-offs that the investigator will have to evaluate. If one uses the scheme of double measurements at beginning and end, and if no terminal measurements are available because a subject drops out early, then all data for that subject are lost. Multiple, evenly spaced measurements minimize that loss, but at a cost of lower precision for everyone in the study. There are intermediate stratagems that may be better. If subject rapport is likely to be good, it may be possible for investigators to obtain terminal measurements at time of withdrawal, whenever that may occur, particularly if the need for such measurement is spelled out in the briefing

ROBERT P. HEANY

and consent process and become a part of the quasi-contract between subject and investigator. Or, it may turn out to be worth the extra cost to build interim measurements into the protocol in the last half of the study (on the expectation that, for some subjects, one of those measurements will be their terminal value). However, such interval measurements in the first half of study are likely to be a poor investment of time and resources. This is because dropouts are fewer then and because the necessary weighting of data by study duration (see Weighting Data, below) sharply discounts the value of salvaged short-term data. At stake is the power of the investigation to find small differences that may, nonetheless, be clinically important. Sample losses reduce power, as we have seen. But so do imprecise estimates of the outcome measure (rate of change in bone mass in this case). No general solution exists. Power calculations should be made for each of the timing options, along with an estimate of likely losses of data with each, and the choice made accordingly.

D. Weighting Data by Duration of Observation Because measurement uncertainty for bone mass is generally greater than the short-term change likely to be produced by treatment, it is highly desirable that observations be weighted by duration of observation. For any given underlying rate of change, slopes computed over study durations of 1 year or less will exhibit relatively huge dispersion values, while slopes computed over 4 years will be much more tightly clustered. Table 2 illustrates this point. The table presents mean and standard deviation values for slope estimates derived over 1, 2, 3, and 4 years, in a series of simulations in which the true slope was a constant 1.0% year and the coefficient of variation for replicate measurements was 1.5% at each measurement point. (Note that this is probably better precision than can usually be obtained at spine or hip in patients with osteoporosis). As can be seen, the 4-year slope value has a precision that is  5 times better than a 1-year estimate, and  2.5 times better than a 2 year estimate. If, in computing average response at end of study, one averages all data without weighting, the precision of the final estimate will be unnecessarily broadened, and the TABLE 2

5 We already know about the remodeling transient and need not be measuring across it if we are interested mainly in steady-state effects. While assessing measurement stability over time can be important, if a study is not otherwise planned specifically to do that, then spacing of measurements may be useless in that regard. Furthermore, a good quality assurance system should be in place in the laboratory to assure long-range measurement stability, so even that goal need not be a reason for building multiple measurements into the study design.

Slope Estimates for Various Study Durationsa

Mean slope Standard deviation

1 Year

2 Years

3 Years

4 Years

 0.538

 0.993

 1.037

 1.034

2.805

1.363

0.875

0.564

a Results of a simulation of 40 subjects; true slope value   1.0 year1; 1.5% measurement precision at each study point; measurements spaced at 1-year intervals

CHAPTER 64 Design Considerations for Clinical Investigations

power to detect treatment effects correspondingly diminshed. The gain in numbers of subjects can be more than offset by the increase in sample variance. A useful starting strategy is simply to weight observations by study duration, although some nonlinear weighting scheme may prove to be better. That way we use all the data and at the same time do not permit the less precise, short-term observations to broaden the uncertainly range of our estimate unduly.

E. What Should We Measure? The main output of most bone absorptiometry instruments (specifically SPA, DPA, pDXA and DXA) is a value called bone mineral density (BMD see also Chapter 59). It is technically an “areal” density, i.e., the quantity of bone mineral lying behind the projected silhouette of a bony region or part. To produce this result, the instrument must detect the edges of the region concerned, compute its area, and measure the mineral it contains. BMD has been emphasized in the development of the technology for mainly commercial reasons. It minimizes differences between large-boned and small-boned individuals and hence facilitates assessment of what is “normal” in the process of screening, which has been a principal target market for the technology. Unfortunately, while useful in screening, density is a conceptually weak end point for the study of bone. For the same microarchitecture, the strength of a bony part will be a straightforward function of its mass (BMC) and shape, while the relation to density will be complex and sometimes the opposite of what one might suppose. (A small, high-density bone will often be structurally weaker than a larger bone with a lower BMD.) Also, in most investigations of therapeutic, preventive, or causative factors, we will have longitudinal measurements in the same individuals; hence there will usually be no reason or need to minimize interindividual differences. Prentice et al. [23] have gone so far as to advocate the abandonment of BMD entirely in epidemiological research, advocating use of BMC and bone size as distinct independent variables. In the original, single-photon absorptiometry instruments, the determinations of BMC and area were independent, with the result that BMC was usually a more precise measure than BMD (i.e., with a smaller coefficient of variation for replicate measurements), and it was therefore not only conceptually preferable, but analytically better as well for longitudinal studies. Unfortunately, with DPA and DXA, the measurements of area and BMC are linked in the software, and BMD emerges as a more precise measure than BMC. That can create a seeming dilemma for the investigator, which should, nevertheless, generally be resolved by coming down on the side of the more apposite measure, irrespective of its precision.

529 Carter et al. [24] have presented a theoretical analysis of this general problem and have proposed a useful alternative approach to BMD. They modify values produced by current instruments by using a readily measurable surrogate for the third dimension missing in areal densities. Their measure, called BMAD (for bone mineral apparent density), is both theoretically and pragmatically preferable to the more usual BMD. It is doubtful, however, that BMAD adds as much as a straightforward combination of BMC and bone size. One example of a situation in which BMD is plainly the wrong measure is a study involving growth. Growth involves increases in all three dimensions of a bone, and BMD eliminates two of those three (leaving only change in the Z-axis). A cube increasing by 6% in mass (and volume) over the course of a study, but without change in true density, will have an apparent increase in density of less than 2% if BMD is used, while BMC will detect the full 6% increase in mass. Thus BMD finds a spurious increase in density, and at the same time misses two-thirds of the real change in mass. It is distressing to note the number of published reports describing studies in children, adolescents, and young adults, in which BMD has been used as the principal outcome measure. It is hardly surprising that many such studies have reported cessation of bony accumulation at relatively young ages, or that they contain such obviously discrepant findings as pubertal children achieving 80% of adult bone values at a time when it is known from physical anthropology that better than half the skeleton is still to be amassed. A recent exchange of correspondence (25,26) highlights the clear superiority of BMC over BMD in such studies. Dimensional changes are most obvious during growth, but they are not confined to that period. For many bones there is slow periosteal expansion throughout life. This has been shown for the femur shaft [27] and vertebral bodies [28 – 30], as well as for the skull [30 – 32]. Matkovic et al., in a study in which bone area, BMD, and BMC were all measured, showed that the slow decline in vertebral BMD commonly reported for women between ages 20 and 50 was actually due to an increase in vertebral size, with BMC remaining constant over that period [30]. This is an instructive example of how using BMC and BMD can lead to quite different conclusions. For short-term treatment studies in mature adults, BMD and BMC will vary in parallel, and when that can be safely predicted, the distinctions discussed here will make little or no practical difference. If DXA is the measurement technology, it will often be preferable, in fact, to use BMD, simply because of its superior precision. Nevertheless, the point of this discussion is that BMC and BMD do not measure the same thing; they are not always interchangeable; and the investigator needs to be quite clear about what it is he or she wishes to investigate.

530

ROBERT P. HEANY

Incidentally, there are research questions for which using a density measure would be precisely the right approach. For example Gilsanz et al. [33] have shown an interesting increase in true trabecular density of the vertebral bodies in girls as they cross puberty, using quantitative computed tomography (QCT). Probably this is the opposite of the corresponding decrease in density which occurs at the time of menopausal cessation of estrogen production. There are interesting biological questions associated with these changes. In general, if the investigator is explicitly concerned to answer a true density question, then probably a QCT-based measure would be preferable to the hybrid, areal density afforded by the DXA BMD approach.

F. Estimating Nutrient Intakes6 At least 10 distinct nutrients are known to influence skeletal status, either positively or negatively (see Chapter 27). In planning a study, an investigator may wish to assess the intake of one or more of these nutrients, either because they may be formally a part of the research question or because of a desire to adjust for confounding influences exerted by uncontrolled variations in intake of these nutrients. The traditional way of doing this is by asking people what they eat (or ate). Methods include food frequency questionnaires, diet records, diet recall, and diet diaries. Usually these methods contain substantial inaccuracies, both random and systematic. Random errors are introduced (1) because database values for nutrient content of food are means, and do not reflect the variability in actual nutrient content; (2) because current intake is only weakly associated with past intake; and (3) because of over- and underestimates of portion size. Systematic biases are introduced by virtue of the documented tendency to understate intake of certain foods and to overstate intake of others. The suitability of these approaches varies widely, depending upon the research question being asked. For example, the random component of the errors involved in intake estimates will not substantially distort estimated mean values for a population, since the errors in opposite directions tend to cancel one another. However, exactly the same random errors bias correlational analyses toward the null, because in such analyses it is the individual values that are correlated, not their mean (and it is the individual values that contain the error). For both types of question the systematic biases, of course, distort the findings, producing erroneous estimates of the population mean as well as giving rise to spurious associations or hiding real ones.

6 See also “Problems in the investigation of the Effects of Nutrition on Bone” in Chapter 27.

One commonly reads, in grant applications and in manuscripts, that the method selected has been “validated.” Often no details are given, and the references cited will describe statistically significant correlation between the results obtained with the method in question and some other method. Both features of such “validation” are irrelevant. What is needed is an indication of the adequacy of the substitution of the value obtained from the method selected for actual nutrient intake. Statistically significant correlation between methods is beside the point; even so, usually no more than half the interindividual variation obtained with one method can be captured by values produced by the other one, purportedly “validated” by correlation with the first. At some point, nutrient intake estimates have to be validated against chemical analyses of the foods concerned. This can be difficult and expensive, and one leading epidemiological textbook goes so far as to suggest that such a “gold standard” does not exist (34). That is, of course, incorrect. Such validation can and must be done. Moreover, it must be done individually, since nutrient content of foods varies regionally, by brand, and over time. An investigator cannot safely rely upon someone else, at some other research center, to do that for him/her. There are at least four key components of a nutrient intake estimate: (1) nutrient content per gram of food; (2) number of grams of food ingested; (3) recall of food items eaten; and (4) long-term consistency of intake, i.e., adequacy of substitution of one day’s intake for a several-day average and stability of intake over time. Additionally, for some nutrients, bioavailability must also be taken into consideration.7 Most instruments rely on various food databases for nutrient content of food, such as the USDA Handbook 8 However, what many investigators fail to recognize is that the values given in such databases are averages and that actual content can vary substantially from the value given.

7

It has long been recognized, for example, that heme iron is better absorbed than nonheme iron; thus the adequacy of a given iron intake will depend strongly upon whether the iron is derived from plant or animal sources. A similar distinction applies to several of the nutrients important for bone health, notably calcium and vitamin D. The calcium of the Brassica sp. vegetables (e.g., kale, collards, bok choy, broccoli, etc.) exhibits about 8– 10  greater bioavailability than that of the calcium in spinach; and for the same ingested calcium load, the calcium of common beans is only half as bioavailable as the calcium of milk. Rhubarb and sweet potatoes, both relatively high in calcium content, exhibit poor bioavailability. Vitamins D2 and D3, once considered equivalent and used interchangeably in milk fortification, have been shown by Vieth [35] to have different potencies, and while this difference is not technically a matter of bioavailability, the net effect is similar. Both instances illustrate the difficulty in estimating effective intake. 8 The usual assumption that nutritionists make is that overestimation for one food will be compensated by underestimation for another, an assumption of dubious validity for individuals.

CHAPTER 64 Design Considerations for Clinical Investigations

Different varieties of the same vegetable will have different nutrient concentrations; soil, weather, season, and region also contribute significant variability, as does loss of moisture during transportation and storage. Charles, for example, found that, when he analyzed weighed quantities of food fed to subjects on a metabolic unit, the database value had captured only 76% of the actual intake variability [36]. While that might sound good if all one were looking for was an association, the residual variability is large enough to vitiate any metabolic balance study, despite the fact that, in a metabolic balance environment, actual food intake (both quality and quantity) is known because it is investigator-controlled. Additionally the instrument must be validated by determining its ability to reproduce portion size estimates. This will require use of food models, weighed servings, and volumetric measures in each subject, usually several times during a long-running project. As is evident, this effort greatly adds to the cost of an investigation, but there is little value in testing an hypothesis under conditions in which the costs are less but exposure to the independent variable can be only approximated. In observational studies, estimates of both portion size and items actually consumed are dependent upon subjects’ self reports. It is a matter of everyday experience that two people do not have the same image of what a “normal” or “usual” portion may be. These problems and their impact on nutrient intake estimates have been reviewed elsewhere (37) and need not be discussed further here, where the emphasis must be on designing around the difficulties.

VII. BURDEN OF PROOF AND THE NULL HYPOTHESIS The null hypothesis is, as noted above, the bedrock from which we make most inferences about phenomena subject to variation. Two samples taken randomly from a single population are likely to differ to some extent, but large differences will be less common than small. With knowledge of the distribution of such observed differences under the null hypothesis for any given sample size, we can quantify how often pure random chance would produce a result with any given degree of difference between two study groups. We thereby estimate the chances of being wrong if we say two treatments differ substantially. When evaluating a new treatment for osteoporosis, our hypothesis takes some such form as: the agent will slow bone loss, or it will reduce fractures, etc. The null hypothesis is the logical opposite of these statements; i.e., the agent does not slow bone loss or the agent does not reduce fractures. Typically in biomedical research we presume the truth of the null hypothesis and shift the burden of proof to those who propose the hypothesis. However, we have already

531 touched upon one situation where presuming the correctness of the null hypothesis may not be appropriate, namely in dealing with subjects dropping out of an investigation prior to planned study termination. We noted here that assuming no differences between the lost and the retained subjects created a bias toward finding the subject losses random. This, as we have noted, was because investigators rarely have power to find differences that might be important and almost never know exactly what differences to test for. Accordingly, when we do test, we almost always find no significant difference. The hypothesis we would like to see supported in this context is, of course, that the lost subjects are the same as those who remained in study; the burden of proof ought to fall on us to show that that is correct. The problem here is that it is very difficult to test such an hypothesis (since, in actuality, it is not null). There are other, analogous situations in the investigation of osteoporosis which give rise to the same kind of dilemma. It is worth touching on them briefly, as they provide a window onto certain psychological quirks and prejudices that the investigator needs to recognize. They affect research approaches and outcomes as surely as does a biased sample. In each instance, they influence where the burden of proof is placed. An example is provided by the hypothesis that, in addition to mass density, fatigue damage in bone contributes importantly to skeletal fragility. The null hypothesis here is straightforward: fatigue damage does not contribute importantly to skeletal fragility. While logically impeccable, that position is at best implausible in light of what is known from other source, particularly from materials science. Skeletal science does not exist in a vacuum. Structure engineers have firmly established that fatigue damage is an important fragility factor for essentially every known structure or device subject to loading. Thus, to maintain the null hypothesis in this instance is to say, in effect, that bone is the only known structural material in which fatigue damage is not a significant fragility factor. (In this instance, objectivity has been further clouded by strong economic interests arrayed on one side of this issue. While not a component of investigational design, per se, such forces can be very much a part of the investigative climate). Given the collateral evidence from materials science, it would seem that the burden of proof should have fallen on those who asserted that fatigue damage was not a factor, rather than upon those who said it was. Yet another example, presenting the same sort of ambiguity, is offered by the relationship between health and nutrient intake, for example, in our context, the matter of bone health and calcium or vitamin D intake (38). Calcium is a nutrient, not a drug, and hence there can be no strict parallel with a pharmacologic agent. There is no exposure/nonexposure dichotomy. Instead, we have a continuum of intakes. One can contrive an hypothesis/null hypothesis

532

ROBERT P. HEANY

formulation if one picks any given intake and then states that more would be beneficial (or less would be harmful); the corresponding null hypothesis would be that more calcium would not be beneficial (or less would not be harmful). But what intake do we select as our reference point? On what intake do we confer the privileged position of not having to prove its correctness, while requiring all other claimants to prove they are better (or worse)? These situations are more common in clinical investigation than is generally recognized, and until the dust settles on any one such issue, the choice of a starting position should be a matter of prudential judgement, since it can have important consequences for public health and for resource utilization. Current practice in a field, or prior understanding of the pertinent mechanisms, should enjoy at best only a very limited privilege. We might rather ask: Which position will do the least harm while we are waiting for definitive evidence one way or the other? Such a “do no harm” approach can be a useful guide in a nutritional context, but it would be less helpful for an issue such as fatigue damage. Here, plausibility, as inferred from collateral evidence, may be the most reliable guide we can find.

References 1. D. Mainland, “Elementary Medical Statistics,” 2nd ed. Saunders, Philadelphia, 1963. 2. M. J. Gardner and D. G. Altman, Statistics with confidence. Br. Med. J. (1989). 3. J. J. Schlesselman, “Case – Control Studies. Design, Conduct, Analysis.” Oxford University Press, New York, 1982. 4. S. J. Pocock, “Clinical Trials. A Practical Approach.” Wiley, New York, 1983. 5. R. Peto, M. C. Pike, P. Armitage, et al., Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Br. J. Cancer 34, 585 – 612 (1976). 6. E. Barrett-Connor, Postmenopausal estrogen and prevention bias. Ann. Intern. Med. 115, 455 – 456 (1991). 7. M. R. Stegman. R. P. Heaney, R. R. Recker, D. Travers-Gustafson, and J. Leist, Velocity of ultrasound and its association with fracture history in a rural population. Am. J. Epidemiol. 139, 1027 – 1034 (1994). 8. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Incidence of clinically diagnosed vertebral fractures: A population-based study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 9. F. H. Lahey and H. Hare, Malignancy in adenomas of the thyroid. JAMA 145, 689 – 695 (1951). 10. F. W. Lafferty and M. E. Fiske, Postmenopausal estrogen replacement: A long-term cohort study. Am. J. Med. 97, 66 – 77 (1994). 11. G. Kochersberger, N. J. Buckley, G. S. Leight, S. Martinez, S. Studenski, J. Vogler, and K. W. Lyles, What is the clinical significance of bone loss in primary hyperparathyroidism? Arch. Intern. Med., 147, 1951 – 1953 (1987). 12. 21 Code of Federal Regulations Ch. 1314.126 (4/1/89 edition). 13. S. Wolf, The pharmacology of placebos. Pharmacol. Rev. 11, 689 – 704 (1959). 14. H. K. Beecher, The powerful placebo. JAMA 159, 1602 – 1606 (1955). 15. J. S. Goodwin, J. M. Goodwin, and A. V. Vogel, Knowledge and use of placebos by house officers and nurses. Ann. Intern. Med. 91, 106 – 110 (1979).

16. A. R. Feinstein, Epidemiologic analyses of causation: The unlearned scientific lessons of randomized trials. J. Clin. Epidemiol. 42, 481 – 489 (1989). 17. E. L. Korn and S. Baumrind, Randomized clinical trials with clinician-preferred treatment. Lancet 337, 149 – 152 (1991). 18. R. M. Veatch, Justice and research design: the case for a semi-randomization clinical trial. Clin. Res. 31, 12 – 22 (1983). 19. W. M.K. Trochim, “Research Design for Program Evaluation: The Regression-Discontinuity Approach.” Sage, Beverly Hills, CA, 1984. 20. S. M. Farley, J. E. Wergedal, J. R. Farley, G. N. Javier, E. E. Schulz, J. R. Talbot, et al., Spinal fractures during fluoride therapy for osteoporosis: Relationship to spinal bone density. Osteoporosis Int. 2, 213 – 218 (1992). 21. R. P. Heaney, The bone remodeling transient: Implications for the interpretation of clinical studies of bone mass change. J. Bone Miner. Res. 9, 1515 – 1523 (1994). 22. R. P. Heaney, En recherche de la difference (P  .05). Bone Miner. 1, 99 – 114 (1986). 23. A. Prentice, T. J. Parsons, and T. J. Cole, Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants. Am. J. Clin. Nutr. 60, 837 – 842 (1994). 24. D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone Miner. Res. 7, 137 – 145 (1992). 25. R. P. Heaney, Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 10, 172 (1995). 26. H. C. G. Kemper, D. Welten, and J. Twisk, Reply (to 25). J. Bone Miner. Res. 10, 173 (1995). 27. R. W. Smith, Femoral expansion in aging women: Implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 28. J. S. Arnold, External and trabecular morphologic changes in lumbar vertebrae in aging. In “Progress in Methods of Bone Mineral Measurement” (G. D. Whedon and J. R. Cameron, eds.) pp. 352 – 410 U.S. Department of Health, Education and Welfare, Washington, DC (1970). 29. M. F. Ericksen, Some aspects of aging in the lumbar spine. Am. J. Phys. Anthropol. 45, 575 – 580 (1976). 30. V. Matkovic, T. Jelic, G. M. Wardlaw, J. Z. Ilich, P. K. Goel, J. K. Wright, M. B. Andon, K. T. Smith, and R. P. Heaney, Timing of peak bone mass in Caucasian females and its implication for the prevention of osteoporosis. J. Clin. Invest. 93, 799 – 808 (1994). 31. H. Israel, Continuing growth in the human cranial skeleton. Arch. Oral. Biol. 13, 133 – 137 (1968). 32. C. Susanne, Ageing, continuous changes of adulthood. In “Human Physical Growth and Maturation. Methodologies and Factors” (F. E. Johnston, A. F.Roche, and C. Susanne, eds.), pp. 161 – 175 Plenum Press, New York, 1979. 33. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechat, H. K. Huang, E. E. Schulz, C. R. Libanati, and C. C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). 34. W. Willett, “Nutritional Epidemiology,” 2nd ed. Oxford Univ. Press, New York, 1998. 35. R. Vieth, Vitamin D supplementation, 25 – hydroxyvitamin D concentrations, and safety. Am. J. Clin. Nutr. 69, 842 – 846 (1999). 36. P. Charles, Metabolic bone disease evaluated by a combined calcium balance and tracer kinetic study. Dan. Med. Bull. 36, 463 – 479 (1989). 37. R. P. Heaney, Nutrient effects: Discrepancy between data from controlled trials and observational studies. Bone 21, 469 – 471 (1997). 38. Consensus Development Conference: Optimal Calcium Intake. JAMA 272, 1942 – 1948 (1994).

CHAPTER 65

Development and Evaluation of New Drugs for Osteoporosis HENRY BONE

I. II. III. IV.

Michigan Bone and Mineral Clinic, Detroit, Michigan 48236

Introduction Selection of Drugs for Clinical Development Role of Preclinical Testing Relationship of Bone Mass to Strength

V. Animal Testing VI. Human Testing References

I. INTRODUCTION

relationship between skeletal fluoride concentration and bone strength [2]. As a result of this disparity, clinical use of fluoride has shown no benefit, and perhaps even an adverse effect, on fracture rates [3]. Similar concerns arose about etidronate, an early bisphosphonate, which can cause osteomalacia if exposure is sufficient [4,5]. Drug development guidelines have evolved to take such experience into account. Current guidelines are well-developed for the evaluation of estrogens and other antiresorptive drugs for postmenopausal osteoporosis, but less specific concerning additional indications and anabolic agents.

Development of new drugs for osteoporosis follows the general principles fundamental to drug development as well as specific principles pertinent to osteoporosis. The objective is to evaluate drugs systematically so that effective, well-tolerated medications become available expeditiously, and ineffective or harmful medications are identified and discarded as early in their development as possible. Regulatory authorities such as the Food and Drug Administration (FDA) and the Committee on Proprietary Medical Products (CPMP) of the European Medicines Agency have issued guidelines for the development and evaluation of drugs for osteoporosis, as has a study group of the World Health Organization (WHO). Drug development guidelines for osteoporosis must take into account the primary relationship between bone mass and bone strength, while acknowledging that certain drugs might alter this relationship. For example, fluoride can clearly increase radiographically measured bone mass [1] but animal studies have shown an inverse

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. SELECTION OF DRUGS FOR CLINICAL DEVELOPMENT When therapeutic objectives are selected in a “top-down” manner, based on perceived medical need and scientific opportunity, a particular desired drug action is typically selected as a target. Assays for the target activity are developed and compounds are screened for the desired

533

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

534 biological effects. In other situations, observations of natural phenomena provide therapeutic leads. Naturally occurring compounds may be identified as potential therapeutic agents, and such compounds or compounds mimicking their actions may be developed. Once compounds are identified that have suitable characteristics including binding, selectivity and solubility, drug effects are further characterized in appropriate animal systems. For instance, preclinical testing of drugs for postmenopausal osteoporosis is usually done in female animals that have undergone oophorectomy. If the desired effect is confirmed in animals at well-tolerated doses, consideration is given to testing in humans.

III. ROLE OF PRECLINICAL TESTING Animal model systems serve multiple purposes in drug development. First, such systems help elucidate physiological and pathophysiological mechanisms. Second, they permit study of the effects of drugs on those processes, thereby demonstrating therapeutic mechanisms of action. Third, animal models permit assessment of drug adverse effects, both on target tissues and on other organ systems. Several points should be made about the employment of animal models in the study of drugs for osteoporosis. At an earlier stage of our understanding of osteoporosis, the disease was often defined by the occurrence of spontaneous a traumatic fractures. Animal models completely reproducing this clinical syndrome were difficult to come by. However, our present understanding is that osteoporosis is a state of bone tissue in which mass or density is reduced in a way in which microstructural elements are also lost or compromised, resulting in fragility [6]. There are several systems that simulate this state well enough to provide highly useful information. Of these, the postoophorectomy rat model is surely the most extensively studied. It is not necessary for an animal to perfectly reproduce the human disease to be useful. It is necessary only that it answer the specific question or questions being studied. It is particularly noteworthy that animal systems have accurately identified all of those drugs with undesirable properties that undermine the normal relationship between bone mass and strength.

IV. RELATIONSHIP OF BONE MASS TO STRENGTH Because of the fundamental relationship between bone mass and bone strength, a central issue in the development of drugs for osteoporosis is the possibility that a drug might increase bone mass, but fail to strengthen, or even weaken, the

HENRY BONE

bone. If it were not for this concern, we could simply rely on the basic principles of engineering that tell us that if the mass of a structure is increased with no harm to its material properties or architecture, it will be strengthened. Thus, detection of any adverse effect of a drug on the material properties of bone is an essential purpose of preclinical investigation. Early in the course of treatment, an anti-resorptive drug may actually strengthen bone out of proportion to its overall effects on bone mass [7]. The explanation for this observation may be that remodeling pits act as flaws in the structure, thereby accounting for weakness out of proportion to the actual volume of the remodeling space. If so, increasing bone mass by reducing the number and/or size of the pits may in turn increase strength by a greater amount than if the increase were diffuse and uniform throughout the bone. Once the remodeling space is in a steady but contracted state, incremental effects on bone mass would be expected to further strengthen the bone, but the relative effects of this accretive increase might well be smaller than the essentially reparative effects of the initial remodeling transient. It is important that the preclinical (and clinical) assessment of drug effects be carried out over a period of exposure significantly long that any chronic harmful qualitative effects are not masked by the remodeling transient (see also Chapter 64). The process of drug development for osteoporosis has been greatly influenced by the concern of drug developers and regulators about drugs that weaken bone. Several such drugs have been studied. The adverse effects of these drugs on the quality of bone formed under their influence fall into several categories: production of abnormal matrix which lacks normal lamellar structure often referred to as “woven” bone; impairment of mineralization (or osteomalacia in the extreme case), as seen with etidronate; and alteration of mineral structure, of which the fluoroapatite in fluoridetreated bone is a prime example. Fortunately, all these problems are readily detected by adequate animal testing, as described below.

V. ANIMAL TESTING A. General Safety Testing For any new drug, systematic toxicological evaluation must be carried out, normally in at least two species, prior to testing in humans. The duration of this animal safety testing depends on the period of anticipated human exposure. Thus, for singledose studies or short periods of exposure in humans, animal safety testing for 30 to 90 days may be sufficient, whereas longterm exposure of humans may require 1 to 2 years of prior animal safety testing. Longterm safety testing is usually carried out at doses up to 10 times the maximum anticipated human exposure. Blood and urine tests are performed at specified time

535 FIGURE 1 New drug development. In the US, human testing is carried out under an “Investigational New Drug Exemption” or IND, granted by the Food and Drug Administration. The application for permission to market a drug is called a “New Drug Application” or NDA. Diagram adapted with permission from M. Lumpkin, FDA, personal communication.

536

HENRY BONE

points for evidence of toxicity. The animals are sacrificed at designated intervals and at the end of the study in order to permit examination of the organs. In addition to the basic toxicological testing, additional studies are performed to evaluate drugs for their carcinogenic and teratogenic potential. These are also long-term studies with well-specified endpoints. It is important that the testing conditions bear a reasonable resemblance to human exposure, albeit at higher doses. In other words, the schedule of exposure should take the intended clinical use into account.

B. Testing for Bone-Specific Adverse Effects In the specific case of drugs (other than estrogens) intended for the prevention or treatment of osteoporosis, special bone testing is required. Such testing includes histological examination and histomorphometry to determine whether abnormalities of bone quality are produced under the influence of the drug, and to evaluate the drug’s effects on bone remodeling. To assure that the strength – mass relationship is preserved, bone strength is evaluated by mechanical testing of bone from sacrificed animals that have been treated with the drug at optimal and excess dosages for specified periods of time. Bone healing is evaluated invivo by observing experimental fractures under controlled conditions. This advanced stage of animal testing is often completed in parallel with early clinical studies.

VI. HUMAN TESTING After animal safety testing, human testing of new drugs generally follows a well-established sequence. In phase I, a range of tolerability is defined and pharmacokinetic characteristics are described, typically in normal volunteers. In phase II, clinical efficacy and tolerability are demonstrated with the development of a dose – response relationship for efficacy. Often, intermediate or surrogate endpoints are used in phase II. Based on phase II results, a dose or restricted range of doses is selected for phase III evaluation, which is generally longer-term and larger-scale, providing the “pivotal” studies upon which registration is based. The desirable characteristics of the dose(s) ultimately selected for clinical use include full efficacy and good tolerability. Generally, it is desirable for antiosteoporotic drugs to be on the plateau of the efficacy dose – response curve in order to deliver the full benefit under most conditions, provided that such dosing is not limited by adverse effects.

A. Phase I Following completion of the basic toxicological testing, human testing can begin. As noted, short-term human studies

normally begin before completion of the full long-term toxicological program. Phase I testing is intended to evaluate the tolerability of the drug over a wide range of doses and to characterize the drug’s absorption, metabolism, and excretion. Phase I testing is typically performed in normal volunteers, but if it is anticipated that age, sex, or racial characteristics of the patient population may have a major effect on the drug’s pharmacokinetics, a subject group resembling the intended target patient population may be selected. Thus, one might either perform phase I testing of a drug for osteoporosis in healthy young men, in postmenopausal women, or in both groups. Paradoxically, some drugs ultimately intended for the treatment of postmenopausal osteoporosis might be tested initially in men, because of the reluctance to expose elderly subjects or women of childbearing age to drugs which have not yet been fully characterized with regard to safety. Obviously, feminizing sex-steroid hormones would not be tested in men. Normally, the initial dose of the drug is small and is escalated in a stepwise manner until the limits of tolerability are reached. “No effect” and maximum tolerated doses are identified in this way. Pharmacodynamic testing can be included in phase I. In the case of drugs for osteoporosis, this would include measurements of effects on serum Ca, PO4 and parathyroid hormone (PTH) and bone turnover markers, for example.

B. Phase II Phase II testing is typically carried out in individuals affected by the condition of interest. Bone mineral density (BMD) entry criteria for osteoporosis treatment studies are similar to those described for phase III (below). The purpose of phase II is to determine whether the drug has the intended effect on patients with the target disease, to determine the tolerability range in patients and to characterize the dose – response relationships for the drug. Clinically useful dose – response curves for efficacy and adverse effects are the essential products of phase II. Phase IIa is carried out over periods of time which are relatively short but expected to be suffi cient for the detection of the effects of interest. For example, in phase IIa, biochemical markers of bone formation and resorption may be used as indicators of drug effect. Obviously, the observation period on treatment should be long enough to detect the effects of interest. Thus, effects of an antiresorptive drug on markers of bone resorption might be seen quickly, but effects on formation markers would appear more slowly. In contrast, the effects of a formation stimulant should be rapid. Based on phase IIa, dosages are selected for further testing in phase IIb, a somewhat longer-term, more definitive dose-finding process intended to characterize the “therapeutic window” for long-term treatment. The end points for drug effect in phase IIb are expected to be “harder” than the relatively short-term end points identified in phase

537

CHAPTER 65 Development and Evaluation of New Drugs

IIa. For instance, the US guidelines for the development of drugs for postmenopausal osteoporosis specify that phase IIb last at least 1 year and use bone mineral density (BMD) as the primary end point.

C. Phase III 1. TREATMENT OF OSTEOPOROSIS— GENERAL APPROACH The purpose of phase III is to provide longer-term, larger-scale proof of efficacy and safety. Typically, phase III is conducted with a restricted range of dosages, or perhaps only one, selected on the basis of phase II studies. The active dosage or dosages are compared with a control group in randomized trials. While phase I studies will usually include dozens of subjects and phase II studies may require hundreds, phase III often requires participation by thousands of patients. Selection of subjects for studies of osteoporosis treatment is crucial. For BMD end point trials, subjects with density measurements at the spine or hip that are 2.0 or 2.5 SD below the mean for young adult women are usually selected. The primary efficacy variable is usually lumbar spine BMD by densitometry, because that site generally gives the greatest separation between different treatment groups. Selection of subjects for fracture end point trials is more problematic. Subjects should have low bone mass and/or prior osteoporotic fracture. Selection of patients with the highest risk for future osteoporotic fractures (very low BMD, history of several fractures) minimizes the sample size and trial duration, but raises ethical issues in placebo-controlled trials. On the other hand, selection of subjects at lower risk can greatly increase the required number of patient-years of exposure. Either approach will require a similar total number of fractures in the control group, depending on the expected effect of the treatment. For development of anti-osteoporotic drugs, the minimum duration of phase III trials is usually specified at 3 years. The US and European requirements differ with regard to the end points upon which registration applications can be submitted. In Europe, at the time of writing, definitive anti-fracture efficacy must be demonstrated prior to initial registration of drugs for the treatment of postmenopausal osteoporosis. This current position of the CPMP [8] differs considerably from the recommendation of the WHO Study Group [9]. The latter organization based its recommendation on the fundamental physical relationship between the mass of an object and its strength, with heavy dependence on preclinical testing to assure that bone architecture and material properties are maintained. Thus, for drug substances which have been shown to

have no adverse effect on bone quality, the WHO guidelines recommend that registration be granted on the basis of a favorable effect on bone mass. The position of the FDA guidelines is intermediate between these two points of view. The current US guidance [10] takes into account prior guidelines and the extensive discussions of their revision, which occurred in 1993 – 1994 as well as subsequent experience with a number of drugs evaluated over the years between the 1985 and 1994 guidelines. Under the US guidance, drugs which have been thoroughly tested in animals and demonstrated to have no adverse effect on bone quality (including strength) in animals or humans may be approved for initial registration based on 3 year studies employing bone densitometry as the primary end point. However, such an approval requires demonstration of at least a favorable trend by interim analysis of the fracture rate in ongoing trials prior to the initial registration, typically 3 years into the trial. The definitive fracture rate studies must subsequently be completed. Once anti-fracture efficacy is confirmed, a stronger claim will be allowed. If preclincial or clinical studies indicate an adverse effect of a drug upon the quality of bone, registration based on bone densitometry effects is not allowed. If development is continued, initial registration must depend on demonstration of anti-fracture efficacy. This approach provides for expeditious drug evaluation with minimal risk that a marketed drug will fail to maintain a good relationship between bone mass and strength. 2. TREATMENT WITH ESTROGENS Criteria are modified somewhat in the case of estrogens. Because of the extensive experience with estrogens prior to the development of guidelines, and because there appear to be no adverse effects of estrogens on bone quality, evidence of anti-fracture efficacy is not required for registration of estrogen for prevention or treatment indications. This reasonable exemption largely accounts for the lack of recent fracture end point trials for estrogens in osteoporosis. 3. PREVENTION OF OSTEOPOROSIS Guidelines distinguish between treatment and prevention of osteoporosis as indications for therapy. Estrogens and related drugs may be registered initially for prevention, but other drug classes must generally achieve approval for treatment first. This serves to ensure that they actually do preserve the relationship between mass and strength. 4. PHASE IV This somewhat elastic term covers the period after initial registration. It sometimes includes new indications, but usually refers to supplementary studies, such as those in combination with other drugs, in special clinical situations,

538 or those with different end points, such as pharmacoeconomic or compliance studies.

D. Expanded Indications and Novel Agents Formal guidance was first developed for postmenopausal osteoporosis because of its prevalence. Regulatory authorities are now considering guidance for glucocorticosteroidinduced osteoporosis and osteoporosis in men. Additional areas of interest for guidelines development include osteoporosis syndromes other than postmenopausal, male, and steroid-related bone loss and the evaluation of anabolic agents such as parathyroid hormone as well as agents intended for short-term use. The general principles of development described above will always apply, including the need for rigorous preclinical testing, thorough dose-finding and adequate clinical observation periods. Although the development of anabolic and other novel agents, such as PTH, will employ these basic principles, there will doubtless be specific variations. The formulation of guidelines for the development of such agents will be a work in progress over the coming years, and the evaluation of the various possible drug combinations will be vary challenging indeed. Expansion of the osteoporosis indications beyond the scope of postmenopausal women is somewhat more straightforward than the initial registration. In most cases, drugs would be registered initially for postmenopausal osteoporosis. In such a situation, the demonstration of a favorable effect on bone mass may suffice for registration for the later additional indications, because the response of bone to treatment is generally qualitatively consistent. Unless there is a specific reason for concern about the relationship of mass and strength in a particular application, trials with bone density end points should be sufficient for initial registration in expanded indications, such as osteoporosis in men. Stronger claims will presumably be allowed if anti-fracture efficacy is demonstrated in the additional indications.

HENRY BONE

E. The End Product of Drug Development The majority of new drugs entering clinical trials fail prior to completion of phase III. However, at the end of a well organized, successful development program for a new drug for the treatment and prevention of osteoporosis, the preclinical studies will have demonstrated a proper relationship between increases in bone mass and increases in bone strength, as measured in experimental fracture studies; the dose – response relationships will be very well characterized; bone mass will be significantly increased or protected in humans; and ultimately, anti-fracture efficacy will be demonstrated.

References 1. C. Rich and P. Ivanovich, Response to sodium fluoride in severe primary osteoporosis. Ann. Intern. Med. 63, 1069 – 1074 (1965). 2. C. H. Turner, L. P. Garetto, A. J. Dunipace, W. Zhang, M. E. Wilson, M. D. Grynpas, D. Chachra, R. McClintock, M. Peacock, and G. K. Stookey, Fluoride treatment increased serum IGF-1, bone turnover, and bone mass, but not bone strength, in rabbits. Calcif. Tissue Int. 61, 77-83 (1997). 3. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. Chao, H. W. Wahner, J. M. Muhs, S. L. Cedel, and L. J. Melton, Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802-809 (1990). 4. L. Flora, G. S. Hassing, G. G. Cloyd, J. A. Bevan, A. M. Parfitt, and A. R. Villanueva, The long-term skeletal effects of EHDP in dogs. Metab. Bone Dis. Relat. Res. 3, 289-300 (1981). 5. H. G. Bone, J. E. Zerwekh, F. Britton, and C. Y. Pak, Treatment of calcium urolithiasis with diphosphonate: Efficacy and hazards. J. Urol. 121, 568-571 (1979). 6. National Institutes of Health Consensus Development Conference “Osteoporosis Prevention, Diagnosis and Therapy” March 27– 29, 2000. 7. R. D. Wasnich and P. D. Miller, Antifracture efficacy of antiresorptive agents are related to changes in bone density. J. Clin. Endocrinol. Metab. 85, 231-236 (2000). 8. Committee for Proprietary Medicinal Products (CPMP), Efficacy Working Party Note for Guidance on Involutional Osteoporosis in Women CPMP/EWP/552/95, 1999. 9. World Health Organization (WHO) Study Group “Guidelines for Preclinical Evaluation and Clinical Trials in Osteoporosis,” World Health Organization Technical Series 843, 1998. 10. Food and Drug Administration (FDA), “Preclinical and Clinical Evaluation of Agents Used in the prevention or Treatment of Postmenopausal Osteoporosis,” April 1994.

CHAPTER 66

Evidence-Based Decision Making in Osteoporosis ANN B. CRANNEY GORDON H. GUYATT

I. II. III. IV. V.

Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada K1Y 1J8 Department of Medicine and Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada L8N 3Z5

VI. VII. VIII. IX.

Introduction Clinical Decision Making in Osteoporosis Collecting the Evidence Strength of the Evidence/Study Design Magnitude of the Treatment Effect

I. INTRODUCTION

that we search for the highest level of evidence available, integrate this evidence with our clinical experience and judgement, and acknowledge the value judgements implicit in moving from evidence to action.

Clinicians make decisions daily on whether to treat patients with osteoporosis. In both the prevention and the treatment of osteoporosis, a patient may be required to take a medication for many years to prevent a hip or vertebral fracture. Unfortunately, these treatments may involve lifestyle changes and unpleasant side effects. The treatment of osteoporosis has become more complex, given the increased number of options available. In addition, each therapy has its unique risk – benefit profile. One method of handling this complex decision making process in the clinical setting is to apply the principles of evidence-based medicine [1]. Evidence-based medicine suggests a hierarchy of evidence beginning with systematic clinical observations with strong study designs developed to reduce bias, through systematic clinical observations with weaker designs, to generalization from physiological studies and unsystematic clinical observations. Evidence-based medicine recommends

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Osteoporosis Outcomes and Their Relevance to Patients Threshold Level of Treatment Consideration of Patient Values Conclusions References

II. CLINICAL DECISION MAKING IN OSTEOPOROSIS Clinical decision making involves identifying the alternative courses of action, evaluating the evidence regarding the outcomes associated with each alternative, and considering the impact of these outcomes on the patient’s life. For most individual decisions, the clinician relies on an intuitive gestalt about the evidence, and an instinctive sense of the impact on the outcomes. Clinicians are confronted with more general clinical questions that apply to whole groups of patients. These questions include “What treatments do I recommend to a

539

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

540 patient with a prevalent vertebral fracture?” and “For which patients with osteoporosis defined by bone densitometry do I recommend treatment?” For these questions, a more rigorous clinical decision making process should occur. This process involves collecting the evidence, weighing the alternate courses of action, and considering the impact of each outcome on the patient. This process should ideally include a consideration of the patient values in the treatment decision. For example, your patient may place a high value on avoiding a hip fracture, and would accept the increased risk of a venous thromboembolism associated with hormone replacement therapy. Individual clinicians cannot themselves conduct the time- and energy-intensive process of systematically collecting and reviewing the evidence bearing on these important clinical issues. What individual clinicians can do, however, is to understand this process and be able to distinguish between evidence summaries and recommendations that are likely to give an unbiased picture of the evidence and those that are not. There are a number of sources that facilitate this process including secondary journals such as the ACP Journal Club and Evidence-Based Medicine, which review a large number of journals which have already passed a methodological filter.

III. COLLECTING THE EVIDENCE An accurate assessment of the impact of a diagnostic test or treatment requires a rigorous, systematic search for the evidence addressing a clinical problem. The highest level of evidence comes from a systematic overview [2]. Systematic denotes overviews that meet the following five standards: The overview addresses a focused clinical question, uses appropriate criteria to select studies for inclusion, conducts a comprehensive search, appraises the validity of the individual studies, and applies appropriate statistical methods to summarize the data. Treatment recommendations based on a review that meets these five criteria are stronger than those that do not. For example, we recently summarized the evidence suggesting that alendronate reduces the incidence of vertebral and nonvertebral fractures in postmenopausal women [3]. In this overview, we selected randomized control trials (RCTs) of postmenopausal women and succeeded, through collaboration with the drug’s manufacturers, in obtaining complete data on both published and unpublished studies. We pooled data widely across patient subgroups (mild osteoporosis versus established osteoporosis with fractures), different drug doses and duration of therapy. Wider pooling across subgroups increases the precision of our estimate and enhances the generalizability of the results. We then tested

CRANNEY AND GUYATT

whether the effects differed across patient subgroups such as treatment versus prevention, duration of therapy, cointerventions, and methodological quality. Summary treatment effects for fractures and bone density were calculated using a random effects model. The random effects model assumes that we are interested in all trials that are conducted and it incorporates between-study variability in the error variance, resulting in wider, more conservative estimates. A fixed-effects model incorporates within-study but not the between-study variability when estimating the error of measurement. One limitation of using random effects model is that it may place greater weight on smaller trials and result in an implausibly wide confidence interval which was noted in our raloxifene meta-analysis [10]. An evidence-based clinician first looks for a systematic review of the effects of treatments for osteoporosis. The estimate of a treatment effect from a systematic review is likely to be less biased than that of an unsystematic review.

IV. STRENGTH OF THE EVIDENCE /STUDY DESIGN The strength of inferences concerning the magnitude of treatment effects that we make from a systematic overview of randomized trials will depend on the consistency of the results from study to study. When different studies addressing a similar question yield different estimates of efficacy (heterogeneity of results) then it is important to account for this heterogeneity. It may be explained by differences in patient population (osteoporotic versus osteopenic), methodological quality of the trials, or concurrent treatments (calcium/vitamin D) used. An alternative to each of these explanations is that the play of chance is responsible for heterogeneity in study results. If the variation in studies is not explained, then any inferences they make about treatment effect will be weaker. Therefore, systematic reviews without statistically significant heterogeneity are ranked higher than those with statistically significant heterogeneity. In our meta-analysis of alendronate, we did not find heterogeneity of the treatment effect across trials for the outcome of vertebral fracture. The magnitude of the effect was similar across both treatment and prevention studies. However, in the etidronate meta-analysis there was significant heterogeneity across studies for outcomes of lumbar spine and femoral neck bone mineral density (BMD). When we explored this heterogeneity further, we discovered that one trial that did not use an intention-to-treat analysis showed a much larger treatment effect. When the analyses were repeated without this trial, we did not find any important heterogeneity [3].

541

CHAPTER 66 Evidence-Based Decision Making

In some cases, a systematic search may produce only studies of weak design. The potential for bias is much greater in cohort and case – control studies (observational studies in which patients receive treatment at their own, or their physicians’ discretion) than in randomized trials. Randomized trials minimize bias through allocation concealment, the use of a placebo and double blinding. Allocation concealment protects the assignment sequence until the point of randomization of the participant to the treatment or control group. In other words, the person deciding on whether the potential participant is, or is not, eligible is unaware of whether the patient will be allocated to treatment or control. Allocation concealment protects against the preferential enrollment of specific patients with a different prognosis into one treatment group over another [4]. Blinding addresses whether participants, investigators and those responsible for measuring the outcome (that is measuring bone density, or determining whether or not a vertebral fracture has occurred) are blind to whether patients are receiving experimental or control treatment. Trials of lower methodological quality have been shown to yield larger treatment effects. There are now standardized reporting criteria for metaanalysis of RCTs [5]. The Heart and Estrogen/Progestin Replacement Study (HERS), which was designed to evaluate the efficacy of hormone replacement therapy for secondary prevention of coronary heart disease (CHD) in postmenopausal women, is an example of how a RCT may contradict findings from previous observational studies [6]. Numerous observational studies consistently found decreased rates of CHD in women who took hormone therapy (HRT) compared to nonusers [7]. However, HERS was the first RCT with a sample size of 2763 women with coronary artery disease, large enough to reliably assess the impact of HRT on myocardial infarction and death. The trial showed no effect of HRT on the overall rate of CHD events in women treated with HRT. While randomized trials provide the most valid estimates of true effect of an intervention, they report average treatment effects. Clinicians must decide how to extrapolate the results to individual patients. Recommendations from overviews combining observational studies will be much weaker. Randomized trials often yield smaller treatment estimates of efficacy than observational studies and observational studies provide weaker levels of evidence (Table 1) [4,8]. When we evaluate whether the treatment of osteoporosis can prevent hip fractures, we find randomized trials that demonstrate hip fracture reduction for alendronate (Grade A). For hormone replacement the only studies that show statistically significant hip fracture reductions are observational in design. Thus any recommendation for use of HRT to prevent fracture would be Grade C.

TABLE 1

Proposed Levels of Evidence for Treatment Recommendation in Osteoporosis

Strength of the study design Level A: RCTs, no clinically important or statistically significant heterogeneity Level B: RCTs, significant heterogeneity Level C: Observational studies Strength of the outcome Level A: Major importance to patients (pain, functional limitation, fracture of long bones) Level B: Of questionable importance to patients (vertebral fractures) Level C: Important to patients only through relation to other variables (bone density) Precision of the outcome Level A: Entire confidence interval below threshold number needed to treat Level B: Confidence interval overlaps threshold number needed to treat

V. MAGNITUDE OF THE TREATMENT EFFECT When deciding whether to recommend a therapy to patients, we should consider the magnitude of the treatment effect. In osteoporosis, we evaluate the extent to which a therapy reduces the risk of vertebral fractures or nonvertebral fractures. The size of the treatment effect will play an important role in the decision to administer therapy and guide in the selection of therapy. Authors can present the magnitude of the treatment effect in different ways. These include the relative risk which is the ratio of the risk of an adverse event in treated patients to the risk of an adverse event in the untreated patients, the relative risk reduction (1-relative risk), the absolute risk reduction, which is the difference in the absolute risk of the adverse outcome between treatment and control groups. Another way of presenting the treatment effect that has gained widespread acceptance is the the number need to treat (NNT) which is the inverse of the absolute risk reduction [9]. The NNT reflects the baseline risk and the relative risk reduction and therefore becomes smaller as the patients’ risk of an adverse outcome rises and may be very large when patients are at very low risk. For example, the number of patients with normal or near normal bone density whom we have to treat for 2 years with alendronate to prevent one vertebral fracture is almost 2000. To take another example, a 60-year-old woman whose femoral neck bone density is 2.5 SD below the mean has an approximate 6.85% risk of nonvertebral fracture over the next 2 years. Results from our meta-analysis suggest that over this 2-year period, the risk of nonvertebral fractures is reduced by 49% with alendronate (95% CI 0.38 to 0.69). Therefore, with alendronate, the number needed to treat to

542

CRANNEY AND GUYATT

prevent one nonvertebral fracture is 24. However, in patients that are at low risk of a nonvertebral fracture, the number needed to treat is 80. One must treat much larger numbers of patients to prevent one hip fracture because of the much lower prevalence. In high-risk patients, the NNT to prevent one hip fracture with alendronate is 237. The average NNT reported in a systematic review or RCT may not be directly applicable to the individual patient (due to differences in baseline risk) and the clinician is faced with deciding what is their patient’s baseline risk of the target outcome and whether their patient’s RRR is likely to be different from the group average. The number needed to harm (NNH) is an expression of the number of patients who would need to receive an intervention to cause one additional adverse event. The NNH is the inverse of the absolute difference in adverse event rates between the experimental and control arms. A meta-analysis of raloxifene for postmenopausal osteoporosis demonstrated an increased risk of hot flashes in data pooled across four trials. The relative risk of hot flashes was approximately 1.5. The proportion of control group women with hot flashes in these trials varied from 6 to 27%. Thus, the percentage of women in whom raloxifene would be responsible for hot flashes might vary from 3% in a low-riskgroup to 13.5% in a high-risk group. This translates into a NNH of 33 for low-risk women and an NNH of approximately 7 for high-risk women. In women at an intermediate risk of 12.5% the NNH would 16 (Table 2) [10]. Treatment recommendations will be strengthened if they incorporate: (1) risk of target adverse outcomes, which patients can anticipate if they are not treated, (2) the extent to which treatment can reduce the risk of target adverse outcomes, (3) the adverse effects and costs that are a consequence of treatment, and (4) the relative values placed on avoiding the target outcome or avoiding the adverse events of therapy.

morphometric vertebral fractures, both of which are surrogate outcomes [11]. Use of surrogate outcomes of bone density would be reasonable if we could demonstrate a direct relationship between BMD and fracture reduction. There is substantial evidence that there is a moderate association between bone density and fracture [12]. However, previous experience with fluoride has shown that increases in bone density may result in an increase in nonvertebral fracture rates [13]. Similarly, treatment with medications which are less potent anti-resorptive agents such as raloxifene results in increases of lumbar spine BMD of 2.6% over 2 years. Applying results from previous observational trials this increase would predict a fracture reduction of 12%. However, the actual reduction in vertebral fracture incidence in a 3-year randomized trial was 30 – 50% [14]. A study of calcium and vitamin D in the elderly showed almost no change in bone density, but a reduction in fracture risk of approximately 50% [15]. We know that changes in bone quality and suppression of bone turnover are not captured in bone density measurements, but may be important determinants of fracture reduction. Therefore, basing treatment decisions on the effect of an intervention on bone density alone is fraught with risk. Outcomes such as clinical fracture reduction, decreased pain and improved quality of life/function are clinically more relevant to patients than increases in bone density. Fortunately for patients, but unfortunately for clinical trialists hip fractures are rare events, especially in populations that are at lower risk. As a result, morphometric vertebral fractures or bone density are used as surrogate measures in many trials. Incident vertebral fractures are associated with increased back pain and disability. However, most clinical trials have not directly measured improvements in back pain, function and quality of life [16]. Nevitt et al. recently reported that treatment of postmenopausal women with prevalent vertebral fractures with alendronate reduced the number of days of bed disability caused by back pain [17]. There are a number of disease-specific quality of life instruments available which have been validated and are included as outcomes in recent osteoporosis clinical trials [18 – 22]. These include the Osteoporosis Assessment Questionnaire (OPAQ), the Osteoporosis Quality of Life Assessment

VI. OSTEOPOROSIS OUTCOMES AND THEIR RELEVANCE TO PATIENTS Until recently, the emphasis in randomized controlled trials has concentrated on bone mineral density and

TABLE 2 Summary Measure of Benefits and Adverse Events for Raloxifene Outcome event rate (%)

Control event rate (%)

Experimental reduction/increase (%)

Absolute risk

Number needed to

treat (NNT) Vertebral fractures

9.0

5.3

3.7

27

Hot flashes

7.8

12.5

6.25

NNH Number need to harm 16

543

CHAPTER 66 Evidence-Based Decision Making

Questionnaire (OQLQ) and the Quality of Life Questionnaire for the European Foundation for Osteoporosis. The OQLQ, for example, is a 30-item questionnaire which has five domains: emotions, symptoms, physical function, activities of daily living, and leisure. These issues lead us to suggest another way of evaluating the strength of recommendations coming from trials of osteoporosis therapies. Grade A recommendations come from trials that assessed long-bone fractures or measure pain and disability directly, Grade B recommendations focus on morphometric vertebral fractures and Grade C on bone density (Table 1).

VII. THRESHOLD LEVEL OF TREATMENT Inclusion of the adverse effects of therapy implies that there is a specific number or threshold above which it is no longer valid to treat. The threshold NNT is the dividing line between when treatment is warranted (the NNT is low enough that the benefits outweigh the costs and risks) and when it is not. If the range of number needed to treat exceeds this threshold, the treatment recommendations lose strength. Our confidence in a therapy would be strengthened if the tradeoff between benefits and adverse consequences of therapy is such that the number needed to treat falls below the threshold value above which treatment is no longer valid. This depends in part on the precision of the estimate of the treatment effect. We can use our meta-analysis of alendronate as an example and decide that our threshold number needed to treat for vertebral fractures is 35. The average 50-year-old woman has an approximately 15% lifetime risk of fracturing her spine. With alendronate the pooled relative risk is 0.52 (95% CI 0.43 to 0.65). This means that instead of 15 of 100 women having a fracture it will be 7 of 100, a difference of 8 in 100. We must therefore treat 13 women for 30 years to prevent one vertebral fracture which is well below our threshold. In this case, the confidence intervals for the relative risk reduction are narrow 0.43 to 0.65, so even the upper boundary results in a 35% reduction in RR of new vertebral fractures which still results in an NNT below our treatment threshold. In contrast for etidronate we obtained a pooled relative risk for reduction of morphometric vertebral fractures of 0.63, 95% CI (0.44 to 0.92). This suggests that etidronate reduces the relative risk by 37% [23]. However the confidence interval is wide and the boundary representing the smallest effect shows a relative risk reduction of only 8%. If the 8% relative risk reduction represented the underlying truth about etidronate treatment, we would need to treat 83 average 50-year-old women for 30 years to prevent a single fracture which is clearly above our defined threshold.

Under these circumstances, few women would be likely to choose etidronate treatment. A recommendation to treat gains strength when the precision of the estimate of the treatment effect is such that the highest possible number needed to treat is lower than the threshold above which we would not treat (Table 1).

VIII. CONSIDERATION OF PATIENT VALUES The decision making process involves the disclosure of the risks and benefits of the therapeutic alternatives, the exploration of the patient’s values about the therapy and health outcomes and the actual decision. A 65-year-old woman with a strong family history of osteoporosis and a prevalent vertebral fracture may elect to take alendronate despite the risk of esophageal ulceration once she is aware of her personal risk of further osteoporotic fractures. Patients vary in their desired level of involvement in the decision making process. Nevertheless, if decisions are to reflect individual patient values, the clinician must evaluate patients’ preferences about therapeutic options and potential health outcomes.

IX. CONCLUSIONS Evidence-based medicine provides a framework to facilitate unbiased estimates of the magnitude of benefits and adverse consequences of osteoporosis therapy. Evidencebased approaches encourage clinicians to be more quantitative, considering a patient’s baseline risk of fracture and the magnitude of the reduction in relative risk of fracture when considering the impact of a decision to treat. Finally, evidence-based practitioners recognize how patients’ values influence the decisions as they translate evidence to action.

References 1. P. Glaziou, G. H. Guyatt, A. L. Dans, A. F. Dans, S. Straus, and D. L. Sackett, Applying the results of trials and systematic reviews to individual patients. Evidence-Based Med. 3165 – 3166 (1998). 2. A. D. Oxman, D. J. Cook, and G. H. Guyatt for the Evidence-Based Medicine Working Group, Users’ guide to the medical literature VI. How to use an overview. JAMA 272, 1367 – 1371 (1994). 3. A. Cranney, G. H. Guyatt, and A. Willan et al., “Meta-analysis of Alendronate for the Treatment of Postmenopausal Osteoporosis.” Endo. Rev. (submitted). 4. D. Moher, B. Pham, A. Jones et al., Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancer 22, 609 – 613 (1998). 5. CONSORT: An evolving tool to help improve the quality of reports of randomized controlled trials. Consolidated Standards of Reporting Trials. JAMA 279(18), 1489 – 1491 (1998).

544 6. S. Hulley, D. Grady, and T. Bush et al., for the HERS Research Group, Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. JAMA 280(7), 605 – 613 (1998). 7. F. Grodstein, M. J. Stampfer, J. E. Manson et al., Postmenopausal estrogen and progestin use and the risk of cardiovascular disease. N. Engl. J. Med. 335, 453 – 461 (1996). 8. T. C. Chalmers, P. Celano, H. S. Sacks, and H. Smith Jr., Bias in treatment assignment in controlled clinical trials. N. Engl. J. Med. 309, 1358 – 1361 (1983). 9. A. Laupacis, D. L. Sackett, and R. S. Roberts, An assessment of clinically useful measures of the consequences of treatment. N. Engl. J. Med. 318, 1728 – 1733 (1988). 10. A. Cranney, G. Guyatt, N. Krolicki et al., “Meta-analysis of Raloxifene for the Prevention and Treatment of Postmenopausal Osteoporosis,” submitted. 11. H. C. Bucher, G. H. Guyatt, D. C. Cook, A. Holbrook, and F. A. McAlister, Users’ guides to the medical literature. XIX. Applying clinical trial results. A. How to use an article measuring the effect of an intervention on surrogate end points. Evidence-Based Medicine Working Group. JAMA 282(2), 771 – 778 (1999). 12. D. Marshall, O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312, 1254 – 1259 (1996). 13. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. Chao, H. W. Wahner, and J. M. Muhs, The effect of fluoride treatment on fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). 14. B. Ettinger, D. M. Black, and B. H. Mitlak et al., Effects of raloxifene on bone mineral density, serum cholesterol concentrations and uterine endometrium in postmenopausal women. N. Engl. J. Med. 282, 637 – 646 (1999).

CRANNEY AND GUYATT 15. M. C. Chapuy, M. E. Arlot, and F. Doboeuf et al., Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 16. M. C. Nevitt, B. Ettinger, and D. M. Black et al., The association of radiographically detected vertebral fractures with back and function: A prospective study. Ann. Intern. Med. 793 – 800 (1998). 17. M. C. Nevitt, D. Thompson, and D. M. Black et al., Effect of alendronate on limited-activity days and bed-disability days caused by back pain in postmenopausal women with existing vertebral fractures. Arch. Intern. Med. 160, 77 – 85 (2000). 18. D. J. Cook, G. H. Guyatt, and J. D. Adachi et al., Quality of life issues in women with vertebral fractures due to osteoporosis. Arthritis Rheum. 36, 750 – 756 (1993). 19. P. Lips, C. Cooper, and D. Agnusdei et al., Quality of life in patients with vertebral fractures: Validation of the quality of life questionnaire of the European Foundation for Osteooporosis. (QUALEFFO). Osteoporosis Int. 10(2): 150 – 60 (1999). 20. S. E. Gabriel, T. S. Kneeland, L. J. Melton 3rd, M. M. Moncur, B. Ettinger, and A. N. Tosteson, Health-related quality of life in economic evaluations for osteoporosis: Whose values should we use? Med. Decis. Making 19(2), 141 – 198 (1999). 21. A. G. Randell, N. Bhalerao, T. V. Nguyen, P. N. Sambrook, J. A. Eisman, and S. L. Silverman, Quality of life in osteoporosis: Reliability, consistency, and validity of the Osteoporosis Assessment questionnaire. J. Rheumatol. 25(6), 1171 – 1179 (1998). 22. C. R. Kessenich, G. H. Guyatt, and C. J. Rosen, Health-related quality of life and participation in osteoporosis clinical trials. Calcif. Tissue Int. 62(3), 189 – 192 (1998). 23. A. Cranney, G. Guyatt, N. Krolicki, V. Welch, L. Griffith, J. D. Adachi, et al., “A Meta-analysis of Etidronate for the Treatment of Postmenopausal Osteoporosis.” Osteoporos. Int. (in press) (2001).

CHAPTER 67

The Role of Calcium in the Treatment of Osteoporosis BESS DAWSON-HUGHES

Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, Massachusetts 02111

III. Calcium Sources References

I. Calcium Absorption II. Calcium and Bone

initiate the synthesis of separate calcium and phosphorus transport proteins. Passive diffusion involves movement of calcium between enterocytes via the paracellular shunt pathway. The amount of calcium absorbed by passive diffusion is proportional to the luminal:serosal calcium concentration gradient. Absorption of calcium by solvent drag also utilizes the paracellular shunt pathway and occurs after the ingestion of an osmotic load. The relationship between calcium intake and absorbed calcium is nonlinear (Fig. 1) [1]. The change in slope at an intake of about 500 mg per day reflects the intake level at which active transport, the major mechanism of absorption at low intake levels, becomes saturated. The increase in absorbed calcium, as intake exceeds 500 mg per day, occurs by passive diffusion. In most individuals, the proportion of calcium absorbed by solvent drag is small. With dietary restriction, absorbed calcium declines but the fraction of a calcium load absorbed increases [2 – 4]. This is illustrated in Fig. 2 [4]. Calcium restriction induces a subtle decrease in blood ionized calcium concentration, a rise in parathyroid hormone (PTH) secretion, stimulation of renal, 1,25(OH)2D production, and increased calcium absorption by active transport. This intestinal adaptation occurs within 1 week of calcium restriction and stabilizes after 2 weeks in healthy adult women [5].

Calcium intake is one of many environmental or lifestyle factors that influences bone loss and risk of osteoporosis. For the individual, the benefit to be expected from increasing calcium intake will depend upon absorptive status, self-selected calcium intake, estrogen status, heredity, and other variables. In this chapter I will focus on the physiology of calcium absorption and the role of calcium in preventing bone loss and osteoporotic fractures. Calcium alone is not considered adequate treatment for patients with established osteoporosis. It is, however, an essential adjunct to treatment with antiresorptive agents and drugs that stimulate bone formation. The specific role of calcium as an adjuvant therapy will be considered in subsequent chapters that address individual drug therapies.

I. CALCIUM ABSORPTION Calcium absorption occurs by several different mechanisms, active transport, passive diffusion, and solvent drag. Active transport involves movement of calcium from the intestinal lumen into enterocytes and then out on the serosal side. This process requires the active vitamin D metabolite, 1,25-dihydroxyvitamin D (1,25(OH)2D), a compound that acts on enterocyte nuclear receptors to

OSTEOPOROSIS, SECOND EDITION VOLUME 2

545

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

546

FIGURE 1

Calcium absorption from diet (g/day) as a function of dietary calcium intake. Reprinted with permission from Heaney et al. [1].

Calcium absorption is influenced by season, race, and, in women, by estrogen status. In the temperate zone, absorption is greater in summer than in winter [6,7], probably because of seasonal changes in 25-hydroxyvitamin D levels. Abrams et al. [8] found calcium absorption to be greater in black than in white girls. Among adult women, we found similar levels of calcium absorption but higher concentrations of 1,25(OH)2D in blacks than in whites, suggesting that black women may have a gut resistance to the action of active vitamin D [5]. Estrogen stimulates 1hydroxylase activity in the kidney and also appears to have a direct positive effect on intestinal calcium absorption [9]. Thus, in women, loss of estrogen at menopause causes a decline in calcium absorption efficiency. Calcium absorption declines with age in men and women [10 – 12]. Several mechanisms may be involved. Ebeling et al. [13] postulated an age-related resistance to the action of 1,25(OH)2D at the gut after finding an age-related

FIGURE 2

Fractional calcium absorption as a function of calcium intake. Reprinted with permission from Heaney et al. [4].

BESS DAWSON-HUGHES

decline in the 1,25(OH)2D receptor (VDR) concentration in human duodenal mucosal biopsy specimens. Consistent with this are the findings of several investigators [13,14] that blood levels of 1,25(OH)2D in women rise with age. There is no concensus that gut resistance to 1,25(OH)2D accounts for declining calcium absorption. Others propose that calcium absorption declines as a result of an age-related decline in 1,25(OH)2D concentration. Two groups have reported an age-related decrease in renal production of 1,25(OH)2D in response to infusion of a single dose of parathyroid hormone (PTH) (1 – 34) [15,16]. Several have reported an age-related decline in fasting serum 1,25(OH)2D concentrations in women [12,16,17]. Orwoll et al. identified no association between serum 1,25(OH)2D concentration and age in men [18]. Examination of the renal 1-hydroxylase response to varying doses of a stimulant, such as PTH(1 – 34), in young and older individuals on calcium and vitamin D replete diets would be one approach to distinguishing the influences of aging on ambient 1,25(OH)2D concentrations and on renal 1,25(OH)2D production capacity. Little is known about whether gender influences calcium absorption efficiency. From one large study that included adult men and women [11], it is difficult to draw conclusions about gender because calcium intake levels were not reported. Food intake surveys generally reveal that men consume more calcium than women, and this would be expected to create gender differences in both total and fractional calcium absorption. Calcium absorption capacity is linked to bone status. Calcium absorption is lower in osteoporotics than in agematched controls [10,19,20]. Nordin et al. [21] demonstrated a positive correlation between rate of calcium absorption and vertebral bone density in osteoporotic women. Francis et al. [22] found blunted calcium absorption responses to vitamin D replacement in vitamin D-deficient osteoporotics compared to those of similarly deficient nonosteoporotics. Krall and Dawson-Hughes reported a significant correlation between whole body 47Ca retention, an index of calcium absorption, and rates of bone loss in healthy postmenopausal women [7]. There is some recent evidence that calcium absorption has a hereditary component. Alleles of the VDR were originally reported by Morrison et al. [23] to be associated with bone mineral density in adult twins and in unrelated postmenopausal women (see Chapter 26). Women with alleles designated bb had higher bone mineral density than women with the alleles designated BB. That association appears to be stronger in women with very low calcium intakes [24]. Because the 1,25(OH)2D receptor plays a central role in intestinal calcium absorption, this led to the hypothesis that VDR alleles influence bone at least in part by affecting calcium absorption efficiency at low to moderate calcium intake levels, those at which the active transport mode of

CHAPTER 67

547

The Role of Calcium in the Treatment of Osteoporosis

teocalcin concentration by 9% in the men and 14% in the women throughout the 3-year intervention period. Several studies have shown similar reductions in postmenopausal women treated with calcium alone [31,32]. A smaller reduction was seen in another study [33], probably because the mean calcium intake of the women at entry was higher.

B. Bone Loss The most convincing information on the bone density response to treatment with calcium comes from randomized

FIGURE 3

Changes in fractional 45Ca absorption and circulating 1,25 (OH)2D after calcium restriction, adjusted for differences in initial (high calcium intake) values expressed as a percentage of the initial values. The star indicates a significant difference between the genotypes (P  0.016). Reprinted with permission from Dawson-Hughes et al. [25]. Copyright © 1995 The Endocrine Society.

absorption predominates. This hypothesis gained support with the demonstration that postmenopausal women with BB and bb alleles had similar levels of fractional calcium absorption on high (1500 mg) calcium intakes but that women with the BB genotype had blunted increases in fractional calcium absorption when calcium intake was lowered to 300 mg per day (Fig. 3) [25]. This study suggests the presence of a hereditary – environmental interaction. The Fok I polymorphism at the VDR transcription site also predicts calcium absorption and BMD in children [25a].

II. CALCIUM AND BONE A. Bone Remodeling Calcium is a substrate for bone mineralization; it also has an antiresorptive effect on bone. Recent evidence indicates that bone density and the bone remodeling rate are independent predictors of fracture [26]. Increases in calcium intake, within the range usually consumed, induce modest reductions in blood PTH concentration [27 – 29]. These, in turn, lower the bone remodeling rate. The magnitude of the change in remodeling induced by added dietary calcium is inversely related to the usual calcium intake; and it appears to be greater in the elderly, who as a group have higher remodeling rates than younger people with similar calcium intakes. The effects of supplementation with 500 mg of elemental calcium and 700 IU of vitamin D on serum concentrations of PTH and osteocalcin, a marker of bone turnover, in older men and women are shown in Fig. 4 [30]. In these subjects with a habitual calcium intake of about 750 mg per day, supplements lowered the mean serum os-

FIGURE 4 Mean  SEM laboratory values in placebo-(-----) and calcium-plus-vitamin-D-treated (—) subjects. At 6 months and thereafter, mean values for each analyte differed by treatment group at a significance level of P<0.005. Reprinted with permission from Dawson-Hughes et al. [30].

548 studies in postmenopausal women. Calcium, by reducing the bone remodeling rate, produces an increase in bone mineral density or a reduction in the rate of bone loss in the first 6 to 12 months of therapy. This bone remodeling transient was described originally by Frost [34] and confirmed subsequently by others. The increase in density that occurs with closure of the remodeling space cannot be used to estimate cumulative effects of long-term supplementation on bone mineral density. To predict long-term effects of added calcium, attention should be drawn to rates of bone change during the second and subsequent treatment years. Most published intervention trials were not designed and do not have the statistical power to detect effects of calcium within the second and subsequent treatment years that, while modest, will likely be significant over time. Nonetheless, it may be useful to examine separately the short-term (year 1) and longer-term (years 2 and beyond) effects of calcium in several randomized calcium trials. 1. EARLY POSTMENOPAUSAL WOMEN Several placebo-controlled calcium-intervention trials have been conducted in women within the first 5 years of menopause [28,31,33,35,36]. Mean dietary calcium intakes of the women ranged from 400 to 1150 mg per day and calcium supplement dosages also varied widely, from 500 to 2000 mg per day. In the Aloia study [35], both the placebo and the calcium group received 400 IU per day of vitamin D. At the spine, calcium had a positive effect on bone change in year 1 but no added effect in years 2 or 3. A similar pattern was seen in the Elders study [33]. At the proximal radius, a site rich in cortical bone, early postmenopausal women have demonstrated modest but consistently positive responses to calcium in the first year of treatment and beyond [28,31,33,35]. Only two investigators measured the effect of calcium on bone density change at the femoral neck [28,35]. Aloia et al. [35] found a sustained benefit from calcium (Fig. 5), whereas Dawson-Hughes et al. [28]

BESS DAWSON-HUGHES

did not. The difference can probably be attributed to the higher dose of supplemental calcium (1700 vs 500 mg per day) and to the 400 IU vitamin D supplement used in the Aloia study. 2. LATE POSTMENOPAUSAL WOMEN Two calcium intervention studies have been conducted in women 6 or more years after menopause [28,37]. In both, calcium had a positive influence at the spine in year 1 but not subsequently. At the proximal radius, 500 mg per day of calcium as the highly absorbable calcium citrate malate reduced bone loss in years 1 and 2, whereas a similar dose of calcium as carbonate did not. At the femoral neck, Dawson-Hughes et al. [28] found a sustained benefit from calcium only in the women with low dietary calcium intakes. In the Reid study [37], there was a positive trend for calcium at the femoral neck in year 1 only. There was, however, a cumulative effect on total body bone mineral density. 3. MEN The effect of calcium alone on bone mass has not been evaluated in men but two studies have evaluated the effects of calcium and vitamin D on rates of bone loss. DawsonHughes et al. found that in older men with usual calcium intakes of about 750 mg per day, supplementation with 500 mg of calcium and 700 IU of vitamin D significantly reduced bone loss from the spine, femoral neck, and total body during the first year and significantly reduced loss from the total body in years 2 and 3 [30]. Figure 6 illustrates the changes in BMD in 176 men and 213 women in that study [30]. In the other study in men, calcium and vitamin D supplementation with higher doses (1000 mg of calcium and 800 IU of vitamin D) had no significant shortor long-term impact on rates of bone loss in men who had usual calcium intakes of about 1200 mg per day [38]. These findings are consistent with the concept that exceeding a threshold calcium intake of about 1200 mg per day adds little additional value. The effect of stopping supplements on bone mineral density and bone turnover has recently been examined in men and women. The effects of supplementation on bone remodeling and bone density are lost within 2 years of stopping the supplements [39]. This suggests that an adequate intake of these nutrients must be sustained in order to retain the maximal benefit.

C. Fractures

FIGURE 5

Rates of change in bone mineral density of the femoral neck. Calcium supplementation had a sustained beneficial effect compared with that of placebo. Reprinted with permission from Aloia et al. [35].

Some information on the effect of added calcium on fracture incidence is now available (Table 1). Most of these studies were designed to detect changes in BMD and were far too small to determine the magnitude of the effect

CHAPTER 67

549

The Role of Calcium in the Treatment of Osteoporosis

Two studies have examined the effect of combined calcium and vitamin D supplementation on fracture incidence. Chapuy et al. [41] studied over 3000 very elderly institutionalized women who had an average of about 500 mg of calcium per day in their diets and had very low serum 25-hydroxyvitamin D levels. They were treated for 3 years with placebo or 1200 mg of calcium and 800 IU of vitamin D per day. The supplemented group had 30% fewer hip fractures and other nonvertebral fractures than did the women treated with placebo. In 389 men and women, age 65 and older, supplementation with 500 mg of calcium and 700 IU of vitamin D daily for 3 years significantly lowered nonvertebral fracture rates (relative risk 0.5; 95% confidence interval, 0.2 to 0.9; P  0.02) [30]. The small size of this study did not allow us to estimate the magnitude of the reduction with much precision. As expected, most of the fractures occurred in the women. The individual contributions of calcium and vitamin D in these two studies cannot be determined but they are probably dependent upon the starting calcium and vitamin D intakes of the specific study population. On the basis of the studies cited above and other evidence, The National Academy of Sciences revised their recommendation for calcium intake from 800 to 1200 mg per day for calcium for all healthy men and women age 51 and older [42].

D. Conclusions and Recommendations

FIGURE 6

Rates of change in bone mineral density (BMD) in men and women treated with placebo (-----) or calcium plus vitamin D (—) for 3 years. Results are expressed as mean ± SEM percentage change from baseline values. Three-year differences between the treatment groups were significant in men at the femoral neck (P = 0.011), spine (P = 0.026), and total body (P<0.001) and in women at the spine (borderline significance, P = 0.086) and total body (P<0.001). Reprinted with permission from DawsonHughes et al., [30].

of calcium on fracture rates. In the largest study [40], postmenopausal women with a mean dietary calcium intake of 433 mg per day were randomly assigned to treatment with 1200 mg of elemental calcium per day or placebo. Spine X-rays were taken at the beginning and end of the study. Calcium supplementation reduced the vertebral fracture rate in a high-risk subset of women with preexisting spine fractures but not in the women without preexisting spine fractures. Overall, the results of these studies are mixed.

Increasing calcium intake to recommended levels, in the presence of adequate vitamin D, induces a sustained lowering of rates of bone remodeling and bone loss in elderly men and women. In older women, these changes are accompanied by a significant reduction in fracture incidence. Based on their similar bone remodeling and bone density responses to added calcium and vitamin D, supplementation is likely to lower fracture rates in men as well, but this has not yet been demonstrated. The fracture reductions seen in women cannot be accounted for by the modest effects of supplementation on bone mineral density; other factors including the supplement induced lowering of the bone remodeling rate are likely to be involved. Finally, sporadic supplement use does not appear to provide long-term effects on the skeleton. This suggests that an adequate continuous intake of calcium (and vitamin D) is needed to achieve lasting skeletal benefit.

III. CALCIUM SOURCES The preferred approach to achieving optimal calcium intake is through the diet. In contrast to most supplements, foods provide a variety of important nutrients. Obtaining

550

BESS DAWSON-HUGHES

TABLE 1

Calcium Intervention Trials and Fracture Incidence

Calcium intake (mg/day) Study

Diet

Persons with new fracture

Supplement

N

Site

Statistically significant difference

18

Vertebra

Yes Yes

Chevalley [32]

600

800

Recker [40]

433

1200

With prior fracture

36

Vertebra

No prior fracture

25

Vertebra

No

Reid [37]

700

1000

9

All sites

Yes

Riggs [53]

700

1600

40

All sites

No

Dawson-Hughes [30]

700

500

37

All nonvertebral

Yes

( 700 IU vit D) Chapuy [41]

500

1200

315

Hip

Yes

( 800 IU vit D)

563

All nonvertebral

Yes

calcium from food sources may also reduce the probability of significant interference with absorption of trace minerals, such as iron and zinc, that may occur with use of large doses of calcium supplements. Calcium supplements are important for individuals who cannot or will not achieve optimal calcium intakes from food. Many different supplements including calcium acetate, carbonate, citrate, citrate malate, glubionate, lactate, lactogluconate, and tricalcium phosphate are available. Use of bone meal, dolomite, and other unrefined calcium sources is discouraged because these sources may contain lead and other toxic contaminants [43].

A. Absorbability Absorbability of calcium is greatly influenced by the study conditions, such as the presence or absence of a test meal and the size of the test calcium load. Although supplement solubility (in water at neutral pH) is only weakly related to absorbability [44], disintegration of tablets is required for absorption. Several investigators have studied absorbability under similar test conditions (a standardized breakfast meal containing 250 mg of elemental calcium) and their results can be compared [44 – 47]. Fractional calcium absorption was a little higher from calcium citrate malate and similar from carbonate, tricalcium phosphate, and milk. Differences in absorbability of these supplement sources are generally offset by differences in their calcium contents, so that the amounts of calcium absorbed per gram of supplement are similar. The studies were conducted in healthy volunteers under age 30 and it is unknown whether absorbability from these sources in the elderly is proportional.

B. Dosage and Schedule Absorbed calcium is dependent upon the calcium test dose. Absorption rises rapidly as the test load approaches 500 mg, and more gradually (as the active transport mechanism becomes saturated) with increases above this level (Fig. 1). Thus, to maximize utilization of calcium from supplements or from calcium-rich foods, calcium should be taken in doses of 500 mg or less [1,48]. In healthy individuals, absorption of calcium from carbonate is more consistent and reproducible when this supplement is taken with a meal than when it is taken during a fast [48]. The timing of calcium carbonate use is especially important for individuals with reduced gastric acid production. These individuals absorb calcium from carbonate poorly when fasting, but normally in a meal setting [49]. The prevalence of this asymptomatic condition rises with aging, reaching about 30% by age 60. A dose of calcium at bedtime reduces the nocturnal rise in the bone resorbing agent, PTH [50]. It is uncertain, however, whether supplement use at bedtime favors bone conservation more than use at other times, since intermittent increases in blood PTH concentration under selected conditions are anabolic to bone.

C. Safety Recommended calcium intakes pose little risk of adverse effects. High intakes of calcium from food sources have been associated with lower incidence of first renal stones in men [51] and women [52]. High intakes from supplements have been associated with a 20% increase in risk of first stone in women [52] but not in men [51]. Based on these

CHAPTER 67

and other data, the National Academy of Sciences identified a safe upper limit for calcium of 2500 mg per day [42]. Individuals may develop constipation or bloating from use of calcium carbonate but this can be resolved by changing to a different calcium source. Some calcium sources may interfere with the absorption of iron, zinc, and trace minerals. Further study is needed to identify and define the clinical impact of nutrient – nutrient interactions.

References 1.

2.

3.

4.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14.

551

The Role of Calcium in the Treatment of Osteoporosis

R. P. Heaney, P. D. Saville, and R. R. Recker, Calcium absorption as a function of calcium intake. J. Lab. Clin. Med. 85, 881 – 890 (1975). F. Bronner and R. S. Harris, Absorption and metabolism of calcium in human beings, studied with calcium45. Ann. NY Acad. Sci. 64, 314 – 325 (1956). H. Spencer, I. Lewin, J. Fowler, and J. Samachson, Influence of dietary calcium intake on Ca47 absorption in man. Am. J. Med. 46, 197 – 205 (1969). R. P. Heaney, R. R. Recker, M. R. Stegman, and A. J. Moy, Calcium absorption in women: Relationships to calcium intake, estrogen status, and age. J. Bone Miner. Res. 4, 469 – 475 (1989). B. Dawson-Hughes, S. Harris, C. Kramich, G. Dallal, and H. M. Rasmussen, Calcium retention and hormone levels in black and white women on high and low calcium diets. J. Bone Miner. Res. 8, 779 – 787 (1993). O. J. Malm, Calcium requirement and adaptation in adult men. Scand. J. Clin. Lab. Invest. (Suppl.) 36, 1 – 280 (1958). E. A. Krall and B. Dawson-Hughes, Relation of fractional 47Ca retention to season and rates of bone loss in healthy postmenopausal women. J. Bone Miner. Res. 6, 1323 – 1329 (1991). S. A. Abrams, K. O. O’Brien, L. K. Liang, and J. E. Stuff, Differences in calcium absorption and kinetics between black and white girls age 5 – 16 years. J. Bone Miner. Res. 10, 829 – 833 (1995). C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D3 in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). L. V. Avioli, J. E. McDonald, and S. W. Lee, The influence of age on the intestinal absorption of 47Ca in women and its relation to 47Ca absorption in postmenopausal osteoporosis. J. Clin. Invest. 44, 1960 – 1967 (1965). J. R. Bullamore, R. Wilkinson, J. C. Gallagher, B. E. C. Nordin, and D. H. Marshall, Effects of age on calcium absorption. Lancet 2, 535 – 537 (1970). J. C. Gallagher, B. L. Riggs, J. Eisman, D. Hamstra, S. B. Arnaud, and H. F. DeLuca, Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. Effect of age and dietary calcium. J. Clin. Invest. 64, 729 – 736 (1979). P. R. Ebeling, M. E. Sandgren, E. P. DiMagno, A. W. Lane, H. F. DeLuca, and B. L. Riggs, Evidence of an age-related decrease in intestinal responsiveness to vitamin D: Relationship between serum 1,25-dihydroxyvitamin D3 and intestinal vitamin D receptor concentrations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). R. Eastell, A. L. Yergey, N. E. Vieira, S. L. Cedel, R. Kumar, and B. L. Riggs, Interrelationships among vitamin D metabolism, true calcium absorption, parathyroid function, and age in women: Evidence of an age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J. Bone Miner. Res. 6, 125 – 132 (1991).

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

25a.

26.

27.

28.

29.

30.

31.

32.

D. N. Slovik, J. S. Adams, R. M. Neer, M. F. Holick, and J. T. Potts, Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 – 374 (1981). K. S. Tsai, H. Heath III, R. Kumar, and B. L. Riggs, Impaired vitamin D metabolism with aging in women. Possible role in pathogenesis of senile osteoporosis. J. Clin. Invest. 73, 1668 – 1672 (1984). Y. Fujisawa, K. Kida, and H. Matsuda, Role of change in vitamin D metabolism with age in calcium and phosphorus metabolism in normal human subjects. J. Clin. Endocrinol. Metab. 59, 719 – 726 (1984). E. S. Orwoll and D. E. Meier, Alterations in calcium, vitamin D, and parathyroid hormone physiology in normal men with aging: Relationship to the development of senile osteoporosis. J. Clin. Endocrinol. Metab. 63, 1262 – 1269 (1986). J. C. Gallagher, J. Aaron, A. Horsman, D. H. Marshall, R. Wilkinson, and B. E. C. Nordin, The crush fracture syndrome in postmenopausal women. Clin. Endocrinol. Metab. 1, 293 – 315 (1973). H. A. Morris, A. G. Need, M. Horowitz, P. D. O’Loughlin, and B. E. C. Nordin, Calcium absorption in normal and osteoporotic postmenopausal women. Calcif. Tissue Int. 49, 240 – 243 (1991). B. E. C. Nordin, A. Robertson, R. F. Seamark, A. Bridges, J. C. Philcox, A. G. Need, M. Horowitz, H. A. Morris, and S. Dean, The relation between calcium absorption, serum dehydroepiandrosterone, and vertebral mineral density in postmenopausal women. J. Clin. Endocrinol. Metab. 60, 651 – 657 (1985). R. M. Francis, M. Peacock, G. A. Taylor, J. H. Storer, and B. E. C. Nordin, Calcium malabsorption in elderly women with vertebral fractures: Evidence for resistance to the action of vitamin D metabolites on the bowel. Clini. Sci. 66, 103 – 107 (1984). N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). E. A. Krall, P. Parry, J. B. Lichter, B. Dawson-Hughes, Vitamin D receptor alleles and rates of bone loss: Influences of years since menopause and calcium intake. J. Bone Miner. Res. 10, 978 – 984 (1995). B. Dawson-Hughes, S. S. Harris, and S. Finneran, Calcium absorption on high and low calcium intakes in relation to vitamin D receptor genotype. J. Clin. Endocrinol. Metab. 80, 3657 – 3661 (1995). S. K. Ames, K. J. Ellis, S. K. Munn, K. C. Copeland, and S. A. Abrams, Vitamin D receptor bone Fok-1 polymorphism predicts calcium absorption and bone mineral density in children. J. Bone Miner. Res. 14, 740– 746 (1999). P. Garnero, E. Hausherr, M.-C. Chapuy, C. Marcelli, H. Grandjean, C. Muller, et al., Markers of bone resorption predict hip fracture in elderly women: The EPIDOS prospective study. J. Bone Miner. Res. 11, 1531 – 1538 (1996). M.-C. Chapuy, P. Chapuy, and P. J. Meunier, Calcium and vitamin D supplements: Effects on calcium metabolism in elderly people. Am. J. Clin. Nutr. 46, 324 – 328 (1987). B. Dawson-Hughes, G. E. Dallal, E. A. Krall, L. Sadowski, N. Sahyoun, and S. Tannenbaum, A placebo-controlled trial of calcium supplementation in postmenopausal women. N. Engl. J. Med. 323, 878 – 883 (1990). G. Kochersberger, C. Bales, B. Lobaugh, and K. W. Lyles, Calcium supplementation lowers serum parathyroid hormone levels in elderly subjects. J. Gerontol. 45, M159 – M162 (1990). B. Dawson-Hughes, S. S. Harris, E. A. Krall, and G.E. Dallal. Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. N. Engl. J. Med. 337, 670 – 676 (1997). B. Riis, K. Thomsen, and C. Christiansen, Does calcium supplementation prevent postmenopausal bone loss. N. Engl. J. Med. 316, 173 – 177 (1987). T. Chevalley, R. Rizzoli, V. Nydegger, D. Slosman, C.-H. Rapin, J.P. Michel, et al., Effects of calcium supplements on femoral bone

552

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

BESS DAWSON-HUGHES mineral density and vertebral fracture rate in vitamin-D-replete elderly patients. Osteoporosis Int. 4, 245 – 252 (1994). P. J. M. Elders, J. C. Netelenbos, P. Lips, F. C. Ginkel, E. Khoe, O. R. Leeuwenkamp, W. H. L. Hackeng, and P. F. van der Stelt, Calcium supplementation reduces vertebral bone loss in perimenopausal women: A controlled trial in 248 women between 46 and 55 years of age. J. Clin. Endocrinol. Metab. 73, 533 – 540 (1991). H. M. Frost, The origin and nature of transients in human bone remodeling dynamics. In “Clinical Aspects of Metabolic Bone Disease” (B. Frame, A. M. Parfitt, and H. Duncan, eds.), International Conference Series No. 270, pp. 124 – 137. Excerpta Medica, Amsterdam, 1973. J. F. Aloia, A. Vaswani, J. K. Yeh, P. L. Ross, E. Flaster, and F. A. Dilmanian, Calcium supplementation with and without hormone replacement therapy to prevent postmenopausal bone loss. Ann. Intern. Med. 120, 97 – 103 (1994). R. L. Prince, M. Smith, I. M. Dick, R. I. Price, P. G. Webb, N. K. Henderson, and M. M. Harris, Prevention of postmenopausal osteoporosis. A comparative study of exercise, calcium supplementation, and hormone-replacement therapy. N. Engl. J. Med. 325, 1189 – 1195 (1991). I. R. Reid, R. W. Ames, M. C. Evans, G. D. Gamble, and S. J. Sharpe, Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women: A randomized controlled trial. Am. J. Med. 98, 331 – 335 (1995). E. S. Orwoll and S. K. Oviatt, The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann. Int. Med.112, 29 – 34 (1990). B. Dawson-Hughes, S. S. Harris, and G. E. Dallal, Effect of withdrawal of calcium and vitamin D supplements on bone mass in elderly men and women. Am. J. Clin. Nutr. 72, 745 – 750 (2000). R. R. Recker, S. Hinders, K. M. Davies, R. P. Heaney, M. R. Stegman, and J. M. Lappe, D. M. Kimmil, Correcting calcium nutritional deficiency prevents spine fractures in elderly women. J. Bone Miner. Res. 11, 1961 – 1966 (1996). M.-C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994). Standing Committee on the Scientific Evaluation of Dietary Reference Intakes, Institute of Medicine, “Dietary Reference Intakes: Cal-

43.

44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

cium, Phosphorus, Magnesium, Vitamin D, and Fluoride.” National Academy Press, Washington, DC, 1997. B. P. Bourgoin, D. R. Evans, J. R. Cornett, S. M. Lingard, and A. J. Quattrone, Lead content in 70 brands of dietary calcium supplements. Am. J. Public Health 83, 1155 – 1160 (1993). R. P. Heaney, R. R. Recker, and C. M. Weaver, Absorbability of calcium sources: The limited role of solubility. Calcif. Tissue Int. 46, 300 – 304 (1990). K. T. Smith, R. P. Heaney, L. Flora, and S. M. Hinders, Calcium absorption from a new calcium delivery system (CCM). Calcif. Tissue Int. 41, 351 – 352 (1987). J. Z. Miller, D. L. Smith, L. Flora, C. Slemenda, X. Jiang, and C. C. Johnston, Calcium absorption from calcium carbonate and a new form of calcium (CCM) in healthy male and female adolescents. Am. J. Clin. Nutr. 48, 1291 – 1294 (1988). R. P. Heaney, K. T. Smith, R. R. Recker, and S. M. Hinders, Meal effects on calcium absorption. Am. J. Clin. Nutr. 49, 372 – 376 (1989). J. A. Harvey, M. M. Zobitz, and C. Y. C. Pak, Dose dependency of calcium absorption: A comparison of calcium carbonate and calcium citrate. J. Bone Miner. Res. 3, 253 – 258 (1988). R. R. Recker, Calcium absorption and achlorhydria. N. Engl. J. Med. 313, 70 – 73 (1985). M. S. Calvo, R. Eastell, K. P. Offord, E. J. Bergstralh, and M. F. Burritt, Circadian variation in ionized calcium and intact parathyroid hormone: Evidence for sex differences in calcium homeostasis. J. Clin. Endocrinol. Metab. 72, 69 – 76 (1991). G. C. Curhan, W. C. Willett, E. B. Rimm, and M. J. Stampfer, A prospective study of dietary calcium and other nutrients and the risk of symptomatic kidney stones. N. Engl. J. Med. 328, 833 – 838 (1993). G. C. Curhan, W. C. Willett, F. E. Speizer, D. Spiegelman, and M. J. Stampfer, Comparison of dietary calcium with supplemental calcium and other nutrients as factors affecting the risk for kidney stones in women. Ann. Intern. Med. 126, 497 – 504 (1997). B. L. Riggs, W. M. O’ Fallon, J. Muhs, M. K. O’Connor, R. Kumar, and L. J. Melton, III. Long-term effects of calcium supplementation on serum parathyroid hormone level, bone turnover, and bone loss in elderly women. J. Bone Miner. Res. 13, 168 – 174 (1998).

CHAPTER 68

Vitamin D and Its Metabolites in the Management of Osteoporosis IAN R. REID

I. II. III. IV.

Department of Medicine, The University of Auckland, Auckland 1, New Zealand

V. Other Osteoporoses VI. Conclusions References

Introduction Vitamin D in the Pathogenesis of Osteoporosis Animal Models Postmenopausal Osteoporosis

I. INTRODUCTION

calcium absorption is a limiting factor in the bone balance of osteoporotic subjects. The other potential role for vitamin D in the management of osteoporosis is as a physiological supplement. This would be appropriate if there were evidence that deficiencies of either the parent vitamin or its metabolites occurred in osteoporotic subjects and that this contributed to their bone loss. This distinction between pharmacological and physiological uses of vitamin D is important because the doses involved and their safety are substantially different, as is the metabolite likely to be used. It should also be remembered that osteoporosis is not a single entity with a single pathogenesis. Thus, the therapeutic role of vitamin D must be assessed in each of the different classes of osteoporosis (e.g., postmenopausal, steroid-induced, male osteoporosis). Even within each of these categories there is likely to be heterogeneity among

The discovery in the first half of the 20th century that vitamin D played a pivotal role in the regulation of intestinal calcium absorption has been followed by a sustained interest in the possibility of using this compound and its metabolites as therapies for osteoporosis. Intestinal calcium absorption is clearly one of the key factors in the determination of an individual’s calcium balance, so vitamin D is an attractive candidate for osteoporosis therapy. However, increasing intestinal calcium absorption will not inevitably result in an improvement in bone density. It is only one of several factors influencing calcium balance, which is quite distinct from bone balance. Bone balance is the difference between rates of matrix synthesis and degradation. Thus, an evaluation of the therapeutic potential of vitamin D requires an assessment of the extent to which intestinal

OSTEOPOROSIS, SECOND EDITION VOLUME 2

553

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

554

IAN R. REID

patients. Thus, vitamin D deficiency might be expected to be a more common problem in individuals living at high latitude or in those who spend their time predominantly indoors. Similarly, patients with other medical problems (e.g., mild renal failure) will be more likely to have abnormal levels of vitamin D metabolites. Thus, the results of studies of both pathogenesis and treatment should not be generalized without a careful consideration of the similarities and dissimilarities between the group of subjects in whom the data were collected and the group of subjects to whom its conclusions are going to be applied. Bearing these cautions in mind, the evidence suggesting a role of vitamin D in the pathogenesis of osteoporosis will now be reviewed, followed by a consideration of its therapeutic potential (see also Chapters 9 and 40).

A. Nomenclature Cholecalciferol or vitamin D3 is produced in the skin as a result of ultraviolet irradiation of 7-dehydrocholesterol. Ergocalciferol or vitamin D2 is produced by ultraviolet irradiation of the plant sterol ergosterol and is used as a vitamin D supplement. Doses of the calciferols are often expressed in international units (IU), there being 40 IU/g of cholecalciferol and 38.8 IU/g of ergocalciferol (for a description of the origins of these units the reader should consult Norman [1]). “Calciferol” refers to both these compounds. The two calciferols have been regarded as having comparable effects, but recent studies have indicated that the vitamin D3 series may have greater biological potency [2 – 4]. The metabolites of cholecalciferol, 25-hydroxyvitamin D3 [25(OH)D3], 1-hydroxyvitamin D3 [1(OH)D3], and 1,25-dihydroxyvitamin D3 [1,25(OH)2D3], are used as therapies and therefore have pharmaceutical names, calcifediol, alphacalcidol, and calcitriol, respectively. This chapter will generally follow the convention of using these pharmaceutical names when the compound is being administered to subjects but of using the abbreviated chemical name when referring to the substance in any other context.

II. VITAMIN D IN THE PATHOGENESIS OF OSTEOPOROSIS A. Calcium Malabsorption and Osteoporosis Intestinal absorption of calcium declines with age particularly after the age of 70 years [5,6]. This has been suggested to contribute to the decline in bone density with age which occurs in both sexes. Several groups have compared calcium absorption in osteoporotic patients with that in

aged-matched normal subjects. Most [6,7], but not all [8], studies have found that calcium absorption is lower in osteoporotic subjects, the reason for which is still the subject of debate. Gallagher et al. [6] found that circulating calcitriol concentrations were lower in osteoporotic subjects than in age-matched controls and a similar trend was observed by Morris et al. [7]. However, data in the latter paper suggested that this trend was not adequate for explaining the difference in intestinal calcium absorption, suggesting that there may be an intrinsic gut defect also contributing to the latter phenomenon. In contrast to these findings, Tilyard [9] reported lower fracture rates in calcium malabsorbers, whether treated with calcium or calcitriol. Some cautions are necessary in interpreting these findings. Most studies of calcium absorption in osteoporotic subjects have dichotomized the population as either having or not having osteoporosis by defining osteoporosis as the presence of a vertebral fracture. The recent widespread use of bone densitometry has made it clear that osteoporosis represents a continuum rather than a water-tight diagnostic category. What the relationship is between intestinal calcium absorption and bone density itself is unknown. Secondly, a finding of calcium malabsorption in osteoporotic subjects does not necessarily establish malabsorption of calcium as a contributor to osteoporotic bone loss. It could represent a normal homeostatic response to the increased rates of bone resorption found in osteoporosis [7] or it could be attributable to lower circulating vitamin D in osteoporotic subjects consequent to their reduced mobility and sunlight exposure.

B. The Role of Vitamin D Deficiency A detailed review of vitamin D deficiency is beyond the scope of this chapter. There is not consistent evidence that subjects with vertebral fractures have lower 25(OH)D concentration than age-matched controls [6 – 8]. In contrast, a number of studies have found such a difference in patients sustaining hip fractures [10]. One obvious difference between these two fracture groups is age, and these findings are consistent with a growing amount of data demonstrating a high prevalence of vitamin D deficiency in the elderly, particularly those not living independently [11 – 18]. The prevalence of vitamin D deficiency depends on sunlight intensity, sunlight exposure and dietary practices [10,12,16,19], and appears to be more common in Europe than in North America [20]. Hypovitaminosis D in the elderly is associated with secondary hyperparathyroidism [15,21] and reduced vertebral bone density [21]. Most reports in younger subjects, however, show no relationship between the circulating 25(OH)D and bone density

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

[22,23], though Khaw et al. [24] found spinal and proximal femoral bone density to be directly related to serum 25(OH)D concentrations and inversely related to those of parathyroid hormone (PTH) in middle-aged women. Two studies reported a positive association between dietary intake of vitamin D and bone density in both young adults [25] and perimenopausal women [26]. In the latter study, there was no association between circulating 25(OH)D and bone density despite the association with dietary intake. This may be explained by the finding that dietary intake of vitamin D correlates closely with that of a number of other nutrients including calcium [27] and thus does not necessarily support a role for vitamin D nutrition itself influencing bone density in normal subjects. There is some evidence of fluctuations in bone density with season [28]. Thus, in the northern United States, winter is associated with a fall in 25(OH)D concentrations, an increase in PTH concentrations, and a fall in bone density. Some of the change in bone density may reflect a change in the remodeling space and does not necessarily represent irreversible bone loss.

C. The Role of Altered Vitamin D Metabolism The principal index of altered vitamin D metabolism which has been studied in osteoporotic subjects is the circulating concentration of 1,25(OH)2D. Several groups have found levels of this hormone to be reduced in subjects with vertebral fractures [6,29], whereas others have found them to be normal [8,30]. One study found that radial bone mass was inversely related to circulating concentrations of 1,25(OH)2D [31]. Among normal subjects, there appears to be a decline in circulating concentrations of 1,25(OH)2D with age [32] but this is not a universal finding [29]. A more sophisticated approach to understanding vitamin D metabolism in osteoporosis has been the assessment of the increase in 1,25(OH)2D concentration produced by infusions of PTH. Riggs et al. [33] have shown that the response to PTH infusion in subjects with vertebral fractures (mean age 67 years) is normal but that it is subnormal in patients with hip fractures (mean age 78 years) [32]. In the latter study, the increase of serum 1,25(OH)2D declined with advancing age in normal subjects and was directly related to the glomerular filtration rate. Slovik [34] carried out similar studies comparing the 1,25(OH)2D response in a small number of men and women who had vertebral fractures with that in normal younger subjects. They also found a diminished rise in 1,25(OH)2D in the older osteoporotic subjects, though their design did not allow a determination of whether this difference was associated with increased age or the presence of osteoporosis.

555

While the literature is not unanimous on any of these issues, it does suggest a decline in circulating levels of 1,25(OH)2D in extreme old age which is related to loss of renal function. This decline may be relevant to the pathogenesis of senile (type 2) osteoporosis. There is also a tendency for 1,25(OH)2D values to be lower in the younger group of patients with vertebral (type 1) osteoporosis. However, their normal response to PTH implies normalcy of their 1-hydroxylase capacity. This is consonant with the finding that circulating PTH is lower in subjects with type 1 osteoporosis [6]. The most likely explanation for the decline in PTH is the increased rates of bone resorption associated with this condition [7,35,36]. These findings suggest that the changes in vitamin D metabolism found in type 1 osteoporosis may well represent an appropriate physiological response to increased bone resorption, and are mediated via suppression of PTH secretion. These data do not provide a rationale for vitamin D therapy in treating early postmenopausal bone loss. However, the clear effects of advanced old age on renal function and thus 1,25(OH)2D production and the increasing prevalence of vitamin D deficiency in some elderly populations do provide a sounder basis for intervention with these compounds in the elderly.

III. ANIMAL MODELS A detailed review of the use of vitamin D metabolites in animal models of osteoporosis is not appropriate to this clinically oriented account. Indeed, for the principal metabolites (calcitriol, alphacalcidol, and calciferol itself) such animal studies have been superseded by clinical studies. However, it is worth noting that calcitriol has generally produced positive effects on bone mass in animals, most commonly the rat. For example, Faugere et al. [37] showed that daily injections of calcitriol over 14 weeks increased cancellous and cortical bone mass in both sham-operated and oophorectomized rats. However, the nearly threefold increase in cancellous bone mass which calcitriol treatment produced in oophorectomized rats did not restore bone mass to the levels seen in the nonoophorectomized control animals. Similar results have been reported by other groups [38,39]. Histomorphometric and biochemical assessments from these studies are consistent with calcitriol having an antiresorptive action, similar to observations in clinical studies. Wronksi et al. [40] performed similar studies in male rats. Although trabecular bone volume increased with calcitriol treatment, they also demonstrated significant osteoid accumulation, implying either reduced mineralization or increased turnover. However, these studies lasted only 13 days and may thus represent a transient phenomenon.

556

IAN R. REID

IV. POSTMENOPAUSAL OSTEOPOROSIS A. Calciferol 1. PHARMACOLOGICAL USE The use of calciferol in the management of osteoporosis has passed through at least two phases. In the 1970s, awareness that intestinal calcium absorption declined with age and the possibility that it was lower in osteoporotic subjects than in other elderly individuals, led to an interest in using pharmacological doses of calciferol (the only available vitamin D metabolite at that time) to increase intestinal calcium absorption. One of the first groups to explore this approach was that of Buring et al. [41]. They reported an uncontrolled study of 53 women with back pain and radiological evidence of spinal osteoporosis, who were treated with vitamin D2 35,000 IU/day plus 1 g calcium over a period of 1 year. Back pain was relieved in two-thirds of these subjects but there was no significant change in the bone mineral density of the forearm. Absence of a control group makes interpretation of these data difficult and it is possible that the observed stability of bone mineral density was a superior outcome to what would have been expected in untreated patients. In 1980, Nordin et al. [42] included calciferol in doses of 10,000 – 50,000 IU daily with or without calcium as two of the regimens used in a nonrandomized study of women with vertebral fractures. Loss of metacarpal cortical area was, if anything, accelerated by the use of vitamin D alone, whereas patients treated with calcium plus vitamin D showed much the same bone loss as the untreated patients. Patients treated with calcium alone fared better still, having no detectable bone loss. The patients treated with vitamin D alone were the only group out of seven different study groups, in which there was a significant increase in the number of vertebral fractures during the mean treatment period of 2 – 3 years. Calcium balance was unaffected by vitamin D treatment — only patients treated with hormone replacement therapy (HRT) showed significant positive changes in calcium balance over the study period. This study was not randomized, but unlike that of Buring, it did include comparison groups. The conclusions of the study were similar to those of the Buring study — calciferol alone in pharmacological doses was not a promising therapy for postmenopausal osteoporosis. In 1982 Riggs et al. published a similar case series in which some patients had received vitamin D in doses of 50,000 IU once or twice weekly [43]. Patients also received one or more of the other therapies assessed: calcium, HRT, and fluoride. In those receiving sodium fluoride, there was evidence of a beneficial effect on vertebral fracture rate from the coadministration of calciferol. In the group as a

whole, there was no difference in fracture rates between those taking calciferol and the others, and, if the fluoridetreated subjects were excluded from the analysis, use of vitamin D was associated with a nonsignificant trend toward higher fracture rates. The authors concluded that there was no evidence to support the use of pharmacological doses of vitamin D other than as cotherapy with fluoride. 2. PHYSIOLOGICAL SUPPLEMENTATION More recently, interest has centered on the use of physiological doses of calciferol to correct subclinical vitamin D deficiency. When using calciferol in this way, it is necessary to define the optimal circulating concentration of 25(OH)D. This has recently been addressed by Malabanan et al. [44], who demonstrated that vitamin D supplementation suppressed PTH levels only in subjects whose baseline serum 25(OH)D was less than 50 nmol/liter (20 ng/ml), suggesting that this is an appropriate target concentration, though some cross-sectional studies suggest it may be as high as 100 nmol/liter (40 ng/ml) [45]. The most physiological way to replace vitamin D in housebound elderly is sunlight exposure. This approach has been described by Reid et al. [46]. Fifteen elderly rest home residents were randomly assigned to three groups: (i) no intervention; (ii) spending 15 min per day outdoors; (iii) spending 30 min per day outdoors. The study was carried out during spring. There were dose-related increases in serum 25(OH)D levels which had not plateaued by the end of the 1-month study. In subjects spending 30 min daily outdoors, serum 25(OH)D values increased by more than 30%. Changes in the intervention groups were associated with significant increases in intestinal strontium absorption (used as a surrogate for intestinal calcium absorption) and declines in serum alkaline phosphatase activity. Circulating concentrations of 1,25(OH)2D did not change during the study period. These data imply that modest changes in vitamin D status in the frail elderly can produce significant beneficial effects on calcium metabolism. Larger changes in circulating 25(OH)D can be produced with oral supplementation, and this has recently been reviewed by Vieth [47]. The dose-response is relatively flat up to intakes of several thousand IU/day, as demonstrated in Fig. 1. In young men receiving oral treatment for 8 weeks, daily calciferol doses of 1000, 10,000, and 50,000 IU increased serum 25(OH)D concentrations by 13, 137, and 883 nmol/liter (5.2, 54.8, and 353 ng/ml), respectively [48]. Honkanen [11] demonstrated a doubling of 25(OH)D concentration in elderly subjects receiving 1800 IU of cholecalciferol daily over an 11-week winter period, during which time the control group showed declines of 30 – 50% in circulating levels of this metabolite. Comparable effects have been demonstrated by others, and have been associated with reductions in serum PTH concentrations [49,50]. Similar effects can be produced with a variety of dosing

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

FIGURE 1

Dose – response for the effect of calciferol supplementation on serum 25(OH)D concentration in adults receiving the indicated dose for at least 1 month. Circles are means from groups and crosses are individual data from patients reported with vitamin D intoxication. The arrowed individual had been receiving 300,000 IU/month. Reprinted with permission from Vieth et al., [47].

regimens. Zeghoud [51] maintained normal circulating concentrations of 25(OH)D by 3-monthly oral administration of 100,000 IU of cholecalciferol, and Heikinheimo et al. [52] found that annual intramuscular injection of 150,000 IU of ergocalciferol achieved a similar end. Vitamin D deficiency can be effectively treated with 500,000 IU of calciferol, either given in a single dose or spread over days or weeks [53]. The addition of calciferol to foods such as milk also results in maintenance of normal vitamin D status [14,54]. Some studies have suggested that cholecalciferol results in larger increments in circulating 25(OH)D concentration than does ergocalciferol [2,4]. The effects of physiological vitamin D supplementation on bone density has been studied in a number of contexts. In normal early postmenopausal women it appears to have small if any effects on bone density, probably because such women are already optimally supplied with the compound. For example, Christiansen et al. followed the radial bone mineral content (BMC) of early postmenopausal women over a 2-year period during which the subjects received either vitamin D3 or placebo [55]. There was no difference in rates of bone loss between the groups. Studies in Finnish women have produced similar results in terms of bone density [56,57], though there was a downward trend in numbers of fractures at 5 years in those receiving 300 IU/day of calciferol alone, compared with placebo [58]. However, these women with baseline 25(OH)D concentrations of 24 – 30 nmol/liter (9.6 – 12 ng/ml) [59], would be considered to be vitamin D deficient by most standards.

557

In older normal women there is more evidence of efficacy. Nordin et al. [60] conducted a randomized placebocontrolled trial of vitamin D2 15,000 IU weekly in normal women ages 65 – 74 years. Metacarpal cortical area was monitored over a 2-year treatment period. Serum concentrations of 25(OH)D rose to the young normal range with this intervention and there was an upward trend in bone area in treated subjects in the face of a significant decline in controls (P  0.01). Ooms et al. carried out a randomized controlled trial of calciferol 400 IU daily versus placebo, and found an increase in serum 25(OH)D concentrations from 27 to 62 nmol/liter (10.8 to 24.8 ng/ml), suppression of serum PTH concentrations, and beneficial effects on femoral neck bone mineral density of about 2% at 2 years [61]. Dawson-Hughes et al. [62] showed a benefit to bone density at the femoral neck of 1.5% over 2 years from the use of a daily calciferol supplement of 700 IU, which increased serum 25(OH)D concentrations to 100 nmol/liter (40 ng/ml), in comparison with concentrations of 66 nmol/liter (26.4 ng/ml) in the control group. In an earlier 1-year study, this group found a similar benefit in the lumbar spine [63]. Baeksgaard et al. [64] found increases in spinal and femoral neck densities of about 1% at 2 years in healthy women with an average age of 63 years randomized to receive calciferol 560 IU/day or placebo. One uncontrolled study of patients with vitamin D deficiency suggested increases of up to 4% in spine and hip bone density could be achieved with calciferol treatment alone [53]. There have now been two large studies of the effects of calciferol administration alone on fractures in the elderly. Heikinheimo et al. [65] studied almost 800 elderly subjects in Finland. Approximately two-thirds were living in their own homes and the remainder were in a municipal home for the elderly. Subjects were randomized to receive 150,000 IU vitamin D2 annually (and in 1 of the 5 years 300,000 IU was given) or to act as controls. About a quarter of the subjects were males and the mean age of the subjects was between 86 and 87 years. Follow-up was from 2 to 5 years, the mean follow-up period being just over 3 years. Circulating levels of 25(OH)D were 31 and 14 nmol/liter (12.1 and 5.6 ng/ml) in the control subjects living independently or in a municipal home, respectively. These were normalized by the intervention (respective means, 49 and 45 nmol/liter (19.6 and 18 ng/ml)) and serum calcium concentrations remained normal in treated subjects. There was a fall in serum alkaline phosphatase activity only in the year during which the double dose of calciferol was given. Symptomatic fractures (confirmed by radiographs) were the principal end-point of the study. Fracture numbers were reduced by 25% in the vitamin D-treated subjects (P  0.03). In contrast, Lips et al. [66] showed no difference in fracture incidence in 2578 independently living men and women over the age of 70 years randomized to receive calciferol 400 IU/day or placebo over a period of up to 3.5

558

IAN R. REID

FIGURE 2 Cumulative probability of hip fracture and other nonvertebral fracture in elderly women treated with placebo (open squares) or vitamin D3 plus calcium (closed circles). Reprinted with permission from Chapuy et al., [67]. years. Mean serum 25(OH)D concentration in the third year of the study was 23 nmol/liter (9.2 ng/ml) in the placebo group and 60 nmol/liter (24.0 ng/ml) in the vitamin D group. There have also been two large studies of combined treatment with calcium and calciferol in the elderly. In the 3-year study of Chapuy et al. [67,68], more than 3000 women ages 69 – 106 years living in institutions for the elderly were randomly allocated to take placebo or 1.2 g of elemental calcium plus 800 IU of vitamin D3 daily. Nonvertebral fractures were recorded, and the cumulative probabilities of fracture in the two groups at 18 months are shown in Fig. 2. There were 32% fewer nonvertebral fractures in those receiving active treatment (P  0.02) and 43% fewer hip fractures (P  0.04). There were more than 350 fractures in total, 99% of which resulted from a fall. The data in Fig. 2 suggest that there was an increase in fracture rate over time in the placebo-treated subjects and this was prevented by the therapeutic regimen used. Bone density measurements taken in a subset of the patients were consistent with this. Proximal femoral bone mineral density (BMD) increased 2.7% in those receiving active treatment, but declined 4.6% in the placebo group (P  0.001). At the end of 3 years of treatment, the probabilities of non-vertebral fractures and hip fractures were reduced by 24 and 29%, respectively (P  0.001), in those receiving active therapy. Calcium-related biochemical indices were assessed in a subgroup of the trial subjects. Baseline serum 25(OH)D concentrations were 33 and 40 nmol/liter (13.2 and 16 ng/ml) in the placebo and active treatment groups, respectively. This implies that a substantial proportion of patients had significant vitamin D deficiency. Levels of 25(OH)D were stable in the placebo group throughout the first 18 months of the study but rose to 100 – 105 nmol/liters (40 – 42 ng/ml) in those receiving active therapy. This was accompanied by a fall in serum PTH concentrations from 54 to 30 pg/ml, whereas PTH rose by a

small but statistically significant amount in placebo-treated subjects. Baseline PTH concentrations were inversely related to femoral bone mineral density. Serum alkaline phosphatase activity remained stable in the active group but rose signifi cantly in placebo-treated subjects. Serum concentrations of 1,25(OH)2D were stable in both groups over the trial period. Dawson-Hughes et al. [69] reported similar findings. They randomized 389 men and women aged over 65 years to treatment with either 500 mg of calcium plus 700 IU of calciferol per day or placebo. At the end of 3 years, there were nonvertebral fractures in 26 subjects in the placebo group and in 11 in the calcium-vitamin D group (P  0.02). Thus, this study reproduces in an American cohort living at home the results of the Chapuy study [67,68]. The data from these studies indicate that vitamin D deficiency and secondary hyperparathyroidism are common in frail elders and suggest that these changes contribute to the progressive reduction in bone density which occurs in this age group. These biochemical abnormalities are reversible with physiological doses of vitamin D and calcium, and this results in beneficial effects on bone density and, more importantly, on fracture rates. Not only does this lead to a substantial reduction in morbidity in the subjects, but is likely to be associated with a significant prolongation of life, since there was a 24% mortality among patients developing hip fractures in the Chapuy study, [67,68]. There remains the question of which component of the therapy contributed to the therapeutic benefit seen in these studies or whether the combination of agents is necessary. The apparently contradictory results from the Heikinheimo and Lips studies do not help address this question. However, three recent studies have suggested an anti-fracture efficacy for calcium supplementation alone [70 – 72], indicating that a significant contribution from the calcium should not be ruled out.

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

3. SAFETY The use of replacement doses of calciferol is a very safe intervention, as would be expected since the intention is to restore circulating concentrations of 25(OH)D to those which are present in the ambulant population. Thus, in the study of Chapuy in which more than 1600 women were treated with calciferol, the only individual who developed hypercalcemia was subsequently found to have primary hyperparathyroidism. A similar zero-incidence of significant hypercalcemia has been reported by other investigators using either continuous or intermittent low-dose regimens of calciferol administration [11,51,73]. It is difficult to assess the safety of pharmacological doses of calciferol from the published series because some contain small numbers of patients [41] and others make no comment on side effects [42,43]. However, there are sufficient case reports of severe hypercalcemia, often of long duration and sometimes associated with renal failure, to counsel great caution in the use of these regimens [74,75]. Vieth [47] has recently reviewed this issue and concluded that the relationship of vitamin D dose to serum 25(OH)D concentration is relatively flat up to a daily calciferol dose of 10,000 IU (Fig. 1). Probably doses of up to 10,000 IU/day are safe in individuals without conditions which predispose them to hypercalcemia (e.g., primary hyperparathyroidism, sarcoidosis), and fully documented cases of toxicity have occurred only with intakes of 40,000 IU/day or more. The hypercalcemia of vitamin D intoxication has usually been attributed to increased intestinal calcium absorption, but Rizzoli et al. [75] recently demonstrated that bisphosphonates are effective in treating some cases, implying that increased bone resorption also contributes. Hyperresorption of bone has certainly been demonstrated in animal models of hypervitaminosis D [76]. The long duration of hypercalcemia associated with calciferol intoxification means that this compound is substantially less safe in high doses than its more active and shorter half-life metabolites such as alphacalcidol and calcitriol. Schwartzman and Franck [74] described four cases of vitamin D toxicity with a mean duration of hypercalcemia of 7 weeks. One of their patients was hypercalcemic for 16 weeks. This potential for significant toxicity together with the lack of any demonstration of efficacy suggest that pharmacological doses of vitamin D should not be used in the management of osteoporosis.

B. Calcitriol 1. EFFECTS ON BIOCHEMICAL INDICES The most consistent biochemical effect of calcitriol is a stimulation of intestinal calcium absorption. This is doserelated and is seen with doses as small as 0.25 g/day [77]. The effects of calcitriol on serum calcium concentrations

559

vary between reports. In general, serum calcium is not significantly elevated in patients taking doses less than 0.5 g/day. The tendency to hypercalcemia with doses from 0.5 to 1.0 g/day is quite marked in some studies [78,79] but not in others [80]. This variability of response may reflect differences in dietary calcium intakes between study populations. Fasting urine calcium excretion is little affected by calcitriol 0.25 g/day but rises significantly with higher doses [77]. These changes have been associated with suppression of circulating PTH in some studies [81,82] but not in others [78,83]. Similarly, a few studies found suppressed bone turnover in calcitriol-treated patients [77,81,83] whereas others show the reverse [84]. Higher doses of calcitriol (e.g. 2 g/day) consistently increase bone resorption reflecting the direct bone resorbing effects of this compound [82]. The action of low-dose calcitriol (in some studies with calcium) as an antiresorptive results from a net increase in intestinal calcium absorption with a resulting suppression of PTH stimulation of bone resorption. Serum osteocalcin concentration, an index of bone formation, is consistently elevated by calcitriol administration despite the suppression of other markers such as serum alkaline phosphatase activity [82,85,86]. This occurs because calcitriol directly regulates transcription of the osteocalcin gene [87] and it does not reflect a generalized increase in bone formation activity. 2. RANDOMIZED CONTROLLED STUDIES Table 1 summarizes the major published randomized controlled trials of calcitriol therapy in the management of postmenopausal osteoporosis. One of the first studies to assess the effects of calcitriol on bone density was that of Christiansen et al. [88], who randomized normal women 2.5 to 5 years after the menopause to receive calcitriol, HRT, both, or neither. Calcitriol was given in a dose of 0.25 g/day and treatment effects on distal forearm BMC were evaluated over a 12-month period. Subjects taking HRT with or without calcitriol showed increases in bone mass of about 1% over 1 year. In contrast, the placebo group and the calcitriol group experienced a 2% decline in BMC. Differences between hormone-treated subjects and both other groups were statistically significant. These findings suggest that calcitriol in this dose (0.25 g/day) has no role in the prevention of early postmenopausal bone loss. The following year, Gallagher et al. [83] performed more detailed studies of the effects of calcitriol (0.5 g/day) or placebo in women with at least one atraumatic vertebral fracture. After 6 – 8 months, the calcitriol-treated subjects showed decreases in bone turnover (by radiocalcium kinetics) and a positive change in calcium balance, but no changes in bone biopsy parameters nor in BMC of the distal radius. In subjects who continued on calcitriol through to 2 years, the positive change in calcium balance

TABLE 1 Randomized Controlled Trials of Calcitriol in Postmenopausal Osteoporosis Studya

nb

Dosec (g/day)

td

Entry criteria

Treatments

Christiansen et al., 1981 [88]

Healthy 2.5 – 5 years postmenopausal

Calcitriol HRT HRT  calcitriol Placebo

84

0.25

12

Gallagher et al., 1982 [83]

Vertebral fracture

Calcitriol Placebo

18

 0.5

6–8

End point(s)

Results

Comments

Distal forearm BMC

No effect of calcitriol, beneficial effect of HRT

Double- blind

Calcium balance Calcium kinetics Bone biopsy

Calcitriol qcalcium balance, pbone resorption. No changes in BMC

No between-groups comparisons, small numbers

Distal forearm BMC

560

Jensen et al., 1982, 1985 [89,90]

Healthy 70year-olds

Calcitriol HRT HRT calcitriol Placebo

74

0.42

12

Distal forearm BMC Metacarpal cortical area Vertebral height

No reduction in bone loss with calcitriol group which tended to lose more bone than control Beneficial effect of HRT Vertebral height decreased more in those taking calcitriol

All groups received Ca Double-blind

Falch et al., 1987 [91]

Recent forearm fracture

Calcitriol Calciferol, 400 IU/day

76

 0.5

36

Forearm BMC Vertebral fracture

No differences between groups

Low precision of BMC measurement

Aloia et al., 1988 [78]

Vertebral fracture

Calcitriol Placebo

27

0.8

24

Total body calcium Distal radius BMC Lumbar spine BMD Metacarpal density Vertebral fracture

Increases in all indices of bone mass with calcitriol No difference in fracture rates

Double-blind One-tailed tests used High incidence of hypercalcuria and hypercalcemia

Gallagher et al., 1989 [92]

Vertebral fracture

Calcitriol Placebo

62

 0.5

12

Vertebral fracture

Reduced fracture rates at one of two centers

Double-blind One-tailed tests used

Ott and Chestnut, 1989 [79]

Vertebral fracture

Calcitriol Placebo

72

0.43

24

Total body calcium Distal radius BMC Lumbar spine BMD Vertebral fracture Bone biopsy

Significantly more rapid loss of distal radius BMC on calcitriol

Double-blind Trend to more fractures in the calcitriol group

561

Fujita et al., 1989 [119]

Osteoporosis

Calcitriol Alphacalcidol

415

0.50

7

Metacarpal BMD

Maintenance of BMD in all groups

No placebo group

Arthur et al., 1990 [96]

Osteopenia

Calcitriol Calciferol

10

0.5

12

Vertebral BMD Bone biopsy

No significant treatment effects

All subjects received calcium Trend to larger BMD increase in calciferol group

Gallagher et al., 1990 [93]

Vertebral fracture

Calcitriol Placebo

40

0.62

24

Total body BMD and its subregions Metacarpal cortical width Vertebral fracture

Between-groups difference in total body BMD and the spine subregion

Double-blind Difference in fractures

Tilyard et al., 1992 [9]

Vertebral fracture

Calcitriol Calcium

432

0.5

36

Vertebral fracture

Fewer fractures in calcitriol group

Not double-blind Progressive rise in fracture rate in calcium group

Masud et al., 1998 [136]

Vertebral fracture or Z score

Etidronate Etidronate  calcitriol

47

0.5

12

Lumbar spine and femoral neck BMD

Changes in BMD 2.5% more positive in combination group

Not blinded

Calcium Calcitriol Alendronate Calcitriol  alendonate

102

0.5

24

Total body BMD

Effects on BMD were: combined therapy  alendronate  calcitriol  calcium

Not blinded

  1.5 Frediani et al., 1998 [84]

a

Total body BMD T score   2.5

Reference number of study. Total number completing study c Average dose of calcitriol in ‘active’ group d Duration in months b

562 was not sustained despite maintenance of high intestinal calcium absorption. Trabecular bone volume at the iliac crest increased from 11.3 to 16.0% but distal radius BMC did not change. This study has been regarded as generally supportive of the use of calcitriol in osteoporosis but it is by no means conclusive. Its placebo-controlled phase lasted only 6 – 8 months. The only differences found at the end of that time were in biochemical parameters. There was an increase in trabecular bone volume at 2 years in the calcitriol group but there are no control data from this time point. It should be noted in the controlled phase of the trial, that trabecular bone volume rose by more in the placebo group than it did in the calcitriol-treated subjects, though the difference was not statistically significant. In the same year, Jensen et al. [89] published the first report of a double-blind randomized controlled trial in healthy elderly (70-year-old) women. A factorial design with HRT was used, similar to that which the same group had used in their earlier study of perimenopausal women [88], but the initial dose of calcitriol was 0.5 g/day. Dose reductions during the course of the 12-month study resulted in an average calcitriol dose of 0.42 g/day. All subjects received a calcium supplement. HRT had consistently positive effects on bone mineral content but calcitriol did not. Bone loss tended to be more marked in the calcitriol-treated group than in the placebo group and at one distal forearm site there was significant loss only in the calcitriol-treated subjects. In a subsequent publication [90] it was reported that significant loss of vertebral height occurred only in patients receiving calcitriol, whether or not they were also taking HRT. In subjects receiving HRT alone or placebo, there was no height loss. This study suggests that calcitriol does not have beneficial effects on appendicular bone in healthy elderly women. The suggestion that it increases vertebral height loss in normal elderly women is of concern. In 1987 Falch et al. [91] randomized a large group of women who had recently suffered a forearm fracture to calcitriol 0.5 g/day or to a physiological replacement dose of calciferol. Forearm bone mineral content and vertebral fractures were monitored over the following 3 years. No differences in rates of bone loss or fracture incidence were found between the two groups. It should be noted, however, that the forearm densitometer used had a precision error of 3.7% at one site and 3.0% at the other. Thus, the study lacked power to detect small treatment effects. In 1989, workers from Creighton University and the Mayo Clinic published a combined report of trials they had carried out some years previously [92]. Sixty-two subjects completed the 12-month double-blind section of this study during which they were randomized to calcitriol 0.5 g/day with subsequent dose escalation, or placebo. Vertebral fracture was the criterion for entry and also the principal outcome variable. After only 12 months of treatment there was a substantial decrease in the fracture rate

IAN R. REID

of patients treated at Creighton University but no significant change in the Mayo Clinic patients. At the end of 12 months all subjects remaining in the study were treated with calcitriol, and subjects from both centers demonstrated lower fracture rates over this period than the placebo group had shown in year 1. These studies are highly suggestive of a beneficial effect of calcitriol therapy on fracture rate. However, the 2- and 3-year data are uncontrolled and one-tailed tests of statistical significance were used throughout, which is of dubious validity considering published data suggesting that calcitriol might also increase fracture rate and accelerate bone loss. The reason for the more gradual reduction in fracture rate in Mayo Clinic patients in comparison with those from Creighton is not apparent. In 1988 – 1989, three American studies appeared [78,79,93] all of which had followed a similar protocol. Subjects entering these studies had at least one vertebral fracture and were randomly allocated to receive therapy with calcitriol or placebo. Subjects on active therapy were started on calcitriol 0.5 g/day with dose escalation until hypercalciuria or hypercalcemia occurred. This dose titration was handled differently in each center with the result that the patients of Ott and Chesnut received a mean dose of 0.43 g/day, those of Gallagher 0.62 g/day, and those of Aloia 0.8 g/day. Subjects were followed for 2 years with measurements of total body calcium (either by neutron activation analysis or by dual-photon absorptiometry) and vertebral fracture. Despite design similarities among these studies, results were quite different. Aloia et al. maintained high intakes of calcitriol at the price of significant hypercalciuria and hypercalcemia. There were significant increases in total body calcium, distal radius BMC, lumbar spine BMD and metacarpal density in subjects receiving calcitriol but no changes in these indices in the placebo group. Again, the analyses were carried out using one-tailed statistical tests. Fracture rates were slightly (but not significantly) higher in the placebo group, but these patients had a higher number of fractures at baseline, indicating that the groups were not well matched. Calcitriol-treated patients showed increases in total body calcium and radius BMC of approximately 1% per annum, with the placebo group showing losses of 1 and 2% at the respective sites. In the spine, the placebo group lost 4% per annum but there was no change in the calcitriol group. While concluding that the effect of calcitriol in these subjects was beneficial, the investigators did not advocate that this treatment regimen be adopted because of the high incidence of side effects. Instead, they recommended the investigation of lower dose regimens and regimens involving parenteral administration of calcitriol. In the study of Gallagher et al. [93], the patients did not have major problems with hypercalcemia following the

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

initial dose titration phase, though dietary calcium intake was reduced to 600 mg per day. Total body calcium, measured by dual-photon absorptiometry, remained stable over the 2-year study period in the subjects receiving calcitriol and fell by approximately 2% in those receiving placebo. The spine subregion of the total body scans demonstrated an increase in BMD of almost 2% in those on calcitriol and a decrease of slightly greater magnitude in those receiving placebo. For metacarpal cortical width, however, the trend was the opposite though effects at this site were not statistically significant. There was no difference between the groups in the numbers of new fractures occurring during the study. This study added significantly to the data published by Aloia, in that it indicated that positive effects on bone density could be achieved with doses of calcitriol that were safe. The study of Ott and Chesnut [79] was the largest of this trio and used the lowest average dose of calcitriol (0.43 g/day at the trial’s conclusion). Subjects were initially placed on a calcium intake of 1000 mg/day which was adjusted according to subsequent urine calcium measurements. The placebo group thus maintained a calcium intake 400 mg higher than that in the calcitriol group throughout most of the study period. Bone density changes tended to be more positive in the placebo group at the three sites assessed, though between-groups differences were not significant. However, there was a significant decrease in distal radius BMC in the calcitriol group which did not occur in the placebo group. There were no significant changes in bone biopsy parameters. Fractures tended to be more common in the placebo group than in the calcitriol group, but this was not statistically significant. The authors concluded that calcitriol was ineffective in the treatment of established postmenopausal osteoporosis. Ott and Chesnut subsequently published a reanalysis of these data in which the calcitriol group was subdivided according to the average dose of drug taken during the study [94]. The 7 subjects taking  0.6 g/day of calcitriol showed more positive changes in all indices of bone density than did the 11 patients receiving 0.5 g/day, who had substantial losses of bone density. The numbers in each group of this post hoc analysis are small but the results are consistent with data from the Aloia and Gallagher studies suggesting that higher doses of calcitriol produce more beneficial effects on bone density. However, this does not necessarily mean that in a given individual a higher dose of calcitriol is superior. It may be that subjects with more severe malabsorption of calcium are more able to tolerate higher doses of calcitriol and these are the individuals who are most likely to benefit from this therapy. Fujita [95] and colleagues conducted a large, doubleblind trial comparing calcitriol 0.5 g/day with alphacalcidol 1 g/day. Six percent of the patients in this study were male. The two therapies appeared to have equivalent effects

563

on metacarpal BMD over a period of 7 months. Arthur et al. [96] reported a small study comparing calcitriol 0.5 g/day with calciferol 50,000 units twice a week. Patients in both groups received 1 g of calcium daily. Not surprisingly in view of the size of the study, there were no significant differences between the groups in their response to therapy. However, the changes in BMD tended to be more positive in the calciferol-treated subjects. Perhaps the greatest encouragement to the use of calcitriol in the therapy of osteoporosis has come from the study of Tilyard et al. [9]. Six hundred and twenty-two women with at least one vertebral fracture were randomly assigned to take calcitriol 0.5 g/day or calcium 1 g daily and were eventually followed over a 3-year period. The only end point was vertebral fracture, which was defined as a decrease in anterior or posterior height of the vertebral body of 15% or greater. Four hundred thirty-two women completed the study. There were significantly more fractures in the calcium-treated subjects than in the calcitrioltreated subjects in both years 2 and 3. The study has been loosely described as demonstrating a 50% reduction in fracture rates in women treated with calcitriol in comparison with those treated with calcium. The data do not really confirm this but rather show a stable fracture rate in the calcitriol-treated subjects throughout the study with a threefold increase in incidence of new fractures over the study period in those taking calcium. Thus, the study appears to demonstrate a deleterious effect of calcium supplementation on vertebral strength rather than a beneficial effect of calcitriol. The lower fracture rates in calcitriol-treated subjects were seen only in those who had fewer than six fractures at baseline. While the scale of this trial and its reliance on fracture as its end point are impressive, there are a number of problems with the study. Unlike most of the other controlled trials, the patients and their doctors were not blinded to the therapy allocation. There was a large number of withdrawals in year 1, and it was only after this period that the fracture rates of the groups diverged. It is possible that the groups continuing beyond the end of year 1 were no longer well matched but this cannot be discerned from the manuscript. It is likely that a significant number of the subjects were vitamin D deficient. Circulating 25(OH)D concentrations were in the range 10 – 80 nmol/liter (4 – 20 ng/ml) [97], which would suggest that approximately half of the population were vitamin D deficient. From the more recent evidence of the effectiveness of treating vitamin D deficiency in the elderly, it might be speculated that comparable effects could have been produced with calciferol [98]. Finally, discrepancies exist between the data in a preliminary report of this study [99] and those that appear in the final paper. In the preliminary report, 226 calcium-treated patients had reached the 2-year end point and 65 fractures had occurred in this group. In the calcitriol group, 224 patients

564 had completed with 46 fractures. In the final report a further 12 patients had reached 2 years in the calcitriol group with a decline in the total number of fractures of one, whereas 14 patients had been added to the calcium group with an increase in fracture number of 21. These discrepancies seem to have arisen from remeasurement of X-rays following the writing of the preliminary report. The asymmetry of the resulting changes in fracture numbers between the groups remains unexplained. This concern, together with uncertainty regarding subject exclusion on the grounds of non-compliance, the possibility that a substantial proportion of the patients were vitamin D deficient, lack of blinding, and the inexplicable changes in fracture rate in the calcium-treated group, mean that this trial does not definitively establish calcitriol as having anti-fracture efficacy in postmenopausal osteoporosis. At the time of writing, several new studies have been presented only in abstract form. Fenton et al. [100] randomized postmenopausal women to receive HRT or calcitriol 0.5 g/day. At 1 year, the HRT group showed BMD increases of 6.2% at the spine and 3.6% at the hip, with no change in those receiving calcitriol. In a similar study, Gallagher and Fowler [101] randomized 489 elderly women to HRT, calcitriol, neither, or both. At 3 years, increases in BMD were about twice as great with HRT compared with those seen in calcitriol-treated patients, and combination therapy tended to produce the greatest increments in BMD. There was a trend for fracture rates to be lower in the calcitriol groups. 3. OTHER STUDIES Several other groups have described the effects of calcitriol therapy in osteoporosis in uncontrolled studies. Caniggia et al. [102] described 62 osteoporotic women treated with calcitriol 0.5 g twice daily for 4 – 30 months. This therapy apparently caused no change in plasma calcium and resulted in stability of the BMC of the ulna and radius. Subsequently, these authors reported on 270 women treated with this regimen for 1 – 8 years [80]. Total body BMD was measured in 56 of these subjects, the majority showing increases in this index over an 18- to 24-month period. Vertebral fracture rates were lower during the period on treatment than retrospective estimates of pretreatment fracture rates. Nuti et al. [103] used the same regimen in a series of 35 Italian women with a vertebral fracture, and compared their outcome with that in a nonrandomized control group. At 2 years, total body BMD increased 0.85% in those receiving calcitriol, whereas it declined 2.15% in the control group. There was a high incidence of hypercalcemia and hypercalciuria despite the dietary calcium intake being 500 mg/day. A further group to describe a number of patients undergoing calcitriol therapy is that of Nordin and Need. They

IAN R. REID

have advocated selective use of calcitriol 0.25 g/day with calcium supplementation in patients with demonstrable calcium malabsorption and have suggested that the benefits are greatest in those with most marked malabsorption [104]. This regimen has clear-cut antiresorptive effects on biochemical indices [77] and forearm bone mineral content is stabilized in retrospective case series [105 – 107]. 4. SAFETY The above studies are generally reassuring regarding the safety of calcitriol use, particularly when doses do not exceed 0.5 g/day. This dose, when used in the absence of calcium supplementation, causes only modest hypercalciuria. None of the more than 200 women treated over a 3year period by Tilyard et al. [9] developed renal colic, and in the 60 subjects who had renal ultrasonography after 2 years of treatment, no evidence of calcium deposition was seen. In most studies, serum calcium concentrations have remained stable throughout the trial period, though there have been reports of hypercalcemia in routine clinical use of calcitriol [108]. However, it is quite clear that a combination of calcitriol with calcium supplementation, or progressive escalation of the calcitriol dose will result in significant hypercalciuria and hypercalcemia, such patients requiring frequent monitoring. The finding in two of the studies discussed above, that either bone loss or loss of vertebral height were greater in those treated with calcitriol than in control subjects, provokes concern. It is possible that these findings arose by chance, though the fact that this has occurred in two separate studies makes this less likely. It is not apparent why these subjects should have experienced a deleterious treatment effect when others have apparently derived benefi from calcitriol. But, the fact that they did, suggests that calcitriol be used with great circumspection, particularly in those who do not have established osteoporosis. The recent adverse outcome of a trial in male osteoporosis (vide infra) reinforces this caution.

C. Alphacalcidol 1. EFFECTS ON BIOCHEMICAL INDICES Alfacalcidol (1-hydroxycholecalciferol) is a synthetic vitamin D3 compound hydroxylated in position 1. It differs from calcitriol in that it has not been hydroxylated at position 25, but this conversion takes place rapidly in the liver following administration of alphacalcidol in humans [109]. Thus, its biochemical effects are very similar to those of calcitriol. It stimulates intestinal calcium absorption [110,111]. This is associated with increases in serum calcium concentrations when a dose of 1 g/day or more is used [111,112] and some authors have noted increases

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

in serum calcium levels with doses of only 0.5 g/day [111,113,114]. Urine calcium excretion increases [111,112,114,115] and several groups have observed suppression of circulating PTH concentrations with doses of 1 g/day or more [111,112]. Indices of bone turnover are usually suppressed by alphacalcidol [113,116] though this is not always observed [112,115]. In high doses, however, alphacalcidol (like calcitriol) stimulates bone resorption [110]. 2. RANDOMIZED CONTROLLED STUDIES Table 2 summarizes the principal randomized controlled studies of alphacalcidol in the management of postmenopausal osteoporosis. Many of these were carried out in Japan and may not be generalizable to European populations since substantial genetic and lifestyle differences distinguish these groups. In particular, dietary calcium intake is very much lower in Japan. One of the first studies of alphacalcidol in established osteoporosis was carried out in Danish women, many of whom also appear to have had osteomalacia [112]. There were substantial increases in forearm bone mineral content, though there was no change in mineralized bone volume of iliac crest biopsies. There was a substantial reduction in the severity of back pain in the treated patients. Several years later, Hoikka et al. [117] carried out a similar study in patients recovering from hip fractures. There were no significant benefits to radial bone density or bone mass at the iliac crest, but  20% of subjects developed hypercalcemia, probably because of the concomitant use of calcium supplementation. This study had the power to detect only a substantial and rapid treatment effect. Chritiansen [55] included a low dose of alphacalcidol as one of the arms in a large double-blind study of osteoporosis prevention in early postmenopausal women. There was no evidence of any slowing of bone loss in the women receiving alphacalcidol, whereas HRT and thiazide diuretics showed significant beneficial effects on forearm BMC at the end of the 2-year study period. Fujita and colleagues reported several controlled studies of vitamin D metabolites in osteoporosis. Unfortunately, detailed accounts of these are not readily accessible [118, 119], though their findings have also been presented in summary form [95]. In the first of these, osteoporotic subjects were randomly allocated to take alphacalcidol 1 g/day or placebo and radial and spinal bone density was followed over a 12-month period. At both sites, there were increases in bone density of 7% at 6 months and 9% above baseline at 12 months in those on active therapy. Placebo treatment, on the other hand, was associated with a 12% decline in bone density at both sites in the first 6 months, though there was a positive trend in bone density in the second 6 months. The positive changes seen in the treated group were substantial and they are comparable to the

565

forearm BMC changes reported by Sorenson et al. [112]. However, the relative size of the changes at the spine and forearm is not typical of the effects of other antiresorptive agents (such as HRT or bisphosphonates) which usually produce substantially more positive effects at the metabolically active trabecular bone of the spine than are seen in cortical appendicular bone. Perhaps the most surprising finding of this study, however, is the substantial decline in the bone density of the placebo group during the first 6 months of therapy. Untreated patients should be in a relatively stable state and it is obviously impossible for patients to sustain rates of bone loss of this magnitude longterm. This implies that there may have been major imprecision in the measurement of bone density in this study. A further large multicenter study from Fujita’s group [95] randomized 300 osteoporotic patients to receive alphacalcidol 0.75 g/day or placebo, for 7 months. By 3 months, a more than 1% difference in metacarpal thickness had developed between groups, and this was maintained through 7 months. Mineral density of the second metacarpal was almost 5% higher at 3 months in those treated with alphacalcidol, and the between-groups difference increased a further 2% in the second half of the study. The between-groups differences found by both measurement techniques were statistically significant. These outcomes suggest a real beneficial effect of alphacalcidol treatment on the appendicular bone of osteoporotic Japanese women. In 1985, Shiraki et al. [120] compared the effects of two doses of alphacalcidol (0.5 and 1 g) with those of two isomers of 1,24(OH)2D and placebo. Both doses of alphacalcidol and the 1,24R(OH)2D isomer produced beneficial effects on radial bone mass, with the higher dose of alphacalcidol having the greatest effect. The sizes of the groups varied widely, suggesting that the assignment of subjects was nonrandom. In 1987, Orimo et al. [121] reported a study of 86 women. The initial report describes the study as being randomized, whereas subsequently it is said to be retrospective [115]. This may explain why the numbers of subjects in each group were not the same and there were variations in the duration of therapy. Placebos were not used. The groups were comparable in the number of baseline fractures apart from the group treated with calcium alone, which had a significantly greater number of fractures at a time of trial entry. After approximately 2 years of study, the fracture rates were: control group, 960/1000 patient-years; calcium group, 650/1000 patient-years; alphacalcidol group, 350/1000 patient-years; combined therapy with calcium and alphacalcidol, 140/1000 patient-years. Both groups of alphacalcidol-treated patients had significantly lower fracture rates than the controls. This study, thus, demonstrates an impressive reduction in fracture rate though its lack of

TABLE 2 Randomized Controlled Trials of Alphacalcidol in Postmenopausal Osteoporosis Studya

Entry criteria

nb

Treatments

Dosec (g/day)

td

End point(s)

Results

Comments

Vertebral fractures

1(OH)D Placebo

26

2

3

Forearm BMC Bone biopsy

BMCq14% posteoid

Both groups received Ca qosteoid at baseline

Hoikka et al., 1980 [117]

Hip fracture

1(OH)D Placebo

37

1

6

Radial BMD Bone biopsy

No treatment effect

Both groups received Ca

Christiansen et al., [55]

Normal, early postmenopausal

1(OH)D Placebo and others

126

0.25

24

Forearm BMC

No treatment effect of 1(OH)D

All subjects took 0.5 g/day Ca Low dose of 1(OH)D

Itami et al., 1982 [118]

Osteoporosis

1(OH)D Placebo

25

1

12

Radial BMD Spine BMD (QCT)

Significant benefits at both sites with 1(OH)D

LargepBMD in controls

Shiraki et al., 1985 [120]

Osteoporosis

1(OH)D Placebo 1,24(OH)2D

78

0.5 and 1

6

Radial BMD

Significant benefits with 1(OH)D and 1,24(OH)2D

? Strictly randomized

Fujita et al., 1990 [95]

Osteoporosis

1(OH)D Placebo

299

0.75

7

Metacarpal BMD and thickness

Significant benefits in both indices with 1(OH)D

Orimo et al., 1987 [121]

Vertebral fracture

1(OH)D 1(OH)D+Ca Ca Control

86

1

 24

Vertebral fracture

pfracture rate with 1(OH)D and 1(OH)D  Ca

Small numbers for fracture study Probably not a randomized study

Fujita et al., 1993 [122]

Osteoporosis

1(OH)D Cyclical EHDP 200 or 400 mg

406

1

11

Spine BMD (DXA) Vertebral fracture

BMDq2 – 3% in etidronate groups — no change with 1(OH)D pFractures with etidronate

No placebo group

566

Sorensen et al., 1977 [112]

567

Menczel et al., 1994 [114]

Osteoporosis

1(OH)D Placebo

46

0.5

36

Distal radial BMC

BMCq0.3%/year with 1(OH)D, p2.6%/year with placebo

Both groups received Ca

Orimo et al., 1994 [115]

Osteopenia  fracture

1(OH)D Placebo

74

1

12

Spine and femur BMD (DXA) Vertebral fracture

Significant benefit of 1(OH)D in spine. p1(OH)D fractures

Both groups received Ca

Lyritis et al., 1994 [123]

Vertebral fracture

1(OH)D Nandrolone

64

1

12

Distal radius BMC

Nandrolone qBMC 5%, 1(OH)D pBMC 2.8% ppain andqmobility with nandrolone

Double-blind

Shiraki et al., 1996 [141]

Osteoporosis

1(OH)D Placebo

43

0.75

24

Spine and total body BMD (DXA)

Spine BMD q2.3% with 1(OH)D, p0.3% with placebo – differences not significant

Double blind  50% noncompleters

Chen et al., 1997 [125]

Spine BMD T score  1.5

1(OH)D Calcium

45

0.75

12

Spine BMD

BMDq2.1% with 1(OH)D p2.1% with calcium

Not blinded

Itoi et al., 1997 [126]

Normal, early postmenopausal

Estriol CE 1(OH)D Calcium

64

1

24

Spine BMD (QCT)

Comparable bone loss with calcium and 1(OH)D (12%), none with hormones

? Blinding

Gorai et al., 1999 [127]

Normal, early postmenopausal

1(OH)D CE CE1(OH)D Control

79

1

24

Spine and femur BMD (DXA)

Spine BMD: CE  2%; 1(OH)DCE  3%; 1(OH)D and control 3%

? Number of dropouts

CE, conjugated estrogens; EHDP, etidronate. a Reference number of study. b Total number completing study. c Average dose of alphacalcidol in ‘active’ group. d Duration in months.

568 blinding, small group sizes and imperfect matching at baseline are significant weaknesses. Fujita’s group have also published a large, double-blind study comparing changes in bone density and fracture rate between patients treated with alphacalcidol and two different doses of etidronate [122]. Patients were treated with alphacalcidol 1 g/day, etidronate 200 mg/day for 2 weeks followed by a 10-week break, or the same regimen using etidronate 400 mg/day for the 2-week treatment period. The changes in biochemical indices confirmed that all three regimens have an antiresorptive effect but both doses of etidronate were clearly more potent antiresorptive agents than alphacalcidol (changes in serum alkaline phosphatase activity: etidronate 200 mg, 11.6%; etidronate 400 mg, 17.5%; alphacalcidol,  4.9%). Urine hydroxyproline excretion declined significantly in both etidronate groups and there was a similar but nonsignificant trend in the alphacalcidol group. At the end of the 48-week treatment period, lumbar spine bone density increased 2.4% (P  0.001) in those receiving the 200-mg etidronate regimen, by 3.4% (P  0.001%) in those receiving the etidronate 400 mg regimen, and showed a nonsignificant decline ( 0.5%) in those receiving alphacalcidol. The changes in bone density were not significantly different between the two etidronate groups but both these groups showed more positive changes in bone density than the subjects receiving alphacalcidol (P  0.001). Metacarpal cortical thickness and mineral density did not differ between groups. Fracture rates were 6.9 per hundred patients in the low-dose etidronate group, 5.4 per hundred patients in the high-dose etidronate group, and 15 per hundred patients in the alphacalcidol group. The difference between high-dose etidronate and alphacalcidol groups was significant. Similar trends in fracture incidence were apparent when subjects in each group were divided according to the presence or absence of fractures at trial entry. There was significantly greater relief of pain in the etidronate-treated subjects. Both therapies were equally well tolerated. This study is impressive for its size and because it compares two different therapies. It suggests that a weak bisphosphonate has a greater therapeutic effect than a potent vitamin D metabolite. It is interesting that there were no changes in cortical bone mass in the alphacalcidol-treated subjects in this study, whereas earlier reports from the same authors had found quite substantial increments with this therapy. This might be accounted for by differences in the patient population (e.g., their baseline serum 25(OH)D concentrations) but there is insufficient detail in the report to explore this possibility. Since there was no placebo group in this study, it provides no information on the absolute effects of alphacalcidol, but the results demonstrate a therapeutic superiority of etidronate. Menczel et al. [114] compared alphacalcidol with placebo over a 3-year period. The alphacalcidol was taken

IAN R. REID

in a dose of 0.25 g twice a day and all subjects took a calcium supplement of 500 mg twice a day. Distal radial BMC was followed over a 3-year period. BMC remained stable in the alphacalcidol-treated subjects but declined at an average rate of 2.6%/year in those receiving calcium plus placebo (significant between-groups difference, P  0.05). There was a larger number of subjects in the control group than in the alphacalcidol group because the control group was pooled with that from a similar study recruited at the same time. It is stated that the two control groups were comparable but no data are presented to allow comparisons solely within the original randomized cohort. Lyritis et al. [123] compared the anabolic steroid, nandrolone, with alphacalcidol in a group of osteoporotic women. They found ongoing bone loss with the vitamin D metabolite but a 5% gain in forearm bone mineral content with nandrolone. There was no untreated comparator group. Pain and mobility were scored throughout the trial and a beneficial effect of nandrolone on both indices was observed. A number of small- to medium-sized studies have now assessed alphacalcidol using modern techniques of axial bone densitometry. Orimo et al. [115] reported a prospective, double-blind, placebo-controlled study of alphacalcidol 1 g/day or placebo, over a 12-month period. Bone density was measured by dual-energy X-ray absorptiometry of the lumbar spine and femur. Seventy-four patients completed the study but as few as 34 were included in some analyses (e.g., femoral bone density) because of missing data. In the lumbar spine, the change in bone density was 0.65% in the alphacalcidol group and  1.14% in the placebo-treated subjects (between-groups difference, P  0.04). In the femoral trochanter, bone density increased by 4.20% in the alphacalcidol group and declined by 2.37% in the control subjects (P  0.06). Similar trends were seen in the femoral neck and Ward’s triangle but the between-groups comparisons were not significant. Of the 25 alphacalcidol-treated subjects in whom fracture data were available, two new fractures occurred. There were 28 evaluable control subjects in whom eight fractures occurred. These data yield fracture rates of 75/1000 patient-years in the alphacalcidol group and 277 in the placebo group, a significant difference (P  0.03). However, the number of fractures at baseline was 50% higher in the control group. Though this was not statistically significant, it does suggest that the control group may have had a higher a priori fracture risk. Shiraki et al. [124] and Chen et al. [125] have also studied women with osteoporosis, and reported modest increases in spine BMD of about 2% at 1 – 2 years. In the Shiraki study, the numbers of subjects with data for each parameter reported varied widely because not all subjects underwent all assessments and because almost half of the bone density measurements were judged to be technically unsatisfactory. The Chen study achieved a 90% retention of

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

subjects, but was not blinded. Despite these differences, the results of these studies are very similar. There have been two recent studies of alphacalcidol in the prevention of early postmenopausal bone loss, both including groups treated with estrogen [126,127]. Neither demonstrated any advantage of alphacalcidol over calcium or control. In contrast, HRT prevented bone loss in both studies, and there was a nonsignificant trend for combination therapy to be superior, in the study of Gorai et al. [127]. A further double-blind, randomized, placebo-controlled trial has not been included in Table 2 since it is not strictly in postmenopausal osteoporosis [128]. This was carried out in 86 elderly patients (51 women) with Parkinson’s disease who were randomized to alphacalcidol 1 g/day or to placebo. The majority were vitamin D deficient at baseline. At 18 months, density of the second metacarpal decreased 1.2% in the alphacalcidol group compared with a loss of 6.7% in the placebo group (P  0.0001). Much of this response may represent the effect of treating vitamin D deficiency. 3. OTHER STUDIES Several other groups have studied alphacalcidol in nonrandomized patients. Lund et al. [116] described the successful treatment of a cohort of elderly patients whose main problem was probably osteomalacia. Marshall and Nordin [129] reported the effects of alphacalcidol 1 – 2 g/day in comparison or in combination with ethinyl estradiol 25 – 50 g/day. These studies were carried out in 33 postmenopausal women, most of whom had vertebral or proximal femoral fractures. Alphacalcidol 1 g/day produced no change in calcium balance but there was a trend toward positive calcium balance in patients treated with 2 g/day of the drug. Interestingly, this higher dose tended to increase urine hydroxyproline, whereas the lower dose significantly reduced it. Patients with low pretreatment intestinal calcium absorption showed more positive changes in calcium balance with alphacalcidol therapy. HRT consistently improved calcium balance whatever the pretreatment intestinal calcium absorption status, and the addition of alphacalcidol to hormone treatment did not improve the outcome. Subsequently, this group published a case series of women with vertebral fractures treated with a variety of regimens including alphacalcidol 1 – 2 g/day and alphacalcidol plus ethinyl estradiol 25 g/day. The treatment regimens were compared using sequential measurements of metacarpal cortical area. Changes in this index during treatment with alphacalcidol were indistinguishable from those seen in patients receiving no treatment. In contrast to the loss of cortical bone which occurred in these two groups, there were increases in cortical bone area in patients treated with hormones. The addition of alphacalcidol to hormone

569

treatment did not influence the effect of the hormones. This bone mass data, thus, produced essentially the same findings as the calcium balance studies – that hormone therapy produces substantial benefit and that alphacalcidol use is without effect. More recently, two further case series have been presented. Pouilles et al. [130] have demonstrated stability of spinal bone density assessed by dual-photon absorptiometry in 25 patients treated with alphacalcidol 1 g/day for 2 years. Control subjects lost bone over this period. Shiraki et al. [113] have published a retrospective case – control study from their osteoporosis clinic. They have compared changes in radial BMD over periods of up to 5 years in 26 patients treated with alphacalcidol (0.5 to 1.0 g/day) with those in age-matched subjects attending the same clinic who proved intolerant of all therapies offered to them. These untreated patients lost approximately 2% of their radial mineral density per annum, whereas the alphacalcidoltreated subjects gained approximately 1% per annum. The differences between the groups were statistically significant. Whereas the change in bone density in the control group was more or less linear over the 5-year period, most of the gain in the treated subjects occurred in the first 12 months, with stability of bone density subsequently. There is clearly the potential for major bias in such a highly selected retrospective series, since the factors that made the control subjects intolerant of all available therapies may also make them not comparable to the treated group. However, this series is informative in that it provides longer term data regarding effects on bone mass than are available from other sources. Since many of the prospective studies are very short-term this delineation of the pattern of change of bone density over time is of value. The pattern of change is consistent with that seen with other antiresorptive treatments, such as HRT and the bisphosphonates. It is in marked contrast to the progressive increases in bone density seen with a formation stimulating regimen such as fluoride, supporting the biochemical evidence that formation stimulation is not the mode of action of the vitamin D metabolites. 4. SAFETY As with calcitriol, the principal safety issue with alphacalcidol is the risk of hypercalciuria and hypercalcemia during treatment. The incidence of these problems will be influenced by the dose used, concomitant use of calcium supplements, dietary intake of calcium, and, possibly, pretreatment intestinal calcium absorption. With the regimen of alphacalcidol 2 g/day plus calcium 1 g/day used by Lund et al. [116], hypercalcemia occurred in half the treated subjects. This frequency was confirmed by Marshall and Nordin [129], who observed hypercalcemia in 3 of 7 patients treated with alphacalcidol 2 g/day but in only 3 of 11 patients treated with half that dose. Sorenson et al. [112]

570

IAN R. REID

observed serum calcium concentrations greater than 3 mmol/liter in ten 10 of 15 patients receiving 2 g/day but in 4 of 18 patients receiving 1 g of alphacalcidol daily. Both groups of patients received 1 g/day of supplemental calcium. Menczel et al. [114] using 0.5 g/day plus calcium, observed hypercalcemia in 46% of alphacalcidoltreated patients, but also in 36% of those receiving calcium alone, suggesting that blood samples were not taken in the fasting state. However, there was a more marked betweengroups difference in hypercalciuria. Forty-eight percent of alphacalcidol-treated patients had a 24-h urine calcium excretion  250 mg, whereas this occurred in only 11% of the control group. Despite the frequency of this problem, no patients developed symptomatic renal calculi during the study. Japanese patients tolerate alphacalcidol well and 1 g/day does not elevate their serum calcium concentration unless it is taken in combination with a calcium supplement [113,121]. Thus, only 1 patient of 38 treated with alphacalcidol 1 g/day developed hypercalcemia in one study [115]. This dose does significantly increase urine calcium excretion in Japanese women, but in the study of Orimo et al., all values remained within the normal range, consistent with the observation of only one renal stone in 8000 Japanese women using alphacalcidol over a period of 6 years [95].

D. Other Vitamin D Metabolites Few studies have been done in humans with the newer vitamin D metabolites. One exception is 1-hydroxyvitamin D2. Gallagher et al. [3] administered increasing doses of this compound to 15 postmenopausal women. In short-term studies they found a lower incidence of hypercalciuria and hypercalcemia than would be expected with comparable doses of alphacalcidol. Circulating osteocalcin was increased and other metabolic indices were unchanged. No bone density measurements were made. These data suggest that this compound is a weak vitamin D agonist and there is not, at the present time, any indication that it will offer significant advantages over the other metabolites already available. The study of Shiraki et al. [120] has been discussed above with respect to alphacalcidol. That study also demonstrated a beneficial effect from the use of 1,24R(OH)2D 1 g/day, but the effect was substantially less than that of the same dose of alphacalcidol. Calcifediol [25(OH)D] has been used in the therapy of steroid-induced osteoporosis and as a form of vitamin D replacement in the elderly [131], but is not generally available. A number of novel vitamin D-derived compounds are coming into use in the management of the secondary hyperparathyroidism of renal failure, but they have not been assessed in osteoporosis therapy.

E. Combination Regimens Vitamin D is often combined with calcium as cotherapy with fluoride. Historically this has been done because high doses of fluoride cause osteomalacia and vitamin D is an effective therapy for some forms of reduced bone mineralization. As discussed above, there is evidence from a nonrandomized retrospective report suggesting that calciferol is of value in this situation [43]. Recently it has been suggested that the role of vitamin D metabolites in this context may be to increase the supply of calcium to a skeleton that is rapidly increasing its mass in response to fluoride therapy. Vitamin D prevents the development of secondary hyperparathyroidism and the resulting loss of cortical bone [132]. There has been little systematic evaluation of combination regimens involving vitamin D metabolites in postmenopausal osteoporosis. Christiansen’s group published two studies, discussed above, in which calcitriol was combined with HRT [88,89]. Nordin also described the combinations of calciferol with calcium and alphacalcidol with HRT [42]. None of these studies provided evidence that combining the vitamin D metabolite with hormone therapy increased the benefit produced by HRT alone. Eriksson and Lindgren [133] assessed combination therapy with calcitonin and calcitriol and found no evidence of benefit from the combination, though the number of patients was small. Giannini et al. [134] reported a randomized comparison of cyclical clodronate therapy, with or without intermittent calcitriol, in an ADFR regimen. Both treatment protocols proved superior to no therapy, though the addition of calcitriol conferred no benefit over clodronate alone. In contrast, combining calcitriol with intermittent injections of PTH appeared in one study to be more effective than PTH alone, but the number of patients studied to date remains small [135]. Recently, the combination of calcitriol with bisphosphonates in patients with established osteoporosis has been assessed. Frediani [84] et al. randomized women to take calcium, calcitriol 0.5 g/day, alendronate (10 mg/day), or both (Table 1). At 2 years, the approximate changes in total body BMD were  2%,  2%,  4%, and  6%, respectively, the combination therapy being significantly better than any of the other interventions. Masud et al. [136] compared cyclical etidronate with this regimen plus calcitriol (Table 1). Again, there was a benefit of more than 2% in the BMD changes at both the spine and the hip. Similar results have been reported using a combination of calcitriol and HRT [137]. These findings would be consistent with the hypothesis that vitamin D metabolites can consistently have small positive effects on bone density when their property of stimulating bone resorption is blocked by the coadministration of an anti-resorptive agent.

571

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

In summary, the use of one of the potent short-acting metabolites of vitamin D is an attractive adjunctive therapy in any context in which it appears likely that intestinal calcium absorption will be a factor limiting the overall efficacy of an osteoporosis treatment regimen. There is room for much more research in this area.

V. OTHER OSTEOPOROSES There are few data regarding the effect of vitamin D or its metabolites in other forms of osteoporosis, except for steroid-induced osteoporosis which is dealt with elsewhere in this volume (Chapter 44). Orwoll et al. [138] studied the effects of calciferol 1000 IU/day plus calcium 1000 mg/day in a placebo-controlled trial in normal men ages 30 – 87 years. Seventy-seven men were studied over a 3-year period. There was no difference in rates of change of either radial or vertebral bone mineral density between the two groups. However, some of the more recent studies of vitamin D replacement have included men [65,69] and they suggest that the beneficial effects are uniform between the sexes (see Chapter 11). Ebeling [139] recently presented a preliminary account of a randomized, double-blind, placebo-controlled trial of calcitriol 0.5 g/day versus calcium 1 g/day in osteoporotic men with at least one baseline fracture. The calcium group showed transient positive changes in bone density in the hip and spine, though at 2 years there were no differences between the groups. Over the 2 years of the study, there were 15 vertebral and 6 nonvertebral fractures in the calcitriol group but only a single vertebral fracture in those taking calcium (P  0.03). Clearly, calcitriol should not be used in idiopathic male osteoporosis. There have been several case reports of calcitriol use in idiopathic juvenile osteoporosis [140]. Some have suggested a beneficial effect of calcitriol therapy. It is extremely difficult to infer cause and effect from these reports, however, since they are not controlled studies. Furthermore, unlike other forms of osteoporosis, bone density tends to return spontaneously to normal in idiopathic juvenile osteoporosis, making interpretation of these data even more difficult.

VI. CONCLUSIONS Among elderly subjects there is consistent evidence that vitamin D deficiency resulting in secondary hyperparathyroidism and accelerated bone loss is common. The treatment/prevention of this problem with physiological doses of calciferol (and possibly with calcium) confers benefits on bone mass and fracture incidence and should

be vigorously promoted as a general principle of patient care. The only thing that can be concluded with certainty with respect to vitamin D metabolite therapy for osteoporosis is that it remains an area of controversy. No evidence supports its use in the prevention of bone loss in normal postmenopausal women. In osteoporotic women, trials have produced variable results, except in Japan, where a consistent body of evidence demonstrates the efficacy of alphacalcidol on BMD. In those studies, beneficial effects on BMD have generally been less than are seen with HRT or bisphosphonates. Vitamin D metabolites may have additive effects on BMD to those of other antiresorptive agents, but there is no conclusive evidence for anti-fracture efficacy of these compounds. Remaining uncertainties regarding the active metabolites of vitamin D will require considerably more data for their resolution, but whether these studies will be completed or whether the vitamin D metabolites will be bypassed as more promising novel therapies are developed is unclear at the present time.

References 1. A. W. Norman, Problems relating to the definition of an international unit for vitamin D and its metabolites. J. Nutr. 102, 1243 – 1246 (1972). 2. L. Tjellesen, L. Hummer, C. Christiansen, and P. Rodbro, Serum concentration of vitamin D metabolites during treatment with vitamin D2 and D3 in normal premenopausal women. Bone Miner. 1, 407 – 413 (1986). 3. J. C. Gallagher, C. W. Bishop, J. C. Knutson, R. B. Mazess, and H. F. Deluca, Effects of increasing doses of 1 alpha-hydroxyvitamin D2 on calcium homeostasis in postmenopausal osteopenic women. J. Bone Miner. Res. 9, 607 – 614 (1994). 4. J. L. Sebert, M. Garabedian, C. De Chasteigner, and D. Defrance, Correction of vitamin D deficiency in the elderly: A comparative study of vitamin D2 and vitamin D3. J. Bone Miner. Res. 6 (Suppl. 1), S123 (1991). 5. J. R. Bullamore, R. Wilkinson, J. C. Gallagher, B. E. C. Nordin, and D. H. Marshall, Effects of age on calcium absorption. Lancet 2, 535 – 537 (1970). 6. J. C. Gallagher, L. B. Riggs, J. Eisman, A. Hamstra, S. B. Arnaud, and H. F. DeLuca, Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. J. Clin. Invest. 64, 729 – 736 (1979). 7. H. A. Morris, A. G. Need, M. Horowitz, P. D. O’Loughlin, and B. E. Nordin, Calcium absorption in normal and osteoporotic postmenopausal women. Calcif. Tissue Int. 49, 240 – 243 (1991). 8. R. M. Francis, M. Peacock, G. A. Taylor, J. H. Storer, and B. E. C. Nordin, Calcium malabsorption in elderly women with vertebral fractures: evidence for resistance to the action of vitamin D metabolites on the bowel. Clin. Sci. 66, 103 – 107 (1984). 9. M. W. Tilyard, G. F. Spears, J. Thomson, and S. Dovey, Treatment of postmenopausal osteoporosis with calcitriol or calcium. N. Engl. J. Med. 326, 357 – 362 (1992). 10. P. Lips, F. C. Van Ginkel, M. J. M. Jongen, F. Rubertus, W. J. F. Van der Vijgh, and J. C. Netelenbos, Determinants of vitamin

572

11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

25.

26.

27. 28.

IAN R. REID D status in patients with hip fracture and in elderly control subjects. Am. J. Clin. Nutr. 46, 1005 – 1010 (1987). R. Honkanen, E. Alhava, M. Parviainen, S. Talasniemi, and R. Monkkonen, The necessity and safety of calcium and vitamin D in the elderly. J. Am. Geriatr. Soc. 38, 862 – 866 (1990). A. R. Webb, C. Pilbeam, N. Hanafin, and M. F. Holick, An evaluation of the relative contributions of exposure to sunlight and of diet to the circulating concentrations of 25-hydroxyvitamin D in an elderly nursing home population in Boston. Am. J. Clin. Nutr. 51, 1075 – 1081 (1990). F. M. Gloth, 3d, J. D. Tobin, S. S. Sherman, and B. W. Hollis, Is the recommended daily allowance for vitamin D too low for the homebound elderly? J. Am. Geriatr. Soc. 39, 137 – 141 (1991). E. M. Keane, A. Rochfort, J. Cox, D. McGovern, D. Coakley, and J. B. Walsh, Vitamin-D-fortified liquid milk — A highly effective method of vitamin D administration for house-bound and institutionalised elderly. Gerontology 38, 280 – 284 (1992). C. T. McMurtry, S. E. Young, R. W. Downs, and R. A. Adler, Mild vitamin D deficiency and secondary hyperparathyroidism in nursing home patients receiving adequate dietary vitamin D. J. Am. Geriatr. Soc. 40, 343 – 347 (1992). K. J. O’Dowd, T. L. Clemens, J. L. Kelsey, and R. Lindsay, Exogenous calciferol (vitamin D) and vitamin D endocrine status among elderly nursing home residents in the New York City area. J. Am. Geriatr. Soc. 41, 414 – 421 (1993). R. P. J. Vanderwielen, M. R. H. Lowik, H. Vandenberg, L. C. P. G. M. Degroot, J. Haller, O. Moreiras, and W. A. Vanstaveren, Serum vitamin D concentrations among elderly people in Europe. Lancet 346, 207 – 210 (1995). F. M. Gloth, C. M. Gundberg, B. W. Hollis, J. G. Haddad, and J. D. Tobin, Vitamin D deficiency in homebound elderly persons. JAMA 274, 1683 – 1686 (1995). L. M. Salamone, G. E. Dallal, D. Zantos, F. Makrauer, and B. Dawson-Hughes, Contributions of vitamin D intake and seasonal sunlight exposure to plasma 25-hydroxyvitamin D concentration in elderly women. Am. J. Clin. Nutr. 58, 80 – 86 (1993). M. J. McKenna, Differences in vitamin D status between countries in young adults and the elderly. Am. J. Med. 93, 69 – 77 (1992). D. T. Villareal, R. Civitelli, A. Chines, and L. V. Avioli, Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J. Clin. Endocrinol. Metab. 72, 628 – 634 (1991). D. E. Meier, M. M. Luckey, S. Wallenstein, T. L. Clemens, E. S. Orwoll, and C. I. Waslien, Calcium, vitamin D, and parathyroid hormone status in young white and black women: Association with racial differences in bone mass. J. Clin. Endocrinol. Metab. 72, 703 – 710 (1991). I. R. Reid, R. Ames, M. C. Evans, S. Sharpe, G. Gamble, J. T. France, T. M. T. Lim, and T. F. Cundy, Determinants of total body and regional bone mineral density in normal postmenopausal women — A key role for fat mass. J. Clin. Endocrinol. Metab. 75, 45 – 51 (1992). K. T. Khaw, M.-J. Sneyd, and J. Compston, Bone density parathyroid hormone and 25-hydroxyvitamin D concentrations in middle aged women. Br. Med. J. 305, 273 – 277 (1992). A. M. Fehily, R. J. Coles, W. D. Evans, and P. C. Elwood, Factors affecting bone density in young adults. Am. J. Clin. Nutr. 56, 579 – 586 (1992). B. Lukert, J. Higgins, and M. Stoskopf, Menopausal bone loss is partially regulated by dietary intake of vitamin D. Calcif. Tissue Int. 51, 173 – 179 (1992). T. L. Holbrook and E. Barrett-Connor, Calcium intake: covariates and confounders. Am. J. Clin. Nutr. 53, 741 – 744 (1991). C. J. Rosen, A. Morrison, H. Zhou, D. Storm, S. J. Hunter, K. Musgrave, T. Chen, A. Wenwei, and M. F. Holick, Elderly women in

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44. 45.

46.

47.

northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner. 25, 83 – 92 (1994). B. J. Lund, O. H. Sorensen, B. I. Lund, and E. Agner, Serum 1,25dihydroxyvitamin D in normal subjects and in patients with postmenopausal osteopenia. Influence of age, renal function and oestrogen therapy. Hormone Metab. Res. 14, 271 – 274 (1982). R. J. Duda Jnr, R. Kumar, K. I. Nelson, A. R. Zinsmeister, K. G. Mann, and B. L. Riggs, 1,25-Dihydroxyvitamin D stimulation test for osteoblast function in normal and osteoporotic postmenopausal women. J. Clin. Invest. 79, 1249 – 1253 (1987). M. R. Sowers, R. B. Wallace, and B. W. Hollis, The relationship of 1,25-dihydroxyvitamin D and radial bone mass. Bone Miner. 10, 139 – 148 (1990). K.-S. Tsai, H. Heath III, R. Kumar, and B. L. Riggs, Impaired vitamin D metabolism with ageing in women. Possible role in pathogenesis of senile osteoporosis. J. Clin. Invest. 73, 1668 – 1672 (1984). B. L. Riggs, A. Hamstra, and H. F. DeLuca, Assessment of 25-hydroxyvitamin D 1-hydroxylase reserve in postmenopausal osteoporosis by administration of parathyroid extract. J. Clin. Endocrinol. Metab. 53, 833 – 835 (1981). D. M. Slovik, J. S. Adams, R. M. Neer, M. F. Holick, and J. T. J. Potts, Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 – 374 (1981). J. P. Brown, P. D. Delmas, M. Arlot, and P. J. Meunier, Active bone turnover of the cortico-endosteal envelope in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 64, 954 – 955 (1987). C. De la Piedra, R. Torres, A. Rapado, M. D. Curiel, and N. Castro, Serum tartrate-resistant acid phosphatase and bone mineral content in postmenopausal osteoporosis. Calcif. Tissue Int. 45, 58 – 60 (1989). M.-C. Faugere, S. Okamoto, H. F. DeLuca, and H. H. Malluche, Calcitriol corrects bone loss induced by oophorectomy in rats. Am. J. Physiol. 250, E35 – E38 (1986). B. Kohn, R. G. Erben, H. Weiser, W. A. Rambeck, and H. Zucker, Osteopenia caused by ovariectomy in young female rats and prophylactic effects of 1,25-dihydroxyvitamin D3. Zentralbl. Veterinarmed. A. 38, 54 – 60 (1991). R. G. Erben, H. Weiser, F. Sinowatz, W. A. Rambeck, and H. Zucker, Vitamin D metabolites prevent vertebral osteopenia in ovariectomized rats. Calcif. Tissue Int. 50, 228 – 236 (1992). T. J. Wronski, B. P. Halloran, D. D. Bikle, R. K. Globus, and E. R. Morey-Holton, Chronic administration of 1,25-dihydroxyvitamin D3: Increased bone but impaired mineralization. Endocrinology 119, 2580 – 2585 (1986). K. Buring, A. G. Hulth, B. E. Nilsson, N. E. Westlin, and P. E. Wiklund, Treatment of osteoporosis with vitamin D. Acta Med. Scand. 195, 471 – 472 (1974). B. E. C. Nordin, A. Horsman, R. G. Crilly, D. H. Marshall, and M. Simpson, Treatment of spinal osteoporosis in postmenopausal women. Br. Med. J. 1, 451 – 454 (1980). B. L. Riggs, E. Seeman, S. F. Hodgson, D. R. Taves, and W. M. O’ Fallon, Effect of the fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporosis. N. Engl. J. Med. 306, 444 – 450 (1982). A. Malabanan, I. E. Veronikis, and M. F. Holick, Redefining vitamin d insufficiency. Lancet 351, 805 – 806 (1998). B. Dawson-Hughes, S. S. Harris, and G. E. Dallal, Plasma calcidiol, season, and serum parathyroid hormone concentrations in healthy elderly men and women. Am. J. Clin. Nutr. 65, 67 – 71 (1997). I. R. Reid, D. J. A. Gallagher, and J. Bosworth, Prophylaxis against vitamin D deficiency in the elderly by regular sunlight exposure. Age Ageing 15, 35 – 40 (1986). R. Vieth, Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety. Am. J. Clin. Nutr. 69, 842 – 856 (1999).

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis 48. M. J. Barger-Lux, R. P. Heaney, S. Dowell, T. C. Chen, and M. F. Holick, Vitamin D and its major metabolites: Serum levels after graded oral dosing in healthy men. Osteoporosis Int. 8, 222 – 230 (1998). 49. M.-C. Chapuy, P. Chapuy, and P. J. Meunier, Calcium and vitamin D supplements: Effects on calcium metabolism in elderly people. Am. J. Clin. Nutr. 46, 324 – 328 (1987). 50. P. Lips, A. Wiersinga, F. C. Van Ginkel, M. J. M. Jongen, J. C. Netelenbos, W. H. L. Hackeng, P. D. Delmas, and W. J. F. Van der Vijgh, The effect of vitamin D supplementation on vitamin D status and parathyroid function in elderly subjects. J. Clin. Endocrinol. Metab. 67, 644 – 650 (1988). 51. F. Zeghoud, A. Jardel, M. Garabedian, R. Salvatore, and R. Moulias, Vitamin D supplementation in institutionalized elderly. Effects of vitamin D3 (100,000 IU) orally administered every 3 months on serum levels of 25-hydroxyvitamin D. Rev. Rhum. Mal. Osteoarticlaires 57, 809 – 813 (1990). 52. R. J. Heikinheimo, M. V. Haavisto, E. J. Harju, J. A. Inkovaara, R. H. Kaarela, L. A. Kolho, and S. A. Rajala, Serum vitamin D level after an annual intramuscular injection of ergocalciferol. Calcif. Tissue Int. 49 (Suppl.), S87 (1991). 53. J. S. Adams, V. Kantorovich, C. Wu, M. Javanbakht, and B. W. Hollis, Resolution of vitamin D insufficiency in osteopenic patients results in rapid recovery of bone mineral density. J. Clin. Endocrinol. Metab. 84, 2729 – 2730 (1999). 54. M. J. McKenna, R. Freaney, R. Byrne, Y. McBrinn, B. Murray, M. Kelly, B. Donne, and M. Obrien, Safety and efficacy of increasing wintertime vitamin d and calcium intake by milk fortification. Mon. J. Assoc. Physicians 88, 895 – 898 (1995). 55. C. Christiansen, M. S. Christensen, P. McNair, et al., Prevention of early postmenopausal bone loss: Controlled 2-year study in normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). 56. M. Komulainen, M. T. Tuppurainen, H. Kroger, A. M. Heikkinen, E. Puntila, E. Alhava, R. Honkanen, and S. Saarikoski, Vitamin D and HRT — No benefit additional to that of HRT alone in prevention of bone loss in early postmenopausal women — A 2.5-year randomized placebo-controlled study. Osteoporos. Int. 7, 126 – 132 (1997). 57. A. M. Heikkinen, M. Parviainen, L. Niskanen, M. Komulainen, M. T. Tuppurainen, H. Kroger, and S. Saarikoski, Biochemical bone markers and bone mineral density during postmenopausal hormone replacement therapy with and without vitamin d-3 – A prospective, controlled, randomized study. J. Clin. Endocrinol. Metab. 82, 2476 – 2482 (1997). 58. M. H. Komulainen, H. Kroger, M. T. Tuppurainen, A. M. Heikkinen, E. Alhava, R. Honkanen, and S. Saarikoski, HRT and Vit D in prevention of non-vertebral fractures in postmenopausal women: A 5 year randomized trial. Maturitas 31, 45 – 54 (1998). 59. A. M. Heikkinen, M. T. Parviainen, M. T. Tuppurainen, L. Niskanen, M. H. Komulainen, and S. Saarikoski, Effects of postmenopausal hormone replacement therapy with and without vitamin d-3 on circulating levels of 25-hydroxyvitamin d and 1,25-dihydroxyvitamin d. Calcif. Tissue Int. 62, 26 – 30 (1998). 60. B. E. C. Nordin, M. R. Baker, A. Horsman, and M. Peacock, A prospective trial of the effect of vitamin D supplementation on metacarpal bone loss in elderly women. Am. J. Clin. Nutr. 42, 470 – 474 (1985). 61. M. E. Ooms, J. C. Roos, P. D. Bezemer, W. J. F. Vandervijgh, L. M. Bouter, and P. Lips, Prevention of bone loss by vitamin D supplementation in elderly women: A randomized double-blind trial. J. Clin. Endocrinol. Metab. 80, 1052 – 1058 (1995). 62. B. Dawson-Hughes, S. S. Harris, E. A. Krall, G. E. Dallal, G. Falconer, and C. L. Green, Rates of bone loss in postmenopausal women randomly assigned to one of two dosages of vitamin D. Am. J. Clin. Nutr. 61, 1140 – 1145 (1995).

573

63. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, S. Harris, L. J. Sokoll, and G. Falconer, Effect of vitamin D supplementation on wintertime and overall bone loss in healthy postmenopausal women [see Comments]. Ann. Intern. Med. 115, 505 – 512 (1991). 64. L. Baeksgaard, K. P. Andersen, and L. Hyldstrup, Calcium and vitamin D supplementation increases spinal BMD in healthy, postmenopausal women. Osteoporosis. Int. 8, 255 – 260 (1998). 65. R. J. Heikinheimo, J. A. Inkovaara, E. J. Harju, M. V. Haavisto, R. H. Kaarela, J. M. Kataja, A. M. Kokko, L. A. Kolho, and S. A. Rajala, Annual injection of vitamin D and fractures of aged bones. Calcif. Tissue Int. 51, 105 – 110 (1992). 66. P. Lips, W. C. Graafmans, M. E. Ooms, P. D. Bezemer, and L. M. Bouter, Vitamin D supplementation and fracture incidence in elderly persons — A randomized, placebo-controlled clinical trial. Ann. Intern. Med. 124, 400 – 406 (1996). 67. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in the elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 68. M. C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994). 69. B. Dawson-Hughes, S. S. Harris, E. A. Krall, and G. E. Dallal, Effect of calcium and vitamin D supplementation on bone, density in men and women 65 years of age or older. N. Engl. J. Med. 337, 670 – 676 (1997). 70. R. R. Recker, D. B. Kimmel, S. Hinders, and K. M. Davies, Antifracture efficacy of calcium in elderly women. J. Bone Miner. Res. 9 (Suppl. 1), s154 (1994). 71. I. R. Reid, R. W. Ames, M. C. Evans, G. D. Gamble, and S. J. Sharpe, Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women- a randomized controlled trial. Am. J. Med. 98, 331 – 335 (1995). 72. T. Chevalley, R. Rizzoli, V. Nydegger, D. Slosman, C. H. Rapin, J. P. Michel, H. Vasey, and J. P. Bonjour, Effects of calcium supplements on femoral bone mineral density and vertebral fracture rate in vitamin-D-replete elderly patients. Osteoporosis Int. 4, 245 – 252 (1994). 73. D. J. Hosking, G. A. Campbell, J. R. Kemm, R. E. Cotton, and R. V. Boyd, Safety of treatment for subclinical osteomalacia in the elderly. Br. Med. J. 289, 785 – 787 (1984). 74. M. S. Schwartzman and W. A. Franck, Vitamin D toxicity complicating the treatment of senile, postmenopausal, and glucocorticoidinduced osteoporosis. Am. J. Med. 82, 224 – 230 (1987). 75. R. Rizzoli, C. Stoermann, P. Ammann, and J. P. Bonjour, Hypercalcemia and hyperosteolysis in vitamin D intoxication — Effects of clodronate therapy. Bone 15, 193 – 198 (1994). 76. Y. B. Jiang, Y. Z. Wang, J. Zhao, G. Marchal, Y. X. Wang, Y. L. Shen, S. Z. Xing, R. G. Li, and A. L. Baert, Bone remodeling in hypervitaminosis D3. Radiologic – microangiographic – pathologic correlations. Invest. Radiol. 26, 213 – 219 (1991). 77. A. G. Need, M. Horowitz, J. C. Philcox, and B. E. C. Nordin, 1,25Dihydroxycalciferol and calcium therapy in osteoporosis with calcium malabsorption. Miner. Electrolyte Metab. 11, 35 – 40 (1985). 78. J. F. Aloia, A. Vaswani, J. K. Yeh, K. Ellis, S. Yasumura, and S. H. Cohn, Calcitriol in the treatment of postmenopausal osteoporosis. Am. J. Med. 84, 401 – 408 (1988). 79. S. M. Ott and C. H. Chesnut, Calcitriol treatment is not effective in postmenopausal osteoporosis. Ann. Intern. Med. 110, 267 – 274 (1989). 80. A. Caniggia, R. Nuti, F. Lore, G. Martini, V. Turchetti, and G. Righi, Long-term treatment with calcitriol in postmenopausal osteoporosis. Metabolism 39, 43 – 49 (1990). 81. B. L. Riggs and K. I. Nelson, Effect of long term treatment with calcitriol on calcium absorption and mineral metabolism in

574

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95. 96.

97.

98. 99.

IAN R. REID postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 61, 457 – 461 (1985). J. E. Zerwekh, K. Sakhaee, and C. Y. C. Pak, Short-term 1,25-dihydroxyvitamin D3 administration raises serum osteocalcin in patients with postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 60, 615 – 617 (1985). J. C. Gallagher, C. M. Jerpbak, W. S. S. Jee, K. A. Johnson, H. F. DeLuca, and B. L. Riggs, 1,25-Dihydroxyvitamin D3:Short and long-term effects on bone and calcium metabolism in patients with postmenopausal osteoporosis. Proc. Natl. Acad. Sci. USA 79, 3325 – 3329 (1982). B. Frediani, A. Allegri, S. Bisogno, and R. Marcolongo, Effects of combined treatment with calcitriol plus alendronate on bone mass and bone turnover in postmenopausal osteoporosis two years of continuous treatment. Clin. Drug Invest. 15, 235 – 244 (1998). I. R. Reid, G. E. Chapman, T. R. C. Fraser, A. D. Davies, A. S. Surus, J. Meyer, N. L. Huq, and H. K. Ibbertson, Low serum osteocalcin levels in glucocorticoid-treated asthmatics. J. Clin. Endocrinol. Metab. 62, 379 – 383 (1986). A. Caniggia, R. Nuti, M. Galli, F. Lore, V. Turchetti, and G. A. Righi, Effect of long-term treatment with 1,25-dihydroxyvitamin D3 on osteocalcin in postmenopausal osteoporosis. Calcif. Tissue Int. 38, 328 – 332 (1986). N. A. Morrison, J. Shine, J. C. Fragonas, V. Verkest, M. L. McMenemy, and J. A. Eisman, 1,25-Dihydroxyvitamin D-responsive element and glucocorticoid repression in the osteocalcin gene. Science 246, 1158 – 1161 (1989). C. Christiansen, M. S. Christensen, P. Rodbro, C. Hagen, and I. Transbol, Effect of 1,25-dihydroxy-vitamin D3 in itself or combined with hormone treatment in preventing postmenopausal osteoporosis. Eur. J. Clin. Invest. 11, 305 – 309 (1981). G. F. Jensen, C. Christiansen, and I. Transbol, Treatment of postmenopausal osteoporosis. A controlled therapeutic trial comparing oestrogen/gestagen,1,25-dihydroxy-vitamin D3 and calcium. Clin. Endocrinol. 16, 515 – 524 (1982). G. F. Jensen, B. Meinecke, J. Boesen, and I. Transbol, Does 1,25(OH)2 D3 accelerate spinal bone loss? Clin. Orthop. 192, 215 – 221 (1985). J. A. Falch, O. R. Odegaard, M. Finnanger, and I. Matheson, Postmenopausal osteoporosis: No effect of three years treatment with 1,25dihydroxycholecalciferol. Acta Med. Scand. 221, 199 – 204 (1987). J. C. Gallagher, B. L. Riggs, R. R. Recker, and D. Goldgar, The effect of calcitriol on patients with postmenopausal osteoporosis with special reference to fracture frequency. Proc. Soc. Exp. Biol. Med. 191, 287 – 292 (1989). J. C. Gallagher and D. Goldgar, Treatment of postmenopausal osteoporosis with high doses of synthetic calcitriol. A randomized controlled study. Ann. Intern. Med. 113, 649 – 655 (1990). S. M. Ott and C. H. Chestnut, Tolerance to dose of calcitriol is associated with improved density in women with postmenopausal osteoporosis. J. Bone Miner. Res. 5 (Suppl. 2), s186 (1990). T. Fujita, Studies of osteoporosis in Japan. Metabolism 39 (Suppl. 1), 39 – 42 (1990). R. S. Arthur, B. Piraino, D. Candib, L. Cooperstein, T. Chen, C. West, and J. Puschett, Effect of low-dose calcitriol and calcium therapy on bone histomorphometry and urinary calcium excretion in osteopenic women. Miner. Electrolyte Metab. 16, 385 – 390 (1990). M. W. Tilyard, G. F. S. Spears, J. Thomson, and S. Dovey, Calcitriol or calcium for postmenopausal osteoporosis. N. Engl. J. Med. 327, 284 (1992). A. Fournier, and J. L. Sebert, Calcitriol or calcium for postmenopausal osteoporosis. N. Engl. J. Med. 327, 284 (1992). M. W. Tilyard, 1,25-Dihydroxyvitamin D vs calcium in the treatment of established postmenopausal osteoporosis. J. Bone Miner. Res. 5 (Suppl. 2), s275 (1990).

100. A. J. Fenton, N. Gilchrist, B. Shand, J. Aitken, W. Gilchrist, J. Singh, and N. Graham, A comparison of calcitriol and hormone replacement therapy in post-menopausal women with osteoporosis. Bone 23 (Suppl. 1), s641 (1998). 101. J. C. Gallagher and S. Fowler, Effect of estrogen, calcitriol and a combination of estrogen and calcitriol on bone mineral density and fractures in elderly women. J. Bone Miner. Res. 14 (Suppl. 1), s209 (1999). 102. A. Caniggia, R. Nuti, F. Lore, and A. Vattimo, The hormonal form of vitamin D in the pathophysiology and therapy of postmenopausal osteoporosis. J. Endocrinol. Invest. 7, 373 – 378 (1984). 103. R. Nuti, G. Martini, R. Valenti, and S. Giovani, Open-label, controlled study on the metabolic and absorptiometric effects of calcitriol in involutional osteoporosis. Clin. Drug Invest. 11, 270 – 277 (1996). 104. A. G. Need, H. A. Morris, M. Horowitz, and B. E. C. Nordin, The response to calcitriol therapy in postmenopausal osteoporotic women is a function of initial calcium absorptive status. Calcif. Tissue Int. 61, 6 – 9 (1997). 105. B. E. C. Nordin, Osteoporosis. In “Metabolic Bone and Stone Disease” (B. E. C. Nordin, ed.). Churchill Livingstone, Edinburgh, 1993. 106. A. G. Need, B. E. C. Nordin, M. Horowitz, and H. A. Morris, Calcium and calcitriol therapy in osteoporotic postmenopausal women with impaired calcium absorption. Metabolism 39S1, 53 – 54 (1990). 107. B. E. C. Nordin, and H. A. Morris, Osteoporosis and vitamin D. J. Cell. Biochem. 49, 19 – 25 (1992). 108. G. Mathur, P. Clifton-Bligh, G. Fulcher, J. Stiel, and A. McElduff, Calcitriol in osteoporosis — is it used safely. Aust. NZ J. Med. 28, 348 – 349 (1998). 109. S. J. Gallacher, R. A. Cowan, W. D. Fraser, F. C. Logue, A. Jenkins, and I. T. Boyle, Acute effects of intravenous 1 alpha-hydroxycholecarciferol on parathyroid hormone, osteocalcin and calcitriol in man. Eur. J. Endocrinology 130, 141 – 145 (1994). 110. M. Peacock, Action of 1-alpha hydroxyvitamin D3 on calcium absorption and bone resorption in man. Lancet 1, 385 – 389 (1974). 111. N. H. Cohen, D. Farrah, I. Fogelman, C. C. Goll, G. H. Beastall, W. B. McIntosh, M. Fletcher, and I. T. Boyle, A low dose regime of 1 hydroxyvitamin D3 in the managment of senile osteoporosis: A pilot study. Clin. Endocrinol. 12, 537 – 542 (1980). 112. O. H. Sorenson, R. B. Andersen, M. S. Christensen, T. Friis, L. Hjorth, F. S. Jorgensen, B. Lund, F. Melsen, and L. Mosekilde, Treatment of senile osteoporosis with 1 -hydroxyvitamin D3. Clin. Endocrinol. 7 (Suppl.), 169s – 175s (1977). 113. M. Shiraki, H. Ito, and H. Orimo, The ultra long-term treatment of senile osteoporosis with 1 alpha-hydroxyvitamin D3. Bone Miner. 20, 223 – 234 (1993). 114. J. Menczel, J. Foldes, R. Steinberg, I. Leichter, B. Shalita, T. Bdolahabram, S. Kadosh, Z. Mazor, and D. Ladkani, Alfacalcidol (Alpha D3) and calcium in osteoporosis. Clin. Orthop. 300, 241 – 247 (1994). 115. H. Orimo, M. Shiraki, Y. Hayashi, T. Hoshino, T. Onaya, S. Miyazaki, H. Kurosawa, T. Nakamura, and N. Ogawa, Effects of 1 alpha-hydroxyvitamin D-3 on lumbar bone mineral density and vertebral fractures in patients with postmenopausal osteoporosis. Calcif. Tissue Int. 54, 370 – 376 (1994). 116. B. Lund, I. Kjaer, L. Hjorth, I. Reimann, R. B. Andersen, T. Friis, and O. H. Sorensen, Treatment of osteoporosis of ageing with 1 hyroxycholecalciferol. Lancet 2, 1168 – 1171 (1975). 117. V. Hoikka, E. M. Alhava, A. Aro, P. Karjalainen, and V. Rehnberg, Treatment of osteoporosis with 1-alpha-hydroxycholecalciferol and calcium. Acta Med. Scand. 207, 221 – 224 (1980). 118. Y. Itami, T. Fujita, T. Inoue, et al., Effect of alphacalcidol on osteoporosis. J. Clin. Exp. Med. 123, 958 – 973 (1982). 119. T. Fujita, T. Inoue, H. Orimo, et al., Clinical evaluation of the effect of calcitriol on osteoporosis. Multicenter double-blind study using

CHAPTER 68 Vitamin D and Its Metabolites in the Management of Osteoporosis

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

alphacalcidol as the control drug. J. Clin. Exp. Med. 148, 833 – 857 (1989). M. Shiraki, H. Orimo, H. Ito, I. Akiguchi, J. Nakao, R. Takahashi, and S. Ishizuka, Long-term treatment of postmenopausal osteoporosis with active vitamin D3,1-alpha-hydroxycholcalciferol (1alphaOHD3) and 1,24 dihydroxycholcalciferol (1,24(OH)2D3). Endocrinol. Jpn. 32, 305 – 315 (1985). h. Orimo, M. Shiraki, T. Hayashi, and T. Nakamura, Reduced occurrence of vertebral crush fractures in senile osteoporosis treated with 1(OH)-vitamin D3. Bone Miner. 3, 47 – 52 (1987). T. Fujita, H. Orimo, T. Inoue, et al., Double-blind multicenter comparative study of alfacalcidol with etidronate disodium (EHDP) in involutional osteoporosis. Clinical Evaluations 21, 261 – 302 (1993). G. P. Lyritis, C. Androulakis, B. Magiasis, Z. Charalambaki, and N. Tsakalakos, Effect of nandrolone decanoate and 1-alpha-hydroxy-calciferol on patients with vertebral osteoporotic collapse. A double-blind clinical trial. Bone Miner. 27, 209 – 217 (1994). I. Paspati, A. Galanos, and G. P. Lyritis, Hip fracture epidemiology in greece during 1977 – 1992. Calcif. Tissue Int. 62, 542 – 547 (1998). J. T. Chen, M. Shiraki, K. Hasumi, N. Tanaka, K. Katase, T. Kato, Y. Hirai, T. Nakamura, and E. Ogata, 1-Alpha-hydroxyvitamin d-3 treatment decreases bone turnover and modulates calcium-regulating hormones in early postmenopausal women. Bone 20, 557 – 562 (1997). H. Itoi, H. Minakami, and I. Sato, Comparison of the long-term effects of oral estriol with the effects of conjugated estrogen, 1-alphahydroxyvitamin d-3 and calcium lactate on vertebral bone loss in early menopausal women. Maturitas 28, 11 – 17 (1997). I. Gorai, O. Chaki, Y. Taguchi, M. Nakayama, H. Osada, N. Suzuki, N. Katagiri, Y. Misu, and H. Minaguchi, Early postmenopausal bone loss is prevented by estrogen and partially by 1 alpha-OH-vitamin D-3: Therapeutic effects of estrogen and or 1 alpha-OH-vitamin D-3. Calcif. Tissue Int. 65, 16 – 22 (1999). Y. Sato, S. Manabe, H. Kuno, and K. Oizumi, Amelioration of osteopenia and hypovitaminosis D by 1 alpha-hydroxyvitamin D3 in elderly patients with Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry 66, 64 – 68 (1999). D. H. Marshall and B. E. C. Nordin, The effect of 1-hydroxyvitamin D3 with and without oestrogens on calcium balance in postmenopausal women. Clin. Endocrinol. 7, 159s – 168s (1977). J. M. Pouilles, F. Tremollieres, and C. Ribot, Prevention of postmenopausal bone loss with 1 alpha-hydroxy vitamin D3. A threeyear prospective study. Clin. Rheumatol. 11, 492 – 497 (1992). M. Peacock, S. Hui, G. Liu, M. Carey, W. Ambrosius, and C. C. Johnston, Effect of oral calcium and vitamin D on bone density at the hip in elderly men and women. J. Bone Miner. Res. 12 (Suppl. 1), s129 (1997).

575

132. B. A. Dure-Smith, S. M. Farley, S. G. Linkhart, and D. J. Baylink, Fluoride treated patients become calcium deficient despite calcium supplements: Correction with 1,25 vitamin D3. In “Proceedings 1993 Fourth International Symposium on Osteoporosis and Consensus Development Conference. Fourth International Symposium on Osteoporosis, Rodovre, 1993.” (C. Christiansen and B. Riis, eds.). 1993. 133. S. A. Eriksson and J. U. Lindgren, Combined treatment with calcitonin and 1,25-dihydroxyvitamin D3 for osteoporosis in women. Calcif. Tissue Int. 53, 26 – 28 (1993). 134. S. Giannini, A. D’Angelo, L. Malvasi, R. Castrignano, T. Pati, R. Tronca, L. Liberto, M. Nobile, and G. Crepaldi, Effects of oneyear cyclical treatment with clodronate on postmenopausal bone loss. Bone 14, 137 – 141 (1993). 135. D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. T. Potts, M. A. Daly, J. A. Campbell, and R. M. Neer, Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1-34) and 1,25 dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 381 (1986). 136. T. Masud, B. Mulcahy, A. V. Thompson, S. Donnelly, R. W. Keen, D. V. Doyle, and T. D. Spector, Effects of cyclical etidronate combined with calcitriol versus cyclical etidronate alone on spine and femoral neck bone mineral density in postmenopausal osteoporotic women. Ann. Rheum. Dis. 57, 346 – 349 (1998). 137. D. H. Gutteridge, M. Holzherr, R. Will, G. O. Stewart, R. I. Price, R. L. Prince, R. W. Retallack, L. Tran, B. G. A. Stuckey, P. Drury, D. Faulkner, R. A. Criddle, C. I. Bhagat, G. N. Kent, K. Jamrozik, and S. S. Dhaliwal, Postmenopausal vertebral fractures — Advantage of HRT plus calcitriol, over HRT alone, at total body and hip in malabsorbers and normal absorbers of Ca. Bone 23 (Suppl.), S527 (1998). 138. E. S. Orwoll, S. K. Oviatt, M. R. McLung, L. J. Deftos, and G. Sexton, The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann. Intern. Med. 112, 29 – 34 (1990). 139. P. R. Ebeling, S. Yeung, C. Poon, N. Salehi, and M. A. Kotowicz, Effects of baseline active calcium absorption on bone mineral density (BMD) responses to calcitriol or calcium treatment in men with idiopathic osteoporosis. J. Bone Miner. Res. 14 (Suppl. 1), s414 (1999). 140. G. Saggese, S. Bertelloni, G. I. Baroncelli, G. Perri, and A. Calderazzi, Mineral metabolism and calcitriol therapy in idiopathic juvenile osteoporosis. Am. J. Dis. Child. 145, 457 – 462 (1991). 141. M. Shiraki, K. Kushida, K. Yamazaki, T. Nagai, T. Inoue, and H. Orimo, Effects of 2 years treatment of osteoporosis with 1-alphahydroxy vitamin D3 on bone mineral density and incidence of fracture — A placebo-controlled, double-blind prospective study. Endocr. J. 43, 211 – 220 (1996).

CHAPTER 69

Estrogens and Osteoporosis ROBERT LINDSAY AND FELICIA COSMAN Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993

I. Introduction II. Epidemiology III. Pathophysiology

IV. Effects of Estrogen Intervention V. Summary References

I. INTRODUCTION Although information about fractures occurring as consequence of osteoporosis can be found throughout medical history, delineation of the changes that occur in calcium homeostasis resulting in bone loss consequent upon loss of sex steroids belongs to Fuller Albright [1]. Albright demonstrated that vertebral fractures were more common among women ovariectomized prior to the expected age of menopause. He showed that those women who presented with fractures were in negative calcium balance, and that calcium balance could be corrected by the use of estrogens [2]. Although there have now been several large clinical trials performed since that time, it is only recently that we have become more knowledgeable about estrogens and their effects on the skeleton than we were at the conclusion of Albright’s career. Some areas that are poorly understood persist. We are still in doubt about the cellular target or targets, the second messengers involved, and the precise effects of intervention on fracture. Despite gaps in our understanding as fundamental as these it is clear that estrogen deficiency is a major cause of fractures. Historically, intervention with estrogens or hormone replacement therapy (HRT) has been considered to be the gold standard for prevention and treatment of osteoporosis among postmenopausal women. The relative paucity of fracture outcome data from randomized clinical trials, however, led to the recent decision by the FDA to rescind estrogen’s osteo-

OSTEOPOROSIS, SECOND EDITION VOLUME 2

577

porosis treatment indication. With respect to bone, estrogens are now indicated in the United States only for osteoporosis prevention. There is also an increasing awareness that estrogens may be important in osteoporosis among males. The development of novel effective therapeutic approaches to osteoporosis underscores the importance of understanding both the mechanism of estrogen’s effects and also the possible effectiveness of estrogen intervention in fracture reduction,

II. EPIDEMIOLOGY In countries such as the United States and in many European countries fractures as a consequence of osteoporosis occur more commonly among women than men [3] (Fig. 1) (see also Chapters 38 and 42).Women are also affected at a younger age than men, there being about a 10-year difference in the age-related increment in hip fracture for example [4]. Vertebral fractures are commonly stated to be 2 – 8 times more common among women [3], and indeed any clinical practice that specializes in the management of patients with osteoporosis may see as many as 10 women for every affected male. However, some recent data from Europe have suggested that in a population base, vertebral fractures are more common among males [5]. Also evaluations from societies with differing lifestyles from the usual in either the United States or most of Europe indicate that the sexual Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

578

LINDSAY AND COSMAN

FIGURE 1

Epidemiology of vertebral, hip, and Colle’s fractures with age in men and women from Rochester, Minnesota. Reprinted with permission from Melton [3].

dichotomy may be mitigated perhaps by lifestyle features such high level of physical activity throughout the lifespan [6]. Data from other countries including parts of Asia also suggest that sex differences in fracture risk are less than commonly assumed, and may be considerably less than the variability seen among fractures across cultures [5]. Hip fracture risk is also greater among women than men in most, but not all, cultures [6 – 8]. In part this variability can be explained by differences in the numbers at risk due to different life expectancies [8]. In developed countries the ratio is approximately 2 – 3:1 in favor of women [3,9]. Colles’ fractures are also more common among women than men, although the epidemiology of wrist fractures differs from that of hip and spine fractures. The risk of wrist fractures increases among women prior to age of menopause, and plateaus in the late 50s without a further age related increase. There is little, if any, age-related increase in risk of wrist fractures among males [10]. Whether these epidemiological findings for the major osteoporosis related fractures are related purely to estrogen deficiency at time of menopause seems unlikely. The likelihood that a bone will fracture depends upon its mass, its configuration (size), and the trauma to which it is subjected. In general the female skeleton is smaller than that of the male, and gender differences in bone density disappear or lessen substantially when appropriate corrections are made for differences in bone size [11]. It is clear that there are differences in bone loss after age 50 that account in part for the sexual dichotomy in fractures, but other factors also play a significant role in determining bone loss. Thus, to expect that introduction of estrogen in replacement doses to postmenopausal women will eliminate the problem of fractures among aging females is naive in the extreme. Nor does the epidemiology of Colles’ fracture lend itself

easily to the estrogen deficiency hypothesis, since the increased risk emerges before menopause. Whether this is due to increased remodeling (during the few years prior to menopause), increased falling, or some other factor remains unexplained [10].

III. PATHOPHYSIOLOGY A. Hormonal Changes From roughly the age of 40 there is a gradual decline in ovarian function directly resulting from a decline in the number of follicles available for maturation [12]. The result is a gradual increase in the number of anovulatory cycles, and a gradual failure of ovarian secretion of estrogen and progesterone [13]. As estrogen concentrations fall, there is a compensatory increase in pituitary secretion of folliclestimulating hormone (FSH) in a vain attempt to drive the ovary to produce more estrogen. Menopause occurs when the ovary finally cannot muster follicular development. All women experience a dramatic decline in sex hormone status at the time menopause (average age, 51 years). The outward sign of this is cessation of menses, which is a marker for women and physicians that the woman has progressed into the postmenopausal period of life. It must be appreciated, however, that this is a gradual process beginning several years before the overt sign, and indeed continuing even after apparent menopause, when ovarian development of a remaining follicle may occur sporadically, giving rise to vaginal bleeding months or even a year or more later [14]. Ovariectomy performed prior to overt menopause produces a rapid progression from the premenopausal state to postmenopausal.

579

CHAPTER 69 Estrogens and Osteoporosis

The transition to postmenopause results in significant alterations in sex hormone supply [13]. Most dramatic are the reductions in circulating estradiol and progesterone. Cyclical production of estradiol by the ovary in premenopausal women maintains serum estradiol concentrations above 50 pg/ml with ovulatory peaks above 200 pg/ml. In postmenopausal women serum estradiol is usually below 20 pg/ml [13]. Estrone values also decline, but to a lesser extent, such that in the postmenopause estrone becomes the major circulating estrogen [13]. The postovulatory increase in progesterone secretion gradually fails with ovulatory failure, and there is modest, if any, secretion of progesterone in postmenopausal women [16]. Androgen concentrations also decline across menopause, although androgen secretion by the ovarian stroma continues. The adrenal supplies the majority as androstenedione, with some secretion of testosterone also [13,17]. These androgens are the major source of estrogen supply in postmenopausal women. Peripheral conversion to estrone and estradiol, respectively, occurs primarily in fat but also in a variety of other tissues, including bone and marrow [18 – 20]. Ovariectomy results in a greater decline in androgen supply than the transition through menopause. The failure of the negative feedback loop primarily results in high circulating levels of FSH and LH [21], although other mechanisms may also contribute [22].

of bone tissue by eliminating the template upon which new bone can be laid down. With both greater activation frequency and excessive osteoclast resorption there is a higher chance of trabecular perforation (Fig. 2). The extent to which these processes contribute to continued bone loss is likely to be highly variable among individuals and influenced by a variety of factors including calcium supply, physical activity, cigarette consumption, alcohol intake, concomitant illnesses, and the drugs used to treat them. It should come as no surprise that the extent and duration of bone loss must also vary among individuals and populations as will the clinical expression of the disease as fracture. 1. TARGET CELLS The mechanisms by which these effects occur are still not well understood. A large number of effects of estrogen on a variety of cells that might be incriminated have been described. However, the exact target cell or cells remains an

B. Skeletal Homeostasis and Calcium Balance The transition across menopause results in profound alterations in skeletal homeostasis. Albright described this as an increased supply of calcium from the skeleton to balance the increased renal calcium loss not offset by increased absorption of calcium across the intestine [1]. We now recognize these changes in calcium homeostasis as signs of increased activation of remodeling within the skeleton [23]. Since remodeling is a phenomenon that is initiated on the surface of bone, the effects of increased activation are most clearly evident in trabecular bone, where the surface to volume ratio is significantly greater than in cortical bone. The consequence of increased activation is a reduction in bone mass. The decline mediated in this fashion is transient and due to a greater portion of the skeleton undergoing resorption at any time point [24]. However, bone loss is an ongoing process among postmenopausal women, suggesting that in addition to increased activation there must also be an inherent defect within the remodeling process. By some mechanism there must be more bone–removed than is subsequently synthesized. Histomorphometry suggests that there is excessive osteoclast action that is not compensated by increased osteoblast activity. Consequently, there is a net deficit at the end of each remodeling cycle [23]. Increased activation of remodeling can also contribute to ongoing loss

FIGURE 2

Scanning electron micrographs of normal and osteoporotic cancellous bone. (Top) Normal cancellous bone has trabecular plates interconnecting with small regular marrow spaces. (Bottom) As bone is lost the trabeculae are penetrated, become rod-like, and eventually are completely resorbed. All stages of this process can be observed. Reprinted with permission from Dempster et al. [119].

580

LINDSAY AND COSMAN

enigma. Multiple candidates exist, including osteoblasts and their relatives, the cells lining inactive bone surfaces presumed to be osteoblasts in an inactive state, osteoclasts, osteocytes, and myriad cells within marrow including precursor cells of both bone and hemopoietic systems, mast cells and other cells of the immune system, fibroblasts, and perhaps even adipocytes. In many cases cells from these lineages express estrogen receptors and can mount cellular responses to estrogen exposure.These vary from modifications of cell growth to synthesis and secretion of candidate local mediators of estrogen action. 2. CYTOKINES A large number of factors normally thought of as controlling the immune and hemopoietic systems may also be involved in the local control of bone remodeling (Fig. 3). Initially, interest focused on interleukin-1 (IL-1] and tumor necrosis factor (TNF), after the observation that there may be a relationship between bone turnover and IL-1 production by monocytes [25,26]. Estrogen loss was associated with increased production of both IL-1 and TNF [27] and more recently of macrophage colony-stimulating factor (M-CSF) [28]. The increment in the latter appears to temporarilly precede the increased production in IL-1 or TNF, suggesting that the prime response to estrogen loss is increased production of M-CSF. Monocytes are the major source of Il-1 and TNF in marrow, and they are clearly

anatomically relevant to the process of remodeling. The Il-1 system is complex and the precise meaning of in vitro experiments is unclear. Even the estimation of IL-1 production by monocytes incubated in vitro is fraught with interpretive difficulty. The capacity of such cells to respond to stimuli must depend on the microenvironment in which they are studied, and in vitro or ex vivo experiments do not adequately recon Fig. that environment. Consequently it is not surprising that a large and confusing literature now exists for the effects of estrogen on a variety cytokine responses of relevant cells including osteoblasts. Dependent upon the experimental design, estrogens (usually estradiol) have been shown to increase IGF-I and the binding protein IGFBP4, stimulate TGF-1, and TNF- and to suppress (or not change) IL-1 (alpha and beta), Il-6, TNF- or osteocalcin synthesis and secretion [29 – 39]. Several gene transcription factors are also switched on by estrogen including c-fos, c-jun, TIEG, NF-B, and c/EDP-. Other genes regulated in vitro in osteoblast cell lines include the receptors for progesterone, 1,25-vitamin D, and parathyroid hormone (PTH), as well as transferrin, protein kinase C, and ERE Cat [40]. The somewhat bewildering and often conflicting responses may be related to differences in species, skeletal sites from which cells were obtained, or differences in the degree of differentiation of the cells. Cells of the monocyte lineage express both ER- and- as do osteoblasts [41]. Estrogen-mediated effects vary

Potential pathways in the estrogen regulation of bone resorption. The IL-1, IL-6, and TNF- inhibit apoptosis and extend the life span of OCLs. In contract, TGF- increases osteoclast apoptosis. Reprinted with permission from Spelsberg et al.

FIGURE 3

581

CHAPTER 69 Estrogens and Osteoporosis

dependent upon the receptor type. Thus, effects mediated by activation of the traditional receptor (ER-) activate AP-1 regulatory elements, while inhibition appears to occur after activation of ER-. Using the knockout mouse, elimination of ER- produces only a modest reduction in bone mass (in contrast to the human male with homozygous null mutation of ER-) [42], while ER- null animals have little skeletal abnormality, except greater cortical bone mass [99]. Whether some of the variability reported in responses relates to species differences either in skeletal response, distribution of receptor, coactivators, or repressors is not clear.

FIGURE 4

3. CONTROL OF BONE REMODELING Estrogens, in vivo, clearly reduce bone turnover and activation of remodeling. While the specific cell-mediated effects in vitro are interesting and provide targets for further dissection of estrogen action they do little to determine the effects of estrogen in the complex mechanisms involved in skeletal remodeling. The remodeling unit (the basic multicellular unit originally described by Frost [43] consists of teams of both osteoblasts and osteoclasts that primarily function to build and destroy bone, respectively (Fig. 4). However, the recruitment, behavior, and lifespan

Bone remodeling and the basic molecular unit, moving along a cancellous surface. Each step represents  10 days, and the BMU moves at about 10 m each day. (A) Origination of BMU, lining cells contract to expose collagen and attract preosteoclasts. (B) Osteoclasts fuse into multinucleated cells which resorb cavity. (C) Monuclear cells continue resorption; preosteoblasts are stimulated to proliferate. (D) Osteoblast team forms at bottom of cavity and starts forming osteoid. (E) Osteoblasts continue forming osteoid (black) and previous osteoid starts to mineralized (horizontal lines). (F – H) Osteoblasts turn into lining cells; bone remodeling at initial surface (left of drawing) is complete, but BMU is still advancing (to the right). Reprinted with permission from S. M. Ott, Theoretic and methodological approach to bone remodeling. In “Principles of Bone Biology” (J.P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), p. 233. Academic Press, San Diego (1996).

582

LINDSAY AND COSMAN

of these cells can be significantly modified by their environment, by their exposure local stress, cytokines, and growth factors, and the circulating hormones that provide tonic control of remodeling. From our current knowledge base it appears as though the factors that control the development of these cells and their lifespan have more profound effects on remodeling than any effects that might be mediated through alterations in work rate of either teams of cells or individual cells. The effects of estrogen in vivo can be explained, at least in part, by modifications in the lifespan or numbers of the cells involved in the remodeling process. If estrogens reduce osteoclast recruitment or shorten their lifespan by stimulating early apoptosis, then bone resorption and bone mineral density BMD loss would diminish [44]. Also, estrogens might modify the birth-rate of teams of osteoclasts, although this still has to be convincingly demonstrated. Thus, in estrogen deficiency, increased production of osteoclasts, increased activation of remodeling, plus a longer osteoclast lifespan could effectively produce the in vivo results seen after ovarian failure. The changes that occur in skeletal remodeling as ovarian function declines can be monitored clinically. A variety of biochemical markers of bone remodeling, which can be measured either in serum or urine, have been shown to increase across the menopause, the consequence of increased activation of skeletal remodeling [45 – 53] (Fig. 5). The increments vary, as would be expected, since many of the markers measure different aspects of the remodeling cycle. Osteocalcin, bone-specific alkaline phosphatase, and the propeptide of type-I collagen, all measuring different aspects of osteoblast development or function, increase gradually to about 50% or more above the premenopausal mean. Markers associated more closely with resorption such as the clipped peptides from either the C or the N terminus of mature collagen with their pyridinoline links are thought to be related to rates of collagen digestion, while tartrate-resistant

acid phosphatase is perhaps an indicator of osteoclast load. All increase with estrogen deficiency. The majority of the data examining this are cross-sectional, and in those data sets many estrogen-deficient women have circulating or excretion levels of marker that are within the premenopausal range. There are two possibilities for this finding. Either not all women exhibit increased remodeling with estrogen defi ciency, or these women still within the premenopausal range did show increased remodeling but less than their peers. Prospective data will be required to evaluate this. Even here the noted marked variability in biochemical markers within individuals may confound the results. It is noteworthy in this regard that biochemical markers correlate only poorly with the rate of bone loss in individuals [54 – 57].

C. Bone Loss with Estrogen Deficiency In general, bone loss follows menopause (Figs. 6A and 6B), although there is also some premenopausal bone loss especially in the fifth decade [58 – 72]. There is considerable heterogeneity in the rates of bone loss among individuals, as well as across skeletal sites within individuals. The primary function of the skeleton is mechanical, and as such it must be responsive to the amount of strain to which it is constantly subjected. Higher strains demand stronger bones [73]. Thus, logic argues that those who are more physically active are likely to suffer less from bone loss consequent upon estrogen deprivation. Data are lacking on this issue and are difficult to obtain because they require the control of other factors that might influence the skeleton at this point in life.

D. Other Factors Contributing to Bone Loss 1. CALCIUM INTAKE The skeleton is also the major repository for calcium, an element critical for life, in an environment that is chroni-

FIGURE 5 Mean urinary concentrations of the hydroxypyridinium crosslinks pyridinoline (PYD) and deoxypyridinoline (DPD) in postmenopausal (post) healthy women and untreated (UTO) and estrogen-treated (ETO) women with postmenopausal osteoporosis compared to premenopausal (pre) healthy women. Column inserts denote percentage change of mean value compared to normal premenopausal controls (Pre). Bars represent standard error of mean. Asterisks indicate P values  0.01. UTO levels of both PYD and DPD were higher than Post (p.01). Reprinted with permission from Seibel et al. [45].

583

CHAPTER 69 Estrogens and Osteoporosis

FIGURE 6 (A) Long-term prevention of bone loss by estrogen. The hatched areas represent the placebo using (mean  SD) a prospective controlled study in oophorectomized women. The three lines show the mean values only for three estrogen-treated groups; treatment initiated at the time of oophorectomy (squares), 3 years after (circles), or 6 years after oophorectomy (triangles). Bone loss is prevented in all three situations. However, the earlier treatment is begun, the better the outcome, in terms of bone mass after 10 years of therapy. Reprinted with permission from Lindsay et al. [72] (B) Relationship of BMD in the total proximal femur versus years since menopause (YSM). The values for YSM  0 are the mean of the premenopausal values. The best exponential fit is shown. Reprinted with permission from Lindsay et al. [72]. cally deprived of calcium (see Chapter 27). It has been established in several population-based studies in the United States that the average calcium intake is around 600 mg/day [74]. Balance studies suggest that the human organism requires about 800 – 1200 mg/day for zero balance [75 – 76]. That is, for intake to balance the calcium losses from the body, at least 800 mg needs to be supplied each day. The effect of menopause is to increase calcium loss, mainly because of increased skeletal remodeling, but perhaps also because of an increase in the obligatory loss of calcium in urine, a speculated direct effect of estrogen deprivation [77]. In the estrogen-deprived individual calcium requirements increase to about 1200 – 1500 mg/day [76]. Thus, in general postmenopausal women are calcium deprived (as well as estrogen deprived) and the degree of calcium deprivation will depend not only on intake but also the efficiency with which calcium is utilized [78]. As age increases there is a decline in intestinal absorption efficiency in part related to decline in the receptor density for 1,25(OH)2D [79). The consequence is a homeostatic increase in secretion of parathyroid hormone, attempting to drive 1-hydroxylation of 25(OH)D in the kidney and to improve indirectly intestinal calcium absorption [80]. These modifications in calcium homeostasis may be at least partly estrogen dependent [81]. In addition, estrogen deprivation reduces the resistance of the skeleton to the bone resorbing effects of PTH [82,83]. Estrogen intervention restores resistance to PTH [83].

2. SMOKING Additional factors that might influence the response of the skeleton to estrogen deficiency include cigarette use and alcohol intake (see also Chapter 31). Cigarette smokers appear to transition through the menopause at an earlier age than nonsmokers [84,85]. There is also evidence that there is increased metabolism of estrogens through the 3-methoxy pathway in cigarette smokers, essentially reducing circulating levels of active estrogens such as estradiol, although this effect, which is mediated in the liver, is seen only in women using HRT/estrogen replacement therapy (ERT) [86]. Sex hormone binding globulin (SHBG) has been reported to be increased in smokers [87]. However, the effects of cigarette consumption are probably minimal unless consumption is quite high (1 pack per day]. Nicotine has also been attributed to have direct toxic effects on osteoblasts [88]. Finally, increased morbidity and reduced physical activity among users of cigarettes as they age undoubtedly contributes to bone loss in this population. Thus, there are several potential mechanisms by which interaction with cigarette use and bone loss as a consequence of estrogen reduction could occur. 3. ALCOHOL Alcohol intake is commonly cited as a risk factor for osteoporosis [89]. However, it is not entirely clear whether this is due to the effects of alcohol on the skeleton or an

584 increased risk of falls among those who drink excessively. In most situations it appears as though excessively high intakes account for the proposed effects, likely related to the malnutrition and changes in sex hormone metabolism associated with such levels of alcohol consumption. Acute alcohol ingestion appears to modify the pharmacokinetics of exogenous estrogen delivery, with effects dependent on the route of estrogen administration. Acute alcohol intake increases estradiol concentrations 20% following transdermal delivery and triples circulating eestradiol following oral administration [90 – 92]. Estrone concentrations do not appear to be affected. Chronic alcohol intake at moderate levels has been reported to increase serum estradiol concentrations, perhaps by promoting aromatization of androgens [93]. Since osteoblasts have aromatase, there is a potential for increased estrogen supply to bone. Intake levels providing such beneficial effects are likely little more than one standard unit of alcohol per day. Alcohol is also directly toxic to osteoblasts, and the bone disease associated with chronic high intakes is one of impaired bone formation [94,95]. The effect may be mediated by increased apoptosis or reduction in DNA activity associated with decreased polyamine levels [95]. Alcohol could also interfere with phosphorylation of the IGF-I receptor [96], thereby contributing to the reduced osteoblast cell number seen in patients with alcohol-inducedosteoporosis [94]. 4. PHYSICAL ACTIVITY The effects of estrogen deficiency are also likely compounded by the changes in physical activity that occur with increasing age. Whether these are sufficient in the general population to jeopardize skeletal status is not at all clear. However, the risk of falling increases with age, as do estimates of frailty, and these will have an impact upon the risk of fracture independently of bone mass [97]. Intervention studies of activity are notoriously difficult to perform and impossible to conduct in a double-blind fashion, as is required in the rigorous studies of pharmacological therapy for current registration in osteoporosis. Some evidence does suggest that, with HRT, increasing physical activity will produce an added benefit in terms of bone density [98], but no such data exist for fractures as the primary outcome. 5. CHRONIC DISEASES AND MEDICATIONS In addition to the effects of these lifestyle features, the effects of estrogen deprivation are likely to be exacerbated by any intercurrent disease process known itself to negatively affect the skeleton. The reader is referred to Chapters 44, 51, and 54 for detailed descriptions of these secondary causes of osteoporosis. In addition the use of medications may negatively impact the skeleton, especially those that act by mechanisms that differ from estrogen’s effects for example glucocorticoids. Evidence gleaned from recent controlled studies examining the ef-

LINDSAY AND COSMAN

fects of bisphosphonates on bone loss and fracture among patients being treated with glucocorticoids for a variety of diseases, suggests strongly that fracture risk is low among premenopausal women, but increases substantially among postmenopausal women [99,100].

E. The Duration of Estrogen Deficiency Induced Bone Loss It is commonly stated that the effects of estrogen defi ciency are mostly seen during the early years (1 – 3 years) after the onset of menopause. Indeed, since the technology has evolved, allowing investigators to examine vertebral bone mass, there is clearly seen a period of rapid bone loss of approximately 1 – 3 years in duration [100 – 105]. However, since bone remodeling is a surface phenomenon, it would be expected that increased remodeling would affect cancellous bone disproportionately, as cancellous bone has a high surface to volume ratio. With loss of bone tissue, surface area must decline, producing a predictable decline in the rate of loss. In the older, but relatively few, studies that have examined cortical bone, bone loss continues because there is a much lesser change in the surface area available for remodeling. Recent data in which the role of estrogen deficiency in bone loss among elderly women has been studied have allowed the conclusion that even in the very old the rate of loss somewhat depends on estrogen supply [106,107].

F. Estrogens and Males Estrogens may also play important and perhaps key roles in modulating skeletal metabolism in males as well as in females. Smith et al. described a young man, age 28 years, with homozygous null mutations in the estrogen receptor gene (presumed ER-) [108]. This eunuchoid individual had normal serum testosterone and increased estradiol concentrations, unfused epiphyses, markedly reduced bone mass, and elevated bone remodeling. A somewhat similar phenotype was described in a man with homozygous null mutations in the aromatase gene [109]. In one individual, testosterone had little effect, while estrogen therapy lowered bone turnover and increased bone mass. In older individuals Slemenda et al. demonstrated that the rate of bone loss correlated with estradiol circulatry, and estradiol (but not testosterone) values have been reported to be low in some men with idiopathic osteoporosis [110].

G. Effects on Bone Structure and Fracture Risk Loss of bone mass increases the likelihood that fracture will occur. However, relatively modest changes in mass pro-

585

CHAPTER 69 Estrogens and Osteoporosis

duce marked increases in the risk of fracture. A reduction in mass of about 10 – 12% produces about a 50 – 100% increase in fracture risk [111 – 117]. It has been argued therefore that in addition to the changes in mass that occur the alterations in structure that result from the progressive elimination of trabeculae in particular must contribute to the change in fracture risk [118]. Several years ago we showed scanning electron micrographs that pictorially confirmed these structural modifications [119]. Similar descriptions have been presented by other investigators including confirmation that trabecular knockout was evident in vertebral bodies, which would then be expected to be particularly sensitive to compressive forces [120]. The agerelated increase in vertebral fractures occurs earlier in the postmenopausal period and has been more closely associated with the duration of estrogen deficiency than other fractures. The later increase in fractures of the hip and the greater difficulty in demonstrating a relation to the duration of the postmenopausal period are perhaps related in part to the slower structural changes and the different type of forces to which the neck of femur is subjected. It must also be remembered that other influences will have a longer period of time over which they can influence the bone loss process. 1. ESTROGEN FROM ENDOGENOUS SOURCES The decline in circulating estradiol across menopause is dramatic, and it has been assumed until recently that it was sufficient to induce bone loss in all individuals. However, bone density studies indicate that not all postmenopausal women lose bone over 2 – 3 years of observation [121,122]. Data from the placebo arms of recently competed controlled clinical trial suggest that the average change in bone density in the spine in women in their 60s or older is zero [123,124]. While some of the lack of spinal bone loss in this age group undoubtedly represents artifact related to increased extraosseous calcification around facet joints and in osteophytes, it may, at least for some individuals, be true. It is routine in such studies to supplement with calcium and vitamin D, and other data confirm that calcium supplementation can slow bone loss and reduce fracture risk among older individuals, although approximately one-half of calcium supplemented older postmenopausal women continue to lose bone mass [124,125]. This variability in age-related bone loss may be related to endogenous estrogen supply. Recent data from the Study of Osteoporotic Fractures (SOF) have suggested that even very low circulating estradiol concentrations can significantly influence bone remodeling rate [106]. This is an attractive hypothesis. However, the difference in circulating estradiol among individuals with and without fractures was 1 – 2 pg/ml, which is extremely small, probably within the interassay error, and certainly within the day-to-day variability within individual postmenopausal women. Among obese women, there is increased aromatization of androgens to estrogens and obese

individuals are likely to have a significant biological effect of estrogen on bone [121] with increased bone mass and reduced risk of fracture. 2. ESTROGEN-DEFICIENT STATES OTHER MENOPAUSE

THAN

While menopause remains the most important time when estrogen levels decline, mainly because it occurs in every woman around the same age, it is by no means the only cause of estrogen deficiency. It is important for the clinician to understand that estrogen deficiency at any age increases bone remodeling and may cause bone loss. Perhaps the best described of these is exercise-induced amenorrhea. It is now well established that female athletes, especially runners, who train excessively will induce hypogonadism probably of hypothalamic origin [122 – 128]. Such individuals have lower bone mass than athletes of similar age whose gonadal function is normal. However, both groups are usually found to have higher bone mass than individuals who do not exercise. The clinical significance of hypogonadism is related to the higher incidence of stress fracture in that group [129]. In hyperprolactinemia, low bone mass does not occur until prolactin levels are sufficient to induce amenorrhea [130]. In anorexic individuals the effects of estrogen deficiency are compounded by the nutritional deprivation of the primary eating disorder and such patients often have osteoporotic fractures at a young age [131 – 133]. Similarly in Turner’s Syndrome, primary amenorrhea results in failure to achieve peak skeletal mass and if estrogen intervention is not initiated osteoporotic fractures may occur at a young age. The problem in Turner’s Syndrome may be compounded by the phenotypic expression of the chromosomal deletion, which may directly affect skeletal development [134 – 139].

IV. THE EFFECTS OF ESTROGEN INTERVENTION A. Bone Remodeling and Turnover In his original studies of individual patients Albright demonstrated that estrogen intervention reversed the negative calcium balance in postmenopausal or ovariectomized women with osteoporosis [2]. Such studies are extremely difficult and tedious to perform, but the results were quite clear, requiring no statistical tests for confirmation of their biological importance. In early studies in which markers of bone remodeling were measured, estrogen was shown to reduce turnover as estimated by alkaline phosphatase and the urinary excretion of hydroxyproline [140,141]. In those studies there was also a marked reduction in the excretion of calcium in urine, which as noted above may be related to both reduction in bone remodeling and also a direct

586

LINDSAY AND COSMAN

renal effect of estrogen inhibiting urinary calcium loss. These data have been confirmed in several studies in which the effects of estrogen on the more recently developed markers of bone remodeling have been measured [45 – 49,56]. Thus, estrogen intervention reduces the excretion of pyridinoline, deoxypyridinoline, and the pyridinoline-linked peptides n-telopeptide and c-telopeptide and reduces serum levels of osteocalcin, bone-specific alkaline phosphatase, and tartrate-resistant acid phosphatase. In general, the belief is that adequate intervention with estrogen will return biochemical markers to the expected means for premenopausal women, although in one recent study that recruited women on long-standing HRT into a clinical trial, mean values of biochemical markers at the time of recruitment were at the upper limit of normal for premenopausal values [142]. Whether this is related to inadequate dosing, poor compliance, or represents the true state of women on long-term HRT is not clear.

B. Bone Mass As noninvasive technology was developed for evaluation of bone mineral density, controlled clinical trials began to appear, demonstrating conservation of bone, at all skeletal sites measured with the use of estrogen intervention [97,99 – 130,141,143 – 144]. The earliest reported studies often examined the effects of estrogen in women following ovariectomy, and measured peripheral cortical bone [141 – 143]. More recently, many studies have looked at combination therapy in which estrogen is given in combination with a progestin, and have been able to examine the effects of sex steroids on vertebral and hip bone mass, the two sites of fracture importance [56,170]. The largest of the

FIGURE 7

studies among healthy, early postmenopausal women [56, 170] the Postmenopausal Estrogen and Progestin Interventions (PEPI) trial confirmed the estrogen effects previously demonstrated in a number of studies of smaller size. In addition, PEPI demonstrated that the addition of a C-21 progestin did not affect the influence of estrogen on bone mass (Fig. 7). In PEPI two progestins were used, medroxyprogesterone acetate (either 2.5 mg every day or 5 mg for 12 days per month) and micronized progesterone (200 mg/day for 12 days/month). Data from other studies, in which the effects of medroxyprogesterone (MPA) alone on bone mass were assessed, suggest that at least 20 mg per day are required before MPA exerts an inhibitory effect on bone remodeling with resultant prevention of bone loss [151,175]. Smaller studies have looked at the effects of 19-nortestosterone-derived progestins [153,171]. Here there is the suggestion that norethindrone, and perhaps norgestrel both add to the estrogen effect on the skeleton, particularly when estrogen is provided in a less than effective dose [176]. While suggestive, the data are far from convincing, and certainly do not address the question of whether such combinations will have a measurably greater effect on fracture reduction than might be expected from an estrogen by itself. Norethindrone may be aromatized to ethinyl estradiol, which is highly active in inhibiting bone remodeling, and may be the source of the apparent additive effect of norethindrone [177]. The majority of studies examining the effects of estrogen intervention on bone mass are of 1 – 3 years duration. Except when estrogens are initiated very shortly after menopause, there is usually a transient increase in apparent density that occurs during the first 1 to 2 years. This increment is most obvious in measurements of the spine, somewhat less so at the hip, and often difficult to see in

Results of various HRT regimens on BMD of the spine (left) and hip (right). Unadjusted mean percentage change in bone mineral density in the hip by treatment assignment and visit: adherent PEPI participants only. Results from the Postmenopausal Estrogen/progestin/Interventions (PEPI) Trial. CEE, conjugated equine estrogen 0.625 mg/day; CEE-MPA (cyc), CEE medroxy progesterone acetate 10 mg/day for 12 day/month; CEE-MPA (con) CEE medroxy progesterone acetate 25 mg daily; CEE-MP (cyc), CEE micronized progesterone 200 mg/day for 12 days/month. Reprinted with permission from Bush et al. [170].

587

CHAPTER 69 Estrogens and Osteoporosis

measurements of peripheral bone. Since the magnitude of the effect mirrors the proportion of cancellous bone present, it has been argued that this increment is related to the decline in the activation of new remodeling sites [174,178]. Since remodeling activity is greatest in cancellous bone, any reduction would be expected to produce a greater effect on measured bone mass at that site. This transient is referred to as refilling the remodeling space [179] (see also Chapter 27). When remodeling has reached a new steady state the transient increase would be expected to plateau. If, in subsequent remodeling activity, bone resorption and bone formation are in balance within each remodeling cycle, then measured bone density should remain constant, provided that there is adequate estrogen supply. In the few studies of longer duration there is in fact a more heterogeneous response. In PEPI, for example, during the third year of the study bone density continued to drift upward [170]. In earlier long-term studies in which peripheral bone was measured, bone density was stable out to 10 years [141]. In other studies, primarily, but not only involving the use of estrogen implants there also appears to be a continued increase in bone density over several years [150,171]. The explanation for the behavior of bone mass after prolonged treatment is far from established. The continued increase has been taken as evidence that estrogens are relatively anabolic (a remodeling imblance within each remoding unit in favor of formation). An alternate explanation may be prolongation of the phase of secondary mineralization resulting in a higher mineral content within already synthesized bone. Stable bone implies reinstitution of the premenopausal state of (almost) perfect skeletal balance. Bone loss at this phase of treatment might imply inadequate estrogen dosing. However, it must also be remembered that other factors influence skeletal homeostasis (as noted above) and that these may well influence the outcome of an estrogen trial. Finally, there is the inherent problem of long-term data accumulation. In studies of relatively long duration, e.g., several years, the study population changes as individuals drop from the study. If, for example, there was an excessive loss of individuals who did not appear to be responding, the apparent result would be a more continuous increase in density. This type of problem besets all large long-term clinical studies, not just those of HRT, or in the bone field, and resultant biases may vary depending on the size of the problem and the reasons for discontinuation. 1. SPECIFIC ESTROGEN PREPARATIONS Until fairly recently there were somewhat limited data that addressed the issue of which estrogens were able to retard bone loss in estrogen deficient women. Studies have now been conducted to evaluate conjugated equine estrogens (CEE), estradiol (either as valerate or micronized),

estrone (as the piperazine sulfate), esterified estrogens, ethinylestradiol, and its 3-methyl ether mestranol, and estriol [145 – 174]. It appears that all estrogens can prevent bone loss. Indeed the critical factor is the dose that is given. Provided that sufficient estrogen can be delivered, then a skeletal benefit can be observed. Only for estriol are the data in doubt, likely related to the poor pharmacodynamics and pharmacokinetics of this agent [179,180]. Estriol binds only weakly to the estrogen receptor and is rapidly cleared from circulation. 2. ROUTE OF ADMINISTRATION There is now ample evidence for the conclusion that estrogens are effective in reducing bone turnover and preserving bone mass by whichever route they are administered. Clinical trial data in which estrogens have been administered orally [153 – 166], intransally [181], vaginally [182], percutaneously [183], subcutaneously [171], and transdermally [172,173], have all demonstrated efficacy. Thus for preservation of bone mass in postmenopausal women, there are many choices, and the choice is made for other reasons, including patient preference. For both oral and transdermal routes of administration combined preparations are now available in many countries, obviating the problem of taking two tablets or using a patch and an oral progestin. 3. DOSE OF ESTROGEN Until fairly recently it was thought that the minimum effective dose of estrogen for skeletal protection was 0.625 mg of conjugated equine estrogens or its equivalent. This conclusion was based upon two studies published in the early 1980s [156,185]. Both were relatively small by current standards, and have been interpreted as demonstrating that doses of CEE of less than 0.625 mg/day were totally lacking in skeletal effects, and that all individuals required this dose. Finally, they were construed as saying that increasing the dose would add little (and consequently might negatively affect the safety effectiveness relationship). However, these conclusions seem unlikely to be correct. First, there are very few situations in medicine where “one size fits all.” Second, for other target tissues, for example menopausal symptoms, the doses required for benefit certainly vary. Third, we know that estrogen kinetics and dynamics vary considerably among individuals. Thus, interpreting what might be necessary for individual patients based upon mean changes in trials is difficult. The doses of estrogens that are approved for osteoporosis prevention in the United States are shown in Table 1. For oral estrogens 0.3 mg/day is the recommended dose for esterified estrogens, while 0.625 mg/day is recommended for conjugated equine estrogens. This is probably only because recent data from a controlled

588

LINDSAY AND COSMAN

TABLE 1

Brand name

Generic name

Preparations of Estrogen and Progestin for HRT Min. dose for preventive therapy

Upper-end dose (mg)

FDA approval for prevention/treatment of osteoporosis

Comments

Premarin

Conjugated equine estrogens

0.3 mg 0.625 mg

1.25

Prevention (and management)

Usual dose  0.625 mg, but 2.5 mg sometimes necessary to control hot flashes in young women

Ogen Ortho-Est

Estropipate

0.625 mg

1.25

Prevention

2.5 mg sometimes necessary to control hot flashes in young women

Estratab Menest Estratest HS Estratest

Esterified estrogen (Estrone, equilin) Esterified estrogens Methytestosterone

0.3 mg

2.5

Prevention

Derived from plant sterol precursers Contains androgens

Estrace

Micronized estradiol

0.5 mg

Alora Climara Estraderm Vivelle

Transdermal estradiol

Combipatch

0.625 mg 1.25 mg

1.25/2.5

Prevention

0.5 mg effective for bone preservation

0.05 – 0.1 mg 0.025 mg (Climara)

No indication No indication Prevention Prevention

Patches applied once or twice a week depending on manufacturer

Estradiol Norethindrate

0.62 – 0.81 mg 2.7 – 4.8 mg

No indication

Patch applied twice a week

Estrace

Micronized estradiol

0.5 or 1.0 mg

Prevention

0.5 mg effective for bone preservation

Prempro

Conjugated equine estrogens/ medroxyprogesterone acetate (MPR)

0.625 mg/2.5 mg or 5 mg 0.625 mg/5 mg

Prevention (and management) Prevention (and management)

If excessive bleeding, may consider increasing MPR dose to 5 mg

Prometrium

Micronized progestin

100 mg (daily dose) 200 mg (cyclic dose)

Uterine protection when used with estrogen

Does not attenuate lipid effects on estrogen

Provera Cycrin Amen

Medroxyprogesterone acetate

5 or 10 mg (cyclic dose) 2.5 mg (daily dose)

Uterine protection when used with estrogen

Aygestin

Norethindrone

2.5 – 10 mg

Uterine protection when used with estrogen

Fem HRT

Ethinyl estradiol Norethindrone

5 g 1 mg

Prevention

Premphase

study of esterified estrogens were used to obtain approval [186]. At the lower dose esterified estrogens prevented bone loss (Fig. 8) although there was greater mean change at double the dose. Based upon a comparison with the PEPI data, it appears as though the two preparations are similar in terms of the bone density response, although no direct head-to-head comparisons have been performed. For transdermal estrogen the commonly used dose supplies 50 g estradiol per day, but now some data indicate that a lower dose may be appropriate for some

2.0

No indication for osteoporosis

2.0

individuals [187]. No good dose response data are available for other routes of administration. Low dose estrogen intervention may be important in the treatment of the older women [187a]. It is commonly recognized that the introduction of estrogen to women over the age of 60 years with a long history of estrogen deprivation is frought with problems especially if estrogens are begun at the generally stated full anti-osteoporosis doses [188 – 190]. Here, symptoms such as breast pain commonly result in early discontinuation. Recently it has been suggested that

589

CHAPTER 69 Estrogens and Osteoporosis

C. Effects on Fractures

FIGURE 8

Percentage change in spine bone mineral density (BMD) (bottom), femoral neck bone mineral density (middle), and total-body bone mineral content (BMC) (top) in the hormone replacement therapy group (solid circles) and placebo group (open squares). The effect of hormone replacement therapy was significant at all sites (P 0.01) except the femoral neck (P  0.19). Bone mass remained the same or increased in the placebo group at the spine, total body, and femoral neck and decreased at the forearm. Error bars represent SEs. Reprinted with permission from Recker et al. [187a].

even low levels of endogenous estrogen result in some skeletal protection among older individuals [191,192]. Thus, smaller doses of estrogen may be required to protect the skeleton of older individuals, which in part may explain the differences in the clinical trial results, since the early dose – response studies were generally conducted in younger women. This cannot explain the phenomenon entirely, however, and it is indeed possible that the elderly skeleton is more, rather than less, sensitive to hormonal manipulation. An additional complication associated with the introduction of estrogens to an older age population is the possible inccreased risk of cardiovascular events in the first year of use, as seen in the HERS trial (see below).

There are few clinical trials that have examined the effects of estrogen using fracture as the primary outcome. Vertebral fracture outcomes have been evaluated in only a small number of studies and have usually been secondary outcomes (with bone density being the primary outcome) [143,193]. Several of these studies were initiated a number of years ago and were not for regulatory purposes. For estrogens to be approved in the United States for osteoporosis only bone density data were required, unlike the situation for novel chemical entities approved for osteoporosis treatment in the past decade such as bisphosphonates, for which fracture data were and are currently required [194, 195]. Thus, there was no impetus for fracture studies involving estrogens. Only one clinical trial comparable to the pivotal studies required prior to marketing for bisphosphonates is currently underway for estrogen, namely the Women’s Health Initiative (WHI) supported by the National Institutes of Health [196]. The initial study that suggested a fracture benefit (at least in terms of vertebral fractures) is an observational follow-up of many of the patients initially treated by Albright [197]. Height loss was used as the surrogate, and estrogen appeared to stop progressive height loss while untreated patients continued to lose height. Interestingly, similar data were produced from the phase 3 studies of the bisphosphonate alendronate, confirming that height loss is an external sign of vertebral fracture and that prevention of fractures equates to preservation of height. The first controlled clinical trial that formally evaluated vertebral deformation was published in 1980 [143]. This was a relatively small study (by the standards of today) but of long duration. The trial was a placebo-controlled double-blind study of ovariectomized women and was conducted over a 10year period. The primary outcome was bone density and vertebral deformation assessed after 9 – 10 years in crosssectional fashion. Since this was a prevention study, there were not many fractures, but sufficient to determine that, while vertebral deformity was evident in placebo-treated subjects, it was virtually absent in the estrogen-treated arm (Table 2). In a somewhat similar small study of equal duration, Nachtigall found similar results in nursing home patients, not selected on the basis of osteoporosis [149]. The first study demonstrating that estrogen might reduce the risk of vertebral fracture in patients being treated for osteoporosis was again a relatively small study. Lufkin et al. demonstrated after only 1 year that transdermal estrogen reduced the risk of vertebral fracture by 70%, at a relatively high dose of 100 g/day [198] (Table 3). For prevention of nonvertebral fractures in general and hip fractures in particular, the data that are available are all observational, with one exception. Several large epidemiological studies have shown reduction in the risk of fractures

590

LINDSAY AND COSMAN

TABLE 2

Effect of Estrogen on Height Loss and Spinal Deformity Prevalence of height loss

Degree of height loss (cm)

Mean spine score a

Ratio of central to posterior height

Estrogen group

4% (n  2)

0.0 cm

0.35

0.89

Placebo group

38% (n  16)

0.9 cm

1.62

0.79

P  0.01

P  0.06

Significance

—

—

Note. Adapted with permission from Lindsay et al. [143]. a Index of wedged and crushed vertebrea using standard definitions (T9 and L2).

of the hip and forearm associated with the use of estrogens among postmenopausal women [199 – 211a] (Table 4 and Fig. 9). In general the reduction in risk is about 50% even after correction for other risk factors (corrections vary among the published data). These studies vary in design and size, with some being retrospective, others prospective, some with population controls, and others using hospital controls. Many evaluated only fractures of the hip, while others also looked at wrist fractures, with a few evaluating the global effects of estrogen on osteoporotic fractures. Many were able to evaluate only estrogen intervention, especially those in which the observation period was in the 1970s and in the United States, when use of ERT was much more prevalent than HRT. The largest of the available studies is from Uppsala, Sweden, which evaluated the effects of HRT on fractures in the Uppsala Health Care Region [209]. The cohort of some 23,000 individuals included all women over 35 years in the region. The authors compared the rates of hip fracture in women who had filled at least one prescription for HRT between 1977 and 1980 and followed the cohort until 1983 (mean duration of observation almost 6 years). HRT use was ascertained from a random sample of 735 women, of whom 9% never took the prescribed agent, and one half had begun before the study period. The incidence of hip fracture was compared to that of the background population of women of the same age range in the region. There was a statistically significant reduction in the risk of hip fracture [RR  0.79] in those who took estradiol or conjugated estrogens, with no

TABLE 3

effect in those taking estriol. The effect was strongest in those who began therapy within 10 years of menopause [RR  0.55]. Women who began therapy after age 60 years were not protected. In the Study of Osteoporotic Fractures (SOF) a cohort of 9704 women over age 65 years are being followed prospectively in four centers in the United States [200]. Women in the cohort who were currently using HRT had a lower risk of wrist fractures [RR  0.46 and all nonspine fractures [RR  0.69]. Previous use of HRT was not protective. The effect of HRT was greater in those who began within 5 years of menopause and continued for more than 10 years. Adjustment for risk factors did not alter the conclusions. The conclusions from these studies are similar to the findings in the many observational studies now published. The effects of estrogen on hip fracture risk may be most evident in smokers, thin women, and alcohol users and less obvious in physically active older women. Only one clinical trial has evaluated the effects of HRT on nonvertebral fractures [212]. In this study 464 early postmenopausal women were randomized to HRT, vitamin D, a combination of both, or placebo. The study was 5 years in duration and 368 women completed the observation period. The estimated risk of new nonvertebral fractures in the HRT group was 0.29 [CI 0.1 – 0.9], the only group significantly reduced from placebo (Fig.10). However, there were only 39 fractures in the entire study, resulting from the small sample size and the relatively young age of the participants.

Effect of Transdermal Estrogen on BMC and Vertebral Fracture Change in BMC after 1 year (%)

Transdermal estrogen  medroxyprogesterone acetate(n  36)

New vertebral

Lumbar

Midradius

Trochanter

fractures

 5.3

 1.0

 7.6

8

Placebo (n  39)

 0.2

 2.6

 2.1

20

P value

0.007

 0.001

0.03

0.04

Note. Adapted with permission from Lufkin et al. [201].

591

CHAPTER 69 Estrogens and Osteoporosis

TABLE 4

Postmenopausal Hormone Use and Hip Fracture Ever use relative risk

Case – control (Year) Hutchinson (1979)

0.2*†

Weiss (1980)

0.4*‡§

Johnson (1981)

0.7

Paganini-Hill (1981)

0.7†

Kreiger (1982)

0.4*†

Williams (1982)

0.4*†

Cohort (Year) Hammond (1979)

0.5*†

Ettinger (1985)

0.4*†

Kiel (1978)

0.6*

Naessén (1990)

0.8*

Paganini-Hill (1991)

1.0

Note. Adapted with permission from Grady et al. * P  0.05. ‡ The relative risk of P value or both are estimated from data provided in the published study. † Risk estimate for hip and distal radius fractures combined. § Current estrogen use.  41% of women 60 years and 20% of women  60 years used combined estrogen plus progestin regimen. Cases are compared with population-based fracture rates in Uppsala, Sweden.

Questions have been raised about the validity of the HRT effect on nonvertebral fractures since the publication of the HERS study (Heart and Estrogen/progestin Replacement Study) [213] (Fig. 11). That study was a secondary prevention trial designed to examine the effects of CEE plus medroyprogesterone acetate (MPA) on nonfatal myocardial infarction (MI). Patients (n  2763) were recruited with established coronary artery disease (evidence of one or more of the following: MI, CABG, percutaneous revascularization of coronaries, or at least 50% block in one or more coronary arteries) [213]. While there was no overall benefit in terms of reduction in MI, there was also no clinical fracture benefit (either hip or other clinical fractures). There are several possible reasons why no fracture benefit was seen. First the patients were recruited on the basis of coronary artery disease, and not osteoporosis (bone density was not universally performed). Second, the patients were older (mean age 67 years) and not within the group who appear to obtain benefit in the observational data. Third, falls were not recorded, and we do not know the fall status of this highly medicated population. Finally, the wide confidence intervals suggest insufficient power to obtain significance down to a 50% reduction in hip fracture or 27% reduction in other fractures. It is worth pointing out in this context that in a population of patients recruited with low bone density above the osteoporosis range, the reduction in

nonvertebral clinical fractures with a potent bisphosphonate was only 16% and not statistically significant. Nonetheless it is clear that a clinical trial should be performed to confirm the fracture effects seen in observational studies. The currently ongoing Women’s health Initiative may provide at least some of the answers for CEE and MPA.

D. The Effects of Other Steroids As noted above, all estrogens (with the exception of estriol) appear to be able to exert a restraining effect on the skeleton. The majority of fracture data come primarily form the evaluation of conjugated equine estrogens (in the United States) and oral estradiol (in Europe). The one prospective prevention study of vertebral fractures used ethinylestradiol, and the one treatment study of ERT and vertebral fractures used transdermal estradiol. There are almost no data on other estrogens or combination treatment (with the exception of HERS) in which fractures have been evaluated. In general it has been presumed that if an estrogen, or estrogen-like substance can show similar effects on bone mass and turnover to conjugated estrogens or estradiol then fracture efficacy can be assumed. Recently, a related steroid tibolone has also been shown to reduce bone remodeling and prevent bone loss [214]. Tibolone is steroid derived from 19-nortestosterone that has androgenic, progestogenic, and estrogenic properties. Its effects on the skeleton appear to be very similar to those of estrogens, although whether this is mediated through the estrogen receptor is not clear. Although several studies have

FIGURE 9

Relative risk of hip fracture in estrogen users from study of osteoporotic fractures. Adapted with permission from Cauley et al.

592

LINDSAY AND COSMAN

diol alone strongly suggests the latter mechanism. As noted above, norethindrone may be aromatized to ethinylestradiol in vivo and even at low conversion rates might be able to supply the equivalent of microgram quantities of ethinylestradiol, enough to add to the somewhat less than adequate doses in the arms of the study in which there appeared to be an additive effect.

E. Combination Therapy

FIGURE 10

Cumulative fracture-free survival as a function of time to the first nonvertebral fracture. The reduction in the risk of nonvertebral fractures was significant in the HRT group (pooled) (P  0.042) and nonsignificant in the vitamin D group (P  0.229, Cox proportional hazards model). HRT is shown as a bold line. Reprinted with permission from Komalainen et al. [212].

now been published evaluating the effects of tibolone on bone density [214 – 221], there are no fracture data with this compound. The advantage with this agent is that it may prevent bone loss at doses that do not stimulate the endometrium [222 – 225]. If this proves to be so in large scale trials, and the requirement for a progestin can be eliminated, this provides a significant advantage for this agent. The long-term benefits and risks of this compound on the breast and heart are currently being evaluated. PEPI suggested that C-21 progestins (specifically progesterone and MPA) do not interact with the dominant estrogen effect on bone mass [56, 170]. Other data also suggest that MPA has very little effect on the skeleton at least up to 20 mg/day [226]. Norethindrone, a 19-nortestosterone derivative, reduces bone remodeling and turnover at doses of 2.5 and 5 mg/day [227,228]. Some data suggest that there may be some interaction between estrogen’s effects on remodeling and norethindrone. A dose – response study examining the effects of ethinylestradiol on bone mass also looked at several combinations of ethinylestradiol and norethindrone [176]. The data suggest that at doses of ethinylestradiol that were suboptimal for skeletal effects, norethindrone had an additive effect. It is not totally clear whether this is likely to be an androgenic (and hence presumably anabolic) effect of this progestin or whether it is an additional anti-resorptive property of norethindrone. A further suppression of bone turnover of associated with the estradiol/norethindrone combination compared with estra-

Combining estrogen with progestins to protect the endometrium from chronic stimulation by estrogen has become standard. As noted above, apart for the possible interaction with norethindrone, there appears to be little further beneficial effect for the skeleton. However, it is also worth noting that neither C-21 steroids nor 19-nortestosterone derivatives blunt the effects of estrogen on bone. Estrogens have also been given in combination with androgens. In open studies using implants there appears to be an additive effect on the skeleton [171]. In one controlled study in which oral estrogens were compared with the same dose of estrogen with added oral methyltestosterone, there again appeared to be an additive effect on bone mass associated with a relative stimulation of bone formation. Whether this results in further fracture benefit is not known [229]. Estrogens or HRT have also been studied in patients with osteoporosis in combination with other skeletally active drugs. Several studies which have evaluated the effects of HRT and bisphosphonates, as well as raloxifene with alendronate, have all demonstrated a further increase in bone mass with etidronate, alendronate, and risedronate [142,230,231]. The outcome of all studies was bone density, and again no fracture data are available. In one study, there appeared to be marked reductions in bone remodeling with the combination of HRT and alendronate based upon

FIGURE 11 Fracture occurrence in the HERS Trial. No significant difference in the incidence of clinical fractures or specifically hip fractures in estrogen-treated women from the HERS Trials. Adapted with permission from Hulley et al. [213].

CHAPTER 69 Estrogens and Osteoporosis

evaluation of bone biopsies [231]. The long-term consequences of this are not known, but it would appear to be prudent to avoid this combination in clinical practice. Estrogens have also been used in combination with parathyroid hormone [232]. PTH stimulates bone formation and bone remodeling and thus has the opposite effect to anti-resorptive agents such as estrogens and bisphosphonates (see also chapter 77). In the original study evaluating the combination, the hypothesis was that estrogen would protect cortical bone from the potentially deleterious effects of PTH, seen in primary hyperparathyroidism. Indeed marked increases in vertebral bone mass were evident through a 3-year treatment program with subcutaneous PTH. The vertebral increases significantly exceeded (approximately double) those seen with even the most potent of the antiresorptive agents. Increments in total body bone mass and at the hip and forearm were also seen although the magnitude of increase was somewhat less than that seen at the spine. Although this study was of modest size, significant reduction in the risk of vertebral fractures was evident in the PTH group in comparison to the HRT alone group. Following discontinuation of PTH, HRT was able to maintain bone mass for at least 1 year [233]. Other studies have now confirmed these effects [234,235].

F. Other Effects of Estrogens Estrogens are capable of multiple effects in the body involving many target tissues. Consequently the prescription of estrogen to postmenopausal women is complex and often involves a difficult decision making process. Estrogens are still most commonly prescribed for menopausal symptoms [236]. For the most part, therefore, the duration of treatment is relatively short, on the order of a few months. When the menopausal symptoms have abated, treatment is often gradually discontinued. No long-term benefits can be expected from such brief treatment periods, and it is reckoned that 50% of the women who initiate HRT continue for less than 1 year [236]. The proposed beneficial effects of long-term use of estrogens on disease processes among the postmenopausal population include effects on heart, brain, and the urogenital system. Potential detrimental effects include increasing risk of breast cancer, endometrial cancer, venous thrombo-embolic events, gallbaldder disease, and possibly a short-term early rise in cardiovascular events in some patients. 1. ESTROGEN AND CARDIOVASCULAR DISEASE Cardiovascular disease is the most common cause of death for women as it is for men [237]. However, there is a lag of about 10 years in the age-related rise in cardiac mortality in women. From a variety of data it has been suggested that this lag is due to the protective effect of es-

593 trogen during the premenopausal years and that it is the loss of estrogen at menopause that results in the increased risk of cardiovascular disease that begins in women in the late 50s. Numerous prospective epidemiological studies support the concept that estrogen replacement therapy reduces the risk of cardiovascular disease among postmenopausal women. The Nurses Health Study has consistently reported a protective effect of estrogen of the order of 40 – 50%, confirming data from the earlier Framingham Study [238]. Clinical trial data that have evaluated surrogate end points support this concept, with estrogen producing favorable effects on serum lipids (reduction in low-density lipoprotein (LDL) and increased high-denisty lipoprotein (HDL)), cholesterol uptake into arterial walls, cholesterol oxidation, improved arterial wall compliance, and in primate models reversal of the constriction mediated by acetyl choline in estrogen deprived coronary arteries [23 9 – 241]. Only one controlled clinical trial of the effects of HRT (conjugated estrogens plus medroxyprogesterone acetate) has been published [213]. This study (HERS) randomized 2762 women with preexisting coronary artery disease to HRT or placebo and followed them for about 4 years. There was a higher incidence of coronary events in the HRT arm during the first year of the study, which gradually reversed toward the end with a lower incidence in the 4th and 5th years although there was no overall beneficial effect. The detrimental effect appeared greatest in those with a recent history of myocardial infarction. This has been confirmed by data from the nurses health study, suggesting that estrogen treatment is associated with an increased relative risk of MI in the first year following a myocardial infarction. These data indicate that there may be at least two mechanisms involved in HRT effects on cardiovascular disease. The trend toward a long-term beneficial effect in HERS is encouraging and supportive of the epidemiological data, although longer studies may be needed to prove that this latter apparent benefit is not related simply to an “attrition of susceptibles’” phenomenon. The early increased risk may be related to the effects of HRT on clotting or on inflammatory markers [242]. Where markers of the clotting cascade have been examined, estrogens, especially when given by the oral route, increase hepatic synthesis of clotting factors, although fibrinogen levels decline. In addition, medroxyprogesterone acetate might blunt some of the beneficial effects of estrogen, for example on HDL. At present HRT should probably not be administered to postmenopausal women with coronary artery disease (at least those who are symptomatic, those with unstable plaque, or those with a propensity for thrombosis). However, the effects of HRT on cardiovascular events in HERS cannot be extrapolated to asymptomatic postmenopausal women; clearly it has become important to confirm the epidemiological data with clinical trials. The ongoing Women’s Health Initiative may answer that specific question with

594 some 27,000 women randomized to estrogen, estrogen plus progestin, or placebo. These women will be followed for 9 years and the primary end points will be cardiovascular mortality and morbidity. 2. ESTROGEN AND THE BRAIN There has been increasing interest in the possible beneficial effects of estrogen on cognitive function in postmenopausal women. Initial interest was stimulated by a series of studies that evaluated the effects of estrogen on cognitive function in healthy postmenopausal women. These controlled short-term studies suggested that certain aspects of central nervous system (CNS) functioning could be improved. In particular, delayed but not immediate recall was improved and visual memory only was offered in one study [243] but verbal memory only in another [244]. The exact role of estrogen on cognitive function in normal women is not entirely clear, although distinct functional modifications of specific areas of the brain are evident on MRI [245]. Preliminary evidence mostly of an observational nature suggests that estrogen may retard the onset of Alzheimer’s disease (AD) or may improve cognitive function in patients with AD [246 – 250]. Several more basic studies indicate marked neurotrophic and neuroprotective effects of estrogen [251 – 253]. Small clinical trials have now suggested that there may be positive short-term effects of estrogen on attention and verbal memory in patients with AD [254,255]. However, no long-term large-scale studies have been performed as yet, in an arena notoriously difficult to study. The preventive effects of estrogen on declining cognitive function may be observed in the Women’s Health Initiative Memory Study that is part of the NIH-sponsored women’s health study referred to already. 3. ENDOMETRIAL CANCER There is no doubt that long-term therapy with unopposed estrogen increases the risk of endometrial hyperplasia and cancer [256]. The effect is both dose- and duration-dependent and may persist for many years after discontinuation of treatment [257 – 261]. The addition of a progestin in HRT regimens is primarily to protect the endometrium, and unopposed estrogen is now typically used in women following hysterectomy. Recent data suggest that the protection may be less than assumed and that there may be a small but significant occurrence of endometrial cancer in patients who use combined continuous preparations [262]. This is an area in which more data are clearly required. 4. BREAST CANCER The most serious concern of women considering longterm estrogen intervention for osteoporosis is the possible effects of estrogen on breast cancer. The rationale

LINDSAY AND COSMAN

that estrogen might be a risk factor for breast cancer is based upon epidemiological observations that suggest a downturn in the age-related increase in breast cancer coincident with the age of menopause, an increased risk of breast cancer in women with a long menarche-tomenopause duration, the protective effects of early pregnancy and in particular lactation, the high breast cancer risk in patients with high estrogenic states (PCO), and the fact that  80% of breast cancers can be shown to express estrogen receptors [263]. Observational studies of patients taking estrogen suggest a small but significant increase in risk of breast cancer with increased duration of exposure [264]. A recent metaanalysis of 51 epidemiological studies suggested an increase of 2.3% in breast cancer risk per year of estrogen exposure, with a relative risk of 1.35 after 5 years or more of estrogen exposure. The risk diminished after cessation of estrogen use and was at baseline in 5 years. The apparent increase in risk may be more apparent in thin women and in those given progestins versus those receiving estrogens alone. A more recent study suggested no increase in ductal carcinoma in situ or invasive ductal or lobular carcinoma but found an increase in a more rare type of breast carcinoma [265]. There is also a suggestion that the tumors that develop during estrogen administration are less aggressive. Progestins may produce another effect [266,267]. 5. OTHER CONCERNS The HERS trial confirmed prior observational studies suggesting that estrogen and HRT increase the risk of deep vein thrombosis and pulmonary embolus. The relative risk is about threefold for DVT, corresponding to a 0.6% annual incidence in the estrogen-treated HERS population vs 0.2% in the placebo group of HERS [213]. Deep vein thrombosis is a more unusual phenomenon among healthier older women without cardiovascular disease or other predisposing factors such as immobility, malignancy, and Leiden factor V (0.1% annual incidence in placebo-treated patients from the MORE study) [268]. Oral estrogens increase the relative risk of gallbaldder disease by approximately twofold [213]. Patients often state that estrogens are responsible for weight gain, but well controlled clinical trials have failed to confirm that. Breast tenderness is a common complaint on initiation of HRT. It often subsides and can be diminished by initiating therapy at a low dose and increasing the dose gradually. Vaginal bleeding occurs even with combined continuous therapy and is seen by patients’ as a side effect even if carefully explained. With many combined continuous regimens, if patients can tolerate the early intermittent variable bleeding that occurs, amennorrhea eventually is obtained in over 80% of patients.

595

CHAPTER 69 Estrogens and Osteoporosis 16.

V. SUMMARY Estrogen in many forms and doses can improve or maintain bone mass at all stages in postmenopausal life. Clinical trial data consistent with this effect include several small studies evaluating vertebral deformity and one study evaluating nonvertebral fracture. Two studies in patients with established cardiovascular disease have shown conflicting results on estrogen’s ability to reduce fracture risk. Use of estrogen to prevent osteoporotic fracture must be considered in light of multisystemic effects including the cardiovascular, central nervous, urogynecologic, and gastrointestinal systems as well as breast.

17.

18.

19.

20.

21.

References 22. 1.

F. Albright, F. Bloomberg, and P. H. Smith, Postmenopausal osteoporosis. Trans. Assoc. Am. Phys. 55, 298 – 305 (1940). 2. F. Albright, The effect of hormones on osteogenesis in man. Recent Prog. Horm. Res. 1, 293 – 353 (1947). 3. L. J. Melton III, Epidemiology of fractures. In Osteoporosis: Etiology, Diagnosis, and Management, (B. L. Riggs and L. J. Melton, eds.), 2nd ed Lippincott-Raven, NY, pp. 225 – 247 (1995). 4. J. A. Barrett, J. A. Baron, M. R. Karagas, and M. L. Beach, Fracture risk in the US Medicare population. J. Clin. Epidemiol. 52, 243 – 249 (1999). 5. T. W. O’Neill, D. Felsnberg, J. Warlow, C. Cooper, J. A. Kanis, and A. J. Silman, The prevalence of vertebral deformity in European men and women: The European Vertebral Osteoporosis Study. J. Bone Miner. Res. 11, 1010 – 1018 (1996). 6. O. Johnell, B. Gullberg, E. Allander, and J. A. Kanis, The apparent increase of hip fracture in Europe: A study of national registry sources. Osteoporosis Int. 2, 298 – 302 (1992). 7. S. Scott and D. H. Grey, The incidence of femoral neck fractures in New Zealand. N. Z. Med. J. 91, 6 – 9 (1980). 8. A. V. Schwartz, J. L. Kelsey, S. Maggi, et al. International variation in the incidence of hip fracture: Cross-national project in osteoporosis for the WHO. Osteoporosis Int. 9, 242 – 253 (1999). 9. R. A. Owen, L. J. Melton III, K. A. Johnson, D. M. Ilstrup, and B. L. Riggs, Incidence of Colles’ fracture in a North American community. Am. J. Public Health 72, 605 – 607 (1982). 10. L. J. Melton III, Epidemiology of age-related fractures. In “The Osteoporotic Syndrome: Detection, Prevention, and Treatment” (L. V. Avioli, ed.), pp.7 – 38. Wiley-Liss, New York, (1993). 11. V. Gilsanz, A. Kovanlikaya, G. Costin, T. F. Roe, J. Sayre, and F. Kaufman, Differential effect of gender on the size of the bones in the axial and appendicular skeletons. J. Clin. Endocrinol. Metab. 82, 1603 – 1607 (1997). 12. S. J. Richardson and J. F. Nelson, Follicular depletion during the menopausal transition. Ann. NY Acad. Sci. 592, 13 – 20 (1990). 13. C. Longcope, The endocrinology of the menopause. In “Treatment of the Postmenopausal Women: Basic and Clinical Aspects” (R. A. Lobo, eds.), Vol. 4, pp. 35 – 42. Lippincott Williams & Wilkins, Baltimore, (1999). 14. R. B. Wallace, B. M. Sherman, J. A. Beau, A. E. Treloar, and L. Schlabaugh, Probability of menopause with increasing duration of amenorrhea in middle-aged women. Am. J. Obstet. Gynecol. 135, 1021 – 1024 (1979). 15. C. Longcope, Adrenal and gonadal androgen secretion in normal females. J. Clin. Endocrinol. Metab. 15, 213 – 228 (1986).

23.

24.

25.

26.

27.

28.

29. 30. 31.

32. 33.

34.

35.

36.

M. G. Metcalf, R. A. Donald, and J. H. Livesey, Pituitary-ovarian function before, during and after the menopause: a longitudinal study. Clin. Endocrinol. 17, 489 – 494 (1982). B. L. Dennefors, P. O. Janson, F. Knutson, and L. Hamberger, Steroid production and responsiveness to gonadotropin in isolated stromal tissue of human postmenopausal ovaries. Am. J. Obstet. Gynecol. 136, 997 – 1002 (1980). P. K. Siiteri and P. C. MacDonald, Role of extraglandular estrogen in human endocrinology. In “Handbook of Physiology: Endocrinology” (R. O. Greep and E. Astwood, eds.), pp. 615 – 629. Vol. 2. American Physiology Society, Washington, DC, (1973). C. Longcope, The significance of steroid production by peripheral tissues. In “Endocrinology of the Ovary” (R. Schollar, ed.), pp. 23 – 25. Editions SEPE, Paris, (1978). R. Lindsay, J. R. T. Coutts, A. Sweeney, and D. M. Hart, Endogenous oestrogen and bone loss following oophorectomy. Calcif. Tissue Res. 22, 213 – 216 (1977). P. O. Kohler, G. T. Ross, and W. D. Odell, Metabolic clearance and production rates of human luteinizing hormone in pre- and postmenopausal women. J. Clin. Invest. 447, 38 – 47 (1968). A. R. Genazzani and F. Petraglia, Opioid control of luteinizing hormone secretion on humans. J. Steroid Biochem. 33, 751 – 755 (1989). D. Dempster Bone remodeling. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 355 – 380. Raven Press, NY, (1992). A. M. Parfitt, C. H. E. Matthews, A. R. Villanuea, M. Kleerekoper, B. Frame, and D. S. Rao, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis: Implications for the microanatomic and cellular mechanisms of bone. J. Clin. Invest. 72, 1396 – 1409 (1983). R. Pacifici, et al. Ovarian steroid treatment blocks a postmenopausal increase in blood monocytes interleukin-1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). R. Pacifici, et al. Effect of surgical menopause and estrogen replacement on cytokine release from human mononuclear cells. Proc. Natl. Acad. Sci. USA 88, 5136 – 5138 (1991). R. Kimble, S. Bani, and R. Pacifici, The functional block of TNF but not of IL-6 prevents bone loss in ovariectomized mice. J. Bone Miner. Res. 12, 935 – 941 (1997). S. Svristava, et al. Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). R. L. Jilka, Cytokines, bone remodeling, and estrogen deficiency: A 1998 update. Bone 23, 75 – 18 (1998). R. Pacifici, Estrogen, cytokines, and pathogenesis of postmenopausal osteoporosis. J. Bone Miners Res. 11, 1043 – 1051 (1996). M. L. Brandi, C. Cresciolo, A. Tanin, U. Frediani, D. Agnusdei, and C. Gennari, Bone endothelial cells as estrogen targets. Calcif. Tissue Int. 53, 312 – 317 (1993). J. A. Aubin and F. Liu, The Osteoblast Lineage. Principles Bone Biol. 51 – 67 (1996). M. Kasssem, R. Okazaki, S. A. Harris, T. C. Spelsberg, C. A. Conoer, and B. L. Riggs, Estrogen effects on insulin-like growth factor gene expression in a human osteoblastic cell line with high levels of estrogen receptor. Calcif. Tissue Int.; 62, 60 – 66 (1998). P. E. Keeting, R. E. Scott, D. S. Colvard, I. K. Han, T. C. Spelsberg, and B. L. Riggs, Lack of a direct effect of estrogen on proliferation and differentiation of normal human osteoblast-like cells. J. Bone Miner. Res. 6, 297 – 304 (1991). T. K. Gray, T. C. Flynn, K. M. Gray, and L. M. Namell, 17-Estradiol acts directly on the clonal osteoblastic cell line UMR106. Proc. Natl. Acad. Sci. USA 84, 6267 – 6271 (1987). M. Ernst, C. Schmid, and Er. Froesch, Enhanced osteoblast proliferation and collagen gene expression by estradiol. Proc. Natl. Acad. Sci. USA 85, 2307 – 2310 (1988).

596 37.

38.

39.

40.

41. 42.

43. 44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

LINDSAY AND COSMAN M. Ernst, J. K. Health, and G. A. Rodan, Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bone. Endocrinology 125, 825 – 833 (1989). D. E. I. Hughes, A. Dai, J. C. Tiffee, H. Lin H, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-. Nature Med. 2, 1132 – 1136 (1996). M. J. Oursler, P. Osdoby, J. Z. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Natl. Acad. Sci. USA 88, 6613 – 6617 (1991). J. E. Aubin, and F. Liu, The osteoblast lineage. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 51 – 67. Academic Press San Diego, CA, 1996 . Y. Onoe, C. Miyaura, H. Ohta, S. Nozawa, and T. Suda, Expression of estrogen receptor in rate bone. 138, 4509 – 4512 (1997). O. Vidal, S. Windahl, G. Andersson, J. A. Gustavsson, and C. Ohlsson, Increased cortical bone mineral content but unchanged trabecular bone mineral density in adult female mioce lacking estrogen receptor-. J. Bone Miner. Res. 14, 1, S141 (1999). H. M. Frost, “Bone Remodeling and Its Relationship to Metabolic Bone Disease.” Charles Thomas, Springfield, MA (1973) S. C. Manologas, Birth and death of bone cells: Basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocr. Rev. 21, 115 – 137 (2000) M. Seibel, F. Cosman, V. Shen, et al. Urinary hydroxypyridinium crosslinks of collagen as markers of bone resorption and estrogen efficacy in postmenopausal osteoporosis. J. Bone Miner. Res. 8, 881 – 889 (1993). R. Hesp, P. Hulme, D. Williams, and J. Reeve, The relationship between changes in femoral bone density and calcium balance in patients with involutional osteoporosis treated with human PTH fragment 1 – 34. Metab. Bone. Dis. Relat. Res. 2, 331 – 334 (1981). D. Alexandersen, C. Hassafe, and B. J. Riis, The effect of menopuase and hormone replacement therapy in bone alkaline Phosphatase. Scand. J. Clin. Lab. Invest. 55, 571 – 576 (1995). C. Hassager, A. Colwell, A. M. A. Assiri, R. Eastell, R. G. G. Russell, and C. Christiansen, Effect of menopause and hormone replacement therapy on urinary excretion of pyriidinium cross-links: A longitudinal and cross-sectional study. Clin. Endocrinol. 37, 45 – 50 (1992). C. Hassager, J. Ristel, I. Ristel, and C. Christiansen, Effect of menopause and hormone replacement therapy on the carboxy-terminal pyridinoline cross-linked telopeptide of type 1 collagen. Osteoporosis Int. 4, 349 – 352 (1994). I. Gerai, Y. Tapuchi, O. Chaki, et al. Specific changes of urinary excretion of cross-linked N-telopeptide of Type 1 Collagen in pre and postmenopausal women: Correlation with other markers of bone turnover. Calcif. Tissue Int. 60, 317 – 322 (1997). P. Ebeling, L. Atley, J. Guthrie, et al. Bone turnover markers and bone density across the menopausal transition. J. Clin. Endocrinol. Metab. 81, 3366 – 3371 (1996). P. Perchl, A. Griesmachei, et al. Association between female sex hormones and biochemical markers of bone turnover in pre and postmenopausal women. Calcif. Tissue Int. 62, 388 – 394 (1998). H. Hoshino, M. Takaholhi, et al. Urinary excretion of type 1 collagen degradation products in healthy women and osteoporotic patients with vertebral and hip fractures. Calcif. Tissue Int. 62, 36 – 39 (1998). A. Looker, D. C. Bauer, C. H. Chestnut, et al. Clinical use of biochemical markers of bone Remodeling. Osteoporosis Int. 11, 467 – 480 (2000). D. C. Bauer, D. M. Black, et al. Biochemical markers predict spine but not hip response to alendronate: The Fracture Intervention Trial (FIT) J. Bone Miner. Res. 12, (SI) 5150 (1997).

56.

57.

58. 59. 60. 61.

62. 63. 64.

65.

66.

67. 68.

69.

70.

71. 72.

73.

74. 75.

77. 78. 79. 80.

R. Marcus, L. Holloway, B. Wells, et al. The relationship of biochemical markers of bone turnover to bone density changes in postmenopausal women: Results from the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. J. Bone Miner. Res. 14, 1583 – 1595 (1999). F. Cosman, J. Nieves, C. Wilkinson, D. Schnering, V. Shen, and R. Lindsary, Bone density change and biochemical indices of skeletal turnover. Calcif. Tissue Int. 58, 236 – 243 (1996). H. E. Meema, Cortical bone atrophy and osteoporosis as a manifestation of aging. A. J Radiol. 89, 1287 – 1295 (1963). B. E. C. Nordin, Clinical significance and pathogenesis of osteoporosis. Br. Med. J. 1, 571 – 576 (1971). S. M. Garn, “In the Earlier Gain and Later Loss of Cortical Bone. Charles Thomas, Springfield, IL, (1970). B. E. C. Nordin, J. MacGregor, and D. A. Smith, The incidence of osteoporosis in normal women: Its relation to age and the menopause. Q. J. Med. 35, 25 – 38 (1966). S. M. Garn, C. G. Rohmann, and B. Wagner, Bone loss as a general phenomenon in men. Fed. Proc. 25,1729 – 1736. R. B. Mazess, On aging bone loss. Age Ageing 165, 239 – 252 (1982). S. H. Cohn, (ed.) “Non-invasive Measurement of Bone Mass and Their Clinical Application.” CRC Press, West Palm Beach, FL, (1980). H. Firooznia, C. Golumbu, M. Rafi, M. S. Schwartz, and E. R. Alterman, Quantative computed tomography assessment of spinal trabecular bone. 1. Age-related regression in normal men and women. J. Comput. Assist. Tomogr. 8, 91 – 97 (1984). P. Guesens, J. Dequeker, A. Verstraeten, and J. Nijs, Age and sexand menopause-related changes of vertebral and peripheral bone in population study using dual and single photon absorptiometry and radiogrammetry. J. Nucl. Med. 27, 1540 – 1549 (1986). C. Jasani, B. E. C. Nordin, D. A. Smith, and I. Swanson, Spinal osteoporosis and the menopause. Proc. R. Soc. Med. 58, 441 – 444 (1965). B. Krolner, and S. P. Nielsen, Bone mineral content of the lumbar spine in normal and osteoporotic women. Clin. Sci. 62, 329 – 336 (1982). M. Madsen, “Vertebral and peripheral bone mineral content by photon absorptiometry” (J. Menczel et al., eds.), pp. 80 – 86. Wiley Osteoporosis Chichester, UK, (1982). H. E. Meema, Menopausal and aging changes in muscle mass and bone mineral content. J. Bone Jt. Surg. Am. 48A, 1138 – 1144 (1966). L. Solomon, Bone density in aging Caucasian and African pouplations. Lancet 1326 – 1330 (1979I. R. Lindsay, F. Cosman, B. S. Herrington, and S. Himmelstein, Bone mass and body composition in normal women. J. Bone Miner. Res. 7, 55 – 63 (1992). H. M. Frost, The skeletal intermediary organization. A synthesis. In “Bone and Mineral Research” (W. A. Peck, ed.), Vol. 3, pp. 49 – 107. Elsevier Amsterdam, (1988). “Dietary Reference Intakes for Calicum, Phosphorus, Magnesium, Vitamin D, and Fluoride (DRIs)” National Academic Press. R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in calcium balance performance. J. Lab. Clin. Med. 92, 953 – 963 (1990). B. E. C. Nordin, Metabolic changes of the menopause. Calcif. Tissue Int. 41, 51 – 59 (1987). R. P. Heaney and R. R. Recker, Distribution of calcium absorption in middle-aged women. Am. J. Clin. Nutr. 43, 299 – 305 (1986). S. Epstein, G. Bryce, J. W. S. Hinman, et al. The influence of age on bone mineral regulating hormones. Bone 7, 421 – 425. (1986) J. M. Quesada, W. Coopmans, B. Ruiz, P. Aljama, I. Jans, and R. Bouillon, Influence of vitamin D on parathyroid function in the elderly. J. Clin. Endocrinol. Metab. 75, 494 – 501 (1992).

597

CHAPTER 69 Estrogens and Osteoporosis 81.

82. 83.

84. 85. 86. 87.

88.

89. 90.

91.

92.

93. 94.

95. 96. 97.

98.

99.

100.

101.

102.

103.

S. Pattanaungkul, B. L. Riggs, A. L. Yergey, N. E. Vieira, W. M. O’Fallon, and S. Khosla, Relationship of intestinal calcium absorption to 1,25-dihydroxyvitamin D [1,25 (OH)2D] levels in young versus elderly women: Evidence for age-related intestinal resistance to 1,25(OH)2D action. J. Clin. Endocrinol. Metab. 85, 4023 – 4027 (2000). R. P. Heaney, A unified concept of osteoporosis. Am. J. Med. 39, 377 – 380 (1985). F. Cosman, V. Shen, F. Xie, M. Seibel, A. Tatcliffe, and R. Lindsay, A mechanism of estrogen action on the skeleton: protection against the resorbing effects of (1-34)hPTH infusion as assessed by biochemical markers. Ann. Intern. Med. 118, 337 – 343 (1993). W. Willett, M. J. Stampfer, C. Bem, et al. Cigarette smoking relative weight, and menopause. Am. J. Epidemiol. 117, 651 – 658 (1983). A. S. Midgette, and J. A. Baron, Cigarette smoking and the risk of natural menopause. Epidemiology 1, 480 (1990). Am. J. Obstet. Gynecol. 276, 1747 (1992). D. L. Cassidenti, A. G. Vijod, M. A. Vijod, Stanczyk, and R. A. Lobo, Short-term effects of smoking on the pharmacokinetic profiles of micronized estradiol in postmenopausal women. Am. J. Obstet. Gynecol. 163, 1953 – 1960 (1990). G. D. Riebel, S. D. Boden, T. E. Whitesides, and W. C. Hutton, The effect of nicotine on incorporation of cancellous bone graft in an animal model. Spine 20, 2198 – 2202 (1995). “NOF Guidelines” E. S. Ginsberg, B. W. Walsh, X. Gao, et al. The effect of acute ethanol ingestion on estrogen levels in postmenopausal women using transdermal estradiol. J. Soc. Gynecol. Invest. 2, 26 – 29 (1995). E. S. Ginsberg, B. W. Walsh, B. F. Shea, et al. The effect of ethanol on the clearance of estradiol in postmenopausal women. Fertil. Steril. 63, 1227 – 1230 (1995). E. S. Ginsberg, N. K. Mello, and J. H. Mendelson, Effects of alcohol ingestion on estrogen in postmenopausal women. JAMA 276, 1747 – 1751 (1996). E. Seeman, The effects of tobacco and alcohol use on bone. Osteoporosis 577 – 597 (1996). J. R. Farley, R. Fitzsimmons, A. K. Taylor, W. M. Jorch, and K.-H. W. Lau, Direct effects of ethanol on bone resorption and formation in vitro. Arch. Biochem. Biophys. 238, 305 – 314 (1985). K. E. Friday and G. A. Howard, Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40, 562 – 565 (1991). Deleted in proof A. Hoffman, D. E. Grobbee, P. T. de Jong, and F. A. van den Ouweland, Determinants of disease and disability in the elderly: The Rotterdam Elderly Study. Eur. J. Epidemiol. 7, 403 – 422 (1991). W. M. Kohrt, D. B. Snead, E. Slatopolsky, and S. J. Birge, Jr. Additive Effects of Weight-Bearing Exercise and Estrogen on Bone Mineral Density in Older Women. J. Bone Miner. Res. 10, 9 (1995). D. M. Reid, R. A. Hughes, R. F. J. M. Laan, N. A. Sacco-Gibson, D. H. Wenderoth, S. Adami, R. A. Eusebio, J. P. Devogelaer, for the European Corticosteriod-Induced Osteoporosis Treatment Study. Efficacy and Safety of daily risedronate in the treatment of low bone mass in men and women receiving corticosteroids: a randomized trial. J. Bone Miner. Res. 15, 1006 – 1013 (2000). K. G. Saag, R. Emkey, T. Schnizer, et al. Alendronate for the treatment and prevention of glucocorticoid-induced osteoporosis. N. Engl. J. Med. 339, 292 – 299 (1998). C. E. Cann, H. K. Ganant, B. Ettinger, and G. S. Gordan, Spinal mineral loss in oophorectomized women. Determination by quantitative computed tomography. JAMA 244, 2056 – 2059 (1980). J. E. Block, R. Smith, C. C. Glueer, P. Steigere, B. Ettinger, and H. K. Genant, Models of spinal trabecular bone loss as determined by quantitative computed tomography. J. Bone Miner. Res. 4, 249 – 257 (1989). P. J. M. Elders, J. C. Netelenbos, P. Lips, M. D. vanGinkel Felicia Cosman, P. T. van der Stelt, Accelerated vertebral bone loss in relation to

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118. 119.

120. 121.

the menopause: A cross-sectional study on lumbar bone density in 286 women of 46 to 55 years of age. Bone Miner. 5, 11 – 19 (1988). H. Ohta, K. Makita, Y. Suda, T. Ikeda, T. Masuzawa, S. Nozawa, Influence of oophorectomy on seerum levels of sex steroids and bone metabolism and assessment of bone mineral density in lumbar trabecular bone by QCT-C value. J. Bone Miner. Res. 7, 659 – 665 (1992). F. Pansini, B. Bagni, G. Bonaccorsi, P. Albertazzi, L. Zanoti, A. Farina, C. Campobasso, R. Orlandi, and G. Mollica, Oophorectomy and spine bone density: Evidence of a higher rate of bone loss in surgical compared with spontaneous menopause. Menopause 2, 109 – 115 (1995). B. Ettinger, A. Pressman, P. Sklarin, D. C. Bauer, J. A. Cauley, and S. R. Cummings, Associations between low levels of serum estradiol, bone density, and fractures among elderly women: The study of osteoporotic fractures. J. Clin. Endocrinol. Metab. 83, 2239 – 2243 (1998). L. H. Kuller, J. A. Cauley, L. Lucas, S. R. Cummings, and W. S. Browner, Sex steroid hormones, bone mineral density, and risk of breast cancer. Environ. Health Perspect. 105 (Suppl. 3), 593 – 599 (1997). E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, K. S. Korach, Estrogen resistance caused by a mutation in the estorgen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effects of testosterone and estradiol in a man with aromatas deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). C. Slemenda, C. Longcope, L. Zhou, S. L. Hui, M. Peacock, C. C. Johnston, Sex steroids and bone mass in older men. Positive associations with serum estrogen, and negative associations with androgens. J. Clin. Invest. 100, 1755 – 1759 (1997). S. I. Hui, C. W. Slemenda, and C. C. Johnston, Age and bone mass as predictors of fracture in a prespective study. J. Clin. Invest. 81, 1804 – 1809 (1988). S. R. Cummings, D. m. Black, M. C. Nevitt, et al. Appendicular bone density and age predict hip fracture in women. JAMA 262, 665 – 668 (1989). L. J. Melton, E. J. Atkinson, W. M. O’Fallan, H. W. Wahner, B. L. Riggs, Long-term fracture prediction by bone mineral assessed at different skeletal sites. J. Bone Miner. Res. 10, 1227 – 1233 (1993). S. l. Hui, C. W. Slemenda, and C. C. Johnston, Baseline measurement of bone mass predicts fractures in white women. Ann. Intern. Med. 111, 355 – 361 (1989). R. Wasnich, P. D. Ross, L. K. Heibrun, and J. M. Vogel, Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am. J. Obstet. Gynecol. 153, 745 – 751 (1985). S. R. Cummings, D. M. Black, M. C. Nevitt, et al. Bone density at various sites for the prediction of hip fractures. Lancet 341, 72 – 75 (1993). P. Gardsell, O. Johnell, B. E. Nilsson, and B. Gullberg, Predicting various fragility fractures in women by forearm bone densitometry. A follow up study. Calcif. Tissue Int. 52, 348 – 353 (1993). R. P. Heaney, Is there a role for bone quality in fragility fractures? Calcif. Tissue Int. 53, 3 – 6 (1993). D. W. Dempster, E. S. Shane, W. Horbert, and R. Lindsay, A simple method for correlative light and scanning electronmicroscopy of human iliac crest bone biopsies: Qualitative observations in normal and osteoporotic subjects. J. Bone and Miner. Res. 1, 15 – 21 (1986). L. Mosekilde, Vertebral Structure and Strength In Vivo and In Vitro. Calcif. Tissue Int. 53 (Suppl.1), S121 – S126 (1993). R. Lindsay, J. R. T. Coutts, and D. M. Hart, The effects of endogenous estrogen on plasma and mineral calcium and phosphate in oophorectomized women. Clin. Endocrinol. 6, 87 – 93 (1977).

598 122. J. S. Lindberg, W. B. Fears, and M. M. Hunt, Exercise induced amenorrhea and bone density. Ann. Intern. Med. 101, 647 – 648 (1984). 123. S. L. Linnel and M. M. Stager, Blue Postmenopausal women Bone mineral content and menstrual regularity in female runners. Med. Sci. Sprots Exerc. 16, 343 – 348 (1989). 124. R. Marcus, C. Cann, and D. Madvig, Menstrual function and bone mass in elite women distance runners. Ann. Intern. Med. 102, 158 – 163 (1985). 125. C. F. Sanborn, B. J. Martin, W. W. Wagner, Is athletic amenorrhea specific to runners. Am. J. Obstet. Gynecol. 143, 859 – 861 (1982). 126. C. E. Cann, M. C. Martin, and H. Genant, Decreased spinal mineral content in amenorrheic women. JAMA 251, 626 – 629 (1984). 127. B. L. Drinkwater, K. Nilson, C. H. I. Chestnut, et al. Bone mineral content of amenorrhea and eumenorrheic athletics. N. Engl. J. Med. 311, 277 – 281 (1984). 128. E. R. Gonzalez, Premature bone loss in some non-menstruating sportwomen. JAMA 248, 513 – 514 (1982). 129. M. P. Warren, J. Brooks-Gunn, L. H. Hamilton, L. J. Warrant, and W. G. Hamilton, Lack of bone accretion and amenorrhea: evidencce for a relative osteopenia in weight-bearing bones. J. Clin. Endocrinol. Metab. 72, 847 – 853 (1991). 130. A. Klibanski, R. M. Neer, I. Z. Beitins, C. Ridgeay, N. T. Zervas, and J. MacArthur, Decreased bone density in hyperprolactenimic women. N. Engl. J. Med. 303, 1511 – 1514 (1980). 131. N. A. Rigotti, S. R. Nussbaum, D. B. Herzog, et al. Osteoporosis in women with anorexia nervosa. N. Engl. J. Med. 311, 1601 – 1606 (1984). 132. M. F. Warren, E. Shane, M. J. Lee, et al. Osteonecrosis of the hip associated with anorexia nervosa in a 20 year old ballet dancer. Clin. Orthop. 251, 171 – 176 (1990). 133. M. Putukian, The female triad eating disorders, amenorrhea and osteoporosis. Med. Clin. North Am. 78, 345 – 356 (1994). 134. F. Albright, P. H. Smith, and R. W. Fraser, A syndrome characterized by primary ovarian insufficiency and decreased stature: report of 11 cases with digression on hormonal control of axillary and pubic hair. Am. J. Med. Sci. 204, 625 – 648 (1942). 135. R. K. Beals, Orthopedic aspects of the XO (Turner’s) syndrome. Clin. Orthop. Relat. Res. 97, 19 – 30 (1973). 136. M. A. Smith, J. Wilson, and W. H. Price, Bone demineralization in patients with Turner’s syndrome. J. Med. Gen. 19, 100 – 103 (1982). 137. D. G. Barr, Bone deficiency in Turner’s syndrome measured by metacarpal dimensions. Arch. Dis. Child. 49, 821 – 822 (1974). 138. R. M. Shore, R. W. Chesney, R. B. Mazess, P. G. Rose, and G. H. Bargman, Skeletal demineralization in Turner’s syndrome. Calcif. Tissue Int. 34, 519 – 522 (1982). 139. D. M. Brown, J. Jowsey, D. S. Bradford, and T. Brincar, Osteoporosis in ovarian dysgenesis. J. Pediatr. 84, 816 – 820 (1974). 140. J. M. Aitken, D. M. Hart, J. B. Anderson, R. Lindsay, D. A. Smith, and C. F. Speirs, Osteoporosis after Oophorectomy for nonmalignant disease in postmenopausal women. Br. Med. J. I, 325 – 328 (1973). 141. J. M. Aitken, D. M. Hart, and R. Lindsay, Oestrogen replacement therapy for prevention of osteoporosis after oophorectomy. Br. Med. J. ii, 515 – 518 (1973). 142. R. Lindsay, F. Cosman, A. L. Rogeris, B. W. Walsh, S. T. Harris, J. E. Reagan, C. L. Liss, M. E. Melton, and C. A. Byrnes, Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84 (1999). 143. R. Lindsay, D. M. Hart, C. Forrest, and C. Baird, Prevention of spinal osteoporosis in oophorectomized women. Lancet 2, 1151 – 1153 (1980). 144. F. Al-Azzawi, D. M. Hart, and R. Lindsay, Long-term effect of oestrogen replacement therapy on bone mass as measured by dual photon absorptiometry. Br. Med. J. 294, 1261 – 1262 (1987).

LINDSAY AND COSMAN 145. A. Horsman, J. C. Gallagher, J. Simpson, and B. E. C. Nordin, Prospective trial of oestrogen and calcium in postmenopausal women. Br. Med. J. 2, 789 – 792 (1977). 146. R. R. Recker, P. D. Saville, and R. P. Heaney, Effects of estrogens and calcium carbonate on bone loss in postmenopausal women. Ann. Intern. Med. 87, 649 – 655 (1977). 147. C. Christiansen, M. S. Christensen, P. McNair, C. Hagen, K. E. Stocklund, and I. Transbol, Prevention of early postmenopausal bone loss: Controlled 2 year study in 315 normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). 148. H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gordon, Quantitative computed tomography of vertebral spongiosa: A sensitive method for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). 149. L. E. Nachtigall, R. H. Nachtigall, R. D. Nachitgall, and E. M. Beckman, Estrogen replacement therapy. I. A 10 – year prospective study in the relationship to osteoporosis. Obstet. Gynecol. 53, 277 – 281 (1979). 150. N. Munk-Jensen, S. P. Nielsen, E. B. Obel, and P. B. Eriksen, Reversal of postmenopausal vertebral bone loss by oestrogen and progestogen: A double blind placebo controlled study. Br. Med. J. 296, 1150 – 1152 (1988). 151. J. C. Gallagher, W. T. Kable, and D. Goldgar, Effect of progestin therapy on cortical and trabecular bone: Comparison with estrogen. Am. J. Med. 90, 171 – 178 (1991). 152. C. Christiansen and B. J. Riis, 17 beta-estradiol and continuous norethisterone: A unique treatment for established osteoporosis in elderly women. J. Clin. Endocrinol. Metab. 71, 836 – 841 (1990). 153. R. Lindsay, D. M. Hart, P. Purdie, M. M. Ferguson, A. C. Clark, and A. Kraszewski, Comparative effects of oestrogen and a progestogen on bone loss in postmenopausal women. Clin. Sci. Mol. Med. 54, 193 – 195 (1978). 154. C. Christiansen and P. Rodbro, Does postmenopausal bone loss respond to estrogen replacement therapy independent of bone loss rate. Calcif. Tissue 35, 720 – 722 (1983). 155. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen therapy prevents bone loss and fracture. Ann. Intern. Med. 102, 319 – 324 (1985). 156. R. Lindsay, D. M. Hart, and D. M. Clark, The minimum effective does of estrogen for prevention of postmenopausal bone loss. Obstet. Gynecol. 63, 759 – 763 (1984). 157. M. E. T. Quigley, B. L. Martin, A. M. Burnier, and P. Brooks, Estrogen therapy arrests bone loss in elderly women. Am. J. Obstet. Gynecol. 156, 1516 – 1523 (1987). 158. S. Meema, M. L. Bunker, and H. E. Meema, Preventive effect of estrogen on postmenopausal bone loss. Arch. Intern. Med. 135, 1436 – 1440 (1975). 159. G. F. Jensen, C Christiansen, and I. Transbol, Treatment of postmenopausal osteoporosis: a controlled therapeutic trial comparing oestrogen/gestagen, 1,25-dihydroxy-vitamin D3 and calcium. Clin. Endocrinol. 15, 515 – 524 (1982). 160. A. Caniggia, G. Delling, R. Nuti, and A. Vattimo, Clinical, biochemical and histological results of a double blind trial with 1,25 – dihydroxy-vitamin D3, estradiol and placebo in postmenopausal osteoporosis. Acta Vitaminol. Enzymol. 6, 117 – 128 (1984). 161. B. Ettinger, H. K. Genant, and C. E. Cann, Postmenopausal bone loss is prevented by treatment with low-dosage estrogen and calcium. Ann. Intern. Med. 106, 40 – 45 (1987). 162. M. J. Finn-Hensen, C. Christiansen, and I. Transbol, Postmenopausal bone loss is prevented by treatment with low-dosage estrogen with calcium. Obstet. Gynecol. 60, 493 – 496 (1982). 163. A. Horsman, M. James, and R. Francis, The effect of estrogen dose on postmenopausal bone loss. N. Engl. J. Med. 309, 1405 – 1407 (1983).

CHAPTER 69 Estrogens and Osteoporosis 164. F. N. Lafferty and D. O. Helmuth, Postmenopausal estrogen replacement: the prevention of osteoporosis and systemic effects. Maturitas 7, 147 – 159 (1985). 165. J. Linde, T. Friis, and E. Ostergaard, Oestrogens and postmenopausal bone loss. Calcif. Tissue Res. 21, 478 – 483 (1976). 166. S. T. Harris, H. K. Genant, D. J. Baylink, et al. The effect of estrogen (Ogen) on spinal bone density of postmenopausal women. Arch. Intern. Med. 151 (1980). 167. J. F. Aloia, A. Vaswani, J. K. Yeh, P. L. Ross, E. Flaster, and F. A. Dilmanian, Calcium supplementation with and without hormone replacement therapy to prevent postmenopausal bone loss. Ann. Intern. Med. 120, 7 – 103 (1994). 168. S. P. Nielsen, O. Barenholdt, J. F. Hermansen, and N. Munk-Jensen, Magnitude and pattern of skeletal response to long term continuous and cyclic sequential oestrogen/progestin treatment. Br. J. Obstet. Gynaecol. 101, 319 – 324 (1994). 169. D. T. Felson, Y. Zhang, M. T. Hannan, D. P. Kiel, W. F. Wilson, and J. J. Anderson, The effect of postmenopausal estrogen therapy on bone density in elderly women. N. Engl. J. Med. 329, 1141 – 1146 (1993). 170. T. L. Bush, et al. for the PEPI writing group. Effects of hormone therapy on bone mineral density JAMA 276, 1389 – 1396 (1996). 171. M. Savvas, J. W. W. Studd, S. Norman, A. T. Leather, and T. J. Garnett, Increase in bone mass after one year of percutaneous oestradiol and testosterone implants in postmenopausal women who have previously received long-term oestrogens. Br. J. Obstet. Gynecol. 99, 757 – 760 (1992). 172. M. Peacock and P. Selby, The effect of transdermal oestrogen on bone, calcium regulating hormones and lier in postmenopausal women. Clin. Endocrinol. 25, 543 – 547 (1986). 173. J. C. Stevenson, M. P. Cust, K. F. Gangar, T. C. Hillard, B. Lees, and M. I. Whitehead, Effects of transdermal versus oral hormone replacement therapy on bone density in spine and proximal femur in postmenopausal women. Lancet 336, 265 – 269 (1990). 174. T. Steiniche, C. Hasling, P. Charles, et al. A randomized study of the effects of estrogen-gestagen or high does oral calcium on trabecular bone remodeling in postmenopausal osteoporosis. Bone 10, 313 – 320 (189). 175. S. L. Nand, B. G. Wren, B. A. Gross, and G. Z. Heller, Bone density effects of continuous estrone sulfate and varying doses of medroxyprogesterone acetate: Hormones and bone density. Obstet. Gynecol. 93(6) (1999). 176. L. Speroff, J. Rowan, J. Symons, H. Genant, and W. Wilborn, The comparative effect of bone density, endometriu, and lipids of continuous hormones as replacement therapy (CHART Study). JAMA 276(17)1397 – 1403 (1996). 177. W. Kuhnz, A. Heuner, M. Hümpel, W. Seifert, and K. Michaelist, In vivo conversion of norethisterone and norethisterone acetate to ethinyl estradiol in postmenopausal women. Contraception 56, 379 – 385 (1997). 178. Deleted in proof 179. A. M. Parfitt, The physiologic and pathogenetic significance of bone histomophometric data. In “Disorders of Bone and Mineral Metabolism,” (F. L. Coe, M. J. Favus, eds.), Vol. 22, pp. 475 – 489 Raven Press, NY, (1992). 180. C. Christiansen, P. Rodboro, Does oestriol add to the beneficial effect of combined hormonal prophylaxis against early postmenopausal osteoporosis. Br. J. Obstet. Gynecol 91, 489 – 493 (1984). 181. P. Garnero, Y. Tsouderos, I. Marton, C. Pelissier, C. Varin, and P. d. Delmas, Effects of Intranasal 17-Estradiol on Bone Turnoer and Serum Insulin-Like Growth Factor I in Postmenopausal Women. J. Clin. Endocrinol. Metab. 84(7) (1999). 182. E. Farish, D. M. Hart, T. M. Jackanicz, C. W. Bardin, and R. Lindsay, Hormone replacement therapy by vaginal ring pessaries: Effects

599 on biochemical indices of bone metabolism. International Symposium on Osteoporosis, Copenhagen, Denmark (1984). 183. B. J. Riis, K. Thomsen, V. Strom, and C. Christiansen, The effect of percutaneous estradiol and progesterone on postmenopausal bone loss. Am. J. Obstet. Gynecol. 156, 61 – 65 (1987). 184. C. Cooper, J. A. Stakkestad, S. Rdowicki, P. Hardy, C. Pilate, M. P. Dain, P. D. Delmas, for the International Study Group Matrix delivery transeermal 17-estradiol for the prevdention of bone loss in postmenopausal women. Osteoporosis Int. 9, 358 – 366 (1999). 185. H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gordon, Quantitative computed tomography of vertebral spongiosa: A sensitive metnod for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). 186. N. B. Watts, J. C. Nolan, J. J. Brennan, and H. M. Yang, Esterified estrogen therapy in postmenopausal women. Relationships of bone marker changes and plasma estradiol to BMD changes: A two-year study. Menopause 7, 375 – 382 (2000). 187. S. R. Weiss, H. Ellman, and M. Dolker, A randomized controlled trial of four doses of transdermal estradiol for preventing postmenopausal bone loss. Obstet. Gynecol. 94, 330 – 336 (1999). 187a.R. R. Recker, K. M. Davies, R. M. Dowd, and R. P. Heaney, The effect of low-dose continuous estrogen and progresterone therapy with calcium and vitamin D on bone in elderly women: A randomized, controlled trial. Ann. Int. Med. 130, 897 – 904 (1999). 188. J. A. Kotzan, B. C. Martin, and W. E. Wade, Persistence with estrogen therapy in postmenopausal medical population. Pharmacotherapy 19(3), 363 – 369 (1999). 189. N. L. Keating, P. D. Cleary, A. S. Rossi, A. M. Zaslavsky, and J. Z. Ayanian, Use of hormone replacement therapy by postmenopausal women in the United States. Ann. Intern. Med. 130(7), 545 – 553 (1999). 190. J. A. Cauley, S. R. Cummings, D. M. Black, S. R. Mascioli, and D. G. Seeley, Prevalence and determinants of estrogen replacement therapy in elderly women. Am. J. Obstet. Gynecol. 163(5 Pt 1), 1438 – 1444 (1990). 191. J. Y. Reginster, N. Sarlet, R. Deroisy, A. Albert, U. Gaspard, and P. Franchimont, Minimal levels of serum estradiol prevent postmenopausal bone loss. Calcif. Tissue Int. 51, 340 – 343 (1992). 192. D. C. Bauer, J. A. Cauley, K. E. Ensrud, M. C. Nevitt, K. Stone, and S. R. Cummings, Women with low serum estradiol have an increased risk of vertebral fracture. Osteoporosis Int. 6(Suppl. 1), 92 (1996). 193. L. E. Nachtigall, R. H. Nachtigall, R.D. Nachtigall, and E. M. Beckman, Estrogen replacement therapy I. A 10-year prospective study in the relationship to osteoporosis. Obstet. Gynecol. 53(3), 277 – 281 (1979). 194. U. A. Liberman, S. R. Weiss, J. Bröll, H. W. Minne, H. Quan, N. H. Bell, J. Rodriguez-Porales, R. W. Downs Jr, J. Dequeker, M. Favus, E. Seeman, R. R. Recker, T. Capizzi, A. C. Santora II, A. Lombardi, R. V. Shah, L. J. Hirsch, and D. B. Karpf, Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N. Engl. J. Med. 333, 11437 – 1443 (1995). 195. S. T. Harris, N. B. Watts, H. K. Genant, C. D. McKeever, T. Hangartner, M. Keller, C. H. Chestnut III, Brown Jacques, E. F. Eriksen, M. S. Hoseyni, D. W. Axelrod, and P. D. Miller, Effects of residronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis. JAMA 282, 1344 – 1352 (1999). 196. J. E. Rossouw, L. P. Finnegan, W. R. Harlan, V. W. Pinn, C. Clifford, and J. A. McGowan, The evolution of the Women’s Health Initiative: perspectives from the NIH. J. Am. Med. Womens. Assoc. 50, 50 –55 (1995). 197. S. Wallach and P. H. Henneman, Prolonged estrogen therapy in postmenopausal women. JAMA 113, 1637 – 1642 (1959). 198. E. G. Lufkin, H. W. Wahner, W. M. O’Fallon, et al. Treatment of postmenopausal osteoporosis with transdermal estrogen. Ann. Intern. Med. 117, 1 – 9 (1992).

600 199. S. Høidrup, M. Grønbaek, A. T. Pedersen, J. B. Lauritzen, A , M. Schroll, Hormone replacement therapy and hip fracture risk: Effect modification by tobacco smoking, alcohol intake, physical activity, and body mass index. Am. J. Epidemiol. 150(10), 1085 – 1093 (1999). 200. J. A. Cauley, D. G. Seeley, K. Ensrud, et al. Estrogen replacement therapy and fractures in older women. Ann. Intern. Med. 122, 9– 16 (1995). 201. K. Michaëlsson, J. A. Baron, B. Y. Farahmand, et al. Hormone replacement therapy and risk of hip fracture: population based casecontrol study. Br. Med. J. 316, 1858 – 1863 (1998). 202. J. A. Kanis, O. Johnell, B. Gullberg, et al. Evidence for efficacy of drugs affecting bone metabolism in preventing hip fracture. Br. Med. J. 305, 1125 – 1128 (1992). 203. N. S. Weiss, C. L. Ure, J. H. Ballard, A. R. Williams, and J. R. Daling, Decreased risk of fractures of the hip and lower forearm with postmenopausal of estrogen. N. Engl. J. Med. 303(21), 1195 – 1198 (1980). 204. N. Kreiger, J. L. Kelsey, T. R. Holford, et al. An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 114 – 118 (1982). 205. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Intern. Med. 102, 319 – 324 (1985). 206. D. P. Kiel, D. T. Felson, J. J. Andersen, P. W. F. Wilson, and M. A. Moskowitz, Hip fracture and the use of estrogens in postmenopausal women. The Framingham study. N. Engl. J. Med. 317, 1169 – 1174 (1987). 207. G. S. Gordan, J. Picchi, and B. S. Roof, Antifracture efficacy of long-term estrogens for osteoporosis. Trans. Assoc. Am. Physicians 86, 326 – 332 (1973). 208. T. A. Hutchinson, S. M. N. Polansky, and A. R. Feinstein, Postmenopausal oestrogens protect against fractures of hip and distal radius. Lancet ii, 706 – 710 (1979). 209. T. Naessén, I. Persson, H.-O. Adami, et al. Hormone replacement therapy and the risk for first hip fracture. Ann. Intern. Med. 113, 95 – 103 (1990). 210. E. R. Johnson and E. E. Specht, The risk of hip fracture in postmenopausal females with and without estrogen drug exposure. Am. J. Public Health 71, 138 – 144 (1981). 211. A. R. Williams, N. S. Weiss, C. L. Ure, J. Ballard, and J. R. Daling, Effect of weight, smoking, and estrogen use of the risk of hip and forearm fractures in postmenopausal women. Obstet. Gynecol. 60, 695 – 699 (1982). 211a.D. Grady, S. M. Rubin, D. B. Petitti, C.S. Fox, D. Black, B. Ettinger, V. I. Ernster, and S. R. Cummings, Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann. Intern. Med. 117, 1016 – 1037 (1992). 212. M. H. Komulainen, H. Kröger, M. T. Tuppurainen, A.-M. E. Heikkinen,, R. Honkanen, S. Saarikoski, Kuopio Univ Hosp, Kuopio, Finland; Univ Tromsö, Tromsö, Norway, H. R. T. and Vit D in prevention of non-vertebral fractures in postmenopausal women: A 5 year randomized trial. Maturitas 31, 45 – 54 (1998). 213. S. Hulley, D. Grady, T. Bush, C. Furberg, D. Herrington, B. Riggs, and E. Vittinghoff, Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. JAMA 280, 605 – 613 (1998). 214. R. Lindsay, D. M. Hart, and A. Kraszewski, Prospective doubleblind trial of synthetic steroid (Org OD14) for preventing postmenopausal osteoporosis. Br. Med. J. 280, 1207 – 1209 (1980). 215. J. Rymer, M. G. Chapman, and I. Fogelman, Effect of tibolone on postmenopausal bone loss. Osteoporosis Int. 4, 314 – 319 (1994). 216. N. Bjarnason, K. Bjarnason, J. Haarbo, et al. Tibolone: Prevention of bone loss in late postmenopausal women. J. Clin. Endocrinol. Metab. 81, 2419 – 2422 (1996). 217. B. Berning, C. Kuijk van, J. W. Kuiper, et al. The effects of two doses of tibolone on trabecular and cortical bone loss in early post-

LINDSAY AND COSMAN

218.

219.

220.

221.

222.

223.

224.

225. 226.

227.

228.

229.

230.

231.

232.

233.

234.

menopausal women: A two year randomized placebo controlled study. Bone 19, 395 – 399. D. Thiebaud, J. M. Bigler, S. Renteria, et al. A 3-year study of prevention of postmenopausal bone loss: Conjugated equine estrogens plus medroxyprogesterone acetate versus tibolone. Climacteric 1, 202 – 210 (1998). P. Geusens, J. Dequeker, J. Gielen, et al. Non-linear increase in vertebral density induced by a synthetic steroid (Org OD14) in women with established osteoporosis. Maturitas 13, 155 – 162 (1991). J. Studd, I. Armala, P. M. Kicovic, et al. A randomized study of tibolone on bone mineral density in osteoporotic postmenopausal women wiwth previous fractures. Obstet. Gynecol. 92, 574 – 579 (1998). P. W. Pavlov, J. Ginsburg, P. M. Kicovic, D. B. vand der Schaaf, G. Prelevic, and H. J. T. Coelingh Bennink, Double-blind, placebo-controlled study of the effects of tibolone on bone mineral density in postmenopausal osteoporotic women with and without previous fractures. Gynecol. Endocrinol. 13, 230 – 237 (1991). P. M. Kicovic, J. Cortes-Prieto, M. Luisi, et al. Placebo controlled cross-over study of effects of Org OD14 in menopausal women. Reproduccion 6, 81 – 91 (1982). L. J. Benedek-Jaszmann, Long-term placebo-controlled efficacy and safety study of Org OD14 in limacteric women. Maturitas(Suppl.) 1, 25 – 33 (1987). R. Punnonen, P. Liukko, J. Cortes-Prieto, et al. Multicentre study of effects of Org OD14 on endometrium, vaginal cytology and cervical mucus in post-menopausal and oophorectomized women. Maturitas 5, 281 – 286 (1984). A. R. Genazzani, L. J. Benedek-Jaszmann, D. M. Hart, et al. Org OD14 and th endometirum. Maturitas 13, 243 – 251 (1991). J. C. Gallagher, W. T. Kable, and D. Goldgar, Effect of progestin therapy on cortical and trabecular bone: Comparison with estrogen. Am. J. Med. 90, 171 – 178 (1991). R. Lindsay, D. M. Hart, D. Purdie, M. M. Ferguson, A. C. Clark, and A. Kraszewski, Comparative effects of oestrogen and a progestogen on bone loss in postmenopausal women. Clin. Sci. Mol. Med. 54, 193 – 195 (1978). H. I. Abdalla, D. M. Hart, R. Lindsay, I. Leggate, and A. Hooke, Prevention of bone loss in postmenopausal women by norethisterone. Obstet. Gynecol. 6, 789 – 792 (1985). S. R. Davis, P. McCloud, B. J. G. Strauss, H. Burger, Testosterone enhances estradiol’s effects on postmenopausal bone density and sexuality. Maturitas 21, 227 – 236 (1995). S. J. Wimalawansa, Prevention and treatment of osteoporosis: Effi cacy of combination of hormone replacement therapy with other antiresorptive agents. J. Clin. Densitom. 3(2), 187 – 201 (2000). H. G. Bone, S. L. Greenspan, C. McKeever, N. Bell, M. Davidson, R. W. Downs, R. Emkey, P. J. Meunier, S. S. Miller, A. L. Mulloy, R. R. Recker, S. R. Weiss, N. Heyden, T. Musliner, S. Suryawanshi, A. J. Yates, Lombardi. Alendronate and estrogen effects in postmenopausal women with low bone mineral density. J. Clin. Endocrinol. Metab. 720 – 726 (2000). R. Lindsay, J. W. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, D. W. Dempster, and F. Cosman, Randomised controlled study of effect of parathyroid hormone on vertebral bone mass and fractaure incidence among postmenopausal women on estrogen with osteoporosis. Lancet 350, 550 – 555 (1997). F. Cosman, J. Nieves, L. Woelfert, C. Formica, S. Gordon, V. Shen, and R. Lindsay, Parathyroid hormone added to established hormone therapy: Effects on vertebral fracture and maintenance of bone mass after PTH withdrawal. J. Bone Mineral Res., in press (2001). N. E. Lane, S. Sanchez, G. Modin, H. K. Genant, E. Pierini, and C. D. Arnaud, Parathyroid hormone treatment can reverse corticosteroid-induced osteoporosis. Results of a randomized controlled clinical. J. Clin. Invest. 102, 1627 – 1633 (1998).

CHAPTER 69 Estrogens and Osteoporosis 235. E. B. Roe, S. D. Sanchez, G. A. Del Puerto, E. Pierini, P. Bacchetti, C. E. Cann, and C. D. Arnaud, Parathyroid Hormone 1-35 (hPTH 134) and Estrogen Produce Dramatic Bone Density Increases in Postmenopausal Osteoporosis — Results from a Placebo-Controlled Randomized Trial. J. Bone Miner. Res. 14 (Suppl. 1), S137 (1999). 236. B. Ettinger, A. Pressman, and P. Silver, Effect of age on reasons for initiation and discontinuation of HRT. Menopause 6, 282 – 289 (1999). 237. G. I. Grodeski and W. H. Utian, Epidemiology and risk factors of cardiovascular disease in postmenopausal women. Treat. Postmenopausal Women 30, 331 – 359 (1999). 238. M. J. Stampfer and F. Grodstein, Role of hormone replacement in cardiovascular disease. Treat. Postmenopausal Women 31, 361 – 367 (1999). 239. R. M. Krauss, Lipids and lipoproteins and effects of hormone replacement. Treat. Postmenopausal Women 32, 369 – 376 (1999). 240. M. R. Adams, S. A. Washburn, J. D. Wagner, J. K. Williams, and TB. Clarkson, Arterial changes: Estrogen deficiency and effects of hormone replacement. Treat. Postmenopausal Women 33, 377 – 384 (1999). 241. P. Collins, Estrogen — Blood flow and vasomotion. Treat. Postmenopausal women 34, 385 – 389 (1999). 242. L. Speroff, The heart and estrogen/progestin replacement study (HERS). Maturitas 31, 9 – 14 (1998). 243. S. M. Resnick, E. J. Metter, and A. B. Zonderman, Estrogen replacement therapy and longitudinal decline in visual memory. Am. Acad. Neurol. 1491 – 1497 (1997). 244. D. L. Kempen and B. B. Sherwin, Estrogen use and vertebral memory in healthy postmenopausal women. Obstet. Gynecol. 83, 979 – 983 (1994). 245. S. E. Shaywitz, B. A. Shaywitz, K. R. Pugh, R. K. Fulbright, P. Skudlarski, W. E. Mencl, R. T. Constable, F. Naftolin, S. F. Palter, K. E. Marchione, L. Katz, D. P. Shankweiler, J. M. Fletcheer, C. Lacadie, M. Keltz, and J. C. Gore, Effect of Estrogen on Brain Activation Patterns in Postmenopausal Women During Working Memory Tasks. JAMA 281, 1197 – 1202 (1999). 246. A. Paganini-Hill, Does estrogen replacement therapy protect against Alzheimer’s disease? Osteoporosis Int. (Suppl. 1), S12 – S17 (1997). 247. A. J. C. Slooter, J. Bronzova, J. C. M. Witteman, C. Van Broeckhoven, A. Hofman, and C. M. vanDuijn, Estrogen use and early onset Alzheimer’s diease: A population-based study. J. Neurol. Neurosurg. Psychiatry 67, 779 – 781 (1999). 248. S. C. Waring, W. A. Rocca, R. C. Petersen, P. C. O’Brien, E. G. Tangalos, E. Kokmen, Postmenopausal estrogen replacement therapy and risk of AD. Am. Acad. Neurol. 965 – 970 (1999). 249. K. Yaffe, G. Saway, I. Lieberburg, and D. Grady, Estrogen therapy in postmenopausal women. JAMA 279 (9), 688 – 695 (1998). 250. V. W. Henderson, Estrogen replacement therapy for the prevention and treatment of Alzheimer’s disease. CNS Drugs 8(5), 343 – 351 (1997). 251. B. McEwen, S. Alves, K. Bulloch, and N. Weiland, Ovarian steroids and the brain: Implications for cognition and aging. Neurology 48 (Suppl. 7), S8 – S15 (1997). 252. B. McEwen, A. Bigeon, C. Fishetee, V. Luine, B. Parsons, and T. Rainbow, Toward a neurochemical basis of steroid hormone action. In “Frontiers in Neuroendocrinology” (L. Martin and W. Ganog, eds.), pp. 153 – 176 Raven Press, New York (1984). 253. B. McEwen and C. Woolley, Estradiol and progesterone regulate neuronal structure and synaptic connectivity in adult as well as devewloping brain. Exp. Gerontol. 29, 431 – 436 (1994).

601 254. V. W. Henderson, et al. Short-term estrogen therapy does not appear to improve the symptoms of Alzheimer’s disease (AD). Neurology 54, 295 – 301 (2000). 255. S. Asthana, S. Craft, L. D. Baker, M. A. Raskind, R. S. Birnbaum, C. P. Lofgreen, R. C. Veith, and S. R. Plymate, Cognitive and neuroendocrine response to transdermal estrogen in postmenopausal women with Alzheimer’s disease: Results of a placebo-controlled, doubleblind, pilot study. Psychoneuroendocrinology 24, 657 – 677 (1999). 256. J. D. Woodruff and J. H. Pickar, for the Menupause Study Group, Incidence of endometrial hyperplasia in postmenopausal women taking conjugated estrogen (Premarin) with medroxyprogesterone acetate or conjugated estrogens alone. Am. J. Obstet. Gynecol. 170(5) 1213 – 1223 (1994). 257. K. L. Cushing, N. S. Weiss, L. F. Oigt, B. McNight, and S. A. A. Beresford, Risk of endometrial cancer in relation to use of low-dose, unopposed estrogens. Obstet. Gynecol. 91, 35 – 39 (1998). 258. The Writing Group for the PEPI Trial, Effects of hormone replacement therapy on endometrial histology in postmenopausal women. JAMA 275, 370 – 376 (1996). 259. S. Shapiro, J. P. Kelly, L. Rosenberg, D. W. Kaufman, S. P. Helmrich, N. B. Rosenshein, J. L. Lewis Jr., R. C. Knapp, and P. D. Stolley, D. Schottenfeld, Risk of Localized and widespread endometrial cancer in relation to recent and discontinued use of conjugated estrogens. N. Engl. J. Med. 313, 969 – 962 (1985). 260. S. S. Jick, A. M. Walker, and H. Jick, Estrogens progesterone, and endometrial cancer. Epidemiology 4, 20 – 24 (1993). 261. A. Paganini-Hill, R. K. Ross, and B. E. Henderson, Endometrial cancer and patterns of use of oestrogen replacement therapy: A cohort study. Br. J. Cancer 59, 445 – 447 (1989). 262. A. T. Leather, M. J. Savvas, and J. W. Studd, Endometrial histology and bleeding patterns after 8 years of continuous comibned estrogen and progestogen therapy in postmenopausal women. Obstet. Gynecol. 78, 1008 – 1010 (1991). 263. B. Zumoff, Does postmenopausal estrogen administration increase the risk of breast cancer? Contributions of animal, biochemical, and clinical investigataive studies to a resolution of the controversy (44202). Soc. Exp. Biol. Med. 30 – 37 (1998). 264. Collaborative Group on Hormonal Factors in Breast Cancer, Breast cancer and hormonal contraceptives: Collaborative reanalysis of individual data on 53 297 women with breast cancer and 100 239 women without breast cancer from 54 epidemiological studies. Lancet 347, 1713 – 1727 (1996). 265. S. M. Gapstur, M. Morrow, and T. A. Sellers, Hormone replacement therapy and risk of breast cancer with a favorable histology. JAMA 281, 2091 – 2097 (1999). 266. C. Schairer, J. Lubin, R. Troisi, S. Sturgeon, L. Brinton, and R. Hooveer, Menopausal estrogen and estrogen-progestin replacement therapy and breast cancer risk. JAMA 283, 485 (2000). 267. C. Magnusson, J. A. Baron, N. Correia, R. Bergström, and H. O. Adami, I. Persson. Breast-cancer risk following long-term oestrogen and oestrogen-progestin-replacement therapy. Int. J. Cancer 81, 339 – 344 (1999). 268. B. Ettinger, D. M. Black, B. H. Mitlak, R. K. Knickerbocker, T Nickelsen, H. K. Genanat, C. Christiansen, P. D. Delmas, J. R. Zanchetta, J Stakkestad, C. C. Glüer, K. Kureger, F. J. Cohen, S. Eckert, K. E. Ensrud , L. V. Avioli, P. Lips, and S. R. Cummings, for the Multiple Outcomes of Raloxifene Evaluation (MORE) Investigators, Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene. Results from a 3-year randomized clinical trial. JAMA 282, 637 – 645 (1999).

CHAPTER 70

Selective Estrogen Receptor Modulators (SERMs) ETHEL S. SIRIS DOUGLAS B. MUCHMORE

Columbia University College of Physicians and Surgeons, and Toni Stabile Center for the Prevention and Treatment of Osteoporosis, Columbia – Presbyterian Medical Center, New York, New York 10032 Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285

I. Introduction II. Mechanism of Action of SERMs

III. Clinical Application of SERMs References

I. INTRODUCTION

nuclear transcription factors, with over 150 members having been identified to date. The superfamily is composed of three subclasses of nuclear receptors: type I receptors mediate steroid hormone action (estrogens, androgens, progestagens, glucocorticoids, and mineralocorticoids); type II receptors mediate the actions of thyroid hormones, vitamin D, and retinoic acids; and the remaining receptors are “orphans,” for which cognate ligands have not yet been identified [1]. All of the nuclear receptors share certain structural analogies, including conservation of a domain organization. The N-terminal sequence of these proteins (the A/B domain) constitutes a transactivation function region (AF1) which may act to modulate gene transcription independent of ligand occupancy of the receptor. The C domain (a highly conserved region) along with the D domain confer specific ability of the receptor to bind to consensus estrogen response element (ERE) DNA sequences. As will be discussed in greater detail below, this binding interaction is essential for the “classical” gene transcriptional activities of ligand-bound receptors. Ligand binding occurs in the E domain of the protein, which forms a hydrophobic pocket in the interior region of the receptor. The E domain also contributes to the structure of a second transactivation

Selective estrogen receptor modulators (SERMs) (also referred to as tissue-selective estrogens) have been introduced into clinical practice for the treatment of breast cancer (tamoxifen, toremifene), the reduction in risk of breast cancer in high-risk women (tamoxifen) and the prevention and treatment of osteoporosis in postmenopausal women (raloxifene). These diverse clinical activities are tied together by virtue of the mechanisms of action of these compounds, which center on their interaction with estrogen receptors. This chapter will review recent advances in understanding these mechanisms, followed by a review of data that supports the clinical use of SERMs.

II. MECHANISM OF ACTION OF SERMS A. Background By definition, SERMs exert their biological effects through specific, high-affinity interaction with estrogen receptors. Estrogen receptors belong to a superfamily of

OSTEOPOROSIS, SECOND EDITION VOLUME 2

603

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

604

SIRIS AND MUCHMORE

function region (AF-2) which interacts with other transcriptional proteins (coactivators and corepressors) to modulate gene transcription after ligand binds to the receptor. Another important function of the E domain is to facilitate receptor dimerization after ligand binding occurs. Finally, some members of the nuclear receptor superfamily (including estrogen receptors) contain a C-terminal F domain. A portion of the F domain includes the C-terminal alpha helix of the receptor, and, following ligand binding, this structure folds across the entrance to the ligand binding pocket, appearing to “lock” the ligand in place. As will be discussed below, the position of this helix is altered when estrogen antagonists are bound to the receptor, and it appears to be crucial in interfering both with receptor dimerization and with interaction of the ligand-bound receptor with transcriptional coactivators. These insights, along with other advances in nuclear receptor biology, have elucidated new layers of complexity in the systems responsible for mediating estrogen effects. It is now possible to develop coherent although incomplete explanations for the seemingly contradictory properties that allow SERMs to act as estrogen agonists in some tissues and as estrogen antagonists in others (see also Chapter 10).

sion that differing ligands could have differing actions (either estrogen agonist or estrogen antagonist) in different tissues. The discussion to follow will review recent progress in understanding estrogen receptor biology as it relates to SERM action.

C. Antagonist Mechanism of SERMs X-ray crystallographic evidence demonstrates that the ligand binding pocket of the estrogen receptor is relatively spacious and can thus accommodate a number of different ligands in a promiscuous manner [8]. As shown in Fig. 1 (see also color plate), estradiol and raloxifene bind to the same structural domain of the estrogen receptor. Both ligands contain critical hydroxyl groups which reside at opposite ends of their respective core structures, and these substituents coordinate to the same amino acids in the receptor pocket [9]. However, the basic side chain of receptor-bound raloxifene causes a displacement of the C-terminal (-helix of the estrogen receptor (helix 12, a part of the F domain). This element then rotates away from its usual location to a new, stable position where it blocks access to

B. Mechanism of Estrogen Action Understanding the mechanisms of SERM action depends on knowledge of the mechanism of estrogen action. The paradigm of classical estrogen action emerged over the past three decades. Estrogen action begins with high affinity binding of ligand to the receptor. As a result of this interaction, conformational changes in the receptor occur, leading to dissociation of the inactive receptor from some proteins (e.g., heat shock protein or hsp-90) and interaction with other proteins (such as proteins from the steroid receptor coactivator, or SRC, family), leading to receptor dimerization. This is then followed by activation of gene transcription at DNA promoter sites containing the ERE consensus sequence [2]. The ERE promoter activates gene transcription of numerous classical estrogen target genes such as vitellogen or progesterone receptor. These gene products are generally associated with reproductive tissues and their functions (see Chapter 10). It is now known that several distinct gene promoter sequences, including the AP-1 promoter [3], the retinoic acid receptor-1 promoter [4], the TGF- promoter [5], and the SF-1 response element [6], are activated by ligand-bound estrogen receptors. A panoply of transcription modulators (both coactivators and corepressors) have recently been described, as reviewed by McKenna et al. [7]. Given this, and also that a variety of transcriptional response element pathways for estrogen action are known, it is possible to envi-

FIGURE 1 Schematic representations of the crystal structure of estrogen receptor alpha highlighting the molecular anatomy of SERM antagonist action. The conformation of the receptor’s ligand-binding domain is shown in the presence of estradiol (left) and raloxifene (right). Ligands are drawn in space-filling form and colored light blue. The red and green cylindrical elements schematically represent the alpha-helices of the receptor. The majority of the structure (helices 1–11; red cylinders) adopts an identical conformation regardless of the bound ligand. When estradiol binds to the receptor, helix 12 (green cylinder) adopts a position across the entrance of the ligand binding pocket, sealing the ligand within. In this orientation, helix 12 completes the formation of the receptor’s AF-2 transactivation function region (area between helices 3, 4, 5, and 12) and enables recruitment of coactivator proteins. Interaction between receptor and coactivator is mediated by the coactivator’s LxxLL interaction module, shown here as a purple helix, which binds along AF-2. In contrast, when raloxifene is bound, helix 12 shifts rightward to block the AF-2 site, preventing both coactivator binding and subsequent gene transcription. Printed with permission from Ashley C. W. Pike, University of York. (See also color plate.)

CHAPTER 70

605

Selective Estrogen Receptor Modulators (SERMS)

a groove that is formed along one side of the receptor. This groove constitutes the binding site for coactivator proteins [10] and is the mediator of transactivation function 2 (AF2) activity. The ligand-dependence of the conformation of helix 12 and its relationship to the AF-2 region are shown in Fig. 1. The AF-2 region is a conserved sequence common to many members of the nuclear receptor family and it is important for modulation of ligand-dependent gene transcription [11]. Mutagenesis experiments have demonstrated that modifications of the estrogen receptor in this region can convert agonists to antagonists, and vice versa [12]. Given the role of AF-2 in facilitation of transcriptional responses to ligand binding, raloxifene’s effect on AF-2 accessibility would be expected to be responsible for its estrogen antagonist action, at least insofar as AF-2-dependent transcriptional activation is concerned. Further support for these inferences from crystallographic studies and transcriptional assays regarding the role of the C-terminal -helix of the estrogen receptor protein in mediating differential effects of agonists and antagonists is derived from experiments which have subjected the C-terminal -helix of the ER to mutational disruptions. Using a fusion protein assay system, Nichols and colleagues [13] tested the influence of C-terminal mutations to modify the ability of the estrogen receptor to interact with agonists and antagonists. The assay uses ligands to bind to and modify the function of a fusion protein. The fusion protein consists of a recombinase component coupled to selected components of the estrogen receptor. After ligand binds to the estrogen receptor portion of the fusion protein, the recombinase portion of the protein is activated to mediate a recombination event. This event results in a change in phenotype of the yeast, which have been engineered to signal the recombination by changing colony color. This system thus allows insight into early functional consequences of agonist or antagonist interactions with receptors at a step that is proximal to the later complex interactions of the receptor with transcription factors and promoter sequences. When ligands are bound to a fusion protein derived from wild-type estrogen receptor sequence, which includes the D, E, and F domains of the receptor, both agonists and antagonists activate the recombinase. This is consistent with the fact that agonists and antagonists bind in the same ligand binding pocket of the receptor, and it implies that both ligand binding and initial functional changes following ligand binding occur similarly for both antagonists and agonists. However, agonist and antagonist activity could be discriminated in this system by using fusion proteins that included deleted or mutated sequences of the estrogen receptor. Thus, deletion of the D “spacer” domain from the fusion protein resulted in activation of recombinase by agonists but not by antagonists, whereas deletion of all but the E domain restored activation by antagonists. Experiments

in which the F domain was subjected to mutation implicated the conformational positioning of this region of the receptor as a critical component in mediating differential effects of agonists and antagonists [13]. This observation provided confirmation in a functional assay of the regulatory importance of the C-terminal F domain as inferred by crystallographic studies [9]. Other portions of the estrogen receptor apart from the AF-2 region also play key roles in regulating antagonist actions of SERMs. Substitution of tyrosine for aspartate at residue 351 of the E domain (a portion of the ligand binding pocket) results in conversion of raloxifene from an antagonist to an agonist [14]. Interestingly, this mutational substitution has been observed to occur under conditions whereby tumors have become dependent upon tamoxifen for growth [15]. Not surprisingly, the aspartate 351 residue is seen in crystallographic studies to be involved with coordination of the basic side chain of raloxifene to the receptor [9], lending structural credence to the functional observation. From the above discussion it is clear that the C-terminal -helix of the receptor plays a key role in mediating estrogen antagonist activity of SERMs. When raloxifene binds to the receptor it assumes a conformation in which the AF-2 site is blocked from interacting with transcriptional copromoters. Since raloxifene and estradiol compete for highaffinity binding to the same occupancy site, raloxifene thus acts as a competitive antagonist for those estrogen actions that depend on transcriptional activation of the classical estrogen pathway. Estrogen receptors depend on dimerization after ligand binding in order for subsequent gene activation of classical estrogen targets. This provides another potential site for SERM action. Tamoxifen results in destabilization of estrogen receptor dimers, and this interferes with estrogenstimulated gene transcription in a yeast model [16].

D. Agonist Mechanisms of SERMs In addition to its modulating transcription of its classical target genes via the ERE promoter, estrogen influences transcription of a number of alternate pathways. These include transforming growth factor- [5], the fos – jun oncogene promoter known as AP-1 [3], the retinoic acid pathway [4], and the SF-1 response element [6]. TGF-, which is known to have antiresorptive activity in bone, is stimulated in vivo by estrogen, making it a candidate gene for estrogen action in bone [17]. For this reason, Yang et al. [5] investigated the molecular regulation of TGF- transcription. Of particular interest to these investigators was whether or not SERMs such as raloxifene, which were known to mimic estrogen action in bone, shared estrogen’s ability to activate TGF-3 transcription.

606

SIRIS AND MUCHMORE

Using an in vitro transient transfection system, they demonstrated that raloxifene, in the presence of estrogen receptors, could activate TGF-3 transcription [5]. Interestingly, although estradiol itself had little agonist activity, a variety of endogenous estrogen metabolites were active. These investigators then studied raloxifene’s agonist potential against a panel of mutant estrogen receptors. They found that agonist actions of raloxifene occurred even when they used estrogen receptor mutants which lack the DNA binding domain that is necessary for the activation of transcription of ERE dependent genes. This suggested that a non-ERE promoter sequence may be involved in mediating this estrogenic action of raloxifene. Deletion of various other estrogen receptor domains had little impact on raloxifene’s agonist activity in this system. In fact, deletion of the ligand binding domain of the receptor was required for elimination of agonist activity. These results are consistent with the conceptualization that raloxifene, acting by binding to the ligand binding domain of the estrogen receptor, activates alternate gene pathways such as TGF-3 through molecular mechanisms that are distinct from those involved in the actions of estradiol to activate classical estrogen target genes. The fact that endogenous estrogen metabolites activate these same pathways underscores the potential physiologic relevance of these alternate response pathways. Further, the lack of activity of estradiol itself points to additional potential control steps in the regulation of estrogen effects on bone in vivo in that these pathways may be regulated by the production and distribution of various estradiol metabolites. Apart from the existence of non-ERE pathways that respond differently to different ligands, there can also be heterogeneity in responses of ERE genes to different ligands. Thus tamoxifen and estradiol are equally effective agonists in a CAT reporter system when the reporter is constructed using the -globulin promoter whereas tamoxifen is ineffective when the reporter is constructed using a thymidylate kinase promoter. On the other hand, raloxifene is essentially without agonist activity using either promoter construct [18]. These results suggest that promoter context may modulate specific ligand responsiveness, thus offering another mechanism of SERM agonist specificity.

E. Role of Other Accessory Factors Integral to the process of transcriptional stimulation by liganded estrogen receptor is modulation of the process by accessory proteins. The roles of coactivators and corepressors in modulating steroid receptor activities has become a field of accelerating discovery. Many such modifiers of transcriptional activation have been identified, and recent work shows how some of these act at the molecular level. For instance, the steroid receptor coactivator-1 (SRC-1) family shares a common pentapeptide sequence (LxxLL)

which appears to be a key component in the interaction of these factors with the AF-2 region [7]. Recent work by Paige et al. [19] has demonstrated that peptides containing the LxxLL motif display different estrogen receptor binding affinity patterns depending on which ligand (various estrogens or SERMs) are bound to the receptor. By testing a panel of different peptides for binding affinity to liganded estrogen receptors, these workers demonstrated a variety of “fingerprint” patterns of peptide binding affinity. This provided indirect confirmation of the hypothesis presented by McDonnell and colleagues [20] that different SERMs act in divergent ways as a result of ligand-specific differences in estrogen receptor conformation. The functional importance of this was further underscored when these same investigators demonstrated that estrogen-responsive cells which are cotransfected with expression vectors that code for peptides containing the LxxLL sequence are able to modulate estrogen and SERMmediated transcription [21]. As shown in Fig. 1, peptides containing the LxxLL motif bind to the AF-2 region of the receptor, and this provides a direct structural explanation for the functional interaction between LxxLL coactivators (such as SRC) and the ligandbound estrogen receptor. Nuclear receptor corepressors were initially described for nonestrogenic receptors such as thyroid, retinoic acid, and progesterone receptors [reviewed in 7 and 22], but corepressors of estrogen action have now also been identified with certainty. One such factor is repressor of estrogen receptor activity (REA), and this 37-kDa protein potentiates the action of SERMs while inhibiting estrogen mediated transcription [23]. As with coactivators, the roles that these factors play in modulating tissue- and pathway-specific actions of SERMs is largely speculative, but information on the in vivo relevance of these potential control mechanisms will no doubt be forthcoming. Further complexity of estrogen response systems is exemplified by the interactions between protein kinases and estrogen receptors. Mitogen-activated protein kinase (MAPK) can serine phosphorylate the estrogen receptor in the AF-1 region [24]. This phosphorylation in turn modulates both estrogenand SERM-mediated transcription. Recent evidence shows that AF-1 phosphorylation aids in the recruitment of SRC-1 binding to ER, and this occurs independent of ligand binding [25]. The extent to which this mechanism may modulate SERM activity is speculative, but it underscores the convoluted interactions that may occur in these systems.

F. Role of Estrogen Receptor Isoforms It is now recognized that estrogen receptors are present in at least two major homologous isoforms, ER and ER (see Chapter 10). Recently, a subtype of ER was further

CHAPTER 70

607

Selective Estrogen Receptor Modulators (SERMS)

described [26]. The major receptor isoforms are distributed differently among tissues [27,28], providing a potential means for tissue selectivity of estrogen action. Indeed, considerable evidence supports differential tissue- and ligand-specific actions of the different receptor isoforms. At the whole animal level, ER knockout mice exhibit skeletal abnormalities which include decrease in appendicular growth but increase in axial growth, effects which are associated with nearly 25% reductions in IGF-I concentrations [29]. ER knockout mice demonstrate enhanced cortical bone radial growth and increased cortical bone mineral content [30]. Although perhaps not fully indicative of estrogen receptor isoform effects in bone in other species, these findings certainly support differential target tissue effects of the different isoforms in vivo. At the level of ligand-specific receptor isoform effects, the ability of raloxifene, estradiol, or tamoxifen to stimulate transcription at the AP-1 promoter site is highly dependent upon the receptor isoform used in the assay [3]. Thus, estradiol and tamoxifen exhibit strong estrogen agonism at the AP-1 site when ER (but not ER) is used in a cotransfection assay in which estrogen receptor and an AP-l reporter construct are employed. However, raloxifene is nonstimulatory when ER is employed although when ER is employed, raloxifene, and the antiestrogen ICI 164381 are strongly stimulatory. Experiments using estrogen receptor chimeras (in which a portion of the ER receptor is engineered to contain a portion of the ER sequence and vice versa) have shed further light on the specificity of ligand-induced transcriptional responses. This approach has implicated the AF-1 region (i.e., A/B domains) of the estrogen receptor to be important in both promoter- and ligand-specific response differences between the two major receptor isoforms [31]. Not only do the estrogen receptor isoforms modulate different effects with different ligands in different target systems but they also directly modulate each other’s actions. Under conditions of subsaturating hormone concentrations, coexpression of ER results in a reduction in ER-mediated estrogen response in a cotransfection assay [32]. Whether or not this effect depends upon the known ability of ER to dimerize with ER [33] is unknown. Regardless of the mechanism, ER is a transdominant inhibitor of ER activity, and the relative expression of the two receptors in a single cell may be important in determining the extent of response to ligand. The clinical utility of compounds that selectively target one or another of the estrogen receptor isoforms is not known but an expected outcome of further investigations of this type would be identification of new SERMs with distinctive pharmacological profiles. Indeed, Sun and colleagues [34] have described a family of ER isoform-specific compounds which selectively stimulate transcription when ER is cotransfected into a cell system, whereas these compounds are devoid of activity when ER is used.

G. Additional Considerations As the field of estrogen receptor biology becomes more complex additional interactions need to be considered in the mechanisms of SERMs. For instance, the orphan receptor known as short heterodimer partner (SHP) has been known to interact with and inhibit transcriptional activation of nonsteroidal pathways (e.g., thyroid hormone receptor). Recently it has also been shown that SHP acts as a negative coregulator for estrogenic pathways, including both ERand ER-mediated transcriptional activity [35]. This appears to occur via blockade of the AF-2 region. Although SHP has a ligand-binding domain, the role of ligand in modulating its effects is unknown. Likewise, the influence that SHP may have on SERM action (agonist or antagonist) is unknown, but it is certainly conceivable that one may modulate activity of the other.

H. Summary of SERM Mechanisms The number of potential mechanisms that may explain the tissue-selective actions of SERMs is growing steadily as new findings are uncovered. Central to all of these mechanisms are the differential effects that SERMs exert on the three-dimensional conformation of the ligand-bound receptor. The specifics of the roles of the various domains of the receptor are beginning to unfold. In particular, the molecular mechanisms underlying the function of the AF-2 region have become increasingly clear with publication of X-ray crystallographic depictions of the receptor bound to various ligands and peptide fragments of coactivator proteins. Emerging understanding of the tissue distribution and functional biology of the estrogen receptor isoforms has filled in even more of the complex picture. These evolving insights should provide the platform for even better understanding of estrogen and SERM action while providing the basis for continued rational drug discovery.

III. CLINICAL APPLICATION OF SERMS A. Introduction From the clinical perspective, the ideal SERM would offer the postmenopausal woman all the benefits of estrogen replacement without the potential estrogen-mediated adverse effects: it would relieve menopausal symptoms, including vaginal dryness and hot flashes, preserve bone and reduce the risk of osteoporotic fractures, reduce the risk of coronary heart disease, cause no vaginal bleeding with no increase in risk of endometrial hyperplasia or carcinoma, produce no mastalgia, and afford protection against the

608

SIRIS AND MUCHMORE

development of breast cancer. Clearly, the ideal SERM does not exist as yet. In the sections that follow, the data are reviewed with respect to many of these issues for tamoxifene and raloxifene and, to a more limited degree, toremifene.

B. Tamoxifen 1. BACKGROUND Tamoxifen was the first SERM to be used clinically in large numbers of women. It is a triphenylethylene derivative (Fig. 2) developed in England in 1966 [36], and it has been widely used in the treatment of breast cancer with more than 6 million patient years of exposure. Originally called an “antiestrogen,” tamoxifen has only recently been recognized as a SERM. The initial development of tamoxifen was as an antifertility agent, but it had the paradoxical effect of inducing ovulation in infertile women [37,38]. Its potential as a breast cancer treatment became apparent after animal studies showed that tamoxifen had striking antitumor activity in carcinogen-induced mammary tumor models in the rat [39,40], and by 1971 the first clinical trials reporting its benefit in women with metastatic breast cancers were published [41]. Tamoxifen was shown to exhibit a high antitumor potency with a low adverse effects profile, and it was both better toler-

FIGURE 2

ated and more effective than two of the standard hormonal treatments for advanced breast cancer in use at the time, high-dose estrogen or androgen therapy. Tamoxifen was approved for the treatment of breast cancer in the United Kingdom in 1973 and for the treatment of advanced breast cancer in postmenopausal women in the United States in 1977. Thereafter it also became a standard initial endocrine therapy for estrogen receptor positive (ER  ) disseminated breast cancer in premenopausal women as well as for metastatic breast cancer in men. In 1998 the findings of the NSABP P-1 Breast Cancer Prevention Trial [42] led to the approval of tamoxifen in the United States for the reduction in incidence of breast cancer in high-risk women. In this section, the SERM profile of tamoxifen is considered with respect to its effects on bone in animal models and in human subjects. In terms of tamoxifen’s skeletal effects in patients, the preponderance of the information comes from studies in women either with breast cancer or at risk of developing breast cancer and includes data from both pre- and postmenopausal subjects. Effects of tamoxifen on bone mineral density (BMD) and fracture reduction are reviewed. Additionally, this section briefly considers the effects of tamoxifen on other organ systems including the cardiovascular system, breast, and uterus and discusses the current adverse events profile of this widely used anti-cancer agent.

Structures of selected SERMs.

CHAPTER 70

Selective Estrogen Receptor Modulators (SERMS)

2. ANIMAL STUDIES IN BONE Initial studies indicated that in the rat, tamoxifen inhibited PTH-induced bone resorption in fetal long bones [43], suggesting an estrogen agonist effect. However, conflicting reports followed. A reduction in bone mass as measured by femur ash weight and X-ray densitometry was found in intact rats given high doses of tamoxifen and was comparable to that seen in untreated but ovariectomized rats [44]. Conversely, tamoxifen was shown by others to maintain bone mass in both mature ovariectomized and young growing rats [45]. Further conflicting results emerged from studies of bone remodeling: in one study tamoxifen failed to prevent calcitriol-stimulated bone resorption in vivo in rats [46], while another early study demonstrated tamoxifen-induced inhibition of bone resorption in the ovariectomized rat, preventing the increased osteoclast number and resorbing surface length characteristic of estrogen deficiency [47]. Subsequent studies illustrated the potential antiresorptive effects of tamoxifen, but further trials would be needed to support the emerging hypothesis that tamoxifen acted as an estrogen agonist at bone, preventing the skeletal alterations resulting from ovarian hormone deficiency [43,48]. 3. HUMAN STUDIES OF TAMOXIFEN a. Bone Mineral Density, Bone Metabolism, and Bone Turnover Markers As tamoxifen gained wider application in the treatment of women with breast cancer, and particularly in the setting of the early and conflicting data from animal studies in the 1980s, concern arose that women receiving this agent for breast disease might be subject to increased bone loss. Several early studies, primarily crosssectional uncontrolled retrospective or small prospective analyses generally suggested no negative effect and a positive effect on bone mass [49–56]. A prospective, randomized, placebo-controlled 2-year-long clinical trial of 140 postmenopausal women with node negative breast cancer reported a significant improvement in spinal BMD measured by dual photon absorptiometry with tamoxifen, 10 mg twice daily, versus placebo [57]. Lumbar spine BMD increased by 0.61% per year with tamoxifen but decreased 1.00% per year in the placebo group (P 0.001). Two markers of bone turnover, serum osteocalcin and total alkaline phosphatase, both decreased significantly, in response to tamoxifen. A subsequent report from this same cohort after a total of 5 years of tamoxifen described the persistence of the lumbar spine BMD effect in the 62 patients who remained in the study on the agent to which they were initially randomized: lumbar spine BMD was increased by 0.8% above baseline with tamoxifen and decreased by 0.7% with placebo (P 0.06) [58]. Lower levels of serum turnover indices in the tamoxifen group also persisted [58].

609 Further support for the beneficial effects of tamoxifen on the bone of postmenopausal women with breast cancer came from a study utilizing histomorphometry [59]. Forty-one women underwent transiliac bone biopsy; 21 of these patients had received a minimum of 15 months of tamoxifen and 19 were untreated. A significantly lower tissue-based formation rate and a longer remodeling period were noted in the treated women. Tamoxifen patients had reduced mean and maximum resorption cavity depth compared with untreated patients. The data indicated a trend toward greater trabecular connectivity in the tamoxifen-treated women, although overall the calculated and directly measured indices of cancellous structure were similar in the two groups. The effect of tamoxifen on BMD and bone metabolism has been studied prospectively in postmenopausal women without breast cancer [60,61]. In one study 57 women were randomly assigned to 20 mg tamoxifen or placebo daily for 2 years. Mean BMD of the lumbar spine increased by 1.4% with tamoxifen and decreased 0.7% with placebo (P 0.1); the tamoxifen effect was maximal after 1 year with no separation of the groups after this. No significant group BMD differences were observed at the proximal femur. Reductions in serum turnover markers were shown with tamoxifen, however, with significant decreases in serum alkaline phosphatase and in urinary hydroxyproline, N-telopeptide, and calcium excretion, all consistent with results from studies of tamoxifen in postmenopausal women with breast cancer [60]. In a recent breast cancer chemoprevention trial, 54 healthy postmenpausal women were randomized to 20 mg/day tamoxifen or placebo for 3 years; tamoxifentreated women experienced 2 – 3% improvements in BMD at spine and hip by 3 years compared to small losses in women receiving placebo (P 0.002, spine; P 0.05, hip) [61]. Another group of 38 postmenopausal women without breast cancer was evaluated in a study comparing tamoxifen alone to tamoxifen in combination with estrogen. Use of tamoxifen led to annual increases in spine and hip BMD of 1.5 and 2.0%, respectively, compared with placebo. When hormone replacement therapy was added to tamoxifen, an additional 2.0% annual increase in hip BMD but no change in spine BMD occurred [62]. The effect of tamoxifen on BMD in premenopausal women has been studied in women with and without breast cancer. In an uncontrolled study of premenopausal breast cancer patients, loss of BMD at radius, spine and hip was observed [63]. In the chemoprevention trial mentioned above [61], 125 healthy premenopausal women also were randomized to either tamoxifen or placebo for 3 years. Both lumbar spine and hip BMD decreased in the tamoxifen subjects and remained stable or increased in the placebo patients. The mean annual loss at the lumbar spine in the tamoxifen-treated women was 1.44%, compared with an annual gain of 0.24% in the placebo-treated women. The general findings from the chemoprevention trials in healthy

610

SIRIS AND MUCHMORE

women indicate that tamoxifen use is associated with prevention of bone loss or small gains in BMD in postmenopausal women and with bone loss in premenopausal women. These results support the concept that tamoxifen has an estrogen agonist effect on bone in settings where circulating estrogen concentrations are low, but estrogen antagonist effects when endogenous levels of estrogen are high. A potential positive BMD effect from combined tamoxifen and estrogen in healthy postmenopausal women [62] is therefore puzzling and requires additional study. b. Effects of Tamoxifen on Fracture Risk Two studies have examined the effect of tamoxifen on fracture risk. The first of these is the Danish Breast Cancer Cooperative Group [64], in which 1716 high risk women were randomized to no treatment or to radiation therapy and tamoxifen, 30 mg per day. The study examined the occurrence of hip fractures during the first year of treatment; there were no data obtained on vertebral deformities or other nonvertebral fractures. Although 51 control subjects and 64 tamoxifen subjects reportedly sustained a femoral fracture during the year of observation, a nonsignificant difference, it was found that 27 tamoxifen patients and 11 control patients sustained their hip fractures in the trochanteric region, with a relative risk of 2.12, 95% CI 1.12 – 4.01, for tamoxifentreated patients. The second study to evaluate fracture risk is the NSABP P-1 Trial, the Breast Cancer Prevention Trial [42]. This very large clinical trial enrolled 13,388 pre- and postmenopausal women at high risk for developing breast cancer, randomizing the subjects to either placebo (n  6707) or 20 mg/day tamoxifen (n  6681). A secondary aim of this study was to assess the effects of tamoxifen on fracture risk. During the average 36-month follow-up period (this 5year study was terminated early because of conclusive reTABLE 1

c. Effects of Tamoxifen on Surrogate Markers for Coronary Heart Disease and Risk of Ischemic Heart Disease Events Most of the studies on surrogate markers come from large clinical trials of women with breast cancer. Tamoxifen has been characterized as an anti-oxidant [65], and its use has been associated with reductions in total cholesterol of 7 – 17% and in low-density lipoprotein (LDL) cholesterol of 6 – 29% [66 – 69]. In most instances the effect on high-density lipoprotein (HDL) cholesterol has been neutral [68,70,71]. Reductions in plasma fibrinogen [72] and in lipoprotein “Little a” [Lp(a)] [70] and either no effect [67] or increases in circulating triglycerides [73] have been reported. With the exception of the possible increase in triglycerides, all of the other effects on surrogate markers above are either neutral or beneficial. Information on the effect of tamoxifen on myocardial infarction (MI) and other coronary events comes from three

Annual Rates for Fracture Events among Participants in the P-1 Trial [42] No. of events

Site of fracture

sults showing tamoxifen reduced the incidence of new cancers, see below) 955 women sustained fractures, including 483 in the placebo group and 472 in the tamoxifen group. When the types of fractures most likely to be the result of osteoporosis — combined hip, lower radius, and clinically symptomatic spine fractures — were counted there was a 19% reduction in fractures in the group receiving tamoxifen: 111 fractures occurred in the tamoxifen subjects and 137 in the placebo subjects, RR  0.81, 95% CI  0.63 – 1.05, a reduction that closely approaches statistical significance. Table 1 shows the results for hip, spine, Colles’, and other lower radial fractures. The overall reduction was greater in the group older than 50 years at entry, suggesting that the inclusion of premenopausal subjects who were most likely to lose bone on tamoxifen might have caused an underestimate of any beneficial effect of tamoxifen on fracture risk.

Rate per 1000 women

Placebo

Tamoxifen

Placebo

Tamoxifen

Risk ratio

95% Confidence interval

Hip

22

12

0.84

0.46

0.55

0.25 – 1.15

Spine

31

23

1.18

0.88

0.74

0.41 – 1.32

Radius, Colles’

23

14

0.88

0.54

0.61

0.29 – 1.23 0.73 – 1.51

Other lower radiusa

63

66

2.41

2.54

1.05

137b

111c

5.28

4.29

0.81

0.63 – 1.05

 49 years of age at entry

23

20

2.24

1.98

0.88

0.46 – 1.68

 50 years of age at entry

114

91

7.27

5.76

0.79

0.60 – 1.05

Total

a

Excludes women who had a Colles’ fracture. One woman had a hip fracture and a Colles’ fracture and one woman had a hip fracture and another lower radial fracture c One woman had a hip fracture and a Colles’ fracture, and one woman had a hip fracture and a spine fracture, and two women had hip fractures and other lower radial fractures. Note: Reprinted with permission from [42]. b

CHAPTER 70

Selective Estrogen Receptor Modulators (SERMS)

major breast cancer studies, the Scottish Cancer Trials Breast Group [74], the Stockholm Breast Cancer Study Group [75], and the NSABP B-14 trial [76], as well as from the NSABP P-1 prevention trial [42]. In the Scottish studies there was evidence of a significantly reduced risk of MI and a trend toward reduction of other ischemic events in tamoxifen “ever” versus “never” users [74]. The Swedish study reported a significant reduction in cardiac disease with tamoxifen [75]. The results from tamoxifen use in the NSABP breast cancer treatment study were similar to those from Europe with respect to reduced fatal heart disease, but the 34% reduction in risk did not reach statistical significance [76]. Finally, results from the P-1 prevention trial failed to show a reduced incidence of ischemic heart disease in that population, including women with and without heart disease, although there was also no increase in risk [42,42a]. Although suggestive, results to date do not clearly indicate a benefit for tamoxifen on cardiac outcomes. It is unclear whether tamoxifen affords a beneficial or detrimental effect on the incidence of cerebrovascular events. In the various breast cancer studies examining this question there does not appear to be a significant difference in the incidence of cerebrovascular disease between those receiving tamoxifen and those not [77]. d. Effects of Tamoxifen on the Breast In 1998, an overview analysis was published [78] summarizing the findings from data on 37,000 women enrolled in 55 randomized clinical trials evaluating the effects of tamoxifen in patients with breast cancer. This was a 10-year analysis that reviewed 87% of the worldwide evidence. The data included information on both pre- and postmenopausal women and included information on ER status. Among the findings was the recognition that both node negative and node positive breast cancer patients had the same proportional reductions in recurrence rates and mortality in response to tamoxifen. A second critical observation was the importance of duration of treatment: there was a clear superiority of 5 years of treatment with tamoxifen over shorter (1- or 2-year) periods in both pre- and postmenopausal women with ER or non-ER-poor tumors. Tamoxifen was found to be without benefit in ER and ER-poor tumors. A third finding was that 5 years of treatment with tamoxifen confers a 47% reduction in the occurrence of contralateral breast tumors, a complication for which women with one primary breast cancer are at increased risk. Five years of treatment was shown to continue to afford protection against contralateral disease for the 5 years following the completed treatment course. The data showing the benefit of tamoxifen in the prevention of contralateral tumors provided supportive evidence for the principle that tamoxifen might have benefit in reducing the incidence of breast cancer in healthy women at high risk for developing breast cancer. The NSABP P-1 trial [42] re-

611 vealed that tamoxifen reduced the risk of invasive breast cancers by 49% and noninvasive cancers by 50%; this magnitude of effect was similar in women less than 49 years of age, those 50 – 59, and those 60 years and older. There was no difference between placebo and tamoxifen in the incidence of ER tumors. As a consequence of these data, tamoxifen was recently approved by the US FDA for the reduction in the incidence of breast cancer in women at high risk. Two smaller studies, one from England and one from Italy, did not find a benefit from tamoxifen in prevention of breast cancer [79,80]. These studies had different populations, differences in age, reproductive status, and family history status (greater numbers of women with strongly positive family histories as was the case in the English study may increase the likelihood of detecting greater numbers of ER tumors), concomitant use of estrogen in some cases, and high drop-out rates in the Italian study. Such factors complicate direct comparisons with the US/Canadian study and suggest the need for further investigation in differing populations. e. Effects of Tamoxifen on the Uterus Tamoxifen increases the risk of uterine bleeding and of both benign and malignant disease of the uterus [42,77,81 – 83]. Fibroid tumors, adenomyosis, endometrial hyperplasia, and benign polyps are more common in women receiving tamoxifen. Among various studies of women with breast cancer or those in prevention studies, the relative risk of endometrial cancer ranged from 2.5 to 7.5 [77]. It is important to note that the various large trials do not provide any evidence that the uterine cancers that occur in the setting of tamoxifen use are any different from those that occur in women on estrogen replacement therapy or on no added treatment; i.e., they are not of a higher grade of malignancy. In the P-1 trial, with a relative risk of uterine cancer of 2.53 with tamoxifen, the malignancies were all Stage I [42]. This risk was restricted to women over age 50, suggesting that it is operative only in postmenopausal women. f. Other Effects of Tamoxifen Tamoxifen increases the occurrence of hot flashes in both pre- and postmenopausal women. Its effects on cognitive function are not known. Several large breast cancer studies have reported an increased risk of venous thromboembolic disease with tamoxifen [84,85]. Up to a 3.0-fold increased risk of pulmomary embolus and 1.6-fold increased risk of deep vein thrombosis was reported in the U.S. breast cancer prevention trial [42]. There also appears to be increased risk (RR 1.14, 95% CI 1.01 – 1.29) of cataract development with tamoxifen [42], but the preponderence of the evidence does not support an increase in macular degeneration or retinal toxicity as was initially thought [77].

612

SIRIS AND MUCHMORE

C. Raloxifene 1. BACKGROUND Raloxifene is a benzothiophene derivative (Fig. 2). Preclinical models demonstrated its capacity to inhibit the binding of estradiol to the estrogen receptor and to block the proliferation of MCF-7 breast cancer cells, an estradioldependent process [86]. Other studies with raloxifene performed in rodents showed inhibitory effects on carcinogeninduced mammary tumors similar to those seen with tamoxifen [87]. However, raloxifene did not demonstrate an anti-tumor effect in a small number of women with advanced breast cancer [88]. Other preclinical studies with raloxifene confirmed that the agent exhibited many of the desirable properties of a SERM: it inhibited bone loss, lowered cholesterol, caused no breast stimulation, and —in distinct contrast to tamoxifen —produced no stimulation of uterine endometrial tissue. As discussed below, studies in postmenopausal women have shown the ability of raloxifene to preserve bone mass in the prevention of osteoporosis and to reduce the risk of vertebral fractures in established osteoporosis. These and other studies also evaluated the effects of raloxifene on surrogate markers of coronary disease and its safety profile at the uterus and breast. The studies on bone resulted in the approval of raloxifene for the prevention of osteoporosis in 1997 and for the treatment of postmenopausal osteoporosis in 1999. Large clinical trials investigating raloxifene effects on nonbone endpoints have been initiated. The first, called the STAR trial (Study of Tamoxifen and Raloxifene), compares raloxifene with tamoxifen for the incidence of breast cancer in women at high risk. The second, designated RUTH (Raloxifene Use for the Heart), compares raloxifene with placebo to evaluate the effects on cardiovascular outcomes in women either currently with or at high risk for coronary heart disease. Both of these studies are ongoing. In the sections that follow, animal studies and the various clinical studies in postmenopausal women that describe the effects of raloxifene on BMD and inhibition of bone loss, on bone turnover markers, and on fracture incidence are reviewed. Effects of raloxifene on surrogate markers for cardiovascular disease and on the breast and uterus are discussed, as is the adverse events profile of this relatively new osteoporosis therapy. 2. ANIMAL STUDIES OF RALOXIFENE IN BONE A series of studies have examined the effects of raloxifene on bone mass, quality, and architecture in ovariectomized rats. Studies extending over 12-month periods showed raloxifene to be comparable to both tamoxifen and estrogen [89] in the inhibition of cancellous bone loss as

measured with bone densitometry [89,90] and histomorphometric volume analysis [91]. Two measures of bone turnover, urinary collagen cross-links and serum osteocalcin, were both reduced with raloxifene to sham-operated levels comparable to those seen with estrogen [92]. Both raloxifene and estrogen inhibited the increases in osteoclast number, eroded perimeter, trabecular separation and bone turnover that occur shortly following ovariectomy, as demonstrated by histomorphometric analysis of the rat tibia; although recently ovariectomized animals treated with estrogen exhibited a substantially reduced cancellous bone formation rate, raloxifene-treated animals did not [93]. In other studies, however, addition of raloxifene following established bone loss produced similar reductions in both bone resorption and formation to those found with estrogen. Biomechanical strength was evaluated in lumbar vertebrae and femurs of ovariectomized rats and raloxifene proved similar to estrogen in preserving both strength and mechanical integrity [94]. Finally, raloxifene inhibited bone loss in female rats treated with LH-RH agonists [95] and increased bone mass and reduced bone turnover in ovariectomized cynomologous monkeys [96]. 3. HUMAN STUDIES OF RALOXIFENE a. Effects of Raloxifene on Bone Mineral Density, Bone Metabolism, and Bone Turnover Markers Human studies with raloxifene have examined its effects on bone remodeling, BMD, bone turnover markers and reductions in vertebral fracture risk in postmenopausal women. Bone remodeling in early postmenopausal women was examined using calcium tracer kinetics under constant diet and metabolic balance conditions [97]. Subjects received 60 mg of raloxifene, cyclic hormone replacement therapy, or no treatment, and were evaluated at baseline, 4 weeks, and 31 weeks to assess both early and later changes in remodeling. Estrogen and raloxifene each led to significant positive shifts in calcium balance and significant decreases in bone resorption at both 4 and 31 weeks, but the effect of estrogen on bone resorption was greater than that of raloxifene at 31 weeks. Neither agent changed bone formation at 4 weeks, while at 31 weeks estrogen, but not raloxifene, reduced formation. At 31 weeks, therefore, standard estrogen doses suppressed bone remodeling to a greater extent than 60 mg raloxifene, although the remodeling balance was the same for the two agents. Studies evaluating the effect of raloxifene on BMD and bone turnover markers have been performed in postmenopausal women with and without osteoporosis, including osteoporotic women with and without prior vertebral fractures. The first large clinical trial for the prevention of osteoporosis compared the effects of placebo or raloxifene 30, 60 or 150 mg daily for 24 months on BMD at the lumbar spine, total hip, and total body [98]. Subjects were 601

CHAPTER 70

Selective Estrogen Receptor Modulators (SERMS)

European postmenopausal women with baseline lumbar spine BMD between 2.0 SD above and 2.5 SD below the mean value in young normal women. These women received supplemental calcium, 400 – 600 mg per day, as part of the study protocol. All doses of raloxifene produced small but significant increases in BMD at all three sites in contrast to losses at all sites with placebo. At 24 months the differences in BMD between raloxifene 60 mg and placebo were 2.4% at the lumbar spine, 2.4% at the total hip, and 2.0% at the total body (P  0.001 for all comparisons). In this study each of the raloxifene doses also significantly reduced markers of bone turnover; at month 24 the 60 mg raloxifene group had decreases in the median values of serum bone specific alkaline phosphatase, serum osteocalcin and urinary C-telopeptide of type 1 collagen of 15, 23, and 34%, respectively. These changes represented reductions from mean baseline values seen in postmenopausal women to mean values typical of premenopausal women. A smaller 6-month prevention study with 51 older postmenopausal women (mean of 18 years postmenopause) evaluated BMD at the lumbar spine and hip comparing raloxifene 60 mg, conjugated estrogen 0.625 mg, and placebo [99]. Both raloxifene and estrogen led to significant increases in BMD at the lumbar spine at 6 months, but the increment with raloxifene was significantly less than that with estrogen (1.3% vs 3.2%, P  0.029). Raloxifene significantly increased the BMD of the femoral neck (2.8%), compared with a non-significant increase with estrogen (1.6%) but the effects of the two drugs at the hip did not differ significantly from each other in this small study. A study of 129 postmenopausal French women (mean age 60 years) with low bone density or osteoporosis (mean lumbar spine BMD T score 2.8) compared the effects of daily raloxifene 60 or 150 mg or placebo for 24 months on bone density at the lumbar spine and hip as well as on markers of bone turnover [100]. All patients received daily supplemental calcium 1000 mg as well as 300 IU vitamin D3. At 24 months the group receiving 60 mg of raloxifene had significant increases in BMD at the spine (3.2%), trochanter (2.7%), femoral neck (2.1%), and total hip (1.6%) compared to the placebo group (P 0.05). The results in the women taking the 150 mg raloxifene dose were similar to those seen with 60 mg. Pooled raloxifene groups experienced a 39% reduction in the urinary type 1 collagen/creatinine ratio compared to placebo (P 0.01), and reductions of 19 and 26%, respectively, in serum bone-specific alkaline phosphatase and osteocalcin (P 0.001), at 24 months. The markers had decreased significantly by 3 – 6 months of raloxifene therapy and remained at the lowered levels for the duration of the study. A shorter duration 1-year study of 143 postmenopausal osteoporotic women living in the United States, mean age 68 years, similarly evaluated the bone density and bone

613 turnover marker changes associated with daily administration of 750 mg calcium and 400 IU vitamin D alone or 60 or 120 mg of raloxifene [101]. To be enrolled in this study subjects had to have a baseline BMD at the lumbar spine or proximal femur of less than or equal to the tenth percentile for premenopausal females and one or more vertebral defomities, defined as a decrease in vertebral height greater than or equal to 15% compared with adjacent vertebrae. BMD increased significantly at the total hip (1.66%) with 60 mg of raloxifene and at the ultradistal radius with both 60 and 120 mg (2.92 and 2.50%, respectively) compared with placebo. Raloxifene 120 mg was associated with a nonsignificant trend toward increased BMD at the lumbar spine and hip. The changes in markers of bone turnover at the end of 12 months for raloxifene, 60 and 120 mg, respectively, were significant for serum bone-specific alkaline phosphatase, reductions of 14.9 and 8.87%; serum osteocalcin, 20.7 and 17.0%; and urinary C-telopeptide fragment of type 1 collagen/creatinine, 24.9 and 30.8%. Preliminary data have been reported comparing the effects of 1 year of treatment with placebo, raloxifene 60 mg, alendronate 10 mg or combined raloxifene and alendronate on lumbar spine and femoral neck BMD and on bone turnover markers in 330 postmenopausal women with a baseline femoral neck T score of less than or equal to 2 [102]. Lumbar spine BMD increased 0.06% with placebo, 2.06% with raloxifene, 4.31% with alendronate, and 5.26% with combined raloxifene and alendronate; femoral neck BMD changes in the same groups were 0.31, 1.71, 2.71, and 3.70%, respectively. All BMD changes with active treatment were significantly different from placebo, P 0.05. The increase in femoral neck BMD was greater in the combined treatment group than in the alendronate group. Bone markers included NTX, CTX, osteocalcin, and bone-specific alkaline phosphatase; all markers decreased significantly compared to placebo with all active treatments, with the magnitude of change being significantly greater for combined raloxifene and alendronate than for raloxifene alone, P 0.05. The Multiple Outcomes of Raloxifene Evaluation (MORE) study is the largest study to evaluate the effects of raloxifene on bone density, bone turnover markers, and vertebral fracture risk, involving 7705 women from 25 countries [103]. All subjects were at least 2 years postmenopause, mean age 67, divided into two groups of those with (about one-third of the total) and those without (about two-thirds of the total) a prevalent vertebral fracture at entry. All subjects without fractures had a BMD T score at the femoral neck or lumbar spine of that was 2.5 or lower. Low bone density or osteoporosis was present in all with a prior vertebral fracture. Within each group, patients were randomly assigned to daily raloxifene 60 or 120 mg or placebo, and the skeletal effects of these agents were evaluated at the end of 36 months of treatment. All subjects

614

FIGURE 3 Percent change in bone mineral density (BMD) at the lumbar spine and femoral neck in patients in the MORE study, comparing placebo with two doses of raloxifene. At both sites the two raloxifene doses led to increases in BMD that were significantly greater than those with placebo. Reprinted with permission from [103].

received 500 mg of calcium and between 400 and 600 IU of vitamin D during the study. Changes in bone mineral density are shown in Fig. 3. Compared with the placebo group, subjects receiving 60 mg of raloxifene had a 2.1 and 2.6% increase at the femoral neck and lumbar spine, respectively; those receiving 120 mg had increases of 2.4 and 2.7% at the femoral neck and lumbar spine, P  0.001 for all comparisons. The bone density in raloxifene groups peaked at the hip at 24 months and remained stable between 24 and 36 months at the lumbar spine. The median decreases in bone turnover markers at 36 months for the groups assigned to placebo, raloxifene 60 mg, and raloxifene 120 mg, respectively, were as follows: serum osteocalcin, 8.6, 26.3, and 31.1%; urinary C-telopeptide excretion, 8.1, 34.0, and 31.5% (P  0.001 for each raloxifene dose compared with placebo). b. Effects of Raloxifene on Vertebral Fracture Risk As noted above, the MORE study was designed to evaluate vertebral fracture risk as its primary end point [103]. The results depicting the reduction in new vertebral fractures with raloxifene 60 or 120 mg compared with placebo are shown in Fig. 4. Among the 6828 women who completed the MORE trial and had evaluable spine radiographs at 36 months, women receiving raloxifene had significantly fewer new vertebral fractures than women receiving placebo, regardless of raloxifene dose and regardless of whether or not the women had prevalent vertebral fractures at the onset of the study. In the group of women without prevalent vertebral fractures, 4.5% of placebo patients sustained a new vertebral fracture in contrast to 2.3% of raloxifene 60 mg patients and 2.8% of raloxifene 120 mg patients, RR, 0.5; 95% CI, 0.4–0.8 (60 mg), and CI, 0.4–0.9 (120 mg). Among the women who had experienced at least one vertebral fracture prior to entering the study, 21.2% of

SIRIS AND MUCHMORE

placebo treated patients had at least one new vertebral fracture compared with 14.7% of patients given 60 mg raloxifene and 10.7% of those receiving 120 mg of raloxifene. The reduction in risk of new vertebral fracture, RR and 95% CI, for the 60-mg patients was 0.7 (0.6–0.9) and for 120 mg patients 0.5 (0.4–0.7). Risk of nonvertebral fracture did not differ significantly between placebo and raloxifene groups: 9.3% of women on placebo and 8.5% of women on raloxifene sustained a nonvertebral fracture, RR, 0.9 CI, 0.8–1.1. Of note, 3.6% of placebo treated patients versus 1.1% of 60 mg and 0.9% of 120 mg raloxifene treated patients withdrew from the study due to multiple fractures or excessive loss of bone density during the study; it was suggested by the investigators that the loss of these higher risk patients because of vertebral fractures might have decreased the ability to detect a significant reduction in nonvertebral fracture risk. The 1-year study of U.S. women [101] also evaluated reduction in vertebral fracture risk in 143 women with prevalent vertebral fractures, randomized to calcium and vitamin D alone or raloxifene 60 or 120 mg. When the diagnosis of a new vertebral fracture was based on a radiographic cut point of a 30% reduction in vertebral height, there was a dose-related reduction in vertebral fracture for the raloxifene groups compared to the calcium/vitamin D group (P  0.047), but this effect was not seen if the cut point was a 15% decrease in height. c. Effects of Raloxifene on Surrogate Markers for Coronary Heart Disease and on Risk of Ischemic Heart Disease Events Several studies designed to examine the skeletal effects of raloxifene in postmenopasual women have also evaluated serum lipids and coagulation factors. In an early 8-week study of raloxifene 200 or 600 mg per day involving 251 women, serum LDL cholesterol decreased by 9.5

FIGURE 4

Reduction in new vertebral fractures among 6828 women completing the MORE study. RR indicates relative risk, numbers in parentheses the 95% confidence interval. Reprinted with permission from [103].

CHAPTER 70

Selective Estrogen Receptor Modulators (SERMS)

and 12.6%, respectively, compared to baseline; there was no increase in HDL [104]. The osteoporosis prevention study of 601 women [98] in which 60 or 150 mg of raloxifene were provided daily for 2 years reported significant median percentage reductions from baseline at 2 years for total cholesterol (6.4 and 9.7%, respectively) and LDL cholesterol (10.1 and 14.1%); neither HDL nor triglycerides increased with either dose of raloxifene. In the 1-year study in established osteoporosis [101], total cholesterol decreased significantly by 7% and LDL by 11.4% with 60 mg of raloxifene. Once again there was no change in either HDL or triglycerides. Finally, the 2-year study in France [100] reported significant reductions in both total and LDL cholesterol at study end with 60 mg of raloxifene (6.4% and 11.4%, respectively) and with 150 mg of raloxifene (7.2 and 13.9%, respectively). There was no increase in HDL, but the 60-mg group had an 11.55% decrease in triglycerides (P  0.054). The first large clinical trial designed specifically to study the effects of raloxifene on cardiovascular surrogate markers in healthy postmenopausal women was a 6 month study of 390 subjects, comparing placebo, raloxifene 60 or 120 mg, and conjugated estrogen 0.625 mg plus medroxyprogesterone 2.5 mg (HRT), all daily [105]. Both doses of raloxifene significantly reduced LDL cholesterol by 12% compared with placebo, similar to a 14% fall with HRT. HDL rose a significant 11% with HRT; it did not change with raloxifene. HDL-2 increased significantly by 15 and 17% with raloxifene 60 mg and 120 mg, compared with a significantly greater increase of 33% with HRT. A similar pattern was seen for Lp(a): there were significant 4% decreases with both doses of raloxifene, and a greater 16% reduction with HRT. Triglycerides rose significantly by 20% with HRT, an undesirable effect, and fell by 4% with 60 mg raloxifene. Finally, raloxifene significantly lowered fibrinogen (12 and 14%), in contrast to no effect with HRT, but it did not change the levels of plasminogen activator inhibitor-1, while HRT significantly decreased PAI-1 by 29%. A recent report indicated that raloxifene lowers homocysteine concentrations significantly and similarly to HRT, but unlike HRT it does not increase c-reactive protein levels [106]. The overall results from studies of these surrogate markers indicate that raloxifene is generally beneficial or neutral with respect to its effects; in some instances the magnitude of the positive effects is less that that with estrogen (e.g., HDL, HDL-2, Lp(a)), but in some instances it is similar to or greater than HRT (e.g., LDL, fibrinogen). The clinical impact of the effects of raloxifene on coronary events remains to be determined; the patients in MORE had too few cardiovascular events to determine any decrease or increase in coronary heart disease risk. It is anticipated that the RUTH study will provide more definitive information on the effects of raloxifene on the heart.

615 d. Effects of Raloxifene on the Breast Virtually all of the clinical trials in osteoporosis [98,100,101,103] and the study of cardiovascular surrogate markers [105] have shown no increase in mastalgia with raloxifene above that seen with placebo and a lack of breast abnormalities overall. The MORE study monitored patients for the occurrence of breast cancer as a secondary end point of that very large osteoporosis clinical trial. These data [107] indicated that within a median of 40 months of follow-up, 54 women from the initial enrollment of 7705 women were determined to have developed breast cancer. There were 13 cases of invasive breast cancer in the 5129 women in the raloxifene group and 27 cases in the 2576 women in the placebo group, so that raloxifene reduced the risk of invasive breast cancer by 76%: RR, 0.24, 95% CI, 0.13 – 0.44. The results were similar if all cases of breast cancer were counted (invasive, noninvasive, uncertain invasiveness): RR, 0.35, CI, 0.21 – 0.58, and were the same for both the 60- and 120-mg doses of raloxifene. Estrogen receptor status was available for 35 of the 40 invasive cases: 24 were ER and 11 were ER–. Raloxifene did not reduce the risk of ER cancers (RR, 0.88, CI, 0.26 – 3.00), but reduced the risk of ER cancers by 90% (RR, 0.10, CI, 0.04 – 0.24). At the end of 4 years the reduction in risk of ER invasive breast cancer was 84% (RR, 0.16, CI, 0.09, 0.30) [107a]. The duration of this reduction of risk is currently being evaluated by a 4-year extension of the MORE trial. As noted earlier, a comparison of breast cancer protective effects of both tamoxifen and raloxifene is underway in the STAR trial. e. Effects of Raloxifene on the Uterus The results of several clinical trials with raloxifene [98,101,105] were recently summarized [108] to review its effects on the uterus in 1157 postmenopausal women. At the 60-mg dose raloxifene was no different from placebo with respect to the incidence of vaginal bleeding, change in endometrial thickness measured by ultrasonography from baseline to endpoint in each trial, or the percentage of women having an increase in endometrial thickness above baseline after 12 or 24 months of treatment. No cases of endometrial hyperplasia or cancer were diagnosed in either the placebo or the raloxifen 60 mg/day groups. In the MORE trial [107] there were 5957 women with an intact uterus; there were 4 cases of endometrial cancer in women in the placebo group and 6 cases in the larger combined raloxifene group by 40 months of follow-up: RR, 0.8, 95% CI, 0.2 – 2.7. Among women who were examined by transvaginal ultrasound, there was a very small (0.3 mm) but statistically significant increase in endometrial thickness in raloxifene-treated versus placebo-treated women, as well as the finding of an endometrial thickness exceeding 5 mm in 4.1% more of those in the raloxifene group than in the placebo group. There was no difference in the incidence of endometrial hyperplasia between treatment groups. Fluid

616

SIRIS AND MUCHMORE

in the endometrial cavity, also seen with tamoxifen [109], was detected in 2.7% more raloxifene-treated women than placebo-treated women; this was viewed by the investigators as a benign finding. The investigators concluded that in the absence of any evidence that raloxifene use is associated with endometrial hyperplasia or cancer there is no indication for routine ultrasonography or biopsy to monitor women receiving raloxifene [107]. However, as there is no reason to believe that raloxifene protects the uterus from any pathology, the same routine surveillance measures should be regularly performed as for any postmenopausal woman who is not receiving HRT. f. Other Effects of Raloxifene Raloxifene increases the risk of venous thromboembolic disease by about threefold [107], an increase similar to that seen with both estrogen [110 – 112] and tamoxifen [42,85,113]. Hot flashes and leg cramps [98,103] as well as peripheral edema [103] are more common with raloxifene than placebo. There are no long-term data on any effect of raloxifene on cognitive function, but no deterioration of function as measured by psychometric testing at baseline and 6 and 12 months was shown with raloxifene in the evaluation of 143 postmenopausal women who received either raloxifene 60 or 120 mg or placebo [114].

D. Toremifene Toremifene (Fig. 2) is approved in several countries, including the United States, for use in postmenopausal women with ER advanced breast cancer. In that disease it has been compared directly with tamoxifen and has been shown to have both a similar efficacy and side effects profile [115]. Limited studies in postmenopausal breast cancer patients have examined the effects of toremifene on bone [116], showing the agent to be similar to, but weaker than, tamoxifen with respect to decreasing bone turnover markers. After 1 year of treatment, urinary NTX fell 33% with tamoxifen, 16% with toremifene; serum osteocalcin decreased 25% with tamoxifen, but was unchanged with toremifene. In the same study, tamoxifen increased spinal and hip bone density by 2 and 1% respectively, but there was no increase in BMD at either site with toremifene. Other effects of toremifene include reductions in total and LDL cholesterol and Lp(a) that were similar to those with tamoxifen, but toremifene increased HDL by 14% as compared to no effect with tamoxifen [68]. Adverse effects from toremifene include vaginal discharge and vaginal bleeding, hot flashes, and an increased risk of venous thromboembolic events [package insert, Searle], all similar to that seen with tamoxifen.

E. Other SERMs Initial clinical studies with idoxifene (Fig. 2) indicated its ability to decrease bone turnover markers by 25% compared to placebo in early postmenopausal women [117], but all studies with this compound are currently inactive. Similarly, clinical studies with droloxifene and levormeloxifene (Fig. 2) have been terminated. Miproxifene is in phase III trials for advanced breast cancer. Lasofoxifene (Fig. 2) has been shown to have skeletal effects similar to estradiol in ovariectomized rats without estrogen-like uterine stimulation [118]. Clinical trials with this agent in osteoporosis are anticipated. Clearly, additional SERMs will emerge in the future, as the search for the ideal SERM or SERMs continues to evolve. In the meantime, selected SERMs have proven value in reduction in risk of and in treatment of breast cancer, treatment of anovulation, and the prevention and treatment of osteoporosis.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

M. J. Tsai and B. W. O’Malley, Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu. Rev. Biochem. 63, 451 – 486 (1994). J. H. Clark, W. T. Schrader, and B. W. O’Malley, In “Textbook of Endocrinology” (J. D. Wilson and D. W. Foster, Eds.), pp. 35 – 90. Saunders, New York, 1992. K. Paech, P. Webb, G. G. J. M. Kuiper, S. Nilsson, J.-A. Gustafsson, P. Kushner, and T. S. Scanlan, Differential ligand activation of estrogen receptors ER and ER at AP1 sites. Science 277, 1508 – 1510 (1997). M. G. Elgort, A. Zou, K. B. Marschke, and E. A. Allegretto, Estrogen and estrogen receptor antagonists stimulate transcription from the human retinoic acid receptor-alpha 1 promoter via a novel sequence. Mol. Endocrinol. 10, 477 – 487 (1996). N. N. Yang, M. Venugopalan, S. Hardikar, and A. Glasebrook, Indentification of an estrogen response element activated by metabolites of 17-estradiol and raloxifene. Science 73, 1222 – 1225 (1996). J.-M. Vanacker, K. Pettersson, J.-A. Gustafsson, and V. Laudet,Transcriptional targets shared by estrogen receptor-related receptors (ERRs) and estrogen receptor (ER) , but not by ER. EMBO J. 18, 4270 – 4279 (1999). N. J. McKenna, J. Xu, Z. Nawaz, S. Y. Tsai, M.-J. Tsai, and B. W. O’Malley, Nuclear receptor coactivators: Multiple enzymes, multiple complexes, multiple functions. J. Steroid. Biochem. Mol. Biol. 69, 3 – 12 (1999). A. C. W. Pike, A. M. Brzozowski, R. E. Hubbard, T. Bonn, A.-G. Thorsell, O. Engstrom, J. Ljunggren, J.-A. Gustafsson, and M. Carlquist, Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist. EMBO J. 18, 4608 – 4618 (1999). A. M. Brzozowski, A. C. Pike, Z. Dauter, R. E. Hubbard, T. Bonn, O. Engstrom, L. Ohman, G. L. Greene, J.-A. Gustafsson, and M. Carlquist, Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389, 753 – 758 (1997). A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A. Agard, and G. L. Greene, The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927 – 937 (1998).

CHAPTER 70 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

617

Selective Estrogen Receptor Modulators (SERMS)

P. S. Danielian, R. White, J. A. Lees, and M. G. Parker, Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J. 11, 1025 – 1033 (1992). M. M. Montano, K. Ekena, K. D. Krueger, A. L. Keller, and B. S. Katzenellenbogen, Human estrogen receptor ligand activity inversion mutants: Receptors that interpret antiestrogens as estrogens and estrogens as antiestrogens and discriminate among different antiestrogens. Mol. Endocrinol. 10, 230 – 242 (1996). M. Nichols, J. M. Rientjes, and A. F. Stewart, Different positioning of the ligand-binding domain helix 12 and the F domain of the estrogen receptor accounts for functional differences between agonists and antagonists. EMBO J. 17, 765 – 773 (1998). A. S. Levenson and V. C. Jordan, The key to the antiestrogenic mechanism of raloxifene is amino acid 351 (aspartate) in the estrogen receptor. Cancer Res. 58, 1872 – 1875 (1998). A. S. Levenson, W. H. Catherino, and V. C. Jordan, Estrogenic activity is increased for an antiestrogen by a natural mutation of the estrogen receptor. J. Steroid Biochem. Mol.Biol. 60, 261 – 268 (1997). H. Wang, G. A. Peters, X. Zeng, M. Tang, W. Ip, and S. A. Khan, Yeast two-hybrid system demonstrates that estrogen receptor dimerization is ligand-dependent in vitro. J. Biol. Chem. 40, 23322 – 23329 (1995). N. N. Yang, H. U. Bryant, S. Hardikar, M. Sato, R. J. Galvin, A. L. Glasebrook, and J. D. Termine, Estrogen and raloxifene stimulate transforming growth factor-beta 3 gene expression in rat bone: A potential mechanism for estrogen- or raloxifene-mediated bone maintenance. Endocrinology 137, 2075 – 2084 (1996). T. Watanabe, S. Inoue, S. Ogawa, Y. Ishii, H. Hiroi, K. Ikeda, A. Orimo, and M. Muramatsu, Agonistic effect of tamoxifen is dependent on cell type, ERE-promoter context, and estrogen receptor subtype: Functional difference between estrogen receptors alpha and beta. Biochem. Biophys. Res. Commun. 236, 140 – 145 (1997). L. A. Paige, D. U. Christensen, H. Gron, J. Norris, E. B. Gottlin, K. M. Padilla, C.-Y. Chang, L. M. Ballas, P. T. Hamilton, D. P. McDonnell, and D. M. Fowlkes, Estrogen receptor (ER) modulators each induce distinct conformational changes in ER and ER. Proc. Natl. Acad. Sci. USA 96, 3999 – 4004 (1999). D. P. McDonnell, D. L. Clemm, T. Hermann, M. E. Goldman, and J. W. Pike, Analysis of estrogen receptor function in vitro reveals three distinct classes of antiestrogens. Mol. Endocrinol. 9, 659 – 669 (1995). J. D. Norris, L. Paige, D. J. Christensen, C.-Y. Chang, M. R. Huacani, D. Fan, P. T. Hamilton, D. M. Fowlkes, and D. P. McDonnell, Peptide antagonists of the human estrogen receptor. Science 285, 744 – 746 (1999). H. Shibata, T. E. Spencer, S. A. Onate, G. Jenster, S. Y. Tsai, M.-J. Tsai, and B. W. O’Malley, Role of coactivtors and corepressors in the mechanism of steroid/thyroid receptor action. Recent Prog. Horm. Res. 52, 141 – 165 (1997). M. M. Montano, K. Ekena, R. Delage-Mourrous, W. Chang, P. Martini, and B. S. Katzenellenbogen, An estrogen receptor-sensitive coregulator that potentiates the effectiveness of antiestrogens and represses the activity of estrogens. Proc. Natl. Acad. Sci. USA 96, 6947 – 6952 (1999). S. Kato, H. Endoh, Y. Masuhiro, T. Kitamoto, S. Uchiyama, H. Sasaki, S. Masushige, Y. Goto, E. Nishida, H. Kawashima, D. Metzger, and P. Chambon, Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270, 1491 – 1494 (1995). A. Tremblay, G. Tremblay, R. Labrie, and V. Giguere, Ligand-independent recruitment of SRC-1 to estrogen receptor  through phosphorylation of activation function AF-1. Mol. Cell 3, 513 – 519 (1999).

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37. 38. 39.

40.

41.

42.

42a.

D. N. Petersen, G. T. Tkalcevic, P. H. Koza-Taylor, T. G. Turi, and T. A. Brown, Identification of estrogen receptor 2, a functional variant of estrogen receptor  expressed in normal rat tissues. Endocrinology 139, 1082 – 1092 (1998). J. F. Couse, J. Lindzey, K. Grandien, J.-A. Gustafsson, and K. Korach, Tissue distribution and quantitative analysis of estrogen receptor- (ER) and estrogen receptor- (ER) messenger ribonucleic acid in the wild-type and ER-knockout mouse. Endocrinology 139, 4613 – 4621 (1997). G. G. Kuiper, B. Carlsson, K. Grandien, E. Enmark, J. Haggblad, S. Nilsson, and J.-A. Gustafsson, Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 138, 863 – 870 (1997). O. Vidal, M. Lindberg, L. Savendahl, D. B. Lubahn, E. M. Ritzen, J.A. Gustafsson, and C. Ohlsson, Disproportional body growth in female estrogen receptor-alpha-inactivated mice. Biochem. Biophys. Res. Commun. 265, 569 – 571 (1999). S. H. Windahl, O. Vidal, G. Andersson, J.-A. Gustafsson, and C. Ohlsson, Increased cortical bone mineral content but unchanged trabecular bone mineral density in female ER-/- mice. J. Clin. Invest. 104, 895 – 901 (1999). E. M. McInerney, K. E. Weis, J. Sun, S. Mosselman, and B. S. Katzenellenbogen, Transcription activation by the human estrogen receptor subtype  (ER) studied with ER and ER receptor chimeras. Endocrinology 139, 4513 – 4522 (1998). J. Hall and D. P. McDonnell, The estrogen receptor -isoform (ER) of the human estrogen receptor modulates ER transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology 140, 5556 – 5578 (1999). P. Pace, J. Taylor, S. Suntharalingam, R. C. Coombes, and S. Ali, Human estrogen receptor beta binds DNA in a manner similar to and dimerizes with estrogen receptor . J. Biol. Chem. 272, 25832 – 25838 (1997). J. Sun, M. F. Meyers, B. E. Fink, R. Rajendran, J. A. Katzenellenbogen, and B. S. Katzenellenbogen, Novel ligands that function as selective estrogens or antiestrogens for estrogen receptor- or estrogen receptor-. Endocrinology 140, 800 – 804 (1999). L. Johansson, J. S. Thomsen, A. E. Damdimopoulos, G. Spyrou, J.-A. Gustafsson, and E. Treuter, The orphan nuclear receptor SHP inhibits angonist-dependent transcriptional activity of estrogen receptors ER and ER. J. Biol. Chem. 274, 345 – 353 (1999). M. J. K. Harper and A. L. Walpole, Contrasting endocrine activities of cis and trans isomers in a series of substitute triphenyethylenes. Nature 212, 87 (1966). A. Klopper and M. Hall, New synthetic agent for the induction of ovulation: Preliminary trial in women. Br. Med. J. 2, 152 – 154 (1971). J. G. Williamson and J. D. Ellis, The induction of ovulation by tamoxifen. J. Obstet. Gynaecol. Br. Commonwealth 80, 844 – 847 (1973). V. C. Jordan, Antitumor efficacy of the antiestrogen ICI 46,474 (tamoxifen) in the dimethylbenzanthracene (DMBA)-induced rat mammary carcinoma model. J. Steroid Biochem. 5, 354 (1974). R. I. Nicholson and M. P. Golder, The effect of synthetic antiestrogens on the growth and biochemistry of rat mammary tumours. Eur. J. Cancer 11, 571 – 579 (1975). M. P. Cole, C. T. A. Jones, and I. D. H. Todd, A new antioestrogen agent in late breast cancer: An early clinical appraisal of ICI 46,474. Br. J. Cancer 25, 270 – 275 (1971). B. Fisher, J. P. Costantino, D. L. Wickerham, et al., Tamoxifen for the prevention of breast cancer: Report of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J. Natl. Cancer Inst. 90, 1371 – 1378 (1998). S. E. Reis, J. P. Costantino, D. L. Wickerham, E. Tan-Chin, J. Wang, M. Kavanah, Cardiovascular effects of tamoxifen in women with and without heart disease: Breast cancer prevention trial. J. Natl. Cancer Inst. 93, 16 – 21 (2001).

618 43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

SIRIS AND MUCHMORE P. H. Stewart and P. H. Stern, Effects of the antiestrogens tamoxifen and clomiphene on bone resorption in vitro. Endocrinology 118, 125 – 131 (1986). S. Feldman, H. W. Minne, S. Parvizi, M. Pfeifer, U. G. Lempert, F. Bauss, and R. Ziegler, Antiestrogen and antiandrogen administration reduce bone mass in the rat. Bone Miner. 7, 245 – 254 (1989). R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen prevents the skeletal effects of ovarian hormone deficiency in rats. J. Bone Miner. Res. 2, 449 – 456 (1987). A. Goulding, E. Gold, and L. Fisher, Effects of clomiphene and tamoxifen in vivo on the bone resorbing effects of parathyroid hormone and of high oral doses of calcitriol (1,25 (OH)2D3) in rats with intact ovarian function consuming a low calcium diet. Bone Miner. 8, 185 – 193 (1990). R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen inhibits osteoclast-mediated resorption of trabecular bone in ovarian hormone-deficient rats. Endocrinology 122, 1146 – 1150 (1988). T. R. Arnett, R. Lindsay, and D. W. Dempster, Effect of oestrogen and antioestrogens on osteoclast activity in vitro. J. Bone Miner. Res. 1(Suppl.), 99 (1986). R. L. Ward, G. Morgan, D. Dalley, and P. J. Kelly, Tamoxifen reduces bone turnover and prevents lumbar spine and proximal femoral bone loss in early postmenopasual women. Bone Miner. 22, 87 – 94 (1993). B. Kristensen, B. Ejlertsen, P. Dalgaard, L. Larsen, S. N. Holmegaard, I. Transbol, and H. T. Mouridsen, Tamoxifen and bone metabolism in postmenopausal low-risk breast cancer patients: A randomized study. J. Clin. Oncol. 12, 992 – 997 (1994). A. Gotfredson, C. Christiansen, and T. Palshof, The effect of tamoxifen on bone mineral content in premenopausal women with breast cancer. Cancer 53, 853 – 857 (1984). R. R. Love, R. B. Mazess, D. C. Tormey, H. S. Barden, P. A. Newcomb, and V. C. Jordan, Bone mineral density in women with breast cancer treated with adjuvant tamoxifen for at least two years. Breast Cancer Res. Treat. 12, 297 – 302 (1988). T. Fornander, L. E. Rutqvist, H. E. Sjoberg, L. Blomqvist, A. Mattsson, and U. Glas, Long-term adjuvant tamoxifen in early breast cancer: Effect on bone mineral density in postmenopausal women. J. Clin. Oncol. 8, 1019 – 1024 (1990). I. S. Fentiman, M. Caleffi, A. Rodin, B. Murby, and I. Fogelman, Bone mineral content of women receiving tamoxifen for mastalgia. Br. J. Cancer 60, 262 – 264 (1989). S. Turken, E. Siris, D. Seldin, E. Flaster, G. Hyman, and R. Lindsay, Effects of tamoxifen on spinal bone density in women with breast cancer. J. Natl. Cancer Inst. 71, 1086 – 1088 (1989). W. G. Ryan, J. Wolter, and J. D. Bagdade, Apparent beneficial effects of tamoxifen on bone mineral content in patients with breast cancer: Preliminary study. Osteoporosis Int. 2, 39 – 41 (1991). R. R. Love, R. B. Mazess, H. S. Barden, S. Epstein, P. A. Newcomb, V. C. Jordan, P. P. Carbone, and D. L. Demets, Effects of tamoxifen on bone mineral density in postmenopausal women with breast cancer. N. Engl. J. Med. 326, 852 – 856 (1992). R. R. Love, H. S. Barden, R. B. Mazess, S. Epstein, and R. J. Chappell, Effect of tamoxifen on lumbar spine bone mineral density in postmenopausal women after five years. Arch. Intern. Med. 154, 2585 – 2588 (1994). C. D. Wright, N. J. Garrahan, M. Stanton, J. C. Gazet, R. E. Mansell, and J. E. Compston, Effect of long term tamoxifen therapy on cancellous bone remodeling and structure in women with breast cancer. J. Bone Miner. Res. 9, 153 – 159 (1994). A. B. Grey, J. P. Stapleton, M. C. Evans, M. A. Tatnell, R. W. Ames, and I. R. Reid, The effect of the antiestrogen tamoxifen on bone mineral density in mormal late postmenopausal women. Am. J. Med. 99, 636 – 641 (1995).

61.

62.

63.

64. 65. 66.

67.

68.

69.

70.

71.

72.

73.

74. 75.

76.

77. 78.

79.

T. J. Powles, T. Hickish, J. A. Kanis, A. Tidy, and S. Ashley, Effect of tamoxifen on bone mineral density measured by dual energy x-ray absorptiometry in healthy premenopausal and postmenopausal women. J. Clin. Oncol. 14, 78 – 84 (1996). J. Chang, T. J. Powles, S. E. Ashley, R. K. Gregory, V. A. Tidy, J. G. Treleaven, and R. Singh, The effect of tamoxifen and hormone replacement therapy on serum cholesterol, bone mineral density and coagulation factors in healthy postmenopausal women participating in a randomized, controlled tamoxifen prevention study. Ann. Oncol. 7, 671 – 675 (1996). A. Gotfredson, C. Christiansen, and T. Palshof, The effect of tamoxifen on bone mineral content in premenopausal women with breast cancer. Cancer 53, 853 – 857 (1984). B. Kristensen, B. Ejlertsen, H. T. Mouridsen, K. W. Andersen, and J. B. Lauritzen, Breast Cancer Res. Treat. 39, 321 – 326 (1996). H. Wiseman, Tamoxifen as an antioxidant and cardioprotectant. Biochem. Soc. Symp. 61, 209 – 219 (1995). S. Ilanchezhian, M. Thangaraju, and P. Sachdanadam, Plasma lipids and lipoprotein alterations in tamoxifen-treated breast cancer women in relation to menopausal status. Cancer Biochem. Biophys. 15, 83 – 90 (1995). A. B. Grey, J. P. Stapleton, M. C. Evans, and I. R. Reid, The effect of the anti-estrogen tamoxifen on cardiovascular risk factors in normal postmenopausal women. J. Clin. Endocrinol. Metab. 80, 3191 – 3195 (1995). T. Saarto, C. Blomqvist, C. Enholm, M. R. Taskinen, and I. Elomaa, Antiatherogenic effects of adjuvant antiestrogens: A randomized trial comparing the effects of tamoxifen and toremifine on plasma lipid levels in postmenopausal women with node-positive breast cancer. J. Clin. Oncol. 14, 429 – 433 (1996). M. Thangaraju, K. Kumar, R. Ghandirajen, et al., Effect of tamoxifen on plasma lipids and lipoproteins in postmenopausal women with breast cancer. Cancer 73, 659 – 663 (1994). D. A. Shewmon, J. L. Stock, C. J. Rosen, K. M. Heiniluoma, M. M. Hogue, A. Morrison, E. M. Doyle, T. Ukena, V. Weale, and S. Baker, Tamoxifen and estrogen lower circulating lipoprotein(a) concentrations in healthy postmenopausal women. Arterioscler. Thromb. 14, 1586 – 1593 (1994). M. M. Biloma and V. C. Jordan, Is it time to develop an optimal endocrine therapy for premenopausal patients with axillary node positive and negative breast cancer? Semin. Surg. Oncol. 12, 339 – 345 (1996). R. R. Love, T. S. Surawicz, and E. C. Williams, Antithrombin III level, fibrinogen level, and platelet count changes with adjuvant tamoxifen therapy. Arch. Intern. Med. 152, 317 – 320 (1992). R. R. Love, P. A. Newcomb, D. A. Wiebe, T. S. Surawicz, V. C. Jordan, P. P. Carbone, and D. L. Demets, Effects of tamoxifen therapy on the lipid and lipoprotein levels in postmenopausal patients with node negative breast cancer. J. Natl. Cancer Inst. 82, 1327 – 1332 (1990). C. C. McDonald and H. J. Stewart, Fatal myocardial infarction in the Scottish adjuvant tamoxifen trial. Br. Med. J. 303, 435 – 437 (1991). L. E. Rutqvist and A. Mattson, Cardiac and thromboembolic morbidity among postmenopausal women with early stage breast cancer in a randomized trial of adjuvant tamoxifen. J. Natl. Cancer Inst. 85, 1398 – 1406 (1993). J. P. Costantino,YL. H. Kuller, D. G. Ives, B. Fisher, and J. Dignam, Coronary heart disease mortality and adjuvant tamoxifen therapy. J. Natl. Cancer Inst. 89, 776 – 782 (1997) F. Cosman and R. Lindsay, Selective endocrine receptor modulators: clinical spectrum. Endocr. Rev. 20, 418 – 434 (1999). Early Breast Cancer Trialists’ Collaborative Group, Tamoxifen for early breast cancer: an overview of randomized trials. Lancet 351, 1451 – 1467 (1998). T. Powles, R. Eeles, S. Ashley, D. Easton, J. Chang, M. Dowsett, A. Tidy, J. Viggers, and J. Davey, Interim analysis of the incidence of

CHAPTER 70

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

Selective Estrogen Receptor Modulators (SERMS)

breast cancer in the Royal Marsden Hospital tamoxifen randomised chemoprevention trial. Lancet 352, 98 – 101 (1998). U. Veronesi, P. Maisonneuve, A. Costa, V. Sacchini, C. Maltoni, C. Robertson, N. Rotmensz, and P. Boyle, Prevention of breast cancer with tamoxifen: preliminary findings from the Italian randomised trial among hysterectomised women. Lancet 352, 93 – 97 (1998). I. Cohen, D. J. Rosen, M. Altaras, Y. Beyth, J. Shapira, and D. Yigael, Tamoxifen treatment in premenopausal breast cancer patients may be associated with ovarian overstimulation, cystic formations and fibroid overgrowth. Br. J. Obstet. Gynecol. 69, 620 – 621 (1993). B. Fisher, J. P. Costantino, C. K. Redmond, E. R. Fisher, D. L. Wickerham, and W. M. Cronin, Endometrial cancer in tamoxifen-treated breast cancer patients: Findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J. Natl. Cancer Inst. 86, 527 – 537 (1994). L. S. Cook, N. S. Weiss, S. M. Scwartz, E. White, B. M. K.ight, D. E. Moore, and J. R. Daling, Population-based study of tamoxifen therapy and subsequent ovarian, endometrial, and breast cancers. J. Natl. Cancer Inst. 87, 645 – 651 (1995). C. C. McDonald, F. E. Alexander, B. W. Whyte, A. P. Forrest, and H. J. Stewart, Cardiac and thromboembolic morbidity in women receiving adjuvant tamoxifen for breast cancer ina randomised trial. The Scottish Cancer Trials Breast Group. Br. Med. J. 311, 977 – 980 (1995). B. Fisher and C. Redmond, Systemic therapy in node-negative patients: Updated findings from NSABP clinical trials. J. Natl. Cancer Inst. 11, 105 – 116 (1992). A. E. Wakeling, B. Valcaccia, E. Newboult, and L. R. Green, Nonsteroidal anti-estrogens: Receptor binding and biological response in rat uterus, rat mammary carcinoma and human breast cancer cells. J. Steroid Biochem. 20, 111 – 120 (1984). J. A. Clemens, D. R. Bennett, L. J. Black, and C. D. Jones, Effect of a new anti-estrogen, keoxifene (LY 156758), on growth of carcinogen-induced mammary tumors and on LH and prolactin levels. Life Sci. 32, 2869 – 2875 (1983). A. U. Buzdar, C. Marcus, F. Holmes, V. Hug, and G. Hortobagyi, Phase II evaluation of LY 156758 in metastatic breast cancer. Oncology 45, 344 – 345 (1988). L. J. Black, M. Sato, E. R. Rowley, D. E. Magee, A. Bekele, D. C. Willaims, G. J. Cullinan, R. Bendele, R. F. Kauffman, W. R. Bensch, C. A. Frolik, J. D. Termine, and H. U. Bryant, Raloxifene (LY 139481 HCl) prevents bone loss and reduces serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J. Clin. Invest. 93, 63 – 69 (1994). M. Sato, J. Kim, L. L. Short, C. W. Slemenda, and H. U. Bryant, Longitudinal and cross sectional analysis of raloxifene effects on the tibiae from ovariectomized aged rats. J. Pharmacol. Exp. Ther. 272, 1252 – 1259 (1995). G. L. Evans, H. U. Bryant, D. E. Magee, and R. T. Turner, Raloxifene inhibits bone turnover and prevents further cancellous bone loss in adult ovariecotmized rats with established osteopenia. Endocrinology 137, 4139 – 4144 (1996). C. A. Frolik, H. U. Bryant, E. C. Black, D. E. Magee, and S. Chandrasekhar, Time-dependent changes in biochemical bone markers and serum cholesterol in ovariectomized rats: Effects of raloxifene HCl, tamoxifen, estrogen and alendronate. Bone 18, 621 – 627 (1996). G. L. Evans, H. U. Bryant, D. Magee, M. Sato, and R. T. Turner, The effects of raloxifene on tibia histomorphometry in ovariectomized rats. Endocrinology 134, 2282 – 2288 (1994). C. H. Turner, M. Sato, and H. U. Bryant, Raloxifene preserves bone strength and bone mass in ovariectomized rats. Endocrinology 135, 2001 – 2005 (1994). H. U. Bryant, H. W. Cole, E. R. Rowley, et al., Raloxifene and LY 117015 prevent bone loss induced by an LH-RH agonist without

619 estrogen-like stimulation of the uterus. In “10th ICE I,” p. 572 (1996). 96. C. P. Jerome and C. J. Lees, Raloxifene increases bone mass and reduces bone turnover in ovariectomized cynomologous monkeys. J. Bone Miner. Res. 11 (Suppl. 1), S445 (1996). 97. R. P. Heaney and M. W. Draper, Raloxifene and estrogen: Comparative bone remodeling kinetics. J. Clin. Endocrinol. Metab. 82, 3425 – 3429 (1997). 98. P. D. Delmas, N. H. Bjarnason, B. H. Mitlak, A.-C. Ravoux, A. S. Shah, W. J. Huster, M. Draper, and C. Christiansen, Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N. Engl. J. Med. 337, 1641 – 1647 (1997). 99. M. Gunness, K. Prestwood, Y. Lu, et al., Histomorphometric, bone marker and bone mineral density response to raloxifene HCl and Premarin in postmenopausal women. In “79th annual meeting of the Endocrine Society,” p.67 (1997). 100. P. J. Meunier, E. Vignot, P. Garnero, E. Confavreux, E. Paris, S. LiuLeage, S. Sarkar, T. Liu, M. Wong, and M. Draper, Treatment of postmenopausal women with osteoporosis or low bone density with raloxifene. Osteoporosis. Int. 10, 330 – 336 (1999). 101. E. G. Lufkin, M. D. Whitaker, T. Nickelsen, R. Argueta, R. H. Caplan, R. K. Knickerbocker, and B. L. Riggs, Treatment of established osteoporosis with raloxifene: A randomized trial. J. Bone Miner. Res. 13, 1747 – 1754 (1998). 102. O. Johnell, W. Scheele, Y. Lu, and M. Lakshmanan, Effects of raloxifene (RLX), alendronate (ALN) and RLX ALN on bone mineral density (BMD) and biochemical markers of bone turnover in postmenopausal women with osteoporosis. J. Bone Miner. Res. 14 (Suppl. 1), S157 (1999). 103. B. Ettinger, D. M. Black, B. H. Mitlak, et al., Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene. Results from a 3-year randomized clinical trial. JAMA 282, 637 – 645 (1999). 104. M. W. Draper, D. E. Flowers, W. J. Huster, J. A. Neild, K. D. Harper, and C. A. Arnaud, A controlled trial with raloxifene (LY 139481) HCl: Impact on bone turnover and serum lipid profile in healthy postmenopausal women. J. Bone Miner. Res. 11, 835 – 842 (1996). 105. B. W. Walsh, L. H. Kuller, R. A. Wild, S. Paul, M. Farmer, J. B. Lawrence, A. S. Shah, and P. W. Anderson, Effects of raloxifene on serum lipids and coagulation factors in healthy postmenopausal women. JAMA 279, 1445 – 1451 (1998). 106. B. W. Walsh, S. Paul, R. A. Wild, R. A. Dean, R. P. Tracy, D. A. Cox, and P. W. Anderson, The effects of hormone replacement therapy and raloxifene on C-reactive protein and homocysteine in healthy postmenopausal women: A randomized, controlled trial. J. Clin. Endocrinol. Metab. 85, 214 – 218 (2000). 107. S. R. Cummings, S. Eckert, K. A. Krueger, et al., The effect of raloxifene on risk of breast cancer in postmenopausal women. Results from the MORE randomized trial. JAMA 281, 2189 – 2197 (1999). 107a. J. A. Cauley, L. Norton, M. E. Lippman et al. Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4 year results from the MORE trial. Br. Cancer Res. Treat. 65, 125 – 134 (2001). 108. G. C. Davies, W. J. Huster, W. Shen, B. Mitlak, L. Plouffe, A. Shah, and F. J. Cohen, Endometrial response to raloxifene compared with placebo, cyclical hormone replacement therapy, and unopposed estrogen in postmenopausal women. Menopause 6, 188 – 195 (1999). 109. F. R. Dijkhuizen, H. A. Brolmann, B. J. Oddens, et al., Transvaginal ultrasonography and endometrial changes in postmenopausal breast cancer patients receiving tamoxifen. Maturitas 25, 45 – 50 (1996). 110. E. Daly, M. P. Vessey, M. M. Hawkins, J. L. Carson, P. Gough, and S. Marsh, Risk of venous thromboembolism in users of hormone replacement therapy. Lancet 348, 977 – 980 (1996).

620 111. H. Jick, L. E. Derby, M. W. Myers, C. Vasilakis, and K. M. Newton, Risk of hospital admission for idiopathis venous thromboembolism among users of postmenopausal oestrogens. Lancet 348, 981–983 (1996). 112. F. Grodstein, M. J. Stampfer, S. Z. Goldhaber, J. E. Manson, G. A. Colditz, F. E. Speizer, W. C. Willett, and C. H. Hennekens, Prospective study of exogenous hormones and rsik of pulmonary embolismin women. Lancet 348, 983 – 987 (1996). 113. A. Lipton, H. A. Harvey, and R. W. Hamilton, Venous thromboembolism as a side effect of tamoxifen treatment. Cancer Treat. Rep. 68, 887 – 889 (1984). 114. T. Nickelsen, E. G. Lufkin, B. L. Riggs, D. A. Cox, and T. H. Crook, Raloxifene hydrochloride, a selective estrogen receptor modulator: Safety assessment of effects on cognitive function and mood in postmenopausal women. Pyschoneuroendocrinology 24, 115 – 128 (1999).

SIRIS AND MUCHMORE 115. M. Gershanovich, D. F. Hayes, J. Ellmen, et al., High-dose toremifene vs. tamoxifen in postmenopausal advanced breast cancer. Oncology 11, 29 – 36 (1997). 116. M. B. Marttunen, P. Hietanen, A. Tiitinen, and O. Ylikorkala, Comparison of effects of tamoxifen and toremifene on bone biochemistry and bone mineral density in postmenopausal breast cancer patients. J. Clin. Endocrinol. Metab. 83, 1158 – 1162 (1998). 117. S. Weiss, H. Mulder, C. Chesnut, M. Greenwald, P. Delmas, R. Eastell, et al., Idoxifene reduces bone turnover in osteopenic postmenopausal women. “80th Annual Meeting of the Endocrine Society,” Abstract 403, (1998). 118. H. Z. Ke, V. M. Paralkar, W. A. Grasser, et al., Efects of CP-336,156, a new, nonsteroidal estrogen agonist /antagonist, on bone, serum cholesterol, uterus and body composition in rat models. Endocrinology , 139, 2068 – 2076 (1998).

CHAPTER 71

Phytoestrogen and Other Phytochemical Effects on Bone RICHARD PRINCE

I. II. III. IV. V. VI.

Department of Medicine, University of Western Australia, Sir Charles Gairdner Hospital, Perth 6009, Western Australia

VII. Specific Phytoestrogens in Relation to Bone and Mineral Physiology and Anatomy VIII. Conclusions References

Introduction Definition Classification and Biological Action of Phytoestrogens Concentration in Plants Bioavailability The Epidemiology of Phytoestrogens as Neutraceutical Agents

I. INTRODUCTION

unnatural, and at risk of having long-term side effects. Furthermore it has been suggested on epidemiological grounds that a variety of food compounds consumed in oriental countries contribute to a low rate of certain diseases in those countries. Prominent among these food compounds are soy-based foods. Interest in the effects of this food staple has generated interest in a wide range of other plant preparations that are sold as over-the-counter products with pharmacologic properties for the alleviation of human disease. A third reason for the current interest in plant compounds in osteoporosis, in addition to this naturalistic approach to human health, is that there is a long history of the development of effective pharmaceuticals from plant products. The list of pharmacologic agents extracted from plants is a long one and includes such important compounds as digoxin, atropine, cocaine, morphine, and salicylic acid. Finally proponents of

The importance of effects of plant-derived compounds on human health and disease has been recognized since time immemorial. In recent years this interest has intensified. A study in 1993 in Australia has shown that nearly 50% of the population consumes at least one nonmedically prescribed alternative medicine at a cost of over half a billion dollars per annum [1]. In America in 1997 there were approximately 629 million visits to alternative practitioners and only 386 million to primary care physicians [2]. The total cost of alternative medicines has been estimated to have risen 45% from 1990 to US $21.2 billion in 1997. Herbal medicine treatments were prominent alternative interventions. This interest has derived from anxieties about long-term effects of pharmaceutical agents that are seen as manufactured,

OSTEOPOROSIS, SECOND EDITION VOLUME 2

621

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

622

RICHARD PRINCE

TABLE 1 Members of the Flavonoid or Phenolic Phytoestrogens Group Lignans

Isoflavones

Coumestans

Flavones/Flavenols

Metairesinol

Formononetin

Coumestrol

Luteolin

Secoisolariciresinol

Daidzein

Repensol

Quercetin

Biochanin A

Trifoliol

Apigenin

Enterolactone

Genistein

KCA 098

Kaempferol

Enterodiol

Equol

KCA 012

Flavanones

Chalcones

Naringenin

Phloretin

Ipriflavone

alternative or complementary medicine suggest they are popular because they work [3]. In this chapter the scientific data currently available is reviewed with particular emphasis on the skeletal system. Deficiency of estrogen and its metabolites have major effects on the development of skeletal fragility in both women and men. There are now a substantial number of studies of phytoestrogen effects on the bone and mineral systems including epidemiologic studies and in vitro and in vivo studies in animals and humans together with a few clinical trials in humans.

II. DEFINITION Phytoestrogens are compounds that are considered to have biological actions similar to that observed for estrogen (Table 1). As such they may fit the concept of selective estrogen receptor modifiers (SERMs). The compounds currently under study in relation to human disease have been demonstrated to have estrogenic activity at least in part. These concepts derive from early studies in Western Australia that identified subterranean clover as a source of estrogenic plant compounds resulting in infertility in sheep [4], Hence the generic name phytoestrogens that has been applied to this class of compounds. In fact, as will become clear, there are some grounds for considering that these compounds may also act through nonestrogenic mechanisms. Despite similarities in their actions in animal physiology, phytoestrogens are considered to have widely differing roles in plant physiology. Their overall importance to human anatomy and physiology depends on their biological potency in the organ under consideration and the concentration that can be reasonably attained in the circulation. This last issue is complex as it depends on the amount of the compound that can reasonably be consumed, which depends on the concentration in the plant and the amount of that plant consumed, and its metabolism and absorption in the intestine.

III. CLASSIFICATION AND BIOLOGICAL ACTION OF PHYTOESTROGENS Although the primary definition of the compounds constituting the phytoestrogens depends on a recognized biological action similar to that of estrogen the category includes a broad range of chemical structures. They consist of six chemical groupings, the chalcones, flavonols, flavanones, flavones, isoflavones, coumestans, and lignans (Table 1; Figure 1). These groups are variously called phenolic phytoestrogens, flavonoids, or nonsteroidal phytoestrogens. These various terminologies emphasize the various aspects of these compounds, namely that while containing a diphenylpropane ring structure they do not have the typical four-ring steroid structure. The flavonoid classification derives from the Latin yellow and refers to their role as plant pigments. They may also function as insect deterrents and fungicides. A seventh group, the resorcylic acid lactones, is not usually considered a true phytoestrogen as this group consists of compounds that are actually produced by molds which can contaminate grains.

FIGURE 1

The structure of the three main phytoestrogen groups compared to that of estrogen.

623

CHAPTER 71 Phytoestrogen and Other Phytochemical Effects on Bone

Each of these groups contain compounds that have estrogenic activity in in vitro assays in which the readout variable is receptor binding and or activation of a reporter estrogen response element, indicating an effect on binding of the receptor to DNA [5,6]. Many of these compounds are over a thousand times less active than 17 estradiol; nevertheless they can constitute up to 7% of the dry weight of some plants [7]. The principal classes of phytoestrogens currently considered to be important in human biology are the isoflavones, the coumestans, and the lignans. Their relative potency in affinity studies using the uterine estrogen receptor was described some time ago [8 – 10] (Table 2). These studies have recently been repeated with recombinant DNA-produced estrogen receptor [11,12]. The study by Kuiper et al. [11] showed that the isoflavones, coumestans, flavones, flavanones, and chalcones all bind to the estrogen receptor and stimulate transcriptional activity, although the relative activities varied between the alpha or beta receptors. Depending on the system studied the phytoestrogens can be agonists or antagonists. Other effects on estrogen metabolism of lignans, coumestans, isoflavones, and flavenols includes inhibition of aromatase at micromolar concentrations [13 – 15] and inhibition of 17-hydroxysteroid dehydrogenase [16,17]. Both of these effects may act to reduce estrogen production. There are some data to suggest that SHBG levels may rise after exposure to a variety of isoflavones [18], which may reduce free estrogen levels. Another enzyme that may be inhibited by isoflavones is 5 reductase [16,19]. It is important to recognize that flavonoid compounds are considered by some to have their principle mechanism of action by antioxidant effects [20]. This applies particularly to genistein, quercetin, myricetin, kaempferol, luteolin, and apigenin. A further well recognized effect of genistein is as an inhibitor of tyrosine TABLE 2 The Relative Potencies of Various Phytoestrogens as Determined by Their Affinity for the Estrogen Receptor of Various Tissues Ishhikawa cellsa

Sheep uterusb

Human breast cellsc

100.00

100.00

100.00

Coumestrol

0.20

5.00

10.00

Genistein

0.08

0.90

2

Equol

0.06

0.40

—

Dadzein

0.01

0.10

0

Biochanin A

0

0

—

Formononetin

0

0

0.001

Estradiol

a

Markiewicz et al. [10]. Shutt and Cox [9]. c Martin et al. [8] b

TABLE 3

Isoflavones

Approximate Concentrations of Dietary Phytoestrogens

Daidzein

Genistein

Coumestans Flavonols

Lignans

Coumestrol

Soybeans

0.10 g/100 g

Tofu

0.01 g/100 g

Black beans

0.07 g/100 g

Soybeans

0.10 g/100 g

Tofu

0.01 g/100 g

Black beans

0.07 g/00 g

Alfafa sprouts

0.07 g/100 g

Clover sprouts

0.70 g/100 g

Onions

0.35 g/100 g

Apples

0.36 g/100 g

Red wine

0.11 g/100 g

Cereals

0.06 g/100 g

Fruits

0.06 g /100 g

Vegetables

0.06 g/100 g

Flax seed

0.67 g/100 g

Linseed

0.67 g/100 g

Liquorice Chalcones

Apple seeds

kinase activity [21,22]. This may account for some of the antiproliferative actions of genistein and account for effects in cells without demonstrable estrogen receptors. Effects of phytoestrogens on the prostaglandin pathway have also been described [23]. Because of the potential for different modes of action and differing potencies it is difficult to predict the overall biological effects of a particular phytoestrogen. The problem is compounded by uncertainties as to the actual concentrations in foods consumed by an individual and in variation in bioavailablity and metabolism. It must not be forgotten that the endogenous production of estrogen will also significantly influence the activity of any phytoestrogens consumed.

IV. CONCENTRATION IN PLANTS Approximate concentrations of the various dietary phytoestrogens have been measured by HPLC [24] and are shown in Table 3. Subjects consuming macrobiotic diets and diets high in soy products can absorb over 100 mg of phytoestrogen per day. There are some general features of these compounds that should be emphasized here: the concentrations of these compounds in any particular plant is quite variable. Factors determining this variability include the strain of the plant species and more particularly the precise cultivation conditions of the plant. This is especially so because phytoestrogens are produced as stress compounds. An extensive series of studies on clovers showed that a variety of stresses including mineral deficiencies and either

624 water deficit or excess resulted in marked increases in isoflavone concentrations [25]. Furthermore different parts of the plant may have high variability in phytoestrogen concentrations. In clovers the principle concentration of isoflavones is in the leaves. In China, where plant extracts, often prepared by aqueous extraction, play a significant role in health care delivery, identification of the correct plant is undertaken by experts in the area. It is for these reasons that in the West methods of extracting and synthesizing the active compound to allow precise dosing were developed. It should, however, be emphasized that many proponents of the efficacy of phytoestrogens in human disease consider it is the combination of phytoestrogens found in individual plants and combinations that provide the therapeutic effect. While synergy in action or effect between various phytoestrogens is theoretically possible to date there is little evidence for this.

V. BIOAVAILABILITY Variability of absorption of phytoestrogens is a recognized problem that may depend on the bacterial populations in the intestine [26]. This is of particular importance as some of the isoflavones are consumed as phytoestrogens of relatively low potency that are chemically modified in the gut to a more active compound. Specific data on absorption and degradation is included under data on the specific phytoestrogen.

VI. THE EPIDEMIOLOGY OF PHYTOESTROGENS AS NEUTRACEUTICAL AGENTS The concept that the dramatic differences in disease incidence around the world have an environmental rather than a genetic basis has become commonplace [27]. However, confounding due to genetic factors cannot be entirely excluded. Adlercreutz has proposed that the incidence of breast cancer, prostate disease, gastrointestinal cancer, and cardiovascular disease in Asian and Western countries may be related to phytoestrogen consumption [28,29]. The principle source of the epidemiologic data for this theory has been derived from a study of disease prevalence and incidence in Japan, where it is claimed that the low incidence of heart disease, breast cancer, and prostate cancer is due to the phytoestrogen content of the diet principally in the form of soy [30,31]. In the case of skeletal disease the epidemiology of both bone density and fracture does not seem to support the concept that high phytoestrogen intake is bone protective. Epidemiological studies of bone density in Japanese and American subjects have shown that bone density increases

RICHARD PRINCE

when ethnic Japanese and Chinese move to Western countries despite a reduction in phytoestrogen intake [32,33]. Indeed in a 10-year longitudinal follow-up of postmenopausal women in Europe, the urine excretion of the lignan enterolactone was higher in women with the greater bone loss. There was no relationship with daidzein, genistein, or equol [34]. In a cross-sectional study from Japan [35] in which food records were analyzed for their phytostrogen content average isoflavone intake was calculated as 40 mg per day (range 8 to 88 mg). There was no relation to bone density although there was a positive relation with serum creatinine and uric acid. The spine fracture epidemiology reflects the bone density data at least in relation to Japanese living in Hawaii, where there is a lower rate of spine fracture than in Japan, which is partly due to the increased bone density in these individuals [36]. Hip fracture epidemiology does not support the concept that higher phytoestrogen intake in Japan protects against fracture. Although hip fracture rates are lower in ethnic Japanese, despite the lower bone density, the rates are lower in both Japanese living in Hawaii and Japan than in Caucasians living in Hawaii or Minnesota [37]. Thus geographic environment, which may include dietary practices, would appear to have little effect on this important fracture syndrome. It is important to understand, however, that phytoestrogens are but one plant constituent that may influence bone strength and osteoporotic fracture. Others include calcium [38], salt [39], protein, and diet acid. Dissecting out the differing effects of these compounds is diffi cult in epidemiologic studies as co-correlations confuse the approach. Thus more specific interventional experimental approaches are required to examine the importance of phytoestrogens to bone health.

VII. SPECIFIC PHYTOESTROGENS IN RELATION TO BONE AND MINERAL PHYSIOLOGY AND ANATOMY A. Isoflavones 1. METABOLISM Some food sources of this class of phytoestrogen are derived from precursors present in legumes or pulses, some of which are shown in Table 3. However, it must be emphasized that even among one food source there is large variation in content [26]. The best recognized members of this class are genistein and daidzein, both found in soy. Genistein is found in its methylated form known as formononetin. Daidzein is found in its methylated form known as biochanin A. In addition glycosides of all four compounds are found in the legume either as the glycated form (ononin, sissotrin, daidzein, genistein) or as the acetyl or malonyl derivatives.

CHAPTER 71 Phytoestrogen and Other Phytochemical Effects on Bone

A study of the renal excretion of the various compounds in the urine of 12 volunteers before and after soy challenge showed huge variations in excretion of daidzein and genistein as well as their metabolites equol, O-desmethylangolensin and 6-hydroxy-O-desmethylangolensin [40]. Data such as this indicate that in addition to large variations in the isoflavone content of foodstuffs there is large variation in intestinal degradation and absorption. The conversion of precursors occurs in the bowel and involves the action of bowel flora with resulting generation of the aglycones and equol. Equol is considered to be a more potent isoflavone than daidzein (Table 2). There are large interindividual variations in the ability to produce this metabolite [41]. In the absence of bowel flora isoflavones are not absorbed [26]. Indeed, over 85% of dietary isoflavones may be degraded in the bowel [42]. Once absorbed the isoflavones are conjugated with glucuronic or sulphuric acid either in the bowel or liver. The plasma concentration of isoflavones has been measured [43]. Although concentrations are low in subjects consuming diets free of soy, after high levels of soy consumption plasma concentrations may be as high as 2 mol/liter [42 – 4 4 ] . This may contribute to large variations in measured excretion in addition to variation in the rates of absorption of the metabolites. Much further work needs to be done on the pharmacokinetics of these compounds. A further problem relates to the question of whether the effects of soy are due to the phytoestrogen content or whether they are due to the effects of the soy protein [45]. An alternative is that the effects may be due to the removal of other dietary products so that they may be replaced with soy. ON

2. BIOLOGIC ACTION OF ISOFLAVONES BONE AND MINERAL

As indicated above, in addition to effects via classical estrogen receptor-mediated pathways genistein has been shown to have tyrosine kinase inhibitory activity [21,22]. This effect has been implicated in the increase in intracellular calcium induced in the rat osteoclast by genistein. This effect was associated with disruption of the actin ring [46]. Evidence for an apoptotic effect of both genistein and daidzein on rat bone osteoclasts via a calcium signaling pathway has also been presented [47]. In addition to inhibition of osteoclast activity there is evidence that isoflavones may have anabolic effects on bone. In vitro studies of bone cultures from the femoral diaphysis of elderly female rats showed evidence for stimulation of anabolic effects at micromolar concentrations of daidzein and genistein compared to nanomolar concentrations of estradiol [48,49]. Stimulatory effects of high concentrations of genistein on cell proliferation have also been described in human bone cells [50]. These data have been supported by animal data reviewed next.

625 a. Animal Studies A recent study has drawn attention to the fact that rat diets containing soy flour may contain enough daidzein and coumestrol to have biological effects [51]. The implication of this finding in evaluation of the effects of interventions using the rat oophorectomy model is uncertain at the current time but clearly could reduce the impact of oophorectomy if a soy-based diet is fed [52]. A study of genistein in the young developing rat oophorectomy model of osteoporosis showed some evidence of protection against oophorectomy-induced reduction in bone density [53]. Histopathology of the distal femur showed an increase in bone formation rate in the genistein group with no effect on bone resorption. This suggests that the relative preservation of bone mass and structure was due to an osteoblast stimulatory action rather than an osteoclast inhibitory action that would be expected from an analogy with estrogen. These results may be associated with the fact that 2-month-old rats are still growing. Two other studies of young rats showed a reduction in bone loss with a soy diet containing approximately 200 mg isoflavones per 100 g soy. The effect was only statistically significant at the femur site and not the fourth lumbar vertebra and only when soy plus isoflavones was started at the time of oophorectomy rather than 35 days after oophorectomy [54,55]. The mouse has been used to study both B cell regulation and dosing effects of genistein on bone density [56, 57]. B cell production was increased by both genistein and estrogen. Beneficial effects were seen on the preservation of bone architecture and density measured by dual-energy X-ray absorption (DXA) with doses as low as 0.4 mg/day administered parenterally. The ED50 for protection against bone loss was 0.29 mg/day compared to the ED50 for the hypertrophic effect on the uterus of 3.0 mg/day. Plasma concentrations at these levels of administration were approximately 1 mol/liter, which can be achieved by high levels of isoflavone intake in man [42,43]. These data suggest that the tissue sensitivity of bone to genistein may be in the range of dosing obtainable in normal life. Unfortunately, the effects of high doses of genistein on bone density was not studied. This issue is quite important in view of evidence for a biphasic dose – response curve for genistein in the lactating calcium deficient rat model of high turnover osteoporosis, which showed a reduction in femur ash weight at high doses of genistein [58]. Recently the effects of natto, soybean fermented with Bacillus subtilis, with or without added menaquinone have been studied [59,60]. Both studies showed prevention of bone loss in the oophorectomized rat. This preparation of soy contains other metabolites of isoflavones including 6-O succinylated isoflavone glycosides. b. Human Studies One randomized controlled doubleblind short-term study of 66 postmenopausal women

626

RICHARD PRINCE

randomized to a diet supplemented with 40 g casein and nonfat dry milk or a 40-g soy protein supplement with either a low (56 mg/day) or high (90 mg/day) isoflavone content was performed over a 6-month period. Bone density in the lumbar spine increased marginally in the 90 mg/day group but not at the lower dose. There was no effect at other bone sites [61]. 3. CONCLUSIONS Evidence that dietary isoflavones have an effect on the skeletal system is now well founded in the experimental literature. To date the effects demonstrated have been small. Nevertheless, they support evidence from the cardiovascular field [62] that the isoflavone content of soy, if consumed and absorbed in large enough quantities, may have beneficial effects on surrogate end points such as bone density. Whether the effect size will be large enough to reduce fracture is uncertain at the current time. Further longitudinal epidemiological and interventional studies preceded by large dose – response studies in animals will be required to answer this important question. The potential for a biphasic dose response curve means that any recommendation to increase soybean or isoflavone intake to preserve the skeleton would be premature.

B. Ipriflavone Ipriflavone is synthetic isoflavone that was developed in the 1960s. In the 1980s investigations of its value for the management of osteoporosis were commenced largely in Japan and Italy. In chemical structure ipriflavone is a modified isoflavone [63]. The biological actions have been studied to some extent. For example, it is known that ipriflavone, although binding to both estrogen receptors has an affinity equivalent to that of formononetin and quercetin and lower than that of other isoflavones [11]. These data raise important questions as to its mode of action, especially as it has been claimed that no estrogenic actions or side effects have been demonstrated in animals or humans [64,65]. More recently ipriflavone has been shown to be an inhibitor of cytochrome P450, which might account for some drug interactions that have been reported [66]. Whatever the biochemical mechanisms of action there is evidence that ipriflavone inhibits osteoclast formation in murine spleen cells cocultured with murine bone marrow stromal cells [67]. In the fetal long-bone culture system ipriflavone inhibits bone resorption [68]. In addition, there is evidence that ipriflavone, unlike estrogen, stimulates increased bone formation in the oophorectomised rat as measured histologically and biochemically [69]. 1. CLINICAL TRIALS To date at least 12 clinical studies of ipriflavone on bone density and biochemical endpoints have been undertaken.

The standard dose is 600 mg per day, which would be impossible to achieve using dietary sources of isoflavones. An early double-blind randomized trial of 6 months duration comparing ipriflavone with placebo in postmenopausal women showed an increase in radial bone density associated with a reduction in osteocalcin levels [70]. Similar data have been shown in women immediately after bilateral oophorectomy, in which ipriflavone was able to prevent radial bone loss [71]. A 2-year randomized trial in postmenopausal women with low bone mass using intention to treat outcomes showed excellent protection against bone loss at the radius with a significant treatment effect at both the 1- and the 2-year time points [72]. Similar data from two other trials has been published [73]. These data have been confirmed in a 12-month placebo-controlled trial of postmenopausal women in which the bone density at the L2 – L4 lumbar vertebrae was the primary outcome variable [74]. Four studies using very similar study design have also showed significant effects at the spine [75 – 78]. Interestingly, two randomized controlled trials showed that there may be an additive effect of ipriflavone on radial bone density when added to low-dose conjugated estrogen, 0.3 mg/day [79,80], suggesting an additive agonist effect at the bone. Bone biochemistry has been studied in several of these trials. In general bone resorption markers fall [72,73, 75,76], although bone formation markers only showed a reduction in some of these studies. A recent large, multicenter study, presented in abstract form, showed no effect on bone density at any site. The mode of action remains uncertain and may involve mechanisms of action different form those of estrogen including stimulation of bone formation. Whether there are treatment benefits in terms of fracture risk is currently unknown. Furthermore, the side effect profile has not been clearly defined, in particular its effects on reproductive tissue.

C. Coumestans The data that the principal coumestan coumestrol interacts with the estrogen receptor is good and has been discussed already. In most biosassays undertaken to date coumestrol has the highest affinity of any of the phytoestrogens for estrogen receptors (Table 2). Also consistent with other phytoestrogens coumestrol has a higher affinity for estrogen receptor alpha than beta [11]. In this study the transcriptional activity of coumestrol was similar to that of genistein. The in situ localization of coumestrol has been confirmed to be associated with the estrogen receptor. Indeed as a result of this study coumestrol has been recommended as an agent for the localization of the estrogen receptor in histochemical studies [81]. Although food sources of coumestrol have been delineated (Table 3), the precise chemical structures of the coumestans in plants has not been defined.

627

CHAPTER 71 Phytoestrogen and Other Phytochemical Effects on Bone

1. BIOLOGIC ACTION OF COUMESTANS BONE SYSTEM

ON THE

Studies in 1995 showed in vitro effects of coumestrol on bone that included inhibition of bone resorption and stimulation of bone formation [82]. Tsutsumi and colleagues from Kissei Pharmaceuticals in Japan have studied two chemical derivatives of coumestrol, KCA 098 and 012 [83 – 85]. In cultures of fetal rat long bones, both preparations as well as coumestrol inhibited bone resorption. The first direct studies of coumestrol in the prevention of bone loss in the ovariectomized rat were undertaken by Dodge et al. [86]. In a study of a variety of phytoestrogen compounds, Draper et al. found biological effects of coumestrol in reducing bone loss [87]. In view of the relatively high potency of this compound compared to other phytoestrogens it may be possible to achieve doses in humans that may exert biological effects.

significance of these compounds in the normal human physiology of bone has yet to be determined. In view of extensive evidence for the biological importance of low concentrations of endogenous circulating estrogen in postmenopausal women, it may be possible to consume enough plant estrogen from particular foods to have bone effects in this age group. Certainly the concept of therapeutic foods is well accepted in Asian countries. An alternative scenario is to examine chemical extracts or concentrates of plant estrogens for use as treatments for osteoporosis. Such extracts would have to demonstrate similar safety to currently available SERMs. Another approach is to chemically modify plant compounds to increase their potency. Finally, it must be accepted that large numbers of individuals are already consuming herbal preparations for therapeutic ends. It is appropriate to examine these practices to determine whether there is any evidence for effectiveness in view of the huge interest in this topic.

D. Lignans References Lignans are found in high concentration in flax seed as metairesinol and secoisolariciresinol. These are broken down in the bowel to the products found in urine, enterodiol and enterolactone [41]. Enterolactone, but not enterodiol, was able to stimulate an estrogen-like effect on MCF-7 cells [88]. In vitro studies in human preadipose cell cultures showed evidence that the lignan enterolactone inhibits aromatase activity [15]. To date no direct studies of these weak estrogens have been undertaken in relation to the skeletal system. However, in an epidemiological study of the excretion of phytoestrogen metabolites, a positive relationship between lignan excretion and bone loss was reported [34]. This suggests a possible action as an antiestrogen in that residual estrogen action may be protective against bone loss and fracture after the menopause [89].

E. Other Agents There are a variety of reports on other plant extracts that have effects on the skeletal system in the ovariectomized rat, suggesting a possible estrogen-like effect. These include an extract of the stem of Sambucus sieboldina and Tochu bark [90,91]. Finally a prenyl derivative of the flavanonone naringenin 8-isopentenylnaringenin has been shown to also reduce bone loss in the rat ovariectomy model [92].

VIII. CONCLUSIONS It is clear from the data presented above that phytochemicals and phytoestrogens must be considered as having potentially important effects on the skeletal system. The exact

1. A. H. MacLennan, D. H. Wilson, and A. W. Taylor, Prevalence and cost of alternative medicine in Australia. Lancet 2, 569 – 573 (1996). 2. D. M. Eisenberg, R. B. Davis, S. L. Ettner, S. Appel, S. Wilkey, M. Van Rompay, et al. Trends in alternative medicine use in the United States 1990 – 1997: Results of a follow-up national survey. JAMA 280, 1569 – 1575 (1998). 3. A. Bensoussan, Complementary medicine — Where lies its appeal? Med. J. Aust. 170, 247 – 248 (1999). 4. R. B. Bradbury and D. E. White, Estrogens and related substances in plants. Vitam. Horm. 12, 207 (1954). 5. R. J. Miksicek, Commonly occurring plant flavonoids have estrogenic activity. Mol. Pharmacol. 44, 37 – 43 (1993). 6. R. J. Miksicek, Interaction of naturally occuring nonsteroidal estrogens with expressed recombinant human estrogen receptor. J. Steroid Biochem. 49, 153 – 160 (1994). 7. J. Kuhnau, The flavenoids. A class of semi-essential food components: Their role in human nutrition. World Rev. Nutr. Diet 24, 117 – 191 (1976). 8. P. M. Martin, K. B. Horwitz, D. S. Ryan, and W. L. McGuire, Phytoestrogen interaction with estrogen receptors in human breast cancer cells. Endocrinology 103, 1860 – 1867 (1978). 9. D. A. Shutt and R. I. Cox, Steroid and phyto-oestrogen binding to sheep uterine receptors in vitro. J. Endocrinol. 52, 299 – 310 (1972). 10. L. Markiewicz, J. Garey, H. Adlercreutz, and E. Gurpide, in vitro bioassays of non-steroidal phytoestrogens. J. Steroid Biochem. Mol. Biol. 45, 399 – 405 (1993). 11. G. G. Kuiper, J. G. Lemmen, B. Carlsson, J. C. Corton, S. H. Safe, P. T. van der Saag et al., Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 139, 4252 – 4263 (1998). 12. M. Casanova, L. You, K. W. Gaido, S. Archibeque-Engle, D. B. Janszen, and Hd’ A. Heck , Developmental effects of dietary phytoestrogens in Sprague – Dawley rats and interactions of genistein and daidzein with rat estrogen receptors alpha and beta in vitro. Toxicol. Sci. 51, 236 – 244 (1999). 13. A. R. Ibrahim and Y. J. Abul-Hajj, Aromatase inhibition by flavenoids. J. Steroid Biochem. Mol. Biol. 37, 257 – 260 (1990). 14. H. Aldercreutz, C. Bannwart, K. Wähälä, T. Mäkelä, G. Brunow, T. Hase, et al. Inhibition of human aromatase by mammalian lignins

628

15.

16.

17.

18.

19.

20.

21.

22. 23.

24.

25. 26.

27.

28.

29. 30.

31.

32.

33.

34.

35.

RICHARD PRINCE and isoflavonoid phytoestrogens. J. Steroid Biochem. Mol. Biol. 44(2), 147 – 153 (1993). C. Wang, T. Makela, T. Hase, H. Adlercreutz, and M. S. Kurzer, Lignans and flavenoids inhibit aromatase enzyme in human preadipocytes. J. Steroid Biochem. Mol. Biol. 50, 205 – 212 (1994). B. A. Evans, K. Griffiths, and M. S. Morton, Inhibition of 5 alpha-reductase in genital skin fibroblasts and prostate tissue by dietary lignans and isoflavenoids. J. Endocrinol. 147, 295 – 302 (1995). J. C. Le Bail, Y. Champavier, A. J. Chulia, and G. Habrioux, Effects of phytoestrogens on aromatase, 3 beta and 17 beta-hydroxysteroid dehydrogenase activities and human breast cancer cells. Life Sci. 66, 1281 – 1291 (2000). M. Loukovaara, M. Carson, A. Palotie, and H. Adlercreutz, Regulation of sex hormone-binding globulin production by isoflavenoids and patterns ofisoflavenoid conjugation in HepG2 cell cultures. Steroids 60, 656 – 661 (1995). W. M. Keung, Dietary estrogenic isoflavones are potent inhibitors of beta-hydroxysteroid dehydrogenase of P. tetosteronii. Biochem. Biophys. Res. Commun. 215,1137 – 1144 (1995). H. C. Jha, G. Von Recklinghausen, and F. Zilliken, Inhibition of in vitro microsomal lipid peroxidation by isoflavonoids. Biochem. Pharmacol. 9, 1367 – 1369 (1985). T. Akiyama, J. Ishida, S. Nakagawa, H. Ogawara, S. Watanabe, N. Itoh, et al. Genistein a specific inhibitor of tyrosine-specific protein kinases. J. Biol. Chem. 262, 5592 – 5595 (1987). G. Peterson, Evaluation of the biochemical targets of genistein in tumor cells. J. Nutr. 125, 784S – 789S (1995). G. H. Degen, Interaction of phytoestrogens and other environmental estrogens with prostaglandin synthase in vitro. J. Steroid Biochem. 35, 473 – 479 (1990). A. A. Franke, L. J. Custer, C. M. Cerna, and K. Narala, Rapid HPLC analysis of dietary phytoestrogens from legumes and from human urine. Proc. Soc. Exp. Biol. Med. 208, 18 – 26 (1995). R. C. Rossiter, Factors affecting the oestrogen content of subterranean clover pastures. Aust. Vet. J. 46, 141 – 145 (1970). K. D. R. Setchell, Phytoestogens: The biochemistry, physiology, and implications for human health of soy isoflavones. Am J. Clin. Nutr. 68S, 1333S – 1346S (1998). J. H. Weisburger, Approaches for chronic disease prevention based on current understanding of underlying mechanisms. Am. J. Clin. Nutr. 71(S6), 1710S – 1714S (2000). H. Adlercreutz, Phytoestrogens: Epidemiology and a possible role in cancer protection. Environ. Health Perspect. 103 (Suppl. 7), 103 – 112 (1995). H. Adlercreutz and W. Mazur, Phytoestrogens and Western diseases. Ann. Med. 29, 95 – 120 (1997). H. Adlercreutz, H. Honjo, A. Higashi, T. Fotsis, E. Hamalainen, T. Hasegawa, et al. Urinary excretion of lignans and isoflavonoid phytoestrogens in Japanese men and women consuming a traditional Japanese diet. Am. J. Clin. Nutr. 54, 1093 – 1100 (1991). H. Adlercreutz, H. Markkanen, and S. Watanabe, Plasma concentrations of phytooestrogens in Japanese men. Lancet 342, 1209 – 1210 (1993). P. D. Ross, H. Orimo, R. D. Wasnich, J. M. Vogel, C. J. MacLean, J. W. Davis, et al. Methodological issues in comparing genetic and environmental influences on bone mass. Bone Miner. 7, 67 – 77 (1989). Q. Wang, P. Ravn, S. Wang, K. Overgaard, C. Hassager, and C. Christiansen, Bone mineral density in immigrants from southern China to Denmark. A cross-sectional study. Eur. J. Endocrinol. 134, 163 – 167 (1996). A. F. Kardinaal, M. S. Morton, I. E. Bruggemann-Rotgans, and E. C. Van Beresteijn, Phytooestrogen excretion and rate of bone loss in postmenopausal women. Eur. J. Clin. Nutr. 52, 850 – 855 (1998). M. Kimara, Y. Arai, K. Shimoi, and S. Watanabe, Japanese intake of flavenoids and isoflavenoids from food. J. Epidemiol. 8, 168 – 175 (1998).

36. C. Huang, P. D. Ross, S. Fujiwara, J. W. Davis, R. S. Epstein, K. Kodoma, et al. Determinants of vertebral fracture prevalence among native Japanese women and women of Japanese descent living in Hawaii. Bone 18, 437 – 442 (1996). 37. P. D. Ross, H. Norimatsu, J. W. Davis, K. Yano, R. D. Wasnich, S. Fujiwara, et al. A comparison of hip fracture incidence among native Japanese, Japanese Americans and American caucasians. Am. J. Epidemiol. 133, 801 – 809 (1991). 38. R. G. Cumming and M. C. Nevitt, Calcium for the prevention of osteoporotic fractures in postmenopausal women. J. Bone Miner. Res. 12, 1321 – 1329 (1997). 39. A. Devine, R. A. Criddle, I. M. Dick, D. A. Kerr, and R. L. Prince, A longitudinal study of the effect of sodium and calcium intakes on regional bone density in postmenopausal women. Am. J. Clin. Nutr. 62, 740 – 745 (1995). 40. G. E. Kelly, C. Nelson, M. A. Waring, G. E. Joannou, and A. Y. Reeder, Metabolites of dietary (soya) isoflavones in human urine. Clin. Chim. Acta. 223, 9 – 22 (1993). 41. J. W. Lampe, M. C. Martini, M. S. Kurzer, H. Adlercreutz, and J. L. Slavin, Urinary lignan and isoflavonoid excretion in premenopausal women consuming flaxseed powder. Am. J. Clin. Nutr. 60(1), 122 – 128 (1994). 42. X. Xu, H. J. Wang, P. A. Murphy, L. Cook, and S. Hendrich, Daidzein is a more bioavailable soymilk isoflavone than is genistein in adult women. J. Nutr. 124(6), 825 – 832 (1994). 43. M. S. Morton, G. Wilcox, M. L. Wahlqvist, and K. Griffiths, Determination of lignans and isoflavonoids in human female plasma following dietary supplementation. J. Endocrinol. 142(2), 251 – 259 (1994). 44. S. Watanabe, M. Yamaguchi, T. Sobue, T. Takahashi, T. Miura, Y. Arai, et al. Pharmacokinetics of soybean isoflavones in plasma, urine and feces of men after ingestion of 60g baked soybean powder (kinako). J. Nutr. 128, 1710 – 1715 (1998). 45. B. H. Arjmandi, L. Alekel, B. W. Hollis, D. Amin, M. Stacewicz-Sapuntzakis, P. Guo, et al. Dietary soybean protein prevents bone loss in an ovariectomized rat model of osteoporosis. J. Nutr. 126, 161 – 167 (1996). 46. H. Kajiya, K. Okabe, F. Okamoto, T. Tsuzuki, and H. Soeda, Protein tyrosine kinase inhibitors increase cytosolic calcium and inhibit actin organisation as resorbing in rat osteoclasts. J. Cell Physiol. 183, 83 – 90 (2000). 47. Y. H. Gao and M. Yamaguchi, Suppressive effect of genistein on rat bone osteoclasts: Apoptosis is induced through Ca2 – signaling. Biol. Pharm. Bull. 22, 805 – 809 (1999). 48. Y. H. Gao and M. Yamaguchi, Anabolic effect of daidzein on cortical bone in tissue culture: Comparison with genistein effect. Mol. Cell. Biochem. 194, 93 – 97 (1999). 49. M. Yamaguchi and Y. H. Gao, Anabolic effects of genistein and genestin on bone metabolism in the femoral-metaphyseal tissues of elderly rats: The genistein effect is enhanced by zinc. Mol. Cell. Biochem. 178, 377 – 382 (1998). 50. H. K. Yoon, K. Chen, D. I. Baylink, and K. H. Lau, Differential effects of two protein tyrosine inhibitors, tyrphostin and genistein, on human bone cell proliferation as compared with differentiation. Calcif. Tissue Int. 63, 243 – 249 (1998). 51. J. E. Thigpen, K. D. Setchell, K. B. Ahlmark, J. Locklear, T. Spahr, G. F. Caviness, et al. Phytoestrogen content of purified, open- and closed-formula laboratory animal diets. J. Nutr. Sci. Vitam. 49, 530 – 536 (1999). 52. E. Harrison, A. Adjei, C. Ameho, S. Yamamoto, and S. Kono, The effect of soybean protein on bone loss in a rat model of postmenopausal bone loss. J. Nutr. Sci. Vitam. 44, 257 – 268 (1998). 53. P. Fanti, M. C. Monier-Faugere, Z. Geng, J. Schmidt, P. E. Morris, D. Cohen, et al. The phytoestrogen genistein reduces bone loss in short-term ovariectomised rats. Osteoporosis Int. 8, 274 – 281 (1998). 54. B. H. Arjmandi, R. Birnbaum, N. V. Goyal, M. J. Getlinger, S. Juma, L. Alekel, et al. Bone-sparing effect of soy protein in ovarian hor-

CHAPTER 71 Phytoestrogen and Other Phytochemical Effects on Bone

55.

56.

57.

58.

59.

60.

61.

62.

63.

64. 65.

66.

67.

68.

69.

70.

71.

72.

mone-deficient rats is related to its isoflavone content. Am. J. Clin. Nutr. 68, 1364S – 1368S (1998). B. H. Arjmandi, M. J. Getlinger, N. V. Goyal, L. Alekel, C. M. Hasler, S. Juma, et al. Role of soy protein with normal or reduced isoflavone content in reversing bone loss induced by ovarian hormone deficiency in rats. Am. J. Clin. Nutr. 68, 1358S – 1363S (1998). Y. Ishimi, C. Miyaura, M. Ohmura, Y. Onoe, T. Sato, Y. Uchiyama, et al. Selective effects of genistein, a soybean isoflavone, on B-lymphopoiesis and bone loss caused by estrogen deficiency. Endocrinology 140, 1893 – 1900 (1999). Y. Ishimi, N. Arai, X. Wang, J. Wu, K. Umegaki, C. Miyaura, et al. Difference in effective dosage of genistein on bone and uterus in ovariectomized mice. Biochem. Biophys. Res. Commun. 274, 697 – 701 (2000). J. J. Anderson, W. W. Ambrose, and S. C. Garner, Biphasic effects of genistein on bone tissue in the ovariectomized, lactating rat model. Proc. Soc. Exp. Biol. Med. 217, 345 – 350 (1998). T. Toda, T. Uesugi, K. Hirai, H. Nukaya, K. Tsuji, and H. Ishida, New 6-0-acyl isoflavone glycosides from soybeans fermented with Bacillus subtilis (natto). 6-0-succinylated isoflavone glycosides and their preventive effects on bone loss in ovariectomised rats fed a calcium deficient diet. Biol. Pharm. Bull. 22, 1193 – 1201 (1999). M. Yamaguchi, H. Kakuda, Y. H. Gao, and Y. Tsukamoto, Prolonged intake of fermented soybean (natto) diets containing vitamin K2 (menaquinone-7) prevents bone loss in ovariectomised rats. J. Bone Miner. Res. 18, 71 – 76 (1999). S. M. Potter, J. A. Baum, H. Teng, R. J. Stillman, N. F. Shay, and J. W. Erdman Jr, Soy protein and isoflavones: their effects on blood lipids and bone density in postmenopausal women. Am. J. Clin. Nut 68, 1375S – 1379S (1998). J. W. Anderson, B. M. Johnstone, and M. E. Cook-Newell, Metaanalysis of the effects of soy protein intake on serum lipids. N. Engl. J. Med. 333, 276 – 282 (1995). J. Y. L. Reginster, Ipriflavone: Pharmacological properties and usefulness in postmenopausal osteoporsis. Bone Miner. 23, 223 – 232 (1993). I. Yamazaki, Effects of ipriflavone on the reponse of uterus and thyroid to estrogen. Life Sci. 38, 757 – 761 (1986). G. B. Melis, A. M. Paoletti, A. Cagnacci, L. Bufalino, A. Spinetti, M. Gambacciani, et al. Lack of any estrogenic effect of ipriflavone in postmenopausal women. J. Endocrinol. Invest. 15, 755 – 761 (1992). K. Monstory, L. Vereczkey, F. Levai, and I.Szatmari, Ipriflavone as an inhibitor of human cytochrome P450 enzymes. Br. J. Pharmacol. 123, 605 – 610 (1998). I. Morita, K. Sakaguchi, T. Kurachi, and S. Murota, Ipriflavone inhibits murine osteoclast formation in vitro. Calcif. Tissue Int. 51, S7 – S10 (1992). M. Tsuda, T. Kitazaki, T. Ito, and T. Fujita, The effect of ipriflavone (TC-80) on bone resorption in tissue culture. J. Bone Miner. Res. 1, 207 – 211 (1986). B. H. Arjmandi, R. S. Birnbaum, S. Juma, E. Barengolts, and S. C. Kukreja, The synthetic phytoestrogen, ipriflavone, and estrogen prevent bone loss by different mechanisms. Calcif. Tissue Int. 66, 61 – 65 (2000). D. Agnusdei, F. Zacchei, S. Bigazzi, C. Cepollaro, P. Nardi, M. Montagnini, et al. Metabolic and clinical effects of ipriflavone in established post-menopausal osteoporosis. Drugs Exp. Clin. Res. 15, 97 – 104 (1989). M. Gambacciani, A. Spinetti, B. Cappagli, F. Taponeco, R. Felipetto, D. Parrini, et al. Effects of ipriflavone administration on bone mass and metabolism in ovariectomised women. J. Endocrinol. Invest. 16, 333 – 337 (1993). S. Adami, L. Bufalino, R. Cervetti, C. Di Marco, O. Di Munno, L. Fantasia, et al. Ipriflavone prevents radial bone loss in postmenopausal wmen with low bone mass over 2 years. Osteoporosis Int. 7, 119 – 125 (1997).

629 73. D. Agnusdei and L. Bufalino, Efficacy of ipriflavone in established osteoporosis and long-term safety. Calcif. Tissue Int. 61, 23 – 27 (1997). 74. A. B. Kovacs, Efficacy of ipriflavone in the prevention and treatment of postmenopausal osteoporosis. Agents Actions 41, 86 – 87 (1994). 75. C. Gennari, S. Adami, D. Agnusdei, L. Bufalino, R. Cervetti, G. Crepaldi, et al. Effects of chronic treatment with ipriflavone in postmenopausal women with low bone mass. Calcif. Tissue Int. 61, S19 – S22 (1997). 76. D. Agnusdei, G. Crepaldi, G. Isaia, G. Mazzuoli, S. Ortolani, M. Passeri, et al. A double-blind, placebo-controlled trial of ipriflavone for prevention of postmenopausal spinal bone loss. Calcif. Tissue Int. 61, 142 – 147 (1997). 77. C. Gennari, D. Agnusdei, G. Crepaldi, G. Isais, G. Mazzuoli, S. Ortolani, et al. Effect of ipriflavone a synthetic derivative of natural isoflavones — On bone mass loss in the early years after menopause. Menopause 5, 9 – 15 (1998). 78. H. Ohta, S. Komukai, K. Makita, T. Masuzawa, and S. Nozawa, Effects of 1-year ipriflavone treatment on lumbar bone mineral density and bone metabolic markers in postmenopausal women with low bone mass. Horm. Res. 51, 178 – 183 (1999). 79. G. B. Melis, A. M. Paoletti, R. Bartolini, M. Tosi Balducci, G. B. Massi, V. Bruni, et al. Ipriflavone and low doses of estrogens in the prevention of bone mineral loss in climacterium. Bone Miner. 19, S49 – S56 (1992). 80. M. Gambacciani, M. Ciaponi, B. Cappagli, L. Piaggesi, and A. R. Genazzani, Effects of combined low dose of the isoflavone derivative ipriflavone and estrogen replacement on bone mineral density and metabolism in postmenopausal women. Maturitas 28, 75 – 81 (1997). 81. R. J. Miksicek, In situ localization of the estrogen receptor in living cells with the fluorescent phytoestrogen coumestrol. J. Histochem. Cytochem. 41, 801 – 810 (1993). 82. N. Tsutsumi, Effect of coumestrol bone metabolism in organ culture. Biol. Pharm. Bull. 18, 1012 – 1015 (1995). 83. N. Tsutsumi, K. Kawashima, N. Arai, H. Nagata, M. Kojima, A. Ujiie, et al. In vitro effect of KCA-098, a derivative of coumestrol, on bone resorption of fetal rat femurs. Bone Miner. 24, 201 – 209 (1994). 84. M. Kojima, N. Tsutsumi, H. Nagata, F. Itoh, A. Ujiie, K. Kawashima, et al. Effect of KCA-098, a new benzofuroquinolone derivative, on bone mineral metabolism. Biol. Pharm. Bull. 17, 504 – 508 (1994). 85. N. Tsutsumi, K. Kawashima, H. Nagata, A. Ujiie, and H. Endo, Effects of KCA-012 on bone metabolism in organ culture. Jpn. J. Pharmacol. 67, 169 – 171 (1995). 86. J. A. Dodge, A. L. Glasebrook, D. E. Magee, D. L. Phillips, M. Sato, L. L. Short, et al. Environmental estrogens: Effects on cholesterol lowering and bone in the ovariectomised rat. J. Steroid Biochem. Mol. Biol. 59, 155 – 161 (1996). 87. C. R. Draper, M. J. Edel, I. M. Dick, A. G. Randell, G. B. Martin, and R. L. Prince, Phytoestrogens reduce bone loss and bone resorption in oophorectomised rats. J. Nutr. 127, 1795 – 1799 (1997). 88. N. Sathyamoorthy, T. T. Y. Wang, and J. M. Phang, Stimulation of PS2 expression by diet-derived compounds. Cancer Res. 54(4), 957 – 961 (1994). 89. S. R. Cummings, W. S. Browner, D. Bauer, K. Stone, K. Ensrud, S. Jamal, et al. Endogenous hormones and the risk of hip and vertebral fractures among older women. N. Engl. J. Med. 339, 733 – 738 (1998). 90. H. Li, J. K. Prasain, Y. Tezuka, T. Namba, T. Miyahara, S. Tonami, et al. Antiosteoporotic activity of the stems of Sambucus sieboldiana. Biol. Pharm. Bull. 21, 594 – 598 (1998). 91. L. J. Jiu, N. Morikawa, N. Omi, I. Ezawa, The effect of tochu bark on bone metabolism in the rat model with ovariectomised osteoporosis. J. Nutr. Sci. Vitam. 40, 261 – 273 (1994). 92. M. Miyamoto, Y. Matsushita, A. Kiyokawa, A. Fukuda, C. Fukuda, Y. Iijima, et al. Prenylflavonoids: A new class of non-steroidal phytoestrogens (Part 2). Estrogenic effects of 8-isopentenylnaringenin on bone metabolism. Planta Med. 64, 516 – 519 (1998).

CHAPTER 72

Bisphosphonates in the Management of Postmenopausal Osteoporosis SOCRATES E. PAPAPOULOS

I. II. III. IV. V. VI.

Department of Endocrinology and Metabolic Diseases, University of Leiden Medical Center, 2333 ZA Leiden, The Netherlands

Introduction Pharmacokinetics – Pharmacodynamics Treatment Protocols Effects on Bone Mineral Density Antifracture Efficacy Special Issues of Bisphosphonate Therapy and Unanswered Questions

VII. Long-Term Effects on Bone Metabolism VIII. Adverse Effects IX. Conclusions References

I. INTRODUCTION

resorption such as Paget’s disease of bone and malignancyassociated hypercalcemia (see also Chapter 16). Increasing understanding of bone physiology and pathology as well as of the pharmacological properties of bisphosphonates led to their application to the treatment of other skeletal disorders (Table 1). Bisphosphonates were first administered to patients with osteoporosis in the early 1970s, but this application started to be systematically explored in the mid-1980s. Although the indolent nature of the disease, requiring long-term observations, is partly responsible for this delay, issues related to bisphosphonate pharmacology and to the complexity of osteoporosis have contributed to it. Bisphosphonate pharmacology

Bisphosphonates are synthetic compounds that have high affinity for calcium-containing crystals, concentrate preferentially in the skeleton, and affect bone surfacerelated processes. The first bisphosphonate was synthesized in the 19th century but the relevance of these compounds to clinical medicine was recognized in the 1960s. They were initially developed as inhibitors of growth and dissolution of calcium crystals but were subsequently found to inhibit osteoclast-mediated bone resorption. Because of this action, bisphosphonates were used initially in the management of conditions characterized by excessive osteoclastic bone

OSTEOPOROSIS, SECOND EDITION VOLUME 2

631

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

632 TABLE 1

SOCRATES E. PAPAPOULOS

Established and Potential Clinical Indications of Bisphosphonate Therapy

Paget’s disease of bone Malignancy–associated hypercalcemia Metastatic bone disease (breast, prostate, other) Multiple myeloma Osteoporosis Primary (postmenopausal, idiopathic, male, juvenile) Steroid–induced Immobilization Posttransplanation Osteogenesis imperfecta Mastocytosis Hyperparathyroidism Rheumatoid arthritis Algodystrophy, transient osteoporosis of the hip Asceptic loosening of hip prostheses Fibrous dysplasia Sterno–costo–clavicular hyperostosis Gaucher’s disease Engelman’ disease Periodontitis Nonskeletal disorders Ectopic calcification/ossification Urolithiasis

presents a challenge to basic scientists and clinicians due to the specific properties of these compounds. Bisphosphonates concentrate on active bone surfaces, bind strongly to hydroxyapatite crystals, are locally released from the skeleton, are taken up by bone cells, can remain embedded in bone for long periods, and can be again slowly released with bone remodeling. The preclinical pharmacology of the bisphosphonates was reviewed in Chapter 16. In this chapter pharmacological issues relevant to their application to the management of postmenopausal osteoporosis are discussed. Their use in other forms of osteoporosis (e.g., steroid-induced, male, childhood) is reviewed in the respective Chapters.

II. PHARMACOKINETICS – PHARMACODYNAMICS Classically, for the development of pharmacological interventions for clinical use the pharmacokinetics of the compounds of interest should be defined. The intestinal absorption of orally administered bisphosphonates is poor, accounting for about 0.6 to 3% of the administered dose. Absorption occurs across intestinal cells, probably by the paracellular route (along the cells) rather than transcellu-

larly (across the cells) [1,2]. This low absorption is attributed to the highly negative charge of phosphonates which inhibits their diffusion through lipophilic membranes [3]. Bisphosphonates have a short plasma half-life. Plasma disappearance is multiexponential and practically all bisphosphonate is cleared from the circulation within 6 to 10 h after administration [4 – 6]. About half of the administered dose concentrates in the skeleton. The fractional skeletal retention of bisphosphonate depends on the rate of bone turnover and the availability of active sites as well as on the structure, with hydroxybisphosphonates generally showing higher binding. The remaining bisphosphonate is excreted unaltered in urine. Up until now no bisphosphonate metabolites have been identified in vivo. Bisphosphonates are embedded in the skeleton where they remain for long periods. Elimination from this compartment is extremely slow. The terminal half-life has been calculated for a number of them and can be as long as 10 years in humans [7,8]. There are certain limitations in the assessment of bisphosphonate pharmacokinetics. For example, measurement of their concentrations in biological fluids has been technically difficult due to their very low circulating concentrations because of poor intestinal absorption and the low doses, especially of newer bisphosphonates, given to patients with osteoporosis. Even with sensitive assays or use of radiolabeled bisphosphonates, classical pharmacokinetic concepts and analyses cannot be applied because of their long retention in the skeleton, which makes a pharmacokinetic steady state difficult to obtain. In addition, in contrast to the bisphosphonate present on the bone surface, that embedded in bone is biologically inert and at present there are no valid models which can distinguish between these two compartments in vivo, particularly if the amount of locally released bisphosphonate by resorption from bone mineral is also accounted for. Therefore, analysis of the pharmacological properties of bisphosphonates and design of appropriate therapeutic regimens for bone diseases depend to a large extent on the interpretation of pharmacodynamic information. This can be accurately obtained in vivo due to the specific property of bisphosphonates to concentrate selectively in bone and to suppress bone resorption, a biological response that can be readily quantified by measuring biochemical indices of bone resorption [9]. Suppression of bone resorption depends on the dose, the route of administration, and the potency of the bisphosphonate as well on the nature of the skeletal disease. With intravenous administration of potent bisphosphonates, the earliest change is detected after 24 to 48 h. Figure 1 depicts an example of a patient with Paget’s disease and increased bone turnover treated with daily infusions of bisphosphonate for 10 days. Treatment induced a rapid decrease in urinary hydroxyproline, cross-linked N-telopeptide of collagen type I, and calcium excretion indicating effective suppression of bone resorption. During this short period, bone formation

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

633

and the intestinal absortion of calcium (calcitriol). During this period calcium balance increases greatly and the magnitude of this increase is limited only by the capacity of the intestine to absorb calcium. As equilibrium between bone resorption and bone formation is restored, the transient changes in calcium homeostasis revert to baseline and a new steady state is obtained at a lower level of bone turnover (Fig. 2). Despite the often marked reductions in bone resorption during treatment of patients with high bone turnover with suppressive doses of potent bisphosphonates, the adaptive changes of calcium metabolism prevent the development of symptomatic hypocalcemia. Analysis of the kinetics of bone resorption and formation following bisphosphonate therapy not only simplified the treatment of patients with Paget’s disease but also emphasized the importance of dose and duration of treatment for long-term responses [16]. In addition, the magnitude of the adaptation of calcium metabolism to the needs of skeletal homeostasis was demonstrated

FIGURE 1 Changes in biochemical indices of bone turnover in a patient with Paget’s disease treated with daily intravenous infusions of a nitrogen -containing bisphosphonate for 10 days. U-OHP urinary hydroxyproline; U-NTx, urinary crosslinked N-telopeptide of collagen type I; S-Alk Phosph, serum alkaline phosphatase activity; AP, serum alkaline phosphatase activity; OC, osteocalcin. During the 10-day treatment period urinary calcium excretion decreased from 4.1 to 0.83 mmol/24h. was not affected, as illustrated by serum alkaline phosphatase activity, which did not change. Serum osteocalcin values even increased, as previously shown to occur during the early phase of treatment of Paget’s disease with bisphosphonate [10]. Serum alkaline phosphatase activity (and osteocalcin) will decrease later as a result of the coupling between bone resorption and bone formation so that after 3 to 6 months a new equilibrium between resorption and formation will be reached. Thus, bisphosphonate therapy induces a secondary suppression of bone formation even if treatment is given only for a few days [12 – 15]. In the initial phase of bisphosphonate treatment there is a dissociation between bone resorption and formation that leads to positive bone balance. The forced uncoupling of bone resorption and bone formation and the resulting positive bone balance are associated with marked changes in calcium metabolism. There is a fall in serum calcium concentrations which stimulates the secretion of PTH and the renal production of calcitriol. These in turn increase the renal tubular reabsorption of calcium (PTH)

FIGURE 2

Changes in serum alkaline phosphatase activity (APase), urinary hydroxyproline excretion (OHP), serum calcium concentrations, urinary calcium excretion, and calcium balance during treatment of patients with Paget’s disease with oral pamidronate. Note the logarithmic scale of the horizontal axis. Reprinted with permission from Frijlink et al. [11].

634 [17]. In this respect, bisphosphonates need also be considered as potent pharmacological tools for the study of bone metabolism. It should be noted that the changes in calcium metabolism following bisphosphonate therapy are not usually seen after treatment with etidronate, which has a weak effect on calcium homeostasis and affects the renal handling of phosphate differently [18,19]. During treatment of high bone turnover diseases with potent bisphosphonates there is an additional transient decrease in serum phosphate concentrations and in renal tubular reabsorption of phosphate due to the secondary increases in serum PTH concentrations. Etidronate therapy may induce an increase rather than a decrease in serum phosphate concentration, probably due to a direct effect on renal phosphate transport and a lack of PTH response. Such changes during etidronate treatment have been interpreted as early evidence of unwanted effects on the mineralization of newly formed bone [20]. In osteoporosis there is an imbalance between bone formation and bone resorption resulting in bone loss with every remodeling cycle. When this is accompanied by an increase in the activation of new bone remodeling units (high bone turnover), more bone will be lost within the same period. In addition, this will adversely affect the structure of the trabeculae and will consequently increase further the risk of fracture. It is, therefore logical that agents which suppress bone resorption and reduce the rate of bone turnover, such as the bisphosphonates, will benefit patients with osteoporosis. This rationale must, however, be put into perspective and be considered together with our knowledge of the pathophysiology of osteoporosis, of the cellular dynamics of bone tissue, and of the specific properties of the bisphosphonates. In an early study radiocalcium kinetics and calcium balance were assessed in patients with osteoporosis before, 6 months after treatment with hPTH (1 – 34) (100 g/day, sc) and after an additional 6 months of continuous treatment with oral pamidronate 150 mg/day [21]. PTH increased the rate of bone turnover but had no clear effect on calcium balance (Fig. 3). In contrast, the bisphosphonate treatment that followed reduced the rate of bone turnover and induced a marked increase in calcium balance. This is the response expected by an antiresorptive agent given under these conditions and resembles that obtained in patients with Paget’s disease and high rates of bone turnover. An increased rate of bone turnover is, however, not a universal finding in elderly patients with osteoporosis. Although bone loss occurs in such patients, this is usually not excessive compared to that occurring immediately after the menopause or in some patients who, for reasons not yet clear, present with high rates of bone turnover (the so-called high turnover osteoporosis). Furthermore, suppression of bone resorption by bisphosphonates will be followed by secondary suppression of bone formation leading to a state of lower bone turnover. Such changes will attenuate bone loss and will induce transient

SOCRATES E. PAPAPOULOS

FIGURE 3

Bone turnover assessed by radiocalcium technique (top) and calcium balance (bottom) in patients with osteoporosis before treatment (basal), after 6 months treatment with hPTH (1 – 34) (PTH), and after a further 6 months of treatment with oral pamidronate (APD). Means  SEM are shown. Adapted with permission from Vismans [21].

gains in bone mass but may also carry certain risks if suppression is excessive. It has been shown, for example, that rigorous long-term suppression of bone turnover in animals by bisphosphonate treatment was associated with increased risk of spontaneous fractures [22]. Thus, in the treatment of osteoporotic patients with bisphosphonates issues related to the dose of these agents as well as to the magnitude and duration of the response need to be carefully considered. In addition, the long-term retention of bisphosphonate in the skeleton, which may in theory lead to long-term cumulative effects on bone metabolism, needed to be addressed. These considerations, which just highlight the complexity of patient treatment, necessitated the design of extensive and prolonged pilot studies which helped to formulate our current approaches to the use of bisphosphonates in osteoporosis.

III. TREATMENT PROTOCOLS The surprising finding of the early studies of patients with Paget’s disease treated daily with oral pamidronate was that the calcium balance remained higher than basal after the

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

new equilibrium between bone resorption and bone formation had been reached (see Fig. 2). This finding implied persistence of the effect of treatment. In an initial study seven patients with osteoporosis were treated with oral pamidronate, 600 mg/day [23]. There was suppression of bone resorption associated with a significant increase in calcium balance (from 0.49  1.0 to 4.99  0.53 mmol/day) after 15 days of treatment. Thus the short-term effect of bisphosphonate therapy could be reproduced in patients with normal bone turnover as a result of the forced transient uncoupling of bone resorption and formation. These doses were, however, high for long-term treatment of patients in whom bone turnover was not high. Heaney and Saville [24] and Jowsey et al. [25] had earlier examined the effect of high dose etidronate (up to 20 mg/kg/day) on calcium balance and bone histology in patients with osteoporosis. Bone turnover was suppressed by about 50% and although there was a slight, but significant increase in calcium balance after 1 year this was associated with histological evidence of osteomalacia. The issue of dose and mode of bisphosphonate administration was addressed in animal studies. Reitsma et al. [26] treated growing rats with daily subcutaneous injections of pamidronate and followed the changes in bone resorption and in calcium balance. Treatment suppressed bone resorption dose-dependently and significantly increased calcium retention. More importantly, with all doses used, suppression of resorption reached a plateau that was also dose-dependent, but did not decrease further despite the continuous administration of the drug (Fig. 4). These results suggested for the first time that it may be possible to suppress bone resorption mildly and to induce significant increases in calcium balance with low-dose bisphosphonate given daily. Moreover, they demonstrated that the daily administration of the bisphosphonate is not necessarily accompanied by progressive suppression of bone resorption and thus the accumulation of the bisphosphonate in the skeleton is not associated with a cumulative effect on bone metabolism. This pattern of response has now been repeatedly shown in humans treated for prolonged periods with oral bisphosphonates given daily. In a series of pilot studies in Leiden, 14 osteoporotic patients were treated with lower oral doses of pamidronate (150 mg/day). Treatment suppressed bone resorption mildly (decrease in urinary hydroxyproline excretion by 25 to 30%) and increased calcium balance from  0.72  0.59 to 1.33  0.87 mmol/day after 1 year. This positive balance may seem small but it corresponds to a yearly gain in bone mass of about 3% in a patient with an osteoporotic skeleton, confirmed by measurements of the BMC of the lumbar spine [23]. These studies suggested that in the management of patients with osteoporosis doses considerably lower than those given in other conditions (e.g., Paget’s disease,

635

FIGURE 4

Changes in urinary hydroxyproline excretion (OHP) in rats treated with daily injections of three different doses of pamidronate (closed circles) or vehicle (open circles). Reprinted with permission from Reitsma [26].

metastatic bone disease) can be used effectively and that the evaluation of the effects of the bisphosphonates should be of long duration. In studies of short duration, interpretation of remodeling transients induced by treatment rather than the effects of bisphosphonates on the steady-state balance of resorption and formation at the bone remodeling unit may lead to wrong conclusions about their efficacy and mechanism of action. Consequently, this protocol, initiated by Bijvoet in Leiden, was used in the development of the more potent nitrogen-containing bisphosphonates, alendronate and risedronate. The second proposed treatment protocol consisted of the intermittent administration of the bisphosphonate. Intermittent therapy was originally developed on the basis of the ADFR concept. Anderson et al. [27] treated five patients for 9 to 24 months according to this regimen and reported improvement in trabecular bone volume on iliac crest biopsies and in bone remodeling dynamics. They used oral phosphate as activator of osteoclast activity (A), followed by etidronate for 2 weeks as depressor (D), followed by 70-day

636 treatment-free period (F). The cycle was then repeated (R). Subsequent studies, however, with more patients from the same institution did not reach the same conclusions and it was suggested that this treatment regimen results in shortterm improvement in trabecular bone mass while showing no long-term improvement in bone remodeling [28]. This view was confirmed in a clinical trial which failed to show any additional benefit on bone mineral density when a stimulator of osteoclast activity was added to a treatment regimen with intermittent etidronate [29]. Although it may argued that the latter study did not apply the ADFR concept as formulated by Frost [30], there is at present no strong evidence to suggest that bisphosphonates as part of an ADFR regimen will confer any advantage over the use of the drug alone. Independently of the ADFR concept, the idea of discontinuous administration of bisphosphonates led to the design of various treatment regimens with different compounds; the theoretical background being that a short period of bisphosphonate therapy will suppress bone resorption that will be followed by a period during which bone formation will proceed unopposed, resulting in positive bone balance. This approach was also thought to have some practical advantages as the skeleton will be exposed for shorter periods to the bisphosphonate and hence less compound will concentrate in bone, reducing potential risks associated with drug accumulation. This mode of application is particularly relevant for etidronate because in this way the chances to induce defective mineralization of newly formed bone could be reduced. Both these protocols have been tested in the management of patients with osteoporosis and sufficient data are currently available that allow conclusions about their validity.

IV. EFFECTS ON BONE MINERAL DENSITY Several bisphosphonates were or are being developed for the management of osteoporosis. Independent of potency, mechanism of action, or mode of administration, bisphosphonate treatment regimens induce variable suppression of bone turnover accompanied by increases in bone mineral density predominantly of the spine. The most consistent results were obtained with the daily oral administration of nitrogen-containing bisphosphonates. Early randomized or open studies with cyclic etidronate, as part or not of a coherence therapy regimen, gave conflicting results on BMD; increases, decreases, or no changes were reported [31 – 33]. However, the first placebo-controlled studies of cyclic etidronate in postmenopausal women with osteoporosis provided consistent results [29,34]. There was a significant modest increase in lumbar spine BMD of about 5% that occurred mainly during the first year of treatment, consistent with the reduction of the remodeling space expected in response to an antiresorptive agent. This

SOCRATES E. PAPAPOULOS

gain was maintained for at least 5 years [35,36], though extensions of these studies reported further small increases up to 7 years of treatment [37]. The gain in BMD at the spine was not due to redistribution of calcium in the skeleton as BMD of primarily appendicular sites either did not change or showed small increases. Similar results were reported in studies of shorter duration with other bisphosphonates given intermittently by the oral route (e.g., clodronate, pamidronate) [38 – 43]. Bisphosphonates also were given intermittently by the intravenous route. Passeri et al. [44] reported an increase in spine BMD of about 9% after 1 year with intravenous alendronate 5 mg per day for 2 consecutive days every 3 months, while alendronate 30 mg over 4 days increases spine BMD by 5.3% after 1 year [8]. Thiebaud et al. [45] in a randomized study comparing pamidronate to sodium fluoride, treated 16 osteoporotic patients with a single intravenous infusion of 30 mg pamidronate every 3 months. This regimen induced signifi cant increases in BMD of the spine and femoral neck of about 10.5 and 4.5%, respectively, after 2 years. Clodronate given once every 6 months (1800 mg) induced a small increase in BMD after 2 years [40]. Ibandronate was given by intravenous injections at various doses for 1 year or 0.5 and 1 mg every 3 months for 3 years in a large placebo-controlled study. All ibandronate regimens with doses higher than 0.5 mg every 3 months induced signifi cant increases in BMD [46,47]. Numerous studies with nitrogen-containing bisphosphonates given daily to osteoporotic women for periods up to 7 years showed increases in BMD at all relevant skeletal sites. Most of the results have been obtained with alendronate, with which there is also the longest experience from controlled trials, and with risedronate. There are also data with oral ibandronate and pamidronate. Optimal reported oral doses for BMD effects are pamidronate, 150 mg/day, alendronate, 10 mg/day, risedronate, 5.0 mg/day and ibandronate, 2.5 mg/day. It should be noted that the dose range tested for pamidronate and risedronate was limited. Comparisons among bisphosphonates for their effect on BMD cannot be made because they were never tested against each other. It appears, however, that alendronate, 10 mg/day, and pamidronate, 150 mg/day, induce the greatest increases.

V. ANTIFRACTURE EFFICACY A. Trial Design and Evaluation of Results The aim of any pharmacological intervention in osteoporosis is the decrease in the risk of fractures in patients who have not yet sustained an osteoporotic fracture or of the progression of the disease in patients with fragility fractures [48]. When interpreting the results of studies examining the

637

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

antifracture efficacy of bisphosphonates (and other antiosteoporotic treatments) certain issues related to the quality of the trials need to be considered [49] (See also Chapter 64). First is the design of the study; randomized clinical trials (RCTs) with blinding of patient and investigator being the optimal design. In addition, the hypothesis to be tested and the type of analysis of the fracture incidence should be prespecified. An intention-to-treat analysis is very conservative but is also the most objective and statistically sound. Second is the duration of the clinical study; this should take into consideration the pathophysiology of the disease, in particular the characteristics of rates of bone loss and of fractures. The duration should be at least 2 years, preferably 3 or more years. Trials of shorter duration can provide results which are due to transient changes in bone remodeling rather than being representative of the steady state of bone metabolism achieved with treatment. Third, the number of patients should be sufficient to provide not only statistically, but also clinically relevant differences between treatments. Fourth, documentation of fractures should be done in an objective way. Nonvertebral fractures are easier to document because they are associated with symptoms; patients present to physicians and X-rays are made. This is not the case with vertebral fractures, two-thirds of which are usually asymptomatic. A number of studies, however, have documented that vertebral deformities increase the risk of new osteoporotic fractures and are associated with increased morbidity and mortality [50,51]. A decrease in any height of the vertebral body by 20%, reducing it by at least 4 mm, is currently considered the gold standard for defining an incident vertebral deformity. Fifth, the number of patients with fractures rather than the number of fractures should be counted; counting the latter may give a false impression of effectiveness and inflate the result because fractures are interrelated events. Finally, an issue which has attracted rather limited attention in the field of osteoporosis is the drop-out rate. For continuous variables such as BMD and biochemical indices of bone turnover, drop-out rates may not significantly

modify the results. However, for discrete variables such as fractures, high dropout rates can inflate the positive effect of a treatment even if these rates are the same in both groups. Currently there is no clear recommendation about a maximally acceptable dropout rate and there are suggestions that this should be less than 20% for the duration of the trial. This is, however, relative because it does not account for studies in special populations in which high dropout rates are expected independently of all efforts for retaining patients in the study. Such an example is the study of Chapuy and colleagues [52] which examined the effect of vitamin D and calcium on the incidence of hip fractures in women residing in nursing homes with a mean age of 84 years. In such populations, dropout rates higher than 20% are inevitable. However, this should not be the case in studies of younger women with osteoporosis living in the community. Randomized controlled trials being the best experimental design, metaanalyses of such trials provide the best evidence of the efficacy and, perhaps more importantly, the consistency of the effect of an intervetion.

B. Effects on Fracture Risk Studies in various animal models showed that bisphosphonates preserve bone quality and improve the biomechanical properties of bone. At present RCTs with assessment of fracture incidence, involving about 30,000 women with postmenopausal osteoporosis, have been reported with bisphosphonates that differ considerably in potency and mechanism of action (Table 2). Inadequate to excellent studies with variable design and end point (primary or secondary) the reduction in the incidence of osteoporotic fractures have been performed allowing the application of the principles of evidence-based medicine to the assessment of their antifracture efficacy. Results pertinent to studies with specific bisphosphonates for which fracture rates have been assessed under controlled conditions are summarized below.

TABLE 2 Randomized Clinical Trials of Bisphosphonates (BP) in Women with Postmenopausal Osteoporosis with Fracture Assessment BP

Patients (no.)

Oral

IV

Mode of administration

Clodronate

593



Daily

Etidronate

495



Intermittently

Tiludronate

2316



Intermittently

Alendronate

9969



Ibandronate

2862

Pamidronate

297



Daily/intermittent

Risedronate

13,015



Daily

Daily 

Intermittently

638 1. NONNITROGEN-CONTAINING BISPHOSPHONATES a. Clodronate The first RCT with this bisphosphonate with assessment of fracture incidence was recently reported [53]. Oral clodronate 800 mg/day or placebo were given daily to 593 women with osteoporosis. They all received calcium 500 mg/day. After 3 years, 23% of women in the placebo group had incident vertebral fractures compared to 12.7% of women in the clodronate group; a significant relative risk reduction of 46%. Treatment was effective in women with and without prevalent vertebral fractures. Clodronate did not reduce the incidence of nonvertebral fractures, but the numbers were small. The quality of the trial cannot be yet assessed as a full report of the study is pending. b. Etidronate The efficacy of cyclic etidronate to reduce the incidence of vertebral fractures in women with postmenopausal osteoporosis was examined in two studies [29,34]. Assessment of fracture incidence was not a primary endpoint of these studies and there was no optimal assessment of new vertebral fractures. The major controlled trial failed to show concrete evidence of antifracture efficacy after 3 years of treatment [36]. A post hoc analysis, however, revealed that this regimen effectively reduced the incidence of new vertebral fractures in postmenopausal women with severe osteoporosis. This was confirmed by a recent metaanalysis, presented at the NIH Consensus Conference in Washington DC (2000). The lower incidence of vertebral fractures was reported to be sustained for up to 7 years of treatment [37]. There are no prospective data on the effect of cyclic etidronate in reducing the risk of nonvertebral fractures but a postmarketing survey suggested that this may be the case [54]. Although the etidronate studies can easily be criticized, it should be mentioned that at the time these were planned and performed they were the best conducted intervention trials in osteoporosis and provided valuable information about the optimal design of such studies. c. Tiludronate The efficacy of oral tiludronate to reduce the incidence of vertebral fractures in women with postmenopausal osteoporosis was examined in two RCTs [55]. Tiludronate was given intermittently 50 or 200 mg per day for 1 week every month to 2316 women. There were minor changes in biochemical indices of bone turnover and in the BMD of the spine. Results were disappointing with both regimens failing to show any antifracture efficacy. This resulted in the withdrawal of this bisphosphonate from any further development in osteoporosis. The tiludronate studies illustrate, in addition, the importance of careful and extensive clinical phase II dose-finding studies and of the danger of directly extrapolating results from animal studies with intermittent administration of bisphosphonates to humans.

SOCRATES E. PAPAPOULOS

2. NITROGEN-CONTAINING BISPHOSPHONATES a. Alendronate When given in different doses to women with postmenopausal osteoporosis, 20% of whom had prevalent fractures, alendronate reduced significantly (by about 50%) the incidence of vertebral fractures after 3 years [56]. Pooling the data of all doses used, which was preplanned, was required to demonstrate this effect. An overall efficacy of alendronate in reducing the risk of osteoporotic fractures was further supported by a metaanalysis of five RCTs [57]. Its antifracture efficacy was, however, demonstrated in a study specifically designed to examine the effect of alendronate on vertebral and nonvetebral fractures in postmenopausal women with low bone mass (Fracture Intervention Trial, FIT). This study had two arms. In the first (the vertebral fracture arm), 2023 postmenopausal women, mean age 71 years, with at least one prevalent vertebral fracture and femoral neck BMD less than 1.6 SD of the mean of healthy premenopausal women were randomized to receive alendronate 5 mg/day or placebo for 3 years [58]. The majority also received calcium 500 mg/day and vitamin D 250 IU/day. The dose of alendronate was increased to 10 mg/day after the second year because in parallel trials this dose was shown to have optimal effects on BMD. The trial was stopped just before its completion (at 2.9 years) because the independent Data Safety Monitoring Board considered its continuation unethical in view of the positive effect of treatment. New vertebral fractures occurred in 15% of women in the placebo group and in 8% of women in the alendronatetreated group (47% risk reduction). Active treatment reduced, in addition, the risk of multiple (by 90%) and clinical (by 55%) vertebral fractures as well as of forearm fractures (by 48%). More importantly, this was the first study to demonstrate a significant reduction in the incidence of hip fractures (by 51%) in calcium and vitamin D-replete, freeliving women with osteoporosis. In the second arm of FIT (clinical fracture arm) 4432 women, mean age 68 years, with femoral neck BMD less than 1.6 SD from the mean of healthy premenopausal women and no prevalent vertebral fractures, were randomized to receive alendronate or placebo for 4 years [59]. Again the dose of alendronate was increased from 5 to 10 mg/day after the second year. Active treatment reduced the incidence of new vertebral fractures by 44%. The risk of clinical fractures was reduced by 14%, a nonsignificant result. In contrast to the effect on vertebral fractures, the efficacy of alendronate on clinical fractures depended on initial femoral neck BMD. In women with osteoporosis (BMD T score   2.5) alendronate reduced significantly the risk of all clinical fractures by 36% and that of hip fractures by 56%. In this group a reduction of the risk of wrist fractures by 12% was not significant. In a multinational study (FOSIT), 1908 women with osteoporosis were treated with placebo or alendronate 10 mg/day for 1 year [60]. The incidence of nonvertebral fractures which were captured as adverse events was reduced significantly by 47%

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

by alendronate. The availability of multiple RCTs in postmenopausal osteoporosis with this bisphosphonate allowed the performance of metaanalysis. These showed that alendronate reduces significantly the risk of vertebral fractures by about 50% in women with a femoral BMD T score   1.6 with or without prevalent vertebral deformities. Similarly, alendronate reduced the risk of hip fractures in women with osteoporosis (prevalent vertebral fractures or femoral neck BMD T score   2.5) also by about 50%. The effect of alendronate on clinical fractures (e.g., spine, hip, multiple spine) was observed early during treatment becoming significant between 12 and 18 months. In addition, treatment was associated with improvement in some parameters of the quality of life of the patients [61]. From these results it can be concluded that alendronate therapy decreases: (1) the risk of vertebral fractures in postmenopausal women with variable risk (RCTs and metaanalysis) and (2) the risk of clinical fractures, including those of the hip, in postmenopausal women with osteoporosis defined by the presence of prevalent vertebral fractures or a BMD T score   2.5. b. Ibandronate Ibandronate is the most potent nitrogen-containing bisphosphonate under investigation for the treatment of osteoporosis. Phase III RCTs with orally administered ibandronate are near completion. Because of its increased potency, low doses of ibandronate can be given which made possible the intravenous administration of the bisphosphonate by injection rather than by infusion. Ibandronate was given by intravenous injections once every 3 months to 2862 patients with postmenopausal osteoporosis in a randomized, placebo-controlled study [47]. Two doses were tested, 0.5 or 1.0 mg every 3 months for 3 years. Preliminary results of this study indicated that this regimen with ibandronate was insufficient to significantly reduce the incidence of new vertebral fractures. Intermittent bisphosphonate treatment with intravenous injections is certainly an attractive management approach but the current evidence indicates that for fracture prevention the optimal dose of ibandronate as well as the dosing intervals need to be established. c. Pamidronate Pamidronate was the first nitrogen-containing bisphosphonate to be given orally to patients with osteoporosis and the results obtained in open studies helped to formulate the current principles of the daily bisphosphonate administration. However, due to increased incidence of gastrointestinal (GI) side effects in the original trials, there was no further development of oral pamidronate for the treatment of osteoporosis at least in Europe and the United States. Two small placebo-controlled trials examined the effect of oral pamidronate, 150 mg/day, for 2 and 3 years, respectively, on BMD. Both studies were not designed to assess antifracture efficacy. However, vertebral fractures were assessed as secondary outcome measure. In the first study a nonsignificant

639

reduction in the incidence of new vertebral fractures was observed after 2 years [62]. In the second study [63], which included 101 men and postmenopausal women with at least one prevalent vertebral fracture, pamidronate, 150 mg/day, decreased significantly the risk of new vertebral fractures (incidence in the placebo group, 33%, and in the pamidronate group, 11%; relative risk reduction, 67%). Another small RCT with intermittent oral pamidronate in osteoporotic women and men, showed no evidence of antifracture efficacy (47). There is no information on the efficacy of pamidronate in reducing the risk of nonvertebral fractures as well as of the efficacy of intermittent intravenous regimens on any type of osteoporotic fractures. d. Risedronate In two studies (one in North America and one multinational) women with osteoporosis were treated with risedronate, 2.5 or 5 mg/day for 3 years. In both trials calcium and vitamin D were given according to predefined protocol requirements. In the first study 2458 postmenopausal women, mean age 69 years, with at least one prevalent vertebral fracture were randomized to receive placebo, risedronate, 2.5 mg/day, or risedronate, 5.0 mg/day, for 3 years [64]. The risedronate, 2.5 mg/day, arm was discontinued after 1 year. Risedronate, 5 mg/day, reduced significantly the cumulative incidence both of new vertebral fractures by 41% (incidence in the placebo group, 16.3%, and in the risedronate group, 11.3%) and of non-vertebral fractures by 39%. In the second study 1226 postmenopausal women, mean age 71 years, with at least two prevalent vertebral fractures were randomized to receive placebo, risedronate, 2.5 mg/day, risedronate, 5.0 mg/day, for 3 years [65]. The risedronate, 2.5 mg, arm was discontinued after 2 years. Risedronate, 5 mg/day, reduced significantly the cumulative incidence of new vertebral fractures by 49% (incidence in the placebo group 29% and in the risedronate group 18.1%) and that of nonvertebral fractures by 33% (95% CI 0.44, 1.04). In both studies the effect of risedronate was rapid, inducing significant decreases in the risk of new vertebral fractures already after one year of treatment. The most important study, however, is a large RCT involving 9331 women with primary outcome the efficacy of risedronate in reducing the risk of hip fractures [66]. The trial had two arms. In the first arm 5445 women, mean age 74 years, with femoral neck BMD T score   3.0 and at least one clinical risk factor for hip fracture or a BMD T score   4.0 were randomized to receive placebo, risedronate, 2.5 mg/day, or risedronate, 5.0 mg/day, for 3 years. The results of the two risedronate doses were pooled, as preplanned. Active treatment reduced the cumulative incidence of hip fractures by 40%. The effect of risedronate was more prominent with increasing severity of osteoporosis. In contrast, in the second arm of the trial in which older women (mean age 84 years) were recruited mainly on the basis of clinical risk factors for hip fracture, the effect of risedronate on the risk of hip

640

SOCRATES E. PAPAPOULOS

fractures was small and not significant. This study illustrated, in addition, that even at an advanced age, patients who will most likely benefit from antiosteoporotic therapy should be identified by means other than clinical risk factors only.

VI. SPECIAL ISSUES OF BISPHOSPHONATE THERAPY AND UNANSWERED QUESTIONS A. Long-Term Increases in BMD Originally in the open studies with pamidronate, and later in the controlled studies with alendronate, it was shown that the increase in BMD with treatment is not confined to the first 2 years, as expected by the reduction in remodelling space induced by these agents (Fig. 5). In contrast, BMD increases further, albeit at a slower rate, for at least 7 years of therapy [24,67,68]. The reason for the continuous increase in BMD with treatment is not clear at present. Various hypotheses have been proposed to explain this, including a possible anabolic effect of nitrogen-containing bisphosphonates on bone, a direct effect on osteogenic cells, a readjustment of bone mass homeostasis or an improvement in the mineralization profile of bone [69]. In vitro studies have shown that, under certain conditions, low concentrations of nitrogen-containing bisphosphonates can stimulate the formation of osteoblasts and more recently that they can reduce the rate of apoptosis of osteoblasts induced by glucocorticoids or other apoptic

FIGURE 5 Patterns of spine BMD changes during treatment of patients with osteoporosis with bisphosphonates. Continuous line, combination of actual data of studies with alendronate, 10 mg/day (controlled studies for 7 years), or pamidronate, 150 mg/day (controlled studies for 5 years and open studies for 7 years). Dashed line, changes expected from a reduction in the bone remodeling space and data obtained in some studies with oral bisphosphonates given intermittently.

agents [70,71]. In addition, dose-dependent effects of bisphosphonates on osteogenic cells that were not directly related to their action on bone resorption have been reported. It may therefore be that bisphosphonates — depending on structure, dose, and mode of administration — may exert skeletal effects beyond those on bone resorption. Alternatively, it may be that such an effect is exerted by treatmentinduced changes in the production of systemic factors, such as parathyroid hormone (PTH), a known anabolic agent. However, chronic increases in the rate of secretion in PTH have been difficult to demonstrate in clinical studies with bisphosphonates [72]. In addition, in vivo studies failed to produce any strong evidence in support of a positive remodeling balance during bisphosphonate treatment. Bone biopsies from animals and humans treated with alendronate or pamidronate showed some increase in the mean wall thickness of trabeculae (an indication of an anabolic action) but these findings have not been consistent [73 – 75]. This issue remains currently open. A more attractive hypothesis is the increase in mineralization of bone with treatment. Meunier and Boivin [76], originally in microradiographic studies of bone biopsies from baboons and more recently from osteoporotic women treated with alendronate [77], showed a shift of the mineralization profile toward normal with treatment. They proposed that treatment increases the secondary mineralization of bone by slowing bone turnover and providing more time for calcium deposition to occur.

B. The Relation between Changes in BMD and Antifracture Efficacy Relevant to the above-mentioned issue is the relationship between increases in BMD and antifracture efficacy of bisphosphonates. Particularly the question whether the modest increases in BMD induced by treatment can account for the large decreases in fracture incidence. In studies with alendronate, patients with greater increases in BMD had a lower incidence of new vertebral deformities [78]. To examine this relationship in more detail, data from studies of various antiresorptive agents, including bisphosphonates, were combined and analyzed [79]. There was substantial variability in antifracture efficacy at any given level of change in BMD. Overall, trials that reported larger increases in BMD appeared to observe greater reductions in the risk of vertebral fractures. This relationship was quantified and a model was proposed that predicted that treatments that increase spine BMD by 8% would reduce fracture risk by 54%. Somehow different conclusions were reached by other authors [80]. There is currently a tendency to oversimplify (positively or negatively) the contribution of BMD changes to antifracture efficacy of antiresorptive agents. Independently of the way data are analyzed, it is

641

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

clear that increases in BMD contribute to the reduction in fracture risk but do not explain it completely. This is logical when considering that BMD is not the only determinant of fracture risk and additional mechanisms must be also operating. For example, the early antifracture effect of treatment (after 1 year) shown now for both alendronate and risedronate should be most probably attributed directly to the antiresorptive action of these agents rather than to early increases in BMD. It may be that treatment prevents the resorption, thinning, and perforation of trabeculae decreasing thus their risk of breaking. In vitro studies have also suggested increased viability of osteocytes by bisphosphonates as a possible additional mechanism of antifracture action independent of BMD [71]. In my view, the question which arises from these studies and analyses is whether it is scientifically correct to lump results of compounds with fundamentally different pharmacology and mechanism of action to explain the relationship between BMD and fracture risk on treatment before first clarifying this relation within every individual class of drugs. There are now enough data to allow such analyses for the bisphosphonates.

C. Duration of Treatment Open and controlled studies with pamidronate, etidronate, and more recently with alendronate showed that these bisphosphonates can be safely given to patients with osteoporosis for up to 7 years. However, the exact duration of treatment with bisphosphonates is not yet known. To address this question properly, it is important to establish first the changes occurring after treatment arrest. Stopping therapy after 1 or 2 years is generally followed by an increase in the rate of bone turnover to baseline within 1 year associated with decreases in BMD at a rate comparable to that of placebo-treated patients. Thus, no catch-up bone loss occuring as has been the case with estrogens. These observations suggested that the bisphosphonate should be given longterm. In open studies discontinuation of oral pamidronate after a mean period of 6.5 years was followed by a slow increase in the rate of bone resorption toward baseline over 2 years. There was, however, no change in either spine or femoral neck BMD during this period [81]. Preliminary results of the extension of the long-term alendronate studies showed similar results on bone resorption and BMD 2 years after stopping a 5-year treatment [68]. These observations raise questions about the mechanism underlying the changes in bone turnover and BMD following discontinuation of long-term therapy. It may be that after stopping treatment the bisphosphonate, which is embedded in bone, is slowly released in quantities that have a partial effect on bone turnover, helping to keep BMD constant but not increasing it further. The significance of these findings for the longterm antifracture efficacy of bisphosphonates is not known.

It is worth mentioning that in a small, open, long-term study with oral pamidronate the incidence of new vertebral fractures during the 2 years following treatment arrest was not different from that observed during treatment. These important practical issues need to be addressed in properly designed studies. An ongoing extension of the Fracture Intervention Trial with alendronate will help answer this question.

D. Timing of Oral Administration All bisphosphonates are poorly absorbed from the intestine and this absorption is further decreased in the presence of food or calcium which bind the drug in the gut. They should therefore be given in the fasting state at least 1/2 before a meal. In the fracture studies with alendronate and risedronate the bisphosphonates were given in the morning before breakfast. This strict regimen can be inconvenient and may decrease adherence to treatment. In our open and controlled studies with oral pamidronate we have followed a more flexible regimen giving the bisphosphonate always on an empty stomach but at different time points during the day. The effects on BMD were indistinguishable from those reported for alendronate, suggesting that such strict intake instructions may not be mandatory. Such approach may increase compliance and perhaps also reduce the incidence of GI irritation. A study specifically designed to address this issue is currently underway with alendronate [82]. Preliminary results indicated that the suppression of bone resorption after 6 months of treatment was similar in patients taking the bisphosphonate before breakfast or 1 hour before lunch or dinner.

E. Mode of Administration As discussed above, various bisphosphonates were given intermittently by intravenous infusions or intravenous injections in the case of ibandronate. This approach that bypasses the intestine and increases compliance is popular among some physicians. However, caution is needed in using such regimens. Although all bisphosphonates given intermittently by the parenteral route increase BMD, there are no data about their antifracture efficacy. Such data were recently presented for intravenous ibandronate given every 3 months for 3 years and were negative. It is frequently assumed that the total dose of bisphosphonate delivered to the skeleton rather than the way that this is given will determine the biological response. There are various examples that this is not true in osteoporosis. For example, the pattern of suppression of bone resorption with oral ibandronate, 2.5 mg/day, is similar to that observed with all potent bisphosphonates given daily by

642

SOCRATES E. PAPAPOULOS

mouth [83]. Assuming a 0.6% intestinal absorption of the drug, the bioavailable dose after 1 year will be about 5.4 mg. Intravenous injections of 1 mg of the bisphosphonate given every 3 months will deliver 5 mg to the patient after 1 year (5 injections). Yet the changes in bone resorption with this intravenous regimen follow a pattern different from that of the continuous administration and do not decrease fracture risk. There is greater initial suppression of resorption, followed, however, by recovery toward baseline by the time of the next injection [46]. Similar responses have been described with intravenous pamidronate and alendronate [84,85]. Thus, although the continuous suppression of bone resorption has been repeatedly shown to be associated with antifracture efficacy, the same cannot be said for the intermittent administration. This mode of 3monthly intermittent administration should be distinguished from protocols giving the bisphosphonate more frequently. Bone biology and animal data indicate that if a dose of bisphosphonate equal to the sum of the daily doses is given within a period of 2 weeks, results obtained with the two regimens will be identical [86]. This hypothesis was tested for alendronate in a clinical trial [87]. Alendronate was given once weekly at a dose corresponding to the total dose given daily (70 mg/week instead of 10 mg/day). The changes in bone resorption and in BMD induced with the weekly regimen were indistinguishable from those induced by the daily administration which has a proven antifracture efficacy. Thus, the total dose of a bisphosphonate may be considered therapeutically equivalent to that of daily dosing as long as it is given within a period not exceeding 2 weeks. For longer intervals doses and dosing intervals should be calculated and clinically tested for every individual bisphosphonate before offering it to osteoporotic patients.

F. Combination Therapies Several studies have explored the possibility of combining bisphosphonate with other therapies used in osteoporosis. Combining antiresorptive agents with different mechanisms of action may theoretically be of greater benefit to the patient than a single agent given alone. Such approaches are routinely used in the management of other chronic diseases. However, the potential advantage, higher antifracture efficacy, has never been demonstrated and is overshadowed by the potential disadvantages of the combination therapy. These include more side effects, lower compliance, higher suppression of bone turnover, and higher costs. There are, however, clinicians who may not be convinced by the evidence of the antifracture efficacy of estrogens but do not want to stop it because of its effects on postmenopausal symptoms and would like to offer additional treatment to their patients. A RCT examined the

effect of the addition of alendronate to ongoing estrogen therapy and demonstrated a significant increase in BMD in women taking the combination compared to those taking only estrogens after 1 year [88]. Similar responses have been reported in other studies as well as after combining alendronate with raloxifene [89 – 91]. Much more interesting and clinically promising especially for patients with severe osteoporosis is the combination of a bisphosphonate with PTH (see also Fig. 3). PTH given to women with postmenopausal osteoporosis increases dramatically spine BMD and reduces the risk of osteoporotic fractures within 2 years of treatment [92]. Treatment should be given, however, by daily subcutaneous injections and there are no data about longer term efficacy or safety. In addition, the effect is lost after stopping treatment. Consolidation of these effects by bisphosphonate therapy presents a very interesting option and initial studies with alendronate support this notion. PTH, however, has not yet been approved for the treatment of postmenopausal osteoporosis and there are ongoing trials that explore in more detail the combination of PTH not only with alendronate but also with other antiresorptive agents [93 – 95] (See Chapters 77 and 79).

VII. LONG-TERM EFFECTS ON BONE METABOLISM Because of the long retention of the bisphosphonates in the skeleton and their property to inhibit the growth of calcium crystal in vitro, establishing skeletal safety forms an extremely important part of the development of these compounds for the management of osteoporosis. The preclinical evidence for that is reviewed in Chapter 16 (Fleisch). The long-term effects of bisphosphonate treatment on bone metabolism and quality in humans have been evaluated by biochemical indices of bone turnover and bone histology. In all studies with nitrogen-containing bisphosphonates given daily, the pattern of response was similar to that originally obtained in the animal study (Fig. 6). A plateau of bone resorption is obtained which is not suppressed further despite the continuing administration of the bisphosphonate. This pattern of response has been repeatedly shown with all nitrogen-containing bisphosphonates. In addition, biochemical indices of bone turnover decrease to the average value of premenopausal women, shown for alendronate, and remain at this level for the duration of treatment [96]. These observations underscore the conclusions formulated above that the biologically active bisphosphonate is the one present on the bone surface rather than that embedded in bone. It has been hypothesized that, in addition to their effect on bone remodeling, bisphosphonates may affect calcium exchange at quiescent bone surfaces by their strong binding to bone mineral. This is not however, the case as long-term therapy with oral pamidronate did not

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

FIGURE 6 Percentage change in biochemical indices of bone turnover in 75 patients with osteoporosis treated with oral pamidronate, 150 mg/day, for 6 years. AP, serum, alkaline phosphatase activity; OHP/Cr, urinary hydroxyproline to ceatinine ratio.

impair the rapid mobilization of calcium to correct an acute hypocalcemia induced by EDTA infusions [97]. In addition, long-term pamidronate treatment did not modify the responsiveness of the parathyroid glands to acutely induced hypocalcemia (Fig. 7) which contrasts findings during estrogen treatment; estrogens were reported to decrease the sensitivity of the parathyroid glands to stimulation by EDTA-induced hypocalcemia [98]. Bone histology has been reported in biopsies taken from osteoporotic patients treated for variable periods with different bisphosphonates. With cyclic etidronate there is a reduction in the activation frequency after 1 to 2 years of treatment. Impairment of bone mineralization that has been reported in animals and humans given high doses of etidronate was not encountered with the intermittent regimen [99 – 101]. There are only anecdotal reports of histologically confirmed osteomalacia with cyclic etidronate [102,103]. Bone biopsies of patients with osteoporosis treated with alendronate, 5 or 10 mg/day, or risedronate, 5 mg/day, for 3 years and pamidronate, 150 mg/day, for 2 years showed the expected decrease in the rate of bone turnover with no evidence of a mineralization defect or qualitative abnormalities of bone collagen or bone marrow [64,74,75,104 – 106]. Additional supportive data of the skeletal safety of nitrogen-containing bisphosphonates has been obtained in limited studies in children [107 – 109]. The use of bisphosphonates in such patients provides the opportunity to address questions of their pharmacology that are impossible to study in adults. The growing skeleton is particularly sensitive to factors which can adversely affect bone metabolism and any potential deleterious effect of long-term bisphosphonate therapy may be best identified in children. In children treated long-term with nitrogen-containing bisphosphonates (mainly pamidronate) the pattern of linear

643

growth was normal, there were marked increases in calcium balance which were sustained after the first year of treatment and there were increases in BMD at all skeletal sites with slopes different from those of their healthy peers and a tendency of values to reach normal ranges, especially when treatment is given before puberty. Radiological changes were striking and consisted of band-like metaphyseal sclerosis and concentric epi- and apophyseal sclerosis (Fig. 8). The extent of sclerosis depended on the dose, duration, of treatment and growth activity. Complete reversal of vertebral deformities was also observed in some prepubertal patients. These observations strengthen the notion that the accumulation of the bisphosphonate in the skeleton is not accompanied by long-term adverse effects on bone metabolism.

FIGURE 7 Changes in serum calcium concentrations (top) and in plasma PTH (bottom) in 10 patients with osteoporosis treated with oral pamidronate 150 mg/day, for at least 5 years (closed circles) and in 10 patients with osteoporosis who had never received any treatment (open circles) after a short infusion with sodium EDTA. Mean and SEM are shown; bar indicates the time of infusion. Reprinted with permission from Landman et al. [97].

644

SOCRATES E. PAPAPOULOS

FIGURE 8 Radiographs of the knees of a child with juvenile osteoporosis before (a) and after 3 (b) and 6 (c) months of treatment with daily oral pamidronate. Note the progressive increase in metaphyseal sclerosis during treatment which is remarkably similar to change observed in growing animals.

VIII. ADVERSE EFFECTS Little toxicity is associated with bisphosphonate therapy as long as the general and the individual properties of the compunds are taken into consideration. Skeletal safety was discussed above. Mild GI complaints have been reported with variable frequency with all bisphosphonates. Aminobisphosphonates, however, given orally can induce more severe effects such as heartburn, nausea, or vomiting [110]. Some cases of severe esophagitis have been reported with oral alendronate and pamidronate [111,112]. With pamidronate GI intolerance is dose-dependent. With 150 mg/day there was no difference between actively and placebo-treated patients [63]. With 300 mg/day the incidence of GI side effects was higher in actively treated patients (8% in placebo-treated versus 28% in amidronate-treated) [106]. Doses higher than 300 mg/day induce GI side effects of varying severity in up to

50% of patients and doses above 600 mg/day are hardly tolerated by the majority of patients. In all controlled trials in osteoporosis with oral alendronate the incidence of GI side effects did not differ between alendronate- and placebo-treated patient [113]. Doses of 40 mg/day or higher are associated with a higher discontinuation rate. In clinical practice the incidence of discontinuation of alendronate treatment due to GI adverse experiences is between 10 and 15%, as estimated in a large pharmacovigilance study in the Netherlands. Proper instructions of taking the medication (full glass of water and not lying down) helps to avoid esophageal irritation. Primary GI intolerance to aminobisphosphonates occurs usually within the first 4 weeks of treatment. Patients should be informed about this possibility and, in case of severe symptoms, treatment should be stopped. After reviewing with the patient the way the bisphosphonate was taken, treatment can be reinstituted after 3 to 4 weeks. In case of

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

FIGURE 8 (continued) reappearance of the symptoms, treatment should be stopped. Similarly, in the published studies with risedronate there were no differences in GI toxicity between active and placebo-treated patients. The incidence in daily practice should await the general use of this bisphosphonate following its recent approval. The mechanism of GI irritation is not known. Furthermore, the question whether the nitrogen functionality (as in all nitrogen-containing biphosphonates) or the aminogroup (as in alendronate and pamidonate) is primarily associated with GI side effects has not been answered yet. Some preliminary in vitro evidence obtained in a cellular model similar to esophageal epithelium, suggested that in these cells nitrogen-containing bisphosphonates act by a mechanism similar to that of their action on osteoclasts, by inhibiting enzymes of the mevalonate pathway [114]. These observations suggest that GI irritation may be a class effect but the appropriate clinical studies to confirm or refute this suggestion need still to be performed. Recently in an animal model of esophageal irritation it was shown that alendronate is devoid of any damage to the esophageal epithelium if given once weekly even at high doses. In clinical studies it was even demonstrated that

645

weekly administration of 70 mg alendronate was associated with less GI side effects than the daily administration of 10 mg [87]. Thus, this mode of alendronate administration may prove to have the same efficacy as the daily one with a more favorable toxicity profile, being at the same time more convenient for chronic treatment. Local irritation at the site of bisphosphonate infusion and mild thrombophlebitis can occur with intravenous aminobisphosphonates but not with clodronate. Aminobisphosphonates should never be given intramuscularly. When infused rapidly in high concentrations, bisphosphonates may chelate calcium in the circulation. Calcium – bisphosphonate complexes may be formed which can be nephrotoxic and they should be, therefore, administered by slow infusion [115]. The duration of the infusion does not affect the therapeutic response but the minimal duration should be determined for every individual bisphosphonate. Newer, very potent nitrogen-containing bisphosphonates, such as ibandronate and zolendronate, can be safely given, however, by intravenous bolus injections. During treatment with intravenous pamidronate some cases of conjunctivitis or uveitis have been reported. There is no obvious causal relationship and it is not known whether this is structure-related. Earlier fears that clodronate may predispose to leukemia were not substantiated by long-term studies and careful analysis of the data. Some allergic skin reactions have been reported for practically all bisphosphonates. In some patients treated for the first time with nitrogencontaining bisphosphonates there is an increase in body temperature within the first 3 days of treatment associated with flu-like symptoms (e.g., malaise and myalgia). This effect is transient, reverses without specific treatment and is structure- and dose-related. It has been reported with several nitrogen-containing bisphosphonates but not with etidronate or clodronate [116 – 119]. Furthermore, it has never been described during treatment with the low doses used in the management of osteoporosis. The clinical and biochemical picture resembles an acute phase reaction. Along with the rise in temperature, there is a transient decrease in lymphocyte count and a transient rise in acute phase reactants in serum, such as C-reactive protein. The exact incidence of this response cannot be estimated in relation to bisphosphonate dose and structure, as in most studies the exact level of temperature increase is not mentioned. In a detailed analysis of patients with Paget’s disease treated for the first time with bisphosphonate and followed daily in a metabolic ward, Harinck et al. [12] reported that 55% of patients treated with oral pamidronate (600 mg/day) and 63% of those treated intravenously (20 mg/day had an increase in body temperature of more than 0.5°C, while in similar studies with olpadronate the incidence of this reaction was 18.6%. In all studies the febrile response was transient and temperature returned to basal

646

SOCRATES E. PAPAPOULOS

FIGURE 9 Early changes in plasma IL–6 bioactivity in two patients with Paget’s disease of bone treated with the nitrogen–containing bisphosphonate olpadronate and showing an acute phase reaction. Temperature is shown at the bottom of the figure. Note the rapid early increase in IL–6 bioactivity. Reprinted with permission from Schweitzer et al. [119].

despite continuation of treatment. It is also of interest to note that this reaction occurs during the first exposure to a nitrogen-containing bisphosphonate and patients who had been previously treated with the same or another nitrogencontaining compound do not in general develop it. An acute phase response is mediated primarily by proinflammatory cytokines, such as IL-6 and TNF-, and there is evidence to support a role of both these cytokines in the bisphosphonate-induced acute phase reaction [119, 120] (Fig. 9). The cellular basis of the response is unclear. A general side effect of treatment seems unlikely because it occurs only on first exposure to the bisphosphonate even if the interval between the two treatments is years apart. It may be that this reaction is related to the action of the drug at the bone – bone marrow interface.

IX. CONCLUSIONS Bisphosphonates, in particular the nitrogen-containing compounds, are currently the most extensively studied antiosteoporotic treatment under controlled conditions. When given daily, they reduce the risk of vertebral fractures. More importantly alendronate and risedronate are currently the only treatments shown in prospective, controlled studies to decrease the incidence of non vertebral fractures, including those of the hip, in vitamin D- and calcium-replete women with postmenopausal osteoporosis.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

References 1. X. Boulenc, E. Marti, H. Joyeux, C. Roques, Y. Berger, and G. Fabre, Importance of the paracellular pathway for the transport of a new bisphosphonate using the human CACO-2 monolayers model. Biochem. Pharmacol. 46, 1591 – 1600 (1993). 2. I. M. Twiss, R. de Water, J. den Hartigh, R. Sparidans, W. RampKoopmanschap, H. Brill, M. Wijedeveld, and P. Vermeij, Cytotoxic effects of pamidronate on monolayers of human intestinal epithelial

13.

14.

(Caco-2) cells and its epithelial transport. J. Pharmaceut. Sci. 83, 699 – 703 (1994). A. D. Geddes, S. M. D’Souza, F. H. Ebetino, and K. J. Ibbotson, Bisphosphonates: Structure-activity relationships and therapeutic implications. In “Bone and Mineral Research” (J. N. M. Heersche and J. A. Kanis, eds.), Vol. 8, pp. 265 – 306. Elsevier Science, Amsterdam, 1994. A Fitton and D. McTavish, Pamidronate, a review of its pharmacological properties and therapeutic efficacy in resorptive bone disease. Drugs 41, 289 – 318 (1991). G. L. Plosker and K. L. Goa, Clodronate, a review of its pharmacological properties and therapeutic efficacy in resorptive bone disease. Drugs 47, 9459 – 82 (1994). J. H. Lin, D. E. Duncan, I. W. Chen, and R. L. Ellsworth, Physiological deposition of alendronate, a potent anti-osteolytic bisphosphonate, in laboratory animals. Drug Metab. Dispos. 19, 926 – 932 (1991). J. A. Kanis, B. J. Gertz, F. Singer, and S. Ortolani, Rationale for the use of alendronate in osteoporosis. Osteoporosis Int. 5, 1 – 13 (1995). S. A. Kahn, J. A. Kanis, S. Vasikaran et al., Elimination and biochemical responses to intravenous alendronate in postmenopausal osteoporosis. J. Bone Miner. Res. 12, 1700 – 1707 (1997). S. E. Papapoulos, Pharmacodynamics of bisphosphonates in man: Implications for treatment, In “Bisphosphonates on Bones” (O. L. M. Bijvoet, H. A. Fleisch, R. E. Canfield, and R. G. G. Russell, eds.) pp. 231 – 263. Elsevier Science, Amsterdam, 1995. S. E. Papapoulos, M. Frolich, A. H. Mudde, H. I. J. Harinck, H. van de Berg, and O. L. M. Bijvoiet, Serum osteocalcin in Paget’s disease of bone: Basal concentrations and response to bisphosphonate treatment. J. Clin. Endocrinol. Metab. 65, 89 – 94 (1987). W. B. Frijlink, J.te Velde, O. L. M. Bijvoet, and G. Heynen, Treatment of Paget’s disease with (3amino–1–hydroxypropylidene)–1,1–bisphosphonate (APD). Lancet I, 799 – 803 (1979). H. I. J. Harinck, S. E. Papapoulos, H. J. Blanksma, A. J. Moolenaar, P. Vermeij, and O. L. M. Bijvoet, Paget’s disease of bone: Early and late responses to three different modes of treatment with aminohydroxypropylidene bisphosphonate (APD). Br. Med. J. 295, 1301 – 1305 (1987). D. Thiebaud, P. Jaeger, C. Gobelet, A. F. Jacquet, and P. Burckhardt. A single infusion of the bisphosphonate AHPrBP (APD) as treatment of Paget’s disease of bone. Am. J. Med. 85, 207 – 212 (1988). D. P. O’Doherty, D. R. Bickerstaff, E. V. McCloskey, N. A. T. Hamdy, M. N. C. Beneton, S. Harris, M. Mian, and J. A. Kanis,

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31.

Treatment of Paget’s disease of bone with aminoxydroxybutylidene bisphosphonate. J. Bone Miner. Res. 5, 483 – 491 (1990). D. H. Schweitzer, A. H. Zwinderman, P. Vermeij, O. L. M. Bijvoet, and S. E. Papapoulos, Improved treatment of Paget’s disease with dimetylaminohydroxypropylidene bisphosphonate. J. Bone Miner Res. 8, 175 – 182 (1993). S. E. Papapoulos, Paget’s disease of bone: Clinical, pathogenetic and therapeutic aspects. Baillieres Clin. Endocrinol. Metab. 11, 117 – 143 (1997). O. L. M. Bijvoet, S. E. Papapoulos, C. W. G. M. Lowik, and H. I. J. Hainck, Manipulation of cell – cell interaction in bone with biphosphonate in the interpretation of boen and calcium homeostasis in man. In “Calcium Regulation and Bone Metabolism — Basic and Clinical Aspects” (D. V. Cohn, T. J. Martin, and P. J. Meunier, eds.), Vol. 9, pp. 93 – 104. Excerpta Medica, Amsterdam, 1987. R. R. Recker, G. S. Hassing, J. R. Lau, and P. D. Saville, The hyperphosphatemic effect of disodium ethane–1 – hydroxy–1, 1–diphosphonates (EHDP, TM): Renal handling of phosphate and the renal response to parathyroid hormone. J. Lab. Clin. Med. 81, 258 – 266. R. J. Walton, R. G. G. Russell, and R. Smith, Changes in the renal and extrarenal handling of phosphate induced by disodium–etidronate (EHDP) in man. Clin. Sci. Mol. Med. 49, 45 – 56 (1975). H. Fleisch, Bisphosphonates, pharmacology and use in the treatment of tumor–induced hypercalcemia and metastatic bone disease. Drugs 42, 919 – 944 (1991). F–J. F. E. Vismans, “APD, Parathyroid Hormone and Osteoporosis: Study of the Effects on Calcium Homeostasis.” PhD thesis, Univ. of Leiden, 1984. L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. Metab. Bone Dis. Relate. Res. 2, 389 – 407 (1980). R. Valkema, F–J. F. E. Vismans, S. E. Papapoulos, E. K. J. Pauwels, and O. L. M. Bijvoet, Maintained imrovement in calcium balance and bone mineral content in patients with osteoporosis treated with the bisphosphonate APD. Bone Miner. 5, 183 – 192 (1989). R. P. Heaney. and P. D. Saville, Etidronate disodium in postmenopausal osteoporosis. Clin. Pharmacol. Ther. 20, 593 – 604 (1976). J. Jowsey, B. L. Riggs, P. J. Kelly, D. L. Hossman, and P. Bordier, The treatment of osteoporosis with disodium ethane–1–hydroxy–1,1–diphosphonate. J. Lab. Clin. Med. 78, 574 – 584 (1971). P. H. Reitsma, O. L. M. Bijvoet, H. Verlinden–Ooms, and L. J. A. Wee van der Plas, Kinetic studies of bone and mineral mineral metabolism during treatment with (3–amino–1–hydroxypropylidene)–1,1–bisphosphonate (APD) in rats. Calcif. Tissue Int. 32, 145 – 157 (1980). C. Anderson, R. D. Cape, R. G. Crilly, A. B. Hodsman, and B. M. Wolfe. Preliminary observations of a form of coherence therapy for osteoporosis. Calcif. Tissue Int. 36, 341 – 343 (1984). A. B. Hodsman, Effects of cyclical therapy for osteoporosis using an oral regimen of inorganic phosphate and sodium etidronate: A clinical and bone histomorphometric study. Bone Miner. 5, 201 – 212 (1989). N. B. Watts, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Miller, R. D. Jackson, A. A. Licata, P. Ross, G. C. Woodson, M. Y. Yanover, W. J. Mysiw, L. Kohse, M. B. Rao, P. Steiger, B. Richmond, and C. H. Chesnut III, Intermittent cyclical etidronate treatment of postmenopausal osteoporosis. N. Engl. J. Med. 323, 73 – 79 (1990). H. M. Frost, Treatment of osteoporosis by manipulation of coherent bone cell populations. Clin. Orthop. 143, 227 – 244 (1979). R. Pacifici, C. McMurtry, I. Vered, R. Rupich, and L. V. Avioli, Coherence therapy does not prevent axial bone loss in osteoporotic women: A preliminary comparative study. J. Clin. Endocrinol. Metab. 66, 855 – 866 (1988).

647

32. L. E. Mallette, A. D. LeBlanc, J. L. Pool, and J. L. Mechanick, Cyclic therapy of osteoporosis with neutral phosphate and brief, high–dose pulses of etidronate. J. Bone Miner. Res. 4, 143 – 148 (1989). 33. T. Steiniche, C. Hasling, P. Charles, E. F. Eriksen, F. Melsen, and L. Moselkide, The effect of etidronate on trabecular bone remodeling in postmenopausal spinal osteoporosis: A randomized study comparing intermittent treatment and a ADFR regime. Bone 12, 155 – 163 (1991). 34. T. Storm, G. Thamsborg, T. Steiniche, H. K. Genant, O. H. Sorensen, Effect of intermittent cyclical etidronate on bone mass and fracture rate in women with postmenopausal osteoporosis. N. Engl. J. Med. 322, 1256 – 1271 (1990). 35. T. Storm, G. Kollerup, G. Thamsborg, H. K. Genant, and O. H. Sorensen, Five years of clinical experience with intermittent cyclical etidronate for postmenopausal osteoporosis. J. Rheumatol., 23, 1560 – 1564 (1996). 36. S. T. Harris, N. B. Watts, R. D. Jackson, H. K. Genant, R. D. Wasnich, P. D. Miller, A. A. Licata, and C. H. Chesnut III, Four–year study of intermittent cyclic etidronate treatment of postmenopausal osteoporosis. Three years of blinded therapy followed by one year of open therapy. Am. J. Med. 95, 557 – 567 (1993). 37. P. D. Miller, N. B. Watts, A. A. Licata, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Ross, R. D. Jackson, M. S. Hoseyni, S. L. Schoenfeld, D. J. Valent, and C. H. Chesnut III, Cyclical etidronate in the treatment of postmenopausal osteoporosis: Efficacy and safety after seven years of treatment. Am. J. Med. 103, 468 – 476 (1997). 38. S. Giannini, A. D’Angelo, L. Malvasi, R. Castrignano, R. Tronca, M. Nobile, and G. Crepaldi, Effects of one year cyclical treatment with clodronate on bone mass and mineral metabolism in postmenopausal osteoporosis. Bone 14, 137 – 141(1993). 39. P. Filipponi, M. Pedetti, L. Fedeli, L. Cini, R. Palumbo, S. Boldrini, C. Massoni, and S. Cristallini, Cyclical clodronate is effective in preventing postmenopausa l bone loss: A comparative study w i t h t r a n scutaneous hormone replacement therapy. J. Bone Miner. Res. 10, 697 – 703 (1995). 40. P.Filipponi, S. Cristallini, G. Policani, M. F. Schifini, C. Casciari, and P. Garinei, Intermittent versus continuous clodronate administration in postmenopausal women with low bone mass. Bone 26, 269 – 274 (2000). 41. J. P. Devogelaar, and C. Nagant de Deuxchaisnes, Treatment of involutional osteoporosis with the bisphosphonate APD (disodium pamidronate): Non-linear increase of lumbar bone mineral density. J. Bone Miner. Res. 5, S97 (1990). 42. S. E. Papapoulos, J. O. Landman, O. L. M. Bijvoet, C. W. G. M. Lowik, R. Valkema, E. K. J. Pauwels, and P. Vermeij, The use of bisphosphonates in the treatment of osteoporosis. Bone 13, S41 – S49 (1992). 43. P. J. Ryan, G. M. Blake, M. Davie, M. Haddaway, T. Gibson, and I. Fogelman, Intermittent oral disodium pamidronate in established osteoporosis: A 2–year double–masked placebo–controlled study of efficacy and safety. Osteoporosis Int. 11, 177 – 187 (2000). 44. M. Passeri, M. C. Baroni, M. Pedrazzoni, M. Pioli, M. Barbagallo, D. Costi, M. Biondi, G. Girasole, B. Arlunno, and E. Palummeri, Intermittent treatment with intravenous 4–amino–1–hydroxybutilydene–1,1–bisphosphonate (AHBuBP) in the therapy of postmenopausal osteoporosis. Bone Miner. 15, 237 – 248 (1991). 45. D. Thiebaud, P. Burckhardt and J. Melchior, Two years’ effectiveness of intravenous pamidronate (APD) versus oral fluoride for osteoporosis occurring in the menopause. Osteoporosis Int. 4, 76 – 83 (1994). 46. D. Thiebaud, P. Burckhardt, H. Kriegbaum, H. Huss, H. Mulder, J. R. Jutmann, and K. H. Schoter, Three monthly intravenous injections of ibandronate in the treatment of postmenopausal osteoporosis. Am. J. Med. 103, 298 – 307 (1997).

648 47. R. R. Recker, J. A. Stakkestad, D. Falsenberg, C. H. Chesnut III, C. Christiansen, M. P. Ettinger, A. Nordby, S. R. Weiss, H. Huss, T. Von Stein, and D. B. Karpf, A new treatment paradigm: Quarterly injections of ibandronate reduce the risk of fractures in women with postmenopausal osteoporosis (PMO): Results of a 3–year trial. Osteoporosis Int. 11 (Suppl. 2), S209 (2000). 48. T. O’Neill and S. Papapoulos, Can we prevent fractures? Bailliere’s Clin. Rheumatol. 11, 565 – 582 (1997). 49. S. E. Papapoulos, Pharmacological management of osteoporosis. Aging Clin. Exp. Res. 10, 263 – 268 (1998). 50. D. M. Black, N. K. Arden, L. Palermo, J. Pearson, and S. R. Cummings, Prevalent vertebral deformities predict hip fractures and new vertebral deformities but not wrist fractures. J. Bone Miner. Res. 14, 821 – 828 (2000). 51. K. E. Ensrud, D. E. Thompson, J. A. Cauley, M. C. Nevitt, D. M. Kado, M. C. Hochberg, A. C. Santora, and D. M. Black, Prevalent vertebral deformities predict mortality and hospitalization in older women with low bone mass. J. Am. Geriatr. Soc. 48, 1 – 9 (2000). 52. M. C. Chapuy, M. E. Arlot, F. Duboef, et al. Vitamin D and calcium to prevent hip fractures in the elderly. N. Engl. J. Med. 327, 1637 – 1642 (1992). 53. E. V. McCloskey, P. Selby, M. Davies, R. Francis, J. Robinson, K. Kayan, M. Beneton, T. Jalava, L. Pylkkanen, J. Kenrali, and J. Kanis, Oral clodronate significantly reduces fracture risk in women with postmenopausal or secondary osteoporosis. J. Bone Miner. Res. 15 (Suppl. 1), S227 (2000). 54. T. P. van Staa, L. Abenhaqim, and C. Cooper, Use of cyclical etidronate and prevention of non–vertebral fractures. Br. J. Rheumatol. 37, 87 – 94 (1998). 55. H. K. Genant, C. H. Chesnut III, J. A. Eisman, et al., Chronic intermittent cyclical administration of tiludronate in postmenopausal osteoporosis: Report of two multicenter studies in 2316 patients. Bone 23 (Suppl.), S175 (1998). 56. U. A. Liberman, S. R. Weiss, J. E. T. Bröll, et al., Effect of oral alendronate on bone mineral density and incidence of fractures in postmenopausal osteoporosis. N. Engl. J. Med. 333, 1437 – 1443 (1995). 57. D. B. Karpf, D. R. Shapiro, E. Seeman, et al., Prevention of nonvertebral fractures by alendronate: a meta-analysis. JAMA 277, 1159 – 1164 (1997). 58. D. M. Black, S. R. Cummings, D. B. Karpf, et al., Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 348, 1535 – 1541 (1996). 59. S. R. Cummings, D. M. Black, D. E. Thompson, et al., Effect of alendronate on risk of fractures in women with low bone density but without vertebral fractures: Results from the fracture intervention trial. JAMA 280, 2077 – 2082 (1998). 60. H. A. P. Pols, D. Falsenberg, D. A. Hanley, et al., Multinational, placebo-controlled randomized trial of the effects of alendronate on bone density and fracture risk in postmenopausal women with low bone mass: Results of the FOSIT study. Osteoporosis Int. 9, 461 – 468 (1999). 61. M. C. Nevitt, D. E. Thompson, D. M. Black, et al., Effect of alendronate on limited-activity days and bed-disability days caused by back pain in postmenopausal women with existing vertebral fractures. Arch. Intern. Med. 160, 77 – 85 (2000). 62. I. R. Reid, D. J. Wattie, M. C. Evans, G. D. Gamble, J. P. Stapleton, and J. Cornish, Continuous therapy with pamidronate, a potent bisphosphonate, in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 79, 1595 – 1599 (1994). 63. C. Brumsen, P. Lips, N. A. T. Hamdy, et al., Oral pamidronate progressively increases lumbar spine BMD over 5 years in women and men with osteoporosis and reduces the risk of new vertebral fractures. Bone 23(Suppl.), S310 (1998).

SOCRATES E. PAPAPOULOS 64. S. T. Harris, N. B. Watts, H. K. Genant et al., Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis. JAMA 282, 134 – 1352 (1999). 65. J–Y Reginster, H. W. Minne, O. H. Sorensen, et al., Randomized trial of the effects of risedronate on vertebral fractures in women with established osteoporosis. Osteoporosis Int. 11, 83 – 91 (2000). 66. P. Geusens, S. Adami, W. Bensen, et al, Risedronate reduces risk of hip fractures in elderly women with osteoporosis. Calcif. Tissue, 66 (Suppl.), S67 (2000). 67. M. Favus, R. Emkey, M. Leite, J. P. Devogelaer, J. Rodriguez, C. Peverly, A. Kauf, and A. Santora, Five-year treatment of osteoporosis in postmenopausal women with oral alendronate: Effects on bone mass and turnover and safety. J. Bone Miner. Res. 12 (suppl), S150 (1997). 68. R. P. Tonino, P. J. Meunier, R. D. Emkey, J. A. Rodriguez-Portales, C. J. Menkes, R. D. Wasnich, H. G. Bone, A. C. Santora, M. Wu, R. Desai, and P. D. Ross, Long–term (seven year) efficacy and safety of alendronate in the treatment of osteoporosis in postmenopausal women. Osteoporosis Int. 11 (Suppl. 2), S202 (2000). 69. G. A. Rodan, Bone mass homeostasis and bisphosphonate action. Bone 21, 373 – 377 (1997). 70. N. Giuliani, M. Pedrazzoni, G. Negri, G. Passeri, M. Impicciatore, and G. Girasole, Bisphosphonates stimulate formation of osteoblast precursors and mineralized nodules in murine and human bone marrow cultures in vitro and promote early osteoblastogenesis in young and aged mice in vivo. Bone 22, 455 – 461 (1998). 71. L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Robertson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999). 72. J. O. Landman and S. E. Papapoulos, Uninterrupted oral bisphosphonate (pamidronate) therapy of patients with osteoporosis is not associated with chronic stimulation of parathyroid hormone secretion. Osteoporosis Int. 5, 93 – 96 (1995). 73. R. Balena, B. C. Toolan, M. Shea et al., The effect of 2–year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry and bone strength in ovariectomized nonhuman primates. J. Clin. Invest. 92, 2577 – 2586 (1993). 74. P. M. Chavassieux, M. E. Arlot, C. Reda, L. Weil, A. J. Yates, and P. J. Meunier, Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. J. Clin. Invest. 100, 1475 – 1480 (1997). 75. N. Bravenboer, S. E. Papapoulos, P. Holzman, N. A. T. Hamdy, J. C. Netelenbos, and P. Lips, Bone histomorphometric evaluation of pamidronate treatment in clinically manifest osteoporosis. Osteoporosis Int. 9, 489 – 493 (1999). 76. P. J. Meunier, and G. Boivin, Bone mineral density reflects bone mass but also the degree of mineralization of bone: Therapeutic implications. Bone 21, 373 – 377 (1997). 77. G. Boivin, A. C. Santora, J. Yates, and P. J. Meunier, In postmenopausal osteoporotic women, alendronate augments bone strength by increasing the degree of mineralization of bone tissue. Osteoporosis Int. 11 (Suppl. 2), S209 (2000). 78. M. Hochberg, P. D. Ross, D. Black et al., larger increases in bone mineral density during alendroante therapy are associated with lower risk of new vertebral fractures in women with potmenopausal osteoporosis. Arthritis Rheum. 42, 1246 – 1254 (1999). 79. R. D. Wasnich and P. D. Miller, Antifracture efficacy of antiresorptive agents are related to changes in bone density. J. Clin. Endocrinol. Metab. 85, 231 – 236 (2000). 80. P. D. Delmas, How does antiresorptive therapy decrease the risk of fracture in women with osteoporosis? Bone 27, 1 – 4 (2000). 81. J. O. Landman, N. A. T. Hamdy, E. K. J. Pauwels, and S. E. Papapoulos, Skeletal metabolism in patients with osteoporosis after discontinuation of long-term treatment with oral pamidronate. J. Clin. Endocrinol. Metab. 80, 3465 – 3468 (1995).

CHAPTER 72 Bisphosphonates in the Management of Postmenopausal Osteoporosis 82. P. D. Delmas, E. Confacreux, G. Genolet, P. Gamero, B. Gibelin, C. Folens, and J. Yates, Can alendroante be taken before lunch or dinner? A randomized trial in osteoporotic women. Osteoporosis Int. 11 (Suppl. 2), S208 (2000). 83. P. Ravn, B.Clemmenson, B. J. Riis, and C. Christiansen, The effect on bone mass and bone markers of different doses of ibandronate: A new bisphosphonate for prevention and treatment of postmenopausal osteoporosis: A 1–year, randomized, double-blind, placebo-controlled dose-finding study. Bone 19, 527 – 533 (1996). 84. J. C. Netelenbos, F. C. van Ginkel, P. Lips, O. R. Leeuwenkamp, R. Barto, W. J. F. van der Vijgh, H van der Wiek, and W. H. L. Hacken, Effect of single infusion aminohydroxypropylidene on calcium and bone metabolism in healthy volunteers monitored during 2 months. J. Clin. Endocrinol. Metab. 72, 223 – 228 (1991). 85. S. Adami, N. Zamberlan, M. Mian, R. Dorizzi, M. Rossini, B. Braga, D. Gatti, F. Bertoldo, and V. Locascio, Duration of the effects of intravenous alendronate in postmenopausal women and in patients with primary hyperparathyroidism and Paget’s disease of bone. Bone Miner. 25, 75 – 82 (1994). 86. H. G. Bone, S. Adami, R. Rizzoli, M. Favus, P. D. Ross, A. Santora, S. Prahalada, A. Daifotis, J. Orloff, and J. Yates, Weekly administration of alendronate: Rationale and plan for clinical assessment. Clin. Ther. 22, 15 – 28 (2000). 87. T. Schnitzer, H. G. Bone, G. Crepaldi, S. Adami, M. McClung, D. Kiel, D. falsenberg, R. R. Recker, R. P. Tonino, C. Roux, A. Pinchera, A. J. Foldes, S. L. Greenspan, M. A. Levine, R. Emkey, A. C. Santora, A. Kaur, D. E. Thompson, J. Jates, and J. J. Orloff, Therapeutic equivalence of alendronate 70 mg once–weekly and alendronate 10 mg daily in the treatment of osteoporosis. Aging Clin. Exp. Res. 12, 1 – 12 (2000). 88. R. Lindsay, F. Cosman, R. A. Lobo, et al., Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84, 3076 – 3081 (1999). 89. H. G. Bone, S. L. Greenspan, C. McKeever, N. Bell, M. Davidson, R. W. Down, R. Emkey, P. J. Meunier, S. S. Miller, A. L. Mulloy, R. R. Recker, S. R. Weiss, N. Heyden, T. Musliner, S. Suryawanshi, A. J. Yates, and A. Lombardi, Alendronate and estrogen effects in postmenopausal women with low bone mineral density. J. Clin. Endocrinol. Metab. 85, 720 – 726 (2000). 90. S. J. Wimalawansa, A four-year randomized controlled trial of hormone replacement therapy and bisphosphonate, alone or in combination, in women with postmenopausal osteoporosis. Am. J. Med. 104, 219 – 226 (1998). 91. O. Johnell, Y. Lu, E. Seeman, J. Reginster, and W. Scheele, Effects of raloxifene and alendronate on bone mineral density and bone turnover markers in postmenopausal women with osteoporosis. Osteoporosis Int. 11 (Suppl. 2), S184 (2000). 92. R. M. Neer, C. Arnaud, J. R. Znchetta, R. Prince, G. A. Gaich, J. Y. Reginster, A. B. Hodsman, E. F. Eriksen, D. Mellstrom, S. IshShalom, E. S. Oefjord, E. Marcinowska-Suchowierska, J. Salmi, L. Gaspar, H. Mulder, J. Halse, A. Z. Sawicki, H. Genant, O. Wang, and B.H. Mitlak, Recombinant human PTH [rhPTH (1 – 34)] reduces the risk of spine and non-spine fractures in postmenopausal osteoporosis. Abstr. Endocrinol. Soc. pp. 42 (2000). 93. R. S. Rittmaster, M. Bolognese, M. P. Ettinger, D. A. Hanley, A. B. Hodsman, D. L. Kendler, and C. J. Rosen, Enhancement of bone mass in osteoporotic women with parathyroid hormone followed by alendronate. J. Clin. Endocrinol. Metab. 85, 2129 – 2134 (2000). 94. R. Lindsay, J. Nieves, C. Formica, et al., Randomized controlled study of effect of parathyroid hormone on vertebral bone mass and fracture incidence among postmenopausal women on estrogen with osteoporosis. Lancet 350, 550 – 555 (1997). 95. M. Horwitz, A. Stewart, and S. L. Greenspan, Editiorial: Sequential parathyroid hormone/alendronate therapy for osteoporosis —Rob-

649

bing Peter to pay Paul ? J. Clin. Endocrinol. Metab. 85, 2127 – 2128 (2000). 96. P. Garnero, W. J. Shih, E. Gineyts, D. B. Karpf, and P. D. Delmas, Comparison of new biochemical markers of bone turnover in postmenopausal osteoporotic women in response to alendronate treatment. J. Clin. Endocrinol. Metab. 79, 1693 – 1700 (1994). 97. J. O. Landman, D. H. Schweitzer, M. Frolich, N. A. T. Hamdy, and S. E. Papapoulos, Recovery of serum calcium concentrations following acute hypocalcemia in patients with osteoporosis on long–term oral therapy with the bisphosphonate pamidronate. J. Clin. Endocrinol. Metab. 80, 524 – 528 (1995). 98. F. Cosman, J. Nieves, J. Horton, V. Shen, and R. Lindsay, Effects of estrogen on the response to edetic acid infusion in postmenopausal osteoporotic women. J. Clin. Endocrinol. Metab. 78, 939 – 943 (1994). 99. T. Storm, T. Steiniche, G. Thamsborg, and F. Melsen, Changes in bone histomorphometry after long–term treatment with intermittent cyclic etidronate for postmenopausal osteoporosis. J. Bone Miner. Res. 8, 199 – 208 (1993). 100. S. M. Ott, G. C. Woodson, W. E. Huffer, P. D. Miller, and N. B. Watts, Bone histomorphometric changes after cyclic therapy with phosphate and etidronate disodium in women with postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 78, 968 – 972 (1994). 101. D. W. Axelrod, and S. L. Teitelbaum, Results of long-term cyclical etidronate therapy: Bone histomorphometric and clinical correlates. J. Bone Miner. Res. 9 (Suppl.) S136(1994). 102. T. Thomas, M. H. Lafage, and C. Alexandre, Atypical osteomalacia after 2 year etidronate intermittent cyclic administration in osteopororsis. J. Rheumatol. 22, 2183 – 2185 (1995). 103. S. J. Wimalawansa, Combined therapy with estrogen and etidronate has an additive effect on bone mineral density in the hip and the vertebrae: Four year randomized study. Am. J. Med. 99, 36 – 42 (1995). 104. H. G. Bone, R. W. Downs Jr., J. R. Tucci et al., Dose response relationships for alendronate treatment in osteoporotic elderly women. J. Clin. Endocrinol. Metab. 82, 265 – 274 (1997). 105. M. McClung, B. Clemmensen, A. Daifotis et al., Alendronate prevents postmenopausal bone loss in women with osteoporosis. Ann. Int. Med. 128, 253 – 261 (1998). 106. F. Eggelmeijer, S. E. Papapoulos, H. C. van Passen, B. A. C. Dijkmans, R. Valkema, M. L. Westedt, J. O. Landman, E. K. J. Pauwels, and F. C. Breedveld, Increased bone mass with pamidronate treatment in rheumatoid arthritis: Results of three–year randomized, double blind trial. Arthritis Rheum. 39, 396 – 402 (1996). 107. E. L. van Persijn-van Meerten, H. M. Kroon, and S. E. Papapoulos, Epi- and Metaphyseal changes in children caused by administration of bisphosphonates. Radiology 184, 249 – 254 (1992). 108. C. Brumsen, N. A. T. Hamdy, and S. E. Papapoulos, Long-term effects of bisphosphonates on the growing skeleton: Studies of young patients with severe osteoporosis. Medicine 76, 266 – 283 (1997). 109. F. H. Glorieux, N. J. Bishop, H. Plotkin, G. Chabot, G. Lanogu, and R. Travers, Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N. Engl. J. Med. 339, 947 – 952 (1998). 110. S. Adami and N. Zamberlan, Adverse effects of bisphosphonates. A comparative review. Drug Safety 14, 158 – 170 (1996). 111. E. G. Lufkin, R. Argueta, M. D. Whitaker, A. L. Cameron, V. H. Wong, K. S. Egan, W. M. O’Fallon, and B. L. Riggs, Pamidronate: An unrecognized problem in gastrointestinal tolerability. Osteoporosis Int. 4, 320 – 322 (1994). 112. P. C. DeGroen, D. F. Lubbe, L. J. Hirsch, A. Daifotis, W. Stephenson, D. Freedholm, S. Pryor-Tillotson, M. J. Seleznick, H. Pinkas, and K. K. Wang, Esophagitis associated with the use of alendronate. N. Engl. J. Med. 335, 1016 – 1021 (1996). 113. D. C. Bauer, D. Black, K. Ensrud, D. Thompson, M. Hochberg, M. Nevitt, T. Musliner, and D. Freedholm, Upper gastrointestinal tract safety profile of alendronate. Arch. Intern. Med. 160, 517 – 525 (2000).

650 114. A. A. Reszka, J. M. Halasy, and G. A. Rodan, Alendronate and risedronate arrest keratinocyte cell growth in a model for esophageal irritation via effects on cyclin-dependent kinases. Osteoporosis Int. 11 (Suppl 2), S57 (2000). 115. H. M. Bounameaux, J. Sheifferli, J-P. Montani, A. Jung, and F. Chatelanat, Renal failure associated with intravenous bisphosphonates. Lancet I, 471 (1983). 116. O. L. M. Bijvoet, W. B. Frijlink, K. Jie, H. van der Linden, C. J. L. M. Meijer, H. Mulder, H. C. van Passen, P. H. Reitsma, J. te Velde, E. de Vries, and J. P. van der Wey, APD in Paget’s disease of bone. Role of the mononuclear phagocyte system. Arthritis Rheum. 23, 1193 – 1204. (1980). 117. S. Adami, G. Salvagno, G. Guarrera, M. Montesant, S. Garavelli, S. Rosini, and V. Locascio, Treatment of Paget’s disease of bone with intravenous 4–amino–1–hydroxybutilydene–1,1–bisphosphonate. Calcif. Tissue Int. 39, 226 – 229 (1986).

SOCRATES E. PAPAPOULOS 118. S. Adami, A. K. Bhalla, R. Dorizzi, F. Montesant, S. Rosini, G. Salvagno, and V. Locascio, The acute phase response after bisphosphonate administration. Calcif. Tissue Int. 41, 326 – 331 (1987). 119. D. H. Schweitzer, M. Oostendorp-van de Ruit, G. van der Pluijm, C. W. G. M. Lowik, and S. E. Papaoulos, Interleukin-6 and the acute phase response during treatment of patients with Paget’s disease with the nitrogen–containing bisphosphonate dimethylaminohydroxypropylidene bisphosphonate. J. Bone Miner. Res. 10, 1 – 7 (1995). 120. A. Sauty, M. Pecherstorfer, I. Zimmer-Roth, P. Fioroni, L. Juillerat, M. Markert, H. Ludwig, P. Leuuenberger, P. Burckhardt, and D. Thiebaud., Interleukin-6 and tumor necrosis factor  levels after bisphosphonate treatment in vitro and in patients with malignancy. Bone 18, 133 – 139 (1996).

CHAPTER 73

Calcitonin for Treatment of Osteoporosis ROBERTO CIVITELLI

Division of Bone and Mineral Diseases, Washington University School of Medicine, and Barnes-Jewish Hospital, St. Louis, Missouri 63110

IV. Analgesic Action V. Future Directions References

I. Introduction II. Clinical Efficacy in Osteoporotic Syndromes III. Clinical Pharmacology

I. INTRODUCTION

anti-resorptive agent in the treatment of osteoporosis is based on the assumption that the inhibition of the resorptive phase of bone remodeling results in filling of the remodeling space, with consequent increased bone mass [2,3]. The same rationale can be applied to most of the currently available anti-resorptive agents used for the treatment of osteoporosis, namely estrogen and bisphosphonates. This paradigm leads to the presumption that the therapeutic effect of resorption inhibitors, including calcitonin, should be transient and limited by the extent of the remodeling space being filled. Once this process is complete, formation and resorption recouple, and no further changes in bone mass should occur. The experimental demonstration of this assumption by long-term studies wherein calcitonin was administered continuously has subsequently led to the design of therapeutic regimens utilizing intermittent administration of calcitonin, alone or in combination with other drugs. As an additional consequence of its action as a bone resorption inhibitor, one should also expect calcitonin to be more effective in individuals with high turnover forms of osteoporosis than those with less rapid remodeling rates. In fact, this hypothesis has been substantiated in some clinical trials, and a similar heterogeneity of clinical responsiveness may be anticipated for estrogen and bisphosphonates. These findings may be important for identifying the most appropriate candidates for anti-resorptive therapy.

Calcitonin was identified in 1962 as a hormonal factor that could lower plasma calcium, and it was named after this effect [1]. The structure, biosynthesis, and physiology of this calciotropic hormone have been defined, and its action as direct inhibitor of osteoclast bone resorption established (see Chapter 8). Consequently, the hormone has been utilized for more than 15 years to effectively inhibit or reverse bone loss in many pathologic conditions characterized by increased bone resorption. Reports of therapeutic responses vary from those demonstrating no substantial effects or simple normalization of accelerated rates of bone loss to those demonstrating reversal of the osteoporotic process, with increased bone mass.

II. CLINICAL EFFICACY IN OSTEOPOROTIC SYNDROMES A. Rationale for the Use of Calcitonin in Osteoporosis Because of its ability to inhibit osteoclastic activity, calcitonin has been used as a remedy in several forms of osteoporosis for the past 15 to 18 years. The use of an

OSTEOPOROSIS, SECOND EDITION VOLUME 2

651

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

652

ROBERTO CIVITELLI

Observations of abnormal circulating calcitonin levels or reduced secretory reserve of calcitonin in women with osteoporosis theoretically lend additional support to the use of calcitonin in osteoporosis. However, this hypothesis has never been conclusively proven, and considering that the pharmacologic concentrations required to achieve significant effects on bone density are much higher than the physiologic levels of the hormone, it is not very likely that the putative pathogenetic role of calcitonin in estrogen-dependent bone loss constitutes a therapeutic rationale for the use of this drug in postmenopausal osteoporosis.

B. Postmenopausal Osteoporosis Difference in total body calcium (mean  SEM) between women treated with injectable salmon calcitonin and control subjects, at different times after initiation of therapy. Reprinted with permission fom Gruber et al. [9].

FIGURE 1 1. TREATMENT OF ESTABLISHED OSTEOPOROSIS The first clinical studies that demonstrated the potential beneficial effect of calcitonin on age-related osteoporosis were performed in the early 1970s using radiocalcium kinetics and low doses of the hormone. In most, though not in all, subjects treated with calcitonin, calcium balance improved, bone accretion rate increased, and bone resorption decreased [4 – 6]. Positive effects on trabecular bone volume were also observed in a small number of patients treated with subcutaneous salmon calcitonin and oral phosphate [7], suggesting that at least in certain circumstances calcitonin could improve bone mass in osteoporotic subjects. In contrast to these observations, Jowsey et al. [8] did not detect any benefits on bone mass — measured by videodensitometry of forearm radiographs — by adding calcitonin to a high calcium and vitamin D regimen over a 15month period of time. Subsequent studies were performed in a more controlled fashion, utilizing more reliable and precise techniques to assess the action of calcitonin on bone mineral density. The milestone study of Gruber et al. [9] compared the effect of a 2-year treatment with 100 IU subcutaneous salmon calcitonin with that of calcium and vitamin D (400 IU/day) alone. Although the average total body calcium, measured by neutron activation analysis, was not different between treated and control groups throughout the study, the changes normalized to baseline values were significantly positive in the calcitonin-treated group compared to the control population up to 18 months of therapy. However, there was a tendency toward a decrease thereafter (Fig. 1). While this finding has often been interpreted as a consequence of clinical resistance to calcitonin (see section III), it may simply confirm the assumption that the effect of an anti-resorptive agent is limited to the filling of the available remodeling space, as discussed above. Since urinary calcium increased during calcitonin treatment, the authors concluded that the increase in total body calcium was probably mediated by an increase in intestinal absorption of calcium. The results were convincing enough for the United

States Food and Drug Administration to approve the use of injectable calcitonin for the treatment of osteoporosis in 1984. These results were subsequently confirmed and extended by several European studies. Significant increases of bone density measured by dual-energy absorptiometry (DPA) were reported at the spine, femoral diaphysis [10], and nondominant forearm [11,12] in subjects treated for 1 year with 100 IU daily of subcutaneous salmon calcitonin. Intermittent administration of injectable salmon calcitonin (100 IU every other day) resulted in less pronounced effects [10]. Because of their better tolerability and improved compliance, most of the later clinical trials on calcitonin have been conducted using nasal spray preparations. At the dose of 200 IU/day, salmon calcitonin nasal spray prevented bone loss at the proximal and distal radius in women with established osteoporosis [13]. In a longer term, dose-finding study the same investigators observed significant, doserelated increase in spinal bone density (Fig. 2), but not at the forearm [14]. Likewise, another group of Danish investigators detected significant increments in bone density of postmenopausal women with Colles’ fractures only with daily doses of 200 IU intranasal salmon calcitonin, and only at the spine, with no effect at the distal radius [15]. Furthermore, the increase in lumbar spine bone density in the calcitonin-treated groups was significant compared to baseline, but not compared to the placebo group [16]. Protection from spinal bone loss was also observed by Ellerington et al. [17] in postmenopausal women with osteopenia using the same daily dose of 200 IU salmon calcitonin nasal spray. A thrice weekly dose was totally ineffective [17]. However, intermittent regimens of 200 IU salmon calcitonin nasal spray given daily, 1 month on and 1 month off [18] or 2 months on and 2 months off [19], to postmenopausal women were sufficient to induce significant increases in axial and appendicular bone density after 2 years

653

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

FIGURE 2 Percentage changes in bone mineral content of the lumbar spine (mean on left; mean and SEM at 1 and 2 years on right) during a 2-year treatment with different daily doses of intranasal salmon calcitonin (– – – –, 50 IU; — — — , 100 IU; ——— , 200 IU) or placebo (—— ). P values are relative to the correlation between calcitonin dose and response. Reprinted with permission from Overgaard et al. [14]. of treatment. In one of these studies, the efficacy of cyclic calcitonin regimens on bone density was demonstrated by both dual-energy X-ray absorptiometry (DXA) and ultrasonographic methods (Fig. 3) [18]. However, the recently completed, large multicentric “PROOF” study on the effect of calcitonin on vertebral fractures — which will be discussed more in-depth later — failed to show any effects of different doses of salmon calcitonin nasal spray on proximal femur bone density [20]. Although with some degree of variability and a few discrepancies, most of the studies reviewed above would suggest that calcitonin can be effectively used to inhibit bone loss or to increase bone density at least in the spine of patients with established osteoporosis. In quantitative terms, the increases in vertebral bone density reported in subjects exposed to calcitonin nasal spray rarely exceed 2% of baseline, at least in the trials wherein DXA techniques were used to measure bone density. A similar degree of change has been observed in the PROOF study, with approximately 1.5% increase in spine bone density in women treated with 200 IU daily salmon calcitonin nasal spray [20]. The effect was evident as early as after 1 year without further changes over subsequent 4 years of continuous treatment. Thus, the magnitude of calcitonin effect on bone density would appear substantially lower compared to the changes reported in subjects treated with alendronate [21] or estrogen [22], but not much lower than with raloxifene [23]. However, such comparisons should not be overinterpreted, since the different outcomes derive from independent studies whose patient populations differ in initial bone density, prevalent fractures, and other critical features. To date, there are only a few head-to-head comparative studies of calcitonin and alendronate on postmenopausal women with osteopenia,

and these are not overly positive as far as the action of calcitonin on bone density. An earlier Italian study showed no significant effects of calcitonin nasal spray on vertebral bone density, although the dose employed, 100 IU daily, was 50% lower than that currently used in the United Stated for osteoporosis treatment [24]. In two larger, more recent trials testing the two drugs at the doses available in the market there was no difference in vertebral bone density between the calcitonin and placebo groups, whereas alendronate-treated women experienced more than 4% increment in bone density [25,26]. Intriguingly, in the femoral neck calcitonin appeared to have an intermediate effect between placebo and alendronate [26]. Because of their preliminary nature, one should exercise caution before drawing definitive conclusions on these comparative studies. One critical factor that determines the response to an antiresorptive agent is the initial level of bone density, that is the severity of osteoporosis. Women with lower bone densities experience larger bone gains after treatment with estrogen relative to women with higher bone densities [27]. Similarly, protection from osteoporotic fractures was obtained with alendronate therapy in women with vertebral fractures or low bone density, but not in those with osteopenia or normal bone density [28]. The same paradigm applies to calcitonin. In the study of Ellerigton et al. [17], calcitonin proved effective in preventing bone loss in late postmenopausal women (3.1% difference with placebo at 2 years), but not in early postmenopausal subjects with higher bone density. These findings may be in part related to correction of a higher degree of bone turnover in subjects with lower bone density, since individuals with accelerated bone remodeling tend to loose bone at a faster rate than those with normal turnover

654

ROBERTO CIVITELLI

Percentage changes (mean  SEM) in bone mineral density of the lumbar spine (BMD-LS), speed of sound (SOS), stiffness, and broadband ultrasound attenuation (BUA) in osteoporotic patients treated with either salmon calcitonin nasal spray and calcium (), or calcium alone () for 2 years. *P  0.05. **P  0.01, ***P  0.001 versus time 0; #P0.01 between groups. Reprinted with permission from Gonnelli et al. [18].

FIGURE 3

[29,30]. The results of a study from our group provided proof of this hypothesis. Postmenopausal women with osteoporosis and high bone turnover responded to treatment with subcutaneous salmon calcitonin, 50 IU every other day, with significant gains in vertebral bone density, whereas no changes were observed in individuals with normal bone remodeling [31]. There was also a significant correlation between changes in bone density and changes in bone turnover, assessed as the whole body retention of injected 99mTc-methylenbisphosphonate (Fig. 4). The wide fluctuations in bone density observed in this study may be related to both the wide range of turnover rates in these patients, and the relatively lower long-term precision of the DPA technology, compared to the currently used DXA. These observations had been later confirmed by Overgaard and coworkers [32], who were able to predict the response to calcitonin therapy (200 IU/day nasal spray) in terms of changes in bone density using baseline measurements of biochemical markers of bone turnover. The clinical relevance of these studies is obvious, but the implementation of an effective screening for identifying those subjects that would better respond to calcitonin treatment relies entirely upon our ability to precisely predict bone loss by a single

assessment of bone turnover in individual patients, which is far from being optimal (Chapter 60). The current lack of valid alternatives to anti-resorptive medications for the treatment of osteoporosis somewhat reduces the value of

FIGURE 4

Correlation of changes in bone mineral content at the lumbar spine (BMC-LS) and changes in whole body retention of 99mTc-methylenbisphosphonate (percentage from baseline) in 50 postmenopausal patients with high turnover () or normal turnover () osteoporosis after 1-year therapy with injectable salmon calcitonin. Reprinted with permission from Civitelli et al. [31].

655

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

screening programs based on bone turnover for long-term therapeutic decisions.

TABLE 1 Effect of Calcitonin on Vertebral Fracture Rate: Number of New Fractures during the Observation Period

2. PREVENTION OF POSTMENOPAUSAL BONE LOSS In 1987, Reginster, et al. [33] reported that intranasal calcitonin (50 IU five times a week) effectively prevented vertebral bone loss in women who had been menopausal for no more than 36 months and as such most likely to have high turnover bone remodeling. There was in fact a remarkable loss in the placebo group in the initial 12 months of observation (approx. 4%). This sharp bone loss, despite calcium supplementation, may raise some concern about a possible selection bias toward enrollment of subjects with the highest turnover rates in this study. Interestingly, the fast bone loss did not continue at the same rate in a 3-year follow-up, although the protective action of calcitonin was confirmed [34]. Except for two other small-sized studies showing protective effects from bone loss with low doses of either injectable human calcitonin (0.1 mg, equivalent to 20 IU) [35] or subcutaneous eel calcitonin (40 IU twice a week) [36], significant effects on bone density have not been confirmed by other investigators with low-dose nasal spray calcitonin in early postmenopausal women [14,15,37]. Overgaard et al. [38], using a double-blind, placebo-controlled protocol in healthy women who had been menopausal for no more than 5 years, demonstrated that a daily dose of 100 IU salmon calcitonin nasal spray prevented vertebral bone loss for 2 years, but they were not able to detect an effect on the appendicular skeleton. In a second study, the same investigators were unable to detect significant effects with 100 IU salmon calcitonin nasal spray daily. Although higher doses (200 and 400 IU daily) were effective in preventing vertebral bone loss in younger women, they also had no effect on forearm bone density [37]. The discrepancy between the results obtained on the vertebrae (mostly trabecular bone) and those on the distal radius (mostly cortical bone) has been attributed to the higher bone turnover in the trabecular bone, as compared to the cortical bone [37]. Gennari et al. [39], using salmon calcitonin nasal spray in a group of early postmenopausal women with high bone turnover at a dose of 200 IU on alternate days reported increments in vertebral bone mass after 12 months in the calcitonin-treated patients, as opposed to a control group of untreated patients who demonstrated significant bone loss. In summary, although there is reasonable evidence suggesting that calcitonin, at the doses used to treat osteoporosis, may also prevent bone loss in early postmenopausal women, lack of reproducible data demonstrating positive effects on both vertebral and appendicular bone, and more importantly, lack of large, controlled prevention studies mitigate the enthusiasm for the use of calcitonin in nonosteoporotic populations. As noted above, the effectiveness of calcitonin on bone density is not all that evident in early

Study

Duration (years)

Control Patients

Treated

Fractures

Patients

Fractures

Gruber [9]

2

21

2

24

6

Gennari [10]

1

15

7

30

7

Rico [40]

2

27

23

28

5

Overgaard [14]

2

40

6a

124

4a

a

New patients with fractures.

postmenopausal women subjects with moderate osteopenia [17], or in osteoporotic patients without accelerated turnover rates [31]. Accordingly, this drug has not been registered with the Food and Drug Administration for use in the prevention of early postmenopausal bone loss. 3. PREVENTION OF OSTEOPOROTIC FRACTURES The incidence of vertebral fractures had been followed during calcitonin treatment in four placebo-controlled studies [9,10,14,40]. Table 1 summarizes the fracture data obtained in these early studies conducted for 1 year or longer and with a control group treated with calcium only. Although all these studies report lower fracture events in calcitonin-treated patients relative to placebo, they were either not designed to test the effect of calcitonin on fracture rate [9,10], or were grossly underpowered to assess such an outcome [40]. In one case [40], the unusually high incidence of fractures (40 fractures/100 person-years at two years) in the calcium-treated subjects suggests enrollment of a population at very high risk of fracture and therefore not representative of postmenopausal women in general. A larger population was followed by Overgaard et al. [14], although the dose – response design required allocation of three-fourths of the study population to treatment with active medication, with consequent loss of power. A significantly lower rate of new vertebral fractures was calculated for the calcitonin-treated subjects compared to the controls. However, since the number of new patients with fractures was very small in both groups (Table I), the validity of the results when expressed as fracture rates must be interpreted with caution. The PROOF (Prevent Recurrence of Osteoporotic Fractures) study was a randomized, double-blind, placebo-controlled trial designed to test the efficacy of different doses of calcitonin nasal spray (100, 200, and 400 IU daily) on incidence of new vertebral fractures [20]. A total of 1255 postmenopausal women with established osteoporosis were enrolled in this multicenter study. Results of the 5-year data (Table 2) indicate that the relative risk of developing a new vertebral fracture in the group taking 200 IU nasal calcitonin was reduced 36% (P  0.03) compared to placebo;

656

ROBERTO CIVITELLI

TABLE 2

Patients with 1 new vertebral fracture [n (%)] Relative risk (95% CI) Patients with 2 new vertebral fracture [n (%)] Odds ratio (95% CI) New vertebral fractures/ 1000 patients-X-ray-years

Vertebral Fracture Analyses for the Entire PROOF Study Cohort Placebo n  270

100 IU n  273

Salmon calcitonin nasal spray 200 IU n  287

400 IU n  278

70 (25.9)

59 (21.6)

51 (17.8)

61 (21.9)

1

0.85 (0.60 – 1.21)

0.67 (0.47 – 0.97)

0.84 (0.59 – 1.18)

33 (12.2)

34 (12.5)

24 (8.4)

30 (10.8)

1

1.02 (0.64 – 1.88)

0.65 (0.38 – 1.14)

0.87 (0.41 – 1.30)

131

129

78

111

and that spinal deformity index, a semiquantitative method of assessing vertebral fractures, was also reduced 52% [41]. Accordingly, the number of new vertebral fractures/1000 patient-X-ray-years was reduced 40% (P  0.02). On the other hand, the lower dose of calcitonin had no effect, and there was only a minor, nonsignificant reduction of relative fracture risk in the 400-IU group. The same results were obtained either with an intention-to-treat analysis or in a 3-year analysis of valid completers. In this latter population, an analysis based on the Kaplan-Meier survival curve showed that 11 patients were needed to be treated for 3 years with 200 IU salmon calcitonin nasal spray to prevent one fracture [20]. While it is reassuring that the calcitonin nasal spray preparation currently used for the treatment of osteoporosis is effective in reducing incident vertebral fractures, the lack of dose – response on fractures and on bone density is a bit disturbing. Although one can argue that 400 IU daily is too high a dose of calcitonin and it could produce compensatory changes in parathyroid hormone secretion, the same dose was found to be as effective as 200 IU in a previous study [37], and it had the same biologic activity as the 200-IU dose in the PROOF cohort [20]. A second problem with the PROOF study is the unexpectedly high dropout rate (59%) at 5 years. However, the different treatment groups were well matched, and the dropouts in the placebo group had a higher decrease in bone density than those who had taken the active medication. Thus, it is unlikely that the high attrition may have significantly biased the results. Finally, one has to contend with the significant reduction in vertebral fracture incidence in spite of marginal changes in bone density. The same issue applies to a similar drug effectiveness trial, the MORE study on raloxifene [23]. While it is certainly possible that factors other than bone density play a role in the action of a drug on the bone tissue, the apparent discrepancy in effect sizes on bone density and fracture rates can be explained at least in part by the different fracture risk in the populations enrolled. The correlation between bone density and fracture risk is exponential, not linear, so that at low bone density levels,

modest changes in density result in substantial reductions of fracture risk [42]. Furthermore, subjects with a prevalent fracture have a much higher risk of experiencing a new fracture than subjects without one fracture [43]. Women enrolled in the PROOF study had spine bone densities below 2 T scores and most of them (73%) had one radiologically confirmed prevalent fracture at enrollment; thus they were at high risk of fracturing. Consequently, even a modest change in bone density (2%) may have been sufficient to cause a 36% reduction in incident fractures. Nonetheless, the problem remains as to why with the same degree of bone density changes did the 400-IU calcitonin formulation not protect from vertebral fractures; and why did subjects with one to five prevalent vertebral fractures — thus at higher risk of fracture — experience the same reduction in vertebral fractures as the entire cohort [20]. A retrospective European study designed to examine the effect of taking drugs that affect bone metabolism on the risk of hip fractures and the PROOF study are the only available sources for data on the effect of calcitonin on hip fractures. The Mediterranean Osteoporosis (MEDOS) study was a retrospective analysis on fractures based on questionnaires involving 14 European centers. The inclusion of a large number of subjects (more than 5500) and the case – control design afforded a strong power (80%) to detect statistical differences. The results on hip fracture prevalence on this large cohort of women showed that taking either calcitonin, or estrogen, or calcium alone significantly decreased the risk of hip fractures [44]. After adjustment for other risk factors, including use of estrogen or calcium supplements, the relative risk of hip fractures was 0.69 (0.51 to 0.92 confidence interval) in women taking calcitonin therapy for a median duration of 2 years [44]. Thus, within the limitations imposed by its retrospective nature and by the differences among European countries in record keeping, data collection, and use of drugs for osteoporosis, this report seems to provide comforting results for the use of calcitonin in established osteoporosis. However, data obtained from retrospective analyses must not be overinterpreted. The

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

recent negative experience with the HERS study on the effect of estrogen replacement therapy on secondary prevention of cardiovascular disease demonstrates that conclusions from observational surveys can be challenged by prospective, controlled studies [45]. Unfortunately, prospective hip fracture studies are not available for calcitonin, and the PROOF study showed a nonsignificant trend in reduced hip fractures in the 200 IU calcitonin nasal spray group, although the trial was not powered to test the effect of the drug on nonvertebral fractures. Intriguingly, in the PROOF study a 36% reduction of all nonvertebral fractures was demonstrated only with the lowest (100 IU) dose, which was otherwise ineffective on vertebral fractures [20]. The biologic significance of this latter finding remains uncertain. 4. COMBINATION THERAPY As an anti-resorptive agent, calcitonin could theoretically be used in combination with “activators” of the remodeling cycle, according to the ADFR (activate, depress, free, repeat) concept. Indeed, some attempts have been made in this direction, but the use of calcitonin in ADFR regimens has not generated as much enthusiasm as other resorption inhibitors, such as bisphosphonates. Patient acceptance and tolerability for the parenteral preparations used initially and the increasing competition from powerful bisphosphonates represent the major disadvantages for calcitonin in this regard. In addition, the momentum gained in the past few years by parathyroid hormone analogs as anabolic therapy for osteoporosis have contributed to decrease the enthusiasm for multidrug regimens for osteoporosis. The first attempt at a “coherence” approach employed a combined regimen with oral phosphate — given as an activator of parathyroid hormone secretion — and subcutaneous calcitonin in postmenopausal osteoporotic women [7,46]. A 20 – 30% increase in trabecular bone volume was observed on bone histomorphometry only in those patients treated for 6 months with both phosphate and calcitonin [46]. Mean wall thickness, an index of bone-forming surfaces, increased only in the groups treated with phosphate; whereas the mean interstitial bone thickness, which is inversely related to the depth of the resorbing cavities, increased only in the calcitonin-treated subjects. Although this study was short-term and used a low dosage of calcitonin (50 IU/day, 5 days every third week), it proved that the combination of an activator and a resorption inhibitor could produce substantial increases of bone mass. Subsequently, Hesch and coworkers [47] used a complex regimen of intermittent parathyroid hormone (1 – 38 fragment) and calcitonin nasal spray in eight patients with low turnover osteoporosis, and reported an increase in bone mineral density in all of them after 14 months. However, no clear data on controls were available, and the results remain inconclusive as to the effect of such a combination regimen compared to each drug alone.

657 A more typical ADFR approach has been pursued by Hodsman and coworkers [48,49], who employed a synthetic parathyroid hormone fragment (1 – 38) as a direct activator of the remodeling cycle, given for 28 days (800 IU/day) followed by calcitonin (75 IU/day) for 42 days as a depressor. Histomorphometric and biochemical parameters of bone turnover were increased with parathyroid hormone only and declined somewhat under treatment with calcitonin, thus indicating that the expected biologic responses were achieved with this regimen [48,49]. In the lumbar spine, bone density increased in all subjects treated with parathyroid hormone, regardless of whether they were taking calcitonin. By contrast, there was a trend toward a decline in bone density on the femoral neck in the patients receiving sequential combination therapy and no changes in those taking parathyroid hormone alone, though there was no statistically significant group effect [49]. The conclusion is that there seems to be no particular advantage in a combined sequential therapy over the benefits that can be achieved with parathyroid hormone alone. Likewise, no improvement in bone density was obtained after 1 year of a combined regimen with calcitriol (0.5 g/day) and subcutaneous human calcitonin (0.5 mg/ day), compared to subjects treated with only calcium supplements at any of the skeletal sites measured (spine, distal radius, and proximal femur) [50]. Combination of calcitonin nasal spray (400 IU daily) and nadrolone decanoate, an anabolic steroid, may even be antagonistic on bone density. In postmenopausal women with fractures, no effect was observed with the combination regimen on vertebral bone density, whereas positive significant changes were noted in subjects taking individual drugs [51]. Calcitonin has been used in sequential regimens with growth hormone (GH), a multidrug approach also in line with the ADFR concept. Aloia et al. [52] used alternate doses of calcitonin (100 IU 4 days a week) and human pituitary-derived GH (6 IU 3 days a week), and compared the effects with those obtained with calcitonin alone at the same dose. After 2 years of therapy, there was no difference between the two groups in rates of change of total body calcium, whereas a clear loss was evident at the forearm in the group treated with growth hormone, implying that GH may in fact be detrimental to cortical bone [52]. The availability of recombinant GH preparations has allowed the design of larger studies, although the results have not been more comforting. Holloway et al. [53] treated women with osteopenia with cycles of GH (20 g/kg/day) for 7 days followed by five daily injections of calcitonin (100 IU/day) and then 44 days on calcium supplements only for 2 years, corresponding to 12 treatment cycles. They found significant increases of bone density at the lumbar spine in women taking GH with or without calcitonin and a nonsignificant increase in those taking calcitonin alone. In the proximal femur (total hip), there were marginal increases in all treated groups compared to placebo, but there was no

658

ROBERTO CIVITELLI

evidence of additive effects of the combination regimen (Fig. 5). Likewise, Gonnelli et al. [54] using a similar sequential protocol (GH, 12 IU/day for 7 days, followed by injectable salmon CT, 50 IU/day for 21 days, and by 61 days without treatment) did obtain stabilization of bone density with the combined therapy at the lumbar spine and forearm, but no additive effects after 2 years. In fact, patients treated with GH lost bone in the femoral shaft regardless of calcitonin, reinforcing the disturbing hypothesis that GH may cause cortical bone loss. In summary, cyclic therapy with GH with or without calcitonin does not seem particularly useful for therapy of postmenopausal osteoporosis.

Percentage changes (mean  SEM) in bone mineral content (BMC) of the forearm in women after oophorectomy treated with either injectable salmon calcitonin (), or a placebo (). The placebo group was crossed over to active treatment after 6 months. Reprinted with permission from Mazzuoli et al. [55].

FIGURE 6

C. Osteoporosis in Surgically-Induced Menopause

FIGURE 5 Percentage changes (mean  SEM) in lumbar spine (top) and proximal femur (total hip, bottom) bone density in women with osteopenia treated for 2 years with repeated cycles of either growth hormone and injectable calcitonin or their respective placebos, as indicated. Reprinted with permission from Holloway et al. [53].

The ovariectomy-induced model of estrogen-dependent bone loss has not been used as extensively as other forms of osteoporosis in clinical studies on calcitonin. Mazzuoli et al. [55] treated a small group of women with either injectable salmon calcitonin, 100 IU every other day, or placebo for 1 year starting from a week after ovariectomy. They observed a prophylactic action of the drug on the fast bone loss that occurred in the placebo group within 6 months after surgery (Fig. 6). Salmon calcitonin nasal spray at the dose of 200 IU daily, given either continuously or intermittently (3 months on, 3 months off) has also been shown to prevent loss of forearm bone density in ovariectomized women [56]. It should be noted that in neither study were data on vertebral bone mass available. According to a more recent study, calcitonin seems to provide weak protection against ovariectomy-induced bone loss at the lumbar spine. Treatment with 100 IU salmon calcitonin nasal spray only partially prevented bone loss at the spine, and the treated group still lost 2% bone density at the end of the 2-year treatment period [57]. It should be noted that the dose of nasal spray calcitonin used in this study is lower than that used in postmenopausal osteoporosis. Interestingly, although urinary hydroxyproline consistently declined in the calcitonin-treated subjects in all these studies, serum osteocalcin, a marker of bone formation, decreased only transiently and tended to increase with time [56,57]. This may reflect a favorable uncoupling of the remodeling cycle by the drug. Nonetheless, the data available remain rather scanty and inconclusive. In the absence of more convincing data, other alternatives should be

659

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

considered to prevent bone loss in ovariectomized women who are not good candidates for estrogen replacement therapy.

D. Glucocorticoid-Induced Osteoporosis The mechanisms by which glucocorticoids affect bone remodeling and cause rapid bone loss are reviewed in Chapter 44. Although glucocorticoid-induced osteoporosis is essentially an osteoblast disease, the relative uncoupling of the remodeling cycle and the mild secondary hyperparathyroidism present in this condition represent a rationale for using antiresorptive medications to prevent or reduce glucocorticoid-induced bone loss. Unfortunately, only a few reliable studies are available on the clinical use of calcitonin in this condition. Continuous 1-year treatment with 100 IU subcutaneous salmon calcitonin reversed bone loss in 18 patients with chronic obstructive pulmonary disease compared to an equal number of matched individuals under the same conditions, taking only calcium supplements [58]. However, bone density was measured only at the radius, and the average dose of steroids during the observation period declined more in the calcitonin than in the control group. A similar regimen of injectable calcitonin (3 times per week) increased vertebral bone density after 1 year in another series of patients with asthma on a slightly lower dose of steroids [59]. Age-matched patients experienced a significant loss. However, there was an alarming dropout rate (35%) because of side effects and poor compliance [59]. A longer-term study on patients with active rheumatoid arthritis receiving low doses of prednisolone (median dose 8.7 mg/day) demonstrated a protective effect of continuous daily doses of 100 IU calcitonin nasal spray on proximal femur bone density (Fig. 7) [60]. Interestingly, no detectable bone loss occurred in the lumbar spine of these patients, perhaps reflecting the low dose of corticosteorids they were taking. Unfortunately, there was a tendency toward a decline in bone proximal femur bone density after 2 years of treatment, raising doubts about the efficacy of this calcitonin regimen for longer periods of time. Curiously, opposite results were obtained in a small cohort of patients with polymyalgia rheumatica who were treated with higher doses of glucocorticoids [61]. Prevention of bone loss was observed only in the lumbar spine but not in the proximal femur with 200 IU daily of nasal spray calcitonin, but the baseline difference in vertebral bone density between treatment groups was greater than the bone loss at 2 years . A few studies have evaluated whether calcitonin can prevent bone loss when started simultaneously with glucocorticoids. Montemurro et al. [62] were able to prevent bone loss in a small number of sarcoid patients on glucocorticoid therapy with intermittent dosing schedules of par-

FIGURE 7

Changes in bone mineral density (mg/cm2) in the lumbar spine and proximal femur in patients with rheumatoid arthritis on lowdose glucocorticoids, treated for 1 year with either 100 IU daily salmon calcitonin nasal spray and calcium () or calcium alone (). Data represent averages; all SD are higher than 25 mg/cm2 and are not indicated in the figure. Treatment effect was significant only for the femoral neck and Ward’s triangle at 12 months. Reprinted with permission from Kotaniemi et al. [60].

enteral and nasal forms of salmon calcitonin for a 2-year period. The protective effect was quite remarkable, considering the impressive 15% loss of trabecular bone in the spine (assessed by quantitative computed tomography QCT) after 1 year in patients treated with high doses of prednisone [62]. Larger studies in different types of pathologic conditions have reported less positive results. In patients starting long-term glucocorticoid therapy for a variety of reasons, addition of calcitonin (400 IU/day nasal spray) to calcitriol (0.5 to 1.0 g daily) did not provide a significant advantage over calcitriol alone, so long as active treatment was continued [63]. In fact, calcitonin appeared to prolong the protective effect of calcitriol for an additional year, whereas the calcitriol-only subjects lost as much vertebral bone as the calcium-treated controls during the second year, although these patients also received more steroids (Fig. 8). Unfortunately, whether calcitonin alone

660

ROBERTO CIVITELLI

after 2 years. However, no significant bone loss was observed in any of these patients, perhaps suggesting that calcium and vitamin D supplementation was sufficient to prevent bone loss in these patients who received low amounts of prednisone [65]. Interestingly, despite no detectable bone loss, there was a three-fold higher fracture incidence in this cohort compared to the expected incidence of postmenopausal women, but calcitonin had no effect of fracture rate [65].

E. Other Forms of Osteoporosis

FIGURE 8 Average changes of bone mineral density (expressed as percentage changes per year) at three skeletal sites in corticosteroid-treated patients. Group 1 received also calcitriol, intranasal calcitonin and calcium; group 2 received calcitriol, and calcium; group 3 calcium alone. Calcitriol and calcitonin were given for the first year only. At the lumbar spine, significant differences were observed between groups 1 and 3 (P  0.001) and groups 1 and 2 (P  0.026) at 1 year; and between groups 1 and 2 (P  0.014) at 2 years. No significant differences were detected at the other sites. Reprinted with permission from Sambrook et al. [63]. could have produced the same effect as calcitriol alone cannot be established because a group treated with calcitonin only was not included. Since the peptide does not accumulate in tissues, the mechanism for a putative prolonged action of calcitonin remain unexplained, although prolonged depression of bone turnover has been reported 6 – 8 months after discontinuation of calcitonin therapy in elderly women with vertebral fractures [64]. A more recent study on patients with temporal arteritis and polymyalgia rheumatica on low dose glucocorticoids does not help in this regard. There was no difference in lumbar spine bone density between calcitonin-treated and untreated subjects

Prolonged immobilization as a consequence of spinal injuries or in patients bedridden for other reasons, as well as weightlessness conditions of space flights leads to rapid and dramatic acceleration of bone remodeling and results in rapid and profound bone loss (see Chapter 46). Bone resorption inhibitors have been used under such conditions in the attempt to counteract the consequences of the increased bone turnover and resorption induced by immobilization. Initial studies with calcitonin in animal models of immobilization were controversial and inconclusive, and very few studies are available on immobilization or disuse osteoporosis in man. In an early study in normal men subjected to prolonged bed rest for 6 weeks, 100 IU injectable salmon calcitonin could not prevent the developing negative calcium balance [66]. Likewise, administration of 200 IU calcitonin nasal spray twice a day for only 18 days to patients immobilized for a prolapsed intervertebral disk resulted in a significant reduction of the increased urinary excretion of calcium and bone turnover markers compared to untreated subjects at absolute bed rest control [67]. These results suggest that calcitonin is potentially useful during the early phases of immobilization, when the abnormalities of bone remodeling are more pronounced [68]. However, active treatment may not significantly affect bone density, even if calcitonin is administered up to 3 months [69]. Therefore, the usefulness of calcitonin in immobilization osteoporosis remains unproven. Patients with rheumatoid arthritis develop generalized osteopenia and increased risk for vertebral and hip fracture even before undergoing glucocorticoid therapy [70,71]. Significant inhibition of bone resorption was obtained in patients on nonsteroidal anti-inflammatory drugs using a short (3 months) therapy with 200 IU thrice weekly salmon calcitonin nasal spray [72]. In a longer-term study which was double blind and placebo-controlled for 4 months and open for the next 36 months, patients with rheumatoid arthritis treated only with nonsteroidal anti-inflammatory drugs lost vertebral bone at a rate of 2% per annum and appendicular bone (distal radius) at 4.8% per annum. In contrast, the patients receiving calcitonin gained 1% in bone mineral density at the lumbar spine and lost no bone in the

661

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

distal radius [73]. The difference compared to the controls remained significant during the second year of treatment, even though the gain in bone mass was not maintained. Therefore, although calcitonin cannot alter the progression of articular bone erosion [73], this study suggests that the drug may be useful in patients with rheumatoid arthritis who develop rapid bone loss as an adjunct to anti-inflammatory or immunosuppressive therapy. Multiple myeloma is usually associated with generalized osteopenia, with or without hypercalcemia, and focal lytic/sclerotic lesions which may cause pathologic fractures. Antiresorptive medications may aid in managing the metabolic bone disorder. In a small series of myeloma patients, 3 months’ treatment with intranasal salmon calcitonin (200 IU/day) improved trabecular bone volume and cortical thickness and decreased bone turnover markers and serum calcium [74]. However, the efficacy of longer term calcitonin administration in patients with malignancies remains undetermined. As it occurs with estrogen deficiency, removal of male gonads also results in a rapid increase of bone turnover (see Chapter 42). Hence, a 3-month regimen with nasal spray calcitonin (100 IU/day) significantly reduced bone resorption parameters in orchidectomized men [75]. Calcitonin, like other resorption inhibitors, can be used under other conditions of osteoporosis in man, such as juvenile osteoporosis and other high bone turnover conditions of unknown etiology. In fact, because of the side effects and the yet unproven efficacy of testosterone in nonhypogonadal males with osteoporosis, nonhormonal medications remain the best option for managing male osteoporotic syndromes in most cases. The use of calcitonin is commonly advocated for symptomatic treatment of transient regional osteoporosis, a syndrome characterized by reversible, localized osteopenia, associated with pain, swelling, cutaneous dystrophic lesions, vasomotor instability, and impaired mobility of one extremity, usually triggered by a traumatic event or infection [76,77]. This syndrome has also been called algodystrophy, Sudek’s atrophy, or reactive sympathetic dystrophy when no evident cause is recognizable (see Chapter 56). It is believed, though not proven, that the osteopenia is caused by transiently accelerated bone remodeling, probably as a consequence of an increased blood flow [78]. Treatment with calcitonin nasal spray (100 IU/day) during the acute phase of transient osteoporosis reduced biochemical parameters of bone turnover in a small series of patients affected by this condition, and thus it was felt that it may contribute to the healing process [78]. Unfortunately, there was no control group in that study, and the condition is known to spontaneously subside within 2 – 3 months from the onset of symptoms. A reduction of pain was reported in a placebocontrolled study, but the effect on the patient’s ability to resume working activity was uncertain [79]. Therefore, it is not clear whether calcitonin may indeed affect the natural

course of this disease. In fact, other investigators have failed to demonstrate any effects on the clinical progression of the disorder, even using 400 IU of nasal calcitonin [80]. Calcitonin may nonetheless help reduce the pain and improve the range of motion of the affected extremity when given in the acute phase of transient osteoporosis [81]. There is some anecdotal evidence suggesting that calcitonin may stimulate bone formation, in addition to its antiresorptive activity. If this is the case, then calcitonin may be expected to help in the process of fracture healing. Although calcitonin has been used in the attempt to relieve the acute pain associated with vertebral fractures, no clinical study has been performed to determine whether the healing process itself can be altered by the drug. A recent review of studies of this kind performed in animals (mostly rat) underlined the difficulty in interpreting these results, because of the diverse experimental design and the variety of animal models and calcitonin preparations employed [82]. In most cases, no effects were seen, in others, only very minor biological differences of dubious significance were reported [82]. Sporadic attempts have been made to use calcitonin in osteogenesis imperfecta [83 – 85]. The biologic bases for a clinical effectiveness of calcitonin in this inborn error are not clear. Nonetheless, calcitonin may improve calcium balance and inhibit bone turnover in some patients with osteogenesis imperfecta [83]. Castells et al. [84] treated 48 children with osteogenesis imperfecta with salmon calcitonin (2 IU/kg three times a week) for up to 48 weeks and observed a reduction in annualized fracture rates, as compared to the rates of the pretreatment period. A decrease in fracture rate has also been reported in a series of patients with osteogenesis imperfecta treated sequentially with injectable and intermittent intranasal calcitonin regimens for 22 to 68 months [85]. Although these data are interesting, whether the reported reduced fracture incidence represents an effect of the therapy rather than a natural evolution of the disease remains to be established.

F. Considerations for Use of Calcitonin in Osteoporosis Injectable forms of calcitonin have been tested and used in the prevention and treatment of osteoporosis as well as in other demineralizing bone diseases, as detailed in the previous sections of this chapter. In August of 1995, just about 25 years since its discovery, the nasal spray formulation of salmon calcitonin received approval by the Food and Drug Administration for treatment of osteoporosis in the United States. Later in the same year, alendronate was introduced in the market for osteoporosis treatment, and since then the two drugs have represented valid alternatives for women who cannot or do not want take estrogen for osteoporosis.

662

ROBERTO CIVITELLI

With the approval of raloxifene, in early 1998, and of risedronate in 1999, the pharmacologic armamentarium available to the physician to prevent and treat bone loss has rapidly widened. Undoubtedly, the availability of new therapeutic options represents progress toward better health for patients with osteoporosis. On the other hand, the larger number of drugs of different classes has created some confusion among physicians as to what criteria to use for devising the most appropriate therapeutic regimen for each patient. Although a thorough discussion of this issue goes beyond the scope of this chapter, one should always keep in mind not only the efficacy of the drug on bone density and fracture incidence, but also the safety, tolerability, likelihood of compliance, and potential extraskeletal effects that may be beneficial to postmenopausal women. Based on bone density data, salmon calcitonin nasal spray emerges as the weakest among the currently available options for treating osteoporosis. However, at the dose of 200 IU daily, nasal calcitonin significantly reduces the incidence of new vertebral fractures in women with established osteoporosis. Therefore, at least in these subjects, calcitonin appears an option for prevention of new osteoporotic fractures. The best candidates for calcitonin are elderly women with low bone density who are not good candidates for estrogen replacement therapy, especially those with gastrointestinal intolerance to bisphosphonates. Patients with multiple medical problems necessitating large numbers of oral medications may also consider calcitonin nasal spray over an oral bisphosphonate for practical reasons and to avoid problems with absorption of bisphosphonate. On the other hand, the limited and still inconsistent data on younger women do not encourage the use of calcitonin for the prevention of osteoporosis, especially in consideration of the many options now available in this type of subjects. Although there is no formal FDA approval, salmon calcitonin nasal spray can be considered in patients taking relatively low doses of oral glucocorticoids and who are not good candidates for estrogen replacement or alendronate — the only drug approved by the US Food and Drug Administration for glucocorticoid-induced osteoporosis. Finally, calcitonin may be considered an option for male osteoporosis and other demineralizing syndromes with increased bone turnover, although the lack of strong, convincing evidence on bone density or fracture end-points limits the clinical relevance of such a treatment.

III. CLINICAL PHARMACOLOGY

tive, which improves its chemical stability. Of these, the last three are still used, but only synthetic salmon calcitonin is currently available in the United States. Calcitonin is a relatively conserved molecule, since the rat and human sequences share 30 of the 32 amino acids, and 29 residues are conserved between salmon and eel calcitonin. Despite these structural similarities, calcitonins of different species differ markedly in biologic potencies, which are classically measured by its hypocalcemic effect in the rat. One IU of calcitonin is defined as 1/100 of the amount of peptide necessary to produce a 10% reduction of blood calcium 1 h after intravenous injection to a young, fasting 150-g rat [86]. The international units are essentially the same as the old MRC (Medical Research Council) units, used in early studies. By this method, 1 IU of salmon calcitonin corresponds to 0.25 g of peptide, whereas 1 IU of human calcitonin corresponds to 5 g of peptide. Accordingly, the potency of 100 IU is obtained with either 25 g salmon or 0.5 mg human calcitonin, with a potency ratio of 20:1 [87]. However this ratio, defined in rat assays, may not translate into equivalent biologic effects in humans, because of substantial differences between calcitonin of different species in pharmacokinetic properties and receptor affinities. Although the apparent biologic half-lives after intravenous administration are short and quite similar among the calcitonin species available in therapy, relatively more marked differences of metabolic clearance values have been observed, with human calcitonin being cleared twice as fast as salmon calcitonin [87 – 89]. Different receptor affinities may provide an additional source of biologic variability of calcitonin preparations. For example, the potency rank for binding to kidney or bone rat cells was found to be salmon  porcine  human [90]. Gennari et al. [91] used the hypocalcemic assay in a group of healthy individuals to assess the biologic potency of human and salmon calcitonin in humans. They found that approximately 75 IU of human calcitonin was necessary to obtain the same effect of 50 IU of salmon calcitonin of serum calcium and cAMP stimulation [91]. The same investigators also reported slightly lower biologic potencies of eel calcitonin as compared to salmon calcitonin in man, despite the fact that the former had been found to be more potent in the rat hypocalcemic assay [92]. These relatively small differences ultimately have limited value for the therapeutic potential of each calcitonin preparation, since the choice of the appropriate formulation is based on clinical efficacy on bone density and fracture risk, not on short-term end points.

A. Pharmacologic Preparations

B. Administration Routes

Preparations of calcitonin from four different species have been developed for use in humans, i.e., porcine, human, salmon and eel, the last as amino-suberic acid deriva-

Less than optimal compliance to subcutaneous injections has constituted a serious limitation for a reliable evaluation of effectiveness of calcitonin in long-term studies

663

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

and has severely curtailed its clinical use. To overcome these limitations, nasal preparations have been developed for salmon, eel, and human calcitonin. Rapid biologic effects (transient hypocalcemia, phosphaturia, increase of serum parathyroid hormone, increase of urine calcium and cAMP production) have been consistently described after administration of a single spray of 50 – 400 IU of calcitonin [93 – 97]. As expected, the rise of immunoreactive exogenous calcitonin plasma level is slower than for parenteral injections, but sizable and more prolonged levels of the drug are achieved with nasal preparations [94,95] (Fig. 9). Although the biologic effects and total plasma levels of calcitonin after nasal administration are in general dosedependent, a spray containing 200 IU is required to obtain effects similar to those achieved with 30 – 80 IU of parenterally injected calcitonin [93,95,97]. The dose-corrected, relative bioavailability of nasal calcitonin is not dose-dependent, and it is about 25% of the administered dose, as compared with 70% of intramuscular injections [95]. Thus, the potency ratio of nasal relative to intramuscular calcitonin is 1:2.8 – 1:3.5, which approximates the relative potency of the two preparations reported in clinical studies. There has been some controversy about the use of

“promoters ” — chemicals with tensioactive properties — that would enhance the absorption of calcitonin through the nasal mucosa [98]. Although the bioavailability is certainly improved by promoters, unfortunately they may also cause undesirable effects on the mucociliary transport of the mucosal cells [99] and can significantly reduce patient tolerability. The evident biologic effects obtained with preparations without promoters [93 – 95,97] indicates that these additives are probably not necessary, provided that adequate amounts of the drug are delivered for each spray. As an alternative to the parenteral injections of calcitonin, rectal suppositories have also been developed. However, suppositories have not met with great success and few studies exist using this preparation. Rectal administration of salmon calcitonin results in rapid increases of circulating calcitonin, with peaks that are actually higher than those obtained with equivalent doses of nasal preparations, although clearance is faster [94]. Reversal of the decline in bone density in elderly women with osteoporosis was obtained with 100 IU/day of salmon calcitonin suppositories for 2 years [100]. As with the nasal sprays, this effect was evident at the spine, but it was not so clear at the forearm, although a significant inhibition of bone turnover was

FIGURE 9 Average changes from baseline in total serum calcium (sCa) and -endorphin, plasma cAMP, and salmon calcitonin (sCT) in healthy women after a single administration of salmon calcitonin as 50 IU intramuscular (– – – –), 50 IU intranasal (–––), 100 IU intranasal (….), and 200 IU intranasal (. . . .). Reprinted with permission from Overgaard et al. [95].

664

ROBERTO CIVITELLI

obtained during the study [100]. Unfortunately, there was a dropout rate of 23%, which is almost the same as for the parenteral preparations. Although positive results have been also reported in another study on ovariectomized women [101], calcitonin suppositories do not appear to offer any advantage compared to parenteral preparations, and the nasal spray formulations offer better tolerability with limited side effects.

C. Therapeutic Regimens The earlier studies on age-related or established osteoporosis were conducted using parenteral preparations of calcitonin of different species, with doses ranging from 20 to 100 IU daily (see Section II). Although initially the dose of 100 IU per day was the most widely used, the difficult compliance to a daily parenteral injection in long-term therapies, and the evidence that even lower doses of calcitonin could be effective has led to a progressive decline of the daily dose of injectable calcitonin in both research and in clinical practice. Although calcitonin is currently used primarily as nasal spray, injectable formulations still exist and many physicians still prefer the parenteral administration route. Despite the large body of data in support of the bioavailability of nasal calcitonin formulations, skepticism still persists in the medical community about the clinical reliability of nasally administered drugs. If one should opt for a parenteral administration route, the accumulated evidence indicates that 100 IU of injectable calcitonin on alternate days is appropriate for the treatment of osteoporosis, but a lower dose (50 IU on alternate days) can be used in selected subjects with high remodeling rates. Higher doses can be employed in unusual cases of very high bone turnover disease or for secondary forms of osteoporosis, depending on the clinical response and tolerability. Calcitonin nasal spray preparations have been used in noncontrolled clinical trials at different doses and regimens ranging from 400 IU daily to 50 IU 5 days a week, with 200 IU daily being the most favorite regimen. The convergence on the 200 IU daily dose is probably the consequence of an earlier, dose-finding study demonstrating a dose-related increases in spinal bone density in women with established osteoporosis, but significant changes only for the 200 IU daily dose, the highest used in that study [14]. Subsequent studies confirmed that prevention of postmenopausal bone loss could be obtained with daily administration of 200 and 400 IU of salmon calcitonin nasal spray, but not with 100 IU [15,37] or less frequent weekly dosing of 200 IU [17,102]. Based on this information, the large, prospective PROOF study was designed as a three-arm treatment trial which included treatment groups of 100, 200, and 400 IU daily calcitonin nasal spray, as detailed in section II. As noted above, results from this milestone study,

however, do not seem to corroborate the notion of a dose-related effect of nasal salmon calcitonin on bone density [20]. Before the release of the PROOF study, many investigators explored the possibility of using intermittent regimens of calcitonin nasal spray, in the attempt to overcome the problem of loss of response observed with continuous longterm treatments [9,14], perhaps by allowing reactivation of remodeling cycles during the period off antiresorptive intervention, or by restoring a biologic response via other mechanisms. This line of reasoning is not dissimilar from the concept of “coherence” therapy, which is based on the assumption that recruitment and synchronization of remodeling units should allow an antiresorptive agent to act on a larger remodeling space, when given intermittently for a relatively short period of time [103]. This paradigm has been also used to devise cyclical regimens with etidronate [104,105]. Initial proof of this concept was obtained by Overgaard et al. [32] who reported that a discontinuous regimen with nasal spray calcitonin (200 IU/day) over a 3year interval can produce a net gain bone in both the vertebrae and axial skeleton. The percentage increase of bone mass in the second year of active treatment — started after 1 year off calcitonin — was at least as large as the increase obtained during the first year of therapy [32]. Various cyclical, intermittent regimens with nasal calcitonin have been used with good results on bone density, but the most popular are repeated cycles of 100 or 200 IU per day, 3 months on and 3 months off, or 1 month on and 1 month off [18,19,56,106]. Overgaard and Christiansen [107] reviewed their data of women who had been treated for up to 3 years with nasal salmon calcitonin (200 IU/day) on open-label after having participated to other studies as placebo group for 2 years. Stratifying the responses in terms of bone density changes according to the time on active medication relative to the total time of observation, they calculated that protection from bone loss at the spine and the forearm is achieved with a ratio of active treatment to off treatment periods of 1:2 or 2:3. Higher ratios did not give any further advantage, whereas treating for one-third of the total time was only partially effective (Fig. 10). Obviously, these women were not treated with a real cyclical regimen, and these results have to be interpreted with circumspection. Whether cyclic regimens will bear any clinical advantages over the currently indicated daily administration will probably remain untested, since it is unlikely that rigorous studies in this direction will be pursued in the current environment.

D. Clinical Resistance In patients with Paget’s bone disease, serum alkaline phosphatase and urinary hydroxyproline excretion rapidly decrease after initiation of treatment with calcitonin, but in some patients these parameters tend to return to baseline

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

Percentage changes (mean  SEM) of bone mineral content (BMC) of the spine and radius in groups of women followed for up to 5 years, and stratified by duration of calcitonin treatment. Periods of active treatment are indicated on the bottom as percentage of time on calcitonin over total time of observation (%), and as the number of years on active treatment over the total years of follow-up (years/years). Reprinted with permission from Overgaard and Christiansen [107].

FIGURE 10

within 3 – 9 months [108,109]. The phenomenon of acquired resistance to the therapeutic action of calcitonin, also called the “escape phenomenon,” has also been noted in osteoporotic patients, during a long-term treatment with calcitonin. This phenomenon is probably not related to the apparent waning action on bone density commonly observed after 12 – 18 months of calcitonin therapy [9,11,52,107]. Besides the fact that the time of onset is different in the two conditions (3 – 12 months for resistance to develop in Pagetic patients, whereas bone density may continue to increase up to 12 – 18 months in osteoporosis), the major limitation to a continuous positive action of an antiresorptive agent is constituted by the size of the remodeling space that can be refilled, which is proportional to bone turnover. Therefore, the gradual loss of effect on bone density does not represent a peculiarity of calcitonin’s mechanism of action, but rather a characteristic of its antiresorptive action. Nonetheless, some common mechanisms may be involved in the loss of clinical responsiveness in Paget’s bone disease and in osteoporosis.

665 Initially, it was felt that the hypocalcemic effect of calcitonin could produce a secondary hyperparathyroidism that in turn could counteract the action of calcitonin [5,8]. If this were the case, the increase of parathyroid hormone should be immediate and prevent any action of calcitonin. However, using calcitonin in combination with high doses of vitamin D metabolites, which prevented the secondary hyperparathyroidism [8], also negated the effect of calcitonin [8,50]. In addition, the hypocalcemic effect of the currently used doses of calcitonin is rather small, and would produce only transient peaks of parathyroid hormone. Although significant increases of circulating parathyroid hormone have been reported during long-term calcitonin treatment in some studies [110,111], other have failed to detect such an increase [11,33,52,55], which may not occur when appropriate calcium and vitamin D intakes are achieved. As detailed in other sections of this book (Chapter 77), intermittent administration of parathyroid hormone fragments is now being tested as an anabolic therapy for osteoporosis, and it appears unlikely that a secondary hyperparathyroidism if it does occur, may cause the long-term clinical resistance to calcitonin. Development of resistance to therapy in patients with Paget’s bone disease is usually correlated with appearance of antibodies against the exogenous calcitonin [108,109]. The proportion of osteoporotic patients developing antibodies approximates what could be predicted based on the experience in Pagetic patients [112,113], and in more than a half of positive patients these antibodies are able to neutralize the stimulatory activity of salmon calcitonin on cAMP production [114,115]. However, the presence of anti-calcitonin antibodies has never been clearly linked to the loss of response after long-term treatment of osteoporosis, despite the fact that formation of antibodies against salmon calcitonin has been repeatedly demonstrated in 60 – 75% of subjects treated for 15 months or longer, using either injectable or nasal spray preparations [9,114,116]. This notion has been corroborated by the results of the PROOF study. Calcitonin binding antibodies were found in up to 34% of treated subjects, but there was no correlation between their presence and skeletal response to calcitonin [20]. Notably, circulating antibodies that can neutralize the biologic activity of calcitonin have been also found in subjects treated with human calcitonin, although the finding is occasional [115,117,118]. Therefore, the clinical relevance of neutralizing antibodies is probably minimal, and the current evidence does not induce to advocate an assessment of antibody titers in patients treated with salmon calcitonin. Human calcitonin preparations, where available, can be used in those cases that become refractory to salmon calcitonin [119,120]. Whether intermittent regimens may reduce or accentuate the problem remains to be determined on an experimental basis. Another hypothesis put forth to explain the gradual loss of the skeletal response to calcitonin is

666

ROBERTO CIVITELLI

receptor downregulation. Downregulation of kidney calcitonin receptors has in fact been demonstrated in the rat by autoradiographic studies after continuous infusion of salmon calcitonin at different doses [121]. If this phenomenon also occurs in bone cells, then one should expect that intermittent regimens may overcome the clinical resistance.

E. Adverse Events Symptoms of various nature are reported in all studies employing parenteral injections, and can be estimated to occur in approximately 40 – 60% of all patients, although in most cases discontinuation of the treatment is not necessary, and dropout rates because of side effects vary between 5 and 15%. Less frequent side effects are reported for nasal preparations. A review of side effects reported by patients followed at the Hammersmith Hospital in London for at least 36 months revealed that the most frequent complaints from injectable preparations are flushing and irritation at the injection site [122]. Interestingly, the prevalence of these disturbances was higher in patients taking human calcitonin (69 and 65%) than in those treated with salmon calcitonin (41 and 32%, respectively). Nausea was reported by 22 and 14% of patients taking either human or salmon calcitonin preparations. Urinary symptoms (polyuria, urinary urgency) occurred in 10 – 15% of the cases, whereas headache and vomiting were reported by less than 10% of the patients [122]. In contrast, the overall prevalence of side effects in subjects under treatment with nasal salmon calcitonin was 32%, compared to 64 and 77% prevalence observed for parenteral salmon and human calcitonin, respectively. Flushing was still the most frequent complaint even for the nasal route (20%), followed by nasal congestion and irritation (16%) and rhinitis (8%). Sporadic episodes of epistaxis [15,33] and partial loss of sense of smell [15] have been also reported by other investigators. On the other hand, no significant side effects clearly related to the drug were reported in other studies [32,56]. Interestingly, many of the side effects experienced with the injectable forms do not occur at all with the nasal preparation, suggesting different absorption efficiencies and tissue distribution patterns for the two formulations. Although the pathogenesis of these side effects is not clear, most vasomotor symptoms, including flushing and headache, may be related to an interaction of calcitonin with receptors for calcitonin gene-related peptide [123], a hormone with potent vasoactive properties. These symptoms characteristically occur at the initiation of therapy and are either severe enough to discontinue the therapy or decrease in severity to disappear completely during treatment. In the English series, only one patient that continued treatment beyond the first year dropped out because of persistent side effects [122]. Most individuals

experiencing side effects will discontinue the calcitonin relatively early, and this occurs in 10 – 13% of the patients on parenteral calcitonin preparations [122]. Discontinuation is hardly necessary during nasal spray administration [13,39], and it is consistently below 5% [34]. Dermatologic symptoms may respond to antihistaminic therapy given 20 to 30 min before the subcutaneous injection of calcitonin, but in most cases this is not necessary. Gastrointestinal side effects such as bloating or nausea can be minimized if the drug is administered 4 to 5 h after a meal, preferably at bedtime.

IV. ANALGESIC ACTION A. Analgesia in Metabolic Bone Diseases An improvement of the painful symptomatology usually associated with osteoporosis and Paget’s disease of bone has been repeatedly observed [6,93,110,111,124 – 126]. However, the clinical importance of what is known as the analgesic action of calcitonin has remained controversial, and its biologic basis ill-defined. Rigorous studies in this area are difficult to carry out, and the results may present with interpretative difficulties because of the uncertainties in defining the nature and the severity of the painful condition. Most of the studies aimed at defining the analgesic effect of calcitonin have been performed in patients with acute vertebral fractures. Nonetheless, even if an acute fracture episode can be clearly identified and its evolution followed in time, quantitation of pain and the inherent variability in pain threshold among patients remain major hurdles. Most investigators have used subjective methods, which require the patient to score his/her pain on visualanalog scales [127]. Using these methods, some groups have reported measurable improvement of pain score with calcitonin in subjects with acute vertebral fractures, when the treatment was started within 2 weeks since the acute episode [128 – 131]. This effect was also associated with a decrease in the number of rescue analgesics required by the patients to control pain [128,129] (Fig. 11). Although in all these studies a placebo effect, or a spontaneous partial resolution of the condition occurred, the decrease of the painful symptomatology was faster and more pronounced in those subjects taking calcitonin [128 – 130]. Analgesic effects have been obtained with doses of the drug commonly used for treatment of osteoporosis and with either parenteral or nasal spray preparations. The notion that nasal administration is more effective than parenteral injections in inducing pain relief remains to be confirmed [128,132]. It is also interesting to note that the analgesic action of calcitonin seems to be dissociated from the changes in biochemical indices of bone turnover [129,133]. This implies that the analgesic effect of calcitonin is independent of

667

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

painful conditions of different nature. In some reports, wherein a placebo-controlled design was used, calcitonin appeared to induce pain relief in some neurologic conditions. In patients with migraine, parenteral salmon calcitonin was significantly more efficient than placebo in reducing the frequency of pain episodes [134,135]. Likewise, calcitonin was able to reduce the pain score in patients with the phantom limb pain syndrome in the early postoperative period [136,137] and in mild cases of lumbar spinal stenosis [133]. In this condition, calcitonin has also been reported to improve physical performance and lengthen the walking distance [133,138]. However, in more severe cases of neurogenic compression, the effect of calcitonin was poor [133]. The reported analgesic effect in all this conditions that are not associated with metabolic abnormalities of bone remodeling further indicates that this action of calcitonin is not mediated by its effect on bone turnover. Reduction of pain has also been observed in metastatic tumors with bone localizations [139 – 142]. The mechanism of calcitonin-induced analgesia in these cases may also involve a decrease of bone resorption activity in the metastases, with reduction of bone erosion.

C. Possible Mechanism of Action

FIGURE 11 Pain ratings according to a visual analog scale (top) and consumption of paracetamol tablets (bottom) in women after an acute vertebral fracture, treated with either intramuscular salmon calcitonin (—— ), or placebo (.........). Asterisks represent the difference (P  0.05) between treated and controls at each time point by t test. Reprinted with permissin from Lyritis et al. [129]. its inhibitory action on bone resorption. Considering that the acute pain of vertebral fractures is consequent to a sudden change in the biomechanics of the spine rather than to the cellular activity in bone at the time of the fracture, such a finding should not come as unexpected. This contention is further underlined by the consideration that the chronic pain associated with osteoporosis is seldom directly linked to the metabolic state of the bone tissue, as discussed above. Thus, the analgesic effect of calcitonin must be related to an extraskeletal action.

B. Analgesia in Other Painful Conditions Several short-term, mostly uncontrolled studies have reported analgesic properties of calcitonin in extraskeletal

Many hypotheses have been put forth to explain the analgesic action of calcitonin, but none of them has been proven. The demonstration of calcitonin binding sites, distinct from calcitonin gene-related peptide binding sites, in areas of the brain involved in the regulation of pain perception has raised the possibility that calcitonin may directly modulate nociception in the central nervous system [143,144]. In support of this hypothesis are the findings of immunoreactive calcitonin in the central nervous tissue [145,146]. However, the possibility of obtaining analgesic effects by direct epidural or subaracnoidal injection of calcitonin in humans remains controversial [147 – 149]. Besides the presence of calcitonin binding sites in brain tissue [144,150,151], and the evidence of some physiologic effects of the hormone, such as inhibition of prolactin release [152 – 154], if calcitonin directly acts on the central nervous system it must cross the blood – brain barrier when given parenterally. Even assuming that small amounts of the peptide could pass into the brain, pain relief is reported to occur after a few injections of calcitonin [129,130,136, 137], and it is unlikely that the peptide can substantially accumulate in the brain. The other leading hypothesis links the analgesic effect of calcitonin to the endogenous opiate system, and in particular, -endorphin. -Endorphin is produced in the intermediate lobe of the pituitary by posttranslational processing of a precursor, pro-opiomelanocortin, common to ACTH and MSH [155]. Some studies have reported an increase in peripheral -endorphin levels following intravenous

668 injection of salmon calcitonin [156,157]. Cosecretion of ACTH and cortisol was also observed in these studies [156,157], suggesting a potential modulatory action of calcitonin on the secretory activity of pituitary cells [152,153]. The possibility that calcitonin may serve as neurotransmitter is intriguing, but the problem of how the hormone can cross the blood – brain barrier remains. In addition, a single nasal spray or intramuscular injection of calcitonin increased circulating -endorphin in one study [95] (Fig. 9), but not in others [93,158], despite a significant decrease of prolactin levels [93]. Furthermore, baseline levels of -endorphin were not affected by long-term therapy with calcitonin [135], and centrally administered calcitonin did not affect endogenous opiates in animal studies [159,160]. Considering that the biologic role of peripheral -endorphin is also unclear, the evidence that the analgesic effect of calcitonin is mediated by an interference with the endogenous opiate system remains tenuous. A peripheral analgesic effect has also been proposed, on the basis of an inhibition of thromboxan production by calcitonin [161], and the report of an enhancement of pain threshold by locally injected calcitonin in animals [162]. Modulation of prostanoid metabolism may be thought to play a role in cases of localized pain, or local bone destruction, such as in tumor metastases. This hypothesis also remains conjectural.

V. FUTURE DIRECTIONS At the dawn of the new millennium, the horizon of treatment options for patients with osteoporosis and other demineralizing conditions is undoubtedly brighter than just 5 years ago. In this ever-changing arena, calcitonin, which was introduced in the mid-1980s as the first valid alternative to hormonal replacement therapy for women with osteoporosis, now faces increased competition from very effective drugs, such as bisphosphonates and selective estrogen receptor modulators. Therefore, the therapeutic niche for this safe though less powerful drug has been recast over the past 4 – 5 years. Yet, calcitonin remains a viable option in select patients, primarily because of its safety and relatively ease of use in its nasal spray formulations, as demonstrated by the still substantial market share that calcitonin enjoys in the osteoporosis field. New formulations and administration routes are currently being explored, in particular an inhaler preparation that it is hoped may improve on the limitations of the nasal delivery systems, especially in subjects with acute or chronic sinus problems. On the other hand, the unexciting results obtained with combination regimens with GH or parathyroid hormone have somewhat dampened the enthusiasm for the use of calcitonin with a bone formation stimulator. Combination with estrogen or a bisphosphonate, both inhibitors of bone resorption, does not appear as a viable option at this juncture.

ROBERTO CIVITELLI

Given the modest effects on bone density obtained with calcitonin nasal spray in the most recent large clinical trials, it seems unlikely that additive or synergistic effects with estrogen or bisphosphonates will emerge. Finally, analogs of calcitonin with improved potency and pharmacokynetic profile are being developed. However, the limitations imposed by the peptidic nature of the active compound for the generation of improved drug preparations will continue to burden developments in this direction in the near future.

Acknowledgments The author dedicates this chapter to Dr. Louis V. Avioli, mentor and leader of our Division of Bone and Mineral Diseases of Washington University School of Medicine. Without his efforts, will, and support not only could calcitonin not have been successfully developed as a drug, but the entire field of bone and mineral research would not have come as far as it has. Warm gratitude also goes to Dr. Carlo Gennari for his critical and insightful review of the two editions of this chapter.

References 1.

2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

D. H. Copp, E. C. Cameron, B. A. Cheney, A. F. G. Davidson, and K. G. Henze, Evidence for calcitonin: A new hormone from the parahtyroid that lowers blood calcium. Endocrinology 701, 638 – 649 (1962). A. M. Parfitt, The coupling of bone formation to bone resorption: A critical analysis of the concept and of its relevance to the pathogenesis of osteoporosis. Metab. Bone Dis. Relat. Res. 4, 1 – 6 (1982). R. Marcus, Normal and abnormal bone remodeling in man. Annu. Rev. Med. 38, 129 – 141 (1987). A. Caniggia, C. Gennari, M. Bencini, L. Cesari, and G. Borrello, Calcium metabolism and 47calcium kinetics before and after longterm thyrocalcitonin treatment in senile osteoporosis. Clin. Sci. 38, 397 – 407 (1970). S. H. Cohn, C. S. Dombrowski, W. Hauser, J. Klopper, and H. L. Atkins, Effects of porcine calcitonin on calcium metabolism in osteoporosis. J. Clin. Endocr. 33, 719 – 728 (1971). S. Wallach, S. H. Cohn, H. L. Atkins, K. J. Ellis, R. Kohberger, J. F. Aloia, and I. Zanzi, Effect of salmon calcitonin on skeletal mass in osteoporosis. Curr. Ther. Res. 22, 556 – 572 (1977). H. Rasmussen, P. Bordier, P. J. Marie, L. Auquier, J. B. Eisinger, D. Kuntz, F. Caulin, B. Argemi, J. Gueris, and A. Julien, Effect of combined therapy with phosphate and calcitonin on bone volume in osteoporosis. Metab. Bone Dis. Relat. Res. 2, 107 – 111 (1980). J. Jowsey, B. L. Riggs, P. J. Kelly, and D. L. Hoffman, Calcium and salmon calcitonin in treatment of osteoporosis. J. Clin. Endocrinol. Metab. 47, 633 – 639 (1978). H. E. Gruber, J. L. Ivey, D. J. Baylink, M. Matthews, W. B. Nelp, K. Sisom, and C. H. Chesnut, III, Long-term calcitonin therapy in postemenopausal osteoporosis. Metabolism. 33, 295 – 303 (1984). C. Gennari, S. M. Chierichietti, S. Bigazzi, L. Fusi, S. Gonnelli, R. Ferrara, and F. Zacchei, Comparative effects on bone mineral content of calcium and calcium plus salmon calcitonin given in two different regimens in postmenopausal osteoporosis. Curr. Ther. Res. 38, 455 – 464 (1985). G. Mazzuoli, M. Passeri, C. Gennari, S. Minisola, R. Antonelli, C. Valtorta, E. Palummeri, G. Cervellin, S. Gonnelli, and G. Francini,

669

CHAPTER 73 Calcitonin for Treatment of Osteoporosis

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Effects of salmon calcitonin in postmenopausal osteoporosis: A controlled double-blind clinical study. Calcif. Tissue Int. 38, 3 – 8 (1986). R. Crespo, M. Revilla, E. Crespo, L. F. Villa, and H. Rico, Complementary medical treatment for Colles’ fracture: A comparative, randomized, longitudinal study. Calcif. Tissue Int. 60, 567 – 570 (1997). K. Overgaard, B. J. Riis, C. Christiansen, J. Podenphant, and J. S. Johansen, Nasal calcitonin for treatment of established osteoporosis. Clin. Endocrinol. 30, 435 – 442 (1989). K. Overgaard, M. A. Hansen, S. B. Jensen, and C. Christiansen, Effect of salcatonin given intranasally on bone mass and fracture rates in established osteoporosis: a dose-response study. Br. Med. J. 305, 556 – 561 (1992). G. Thamsborg, T. L. Storm, R. Sykulski, E. Brinch, H. K. Nielsen, and O. H. Sorensen, Effect of different doses of nasal salmon calcitonin on bone mass. Calcif. Tissue Int. 48, 302 – 307 (1991). G. Thamsborg, J. E. Jensen, G. Kollerup, E. M. Hauge, F. Melsen, and O. H. Sorensen, Effect of nasal salmon calcitonin on bone remodeling and bone mass in postmenopausal osteoporosis. Bone. 18, 207 – 212 (1996). M. C. Ellerington, T. C. Hillard, S. I. Whitcroft, M. S. Marsh, B. Lees, L. M. Banks, M. I. Whitehead, and J. C. Stevenson, Intranasal salmon calcitonin for the prevention and treatment of postmenopausal osteoporosis. Calcif. Tissue Int. 59, 6 – 11 (1996). S. Gonnelli, C. Cepollaro, C. Pondrelli, S. Martini, S. Rossi, and C. Gennari, Ultrasound parameters in osteoporotic patients treated with salmon calcitonin: A longitudinal study. Osteoporosis Int. 6, 303 – 307 (1996). P. Peichl, B. Rintelen, W. Kumpan, and H. Broll, Increase of axial and appendicular trabecular and cortical bone density in established osteoporosis with intermittent nasal salmon calcitonin therapy. Gynecol. Endocrinol. 13, 7 – 14 (1999). C. H. Chesnut, III, S. L. Silverman, K. Andriano, H. K. Genant, A. Gimona, S. Harris, D. P. Kiel, M. S. LeBoff, M. Maricic, P. Miller, C. Moniz, M. Peacock, P. Richardson, N. B. Watts, and D. J. Baylink, Prospective, randomized trial of nasal spray salmon calcitonin in postmenopausal women with established osteoporosis: The PROOF study. Am. J. Med. 109, 267 – 276 (2000). U. A. Liberman, S. R. Weiss, J. Broll, H. W. Minne, H. Quan, N. H. Bell, J. Rodriguez-Portales, R. W. Downs, Dequeker Jr., J., and M. J. Favus, Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. The Alendronate Phase III Osteoporosis Treatment Study Group. N. Engl. J. Med. 333, 1437 – 1443 (1995). Anonymous, Effects of hormone therapy on bone mineral density: results from the postmenopausal estrogen/progestin interventions (PEPI) trial. The Writing Group for the PEPI. JAMA. 276, 1389 – 1396 (1996). B. Ettinger, D. M. Black, B. H. Mitlak, R. K. Knickerbocker, T. Nickelsen, H. K. Genant, C. Christiansen, P. D. Delmas, J. R. Zanchetta, J. Stakkestad, C. C. Gluer, K. Krueger, F. J. Cohen, S. Eckert, K. E. Ensrud, L. V. Avioli, P. Lips, and S. R. Cummings, Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. Multiple Outcomes of Raloxifene Evaluation (MORE) Investigators. JAMA. 282, 637 – 645 (1999). S. Adami, M. C. Baroni, M. Broggini, L. Carratelli, I. Caruso, L. Gnessi, M. Laurenzi, A. Lombardi, G. Norbiato, and S. Ortolani, Treatment of postmenopausal osteoporosis with continuous daily oral alendronate in comparison with either placebo or intranasal salmon calcitonin. Osteoporosis Int. 3, S21 – S27, (1993). C. J. Rosen, S. L. Bonnick, P. D. Miller, M. McClung, R. Wasnich, S. R. Weiss, G. C. Woodson, T. J. Schnitzer, P. Lenihan, and R. K. Ross, Treatment of osteoporosis in postmenopausal women: alendronate vs. intranasal calcitonin. J. Bone Miner. Res. 14, S400 (1999).

26.

27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42. 43.

R. W. Downs, M. J. Favus, B. W. Walsh, N. H. Bell, M. Ettinger, L. Wang, B. Mako, M. E. Smith, G. J. Gormley, and L. J. Melton, III, Alendronate vs. intranasal calcitonin for treatment of osteoporosis in postmenopausal women. J. Bone Miner. Res. 14, S401 (1999). R. Armamento-Villareal and R. Civitelli, Estrogen action on bone mass of postmenopausal women is dependent on body mass and initial bone density. J. Clin. Endocrinol. Metab. 80, 776 – 782 (1995). S. R. Cummings, D. M. Black, D. E. Thompson, W. B. Applegate, E. Barrett-Connor, T. A. Musliner, L. Palermo, R. Prineas, S. M. Rubin, J. C. Scott, T. Vogt, R. Wallace, A. J. Yates, and A. Z. LaCroix, Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: Results from the Fracture Intervention Trial. JAMA 280, 2077 – 2082 (1998). C. W. Slemenda, S. L. Hui, C. Longcope, and C. C. Johnston, Jr., Sex steroids and bone mass. A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). C. Christiansen, B. J. Riis, and P. Rødbro, Prediction of rapid bone loss in postmenopausal women. Lancet 1, 1105 – 1108 (1987). R. Civitelli, S. Gonnelli, F. Zacchei, S. Bigazzi, A. Vattimo, L. V. Avioli, and C. Gennari, Bone turnover in postmenopausal osteoporosis: Effect of calcitonin treatment. J. Clin. Invest. 82, 1268 – 1274 (1988). K. Overgaard, M. A. Hansen, V. A. Nielsen, B. J. Riis, and C. Christiansen, Discontinuous calcitonin treatment of established osteoporosis — effects of withdrawal of treatment. Am. J. Med. 89, 1 – 6 (1990). J. Y. Reginster, D. Denis, A. Albert, R. Deroisy, M. P. Lecart, M. A. Fontaine, P. Lambelin, and P. Franchimont, 1-Year controlled randomised trial of prevention of early postmenopausal bone loss by intranasal calcitonin. Lancet 2, 1481 – 1483 (1987). J. Y. Reginster, D. Denis, R. Deroisy, M. P. Lecart, M. De Longueville, B. Zegels, N. Sarlet, P. Noirfalisse, and P. Franchimont, Long-term (3 years) prevention of trabecular postemenopausal bone loss with low-dose intermittent nasal salmon calcitonin. J. Bone Miner. Res. 9, 69 – 73 (1994). I. MacIntyre, J. C. Stevenson, M. I. Whitehead, S. J. Wimalawansa, L. M. Banks, and M. J. R. Healy, Calcitonin for prevention of postmenopausal bone loss. Lancet 1, 900 – 902 (1988). M. Meschia, M. Brincat, P. Barbacini, P. G. Crossignani, and W. Albisetti, A clinical trial on the effects of a combination of elcatonin (carbocalcitonin) and conjugated estrogens on vertebral bone mass in early postmenopausal women. Calcif. Tissue Int. 53, 17 – 20 (1993). K. Overgaard, Effect of intranasal salmon calcitonin therapy on bone mass and bone turnover in early postmenopausal women: A dose-response study. Calcif. Tissue Int. 55, 82 – 86 (1994). K. Overgaard, B. J. Riis, C. Christiansen, M. A. Hansen, and D. Aarksog, Effect of salcatonin given intranasally on early postmenopausal bone loss. Br. Med. J. 299, 477 – 479 (1989). C. Gennari, D. Agnusdei, M. Montagnani, S. Gonnelli, and R. Civitelli, An effective regimen of intranasal salmon calcitonin in early postmenopausal bone loss. Calcif. Tissue Int. 50, 381 – 383 (1992). H. Rico, E. R. Hernandez, M. Revilla, and F. Gomez-Castresana, Salmon calcitonin reduces vertebral fracture rate in postmenopausal crush fracture syndrome. Bone Miner. 16, 131 – 138 (1992). S. L. Silverman, C. H. Chesnut, III, K. Andriano, H. K. Genant, A. Gimona, M. Maricic, J. L. Stock, and D. J. Baylink, Salmon calcitonin nasal spray (NS-CT) reduces risk of vertebral fracture(s) (VF) in established osteoporosis and has continuous efficacy with prolonged treatment: Accrued 5-year worldwide data of the PROOF study. Bone. 23, S174 (1998). R. Eastell, Treatment of postmenopausal osteoporosis. N. Engl. J. Med. 338, 736 – 746 (1998). P. Burckhardt and B. Burnand, The effect of treatment with calcitonin on vertebral fracture rate in osteoporosis. Osteoporosis Int. 3, 24 – 30 (1993).

670 44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58. 59.

ROBERTO CIVITELLI J. A. Kanis, O. Johnell, B. Gullberg, E. Allander, G. Dilsen, C. Gennari, A. A. Lopes Vaz, G. P. Lyritis, G. Mazzuoli, L. Miravet, M. Passeri, R. Perez Cano, A. Rapado, and C. A. Ribot, Evidence for efficacy of drugs affecting bone metabolism in preventing hip fracture. Br. Med. J. 305, 1124 – 1128 (1992). S. Hulley, D. Grady, T. Bush, C. Furberg, D. Herrington, B. Riggs, and E. Vittinghoff, Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 280, 605 – 613 (1998). P. J. Marie and F. Caulin, Mechanisms underlying the effects of phosphate and calcitonin on bone histology in postmenopausal osteoporosis. Bone 7, 17 – 22 (1988). R. D. Hesch, U. Busch, M. Prokop, G. Delling, and E. F. Rittinghaus, Increase of vertebral density by combination therapy with pulsatile 1-38hPTH and sequential addition of calcitonin nasal spray in osteoporotic patients. Calcif. Tissue Int. 44, 176 – 180 (1989). A. B. Hodsman, L. J. Fraher, T. Ostbye, J. D. Adachi, and B. M. Steer, An evaluation of several biochemical markers for bone formation and resorption in a protocol utilizing cyclical parathyroid hormone and calcitonin therapy for osteoporosis. J. Clin. Invest. 91, 1138 – 1148 (1993). A. B. Hodsman, L. J. Fraher, P. H. Watson, T. Ostbye, L. W. Stitt, J. D. Adachi, D. H. Taves, and D. Drost, A randomized controlled trial to compare the efficacy of cyclical parathyroid hormone versus cyclical parathyroid hormone and sequential calcitonin to improve bone mass in postmenopausal women with osteoporosis. J. Clin. Endocrinol. Metab. 82, 620 – 628 (1997). S. A. V. Eriksson and J. U. Lindgren, Combined treatment with calcitonin and 1,25- dihydroxyvitamin D3 for osteoporosis in women. Calcif. Tissue Int. 53, 26 – 28, (1993). L. Flicker, J. L. Hopper, R. G. Larkins, M. Lichtenstein, G. Buirski, and J. D. Wark, Nandrolone decanoate and intranasal calcitonin as therapy in established osteoporosis. Osteoporos. Int. 7, 29 – 35 (1997). J. F. Aloia, A. N. Vaswani, A. Kapoor, J. K. Yeh, and S. H. Cohn, Treatment of osteoporosis with calcitonin, with and withour growth hormone. Metabolism. 34, 124 – 129, (1985). L. Holloway, L. Kohlmeier, K. Kent, and R. Marcus, Skeletal effects of cyclic recombinant human growth hormone and salmon calcitonin in osteopenic postmenopausal women. J. Clin. Endocrinol. Metab. 82, 1111 – 1117 (1997). S. Gonnelli, C. Cepollaro, M. Montomoli, L. Gennari, A. Montagnani, R. Palmieri, and C. Gennari, Treatment of post-menopausal osteoporosis with recombinant human growth hormone and salmon calcitonin: a placebo controlled study. Clin. Endocrinol. (Oxf). 46, 55 – 61 (1997). G. Mazzuoli, S. Tabolli, F. Bigi, C. Valtorta, S. Minisola, D. Diacinti, L. Scarnecchia, G. Bianchi, M. Piolini, and S. Dell’Acqua, Effects of salmon calcitonin on the bone loss induced by ovariectomy. Calcif. Tissue Int. 47, 209 – 214, (1990). P. Fioretti, M. Gambacciani, F. Taponeco, G. B. Melis, N. Capelli, and A. Spinetti, Effects of continuous and cyclic nasal calcitonin administration in ovariectomized women. Maturitas 15, 225 – 232 (1992). O. Grigoriou, I. Papoulias, N. Vitoratos, C. Papadias, S. Konidaris, G. Antoniou, and A. Chryssikopoulos, Effects of nasal administration of calcitonin in oophorectomized women: 2-Year controlled double-blind study. Maturitas. 28, 147 – 151, (1997). J. D. Ringe, and D. Welzel, Salmon calcitonin in the therapy of corticoid-induced osteoporosis. Eur. J. Clin. Pharmacol. 33, 35 – 39 (1987). M. Luengo, C. Picado, L. Del Rio, N. Guañabens, J. M. Montserrat, and J. Setoain, Treatment of steroid-induced osteopenia with calcitonin in corticosteroid-dependent asthma. Am. Rev. Resp. Dis. 142, 104 – 107 (1990).

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

A. Kotaniemi, H. Piirainen, L. Paimela, M. Leirisalo-Repo, K. Uoti-Reilama, P. Lahdentausta, P. Ruotsalainen, M. Kataja, E. Vaisanen, and P. Kurki, Is continuous intranasal salmon calcitonin effective in treating axial bone loss in patients with active rheumatoid arthritis receiving low dose glucocorticoid therapy? J. Rheumatol. 23, 1875 – 1879 (1996). J. D. Adachi, W. G. Bensen, M. J. Bell, F. A. Bianchi, A. A. Cividino, G. L. Craig, W. C. Sturtridge, R. J. Sebaldt, M. Steele, M. Gordon, E. Themeles, P. Tugwell, R. Roberts, and M. Gent, Salmon calcitonin nasal spray in the prevention of corticosteroid- induced osteoporosis. Br. J. Rheumatol. 36, 255 – 259 (1997). L. Montemurro, G. Schiraldi, P. Fraioli, G. Tosi, A. Riboldi, and G. Rizzato, Prevention of cortocosteroid-induced osteoporosis with salmon calcitonin in sarcoid patients. Calcif. Tissue Int. 49, 71 – 76 (1991). P. N. Sambrook, J. Birmingham, P. J. Kelly, S. Kempler, T. V. Nguyen, N. A. Pocock, and J. A. Eisman, Prevention of corticosteroid osteoporosis. A comparison of calcium, calcitriol, and calcitonin. N. Engl. J. Med. 328, 1747 – 1752 (1993). H. Resch, P. Pietschmann, and R. Willvonseder, Estimated long-term effect of calcitonin treatment in acute osteoporotic spine fractures. Calcif. Tissue Int. 45, 209 – 213 (1989). J. H. Healey, S. A. Paget, P. Williams-Russo, T. P. Szatrowski, R. Schneider, H. Spiera, H. Mitnick, K. Ales, and P. Schwartzberg, A randomized controlled trial of salmon calcitonin to prevent bone loss in corticosteroid-treated temporal arteritis and polymyalgia rheumatica. Calcif. Tissue Int. 58, 73 – 80 (1996). D. A. Hantman, J. M. Vogel, C. L. Donaldson, R. Friedman, R. S. Goldsmith, and S. B. Hulley, Attempts to prevent disuse osteoporosis by treatment with calcitonin, longitudinal compression and supplementary calcium and phosphate. J. Clin. Endocrinol. Metab. 36, 845 – 858 (1973). H. E. Van der Lips, P. Wiel, J. Nauta, G. Kwakkel, G. Hazenberg, J. C. Netelenbos, and W. J. Van der Vijgh, Intranasal calcitonin suppresses increased bone resorption during short-term immobilization: a double-blind study of the effects of intranasal calcitonin on biochemical parameters of bone turnover. J. Bone Miner. Res. 8, 1459 – 1465 (1993). A. Chantraine, B. Nusgens, and Ch. M. Lapiere, Bone remodeling during the development of osteoporosis in paraplegia. Calcif. Tissue Int. 38, 323 – 327 (1986). M. M. Petersen, J. B. Lauritzen, P. Schwarz, and B. Lund, Effect of nasal salmon calcitonin on post-traumatic osteopenia following ankle fracture. A randomized double-blind placebo- controlled study in 24 patients. Acta Orthop. Scand. 69, 347 – 350 (1998). R. F. Laan, P. L. Van Riel, and L. B. Van de Putte, Bone mass in patients with rheumatoid arthritis. [Review]. Ann. Rheum. Dis. 51, 826 – 832 (1992). T. D. Spector, G. M. Hall, E. V. McCloskey, and J. A. Kanis, Risk of vertebral fracture in women with rheumatoid arthritis. Br. Med. J. 306, 558 – 558 (1993). H. Kröger, I. Arnala, and E. M. Alhava, Effect of calcitonin on bone histomorphometry and bone metabolism in rheumatoid arthritis. Calcif. Tissue Int. 50, 11 – 13 (1992). A. Sileghem, P. Geusens, and J. Dequeker, Intranasal calcitonin for the prevention of bone erosion and bone loss in rheumatoid arthritis. Ann. Rheum. Dis. 51, 761 – 764 (1992). H. Rico, E. R. Hernandez, J. Diaz-Mediavilla, A. Alvarez, R. Martinez, and D. Espinos, Treatment of multiple myeloma with nasal spray calcitonin: A histomorphometric and biochemical analysis. Bone Min. 8, 231 – 237 (1990). J. J. Stepán, M. Lachman, J. Zverina, V. Pacovsky, and D. J. Baylink, Castrated men exhibit bone loss: Effect of calcitonin treatment on biochemical indices of bone remodeling. J. Clin. Endocrinol. Metab. 69, 523 – 527 (1989).

671

CHAPTER 73 Calcitonin for Treatment of Osteoporosis 76.

77. 78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90. 91.

92.

93.

94.

95.

S. Lakhanpal, W. W. Ginsburg, H. S. Luthra, and G. G. Hunder, Transient regional osteoporosis. A study of 56 cases and review of the literature. Ann. Int. Med. 106, 444 – 450 (1987). S. J. Naides, D. Resnick, and N. J. Zvaifler, Idiopathic regional osteoporosis: a clinical spectrum. J. Rheumatol. 12, 763 – 768 (1985). A. Sawicki, P. Szulc, T. Sobczyk, J. Goliszewski, P. Garnier, and R. Labuszewski, Influence of calcitonin treatment on the osteocalcin concentration in the algodystrophy of bone. Clin. Rheumatol. 11, 346 – 350 (1992). C. Gobelet, J. L. Meier, W. Schaffner, A. Bischof-Delaloye, J. C. Gerster, and P. Burckhardt, Calcitonin and reflex sympathetic dystrophy syndrome. Clin. Rheumatol. 5, 382 – 388 (1986). D. R. Bickerstaff and J. A. Kanis, The use of nasal calcitonin in the treatment of post-traumatic algodystrophy. Br. J. Rheumatol. 30, 291 – 294 (1991). C. Gobelet, M. Waldburger, and J. L. Meier, The effect of adding calcitonin to physical treatment on reflex sympathetic dystrophy. Pain 48, 171 – 175 (1992). P. Paavolainen, T. Taivainen, J. E. Michelsson, M. Lalla, and R. Penttinen, Calcitonin and fracture healing. An experimental study on rats. J. Orthop. Res. 7, 100 – 106 (1989). S. Castells, C. Lu, R. K. Baker, and S. Wallach, Effects of synthetic salmon calcitonin in osteogenesis imperfecta. Curr. Ther. Res. 16, 1 – 14 (1974). S. Castells, C. Colbert, C. Chakabarti, R. S. Bachell, E. G. Kassner, and S. Yasumura, Therapy of osteogenesis imperfecta with synthetic salmon calcitonin. J. Pediat. 95, 807 – 811 (1979). Y. Nishi, K. Hamamoto, M. Kajiyama, H. Ono, M. Kihara, and K. Jinno, Effect of long-term calcitonin therapy by injection and nasal spray on the incidence of fractures in osteogenesis imperfecta. J. Pediatr. 121, 477 – 480 (1992). M. A. Kumar, E. Slack, A. Edwards, H. A. Soliman, A. Baghdiant, G. V. Foster, and I. MacIntyre, A biologic assay for calcitonin. J. Endocrinol. 33, 469 – 475 (1965). C. Gennari, S. Gonnelli, D. Agnusdei, and R. Civitelli, Biological activity of different calcitonins in men. Exp. Gerontol. 25, 339 – 347 (1990). B. L. Riggs, C. D. Arnaud, R. S. Goldsmith, W. F. Taylor, J. T. McCall, and A. D. Sessler, Plasma kinetics and acute effects of pharmacologic doses of porcine calcitonin in man. J. Clin. Endocr. 33, 115 – 127 (1971). R. Huwyler, W. Born, E. E. Ohanaus, and J. A. Fischer, Plasma kinetics and urinary excretion of exogenous human and salmon calcitonin in man. Am. J. Physiol. 236, E15 – E19, (1979). S. J. Marx, C. J. Woodward, and G. D. Aurbach, Calcitonin receptors in kidney and bone. Science 178, 998 – 1001 (1972). C. Gennari, S. M.Chierichietti, C. Vibelli, G. Francini, E. Maioli, and S. Gonnelli. Acute effects of salmon, human and porcine calcitonin on plasma calcium and cyclic AMP levels in man. Curr. Ther. Res. 30, 1024 – 1032 (1981). M. Otani, S. Kitazawa, H. Yamauchi, T. Meguro, and H. Orimo, Stability of the biologic activity of eel calcitonin in rats. Horm. Metab. Res. 10, 252 – 256 (1978). J. Y. Reginster, D. Denis, A. Albert, and P. Franchimont, Assessment of the biologic effectiveness of nasal synthetic salmon calcitonin (SSCT) by comparison with intramuscular (i.m.) or placebo injection in normal subjects. Bone Miner. 2, 133 – 140 (1987). T. Buclin, J. P. Randin, A. F. Jacquet, M. Azria, M. Attinger, F. Gomez, and P. Burckhardt, The effect of rectal and nasal administration of salmon calcitonin in normal subjects. Calcif. Tissue Int. 41, 252 – 258 (1987). K. Overgaard, D. Agnusdei, M. A. Hansen, E. Maioli, C. Christiansen, and C. Gennari, Dose – response bioactivity and bioavailability of salmon calcitonin in premenopausal and postmenopausal women. J. Clin. Endocrinol. Metab. 72, 344 – 349 (1991).

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

A. E. Pontiroli, M. Alberetto, and G. Pozza, Intranasal calcitonin and plasma calcium concentrations in normal subjects. Br. Med. J. 290, 1390 – 1391 (1985). D. P. O’Doherty, D. R. Bickerstaff, E. V. McCloskey, R. Atkins, N. A. Hamdy, and J. A. Kanis, A comparison of the acute effects of subcutaneous and intranasal calcitonin. Clin. Sci. 78, 215 – 219 (1990). A. E. Pontiroli, M. Alberetto, A. Calderara, E. Pajetta, and G. Pozza, Nasal administration of glucagon and human calcitonin to healthy subjects: A comparison of powders and spray solutions and of different enhancing agents. Eur. J. Clin. Pharmacol. 37, 427 – 430 (1989). P. C. Braga, L. Allegra, C. Rampoldi, G. Beghi, A. Ornaghi, G. Caminiti, Y. R. Zheng, and F. Bartucci, Topical tolerability of salmon calcitonin assessed by mucociliary transport velocity investigation. Arzneimittel Forsch. 40, 938 – 941 (1990). K. Overgaard, M. A. Hansen, K. L. Dirksen, and C. Christiansen, Rectal salmon calcitonin for the treatment of postmenopausal osteoporosis. Calcif. Tissue Int. 51, 184 – 188 (1992). F. Polatti, F. Perotti, G. P. Angelini, D. Vassellati, and I. Rapisardi, Effects of salmon calcitonin suppositories in the prevention of bone loss in oophorectomized women. Maturitas. 18, 73 – 76 (1993). A. Gurlek, M. Bayraktar, and O. Gedik, Comparison of calcitriol treatment with etidronate- calcitriol and calcitonin-calcitriol combinations in Turkish women with postmenopausal osteoporosis: a prospective study. Calcif. Tissue Int. 61, 39 – 43 (1997). H. M. Frost, Treatment of osteoporoses by manipulation of coherent bone cell populations. Clin. Orthrop. Relat. Res. 143, 227 – 244 (1979). T. L. Storm, G. Thamsborg, T. Steiniche, H. K. Genant, and O. H. Sorensen, Effect of intermittent cyclical etidronate therapy on bone mass and fracture rate in women with postmenopausal osteoporosis. N. Engl. J. Med. 322, 1265 – 1271 N. B. Watts, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Miller, R. D. Jackson, A. A. Licata, P. D. Ross, G. C. Woodson, M. J. Yanover, W. J. Mysiw, L. Kohse, M. B. Rao, P. Steiger, B. Richmond, and C. H. Chesnut, III, Intermittent cyclical etidronate treatment of postmenopausal osteoporosis. N. Engl. J. Med. 323, 73 – 79 (1990). G. Perrone, P. Galoppi, M. Valente, O. Capri, C. D’Ubaldo, G. Anelli, and L. Zichella, Intranasal salmon calcitonin in postmenopausal osteoporosis: Effect of different therapeutic regimens on vertebral and peripheral bone density. Gynecol. Obst. Invest. 33, 168 – 171 (1992). K. Overgaard and C. Christiansen, Long-term treatment of established osteoporosis with intranasal calcitonin. Calcif. Tissue Int. 49, S60 – S63 (1991). F. R. Singer, R. S. Fredericks, and C. Minkin, Salmon calcitonin therapy for Paget’s disease of bone. Arthritis. Rheum. 23, 1148 – 1154 (1980). J. G. Haddad and J. G. Caldwell, Jr., Calcitonin resistance: Clinical and immunologic studies in subjects with Paget’s disease of bone treated with porcine and salmon calcitonin. J. Clin. Invest. 51, 3133 – 3141 (1972). J. Jowsey, B. L. Riggs, R. S. Goldsmith, P. J. Kelly, and C. D. Arnaud, Effects of prolonged administration of porcine calcitonin in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 33, 752 – 758 (1971). S. Ljunghall, P. Gardsell, O. Johnell, K. Larsson, E. Lindh, K. Obrant, and I. Sernbo, Synthetic human calcitonin in postmenopausal osteoporosis: A placebo-controlled, double-blind study. Calcif. Tissue Int. 49, 17 – 19 (1991). A. Grauer, F. Raue, H. G. Schneider, K. Frank-Raue, and R. Ziegler, in vitro detection of neutralizing antibodies after treatment of Paget’s disease of bone with nasal salmon calcitonin. J. Bone Miner. Res. 5, 387 – 391 (1990).

672 113. F. Levy, R. Muff, S. Dotti-Sigrist, M. A. Dambacher, and J. A. Fischer, Formation of neutralizing antibodies during intranasal synthetic salmon calcitonin treatment of Paget’s disease. J. Clin. Endocrinol. Metab. 67, 541 – 545 (1988). 114. R. Muff, M. A. Dambacher, and J. A. Fischer, Formation of neutralizing antibodies during intranasal synthetic salmon calcitonin treatment of postmenopausal osteoporosis. Osteoporosis Int. 1, 72 – 75 (1991). 115. A. Grauer, H. H. Reinel, R. Ziegler, and F. Raue, Neutralizing antibodies against calcitonin. Horm. Metab. Res. 25, 486 – 488 (1993). 116. J. Y. Reginster, S. Gaspar, R. Deroisy, B. Zegels, and P. Franchimont, Prevention of osteoporosis with nasal salmon calcitonin: Effect of anti-salmon calcitonin antibody formation. Osteoporosis Int. 3, 261 – 264 (1993). 117. A. Grauer, H. H. Reinel, S. Lunghall, E. Lindh, R. Ziegler, and F. Raue, Formation of neutralizing antibodies after treatment with human calcitonin. Am. J. Med. 95, 439 – 442 (1993). 118. F. M. Dietrich, J. A. Fischer, and O. L. M. Bijvoet, Formation of antibodies to synthetic human calcitonin during treatment of Paget’s disease. Acta Endocrinol. 92, 468 – 476 (1979). 119. S. Rojanasathit, E. M. Rosenberg, and J. G. Haddad, Jr., Paget’s bone disease: Response to human calcitonin in patients resistant to salmon calcitonin. Lancet. 2, 1412 – 1415 (1974). 120. R. Muff, M. A. Dambacher, A. Perrenoud, C. Simon, and J. A. Fischer, Efficacy of intranasal human calcitonin in patients with Paget’s disease refractory to salmon calcitonin. Am. J. Med. 89, 181 – 184 (1990). 121. Z. Bouizar, W. H. Roestene, and G. Milhaud, Down-regulation of rat kidney calcitonin receptors by salmon calcitonin infusion evidenced by autoradiography. Proc. Natl. Acad. Sci. USA 84, 5125 – 5218 (1987). 122. S. J. Wimalawansa, Long- and short-term side effects and safety of calcitonin in man: a prospective study. Calcif. Tissue Int. 52, 90 – 93 (1993). 123. D. Goltzman and J. Mitchell, Interactions of calcitonin and calcitonin gene-related peptide at receptor sites in target tissues. Science 227, 1343 – 1345 (1985). 124. F. Shai, R. K. Baker, and S. Wallach, The clinical and metabolic effects of porcine calcitonin on Paget’s disease of bone. J. Clin. Invest. 50, 1927 – 1940 (1971). 125. R. Franceschini, P. Bottaro, C. Panopoulos, and V. Messina, Longterm treatment with salmon calcitonin in postmenopausal osteoporosis. Curr. Ther. Res. 34, 795 – 800 (1983). 126. W. Grotz, J. A. Rump, A. Niessen, H. Schmidt-Gayk, and P. Schollmeyer, Treatment of bone pain after kidney transplantation. Transplant. Proc. 30, 2114 – 2116 (1998). 127. E. C. Huskisson, Masurement of pain. Lancet 2, 1127 – 1131 (1974). 128. C. Gennari, D. Agnusdei, and A. Camporeale, Use of calcitonin in the treatment of bone pain associated with osteoporosis. Calcif. Tissue Int. 49, S9 – 13 (1991). 129. G. P. Lyritis, N. Tsakalakos, B. Magiasis, T. Karachalios, A. Yiatzides, and M. Tsekoura, Analgesic effect of salmon calcitonin in osteoporotic vertebral fractures: A double-blind placebo-controlled clinical study. Calcif. Tissue Int. 49, 369 – 372 (1991). 130. K. K. Pun and L. W. Chan, Analgesic effect of intranasal salmon calcitonin in the treatment of osteoporotic vertebral fractures. Clin. Ther. 11, 205 – 209 (1989). 131. G. P. Lyritis, I. Paspati, T. Karachalios, D. Ioakimidis, G. Skarantavos, and P. G. Lyritis, Pain relief from nasal salmon calcitonin in osteoporotic vertebral crush fractures. A double blind, placebocontrolled clinical study. Acta Orthop. Scand. Suppl. 275, 112 – 114, (1997). 132. A. E. Pontiroli, E. Pajetta, A. Calderara, M. Alberetto, G. Pozza, V. Manganelli, G. Resmini, L. Tessari, and V. Maresca, Intranasal and intramuscular human calcitonin in female osteoporosis and in

ROBERTO CIVITELLI

133.

134.

135.

136. 137. 138. 139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150. 151.

152.

Paget’s disease of bones: A pilot study. J. Endocrinol. Invest. 14, 47 – 51 (1991). A. Eskola, T. Pohjolainen, H. Alaranta, J. Soini, K. Tallroth, and P. Slatis, Calcitonin treatment in lumbar spinal stenosis: a randomized, placebo-controlled, double-blind, cross-over study with oneyear follow-up. Calcif. Tissue Int. 50, 400 – 403 (1992). C. Gennari, S. M. Chierichietti, S. Gonnelli, C. Vibelli, M. Montagnani, and M. Piolini, Migraine prophylaxis with salmon calcitonin: A cross-over double-blind, placebo-controlled study. Headache. 26, 13 – 16 (1986). G. Micieli, A. Cavallini, E. Martignoni, V. Covelli, F. Facchinetti, and G. Nappi, Effectiveness of salmon calcitonin nasal spray preparation in migraine treatment. Headache. 28, 196 – 200 (1988). H. Jaeger and C. Maier, Calcitonin in phantom limb pain: A doubleblind study. Pain. 48, 21 – 27 (1992). C. Kessel and R. Worz, Immediate response of phantom limb pain to calcitonin. Pain 30, 79 – 87 (1987). R. W. Porter and C. Hibbert, Calcitonin treatment for neurogenic claudication. Spine. 8, 585 – 592 (1983). C. Gennari, S. M. Chierichietti, M. Piolini, C. Vibelli, D. Agnusdei, R. Civitelli, and S. Gonnelli, Analgesic activity of salmon and human calcitonin against cancer pain: a double-blind, placebo-controlled clinical study. Curr. Ther. Res. 38, 298 – 308 (1985). J. Szanto, S. Jozsef, J. Rado, E. Juhos, I. Hindy, and S. Eckhardt, Pain killing with calcitonin in patients with malignant tumours. Oncology. 43, 69 – 72 (1986). A. Roth and K. Kolaric, Analgetic activity of calcitonin in patients with painful osteolytic metastases of breast cancer. Results of a controlled randomized study. Oncology. 43, 283 – 287 (1986). J. Szanto, N. Ady, and S. Jozsef, Pain killing with calcitonin nasal spray in patients with malignant tumors. Oncology. 49, 180 – 182 (1992). V. Olgiati, F. Guidobono, C. Netti, and A. Pecile, Localization of calcitonin binding sites in rat central nervous system: Evidence of its neuroactivity. Brain Res. 265, 209 – 215 (1983). F. A. Tschopp, H. Henke, J. B. Petermann, P. H. Tobler, R. Janzer, T. Hîkfelt, J. M. Lundberg, C. Cuello, and J. A. Fischer, Calcitonin gene-related peptide and its binding sites in the human central nervous system and pituitary. Proc. Natl. Acad. Sci. USA 82, 248 – 252 (1985). L. J. Deftos, D. Burton, H. G. Bone, B. D. Catherwood, J. G. Parthermore, R. Y. Moore, S. Minick, and R. Guillemin, Immunoreactive calcitonin in the intermediate lobe of the pituitary. Life Sci. 23, 743 – 748 (1978). D. M. Pavlinac, L. W. Lenhard, J. G. Parthermore, and L. J. Deftos, Immunoreactive calcitonin in human cerebrospinal fluid. J. Clin. Endocrinol. Metab. 50, 717 – 720 (1980). C. E. Fiore, F. Castorina, L. S. Malatino, and C. Tamburino, Antalgic activity of calcitonin: Effectiveness of the epidural and subarachnoid routes in man. Int. J. Clin. Pharm. Res. 3, 257 – 260 (1983). F. Fraioli, A. Fabbri, L. Gnessi, C. Moretti, C. Santoro, and M. Felici, Subarachnoid injection of salmon calcitonin induces analgesia in man. Eur. J. Pharmacol. 78, 381 – 382 (1982). Z. Wiesenfeld-Hallin and A. Persson, Subarachnoid injection of salmon calcitonin does not induce analgesia in rats. Eur. J. Pharmacol. 104, 375 – 377 (1984). R. Maurer, P. Marbach, and R. Mousson, Salmon calcitonin binding sites in rat pituitary. Brain Res. 261, 346 – 348 (1983). J. A. Fischer, P. H. Tobler, M. Kaufmann, W. Born, H. Henke, P. E. Cooper, S. M. Sagar, and J. B. Martin, Calcitonin: regional distribution of the hormone and its binding sites in the human brain and pituitary. Proc. Natl. Acad. Sci. USA 78, 7801 – 7805 (1981). Y. Iwasakii, K. Chiihara, J. Iwasakii, H. Abe, and F. Fakuo, Effect of calcitonin on prolactin release in rats. Life Sci. 25, 1243 – 1245 (1979).

673

CHAPTER 73 Calcitonin for Treatment of Osteoporosis 153. G. V. Shah, W. Wang, C. E. Grosvernor, and W. R. Crowley, Calcitonin inhibits basal and thyrotropin-releasing hormone-induced release of prolactin from anterior pituitary cells: Evidence for a selective action exerted proximal to segretagogue-induced increases in cytosolic Ca2 . Endocrinology. 127, 622 – 629 (1990). 154. V. Sibilia, C. Netti, F. Guidobono, I. Villa, and A. Pecile, Inhibitory effects of centrally administered ASU17eel calcitonin on basal and stimulated prolactin release in rats. J. Endocrinol. Invest. 13, 507 – 511 (1990). 155. S. Nakanishi, T. Inoue, T. Kita, M. Nakamura, A.-C. Chang, S. N. Cohen, and S. Numa, Nucleotide sequence of cloned cDNA for bovine corticotropin releasing -lipotropin precursor. Nature (London) 278, 423 – 429 (1979). 156. L. Laurian, Z. Oberman, E. Graff, S. Gilad, E. Hoerer, and R. Simantov, Calcitonin induced increase in ACTH, -endorphin and cortisol secretion. Horm. Metab. Res. 18, 268– 271 (1986). 157. M. Ustdal, P. Dogan, A. Soyuer, and S. Terzi, Treatment of migraine with salmon calcitonin: Effects on plasma beta-endorphin, ACTH

158.

159.

160.

161. 162.

and cortisol levels. Biomed. Pharmacotherapy. 43, 687 – 691 (1989). P. P. Vescovi, M. Pedrazzoni, G. Gerra, L. Maninetti, Michelini, M., Pioli, G., and Passeri, M., Salmon calcitonin given by nasal spray or by injection does not increase beta-endorphin levels in normal men. Life Sci. 47, 1469 – 1473 (1990). A. Fabbri, C. Santoro, C. Moretti, M. Cappa, F. Fraioli, G. P. Di Julio, T. Galluzzi, and V. La Manna, The analgesic effect of calcitonin in humans: Studies on the role of opioid peptides. Int. J. Clin. Pharmacol. Ther. Toxicol. 19, 509 – 511 (1981). P. C. Braga, S. Ferri, A. Santagostino, V. Olgiati, and A. Pecile, Lack of opiate receptor involvement in centrally induced calcitonin analgesia. Life Sci. 22, 971 – 978 (1978). R. Ceserani, M. Colombo, V. Olgiati, and A. Pecile, Calcitonin and prostaglandin system. Life Sci. 25, 1851 – 1856 (1979). F. Guidobono, C. Netti, P. Villani, P. Bettica, and A. Pecile, Antinociceptive activity of ell calcitonin, injected into the inflamed paw in rats. Neuropharmacology 30, 1275–1278 (1991).

CHAPTER 74

Fluoride Therapy for Osteoporosis K.-H. WILLIAM LAU AND DAVID J. BAYLINK Departments of Medicine and Biochemistry, Jerry L. Pettis Veterans Affairs Medical Center, and Loma Linda University, California 92357

VI. VII. VIII. IX.

I. II. III. IV.

Introduction Molecular Mechanism of Action of Fluoride on Bone Cells Pharmacokinetics and Metabolism of Fluoride Relevance of Serum Fluoride Concentrations to Biological Responses V. Skeletal Responses to Fluoride Therapy

I. INTRODUCTION

of fluoride therapy will be included. Because in-depth understanding of basic aspects of the mechanism of action of fluoride is paramount to the proper use of fluoride in the treatment of osteoporosis and related diseases, information regarding the molecular mechanism of action of fluoride is also discussed. A focused analysis of the clinical trials employing fluoride therapy is presented in Chapter 75.

Fluoride is an essential trace element [1]. The recommended daily fluoride dose is 1.5 to 4 mg [2]. Like all trace elements, fluoride demonstrates biphasic actions in the organism; intakes below the recommended daily dose result in growth and development retardation, while high intakes, such as those associated with long exposure to fluoride from endemic or industrial sources, result in skeletal sclerosis. Pharmacologic doses, 20 to 100 mg/day, cause significant increases in skeletal mass. These observations, along with the recognition that the consequences of osteopenia could be reduced in people who had been exposed to fluoride [3,4], led to the suggestion that fluoride could be of therapeutic value in osteoporosis [5]. This chapter reviews and discusses the clinical use of fluoride for established osteoporosis. Advantages and disadvantages of fluoride therapy are addressed. Because a high benefit-to-risk profile of a drug is essential for efficacy of a therapy, a proposed strategy for improving this profile

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Side Effects of Fluoride Therapy Clinical Efficacy of Fluoride Therapy Strategies for Improving Fluoride Therapy Conclusion References

II. MOLECULAR MECHANISM OF ACTION OF FLUORIDE ON BONE CELLS The first definitive in vitro evidence that fluoride acts directly on chicken bone cells to stimulate cell proliferation was reported in 1983 [6]. These in vitro studies were prompted by earlier bone histomorphometric findings, which indicated that (a) the effect of fluoride treatment to increase bone mass was due entirely to an increase in bone formation and not to a decrease in bone resorption and (b)

675

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

676

FIGURE 1 Fluoride increases human bone cell proliferation in vitro. Human bone cell proliferation was measured by [3H]thymidine incorporation into DNA and by the number of cell population doublings per day. Each data point is the mean  SEM of six replicates. Adapted with permission from Wergedal et al. [9]. the increase in bone formation was characterized by an increase in osteoblast number [7]. A number of laboratories subsequently confirmed the in vitro bone cell mitogenic activity1 of fluoride on bone cells of various species, including humans [8 – 18]. The optimal stimulatory dose of fluoride was approximately 10 M, which is similar to the effective serum fluoride concentrations in fluoride-treated patients (i.e., basal level at 5 – 10 M and peak level at 30 M) [19]. An example of the dose-dependent mitogenic effect of fluoride on human bone cells is shown in Fig. 1, which illustrates that fluoride at clinically relevant concentrations significantly increased DNA synthesis and the cell doubling in human bone cells. Thus, the bone cell mitogenic effect of fluoride is compatible with past bone histomorphometric observations of a fluoride-induced increase in osteoblast proliferation [7,20]. Several unique properties of the bone cell mitogenic activity are relevant to its molecular mechanism: (1) The effective mitogenic dose of fluoride is relatively low (i.e., 10 – 100 M) and is at least two orders of magnitude lower than doses of fluoride required for its effects on other biological systems (i.e., millimolar level) [21]. (2) The in vitro mitogenic activity of fluoride, like its in vivo osteogenic actions, is specific for cells of skeletal origin [6,9]. (3) The bone cell mitogenic activity of fluoride requires the presence of a bone cell growth factor, such as insulin-like growth factor (IGF)-I or transforming growth factor (TGF) [16,22]. (4) Fluoride potentiates the mitogenic action of bone cell growth factors, such as IGF-I, both in vivo [23] and in vitro [21]. (5) The bone cell mitogenic activity of 1

For convenience, we will refer to the ability of fluoride to increase bone cell proliferation (or DNA synthesis) as bone cell mitogenic activity in this chapter.

LAU AND BAYLINK

fluoride is sensitive to changes in the concentrations of inorganic phosphate in culture medium [22,24]. (6) Fluoride acts primarily on osteoprogenitor cells or undifferentiated osteoblasts rather than the highly differentiated, mature osteoblasts [18,25,26]. Finally, (7) the bone cell mitogenic action of fluoride is associated with increases in the overall tyrosine phosphorylation status of several cellular proteins, including mitogen-activated protein kinase (MAPK) [17,21,27 – 29]. Any model for molecular mechanism of the mitogenic action of fluoride must account for these unique characteristics. Several models of the mechanism of bone cell mitogenic activity of fluoride have been proposed. It has been suggested that fluoride (at millimolar concentrations) stimulates the proliferation of L-929 fibroblasts by activating protein kinase C (a protein-serine/threonine kinase) through a heterotrimeric GTP-binding protein (G protein) [30]. Because fluoride treatment of bone cells triggers an acute increase in intracellular calcium levels [28,31] and increases in intracellular calcium concentration are associated with cell proliferation, it has also been proposed that the increase in intracellular calcium in response to fluoride treatment may be, in part, involved in the mechanism of the mitogenic activity of fluoride. Reed and coworkers postulated that the mitogenic activity of fluoride is mediated through modulation of cellular sensitivity to the bone cell growth factor, TGF- [16]. More recently, based on the findings that fluoride at 15 – 50 mM activated phospholipase D in human SaOS-2 osteosarcoma cells, Bourgoin et al. [32] proposed that the mitogenic activity of fluoride may involve phospholipase D activation through a G protein. Although each of these interesting models merits further investigation, none of these proposed mechanisms can account for all of the aforementioned observations, such as low (micromolar) effective doses, the cell- and tissuespecificity, and the absolute requirement of an appropriate growth factor. Lau et al. [21] have advanced an alternative model (summarized in Fig. 2). This model involves stimulation of the MAPK mitogenic signal transduction pathway through inhibition of an osteoblast-specific, fluoride-sensitive, protein-tyrosine phosphatase (PTP) [33]. This model is supported by a large body of strong, albeit circumstantial, evidence [33]. It is tenable and attractive because it accounts for each of the unique properties of the mitogenic activity of fluoride mentioned above. The arguments in favor of this model include the following 7 points: (1) The fluoride inhibition constant of the osteoblastic fluoride-sensitive PTP is in the same low micromolar range as the dose that stimulates bone cell proliferation and bone formation in vitro and in vivo [6,9,19]. (2) The fluoride-sensitive PTP activity is present in osteoblasts but not other cells and tissues, with the exception of kidney [21], which is consistent with the premise that the effect of fluoride (on this enzyme and the

677

CHAPTER 74 Fluoride Therapy for Osteoporosis

FIGURE 2 A proposed molecular mechanism for the mitogenic action of fluoride on bone cells. The binding of a growth factor, such as IGF-I, to its cell surface receptor activates the receptor protein-tyrosine kinase by promoting dimerization and autophosphorylation (top left corner of the graph). The activated receptor protein-tyrosine kinase then initiates a cascade of phosphorylation reactions, leading to the tyrosyl phosphorylation of the ras GTPase activating protein (rasGAP) and the activation of Ras, which in turn phosphorylates and activates Raf. The activated Raf will then phosphorylate and activate MEK, which, in turn, activates MAPK. The activated MAPK can then migrate to the nucleus, where it can phosphorylate and activate the proto-oncogenes and transcriptional factors, which are believed to be responsible for the stimulation of DNA synthesis and cell proliferation. Four potential dephosphorylation sites whereby fluoride could act to inhibit the osteoblastic fluoridesensitive protein-tyrosine phosphatase (PTP) activity are identified in the model with numbers 1 through 4. consequent effect on cell proliferation) is specific for bone cells. (3) A fluoride-dependent inhibition of dephosphorylation of cellular phosphotyrosine proteins can increase their overall tyrosine phosphorylation levels but is only effective when the basal level of phosphorylation has been raised in response to activation of protein-tyrosine kinases (PTKs). Thus, the optimal mitogenic activity of fluoride would require the presence of a PTK-activating growth factor, such as IGF-I, to increase the basal tyrosine phosphorylation level of cellular proteins. (4) Because the mitogenic activity of growth factors, such as IGF-I, is mediated through direct activation of the PTK activity (of their corresponding receptor) and the mitogenic activity of fluoride is presumed to be mediated through an inhibition of the phosphotyrosine dephosphorylation, it follows that fluoride would interact with PTK-activating growth factors to promote bone cell proliferation and bone formation as was reported in vivo [23] and in vitro [21]. (5) Fluoride can act, in coordination with divalent cations, as a transition state analog of inorganic phosphate [34], which is a potent inhibitor of PTPs.

Several known transition-state analogs of inorganic phosphate (i.e., vanadate, molybdate, phenylarsenic oxide), at concentrations that inhibited the fluoride-sensitive PTP, also stimulated bone cell proliferation to the same extent as fluoride [21,35,36]. This might explain why the mitogenic activity of fluoride is sensitive to changes in phosphate concentration in culture medium. (6) Osteoprogenitor cells (and less differentiated bone cells) produce more growth factors [18,37] and appear to contain more of this fluoridesensitive PTP [26] than more differentiated osteoblasts. Thus, it is not surprising that less differentiated osteoprogenitor cells are the preferred target cells for fluoride. (7) Fluoride treatment increases the overall tyrosine phosphorylation level [27 – 29] and activity [28] of MAPK and signaling proteins of the MAPK mitogenic signal transduction pathway in bone cells. Caverzasio and Bonjour [38] have proposed a competing model, which also focuses on the concept that fluoride stimulates bone cell proliferation by increasing the activity of a specific PTK (as opposed to inhibiting a PTP). In contrast to

678

LAU AND BAYLINK

the aforementioned model, these investigators postulated that fluoride complexes with the aluminum ion, forming fluoroaluminate ion (AlF4). AlF4 then acts directly on a specific inhibitory heterotrimeric GTP-binding protein (Gi/o) in bone cells, which subsequently leads to activation of one or more PTKs, resulting in the stimulation of tyrosine phosphorylation of key signaling proteins, including Shc and MAPK. The activation of these signaling proteins is then responsible for the subsequent stimulation of cell proliferation [29,38,39]. This model is supported by their findings that (1) there appeared to be an absolute requirement of 10 – 50 M aluminum ion for fluoride to stimulate cell proliferation and MAPK tyrosine phosphorylation in rodent bone cells [29], (2) the mitogenic activity of AlF4 was blocked by a PTK inhibitor, genistein [29], and (3) pertussis toxin, a presumed specific inhibitor of Gi or Go proteins, completely blocked the mitogenic activity of AlF4 [39]. Although their data are generally consistent with their conclusions, there are alternative interpretations to several of their findings, as previously discussed elsewhere [33]. More importantly, this model fails to account for several important characteristics of the mitogenic activity of fluoride, such as the cell/tissue-specificity, the requirement for growth factor(s), or the sensitivity to phosphate concentration. In summary, although the precise molecular basis of fluoride’s mitogenic activity on osteoblast-line cells has not been definitively established, recent studies indicate that an increased level of tyrosine phosphorylation of a signaling protein that regulates mitosis (e.g., MAPK) is probably involved. The most likely explanation for the increase in tyrosine phosphorylation of key mitogenic signaling proteins is an inhibition of an osteoblast-specific PTP(s) by fluoride. However, regardless of whether the enhancement is due to an inhibition of a unique fluoridesensitive PTP or a stimulation of PTKs through a Gi/o protein, it appears that one may increase bone cell proliferation by enhancing a key signal transduction pathway (e.g., the MAPK pathway). Thus, it opens up a new and exciting area of research in which the MAPK signal transduction pathway may be used as the screening target for discovery of new anabolic drugs for osteoporosis and related bone diseases.

III. PHARMACOKINETICS AND METABOLISM OF FLUORIDE A. Fluoride Salts and Preparations Two fluoride salts are presently available for clinical uses: plain sodium fluoride (NaF) and monofluorophosphate (MFP). MFP has three important advantages over NaF: (1) NaF often causes gastrointestinal discomfort due to gastric absorption and formation of hydrofluoric acid in

the stomach. In contrast, MFP is not absorbed in the stomach, does not form hydrofluoric acid, and is hydrolyzed by alkaline phosphatase to release free fluoride ion for rapid absorption in the duodenum [40,42]. (2) The intestinal absorption of NaF [41], but not MFP [40,42], is reduced by dietary calcium. Thus, MFP, but not NaF, can be taken simultaneously with calcium without concern for decreased absorption. (3) The bioavailability of the fluoride ion from MFP is much higher than that from NaF [43 – 45]. Both fluoride salts are available in enteric-coated and sustained-release preparations. The enteric-coated (galenic) preparations would reduce gastrointestinal discomfort. The sustained-release formulations allow a gradual (slow) release of fluoride ion, allowing the maintenance of the serum fluoride concentration at therapeutic levels without sharp postabsorption peaks. These peaks are not required for an osteogenic effect, but could markedly increase fluoride deposition in bone, which has a deleterious effect on bone quality (see below). The sustained-release preparations also cause less gastrointestinal irritation than the non-sustained-release forms.

B. Fluoride Absorption Fluoride ion is absorbed in the gastrointestinal tract, presumably via passive mechanisms [46,47]. The majority of the fluoride ion is absorbed in the small intestine, primarily in the duodenum and jejunum [48]. However, a considerable amount of fluoride ion is also absorbed in the stomach [49]. The rate of intestinal fluoride absorption depends on: (1) the nature of the fluoride salt and the galenic formulation; (2) the type and amount of other ions in the intestine, such as Ca2 , Mg2 , Al3 , and Cl, (which, particularly Ca2 , reduce the rate and extent of intestinal fluoride absorption, presumably due to formation of poorly absorbed fluoride compounds, e.g., CaF2 [50]), and (3) the physiological and pathological state of patients, such as age, acid – base equilibrium, gastric and urine pH, and renal functions [51].

C. Skeletal Deposition of Fluoride Once absorbed, the circulating fluoride is distributed mainly into two compartments: (a) extracellular fluid and soft tissues, where it has a relatively short half-life of a few hours, and (b) hard tissues (bone, calcified cartilages, and teeth), in which the half-life can be up to several years. The primary deposition site of fluoride is bone, where fluoride ion is exchanged for the hydroxyl group in hydroxyapatite to form fluoroapatite, which is biologically inert. 18F studies have shown that fluoride in bone is

679

CHAPTER 74 Fluoride Therapy for Osteoporosis

not homogeneously distributed but is deposited mostly in areas where active mineralization is taking place (i.e., newly formed bone during the therapy) [52]. Thus, along with the dosage and treatment duration of the fluoride therapy, the amount of bone deposition of fluoride is also determined by the bone formation rate. Fluoride deposited in the skeleton is removed by bone resorption during remodeling. Part of the released bone fluoride recycles into newly formed bone mineral, while the rest is excreted in the urine. Accordingly, the bone fluoride content would only drop slowly, but progressively, after fluoride therapy is stopped [53]. This is responsible for the relatively long half-life of fluoride in the bone.

D. Renal Fluoride Excretion In humans, 75% of the total excreted fluoride is in the urine [54]. Thus, the kidney is the major clearance site of circulatory fluoride ion, where it is filtered freely [55]. The extent of renal fluoride excretion depends on the filtered load (i.e., the glomerular filtration rate  the serum fluoride concentration) and free water clearance (i.e., the greater the free water clearance, the greater the fluoride excretion) [56]. Because diet can affect urinary pH and urinary pH is a key determinant of urinary fluoride clearance, diet may have an effect on renal fluoride clearance. Vegetarian diets, which are alkaline, could result in greater fluoride retention [56]. Some workers have suggested that renal fluoride excretion can be used as a clinical index to predict the subsequent increase in bone density [57,58]. Accordingly, in a group of osteoporotic patients with normal renal function, there was a positive correlation between urinary fluoride excretion and the increase in bone density. However, it is also possible that the increase in urinary fluoride reflected a higher serum concentration, which itself determined the subsequent increase in bone density. This alternative interpretation would be consistent with the results of another study in which “good” and “poor” responders to fluoride were evaluated in terms of fluoride kinetics [59]. In the good responder group, there was a lower renal fluoride clearance and a greater extrarenal clearance (i.e., a greater deposition in bone). Thus, the good responders had slightly impaired renal function and a higher serum level of fluoride, which could have determined their skeletal response. On the other hand, too high levels of fluoride retention could lead to skeletal fluorosis. Therefore, it should be emphasized that, inasmuch as the kidney is the major clearance route of fluoride, even a moderate impairment of renal function could predispose to excessive fluoride retention during fluoride therapy [60]. Increased fluoride retention would increase the risk of skeletal fluorosis. Therefore, fluoride should only be used as a therapy with great caution, or not at all, in patients with renal insufficiency [61].

IV. RELEVANCE OF SERUM FLUORIDE CONCENTRATIONS TO BIOLOGICAL RESPONSES Three pieces of circumstantial evidence led to the conclusion that serum concentration of fluoride is the primary determinant of the skeletal response: first, in vitro studies indicate that the fluoride ion acts directly on osteoblast-line cells. The circulating free fluoride ion is probably the biologically active species, since the bone matrix-bound fluoride (in the form of fluoroapaptite) is biologically inert. Second, the optimal in vitro osteogenic concentrations of fluoride were in the same range as the effective serum fluoride concentrations in fluoride-treated patients. Third, good fluoride responders showed a higher serum fluoride concentration than poor responders [59]. Figure. 3 shows that the osteogenic response (i.e., increases in spinal bone density) correlated positively with the daily oral dose but not with treatment duration [62]. The serum fluoride concentration is determined primarily by the oral dose; thus, this provides additional evidence for the premise that serum fluoride concentration is the major determinant of the skeletal response.

A. Optimal Serum Fluoride Concentrations The optimal serum fluoride concentrations in humans have not been established clearly. Low serum fluoride concentrations would not be clinically effective, whereas high concentrations generally result in cytotoxicity and cause severe side effects. Two problems have precluded a definitive recommendation of an optimal serum fluoride concentration: first, there is a paucity of data on controlled studies of dose – effect relationships of fluoride therapy. Second, determinations of serum fluoride concentrations vary significantly

FIGURE 3

The daily dose of fluoride was related to the rate of change in spinal bone density in 41 osteoporotic patients receiving fluoride (range 15 to 43 mg/day) and calcium (1500 mg/day). Adapted with permission from Dure-Smith et al. [62].

680 among different commercial laboratories, partly due to the fact that the standard curve is not linear and that measurements usually involve extrapolation. Nonetheless, Taves in 1970 [63] proposed that fasting serum fluoride concentrations be maintained between 5 and 10 M for an osteogenic response in humans. Although these proposed limits have frequently been referred to as the theoretical “therapeutic window” for fluoride therapy, they have not been confirmed experimentally. Anecdotal data indicate that serum fluoride concentrations below the lower limit of 5 M rarely produce significant increases in bone formation in humans. Thus, it seems reasonable to set 5 M as the low limit. The determination of the upper limit is more difficult, because risk/benefit data are lacking for serum concentrations above 15 M. In vitro studies (e.g., Fig. 1) show that dose-dependent stimulation of human osteoblast proliferation persists at fluoride concentrations of 10 – 30 M but is linear only up to 10 M [9]. While it is likely that a higher serum fluoride concentration will produce a greater increase in bone formation (Fig. 3), it is also conceivable that the greater increases in bone formation by higher serum fluoride concentrations will be offset by a greater propensity for harmful side effects. In this regard, the side effects of gastrointestinal irritation and peripheral pain syndrome are dose-dependent. The development of osteomalacia associated with fluoride administration is probably also dose-related. Thus, the optimal upper limit should be one that yields the highest benefit-to-risk ratio. Unfortunately, information regarding the relationship between benefit-to-risk data and serum fluoride concentration is lacking. On the other hand, an appropriate fluoride dose should be one that yields a blood fluoride concentration that stays within the upper linear portion of the mitogenic dose – response curve, i.e., 10 M (Fig. 1). Thus, it would seem reasonable to set the upper limit at 10 M. Two additional observations support the rationale for recommending 5 to 10 M as a safe and effective level: (1) a consistent stimulation of bone formation was observed in patients when fasting serum fluoride concentrations were maintained at 5 – 10 M and (2) no serious adverse side effects were observed at a serum fluoride concentration of 5 – 10 M. Consequently, until more information regarding benefit-to-risk ratio of fluoride in humans becomes available, a morning predose serum fluoride concentration of 5 – 10 M would seem to be a reasonable therapeutic serum concentration for adults with established osteoporosis.

B. Dosage and Regimen of Fluoride Therapy Analyses of the results of three randomized, prospective, placebo-controlled studies appear to reinforce the concept that the dosage and the regimen of fluoride treatment each play an important role in determining therapeutic efficacy

LAU AND BAYLINK

[64 – 68]. Two of these three studies (the Mayo Clinic and the Henry Ford Hospital studies) [64,65] followed a protocol in which postmenopausal women were treated with 75 mg/day plain NaF (or 34 mg/day elemental fluoride) together with calcium carbonate at a dose of 1500 mg/day for 4 years. Despite a 35% increase in spinal bone density, neither study showed a significant reduction in vertebral fracture rate compared to the placebo-treated patients [64,65]. However, an extended analysis of a subgroup of 50 patients in the Mayo Clinic study, who, because of side effects, did not tolerate the full fluoride dose and received a dose reduction, revealed that this subgroup of patients exhibited moderate but significant decreases in the vertebral fracture rate, even though they had a smaller increase in spinal bone density (i.e.,  17%) and a lower serum fluoride concentration (i.e.,  8 M) [66]. This post hoc analysis, which lacks the rigor of a controlled study, also revealed that patients who experienced very rapid increases in bone density or large augmentations in serum fluoride concentrations failed to benefit with a decrease in vertebral fracture rate [66]. Thus, it seems that higher fluoride doses, while showing rapid increases in bone density and large increments in serum fluoride concentrations, did not reduce the vertebral fracture rate, whereas lower fluoride doses, which yielded lesser increments in serum fluoride concentrations and bone density, were effective in reducing vertebral fracture rate. The report of severe osteomalacia in the transilial biopsies from subjects of the Mayo Clinic study, even though these patients had received daily supplementation of 1500 mg calcium carbonate [69] provides a potential mechanism that may explain the lack of a beneficial effect on vertebral fractures in patients treated with higher doses of fluoride. Accordingly, if higher serum fluoride concentrations produced a relatively large increase in bone formation, it follows that the higher the increment in serum fluoride concentration, the greater the demand for calcium and the greater calcium demand may lead to development of osteomalacia. Thus, patients with a lower serum fluoride level may have a lesser tendency toward osteomalacia. However, a second explanation relates to the fact that higher fluoride doses (and higher serum fluoride concentrations) result in an increased deposition of fluoride into bone mineral, which can have a deleterious effect on bone quality and strength (see below). Therefore, development of osteomalacia and/or high bone deposition of fluoride in patients treated with higher doses of fluoride may lead to the lack of beneficial effect on fracture rate reduction. The data from the third, randomized, placebo-controlled study from the University of Texas Southwestern Medical Center [67,68] are consistent with the interpretation that lower fluoride doses, rather than higher doses, are beneficial (see Chapter 75). Unlike the other two studies, patients in this study were treated for 4 years with an intermittent slowrelease NaF of 50 mg/day (23 mg/day elemental fluoride)

681

CHAPTER 74 Fluoride Therapy for Osteoporosis

and 800 mg/day calcium citrate supplementation in a dosage regimen consisting of 14-month cycles (12 months receiving fluoride and 2 months off therapy). Patients in this study, unlike those in the other two studies, showed a significant reduction in new vertebral fracture rate accompanying a moderate increase in spinal bone density [67,68]. Two aspects of this trial may have precluded the development of osteomalacia: (1) the use of a lower fluoride dose than was given in the two studies performed by Mayo Clinic and Henry Ford Hospital, and (2) the intermittent design of the trial. It is further speculated that the time off fluoride, together with the use of calcium citrate, which is more soluble and readily absorbed than calcium carbonate, would facilitate the resolution of any potential osteomalacia that might have occurred during the therapy. Accordingly, there was no clear evidence for osteomalacia and secondary hyperparathyroidism in patients in this study [67,68]. The “off” fluoride period could also reduce the amount of fluoride deposited in bone. Thus, the observed reduction in fracture rate in this study may be attributed to the absence of osteomalacia and the reduced fluoride deposition in bone. Consequently, these observations may indicate that regimens containing moderate and cyclic doses of fluoride can have a favorable benefit-to-risk ratio. However, additional work is needed to determine the most appropriate dosage and cycling time of the fluoride therapy.

V. SKELETAL RESPONSES TO FLUORIDE THERAPY

for assessing the response to fluoride therapy. Unfortunately, fluoride therapy can also cause osteomalacia, which is also associated with an increase in serum alkaline phosphatase activity.2 Therefore, in patients receiving fluoride therapy, measurement of serum alkaline phosphatase activity alone does not allow discrimination between the two possible etiologies for a rise in this serum marker, namely an increase in bone formation or a mineralization defect. Hence, it is currently not possible to interpret an increase in serum alkaline phosphatase activity as unambiguous evidence of either increased bone formation or osteomalacia. Other markers may assist in this distinction. Accordingly, when the increase in serum alkaline phosphatase is associated with calcium deficiency and osteomalacia, there will also be changes in other markers typically associated with secondary hyperparathyroidism, such as a decrease in urine calcium, an increase in serum parathyroid hormone (PTH) concentrations, and an increase in urine bone resorption markers. Conversely, there should be no such changes in these markers accompanying the rise in serum alkaline phosphatase activity associated with fluoride-related increases in bone formation. In vitamin D deficiency-associated osteomalacia, there is an increase in skeletal alkaline phosphatase but not an increase in serum procollagen peptides (another serum bone formation marker) [73]. Thus, it is possible that one could measure serum procollagen peptide concentrations to assess the osteogenic action of fluoride or as a means of determining whether the increase in skeletal alkaline phosphatase indicates a mineralization defect.

Clinical responses to fluoride therapy may be assessed by increases in serum bone formation biochemical markers, histomorphometric parameters, and bone density.

A. Biochemical Markers Serum bone formation biochemical markers, such as skeletal alkaline phosphatase and osteocalcin, are useful in assessing bone formation in response to therapy in patients (see Chapter 60). Thus, serial measurements of these bone formation markers are often used as an acute quantitative index of the skeletal response to fluoride. Fluoride administration results in significant increases in serum skeletal alkaline phosphatase activity [70] and osteocalcin concentration [71,72] within weeks. Importantly, increases in serum alkaline phosphatase activity precede improvement of clinical symptoms (e.g., reducing back pain) and bone density gains, and increases in serum skeletal alkaline phosphatase activity correlate with the fluoride-dependent increase in histomorphometric bone formation parameters [70]. Thus, in theory, serum bone formation markers, especially skeletal alkaline phosphatase activity, are useful early indicators

B. Bone Histomorphometry Histomorphometric data from both animal and human studies confirm the in vitro osteogenic action of fluoride by showing increases in osteoid and bone formation histomorphometric parameters [74 – 79]. Bone histomorphometric data in humans are almost exclusively derived from transilial biopsies. Fluoride treatment increases the number of osteoblasts and osteoblast-covered surfaces [7,14,20,74] without direct effects on bone resorption parameters [7,20], indicating that the fluoride-induced bone formation is mediated through an increase in osteoblast proliferation. Therefore, it is generally believed that the rise in the number of osteoblasts through increased osteoblast proliferation leads to subsequent overfilling of the resorption excavation cavity (Fig. 4), and this, in turn, results in an increase in trabecular thickness and, thus, bone volume.

2

Because increased calcium absorption is required during fluoride therapy, vitamin D deficiency should be excluded (i.e., a normal serum 25(OH)D) before fluoride therapy is initiated.

682

FIGURE 4

Model of the effect of fluoride on bone formation. In osteoporosis, the resorption cavities are underfilled, due to impaired bone formation. As a consequence, bone density decreases at a given site with each cycle of bone remodeling. Fluoride corrects this deficit by promoting an overfilling of the resorptive cavity (and thereby increasing mean wall thickness). The activation frequency (the number of active resorptive sites created per unit of bone surface per unit of time in a given unit of bone) is not affected by fluoride [80].

LAU AND BAYLINK

produced by fluoride therapy [86] and because calcium defi ciency is a recognized cause of osteomalacia [87,88], it seems likely that a major cause of osteomalacia would be the calcium deficiency. On the other hand, we cannot overlook the possibility that high fluoride incorporation into bone mineral per se may directly retard the mineralization process through physicochemical interactions and, thus, may in part be responsible for the fluoride-mediated osteomalacia. Regardless of the mechanism, osteomalacia impairs the mechanical properties of the bone. Therefore, the greater incidence of poor bone quality is associated with the greater calcium deficit (and osteomalacia). Consequently, the significance of this observation is that calcium defi ciency-associated osteomalacia may have contributed, to some extent, to the lack of positive effects of fluoride in some past fracture studies.

C. Bone Density Hence, in bone histomorphometric terms, fluoride increases bone volume by increasing mean wall thickness [80]. This concept is supported by the findings that the increase in osteoid volume is due, in large part, to an increase in osteoid surface and, to a lesser extent, to an increase in osteoid thickness [20,78 – 80]. Despite positive reports from two groups [81,82], it has not been established that fluoride promotes trabecular connectivity. Therefore, we cannot assume that fluoride can restore the microarchitecture that may have been disrupted during the development of osteoporosis; the action of fluoride may be limited to a thickening of existing (remaining) trabeculae. There is compelling evidence that fluoride therapy is often associated with a delay in mineralizing newly synthesized osteoid. Early studies recognized that osteomalacia was common in bone biopsy samples from patients receiving fluoride therapy (Fig. 5) [69,78,83,84]. This is clearly evident in the Mayo Clinic trial [64,66] in which fluoridetreated patients showed a 14-fold increase in the mineralization lag time over the placebo control subjects (Fig. 6). The mineralization lag time is a specific measure of the degree of osteomalacia. The osteomalacia effect is related to fluoride dose and, also, to the availability of adequate calcium to mineralize the osteoid matrix [7,78,80,85]. Accordingly, large and rapid fluoride-induced increases in matrix deposition are more likely associated with a mineralization defect than smaller and slower increases. Elderly patients have a reduced intestinal calcium absorption efficiency; thus, the calcium insufficiency due to a rapid and large increase in bone formation in response to fluoride would probably be more pronounced in elderly subjects. Administration of calcium should reduce, but may not eliminate, the fluoride-induced osteomalacia. Because physiological doses of 1,25(OH)2D corrected the calcium deficiency (and, perhaps, also the osteomalacia)

An important end point of fluoride therapy is to increase bone mass and density. Bone density can be evaluated by precise and accurate noninvasive methods, viz. dual-energy X-ray absorptiometry (DXA) and quantitative computed tomography (QCT), in both the axial and the peripheral skeleton. The bone density changes induced by fluoride are of different magnitude in these two skeletal compartments and, thus, the effects of fluoride on the axial and peripheral skeleton, respectively, will be addressed separately. 1. AXIAL SKELETON Fluoride therapy has consistently been shown to increase axial trabecular (lumbar spine) bone density in osteoporotic patients. The fluoride-induced increase in axial bone density is progressive with treatment duration [64,89 – 91], as shown in Fig. 7, which shows that the increase in spinal bone density in postmenopausal women is proportional to treatment time for at least 5 years. The dose – response increase in spinal bone density has not been evaluated rigorously in prospective controlled studies, but there seems to be a dose – response increase in spinal bone density between 5 and 40 mg/day of elemental fluoride (Fig. 3). Thus, the increase in spinal bone density in response to fluoride is proportional to the dosage and treatment duration. Depending on the fluoride dose and the regimen utilized, reported annual increases in axial bone density for groups of patients vary between 4 – 10% (by DXA) and 20 – 30% (by QCT), respectively. A larger percentage increase is seen with QCT, because QCT measures vertebral trabecular bone exclusively, and fluoride therapy preferentially increases trabecular bone mass. Such a significant increase in bone density can correct the spinal bone density deficit which is characteristic of osteoporosis. This unique feature of fluoride therapy, namely, to be able

683

CHAPTER 74 Fluoride Therapy for Osteoporosis

FIGURE 5 Goldner’s stained undemineralized bone sections from iliac crest biopsies from a control subject (A) and a fluoride treated patient (B) (at a dose of 30 mg/day of elemental fluoride for 2 years). Note the wide osteoid seam in B (arrow) and the wide osteoid around the osteocytes, which is typical of calcium deficiency osteomalacia.

to increase a patient’s low bone density to a level that is considerably above that of patients with established osteoporosis (i.e.,  2.5 T score) [92], has maintained the clinical interest in fluoride as a therapy for osteoporosis. The increase in spinal bone density with fluoride therapy is unrelated to the severity of the osteoporosis, the age of the patient, or the cause of the osteoporosis. Elderly osteoporotic

patients may show increases in trabecular vertebral bone density from a very low value of 20 to 200 mg/cm3, a value above the peak bone density. Fluoride therapy is just as effective in producing large increases in spinal bone density in patients with glucocorticoid-induced osteoporosis [93] or idiopathic osteoporosis [94] as in those with postmenopausal osteoporosis. However, the gain in spinal bone density from

684

FIGURE 6

Effect of 4 years of fluoride therapy on the mineralization lag time in iliac crest biopsy samples from the Mayo Clinic prospective placebo-controlled clinical trial [64,66]. Values represent mean  SEM. This presents compelling bone histomorphometric evidence that fluoride therapy causes osteomalacia. Adapted with permission from Lundy et al. [69].

fluoride therapy is not maintained after cessation of the therapy [95]. Moreover, not all patients respond to fluoride therapy. Of the osteoporotic patients, 20 – 25% do not manifest an increase in spinal bone density in response to the therapy and, therefore, are considered nonresponders [7,79,96]. 2. PERIPHERAL SKELETON There is radiological [97] and bone density [64,98,99] evidence that fluoride therapy increases new bone formation in a number of peripheral skeletal sites that contain a large amount of trabecular bones, such as knee, ankle, foot, hip, tibial diaphysis, and femoral condyle. Histomorphometric studies indicate that iliac cortical bone does not show increased porosity as a result of fluoride therapy but does show an increase in osteoid. Thus, it is possible that fluoride therapy may also increase bone formation in cortical bones.

FIGURE 7

The fluoride-dependent increase in spinal bone density as measured by QCT, as a function of time, in 389 osteoporotic patients treated with fluoride (30  8 mg elemental fluoride/day) and calcium carbonate (1500 mg/day) for up to 6 years. Note the apparent linear increase in spinal bone density with fluoride treatment time for up to 6 years. Reprinted with permission from Lau et al. [144].

LAU AND BAYLINK

FIGURE 8 Bone density increases at different site in osteoporotic subjects in response to fluoride therapy. Increases in bone density were measured by QCT in the spine and femoral condyle and by DXA in the hip. (Note that only trabecular bone density was measured by QCT in the spine and femoral condyle, whereas cortical and trabecular bone densities were measured by DXA in the hip. Thus, the different responses in spine and hip are, in part, due to different techniques.) Bone density increases associated with fluoride therapy are typically larger in the spine than in the axial skeleton and are also typically larger in the axial skeleton and are also typically larger in trabecular (compared to cortical) bone. Reprinted with permission from Dure-Smith et al. [62]. The magnitude of increases in bone density at peripheral skeletal sites is much smaller than that seen in the axial skeleton (Fig. 8) [62,100]. However, the osteoporotic deficit in the peripheral skeleton (e.g., hip) is also less severe than in the axial skeleton (e.g., spine). The axial skeleton is made up mostly of trabecular bone, while the peripheral skeleton predominantly consists of cortical bone. Thus, the relative differences in bone density changes between the peripheral and the axial skeleton could, in part, reflect the lower inherent turnover of cortical bone compared to that of trabecular bone. Accordingly, cortical bone density measured at the wrist during fluoride therapy reveals small decreases or no change, and when cortical bone is assessed at the hip or the femoral shaft, the changes range from small decreases to modest increases. In contrast, measurements in regions of the peripheral skeleton rich in trabecular bone, such as the proximal tibia or the femoral condyle, show significant gains in bone density [98,99]. However, the important issue is whether the small increase in bone density of the hip with fluoride therapy would suffice to reduce hip fracture risk. In this respect, there is currently no reliable evidence that fluoride therapy significantly reduces hip fracture risk. Mechanical loading enhances the osteogenic effect of fluoride at peripheral skeletal sites, as fluoride treatment seems to increase bone density of weight-bearing bones more than that of non-weight-bearing bones [93]. Mechanical loading has been shown to increase the local bone cell production of growth factors [101]. Because the model of molecular mechanism of fluoride (Fig. 2) predicts a synergy between bone cell growth factors and fluoride, it could

685

CHAPTER 74 Fluoride Therapy for Osteoporosis

be speculated that the greater increase in bone density in weight-bearing compared to non-weight-bearing skeleton is the result of a synergism between growth factors released by bone cells during mechanical loading and fluoride action on bone formation. The prospective study of the Mayo Clinic [64] reported that there was a highly significant 7.7% per year decrease in bone density of forearm in fluoride-treated patients, in spite of a large increase in spinal bone density. This finding led the investigators to speculate that the large increase in spinal bone density occurs at the expense of the cortical bone of the peripheral skeleton. A potential mechanistic reason for the loss of cortical bone at peripheral bone sites is due to fluoride-induced calcium deficiency. In this respect, patients in the Mayo Clinic study developed severe osteomalacia (Fig. 6) [69], suggesting that they might also have developed calcium deficiency, which leads to development of secondary hyperparathyroidism. The cortical bone of the peripheral skeleton appears to be more susceptible to secondary hyperparathyroidism than the axial skeleton. Consequently, the calcium deficiency and the accompanying secondary hyperparathyroidism associated with the fluoride therapy were, to some extent, responsible for the peripheral cortical bone loss in these patients. Consistent with this speculation, patients in the prospective controlled trial of the Southwestern Medical Center, who exhibited no evidence of calcium deficiency, did not lose cortical bone mass at peripheral bone sites [67,68].

VI. SIDE EFFECTS OF FLUORIDE THERAPY Significant side effects of fluoride therapy for osteoporosis include gastrointestinal irritation, peripheral joint pain, calcium deficiency, stress fractures, and perhaps also hip fractures.

A. Gastrointestinal Irritation As indicated earlier, free fluoride ion can be absorbed through the stomach and also reacts with gastric acid to form hydrofluoric acid, which is an irritant. Therefore, in the past, gastrointestinal irritation was a common side effect of fluoride therapy [55], occurring in more than 25% of the fluoride-treated patients. Although the most common symptoms, epigastric pain, nausea, and vomiting (duodenal ulceration and bleeding rarely developed), could be effectively treated with antacids or H2 blockers, these actions also led to a decrease in fluoride absorption. The severity of the gastrointestinal irritation side effect is dose-related. Thus, this side effect can also be treated by

reducing the dosage or by temporary discontinuation of the therapy. However, with the development and recent use of enteric-coated and galenic-sustained release formulations of fluoride and the use of monofluorophosphate as an alternative fluoride source, which effectively avoids gastric absorption and interaction with gastric acids, the incidence of gastrointestinal side effects has virtually been eliminated.

B. Peripheral Joint Pain The occurrence of periarticular pain, sometimes severe and usually in the lower extremities, has been long recognized as a complication of fluoride therapy. Severe pain, requiring temporary discontinuation of the therapy, develops at the knees, ankles, or feet in 10 – 40% of fluoride-treated patients [55,102]. This side effect is also dose-dependent and is less frequent in patients receiving a low daily dose (e.g., 20 mg of elemental fluoride), especially so when this dose is given cyclically [103]. The etiology of the pain is controversial. Although some investigators argue that it invariably represents the occurrence of stress fractures [104 – 106], there are at least three pieces of circumstantial evidence indicating that the peripheral pain syndrome results from a marked local anabolic skeletal response(s) to fluoride [100,107], a phenomenon that is analogous to the bone pain experienced in rapidly growing children. First, the total body scintigram (Fig. 9) shows that increased uptake is primarily in trabecular-rich regions of the weight-bearing lower skeleton, such as the ankles and femoral condyles, and is diffuse and bilaterally symmetrical [108]. The pain is also often bilateral, which is rare for stress fractures. Second, radiological changes indicative of new bone formation are frequently associated with painful episodes in many fluoride-treated patients [97]. Third, the symptoms resolve within 2 – 3 days of temporary discontinuation of the therapy or lowering the dosage but return when the therapy is restarted. Notwithstanding the fact that fluoride therapy can cause stress fracture [64,104 – 106], it seems likely that the majority of the peripheral joint pain in the lower extremities during fluoride therapy is due to a local rapid increase in periosteal bone formation. If this assumption is confirmed, the peripheral join pain, albeit an undesirable side effect, may indicate a positive osteogenic response. Although patients exhibiting the peripheral pain syndrome frequently are good responders to fluoride treatment in terms of an increase in spinal bone density, there are patients with severe osteoarthritis who show increased susceptibility to the peripheral pain syndrome (i.e., pain in the arthritic joints) and who do not necessarily show a good increase in spinal bone density. Thus, the peripheral pain syndrome seems to have a complex etiology.

686

LAU AND BAYLINK

elderly patients still developed osteomalacia (Figs. 5 and 6) and calcium deficiency [57,84,86]. Therefore, it is particularly important to recognize the development of calcium deficiency in fluoride-treated patients, because this side effect can be completely corrected by administration of calcitriol (or alfacalcidol) and calcium. The possible development of calcium deficiency in patients responding to fluoride can be assessed by comparison of 24-hr urine calcium excretion as well as serum PTH and serum bone formation markers before and after the treatment. If calcium deficiency has developed, the 24-h urinary calcium will be decreased, and serum PTH will be correspondingly increased [86]. More importantly, one must have a high degree of suspicion of potential complication of calcium deficiency when the patient shows a large and rapid increase in bone formation (i.e., a marked increase in bone formation biochemical markers) [86].

D. Stress Fractures

FIGURE 9

Skeletal scintigrams (Tc-99m methylene diphosphonate) before (left) and 11 months after (right) fluoride therapy. Note that marked increased activity in metaphyseal regions of the weight bearing peripheral skeleton which are rich in trabecular bone. These scintigraphic findings are consistent with an increase in bone formation and are associated with increases in bone density as shown in Fig. 11. The increased scintigraphic uptake was probably also exaggerated by calcium deficiency. Reprinted with permission from Schulz et al. [108].

C. Calcium Deficiency Of major concern is the observation that very rapid and large increases in bone formation (e.g., as in response to fluoride therapy) may lead to calcium deficiency. Calcium deficiency is a serious side effect because it can lead to secondary hyperparathyroidism, increased bone resorption, osteomalacia, and also peripheral bone loss [57,86]. Osteomalacia can also impair the healing of microdamage in the bone, thereby leading to an accumulation of microdamage in the skeleton. Earlier studies on fluoride therapy had already raised concerns about the need to assure that enough calcium be available to enable adequate mineralization of newly formed matrix to avoid osteomalacia. Unfortunately, despite the supplementation of large doses of calcium carbonate, many

While the histomorphometric outcomes of severe calcium deficiency are cortical bone loss and osteomalacia, the clinical consequence is bone with poor mechanical performance. The cortical bone loss weakens the bone, and, as indicated above, osteomalacia impairs the healing of microdamage, and the accumulated microdamage increases the risk for stress fractures. Accordingly, fluoride therapy may increase the prevalence of stress fractures in weight-bearing peripheral bones (primarily in the legs) [64]. However, fluoride-related stress fractures is considered a relatively benign complication, because they are predominantly incomplete fractures and they heal after a temporary discontinuation of the therapy [64,91]. Stress fractures are more likely to develop in patients who (a) have more severe osteoporosis at the beginning of therapy and (b) show a rapid increase in serum alkaline phosphatase activity in response to fluoride. It may be postulated that the major underlying cause of stress fracture is the calcium deficiency, which are results of (1) the rapid increase in bone formation in response to the therapy and (2) the aging-related reduction in calcium absorption. Thus, the greater the increase in bone formation in response to fluoride therapy, the greater the calcium deficit and the associated risk for stress fractures. This situation would be further exacerbated by poor calcium absorption, such as that which occurs with aging. Figure 10 illustrates a proposed model of the mechanism(s) whereby stress fractures may develop in some fluoride-treated patients. Accordingly, calcium deficiency causes secondary hyperparathyroidism, which leads to cortical bone loss and trabecular perforation. Calcium deficiency delays mineralization (i.e., osteomalacia), leading to an accumulation of microdamage. These effects together would be expected to increase the risk for stress fractures and, perhaps, hip fractures, also.

687

CHAPTER 74 Fluoride Therapy for Osteoporosis

250 patients for over a 5-year period showed that the therapy did not increase the frequency of hip fractures [112]; (2) the trial of the Southwestern Medical Center, using the intermittent slow-release fluoride protocol, did not show an increase in incidence of hip or stress fractures [67,68]. However, these negative findings should not be interpreted as proof that fluoride could not, under certain conditions, contribute to stress and complete fractures, including hip fractures. As described above, there are theoretical reasons (i.e., fluoride-induced calcium deficiency and the associated osteomalacia) to suspect that fluoride therapy could increase the risk of fractures, including hip fractures. Therefore, much future work is needed to definitively resolve the issue of whether fluoride therapy can cause hip fractures. FIGURE 10 Proposed pathogenic mechanism(s) of stress fracture development in fluoride-treated osteoporotic subjects. The fluoride-induced increase in bone formation in the presence of calcium malabsorption (such as occurs in elderly subjects) results in a state of secondary hyperparathyroidism which, in turn, results in cortical bone resorption and cortical bone loss. The weakened bone develops microdamage for which healing is delayed because of the osteomalacia and, as a result, the patient is at risk for stress fractures. This model emphasizes the importance of assuring adequate calcium absorption in fluoride-treated patients. We should not overlook the fact that a reduction of symptoms (e.g., back pain) in fluoride responders may allow patients to engage in additional physical activities. The increase in physical activity, coupled with the calcium deficiency in patients who respond to fluoride therapy with rapid and large increases in bone formation, could also be a contributing factor to stress fractures, especially in patients with severe osteoporosis.

E. Hip Fractures Two small studies have suggested that fluoride therapy might promote an increased incidence of spontaneous hip fractures in severely osteoporotic patients [109,110] and that fluoride is the cause of the hip fracture [110]. Because of the very serious nature of hip fractures, a collective retrospective analysis of the effect of fluoride therapy on hip fractures was performed on more than 1000 patient-years of fluoride therapy by five international medical centers who have extensive experience in the use of fluoride therapy [111]. This analysis did not confirm the proposal that fluoride therapy increases the risk of spontaneous hip fractures. The results of this multicenter retrospective analysis were interpreted to indicate that, while hip fracture risk is increased in osteoporotic women because of their bone mass deficit, hip fracture risk was not increased by fluoride therapy. Two additional studies also failed to confirm the causal relationship between fluoride therapy and hip fracture risks: (1) an observational, longitudinal fluoride trial of more than

VII. CLINICAL EFFICACY OF FLUORIDE THERAPY The clinical efficacy of fluoride therapy of osteoporosis is assessed from the effect of the treatment on the symptoms, vertebral fracture rate, and bone quality and strength.

A. Symptoms The major symptomatic improvement in response to fluoride therapy is a feeling of less pain and increased strength — a feeling that the patient can perform tasks that she could not accomplish before therapy (e.g., the patient is now able to use a vacuum cleaner without getting a tired back or walk several blocks without pain). There are at least three studies reporting a significant reduction of back pain as a result of fluoride therapy [7,70,113]. This is an important benefit of any treatment of osteoporosis because it enables patients to become more active and participate in physical therapy and exercise programs, and this improves the patients’ quality of life.

B. Vertebral Fracture Rate The most important index of clinical efficacy for fluoride therapy of osteoporosis is the reduction of fracture rates. In this regard, whereas the effect of fluoride therapy to increase vertebral bone density is unambiguous, the issue of whether fluoride therapy has a beneficial effect on vertebral fracture rate is highly controversial [114,115]. This is largely because the reported efficacy of fluoride to reduce vertebral fracture rate in osteoporotic patients varied substantially from no beneficial effects to a highly significant reduction in fractures in a number of controlled and uncontrolled studies.

688

FIGURE 11

Relationship between spinal bone density during fluoride therapy and the spinal fracture rate in 510 osteoporotic patients. The spinal fracture rate decreases as a function of increasing bone density during fluoride therapy, suggesting that the bone accumulated during fluoride therapy has a positive impact on vertebral mechanical strength. Reprinted with permission from Farley et al. [121].

Results of several past uncontrolled studies have suggested that a significant reduction in the vertebral fracture rate is often associated with the increased spinal bone density in fluoride-treated osteoporotic patients [7,103,116 – 120] and that the vertebral fracture rate correlates inversely with spinal bone density [92,121]. For example, a retrospective study in which 510 osteoporotic women treated with an average daily dose of 30 mg elemental fluoride for 5 years showed a marked and progressive increase in spinal bone density [121]. The vertebral fracture rate decreased exponentially as a function of spinal bone density (Fig. 11). Moreover, the vertebral fracture rate in good responders (i.e., with a significant increase in spinal bone density) was 76% less than that in poor responders (Fig. 12).

FIGURE 12

The vertebral fracture incidence in osteoporotic patients who responded to fluoride therapy with a significant increase in spinal bone density (i.e., responders) compared to the vertebral fracture incidence of those osteoporotic patients who failed to respond to fluoride therapy (i.e., nonresponders). Reprinted with permission from Farley et al. [121].

LAU AND BAYLINK

The assumption that fluoride therapy has a beneficial effect on vertebral fracture rates in osteoporotic patients, however, was called into question by the 1990 report of the results of the randomized, placebo-controlled, double-blind study of the Mayo Clinic [64]. In that study, the overall rate of new vertebral fractures was not significantly reduced. In fact, the incidence of peripheral fractures not only was not reduced but it was significantly increased in the fluoride-treated osteoporotic patients. A similar but smaller prospective controlled trial with the same treatment protocol conducted by the Henry Ford Hospital reached the same conclusion [65]. Moreover, the recently completed prospective, randomized, double-blinded, FAVOStudy [122], in which 354 osteoporotic women with vertebral fractures treated with NaF or MFP, supplemented with 1000 mg calcium and 800 IU vitamin D2 for 2 years, also failed to show efficacy of fluoride therapy with respect to prevention of new vertebral fractures. To add to the controversy, the follow-up report of the Mayo Clinic study acknowledged that a subgroup of fluoride-treated patients (i.e., those patients with relatively low serum fluoride concentrations) in that trial had, in fact, exhibited an unequivocal reduction in vertebral fracture rate [66]. In addition, a recent 4-year randomized, controlled trial in patients with mild to moderate osteoporosis also showed that a low dose of MFP (20 mg elemental fluoride/day) plus calcium, given continuously, significantly reduced vertebral fracture rates compared with calcium alone [123]. Consistent with the possibility that the lack of efficacy on vertebral fracture rates may be related to the high fluoride dose, a recent 3year randomized, prospective study on a cyclic protocol (3 months on and 1 month off) showed that patients on a low dose of MFP (an average daily fluoride ion dose of 11.2 mg) plus calcium showed a greater efficacy in reducing vertebral fracture than those women who received a higher dose of MFP (an average daily fluoride ion dose of 20 mg) plus calcium [124]. Patients in the prospective, randomized, placebo-controlled trial of the Southwestern Medical Center, which also used an intermittent regimen of a low dose of slow-release fluoride preparation, exhibited an impressive reduction in vertebral fracture rates along with a moderate increase in spinal bone density [67,68]. In addition, a recent prospective, randomized, controlled trial [124], in which men with idiopathic osteoporosis treated for 3 years with a low dose of MFP on an intermittent protocol (3 months on fluoride, 1 month off therapy) showed a marked reduction in vertebral fracture rate along with a significant improvement of back pain. These prospective controlled studies, contrary to the three aforementioned prospective controlled trials (i.e., Mayo Clinic study, Henry Ford Hospital study, and FAVOStudy), indicate that fluoride therapy is beneficial with respect to vertebral fracture risk. These observations raise the possibility that, under optimal conditions, fluoride therapy can reduce vertebral fracture rate significantly. This view is further developed in Chapter 75.

689

CHAPTER 74 Fluoride Therapy for Osteoporosis

To define the optimal conditions that would allow beneficial effects on vertebral fracture rates, it is essential to understand the cause(s) for the lack of an effect of fluoride therapy on vertebral fracture rates in some prospective, randomized, controlled trials. Theoretical considerations of skeletal responses from fluoride therapy suggest three potential explanations. 1. FLUORIDE-INDUCED CALCIUM DEFICIENCY AND OSTEOMALACIA As pointed out earlier, fluoride-treated patients, particularly those who exhibit a rapid increase in bone density, are prone to developing calcium deficiency, secondary hyperparathyroidism, and osteomalacia. Also as discussed earlier, these side effects have two potential adverse impacts on vertebral fracture rate, despite the fluoride-mediated increase in bone density: (1) calcium deficiency and the consequent secondary hyperparathyroidism cause bone loss and could lead to a decrease in trabecular connectivity, and (2) osteomalacia could impair healing of microdamage, contributing to increased vertebral fragility. In this respect, many patients in the Mayo Clinic trial, who did not show efficacy of fluoride therapy on fracture risk, clearly developed osteomalacia, which appeared to be related to calcium deficiency [69]. Conversely, there is no evidence that patients in the Southwestern Medical Center trial, which clearly demonstrated efficacy regarding vertebral fractures, developed calcium deficiency, secondary hyperparathyroidism, and/or osteomalacia [67,68]. Therefore, it seems reasonable to speculate that fluoride-induced calcium deficiency and osteomalacia may, in part, contribute to the lack of an effect of fluoride therapy on vertebral fracture risk in some of the past trials. This emphasizes, again, the need for avoiding calcium deficiency during fluoride treatment. 2. LOSS OF TRABECULAR NUMBER AND CONNECTIVITY A decrease in the number of vertebral trabeculae and a decrease in their connectivity are characteristics of osteoporosis and may result in a relative mechanical strength deficit in osteoporotic bone (compared to normal bone), even after normal bone density has been restored (e.g., by thickening of remaining trabeculae). Therefore, the issue as to whether fluoride therapy would promote trabecular connectivity and/or restore trabecular number is highly relevant to clinical efficacy with respect to fracture risk. There is currently no compelling evidence that fluoride therapy restores trabecular number or connectivity. If fluoride therapy cannot increase trabecular number or connectivity but only increases the thickness and volume of remaining trabeculae, the increased thickness alone, without reconnecting the lost trabeculae, may not be sufficient to restore the biomechanical integrity of the skeleton. Accordingly, the

thickened and less connected trabeculae in the fluoridetreated skeleton may increase biomechanical strength compared to the osteoporotic trabeculae, but the strength per unit of bone mass would still be weaker than that of normal skeleton, resulting in only a modest improvement in fracture risk. If this hypothesis is valid, then the number and connectivity of the trabeculae present before fluoride therapy would be a significant variable in determining efficacy in reducing fracture risk. In other words, the more severe the disease, the less likely fluoride would be to reduce the fracture risk. The fact that the patients in the Southwestern Medical Center trial [67,68] appeared to have less severe osteoporosis (i.e., higher basal spinal bone density) than those in the Mayo Clinic study [64] and in the Henry Ford Hospital study [65] is consistent with this possibility. On the other hand, a retrospective study of 510 patients suggested this was not the case [121]. The patients with severe osteoporosis, who had large increases in bone density in response to fluoride therapy, seemed to fall on the same line relating final spinal bone density to fracture rate as those patients who had less severe osteoporosis but showed an increase in bone density in response to fluoride [121]. Therefore, this retrospective study seems to suggest that, even if fluoride therapy can only thicken trabeculae and thereby increase spinal bone density back to normal, the therapy may still be beneficial as it may still be able to reduce the fracture risk by partially (albeit not completely) compensating for the decrease in trabecular number and connectivity that occurs in osteoporosis. 3. FLUORIDE DEPOSITION IN BONE MINERAL Deposition of a high level of fluoride in bone can adversely affect bone quality and strength (see next section). Because patients in the Mayo Clinic study, who did not show efficacy on vertebral fracture rates, were on a significantly larger daily fluoride dose than those patients in the Southwestern Medical Center study, who exhibited significant reduction in vertebral fracture rates, and because patients in the Southwestern Medical Center trial contained a 2-month “off fluoride” period, patients in the Mayo Clinic studies received a significantly larger total accumulated fluoride dose than patients in the intermittent trial. More importantly, the slow-release regimen of the Southwestern Medical Center study, which did not produce a large postabsorption peak of serum fluoride concentration, led to a much smaller area under the serum fluoride curve than that with acute release fluoride (as used in the Mayo Clinic and Henry Ford Hospital studies) [126]. Therefore, it may be assumed that the amount of fluoride deposition in bone minerals in patients treated with low doses of intermittent slow-release fluoride would be correspondingly much lower than that found in the patients treated with high doses of continuous acute release fluoride. Figure 13 shows that the bone fluoride content in the fluoride-treated patients of

690

LAU AND BAYLINK

FIGURE 13

Relationship between total elemental fluoride taken orally in clinical trials and the corresponding subsequent bone fluoride contents. The total oral fluoride intake and bone fluoride contents are calculated from data obtained from Refs. [68,69,129], respectively. The putative “maximum safe bone fluoride level” (0.45% in bone ash was suggested by Turner et al. [127]). Reprinted with permission from Lau and Baylink [145].

the Mayo Clinic study (i.e., 0.95%) was above, whereas that in the fluoride-treated patients of the Southwestern Medical Center study was below, the putative “maximum safe level”3 (i.e., 0.45%) [127]. Thus, the differences in bone fluoride content in patients among various trials could, to some extent, account for different efficacy with respect to vertebral fracture risks. Consequently, it would be essential to reduce the amount of fluoride deposition to bone during fluoride therapy to the lowest possible level.

C. Bone Quality and Strength An important and relevant issue concerning fracture risks and, thereby, the efficacy of the therapy is the effect of fluoride treatment on bone quality and strength. Incorporation of fluoride in bone minerals leads to the substitution of fluoride for the hydroxyl group in apatite to form fluoroapatite crystals. Fluoride therapy increases the incorporation of fluoride into bone mineral crystals in both humans and animals. A survey of available data from previous human studies reveals that the amount of fluoride in bone ash is proportional to the total amount of fluoride taken by the patients (Fig. 13), indicating that the dosage and treatment duration of fluoride therapy may determine the amount of fluoride that is deposited in bone minerals. There is an abundance of evidence that deposition of large amounts of fluoride into bone minerals reduces bone quality and strength in rodents

[127,128] and humans [64,129]. Consequently, incorporation of fluoride in bone minerals, especially at high levels, may negatively affect bone quality and strength, which could be, to some extent, responsible for the lack of efficacy on fracture risks, despite large increases in bone density. It should be emphasized that, while deposition of high concentrations of fluoride in bone can have an adverse effect on bone quality which could negatively affect bone strength, the issue of the exact contribution of fluoride-associated reduction in bone quality to the lack of efficacy on fracture risks has not been definitively resolved. For example, there was an exponential decrease in the vertebral fracture rate as a function of the response to fluoride therapy (i.e., increase in spinal bone density) in 510 osteoporotic patients (Fig. 11). If the reduction in bone quality were the primary determining factor of fracture risk, it follows that there should not be an improvement in fracture risk in these 510 fluoridetreated patients. In other words, one would expect to see a plot of fracture rate that is parallel to the horizontal axis (i.e., no reduction in fracture rate) instead of an exponential decrease in fracture rate as shown in Fig. 11. Consequently, although it is reasonable to expect that bone quality would undoubtedly play a role in fracture risk, the concomitant increase in bone density may have, in part, compensated for this adverse effect. Theoretical considerations indicate that a high level of fluoride deposition in bone minerals could act through one or more of the following means to reduce bone quality: (1) increases bone material density; (2) leads to formation of bone of histologically abnormal appearance and structure; and (3) causes a mineralization defect through fluoride-induced calcium deficiency and osteomalacia. 1. INCREASES IN BONE MINERAL DENSITY Incorporation of fluoride into bone minerals improves the apatite crystallinity of bone minerals, decreases the degree of lattice distortion [130], and increases the width and thickness of the mineral crystals (i.e., larger crystals) without an effect on the crystal length [131]. These effects on bone minerals result in an increase in bone mineral density (increased calcium content per unit dry bone weight) and microhardness in cortical bones [132,133]. High bone mineral density could make bone more rigid and, thus, more brittle. Large mineral crystal size may somehow interfere with the interface between the matrix and the mineral and, thus, could affect bone mechanics. 2. PRODUCTION OF HISTOLOGICALLY ABNORMAL BONE

3 Turner et al. [127] observed in rodents and rabbits that deleterious effects of fluoride on cortical bone quality and strength are seen once the bone fluoride content reaches about 0.5% in bone ash (i.e., 5000 to 10,000 ppm). Accordingly, these investigators proposed that the bone fluoride level should be kept below 0.45% (i.e., “the maximum safe level”) to avoid harmful effects on bone quality and strength.

Woven bone has not been seen in fluoride-treated osteoporotic patients, not even in patients of the Mayo Clinic study who received relatively large doses of fluoride. Consequently, there is no evidence that woven bone with poorly organized bone matrix is a problem in fluoride-treated

691

CHAPTER 74 Fluoride Therapy for Osteoporosis

osteoporotic patients, even though bone of woven appearance has been reported in bones of patients with endemic fluorosis. On the other hand, because the fluoride ion does not diffuse in large amounts into preformed hydroxyapatite crystals in highly mineralized bone, the fluoride ion is primarily incorporated in mineral crystals of newly mineralized bone. Hence, most of the fluoride ion in the skeleton is deposited in actively mineralizing areas of the bone [134], thereby, producing distinct focal areas of high mineral density in cortical bones [130 – 133,135]. The combination of hypermineralized and hypomineralized areas produces a picture of mottled bone with low mineral density (but well-organized matrix) and hypomineralized halos around mottled lacunae [52,136,137]. Thus, fluoride treatment may lead to the production of bone with an abnormal histologic appearance. It may be speculated that increased variability of bone mineral content with areas of high and low mineral density could weaken the bone integrity and, thus, contribute to decreased quality of the fluoride-treated bone [138].

TABLE 1

Proposed Strategy to Improve the Benefit-toRisk Profile of Fluoride Therapy

Objective

Strategies

Prevention of calcium

1. Use low doses of fluoride.

deficiency

2. Regularly monitor signs of calcium deficiency. 3. Supplement patients with bioavailable calcium salt. 4. Increase intestinal calcium absorption with calcitriol or alfacalcidol.

Reducing amount of fluoride incorporation into bone minerals

1. Use doses of fluoride that produce morning predose serum fluoride levels of 5 – 10 M. 2. Use slow-release preparations. 3. Use intermittent, cyclic regimen. 4. Avoid calcium deficiency

Minimization of further loss of trabecular connectivity and number

1. Add an antiresorptive agent if bone resorption is elevated.

3. CALCIUM DEFICIENCY AND OSTEOMALACIA As discussed earlier, fluoride therapy can induce calcium deficiency and osteomalacia, which would lead to mineralization defects and an inability to heal microdamage. It is tenable that uncorrected osteomalacia could impair bone strength through at least three mechanisms. First, accumulation of poorly healed or unhealed microdamage (osteomalacia impairs healing of microdamage) within the bone could predispose to microfractures and, eventually, to complete fractures. Second, calcium deficiency can lead to secondary hyperparathyroidism. Third, the unmineralized bone may be less resistant to compressive loading, since undermineralized bone does not have the same mechanical properties as fully mineralized bone. Moreover, there is evidence that calcium deficiency may also promote deposition of fluoride into bone minerals, as there was a much greater amount of fluoride in the bone minerals of rats fed a calcium-deficient diet than rats fed a normal calcium diet [139]. As indicated earlier, high deposition of fluoride in bone minerals could greatly reduce the bone strength. Therefore, the consequences of fluoride-induced calcium deficiency and osteomalacia may produce deleterious effects on bone strength and quality.

VIII. STRATEGIES FOR IMPROVING FLUORIDE THERAPY It appears that most of the negative effects of fluoride therapy that lead to a reduced benefit-to-risk profile are caused primarily by calcium deficiency and high fluoride deposition in bone minerals. Therefore, in order to improve the

benefit-to-risk profile of fluoride therapy for osteoporosis, it would be essential to avoid or reduce the calcium deficiency and high bone fluoride deposition. Accordingly, this chapter proposes a set of strategies (Table 1) to (a) avoid calcium deficiency and (b) reduce fluoride deposition in bone minerals. Because fluoride therapy does not promote trabecular connectivity and further loss of trabecular connectivity and number is detrimental to bone strength, a strategy of using combination therapy of fluoride with an antiresorption therapy is also proposed, especially when the patient has an elevated bone resorption rate. However, it should be noted that fluoride therapy has not yet been approved for treatment of osteoporosis in most countries, including the United States. Therefore, these recommendations should not be considered as treatment guidelines, but rather, they are solely for the purpose of designing additional studies.

A. Strategy for Avoiding Calcium Deficiency The fluoride-induced increase in bone formation rate appears to be dose-dependent (Fig. 3). Accordingly, the use of lower doses of fluoride is recommended. This would produce less of an increase in bone formation rate which, in turn, would reduce the demand for calcium and should decrease the risk of calcium deficiency. Consequently, a way to avoid calcium deficiency is to use low doses of fluoride. It is also highly recommended that the fluoride-treated patients be monitored regularly (e.g., every few months) for signs of calcium deficiency, such as low urinary calcium, high serum PTH, and marked increases in bone formation biochemical markers

692

LAU AND BAYLINK

[86]. Regular monitoring for calcium deficiency should alert physicians to take early necessary precautionary interventions. It is also necessary to provide patients with sufficient amounts of calcium in a bioavailable form (e.g., calcium citrate). In older patients who may have reduced intestinal calcium absorption, it may be necessary to treat with calcitriol (0.250.5 g/day) or alfacalcidol (0.5 – 1.0 g/day) to enhance calcium absorption. These steps, when taken together, should minimize the occurrence of calcium deficiency and the associated secondary hyperparathyroidism and osteomalacia.

B. Strategy for Reducing Fluoride Deposition into Bone Minerals One of the major determinants of fluoride deposition in bone ash is the amount of fluoride intake. Thus, the proper fluoride dosage is an important issue concerning the amount of fluoride deposition. On the one hand, the patient must be given an adequate fluoride dose that would effectively increase bone density. On the other hand, the dosage must not be so high as to cause an excessive accumulation of fluoride in bone minerals. As discussed earlier, the optimal serum fluoride concentrations are restricted to between 5 and 10 M. Therefore, patients should be given an appropriate fluoride dose that yields a morning predose blood fluoride level that is maintained between 5 and 10 M for a desirable clinical outcome (i.e., an increase in spinal bone density with reduced risk for accumulation of high levels of fluoride in bone minerals. The amount of skeletal fluoride deposition is directly proportional to the circulating fluoride concentration, i.e., the larger the area under the serum fluoride curve, the more fluoride will be deposited in bone [64,67,71]. The area under the curve after administration of plain MFP is much greater than that of the slow-release MFP [126]. The use of plain fluoride salts, but not the slow-release fluoride formulations, produces large postabsorption serum fluoride peaks, which are not essential for in vivo efficacy [67,68,71] but markedly increase bone fluoride deposition. Therefore, it is advisable to use slow-release fluoride preparations to limit fluoride deposition in bone minerals. Treatment duration is also another important determinant of the amount of fluoride being deposited into bone minerals. In the rat, the bone formation effect of fluoride on the osteoblast persists for weeks even after fluoride withdrawal [140]. In humans following the discontinuation of fluoride therapy, the half-life for the disappearance of the increase in bone formation, as measured by bone formation markers, was much slower than that for the disappearance of serum fluoride [126]. These findings suggest that brief interruptions of fluoride treatment would probably not affect the overall clinical efficacy significantly, but they would reduce the exposure time of patients to fluoride and,

thereby, reduce the amount of fluoride deposited in the bone. Consequently, cyclic or intermittent regimens would considerably reduce bone fluoride deposition without diminishing the clinical efficacy of the therapy. For instance, in the Southwestern Medical Center intermittent fluoride study [67,68] in which patients were on a 14-month cyclic regimen (12 months on fluoride therapy followed by 2 months off fluoride), the patients probably would obtain the same increase in bone density as if the drug was given continuously for the entire 14 months. However, there would be significantly less (15%) fluoride incorporated in bone for the same amount of bone formation. The benefits of slow-release cyclic fluoride formulation in the treatment of osteoporosis have been convincingly demonstrated by Pak et al. [67,68] and Ringe and coworkers [124,125] and are detailed in Chapter 75. There is also evidence that treatment with intermittent slow-release NaF therapy produced much less histomorphometric abnormality in the bone [141], whereas there is clear evidence of osteomalacia in patients treated with continuous fluoride therapy. Consequently, the use of cyclic therapy instead of continuous therapy is a logical approach to avoiding a high accumulation of fluoride in bone minerals. It is important to point out that, although the regimens used by Dr. Pak and Dr. Ringe successfully produced the desirable reduction in vertebral fracture rates [67,68,124,125], there is no compelling evidence that these regimens (i.e., 14month cycle with 2-month “off fluoride period” and 4-month cycle with 1-month off fluoride, respectively) are the most appropriate. Much additional work is needed to define the “optimum” cyclic regimen. Ideally, the optimum cyclic regimen would be one that contains the shortest possible fluoride treatment period with the longest possible “off fluoride period” without sacrificing therapeutic efficacy. The optimum cyclic regimen depends on the biological half-lives of the serum fluoride level and the fluoride-induced bone formation. The halflife for serum osteocalcin (a marker of bone formation) after 6 – 12 months of fluoride therapy was 39 days, which is much longer that the half-life for decrease in serum fluoride after discontinuation of slow-release MFP ( 2 days) [71]. This raises the possibility that fluoride therapy may be stopped for 1 month or longer after 6 – 12 months of fluoride treatment, during which time the serum fluoride concentration would return to the pretreatment level but the bone formation rate would remain elevated. Nonetheless, a better understanding of appropriate rate constants between the persistence of osteogenic effects of fluoride and the duration of fluoride treatment would help optimize the cyclic regimens. Finally, as indicated earlier, calcium deficiency during fluoride therapy can also enhance fluoride deposition in bone [139]. Therefore, it is important to avoid calcium deficiency during fluoride therapy to further reduce the amount of bone fluoride deposition. If these strategies are followed, it should be possible to minimize the amount of

693

CHAPTER 74 Fluoride Therapy for Osteoporosis

fluoride deposited in bone minerals without significant losses in therapeutic efficacy.

C. Combination Therapy with an Antiresorptive Therapy The bone resorption rate in postmenopausal osteoporotic patients is often elevated above the premenopausal level. Increased bone resorption rates could increase perforations of trabeculae, which results in loss of trabecular connectivity, leading to destruction of the microarchitecture of the bone and diminishing bone strength. The ideal goal of osteoporosis therapy is to increase bone density and, at the same time, preserve trabecular architecture. Thus, if the patient were treated with fluoride to increase bone formation when she also has an elevated bone resorption rate, the fluoride therapy, which does not appear to promote trabecular reconnection or create new trabeculae, would not prevent further loss of trabecular connectivity. Thus, it seems logical to use combination therapy of fluoride with an antiresorptive agent, such as estrogens, bisphosphonates, or calcitonins, especially if the bone resorption rate of the patient is elevated. The combination therapy is expected to prevent further loss of trabecular connectivity by inhibiting bone resorption with an antiresorptive agent and, at the same time, to increase the thickness of the remaining trabeculae with fluoride. Because fluoride acts only at remodeling sites to increase wall thickness and does not promote bone formation at neutral surfaces (Fig. 4), there was a concern about whether the decrease in number of remodeling sites caused by an antiresorptive agent could impair the effect of fluoride to increase density. However, a recent study shows that combination treatment with MFP and estrogen not only did not prevent the stimulatory action of fluoride on bone formation but also produced a synergistic effect on bone mass at the spine as well as in the distal forearm and the hip [142]. Moreover, the combination of fluoride and etidronate also appears to be superior to the therapy of etidronate alone in increasing lumbar spine bone mineral mass and in preserving hip bone mass in corticosteriod-induced osteoporosis [143]. These studies provide a good rationale for combination therapy of fluoride and an anti-resorptive agent.

IX. CONCLUSION Despite intense investigations during the past three decades, there is no consensus concerning the suitability of therapy for treating established osteoporosis. Fluoride treatment has both beneficial and undesirable effects on the skeleton. On the one hand, fluoride is the only orally active

bone formation agent in clinical trials. It causes a large increase in spinal bone density, which is needed to restore bone density to a normal within a reasonable time frame. It increases bone formation at trabecular, as well as corticalendosteal, bone sites, which are sites of bone loss in osteoporosis, and it does not increase bone resorption. Moreover, fluoride acts through a unique mechanism — it inhibits a specific osteoblastic fluoride-sensitive PTP, which leads to potentiation of the osteogenic actions of endogenous bone cell growth factors. Thus, fluoride, in effect, may be considered an inexpensive, orally active, bone-specific growth factor therapy. Accordingly, fluoride therapy would be an attractive anabolic therapy for established osteoporosis. On the other hand, fluoride treatment has several undesirable side effects that significantly reduce its benefit-to-risk profile and, thereby, diminish its therapeutic value. Although most of the side effects of fluoride therapy are manageable, two of these side effects, i.e., calcium deficiency and high fluoride deposition, could have detrimental effects on the architectural integrity and mechanical strength of the bone. Consequently, fluoride therapy is a highly controversial therapy for osteoporosis [114,115]. In order for fluoride therapy to be generally accepted as a form of treatment for osteoporosis, the benefit-to-risk profile of the therapy must be improved. This chapter has proposed a set of strategies to improve the benefit-to-risk profile of fluoride therapy of osteoporosis. The primary goals of these strategies are to minimize the two major deleterious side effects of fluoride therapy: (1) calcium deficiency and the consequent secondary hyperparathyroidism and osteomalacia and (2) high levels of fluoride deposition in bone minerals. If applied appropriately, these strategies should allow the fluoride therapy to increase bone density without producing osteomalacia and without increasing the bone fluoride concentration up to a level that would impair bone quality and strength. In this regard, some of these concepts have already been incorporated into the studies of Dr. Pak [67,68] and Dr. Ringe [124,125], which have provided beneficial results with respect to the reduction in fracture rate (see Chapter 75). Consequently, further studies to optimize fluoride therapy and to further assess the proposed set of strategies are warranted. In conclusion, in light of the advantages of fluoride therapy as being an orally active, bone cell-specific, anabolic drug for bone and the apparent manageability of the side effects, fluoride therapy remains a viable anabolic treatment of established osteoporosis.

Acknowledgments This work was supported in part by research grants from the National Institutes of Health (DE13097) and the Department of Veterans Affairs.

694

LAU AND BAYLINK

References 1. W. R. Faulkner, The trace elements in laboratory medicine. Lab. Manage. July, 21 – 35 (1981). 2. W. Mertz, The essential trace elements. Science 213, 1332 – 1338 (1981). 3. N. C. Leone, et al. A roentgenologic study of a human population exposed to high fluoride domestic water (a ten-year study). Am. J. Roentgenol. 74, 874 – 885 (1955). 4. D. S. Bernstein, N. Sadowsky, D. M. Hegsted, C. D. Guri, and F. J. Stare, Prevalence of osteoporosis in high- and low-fluoride areas in North Dakota. JAMA 198, 499 – 504 (1966). 5. C. Rich, J. Ensinck, and P. Ivanovich, The effects of sodium fluoride on calcium metabolism of subjects with metabolic bone diseases. J. Clin. Invest. 43, 545 – 556 (1964). 6. J. R. Farley, J. R. Wergedal, and D. J. Baylink, Fluoride directly stimulates proliferation and alkaline phosphatase activity of bone forming cells. Science 222, 330 – 332 (1983). 7. D. Briancon and P. J. Meunier, Treatment of osteoporosis with fluoride, calcium, and vitamin D. In “The Orthopedic Clinics of North America” (H. M. Frost, ed.), Vol. 12, pp. 629 – 648. Saunders, Philadelphia, 1981. 8. B. K. Hall, Sodium fluoride as an initiator of osteogenesis from embryonic mesenchyme in vitro. Bone 8, 111 – 116 (1987). 9. J. E. Wergedal, K.-H. W. Lau, and D. J. Baylink, Fluoride and bovine bone extract influence cell proliferation and phosphatase activities in human bone cell cultures. Clin. Orthopaed. Rel. Res. 233, 274 – 282, (1988). 10. M. A. Khokher and P. Dandona, Fluoride stimulates 3H-thymidine incorporation and alkaline phosphatase production by human osteoblasts. Metabolism 39, 1118 – 1121 (1990). 11. C. G. Bellows, J. N. M. Heersche, and J. E. Aubin, The effects of fluoride on osteoblast progenitors in vitro. J. Bone Miner. Res. 5 (Suppl. 1), S101 – S105 (1990). 12. K.-H. W. Lau, A. Yoo, and S. P. Wang, Aluminum stimulates the proliferation and differentiation of osteoblasts in vitro by a mechanism that is different from fluoride. Mol. Cell. Biochem. 105, 93 – 105 (1991). 13. D. J. Simmons, P. Seitz, L. Kidder, G. L. Klein, M. Waeltz, C. M. Gundberg, C. Tabuchi, C. Yang, and R. W. Zhang, Partial characterization of rat marrow stromal cells. Calcif. Tissue Int. 48, 326 – 334 (1991). 14. D. Modrowski, L. Miravet, M. Feuga, F. Bannie, and P. J. Marie, Effect of fluoride on bone and bone cells in ovariectomized rats. J. Bone Miner. Res. 7, 961 – 969 (1992). 15. M. Kassem, L. Mosekilde, and E. F. Eriksen, Effects of fluoride on human bone cells in vitro: Differences in responsiveness between stromal osteoblast precursors and mature osteoblasts. Acta Endocrinol. 130, 381 – 386 (1994). 16. B. Y. Reed, J. E. Zerwekh, P. P. Antich, and C. Y. C. Pak, Fluoridestimulated [3H]thymidine uptake in a human osteoblastic osteosarcoma cell line is dependent on transforming growth factor . J. Bone Miner. Res. 8, 19 – 25 (1993). 17. D. Burgener, J. P. Bonjour, and J. Caverzasio, Fluoride increases tyrosine kinase activity in osteoblast-like cells: Regulatory role for the stimulation of cell proliferation and Pi transport across the plasma membrane. J. Bone Miner. Res. 10, 164 – 171 (1995). 18. M. Kassem, L. Mosekilde, and E. F. Eriksen, Effects of fluoride on human bone cells in vitro: differences in responsiveness between stromal osteoblast precursors and mature osteoblasts. Acta Endocrinol. 130, 381 – 386 (1994). 19. D. J. Baylink, Serum fluoride levels. In “Primers on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), 2nd ed., pp. 262 – 263. Raven Press, New York, 1993.

20. J. E. Harrison, K. G. McNeill, W. C. Sturtridge, T. A. Bayley, T. M. Murray, C. Williams, C. Tam, and V. Fornasier, Three-year changes in bone mineral mass of osteoporotic patients based on neutron activation analysis of the central third of the skeleton. J. Clin. Endocrinol. Metabol. 52, 751 – 758 (1981). 21. K.-H. W. Lau, J. R. Farley, T. K. Freeman, and D. J. Baylink, A proposed mechanism of the mitogenic action of fluoride on bone cells: Inhibition of the activity of an osteoblastic acid phosphatase. Metabolism 38, 858 – 868 (1989). 22. J. R. Farley, N. Tarbaux, S. Hall, and D. J. Baylink, Evidence that fluoride-stimulated 3[H]thymidine incorporation in embryonic chick calvarial cell cultures is dependent on the presence of a bone mitogen, sensitive to changes in the phosphate concentration, and modulated by systemic skeletal effectors. Metabolism 37, 988 – 995 (1988). 23. P. Ammann, R. Rizzoli, J. Caverzasio, and J.-P. Bonjour, Fluoride potentiates the osteogenic effects of IGF-I in aged ovariectomized rats. Bone 22, 39 – 43 (1998). 24. J. R. Farley, S. L. Hall, S. Herring, and M. A. Tanner, Fluoride increases net 45Ca uptake by SaSO-2 cells: The effect is phosphate dependent. Calcif. Tissue Int. 53, 187 – 192 (1993). 25. C. G. Bellows, J. N. M. Heersche, and J. E. Aubin, The effects of fluoride on osteoblast progenitors in vitro. J. Bone Miner. Res. 5 (Suppl. 1), S101 – S105 (1990). 26. J. R. Farley, N. Tarbaux, S. Hall, and D. J. Baylink, Mitogenic action(s) of fluoride on osteoblast line cells: Determinants of the response in vitro. J. Bone Miner. Res. 5 (Suppl. 1), S107 – S113 (1990). 27. A. B. Thomas, H. Hashimoto, D. J. Baylink, and K.-H. W. Lau, Fluoride at mitogenic concentrations increases the steady state phosphotyrosyl phosphorylation level of cellular proteins in human bone cells. J. Clin. Endocrinol. Metabol. 81, 2570 – 2578 (1996). 28. L.-W. Wu, H. K. Yoon, D. J. Baylink, L. M. Graves, and K.-H. W. Lau, Fluoride at mitogenic doses induces a sustained activation of p44mapk, but not p42mapk, in human TE85 osteosarcoma cells. J. Clin. Endocrinol. Metabol. 82, 1126 – 1135 (1997). 29. J. Caverzasio, T. Imai, P. Ammann, D. Burgener, and J.-P. Bonjour, Aluminum potentiates the effect of fluoride on tyrosine phosphorylation and osteoblast replication in vitro and bone mass in vivo. J. Bone Miner. Res. 11, 46 – 55 (1996). 30. T. Kawase and A. Suzuki, Studies on the transmembrane migration of fluoride and its effects on proliferation of L-929 fibroblasts (L cells) in vitro. Arch. Oral Biol. 34, 103 – 107 (1989). 31. J. Zerwekh, A. Morris, P. Padalino, F. Gottschalk, and C. Y. C. Pak, Fluoride rapidly and transiently raises intracellular Ca in human osteoblasts. J. Bone Miner. Res. 5 (Suppl. 1), S131 – S136 (1990). 32. S. G. Bourgoin, D. Harbour, and P. E. Poubelle, Role of protein kinase C, Arf, and cytoplasmic calcium transient in phospholipase D activation by sodium fluoride in osteoblast-like cells. J. Bone Miner. Res. 11, 1655 – 1665 (1996). 33. K.-H. W. Lau and D. J. Baylink, Molecular mechanism of action of fluoride on bone cells. J. Bone Miner. Res. 11, 1660 – 1667 (1998). 34. D. Antonny, J. Bigay, and M.A. Chabre, A novel magnesium-dependent mechanism for the activation of transduction by fluoride. FEBS Lett. 268, 277 – 280 (1990). 35. K.-H. W. Lau, H. Tanimoto, and D. J. Baylink, Vanadate stimulates bone cell proliferation and bone collagen synthesis in vitro. Endocrinology 123, 2858 – 2867 (1988). 36. H. K. Yoon, D. J. Baylink, and K.-H. W. Lau, Protein tyrosine kinase inhibitors block the stimulatory actions of phosphotyrosine phosphatase inhibitors to increase cell proliferation, alkaline phosphatase activity, and collagen synthesis in normal human bone cells. Am. J. Nephrol. 20, 153 – 162 (2000). 37. C. H. Kasperk, J. E. Wergedal, D. D. Strong, J. R. Farley, K. Wangerin, H. Gropp, R. Ziegler, and D. J. Baylink, Human bone cell phenotypes differ depending on their skeletal site of origin. J. Clin. Endocrinol. Metabol. 80, 2511 – 2517 (1995).

CHAPTER 74 Fluoride Therapy for Osteoporosis 38. J. Caverzasio, G. Palmer, and J. P. Bonjour, Fluoride: mode of action. Bone 22, 585 – 589 (1998). 39. J. Caverzasio, G. Palmer, A. Suzuki, and J.-P. Bonjour, Mechanism of the mitogenic effect of fluoride on osteoblast-like cells: Evidence for a G protein-dependent tyrosine phosphorylation process. J. Bone Miner. Res. 12, 1975 – 1983 (1997). 40. Y. Ericsson, Monofluorophosphate physiology: General considerations. Caries Res. 17 (Suppl. 1), 46 – 55 (1983). 41. J. Jowsey and B. L. Riggs, Effect of concurrent calcium ingestion on intestinal absorption of fluoride. Metabolism 27, 971 – 974 (1978). 42. J. R. Farley, N. M. Tarbaux, K.-H. W. Lau, and D. J. Baylink, Monofluorophosphate is hydrolyzed by alkaline phosphatase and mimics the actions of NaF on skeletal tissues in vitro. Calcif. Tissue Int. 40, 35 – 42 (1987). 43. P. D. Delmas, J. Dupuis, F. Duboeuf, M. C. Chapuy, and P. J. Meunier, Treatment of vertebral osteoporosis with disodium monofluorophosphate: Comparison with sodium fluoride. J. Bone Miner. Res. 5 (Suppl. 1), S143 – S147 (1990). 44. A. Rigalli, M. Morosano, and R. C. Puche, Bioavailability of fluoride administered as sodium fluoride or sodium monofluorophosphate to human volunteers. Arzneimittel-Forschung 46, 531 – 533 (1996). 45. P. van Asten, S. A. Duursma, J. H. Glerum, F. F. Ververs, H. J. van Rijn, and A. van Dijk, Absolute bioavailability of fluoride from disodium monofluorophosphate and enteric-coated sodium fluoride tablets. Eur. J. Clin. Pharmacol. 50, 321 – 326 (1996). 46. H. C. Hodge and F. A. Smith, Fluoride, In “Disorders of Mineral Metabolism” (F. Bronner and J. W. Coburn, eds.), Vol. 1, pp. 439 – 481. Academic Press, New York, 1981. 47. H. D. Cremer and W. Bütner, Absorption of fluorides. In “Fluorides and Human Health,” pp. 75 – 91. World Health Organization, Monograph Series, Geneva, 1970. 48. G. M. Whitford, Intake and metabolism of fluoride. Adv. Dent. Res. 8, 5 – 14 (1994). 49. G. M. Whitford, Effects of plasma fluoride and dietary calcium concentrations on GI absorption and secretion of fluoride in the rat. Calcif. Tissue Int. 54, 421 – 425 (1994). 50. A. Richards, J. Kragstrup, and F. Nielsen-Kudsk, Pharmacokinetics of chronic fluoride ingestion in growing pigs. J. Dent. Res. 64, 425 – 430 (1985). 51. S. A. Duursma, J. A. Raymakers, T. M. V. Fakkeldij, and P. van Asten, Pharmacokinetics of monofluorophosphate. Res. Clin. Forums 15, 21 – 31 (1993). 52. G. Boivin, P. Chavassieux, M. C. Chapuy, and P. J. Meunier, Skeletal fluorosis: Histomorphometric analysis of bone changes and bone fluoride content in 29 patients. Bone 10, 89 – 99 (1989). 53. G. Boivin, M. C. Chapuy, C. A. Baud, and P. J. Meunier, Fluoride content in the human iliac bone, results in controls, patients with fluorosis, and osteoporotic patients treated with fluoride. J. Bone Miner. Res. 3, 497 – 502 (1988). 54. R. D. Gabovich and G. D. Ovrutsky, Fluoride in stomatology and hygiene. “NIH Publ. 78 – 785.” National Institutes of Health, Bethesda, MD, 1977. 55. C. Y. C. Pak, Fluoride and osteoporosis. Proc. Soc. Exp. Biol. Med. 191, 278 – 286 (1989). 56. J. Ekstrand, M. Ehrnebo, and L. O. Boreus, Fluoride bioavailability after intravenous and oral administration: Importance of renal clearance and urine flow. Clin. Pharmacol. Ther. 23, 329 – 337 (1978). 57. S. A. Duursma, J. H. Glerum, A. van Dijk, R. Bosch, H. Kerkhoff, J. van Putten, and J. A. Raymakers, Responders and nonresponders after fluoride therapy in osteoporosis. Bone 8, 131 – 136 (1987). 58. S. A. Duursma, J. A. Raymakers, M. E. de Raadt, N. J. G. H. Karsdorp, A. van Dijk, and J. Glerum, Urinary fluoride excretion in responders and nonresponders after fluoride therapy in osteoporosis. J. Bone Miner. Res. 5 (Suppl. 1), S43 – S47 (1990).

695 59. M. E. Kraenzlin, C. Kraenzlin, S. M. G. Farley, R. J. Fitzsimmons, and D. J. Baylink, Fluoride pharmacokinetics in good and poor responders to fluoride therapy. J. Bone Miner. Res. 5 (Suppl. 1), S49 – S52 (1990). 60. J. Ekstrand and C.-J. Spak, Fluoride pharmacokinetics: Its implications in the fluoride treatment of osteoporosis. J. Bone Miner. Res. 5 (Suppl. 1), S53 – S61 (1990). 61. H. Spencer, L. Kramer, C. Gatza, C. Norris, E. Wiatrowski, and V. C. Gandhi, Fluoride metabolism in patients with chronic renal failure. Arch. Intern. Med. 140, 1331 – 1335 (1980). 62. B. A. Dure-Smith, M. E. Kraenzlin, S. M. Farley, C. R. Libanati, E. E. Schulz, and D. J. Baylink, Fluoride therapy for osteoporosis: A review of dose response, duration of treatment, and skeletal sits of action. Calcif. Tissue Int. 49 (Suppl.), S64 – S72 (1991). 63. D. R. Taves, New approach to the treatment of bone disease with fluoride. Fed. Proc. 29, 1185 – 1187 (1970). 64. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. Chao, H. W. Wahner, J. M. Muhs, S. L. Cedel, and L. J. Melton, III. Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). 65. M. Kleerekoper, E. L. Peterson, D. A. Nelson, E. Phillips, M. A. Schork, B. C. Tilley, and A. M. Parfitt, A randomized trial of sodium fluoride as a treatment for postmenopausal osteoporosis. Osteoporosis Int. 1, 155 – 161 (1991). 66. B. L. Riggs, W. M. O’Fallon, A. Lane, S. F. Hodgson, H. W. Wahner, J. Muhs, E. Chao, and L. J. Melton, III. Clinical trial of fluoride therapy in postmenopausal osteoporotic women: Extended observations and additional analysis. J. Bone Miner. Res. 9, 265 – 275 (1994). 67. C. Y. C. Pak, K. Sakhaee, V. Piziak, R. D. Peterson, N. A. Breslau, P. Boyd, J. R. Poindexter, J. Herzog, A. Heard-Sakhaee, S. Haynes, B. Adams-Huet, and J. S. Reisch, Slow-release sodium fluoride in the management of postmenopausal osteoporosis. A randomized controlled trial. Ann. Intern. Med. 120, 625 – 632 (1994). 68. C. Y. C. Pak, K. Sakhaee, B. Adams-Huet, V. Piziak, R. D. Peterson, and J. R. Poindexter, Treatment of postmenopausal osteoporosis with slow-release sodium fluoride. Final report of a randomized controlled trial. Ann. Intern. Med. 123, 401 – 408 (1995). 69. M. W. Lundy, M. Stauffer, J. E. Wergedal, D. J. Baylink, J. D. Featherstone, S. F. Hodgson, and B. L. Riggs, Histomorphometric analysis of iliac crest bone biopsies in placebo-treated versus fluoride-treated subjects. Osteoporosis Int. 5, 115 – 129 (1995). 70. S. M. G. Farley, J. E. Wergedal, L. C. Smith, M. W. Lundy, J. R. Farley, and D. J. Baylink, Fluoride therapy for osteoporosis: Characterization of the skeletal response by serial measurements of serum alkaline phosphatase activity. Metabolism 36, 211 – 218 (1987). 71. A. Battman, H. Resch, C. R. Libanati, D. Ludy, M. Fischer, S. Farley, and D. J. Baylink, Serum fluoride and serum osteocalcin levels in response to a novel sustained release monofluorophosphate preparation, comparison with plain monofluorophosphate. Osteoporosis Int. 7, 48 – 51 (1997). 72. P. Dandona, A. Coumar, D. S. Gill, J. Bell, and M. Thomas, Sodium fluoride stimulates osteocalcin in normal subjects. Clin. Endocrinol. 29, 437 – 441 (1988). 73. P. Charles, L. Mosekilde, L. Risteli, J. Risteli, and E. F. Eriksen, Assessment of bone remodeling using biochemical indicators of type I collagen synthesis and degradation: Relation to calcium kinetics. Bone Miner. 24, 81 – 94 (1994). 74. L. J. Ream, The effect of short-term fluoride ingestion on bone formation and resorption in the rat femur. Cell Tissue Res. 221, 421 – 430 (1981). 75. M. W. Lundy, J. R. Farley, and D. J. Baylink, Characterization of a rapidly responding animal model for fluoride-stimulated bone formation. Bone 7, 289 – 293 (1986).

696 76. P.-T. Cheng and S. M. Bader, Effects of fluoride on rat cancellous bone. Bone Miner. 11, 153 – 161 (1990). 77. P. Chavassieux, P. Pastoureau, D. Boivin, M. C. Chapuy, P. D. Delmas, and P. J. Meunier, Dose effects on ewe bone remodeling of short-term sodium fluoride administration – A histomorphometric and biochemical study. Bone 12, 421 – 427 (1991). 78. A. J. Olah, F. W. Reutter, and M. A. Dambacher, Effects of combined therapy with sodium fluoride and high doses of vitamin D in osteoporosis. A histomorphometric study in the iliac crest. In “Fluoride and Bone” (B. Courvoisier, A. Donath, and C. A. Baud, eds.), pp. 242 – 253. Bern, Huber, 1978. 79. J. E. Harrison, T. A. Bayley, R. G. Josse, T. M. Murray, W. Sturtridge, C. Williams, S. Goodwin, C. Tam, and V. Fornasier, The relationship between fluoride effects on bone histology and on bone mass in patients with postmenopausal osteoporosis. Bone Miner. 1, 321 – 333 (1986). 80. E. F. Eriksen, L. Mosekilde, and F. Melsen, Effects of sodium fluoride, calcium, phosphate, and vitamin D2 on trabecular bone balance and remodeling in osteoporosis. Bone 6, 381 – 389 (1985). 81. A. Vesterby, H. J. Gundersen, F. Melsen, and L. Mosekilde, Marrow space star volume in the iliac crest decreases in osteoporotic patients after continuous treatment with fluoride, calcium and vitamin D2 for five years. Bone 12, 33 – 37 (1991). 82. J. E. Zerwekh, H. K. Hagler, K. Sakhaee, F. Gottschalk, R. D. Peterson, and C. Y. C. Pak, Effect of slow-release sodium fluoride on cancellous bone histology and connectivity in osteoporosis. Bone 16, 691 – 699 (1994). 83. D. J. Baylink and D. S. Bernstein, The effects of fluoride therapy on metabolic bone disease. Clin. Orthop. Relat. Res. 55, 51 – 85 (1967). 84. J. E. Compston, S. Chadha, and A. L. Merrett, Osteomalacia developing during treatment of osteoporosis with sodium fluoride and vitamin D. Br. Med. J. 281, 910 – 911 (1980). 85. J. Jowsey, B. L. Riggs, P. J. Kelly, and D. L. Hoffmann, Effect of combined therapy with sodium fluoride, vitamin D and calcium in osteoporosis. Am. J. Med. 53, 43 – 49 (1972). 86. D. A. Dure-Smith, S. M. Farley, S. G. Linkhart, J. R. Farley, and D. J. Baylink, Calcium deficiency in fluoride treated osteoporotic patients despite calcium supplementation. J. Clin. Endocrinol. Metab. 81, 269 – 275 (1996). 87. M. Stauffer, D. J. Baylink, J. Wergedal, and C. Rich, Decreased bone formation and mineralization, and enhanced resorption in calcium-deficient rats. Am. J. Physiol. 225, 269 – 276 (1973). 88. W. Koo and R. Tsang, Bone mineralization in infants. Prog. Food Nutri. Sci. 8, 229 – 302, (1984). 89. S. M. G. Farley, C. R. Libanati, M. R. Mariano-Menez, L. A. Tudtud-Hans, E. E. Schulz, and D. J. Baylink, Fluoride therapy for osteoporosis promotes a progressive increase in spinal bone density. J. Bone Miner. Res. 5 (Suppl. 1), S37 – S42 (1990). 90. M. Kleerekoper and R. Balena, Fluorides and osteoporosis. Annu. Rev. Nutri. 11, 309 – 324 (1991). 91. M. Kleerekoper, Fluorides and the skeleton. Crit. Rev. Clin. Lab. Sci. 33, 139 – 161 (1996). 92. C. V. Odvina, J. E. Wergedal, C. R. Libanati, E. E. Schulz, and D. J. Baylink, Relationship between trabecular vertebral body density and fractures: A quantitative definition of spinal osteoporosis. Metabolism 37, 221 – 228 (1988). 93. R. Rizzoli, T. Chevalley, D. O. Slosman, and J.-P. Bonjour, Sodium monofluorophosphate increases vertebral bone mineral density in patients with corticosteroid-induced osteoporosis. Osteoporosis Int. 5, 39 – 46 (1995). 94. J. D. Ringe, A. Dorst, C. Kipshoven, L. C. Rovati, and I. Setnikar, Avoidance of vertebral fractures in men with idiopathic osteoporosis by a three year therapy with calcium and low-dose intermittent monofluorophosphate. Osteoporosis Int. 8, 47 – 52 (1998).

LAU AND BAYLINK 95. J. R. Talbot, M. M. Fischer, S. M. Farley, C. Libanati, J. Farley, A. Tabuenca, and D. J. Baylink, The increase in spinal bone density that occurs in response to fluoride therapy for osteoporosis is not maintained after the therapy is discontinued. Osteoporosis Int. 6, 442 – 447 (1996). 96. F. H. Budden, T. A. Bayley, J. E. Harrison, R. G. Josse, T. M. Murray, W. C. Sturtridge, R. Kandel, R. Vieth, A. L. Strauss, and S. Goodwin, The effect of fluoride on bone histology depends on adequate fluoride absorption and retention. J. Bone Miner. Res. 3, 127 – 132 (1988). 97. E. E. Schulz, H. Engstrom, D. D. Sauser, and D. J. Baylink, Osteoporosis: Radiographic detection of fluoride-induced extra-axial bone formation. Radiology 159, 457 – 462 (1986). 98. H. Resch, C. Libanati, S. Farley, P. Bettica, E. Schulz, and D. J. Baylink, Evidence that fluoride therapy increases trabecular bone density in a peripheral skeletal site. J. Clin. Endocrinol. Metab. 76, 1622 – 1624 (1993). 99. M. A. Dambacher, J. Ittner, and P. Ruegsegger, Long-term fluoride therapy of postmenopausal osteoporosis. Bone 7, 199 – 205 (1986). 100. M. Kleerekoper, and D. B. Mendlovic, Sodium fluoride therapy of postmenopausal osteoporosis. Endocr. Rev. 14, 312 – 323 (1993). 101. S. C. Rawlinson, S. Mohan, D. J. Baylink, and L. E. Lanyon, Exogenous prostacyclin but not prostaglandin E2 produces similar responses in both G6PD activity and RNA production as mechanical loading, and increases IGF-II release, in adult cancellous bone in culture. Calcif. Tissue Int. 53, 324 – 329 (1993). 102. B. L. Riggs, Treatment of osteoporosis with sodium fluoride: An appraisal. In “Bone and Mineral Research” (W. A. Peck, Ed.), pp. 366 – 393. Elsevier, Amsterdam, 1983. 103. C. Y. C. Pak, K. Sakhaee, J. E. Zerwekh, C. Parcel, R. Peterson, and K. Johnson, Safe and effective treatment of osteoporosis with intermittent slow-release sodium fluoride: Augmentation of vertebral bone mass and inhibition of fractures. J. Clin. Endocrinol. Metab. 68, 150 – 159 (1989). 104. C. M. Schnitzler, and L. Solomon, Trabecular stress fractures during fluoride therapy for osteoporosis. Skeletal Radiol. 14, 276 – 279 (1985). 105. J. D. O’Duffy, H. W. Wahner, W. M. O’Fallon, K. A. Johnson, J. M. Muhs, J. W. Beabout, S. F. Hodgson, and B. L. Riggs, Mechanism of acute lower extremity pain syndrome in fluoride-treated osteoporotic patients. Am. J. Med. 80, 561 – 566 (1986). 106. T. A. Bayley, J. E. Harrison, T. M. Murray, R. G. Josse, W. Sturtridge, K. P. Pritsker, A. Strauss, R. Vieth, and S. Goodwin, Fluoride induced fractures: Relation to osteogenic effect. J. Bone Miner. Res. 5 (Suppl. 1), S217 – S222 (1990). 107. H. E. Gruber and D. J. Baylink, The effects of fluoride on bone. Clin. Orthop. Relat. Res. 267, 264 – 277 (1991). 108. E. E. Schulz, C. R. Libanati, S. M. Farley, G. A. Kirk, and D. J. Baylink, Skeletal scintigraphic changes in osteoporosis treated with sodium fluoride: Concise communication. J. Nuclear Med. 25, 651 – 655 (1984). 109. L. R. Hedlund and J. C. Gallagher, Increased incidence of hip fracture in osteoporotic women treated with sodium fluoride. J. Bone Miner. Res. 4, 223 – 225 (1989). 110. D. H. Gutteridge, R. I. Price, G. N. Kent, R. L. Prince, and P. A. Michell, Spontaneous hip fractures in fluoride-treated patients: Potential causative factors. J. Bone Miner. Res. 5 (Suppl. 1), 205 – 212 (1990). 111. B. L. Riggs, D. J. Baylink, M. Kleerekoper, J. M. Lane, L. J. Melton, and P. J. Meunier, Incidence of hip fractures in osteoporotic women treated with sodium fluoride. J. Bone Miner. Res. 2, 123 – 126 (1987). 112. J. Farrerons, A. Rodriguez de al Serna, N. Guanabens, L. Armadans, A. Lopez-Navidad, B. Yoldi, A. Renau, and J. Vaque, Sodium fluoride treatment is a major protector against vertebral and nonvertebral fractures when compared with other common treatments of os-

CHAPTER 74 Fluoride Therapy for Osteoporosis

113. 114.

115. 116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

teoporosis: A longitudinal, observational study. Calcif. Tissue Int. 60, 250 – 254 (1997). D. S. Bernstein and P. Cohen, Use of sodium fluoride in the treatment of osteoporosis. J. Clin. Endocrinol. 27, 197 – 210 (1967). C. Y. C. Pak, J. E. Zerwekh, P. P. Antich, N. H. Bell, and F. R. Singer, Slow-release sodium fluoride in osteoporosis. J. Bone Miner. Res. 11, 561 – 564 (1996). M. Kleerekoper, Fluoride: The verdict is in, but the controversy lingers. J. Bone Miner. Res. 11, 565 – 567 (1996). B. L. Riggs, J. F. Hodgson, D. L. Hoffman, P. J. Kelly, K. A. Johnson, and D. Taves, Treatment of primary osteoporosis with fluoride and calcium: Clinical tolerance and fracture occurrence. J. Am. Med. Assoc. 243, 446 – 449 (1980). B. L. Riggs, E. Seeman, S. F. Hodgson, D. R. Taves, and W. M. O’Fallon, Effect of fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporosis: Comparison with conventional therapy. N. Engl. J. Med. 306, 446 – 450 (1982). J. M. Lane, J. H. Healey, E. Schwartz, V. J. Vigorita, R. Schneider, T. A. Einhorn, M. Suda, and W. C. Robbins, Treatment of osteoporosis with sodium fluoride and calcium: Effects on vertebral fracture incidence and bone histomorphometry. Orthop. Clin. North Am. 15, 728 – 745 (1984). N. Mamelle, P. J. Meunier, R. Dusan, M. Guillaume, J. L. Martin, A. Gaucher, A. Prost, G. Zeigler, and P. Netter, Risk benefit of sodium fluoride treatment in primary vertebral osteoporosis. Lancet 2, 361 – 365 (1988). R. P. Heaney, D. J. Baylink, C. C. Johnson, L. J. Melton, III, P. J. Meunier, T. M. Murrary, and C. Nagant de Deuxchaisnes, Fluoride therapy for the vertebral crush fracture syndrome (a status report). Ann. Intern. Med. 111, 687 – 680 (1989). S. M. Farley, J. E. Wergedal, J. R. Farley, G. N. Javier, E. E. Schulz, J. R. Talbot, C. R. Libanati, L. Lindegren, M. Bock, M. M. Goette, S. S. Mohan, P. Kimball-Johnson, V. S. Perkel, R. J. Cruise, and D. J. Baylink, Spinal fractures during fluoride therapy for osteoporosis: relationship to spinal bone density. Osteoporosis Int. 2, 213 – 218 (1992). P. J. Meunier, J.-L. Sebert, J.-Y. Reginster, D. Briancon, T. Appelboom, P. Netter, G. Loeb, A. Rouillon, S. Barry, J.-C. Evreux, B. Avouac, X. Marchandise, and the FAVOStudy Group. Fluoride salts are no better at preventing new vertebral fractures than calcium–vitamin D in postmenopausal osteoporosis: The FAVOStudy. Osteoporosis Int. 8, 4 – 12 (1998). J. Y. Reginster, L. Meurmans, B. Zegels, L. C. Rovati, H. W. Minne, G. Giacovelli, A. N. Taquet, I. Setnikar, J. Collette, and C. Gosset, The effect of sodium monofluorophosphate plus calcium on vertebral fracture rate in postmenopausal women with moderate osteoporosis. A randomized, controlled trial. Ann. Intern. Med. 129, 1 – 8 (1998). J. D. Ringe, C. Kipshoven, A. Coster, R. Umbach, Therapy of established postmenopausal osteoporosis with monofluorophosphate plus calcium: Dose-related effects on bone density and fracture rate. Osteoporosis Int. 9, 171 – 178 (1999). J. D. Ringe, A. Dorst, C. Kipshoven, L. C. Rovati, and I. Setnikar, Avoidance of vertebral fractures in men with idiopathic osteoporosis by a three year therapy with calcium and low-dose intermittent monofluorophate. Osteoporosis Int. 8, 47 – 52 (1998). H. Resch, C. Libanati, J. Talbot, M. Tabuenca, S. Farley, P. Bettica, W. Tritthart, and D. J. Baylink, Pharmacokinetic profile of a new fluoride preparation: sustained-release monofluorophosphate. Calcif. Tissue Int. 54, 7 – 11 (1994). C. H. Turner, K. Hasegawa, W. Zhang, M. Wilson, Y. Li, and A. J. Dunipace, Fluoride reduces bone strength in older rats. J. Dent. Res. 78, 1475 – 1481 (1995).

697 128. C. H. Turner, M. P. Akhter, and R. P. Heaney, The effects of fluoridated water on bone strength. J. Orthop. Res. 10, 581 – 587 (1992). 129. C. H. Sögaard, L. Mosekilde, A. Richards, and L. Mosekilde, Marked decrease in trabecular bone quality after five years of sodium fluoride therapy — assessed by biomechanical testing of iliac crest bone biopsies in osteoporotic patients. Bone 15, 393 – 399 (1994). 130. E. D. Eanes and A. H. Reddi, The effect of fluoride on bone mineral apatite. Metab. Bone Dis. 2, 3 – 11 (1979). 131. L. Singer, W. D. Armstrong, I. Zipkin, and P. D. Frazier, Chemical composition and structure of fluorotic bone. Clin. Orthop. Rel. Res. 99, 303 – 312 (1974). 132. K. Yamamoto, J. E. Wergedal, and D. J. Baylink, Increased bone microhardness in fluoride treated rats. Calcif. Tissue Res. 15, 45 – 54 (1974). 133. J. Franke, H. Runge, P. Grau, F. Fengler, C. Wanka, and H. Rempel, Physical properties of fluorosis bone. Acta Orthop. Scand. 47, 20 – 27 (1976). 134. M. d. Grynpas, Fluoride effects on bone crystals. J. Bone Miner. Res. 5 (Suppl. 1), S169 – S175 (1990). 135. M. W. Lundy, J. E. Russell, J. Avery, J. E. Wergedal, and D. J. Baylink, Effect of sodium fluoride on bone density in chickens. Calcif. Tissue Int. 50, 420 – 426 (1992). 136. B. H. Wiers, M. D. Francis, K. Hovanick, C. K. Ritchie, and D. J. Baylink, Theoretical physical chemical studies of the cause of fluoride-induced osteomalacia. J. Bone Miner. Res. 5 (Suppl. 1), S63 – S70 (1990). 137. D. J. Baylink and D. S. Bernstein, The effects of fluoride therapy on metabolic bone disease. Clin. Orthop. Rel. Res. 55, 51 – 85 (1967). 138. D. R. Carter and G. S. Beaupre, Effects of fluoride treatment on bone strength. J. Bone Miner. Res. 5 (Suppl. 1), S177–S184 (1990). 139. D. Beary, The effects of fluoride and low Ca on the physical properties of the rat femur. Anat. Rec. 164, 305 – 316 (1969). 140. P. Chavassieux, G. Boivin, C. M. Serre, and P. J. Meunier, Fluoride increases rat osteoblast function and population after in vivo administration but not after in vitro exposure. Bone 14, 721 – 725 (1993). 141. C. M. Schnitzler, J. R. Wing, F. J. Raal, M. T. van der Merwe, J. M. Mesquita, K. A. Gear, H. J. Robson, and R. Shires, Fewer bone histomorphometric abnormalities with intermittent than with continuous slow-release sodium fluoride therapy. Osteoporosis Int. 7, 376 – 389 (1997). 142. P. Alexandersen, B. J. Riis, and C. Christiansen, Monofluorophosphate combined with hormone replacement therapy induces a synergistic effect on bone mass by dissociating bone formation and resorption in postmenopausal women: A randomized study. J. Clin. Endocrinol. Metab. 84, 3013 – 3020 (1999). 143. W. F. Lems, J. W. Jacobs, J. W. Bijlsma, G. J. van Veen, H. H. Houben, H. C. Haanen, M. I. Gerrits, H. J. van Rijn, Is addition of sodium fluoride to cyclical etidronate beneficial in the treatment of corticosteroid induced osteoporosis? Ann. Rheum. Dis. 56, 357 – 363 (1997). 144. K.-H. W. Lau et al., Osteogenic actions of fluoride: Its therapeutic use for established osteoporosis. In “Anabolic Treatments for Osteoporosis” (J. F. Whitfield and P. Morley, eds.), pp. 207 – 254. CRE Press, Boca Raton, 12, 1988. 145. K.-H. W. Lau and D. J. Baylink, Fluoride therapy of established osteoporosis. In “The Aging Skeleton” (C. J. Rosen, J. Glowacki, and J. P. Bilezikian, eds.), Chap. 48, pp. 587 – 612. Academic Press, San Diego, 1999.

CHAPTER 75

A New Perspective on Fluoride Therapy CHARLES Y. C. PAK

I. II. III. IV.

Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center, Dallas, Texas 75390

Introduction Pharmacokinetic Studies Long-Term Randomized Clinical Trials Analysis of Spinal Fracture Data from 12 Trials According to Relative Fluoride Absorption

V. Analysis of Appendicular Fracture Data from 12 Trials According to Relative Fluoride Absorption VI. Discussion References

I. INTRODUCTION

sponse. Moreover, an excessive release of fluoride in the stomach may cause gastric bleeding by forming corrosive hydrofluoric acid. On the other hand, a low fluoride exposure or dose may stimulate formation of normally mineralized bone [8], reducing spinal fractures without causing microfractures or appendicular fractures. A limited release of fluoride in the stomach may explain improved gastrointestinal tolerance. In this chapter, the confirmation of the above scheme was sought from a careful review of pharmacokinetic studies and randomized clinical trials. Chapter 74 presents a complete review of fluoride therapy.

Fluoride has been used for the treatment of osteoporosis for several decades. This use is based on the well-known ability of fluoride to stimulate osteoblastic bone formation and to increase spinal bone mass [1]. Unfortunately, the results of fluoride treatment have been conflicting. Some studies have reported no effect on spinal fractures despite a marked increase in spinal bone mass, and exaggeration of microfractures and appendicular fractures [2]. Moreover, gastric ulceration was found to complicate fluoride treatment [3]. Several recent trials, however, have indicated that fluoride treatment could substantially reduce the rate of spinal fractures, without the complications cited above [4,5]. There is increased awareness that the conflicting clinical response to fluoride treatment can be ascribed to the wellknown biphasic action of fluoride. Following excessive exposure or dose of fluoride, fluoride-induced synthesis of bone matrix may outstrip the available calcium. Thus, the bone formed may be poorly mineralized and mechanically incompetent [6,7], accounting for the poor fracture re-

OSTEOPOROSIS, SECOND EDITION VOLUME 2

II. PHARMACOKINETIC STUDIES A. Therapeutic Window for Serum Fluoride The main factor limiting the clinical use of fluoride has been commonly attributed to a narrow “therapeutic window.” Thus, the therapeutic threshold for serum fluoride of 95 ng/ml is not much lower than the toxic threshold of 190

699

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

700

CHARLES Y. C. PAK

TABLE 1 Study no.

Drug daily dose

Comparison of Peak-Basal Variation and AUC F per dose (mg)

Peak-basal (ng/ml)

AUC per dose (ng.h/ml)

Reference

1

MFP 100 mg

13.2

260

1711

Liote et al. [21]

2

MFP 100 mg

13.2

303

1363

Meunier [20]

3

MFP 76 mg

10.0

181

1119

Resch et al. [22]

4

MFP 76 mg

10.0

122

1060

Warneke and Setnikar [23]

5

MFP 76 mg

10.0

369

1150

Warneke and Setnikar [23]

6

MFP 76 mg

10.0

398

1452

Gitomer et al. [19]

7

EC-NaF 25 mg

11.3

200

1202

Liote et al. [21]

8

EC-NaF 25 mg

11.3

229b

1620b

Briancon et al. [24]

a

729

Nagant et al. [25]

730

Meunier [20]

9

EC-NaF 25 mg

11.3

168

10

EC-NaF 25 mg

11.3

135

a

11

EC-NaF 25 mg

11.3

143

893

Devogelaer et al. [26]

12

Plain NaF 30 mg

13.6

185

1042c

Sakhaee and Pak [10]

a

a,c

13

Plain NaF 30 mg

13.6

380

14

Plain NaF 25 mg

11.3

122

1331

999c

Nagant et al. [25] Pak et al. [27]

15

Plain NaF 25 mg

11.3

370

1562

Gitomer et al. [19]

16

SR-NaF 25 mg

11.3

40

343

Sakhaee and Pak [10]

17

SR-NaF 25 mg

11.3

59

315

Meunier [20]

18

SR-NaF 25 mg

11.3

48

375

Pak et al. [27]

19

SR-NaF 25 mg

11.3

80

466

Gitomer et al. [19]

Note. Values were obtained from group means, unless provided in the text. a Calculated by assuming basal serum fluoride to be 24 ng/ml as in the study by Meunier [20]. b Calculated by taking basal serum fluoride value of 30 ng/ml (from the figure in text). c For estimation of skeletal fluoride, AUC was corrected for dose schedule by using a correction factor ( 75/60).

ng/ml. This range was first suggested by Taves [9], from experimental animal studies in which stimulation of new bone formation for the lower range and appearance of abnormal bone for the upper range were detected. Clinical confirmation of this concept has been difficult, because many trials did not consider the circadian fluctuation of serum fluoride after multiple dosing during long-term treatment. By reporting only the trough serum fluoride value, for example, the attainment of toxic concentration by the peak fluoride value could have been missed [2]. Recent trials suggest that the limits for therapeutic window identified above are probably valid. In trials where both trough and peak values for serum fluoride were kept largely within 95 – 190 ng/ml, a positive spinal fracture outcome with stimulation of normal bone formation has been described [4]. In contrast, the trials where peak serum fluoride concentration considerably exceeded the toxic threshold [10] have yielded negative clinical response.

B. Toxic Skeletal Threshold The toxic threshold for skeletal fluoride is believed to be 0.6 – 0.8% bone ash [11,12] based on the appearance of flu-

orotic changes. The toxic threshold values for bone and serum fluoride are probably linked. Thus, the skeletal uptake of fluoride has been shown to be directly correlated with the pharmacokinetically measured fluoride absorption from the gastrointestinal tract and with the peak serum fluoride concentration [12]. Treatment with fluoride preparations of high bioavailability has been associated with toxic fluoride concentration in serum and accumulation in bone and with abnormal bone formation [7].

C. Importance of Single-Dose Bioavailability In various clinical trials, fluoride preparations varied considerably with respect to formulation, in vitro release, and dosage. Thus, the amount of fluoride absorbed from the gastrointestinal tract, the peak concentration attained in circulation, and the quantity accumulating in bone must have varied extensively. The above pharmacokinetic differences may be quantitated by a “single-dose bioavailability” study. From the changes in serum fluoride concentration after a single oral dose of a fluoride preparation, the peak fluoride concentration in serum (Cmax) and fluoride absorption (change in

CHAPTER 75

701

A New Perspective on Fluoride Therapy

FIGURE 1 Estimated peak serum fluoride concentration during long-term treatment with various fluoride preparations. Peak fluoride concentration after multiple dosing was estimated by adding peak-basal variation in serum fluoride from a single dose bioavailability study with the basal value, assumed to be 130 ng /ml. The upper dashed horizontal line is the toxic threshold. The shaded bars above the upper dashed line show the degree by which the toxic threshold is exceeded. The figure was drawn from 19 single-dose pharmacokinetic studies. MFP, monofluorophosphate; EC-NaF, enteric-coated sodium fluoride; SR-NaF, sustained-release sodium fluoride. area-under-the-curve or AUC) can be calculated. From these pharmacokinetic parameters, the peak serum fluoride concentration and skeletal fluoride content after multiple dosing can be estimated.

D. Review of Pharmacokinetic Studies Ten articles on single-dose bioavailability of various fluoride preparations were reviewed (Table 1). They contained 19 studies, comprosed of 6 studies on monofluorophosphate (MFP), 5 on enteric-coated sodium fluoride (NaF), 4 on plain (rapid-release) NaF, and 4 on sustained-release NaF (SR-NaF or Neosten). Fluoride load ranged from 10.0 to 13.6 mg. 1. ESTIMATED PEAK SERUM FLUORIDE CONCENTRATION For SR-NaF, the peak-trough variation of serum fluoride ranged from 40 to 80 ng/ml among four studies, with a mean of 57 ng/ml (Table 1). Much higher values were reported with enteric-coated NaF, plain NaF, and MFP. After long-term treatment with SR-NaF, the fasting serum fluoride concentration in the morning (with the last dose given at bedtime the previous night) was about 130 ng/ml [4]. Assuming this figure to be the baseline value after multiple dosing for all studies, the peak serum fluoride concentration during long-term treatment was calculated by

adding 130 ng/ml to the peak fluoride value obtained after a single dose. For MFP, plain NaF, and enteric-coated NaF, the estimated peak serum fluoride concentration so-derived considerably exceeded the toxic threshold of 190 ng/ml (Fig. 1). However, the value for SR-NaF resided within the therapeutic window or only slightly exceeded the toxic threshold. 2. ESTIMATED SKELETAL FLUORIDE CONTENT 4 YEARS OF TREATMENT

AFTER

For SR-NaF, (AUC of serum fluoride ranged from 315 to 466 ng.h/ml among the four studies, with a mean of 375 ng.h/ml (Table 1). Much higher values were disclosed with other fluoride preparations. Skeletal fluoride content is well known to be directly dependent on fluoride absorption. In two studies, bone fluoride content was directly measured from bone biopsy specimens. After 4 years of treatment with plain NaF at a dose of 30 mg twice daily alternating with thrice daily (corresponding to Study 12; Riggs [2]), bone fluoride content was 1.04% bone ash (or 55 mol/mmol calcium) [7]. Bone fluoride was 0.28% bone ash after 4 years of treatment with SR-NaF at a dose of 25 mg twice daily (Study 16) [12]. The corresponding AUC was 1042 ng.h/ml per dose of plain NaF (30 mg containing 13.6 mg F) and 343 ng.h/ml per dose of SR-NaF (25 mg NaF containing 11.3 mg F). Corrected AUC for plain NaF was 1303 ng.h/ml, assuming that the drug was constantly given twice daily. From the relationship of

702

CHARLES Y. C. PAK

Estimation of skeletal fluoride from AUCs. Shaded region between 0.6 to 0.7% bone ash represents the toxic threshold. Closed symbols indicate directly measured bone fluoride for studies 12 (for plain NaF) and 16 (SR-NaF). Open symbols depict estimated skeletal fluoride from the regression line. Numbers next to symbols correspond to study numbers from Table 1.

FIGURE 2

directly measured bone fluoride content against the corresponding AUC, the line connecting the two points extrapolated close to the value of zero at the origin. From this regression line, bone fluoride content after 4 years of treatment could be estimated from the corresponding AUC for remaining 17 trials (Fig. 2). Since a twice daily schedule was used, no correction for dose schedule was necessary. The estimated bone fluoride considerably exceeded the toxic threshold for plain NaF and MFP and approached or mild-moderately exceeded the toxic threshold for enteric-coated NaF. However, the estimated bone fluoride was well below the toxic threshold for SR-NaF.

III. LONG-TERM RANDOMIZED CLINICAL TRIALS A. Randomized Clinical Trials with Fluoride There have been nine reports of randomized clinical trials with fluoride, in which spinal and appendicular fracture rates were compared between fluoride and control groups (Table 2). 1. DESCRIPTION OF NINE REPORTS In Report 1, Riggs et al. [2] compared plain NaF at an average dose of 33.9 mg F/day given continuously with placebo in 202 women with established postmenopausal osteoporosis (with prevalent spinal fractures). In Report 2,

Kleerekoper et al. [13] compared plain NaF (33.9 mg F/day given continuously) with placebo in 75 women with established postmenopausal osteoporosis. In Report 3, Meunier et al. [14] compared two dosages of MFP (22.6 and 26.4 mg F/day) and enteric-coated NaF (22.6 mg F/day) given continuously with placebo in 316 women with established postmenopausal osteoporosis. In Report 4, Mamelle et al. [14] compared enteric-coated NaF (22.6 mg F/day) given continuously with conventional fluoride treatment in 316 women and men with established primary vertebral osteoporosis. In Report 5, Reginster et al. [16] compared MFP (20 mg F/day) given continuously with placebo in 164 women with postmenopausal osteoporosis without prevalent fractures diagnosed by density criteria. In Report 6, Pak et al. [4] compared SR-NaF (22.6 mg F/day) given intermittently (12 months on, 2 months off in each cycle) with placebo in 99 women with established postmenopausal osteoporosis. In Report 7, Rubin et al. [17] compared SR-NaF (22.6 mg F/day) given intermittently with placebo in 73 elderly women with established postmenopausal osteoporosis. In Report 8, Ringe et al. [18] compared MFP (15 mg F/day) given intermittently (3 months on, 1 month off) with placebo in 60 men with idiopathic osteoporosis (without prevalent vertebral fractures). In Report 9, Ringe et al. [5] compared two dosages of MFP (15 mg F/day given intermittently and 20 mg F/day given continuously) with placebo in 116 women with established postmenopausal osteoporosis.

CHAPTER 75

703

A New Perspective on Fluoride Therapy

TABLE 2 Report no., author [ref]

Trial no.

Features of Randomized Trials with Fluoride Fluoride preparation

Daily F dose (mg/day)

Control group

Diagnosis

1. Riggs [2]

1

Plain NaF

33.9 mg F continuous

Placebo

Postmenopausal osteoporosis

2. Kleerekoper [13]

2

Plain NaF

33.9 mg F continuous

Placebo

Postmenopausal osteoporosis

3. Meunier [14]

3

MFP

22.6 mg F continuous

Placebo

Postmenopausal osteoporosis

3. Meunier [14]

4

MFP

26.4 mg F continuous

Placebo

Postmenopausal osteoporosis

3. Meunier [14]

5

Enteric-coated NaF

22.6 mg F continuous

Placebo

Postmenopausal osteoporosis

4. Mamelle [15]

6

Enteric-coated NaF

22.6 mg F continuous

Conventional treatment

Primary osteoporosis

5. Reginster [16]

7

MFP

20 mg F continuous

Placebo

Postmenopausal osteoporosis

6. Pak [12]

8

Sustained-release NaF

22.6 mg F intermittent

Placebo

Postmenopausal osteoporosis

7. Rubin [17]

9

Sustained-release NaF

22.6 mg F intermittent

Placebo

Postmenopausal osteoporosis

8. Ringe [18]

10

MFP

15 mg F intermittent

Placebo

Idiopathic (male) osteoporosis

9. Ringe [5]

11

MFP

15 mg F intermittent

Placebo

Postmenopausal osteoporosis

9. Ringe [5]

12

MFP

20 mg F continuous

Placebo

Postmenopausal osteoporosis

Note. MFP, monofluorophosphate.

2. DESCRIPTION OF 12 TRIALS One report (Report 3) compared MFP at two doses and enteric-coated NaF at one dose with placebo (Table 2). Another report (Report 9) compared MFP at two doses with placebo. The remaining 7 reports compared one dose of a fluoride preparation with placebo or a control group. Thus, 9 reports contained 12 trials, which compared a fluoride preparation with a placebo or control group (second column, Table 2). Plain NaF was used at an average dose of 75 mg (33.9 mg F)/day in two trials (Trials 1 and 2), enteric-coated NaF at a dose of 50 mg (22.6 mg F)/day in two trials (Trials 5 and 6), and SR-NaF at a dose of 50 mg (22.6 mg F)/day in two trials (Trials 8 and 9) (Table 2). MFP was given at a dose 26.4 mg F/day in Trial 4, 22.6 mg F/day in Trial 3, 20 mg F/day in Trials 7 and 12, and 15 mg/day in Trials 10 and 11. Fluoride was given intermittently in Trials 8 – 11 and continuously in the remaining trials. The control group was composed of patients taking placebo in all trials except in Trial 6, in which it represented conventional nonfluoride treatment. In Trial 7, patients without prevalent spinal fractures had postmenopausal osteoporosis on the basis of T scores of less than 2.5 for spinal

bone density. Trial 10 comprised men with idiopathic osteoporosis, and Trial 6 comprised both women and men with primary spinal osteoporosis. In the remaining 9 trials, study subjects were those with postmenopausal osteoporosis with prevalent spinal fractures.

B. Categorization of Trials into Varying Fluoride Exposure In the analysis of pharmacokinetic data presented previously (Table 1), different fluoride formulations could generally be distinguished from each other on the basis of estimated bone fluoride (Fig. 2). However, there was a considerable variation in estimated skeletal fluoride among different studies employing the same fluoride formulation. This variation was in part due to the differences in study subjects and fluoride loads among various pharmacokinetic studies. In the categorization of 12 randomized clinical trials according to fluoride exposure, we therefore decided to take two single-dose bioavailability studies in which three fluoride formulations had been tested in the same

704

CHARLES Y. C. PAK

subjects. In our study of 12 normal subjects [19], SR-NaF was compared with plain NaF and MFP in a cross-over design. In another study of 12 normal women [20], SRNaF was compared with enteric-coated NaF in the same subjects. 1. CALCULATION OF RELATIVE FLUORIDE ABSORPTION In the study of Gitomer et al. [19], AUC for each milligram of fluoride dose was 138 ng . h/ml for plain sodium fluoride, 145 ng . h/ml for plain NaF, and 41 ng . h/ml for SR-NaF. In the study of Meunier [20], AUC for entericcoated NaF per milligram of fluoride dose was found to be 2.3-fold greater than that of SR-NaF. Thus, a value of 95 ng . h/ml was assigned to enteric-coated NaF. Daily fluoride dose varied from 15.0 to 33.9 mg among the 12 trials (Table 2). Thus, AUC per daily dose of fluoride was calculated by multiplying the value per milligrams F dose (Table 3) by the daily fluoride dose in milligrams (Table 2). Fluoride was given continuously in some trials and intermittently in other trials. For example, SR-NaF was administered in a repeated cycle of 14 months (12 months on, 2 months off) [4]. Thus, AUC was normalized to 14 months of treatment by multiplying the value for daily dose by the actual number of days fluoride was administered over a 14-month period (column 4, Table 3). Relative fluoride absorption was calculated by assigning a value of 1 to SR-NaF, and dividing the 14-month AUC values for other fluoride preparations by 0.34, the value for SR-NaF (Table 3). Relative fluoride absorption for a fluoride treatment therefore represented the number of times its fluoride absorption exceeded that of SR-NaF.

TABLE 3 Trial no., author [ref], fluoride preparation

OF

2. CATEGORIZATION OF TRIALS INTO THREE GROUPS FLUORIDE EXPOSURE

From relative fluoride absorption, various trials were categorized into three groups of varying fluoride exposure. We defined relative fluoride absorption of 4 – 6 as “high fluoride exposure,” a value of 2.5 – 4 as “moderate fluoride exposure,” and value of 1 – 2.5 as “low fluoride exposure.” According to the above definition, Trials 1 – 4 fell into high fluoride exposure (Tables 2 and 3). Trials 1 and 2 utilized optimally bioavailable plain NaF at a high dose of 33.9 mg F/day continuously. Compared with Trials 8 and 9, which used SR-NaF, these trials delivered about six times as much fluoride. Bone biopsy examination after 4 years of treatment had disclosed toxic fluoride accumulation and impaired mineralization [7]. Trials 3 and 4 used equally bioavailable MFP [19], at two doses continuously, that delivered four to five times as much fluoride as SR-NaF. Trials 5 – 7 and 12 were categorized as moderate fluoride exposure (Tables 2 and 3). Trials 5 and 6 utilized moderately bioavailable enteric-coated NaF at the usual dose of 22.6 mg F/day continuously, which delivered 2.7 times as much fluoride as SR-NaF. Trials 7 and 12 used highly bioavailable MFP at a lower dose of 20 mg F/day continuously. Estimated fluoride delivered was 3.6 times that of SR-NaF. Trials 8 – 11 were considered to be low fluoride exposure (Tables 2 and 3). Trials 8 and 9 were two randomized trials with SR-NaF 22.6 mg F/day given for 12 months followed by 2 months of withdrawal in each 14-month cycle. Bone biopsy specimens obtained after up to 4 years of this treatment disclosed structurally normal appearance with subtoxic fluoride content [8]. Trials 10 and 11 utilized MFP at a low dose of 15 mg F/day intermittently (3 months on, 1 month

Relative Fluoride Absorption among 12 Trials

AUC per mg F (ng . h/ml)

AUC per day (g . h/ml)

AUC per 14 mo (mg . h/ml)

Relative fluoride absorption 5.85

1. Riggs [2], plain NaF

138

4.68

1.99

2. Kleerekoper [13], plain NaF

138

4.68

1.99

5.85

3. Meunier [14], MFP

145

3.28

1.39

4.08

4. Meunier [14], MFP

145

3.83

1.63

4.79

5. Meunier [14], EC-NaF

95

2.15

0.91

2.68

6. Mamelle [15],EC-NaF

95

2.15

0.91

2.68

145

2.90

1.23

3.62

7. Reginster [16], MFP 8. Pak [12], SR-NaF

41

0.93

0.34

1

9. Rubin [17], SR-NaF

41

0.93

0.34

1

10. Ringe [18], MFP

145

2.18

0.73

2.14

11. Ringe [5], MFP

145

2.18

0.73

2.14

12. Ringe [5], MFP

145

2.90

1.23

3.62

Note. MFP, monofluorophosphate; EC-NaF, enteric-coated sodium fluoride; SR-NaF, sustained-release sodium fluoride; AUC, delta area-under-the-curve.

CHAPTER 75

705

A New Perspective on Fluoride Therapy

off every 4 months). Estimated fluoride absorbed over 14 months was about twofold greater than with SR-NaF.

IV. ANALYSIS OF SPINAL FRACTURE DATA FROM 12 TRIALS ACCORDING TO RELATIVE FLUORIDE ABSORPTION It was not meaningful to compare actual spinal fracture rates among fluoride-treated patients between various trials, because of differences in methods of fracture detection, study population (postmenopausal vs idiopathic osteoporosis), and severity of osteoporosis (with or without prevalent fractures). We therefore relied on the comparison of fracture rates between the fluoride group and the control group (placebo or conventional treatment) in each trial, expressed as the relative risk and 95% confidence intervals. In high fluoride exposure (Trials 1 – 4), the difference in relapse rates (percentage of patients with incident fractures) between the fluoride and placebo groups was not significant (Table 4). The relative risk (fluoride/placebo) ranged from 0.82 to 1.11. In moderate fluoride exposure, two trials involving patients with established primary or postmenopausal osteoporosis with prevalent spinal fractures showed no significant effect of fluoride in inhibiting spinal fractures with a relative risk of 1.17 (Trial 5), or a moderate significant effect with a relative risk of 0.78 (Trial 6) (Table 5). The remaining two trials (Trials 7 and 12) evaluated patients with milder form of osteoporosis without prevalent fractures. Fluoride caused a dramatic significant TABLE 4

reduction in spinal fracture rate, with a relative risk of 0.24 and 0.44. In low fluoride exposure, relative risk was 0.31 and 0.37 in Trials 8 and 9 with intermittent SR-NaF among patients with established postmenopausal osteoporosis (Table 4). Thus, fluoride significantly reduced spinal fracture rate by 63 – 69%. In Trial 10 involving patients with osteoporosis by density criteria (without prevalent spinal fractures), intermittent MFP produced a significant 75% reduction in spinal fracture rate, with a relative risk of 0.25. In Trial 11, comprising patients with established osteoporosis, intermittent MFP was also found to be effective in inhibiting spinal fractures, with a relative risk of 0.29. The relative risk (fluoride/control-placebo) of relapse rate of spinal fractures was directly and significantly correlated with relative fluoride absorption (Fig. 3). The relative risk progressively declined, from a value of about 1 with high fluoride exposure, to a value much less than 1 with low fluoride exposure.

V. ANALYSIS OF APPENDICULAR FRACTURE DATA FROM 12 TRIALS ACCORDING TO RELATIVE FLUORIDE ABSORPTION A. Hip Fracture Rates Hip fracture data were not available in two trials (Trials 2 and 10) (Table 5). In high fluoride exposure, Trials 1 and 4 with the highest relative fluoride absorption displayed a

Summary of Spinal Fracture Data: Relapse Rate or % Patients with Incident Fractures Established osteoporosis

Relative risk

95% Confidence interval

P value

Yes

0.82

0.55 – 1.23

0.2

High Fluoride Exposure Trial 1, plain NaF Trial 2, plain NaF

Yes

1.11

0.82 – 1.50

0.2

Trial 3, MFP

Yes

0.97

0.70 – 1.33

0.2

Trial 4, MFP

Yes

1.03

0.76 – 1.40

0.2

Trial 5, EC-NaF

Yes

1.17

0.88 – 1.54

0.2

Trial 6, EC-NaF

Yes

0.78

0.61 – 0.99

0.05

Trial 7, MFP

No

0.24

0.05 – 1.09

0.05

Trial 12, MFP

No

0.44

0.27 – 0.73

0.0006

Trial 8, SR-NaF

Yes

0.31

0.15 – 0.65

0.0005

Trial 9, SR-NaF

Yes

0.37

0.16 – 0.83

0.02

Trial 10, MFP

No

0.25

0.08 – 0.80

0.02

Trial 11, MFP

Yes

0.29

0.15 – 0.55

0.0001

Moderate Fluoride Exposure

Low Fluoride Exposure

Note. For Trial 9, fracture data were analyzed by multivariate Cox model. In other trials, Fisher’s exact test was used.

706

CHARLES Y. C. PAK

TABLE 5

Summary of Hip Fracture Rates Hip fracture rate (% patients with fx)

Relative

Fluoride

Control

risk

P value

High fluoride exposure Trial 1, plain NaF

6.9

3.0

2.30

0.2

Trial 3, MFP

0

1.4

0

0.2

Trial 4, MFP

6.0

1.4

4.29

0.08

Moderate fluoride exposure Trial 5, EC-NaF

0

1.4

0

0.2

Trial 6, EC-NaF

3.3

2.9

1.14

0.2

Trial 7, MFP

1.0

1.0

1.0

0.2

Trial 12, MFP

2.7

7.3

0.37

0.2

Trial 8, SR-NaF

0

1.8

0

0.2

Trial 9, SR-NaF

0

0

0

0.2

Trial 11, MFP

5.3

7.3

0.74

0.2

Low Fluoride Exposure

In moderate fluoride exposure, three trials (Trials 5 – 7) gave a nonsignificant, slightly higher rate of total appendicular fractures in the fluoride group (Table 6). In the remaining Trial 12, the total appendicular fracture rate was marginally lower in the fluoride group. In low fluoride exposure, all four trials (Trials 8 – 11) showed a numerically lower total appendicular fracture rate (by 25 – 65%) in the fluoride group compared with the placebo group. The change was statistically significant in Trial 11, but not in others possibly due to low power. The relative risk of total appendicular fracture rate was directly and significantly correlated with relative fluoride absorption (Fig. 4). As with spinal fractures, relative risk was about 1 or higher with high fluoride exposure, about 1 with moderate fluoride exposure, and less than 1 with low fluoride exposure.

VI. DISCUSSION

Note. Fx, fracture.

numerically higher hip fracture rate in the fluoride group than in the placebo group (Table 5). The change was marginally significant in Trial 4 and not significantly different in Trial 1. In Trial 3 with a lower relative fluoride absorption, the fluoride group had no hip fracture, but 1.4% of patients in the placebo group sustained a hip fracture. In moderate fluoride exposure, Trial 6 had a numerically higher hip fracture rate in the fluoride group, and Trial 7 displayed an equivalent rate (Table 5). Remaining trials (Trials 5 and 12) had a numerically lower hip fracture rate. None of these changes were statistically significant. In low fluoride exposure, Trials 8 and 11 displayed a nonsignificant lower hip fracture rate in the fluoride group, and Trial 9 showed an equivalent rate (Table 5).

A thorough review of pharmacokinetic studies and randomized clinical trials with fluoride disclosed a biphasic action of fluoride. From 19 single-dose bioavailability studies, fluoride absorption and peak-to-basal variation in serum fluoride were calculated. From these pharmacokinetic parameters, it was possible to estimate the peak fluoride concentration in serum after multiple dosing and skeletal fluoride content after 4

B. Total Appendicular Fracture Rates Total appendicular fracture data were not available in one trial (Trial 2). Table 6 compares rates of total appendicular fractures between fluoride and placebo-control groups for the remaining trials. Both hip fractures and other nonvertebral fractures were counted but incomplete fractures were excluded. In high fluoride exposure, Trials 1 and 4 with the highest values for relative fluoride absorption revealed a marginally higher rate of total appendicular fractures in the fluoride group than in the placebo group (Table 6). Trial 3 gave a lower rate in the fluoride group, but this change was not significant.

FIGURE 3 Dependence of relative risk of spinal fracture rate on the degree of fluoride exposure. Vertical bars represent relative risks and 95% confidence intervals for each trial. Diagonal line is the regression line. Two vertical dashed lines separate three levels of fluoride exposure. Dashed horizontal line depicts relative risk of 1. r, correlation coefficient.

CHAPTER 75

TABLE 6

707

A New Perspective on Fluoride Therapy

Summary of Total Appendicular Fracture Rates Fracture rate (% patients with fx) Fluoride Control

Relative risk

P value

High fluoride exposure Trial 1, Plain NaF

34.7

21.8

1.59

0.06

Trial 3, MFP

8.8

11.6

0.76

Trial 4, MFP

19.4

11.6

1.67

13.7

11.6

1.18

0.2

Trial 6, EC-NaF

12.8

12.5

1.02

0.2

Trial 7, MFP

12.0

11.0

1.09

0.2

Trial 12, MFP

35.1

53.7

0.65

Trial 8, SR-NaF

3.7

10.7

0.35

0.2

Trial 9, SR-NaF

7.9

10.5

0.75

0.2

Trial 10, MFP

10.0

26.7

0.37

0.18

Trial 11, MFP

21.1

53.7

0.39

0.005

0.2 0.14

Moderate fluoride exposure Trial 5, EC-NaF

0.12

Low fluoride exposure

years of treatment. For sustained-release NaF for which normally mineralized bone formation has been reported [8], the estimated peak fluoride concentration in serum was kept within the therapeutic window and the skeletal fluoride was below the toxic threshold. However, for more bioavailable fluoride preparations (such as plain NaF for which abnormal bone formation has been reported) [7], both peak serum fluoride and bone fluoride exceeded the toxic threshold. A review of 12 randomized clinical trials with fluoride showed a clear dependence of skeletal fracture outcome on fluoride exposure. In trials with fluoride preparations that conferred toxic fluoride exposure, spinal fracture rate was similar between fluoride and placebo groups, and the hip and total appendicular fracture rates were generally numerically higher in the fluoride group. A moderate fluoride exposure yielded a modest decline or no change in spinal fracture rate, and no change in hip or appendicular fracture rates. A low fluoride exposure caused a marked reduction in spinal fracture rate, with lower hip and total appendicular fracture rates. The relative risk of spinal and total appendicular fracture rates was directly and significantly correlated with relative fluoride absorption. The biphasic action of fluoride can explain the conflicting results of two recent trials. In the FAVOStudy by Meunier et al. [14], the inability of fluoride treatment to inhibit spinal fractures could be ascribed to the use of fluoride preparations that conferred moderate to high fluoride exposure. The estimated amount of fluoride absorbed from the gastrointestinal tract was 2.7 – 4.8 times greater than that of

Neosten over 14 months of treatment (corresponding to 1 cycle) (Table 3). It is also noteworthy that hip and total appendicular fracture rates were numerically highest in the group taking the largest dose of fluoride (Trial 4, Tables 5 and 6). Ringe et al. [5] compared the effect of a moderate dose of continuous MFP treatment with a low dose of MFP given intermittently. The former group mimicked the study of Meunier et al. [14] and the latter group simulated low fluoride exposure with SR-NaF [4]. While both MFPtreated groups displayed a significant reduction in spinal fracture rate by fluoride, the intermittent MFP with low fluoride exposure displayed a greater fall than the continuous MFP with moderate fluoride exposure. Moreover, the group with low fluoride exposure displayed a significantly lower total appendicular fracture rate compared with placebo. Thus, the study of Ringe et al. [5] partly explains the negative results of Meunier et al. [14] and provides confirmation of the efficacy of SR-NaF in inhibiting spinal fractures shown previously [4]. It also suggests that the risk of appendicular fractures, which may complicate high fluoride exposure [2], does not prevail with low fluoride exposure. On the contrary, a low fluoride exposure may be protective against appendicular fractures. In conclusion, a review of 19 pharmacokinetic studies and 12 randomized trials with fluoride supports the biphasic fluoride action. The negative effects on fracture rates of plain NaF from the report of Riggs et al. [2], and of MFP and enteric-coated NaF from the report of Meunier et al. [14] could be attributed to excessive fluoride exposure with

FIGURE 4

Dependence of relative risk of total appendicular fracture rate on the degree of fluoride exposure. Solid diagonal line is the regression line. r, correlation coefficient; relative risk, rate in fluoride group divided by rate in control/placebo group; total, hip  other appendicular fractures.

708

CHARLES Y. C. PAK

toxic fluoride content in serum and bone. Positive effects on spinal fractures were obtained with SR-NaF and lowdose MFP used intermittently, the treatment formats that provided a low fluoride exposure and subtoxic fluoride content in serum and bone. The risk of hip or appendicular fractures was not a concern with low fluoride exposure.

13.

14.

15.

References 16. 1.

2.

3.

4.

5.

6.

7.

8.

9. 10.

11. 12.

J. R. Farley, J. E. Wergedal, and D. J. Baylink, Fluoride directly stimulates proliferation and alkaline phosphatase activity of boneforming cells. Science 222, 330 – 332 (1983). B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. S. Chao, et al. Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). B. L. Riggs, E. Seeman, S. F. Hodgson, D. R. Taves, et al. Effect of the fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporosis. N. Engl. J. Med. 123, 401 – 408 (1982). C. Y. C. Pak, K. Sakhaee, B. Adams-Huet, V. Piziak, et al. Treatment of postmenopausal osteoporosis with slow-release sodium fluoride: Final update of a randomized controlled trial. Ann. Intern. Med. 123, 401 – 408 (1995). J. D. Ringe, C. Kipshoven, A. Coster, and R. Umback, Therapy of established postmenopausal osteoporosis with monofluorophosphate plus calcium: Dose-related effects on bone density and fracture rate. Osteoporosis Int. 9, 171 – 178 (1999). P. Fratzl, P. Roschger, J. Eschberger, B. Abendroth, et al. Abnormal bone mineralization after fluoride treatment in osteoporosis: A smallangle x-ray-scattering study. J. Bone Miner. Res. 9, 1541 – 1549 (1994). M. W. Lundy, M. Stauffer, J. E. Wergedal, D. J. Baylink, et al., Histomorphometric analysis of iliac crest bone biopsies in placebo-treated versus fluoride-treated subjects. Osteoporosis Int. 5, 115 – 129 (1995). J. E. Zerwekh, P. P. Antich, K. Sakhaee, J. Prior, et al. Lack of deleterious effect of slow-release sodium fluoride treatment on cortical bone histology and quality in osteoporosis patients. Bone Miner. 18, 65 – 76 (1992). D. R. Taves, New approach to the treatment of bone disease with fluoride. Fed. Proc. 29, 1185 – 1187 (1970). K. Sakhaee and C. Y. C. Pak, Fluoride bioavailability from immediate-release sodium fluoride with calcium carbonate compared with slow-release sodium fluoride with calcium citrate. Bone Miner. 14, 131 – 136 (1991). J. D. Ringe and P. J. Meunier, What is the future for fluoride in the treatment of osteoporosis? Osteoporosis Int. 5, 71 – 74 (1995). C. Y. C. Pak, K. Sakhaee, C. D. Rubin, J. E. Zerwekh, et al. Sustained-release sodium fluoride in the management of established postmenopausal osteoporosis. Am. J. Med. Sci. 313, 23 – 32 (1997).

17.

18.

19.

20. 21.

22.

23.

24.

25. 26.

27.

M. Kleerekoper, E. L. Peterson, D. A. Nelson, E. Phillips, et al. A randomized trial of sodium fluoride as a treatment for postmenopausal osteoporosis. Osteoporosis Int. 1, 155 – 161 (1991). P. J. Meunier, J. L. Sebert, J. Y. Reginster, D. Briancon, et al. Fluoride salts are no better at preventing new vertebral fractures than calcium-vitamin D preparations in postmenopausal osteoporosis: the FAVOStudy. Osteoporosis Int. 8, 4 – 12 (1998). N. Mamelle, P. J. Meunier, R. Dusan, M. Guillaume, et al. Risk-benefit ratio of sodium fluoride treatment in primary vertebral osteoporosis. Lancet 2, 361 – 365 (1988). J. Y. Reginster, L. Meurmans, B. Zegels, L. C. Rovati, et al. The effect of sodium monofluorophosphate plus calcium on vertebral fracture rate in postmenopausal women with moderate osteoporosis. Ann. Intern. Med. 129, 1 – 8 (1998). C. D. Rubin, B. Adams-Huet, and C. Y. C. Pak, The treatment of senile osteoporosis with sustained-release sodium fluoride. J. Bone Miner. Res. 12(1) S243 (1997). J. D. Ringe, A. Dorst, C. Kipshoven, L. C. Rovati, et al. Avoidance of vertebral fractures in men with idiopathic osteoporosis by a three year therapy with calcium and low dose intermittent monofluorophosphate. Osteoporosis Int. 8, 47 – 52 (1998). W. L. Gitomer, K. Sakhaee, and C. Y. C. Pak, Comparison of fluoride bioavailability from a sustained-release NaF preparation (Neosten) and other fluoride preparations. J. Clin. Pharm. 40, 138 – 141 (2000). P. Meunier, P. Pers. commun., 1990. F. Liote, C. Bardin, A. Liou, A. Brouard, et al. (1992). Bioavailability of fluoride in postmenopausal women: comparative study between sodium fluoride and disodium monofluorophosphate-calcium carbonate. Calcif. Tissue Int. 50, 209 – 213. H. Resch, C. Libanati, J. Talbot, M. Tabuenca, et al. Pharmacokinetic profile of a new fluoride preparation: sustained-release monofluorophosphate. Calcif. Tissue Int. 54, 7 – 11 (1994). G. Warneke and I. Setnikar, Effects of meal on the pharmacokinetics of fluoride from monofluorophosphate. Arzneim-Forsch/Drug Res. 43(I), 584 – 590 (1993). D. Briancon, P. D’Aranda, P. Quillet, B. Duplan, et al. Comparison study of fluoride availability following administration of sodium fluoride alone and in combination with different calcium salts. J. Bone Miner. Res. 5, S71 – S73 (1990). C. Nagant de Deuxchaisnes, J. P. Devogelaer, and F. Stein, Fluoride treatment for osteoporosis. Lancet 336, 48 – 49 (1990). J. P. Devogelaer, C. Nagant de Deuxchaisnes, and F. Stein, Bioavailability of enteric-coated sodium fluoride tablets as affected by the administration of calcium supplements at different time intervals. J. Bone Miner. Res. 5, S75 – S79 (1990). C. Y. C. Pak, K. Sakhaee, C. Gallagher, C. Parcel, et al. Attainment of therapeutic fluoride levels in serum without major side effects using a slow release preparation of sodium fluoride in postmenopausal osteoporosis. J. Bone Miner. Res. 1, 563 – 571 (1986).

CHAPTER 76

Androgens and Androgenic Progestins SUNDEEP KHOSLA

I. II. III. IV.

Department of Medicine, Mayo Clinic and Foundation, Rochester, Minnesota 55905

V. Tibolone VI. Conclusions References

Introduction Classification of Androgens and Progestins Androgens Progesterone and Androgenic Progestins

II. CLASSIFICATION OF ANDROGENS AND PROGESTINS

I. INTRODUCTION Since osteoporosis has been perceived, at least until recently, as primarily a disease of women, most of the scientific attention in terms of its pathogenesis and treatment has focused on estrogen. It is clear, however, that androgens also have significant effects on bone. In addition, both endogenous progesterone and androgenic progestins may have significant skeletal effects. Thus, a comprehensive understanding of the role of sex steroids in bone metabolism must include a consideration of androgens and progestins. Therein also lie potential opportunities for improved therapy of osteoporosis, including possible nonskeletal benefits of these agents, as for example on libido and muscle mass. This chapter reviews briefly the potential role of androgens and progesterone on bone metabolism in animals and humans and then focuses on the potential therapeutic uses of these agents in the treatment of osteoporosis. Also see Chapters 11  42 for a discussion of androgen effects on bone cells and osteoporosis in men respectively.

OSTEOPOROSIS, SECOND EDITION VOLUME 2

The major classes of androgens relevant to the subsequent discussion are the testicular androgens (testosterone and 5-dihydrotestosterone (5-DHT)), adrenal androgens (principally dehydroepiandrosterone (DHEA) and its sulfated ester, DHEA-sulfate (DHEAS), as well as androstenedione), and synthetic anabolic steroids (such as nandrolone decanoate and stanozolol). The specific chemical structures of these compounds are shown in Fig. 1. The testicular and synthetic androgens bind directly to the androgen receptor, whereas the adrenal androgens are metabolized in peripheral tissues to testosterone and 5-DHT prior to receptor binding. The synthetic anabolic steroids were developed with the hope of separating the anabolic effects of androgens on target tissues (i.e., muscle, bone, etc.) from their virilizing effects. This effort has met with only modest success, as complete separation of the two activities has not been possible. The progestins can be classified into the natural progestin, progesterone, versus synthetic progestins. The latter, in turn,

709

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

710

SUNDEEP KHOSLA

FIGURE 1

Structures of testicular, adrenal, and commonly used synthetic androgens.

CHAPTER 76 Androgens and Androgenic Progestins

FIGURE 2

Structures of progesterone, synthetic nonandrogenic, and synthetic androgenic progestins.

711

712

SUNDEEP KHOSLA

FIGURE 3 Structure of the mixed estrogenic, progestogenic, and androgenic compound, tibolone. are divided into those that are structurally related to progesterone and hence, have minimal or no androgenic activity (such as medroxyprogesterone acetate and megestrol acetate) versus those that are structurally related to testosterone and therefore have variable androgenic activity (such as norethindrone and norethindrone acetate). Figure 2 depicts the structures of the relevant natural and synthetic progestins. It is important to keep in mind that from a therapeutic standpoint, the effects of the androgenic progestins may differ substantial from these of the nonandrogenic progestins. Finally, synthetic steroids are being developed with mixed activities, as for example tibolone (Fig. 3). Tibolone has mixed estrogenic, androgenic, and progestogenic activities, and represents a promising new approach to hormone replacement therapy for osteoporosis.

III. ANDROGENS A. Effects of Androgens on Bone Cells The regulation of bone cell function by androgens is reviewed in detail in Chapter 11, and will be only briefly summarized here. Both osteoblasts [1] and osteoclasts [2] have been shown to have androgen receptors. In addition, the number of specific androgen binding sites in osteoblasts (1000 – 3000 sites/cell) is similar to the number of binding sites present in other androgen-responsive tissues, such as the prostate [3]. Bone cells have also been shown to have 5-reductase activity [4 – 6], which is responsible for the conversion of testosterone to 5-DHT, the major biologically active metabolite of testosterone in most tissues. In addition, given increasing evidence that estrogen plays an important role in bone metabolism, not just in women but also in men [7], it is of interest that bone cells also possess aromatase activity [8], which can convert androstenedione and

testosterone to estrogens. The presence of estrogen receptors and estrogen action in bone cells is disscused in Chapter 10. Androgens have been shown to stimulate osteoblast proliferation in most [9 – 11], but not all systems [12,13]. In terms of osteoblast differentiation, androgens increase alkaline phosphatase activity in primary cultures of osteoblastlike cells [9,10], although studies using a fetal human osteoblast cell line stably transfected with the androgen receptor found a decrease in alkaline phosphatase activity following exposure to 5-DHT [12]. As in the case of estrogen, it appears that androgens regulate the production of a number of growth factors/cytokines by osteoblastic cells, including interleukin-6 [14,15], transforming growth factor- [11,13,16], and the insulin-like growth factor system [11,17]. DHEA is generally considered a prohormone which is converted in various tissues to androstenedione and then into potent androgens and/or estrogens. However, specific binding sites for DHEA have been described [18], and Bodine et al. [16] showed that DHEA, but not testosterone or 5-DHT, caused a rapid reduction (30 min posttreatment) in c-fos mRNA levels in normal human osteoblastic cells. All three agents significantly increased transforming growth factor- activity in these cells. These results thus suggest that bone cells have the ability to respond to DHEA and that the effects of DHEA or these cells may not always be identical to those of testosterone.

B. Effects of Androgens in Experimental Animals Several studies using gonadectomized (male and female) rodent models have examined the role of androgens on bone metabolism. Turner et al. [19] found that ovariectomy in female rats resulted in an increase in periosteal bone formation rate, whereas orchiectomy in male rats resulted in a decrease in periosteal bone formation rate (Fig. 4). By inference, these data indicate that androgens stimulate, whereas estrogen inhibits, periosteal bone formation in this model. Similar findings were reported by Coxam et al. [20], who found that following ovariectomy in the rat, estrogen treatment markedly decreased periosteal bone formation rates, whereas 5-DHT cotreatment prevented the decrease in periosteal bone formation induced by estrogen alone. In contrast, however, Vanderschueren et al. [21] found that whereas cortical area decreased 4 months following orchidectomy in aged rats, testosterone, 5-DHT, nandrolone, and estrogen were equally effective in preventing this decrease. Moreover, all four agents were also equally effective in preventing the loss of cancellous and cortical bone following orchidectomy. In an attempt to dissect the relative contributions of estrogens versus androgens toward bone metabolism, Vanderschueren et al. [22] compared the effects of orchidectomy to those of the aromatase inhibitor, vorozole, in aged

713

CHAPTER 76 Androgens and Androgenic Progestins

(A) The effect of ovariectomy on periosteal bone formation rate. The mean  SE (vertical bar) and tetracycline labeling period (horizontal line) for intact controls (open circles) and ovariectomized (closed circles) rats are shown as a function of time after ovariectomy. P  0.01 for all ovariectomy time points compared to intact controls. (B) The effect of orchidectomy on periosteal bone formation rate. The mean  SE (vertical bar) and tetracycline labeling period (horizontal line) for intact controls (open triangles) and orchidectomized (closed triangles) rats are shown as a function of time after orchidectomy. P  0.01 for all orchidectomy time points compared to intact controls. Adapted with premission from Turner et al. [19].

FIGURE 4

male rats. As shown in Fig. 5, sham-operated animals given vorazole had similar decreases in calcium content of the first four lumbar vertebrae as orchidectomized animals, consistent with an important role for the aromatization of androgens to estrogens in the maintenance of bone mass in male rats. The potential regulatory role of adrenal androgens in bone metabolism has also been assessed in several animal studies. Thus, Durbridge et al. [23] found that, in female rats, adrenalectomy alone resulted in loss of metaphyseal trabecular bone of an extent similar to that produced by oophorectomy. These investigators interpreted these data as being consistent with an important role for adrenal

androgens in the maintenence of skeletal mass in rats. In addition, Turner et al. [24] showed that treatment with DHEA reduced the loss of cancellous bone following oophorectomy in female rats, indicating that adrenal androgens may prevent the bone loss induced by estrogen deficiency. Collectively, the animal data indicate that both estrogen and androgens are likely important for bone metabolism in both male and female animals. There is some evidence for a preferential effect of androgens on periosteal bone formation, and at least part of the effect of testosterone and other aromatizable androgens on the skeleton is likely mediated via estrogen.

C. Effects of Androgens in Humans 1. RELATIONSHIP OF ANDROGENS TO BONE DENSITY

FIGURE 5

Calcium content of the first four lumbar vertebrae in sham and orchidectomized male rats treated without (open bars) or with (hatched bars) vorazole. Significance: *P  0.05 versus sham; **P  0.01 versus sham; ***P  0.001 versus sham. Adapted with permission from Vanderschueren et al. [22].

Perhaps the most obvious manifestation of androgen effects on bone metabolism is the sexual dimorphism of the skeleton that is evident during adolescence: the male skeleton is larger in most dimensions than the female skeleton, the diameter and cortical thickness of long bones is greater in men than in women, and vertebral size is larger in men [25 – 27]. In part, this is likely due to the stimulation of periosteal bone formation demonstrated in the animal studies, although direct evidence for this in humans is lacking. Consistent with the animal studies using aromatase inhibitors, however, it is clear that estrogens also play a significant role on bone metabolism in men. Thus, Smith et al. [28] reported a young man with homozygous mutations in the ER- gene who had severe osteopenia and unfused epiphyses.

714

SUNDEEP KHOSLA

In addition, two men with complete aromatase deficiency due to mutations in the aromatase gene [29,30] had similar skeletal phenotypes. The results of these “experiments of nature” thus suggest that estrogen likely plays a significant role in bone metabolism in males, at least during growth and adolescence. Several epidemiologic studies have examined the relationship between sex steroids and bone mineral density (BMD) in men and in women, with somewhat conflicting results. Although several studies noted significant correlations between either total or bioavailable testosterone concentrations in aging men and BMD at various sites [31,32], others failed to confirm this relationship [33,34]. To address this issue, we examined these relationships in a population-based, age-stratified sample of 346 men (ages 23–90 years) and 304 women (ages 21–94 years) (Table 1) [35]. Total testosterone values did not correlate with BMD at any site in the men, except weakly at the radius. By contrast, bioavailable (not protein bound to sex hormone binding globulin or other protein) testosterone did correlate significantly with BMD at all sites in the men. Both total and bioavailable testosterone correlated equally well with BMD at all sites in the women. Of interest, total and bioavailable estrogen (estradiol plus estrone) concentrations were also positively associated with BMD in men as well as women (Table 1). By multivariate analysis, however, bioavailable estrogen was a better predictor of BMD in men and postmenopausal women than bioavailable testosterone. In contrast, bioavailable testosterone was a more robust predictor of BMD than were estrogen levels in premenopausal women. Similar findings have been reported by Slemenda et al. [36], who found positive associations between circulating estradiol and BMD in men. In other studies, these investigators also found that bone loss in pre-

menopausal women was associated with lower androgen concentrations, and with lower estrogens and androgens in periand postmenopausal women [37]. In addition to the gonadal androgens, considerable attention has been focused in recent years on the possible relationship of adrenal androgens to a number of age-related diseases. DHEA and its sulfate ester (DHEAS) are the major circulating adrenal androgens [38]. DHEAS is enzymatically converted to DHEA in all tissues [38], with subsequent local conversion of DHEA in various tissues to androgens and estrogens. Serum DHEA concentrations peak by the second decade of life and then steadily decline by an average of 10% per decade [38], with an accelerated decline after age 80. Indeed, no other hormone declines as profoundly with aging as DHEA. Several studies have noted associations between serum DHEAS levels and BMD [39,40], although others [41,42] have failed to find such an association. In addition, postmenopausal women with primary adrenal failure were noted to have decreased circulating DHEA and decreased distal forearm bone density compared with control postmenopausal women [43], further suggesting a role for adrenal androgens in the maintenence of bone mass, particularly in postmenopausal women. Collectively, the data are consistent with a clear role for gonadal androgens in bone metabolism in both genders, although the relative effects of androgens versus estrogens on the skeleton require further clarification. Adrenal androgens may also efect significant exert on bone, although the evidence for this is much less robust. ON

TABLE 1 Correlation Coefficients for Association between BMD and Sex Steroid Levels among Age-Stratified Samples of Rochester, MN, Men and Women BMD

Total T

Bio T

Total E

Bio E

Total body Men Women

 0.01 0.22*

0.22* 0.22*

0.16† 0.42*

0.31* 0.42*

Lateral spine Men Women

0.09 0.27*

0.33* 0.29*

0.20* 0.43*

0.38* 0.45*

Proximal femur Men Women

0.00 0.25*

0.28* 0.30*

0.19* 0.33*

0.38* 0.38*

Radius Men Women

0.15† 0.27*

0.38* 0.28*

0.15† 0.43*

0.34* 0.45*

Note. T, testosterone; bio, bioavailable; E, estrogen (estrone  estradiol). Adapted with permission from Khosla [35]. * P  0.001. † P  0.01.

2. EFFECTS OF EXOGENOUS ANDROGENS BONE TURNOVER AND BMD

Studies using androgen analogs in postmenopausal women have generally noted decreases in fasting urinary calcium excretion, consistent with decreased calcium release from bone, although a direct effect of androgens on renal calcium handling is also possible. Chesnut et al. [44] noted a 32% decrease in urinary calcium excretion in 23 postmenopausal women with osteoporosis following 8 – 32 months of therapy with stanozolol. Similarly, Need et al. [45] indirectly estimated renal tubular calcium reabsorption and found that this increased significantly in 27 women with postmenopausal osteoporosis treated for 3 months with nandrolone decanoate. Studies directly measuring bone resorption markers have generally found a decrease in these markers following androgen therapy. Tenover [46], for example, showed that 3 months of testosterone treatment of 13 healthy men (age 57 – 76 years) with borderline testosterone status at baseline resulted in a 28% reduction in urinary hydroxyproline excretion. Similarly, Anderson et al. [47] treated eugonadal men with osteoporosis with 6 months of intramuscular testosterone and found a 19% reduction in urinary deoxypyridinoline excretion and a 39% reduction in urinary N-telopeptide of type I collagen (NTx) excretion.

CHAPTER 76 Androgens and Androgenic Progestins

In addition to inhibiting bone resorption, androgen treatment may also stimulate bone formation, although data are somewhat conflicting in this regard [70 – 74]. Thus, in the studies noted above, Tenover [46] found that testosterone treatment had no effect on the bone formation marker, serum osteocalcin. Riggs et al. [48] acquired bone biopsies from 29 women with postmenopausal osteoporosis before and after either estrogen or oxandrolone (a synthetic androgen) treatment and found a decrease in both resorption and formation surfaces. In contrast, Raisz et al. [49] found that postmenopausal women treated with estrogen plus 2.5 mg of methyltestosterone had a 24% increase in serum osteocalcin concentration after 3 months of treatment, compared to 40% lower values in women treated with estrogen alone. Similar findings were noted for bone-specific alkaline phosphatase and C-terminal procollagen peptide (Fig. 6). In

The effects of estrogen  testosterone (closed circles) and estrogen alone (open circles) on serum bone formation marker levels, osteocalcin, bone-specific alkaline phosphatase (BSAP), and C-terminal procollagen peptide (PICP), from baseline (B) through treatment weeks 3, 6, and 9, and 3 weeks posttreatment (3Post). †, Significantly different from baseline, P  0.05. *, Significant difference between treatments, P  0.05. Adapted with permission from Raisz et al. [49].

715 addition, nandrolone decanoate therapy of postmenopausal women has also been found to increase bone formation markers [50,51]. Thus, at least over the short term, androgen treatment may have a mixed anabolic/anti-resorptive effect on bone, although the issue is far from resolved and further studies addressing this issue are needed. Studies assessing the BMD response to androgens have also generally found positive results, although the magnitude of response has varied depending on the study population. Most studies examining this issue have been made in men with overt hypogonadism. In a study of 21 adult males (mean age 29 years, range 19 to 53 years) with hypogonadotropic hypogonadism studied before and after restoration of gonadal status, for example, Finkelstein et al. [52] found that in men with open epiphyses, cortical and trabecular BMD increased nearly 13% over 2 years, whereas in men with fused epiphyses, cortical BMD increased only by 4% and trabecular BMD did not change. Similarly, Devogelaer et al. (53) studied 16 male hypogonadal patients (mean age, 31.9 years; range, 18 – 57 years) before and following testosterone replacement and found a significant increase (5.9%) in distal radius bone mineral content. Of interest, an open, prospective trial of intramuscular testosterone treatment of eugonadal osteoporotic men found that spine BMD increased by 5% over 6 months, although hip BMD did not change [54]. In contrast, Snyder et al. [55] could not demonstrate a significant effect of testosterone therapy on BMD in men with low-normal baseline testosterone concentrations, although they did demonstrate by linear regression analysis that the lower the pretreatment serum testosterone concentration, the greater the effect of testosterone on lumbar spine BMD at 36 months (Fig. 7). Finally, Reid et al. [56] showed that in glucocorticoid-treated men who were treated with intramuscular

FIGURE 6

FIGURE 7 The testosterone treatment effect on percent change in bone mineral density during 36 months of testosterone treatment in men over age 65 years as a function of the pretreatment serum testosterone concentration. The testosterone treatment effect was statistically significant (P  0.01) for pretreatment serum testosterone concentrations of 100 – 300 ng/dl. Adapted with permission from Snyder et al. [55].

716 testosterone for 12 months and then crossed over to 12 months of observation alone, testosterone therapy was associated with a 5% increase in lumbar spine BMD, compared to no change during the control period (P  0.05). Studies assessing androgen effects on bone mass in postmenopausal women have exclusively used synthetic anabolic agents with the goal of minimizing virilizing side effects. Chesnut et al. [57] initially reported that 26 months of methandrostenolone treatment of women with postmenopausal osteoporosis resulted in a 2% increase in total body calcium, compared to a 3% decrease in the placebo group (P  0.01 for the difference between groups). They followed this up with a similar study using stanozolol which resulted in a 4.4% increase in total body calcium, compared to no change in the placebo group (P  0.03) [44]. Subsequent to these early studies, the majority of the later studies have used nandrolone decanoate as the anabolic androgen. Geusens et al. [58], compared the effects of 2 years of therapy of 48 women and 12 men with symptomatic osteoporosis with either nandrolone decanoate, 1hydroxyvitamin D3 or intermittent calcium infusions. Nandrolone therapy resulted in a significant increase in bone mineral content at the radius (assessed by single photon absorptiometry) along with reduced endosteal bone loss at the metacarpals. Several other studies using single photon absorptiometry found similar results [59 – 61], even adjusting for possible changes in marrow fat which may have confounded the results [60]. Subsequently, Gennari et al. [62] used dual photon absorptiometry and also showed a beneficial effect of nandrolone decanoate on lumbar spine bone mineral content in postmenopausal women (Fig. 8). Despite these positive effects on BMD, there are at present no

FIGURE 8

Lumbar spine bone mineral content (BMC) variations (mean  SE) during nandrolone decanoate (dashed lines) or placebo (solid lines) treatment in established postmenopausal osteoporotic patients. *P  0.05 versus baseline. Adapted with permission from Gennari et al. [62].

SUNDEEP KHOSLA

prospective randomized studies assessing the effects of androgens on fracture frequency. In contrast to these data on gonadal or synthetic androgens, there are little or no data on the effects of adrenal androgens on bone metabolism in humans. Several investigators have employed either pharmacologic [63] or physiologic [64] doses of DHEA and have found variable effects on blood lipids or body composition, although none of these studies examined the effects of adrenal androgen replacement on the skeleton. Of interest, a preliminary report of DHEA therapy in six patients with systemic lupus erythematosis who were on corticosteroids did suggest that DHEA may have some utility in preventing steroid-induced bone loss [65]. In addition, Labrie et al. [66], in an uncontrolled study, recently assessed the effects of daily applications of 10% DHEA cream in 14 postmenopausal women and noted a significant (2.0%) increase in hip BMD associated with a decrease in bone turnover markers. It is also important to keep in mind that the effects of DHEA may differ from those of testosterone, since their local tissue metabolism may differ. Further, DHEA may also have distinct effects on bone cells [16]. Finally, despite the absence of any clear data indicating beneficial effects of DHEA replacement on any age-related changes, this agent has stimulated a great deal of attention among the media and the public, and a substantial number of individuals currently use DHEA to prevent or reverse these age-related changes. Although this discussion has focused on the skeletal effects of androgens, it should be kept in mind that androgen therapy also results in significant alterations in body composition that may have a beneficial effect on fracture risk [67,68]. In general, androgen therapy is associated with decreases in body fat and increases in lean mass [67,68]. This may, in turn, result in enhanced muscle strength and a reduced risk of falling. In addition, some women find a marked decrease in libido following the menopause, and androgen treatment has recently been shown to have benefi cial effects on libido and sexual interest [69]. Finally, androgen treatment clearly has potential adverse side effects. It is generally associated with reductions in HDL concentrations [67,70 – 72]. In men, testosterone treatment mandates close follow-up for prostate cancer and possible worsening symptoms of prostatism. Treatment with nandrolone decanoate has resulted in increases in serum aspartate aminotransferase (AST) activity, and oral anabolic steroids used in high doses for prolonged periods have been associated with hepatomas and peliosis hepatica [73]. Other adverse reactions to nandrolone in women include increased facial hair, acne, and hoarseness. In summary, androgens clearly have a potential role in the treatment of osteoporosis, although where they fit into the overall therapeutic strategy remains to be defined. The most unequivocal indication is for the use of testosterone in hypogonadal men. Elderly men with decreased (but not

717

CHAPTER 76 Androgens and Androgenic Progestins

clearly hypogonadal) testosterone concentrations may also benefit from testosterone replacement, both from the standpoint of the skeleton and for their general sense of wellbeing, although much additional data are needed before this can be recommended on a routine basis. Anabolic androgens have been studied fairly extensively in postmenopausal women, and while they appear to have beneficial effects on bone mass, their possible adverse metabolic and virilizing side effects have dampened enthusiasm for their widespread use. Studies using adrenal androgens are currently underway, but their use cannot be recommended at this time since there are little or no data on their effects on the skeleton or other tissues. Perhaps the major advance in this area will come with the development of selective androgen receptor modulators (SARMs) which, in a manner analogous to the selective estrogen receptor modulators (SERMs), such as raloxifene, will have beneficial effects on the skeleton and on muscle, without some of the adverse consequences of androgen therapy noted above.

IV. PROGESTERONE AND ANDROGENIC PROGESTINS While estrogen and androgens clearly have significant effects on bone metabolism, the possible skeletal effects of progesterone remain a subject of ongoing controversy. This is reviewed briefly below, along with the available data on the effects of therapy with either progesterone or with androgenic progestins on the skeleton.

A. Potential Role of Progesterone in Skeletal Metabolism The presence of progesterone receptor (PR) mRNA and protein has been demonstrated in primary cultures of human osteoblasts [74] and in several osteoblastic cell lines (HOS TE85, MG-63, and SaOS-2) [75,76], including transcripts for both isoforms of the PR (PR-A and PR-B) [77,78]. PR protein was also demonstrated in human osteoclasts [79]. Thus, both osteoblasts and osteoclasts are at least potential targets for progesterone action. Progesterone has been noted to affect osteoblast and osteoblast precursor cells, although some of the data are conflicting, perhaps related to the different model systems used. Thus, Canalis and Raisz [80] demonstrated that progesterone inhibited proliferation and collagen synthesis by fetal rat calvarial cells. In contrast, Slootweg et al. [81] found that whereas progesterone alone did not affect SaOS-2 proliferation, it had a synergistic effect with estrogen in stimulating proliferation of these cells. Tremollieres et al. [82], however, reported that progesterone alone stimulated proliferation of both TE 85

osteosarcoma and normal osteoblastic cells. Subsequent studies observed stimulatory effects of progesterone on proliferation of normal and transformed osteoblastic cells [83 – 85]. Finally, Ishida et al. [86 – 89] found that progesterone increased the number of alkaline-phosphatase positive colonies derived from osteoprogenitor cells isolated from rat vertebral bone explants. A number of experimental animal studies have also indicated significant effects of progesterone on bone metabolism. Progesterone has been shown to stimulate mineralization of newly induced bone in rats [90], and to increase cortical bone formation rate in spayed Beagle dams [91,92]. Moreover, progesterone prevented ovariectomy-induced bone loss in rats, with inhibition of resorption indices but persistently elevated bone formation rates [93,94]. In addition, Bowman et al. [95] found that the high circulating progesterone concentrations in pseudopregnant rats were associated with preservation of bone mass despite estrogen levels that were comparable to ovariectomized rats. In humans, Prior et al. [96] found that women with ovulatory disturbances related to luteal phase defects lost bone mass over 1 year compared to women with normal menstrual cycles. In addition, the woman showed increases in bone mass in response to medroxyprogesterone therapy during the luteal phase [97]. Other investigators, however, have not confirmed these findings [98,99].

B. Effects of Treatment with Progesterone and Androgenic Progestins Several studies have found that progesterone treatment of postmenopausal women resulted in decreased urinary indices of bone resorption [100,101]. Studies using BMD as an end point, however, have shown variable results. McNeeley et al. [102] found that medroxyprogesterone alone was as effective as estrogen or estrogen plus medroxyprogesterone in the prevention of bone loss in postmenopausal women over 1 year, and Gallagher et al. [103] found that medroxyprogesterone therapy of postmenopausal women either prevented or significantly attenuated bone loss at predominantly cortical sites (total body, radius, and metacarpal cortex), but had no effect in preventing bone loss in predominantly cancellous bone, the spine. Subsequently, Grey et al. [104] demonstrated that combined therapy with estrogen plus medroxyprogesterone resulted in a 6.6% increase in spine BMD over 1 year in postmenopausal women, compared to a 4.0% increase in women treated with estrogen alone (P  0.01). In the largest study addressing this issue, women assigned to continuous estrogen plus medroxyprogesterone therapy in the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial had greater increases in spine BMD (5%) compared to women assigned to placebo, estrogen alone,

718

SUNDEEP KHOSLA

or estrogen plus cyclical progesterone (micronized progesterone or medroxyprogesterone) (average increase, 3.8%) [105], although when the investigators performed a subgroup analysis involving only those women adhering to assigned therapy, there was no significant difference in BMD response among the four active groups. Collectively, the in vitro, animal, and human data do indicate that progesterone may have effects on bone metabolism. Some of the controversy in the literature may be related to the fact that the doses of progesterone required to achieve significant skeletal responses may be considerably higher than those typically used in postmenopausal hormone replacement therapy. Thus, the major role of progesterone on bone may not be when given exogenously (an in estrogen progestin therapy of postmenopausal women at current doses), but rather in states where endogenous progesterone concentrations are relatively high, as in pregnancy or during the luteal phase of the menstrual cycle. More studies are needed to address these issues. In contrast to data with nonandrogenic progestins, studies using androgenic progestins have generally found beneficial skeletal effects of these compounds. Abdalla et al. [106] initially demonstrated that norethisterone prevented radial bone loss in postmenopausal women. Several other studies using norethisterone acetate have found similar results [107 – 111]. Perhaps the most convincing evidence for

a beneficial skeletal effect of norethisterone came from the Continuous Hormones as Replacement Therapy (CHART) study, in which patients were randomized to placebo or one of eight treatment groups consisting of varying doses of either norethisterone acetate combined with ethinyl estradiol or varying doses of ethinyl estradiol alone [112]. Figure 9 summarizes the results of this study, which clearly showed that combined treatment with norethisterone acetate and estradiol produced greater increases in BMD than treatment with estradiol alone. Whereas treatment with estradiol alone was associated with endometrial hyperplasia, the combination of norethisterone acetate and estradiol protected the endometrium against hyperplasia. These data thus indicate that androgenic progestins, when used with estrogen, may have beneficial skeletal effects. The major concern with progestins, however, remain the possible adverse lipid and vascular effects of certain progestins [113], as well as increasing evidence that combined estrogen/progestin therapy may be associated with a greater risk of breast cancer than estrogen therapy alone [114]. Again, the possible development of more selective progestational agents that protect against estrogen-induced endometrial hyperplasia but without adverse effects on the breast or on lipid profiles would provide improved options for hormone replacement therapy in postmenopausal women.

Adjusted mean change (  SE) from baseline in spinal trabecular BMD measure by quantitative computed tomography at the end of 24 months of treatment with either a combination of norethisterone acetate and ethinyl estradiol (A) or ethinyl estradiol alone (B). A statistically significant (P  0.05) linear dose – response relationship was present in the norethisterone acetate and ethinyl estradiol combined groups, but not in the ethinyl estradiol alone groups. Adapted with permission from Speroff et al. [112].

FIGURE 9

719

CHAPTER 76 Androgens and Androgenic Progestins

VI. CONCLUSIONS While androgens and progestins have significant skeletal effects, their role in the management of osteoporosis remains to be clearly defined. There appears to be little doubt that hypogonadal men likely achieve skeletal and other benefits from testosterone replacement. Beyond this, however, much more work is needed to define the potential utility of testosterone replacement in aging men with mild degrees of hypogonadism, the role of androgen replacement in regimens of postmenopausal hormone replacement therapy, and the possible skeletal effects of DHEA. While progesterone may have skeletal effects in certain situations, only the androgenic progestins have clearly been shown to enhance the skeletal response to estrogen replacement therapy in postmenopausal women. The future development of tissue-specific androgenic and progestogenic compounds may significantly improve the risk/benefit profile of these agents. Finally, the development of synthetic steroids, such as tibolone, that can interact somewhat promiscuously with a number of sex steroid receptors will provide ever expanding options for hormone replacement therapy in women, and possibly also in men. FIGURE 10 Densitometric measurements at 3-month intervals, expressed as percentages of initial values (mean  SE). Open circles, 2.5 mg tibolone; closed circles, 1.25 mg tibolone; squares, placebo. BMD spine, BMD of the spine (L2 –L4); BMD forearm, BMD of the distal forearm. P  0.001 for placebo versus treated groups at both sites. Adapted with permission from Bjarnason et al. [118].

V. TIBOLONE Tibolone is a synthetic steroid that possesses weak estrogenic, progestogenic, and androgenic properties. It has about 1/50 the estrogenic activity of ethinyl estradiol and 1/8 the progestational activity of norethisterone [115]. Several studies have examined its effects on BMD [116 – 123], and Fig. 10 shows data representative of these studies. Over 2 years, tibolone resulted in an approximately 6% increase in spine BMD and a 2% increase in distal forearm BMD [118]. In a subsequent study, Lippuner et al. [120] have shown that a 2.5 mg/day dose of tibolone was comparable to 50 g/day of transdermal estrogen or 2 mg/day of oral estradiol in terms of preservation of bone mass. The improvement in BMD is associated with reductions in bone turnover. There are, however, no data at this point on tibolone’s effects on fracture risk. Tibolone appears not to have adverse effects on the uterus [115] and in contrast to SERMs such as raloxifene, it seems to reduce hot flushes [115]. Its major disadvantage is that it may reduce HDL cholesterol concentrations by as much as 26%, although it does also reduce Lp(a) levels and has little or no effect on circulating triglycerides [115].

References 1. D. S. Colvard, E. F. Eriksen, P. E. Keeting, E. M. Wilson, D. B. Lubahn, F. S. French, B. L. Riggs, and T. C. Spelsberg, Identification of androgen receptors in normal human osteoblast-like cells. Proc. Natl. Acad. Sci. USA 86, 854 – 857 (1989). 2. Y. Mizuno, T. Hosoi, S. Inoue, A. Ikegami, M. Kaneki, Y. Akedo, T. Nakamura, Y. Ouchi, C. Chang, and H. Orimo, Immunocytochemical identification of androgen receptor in mouse osteoclastlike multinucleated cells. Calcif. Tissue Int. 54, 325 – 326 (1994). 3. G. Sanchez-Visconti, L. Herrero, M. Rabadan, I. Pereira, and A. RuizTorres, Ageing and prostate: Age-related changes in androgen receptors of epithelial cells from benign and hypertrophic glands compared with cancer. Mech. Ageing Dev. 82, 19 – 29 (1995). 4. H. U. Schweikert, W. Rulf, N. Niederle, H. E. Schafer, E. Keck, and F. Kruck, Testosterone metabolism in human bone. Acta Endocrinol. 95, 258 – 264 (1980). 5. J. Vittek, K. Altman, G. G. Gordon, and A. L. Southren, The metabolism of 7 a 3H-testosterone by rat mandibular bone. Endocrinology 94, 325 – 329 (1974). 6. H.-R. Bruch, L. Wolf, R. Budde, G. Romalo, and H.-U. Schweikert, Androstenedione metabolism in cultured human osteoblast-like cells. J. Clin. Endocrinol. Metab. 75, 101 – 105 (1992). 7. B. L. Riggs, S. Khosla, and L. J. Melton III, A unitary model for involutional osteoporosis: Estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to the continuous phase of bone loss in aging men J. Bone Miner. Res. 13, 763 – 773 (1998). 8. Y. Nakano, I. Morimoto, O. Ishida, T. Fujihara, A. Mizokami, A. Tanimoto, N. Yanagihara, F. Izumi, and S. Eto, The receptor, metabolism and effects of androgen in osteoblastic MC3T3-E1 cells. Bone Miner. 26, 245 – 259 (1994). 9. C. H. Kasperk, J. E. Wergedal, J. R. Farley, T. A. Linkhart, R. T. Turner, and D. J. Baylink, Androgens directly stimulate proliferation of bone cells in vitro. Endocrinology 124, 1576 – 1578 (1989).

720 10. C. H. Kasperk, G. K. Wakley, T. Hierl, and R. Ziegler, Gonadal and adrenal androgens are potent regulators of human bone cell metabolism in vitro. J. Bone Miner. Res. 12, 464 – 471 (1997). 11. C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). 12. L. C. Hofbauer, K. C. Hicok, and S. Khosla, Effects of gonadal and adrenal androgens in a novel androgen-responsive human osteoblastic cell line. J. Cell. Biochem. 71, 96 – 108 (1998). 13. D. J. Benz, M. R. Haussler, M. A. Thomas, B. Speelman, and B. S. Komm, High-affinity androgen binding and androgen regulation of A1(I)-procollagen and transforming frowth factor-beta steady state messesnger ribonucleic acid levels in human osteoblast-like osteosarcoma cells. Endocrinology 128, 2723 – 2730 (1991). 14. T. Bellido, R. L. Jilka, B. F. Boyce, G. Girasole, H. Broxmeyer, S. A. Dalrymple, R. Murray, and S. C. Manolagas, Regulation of interleukin-6, osteoclastogenesis, and bone mass by androgens. J. Clin. Invest. 95, 2886 – 2895 (1995). 15. L. C. Hofbauer, R. M. Ten, and S. Khosla, The anti-androgen hydroxyflutamide and androgens inhibit interleukin-6 production by an androgen-responsive human osteoblastic cell line. J. Bone Miner. Res. 14, 1330 – 1337 (1999). 16. P. V. N. Bodine, B. L. Riggs, and T. C. Spelsberg, Regulation of c-fos expression and TGF- production by gonadal and adrenal androgens in normal human osteoblastic cells. J. Steroid. Biochem. Mol. Biol. 52, 149 – 158 (1995). 17. F. Gori, L. Hofbauer, C. A. Conover, and S. Khosla, Effects of androgens on the insulin-like growth factor system in an androgenresponsive human osteoblastic cell line. Endocrinology 140, 5579 – 5586 (1999). 18. A. W. Meikle, R. W. Dorchuck, B. A. Araneo, J. D. Stringham, T. G. Evans, S. L. Spruance, and R. A. Daynes, The presence of a dehydroepiandrosterone-specific receptor binding complex in murine T cells. J. Steroid Biochem. Mol. Biol. 42, 293 – 304 (1992). 19. R. T. Turner, , K. S. Hannon, L. M. Demers, J. Buchanan, and N. H. Bell, Differential effects of gonadal function on bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). 20. V. Coxam, B. M. Bowman, M. Mecham, C. M. Roth, M. A. Miller, and S. C. Miller, Effects of dihydrotestosterone alone and combined with estrogen on bone mineral density, bone growth, and formation rates in ovariectomized rats. Bone 19, 107 – 114 (1996). 21. D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Schot, and R. Bouillon, Bone and mineral metabolism in aged male rats: Short and long term effects of androgen deficiency. Endocrinology 130, 2906 – 2916 (1992). 22. D. Vanderschueren, E. Van Herck, R. De Coster, and R. Bouillon, Aromatization of androgens is important for skeletal maintenance of aged male rats. Calcif. Tissue. Int. 59, 179 – 183 (1996). 23. T. C. Durbridge, H. A. Morris, A. M. Parsons, L. H. Parkinson, R. J. Moore, S. Porter, A. G. Need, B. E. C. Nordin, and B. VernonRoberts, Progressive cancellous bone loss in rats after adrenalectomy and oophorectomy. Calcif. Tissue Int. 47, 383 – 387 (1990). 24. R. T. Turner, E. T. Lifrak, M. Beckner, G. K. Wakley, K. S. Hannon, and L. N. Parker, Dehydroepiandrosterone reduces cancellous bone osteopenia in ovariectomized rats. Am. J. Physiol. 258, E673 – E677 (1990). 25. H. Rico, M. Revilla, Hernandez, L. F. Villa, and M. Alverez del Buergo, Sex differences in the acquisition of total bone mineral mass peak assessed through dual-energy x-ray absorptiometry. Calcif. Tissue Int. 51, 251 – 254 (1992). 26. A. M. Fehily, R. J. Coles, W. D. Evans, and P. C. Elwood, Factors affecting bone density in young adults. Am. J. Clin. Nutr. 56, 579 – 586 (1992).

SUNDEEP KHOSLA 27. P. J. Kelly, L. Twomey, P. N. Sambrook, and J. A. Eisman, Sex differences in peak adult bone mineral density. J. Bone Miner. Res. 5, 1169 – 1175 (1990). 28. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C., Lubahn, D. B. and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 29. A. Morishima, M. M. Grumbach, E. R. Simpson, C. Fisher, and K. Qin, Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J. Clin. Endocrinol. Metab. 80, 3689 – 3698 (1995). 30. C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effect of testosterone and estradiol in a man with aromatase deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). 31. P. J. Kelly, N. A. Pocock, P. N. Sambrook, and J. A. Eisman, Dietary calcium, sex hormones, and bone mineral density in men. Br. Med. J. 300, 1361 – 1364 (1990). 32. S. Murphy, K.-t. Khaw, A. Cassidy, and J. E. Compston, Sex hormones and bone mineral density in elderly men. Bone Miner. 20, 133 – 140 (1993). 33. P. J. Drinka, J. Olson, S. Bauwens, S. K. Voeks, I. Carlson, and M. Wilson, Lack of association between free testosterone and bone density separate from age in elderly males. Calcif. Tissue Int. 52, 67 – 69 (1993). 34. D. E. Meier, E. S. Orwoll, E. J. Keenan, and R. M. Fagerstrom, Marked decline in trabecular bone mineral content in healthy men with age: Lack of association with sex steroid levels. J. Am. Geriatr. Soc. 35, 189 – 197 (1987). 35. S. Khosla, L. J. Melton III, E. J. Atkinson, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274, (1998). 36. C. W. Slemenda, C. Longcope, L. Zhou, S. L. Hui, M. Peacock, and C. C. Johnston, Sex steroids and bone mass in older men. Positive associations with serum estrogens and negative associations with androgens. J. Clin. Invest., 100, 1755 – 1759 (1997). 37. C. Slemenda, C. Longcope, M. Peacock, S. Hui, and C. C. Johnston, Sex steroids, bone mass, and bone loss. A prospective study of pre-, peri-, and postmenopausal women. J. Clin. Invest. 97, 14 – 21 (1996). 38. A. W. Meikle, R. A. Daynes, and B. A. Araneo, Adrenal androgen secretion and biologic effects. Endocrinol. Metab. Clin. N. Am. 20, 381 – 400 (1991). 39. M. Szathmari, J. Szucs, T. Feher, and I. Hollo, Dehydroepiandrosterone sulphate and bone mineral density. Osteoporos. Int. 4, 84 – 88 (1994). 40. S. Rozenberg, H. Ham, D. Bosson, A. Peretz, and C. Robyn, Age, steroids and bone mineral content. Maturitas 12, 137 – 143 (1990). 41. B. J. Davidson, B. L. Riggs, H. W. Wahner, and H. L. Judd, Endogenous cortisol and sex steroids in patients with osteoporotic spinal fractures. Obstet. Gynecol. 61, 275 – 278 (1983). 42. E. Barrett-Connor, D. Kritz-Silverstein, and S. L. Edelstein, A prospective study of dehydroepiandrosterone sulfate (DHEAS) and bone mineral density in older men and women. Am. J. Epidemiol. 137, 201 – 206 (1993). 43. J. P. Devogelaer, J. Crabbe, and C. N. De Deuxchaisnes, Bone mineral density in Addison’s disease: Evidence for an effect of adrenal androgens on bone mass. Br. Med. J. 294, 798 – 800, (1987). 44. C. H. Chesnut III, J. L. Ivey, H. E. Gruber, M. Matthews, W. B. Nelp, K. Sisom, and D. J. Baylink, Stanozolol in postmenopausal osteoporosis: Therapeutic efficacy and possible mechanisms of action. Metabolism 32, 571 – 580 (1983). 45. A. G. Need, H. A. Morris, T. F. Hartley, M. Horowitz, and B. E. C. Nordin, Effects of nandrolone decanoate on forearm mineral

CHAPTER 76 Androgens and Androgenic Progestins

46. 47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

density and calcium metabolism in osteoporotic postmenopausal women. Calcif. Tissue Int. 41, 7 – 10 (1987). J. S. Tenover, Effects of testosterone supplementation in the aging male. J. Clin. Endocrinol. Metab. 75, 1092 – 1098 (1992). F. H. Anderson, R. M. Francis, R. T. Peaston, and H. J. Wastell, Androgen supplementation in eugonadal men with osteoporosis: Effects of six months treatment on markers of bone formation and resorption. J. Bone Miner. Res. 12, 472 – 478 (1997). B. L. Riggs, J. Jowsey, R. S. Goldsmith, P. J. Kelly, D. L. Hoffman, and C. D. Arnaud, Short- and long-term effects of estrogen and synthetic anabolic hormone in postmenopausal osteoporosis. J. Clin. Invest. 51, 1659 – 1663 (1972). L. G. Raisz, B. Wiita, A. Artis, A. Bowen, S. Schwartz, M. Trahiotis, K. Shoukri, and J. Smith, Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. J. Clin. Endocrinol. Metab. 81, 37 – 43 (1996). C. Hassager, L. T. Jensen, J. Podenphant, B. J. Riis, and C. Christiansen, The carboxy terminal propeptide of type I collagen in serum as a marker of bone formation: The effect of nandrolone decanoate and female sex hormones. Metabolism 39, 1167 – 1169 (1990). C. Hassager, J. T. Jensen, J. S. Johansen, B. J. Riis, J. Melkko, J. Podenphant, C. Christiansen, and J. Risteli, Collagen synthesis in postmenopausal women during therapy with anabolic steroid and female sex hormones. Metabolism 40, 205 – 208 (1991). J. S. Finkelstein, A. Klibanski, R. M. Neer, S. H. Doppelt, D. I. Rosenthal, G. V. Segre, and W. F. Crowley, Increases in bone density during treatment of men with idiopathic hypogonadotropic hypogonadism. J. Clin. Endocrinol. Metab. 69, 776 – 783 (1989). J. P. Devogelaer, S. D. Cooman, and C. Nagant de Deuxchaisnes, Low bone mass in hypogonadal males. Effect of testosterone substitution therapy, a densitometric study. Maturitas 15, 17 – 23 (1992). F. H. Anderson, R. M. Francis, and K. Faulkner, Androgen supplementation in eugonadal men with osteoporosis — effects of 6 months of treatment on bone mineral density and cardiovascular risk factors. Bone 18, 171 – 177 (1996). P. J. Snyder, H. Peachey, P. Hannoush, J. A. Berlin, L. Loh, J. H. Holmes, A. Dlewati, J. Staley, J. Santanna, S. C. Kapoor, M. F. Attie, J. G. Haddad, Jr., and B. L. Strom, Effect of testosterone treatment on bone mineral density in men over 65 years of age. J. Clin. Endocrinol. Metab. 84, 1966 – 1972 (1999). I. R. Reid, D. J. Wattie, M. C. Evans, and J. P. Stapleton, Testosterone therapy in glucocorticoid-treated men. Arch. Intern. Med. 156, 1173 – 1177 (1996). C. H. Chesnut, III, W. B. Nelp, D. J. Baylink, and J. D. Denney, Effect of methandrostenolone on postmenopausal bone wasting as assessed by changes in total bone mineral mass. Metabolism 26, 267 – 277 (1977). P. Geusens and J. Dequeker, Long term effect of nandrolone decanoate, 1-hydroxyvitamin D3 or intermittent calcium infusion therapy on bone mineral content, bone remodeling and fracture rate in symptomatic osteoporosis: A double-blind controlled study. Bone Miner. 1, 347 – 357 (1986). C. Hassager, J. Podenphant, B. J. Riis, J. S. Johansen, J. Jensen, and C. Christiansen, Changes in soft tissue body composition and plasma lipid metabolism during nandrolone decanoate therapy in postmenopausal women. Metabolism 38, 238 – 242 (1989). A. G. Need, M. Horowitz, C. J. Walker, C. J. Chatterton, I. C. Chapman, and B. E. C. Nordin, Cross-over study of fat-corrected forearm mineral content during nandrolone decanoate therapy for osteoporosis. Bone 10, 3 – 6 (1989). J. S. Johansen, C. Hassager, J. Podenphant, B. Q. Rus , D. Hartwell, K. Thomsen, and C. Christiansen, Treatment of postmenopausal osteoporosis. Is the anabolic steroid nandrolone decanoate a candidate? Bone Miner. 6, 77 – 86 (1989).

721 62. C. Gennari, D. AgnusDei, S. Gonnelli, and P. Nardi, Effects of nandrolone decanoate therapy on bone mass and calcium metabolism in women with established post-menopausal osteoporosis: A doubleblind placebo-controlled study. Maturitas 11, 187 – 197 (1989). 63. J. E. Nestler, C. D. Barlascini, J. N. Clore, and W. G. Blackard, Dehydroepiandrosterone reduces serum low density lipoprotein levels and body fat but does not alter insulin sensitivity in normal men. J. Clin. Endocrinol. Metab. 78, 1360 – 1367 (1994). 64. A. J. Morales, J. J. Nolan, J. C. Nelson, and S. S. C. Yen, Effects of replacement dose of dehydroepiandrosterone in men and women of advancing age. J. Clin. Endocrinol. Metab. 78, 1360 – 1367 (1994). 65. R. F. Van Vollenhoven, Increased bone mineral density in patients with SLE treated with DHEA. Arthritis Rheum. 39, S292 (1996). 66. F. Labrie, P. Diamond, L. Cusan, J. L. Gomez, A. Belanger, and B. Candas, Effect of 12-month dehydroepiandrosterone replacement therapy on bone, vagina, and endimetrium in postmenopausal women. J. Clin. Endocrinol. Metab. 82, 3498 – 3505 (1997). 67. C. Hassager, J. Podenphant, B. J. Riis, J. S. Johansen, J. Jensen, and C. Christiansen, Changes in soft tissue body composition and plasma lipid metabolism during nandrolone decanoate therapy in postmenopausal osteoporotic women. Metabolism 38, 238 – 342 (1989). 68. J. F. Aloia, A. Kapoor, and A. Vawani, Changes in body composition following therapy of osteoporosis with methandrostenolone. Metabolism 30, 1076 – 1079 (1981). 69. J. L. Shifren, G. D. Braunstein, J. A. Simon, P. R. Casson, J. E. Buster, G. P. Redmond, R. E. Burki, E. S. Ginsburg, R. C. Rosen, S. R. Leiblum, K. E. Caramelli, and N. A. Mazer, Transdermal testosterone treatment in women with impaired sexual function after oophorectomy. N. Eng. J. Med. 343, 682 – 688 (2000). 70. H. M.. Tagget, D. Applebaum-Bowden, and S. Haffner, Reduction in high density lipoproteins by anabolic steroid (stanozolol) therapy for postmenopausal osteoporosis. Metabolism 31, 1147 – 1152 (1982). 71. J. K. Allen and I. S. Frazer, Cholesterol, high density lipoprotein and danazol. J. Clin. Endocrinol. Metab. 53, 149 – 153 (1981). 72. M. C. Cheung, J. J. Albers, and P. W. Wahl, High density lipoproteins during hypolipidemic therapy. A comparative study of four drugs. Atherosclerosis 35, 215 – 228 (1980). 73. H. A. Haupt and G. D. Rovere, Anabolic steroids: A review of the literature. Am. J. Sports Med. 12, 469 – 484 (1984). 74. E. Eriksen, D. Colvard, N. Berg, M. Graham, K. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 87 (1988). 75. L. L. Wei, M. W. Leach, R. S. Miner, and L. M. Demers, Evidence for progesterone receptors in human osteoblast-like cells. Biochem. Biophys. Res. Commun. 195, 525 – 532 (1993). 76. P. MacNamara, C. O. O’Shaughnessy, P. Manduca, and H. C. Loughrey, Progesterone receptors are expressed in human osteoblast-like cell lines and in primary human osteoblast cultures. Calcif. Tissue Int. 57, 436 – 441 (1995). 77. P. MacNamara and H. C. Loughrey, Progesterone receptor A and B isoform expression in human osteoblasts. Calcif. Tissue Int. 63, 39 – 46 (1998). 78. P. MacNamara and H. C. Loughrey, Messenger RNA transcripts coding for progesterone receptor A and B isoforms are expressed in human osteoblast cells. Biochem. Soc. Transactions. 26, 1 – 7 (1998). 79. J. M. Pensler, J. A. Radosevich, R. Higbee, and C. B. Langman, Osteoclasts isolated from membranous bone in children exhibit nuclear estrogen and progesterone receptors. J. Bone Miner. Res. 5, 797 – 802 (1990). 80. E. Canalis and L. G. Raisz, Effects of sex steroids on bone collagen synthesis in vitro. Calcif. Tissue Res. 25, 105 – 110 (1978).

722 81. M. C. Slootweg, A. G. H. Ederveen, L. P. C. Schot, W. G. E. J. Schoonen, and H. J. Kloosterboer, Oesterogen and progesterone synergistically stimulate human and rat osteoblast proliferation. J. Endocrinol. 133, R5 – R8 (1992). 82. F. Tremollieres, D. Strong, D. Baylink, and S. Mohan, Progesterone and promegestone stimulate human bone cell proliferation and insulin-like growth factor-2 production. Acta Endocr. 126, 329 – 337 (1992). 83. B. A. A. Scheven, C. A. Damen, N. J. Hamilton, H. J. J. Verhaar, and S. A. Duursma, Stimulatory effects of estrogen and progesterone on proliferation and differentiation of normal human osteoblast-like cell in vitro. Biochem. Biophys. Res. Commun. 186, 54 – 60 (1992). 84. H. J. J. Verhaar, C. A. Damen, S. A. Duursma, and B. A. A. Scheven, A comparison of action of progestins and estrogen on the growth and differentiation of normal adult human osteoblast-like cells in vitro. Bone 15, 307 – 311 (1994). 85. D. L. Manzi, C. C. Pilbeam, and L. G. Raisz, The anabolic effects of progesterone on fetal rat calvaria in tissue culture. J. Soc. Gynecol. Invest. 1, 302 – 309 (1994). 86. Y. Ishida, I. Tertinegg, and J. N. Heersche, Progesterone and dexamethasone stimulate proliferation and differentiation of osteoprogenitors for adipocytes and macrophages in cell populations derived from adult rat vertebrae. J. Bone Miner. Res. 11, 921 – 930 (1996). 87. Y. Ishida and J. N. Heersche, Progesterone stimulates proliferation and differentiation of osteoprogenitor cells in bone cell populations derived from adult female but not from adult male rats. Bone 20, 17 – 25 (1997). 88. Y. Ishida, C. G. Bellows, I. Tertinegg, and J. N. M. Heersche, Progesterone-mediated stimulation of osteoprogenitor proliferation and differentiation in cell populations derived from adult or fetal rat bone tissue depends on the serum component of the culture media. Osteoporosis Intl. 7, 323 – 330 (1997). 89. Y. Ishida and J. N. Heersche, Progesterone- and dexamethasone-dependent osteoprogenitors in bone cell populations derived from rat vertebrae are different and distinct. Endocrinology 140, 3210 – 3218 (1999). 90. C. C. Burnett and A. H. Reddi, Influence of estrogen and progesterone on matrix-induced endochondral bone formation. Calcif. Tissue Int. 35, 609 – 614 (1983). 91. G. R. Snow and C. Anderson, The effects of continuous progesterone treatment on cortical bone remodeling activity in beagles. Calcif. Tissue Int. 37, 282 – 286 (1985). 92. G. R. Snow and C. Anderson, The effects of 17-estradiol and progesterone on trabecular bone remodeling in oophorectomized dogs. Calcif. Tissue Int. 39, 198 – 205 (1985). 93. E. I. Barengolts, H. F. Gajardo, T. J. Rosol, J. J. D’Anza, M. Pena, J. Botsis, and S. C. Kukreja, Effects of progesterone on postovariectomy bone loss in aged rats. J. Bone Miner. Res. 5, 1143 – 1147 (1990). 94. E. I. Barengolts, T. Kouznetsova, A. Segalene, P. Lathon, C. Odvina, S. C. Kukreja, and T. G. Unterman, Effects of progesterone on serum levels of IGF-I and on femur IGF-I mRNA in ovariectomized rats. J. Bone Miner. Res. 11, 1406 – 1412 (1996). 95. B. M. Bowman and S. C. Miller, Elevated progesterone during pseudopregnancy may prevent bone loss associated with low estrogen. J. Bone Miner. Res. 11, 15 – 21 (1996). 96. J. C. Prior, Y. M. Vigna, M. T. Schechter, and A. E. Burgess, Spinal bone loss and ovulatory disturbances. N. Engl. J. Med. 323, 1221 – 1227 (1990). 97. J. C. Prior, Y. M. Vigna, S. L. Barr, C. Rexworthy, and B. C. Lentle, Cyclic medroxyprogesterone treatment increases bone density: A controlled trial in active women with menstrual cycle disturbances. Am. J. Med. 96, 521 – 530 (1994). 98. K. Waller, J. Reim, L. Fenster, S. H. Swan, B. Brumback, G. C. Windham, B. Lasley, B. Ettinger, and R. Marcus, Bone mass and sub-

SUNDEEP KHOSLA

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115. 116.

tle abnormalities in ovulatory function in healthy women. J. Clin. Endocrinol. Metab. 81, 663 – 668 (1996). M. J. De Souza, B. E. Miller, L. C. Sequenzia, A. A. Luciano, S. Ulreich, S. Stier, K. Prestwood, and B. L. Lasley, Bone health is not affected by luteal phase abnormalities and decreased ovarian progesterone production in female runners. J. Clin. Endocrinol. Metab. 82, 2867 – 2876 (1997). F. P. Mandel, B. J. Davidson, Y. Erlik, H. L. Judd, and D. R. Meldrum, Effects of progestins on bone metabolism in postmenopausal women. J. Reprod. Med. 27 (8 Suppl.), 511 – 514 (1982). R. A. Lobo, W. McCormick, F. Singer, and S. Roy, Depo-medroxyprogesterone acetate compared with conjugated estrogens for the treatment of postmenopausal women. Obstet. Gynecol. 63, 1 – (1984). S. G. McNeeley, J. S. Schinfeld, T. G. Stovall, F. W. Ling, and B. H. Buxton, Prevention of osteoporosis by medroxyprogesterone acetate in postmenopausal women. Int. J. Gynecol. Obstet. 34, 253 – 256 (1991). J. C. Gallagher, W. T. Kable, and D. Goldgar, The effect of progestin therapy on cortical and trabecular bone: Comparison with estrogen. Am. J. Med. 90, 171 – 178 (1991). A. Grey, T. Cundy, M. Evans, and I. Reid, Medroxypro-gesterone acetate enhances the spinal bone mineral density responses to oestrogen in late post-menopausal women. Clin. Endocrinol. 44, 293 – 296 (1996). The Writing Group for the PEPI Tria. Effects of hormone therapy on bone mineral density. Results from the postmenopausal estrogen/progestin interventions (PEPI) trial. JAMA 276, 1389 – 1396 (1996). H. I. Abdalla, D. M. Hart, R. Lindsay, I. Leggate, and A. Hooke, Prevention of bone mineral loss in postmenopausal women by norethisterone. Obstet. Gynecol. 66, 789 – 792 (1985). C. Christiansen, M. S. Christensen, and I. Transbol, Bone mass in postmenopausal women after withdrawal of oestrogen/ gestagen replacement therapy. Lancet 1, 459 – 461 (1981). G. F. Jensen, C. Christiansen, and I. Transbol, Treatment of postmenopausal osteoporosis. A controlled therapeutic trial comparing oestrogen/gestagen, 1,25-dihydroxyvitamin D3 and calcium. Clin. Endocrinol. 16, 515 – 524 (1982). N. Munk-Jensen, S. P. Nielsen, E. B. Obel, and P. B. Eriksen, Reversal of postmenopausal vertebral bone loss by oestrogen and progestogen: A double blind placebo controlled study. Br. Med. J. 296, 1150 – 1152 (1988). C. Christiansen, L. Nilas, B. J. Riis, P. Rodbro, and L. Deftos, Uncoupling of bone formation and resorption by combined oestrogen and progestagen therapy in postmenopausal osteoporosis. Lancet 2, 800 – 801 (1985). B. J. Riis, C. Christiansen, J. S. Johansen, and J. Jacobson, Is it possible to prevent the bone loss in young women treated with LH – RH agonists. J. Clin. Endocrinol. Metab. 70, 920 – 924 (1990). L. Speroff, J. Rowan, J. Symons, H. Genant, and W. Wilborn, The comparative effect on bone density, endometrium, and lipids of continuous hormones as replacement therapy (CHART Study). JAMA 276, 1397 – 1403 (1996). E. Barrett-Connor, S. Slone, G. Greendale, D. Kritz-Silverstein, M. Espeland, S. R. Johnson, M. Waclawiw, and S. E. Fineberg, The postmenopausal estrogen/progestin interventions study: Primary outcomes in adherent women. Maturitas 27, 261 – 274 (1997). C. Schairer, J. Lubin, R. Troisi, S. Sturgeon, L. Brinton, and R. Hoover, Menopausal estrogen and estrogen-progestin replacement therapy and breast cancer risk. JAMA 283, 485 – 491 (2000). P. Albertazzi, R. Di Micco, and E. Zanardi, Tibolone: a review. Maturitas 30, 295 – 305 (1998). J. Rymer, M. G. Chapman, and I. Fogelman, Effect of tibolone on postmenopausal bone loss. Osteoporosis Int. 4, 314 – 319 (1994).

CHAPTER 76 Androgens and Androgenic Progestins 117. G. P. Lyritis, S. Karpathios, K. Basdekis, O. Grigoriou, T. Katostaras, I. Paspati, G. Siampalioti, and P. G. Lyritis, Prevention of post-oophorectomy bone loss with tibolone. Maturitas 22, 247 –253 (1995). 118. N. H. Bjarnason, K. Bjarnason, J. Haarbo, C. Rosenquist, and C. Christiansen, Tibolone: Prevention of bone loss in late postmeopausal women. J. Clin. Endocrinol. Metab. 81, 2419 – 2422 (1996). 119. B. Berning, C. V. Kuijk, J. W. Kuiper, H. J. Bennink, P. M. Kicovic, and B. C. Fauser, Effects of two doses of tibolone on trabecular and cortical bone loss in early postmenopausal women: A two-year randomized, placebo-controlled study. Bone 19, 395 – 399 (1996). 120. K. Lippuner, W. Haenggi, M. H. Birkhaeuser, J. P. Casez, and P. Jaeger, Prevention of postmenopausal bone loss using tibolone or conventional peroral or transdermal hormone replacement therapy

723 with 17beta-estradiol and dyhydrogesterone. J. Bone Miner. Res. 12, 806 – 812 (1997). 121. N. H. Bjarnason, K. Bjarnason, C. Hassager, and C. Christiansen, The response in spinal bone mass to tibolone treatment is related to bone turnover in elderly women. Bone 20, 151 – 155 (1997). 122. J. Studd, I. Arnala, P. M. Kicovic, D. Zamblera, H. Kroger, and N. Holland, A randomized study of tibolone on bone mineral density in osteoporotic postmenopausal women with previous fractures. Obstet. Gynecol. 92, 574 – 579 (1998). 123. S. A. Beardsworth, C. E. Kearney, and D. W. Purdie, Prevention of postmenopausal bone loss at lumbar spine and upper femir with tibolone: A two year randomised controlled trial. Br. J. Obstet. Gynaecol. 106, 678 – 683 (1999).

CHAPTER 77

Treatment with PTH Peptides LIS MOSEKILDE JONATHAN REEVE

Department of Cell Biology, Institute of Anatomy, University of Aarhus, DK-8000, Aarhus C, Denmark Department of Medicine, University of Cambridge, Cambridge CB2 2QQ, United Kingdom

I. Rationale for the Treatment of Osteoporosis with Parathyroid Peptides or Other Anabolic Agents II. Historical Background III. Biological Basis of the Actions of Parathyroid Hormones

IV. Effects of PTH in Animal Models V. Clinical Studies VI. Conclusions References

I. RATIONALE FOR THE TREATMENT OF OSTEOPOROSIS WITH PARATHYROID PEPTIDES OR OTHER ANABOLIC AGENTS

of previous bone resorption. Bone can accumulate by continuous accretion only on periosteal surfaces, and this capacity declines to very low levels in adults [2]. Therefore, if bone resorption is reduced, eventually fewer sites may become available for the formation of thickened packets of new bone. Although the balance between formation and resorption may remain at a more favorable level at the individual bone remodeling site, antiresorbing drugs also reduce the number of sites, with the commonly observed result that treatment begins with a transient rise in bone density, followed by a plateau and maybe eventually a slow decline. Because the majority of patients present with fractures only after they have lost around half of the cancellous bone in the spine [3] and a third or more elsewhere, this plateau prevents most patients treated with antiresorbing agents from regaining normal bone mass. The implication from studies relating bone density to fracture risk is that such patients, either during or after treatment, remain at considerably increased risk of new fractures in comparison with their peers who have never fractured. The anabolic actions of parathyroid hormone (PTH) on bone have been reviewed previously [4]. The biology of

For many diseases the treatment is to remove the cause. Although the causes of osteoporosis are still the subject of intensive investigation, it is generally agreed that bone resorption in excess of the capacity of the osteoblasts to replace bone previously removed is important at a critical stage in pathogenesis. Hence much effort has been devoted to developing treatments that reduce bone resorption. However, there are sound reasons for expecting only moderately good results from any treatment regimen that depends exclusively on controlling osteoclasts. Once adulthood is reached, the skeleton loses the capacity to form bone de novo in regions formerly occupied by soft tissue. Bone is strengthened (e.g., in response to mechanical stress) by adding packets of new bone that are thicker than those formed previously [1]. Internally, on endosteal and trabecular surfaces, packets of new bone are formed at sites

OSTEOPOROSIS, SECOND EDITION VOLUME 2

725

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

726

MOSEKILDE AND REEVE

PTH is fully discussed in Chapter 7. The emphasis in this chapter will be to examine the future potential for harnessing the anabolic effects of PTH so as to develop effective new treatments for osteoporosis.

II. HISTORICAL BACKGROUND A. Animal Studies with Parathyroid Extract The first indications that parathyroid hormone (PTH) may have an anabolic action on the skeleton emerged over 60 years ago [5 – 9]. Studies in rats using moderate doses of a parathyroid extract (PTE), which was extensively used until the modern era and available from Eli Lilly as “Parathormone,” showed increases in cancellous bone density with various experimental regimens. Investigators following up the results of Bauer et al. [5] were intrigued by the sequential development of Osteitis Fibrosa and “Marble Bone Disease” [6,8], which was not recognized in clinical hyperparathyroidism [6] until some 40 years later and then only as a rarity [10,11]. R. B. Burrows [6] concluded in 1938 that the then competing theories of the Albright school (that PTH had its main action on phosphate transport by the renal tubular cell) and other theories which emphasized a direct bony action of PTH on osteoclasts [8,12,13] or through increased acid production [14] were incapable of explain-

FIGURE 1

ing the experimental results. He postulated that the anabolic actions of PTH were the result of the actions of a second biochemical messenger and concluded that it would be necessary to obtain pure PTH to discover whether this second messenger was an antibody-like substance directed at impurities in PTE, or part of some other intrinsic biological mechanism which for unknown reasons did not routinely find expression in clinical hyperparathyroidism. Later, the low levels of purity of all parathyroid extracts used at that time raised further doubts about the nature of the anabolic constituent(s) and calcitonin was considered a possible candidate. Kalu, Walker, and their colleagues [15, 16] were able to show using purified PTH that it was the PTH which was anabolic to bone. Their work opened the way two decades ago to the first clinical studies with parathyroid peptides in osteoporosis [17 – 19]. At the same time as these first clinical studies were being concluded, the animal studies of Tam et al. [20] and Podbesek et al. [21] showed that the anabolic effects of PTH on bone depended on discontinuous exposures to high blood levels of PTH peptides rather than continuous exposure to moderately increased blood levels such as occurs in human hyperparathyroidism (Fig. 1). While this chapter is written as though the development of PTH as a therapy occurred in the usual sequence of animal studies preceding Phase I clinical studies, as a matter of historical fact there was little other animal work

Bone apposition rate () and serum calcium levels () in rats receiving various doses of bPTH(1 – 84) by intermittent injection or continuous infusion. Reprinted with permission from Tam et al. [20].

727

CHAPTER 77 Treatment with PTH Peptides

documenting the effects of administering hPTH (1 – 34) in therapeutic doses to animals at the time it was first administered to patients. Animal studies (such as those of Podbesek et al. [21,22] were undertaken when safety and comfort considerations (in this case, repeated exposures to ionizing radiations that would have been excessive in man as well as subcutaneous implantation of minipumps containing PTH peptide) precluded the same studies being performed in patients. The contemporary, accepted ethical justification for this approach was that PTH is a hormone, not a conventional drug, and its potentially adverse effects could be closely monitored, with effective remedies available if necessary.

III. BIOLOGICAL BASIS OF THE ACTIONS OF PARATHYROID PEPTIDES PTH exerts many of its most important biologic actions on bone through the PTH1 receptor it shares with the PTHrelated protein (PTHrP). The effects on bone of activating the recently described PTH2 receptor [23] remain to be elucidated. PTH1 receptors are not found in osteoclasts but are abundant in at least some of the stromal cells that generate the osteoblast lineage as well as in the growth plate in young animals. In different circumstances, the intracellular consequences of interactions of PTH with its receptor can modulate concentrations of cyclic adenosine monophosphate (cAMP) and cAMP-dependent protein kinase-A (PKA). At high concentrations, the phospholipase-c (PLC) system can also be activated [24]. This in turn modulates the intracellular concentrations of calcium ions, diacyl glycerol, and hydrogen ions. Increased osteoclastic resorption driven by PTH is believed to be a consequence of secondary signaling, probably involving communication of the cells lining quiescent bone surfaces or other stromal cells with osteoclast precursors. It now seems likely that cells of this lineage may generate the specific osteoclast inhibitor, osteoprotegerin (OPG) [25], which acts in vivo as a decoy receptor for the osteoclast differentiation factor (ODF) [26,27] (there are multiple acronyms: ODF/OPGL/TRANCE/RANKL). In a preliminary report, hPTH [1 – 38] was shown to inhibit OPG synthesis in vitro [28], which has the potential to explain why under some circumstances PTH exposure leads to osteoclast differentiation. Access of osteoclasts to mineralized bone may require both retraction of lining cells and digestion of the thin protective layer of unmineralized osteoidlike material normally interposed between lining cells and mineralized bone. This digestion is accomplished by metalloproteinases, such as collagenase, that are secreted by the lining cells themselves [29]. Further events secondary to activation of the PTH receptor with potential to explain its anabolic actions have been

the focus of increasing attention. These include recruitment to the osteoblastic phenotype, prevention of programmed cell death (apoptosis) of osteoblasts, and augmentation of the capacity of osteoblasts to form new bone. Dobnig and Turner [30] showed that injections of PTH leading to intermittent stimulation lead rapidly to the accumulation of osteoblasts on previously quiescent cancellous rat bone surfaces, without evidence of cell division. They concluded that the previously quiescent lining cells had reverted to their previous osteoblastic phenotype, a conclusion that was consistent with a careful cellular ultrastructural study by Leaffer et al. [31]. In a human study, Hodsman and Steer found further evidence of very rapid osteoblast accumulation in response to PTH injections [32]. More recent work has centered on the duration of action of osteoblasts. Jilka et al. have found evidence that PTH injections reduce the rate of osteoblast (and osteocyte) apoptosis both in vivo and in vitro [33], which may explain the action of PTH therapy to increase the amount of bone formed at individual bone (re)modeling sites [34]. Canalis et al. [35] showed that continuous treatment with PTH inhibited, whereas transient treatment stimulated, calvarial collagen synthesis; and the stimulatory effect was mediated by local production of insulin-like growth factor-I (IGF-I) [36,37]. Concurrently McCarthy et al. showed that PTH enhanced local IGF-I synthesis by increasing IGF-I transcripts [37]. Watson et al. [38] found that cancellous osteoblasts from PTH-treated OVX rats showed twice the IGF-I mRNA abundance seen in OVX controls. The IGFs might act in association with osteoblasts to prolong their bone synthesizing phenotype and augment the formation of packets of new bone with higher than average wall width. The IGFs are increasingly thought of as inhibitors of apoptosis [39], so part of this effect might be through increasing osteoblast survival. Other actions of PTH include modulation of transforming growth factor -1 (TGF-1) [40]. Osteogenic cells also produce prostaglandins in response to PTH; these can stimulate bone formation in vivo [41]. An alternative activator of the PTH receptor, which might in principle have naturally anabolic effects if released transiently, are the locally produced paracrine regulators derived from the PTH-related peptide (PTHrP) gene [35], since nontransformed osteoblasts produce PTHrP in vitro [42,43].

IV. EFFECTS OF PTH IN ANIMAL MODELS A. Animal Models Used Before accepting any agent for clinical osteoporosis trials, the FDA requires that the agent has proven efficacy in a small animal model and also in a larger animal model with

728

MOSEKILDE AND REEVE

known intracortical bone remodeling. The efficacy of the agent should be proven by measurements of bone mass and also of the biomechanical competence of bone. The majority of animal work has employed PTH as its [1 – 34] fragment; so in what follows PTH can be taken to refer to hPTH [1 – 34] unless stated otherwise. 1. SMALL ANIMALS LACKING INTRACORTICAL BONE REMODELING (RATS) It is generally accepted that rats should be used as the animal model for screening new agents for osteoporosis therapy [44,45]. Although Baron et al. [46], Vignery and Baron [47], and Tran Van et al. [48,49] showed several years ago that rats do have trabecular bone remodeling similar to that in humans, it is only recently that rats have been more fully accepted in bone research. However, the advantages of using a rat model for screening are numerous: the study can be conducted under very standardized conditions; rats have lamellar bone and rats have trabecular bone remodeling (Fig. 2). The rat skeleton reacts to hormonal and mechanical stimuli in a manner that is similar to that in the human skeleton. Furthermore, ovariectomy and aging produce alterations in the trabecular network identical to those seen during menopause and aging in the human skeleton (Figs. 3A and 3B). The disadvantages of using a rat model are summarized as follows:

FIGURE 2

1. Rats have a different loading pattern. 2. Rats have open epiphyses to the age 12 – 24 months. 3. Rats have pronounced modeling throughout life, which implies that there is “drift” through space in the individual cortical and trabecular elements. 4. Rats have very little intracortical bone remodeling. These advantages and disadvantages should be borne in mind when considering the description of the effect of PTH in different rat models (see below). Furthermore, as most agents used for the treatment of osteoporosis or for osteoporosis prevention (antiresorptive and anabolic agents) exert their main influence through the remodeling process (including Haversian remodeling of the cortices), a larger animal model is also needed in the preclinical testing procedure. There is still no agreement concerning which large animal model to use. Alternatives include primates, minipigs, dogs, ferrets, and sheep. Both dogs [21] primates [50] and sheep [51] have been used in studies of PTH peptides. 2. DIFFERENT MODELS USED FOR DETERMINING EFFECTS OF PTH AND ITS BIOACTIVE PEPTIDE PRODUCTS THE

a. Intact Rats In 1982, Hefti et al., showed that PTH was capable of increasing whole body calcium and skeletal mass in normal and osteoporotic rats [52]. Shortly afterward Gunness-Hey and Hock [53] showed that human

Scanning electron microscopy (SEM) photograph from vertebral body of a young, intact rat. Deep osteoclastic resorption lacunae are demonstrated at the remodeling site on a trabecular structure (original magnification, 700). Reprinted with permission from Mosekilde, Endocrinology 129 (1), 421 – 428 (1991).

729

CHAPTER 77 Treatment with PTH Peptides

FIGURE 3

SEM photographs showing osteoclastic perforation of trabecular structure. (A) A human vertebral body from a 58-year-old woman (original magnification, 370). (B) A vertebral body from an aged, ovariectomized rat (original magnification,  370).

synthetic PTH increased trabecular bone mass in the distal femur of intact rats. Their study also indicated that PTH had a concomitant anabolic effect on cortical femoral bone. These promising studies on intact animals were soon followed by studies in “manipulated” animal models. b. Manipulated Models The anabolic effect of PTH was demonstrated in 1988 in both ovariectomized and or-

chidectomized rats, showing that the anabolic response to PTH was not dependent on gonadal hormones [54]. In the same year, Gunness-Hey et al. showed that this anabolic effect was not augmented by 1,25-dihydroxyvitamin D3 [55], and in 1990 Hock and Fonseca used a hypophysectomized rat model to demonstrate that the anabolic effect of synthetic PTH depended on growth hormone (GH) [56]. By using these specifically manipulated in vivo models, a large

730 body of information was accumulated concerning the action of PTH on bone. c. The Ovariectomized Rat Model At the same time, bone research had also been concerned with how to obtain the optimal small animal model for osteoporosis research. This had been tackled predominantly by the use of histomorphometric techniques. In 1986, Wronski et al. gave a very detailed description of the histological evidence for osteopenia and increased bone turnover in ovariectomized rats [57], followed by studies describing the time course of osteopenia at different skeletal sites in the ovariectomized rat model [58 – 60]. This osteopenic rat model now formed the basis for testing not only antiresorptive agents [61,62], but also the anabolic agent PTH [63 – 70]. By using the ovariectomized rat model, Hori et al. demonstrated that pulsatile administration of human PTH prevented the development of osteopenia induced by ovariectomy [64]. Tada et al. showed in 1990 that PTH stimulated bone formation independent of bone resorption (suggesting so-called “uncoupling” of the two processes) and that it restored cancellous and cortical bone volumes in osteopenic rats [65]. One year later, Takahashi et al. demonstrated again that PTH raised bone formation activity and greatly minimized the occurrence of osteopenia in lumbar vertebrae in ovariectomized rats [66]. Liu and Kalu [67] and Liu et al. [68] used the ovariectomized rat model in both their prevention study [67] and their intervention study [68]. In the former study they showed that when PTH administration to rats was started immediately following ovariectomy it prevented bone loss due to ovarian hormone deficiency. In the latter study, they were able to show that PTH could substantially augment bone mass after loss due to ovarian hormone deficiency had already occurred. These studies provided a large body of information about the action of PTH on the nonosteopenic and osteopenic rat skeleton. The studies were all based on bone histomorphometry and measurement of bone mass. 3. LARGE ANIMAL MODELS USED FOR DETERMINING EFFECTS OF PTH AND ITS BIOACTIVE PEPTIDE PRODUCTS THE

A potentially negative side effect of intermittent PTH therapy in the treatment of osteoporosis is loss of cortical bone concomitantly with a gain in cancellous bone. In order to address this issue, a larger animal model with proven intracortical bone remodeling is needed. Such a study has recently been conducted in ovariectomized, young adult cynomolgus monkeys [50]. In this study, hPTH (1 – 34) was given subcutaneously, 10 g/kg and 3 days per week for 6 months. The study had multiple end points: bone mass, strength, and turnover in both the axial and the appendicular skeleton. This study showed that PTH treatment led to no change in whole-body bone mass, but a 6.7% increase in

MOSEKILDE AND REEVE

spinal areal bone mineral density. Increases in bone strength were observed in both axial (vertebral) and appendicular (femoral neck) skeletal sites. In conclusion, these results indicated that PTH therapy in the large animal model (cynomolgus monkey) resulted in a net gain of spinal and appendicular cancellous bone with no adverse effect on cortical bone. This very thorough study in a large animal model has, for the first time, shown that PTH treatment increases cancellous bone mass without any cortical steal phenomenon. It also showed that the increase in bone mass leads to a significant increase in bone strength. 4. EFFECT OF DIFFERENT PTH TREATMENT SCHEDULES a. Continuous vs Intermittent Many of the in vivo studies had indicated that only PTH given intermittently had an anabolic effect on bone. However, as early as 1983, Podbesek et al. had shown in a study with adult greyhounds that the increased osteoblastic activity induced in trabecular bone by a daily injection regime was dependent on the noncontinuous nature of the PTH stimulus [21]. Also, Tam et al. had shown in their rat model the differential effects of intermittent and continuous administration of PTH [20] (Fig. 1). These studies very clearly showed that the anabolic effect was dependent on intermittent dosing. Recently, a study by Morley et al. [71] has confirmed the necessity of intermittent treatment. In their study, ovariectomized rats were treated with prolonged low-dose infusion (delivered by Alzet minipumps) without there being any effect on cancellous bone mass or bone formation. The conclusion from this study was, accordingly, that there is an absolute requirement for intermittent dosing in order to achieve osteogenic action of hPTH on bone. Discontinuation of PTH treatment has been shown to cause loss of newly formed bone, due to an inhibition of PTH-stimulated bone formation and an initial transient phase of increased bone resorption [69]. Therefore, the maintenance of the anabolic effect of PTH seemed dependent in this study upon daily administration of the hormone [69]. It should be noted, though, that in 1998 Okimoto et al. could show that in rats an injection frequency of only one per week effectively stimulated bone formation in both trabecular and cortical bone, leading to positive effects on bone mass and structure [72]. b. Dose – Response Studies Few dose – response studies have been conducted. However, in 1984 Gunness-Hey and Hock did show a dose-dependent increase in trabecular bone mass of the distal femur when PTH was given by injection [53]. Later, a study was carried out on young, intact male rats [73 – 75]. This showed a dose-dependent increase in vertebral bone mass, size, and also biomechanical com-

731

CHAPTER 77 Treatment with PTH Peptides

petence [73]; a dose-dependent increase in cortical bone strength [74]; and a dose-dependent increase in cortical bone formation at both the periosteum and the endosteum [75]. The study was conducted in growing rats, and PTH was given for 5 weeks only, so a steady state was not attained. Kimmel et al. showed in their study [76] on adult osteopenic, ovariectomized rats that PTH increased trabecular bone volume and trabecular thickness dose-dependently. Again, this study was short-term (28 days). In the study of Okimoto et al. [72], weekly injections were given with either hPTH (1 – 34) 10 or 90 g/kg for 3 months. Again, a clear dose – response pattern was shown, concerning both bone mass and bone strength.

B. Effects of PTH on Bone Mass and Size 1. TRABECULAR BONE (VOLUME, THICKNESS, AND CONNECTIVITY) The study of Gunness-Hey and Hock [53] demonstrated that PTH induced an increase in trabecular bone mass in the distal part of the femur. In their study, trabecular and cortical bone were separated manually, and dry weight was measured and used as the end point for bone mass changes. In a later study of Hock et al. [77], histomorphometry of the tibial metaphysis was performed on rats treated for 12 days with PTH either alone or in combination with calcitonin for 3 days. This study showed a significant increase in the proportion of the space between cortices occupied by bone after treatment with PTH alone — and an identical significant increase after combination therapy. This demonstrated the powerful anabolic effect of PTH on trabecular bone volume (an effect which was found to be independent of initial bone resorption). The authors also found an almost threefold increase in osteoblast surface and osteoid surface, while the rate at which the individual osteoblast made bone (mineral apposition rate) was unchanged. There was only a slight change in eroded surface as a percentage of total trabecular surface. These impressive data were later confirmed in studies by Liu and Kalu, Tada et al., and Takahashi et al. Liu and Kalu [67] showed that PTH caused an increase in trabecular bone volume and thickness in the tibial metaphyses and lumbar vertebrae in ovariectomized rats. Tada et al. also showed an increase in trabecular volume and thickness in tibia and vertebrae when PTH was given to parathyroidectomized and osteopenic rats [65]. Furthermore, Takahashi et al. showed an increase in trabecular volume and thickness in lumbar vertebra L5 after PTH treatment of ovariectomized rats [66]. These data were confirmed by Wronski et al. [63] and also in biomechanical studies performed on young intact rats, mature ovariectomized rats, and aged osteopenic rats [73,78,79]. None of these studies showed any change in trabecular

“number” (an inverse measure of the spacing of trabeculae) during PTH treatment. Trabecular connectivity was not measured. A study by Shen et al. [80] showed an increase in trabecular connectivity when PTH was given concomitantly with estrogen. Measurements were performed by use of node and strut analysis of the proximal tibia. These data still need confirmation by use of other stereological measurements of connectivity. Most recently, Wronski et al. examined the effects of 5 days/week PTH on the thickness of completed packets of new bone (wall width) in the cancellous bone of the first lumbar vertebra. The animals were 3 months of age at ovariectomy and some began 6 weeks of treatment shortly after surgery while the remainder started 10 weeks of treatment 1 year later [81]. Compared to vehicle, in both groups wall width was increased substantially, a result that provided at least a partial explanation of the anabolic effect of PTH therapy on vertbral cancellous bone volume in young and old ovariectomised rats and was comparable to that obtained by Bradbeer et al. in patients [34]. 2. CORTICAL BONE The study of Gunness-Hey and Hock in 1984 [53] had indicated that cortical bone might be increased by intermittent treatment with PTH (although their data were not statistically significant). This issue is especially interesting, as it had been suggested by human studies that the PTH-induced increase in trabecular bone was achieved at the expense of cortical bone. Liu and Kalu showed in 1990 that PTH did improve both periosteal and endosteal bone formation and thereby increased cortical bone formation in sexually mature ovariectomized rats [67]. In 1993, Wronski et al. showed that PTH increased cortical bone mass by increasing periosteal apposition and endosteal bone formation [63]. This was further verified by Oxlund et al. [75]. Therefore, none of the rat studies to date has shown any sign of loss of cortical bone during intermittent PTH treatment. On the contrary, they have all shown an increase in cortical bone mass and size, which in biomechanical terms is very important. Furthermore, as mentioned earlier, very new data achieved from a large animal model (cynomolgus monkeys) cannot confirm any loss of cortical bone mass due to PTH treatment [50].

C. Effects of PTH on the Biomechanical Competence of Bone Previous studies have clearly demonstrated the anabolic effect of PTH on bone mass in rat models [52 – 54,64, 66 – 68]. This anabolic effect has been demonstrated concerning both trabecular bone and also cortical bone. However, during the past 4 – 5 years, it has become evident that

732 measurements of bone mass in both clinical and preclinical studies have to be accompanied by measurements of bone strength or bone “quality.” In clinical studies, occurrence of new fractures has become the required end point for assessment of efficacy of treatments. In preclinical trials, biomechanical testing of bone from different skeletal sites is now required by the FDA. Recently, several biomechanical studies have been performed concerning the effect of PTH on bone quality in young, intact male rats [73,74], in sexually mature, ovariectomized, and slightly osteopenic rats [78,82 – 84], and also in aged, ovariectomized, osteopenic rats [79]. These studies have elucidated the effect of PTH on bone strength at different skeletal sites: the vertebral bodies [73,78,79], femoral neck [84], and femoral cortical bone [74,82]. 1. VERTEBRAL BODY (NON-LOAD-BEARING SITE) Measurements of vertebral body strength or biomechanical competence in rats are relevant to man, since the vertebral bodies, like those in humans, comprise predominantly a core of central trabecular bone with a cortical shell. The trabecular network is stereologically very oriented and of the same dimensions as the human vertebral trabecular network [73]. In humans, vertebral fractures, alongside wrist fractures, are the most common osteoporotic fractures to occur after the menopause. The very first biomechanical study performed with PTH on rat vertebral bodies was published in 1991 [73]. This study was conducted in young male rats. They had, therefore, a very intact and well-connected trabecular network. The study clearly showed that PTH given intermittently caused a dose-dependent increase in bone strength, whether normalized for bone mass or cross-sectional area. This

MOSEKILDE AND REEVE

clearly demonstrated that PTH had no negative effect on bone material strength. A more clinically relevant study was carried out on sexually mature, ovariectomized rats [78]. This study confirmed the findings from the first study: that PTH given intermittently had a positive effect on bone mass and material strength. Finally, the study on aged, ovariectomized, and osteopenic rats [79] confirmed again the positive effect of PTH on vertebral bone mass and strength. In this study, where connections and whole trabecular structures had been lost in the vertebral cancellous bone core, the effect of PTH was still very positive (Fig. 4). 2. FEMORAL NECK (LOAD-BEARING SITE) In humans, the femoral neck is perhaps clinically the most important site of osteoporotic fractures. The rat femoral neck, like that in the human, is loadbearing. It is placed inside the hip joint and is therefore only partly covered with periosteal tissue; it consists of equal proportions of cortical and trabecular bone [85]. The femoral neck has been tested in the sexually mature, ovariectomized rat model [84]. This study clearly showed a positive effect of PTH on rat femoral neck strength (load). No normalization procedures were performed in this study, so changes in material strength as such could not be assessed. Recently, a short-term study of Toromanoff et al. detected increases in femoral neck strength after only 15 days treatment of young, intact female rats with PTH 50 g/kg day [86]. 3. FEMORAL CORTICAL BONE As it has been suggested several times that the pronounced anabolic effect of intermittent PTH treatment on trabecular bone mass was achieved at the expense of cortical

FIGURE 4 SEM at low magnification ( 20) of whole vertebral bodies from rats treated with PTH. (A) Aged, ovariectomized, osteopenic rat. (B) Aged ovariectomized, osteopenic rat treated with PTH for 24 weeks. It is clearly visualized that PTH maintains trabecular connectivity and increases trabecular thickness an also cortical thickness. Reproduced with permission of Lis Mosekilde, Endocrinology 124 (5), 2126 – 2134 (1994).

733

CHAPTER 77 Treatment with PTH Peptides

TABLE 1 Skeletal site

Vertebral bodies Femoral neck Femoral cortical bone

Loads at Failure (Newtons) in Studies Comparing PTH Treated Rats with Controls Control (N)

Treatment (N)

Difference

p Value

PTH, young, male, 5w [73]

254  10

392  23

54%

p 0.001

PTH, mature, Ovx, 15w [78]

250  15

470  23

88%

p 0.001

PTH, aged, Ovx, 24w [79]

342  33

588  25

59%

p 0.001

89  4

130  6

45%

p 0.001

Study design [reference]

PTH, mature male, Ovx, 15w [84] PTH, young male, 5w [74]

*

PTH, mature, Ovx, 15w [82]

144  4

169  5

17%

p 0.01

1025  33

1218  30

20%

p 0.01

Note. Load values (N) from compression test. *Load values from three-point bending test. Means  SEM. Study referred to in reference list is indicated in brackets.

bone (“cortical steal” phenomenon), studies on cortical bone strength would also be very relevant. One study was performed by use of three-point bending of femora from young intact male rats [74]; and a further study used a compression test on pure cortical, femoral bone from sexually mature, ovariectomized rats [82]. Both of these studies clearly showed that intermittent treatment with PTH in rat models did not induce a “cortical steal” phenomenon — on the contrary, PTH increased cortical bone strength in both studies. To conclude, the effect of PTH on biomechanical competence in these different rat models and at different skeletal sites has been positive. An increase in bone structural and material strength has been shown in all studies and at all skeletal sites investigated (Table 1). Importantly, a recent large animal study with cynomolgus monkeys has proven that this is also the case in animals with intracortical bone remodeling [50]. 4. EFFECTS OF PTH ON RESPONSE TO MECHANICAL LOADING OF BONE Another important aspect is whether there is a role for PTH in the mechanical responsiveness of bone. A study by Chow et al. in 1998 [87] showed that there is a role for PTH in increasing the effect of mechanical stimulation on bone. Chow et al. used a rat model where mechanical stimulation of a caudal vertebra was performed. This gave a specific osteogenic response in normal rats. However, this response could be augmented by a single injection of PTH given 30 – 45 min before loading. Based on these results, the authors suggested that physiological levels of PTH may be necessary for the optimal responsiveness of bone to mechanical loading and that PTH had a key role in sensitizing the strain-sensing mechanism to achieve optimal responsiveness to mechanical loading. Nearly simultaneously, Turner et al. [88], also using a rat model, reached rather different conclusions. In their study PTH was shown to be effective in increasing osteoblast number and bone formation; and also had beneficial effects on bone volume in the absence of weight bearing. The au-

thors concluded that the anabolic effects of PTH on cancellous bone were independent of mechanical usage [88]. Therefore the relationship between PTH treatment and the effects of mechanical loading requires further investigation. 5. EFFECTS OF PTH IN COTHERAPY OR SEQUENTIAL THERAPY WITH ANTI-RESORPTIVE AGENTS It has previously been shown that the anabolic effect of PTH is not dependent on initial bone resorption [20,65,77]. Furthermore, as PTH for some years has been thought to exert its anabolic effect on trabecular bone at the expense of cortical bone, several studies have been conducted to elucidate whether the combination of PTH with an antiresorptive agent (estrogen, bisphosphonate, or calcitonin) would be superior to PTH alone. These studies have again been conducted in sexually mature ovariectomized rats [63,70,78,82,84] and also in aged, osteopenic, ovariectomized rats [79]. In these studies, histomorphometry was performed on tibial, metaphyseal trabecular bone [63], and tibial cortical bone [70], combined with biomechanical testing of vertebral bodies [79], femoral necks [84], and femoral cortical bone [82]. All these investigations showed unequivocally that PTH alone was as effective as PTH in combination therapy concerning outcomes measured as bone mass, structure, and strength. However, a study by Shen et al. [80] has recently shown PTH in combination with estrogen to be slightly more effective than PTH monotherapy. This is especially the case with regard to restoring structural connectivity. In all these studies, the effect of the anti-resorptive agents has also been separately tested. Although they have been capable of maintaining bone mass and strength, none of these agents has been capable of increasing bone strength over the level seen in control animals. In summary, PTH monotherapy seems as effective or almost as effective as PTH cotherapy in maintaining or restoring bone mass and strength in different rat models. It has to be acknowledged, though, that rats have very little intracortical bone remodeling [89]. Rat models are,

734

MOSEKILDE AND REEVE

therefore, not optimal for comparing the effects of PTH mono- and cotherapy in relation to cortical porosity. Another problem which arises is that of bone loss after withdrawal of PTH treatment. In the past 2 – 3 years, several studies have addressed this problem. In the study of Mosekilde et al. PTH was given to ovariectomized (OVX) rats for a period of 2 weeks. The dose was 20 g/kg/day followed by a 10-week withdrawal period [90]. The study showed an increase in bone mass and strength after 2 weeks of treatment, but no loss was seen following withdrawal of treatment — even 10 weeks later. However, concomitantly Ejersted et al. [91] showed that after treatment with PTH (1 – 34) 62 g/kg/day for 8 weeks, withdrawal of treatment led to bone loss and also to loss of bone strength. In their study, the authors could show that treatment with Risedronate 5 g/kg twice a week could prevent loss of bone mass and strength due to withdrawal of PTH therapy. By comparing these two studies it is clear that shortterm, low-dose PTH can increase bone mass without any withdrawal symptoms. However, long-term, high-dose PTH increases bone mass to such an extent that withdrawal is followed by a rapid loss of this newly acquired bone. A recent study by Kishi et al. could further confirm this by showing that skeletal sites with the highest positive response to PTH treatment also suffered the largest loss after withdrawal [92]. Concerning maintenance of bone mass due to PTH treatment, Hodsman et al. have shown that the Raloxifene analog (LY117018) may be useful, but that this strategy allowed only a dose reduction of hPTH(1 – 34) rather than its discontinuation [93]. The above-mentioned studies have all focused on cotherapy or sequential therapy at the time of OVX (prevention) or after OVX-induced bone loss (intervention). However, Shen et al. have used PTH treatment before onset of estrogen deficiency in order to augment “peak bone mass” [94]. Their very elegant study showed that treatment of rats with PTH prior to ovariectomy produces increase in BMD and strength, that these beneficial effects extended to a period of at least three times the duration of treatment, that the BMD lost when PTH was discontinued equated to the amount accrued during PTH treatment, and that estrogen replacement therapy could be used to maintain the bone gained as a result of PTH treatment.

D. Comparison among PTH, Fluoride, and Growth Hormone For several years, fluoride has been used in human clinical trials and has proved itself capable of increasing vertebral bone mass [95]. However, recent studies have indicated that this positive effect was not accompanied by a positive

effect on bone quality — i.e., it did not decrease the incidence of fracture [95,96] to the expected extent at the chosen dose of 75 mg NaF per day (see Chapters 14 and 75). Studies in pigs [97] and in humans [83] have both confirmed this discrepancy between bone mass and bone strength during fluoride therapy. Additionally, recently a study on mature, intact female rats has shown a clear, dosedependent increase in bone mass which was not followed by a similar increase in structural strength. This suggested that bone quality had declined during fluoride therapy [98]. Therefore, PTH and fluoride seem to act differently on bone: although they both have an anabolic effect on bone mass, only PTH has a positive effect on bone strength. In contrast to fluoride, PTH does not have a negative effect on the material quality of bone. Alternatively, the beneficial effects of fluoride on bone strength are critically dose dependent, so that fluoride may have a narrower therapeutic “window” than PTH, an idea supported by the finding of osteomalacia in patients treated with high doses [99]. Growth hormone (GH) treatment has also been suggested as an anabolic agent for the treatment of osteoporosis. In preclinical studies, GH treatment of OVX rats has shown that GH acts only on the periosteal surface and not on the endosteal surface (cortical or cancellous). GH treatment can thereby increase bone size, mass, and strength at several skeletal sites — but it is not capable of increasing cancellous bone mass [100]. Furthermore, even after highdose treatment with GH (5 mg/kg/day) the response in bone mass and strength will vary between 25 and 50% (compared with high-dose PTH, which can increase bone strength 50 – 180%) [101]. GH treatment and exercise have an anabolic additive effect on the rat skeleton [102]. Furthermore, the bone formed during GH treatment has normal material quality. However, GH increases body weight and it has many undesirable side effects — so its therapeutic “window” is also rather narrow.

E. Discussion of Findings in Relation to Clinical Studies All the rat studies mentioned above have consistently shown intermittently dosed PTH to have a positive effect on bone mass, size, structure, and strength. This has now been further confirmed in a large animal study [50]. In clinical studies, postmenopausal and age-related osteoporosis have been characterized by loss of trabecular bone mass and trabecular connectivity, thinning of cortical bone, the so-called “trabecularization” of cortical bone brought about by high rates of unbalanced endosteal remodeling, and also a lack of capability for periosteal apposition (and thereby also a lack of compensatory increase in bone size).

735

CHAPTER 77 Treatment with PTH Peptides

Antiresorptive agents such as estrogen, bisphosphonates, and calcitonin have shown themselves capable of modestly increasing spinal bone mass [103 – 105]. Some studies have also indicated an increase in bone quality (reduction of fracture rate) during treatment with antiresorptive agents [106]. From the rat studies performed to date, PTH seems to have considerable additional advantages when given intermittently. In these studies, PTH has proved itself capable of increasing trabecular bone mass at all skeletal sites [20,76,78,79] and of maintaining trabecular connectivity [76,79 ] — or improving connectivity [80]. At the same time, PTH has proven itself capable of increasing cortical thickness by adding bone both at the endocortical surface and also at the biomechanically more important envelope: the periosteal surface [67,70,73,75]. Further, PTH has been able to increase bone structural strength, including material strength, at all skeletal sites tested.

F. Conclusions Based on preclinical studies performed mainly in rats, but now also in larger animals, it is concluded that intermittent PTH therapy seems very promising. PTH is capable of increasing bone mass at all four envelopes: cancellous, endocortical, intracortical, and periosteal. Also, bone structural and material strength are increased during PTH therapy. All skeletal sites (vertebrae, tibial metaphyseal and cortical bone, femoral cortical bone, and femoral necks) are affected positively during treatment. Furthermore, all animal models (young, growing intact rats; mature ovariectomized rats; orchidectomized rats; aged, osteopenic rats; and cynomolgus monkeys) have responded in an equally positive manner. Therefore, PTH seems to offer very substantial promise in the management of established osteoporosis.

V. CLINICAL STUDIES A. Short-Term Treatment Studies with hPTH Peptides in Osteoporosis 1. FORMULATIONS OF PTH AND ITS PEPTIDES THAT HAVE BEEN USED CLINICALLY Although it is now possible to synthesize the whole parathyroid hormone molecule by recombinant technology [76], only active fragments of PTH, made (like calcitonin) by solid-state technology, were available when human studies began. Initial studies were undertaken with hPTH(1 – 34) because it was thought to be a natural cleavage product and because it retained all the measurable bioactivity of natural PTH in the chick hypercalcemia as-

say. More recently, hPTH(1 – 38) has also been studied because of its potentially greater similarity to natural cleavage products and the possibility that it may have greater potency than hPTH(1 – 34) [107]. There has also been one impressive human study with PTHrP(1 – 36), [108]. PTH and its peptides have been given to patients as subcutaneous injections in aqueous solution, sometimes containing a trace of human serum albumin and, in the case of some recent studies with hPTH(1 – 34), a gelatin stabilizer. 2. SHORT-TERM STUDIES WITH DAILY INJECTION REGIMES Early work showed that the renal effects of daily injections of 1 – 34 peptide were to depress the renal transport maximum for inorganic phosphate for up to 6 h [17]. This fairly brief response reflected the very short biological exposure to subcutaneous hPTH(1 – 34) documented in a volunteer study using the cytochemical PTH bioassay [109]. Slovik et al. studied the effects of the first few weeks of PTH therapy and failed to find large effects on calcium balance or kinetics in the first 2 weeks of treatment [110]. However, by the 31st day of treatment substantial effects on both histological and biochemical indices of bone turnover are evident, as is discussed in section C3. Plotkin et al.’s study with PTHrP(1 – 36) showed that within 2 weeks, doses of PTHrP which were deliberately rather larger on a molar basis than those used recently for longer term studies of hPTH(1 – 34) resulted in a trend upward in osteocalcin (probably the best formation marker in PTH studies, vide infra) with no change in serum calcium or other safety markers and a decline in endogenously produced PTH and nephrogenous cyclic AMP [108]. The most surprising finding was that the bone resorption markers N-telopeptide and deoxypyridinoline both declined. Longer term studies are awaited to see if this potentially beneficial effect on bone resorption is sustained with prolonged PTHrP treatment.

B. Treatment Regimes for Osteoporosis These have all involved parenteral administration accessing the general circulation, with the exception of one study in which the PTH was delivered locally through degradation of a polymer containing plasmid DNA for hPTH(1 – 34) [111]. 1. HPTH AND FRAGMENTS GIVEN ALONE IN THE LONGER TERM Initial studies with hPTH(1 – 34) in humans used a daily-dose regimen, ranging from 133 to 800 U/day (units were derived from an in-house chick hypercalcaemia based assay which, while never adopted as an international standard, remains in current use in a number of laboratories). In humans, the 133 U/day regimen was thought to be

736

MOSEKILDE AND REEVE

almost ineffective, [112] and all studies claiming clinical benefit have been conducted in the range of 400 – 1200 U/injection. The early studies in patients with osteoporosis demonstrated unequivocally that the anabolic effect of hPTH(1 – 34) on cancellous bone volume depended on achieving supranormal rates of bone formation [19] and also showed considerable apparent heterogeneity of response [19,113]. In these early studies hPTH(1 – 34) was given alone as a daily injection, with nutritional supplements where necessary to safeguard an intake of 20 mmol (800 g) calcium daily and a normal intake of vitamin D. In retrospect. it is now known from later HPLC studies (J. M. Zanelli, pers. commun). that some of the hPTH administered to patients in the 1970s was not as pure as peptides that are currently available. There is evidence that for a minority of patients the administered peptide or an impurity was antigenic [19,114 – 116]. As already described, studies with rats [20,117] and dogs [21] disproved the early hypothesis of Parsons and Reit [118] that a better anabolic effect could be obtained by mimicking hyperparathyroidism with a less discontinuous exposure to raised PTH levels than could be obtained with subcutaneous injections. Recently, two important studies have been published in which hPTH has been used alone to prevent or reverse bone loss provoked respectively by GnRH agonist treatment of endometriosis [119] and glucocorticoid therapy [120]. Also, there has been a recent multicenter study to test the possibility that weekly parenteral doses of hPTH(1 – 34) in the range 15 – 60 g peptide/dose can be effective [121]. This is in accordance with the preclinical rat study by Okimoto et al. [72]. 2.

HPTH

GIVEN WITH OTHER AGENTS

Over the past 10 years efforts have been made, by adding a second agent, to develop subcutaneous injection regimens that protect the peripheral skeleton while preserving the anabolic effect on the spine. Two approaches were employed. The first was to add calcitriol to promote calcium absorption [122,123]. The second aimed to suppress bone resorption on the grounds that no quantitative link may exist between the anabolic actions of PTH and previous bone resorption [124]. Humans have been studied with the combination of PTH and oestrogen [34,115,116,125 – 127]. In a relatively short-term study to test the possibility that alendronate might block the effectiveness of hPTH therapy, Cosman et al. gave hPTH(1 – 34) to patients already established on 10 mg/day alendronate therapy and found that cotreatment with the bisphosphonate did not prevent hPTH inducing the increase in bone formation markers normally seen with hPTH mono-therapy (F. Cosman and R. Lindsay, pers. commun.). However, another study performed in rats has shown the opposite, that alendronate does blunt the effect of PTH [128].

3. CYCLICAL AND “ADFR-LIKE” REGIMES The third approach has been to give PTH cyclically. Cycle lengths have varied, with the shortest being 1 week in 4 [129 – 132]. Two of these intermittent treatment regimens involved sequential administration of a second agent (either calcitriol or calcitonin). Unlike treatment studies with sodium fluoride, no consensus supports the concurrent use of large doses of calcium supplements with PTH peptides, although all studies have aimed to ensure a daily minimum of 800 mg (20 mmol) dietary calcium. 4. LOCAL DELIVERY OF PTH Early studies by Barnicot [133] and Chang [134] were interpreted previously as predicting an osteolytic response to most modes of local PTH delivery. However, Bonadio et al. [111], have recently shown that when implanted in bone defects, plasmid genes expressing hPTH(1 – 34) entrapped in a polymer matrix sponge can promote bone repair when taken up and expressed by host fibroblasts.

C. Long-Term Clinical Studies with hPTH Peptides Experimental studies with parathyroid peptides in humans began 25 years ago [18] but trials of antifracture efficacy are only now appearing, in part because of initial difficulties in manufacturing parathyroid peptides of adequate purity in sufficient quantity. A second reason for the slow rate of progress since the 1970s was the result (originally unexpected) that human parathyroid hormone (hPTH) (1 – 34), when used alone, increased bone mass in the axial skeleton at the apparent expense of peripheral bone rather than by increasing gastrointestinal absorption of bone minerals. However, solid evidence now suggests that hPTH peptides increase spinal bone mass perhaps even more rapidly than fluoride salts; strategies for conserving peripheral bone during and after hPTH treatment are showing signs of success. 1. EFFECTS ON SYMPTOMS Although so far it has proved impossible to devise an effective delivery route other than parenterally, all researchers report good cooperation and compliance with daily injection regimens in patients who were suffering continued symptoms from previous fractures or anxiety about future fractures. Three or four patients treated with the highest doses used for periods of up to 1 month or more (800 U/day) have reported dull pains in the tibiae, which resolved quickly on cessation of treatment or reduction of dose (A. Hodsman, pers. Commun.; J Reeve, unpublished data). Symptomatic relief of backache usually has been ascribed to the general effects of physician intervention and the normal tendency in most patients for symptoms to improve between episodes of fracture.

737

CHAPTER 77 Treatment with PTH Peptides

2. SIDE EFFECTS Several patients in earlier studies who developed high levels of plasma hPTH (1 – 34) binding raised the possibility that the antibodies responsible for the binding may interfere with the effectiveness of treatment [115,116]. One particularly interesting patient developed clinical hypoparathyroidism that resolved on withdrawal of treatment and reemerged on rechallenge [114]. She was thought to have developed antiidiotypic antibodies that blocked the PTH receptor. Otherwise, reported side effects have been confined to transient reddening around the injection site and in a few patients transient elevation of plasma calcium above the upper limit of normal at about 6 h after injection. 3. EFFECTS ON BONE REMODELING a. The Whole Skeleton A single subcutaneous injection of hPTH(1 – 34) causes a very brief peak of plasma PTH bioactivity [109] followed by a rather longer effect on maximal renal transport for phosphate, which is nevertheless no longer evident 6 – 8 h after injection [17]. In the long term no tachyphylaxis occurs, with the generation of unchanged amounts of nephrogenous cyclic AMP [116] at the end of long-term treatment. Similarly, the long-term effects of treatment with 500 IU/day and oestrogens on urinary excretion of calcium and phosphate have not been substantial [116]. In contrast, Hodsman et al. [131] showed that by 28 days of treatment with 800 IU/day of hPTH(1 – 34), three biochemical markers of bone formation (alkaline phosphatase, osteocalcin, and the procollagen-1 carboxyterminal extension peptide) as well as urinary hydroxyproline and calcium had increased substantially. Comparable changes were not evident in treatment with smaller doses of hPTH(1 – 34) [110]. When the hPTH was given as a weekly injection, the biochemical markers of bone resorption fell significantly [121]. By 6 months of treatment with hPTH(1 – 34) alone, rates of 47Ca estimated bone mineralization had increased substantially in most patients and were accompanied by comparable rises in rates of 47Ca estimated resorption and elevations in hydroxyproline excretion in many patients [19]. At 1 year, treatment with hPTH(1 – 34 or 1 – 38) plus estrogens resulted in comparable increases in kinetic measures of bone formation but much less impressive increases in bone resorption, because of statistically significant improvements in calcium balance [127]. Reeve et al. [127] showed that the biochemical marker plasma osteocalcin tracked closely with a much more expensive referent radioisotopic method for measuring bone formation during the response to hPTH therapy, so it was of considerable potential importance that Lane et al. found during treatment of glucocorticoid-induced osteoporosis that the early osteocalcin response was a quanititative marker for the later axial bone density increase with hPTH therapy [120].

FIGURE 5 Changes in cancellous bone area (as a percentage of total area between iliac cortices) compared with the wall width of completed packets of new cancellous bone in patients treated with hPTH(1 – 34) plus oestrogen replacement therapy for 1 year. Note that a response in bone area appears statistically to depend on the formation by osteoblasts of more new bone at individual remodeling sites (as with fluoride therapy). Reprinted with permission from Bradbeer et al. [34]. b. Iliac Bone Hodsman et al. [32,131] demonstrated for the first time that treatment with hPTH(1 – 38) has substantial effects on iliac cancellous bone remodeling within 1 month of the start of therapy. By comparison with a historical control group, large increases were seen, on the one hand in both eroded surfaces and osteoclast numbers and on the other hand in formation indices, including the extent of mineralizing surfaces and bone formation rate. At later stages in treatment, there were either smaller increments (at 6 months [19,113]) or minimal changes (at 1 year [116]) in remodeling, whereas wall thickness of new packets of cancellous bone, was considerably increased [34] (Fig. 5). Because of difference in dose levels, it is not possible to be certain that the changes seen by Hodsman were similar to those in the iliac bone of patients treated with lower doses and biopsied at a much later stage of treatment. 4. EFFECTS ON BONE PERFUSION The only studies on bone perfusion in patients on hPTH therapy are those of Wootton et al. Increases in bone blood flow measured with the 18F technique have been moderate [135] compared to those seen with osteomalacia and Paget’s disease of bone [136]. In the only case they studied who suffered from boring pain in her tibiae prior to reduction of the dose, skeletal perfusion increased from 3.9 to 6.4% blood volume per minute which was considered unexceptional (case 23 in Ref. [135] and case 8 in Ref. [137]). 5. EFFECTS ON BONE MASS The only reliable techniques for assessing bone mass at the time of the earliest studies were single-photon

738

MOSEKILDE AND REEVE

FIGURE 7

FIGURE 6

The relationship between final iliac cancellous bone volume after 6 months of treatment with hPTH(1 – 34), 500 IU/day, and the radioisotopically measured 47Ca accretion rate, a measure of whole body bone formation. Reprinted with permission from Reeve et al. [19].

absorptiometry and iliac bone histomorphometry (Fig. 6). In the early multicenter study of Reeve et al. [19], a substantial increase of about 70% in iliac cancellous bone volume (as a proportion of bone plus marrow [BV/TV]) was reported after 6 months of daily injections at an average dose of about 500 IU/day. The final BV/TV was found to be predicted quite closely by the rate of new bone formation, as determined with 47Ca studies of bone mineralization rates (Fig. 6). However, this effect was not reproduced in the iliac cortices, which qualitatively appeared somewhat more porous (consistent with increased remodeling), or in the femoral cortices, which tended to lose bone in the small subgroup of patients studied [138]. Subsequent studies have shown that daily treatment with hPTH peptides also increases spinal bone mass. Reeve et al. [115, 127] found that, when combined with estrogens, hPTH in a dose of about 500 IU/day increased spinal cancellous bone mass substantially (Fig. 7). This increase was sustained in the first year after cessation of PTH therapy during maintenance of hormone replacement therapy with estrogen and (in women with a uterus) a progestogen. The increase in vertebral body cancellous bone was sufficient to account for all the increase in spinal bone mass as assessed by dual photon absorptiometry. Similar results for the spine were obtained by Slovik et al. [123] and Neer et al. [122] in men and women treated with daily injections of hPTH(1 – 34) and 0.25 g oral calcitriol. A second group of women treated in cycles (6 weeks of intermittent hPTH(1 – 34) followed by 6 weeks of calcitriol, 0.25g twice daily) showed slightly smaller responses in spinal bone mass [122].

Changes in spinal bone mineral with treatment. (Left) hPTH(1 – 34), 500 IU/day, plus additional hormone replacement therapy and (right) NaF plus calcium therapy at a mean dose of 37.1 mg/day NaF. Changes in cancellous bone (QCT) and combined cancellous and cortical (DPA) bone density are normalized to the same units as well as the differences between them. Error bars, 1 SD. Reprinted with permission from Reeve et al. [115].

More recently, two further studies [125] (one so far published only in abstract form [126]) have appeared which have confirmed emphatically that hPTH(1 – 34) given to estrogenized postmenopausal women with osteoporosis and between none and several baseline vertebral fractures increases spinal bone density considerably more than can be achieved with estrogens alone in conventional doses. In another study on women with extremely deteriorated vertebral osteoporosis (and having a mean of only 6/15 undamaged vertebrae for evaluation at recruitment) Reeve et al. confirmed the superior effectiveness of this combination therapy for reversing and correcting low bone density [127]. Once weekly hPTH(1 – 34) given alone, at the two higher doses used, had measurable effects equating to 3.6 and 8.1% increases in spinal bone density measured by DXA [121]. Now that DXA and QCT densitometry can be used to measure more precisely bone density in the proximal femur, new data have appeared showing that daily hPTH therapy generally increases bone density in the total hip region in parallel with spinal increases [120,125 – 127, 139]. However, the size of the response is smaller in relative terms than in the spine. In a careful QCT study, Cann et al. were able to show that hPTH increased cancellous bone in the proximal femur, while the volume of cortical bone expanded, particularly in the femoral neck, and at the same time it showed evidence of becoming somewhat more porous [139] presumably due to increased haversian remodeling [19]. Concerning cyclic administration regimes, Hesch et al. [129] studied a group of patients receiving a daily injection

739

CHAPTER 77 Treatment with PTH Peptides

of hPTH(1 – 38) for 70 days with three 14-day periods of nasal calcitonin (200 IU/day), starting on Day 15. This cycle was repeated four times in 14 months. The results were comparable to those obtained with the daily injection regimens of Neer et al. [122] and Reeve et al. [115] with respect to both spinal cancellous bone measured with quantitiative computed tomography and with transiliac biopsies [140]. Hodsman et al. administered hPTH(1 – 38) [130] and hPTH(1 – 34) [131] in 70- and 90-day cycles; hPTH was given in doses of 400 and 800 IU/day, respectively, for the first 14 days and (in the second study) for the first 28 days of each cycle. With the higher-dose regimen, substantial positive effects were seen on biochemical markers of bone formation accompanied by substantial improvements in spinal bone density [141]. Earlier, Reeve et al. [132] found that a short-cycle regimen in which hPTH(1 – 34) was given for 1 week in 4 at even larger daily doses than 800 IU/day had no measurable anabolic effects. This disparity suggests that the PTH stimulus may need to be given for a minimal period longer than 1 week to result in increased bone mass. The results of Slovik et al. [110] also suggested that the induction period may need to be in excess of 2 weeks. 6. EFFECTS ON BIOMECHANICAL COMPETENCE OF BONE The best clinical test of biomechanical competence of bone is a randomized controlled trial which is sufficiently powerful statistically that there is an a priori better than 80% chance at the design stage of the trial showing a positive effect on fractures for the desired reduction in fracture rate. Over 18 months hPTH treatment on average was received by a large cohort of patients participating in a placebo-controlled multicenter randomized trial of hPTH(1 – 34) [142], before the extended period of treatment planned for in the trial was shortened as a precautionary measure. At the time of writing, the FDA has now authorized further simultaneous clinical as well as animal studies following a thorough review of safety concerns. These centered on some observations made on the tumorogenic effect of extremely high doses of hPTH given to a particular strain of rat for over half the animals’ normal lifespan. The study of Lindsay et al. found a statistically significant effect of hPTH(1 – 34) against spine fractures, although it was not powered to do so [125]. The study of Reeve et al. [127], which was very underpowered for this purpose because of the small numbers of intact vertebral bodies at baseline as well as the small numbers of patients, found a nonsignificant effect. Now the adequately powered studies of Neer et al. and Marcus et al. [142,143] have demonstrated very substantial effects amounting to an approximate 65% reduction in spine fractures and 50% reduction in nonspine fractures in postmenopausal women with established osteoporosis. This

appears to confirm rhPTH(1 – 34) as one of, if not the, most efficacious treatments for osteoporosis yet discovered. Over the next few years a series of further studies on the anti-fracture efficacy of PTH-receptor activators is anticipated. 7. EFFECTS ON THE GI TRACT AND RENAL HANDLING OF CALCIUM Daily injections of hPTH(1 – 34) at a dose of about 500 IU/day did not improve overall calcium balance but substantially increased axial cancellous bone volume [19]. In the subgroup of patients in whom it was measured, hPTH therapy given alone did not increase 1,25-dihydroxyvitamin D levels [144] in contrast to some animal studies [22]. It seemed in these early studies that the mineral required for the anabolic effects of hPTH(1 – 34) in the axial skeleton may have been derived from the peripheral skeleton. The lack of effect on calcium absorption of hPTH given alone to patients was confirmed by radiotracer as well as metabolic balance methods, both showing no suggestion of an overall increase in calcium absorption after 6 months of treatment [19,145]. A later study by Podbesek [22] revealed that intact greyhounds responded differently, with marked increases in radiocalcium absorption being observed when they were treated with a similar regimen. In a recent analysis of data derived from metabolic balance studies in patients treated with hPTH(1 – 34 or 1 – 38) and estrogen, Reeve et al. have observed that there was a significantly positive trend in calcium balance and that the size of the hPTH effect on GI calcium absorption was statistically dependent on the size of the increment in axial BMD [127]. It seems that with this combination of treatments there is no evidence that the peripheral skeleton is “robbed” of its bone mineral to permit the observed increased in axial bone density.

D. Prevention of Bone Loss with PTH Peptides The apparently successful treatment of established osteoporosis with PTH fragments demonstrated by themselves and other investigators led Finkelstein et al. [119] to study the usefulness of hPTH(1 – 34) as a preventative agent in a newly important group of patients at risk of developing osteoporosis. In the past decade it has become clear that GnRH agonists are by most criteria the treatment of choice for women with severe endometriosis. Several studies have, however, documented bone loss from the spine over treatment periods as short as 6 months. Bone mass appears to be largely regained after short treatment periods (6 months) but curtailing treatment in this way is frequently to frustrate the main objectives of GnRH therapy, and symptoms often recur quite quickly. In a controlled 12-month trial involving 43 women, Finkelstein et al. found that spinal bone density was significantly increased and that femoral neck, femoral

740

MOSEKILDE AND REEVE

FIGURE 8 Bone mineral density of the lumbar spine measured in the anteroposterior and lateral projections, femoral neck, trochanter, one-third radius, and total body in women receiving nafarelin alone (squares) or nafarelin plus human parathyroid hormone (circles). Values are expressed as the percentage of baseline. Reproduced with permission from Finkelstein et al. [119].

trochanter, and whole body bone loss evident in the control group was prevented by daily injections of 500 units of hPTH(1 – 34) (Fig. 8). Because Nafarelin therapy results in temporary estrogen withdrawal, this study provides a clinical model for the more permanent natural situation of the menopause. Finkelstein’s bone density and biochemical data, which showed a two- to threefold increase in both bone formation and resorption markers were again predictive of the size of the bone density response. These data provide a unique insight into the likely effects of PTH therapy were it to be used in the immediate postmenopause. As with

patients with established osteoporosis, the biochemical data suggested substantial increases in bone formation (Fig. 9) which were far larger than those seen with Nafarelin treatment alone. Interestingly, both markers of bone resorption also showed substantial rises of over 100% above baseline, more in percentage terms than has been reported with similar doses of hPTH(1 – 34) in patients with osteoporosis. It seems possible that in their state of recent estrogen withdrawal, these women were more sensitive to the effects of PTH in promoting bone resorption. This could also account for the more modest increases in spinal bone mass than

741

CHAPTER 77 Treatment with PTH Peptides

formed was fairly typical of remodeled callus together with evident remodeling cartilage formed near the gap center. Given further success with this approach, it seems inevitable that locally delivered PTH will be used in further work aimed at restoring local bone integrity where it has previously been lost due to fracture, or to other forms of local bone disease including osteoporosis.

VI. CONCLUSIONS

FIGURE 9 Serum alkaline phosphatase and osteocalcin concentrations and urinary excretion of hydroxyproline and pyridinolines in women receiving Nafarelin (squares) or Nafarelin and PTH (circles). The values shown are percentages of the baseline values  SE. The P values are for the comparisons of the rates of change in the two groups. The error bars for the serum alkaline phosphatase and osteocalin concentrations in the women treated with nafarelin alone are contained within the symbols. Reproduced with permission from Ref. [119]. those seen in patients with osteoporosis. In future work, it will be desirable to investigate whether in this situation PTH and its fragments (used as a means of preventing bone loss) also allow the preservation of the microstructure and connectivity of the trabecular network, which Compston and coworkers [146] have shown to be damaged in iliac bone by GnRH agonists used alone for 6 months.

E. Promotion of Local Bone Repair with Local PTH Delivery Bonadio et al. studied a beagle tibia or femur critical defect model with implanted polymer matrix sponges containing a plasmid releasing hPTH(1 – 34) gene as the matrix is degraded [111]. Based on prior evidence, it was concluded that this gene transfected the animal’s own fibroblasts, which then secreted the gene product. There was no evidence at the doses used of measurably increased blood levels of the PTH fragment. They found that the bone gaps filled with new bridging bone callus which consolidated (in contrast to the control animals) and under the stimulus of gradually increasing weight-bearing the callus remodeled sufficiently to allow the dogs to walk without fixators or apparent risk of refracture. Histologically, the new bone

The prospects are favorable that in the future the majority of patients with spinal osteoporosis and one or more fractures who need an effective form of secondary fracture prevention might be helped by anabolic agents such as parathyroid hormone or one of its fragments. Up to 65% of first spinal fractures go unreported to the patient’s physician [147] even though the impact of additional fractures on the patient’s health increases approximately in proportion to the number of severe deformities subsequently experienced [148], and undiagnosed fracture have a considerable impact on quality of life [149 – 151]. Patients frequently present after the first, second, or even third spine fracture. Currently, bisphosphonate treatments for spinal osteoporosis have or are becoming available which reduce substantially the rate of new vertebral fracture events [152,153]. Since these agents have more modest effects on bone mass than observed with PTH in the majority of studies, the question arises whether the substantial numbers of patients who are anticipated to be at risk of a “breakthrough” fracture despite bisphosphonate therapy (about 50% of those who would suffer similarly without treatment [152,153]) would benefit from a second-line treatment such as PTH or an analog. The effectiveness of PTH therapy in restoring mineral mass in bone which has been previously treated with bisphosphonates requires further investigation. Since estrogens and selective estrogen receptor modulators (SERMs, e.g., raloxifene), which have similar effectiveness to alendronate against spine fractures, appear to enhance rather than depress the effectiveness of PTH therapy, this might affect the choice of first-line agent in the severely affected patient. Certainly, since PTH can be given only by injection and injectables have been shown to be associated with poor compliance for osteoporosis prophylaxis in those who are painfree [154], an alternative delivery route or injection technique would have to be developed for PTH to become a first-line drug for many patients. However, pain and disability increase with the number of fractures, and with Caucasian postmenopausal women having at age 65 a 1% or greater risk of a new vertebral fracture every year [155] and this risk being increased seven-fold in those who already have a fracture [155 – 157], candidate patients for a second-line treatment upon the perceived failure of bisphosphonate therapy could

742

MOSEKILDE AND REEVE

after 3 years be 1.6% of those with preexisting osteoporosis according to the WHO definition [158] and 11% of those with severe osteoporosis and a pretreatment spine fracture. Also, improved means of delivery are now available. It is therefore likely that a sizeable clinical need will be shown to exist for an effective nonbisphosphonate treatment for severe vertebral osteoporosis (including osteoporosis secondary to glucocorticoid therapy) within a few years. Future areas of intensive investigation will include studies on the effectiveness of PTH therapy in bones which will experience relative mechanical disuse (such as the hip in partly mobile stroke patients), as well as studies of new agents which act through PTH receptor-mediated and novel independent pathways, such as those [159] which promote bone morphogenetic protein 2 (BMP-2) production in bone.

References 1. R. A. Hillam and T. M. Skerry, Inhibition of bone resorption and stimulation of formation by mechanical loading of the modelling rat ulna in vivo. J. Bone Miner. Res. 10, 683 – 689 (1995). 2. S. M. Garn, The Earlier Gain and Later Loss of Cortical Bone in Nutritional Perspective (monograph). C. C. Thomas, Springfield, IL (1970). 3. H. P. J. Kröger, M. Lunt, J. Reeve, J. Dequeker, J. E. Adams, J. C. Birkenhäger, et. al. Bone density reduction in various measurement sites in men and women with osteoporotic fractures of spine and hip: The European Quantitation of Osteoporosis (QAO) Study. Calcif. Tissue Int. 64, 191 – 199 (1999). 4. D. W. Dempster, F. Cosman, M. Parisien, V. Shen, and R. Lindsay, Anabolic actions of parathyroid hormone on bone. Endocr. Rev. 14, 690 – 709 (1993). 5. E. Bauer, J. Aub, and F. Albright, Studies of calcium and phosphorus metabolism. V. A study of the bone trabeculae as a readily available reserve supply of calcium. J. Exp. Med. 49, 145 – 162 (1929). 6. R. Burrows, Variations produced in bones of growing rats by parathyroid extracts. Am. J. Anat. 62, 237 – 290 (1938). 7. L. I. Pugsley and H. Selye, The histological changes in the bone responsible for the action of parathyroid hormone on the calcium metabolism of the rat. J. Physiol. (London) 79, 113 – 117 (1933). 8. H. Selye, On the stimulation of new bone-formation with parathyroid extract and irradiated ergosterol. J. Endocrinol. 16, 547 – 558 (1932). 9. D. H. Shelling, D. E. Asher, and D. A. Jackson, Calcium and phosphorus studies. VII. The effects of variations in dosage of parathormone and calcium and phosphorus in the diet on the concentrations of calcium and phosphorus in the serum and on the histology and chemical composition of the bones of rats. Bull. Johns Hopkins Hosp. 53, 348 – 389 (1933). 10. T. B. Connor, J. Freijanes, R. E. Stoner, L. G. Martin, and J. Jowsey, Generalized osteosclerosis in primary hyperparathyroidism. Trans. Am. Clin. Climatol. Assoc. 85, 185 – 201 (1973). 11. H. K. Genant, J. M. Baron, E. Paloyan, and J. Jowsey , Osteosclerosis in primary hyperparathyroidism. Am. J. Med. 59, 104 – 113 (1975). 12. E. Bülbring, U¨ ber die Beziehungen zwischen Epithelk¨orperchen, Calciumstoffwechsel und Knochenwachstum. Arch. Exp. Pat. Pharm. 162, 209 – 248 (1931). 13. D. L. Thomson and L. I. Pugsley, On the mechanism of parathyroid hormone action. Am. J. Physiol. 102, 350 – 354 (1932).

14. H. L. Jaffé, Hyperparathyroidism (Recklinghausen’s disease of bone). Arch. Pathol. 16, 63 – 112, 236-258 (1933). 15. D. N. Kalu, J. Pennock, F. H. Doyle, and G. V. Foster, Parathyroid hormone and experimental osteosclerosis. Lancet 1, 1363 – 1366 (1970). 16. D. G. Walker, The induction of osteoporotic changes in hypophysectomised thyroparathyroidectomised and intact rats of various ages. Endocrinology 89, 1389 – 1406 (1971). 17. J. Reeve, G. W. Tregear, and J. A. Parsons, Preliminary trial of low doses of human parathyroid hormone 1 – 34 peptide in treatment of osteoporosis. Calcif. Tissue Res. 21 (Suppl.), 469 – 477 (1976). 18. J. Reeve, R. Hesp, D. Williams, P. Hulme, L. Klenerman, J. Zanelli, et al. The anabolic effect of low doses of human parathyroid hormone fragment on the skeleton in postmenopausal osteoporosis. Lancet 1, 1035 – 1038 (1976). 19. J. Reeve, P. J. Meunier, J. A. Parsons, M. Bernat, OLM. Bijvoet, P. Courpron, et al. The anabolic effect of human parathyroid hormone fragment (hPTH 1 – 34) therapy on trabecular bone in involutional osteoporosis: Report of a multi-centre trial. Br. Med. J. 280, 1340 – 1344 (1980). 20. C. S. Tam, J. N. M. Heersche, T. M. Murray, and Parsons J. A. Parathyroid hormone stimulates the bone apposition rate independently of its resorptive action: Differential effects of intermittent and continuous administration. Endocrinology 110, 506 – 512 (1982). 21. Podbesek R. C. Edouard, P. J. Meunier, J. A. Parsons, J. Reeve, R. W. Stevenson, et al. Effects of two treatment regimes with synthetic human parathyroid hormone fragment on bone formation and the tissue balance of trabecular bone in greyhounds. Endocrinology 112, 1000 – 1006 (1983). 22. R. D. Podbesek, E. B. Mawer, G. D. Zanelli, J. A. Parsons, and Reeve J. Intestinal absorption of calcium in greyhounds: the response to intermittent and continuous administration of human synthetic parathyroid hormone fragment 1 – 34 (hPTH 1-34). Clin. Sci. 67, 591 – 599 (1984). 23. T. B. Usdin, C. Gruber, and T. I. Bonner, Identification and functional expression of a receptor selectively recognising parathyroid hormone, the PTH2 receptor. J. Biol. Chem. 270, 15455 – 15458 (1995). 24. H. Schneider, J. H. M. Feyen, K. Seuwen, and N. R. Movva, Cloning and functional expression of a human parathyroid hormone receptor. Eur. J. Pharmacol. 246, 149 – 155 (1993). 25. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, et al. Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 399 (1997). 26. B. R. Wong, J. Rho, J. Arron, E. Robinson, J. Orlinick, M. Chao, et al. TRANCE is a novel ligand of the tumour necrosis factor receptor family that activates c-Jun N-terminal kinase in T cells. J. Biol. Cheml. 272, 25190 – 25194 (1997). 27. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. U.S.A. 95, 3597 – 3602 (1998). 28. J. E. Onyia, R. R. Miles, D. L. Halladay, S. Chandrasekhar, and T. J. Martin, in vivo demonstration that parathyroid hormone (hPTH 1 – 38) inhibits the expression of osteoprotegerin (OPG) in bone with the kinetics of an immediate early gene. J. Bone Miner. Res. 14s 166, (1999). [abstract]. 29. B. M. Thompson, S. J. Atkinson, and A. M. McGarrity, Type I collagen degradation by mouse calvarial osteoblasts stimulated with 1,25 – dihydroxyvitamin D-3: Evidence for a plasminogen-plasminmetalloproteinase activation cascade. Biochim. Biophys. Acta 1014, 125 – 132 (1989).

CHAPTER 77 Treatment with PTH Peptides 30. H. Dobnig and R. T. Turner, Evidence that intermittent treatment with parathyroid hormone increases bone formation in adult rats by activation of bone lining cells. Endocrinology 136, 3632 – 3638 (1995). 31. D. Leaffer, M. Sweeney, L. A. Kellerman, Z. Avnur, J. L. Krstenansky, B. H. Vickery, et al. Modulation of osteogenic cell ultrastructure by RS-23581, an analog of human parathyroid hormone (PTH)-related peptide-(1 – 34) and bovine PTH-(1 – 34). Endocrinology 136, 3624 – 3631 (1995). 32. A. B. Hodsman and B. M. Steer, Early histomorphomorphic changes in response to parathyroid hormone therapy in osteoporosis: Evidence for de novo bone formation on quiescent cancellous surfaces. Bone 14, 523 – 527 (1993). 33. R. L. Jilka, R. S. Weinstein, T. Bellido, P. Robertson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). 34. J. N. Bradbeer, M. Arlot, P. J. Meunier, and J. Reeve, Treatment of osteoporosis with parathyroid peptide (hPTH 1 – 34) and oestrogen: Increase in volumetric density of iliac cancellous bone, may depend on reduced trabecular spacing as well as increased thickness of packets of newly formed bone. Clin. Endocrinol. 37, 282 – 289 (1992). 35. E. Canalis, T. L. McCarthy, and M. Centrella, Differential effects of continuous and transient treatment with parathyroid hormone related peptide (PTHrP) on bone collagen synthesis. Endocrinology 126, 1806 – 1812 1990. 36. E. Canalis, M. Centrella, W. Burch, and T. McCarthy, Insulin-like growth factor 1 mediates selective anabolic effects of parathyroid hormone in bone cultures. J. Clin. Invest. 83, 60 – 65 (1989). 37. T. L. McCarthy, M. Centrella, and E. Canalis, Parathyroid hormone enhances the transcript and polypeptide levels of insulin-like growth factor I in osteoblast-enriched cultures from fetal rat bone. Endocrinology 124, 1247 – 1253 (1989). 38. P. Watson, D. Lazowski, V. Han, L. Fraher, B. Steer and A. Hodsman, Parathyroid hormone restores bone mass and enhances osteoblast insulin-like growth factor I gene expression in ovariectomized rats. Bone 16, 357 – 365 (1995). 39. E. A. Harrington, A. Fanidi, and G. I. Evan, Oncogenes and cell death. Genet. Dev. 4, 120 – 129 (1994). 40. M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, et al. Modulation of transforming growth factor- production in normal human osteoblast-like cells by 17 estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). 41. S. Mori, W. S. S. Jee, and Li XJ. Production of new trabecular bone in osteopenic ovariectomized rats by prostaglandin E2. Calcif. Tissue Int. 50, 80 – 87 (1992). 42. C. A. Walsh, M. A. Birch, W. D. Fraser, R. Lawton, J. Dorgan, S. Walsh, et al. Expression and secretion of parathyroid hormone-related protein by human bone-derived cells in vitro: Effects of glucocorticoids. J. Bone Miner. Res. 1, 17 – 25 (1986). 43. H. L. Guenther, W. Hofstetter, J. M. Moseley, M. T. Gillespie, N. Suda, and T. J. Martin, Evidence for the synthesis of parathyroid hormone-related protein (PTHrP) by nontransformed clonal rat osteoblastic cells in vitro. Bone 16, 341 – 347 (1995). 44. D. N. Kalu, The ovariectomized rat model of postmenopausal bone loss. Bone Miner. 15, 175 – 192 (1991). 45. H. M. Frost and W. S. S. Jee, On the rat model of human osteopenias and osteoporoses. Bone Miner. 18, 227 – 236 (1992). 46. R. Baron, R. Tross, and A. Vignery, Evidence of sequential remodeling in rat trabecular bone: Morphology, dynamic histomorphometry, and changes during skeletal maturation. Anat. Rec. 208, 137 – 145 (1984). 47. A. Vignery and R. Baron, Dynamic histomorphometry of alveolar bone remodeling in the adult rat. Anat. Rec. 196, 191 – 200 (1990).

743 48. P. Tran Van, A. Vignery, and R. Baron, An electron-microscopic study of the bone remodeling sequence in the rat. Cell Tissue Res. 225, 283 – 292 (1982). 49. P. Tran Van, A. Vignery, and R. Baron, Cellular kinetics of the bone remodeling sequence in the rat. Anat. Rec. 202, 445 – 451 (1982). 50. C. P. Jerome, C. S. Johnson, H. T. Vafai, K. C. Kaplan, J. Bailey, B. Capwell, et al. Effect of treatment for 6 months with human parathyroid hormone (1 – 34) peptide in ovariectomized cynomolgus monkeys (Macaca fascicularis). Bone 25, 301 – 309 (1999). 51. P. D. Delmas, P. Vergnand, M. E. Arlot, P. Pastoureau, P. J. Meunier, and M. H. L. Nilsson, The anabolic effect of Human PTH (1 – 34) on bone formation is blunted when bone resorption is inhibited by the bisphosphonate tiludronate — Is activated resorption a prerequisite for the in vivo effect of PTH on formation in a remodeling system? Bone 16, 603 – 610 (1995). 52. E. Hefti, U. Trechsel, J. P. Bonjour, H. Fleisch, and R. Schenk. Increase of whole body calcium and skeletal mass in normal and osteoporotic adult rats treated with parathyroid hormone. Clin. Sci. 62, 389 – 396 (1982). 53. M. Gunness-Hey and J. M. Hock, Increased trabecular bone mass in rats treated with human synthetic parathyroid hormone. Metab. Bone Dis. Relat. Res. 5, 177 – 181 (1984). 54. J. M. Hock, I. Gera, J. Fonseca, and L. G. Raisz, Human parathyroid hormone-(1 – 34) increases bone mass in ovariectomized and orchidectomized rats. Endocrinology 122, 2899 – 2904 (1988). 55. M. Gunness-Hey, I. Gera, J. Fonseca, L. G. Raisz, and J. M. Hock, 1,25 Dihydroxyvitamin D3 alone or in combination with parathyroid hormone does not increase bone mass in young rats. Calcif. Tissue Int. 43, 284 – 288 (1988). 56. J. M. Hock and J. Fonseca, Anabolic effect of human synthetic parathyroid hormone-(1 – 34) depends on growth hormone. Endocrinology 127, 1804 – 1810 (1990). 57. T. J. Wronski, C. C. Walsh, and L. J. Ingaszewski, Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone 7, 119 – 123 (1986). 58. T. J. Wronski, M. Cintron, and L. M. Dann, Temporal relationship between bone loss and increased bone turnover in ovariectomized rats. Calcif. Tissue Int. 43, 179 – 183 (1988). 59. T. J. Wronski, L. M. Dann, K. S. Scott, and M. Cintron, Long-term effects of ovariectomy and aging on the rat skeleton. Calcif. Tissue Int. 45, 360 – 366 (1989). 60. T. J. Wronski, L. M. Dann, and S. L. Horner, Time course of vertebral osteopenia in ovariectomized rats. Bone 10, 295 – 301 (1989). 61. T. J. Wronski, M. Cintron, A. L. Doherty, and L. M. Dann, Estrogen treatment prevents osteopenia and depresses bone turnover in ovariectomized rats. Endocrinology 123, 681 – 686 (1988). 62. T. J. Wronski, L. M. Dann, K. S. Scott, and L. R. Crooke, Endocrine and pharmacological suppressors of bone turnover protect against osteopenia in ovariectomized rats. Endocrinology 125, 810 – 816 (1989). 63. T. J. Wronski, C.-F. Yen, H. Qi, and L. M. Dann, Parathyroid hormone is more effective than estrogen or hisphosphonates for restoration of lost bone mass in overiectomized rats. Endocrinology 132, 823 – 833 (1993). 64. M. Hori, T. Uzawa, K. Morita, T. Noda, H. Takahashi, and J. Inoue, Effect of human parathyroid hormone (PTH (1 – 34)) on experimental osteopenia of rats induced by ovariectomy. Bone Miner. 3, 193 – 199 (1988). 65. K. Tada , T. H. Yamamuro, R. Okamura, Kasai, and H. E. Takahashi, Restoration of axial and appendicular bone volumes by hPTH(1 – 34) in parathyroidectomized and osteopenic rats. Bone 11, 163 – 169 (1990). 66. H. E. Takahashi, T. Tanizawa, M. Hori, and T. Uzawa, Effect of intermittent administration of human parathyroid hormone (1 – 34) on

744

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

MOSEKILDE AND REEVE experimental osteopenia in rats induced by ovariectomy. Cells. Mater. (Suppl. 1). 113 – 117 (1991). C. C. Liu and D. N. Kalu, Human parathyroid hormone-(1 – 34) prevents bone loss and augments bone formation in sexually mature ovariectomized rats. J. Bone Miner. Res. 5, 973 – 982 (1990). C. C. Liu, D. N. Kalu, E. Salerno, R. Echon, B. W. Hollis, and M. Ray, Pre-existing bone loss associated with ovariectomy in rats is reversed by parathyroid hormone. J. Bone Miner Res. 6, 1071 – 1080 (1991). M. Gunness-Hey and J. M. Hock, Loss of the anabolic effect of parathyroid hormone on bone after discontinuation of hormone in rats. Bone 10, 447 – 452 (1989). T. J. Wronski and C.-F. Yen, Anabolic effects of parathyroid hormone on cortical bone in overiectomized rats. Bone 15, 501 – 508 (1994). P. Morley, J. F. Whitfield, G. E. Willick, V. Ross, S. Maclean, R. J. Isaacs, et al. Prolonged low-dose infusion of human parathroid hormone does not increase femoral cancellous bone volume in ovariectomized rats. Eur. J. Endocrinol. 141, 70 – 74 (1999). N. Okimoto, H. Tsurukami, Y. Okazaki, S. Nishida, A. Sakai, H. Ohnishi, et al. Effects of weekly injection of human parathyroid hormone (1 – 34) and withdrawal on bone mass, strength, and turnover in mature ovariectomized rats. Bone 22, 523 – 531 (1998). Li. Mosekilde, C. H. Søgaard, C. C. Danielsen, O. Tørring O, and M. H. L. Nilsson. The anabolic effect of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: A comparison study between hPTH(1 – 34) and hPTH(1 – 84). Endocrinology 129, 421 – 428 (1991). C. Ejersted, T. T. Andreassen, H. Oxlund, P. H. Jurgensen, B. Bak, J. Heggblad, et al. Human parathyroid hormone (1 – 34) and (1 – 84) increase the mechanical strength and thickness of cortical bone in rats. J. Bone Miner. Res. 9, 1097 – 1101 (1993). H. Oxlund, C. Ejersted, T. T. Andreassen, O. Torring, and M. H. L. Nilsson, Parathyroid hormone (1 – 34) and (1 – 84) stimulate cortical bone formation both from periosteum and endosteum. Calcif. Tissue Int. 53, 394 – 399 (1993). D. B. Kimmel, R. P. Bozzato, K. A. Kronis, T. Coble, D. Sindrey, P. Kwong, et al. The effect of recombinant human (1 – 84) or synthetic human (1 – 34) parathyroid hormone on the skeleton of adult osteopenic ovariectomized rats. Endocrinology 132, 1577 – 1583 (1993). J. M. Hock, J. R. Hummert, R. W. Boyce, J. Fonseca, and L. G. Raisz, Resorption is not essential for the stimulation of bone growth by hPTH-(1 – 34) in rats in vivo. J. Bone Miner. Res. 4, 449 – 458 (1989). Li. Mosekilde, C. H. Søgaard, J. E. McOsker, and T. J. Wronski, PTH has a more pronounced effect on vertebral bone mass and biomechanical competence than antiresorptive agents (estrogen and bisphosphonate ) — Assessed in sexually mature, ovariectomized rats. Bone 15, 401 – 408 (1994). Li. Mosekilde, C. C. Danielsen, and J. A. Gasser, The effect on vertebral bone mass and strength of long term treatment with antiresorptive agents (Estrogen and Calcitonin), human parathyroid hormone-(1 – 38), and combination therapy, assessed in aged ovariectomized rats. Endocrinology 135, 2126 – 2134 (1994). V. Shen, D. W. Dempster, R. Birchman, R. Xu, and R. Lindsay, Loss of cancellous bone mass and connectivity in ovariectomized rats can be restored by combined treatment with parathyroid hormone and estradiol. J. Clin. Invest. 91, 2479 – 2487 (1993). T. J. Wronski, S. Pun, and H. Liang, Effects of age, estrogen depletion and parathyroid hormone treatment on vertebral cancellous wall width in female rats. Bone 25, 465 – 468 (1999). Li. Mosekilde, C. C. Danielsen, C. H. S. Søgaard, J. E. McOsker, and T. J. Wronski, The anabolic effects of parathyroid hormone on cortical bone mass, dimensions and strength — Assessed in a sexually mature, ovariectomized rat model. Bone 16, 223 – 230 (1995).

83. C. H. Søgaard, Li. Mosekilde, A. Richards, and Le Mosekilde. Marked decrease in trabecular bone quality after five years of sodium fluoride therapy — assessed by biomechanical testing of iliac crest bone biopsies in osteoporotic patients. Bone 15, 393 – 399 (1994). 84. C. H. Søgaard, T. J. Wronski, J. E. McOsker, and Li. Mosekilde. The positive effect of parathyroid hormone on femoral neck bone strength in ovariectomized rats is more pronounced than that of estrogen or bisphosphonates. Endocrinology 134, 650 – 657 (1994). 85. C. H. Søgaard, C. C. Danielsen, E. B. Thorling, and Li. Mosekilde. Long-term exercise of young and adult female rats: Effect on femoral neck biomechanical competence and bone structure. J. Bone Miner. Res. 9, 409 – 416 (1994). 86. A. Toromanoff, P. Ammann, and J.-L. Riond, Early effects of shortterm parathyroid hormone administration on bone mass, mineral content and strength in female rats. Bone 22, 217 – 223 (1998). 87. J. W. M. Chow, S. Fox, C. J. Jagger, and T. J. Chambers, Role for parathyroid hormone in mechanical responsiveness of rat bone. Am. J. Physiol. Endocrinol. Metab. 37, (1998). 274:E146-54. 88. R. T. Turner, G. L. Evans, J. M. Cavolina, B. Halloran, and E. MoreyHolton, Programmed administration of parathyroid hormone increases bone formation and reduces bone loss in hindlimb-unloaded ovariectomised rats. Endocrinology 139, 4086 – 4091 (1998). 89. W. S. S. Jee, S. Mori, X. J. Li, and S. Chan, Prostaglandin E2 enhances cortical bone mass and activates intracortical bone remodeling in intact and ovariectomized female rats. Bone 11, 253 – 266 (1990). 90. Li. Mosekilde, J. S. Thomsen, and J. E. McOsker, No loss of biomechanical effects after withdrawal of short-term PTH treatment in an aged, osteopenic, ovariectomised rat model. Bone 20, 429 – 437 (1997). 91. C. Ejersted, H. Oxlund, and T. T. Andreassen, Bisphosphonate maintains parathyroid hormone (1 – 34)-induced cortical bone mass and mechanical strength in old rats. Calcif. Tissue Int. 62, 316 – 322 (1998). 92. T. Kishi, H. Hagino, H. Kishimoto, and H. Nagashima, Bone responses at various skeletal sites to human parathyroid hormone in ovariectomized rats: Effects of long-term administration, withdrawal and readministration. Bone 22, 515 – 522 (1998). 93. A. B. Hodsman, P. H. Watson, D. Drost, D. Holdsworth, M. Thornton, J. Hock, et al. Assessment of maintenance therapy with reduced doses of PTH(1 – 34) in combination with raloxifene analogue (LY 117018) following anabolic therapy in the ovariectomized rat. Bone 24, 451 – 455 (1999). 94. V. Shen, X. Birchman, G. Liang, D. D. Wu, D. W. Dempster, and R. Lindsay, Accretion of bone mass and strength with parathyroid hormone prior to the onset of estrogen deficiency can provide temporary beneficial effects in skeletally mature rats. J. Bone Miner. Res. 13, 883 – 890 (1998). 95. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. V. S. Chao, H. W. Wahner, J. M. Muhs, et al. Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). 96. M. Kleerekoper, E. Peterson, D. A. Nelson, E. Philips, M. Schork, B. C. Tilley, and A. M. Parfitt, A randomized trial of sodium fluoride as a treatment for post-menopausal osteoporosis. Osteoporosis Int. 1, 155 – 162 (1991). 97. Li. Mosekilde, J. Kragstrup, and A. Richards, Compressive strength, ash weight, and volume of vertebral trabecular bone in experimental fluorosis in pigs. Calcif. Tissue Int. 39, 69 – 73 (1987). 98. C. H. Søgaard, Li. Mosekilde, W. Schwartz, G. Leidig, H. W. Minne, and R. Ziegler, Effects of fluoride treatment on rat vertebral body biomechanical competence and bone mass. Bone 16, 163 – 169 (1995).

CHAPTER 77 Treatment with PTH Peptides 99. M. W. Lundy, M. Stauffer, J. E. Wegerdal, D. J. Baylink, J. D. B. Featherstone, S. F. Hodgson, et al. Histomorphometric analysis of iliac crest bone biopsies in placebo-treated versus fluoride-treated subjects. Osteoporosis Int. 5, 115 – 129 (1995). 100. Li. Mosekilde, J. S. Thomsen, P. B. Orhii, and D. N. Kalu, Growth hormone increases vertebral and femoral bone strength in osteopenic, ovariectomized aged rats in a dose-dependent and sitespecific manner. Bone 23, 343 – 352 (1998). 101. J. S. Thomsen, J. A. Gasser, and Li. Mosekilde, Long-term therapy of ovariectomy-induced osteopenia with parathyroid hormone analog SDZ. PTS 893 and bone maintenance in retired breeder rats. Bone 25, 561 – 569 (1999). 102. Li. Mosekilde, J. S. Thomsen, P. B. Orhii, R. J. McCarter, W. Mejia, and D. N. Kalu, Additive effect of voluntary exercise and GH treatment on bone strength — Assessed at four different skeletal sites in a rat model. Bone 24, 71 – 80 (1999). 103. G. Thamsborg, T. L. Storm, R. Sykluski, H. K. Nielsen, and O. H. Sørensen, Effect of different doses of nasal salmon calcitonin on bone mass. Calcif. Tissue Int. 48, 302 – 307 (1991). 104. J.-Y. Reginster, Management of high turnover osteoporosis with calcitonin. Bone 13, S37 – S40 (1992). 105. S. T. Harris, H. K. Genant, D. J. Baylink, J. C. Gallagher, S. K. Karp, M. A. McConnell, et al. The effects of estrogen on spinal bone density of postmenopausal women. Arch. Intern. Med. 151, 1980 – 1984 (1991). 106. K. Overgaard, M. A. Hansen, S. B. Jensen, and C. Christiansen, Effect of salcalcitonin given intranasally on bone mass and fracture rates in established osteoporosis: A dose – response study. Br. Med. J. 305, 556 – 561 (1992). 107. J. M. Zanelli, B. Aufmkolk, E. Lane, and C. Ball, The biological activity of the synthetic 1 – 38 sequence of human parathyroid hormone in in vivo and in vitro bioassay systems. In “Calcium Regulation and Bone Metabolism” (D. V. Cohn, T. J. Martin, and P. J. Meunier, eds.), p. 561. Excerpta Medica, Amsterdam, (1987). 108. H. Plotkin, C. Gundberg, M. Mitnick, and A. F. Stewart, Dissociation of bone formation from resorption during 2-week treatment with human parathyroid hormone-related peptide-(1 – 36) in humans: Potential as an anabolic therapy for osteoporosis. J. Clin. Endocrinol. Metab. 83, 2786 – 2791 (1998). 109. G. N. Kent, N. Loveridge, J. Reeve, and J. M. Zanelli, Pharmacokinetics of synthetic human parathyroid hormone 1 – 34 in man measured by cytochemical bioassay and radioimmunoassay. Clin. Sci. 68, 171 – 177 (1985). 110. D. M. Slovik, R. M. Neer, and J. T. Potts Jr, Short-term effects of synthetic human parathyroid hormone (1 – 34) administration on bone mineral metabolism in osteoporotic patients. J. Clin. Invest. 68, 1261 – 1271 (1981). 111. J. Bonadio, E. Smiley, P. Patil, and S. Goldstein, Localised, direct plasmid gene delivery in vivo: Prolonged therapy results in reproducible tissue regeneration. Nat. Med. 5, 753 – 759 (1999). 112. M. Tellez, J. M. Zanelli, R. Wootton, and J. Reeve, Dose variation in the treatment of osteoporosis with synthetic human parathyroid hormone fragment hPTH 1 – 34. Clin. Sci. 60, 26P 1981. [abstract] 113. J. Reeve, M. E. Arlot, M. Bernat, C. Edouard, R. Hesp, D. Slovik, et al. Treatment of osteoporosis with human parathyroid hormone fragment 1 – 34: A positive final tissue balance in trabecular bone. Metab. Bone. Dis. Relat. Res. 2, 349 – 354 (1980). 114. M. Audran, M.-F. Basle, A. Defontaine, P. Jallet, M. T. Bidet, A. Ermias, et al. Transient hypoparathyroidism induced by synthetic hPTH 1-34 treatment. J. Clin. Endocrinol. Metab. 64, 937 – 943 (1987). 115. J. Reeve, U. M. Davies, R. Hesp, E. McNally, and D. Katz, Human parathyroid peptide treatment of osteoporosis substantially increases spinal trabecular bone (with observations on the effects of sodium fluoride therapy). Br. Med. J. 301, 314 – 318 and 477 (1990).

745 116. J. Reeve, J. N. Bradbeer, M. E. Arlot, U. M. Davies, J. R. Green, L. Hampton, et al. hPTH 1 – 34 treatment of osteoporosis with added hormone replacement therapy: Biochemical, kinetic and histological responses. Osteoporosis Int. 1, 162 – 170 (1991). 117. T. Uzawa, M. Hori, S. Ejiri, and H. Ozawa, Comparison of the effects of intermittent and continuous administration of human parathyroid hormone (1 – 34) on rat bone. Bone 16, 477 – 484 (1995). 118. J. A. Parsons and B. Reit, Chronic response of dogs to parathyroid hormone infusion. Nature (London) 230, 254 – 257 (1974). 119. J. S. Finkelstein, A. Klibanski, A. L. Arnold, T. L. Toth, M. D. Hornstein, and R. M. Neer, Prevention of estrogen deficiency-related bone loss with human parathyroid hormone-(1 – 34): A randomised trial. J. Am. Med. Assoc. 280, 1067 – 1073 (1998). 120. N. E. Lane, S. Sanchez, G. W. Modlin, H. K. Genant, E. Pierini, and C. D. Arnaud, Parathyroid hormone treatment can reverse corticosteroid-induced osteoporosis — results of a randomised controlled clinical trial. J. Clin. Invest. 102, 1627 – 1633 (1998). 121. T. Fujita, T. Inoue, R. Morii, H. Norimatsu, H. Orimo, et al. Effect of an intermittent weekly dose of human parathyroid hormone (1 – 34) on osteoporosis: A randomised double-masked prospective study using three dose levels. Osteoporosis Int. 9, 296 – 306 (1999). 122. R. Neer, D. Slovik, M. Daly, C. Lo, J. Potts, and S. Nussbaum, Treatment of postmenopausal osteoporosis with daily parathyroid hormone plus calcitriol. In C. Christiansen, K. Overgaard, (eds.) “Osteoporosis”Osteopress Copenhagen, ApS, 1990 Vol. 3, pp.1314 – 1317. 123. D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. T. Potts Jr, M. Daly, J. A. Campbell, et al. Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1 – 34) and 1,25-dihydroxyvitamin. J. Bone Miner. Res. 1, 377 – 381 (1986). 124. J. M. Hock and I. Gera, Effects of continuous and intermittent administration and inhibition of resorption on the anabolic response of bone to parathyroid hormone. J. Bone Miner. Res. 7, 65 – 72 (1992). 125. R. Lindsay, J. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, et al. Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550 – 555 (1997). 126. E. B. Roe, S. D. Sanchez , G. A. del Puerto, E. Pierini, P. Bacchetti, C. Cann, et al. Parathyroid hormone 1 – 34 (hPTH 1 – 34) and estrogen produce dramatic bone density increases in postmenopausal osteoporosis – Results from a placebo-controlled randomized trial. J. Bone Miner. Res. 14, s137 (1999).[abstract] 127. J. Reeve, A, Mitchell, M. Tellez, P. Hulme, J. R. Green, B. WardleySmith, and R. Mitchell, Treatment with parathyroid peptides and estrogen replacement for severe postmenopausal vertebral osteoporosis: Long-term effects on spine and femur and determinants of magnitude of response. J. Bone Min. Metabol. 19(2), 102 – 114 (2001). 128. Li. Mosekilde, J. S. Thomsen, M. Kneissel, and J. A. Gasser, Alendronate blunts the anabolic effect of a PTH analog SDZ. PTS 893 when administered sequentially. J. Bone Miner. Res. 12(Suppl. 1), s343 (1997). [abstract] 129. R.-D. Hesch, U. Busch, M. Prokop, G. Delling, and E.-F. Rittinghaus, Increase of vertebral density by combination therapy with pulsatile 1 – 38 hPTH and sequential addition of calcitonin nasal spray in osteoporotic patients. Calcif. Tissue Int. 44, 176 – 180 (1989). 130. A. B. Hodsman, B. M. Steer, L. J. Fraher, and D. J. Drost, Bone densitometric and histomorphometric responses to sequential human parathyroid hormone 1 – 38) and salmon calcitonin in osteoporotic patients. Bone Miner. 14, 67 – 83 (1991). 131. A. B. Hodsman, L. J. Fraher, T. Ostbye, J. D. Adachi, and B. M. Steer, An evaluation of several biochemical markers for bone formation and resorption in a protocol utilizing cyclical parathyroid hormone and calcitonin therapy for osteoporosis. J. Clin. Invest. 91, 1138 – 1148 (1993).

746 132. J. Reeve, M. E. Arlot, T. R. Price, C. Edouard, R. Hesp, P. Hulme, et al. Periodic courses of human 1 – 34 parathyroid peptide alternating with calcitriol paradoxically reduce bone remodelling in spinal osteoporosis. Eur. J. Clin. Invest. 17, 421 – 428 (1987). 133. N. A. Barnicot, The local action of the parathyroid and other tissues on bone in intracerebral grafts. J. Anat. (London) 82, 233 – 248 (1948). 134. H. Y. Chang, Grafts of parathyroid and other tissues to bone. Anat. Rec. 111, 23 – 47 (1951). 135. J. Reeve, M. Arlot, R. Wootton, C. Edouard, M. Tellez, R. Hesp, et al. Skeletal blood flow, illac histomorphometry and strontium kinetics in osteoporosis: A relationship between blood flow and corrected apposition rate. J. Clin. Endocrinol. Metab. 66, 1124 – 1131 (1988). 136. R. Wootton, J. Reeve, E. Spellacy, and M. Tellez-Yudilevich, Skeletal blood flow in Paget’s disease and its response to calcitonin therapy. Clin. Sci. Mol. Med. 54, 69 – 74 (1978). 137. J. Reeve, R. Hesp, R. Wootton, G. D. Zanelli, D. Slovik, R. M. Neer, et al. Clinical trial of hPTH in idiopathic osteoporosis. An interim report. In “Endocrinology of Calcium Metabolism: Proceedings of the 6th Parathyroid Conference” (D. H. Copp and T. V. Talmage, eds.), Excerpta Medica, Amsterdam, Vancouver, BC, pp. 71 – 75 (1979). 138. R. Hesp, P. Hulme, D. Williams, and J. Reeve, The relationship between changes in femoral bone density and calcium balance in patients with involutional osteoporosis treated with human parathyroid hormone fragment (hPTH 1 – 34). Metab. Bone Dis. Relat. Res. 2, 331 – 334 (1981). 139. C. E. Cann, E. B. Roe, S. D. Sanchez, and C. D. Arnaud, PTH effects in the femur: Envelope-specific responses by 3DQCT in postmenopausal women. J. Bone Miner. Res. 14, s137 (1999).[abstract] 140. M. Vogel, R. D. Hesch, and G. Delling, Morphologische Untersuchung der Beckenkammspongiosa bei Patienten mit Osteoporose unter einer Kombinationstherapie mit pulsatiler Gabe von Parathormon (1 – 38 hPTH) und sequentieller Verabreichung von CalcitoninNasenspray. Med. Klin. 85, 82 – 86 (1990). 141. A. Hodsman, L. Fraher, P. Watson, T. Ostbye, L. Stitt, J. Adachi, et al. A randomized controlled trial to compare the efficacy of cyclical parathyroid hormone versus cyclical parathyroid hormone and sequential calcitonin to improve bone mass in postmenopausal women with osteoporosis. J. Clin. Endocrinol. Metab. 82, 620 – 628 (1997). 142. R. M. Neer, C. D. Arnaud, J. R. Zanchetta, R. Prince, G. A. Gaich, J. Y. Reginster, A. B. Hodsman, E. F. Eriksen, S. Ish-Shalom, H. K. Genant, O. Wang, B. H. Mitlak, D. Mellstrom, E. S. Oefjord, E. Marcinowska-Sucherowierska, J. Salmi, L. Gasper, H. Mulder, J. Halse, and A. Z. Sawicki, Recombinant human PTH (1 – 34) fragment [rhPTH] reduces the risk of spine and non-spine fractures in post-menopausal osteoporosis. N. Engl. J. Med. 344, 1434 – 1441 (2001). 143. R. Marcus, G. A. Gaich, J. H. Satterwhite, S. L. Myers, O. Wang, B. H. Mitlak, Effects of baseline BMD, age and prevalent vertebral fractures on the response of osteoporotic patients to LY333334 [rhPTH(1 – 34)]. J. Bone Miner. Res. 15(Suppl.1), s194 2000. [abstract] 144. J. A. Parsons, P. J. Meunier, R. M. Neer, R. D. Podbesek, and J. Reeve, Effect of synthetic human parathyroid hormone fragment (hPTH 1-34) on bone mass and bone mineral metabolism. In (H. F de Luca et al. ed.), pp.457 – 465 “Osteoporosis: Recent Advances in Pathogenesis and Treatment” University Park Press, Baltimore, 1981.

MOSEKILDE AND REEVE 145. J. Reeve, O. L. M. Bijvoet, R. M. Neer, D. Slovik, M. Tellez, F. J. F. E. Vismans, et al. A comparison between the balance method and radiotracer methods for measuring calcium absorption in treated and untreated patients with osteoporosis. Metab. Bone Dis. Relat. Res. 2, 233 – 238 (1980). 146. J. E. Compston, K. Yamaguchi, P. I. Croucher, N. J. Garrahan, PC. Lindsay, and R. W. Shaw, The effects of gonadotrophin-releasing hormone agonists on iliac crest cancellous bone structure in women with endometriosis. Bone 16, 261 – 268 (1995). 147. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Incidence of clinically diagnosed vertebral fractures: A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 148. B. Ettinger, D. Black, M. Nevitt, A. C. Rundle, J. A. Cauley, S. R. Cummings, et al. Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). 149. M. C. Nevitt, D. E. Thompson, D. M. Black, S. R. Rubin, K. Ensrud, A. J. Yates, et al. Effect of alendronate on limited-activity days and bed-disability days caused by back pain in postmenopausal women with existing vertebral fractures. Arch. Intern. Med. 160, 77 – 85 (2000). 150. A. Oleksik, W. Shen, A. Dawson, M. Minshal, and P. Lips, The impact of incident vertebral fractures on health-related quality of life in women with prevalent vertebral fractures. J. Bone Miner. Res. 14, (Suppl. I), s262 (1999). [abstract] 151. S. L. Silverman, M. E. Minshall, W. Shen, K. D. Harper, and S. Xie, The inpact of incident vertebral fracture on health-related quality of life in established postmenopausal osteoporosis: Results from the Multiple Outcomes of Raloxifene Evaluation study. J. Bone Miner. Res. 14, (Suppl. I), s159 (1999). [abstract] 152. S. T. Harris, N. B. Watts, H. K. Genant, C. D. McKeever, T. Hangartner, M. Keller, et al. Effects of risedronate treatment on vertebral and non-vertebral fractures in women with postmenopausal osteoporosis. J. Am. Med. Assoc. 282, 1344 – 1352 (1999). 153. D. M. Black, S. R. Cummings, D. B. Karpf, J. A. Cauley, D. E. Thompson, M. C. Nevitt, et al. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 348, 1535 – 1541 (1996). 154. I. MacIntyre, J. C. Stevenson, M. I. Whitehead, S. J. Wimalawansa, L. M. Banks, and M. J. R. Healy, Calcitonin for prevention of postmenopausal bone loss. Lancet 1, 900 – 902 (1988). 155. D. Felsenberg, A. J. Silman, M. Lunt, G. Armbrecht, A. A. Ismail, J. D. Finn et al., Incidence of vertebral fractures in Europe: Results from the European Prospective Osteoporosis Study (EPOS). J. Bone Miner. Res. (in press) (2001). 156. P. D. Ross, J. N. Davis, R. S. Epstein, and R. D. Wasnich, Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 157. P. D. Ross, H. Genant, J. Davis, P. Miller, and R. Wasnich, Predicting vertebral fracture incidence from prevalent fractures and bone density among non-black, osteoporotic women. Osteoporosis Int. 3, 120 – 126 (1993). 158. WHO Study Group, “Assessment of Fracture Risk and its Application to Screening for Postmenopausal Osteoporosis WHO. Technical Report,” Vol. 843. World Health Organisation, Geneva, 1994. 159. G. Mundy, R. Garrett, S. Harris, J. Chan, D. Chen, G. Rossini, et al. Stimulation of bone formation in vitro and in rodents by statins. Science 286, 1946 – 1949 (1999).

CHAPTER 78

Growth Hormone, Insulinlike Growth Factors Potential Applications and Limitations in the Management of Osteoporosis CHRISTIAN WÜSTER CLIFFORD ROSEN

Department of Medicine, Novo Nordisk Pharma, 55127 Mainz, Germany Maine center for Osteoporosis Research and Education, St. Joseph Hospital, Bangor, Maine 04401

I. Introduction II. Physiology of Growth-Hormone-Releasing Hormone (GHRH) – GH – IGF-I III. The Role of GH – IGF-I in Skeletal Physiology IV. Pathophysiology of Osteoporosis: Role of GH – IGFs

V. Growth Hormone Therapy for Osteoporosis VI. IGF-I for the Treatment of Osteoporosis VII. Summary References

I. INTRODUCTION

decade, although mechanisms responsible for this decline have not been clearly defined. The hypothalamic – pituitary axis is similarly affected by aging. Growth hormone (GH) secretion is reduced, resulting in lower levels of circulating insulin-like growth factor-I (IGF-I) [1,2]. Early attempts to link age-related bone loss to a damped GH – IGF-I axis

Biologic aging is a normal physiologic process, part of the continuum from growth to death. Like other organ systems, skeletal homeostasis is maximized during the second and third decades of life. Bone loss begins by the fourth

OSTEOPOROSIS, SECOND EDITION VOLUME 2

747

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

748 spawned considerable interest in growth hormone as a therapeutic tool for osteoporosis [3]. The advent of recombinant gene technology propelled synthetic growth factors into an ever expanding therapeutic domain [4]. Growth hormone treatment for adults became more realistic after 1979. Prior to that time, GH extract could be obtained in the United States only from the National Pituitary Agency on a case-by-case proposition. Most therapeutic applications were restricted to pubertal children with GH deficiency (GHD). With the advent of recombinant human GH (rhGH) large-scale multicentered trials in GH-deficient children were initiated. During that same period, it was recognized that adults who had undergone pituitary surgery or irradiation were either partially or completely GH deficient. GH “replacement” was soon considered not only feasible but desirable by some investigations. After the first trials with rhGH were completed, another therapeutic venue for GH was introduced. rhGH was administered to elders in order to test the widely held thesis that a “somatopause” produced discrete musculoskeletal changes during aging. Data from one small GH trial fueled a growing public desire to test agents which would forestall the aging process [5]. Although initial studies were small and changes relatively minor, novel strategies for treating the senescent skeleton with growth factors began to emerge. This chapter will examine the skeletal effects of two recombinant peptides (GH and IGF-I) and the utility of these growth factors in the treatment of osteoporosis.

II. PHYSIOLOGY OF GROWTHHORMONE-RELEASING HORMONE (GHRH) – GH – IGF-I A. Regulation of the GHRH – GH – IGF System 1. GHRH The regulation of GH secretion from the pituitary is complex and involves elaboration of discrete neurosecretory peptides from the hypothalamus. Hypothalamic releasing factors were postulated to exist for more than five decades, but the exact structure of GHRH was not elucidated until extracts of pancreatic islet cell tumors from two patients with ectopic acromegaly were characterized [6]. Subsequently, two different GHRH peptides, one of 40 amino acids and the other of 44, were isolated from the hypothalamus [7]. GHRH is a potent stimulus for GH release and synthetic analogs are now undergoing clinical trials both for diagnostic and therapeutic purposes in patients with hypothalamic – pituitary disorders. Preliminary investigations examining the utility of these growth-hormone-releasing peptides in elders have also recently been initiated.

WÜSTER AND ROSEN

2. SOMATOSTATIN Somatostatin (SMS) is a small (14-amino-acid) but ubiquitous polypeptide which inhibits GH synthesis and release [8]. In concert with GHRH, SMS regulates GH secretion through a dual control system, the former stimulatory, the latter inhibitory. Several molecular forms of somatostatin, distinct from the native 14-amino-acid peptide, have been isolated. In addition to inhibition of GH release, SMS also inhibits secretion of thyrotropin as well as several pancreatic hormones, including glucagon and insulin. The SMS receptor has been localized to various cell types, especially those of neuroendocrine origin. Localization of this receptor suggests that SMS acts as both an endocrine and a paracrine regulator in diverse tissues. A highly potent synthetic analog of SMS, octreotide, has been used therapeutically in acromegaly and diagnostically (in a radiolabeled form) for scintigraphic visualization of neuroendocrine tumors.

B. Secretion of GH from the Pituitary 1. MECHANISM OF GH SECRETION The secretion of GH is regulated by GHRH and SMS. Both neuropeptides are synthesized in specialized hypothalamic neurons, released from axon terminals in the median eminence, and transported through the hypophyseal portal circulation into the anterior pituitary [9]. GH secretion is pulsatile (due to episodic release of GHRH) and circadian with the highest pulse amplitude occurring between 02:00 and 06:00 [10,11]. Puberty has a dramatic effect on the amplitude of GH pulses, due to changes in the hypothalamic milieu as a result of rising sex steroid concentrations [12]. Apart from tight neuroendocrine regulation of GH secretion by SMS and GHRH, there is a negative feedback loop on GHRH and GH by IGF-I. 2. EFFECTS OF GONADAL STATUS ON GH/IGF-I The pattern of GH secretion in animals and humans depends highly on age and sex [9–12]. Both factors strongly influence the frequency and amplitude of GH pulses, GH basal secretory rates, and the serum concentrations of IGF-I. Characteristic changes during puberty in rats parallel pubertal changes in humans [12]. GH secretion in male and female rats is identical after birth but at puberty, a sexually differentiated pattern of secretion appears with male rats displaying high-amplitude low-frequency pulses, and female rats displaying pulses of high frequency but low amplitude [12]. This sexual dimorphism can be altered by manipulating the gonadal steroid environment, suggesting that sex steroids are important modulators of GH secretion. In humans, sexual differences in GH secretion during puberty are less pronounced, even though administration of gonadal steroids to prepubertal children increases GH pulses

749

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

Comparison of mean  SEM integrated GH curves (IGHC), pulse frequency, amplitutide, and fraction of GH secreted as pulses (FGHP) in 10 young women (ages 18 – 30 years; F; Y) and 10 young men (ages 18 – 30 years; Y; M) and 8 postmenopausal women (ages ears; F; 0) and 8 older men (ages 55 years; O; M) grouped according to sex or age (left) or according to subgroups of young women, older women, and older men (right). Younger women have statistically greater IGHC, amplitudes of GH secretion, and fractions of GH secreted as pulses than older females (right, first two bars). Adapted with permission from Ho et al. [11].

FIGURE 1

and mimics the pubertal milieu of the hypothalamus. Various sampling techniques (profiles versus stimulatory tests) and assays with different sensitivities have produced disparate findings. However, spontaneous and stimulated GH peaks in humans are enhanced during puberty. Matched for age and body mass index, young girls were found to have higher in-

tegrated GH (IGHC) levels than boys [11,13] (Fig. 1). Other secretory characteristics, including pulse amplitude, frequency, and the fraction of GH secreted as pulses (FGHP), were similar in both sexes of the same age. In a preliminary study, black adolescents (males and females) had higher GH secretory rates than age-matched whites [14]. Higher GH

750

WÜSTER AND ROSEN

secretion rates in adolescent blacks could lead to greater acquisition of bone mass. Gonadal steroids affect GH secretion in perimenopausal women. IGHC and mean pulse amplitude of GH and FGHP are lower in older women than in premenopausal women [13] (see Fig. 1). GH secretory indices in postmenopausal women correlate with serum concentrations of estradiol, but not with those of total androgen. During menopause, GH secretion is reduced [11,13]. However, oral administration of estradiol (or conjugated equine estrogens) increases GH secretion as a result of reduced hepatic generation of IGF-I [11,13,15]. On the other hand, transdermal administration of 17- estradiol increases serum IGF-I concentration, suggesting that suppression of IGF-I by oral estrogens is due to “first-pass” hepatic effects [13]. Impaired IGF-I generation in the liver removes a key component of negative feedback on the hypothalamus, resulting in increased GH release [16]. 3. EFFECTS OF AGE ON GH – IGF-I The GH – IGF-I axis undergoes changes over a life span so that elders have lower spontaneous GH secretion rates and serum IGF-I levels than younger people [1,2,17]. These age-related differences are a function of an altered hypothalamic – pituitary set point due in part to changes in lifestyle and nutrition [18]. The GH secretory response to common stimuli such as GHRH, clonidine, L-dopa, physostigmine, pyridostigmine, hypoglycemia, and met-enkephalin, but not arginine, is reduced by aging. Somatotrope responsiveness to GHRH and arginine does not vary with age, implying that the maximal secretory capacity of somatotropic cells is preserved in elderly people [19]. There is fairly strong evidence that GH and IGF-I concentrations decline with advanced age although several large cross-sectional studies have either failed to demonstrate an association between diminished serum IGF-I and age-related bone loss or a weak association between serum IGF-I and bone mineral density[2,19 – 24]. On the other hand, concentrations of IGF-I, IGF-II, and IGFBP-5 in femoral cortical and trabecular bone decline significantly with age [2,19 – 23].

C. Regulation of GH Bioactivity 1. GROWTH HORMONE BINDING PROTEIN Growth hormone exerts a multitude of biological effects on various tissues through the GH receptor. Regulation of GH bioactivity occurs at several pre- and postreceptor levels. Growth hormone binding protein (GHBP) is a plasma binding protein identical to the extracellular domain of the tissue GH receptor [25]. GHBP binds exclusively to GH and most, if not all serum GH, is bound to this carrier protein [27]. Measurements of GH binding protein in serum are relatively stable and reflect the endogenous status of the

GH receptor in responsive tissues [26]. With advanced age, GHBP concentrations increase substantially [24]. 2. INSULIN-LIKE GROWTH FACTORS Most of the effects of GH on hard and soft tissue occur through induction of IGF-I. The presence of this peptide in serum was first postulated by Salmon and Daughaday in 1957 [27]. The nonsuppressible insulin-like activity of these proteins was described in 1963 [28,29]. These factors were called somatomedins primarily because they were considered the sole mediators of growth hormone action [30]. It soon became apparent that the somatomedins had autocrine and paracrine activity independent of GH. Purifi cation of the somatomedins by Rinderknecht and Humbel in 1978 revealed two distinct peptides [31]. Because of their pervasiveness, their diffuse biological activities, and their partial independence from GH, the term somatomedin was abandoned and these peptides were renamed IGF-I and IGF-II [32]. The IGFs are single-chain polypeptides. IGF-I consists of 70 amino-acid residues and IGF-II has 67 amino acids. They have A, B, and C domains similar to proinsulin, but also a D domain which is not found in proinsulin. This D domain may sterically hinder the interaction of the IGFs with the insulin receptor, leading to only weak ligand binding of the IGFs to the insulin receptor. A number of posttranscriptional and posttranslational variants of the IGFs have also been described [33]. These IGFs have variable affinity for insulin-like growth factor binding proteins (IGFBPs) and the IGF receptor. In vitro, these growth factors may have significantly greater activity than native IGFI or IGF-II, especially those that exhibit weak binding to the IGFBPs. IGF-I and IGF-II differ in their ability to promote tissue growth due in part to the presence of distinct IGF receptors (the type I and type II receptors) [34]. The type I dimeric IGF receptor is structurally and functionally homologous to the insulin receptor but binds IGF-I with higher affinity than IGF-II or insulin [34]. It has intrinsic tyrosine kinase activity critical for specific second message generation and indeed, ligand activation of the type I IGF-I receptor stimulates the JAK/STAT- and MAPK (mitogen-activated protein kinase)-signaling pathways [34,35]. The type I IGF-I receptor (IGF-IR), when activated, also has a strong anti-apoptotic effect on cell behavior [34,35]. This characteristic may have prognostic significance in certain malignancies such as breast cancer and may also be important in relation to the fate of osteoblasts. On the other hand, the monomeric type II IGF receptor bears no resemblance to the insulin, or IGF-IR, has a higher affinity for IGF-II than for IGF-I, and cannot bind insulin. The type II receptor exhibits no intrinsic kinase activity but is structurally very similar to the mannose 6-phosphate receptor which is involved in targeting lysosomal enzymes intracellularly [34,35].

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

The IGFs possess tremendous growth potential as endocrine, autocrine, and paracrine factors. However, adequate nutrition is required for the full expression of IGF’s biologic activity, including its critical role in linear growth. For example, during states of malnutrition, growth hormone production increases but hepatic IGF-I generation is severly impaired. Resistance at the hepatic GH receptor reduces serum IGF-I and impairs GH bioactivity. For malnourished children, the result is cessation of linear growth. Growth hormone resistance, to lesser degrees, occurs in other conditions such as diabetes mellitus, acute catabolic stresses, and renal insufficiency. IGFs are produced in virtually every tissue [20]. However, the main source of circulating IGFs is the liver. With acute or chronic hepatic insufficiency both serum IGF-I and serum IGF-II are markedly decreased. Bone is the second richest source of IGFs and may contribute significantly to the circulating IGF pool [20,36]. In the circulation IGFs are bound to serum IGFBPs, with a relatively small but detectable amount of “free” IGF-I which does circulate but has a very short half-life. The distribution of IGFs in the serum pool is determined by the relative saturation of the IGF binding proteins. This may explain why treatment with IGF-I may have different tissue effects than treatment with growth hormone. Infusions of IGF-I produce a transient rise in free IGF-I and suppression of IGF-II, insulin, and endogenous GH [37]. During the course of an IGF-I infusion, however, IGF-I is partitioned into several pools. This is due to the unsaturated nature of the lower molecular weight IGFBPs and the presence of a large (150-kDa) circulating ternary IGF binding complex. This complex, composed of IGF-I (or -II), IGFBP-3, and an acid labile subunit, is the major circulatory reservoir for the IGFs. Normally, the majority of circulating IGF is bound to this saturated intravascular complex. However, with rapid IGF-I infusions, some IGF-I goes into the lower (50 kDa) unsaturated IGFBP fractions where transport into the extravascular space is possible. Partitioning of IGFs into various binding pools is critical to the biologic activity of GH and IGF-I. 3. IGF-BINDING PROTEINS Just as the IGFBPs serve important regulatory functions within the circulation, their role at the tissue level is also critical for the full biologic expression of IGFs. In tissues and the circulation there are six IGFBPs. The predominant binding protein in serum (and bone) is IGFBP-3, a 43-kDa glycosylated peptide. It is present in large concentrations in the serum and is easily measurable by radioimmunoassay (RIA) [38]. As noted earlier, IGFBP-3 is part of a larger saturated ternary complex including IGF-I (or -II) and a 80-kDa acid-labile subunit. The association of these three proteins requires the presence of either IGF-I or IGF-II. In turn, this complex pro-

751 longs the half-life of the IGFs and provides a unique storage site. The concentration of circulating IGFBP-3 is principally controlled by growth hormone [39 – 41]. However, IGFBP3 synthesis outside of the liver is regulated by other endocrine and paracrine factors. At a cellular level, IGFBP-3 has stimulatory or inhibitory effects on IGF-I depending on cell type and the physiologic milieu. IGFBP-3 action at the cell is characterized by its interaction with IGF-I or -II. In vitro, coincubation of IGFBP-3 with IGF-I can block IGF access to the type I receptor [20,42]. Conversely, preincubation of IGFBP-3 in certain cell systems facilitates receptor binding of the ligand by attaching to the cell membrane at a site remote from the receptor. In addition, very recent data suggest that IGFBP-3 may have IGF independent actions on cell action [43]. Although a putative IGFBP-3 receptor has not been cloned, IGFBP-3 has been shown to down regulate cell proliferation in certain cell lines and to enhance p53 production [44]. Further regulation of IGF-I by IGFBP-3 can occur in the extracellular space if IGFBP-3 undergoes proteolysis. Enzymatic degradation of IGFBP-3 produces low-molecularweight IGFBP-3 fragments which differ in their affinity for the IGFs [45,46]. There are numerous IGFBP-3 proteases which are produced by various cell types that can be found in the intra- and extravascular space and are regulated by both endocrine and paracrine factors. Prostate-specific antigen (PSA) is a serine protease which cleaves IGFBP-3 and may be important in defining skeletal metastases with prostate cancer [47]. IGFBP-1, -2, -4, and -5 are also important systemic and local regulators of IGF bioactivity. In contrast to IGFBP-3, these IGFBPs are not fully saturated and easily translocate from the circulation into the extracellular space. IGFBP-1 is a 30-kDa peptide produced primarily in the liver. Serum IGFBP-1 concentrations correlate inversely with circulating insulin and, in poorly controlled insulin-dependent diabetes mellitus, serum IGFBP-1 values are quite high [48]. Hepatic IGFBP-1 production is tightly regulated by insulin and substrate availability. However, unsaturated IGFBP-1 could also serve as a reservoir of binding activity for unbound IGF or could act as the initial binding site for cell secreted IGF, prior to transfer to the more stable, growth-hormone-dependent 150-kDa complex. Shifts in the levels of IGFBP-1 may alter the distribution of the IGFs among the other IGFBPs and thus affect the relative distribution of the IGFs between the intra- and extravascular space. This mechanism could be critical in controlling metabolic and mitogenic activities of the IGFs [49]. In relation to the skeleton there is some in vitro suggestion that IGFBP-1 is synthesized by osteoblasts and could inhibit IGF actions in bone during states of high IGFBP-1 production, such as starvation, and type I diabetes mellitus.

752 Human IGFBP-2 is a 31-kDa protein which preferentially binds IGF-II. It is the major IGFBP in cerebrospinal fluid and likely is produced by neural cells. Insulin and dexamethasone have been shown to decrease production of IGFBP-2 in rat osteoblasts [50]. Recombinant human IGFBP-2 inhibits IGF-I stimulated bone cell proliferation, bone collagen synthesis, and bone formation [51]. However, skeletal concentrations of IGFBP-2 are not nearly as high as IGFBP-3, -4, or -5 [51]. IGFBP-4 is a glycosylated 24-kDa binding protein. It is one IGFBP which is consistently inhibitory for the IGFs-in numerous cell systems. It was originally isolated from skeletal tissue and was found to inhibit IGF mediated bone cell proliferation [52,53]. The expression of IGFBP-4 in bone cells is regulated by cyclic AMP, PTH, and 1,25-dihydroxyvitamin D [53]. In addition, IGF-I stimulates IGFBP-4 proteolysis, thereby providing an autocrine paracrine loop between the ligand and its binding protein [54]. Preliminary evidence suggests that circulating levels of IGFBP-4 may reflect local bone cell regulation. Rosen et al. have shown high serum levels of a 24-kDa IGFBP (likely to be IGFBP4) in elderly women with hip and spine fractures [55]. The relative ligand binding of IGFBP-4 in serum from osteoporotic women closely correlated with circulating concentration of PTH, suggesting that serum changes mirrored local skeletal activity [55]. More recent data from Mohan et al., utilizing a specific RIA for IGFBP-4, have shown an age related increase in this binding protein and a relatively strong correlation between PTH and IGFBP-4 [56]. IGFBP-5 is a nonglycosylated 31-kDa IGFBP produced by osteoblasts and numerous other cell types. It is found in relatively high concentrations, both in bone and in serum where it can be measured by RIA [22]. IGFBP-5 has the unique capacity to bind extracellular matrices, particularly hydroxyapatite. In vitro, IGFBP-5 enhances IGF bioactivity, especially in bone. Its synthesis is increased by PTH and other cyclic AMP analogs [57]. Intact IGFBP-5 can be found circulating in the extracellular space, attached to extracellular matrices, or cleaved into lower molecular weight protein fragments. Intact IGFBP-5’s major role in the skeleton may be as a storage component for the IGFs since both IGF-I and IGF-II have very low binding affinity for hydroxyapatite, but bind avidly to IGFBP-5 [20]. During remodeling, resorption enhances proteolytic cleavage of IGFBP-5. In addition, during formation and mineralization, synthesis and release of IGFBP-5 by bone cells facilitates attachment of IGFs to the newly mineralized matrix [58]. 4. PROTEASES THAT CLEAVE IGFBP’S IGF bioactivity is regulated transcriptionally by hormones and paracrine factors. Tissue activity of the IGFs is also controlled by their respective binding proteins. Tissuespecific proteases provide another form of regulation of IGFs, this time on a posttranslational level. Binding-pro-

WÜSTER AND ROSEN

tein-specific proteases have been identified in serum and in various tissues including bone. These proteases alter the binding capacity of IGFs for the IGFBPs, thereby freeing the IGFs to bind to their respective IGF receptor [20]. Several of these proteases have been tentatively identified including PSA (prostate specific antigen, a serine protease; see below). Bone is a rich source of binding protein proteases which are active against IGFBP-3, -4, and -5. The exact nature of these enzymes which target skeletal IGFBPs has still to be characterized and their regulation remains an area of intense research. However, it is clear that the IGFs can regulate tissue-specific proteases, thereby establishing a complex regulatory loop in which the ligand (IGF) controls its own bioavailability through transcriptional and nontranscriptional means [49]. One protease which has clinical relevance and is under hormonal control is prostate-specific antigen (PSA). This serine protease enhances cleavage of IGFBP-3 into several lower molecular weight fragments and is regulated at least to some extent by testosterone and other androgens [45,58]. Its role in mediating bone formation at the site of metastatic prostate cancer remains to be determined but could provide one mechanism whereby IGFs stimulate mitogenic activity free of one IGFBP.

III. THE ROLE OF GH – IGF-I IN SKELETAL PHYSIOLOGY A. GH – IGF-I Effects on Longitudinal Growth Growth hormone has distinct effects on the skeleton in terms of both linear growth and bone remodeling. However, it has been extremely difficult to ascertain a role for IGF-I independent of growth hormone. The interaction of GH and IGF-I in bone during growth is complex and has been labeled a “dual-effector” process [59]. However, irrespective of the precise mechanisms, GH and IGF-I serve critical functions in skeletal growth and remodeling. Longitudinal growth results from the activity of growth hormone on the skeleton, particularly at the cartilaginous growth plate. In human bone, proliferating chondrocytes express type I IGF receptors and are responsive to paracrine IGFs secreted by differentiated cartilage cells [60]. The target for GH in the growth plate is the differentiated chondrocyte which synthesizes IGF-I in response to GH. Proliferating chondrocytes respond to locally produced IGF-I by differentiation which in turn leads to cartilage expansion and linear growth. Thus, growth hormone’s stimulatory properties on the endochondral growth plate are mediated by induction of IGF-I. GH may have its own effect on linear growth, independent of IGF-I. For example, GH stimulates longitudinal

753

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

bone growth in normal rats but rhIGF-I does not [62]. Similarly, transgenic mice which overexpress GH grow to twice their normal size, even though administration of IGF-I to normal mice does not provoke a similar growth response. These effects may be a result of distinct GH receptors on osteoblasts. The presence of both GH and IGF receptors on bone cells complicates interpretation of GH’s action. In vitro, GH stimulates osteoblastic proliferation, differentiation, and matrix mineralization [63,64]. GH also induces the synthesis and release of IGFBP-3 and IGFBP4 in rodent calvarial cells [65]. The response to GH in human bone cells (hOB) is dependent on cell culture conditions. For subconfluent cultures, GH stimulates cell proliferation, while in confluent hOB cultures, GH induces cell differentiation (as measured by cell alkaline phosphatase and procollagen type I propeptide synthesis) [66].

B. Growth Hormone and IGF-I Effects on Bone Remodeling The effects of GH on bone remodeling are extremely complex. However, it is certain that the skeletal IGFs are critical to this process (see also Chapter 14). Remodeling is the sum of several distinct events beginning with resorption and ending with formation. The osteoblast is the principle site of hormonal regulation over remodeling. Therefore, it is not surprising that the final common pathway for GH activity is through the osteoblast. However, the precise manner in which GH activates the remodeling sequence is uncertain. Induction of IGF-I synthesis in osteoblasts is one potential mechanism for activation of remodeling. Evidence that IGFs are critical to orderly remodeling (and not just osteoblast activity) is derived from several clinical and basic studies. First, large quantities of growth factors are stored in bone and released during active resorption, suggesting that the IGFs could couple formation to resorption [67]. Second, IGFs can stimulate the differentiation and activation of osteoclasts, possibly in concert with certain cytokines [68,69]. Third, administration of IGF-I enhances bone formation and bone resorption to relatively the same degree [37]. Fourth, bone marrow stromal cells which produce osteoclast-activating cytokines, are also rich sources of insulin-like growth factors and IGFBPs [70]. Fifth, at least one protease which cleaves IGFBP-4 is physiologically active only at very low pH [71]. The acidic pH necessary for protease activation is approximately the same pH present within the microenvironment of the osteoclast during its active proton secretion phase. Sixth, preliminary evidence suggests that IGF-I by suppressed OPG (osteoprotegerin), thereby enhancing osteoclastogenesis (J. Rubin, pers. commun.).

GH, IGF-I, and IGF-II all have potent stimulatory effects on bone cell growth [38]. However, GH-induced cell proliferation can be blocked by simultaneous addition of a specific monoclonal antibody to IGF-I [67]. IGF-II, on the other hand, stimulates mitogenesis even if high doses of IGF-I are coadministered. This suggests that IGF-II could regulate osteoblastic proliferation via the IGF type II receptor [72]. Both IGF-I and IGF-II are mitogenic to bone cells and both rapidly increase mRNA expression of the protooncogene, c-fos, 20- to 40-fold in less than 30 min [76]. The IGFs also stimulate type I collagen synthesis, alkaline phosphatase activity, and osteocalcin in human osteoblast-like cells [73 – 75].

IV. PATHOPHYSIOLOGY OF OSTEOPOROSIS: ROLE OF GH – IGFs A. Effects of GH Deficiency on Bone Metabolism Growth hormone deficiency (GHD) in childhood is associated with growth failure and short stature. However, the effects of GHD on bone mineral density (BMD) in prepubertal children have been more difficult to quantitate. This has resulted in a paucity of studies examining bone mineral status in GH-deficient children. By single-photon absorptiometry (SPA) of the wrist, children with GHD have been found to have low bone mass [77]. Serum concentrations of osteocalcin are also reduced in children with GHD, but the response of osteocalcin to GH administration does not correlate with linear growth [78]. In several cross-sectional studies of adults with GHD, lumbar spine BMD is reduced compared to that in age matched controls [79 – 84]. In one group of adult GHD patients, the lowest spinal BMD was found in people who were previously treated with rhGH during childhood [79,85]. This degree of osteopenia was not due to cortisone or thyroxine substitution since the BMDs of patients on hormonal substitution did not differ from those without hormone replacement [85] (Fig. 2). In that same study, Wüster et al. showed an increased prevalence of vertebral osteoporotic fractures among GH-deficient adults [79]. Kaufman et al. confirmed low BMD in GH-deficient adults with or without hormonal deficiencies [86]. However, Kann et al. found no difference in the apparent phalangeal ultrasound transmission velocity of GHD patients compared to age- and sex-matched controls [87]. De Boer et al. noted that low BMD was partly explained by reduced body height, but with correction for body mass index, BMD was still significantly reduced compared to ageand sex-matched controls [77,84]. The cause of low bone mass in adult GHD has been thought to be due to insufficient bone acquisition during the adolescent years [84]. This hypothesis is supported in

754

WÜSTER AND ROSEN

FIGURE 2

Influence of hormonal substitution on bone mineral density (BMD) in patients with pituitary insufficiency involving multiple axes. BMD was measured at the lumbar spine by dual X-ray absorptiometry (Hologic QDR 1000, Waltham, MA). Values are given as means  SD. Z scores show the deviations from the mean of an age- and sex-matched reference population. Left bars represent patients taking no hormonal substitution for the thyroid (left pair of bars) or adrenal axis (right pair of bars). Right bars represent patients fully substituted with L-thyroxine or hydrocortisone. Adapted with permission from Ref. [85].

one study by bone histomorphometry. In 36 men with GHD (primarily of juvenile onset) there were increased eroded surfaces, increased osteoid thickness, and increased mineralization lag time, all indicative of delayed mineralization probably due to changes in the timing of puberty [84]. In support of those histomorphometric changes, low serum concentrations of osteocalcin have been detected in some adult GHD patients [78,89]. This is in sharp contrast to patients with normal GH secretion but multiple pituitary hormone deficiencies, where osteocalcin values are normal but there is markedly increased urinary pyridinoline excretion [90,91]. Although inadequate acquisition of bone mass during childhood may be one explanation for the osteopenia of GHD, the role of gonadal steroids in this process has not been completely clarified. Furthermore, there are no data on hip fractures in GHD adults, and evidence that spinal fractures are more prevalent in GHD is still preliminary. Only longitudinal studies of GHD patients will be able to determine the precise cause of osteopenia in the acquired GHD syndrome.

B. Effects of GH Excess on Bone Mass and Bone Turnover Chronic GH excess in adults (i.e., acromegaly) has been a surrogate model for studying the effects of GH on the

skeleton. However, this disease is complicated by changes in vitamin D metabolism and gonadotropin secretion [92]. Increased bone turnover has been reported in acromegaly by biochemical markers and histomorphometric studies [90,91,93 – 95]. However, bone mass determinations in acromegaly vary according to the site of measurement. Cortical BMD is increased compared to age-matched controls and is directly related to the degree of GH excess [95 – 98]. Trabecular BMD, however, can be high, normal, or low [94,98,99]. In one study, CT measurements of the lumbar spine revealed that trabecular BMD was elevated in only 1 of 14 patients with active acromegaly [94]. This may have been due to hypogonadism in the acromegalics. Wüster et al. recently studied five patients with active acromegaly treated with octreotide for 5 years. All had achieved normal IGF-I levels during therapy. Spinal BMD was initially decreased in all five patients but normalized in three of them with octreotide. All patients remained eugonadal throughout follow-up [85]. As noted above, biochemical markers of bone turnover are altered in acromegaly. Many of these changes can be related to alterations in gonadal status during the disease and its treatment. However, changes in bone turnover with acromegaly reflect persistent coupling of the remodeling cycle with increased resorption and formation. Serum osteocalcin and skeletal alkaline phosphatase concentrations are increased in acromegaly, as are urinary calcium and hydroxyproline excretion [83 – 85]. Although serum calcium,

755

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

FIGURE 3

Serum IGFBP-3, IGF-I, and IGF-II concentrations in patients with osteoporosis, osteoarthritis, and age-matched controls. Adapted from [99] with permission of the author.

total alkaline phosphatase, and phosphorus values are usually normal, there may also be increased synthesis of 1,25dihydroxyvitamin D. This results from significant intracellular phosphate shifts due in part to increased circulating IGF-I. The consequence of this change, however, is not entirely clear.

C. Changes in the GH – I G F - I Axis in Patients with Osteoporosis For several years, attempts have been made to link GH secretory status with low bone mass and osteoporosis. As noted previously, efforts to find a relationship between GH secretion and age-related bone loss have been conflicting at best. However, other investigators have examined the relationship of GH to bone mass in the immediate menopausal period. These efforts gained prominence in 1982 when it was reported that GH secretion in patients with osteoporosis was reduced even after stimulation with L-arginine [100]. More recently, low serum IGF-I, IGF-II, and IGFBP-3 levels (by RIA) were noted in 98 women with postmenopausal osteoporosis compared to normals and patients with osteoarthritis or degenerative bone disease [101] (see Fig. 3). In a cross-sectional study of a large cohort of older postmenopausal women from Framingham, Langlois et al. reported very strong correlations between the lowest quintile of IGF-I and BMD at the spine, hip, and radius [24]. Very recently Bauer et al. reported that in the Study of Osteoporotic Fractures (SOF), women in the lowest quartile for serum IGF-I had a 60% greater likelihood of hip or spine fractures, even when controlling for bone mineral density [102]. Similar results were noted by Ganero et al. [103].

Similar results have been noted in male osteoporotics where serum IGF-I as well as IGFBP-3 concentrations correlated with lumbar BMD [104]. Comparable results have been noted for IGF-I by others [105,107]. Johannsen et al. reported that among healthy males, IGFBP-3 was the best predictor of femoral bone mineral density [106]. Kurland et al. reported that younger males with idiopathic osteoporosis had low serum levels of IGF-I in relation to age-matched controls [107]. Moreover, these men also had low rates of bone turnover by histomorphometry but normal GH dynamics [108]. Of potential pathophysiologic importance is the observation that patients with osteoarthritis have higher concentrations of IGF-II than normal controls [101,109]. Further longitudinal studies will be required to determine the precise relationship between the IGFs, GH, and osteoporosis.

V. GROWTH HORMONE THERAPY FOR OSTEOPOROSIS A. Mechanisms Growth hormone has direct and indirect effects on bone, depending on age and skeletal maturity (see Fig. 4). Indirectly, GH can enhance bone mass through its effects on muscle mass and calcium transport in the gut [110]. In addition, growth hormone can directly stimulate bone remodeling and increase endochondral growth through its actions on the osteoblast. Overall, GH is considered essential for both the growth and maintenance of skeletal mass. GHD, whether it be in childhood or

756

FIGURE 4

Summary of the effects of GH on various systems which influence directly or indirectly bone and calcium metabolism. Adapted with permission from Ref. [156].

adulthood, is associated with reduced bone mass. In one of the largest observational trials to date, KIMS, GHD was linked to a marked increase in fracture risk,compared to age-matched normals [111]. Substantial differences between the direct and indirect effects of GH on the osteoblast partially explain changes in skeletal responsiveness to GH and IGF-I. For example, exogenous GH stimulates longitudinal growth in normal rats, but rhIGF-I does not [62]. Similarly, GH transgenic mice grow to twice their normal size while exogenous administration of IGF-I is far less efficient in stimulating long bone growth [112]. Thus, despite the fact that GH induces IGF-I production in the skeleton and elsewhere, treatments with GH and with IGF-I are not equivalent. In general, skeletal responses to GH and IGF-I depend on the species, the GH status of the animal, and the mode of administration. Even the systemic side effects of rhGH and rhIGF-I therapy may differ substantially. 1. GH TREATMENT FOR GHD CHILDREN Early clinical experiences with rhGH in GH-deficient children provided investigators with a model for studying skeletal responsiveness to somatotropin. Intermittent (daily or three times weekly) injections of rhGH results in a prolonged and sustained GH profile with resultant catch-up growth evident during the first year of treatment [113]. This increase in skeletal growth is accompanied by a rise in serum levels of type I procollagen peptide [114]. Although dosage schemes vary between the United States and Europe (0.1 mg/kg/tiw [US] to 0.7 IU/kg ( 0.23 mg/kg)/week [Europe]), there is a strong dose-related growth response to rhGH [115].

WÜSTER AND ROSEN

The skeletal response to GH depends on several factors including: (i) GH secretory status, (ii) pretreatment IGF-I levels, (iii) pretreatment height velocity, and (iv) GH dosage [115]. The rate of change in serum IGF-I (rather than the absolute level of IGF-I attained by GH treatment) is a relatively nonspecific predictor of growth as are procollagen I and osteocalcin concentrations [78,115]. Serum procollagen III levels do correlate with growth rates during GH treatment [78,114]. Linear growth is a measurable response to exogenous GH, but changes in bone mineral density in children are more difficult to quantify. In some studies, bone mineral content is increased during GH treatment to a greater extent than expected for change in bone size [116]. In one of the longest intervention trials to date, 26 GH-deficient children were given rhGH (0.6 IU/kg per week) for 12 months [117]. Baseline radial bone mineral content (BMC) Z scores (corrected for their chronological, statural, and bone ages) were significantly reduced, as were serum osteocalcin and procollagen peptide levels. Treatment with rhGH six times per week increased BMC and normalized Z scores of the radius in nearly 50% of the subjects. Serum levels of procollagen peptide during the first week of treatment were positively related to growth velocity at 6 and 12 months and radial BMC at 12 months. Although there were striking differences between longitudinal growth in children and remodeling in adults, criteria which determine rhGH responsiveness in children may be relevant for older individuals. It has already been established that biochemical and histomorphometric responses to rhGH in children may differ according to their GH secretory status. The same principle probably holds for adults treated with growth hormone. 2. GH TREATMENT FOR ADULTS Three adult populations have been studied before and after GH in order to examine predictors of skeletal responsiveness: (i) healthy adults, (ii) GH-deficient adults, (iii) elderly men and women with/without osteoporosis. a. GH Administration to Healthy Adults Initial studies with rhGH in adults focused primarily on changes in body composition. Short-term treatment with rhGH leads to a decrease in adiposity and an increase in lean body mass [118]. There is also a marked shift in extracellular water [119]. Detailed analysis of skeletal markers during GH treatment was first reported by Brixen et al. [120]. Twenty male volunteers (ages 22 – 31) were given a relatively large dose (0.1 IU/kg) of rhGH twice daily for 7 days. Serum osteocalcin increased after 2 days of treatment and remained elevated for 6 months. Bone alkaline phosphatase decreased initially (during the 7 days of GH

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

treatment) but then increased slightly over 6 months [120]. Serum calcium and phosphate increased but only during the 7-day treatment phase. Like bone formation indices, urinary markers of bone resorption (urinary calcium/creatinine and hydroxyproline/creatinine) rose during treatment and remained elevated for up to 4 weeks after discontinuation of therapy. Treatment with rhGH stimulates bone remodeling. More importantly, the anabolic effect on bone may persist well beyond discontinuation of growth hormone. Early (2-day) and late (2-week) osteocalcin responses imply that GH can stimulate existing osteoblasts and enhance recruitment of new osteoblasts. Still, it is uncertain if those effects are mediated through IGF-I. For example, Brixen et al. were unable to find a significant correlation between the rise in serum IGF-I and an increase in osteocalcin or bone alkaline phosphatase [120]. The absence of a significant correlation between bone formation markers and serum IGF-I, however, may be due to the low skeletal specificity of serum IGF-I measurements. Skeletal resistance to GH has been considered a possible cause for postmenopausal osteoporosis. On the basis of one double-blinded rhGH trial in postmenopausal women, this is unlikely. Kassem et al. noted that administration of rhGH (0.2 IU/kg/day) for 3 days increased serum IGF-I, osteocalcin, and procollagen type I C-terminal propeptide (PICP) to the same extent in 15 women with severe postmenopausal osteoporosis as in 15 age-matched control women [121]. Serum and urinary markers of bone resorption also did not differ between the two groups. In vitro studies of marrow stromal cells from osteoporotic women demonstrate full GH responsiveness [121]. Therefore, it is unlikely that the osteoporotic skeleton is resistant to rhGH treatment. b. GH Treatment for Growth Hormone-Deficient Adults i. GENERALIZED GH ACTIONS DURING TREATMENT IN ADULT GH DEFICIENCY. Growth hormone deficiency can be documented by provocative stimuli (GHRH, insulin, glucagon) and serial GH measurements. The majority of adult patients treated with rhGH have either idiopathic GHD or a history of previous central nervous system (CNS)/pituitary – hypothalamic tumors. Early trials with rhGH replacement therapy examined changes in muscle mass, muscle strength, and body fat. Daily administration of subcutaneous rhGH to GHD patients produced a marked rise in serum IGF-I and an increase in muscle mass and basal metabolic rate [122]. Some of those anabolic changes were noted soon after the initiation of rhGH. For example, mean nitrogen retention during the first 15 days of rhGH treatment was as much as 2.8 g per day (approximately 20 g of muscle mass) [123]. GH treatment can also increase the total cross-sectional area of thigh muscles and quadriceps as well as improve hip flexors and limb girdle strength [122,124]. At least one group has suggested that rhGH can

757 increase the number of type II muscle fibers. Total fat mass consistently decreases during rhGH treatment [122,124, 125]. Based on these and other studies, the U.S. FDA approved the use of rhGH in patients with established growth hormone deficiency. ii. GH EFFECTS ON BIOCHEMICAL MARKERS OF BONE TURNOVER IN GHD. Several biochemical tests reflect the physiologic action of GH on the skeleton. Serum calcium, osteocalcin, and urinary hydroxyproline all increase, while PTH declines slightly during rhGH treatment [126]. Newer and more sensitive markers of bone turnover also reflect changes during rhGH treatment. Urinary deoxypyridinoline increases threefold and the amino-terminal propeptide of type III procollagen doubles during 4 months of daily rhGH [127,128]. After cessation of rhGH treatment, deoxypyridinoline excretion decreases but type III procollagen levels remain higher than controls for several months [129]. iii. EFFECTS OF GH ON BONE DENSITY IN GHD. If prolonged growth hormone deficiency in adults results in profound changes in the musculoskeletal system, then GH replacement would be expected to enhance muscle performance and bone mass. Fourteen GH-deficient adults given a nightly dose of rhGH (0.5 IU/kg/week) did show increases in exercise capacity, maximum oxygen consumption, and alkaline phosphatase even though quadriceps strength and spinal bone density did not change [130]. When 0.25 IU/kg/week of rhGH was administered to 12 GH deficient adults for 1 year, there was a marked increase in trabecular bone density (measured by single and dual energy quantitative computed tomography(QCT) of the spine) at 6 and 12 months. At 12 months proximal and distal forearm BMC increased, mid-thigh muscle area was greater, and fat cross-sectional area decreased. Since the rise in spine BMD was noted with both single and dual-energy CT measurements of the spine, this increase probably resulted from enhancement in bone mass not a reduction in marrow fat. More recently several groups have performed longer studies with rhGH for replacement in the GHD syndrome. Although changes in BMD are not significant at 12 months, by 24 and 36 months, BMD can increase by as much as 5 – 8% in the spine [131 – 134). Moreover, some investigators have also reported a concomitant increase in muscle strength after 2 years of treatment. It appears that those individuals with earlier onset of GHD, as well as those with the lowest BMD, had the greatest likelihood of showing significant changes in bone mass with rhGH. c. GH Administration to Elders i. GENERALIZED GH EFFECTS. As previously noted, elderly people have lower GH secretory amplitudes and reduced serum levels of IGF-I and IGFBP-3 compared to younger adults [1,2,17,135]. Moreover, the pulse frequency for GH is less in older people. Based on these data, it was assumed that skeletal responsiveness to GH in elders would be identical to

758 that seen in GHD patients. In elderly men one group has reported a blunted serum IGF-I response to 0.1 mg/kg GH (36% lower) compared to that in younger men or adults with GHD [136]. However, Rosen et al. noted recently that generation of IGF-I after various doses of rhGH to frail elders was not associated with growth hormone resistance [137]. Based on some recent data it appears that GH replacement for adult GHD or for pharmacologic treatment results in similar IGF-I responses. ii. EFFECTS OF GH ON BONE DENSITY AND BONE REMODELING IN ELDERS. The most widely publicized growth hormone trial in elders involved 21 men over age 65 randomized to receive 0.03 mg/kg of rhGH three times per week (as a subcutaneous injection) or to no treatment whatsoever. Twelve men received rhGH while 9 men served as observational controls. The men were selected on the basis of a low serum IGF-I (100 mg/ml) concentration [5]. rhGH produced a threefold rise in circulating IGF-I, an increase in lean body mass (as measured by potassium40 analysis) and a decline in total adipose mass. Bone density of the lumbar vertebrae (L1 –L4) as measured by dual photon absorptiometry, increased 1.6% after 6 months in the treatment group while no change was noted in controls. Biochemical markers of bone turnover were not examined and no changes in bone density were detected in the mid- or distal radius or three areas of the hip. Furthermore, the spinal BMD changes at 6 months were not sustained at 1 year [138]. Marcus et al. studied the effects of rhGH in 16 men and women over age 60 [139]. Daily doses of rhGH (0.03, 0.06, or 0.12 mg/kg BW/day) were randomly assigned to each subject and administered once daily for 7 days. Serum IGF-I, osteocalcin, PTH, and calcitriol concentrations all increased during treatment. In this short-term study, there was also a significant rise in urinary hydroxyproline and urinary calcium excretion with a decline in urinary sodium. Holloway et al. conducted a longer randomized doubleblinded placebo-controlled trial of daily rhGH for 1 year in 27 healthy elderly women, 8 of whom took a stable dose of estrogen throughout the study [119]. Thirteen women completed 6 months of treatment and 14 women completed 6 months in the placebo group. Side effects prompted a 50% reduction in the original dose of rhGH (from 0.043 mg/kg body wt or approximately 0.3 mg rhGH/kg/week to 0.02 mg/kg/day) and led to several dropouts in the treatment group. Fat mass and percentage body fat declined in the treatment group but there were no changes in bone density at the spine or hip at 6 or 12 months in other groups [119]. Although bone mass did not change, there were changes in some biochemical parameters. In particular, urinary markers of bone resorption (hydroxyproline and pyridinoline) increased after 6 months of rhGH treatment. The response of bone formation markers was more variable. Osteocalcin increased but type I procollagen peptide levels did not change. For women taking estrogen replace-

WÜSTER AND ROSEN

ment therapy, indices of bone turnover (both formation and resorption) were blunted (see Fig. 5). More recently, Rosen et al., reported a dose-dependent decrease in bone mass after 1 year of rhGH in frail elderly men and women [137]. This occurred despite striking increases in serum osteocalcin and IGF-I concentrations with the highest doses of rhGH (0.01 mg/kg/day). In part, the absence of a positive GH effect on BMD is not surprising since resorption is coupled to formation and GH activates the entire remodeling sequence. Indeed, in the same trial of 132 frail elderly subjects by Rosen and colleagues, urinary N-telopeptide and osteocalcin values both rose to the same extent, suggesting that total bone turnover, not just bone formation, was increased by rhGH therapy [137]. Also, 1 year is too short a time to conclude definitively that rhGH would not increase bone mass. This is best exemplified by the rhGH trials in GHD. However,the relatively high incidence of side effects (weight gain, carpal tunnel syndrome, edema, glucose intolerance) in GH trials, especially in the frail elderly, is particularly troublesome. Ongoing rhGH studies in elders will examine factors such as the age of the person, presence of preexisting conditions, and IGF-I levels after GH to determine if these side effects can be predicted. 3. GH TREATMENT FOR OSTEOPOROTIC PATIENTS Short nonrandomized clinical trials with GH in osteoporosis were attempted well before GH replacement therapy was considered. As early as 1975, two patients with osteogenesis imperfecta and one patient with involutional osteoporosis were treated with GH [140]. Histomorphometric parameters of increased bone formation and resorption were noted. Subsequent studies employed GH with and without anti-resorptive agents. Aloia et al. administered between 2 and 6 IU/day of GH for 12 months to eight patients with postmenopausal osteoporosis (the first 6 months of treatment freatured low-dose GH; the last 6 months consisted of high dose GH (6 IU/day). Radial bone mineral content dropped slightly and histomorphometric parameters did not change during treatment. However, severity of back pain decreased considerably in several people [141]. Daily GH injections (4 IU/day) combined with alternating doses of calcitonin produced an increase in total body calcium (measured by neutron activation analysis) but a decline in radial bone mass after 16 months [142]. In a separate trial, 14 postmenopausal women were given 2 months of GH and then 3 months of calcitonin in a modified form of coherence therapy [143]. Total body calcium increased 2.3%/year and there were few side effects, but there were no changes in bone mineral density or histomorphometric indices. Dambacher et al. administered 16 IU of rhGH every other day along with daily sodium fluoride to six women with postmenopausal osteoporosis [144]. On histomorphometric analysis, there was a significant increase in the

759

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

FIGURE 5

Effects of rhGH on markers of bone turnover. Urinary bone resorption markers are shown in the graphs on the left and indicate results from baseline (black) and 6 months (lightly hatched). Serum bone formation markers are shown in the graphs on the right and indicate results at baseline (black), 3 months (lightly hatched), and 6 months (densely hatched). PICP, type I procollagen extension peptide. Asterisks indicate significant changes from baseline; *P  0.05, ***P  0.001. Bars showing similar letters differ significantly, P  0.05. Adapted with permission from Holloway et al. [106].

number of osteoblasts and osteoclasts but bone mass was unchanged. Johannsen et al. conducted a placebo-controlled doubleblinded crossover trial of rhGH and IGF-I in 14 men with idiopathic osteoporosis [145]. In this 7-day trial with rhGH (2 IU/m2), procollagen peptide and osteocalcin levels increased after treatment as did urinary markers of bone resorption. The changes in osteocalcin were relatively small, however, and were not sustained after discontinuation of growth hormone treatment. There are no GH trials (past or present) which have examined spinal fractures as a therapeutic end point. Therefore it is difficult to judge the potential efficacy of GH in the treatment of osteoporosis. However, GH stimulates bone remodeling activity, thereby leaving open the possibility that GH can be coupled to anti-resorptive agents to improve its effect on bone. This thesis was tested in a 2-year randomized trial by Holloway and colleagues [146]. In that study, cyclic administration of rhGH and nasal calcitonin increased spine BMD by approximately 2%. This result, however, was not much different than that seen with CT alone, and certainly less than what has been seen in very large randomized trials with anti-resorptive agents. Once again, the prevalence of side effects resulted in limited enthusiasm for rhGH as a primary treatment for osteoporosis

[146]. Therefore, it is likely that GH may induce small but significant changes in bone mass which over an extended period could translate into fewer spine fractures. However, there are still no data to support that contention. In the meantime, several very small trials have looked at the effects of GH releasing analogs on bone turnover and bone mass. Not unlike rhGH, however, these studies have been small, and the results somewhat conflicting. However, in contrast to rhGH, GH releasing analogs are NOT associated with significant side effects. Hence, further trials are likely to begin with these analogs.

VI. IGF-I FOR THE TREATMENT OF OSTEOPOROSIS A. Introduction In the late 1980s clinical trials with recombinant human IGF-I for diabetes mellitus were begun. The availability of this recombinant peptide and the absence of other treatments to stimulate bone formation accelerated animal and human studies of rhIGF-I in metabolic bone diseases. Theoretically, there are potential benefits for rhIGF-I compared to rhGH. These include: (i) more direct stimulation

760

WÜSTER AND ROSEN

of bone formation, (ii) bypass of skeletal GH theoretical resistance, (iii) reduction in GH-induced side effects such as carpal tunnel and diabetes mellitus. There are, however, considerably fewer animal and human studies using rhIGF-I than rhGH. Therefore, these advantages have either yet to be fully realized or have not been validated.

B. Animal Studies with rhIGF-I 1. rhIGF-I AND BONE GROWTH IGF-I is not a potent mitogen in most tissues and bone is no exception (see Chapter 14). There are high affinity receptors for IGF-I on osteoblasts and IGF-I can stimulate preosteoblast replication and provoke resting cells to proceed through their growth cycles. IGF-I maintains the differentiated osteoblast phenotype, stimulates collagen synthesis, and prevents collagen degradation. Theoretically, therefore, despite its relatively weak mitogenic properties, IGF-I could have significant anabolic activity on the skeleton. In hypophysectomized rats, growth can be fully restored by administration of either GH or IGF-I but not IGF-II [147,148]. A similar growth response occurs after rhIGF-I in streptozotocin-diabetics rats but not in sex-linked dwarfmutant chickens [149,150]. In normal rats, rhIGF-I administrated either systemically or locally (hindlimb infusions), does not stimulate longitudinal bone growth [151]. In the spontaneously diabetic BB rat, rhIGF-I treatment does not result in changes in epiphyseal width, osteoblast surfaces, or osteocalcin concentration [152]. The skeletal response to rhIGF-I is determined by the growth hormone/IGF-I status of the animal. For example, IGF-I does not increase bone formation in normal rats, whereas it stimulates bone growth and normalizes type I procollagen mRNA levels in hypophysectomized rats [153 – 155]. Similarly, in the spontaneous mouse mutant (lit/lit), absence of GH receptors results in very low levels of IGF-I and skeletal dwarfism. IGF-I treatment restores growth and increases total body water but does not enhance bone mass in these mice [156]. These findings are somewhat similar to the effects of GH on the skeleton in GH-deficient animals. However, rhIGF-I and rhGH differ in their actions on the circulatory IGF regulatory system. Growth hormone stimulates hepatic production of both IGF-I and IGFBP-3 while rhIGF-I administration increases the total circulating pool of IGF-I but suppresses hepatic production of IGFBP-3, primarily through feedback inhibition of GH secretion. It is conceivable that variations in IGF-I biological activity (between direct IGF-I administration and endogenously produced IGF-I as a result of GH treatment) may be due to the relative proportion of IGF-I bound to IGFBP-3.

2. rhIGF-I EFFECTS ON BONE MASS IN ANIMALS WITH ALTERED BONE TURNOVER Several experimental paradigms have been employed to study the effects of IGF-I on bone turnover in animals. These include: (i) oophorectomy, (ii) diabetes mellitus (spontaneous or induced), (iii) immobilization. In each situation, bone remodeling is markedly altered prior to IGF-I treatment in order to study growth factor actions on bone resorption and formation. These experimental models provide useful clinical information since IGF-I has been considered a potential therapeutic agent in conditions similar to those produced experimentally. In oophorectomized rats, administration of rhIGF-I has variable effects on bone remodeling, bone mass, and bone strength. Kalu reported partial restoration of trabecular bone volume after oophorectomy in adult rats treated with rhIGF-I [157]. In older oophorectomized rats, rhIGF-I increasd mid-shaft tibial BMD and enhanced periosteal bone apposition [158]. Six weeks of rhIGF-I (delivery by miniosmotic pump) to older rats caused a dose-dependent increase in bone density in the lumbar spine and proximal femur although bone strength and stiffness did not change. Muller reported that subcutaneous administration of rhIGFI to adult oophorectomized rats stimulated bone formation as evidence by increased osteoid surfaces, osteoblast surfaces, and mineral apposition rates [159]. At high doses of rhIGF-I, osteoclast surface and osteoclast number also increased. In contrast, Tobias et al. found that rhIGF-I (200 mg/kg) administered for 17 days to 15-week-old rats increased longitudinal and periosteal growth but suppressed trabecular bone formation in both oophorectomized and control rats [155]. Bone resorption was also slightly suppressed during rhIGF-I treatment, although not to the extent that bone formation was inhibited. Type I insulin-dependent diabetes mellitus (IDDM) is associated with decreased cortical bone mineral density [160]. Although the pathophysiology of diabetic osteopenia remains unknown, it appears that the duration of diabetes, the extent of diabetic control, and the timing of disease onset are each associated with higher risks of low density [161,162]. Serum markers of bone formation are reduced in type I diabetics, suggesting a possible defect in osteolastic activity [152,163]. Serum IGF-I levels are either normal or low in type I diabetes mellitus, but often are reduced in patients with poor diabetic control. In these same people, serum IGFBP-1 levels are quite high. This has led investigators to believe that changes in the IGF regulatory system during poor metabolic control contribute to impaired growth. Spontaneously diabetic BB rats exhibit osteopenia and therefore provide a useful model for studying the effects of IGF-I on bone remodeling. Even though bone formation is lower in BB than control rats (as measured by serum markers), administration of rhIGF-I does not increase bone

761

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

epiphyseal width, osteoblast surfaces or serum osteocalcin [152]. Thus, despite evidence that circulating levels of IGFI are reduced in some patients with type I IDDM, preliminary animal studies have failed to show that IGF-I administration can correct any inherent defect in bone formation. Chronic immobilization inhibits bone formation and leads to signficant bone loss. The pathophysiology of immobilization caused by bed rest, hindquarter elevation, or spaceflight is unknown, but the bone remodeling unit is uncoupled due to a transient decrease in bone formation and a marked rise in bone resorption [164,165]. Some investigators have proposed that reduced bone formation during immobilization results from resistance to skeletal IGF-I. Immobilization in rats by the hindlimb elevation method causes cessation of bone growth [166]. Paradoxically, mRNA levels for IGF-I and the type I IGF-I receptor are substantially increased in the proximal tibia and distal femur of hindlimb-elevated rats [166]. Infusion of rhIGF-I (200 mg/day) during hindlimb elevation does not reverse the cessation in linear growth induced by immobilization, even though growth and bone formation resume relatively soon after immobilization is stopped [166]. This would suggest that there may be, at least transiently, resistance to IGF-I bioactivity. Other investigators have reported contrasting results during hindlimb elevation. Machwater et al. continuously infused rhIGF-I (1.3 – 2.0 mg/kg/day) for 14 days to 5week-old hindlimb-elevated rats [167]. The decline in bone mineral density of the proximal femur with unloading was blunted by infusions of IGF-I. At the tibial metaphysis of IGF-infused animals, bone formation rate and trabecular number were markedly increased. Marrow stromal cells from unloaded rats exhibit decreased proliferative characteristics, but addition of IGF-I greatly increased alkaline phosphatase positive cell proliferation. rhIGF-I also enhanced serum alkaline phosphatase activity and osteocalcin levels in immobilized rats. Alternate ways of exploiting the anabolic properties of IGF-I in bone have been proposed. IGF-I has been administered by intraarterial infusion or coupled to IGFBP-3. Infusion of rhIGF-I continuously into the arterial supply of the right hindlimb of ambulatory rats for 14 days leads to a 22% increase in cortical and trabecular bone formation in the infused limb [154]. By histomorphometry, the number of osteoblasts (but not osteoclasts) increases. Using an alternative model, Bagi et al. administered rhIGF-I or a complex of IGFI – IGFBP-3 to 16-week-old oophorectomized rats [168]. The IGF-I – IGFBP-3 complex (7.5 mg/kg/day) increased bone formation more than did IGF-I alone, even though both treatments increased longitudinal bone growth. The highest doses of rhIGF-I and rhIGF-I – GFBP-3 enhanced trabecular thickness in the lumbar vertebrae and femoral epiphyses and increased bone

resorption but only in the femoral metaphysis. A similar study contrasting IGF-I with IGF-I – GFBP-3 was performed in 22-week-old oophorectomized rats [169]. Bone mineral density increased in both groups but fewer than 10% of the rats treated with IGF-I – IGFBP-3 complex developed hypoglycemia, compared to nearly 50% with rhIGF-I alone.

C. Human Studies with IGF-I The potential utility of insulin-like growth factors in several disorders has led to trials with IGF-I in humans. To understand how IGF-I affects bone remodeling, three groups of adults have been studied before and after rhIGF-I: (1) normal postmenopausal women, (2) Laron dwarfs, and (3) patients with idiopathic and age-related osteoporosis. 1. IGF-I ADMINISTRATION TO NORMAL POSTMENOPAUSAL WOMEN There is one published study of bone markers which employed rhIGF-I to healthy young postmenopausal women. Doses of rhIGF-I from 30 to 180 mg/kg/day were administered daily by subcutaneous injection for 6 days to older postmenopausal women without fractures and normal bone density [37]. Very significant dose-dependent increases in serum type I procollagen carboxyl-terminal propeptide (PICP), osteocalcin, and urinary deoxypyridinoline were reported. Although the rise in PICP was greater than the increase in collagen breakdown (measured by deoxypyridinoline), it is uncertain whether this meant that formation was stimulated more than resorption. For the two highest doses of rhIGF-I (120 and 180 mg/kg/day), orthostasis, weight gain, edema, tachycardia, and parotid discomfort were noted. At lower doses (30 and 60 mg/kg/day) fewer side effects were reported, but less discrete changes in PICP were noted. As noted below, high- and low-dose rhIGF-I was also administered for 28 days to elderly postmenopausal women and the results are somewhat different in that bone formation was stimulated by lower doses of rhIGF-I [170]. 2. IGF-I ADMINISTRATION TO GH-RESISTANT SHORT STATURE PATIENTS One potential indication for IGF-I might be in the GHresistant short stature syndrome (Laron dwarf). By the end of this century, Sweden and several other European countries had approved rhIGF-I for that purpose. Patients with the Laron dwarf syndrome lack functional growth hormone receptors and thus do not respond to GH; their IGF-I concentrations are very low, growth is slow, and circulating GH concentrations are high due to lack of negative feedback on GH by IGF-I [171]. Underwood treated one such boy (age 9) with 2 weeks of continuous intravenous

762 rhIGF-I [171]. Urinary calcium excretion increased while urinary phosphate and sodium decreased. After a 2-week continuous infusion of rhIGF-I, the patient was treated with twice daily sc rhIGF-I (120 mg/kg) for 2 years. Growth occurred at a rate of 10 cm/year, compared to 5 cm for the 3year period prior to treatment. Subsequently, Underwood and colleagues have treated eight patients in this manner without hypoglycemia while Laron and his group have treated five children [171,172]. More recently,a child with an IGF-I deletion mutation in exon 5 has been reported. This patient had very short stature, mental retardation, and other abnormalities along with very low levels of circulating IGF-I [173]. rhIGF-I treatment led to a marked increase in linear growth and a huge increase in spinal bone mass. However, when corrected for changes in size of the bone, the incremental changes in volumetric bone mass were much less impressive [174]. Hypoglycemia was avoided in these cases by having children eat 3 to 4 h after their IGF-I injection. Of interest, several children had selective growth of adenoidal tissue. Two unique aspects about these IGF-I data challenge previous concepts about the role of GH in skeletal homeostatsis. First, IGF-I can act as a classical endocrine hormone stimulating longitudinal growth independent of GH; second, GH may not be absolutely essential for statural growth; i.e., the stimulatory effect of GH on chondrocytes which permits skeletal responsiveness to IGF-I may not be as critical as once perceived. However, caution is warranted in examining the skeletal effects of rhIGF-I in children as most of the changes in the skeleton reflect linear growth and periosteal enhancement, both of which can contribute to two dimensional changes in BMD as measured by DXA, but represent lesser changes when corrected for size [175].

3. IGF-I ADMINISTRATION IN IDIOPATHIC AND AGE-RELATED OSTEOPOROSIS Idiopathic osteoporosis in men is an ill-defined syndrome of low bone mass and spinal fractures without associated hypogonadism. By histomorphometry, these men often have low bone turnover, suggesting a possible defect in bone formation. Several groups of investigators have suggested that this syndrome is related to low serum IGF-I levels [107,108,176]. Since the therapeutic options in males with osteoporosis are somewhat limited and treatment for low bone turnover states, in general, is frustrating, the therapeutic potential for anabolic agents like IGF-I in this condition should be quite high. In one male with idiopathic osteoporosis and low serum IGF-I, Johansson et al. administered subcutaneous rhIGF-I (160 mg/kg/day) for 7 days [164]. Bone alkaline phosphatase, osteocalcin, and the carboxyterminal peptide of procollagen type I all increased more than 40% over baseline. However, urinary

WÜSTER AND ROSEN

calcium/creatinine and hydroxyproline excretion rose during treatment. In a recent trial rhIGF-I (at doses of 80 mg/kg/day) and rhGH (2 IU/m2/day) in 12 men, serum osteocalcin, serum procollagen peptide, and urinary deoxypyridinoline excretion all increased following 7 days of rhIGF-I treatment [177]. Although there were slight differences in the response of certain biochemical markers to IGF-I and GH, both forms of therapy produced significant increases in bone resorption. Clinical trials provide evidence that IGF-I acts by increasing the birth rate of remodeling osteons, thereby promoting bone resorption and formation. Yet, it is conceivable that low doses of rhIGF-I (30 g/kg/day) may differentially stimulate bone formation. In one trial of 16 healthy elderly women, 60 g/kg/day (high-dose) and 15 g/kg/day (low-dose) of rhIGF-I were tested for 28 days. The high-dose rhIGF-I increased markers of bone resorption and formation. However, low doses of rhIGF-I caused increases in serum osteocalcin and type I procollagen carboxyterminal peptide, but had no effect on total pyridinoline excretion [170]. These data would support the thesis that low doses of rhIGF-I may directly increase osteoblastic function with only a minimal increase in bone resorption. Further studies will be needed to assess the future therapeutic role of low doses rhIGF-I in osteoporosis. Recently, novel approaches to enhancing IGF-I action in bone have been proposed. One strategy is to administer a bone-specific agent which stimulates bone mass such as parathyroid hormone (PTH). Intermittent hPTH increases trabecular bone by stimulating osteoblasts to synthesize IGF-I and other growth factors [178]. Another strategy is to administer IGF-I along with an IGF binding protein. Bagi et al. previously reported that IGF-I/IGFBP-3 complex could enhance bone mass in the metaphysis and epiphysis of rats. One very small randomized trial utilized subcutaneous infusions of IGF-I/IGFBP-3 in 24 older women with hip fractures. Bone loss in the contralateral hip was reduced considerably after 6 months (i.e., from 6 to 1.5%) in those subjects who were given the complex vs those receiving saline [179]. Accompanying that change in BMD, there was also an increase in grip strength in those that received the active agent, while no significant side effects were reported. Larger clinical trials are likely to commence utilizing this or other combinations linking IGF-I to one or more IGFBPs, especially if bone is considered the principle target.

VII. SUMMARY Several lines of evidence suggest that recombinant growth factors may be anabolic for the skeletal remodeling unit. First, both GH and IGF-I stimulate osteoblastic differ-

763

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors

entiation. Second, in animal models, GH and IGF-I enhance longitudinal growth, bone formation, and bone mass. Third, in GHD children and adults with an impaired GH – IGF-I axis, rhGH and rhIGF-I both enhance trabecular and cortical bone mineral density. However, evidence that rhGH or rhIGF-I can differentially stimulate bone formation for a sustained period in older adults is lacking. Therefore, favorable responses in properly controlled clinical trials will be required before rhGH or rhIGF-I can be recommended for treatment of osteoporosis.

References

16.

17.

18.

19. 20.

1. M. Kelijman, Age-related alterations of the GH/IGF-I axis. J. Am. Geriatr. Soc. 39, 295 – 307 (1991). 2. D. V. Rudman, M. H. Vintner, C. M. Rosers, M. F. Lubin, G. H. Fleming, and P. B. Raymond, Impaired GH secretion in the adult population. Relation to age and adiposity. J. Clin. Invest. 67, 1361 – 1369 (1981). 3. H. Rico, A. Del Rio, T. Vila, R. Patino, F. Carrera, and D. Espinos, The role of GH in the pathogenesis of postmenopausal osteoporosis. Arch. Intern. Med. 139, 1263 – 1265 (1979). 4. D. V. Goeddel, H. L. Heyneker, T. Hozumi, R. Arentzen, K. Itakura, D. G. Yansura, M. J. Ross, G. Miozzari, R. Crea, and P. H. Seeburg, Direction expression in E. coli of a DNA sequence coding for hGH. Nature 281, 544 – 548 (1979). 5. D. V. Rudman, A. G. Feller, H. S. Nagrog, G. A. Gergens,P. V. Lalitha, A. F. Goldberg, R. A. Schubler, C. Cohn, I. W. Rudman, and D. E. Mattson, Effect of human growth hormone in men over age 60. N. Engl. J. Med. 323, 52 – 54 (1990). 6. P. Brazeau, W. Vale, and R. Burgus, Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 178, 77 – 79 (1973). 7. J. B. Martin and W. J. Willard, Brain regulation of GH secretion. J. Anim. Sci. 63 (Suppl. 2), 11 (1986). 8. R. Hall, A. V. Schally, and D. Evered, Action of GH release inhibitory hormone in healthy men and acromegaly. Lancet 2, 581 – 584 (1973). 9. M. J. Cronin and M. O. Thorner, Basic studies with GH releasing factor. In “Endocrinology” (L. J. DeGroot, ed.), Vol. 1, pp. 183 – 191. Saunders, Philadelphia, 1989. 10. L. M. Winer, M. A. Shaw, and G. A. Bauman, Basal plasma GH levels in man: New evidence for rhythmicity of GH secretion. J. Clin. Endocrinol Metab. 70, 1678 – 1686 (1990). 11. K. Y. Ho and A. J. Weissberger, Secretory patterns of growth hormone according to sex and age. Horm. Res. 33 (Suppl. 4), 7 – 11 (1990). 12. J. O. Jannson, S. Eden, and O. Isaakson, Sexual dimorphism in the control of growth hormone secretion. Endocr. Rev. 6, 128 – 150 (1985). 13. A. J. Weissberger, K. Y. Ho, and L. Lazarus, Contrasting effects of oral and transdermal routes of estrogen replacement therapy on 24hour growth hormone (GH) secretion, insulin-like growth factor I, and GH-binding protein in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 374 – 381 (1991). 14. N. M. Wright, J. Renault, S. Willi, L. L. Key, and N. H. Bell, Greater growth hormone secretion in black compared with white adult males: Possible factor in greater bone mineral density. J. Bone Miner. Res. 9 (Suppl. 1), S149 (1994). 15. B. Dawson-Hughes, D. Stern, J. Goldman, and S. Reichlin, Regulation of growth hormone and somatomedin-C secretion in post-

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

menopausal women: Effect of physiological estrogen replacement. J. Clin. Endocrinol. Metab. 63, 424 – 432 (1986). M. F. Bellantoni, S. M. Harman, D. Cho, and M. R. Blackman, Effects of progestin-opposed transdermal estrogen administration on GH and IGF-I in postmenopausal women of different ages. J. Clin. Endocrinol. Metab. 72, 172 – 178 (1991). C. J. Rosen, L. R. Donahue, and S. J. Hunter, Age-related changes in insulin-like growth factor binding proteins. J. Clin. Endocrinol. Metab. 71, 575 – 579 (1990). E. Ghigho, E. Arvat, S. Goffi, J. Bellone, M. Nicolosi, M. Procopio, M. Maccario, and F. Cammanni, Neutral control of growth hormone secretion in aged subjects. In “Growth Hormone and Somatomedins during Lifespan” (E. E. Muller, D. Cocchi, and V. Locatelli, eds.), pp. 275–287. Springer-Verlag, Heidelberg, 1993. E. Carpas, M. Harman, and M. R. Blackman, Growth hormone and aging. Endocr. Rev. 14, 20 – 39 (1993). C. J. Rosen, L. R. Donahue, and S. J. Hunter, Insulin-like growth factors and bone: The osteoporosis connection. Proc. Soc. Exp. Biol. Med. 206, 83 – 102 (1994). A. E. Bennett, H. W. Wahner, B. L. Riggs, and R. L. Hintz, IGF-I and -II: Aging and bone density in women. J. Clin. Endocrinol. Metab. 59, 701 – 704 (1994). V. Nicholas, A. Prewett, S. Mohan, R. Finkelman, J. Farley, and D. J. Baylink. An age-related decrease of insulin-like growth factor-binding protein-5 (IGFBP-5) in human cortical bone: Implications for bone loss with aging. J. Bone Miner. Res. 9, S1 (1994). [Abstract 151] V. Nicolas, A. Prewett, P. Bettica, S. Mohan, R. D. Finkelman, D. J. Baylink, and J. R. Farley, Age-related decreases in IGF-I and transforming growth factor-b in femoral cortical bone from both men and women: Implications for bone loss with aging. J. Clin. Endocrinol. Metab. 78, 1011 – 1016 (1994). J. A. Langlois, C. J. Rosen, M. Visser , M. Hannan, T. Harris, P. F. W. Wilson, and D. Kiel. The association between IGF-I and bone mineral density in women and men: the Framingham heart study. J. Clin. Endocrinol. Metab. 83, 4257 – 4262 (1998). D. W. Leung, S. A. Spencer, G. Cachines, R. G. Hammonds, C. Collins, W. J. Henzel, R. Barnard, M. J. Waters, and W. I. Wood, Growth hormone receptor and serum binding protein: Purification, cloning and expression. Nature 330, 537 – 543 (1987). G. Baumann, M. A. Shaw, and K. Amburn, Regulation of plasma growth hormone binding proteins in health and disease. Metabolism 38, 683 – 689 (1989). W. D. Salmon and W. H. Daughaday, A hormonally controlled serum factor which stimulates sulphate incorporation by cartilage in vitro. J. Lab. Clin. Med. 49, 825 – 836 (1957). E. R. Froesch, H. Burgi, E. B. Ramsier, P. Bally, and A. Labhart, Antibody suppressible and non-suppressible insulin-like activities in human serum and their physiological significance. J. Clin. Invest. 42, 1816 – 1834 (1963). N. Dulak and H. Temin, A partially purified polypeptide fraction from rat liver cell conditioned medium and multiplication stimulating activity for embryo fibroblasts. J. Cell. Physiol. 81, 153 – 160 (1973). W. H. Daughaday, K. Hall, M. S. Raben, W. D. Salmon, J. L. Van de Brande, and J. J. Van Wyk, Somatomedin: Proposed designation for sulphation factor. Nature 235, 107 (1972). E. Rinderknecht and R. E. Humbel, The amino acid sequence of human insulin-like growth factor I and its structural homology with proinsulin. J. Biol. Chem. 253, 2769 – 2776 (1978). W. H. Daughaday, K. Hall, W. D. Salmon, J. L. Van den Brande, and J. J. Van Wyk, On the nomenclature of the somatomedins and insulin-like growth factors. J. Clin. Endocrinol. Metab. 65, 1075 – 1076 (1987). M. C. Slootweg, C. M. Hoogerbrugge, T. L. De-porter, S. A. Duursma, and S. C. Van-Buursma-Offers, The presence of classical

764

34. 35. 36. 37.

38.

39.

40.

41.

42.

43. 44.

45.

46. 47.

48.

49.

50.

51.

52.

WÜSTER AND ROSEN IGF type 1 and type 2 receptors on mouse osteoblasts: Autocrine/paracrine growth effect of IGFBPs. J. Endocrinol. 125, 271 – 277 (1990). D. LeRoith, M. Parriza, and V. A. Blakesley. The IGF-I receptor and cancer Endocrine 7, 103 – 105 (1997). C. Sell, R. Baserga, and R. Rubin IGF-I and IGF-I receptor prevent etoposide induced apoptosis. Cancer Res 55, 303 – 306 (1995). S. Mohan and D. J. Baylink, Autocrine-paracrine aspects of bone metabolism. Growth Genet. Horm. 6, 1 – 9 (1990). P. R. Ebeling, J. D. Jones, W. M. O’Fallon, C. H. Janes, and B. L. Riggs, Short term effects of recombinant hIGF-I on bone turnover in normal women. J. Clin. Endocrinol. Metab. 77, 1384 – 1387 (1993). W. F. Blum, M. B. Ranke, K. Kietzman, E. Gauggel, H. J. Zeisel, and J. R. Bierich, A specific radioimmunoassay for the growth hormone (GH)-dependent somatomedin-binding protein: Its use for diagnosis of GH deficiency. J. Clin. Endocrinol. Metab. 70, 1292 – 1298 (1990). J. L. Martin and R. C. Baxter, Insulin-like growth factor-binding proteins (IGF-BPs) produced by human skin fibroblasts: Immunological relationship to other human IGF-BPs. Endocrinology 123, 1907 – 1915 (1988). W. F. Blum and M. B. Ranke, Use of insulin-like growth factorbinding protein 3 for the evaluation of growth disorders. Horm. Res. 33 (Suppl. 4), 31 – 37 (1990). W. F. Blum, K. Albertsson-Wikland, and J. O. L. Jorgensen, Insulinlike growth factor binding protein 3 (IGFBP-3) reflects spontaneous growth hormone (GH) secretion. Horm. Res. 33 (Suppl. 3), 3 – 9 (1990). C. A. Conover, Glycosylation of insulin-like growth factor binding protein-3 (IGFBP-3) is not required for potentiation of IGF-1 action: Evidence for processing of cell-bound IGFBP-3. Endocrinology 129, 3259 – 3268 (1992). P. F. Collett-Solberg and P. Cohen, Genetics, chemistry and function of IGF/IGFB system. Endocrine 12 (2), 121 – 137 (2000). M. Neuberg, L. Buckbinder, B. Seizinger, and N. Kley The p53/IGF-I receptor axis in the regulation of programmed cell death. Endocrine 7, 107 – 109 (1997). P. Cohen, H. C. Graves, D. M. Peehl, M. Kamarei, L. C. Giudice, and R. G. Rosenfeld, PSA is an IGFBP-3 protease found in seminal plasma. J. Clin. Endocrinol. Metab. 75, 1046 – 1051 (1992). S. Mohan, IGFBPs in bone cell regulation. Growth Regul. 3, 67 – 70 (1993). S. E. Nunn, T. Beers Gibson, R. Roopmathy, and P. Cohen, Regulation of prostate cell growth by the IGFBPs and their proteases. Endocrine 7, 115 – 118 (1997). K. Brismar, M. Gutniak, G. Povoa, S. Werner, and K. Hall, Insulin regulates the 35kDa IGFBP in patients with diabetes mellitus. J. Endocrinol. Invest. 11, 599 – 602 (1988). E. Arary, S. Afford, A. J. Strain, P. J. Winwood, M. Arthur, and D. J. Hill, Differential cellular synthesis of IGFBP-1 and IGFBP3 in human liver. J. Clin. Endocrinol. Metab. 79, 1871 – 1876 (1994). C. Schmid, I. Schlapher, M. Waldvogel, P. J. Meier, J. Schwander, M. Boni Schnetzler, J. Zapf, and E. R. Froesch, Differential regulation of IGFBP-2 mRNA in liver and bone cells by insulin and retinoic acid. FEBS Lett. 303, 205 – 209 (1992). J. H. M. Feyen, D. B. Evans, C. Binkert, G. F. Heinrich, S. Geisse, and H. P. Kocher, Recombinant human Cyst 281 IGFBP-2 inhibits both basal and IGF-I stimulated proliferation and collagen synthesis in fetal rat calvariae. J. Biol. Chem. 266, 19469 – 19474 (1991). S. Mohan, C. M. Bautista, J. Wergedal, and D. J. Baylink, Isolation an inhibitory insulin-like growth factor (IGF) binding protein from bone cell-conditioned medium: A potential local regulator of IGF action. Proc. Natl. Acad. Sci. USA 86, 8338 – 8342 (1989).

53. S. H. Scharla, D. D. Strong, S. Mohan, D. J. Baylink, and T. A. Linkhart, 1,25-Dihydroxyvitamin D differentially regulates the production of insulin-like growth factor I (IGF-1) and IGF-binding protein-4 in mouse osteoblasts. Endocrinology 129, 3139 – 3146 (1991). 54. S. K. Durham, B. L. Riggs, and C. A. Conover, IGFBP-4 – IGFBP-4 protease system in normal human osteoblast like cells. J. Clin. Endocrinol. Metab.79, 1752 –1758 (1994). 55. C. J. Rosen, L. R. Donahue, S. J. Hunter, M. J. Holick, H. Kavookjian, A. Kirschenbaum, S. Mohan, and D. J. Baylink, The 24/25kDa serum IGFBP is increased in elderly women with hip and spine fractures. J. Clin. Endocrinol. Metab. 74, 24 – 27 (1992). 56. Y. Honda, E. C. Landale, D. D. Strongg, D. J. Baylink, and S. Mohan, Recombinant synthesis of IGFBP-4: Development, application and validation of a radioimmunoassay for IGFBP-4 in human serum. J. Clin. Endocrinol. Metab. 81, 1389 – 1396 (1996). 57. C. A. Conover, L. K. Bale, J. T. Clarkson, and O. Torring, Regulation of IGFBP-5 mRNA expression and protein availability in rat osteoblast like cells. Endocrinology 132, 2525 – 2530 (1993). 58. M. Koutsilieris and C. Polychroakos, Proteinolytic activity against IGFBPs involved in the paracrine interactions between prostate adenocarcinoma cells and osteoblasts. Anticancer Res. 12, 905 – 910 (1992). 59. H. Green, M. Morikawa, and T. Nixon, A dual effector theory of GH action. Differentiation 29, 195 – 198 (1985). 60. O. G. P. Isaakson, A. Lindahl, A. Nilsson, and J. Isgaard, Mechanism of the stimulatory effect on GH on longitudinal bone growth. Endocr. Rev. 8, 426 – 438 (1987). 61. O. G. P. Isaakson, S. M. Russell, and E. M. Spencer, Local injections of human or rat GH or of purified IGF-I stimulate unilateral tibial epiphyseal growth in hypophysectomized rats. Endocrinology 116, 2563 – 2568 (1985). 62. C. Schmid and M. Ernst, IGFs. In “Cytokines and Bone Metabolism (M. Gowen, ed.), pp. 229 – 259. CRC Press, Boca Raton, FL, 1991. 63. M. Ernst and G. A. Rodan, Increased activity of IGF in osteoblastic cells in the presence of GH. Endocrinology 127, 807 – 814 (1990). 64. G. Maor, Z. Hochberg, K. Mark, D. Heinegard, and M. Silbermann, Human GH enhances chondrogenesis and osteogenesis in tissue culture system of chondroprogenitor cells. Endocrinology 125, 1239 – 1245 (1989). 65. M. Ernst and E. R. Froesch, GH dependent stimulation ofosteoblast like cells in serum free cultures via local synthesisof IGF-I. Biochem. Biophys. Res. Commun. 151, 142 – 147 (1988). 66. M. Kassem, W. Blum, J. Ristelli, L. Mosekilde, and E. F. Eriksen, GH stimulates proliferation and differentiation of normal human osteoblast-like cells in vitro. Calcif. Tissue Int. 52, 222 – 226 (1993). 67. S. Mohan and D. J. Baylink, Bone growth factors. Clin. Orthop. Relat. Res. 63, 30 – 48 (1991). 68. E. G. Slootweeg, W. W. E. Van Beek, L. P. C. Schot, S. E. Papapaoulos, C. W. G. M. Lowik, Osteoclast formation together with interleukin- 6 production in mouse long bones is increased by insulinlike growth factor-I. J. Endocrinol. 132, 433 – 438 (1992). 69. H. Mochizuki, Y. Hakeda, N. Wakasuki, N. Usui, N. S. Akashi, T. Sato, and K. Tanaka, IGF-1 supports formation and activation of osteoclasts. Endocrinology 131, 1075 – 1080 (1992). 70. S. L. Abboud, C. R. Bethel, and D. C. Aron, Secretion of IGF-I and IGFBPs by murine bone marrow stromal cells. J. Clin. Invest. 88, 470 – 475 (1991). 71. C. A. Conover, Insulin-like growth factors as modulators: Post translational regulation. In “The Endocrine Society, 76th Annual Meeting,” p. 177. (abst). Endocrine Society Press, 1994. 72. S. Mohan, T. Linkhart, R. Rosenfeld, and D. J. Baylink, Characterization of the receptor for IGF-II bone cells.J. Cell. Physiol.140, 169 –176 (1989). 73. S. Mohan and D. J. Baylink, The Role of IGF-II in the coupling of bone formation to resorption. In “Modern Concepts of IGFs” (E. M. Spencer, eds.), pp. 169 – 184. Elsevier, New York, 1991.

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors 74. C. Schmid, T. Steiner, and E. R. Froesch, IGF-I supports differentiation of cultured osteoblast-like cells. FEBS Lett. 173, 48 – 5 2 (1984). 75 . C. Chenu, A. Opran-Valentin, P. Chavassieux, S. Saez, P. J. Meunier, and P. D. Delmas, IGF-I hormonal regulation by GH and 1,25 dihydroxyvitamin and activity on human osteoblast-like cells in short term cultures. Bone 11, 81–86(1990). 76. H. L. Merriman, D. LaTour, T. A. Linkhar, S. Mohan, D. J. Baylink, and D. D. Strong, IGF-I and IGF-II induce c fos in mouse osteoblast cells. Calcif. Tissue Int. 46, 258 – 262 (1990). 77. C. Wuster, G. Duckeck, A. Ugurel, M. Lojen, H. W. Minne, and R. Ziegler, bone mineral content of spine and forearm in osteoporosis and in german normals. Eur. J. Clin. Invest. 22, 366 – 370 (1992). 78. P. D. Delmas, P. Chatelain, L. Malaval, and G. Bonne, Serum bone Gla protein in GH deficient children. J. Bone Miner. Res. 1, 333 – 338 (1986). 79. C. Wuster, E. Slenczka, and R. Ziegler, Increased prevalence of osteoporosis and arteriosclerosis in patients with conventionally substituted pituitary insufficiency: Is there a need for additional GH substitution? Klin. Wochenschr. 69, 769 – 773 (1991). 80. A. G. Johansson, P. Burman, K. Westermark, and S. Ljunghall, The bone mineral density in acquired growth hormone deficiency correlates with circulating insulin-like growth factor I. J. Intern. Med. 232, 447 – 452 (1992). 81. S. L. Hyer, D. A. Rodin, J. H. Tobias, A. Leiper, and S. S. Nussey, Growth hormone deficiency during puberty reduces adult bone mineral density. Arch. Dis. Child. 67, 1472 – 1474 (1992). 82. T. Rosen, T. Hansson, H. Granhed, J. Szucs, and B. A. Bengtsson, Reduced bone mineral content in adults with growth hormone deficiency. Acta Endocrinol. 129, 201 – 206 (1993). 83. R. G. Bing-You, M. C. Denis, and C. J. Rosen, Low bone mineral density in adults with previous hypothalamic-pituitary tumors: Correlations with serum GH responses to GHRH, IGF-I and IGFBP-3. Calcif. Tissue Int. 52, 183 – 187 (1993). 84. H. DeBoer, G. J. Blok, A. Van Lingen, G. J. J. Teule, P. Lips, and E. Van Der Veen, Consequences of childhood-onset growth hormone deficiency for adult bone mass. J. Bone Miner. Res. 9, 1319 – 1326 (1994). 85. C. H. Wuster, Growth hormone and bone. Acta Endocrinol. 128, S2, 14 – 18 (1993). 86. J. M. Kaufman, P. Taelman, A. Vermeulen, and M. Vandeweghe, Bone mineral status in growth hormone-deficient males with isolated and multiple pituitary deficiencies of childhood onset. J. Clin. Endocrinol. Metab. 74, 118 – 123 (1993). 87. P. Kann, B. Piepkorn, A. Pfutzner, R. Peipenburg, and J. Beyer, Bone quality in growth hormone deficient adults. Acta Endocrinol. 128 (Suppl. 2), 60 (1993). 88. N. Bravenboer, P. Lips, and G. J. Blok, The effect of GH on bone mass and bone turnover of GHD men. J. Bone Miner. Res. 9, S1, B253 (1994). 89. H. K. Nielsen, J. O. L. Jorgensen, K. Brixen, and J. S. Christiansen, Serum osteocalcin and bone isoenzyme alkaline phsophatase in GHD patients: Dose response studies with biosynthetic GH. Calcif. Tissue Int. 46, 82 – 87 (1991). 90. J. Halse and J. O. Gordeladze, Total and non-dialyzable urinary hydroxyproline in acromegalics and control subjects. Acta Endocrinol. 96, 451 – 457 (1981). 91. C. De La Piedra, J. Larranaga, N. Castro, C. Horcajada, A. Rapada, and J. L. H. Pompo, Correlation among plasma osteocalcin, growth hormone and somatomedin C in acromegaly. Calcif. Tissue Int. 43, 44 – 45 (1988). 92. R. Bouillon, Growth hormone and bone. Horm. Res. 36 (Suppl. 1), 49 – 55 (1991). 93. J. Halse and J. O. Gordeladze, Urinary hydroxyproline excretion in acromegaly. Acta Endocrinol. 89, 483 – 491 (1978).

765 94. S. Ezzat, S. Melmed, D. Endres, D. Eyre, and F. Singer, Biochemical assessment of bone formation and resorption in acromegaly. J. Clin. Endocrinol. Metab. 76, 1452 – 1457 (1993). 95. J. Halse, F. Melsen, and L. Mosekilde, Iliac crest bone mass and remodelling in acromegaly. Acta Endocrinol. 97, 18 – 22 (1981). 96. L. R. Riggs, R. V. Randall, H. W. Wahner, J. Jowsey, P. J. Kelly, and M. Singh, The nature of the metabolic bone disorder in acromegaly. J. Clin. Endocrinol. Metab. 34, 543 – 551 (1972). 97. J. F. Aloia, M. S. Roginsky, J. Lowsey, C. S. Dombrowski, K. K. Shukla, and S. H. Cohn, Skeletal metabolism and body composition in acromegaly. J. Clin. Endocrinol. Metab. 35, 543 – 551 (1972). 98. T. Diamond, L. Nery, and S. Posen, Spinal and peripheral bone mineral densities in Acromegaly. The effects of excess GH and hypogonadism. Ann. Intern. Med. 111, 567 – 573 (1989). 99. E. Seeman, W. H. Wahner, K. P. Offord, R. Kumar, W. Johnson, and B. L. Riggs, Differential effects of endocrine dysfunction on the axial and the appendicular skeleton. J. Clin. Invest. 69, 1302 – 1309 (1982). 100. J. Dequeker, A. Burssens, and R. Bouillon, Dynamics of GH secretion in patients with osteoporosis and in patients with osteoarthrosis. Horm. Res. 16, 353 – 356 (1982). 101. C. H. R. Wuster, W. F. Blum, S. Schlemilch, M. Ranke, and R. Ziegler, Decreased serum levels of insulin like growth factors 1 and 2 and IGF binding protein-3 in patients with osteoporosis. J. Intern. Med. 234, 249 – 255 (1993). 102. D. C. Bauer, C. J. Rosen, J. Cauley, and S. R. Cummings, Low serum IGF-I but not IGFBP-3 predicts hip and spine fractures; Bone 1997 S 561. 103. P. Gamero, E. Sornay-Rendu, and P. D. Delmas, Low serum IGF-1 and occurrence of osteoporotic fractures in postmenopausal women [letter]. Lancet 355 (9207), 898 – 899 (2000). 104. S. Ljunghall, A. G. Johansson, P. Burman, O. Kampe, E. Lindh, and F. A. Karlsson, Low plasma levels of insulin-like growth factor 1 (IGF-1) in male patients with idiopathic osteoporosis. J. Int. Med. 232, 59 – 64 (1992). 105. T. Nakamura, T. Hosi, Y. Mizumo, Y. Ouchi, M. Shiraki, and H. Orimo, Clinical significance of serum levels of insulin like growth factors as bone metabolic markers in postmenopausal women. Bone Miner. 17 (Suppl. 1), 170 (1992). [abstract] 106. A. Johansson, A. Forslund, L. Hambraeus, W. F. Blum, and S. Ljunghall, GH dependent IGFBP is a major determinant of bone mineral density in healthy men. J. Bone Miner. Res 9, 915 – 921 (1994). 107. E. Kurland, C. J. Rosen, F. Cosman, D. McMahon, E. Shane, R. Lindsay, D. Dempster, and Bilezikian Osteoporosis in men:Abnormalities in the IGF-I axis. J. Clin. Endocrinol. Metab. 82, 2799 – 2805 (1997). 108. E. S. Kurland, F. K. W. Chan, C. J. Rosen, and J. P. Bilezikian, Growth hormone axis in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 83, 2576 – 2579 (1998). 109. S. Mohan, J. Dequeker, R. Van Den Eyned, J. Peeters, J. Aerssens, and D. J. Baylink, Increased IGF-I and IGF-II in bone from patients with osteoarthritis. J. Bone Miner. Res. 1, S131 (1991). 110. J. C. Fleet, E. Bruns, J. M. Hock, and R. J. Wood, GH and PTH stimulate intestinal calcium absorption in aged female rats. Endocrinology 134, 1755 – 1760 (1994). 111. T. Rosen, Increased fracture frequency in adult patients with hypopituiarism and GH deficiency. Eur.J.Endocrinol. 137, 240 – 245 (1997). 112. H. P. Guler, J. Zapf, E. Scheiwiller, and E. R. Froesch, Recombinant IGF-I stimulates growth and has distinct effects on organ size in hypophysectomized rats. Proc. Natl. Acad. Sci. USA 85, 4889 – 4893 (1988). 113. R. Rappaport and P. Czernichow, Disorders of GH and prolactin secretion. In “Pediatric Endocrinology” (J. Bertrand, R. Rappapaort, and P. C. Sizonenko, eds.), pp. 220 – 241. Williams and Wilkins, Baltimore, 1993.

766 114. D. E. Carey, B. Goldberg, S. K. Ratzan, K. R. Rubin, and D. W. Rowe, Radioimmunoassay for type I procollagen in GH deficient children before and during treatment with GH. Pediatr. Res. 19, 8 – 11 (1985). 115. S. DeMuinck Keizer-Schrama, B. Rikken, H. J. Wynne, A. C. S. Hokken-Koelega, J. M. Wit, A. Bot, and S. L. Drop, Dose – response study of biosynthetic hGH in GH deficientchildren. J. Clin. Endocrinol. Metab. 74, 898 – 905 (1992). 116. S. E. Inzucchi and R. J. Robbins, Effects of GH on human bone biology. J. Clin. Endocrinol. Metab. 79, 691 – 694 (1994). 117. C. Saggese, G. I. Baroncelli, S. Bertelloni, L. Cinquanta, and G. DiNero, Effects of long term treatment with GH in bone and mineral development in children with GH deficiency. J. Pediatr. 122, 37 – 45 (1993). 118. D. M. Crist, G. T. Peake, P. A. Eagen, and D. L. Waters, Body composition response to exogenous GH during training in healthy conditioned adults. II. Effects on exercise performance. J. Appl. Physiol. 70, 695 – 700 (1991). 119. L. Holloway, G. Butterfield, R. L. Hintz, N. M. Gesundheit, and R. Marcus, Effects of hGH on metabolic indices, body composition, and bone turnover in healthy elderly women. J. Clin. Endocrinol. Metab. 79, 470 – 479 (1994). 120. K. Brixen, H. K. Nielsen, L. Mosekilde, and A. Flyvbjerg, A short course of rhGH stimulates osteoblasts and activates bone remodeling in normal human volunteers. J. Bone Miner. Res. 5, 609 – 618 (1990). 121. M. Kassem, K. Brixen, W. F. Blum, L. Mosekilde, and E. F. Eriksen, Normal Osteoclastice and osteoblastic response to exogenous GH in patients with postmenopausal spinal osteoporosis. J. Bone Miner. Res. 9, 1365 – 1370 (1994). 122. J. O. L. Jorgensen, S. A. Pedersen, L. Thuesen, J. Jorgensen, N. Moller, J. Muller, N. E. Skakkebaek, and J. S. Christiansen, Long term GH treatment in GH deficient adults. Acta Endocrinol. 125, 449 – 453 (1991). 123. N. K. Valk, A. J. Lely, W. W. Herder, J. Lindemans, and S. W. Lamberts, Effects of GH administration in GH adults: A 20 day metabolic ward Study. J. Clin. Endocrinol. Metab. 79, 1070 – 1076 (1994). 124. R. C. Cuneo, F. Salomon, C. W. Wiles, R. Hesp, and P. H. Sonksen, GH treatment in GH deficient adults. I. Effects on muscle mass and strength. J. Appl. Physiol. 70, 688 – 694 (1991). 125. J. O. L. Jorgensen, L. Thuesen, T. Ingemann-Hansen, S. A. Pedersen, J. U. Jorgensen, N. E. Skakkebaek, and J. S. Christiansen, Beneficial effects of GH treatment in GH deficient adults. Lancet 1, 1221 – 1225 (1989). 126. E. A. Van der Veen and J. C. Netelembos, Growth hormone replacement therapy in adults: Bone and calcium metabolism. Horm. Res. 33, S4, 65 – 68 (1990). 127. J. S. Johansen, S. A. Pedersen, and J. O. L. Jorgensen, Effects of GH on plasma bone GLA protein in GHD adults. J. Clin. Endocrinol. Metab. 70, 916 – 919 (1990). 128. J. S. Christiansen, J. O. Jorgensen, S. A. Pedersen, J. Muller, J. Jorgensen, J. Moller, L. Heickendorf, and N. E. Skakkebaek, GH replacement therapy in adults. Horm. Res. 36 (Suppl. 1), 66 – 72 (1991). 129. J. O. L. Jorgensen, S. A. Pedersen, L. Thusen, J. Jorgensen, N. E. Skakkebaek, and J. S. Christiansen, Long-term GH replacement therapy in adult GH deficient patients. Horm. Res. 33 (Suppl. 3), 5 (1990). 130. H. M. Whitehead, C. Boreham, E. M. McIlrath, B. Sheridan, L. Kennedy, A. B. Atkinson, and D. R. Hadden, GH treatment of adults with GH deficiency; results of a 13 month placebo controlled cross-over study. Clin. Endocrinol. 36, 45 – 52 (1992). 131. Y. J. H. Janssen, N. A. T. Hamdy, M. Frolich, and F. Roelfsema. Skeletal effects of two years of treatment with low physiological

WÜSTER AND ROSEN doses of rhGH in patients with adult onset GHD. J. Clin. Endocrinol. Metab. 83, 2143 – 2146 (1998). 132. G. Johannsson, T. Rosen, I. Bosaeus, L. Sjostrom, and B. A. Bengrsson, Two years of GH treatment increases BMD and density in hypopituitary pateints with adult onset GH deficiency. J.Clin.Endocrinol.Metab. 81, 2865 – 2873.(1996). 133. H. B. Baum, B. M. Biller, and J. S. Dinkelstein, Effects of physiologic GH therapy on bone density and body composition in patients with adult onset growth hormone deficiency. Ann. Intern. Med. 125, 883 – -890 (1996). 134.M. A. Papadakis, D. Grady, D. Black, et al. 1996. Growth hormone replacement in healthy older men improves body composition but not functional ability. Ann. Int. Med. 124, 708 – 716. 135. J. W. Finkelstein, H. P. Roffwarg, and R. M. Boyar, Age-related change in the twenty four hour spontaneous secretion of GH. J. Clin. Endocrinol. Metab. 35, 665 – 670 (1972). 136. S. A. Lieberman, A. M. Mitchell, R. Marcus, R. L. Hintz, and A. R. Hoffman, The insulin-like growth factor I generation test: Resistance to GH with aging and estrogen replacement therapy. Horm. Metab.Res. 26, 229 – 233 (1994). 137. C. J. Rosen, J. Friez, D. B. MacLean, K. Berg, D. P. Kiel, The RIGHT study: A randomized placebo controlled trial of recombinant human growth hormone in frail elderly: Dose response effects on bone mass and bone turnover. J. Bone Miner. S 14, 208 (1999). 138. D. V. Rudman, A. G. Feller, K. R. Sully, I. Cohen, I. W. Rudman, and M. W. Draper. Effect of hGH on body composition in elderly men. Horm. Res. 36 (1), 73 – 81 (1991). 139. R. Marcus, G. Butterfield, L. Holloway, L. Gilliland, D. J. Baylink, R. L. Hintz, and B. Sherman, Effects of short term administration of rhGH to elderly people. J. Clin. Endocrinol. Metab. 70, 519 – 525 (1990). 140. H. P. Kruse and F. Kuhlencordt, On an attempt to treat primary and secondary osteoporosis with human growth hormone. Horm. Metab. Res. 7, 488 – 491 (1975). 141. J. F. Aloia, I. Zanz, K. Ellis, J. Jowsey, M. Rogmilin, S. Wallach, and S. H. Cohn, Effects of GH in osteoporosis. J. Clin. Endocrinol. Metab. 43, 992 – 999 (1976). 142. J. F. Aloia, I. Zanzi, A. Vaswani, K. Ellis, and S. H. Cohn, Combination therapy for osteoporosis. Metabolism 26, 787 – 792 (1977). 143. J. F. Aloia, A. Vaswani, P. J. Meunier, C. M. Adouard, M. E. Arlot, J. K. Yeh, and S. H. Cohn, Coherence treatment of postmenopausal osteoporosis with GH and calcitonin. Calcif. Tissue Int. 40, 253 – 259 (1987). 144. M. A. Dambacher, T. Lauffenburger, and H. G. Haas, Vergleich vershiender medikamentoser Therapierformen bet osteoporose. Akt. Rheumatol. 7, 249 – 252 (1982). 145. A. G. Johannson, E. Lindh, W. F. Blum, G. Kollerup, O. H. Sorensen, and S. Ljunghall, Effects of shortterm treatment with IGF-I and GH on markers of bone metabolism in idiopathic osteoporosis. J. Bone Miner. Res. 9, S328 (1994). 146. L. Holloway, L. Kohlmeier, K. Kent, and R. Marcus, Skeletal effects of cyclic recombinant human growth hormone and salmon calcitonin in osteopenic postmenopausal women. J. Clin. Endocrinol. Metab. 82, 1111 – 1117 (1997). 147. E. Schoenle, J. Zapf, R. E. Humbel, and E. R. Froesch, IGF-I stimulates growth in hypophysectomized rats. Nature 296, 252 – 253 (1982). 148. E. Schoenle, J. Zapf, C. Hauri, T. Steiner, and E. R. Froesch, Comparison of in vivo effects of IGF-I and IGF-I and of growth hormone in hypophysectomized rats. Acta Endocrinol. 108, 167 – 174 (1985). 149. J. Zapf, C. Hauri, M. Waldvogel, E. Futo, H. Hasler, K. Binz, H. P. Guler, C. Schmid, and E. R. Froesch, Recombinant human IGF-I induces its own specific carrier protein in hypophysectomized and diabetic rats. Proc. Natl. Acad. Sci. USA 86, 3813 – 3817 (1989).

CHAPTER 78 Growth Hormone, Insulin-like Growth Factors 150. M. Tixier-Bouchard, M. Huybrechts, F. Decuypere, E. R. Kuhn, J. L. Monvoisin, G. Coquerelle, J. Charrier, and J. Simon, Effects of IGF-I infusion and dietary T3 supplementation on growth body composition, and plasma hormone levels in sex-linked dwarf mutant and normal chickens. J. Endocrinol. 133, 101 – 110 (1992). 151. J. Zapf, H. P. Guler, C. H. Schmid, A. Kurtz, and E. R. Froesch, In vivo actions of IGF-I. In “Advances in Growth Hormone and Growth Factor Research” (E. E. Muller, D. Cocchi, and V. Locatelli, eds.), pp. 245 – 260. Pythagora Press, Rome, Milan, Berlin-Heidelberg, 1990. 152. J. Verhaeghe, A. M. Suiker, W. J. Visser, E. Van Herck, R. Van Bree, and R. Bouillon, The effects of systemic insulin, IGF-I and GH on bone growth and turnover in spontaneously diabetic BB rats. J. Endocrinol. 134, 485 – 492 (1992). 153. C. Schmid, H. P. Guler, D. Rowe, and E. R. Froesch, IGF-I regulates type I procollagen mRNA steady state levels in bone of rats. Endocrinology 125, 1575 – 1580 (1989). 154. E. M. Spencer, C. C. Liu, and G. A. Howard, In vivo actions of IGF-I on bone formation and resorption in rats. Bone 12, 21 – 26 (1991). 155. J. H. Tobias, J. W. M. Chow, and T. J. Chambers, Opposite effects of IGF-I on the formation of trabecular and cortical bone in adult female rats. Endocrinology 131, 2387 – 2392 (1992). 156. L. R. Donahue, G. Watson, W. G. Beamer, Regulation of metabolic water and protein compartments by IGF-I and testosterone in growth hormone deficient lit/lit mice. J. Endocrinol. 139, 431 – 439 (1993). 157. D. N. Kalu, C. C. Liu, E. Salerno, M. Salih, R. Echon, M. Ray, and B. W. Hollis, IGF-I partially prevents ovariectomy-induced bone loss: A comparative study with hPTH. J. Bone Miner. Res. 6, 548 (1991). 158. P. Ammann, R. Rizzoli, J. Meyer, D. Siosman, and J. P. Bonjour, Bone mechanical properties and mineral density in IGF-I and pamidronate treated ovariectomized rats. J. Bone Miner. Res. 8, 612 (1993). 159. K. Mueller, R. Cortesi, D. Modrowski, and P. J. Marie, Stimulation of trabecular bone formation by IGF-I in adult ovariectomized rats. Am. J. Physiol. 267, E1 – E6 (1994). 160. J. V. Santiago, W. H. McAlister, and S. K. Ratzan, Decreased cortical thickness and osteopenia in children with diabetes mellitus. J. Clin. Endocrinol. Metab. 45, 845 – 848 (1977). 161. P. McNair, Bone mineral metabolism in type I diabetes mellitus. Dan. Med. Bull. 35, 109 – 121 (1988). 162. R. B. Mazess, Diabetes mellitus and the risk of skeletal fractures. N. Engl. J. Med. 304, 115 – 116 (1981). 163. H. Heath, L. J. Melton, and C. P. Chu, Diabetes mellitus and the risk of skeletal fracture. N. Engl. J. Med. 303, 567 – 570 (1980). 164. J. Tuukkanene, B. Wallmark, P. Jalovaara, et al., Changes induced in growing rat bone by immobilization and remobilization. Bone 12, 113 – 118 (1991).

767 165. R. P. Heaney, Radiocalcium metabolism in disuse osteoporosis in man. Am. J. Med. 33, 188 – 200 (1962). 166. D. D. Bikle, J. Harris, B. P. Halloran, and E. R. Morey-Holton, Skeletal unloading induces resistance to IGF-I. J. Bone Miner. Res. 11, 1789 – 1796 (1994). 167. M. Machwater, E. Zerath, X. Holy, P. Pastoureau, and P. J. Marie, IGF-I increases trabecular bone formation and osteoblastic cell proliferation in unloaded rats. Endocrinology 134, 1031 – 1038 (1994). 168. C. M. Bagi, R. Brommage, L. Deleon, S. Adams, D. Rosen, and A. Sommer, Benefit of systemically administered rhIGF-I and rhIGFI/IGFBP-3 on cancellous bone in ovarectomized rats.J. Bone Miner. Res. 9, 1301 – 1312 (1994). 169. R. Bromage, E. Millerman, E. Swett, E. DeLeon, S. Adams, and C. M. Bagi, Treatment with the rhIGF-I/IGFBP-3 complex increases cortical bone and lean body mass in oophorectomized rats. J. Bone Miner. Res. 9, S1 (1993). 170. L. Ghiron, J. Thompson, L. Halloway, , G. E. Butterfield, A. R. Hoffman, and R. Marcus, Effects of rhGH and IGF-I on bone turnover in elderly women. J. Bone Miner. Res. 10, 1844 – 1852 (1995). 171. C. A. Bondy, Clinical uses of IGF-I. Ann. Intern. Med., 20, 593 – 601 (1994). 172. Z. Laron and S. B. Klipper-Auerbach, Effects of IGF on linear growth, head circumference and body fat in patients with Larontype dwarfism. Lancet 339, 1258 – 1261 (1992). 173. K. A. Woods, C. Camacho-Hubner, M. O. Savage, and A. J. L. Clark, Intrauterine growth retardation and postnatal growth failure associated with deletion of the IGF-I gene. N. Engl. J. Med. 335, 1363 – 1367 (1996). 174. C. Camacho-Hubner, K. A. Woods, F. Miraki-Moud, P. C. Hindsmarsh, A. J. Clark, Y. Hansson, A. Johnston, R. C. Baxter, and M. O. Savage, Effects of recombinant hIGF-I therapy on the GH/IGF system of a patient with a aprtial IGF-I gene deletion. J. Clin. Endocrinol. Metab. 84, 1611 – 1650 (1999). 175. L. K. Bachrach, R. Marcus, S. M. Ott, A. L. Rosenbloom, A. Vasconex, V. Martinex, A. L. Martinex, et al. Bone mineral histomorphometry and body composition in adults with growth hormone receptor deficiency. J. Bone Miner. Res. 13, 415 – 421 (1998). 176. S. Ljunghall, A. G. Johansson, P. Burman, O. Kampe, E. Lindh, and F. A. Karlsson, Low plasma levels of IGF-I in male patients with idiopathic osteoporosis. J. Intern. Med. 232, 59 – 64 (1992). 177. A. G. Johansson, E. Lindh, and S. Ljunghall, IGF-I stimulates bone turnover in osteoporosis. Lancet 339, 1619 (1992). 178. L. R. Donahue and C. J. Rosen, IGFs and bone: The osteoporosis connection revisited. Proc. Soc. Exp. Bio. Med. 219, 1 – 7 (1998). 179. P. Geusens, R. Bouillon, P. Broos, D. M. Rosen, S. Adams, M. Sanders, J. Raus and S. Boonen, Musculoskeletal effects of rhIGFI/IGFBP-3 in hip fracture patients: results from a double blind placebo controlled phase II study. Bone 23 (5), S157 (1998).

CHAPTER 79

New Approaches to Osteoporosis Therapeutics S. AUBREY STOCH, MICHAEL CHOREV, AND MICHAEL ROSENBLATT Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215

VI. VII. VIII. IX. X.

I. Introduction II. Parathyroid Hormone and Parathyroid Hormone-Related Protein III. Calcium Receptor Antagonists IV. Stimulation of Bone Formation and Inhibition of Resorption by Statins V. Inhibition of Bone Resorption with v3 Integrin Antagonists

I. INTRODUCTION Greater use of diagnostic modalities and the emergence of more effective therapies have made the treatment of osteoporosis more satisfying for physicians and patients. Osteoporosis affects 4 to 6 million women and 1 to 2 million men in the United States; an even larger population is at risk with osteopenia or decreased bone mineral density [1]. Fractures — the most deleterious consequence of this disease — consumed $13.8 billion in health-care expenditures in 1995 [2]. With increased awareness made possible by new and more sensitive diagnostic testing modalities, physicians and patients are focusing their attention on novel therapies. We are fortunate that advances in basic scientific research coupled with carefully conducted clinical trials have enriched our therapeutic armamentarium. While more FDA-approved therapies are available than ever before, some patients are still intolerant or do not respond to

OSTEOPOROSIS, SECOND EDITION VOLUME 2

Osteoprotegerin Cathepsin K Inhibitors of Src in Osteoporosis Combined Therapies Summary References

existing treatments. Novel therapies currently under study exploit different targets in the osteoclast – osteoblast remodeling cycle and will provide “fine-tuning” of therapy selection for different patient groups. We have focused our attention on discussing novel therapeutic approaches to treat osteoporosis, and may not have devoted sufficient attention to every candidate that holds therapeutic promise. Wherever possible we have concentrated on entities that are close to or already in clinical trials. We have also included data on certain emerging therapies that take advantage of novel pathophysiological mechanisms (Fig. 1, see also color plate), even though these therapies are not ready for Phase I studies. In addition, agents that diminish bone resorption and stimulate bone formation have also been included, as well as combination therapies. The latter approach has received enthusiastic attention both in scientific literature and in clinical practice.

769

Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.

770

STOCH, CHOREV, AND ROSENBLATT

FIGURE 1

Schematic representation of selected targets in the osteoclast-pathway. These include anti-resorptive agents that may hold promise as future therapies to treat osteoporosis. Modified with permission from Gowen et al. [385a]. (See also color plate.)

II. PARATHYROID HORMONE AND PARATHYROID HORMONE-RELATED PROTEIN A. Background Current osteoporosis therapies reduce bone turnover — resulting in small increments in bone mass during the first 2 – 3 years of therapy with a concomitant fracture rate reduction of up to 50% [3]. Anabolic agents, by contrast, stimulate bone formation, and furthermore may restore bone architecture, leading to increased mechanical strength. Treatment with parathyroid hormone (PTH) or its analogs results in large increases in bone mass. This anabolic effect was initially reported by Albright and coworkers, who injected rats with parathyroid extracts [4]. This same observation was subsequently confirmed by Seyle [5] using parathyroid extract and, many years thereafter, by Reeve and coworkers using a synthetic N-terminal PTH fragment [6]. As a result, the established paradigm accepts intermittent low-dose PTH as an anabolic agent in bone tissue (see also Chapter 77).

The intermittent administration of PTH-(1 – 34) to ovariectomized (OVX) rats stimulates trabecular bone formation in the proximal tibia [7 – 11] femoral neck [12,13], distal femur [8,14 – 16], and vertebrae [16 – 20]. Similar observations have been reported in senile rats with established osteopenia [8,13,19,21,22]. PTH treatment is effective, provided that there are remnants of trabecular bone spicules present [23]. It has been shown that intermittent PTH administration not only increases bone formation in senile rats but also increases mechanical strength, decreases brittleness, and increases “load-to-failure” levels compared to sham and OVX control rats [24]. Furthermore, substantial gains in cortical bone mass have also been reported [25 – 34]. Parathyroid hormone-related protein (PTHrP), the only other known endogenous ligand of the PTH1 receptor, similarly has been shown to elicit anabolic effects on bone [35 – 38]. Hock and coworkers reported that hPTHrP(1 – 34) is less potent and less effective than hPTH-(1 – 34) in inducing an anabolic response in young normal rats [35]. Furthermore, hPTHrP-(1 – 34) is 3- to 10-fold less potent than hPTH-(1 – 34) in inducing renal responses following continuous infusions at a constant dose for 12 h in young healthy men [39]. This difference may be attributed to the shorter apparent half-life (8.3 min versus 10.2 min) and accelerated plasma clearance (4.0 liters/min versus 2.0 liters/min) of the PTHrP compared to the PTH peptide [36]. However, in 4-week-old rats PTHrP-(1 – 74) produces larger anabolic effects than PTH-(1 – 34) [40]. Unlike hPTH-(1 – 31)NH2, PTHrP-(1 – 31)NH2 does not stimulate trabecular growth in the distal femurs of OVX rats [38]. These findings suggest that the residues C-terminal to Ala34 may confer greater metabolic stability and thereby increase the circulating concentration of PTHrPderived peptides that are longer than the 34-amino-acid fragment. The ligand-selectivity of the PTH2 receptor (PTH-Rc) for PTH [41,42] indicates that peptides based on the PTHrP sequence may be more suitable as osteogenic anabolic candidates than those based on the PTH sequence. Furthermore, a potential reduction in side effects may be evident with a PTHrP-derived anabolic compound that does not interact with the PTH2-Rc in tissues and organs which are not related to bone physiology. Therefore, the development of RS-66271 [43], a PTHrP-based anabolic peptide, is of great interest. A major concern related to the therapeutic use of PTH is the existence of a narrow toxic/therapeutic window associated with the duration of exposure to PTH. This problem underscores the importance of correct dosing kinetics during PTH-based therapy. For example, intermittent administration of PTH results in an anabolic effect on bone — whereas infusion of PTH using regimens of 2 h/day and longer are sufficient to induce hypercalcemia

771

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

and bone loss (Fig. 2) [44,45]. Nevertheless, postmenopausal women treated with intermittent PTH experience an anabolic response similar to that seen in animal studies [6,46 – 50]. This well established anabolic effect of intermittent PTH on bone is extensively summarized elsewhere in this volume (Chapter 77) and in numerous reviews [3,51 – 55]. The desirable anabolic effects seen with the intermittent administration of small doses of PTH has been the impetus to design novel PTH analogs that are better tolerated, smaller in size, have greater metabolic stability, and demonstrate distinct signaling selectivity.

FIGURE 2

B. Design of New PTH Analogs 1. METABOLICALLY STABLE ANALOG A novel PTH analog, SDZ PTS 893, which has the structure [Leu8, Asp10, Lys11, Ala16, Gln18, Thr33, Ala34]hPTH(1 – 34), is chemically more stable and has two- to fivefold higher binding affinity and efficacy — in stimulating both adenylyl cyclase and intracellular [Ca 2] transients — than PTH-(1 – 34) [56]. In addition, SDZ PTS 893 is highly anabolic at endocortical and periosteal sites. The intermittent administration of SDZ PTS 893 to OVX retired breeder rats

Subcutaneous and programmed infusion of hPTH for 1 h/day increases osteoblast number, wheras administration of the hormone for 6 h/day and continuous infusion resulted in histological changes in the proximal tibial metaphysis resembling abnormalities reported in patients with chronic hyperparathyroidisim. The photomicrographs are representative of rats infused with saline only (A), hPTH injected sc (B), hPTH infused for 1 h/day (C), hPTH infused for 2 h/day (D), hPTH infused for 6 h/day (E), and continuously infused hPTH (F). The open curved arrows point to trabecular perimeters bounded by osteoblasts. The closed curved arrows point to trabecular perimeters bounded by peritrabecular marrow fibrosis Reprinted with permission from Dobnig and Turner [45].

772

STOCH, CHOREV, AND ROSENBLATT

FIGURE 3

Effect of PTH fragments on distal femur bone density of OVX rats (A) and sham-operated rats (B). One week after operation, 6 month-old female rats were treated with either PTH-(1-38)(), PTH (3-38)(■), or vehicle (O) by SC injection for 4 weeks. A separate group of animals was treated with -estradiol by sc infusion(◆) (see A). Reproduced with permission of the American Society for Bone and Mineral Research [9].

results in increases in bone density and bone strength at the lumbar vertebrae, femoral diaphysis, distal femoral metaphysis, and femoral neck. Withdrawal of SDZ PTS 893 leads to decline in bone density and strength predominantly from the endosteal surfaces. Sequential treatment with alendronate is more effective than estradiol in maintaining bone strength following the withdrawal of SDZ PTS 893 [56]. 2. SIGNALING-SELECTIVE LIGANDS One of the seminal questions in PTH research focuses on the mechanism responsible for the anabolic effect of intermittent PTH administration vs the catabolic response seen after continuous administration of the hormone. Activation of PTH1-Rc elicits dual-signaling pathways, increasing both adenylyl cyclase/protein kinase A (PKA) via Gs and phospholipase C (PLC)/inositol 1,4,5,-triphosphate (IP3)-1,2-diacylglycerol (DAG)/cytosolic transients of [Ca2]i/protein kinase C (PKC) via Gq [57 – 64]. However, the correlation between the activation of these signaling pathways at the cellular level and the in vivo responses to PTH is not fully understood. Understanding the role of cellular processes, including receptor inactivation, internalization, trafficking, and recycling to bone metabolism is only beginning to be elucidated. Furthermore, the linkage of these signaling pathways to the anabolic effect of PTH remains to be established. Much attention is focused on identifying the signaling pathway responsible for the anabolic activity of PTH. This putative anabolic signaling pathway or a part thereof, may be shared by other agents exerting an anabolic effect on bone. Therefore, identifying a differential gene expres-

sion pattern characteristic and common to different bone anabolic agents is of great interest. These regulated genes may provide the clues for understanding the molecular mechanism of the anabolic activity associated with intermittent PTH treatment. Studies carried out in vitro on bone cell and organ cultures suggest that PTH residues 1 – 7 form the cAMP/PKA activation domain [65], while residues 28 – 34 comprise the PKC activation domain [66,67]. The latter sequence also encompasses the region responsible for the mitogenic activity of PTH on cultured osteoblast-like cells (residues 30 – 34) [68,69]. Several conflicting in vitro observations relating the intracellular signaling events to cell proliferation make it diffi cult to link these in vitro activities to the anabolic effect seen in vivo. Cyclic AMP appears to be instrumental in the role PTH plays in bone formation [70] and resorption [71]. PTH analogs that stimulate an increase in cAMP levels have been shown to either inhibit [72 – 74] or stimulate [74 – 76] osteoblastic cell proliferation. At the same time, in TE-85 human osteosarcoma cells, PTH-(1 – 34) stimulates proliferation but does not stimulate increases in intracellular cAMP [77]. Apparently, these effects depend on the particular species used, cell models employed, and the underlying experimental conditions. In general, N-terminally truncated fragments of PTH, which selectively activate PKC [65,78,79] are also mitogenic for osteoblastic cells [80]. Taken together, these observations suggest that the mitogenic N-truncated fragments of PTH which do not stimulate bone resorption [71] may be more effective “anabolic” analogs than peptides with an intact N-terminus. For example, truncation of the two first amino acids from the

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

N-terminus of PTH-(1 – 34) yields PTH-(3 – 34). This fragment displays reduced adenylyl cyclase activation, but maintains significant PKC activation and the mitogenic response in vitro [78]. Unfortunately, none of the N-terminal truncated PTH-derived peptides prove to be effective bone anabolic agents in vivo. Although PTH stimulation of bone resorption in vitro is mediated primarily through cAMP-dependent activation of PKA [81], it may not be the sole “second messenger” pathway involved in resorption [82,83]. One working hypothesis holds that dissociating the two signaling pathways of PTH, adenylyl cyclase and PLC, may separate the anabolic from the catabolic activities of PTH in bone [70,78,84 – 89]. However, recent data makes this hypothesis seem oversimplistic. The initial postulate was if the stimulation of bone resorption in vivo is related to the bone resorption response in vitro, then the in vivo response should be diminished in amino terminal-truncated PTH fragments. However, neither PTH-(3 – 34) nor PTH(3 – 38) (both PKC-selective, N-terminal-truncated analogs of PTH) are active in vivo as bone anabolic agents (Fig. 3) [52,66,70,89 – 92]. Furthermore, desamino-PTH-(1 – 34 ) — which has little ability to stimulate adenylyl cyclase but is equipotent to hPTH-(1 – 34) in stimulating PKC — does not stimulate cortical or trabecular bone growth in OVX rats [70]. Interestingly, in porcine renal epithelial HKRK B7 cells expressing hPTH1-Rc (950,000 Rc/cell), desamino[Gly1]- and desamino-[Ala1]hPTH-(1 – 34) are equipotent to hPTH-(1 – 34) in stimulating adenylyl cyclase but are devoid of PKC activity [93]. Currently, the in vivo effect of these analogs on bone metabolism is not known. Paradoxically, single subcutaneous injection of hPTH(1 – 38) to rats rapidly and transiently downregulates osteoprotegerin (OPG) mRNA levels. Osteoprotegerin is a negative regulator of osteoclast formation and function [94] (see Chapter 3). This effect is reproduced by hPTH(1 – 31), a cAMP/PKA-signaling selective agonist [67,70], but not by bPTH-(3 – 34) and bPTH-(7 – 34 ) — both PKCsignaling selective analogs [65,66,78]. This PTH-induced resorptive activity, reflected by a subtle and transient increase in osteoclast formation and activity, may represent a priming event necessary to prepare the bone for subsequent PTH-induced enhanced bone formation [95]. Surprisingly, hPTH-(1 – 31)NH2 (Ostabolin), an adenylyl cyclase-selective PTH agonist which is equipotent to PTH(1 – 34) in stimulating cAMP production in rat osteosarcoma (ROS) 17/2 cells [67,96], strongly stimulates cortical and trabecular bone growth in OVX rats [70,89,91,92,97]. In this analog the putative PKC-signaling motif Gln 2 8 –His32 is compromised by the elimination of His32 [67]. The short-term effects on biochemical parameters in human subjects following 8 h infusion (8 pmol/kg • h) of hPTH-(1 – 34) or hPTH-(1 – 31) demonstrate significant differences between these two peptides [98]. The 1 – 34 peptide induces hypercalcemia, lowers the level of immunore-

773 active PTH-(1 – 84), and increases urinary levels of type I collagen N-terminal telopeptide (NTx), whereas the 1 – 31 peptide does not induce acute hypercalcemia, does not affect plasma immunoreactive PTH-(1 – 84) levels, and does not affect urinary levels of NTx. Both peptides induce significant phosphaturic and natriuretic responses. Taken together, these observations suggest that PTH-(1 – 31) is more osteogenic than the longer fragments of PTH. Preferential induction of bone formation by PTH-(1 – 31) may be related to its selective stimulation of the cAMP/PKA signaling pathway, while the dual signaling (both the cAMP/PKA and the PKC signaling pathways) of PTH-(1 – 34) may be associated with the induction of both bone formation and bone resorption [98 – 100]. A second generation of adenylyl cyclase-selective analog, c[Glu22, Lys26, Leu27]hPTH(1-31)NH 2 —in which the helical nature of the C-terminus was enhanced by the formation of a side-chain to side-chain lactam ring and the introduction of a hydrophobic residue at position 27 — is a more potent stimulator of femoral trabecular bone growth than either hPTH-(1 – 34)NH2 [38] or its linear parent analog (1.2- to 2-fold) (Fig. 4) [101]. Recently, both hPTH(1 – 31)NH2 and c[Glu22, Lys26, Leu27]hPTH(1 – 31)NH2 were reported to prevent vertebral trabecular bone loss and increase vertebral-trabecular volume and thickness over that of control vehicle-injected sham-operated rats [97]. The action of these analogs on vertebral bone is as effective as that of hPTH-(1 – 34)NH2; however, unlike hPTH(1 – 34)NH2, their effects on pelvic BMD are equivocal. Similar modifications applied to hPTH-(1 – 28)NH2, the shortest PTH fragment that is still able to fully stimulate adenylyl cyclase (EC50  23.9 nM in ROS 17/2 cells), generates c[Glu22, Lys26, Leu27]hPTH(1 – 28)NH2, a more potent analog than the parent linear peptide (EC50  9.6 nM in ROS 17/2 cells) [100]. Intermittent administration of c[Glu22, Lys26, Leu27]hPTH(1 – 28)NH2 (5 nmol/day) to OVX rats significantly stimulates trabecular bone growth in the distal femur and L5 vertebra; this is not the case with the linear parent peptide [100]. It is evident that this shortest cAMP/PKA-selective PTH analog, which has significant in vivo osteogenic activity, offers better opportunity for the development of a nonparenteral anabolic drug in the treatment of osteoporosis. Recently, an alternative approach has been described to identify the structural determinants associated with selective signaling pathway activation. The replacement of Glu19 : Arg, a receptor affinity-enhancing modification, generates [Arg19]PTH-(1 – 28), a potent and full stimulator of adenylyl cyclase and PKC. In addition, substituting Gly for Ala1 generates a PKA-selective agonist, [Gly1, Arg19]hPTH-(1 – 28) (Fig. 5) [93]. The above study concludes that the extreme N-terminus of hPTH constitutes a critical activation domain for coupling to PLC. The C-terminal region, especially hPTH-(28 – 31), contributes to

774

STOCH, CHOREV, AND ROSENBLATT

FIGURE 4

Typical specimens of demineralized distal femurs showing the different abilities of hPTH-(1 – 34) and c[Glu22,Lys26,Leu27]hPTH-(1 – 31)NH2 to stimulate trabecular bone growth in ovariectomized 3-month-old rats. The experiments were carried in two modes: the preventive experiment (A), injections (once daily 6 days/week) started at the end of the second week after ovariectomy and continued for 6 weeks; or restorative experiment (B), injections (once daily 6 days/week) started at the end of the ninth week after ovariectomy and continued for 6 weeks. Reprinted with permission from Whitfield et al. [101].

FIGURE 5 Effects of substitutions at positions 1 and 19 on the properties of hPTH-(1 – 28) in HKRK B7 cells (subclones of LLC-PK1, porcine renal epithelial cells that stably express hPTH1-Rc). Intracellular cAMP accumulation (A), IP3 formation (B), and competitive radioligand binding (C) are depicted for [Ala1,Arg19]hPTH-(1 – 28) () and [Gly1,Arg19]hPTH-(1 – 28) () at the indicated concentrations. Results are presented as precentages of the maximal response to hPTH-(1 – 34) measured in the same experiment. Reprinted with permission from Takasu et al. [93].

775

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

PLC activation through receptor binding, but it is not required for full PLC activation. The N-terminal determinants for adenylyl cyclase and PLC activation in hPTH(1 – 34) overlap but are not identical; subtle modifications in this region may dissociate activation of these two effectors. Another approach to developing better bone anabolic agents was an attempt to generate bone-specific PTH analogs. To this end, [His3]- and [Leu3]hPTH-(1 – 34) were generated and found to be partial agonists of adenylyl cyclase (50 and 20%, respectively) in a kidney cell line, but full agonists in UMR-106 rat osteosarcoma cells [102]. This differential potency, which is attributed to subtle differences in the environment of PTH1 receptors in kidney and bone, has not been substantiated by experimental evidence. However, both analogs are less potent than PTH in vivo in the induction of bone formation. In the course of designing photoreactive PTHrP analogs for mapping the bimolecular ligand – receptor interface, we generated [Bpa1, Ile5, Arg11,13, Try36]PTHrP-(1 – 36)MH2 [103]. This analog binds and stimulates adenylyl cyclase equipotently to the parent analog [Ile5, Arg11,13, Try36]PTHrP(1 – 36)NH2 in HEK-293/C-21 cells which overexpress the human PTH1-Rc (400,000 receptors/cell); however, this analog does not stimulate an increase in intracellular calcium transients or the translocation of -arrestin2-green fluorescent protein (GFP), the fusion protein, from the cytoplasm to the membrane — an effect that is PKC-dependent [104]. In summary, the development of an effective and safe therapeutic agent that would stimulate the formation of new, mechanically competent bone and possibly reconstitute trabecular architecture in osteoporotic patients continues to be a worthy goal. This goal may be approached by analogs that interact with the PTH1-Rc in a signaling-selective manner. Recently, Jilka and coworkers proposed that Gs-mediated anti-apoptotic effects of intermittently administered PTH on osteoblasts may be one of the underlying mechanisms for the anabolic effect of PTH [105]. Since bone formation is largely determined by the number of osteoblasts [106], the anti-apoptotic effect of intermittent PTH administration on osteoblasts will contribute to the increase in their number. In the same study, Jilka and coworkers were also able to demonstrate an anabolic effect in osteoblastogenesis-impaired mice following intermittent treatment with PTH. The increase in the number of osteoblast progenitors obtained from bone marrow of these mice strongly suggests that the anabolic effect of PTH is due to an anti-apoptotic effect on osteoblasts and not due to stimulation of osteoblastogenesis [105]. However, Turner and coworkers, studying the effect of PTH and PTHrP on apoptosis of HEK 293 cells which stably express the hPTH1 receptor, report contradictory results [107]. They found that treatment of these cells with PTH induces apoptosis via a Gq-mediated PLC/Ca2 signaling pathway. The differential apoptotic effects induced

by PTH observed in these two studies may be attributed to the osteoblastic and non-osteoblastic cells employed.

OF

3. CONFORMATION-GUIDED DESIGN PTH AND PTHrP ANALOGS

According to a generally accepted working hypothesis, only a small fraction of the ensemble of dynamically equilibrating conformations presented by linear sequences of native bioactive peptides display the bioactive conformation recognized by receptors. Therefore, introduction of conformational constraints into a hormone analog may stabilize the ligand in receptor-favored bioactive conformations by precluding a wide range of nonproductive conformations. Another hypothesis holds that enhancement of the amphiphilicity of a helix will increase biopotency if this structural feature is important in stabilizing a favored secondary structure and/or providing a productive interaction with the receptor. Such modifications may favor conformations with the highest affinity for the receptor, resulting in enhancement of bioactivity. To this end, several studies incorporated structural modifications that stabilize conformational elements important for populating putative bioactive conformations [108 – 114]. Much attention has been drawn to the amphiphilic nature of the C-terminal helix comprising residues 20 – 34 of hPTH-(1 – 34) based on its role in receptor-binding [115,116]. A Lys27 to Leu substitution in PTH(1 – 34)NH2 and PTH-(1 – 31)NH2 improves the amphipathic character of the C-terminal helical sequence, resulting in about a five- and twofold increase in adenylyl cyclase-stimulating activity over the corresponding nonsubstituted sequences [112,113]. In addition, substitution in PTHrP-(1 – 34)NH2 of the sequence 22 – 31 with a model amphiphatic peptide (MAP; Glu1-Leu-Leu-Glu-Lys-LeuLeu-Glu-Lys-Leu-Lys10), which is highly -helical when incorporated into short peptides [117], generates [(MAP1 – 10)22 – 31]hPTHrP-(1 – 34)NH2 (RS-66271) (Fig. 6, see also color plate) [43]. Important structural features, such as Leu24 and 27, are maintained in this analog; Ile22 and 31 are substituted conservatively by Leu. In aqueous buffer, RS-66271 displays 8- to 9-fold higher helicity than the parent peptide. Detailed conformational analysis of RS66271 in water, employing CD and 1H NMR spectroscopy, confirms the presence of an extensive helical structure encompassing residues 16 – 32 [118]. The absence of a hinge element around Arg19, expected to contribute to high biological activity, may explain the 6- and 10-fold lower adenylyl cyclase and binding affinity, respectively, of RS66271 in ROS17/2.8 cells, as compared to the more flexible and less helical PTHrP-(1 – 34) [117, 119]. Importantly, the preservation of significant in vitro potency in spite of the multitude of substitutions supports the rationale behind the design of RS-66271.

776

STOCH, CHOREV, AND ROSENBLATT

FIGURE 6 Helical wheel representations of hPTHrP, RS-66271, and hPTH for residues 22 – 31, illustrating the amphiphilic nature of the helices. Hydrophobic amino acids are denoted in red and hydrophilic in blue. Reprinted with permission from Pellegrini et al. [118]. (See also color plate.) Despite lower binding affinity and efficacy in vitro, RS-66271 has significantly greater activity than either PTH-(1 – 34) or PTHrP-(1 – 34) in restoring lost trabecular and cortical bone in vivo in OVX osteopenic rats [43,119]. A more recent study confirmed the reduced in vitro activity of RS-66271 but demonstrated anabolic activity in vivo comparable to that of PTH-(1 – 34) [120].

C. Summary Understanding the molecular basis for the differential metabolic effects associated with intermittent versus continuous administration of PTH continues to grow. PTH analogs with enhanced anabolic activity are developed following the paradigm that recognizes signaling via Gs as associated with anabolic activity, in contrast to signaling via Gq, which is associated with catabolic activity. The screening of compound libraries for small nonpeptide PTH-mimetic agents holds promise of discovering orally bioavailable PTH-like agents. Last but not least, better understanding of the PTH – PTH receptor bimolecular interaction at the atomic level will provide the insights necessary for rational drug design leading to the development of nonpeptide drug candidates which will be small, safe, potent, orally available, and anabolic.

III. CALCIUM RECEPTOR ANTAGONISTS A. Background The recent discovery and cloning of the calcium receptor has not only elucidated the structure of an important macromolecule in calcium – mineral metabolism

homeostasis [121,122] but has provided an elegant new target and approach to treating disorders central to mineral metabolism. This important advance in understanding the mechanisms by which levels of plasma calcium regulate PTH secretion has enabled the discovery of a number of compounds. These drug candidates are now in clinical development directed at treating an array of disorders of calcium homeostasis [123 – 126]. These small molecules, also known as “calcilytics,” act on calcium receptors (located on the surface of the parathyroid gland cells) through which they stimulate transient increases in PTH secretion. Since these orally active calcilytics stimulate the intermittent release of endogenous PTH, they provide a potential novel approach to treating osteoporosis. Consequently, calcium receptor antagonists may further enrich our osteoporosis therapeutic armamentarium.

B. The Calcium Receptor The Ca2 receptor (CaR) was initially cloned from parathyroid cells, where it regulates PTH secretion [122]. The human parathyroid CaR cDNA encodes a 1078-aminoacid glycoprotein with striking sequence homology to the bovine parathyroid cell Ca2 receptor [127]. It is composed of a large amino-terminal extracellular domain with 11 potential glycosylation sites, an intramembranous portion containing seven membrane-spanning domains, and a large carboxy-terminal cytoplasmic domain [128]. This receptor has both structural and functional characteristics similar to other G-protein-coupled receptors [122,128]. The CaR has limited homology with metabotropic glutamate receptors (mGluR’s): its overall sequence homology is 25 – 30% with mGluR1 and mGluR5 [128]. The secondary structure of the CaR and mGluR receptors, however, are quite similar as

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

the number and positions of cysteines are conserved [128]. While these receptors exhibit structural homology, the CaR and mGluRs are not affected by the same compounds [128]. The extracellular domain is involved in binding calcium, but it lacks the “EF” hands that are characteristic of high-affinity Ca2 -binding proteins such as calmodulin and other intracellular calcium-binding proteins [128 – 130]. Presumably this is related to its function in sensing calcium present in the extracellular milieu in the millimolar range [128]. Calcium is thought to bind to regions of the receptor conserved across species that are enriched with acidic amino acids [128]. The finding of both activating and inactivating mutations highlights the importance of the extracellular domain in CaR activation [128]. The CaR is not very selective in binding compounds with net positive charges [128]. Furthermore, the CaR has been shown to bind a number of organic and inorganic polycations that inhibit PTH secretion by parathyroid cells in vitro [128,131,132]. Hence, these polyionic compounds may behave as receptor agonists; however, they are not perfect pharmacological probes. A number of compounds have been identified which act as positive allosteric modulators and heighten the sensitivity of the CaR to activation by extracellular calcium. These compounds are referred to as calcimimetics, compounds designed to selectively potentiate or mimic the action of extracellular calcium on the CaR. One such compound is NPS R-568, which currently is in clinical trials to treat conditions such as primary and secondary hyperparathyroidism [124,133]. By contrast, compounds specifically designed to block the actions of extracellular calcium at the CaR are termed calcilytics [128].

C. In Vitro and in Vivo Studies PTH is known to exhibit both catabolic and anabolic effects on bone [51]. The difference in PTH effect on bone depends on the regimen of PTH administration and the dose. Chronic sustained PTH elevation, as seen with severe hyperparathyroidism, is associated with net bone loss [128], but intermittent increases in PTH at low levels results in new bone formation. As discussed above, this anabolic effect may be useful in treating osteoporosis. It has been demonstrated in several animal models and in clinical trials and underscores the interest in developing PTH peptide analogs for treating osteoporosis [128] as well as stimulators of secretion of endogenous PTH. The anabolic effect characteristic of PTH peptide intermittent administration may be achieved through the administration of calcilytics which transiently antagonize the CaR and mimic a decline in extracellular calcium levels — leading parathyroid gland cells to secrete PTH. Prior to the development of calcilytics, the generation of an endoge-

777 nous PTH response akin to inhibiting the CaR was achieved by inducing hypocalcemia; this resulted in transient increases in PTH concentrations similar to those seen with exogenous administration of rat PTH-(1 – 84) [134] or rat PTH-(1 – 34) [135]. A 5 g/kg subcutaneous dose of rat PTH-(1 – 34) was found to parallel the endogenous hormone levels achieved through hypocalcemia [135]. This dose administered daily for 28 days in the ovariectomized rat results in significant increases in trabecular bone content and thickness as well as in indices of bone formation [135]. Hence, amounts of PTH sufficient to stimulate new bone formation can be secreted endogenously in response to requisite stimulation in osteopenic rats. This provides an exciting rationale for the development of calcilytic agents that may be utilized in the treatment of osteoporosis. Promising preliminary data has been obtained showing the stimulation of PTH secretion by calcilytics both in vitro and in vivo [126,128]. NPS 2143 (Fig. 7A), a small molecule with MW 409, antagonizes calcium at the CaR (IC50  40 M) and displaces the binding of [3H]NPS 1377 — an allosteric activator of the CaR which binds within the transmembrane domain of the receptor [128]. NPS 2143 blocks increases in cytosolic calcium levels elicited by extracellular Ca2 in bovine parathyroid cells and stimulates PTH secretion. At doses of 1 mol NPS 2143 does not alter increases in cytosolic calcium levels elicited by bradykinin, thrombin, or ATP in HEK 293 cells. Furthermore, NPS 2143 stimulates PTH secretion with comparable magnitude to PTH secretion stimulated by low (0.5 mM) extracellular Ca2 concentration. The intravenous administration of NPS 2143 to normal rats elicits a rapid, dose-dependent increase in plasma PTH concentration. A single IV bolus dose of 3 mol/kg NPS 2143 results in a 10-fold increase in PTH within 1 min of injection; values return to baseline within 60 min [128]. The oral administration of NPS 2143 to normal rats results in a seven-fold increase in circulating PTH, which reaches a maximum within 15 – 30 min following administration (ED50  90  26 mol/kg) [126], levels remained elevated for at least 4 h (Fig. 7B) [136]. The effect of repeated stimulation of PTH secretion on bone was assessed by administering NPS 2143 (100 mol/kg) by daily gavage over 8 weeks in 10-month old rats that were ovariectomized 3 months earlier. Positive controls were provided by rats injected daily with rat PTH-(1 – 34) (5 g/kg, sc), which achieves a comparable rise in plasma PTH (Fig. 7B). Bone mineral density (BMD) at the distal femur was lower in control OVX rats than in sham-operated animals, and did not increase in OVX rats given NPS 2143; however, BMD increased significantly in those treated with PTH [126,136]. A similar BMD pattern was seen at the proximal tibia (Fig. 8A) [136]. Histomorphometric analysis of cancellous bone in the proximal tibial metaphysis revealed similar increases in indices of bone formation: in

778

STOCH, CHOREV, AND ROSENBLATT

FIGURE 7 Structure of NPS 2143 and its effects on plasma PTH levels. (A) Structure of NPS 2143, a selective antagonist (“calcilytic”) of the parathyroid cell Ca2 receptor. Reprinted with permission from Nemeth et al. [125]. (B) PTH levels were elevated (100 pg/ml) within 30 min after dosing with NPS 2143; levels remained high for at least 4 hours. Plasma levels of PTH after the administration of NPS 2143 (100 mol/kg, po) or rat PTH-(1 – 34) (5 g/kg, sc) to overiectomized (OVX) rats. Both endogeneous and exogeneous PTH levels are of the 1 – 34 peptide. Values are mean  SE; n  3 – 4 per time point. Reprinted with permission from Gowen et al. [136].

OVX rats treated with either PTH or NPS 2143, the NPStreated rats did not show the increase in trabecular thickness or area seen in rats treated with PTH. The differences in duration of PTH response may be explained by the sus-

tained increase (up to 8 h) in NPS 2143 levels following oral administration. A second study was performed in which NPS 2143 was administered daily for 5 weeks in the presence and absence

FIGURE 8 Effects of NPS 2143 om bone mineral density in OVX rats. (A) Bone mineral density (BMD) in the proximal tibia unaffected by daily NPS 2143, but returned to preovariectomy levels after 8 weeks of treatment with PTH. Effect of daily administration of NPS 2143 (100 g/kg, po) or rat PTH(1 – 34) (5 g/kg, sc) on BMD in the proximal tibia of osteopenic ovariectomized rats. Values shown as mean  SE; n  10 – 14 per group. ASignificantly different from sham-vehicle; P 0.05. BSignificantly different from OVX vehicle; P 0.05. CSignificantly different from OVX-NPS 2143; P 0.05. (B) The coadministration of NPS 2143 and 17-estradiol significantly increases BMD at the distal femur in osteopenic ovariectomized rats. Effect of daily treatment of osteopenic ovariectomized rats for 5 weeks with 17-estradiol (continuous sc infusion), NPS 2143 (100 g/kg, po) or NPS 2143 and 17-estradiol on bone mineral density (BMD) in the distal femur. Values are mean  SE; n  9 – 10 per group. BSignficantly different from OVX-vehicle, P 0.05. CSignificantly different from OVX-NPS 2143; P 0.05. Reprinted with permission from Gowen et al. [136].

779

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

of estrogen to assess the effect of this combination on bone mass and bone turnover [136]. The coadministration of NPS 2143 and 17-estradiol alters histomorphometric indices of bone formation and resorption, as evidenced by the elevation in bone formation, resulting in an increase in cancellous bone content [136]. BMD at the distal femur increased 2.9  1.5% in rats given 17-estradiol alone vs 7.2  1.4% in rats given NPS 2143 and 17-estradiol (P 0.02) (Fig. 8B) [136]. It is noteworthy that despite sustained parathyroid gland stimulation, the parathyroid cells did not exhibit hyperplasia in the above animal models [136]. These data support the principle that stimulation of endogenous PTH secretion by antagonizing the parathyroid cell CaR increases bone formation and resorption. Furthermore, the addition of an anti-resorptive agent, such as estrogen, to NPS 2143 leads to an increase in bone mass. While “proof of concept” has been demonstrated in rats, further validation in both nonhuman primate models and thereafter in human subjects is still required. Since calcilytics are in the preclinical development phase, no data is available on their effects in humans. Based on the promising data obtained for calcimimetics in hyperparathyroidism [133], there is reason to be optimistic about the potential value of calcilytics in the treatment of osteoporosis.

D. Potential Development Issues The efficacy of potent and selective orally active CaR antagonists ultimately will rest on eliciting the correct kinetics of endogenous PTH secretion. Both the level and the duration of PTH release are critically important. Clearly, sustained high levels of PTH will enhance osteoclast-mediated bone resorption, while PTH levels which are inadequate or of insufficient duration will be completely ineffective.

IV. STIMULATION OF BONE FORMATION AND INHIBITION OF RESORPTION BY STATINS A. Background A recent report by Mundy and colleagues demonstrates that 3-hydroxy-3-methylglutaryl coenzyme A (HMG Co-A) reductase inhibitors or “statins,” drugs specifically designed to lower serum cholesterol, are able to enhance new bone formation in vitro as well as in vivo in rats and mice [137]. The screen was based on the hypothesis that agents capable of stimulating the expression of bone morphogenic protein (BMP)-2, a member of the BMP family which stimulates osteoblast differentiation [138] would also be effective as

bone anabolic agents. This was achieved by the screening of large libraries of natural products for the stimulation of BMP-2 formation using murine osteoblasts transfected with a luiferase reporter gene driven by the mouse BMP-2 promoter. This screen identified lovastatin from among 30,000 compounds as the only compound capable of specifically increasing luciferase activity. The inhibition of the cholesterol-biosynthetic pathway using HMG Co-A reductase inhibitors, agents specifically designed to treat dyslipidemias, also inhibits osteoclast resorption, as this pathway is also necessary for osteoclast function [139]. Other members of the “statin” class of drugs, including, simvastatin, mevastatin, and fluvastatin are also capable of generating maximal stimulation of BMP-2 promoter-driven luciferase reporter at doses of 5 M. This inhibition can be rescued using mevalonate — a metabolite located downstream to HMG Co-A reductase in the cholesterol pathway. Each of these statins increases new bone formation by twoto threefold in neonatal murine calvarial bones in organ culture. This effect is comparable to the observed effect using other stimulators of osteoblast proliferation, including BMP2 and fibroblast growth factor-1 (FGF-1) [140]. The in vivo administration of simvastatin into subcutaneous tissue overlying the murine calvarium results in a 50% increase in new bone formation as early as 5 days posttreatment (Fig. 9) [137]. Similar effects were observed following treatment with FGF-1 and BMP-2. The oral administration of either simvastatin or lovastatin (5 – 50 mg/kg/day) to OVX rats results in an increase in trabecular bone volume (39 – 94%) and bone formation rates and a decrease in osteoclast number — thus clearly demonstrating strong anabolic and inhibition of resorption effects on bone.

B. Human Data Recent exciting retrospective data involving older women taking statins for dyslipidemia suggests that this therapy is accompanied by higher levels of hip BMD and a lower risk of hip fractures [141]. This effect was not seen in age-matched subjects taking lipid-lowering medication of the nonstatin variety. Chung and colleagues suggest that the effect of HMG-CoA reductase inhibitors on BMD is even more profound in men with type 2 diabetes mellitus than in women with the same disorder [142]. Since statins effectively increase osteoblastic bone formation both in vitro and in vivo [137], they may be more effective in treating male osteoporosis (in which reduced osteoblastic function appears to be the principal mechanism for bone loss) than in treating the increased osteoclastic function seen in postmenopausal women [142]. Several recent retrospective studies report that treatment with statins reduces fracture risk [143 – 145]. Chan and colleagues report

780

STOCH, CHOREV, AND ROSENBLATT

cells (Fig. 10) [139] (see Chapter 16). Osteoclastogenesis in culture is inhibited both by alendronate, a nitrogen-containing bisphosphonate, and lovastatin [146]. This inhibition of osteoclast formation and function by alendronate can be reversed using exogenous geranylgeraniol, thereby enabling conversion of mevalonate to geranylgeranyl diphosphate (GGPP), an essential intermediate in the cholesterol biosynthesis pathway (Fig. 10). Bergstrom and colleagues have recently reported that farnesyl diphosphate synthase is the selective target for N-containing bisphosphonates (e.g., IC50  460 nM for alendronate) [147]. The inhibition of the farnesyl diphosphate synthase reduces the cellular levels of GGPP, which is necessary for the prenylation of GTP-binding proteins that control cytoskeletal reorganization, vesicular fusion, and apoptosis. These aformentioned processes are required for osteoclast activation and survival. Braga reported that the treatment of postmenopausal women with aminobisphosphonates also lowers the serum concentration of LDL cholesterol and increases those of HDL cholesterol [148]. Reszka and coworkers suggest that N-containing bisphosphonates act directly on osteoclasts to block geranylgeranylation and thus the proper intracellular targeting of G proteins. This in turn triggers caspase-induced cleavage of mammalian sterile 20-like (Mst) 1 kinase, thereby yielding a highly active 34-kDa species which is capable of activating caspase and generating an apoptosis autoactivation loop which leads to osteoclast apoptosis [149]. Hence, one may speculate that the statins, which also block the mevalonate pathway upstream to GGPP formation, will also lead to osteoclast apoptosis and consequently to the inhibition of bone resorption (Fig. 10). FIGURE 9

Quantitative effects of daily injections of simvastatin on the width (top) and total area (bottom) of calvaria bones. Simvastatin was injected sc over the murine calvaria. Reprinted with permission from Mundy et al. [137].

that women age 60 years or older who regularly used statins over a 2-year period had a greater than 50% reduction in fracture risk [143]. Meier et al. have also seen a reduction in fracture risk in individuals age 50 years or older who have been treated with statins [144]. Furthermore, Wang has reported that patients, age 65 years or older, treated with statins have a reduced risk of hip fractures [145]. The above data is exciting, but will need to be confirmed in prospective randomized clinical trials. It is of much interest that nitrogen-containing bisphosphonates, an established class of anti-resorptives used to treat osteoporosis, can induce macrophage and osteoclast apoptosis. This is accomplished by inhibiting enzymatic steps in the mevalonate pathway, and interupting the posttranslational prenylation of proteins, including Ras in these

C. Potential Development Issues The doses of statins given to OVX rats in Mundy’s study are on the order of 10-fold higher than those routinely used by patients to lower serum cholesterol. The development of “bone-specific” statins may allow a further reduction in dose and the generation of better anti-resorptive drugs. Distinct targeting of statins to the bone may provide the specificity needed for a safe bone agent devoid of any significant liver toxicity.

V. INHIBITION OF BONE RESORPTION WITH v3 INTEGRIN ANTAGONISTS A. Introduction Integrins comprise a family of cell surface receptors, constructed from transmembrane heterodimeric glycoproteins,

781

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 10 Schematic representation of intracellular effects of bisphosphonates in osteoclasts. Reprinted with permission from Reszka et al. [149]. GGOH, geranylegeraniol; FPP, farnesyl diphosphate; Z-VAD, Z-Val-Ala-Asp (carpase inhibitor). which have a pivotal role in numerous developmental, physiological, and pathological processes. They are the principal mediators of cell-to-extracellular matrix anchorage — which is of fundamental importance to cell function and tissue integrity. Some integrins mediate cell-to-cell contact and almost all appear to be involved in signal transduction. Functional cellular adhesion also provides cues for migration as well as signals for growth and differentiation. In a myriad of roles, integrins promote platelet aggregation, bone resorption, immune function, cell fusion, tumor invasion and metastases, programmed cell death, leukocyte homing and activation, and the response of cells to biomechanical forces [150 – 155]. The potential therapeutic utility of interfering with integrin-mediated events in osteoporosis and other metabolic bone diseases is both exciting and promising.

B. Integrin Chemistry Integrins are composed of noncovalently associated  and  subunits [156]. Both subunits are characterized by a large N-terminal extracellular component, a transmembrane domain, and a short C-terminal tail. To date, 17 different integrin  subunits and eight different  subunits have been iden-

tified. Together they combine to form at least two dozen different, naturally occurring integrin receptors [157]. Integrins are expressed on more than one cell type, and most cells express several integrins. Furthermore, individual integrins can often bind more than one ligand, and a given ligand is often recognized by more than one integrin. Most integrins bind ligands which contain an Arg-Gly-Asp (RGD) sequence [155,158,159]. This three-amino acid motif is found in extracellular matrix proteins including fibronectin, laminin, vitronectin, fibrinogen, von Willebrand’s factor, and osteopontin. Short synthetic peptides containing “RGD” can mimic the biological activity of intact native ligands and competitively inhibit cell adhesion to native ligands present in the extracellular matrix [159]. The essential role of the RGD motif in cell attachment was demonstrated by sitedirected mutagenesis of the RGD triad in fibronectin and vitronectin to RGE (Arg-Gly-Glu), which results in the complete loss of cell attachment to the mutated matrix proteins [156,160]. Throughout life, bone undergoes continuous turnover or “remodeling,” a dynamic process in which some bone is being resorbed while new bone is being formed. Osteoporosis is characterized by a failure to precisely match these two processes: the normal equilibrium of bone remodeling is lost and resorption exceeds formation.

782

STOCH, CHOREV, AND ROSENBLATT

FIGURE 11

Representative structures of v3 antagonists.

783

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

Osteoclasts begin the process of bone resorption when they migrate and attach to mineralized bone matrix. This attachment results in cellular polarization forming three discrete plasma membrane domains: the basolateral, the sealing zone (also called the clear zone), and the ruffled border, which is in close contact with the bone matrix sequestered underneath [61]. The tightly sealed compartment functions as an “extracellular lysosome” into which acid and acid proteases are secreted by the osteoclast. Digestion of the matrix proteins and the solubilization of the mineralized matrix lead to the formation of the resorptive lacunae [161] (see also Chapter 3). Osteoclast adhesion involves several integrins, including v3, 21, and v1 [162]. The highest level of physiological expression of v3 is on the osteoclast with approximately 15  106 receptors/osteoclast [162]. Hence, development of v3 antagonists represents a sound mechanism-based antiresorptive approach to therapy.

C. Design of Integrin Antagonists The importance of the RGD motif to osteoclast-mediated bone resorption has been demonstrated by several means: short synthetic RGD-containing peptides, disintegrins, monoclonal antibodies, and nonpeptide RGD mimetics. All have been shown to inhibit bone resorption in vitro by isolated osteoclasts [155, 163 – 167]. The “disintegrins” are a group of relatively small RGD-containing proteins which include echistatin (1 in Fig. 11), a 49-amino-acid peptide which was isolated from snake venom [168]. Echistatin has been found

to be a highly potent inhibitor of bone resorption both in vitro [168 – 170] and in vivo [165]. Recently, 4 weeks of in vivo echistatin administration (0.26 g/kg/h) to OVX rats was shown to prevent bone loss in the femur with no evidence of side effects, such as bleeding due to inhibition of platelet aggregation through the IIb3 integrin [166]. As further “proof of concept” of the fundamental role that the vitronectin receptor plays in osteoclastic bone resorption, echistatin has been tested in several animal models. Fisher reported effective dose-dependent reversal of PTH-induced hypercalcemia in thyroparathyroidectomized (TPTX) rats (IC50  100 nM) [157]. Three-day treatment with echistatin (30 g/kg/min) prevented bone resorption in mice with low calcium diet-induced secondary hyperparathyroidism (Fig. 12, see also color plate) [167]. Immunochemistry has shown colocalization of echistatin with the v-like subunit at the osteoclast clear zone, suggesting that echistatin blocks bone resorption by interacting with v3 [167]. Unfortunately, echistatin has several drawbacks, including lack of selectivity and its requirement for parenteral administration. In addition, it is only 300-fold more potent as an anti-resorptive agent than as an anti-platelet aggregation agent [168]. The lack of selectivity is attributed to the common -subunit shared by both v3 and IIb3. Nevertheless, studies with disintegrins indicate that integrin selectivity is effected by the amino acids flanking the RGD triad and can be modified by their substitution [171]. 1. PEPTIDES Initial efforts to develop v3 receptor-selective antagonists focused on RGD-containing peptides. Kessler and col-

FIGURE 12 The integrin ligand echistatin prevents bone loss in ovariectomized mice and rats. Photomicrographs of representative distal femora from sham-operated (Sham-op) and ovariectomized (OVX) mice treated with vehicle and OVX mice treated with echistatin (OVX  Echi) for 4 weeks. Longitudinal sections were stained with Masson’s trichrome. The bone trabeculae are stained blue, and the marrow stroma is stained red. Scale bar, 1 mm. Reprinted with permission from Yamamoto et al. [166]. (See also color plate.)

784

STOCH, CHOREV, AND ROSENBLATT

leagues were first to demonstrate that a particular conformation of RGD is essential for v3-binding and contributes to selectivity for v3 versus other integrins [172,173]. This led to the development of cyclo penta- and hexapeptide antagonists of v3 [172,174,175], which were later employed as lead structures in the development of nonpeptide RGD mimetics. Conformational constraints in the form of cyclic pentapeptide 2 [176] and end group-to-end group tethered tetrapeptides 3 [177] yield compounds with 1400- and 900fold greater selectivity for v3 than the IIb3 receptor, respectively (see Fig. 13, see also color plate). The specific conformation of the cyclic peptide around the Gly in the RGD motif dictates integrin receptor selectivity. A turn-extended-turn conformation which extends the Arg side-chain away from the Asp side-chain confers preferential IIb3 antagonism. A type I -turn around Gly, which shortens the distance between the guanidine and carboxyl groups, confers preferential v3 antagonism (Fig. 13) [172,177,178]. 2. BENZODIAZEPINES The search for potent selective v3 antagonists later was extended to include small molecule RGD mimetics with oral bioavailability. Using a benzodiazepine scaffold, researchers at SmithKline Beecham have developed a benzimidazole-containing compound, SB-223245 (4 in Fig. 11), which is highly potent (IC50  2.0 nM) and selective

FIGURE 13

( 10,000-fold selectivity for v3 vs IIb3) in vitro [179]. However, in rodents, SB-223245 has poor bioavailability (3 – 7%) and a short half-life ( 9 – 16 min). A dramatic improvement in the pharmacokinetic profile, with almost 100% oral bioavailability and a longer half-life (t1  192 min), was achieved with SB-265123 (5 in Fig. 11). Unfortunately, this also led to a reduction in potency (IC50  4.0 nM). This modification was accomplished by eliminating the amide bonds, introducing a fused phenyl ring, and replacing the benzimidazole (guanidino mimetic moiety) at the N-terminus with a pyridylamino function [179]. The oral administration of SB-265123 (30 mg/kg b.i.d. for 6 weeks) effectively prevents bone loss in OVX rats [180]. Furthermore, enhancing lipophilicity, as in SB-267268 (6 in Fig. 11), results in higher affinity for the v3 receptor and is more potent in inhibiting cell adhesion activity (K  0.9 nM, IC50  12 nM, respectively). Moreover, compound 6 has a better pharmacokinetic profile (34% oral bioavailability, t12  360 min) [179]. The oral administration of this compound to OVX rats (15 and 60 mg/kg b.i.d.) for 17 weeks results in a statistically significant increase in BMD at the lumbar spine [181]. 3. PHENYLSULFONAMIDES In compounds of the -phenylsulfonamide series, targeted selectivity for v3-mediated versus IIb3-mediated

Overlaid minimum energy conformation of cyclo[RGDRGD], generated by quenched molecular dynamics (QMD). Reprinted with permission from Burgess et al. [204]. (See also color plate.)

785

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

effects can be accomplished by the particular choice of pKa, geometry, and the hydrogen-bonding characteristics of the guanidine-mimetic moieties. Researchers at Merck have demonstrated that, by replacing the disubstituted guanidine moiety at the N-terminus of L-748,415 (7, in Fig. 11) with the moderately basic 5,6,7,8-tetrahydro-[1,8]naphthyridine, they obtain L-767,635 (8, in Fig. 11), which displays enhanced potency and selectivity toward the v3 receptor as opposed to the IIb3 receptor (IC50 IIb3/IC50 v3  55 and 1600 for compounds 7 and 8, respectively) [182, 183]. In vivo, L-767,635 is equipotent to echistatin. In addition, it is threefold more potent than L-748,415 in inhibiting PTH-induced hypercalcemia in TPTX rats (IC50  0.6 M versus 0.2 M, respectively) [183]. Recently, researchers at Searle have developed several glycine-centered RGD mimetics as selective v3 antagonists [184]. SC56631 (9 in Fig. 11), which lacks adequate selectivity between v3 and IIb3, prevents 55% of trabecular bone loss in the proximal tibia of OVX rats following 6 weeks of intravenous administration (0.5 mg/kg/min). Effective reduction of elevated urinary pyridinoline metabolite levels in this animal model was observed even at a lower dose of 9 (0.1 mg/kg/min). Replacement of the pyridin-3-yl in 9 with a more lipophilic 3,5-dichlorophenyl moiety (10 in Fig. 11) results in 2 orders of magnitude improved v3 (IC50  1.1 nM) versus IIb3 (IC50  152 nM) selectivity [185]. Researchers at Hoechst Marion Roussel designed hydantoin-based RGD mimetics, which are potent and selective v3 antagonists [186]. Incorporation of cyclic guanidine moieties at the N-terminus, as in compounds 11 and 12 (Fig. 11), results in enhanced v3 selectivity compared to similar compounds containing noncyclic guanidine moieties. Compounds 11 and 12 have 500- and 200-fold higher affinity for the v3 receptor (IC50  20 and 40 nM, respectively) than the IIb3 receptor (IC50 10 M).

D. Potential Development Issues Remarkable progress has been made in the design of highly potent, orally active v3-selective antagonists. However, it remains to be seen whether chronic treatment with these agents, as is required in the treatment and prevention of osteoporosis, will have an acceptable safety profile. Given the wide distribution of integrins and their involvement in numerous physiological processes, possible side effects may include bleeding and compromised wound-healing. Since the osteoclast expresses more than one integrin receptor, it is possible that blocking the vitronectin receptor alone may not be sufficient to avert bone loss. Last, cross-reactivity may occur with other closely related 3-containing integrins, including IIb3, with the potential to cause side effects.

VI. OSTEOPROTEGERIN A. Introduction The coupling of bone resorption and formation is necessary in the removal of old bone and the synthesis of new bone. This remodeling cycle is a fundamental process in bone physiology. The recent discovery of osteoprotegerin and its ligand (OPG-L or RANKL) enhances our understanding of bone physiology. Furthermore, these naturally occurring substances have the potential to further enrich our osteoporosis therapeutic armamentarium. We will now discuss the skeletal effects of OPG and OPG-L/RANKL, and their potential use as agents for the treatment of osteoporosis.

B. Chemistry and Actions of OPG and OPG-L/RANKL Osteoprotegerin (OPG), also termed osteoclastogenesis inhibitory factor (OCIF), was identified independently by two groups and is intensively discussed in Chapters 12 and 13 [187,188]. It is a novel secreted disulfide-linked dimeric glycoprotein which has been shown to regulate bone resorption [187]. It was initially identified following the screening of a fetal rat intestine cDNA library and found to be a member of the tumor necrosis factor receptor (TNF-R) superfamily [187]. The cDNA encodes a 401-amino-acid polypeptide with features of a secreted glycoprotein, including a hydrophobic leader sequence and four potential sites for Nglycosylation. After cleavage of the signal peptide (21 amino acids), a mature peptide (380 amino acids) is generated [187,188]. OPG is synthesized as a 55-kDa monomer within the cell and is thereafter converted to a disulfidelinked dimer which is the predominant extracellular form of approximately 110 kDa [187]. While these two forms have similar physicochemical properties, the homodimer has higher heparin-binding capacity and greater hypocalcemic potency [189,190]. OPG is a glycoprotein with seven structural domains, including four TNF-R motifs at the N-terminus (domains 1 – 4), two “death domain” homologous regions (domains 5 and 6) which mediate cytotoxicity, and a C-terminus (domain 7) which contains a cysteine residue for homodimer formation [191]. OPG has four random cysteine-rich TNF-R motifs at the N-terminus and its C-terminus is unrelated to any known protein. It appears that all four cysteine-rich motifs are required for biological activity of OPG/OCIF [187,192]. It lacks a transmembrane domain, suggesting that it acts as a secreted cytokine receptor rather than a membrane-anchored receptor characteristic of other members of the TNF-R superframily [191]. OPG/OCIF mRNA has wide tissue distribution, including lung, heart, kidney, liver, stomach, intestine, skin, brain and

786

STOCH, CHOREV, AND ROSENBLATT

FIGURE 14

A model illustrating a mechanism whereby osteoblasts/stromal cells regulate osteoclastogenesis. Three distinct signals stimulated by 1,25(OH)2D3, PGE2/PTH, and IL-11 induce OPGL/ODF/RANKL expression on osteoblasts/stromal cells. OPG-L/ODF/RANKL mediates a signal for osteoclastogenesis through the ODF/RANK receptor expressed on osteoclast progenitors. OPG/OCIF inhibits osteoclastogenesis by acting as a decoy, interrupting the binding of OPGL/ODF/RANKL and ODF/RANK receptor. Reprinted with permission from Yasuda et al. [206].

spinal cord, thyroid, and bone [187,188,193,194]. OPG/OCIF mRNA and protein levels are increased by the cytokines IL1, IL-1, TNF-, TNF- (Fig. 14) [191,195,196], by the osteoblastic differentiating factors, 1,25-(OH)2D3 and BMP 2 [191], and by the anti-resorptive agents, estrogen and TGF- [197 – 199]. By contrast, glucocorticoids, PGE2 and the pure antiestrogen, ICI 182,780, decrease OPG/OCIF mRNA and protein levels [197,200,201]. Osteoprotegerin was so named because it protects bone by suppressing osteoclastogenesis and inhibiting bone resorption. It acts as a soluble receptor for OPG-L/ODF/RANKL and prevents it from binding to and activating the ODFR/RANK on the osteoclast cell surface (Fig. 14) [202]. OPG/OCIF inhibits osteoclast differentiation [187 – 189], suppresses mature osteoclast activation [203,204], and induces osteoclast apoptosis [205]. The overexpression of OPG cDNA in transgenic mice results in a phenotype which appears to be a non-lethal form of osteopetrosis [187,188]. These transgenic mice do not differ in external appearance, body weight or behavior from their normal littermates, but they do exhibit a generalized increase in radiodensity in long bones, vertebrae and the pelvis (Fig. 15, see also color plate) [187]. Furthermore, the severity of the osteopetrosis correlates with the level of OPG mRNA expression [187]. This increase in bone mass signifies therapeutic potential for the use of OPG as a drug, although it may be less effective in the clinical setting. The cognate ligand for OPG (OPG-L) or ODF (osteoclast differentiation factor) or RANKL was also identified independently by two groups [203,206]. Human OPG-L/ODF/RANKL

consists of 317 amino acids and exists as both a cellular, membrane-bound form (40–45 kDa) and a soluble form (31 kDa), both of which are bound by OPG [203]. The molecular sequence of OPG-L/ODF/RANKL has been found to be identical to two previously reported novel members of the TNF-R superfamily: tumor necrosis factor (TNF)-related activation-induced cytokine (TRANCE) [207] and receptor activator of NFKB ligand (RANKL) [208]. Steady-state levels of OPGL/ODF/RANKL mRNA are high in both skeletal (trabecular bone, bone marrow) and lymphoid tissue and low in other organ tissues [203,206]. OPG-L/ODF/RANKL mRNA levels are upregulated by dexamethasone, 1,25-(OH)2D3, IL-1, IL-11, TNF-, PTH, and PGE2 [206,209] and suppressed by TGF- [199]. Both forms of OPG-L/RANKL stimulate osteoclastogenesis and enhance the activity of mature osteoclasts; this strong activation is inhibited by the addition of OPG (Fig. 14) [203]. In addition, OPG-L/RANKL inhibits osteoclast apoptosis [210]. OPG-L /RANKL directly mediates osteoclastogenesis by binding to the RANK receptor found on osteoclast precursor cells [211]. The direct binding of OPG-L/RANKL to mature osteoclasts rapidly (within 30 min) induces actin-ring formation—a cytoskeletal rearrangement required for bone resorption [204]. While OPG-L/RANKL has been shown to stimulate bone resorption both in vitro [203] and in vivo [203,206] it does not lead to osteoclast formation in tissues other than bone [203]. In vitro administration of OPG-L/ODF/RANKL increases various osteoclastic differentiation markers including TRAP (tartrate-resisitant acid phosphatase) and the expression of  integrin, cathepsin K, c-src, and the vitronectin receptor [203,206,212,213].

787

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 15

Increased bone density in OPG transgenic mice. Two 10-week founder mice expressing relatively high (panel 17, bottom row) and low (middle row) levels of OPG transgene mRNA, and a control littermate mouse (panel 18, top row) were subjected to X-ray (left column) or histologic (middle and right columns) analysis of their long bones, pelvis, and vertebrae. Increased bone density is seen in radiographs of transgenics (panels 17 and 11), as are visible increases in bone by histological analysis. The lower power magnification (right column, bar  1 mm) shows the distal femur and growth plate, and the higher power magnification (middle column, bar  100 m) shows the femoral midshaft. In these sections, bone stains deep blue or red, cartilage stains light blue, and marrow stains dark purple. Reprinted with permission from Simonet et al. [187]. (See also color plate.)

The injection of recombinant OPG-L/RANKL into normal mice results in an increase in both size and nuclearity of osteoclasts, profound bone loss, and rapid-onset of hypercalcemia [203]. OPG, by contrast, blocks OPGL/RANKL’s effects on actin ring formation and bone resorption [204]. These effects suggest that OPG-L/RANKL is the osteoclast differentiation and activation factor, whereas OPG is the naturally occurring soluble receptor that counterbalances the biological effects of OPGL/RANKL, serving to preserve bone mass by preventing OPG-L/ODF/RANKL from binding to the RANK receptor found on osteoclast precursors (Fig. 14) [191]. Further support for the protective role of OPG is provided by the finding of osteoporosis, due to increased osteoclastogenesis, in the OPG knockout mouse [214]. In addition, OPGdeficient mice also exhibit medial calcification of the aorta and renal arteries, suggesting that OPG may play a role in vascular calcification [215]. The OPG-L/RANKL knockout mouse, by contrast, has been shown to develop severe osteopetrosis and completely

lack mature osteoclasts as a result of the osteoblast’s inability to support osteoclastogenesis [216]. RANK, the receptor for OPG-L/ODF/RANKL, was originally described as a receptor on both T-cells and dendritic cells [208]. It is sometimes referred to as ODAR, osteoclast differentiation and activation receptor, as it is identical to the receptor located on osteoclasts that binds OPG-L/RANKL with high specificity and affinity [202,211]. It is highly expressed on osteoclasts as well as hematopoeitic osteoclast progenitors, but is not found on other bone cells [211]. “Stimulating” antibodies directed at the extracellular domains of ODAR/RANK promote osteoclastogenesis by mimicking OPG-L/ODF/RANKL action. Inhibitory Fab fragments of this antibody block the process by competing with OPG-L/ODF/RANKL for the cell-bound receptor [211,217]. OPG functions as a decoy receptor by binding OPG-L/ODF/RANKL, thereby blocking its interaction with ODAR/RANK [218]. It has recently been demonstrated that OPG-L/ODF/RANKL may directly activate mature osteoclasts in vitro and in vivo

788

STOCH, CHOREV, AND ROSENBLATT

FIGURE 16 Recombinant OPG administration increases bone density in normal mice. Four-week-old male BDP1 mice received subcutaneous injections of either vehicle (A and B), murine OPG [22 – 401]-Fc protein (10 mg/kg/day) (C and D), or the known anti-resorptive pamidronate (5 mg/kg/day) (E and F) for 7 days. The left panels are photomicrographs of Von Kossa-stained frozen sections of the proximal tibial metaphysis. Mineralized bone matrix is stained black in these micrographs, and shown is a similar visible increase in bone density in OPG-treated and pamidronate-treated mice compared to controls. Bone density was measured in the region highlighted (enclosed rectangles) and found to be increased in OPG and pamidronate-treated mice about two- to three-fold. The right panels are photomicrographs of decalcified Masson’s trichrome-stained sections of the distal femoral metaphysis of vehicle-treated (B), OPG-treated (D), or pamidronate-treated (F) mice. Bars in (A) and (B) indicate intervals of 1 mm and 100 m, respectively. Reprinted with permission from Simonet et al. [187]. (See also color plate.) [204]. It is believed that glucocorticoids exert their deleterious effects on skeletal integrity by both inhibiting OPG and stimulating the expression of OPG-L/ODF/RANKL. This has recently been shown to occur in vitro in human osteoblasts and primary and immortalized bone marrow stromal cells [219]. The interplay between OPG-L/ODF/RANKL, OPG/ OCIF, and ODAR/RANK conforms to a signal (agonist), receptor, and decoy receptor, (antagonist) triad [220]. This

molecular “concert” in which upstream factors impact the skeletal phenotype through a common downstream pathway has led to a “convergence hypothesis” [202]. This hypothesis embraces two levels of osteoclast regulation. The “upstream” cytokines and hormones change the pool size of active osteoclasts by converging on the “downstream” effectors, OPG-L/ODF/RANKL and OPG/OCIF [202]. This interplay serves to maintain a pool of active osteoclasts required for active bone resorption. A change in the upstream

789

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

regulators favoring OPG-L/ODF/RANKL will expand the pool of active osteoclasts while regulators that increase OPG/OCIF will reduce this pool size. This regulatory tilt underscores the importance of this triad in regulating the final common pathway required for skeletal integrity.

C. In Vitro Studies 1. ESTROGEN OPG may be responsible for mediating the anti-resorptive action of estrogen on bone. Recent data have shown that 17estradiol increases OPG mRNA in a both dose- and time-dependent manner in the human osteoblastic cell lines, hFOB/ER-3 and hFOB/ER-9, with OPG mRNA and protein levels increasing to a maximum of 370 and 320% of basal, respectively (P 0.001) [197]. The cotreatment of these cell lines with the anti-estrogen ICI 182,780 completely abrogates this effect. Furthermore, the treatment of the mouse stromal cell line, ST-2, with 17-estradiol significantly increases the steady-state level of OPG mRNA [221]. This effect is detectable at 8 h, reaches a peak at 24 h and persists for 48 h. Neither the 17-estradiol stereoisomer nor testosterone have any such effect, further confirming that 17-estradiol enhances OPG transcription through the estrogen receptor [221]. In addition, 17--estradiol also increases OPG mRNA and protein levels in normal human osteoblasts in a dose-dependent manner [197]. Hence, estrogen stimulates the production of OPG in estrogen-responsive human osteoblastic cell lines as well as in normal human osteoblasts. This effect may in part be mediated by TGF-, since estrogen has been shown to increase TGF- both in osteoclasts and osteoblasts [222 – 224], and TGF- stimulates OPG/OCIF [198,199]. Hence, estrogen appears to act by increasing the secretion of OPG in the bone microenvironment and may reduce bone resorption through a paracrine mechanism. By contrast, estrogen withdrawal may cause a decrease in OPG expression leading to enhanced osteoclastogenesis and bone loss.

occurs in a dose- and time-dependent fashion [219]. Simultaneously, dexamethasone stimulates OPG-L/RANKL mRNA steady-state levels in MS and hFOB cells by two- and fourfold, respectively [219]. This is consistent with the hypothesis that glucocorticoids enhance bone resorption by promoting osteoclastogenesis through inhibition of OPG production and simultaneous OPG-L/RANKL stimulation [219]. Strategies aimed at lowering the OPG-L/OPG ratio during the systemic use of glucocorticoids may be helpful in preventing glucocorticoid-induced osteoporsis [219]. 3. PARATHYROID HORMONE PTH may regulate bone resorption, in part, via its effects on OPG-L/RANKL and OPG expression [229]. PTH has been shown to rapidly increase TRANCE (OPG-L/ RANKL) mRNA levels (at 1 h) and decrease OPG mRNA levels (at 6 h) — thereby demonstrating that this regulation is an early step [229]. In in vitro studies using bone marrow, PTH-(1 – 34) administered for 6 days stimulates the number of osteoclast/well in a dose-dependent manner [229]. While OPG-L/RANKL mRNA increases in a dosedependent manner, OPG mRNA expression is decreased/ abolished depending on the dose of PTH selected [229]. In primary osteoblastic cells, PTH stimulates OPG-L/RANKL mRNA expression fourfold at 2 h and inhibits OPG mRNA expression by 46% [229]. PTH may exert its inhibitory effect on OPG mRNA levels through the cAMP pathway. In vivo, a single dose of PTH-(1 – 38) given subcutaneously leads to a decrease in OPG mRNA expression, in both metaphyseal and diaphyseal bone [230]. This reduction in OPG mRNA levels is evident at 1 h and returns to baseline after 3 h. The analog PTH-(1 – 31) stimulates intracellular cAMP production and inhibits OPG expression, whereas PTH analogs (3 – 34 and 7 – 34) which do not stimulate cAMP production have no effect on OPG expression [230]. PTH rapidly stimulates TRANCE/RANKL mRNA expression (at 1 h) and inhibits OPG mRNA expression (at 6 h) — hence this regulation is an early step [229].

2. GLUCOCORTICOIDS Glucocorticoids are a well-recognized risk factor for both bone loss and fracture occurrence [225,226]. While administration of glucocorticoids systemically increases bone resorption and decrease bone formation [227,228], the putative molecular mechanism is poorly understood (see Chapter 44). Glucocorticoid therapy alters the OPG-L/OPG ratio by both increasing OPG-L levels and decreasing those of OPG. This regulation has been assessed in various human osteoblastic cell lineages using Northern analysis, RT-PCR, and ELISA [219]. Dexamethasone diminishes constitutive OPG mRNA steady-state levels by 70 – 90% in primary (MS) and immortalized stromal cells (hMS), primary trabecular osteoblasts (hOB), immortalized fetal osteoblasts (hFOB), and osteosarcoma cells (MG-63). In hFOB cells, this decline (up to 90%)

D. In Vivo Animal Studies 1. SMALL ANIMALS a. Normal Mice and Rats The administration of recombinant OPG (10 mg/kg/day) for 7 days to normal mice results in a threefold increase in trabecular bone mass compared to controls (31.1% vs 12%, respectively) (P 0.0001) (Fig. 16, see also color plate), [187]. This is accompanied by serum OPG levels of 320% at 24 h posttreatment, which are comparable to those seen in transgenic mice at steady state. Furthermore, treatment with OPG/OCIF (5 mg/kg/day) for 2 weeks leads to a decrease in osteoclastic bone resorption [187]. In addition, an increase in BMD and bone volume accompanied by a reduction in active

790 osteoclast number is seen in normal rats following the administration of recombinant human OPG [188]. b. Osteoporosis Models i. OVARIECTOMIZED RAT. Animal models of osteoporosis are useful for preclinical testing of drugs prior to performing human studies [231]. The OVX rat is an established model for osteoporosis research and has been shown to closely capture the salient clinical features of estrogen depletion on the human skeleton [231]. Used prospectively to evaluate the potential utility of agents, it has proven highly predictive of human therapeutic efficacy (see Chapter 37). When recombinant OPG is tested in the OVX rat model, it produces an increase in bone volume and a decrease in osteoclast number [187]. This effect is comparable to that seen with the established anti-resorptive bisphosphonate, pamidronate (APD) [187]. This anti-resorptive effect of OPG further suggests that estrogen affects bone resorption via the OPG/OPG-L pathway [220]. The N-linked portion of the OPG molecule contains the TNF-R-like domain that is both necessary and sufficient to inhibit osteoclastogenesis. The N-terminal 185 amino acids of OPG are required for activity. Also, truncations that disrupt the SS3 bond of domain 4 eliminate this activity [187]. All active forms of OPG have been shown to inhibit osteoclastogenesis in a dose-dependent manner [187]. ii. DISUSE OSTEOPOROSIS. In a 14-day tail-suspension animal model of local disuse osteopenia, OPG has been shown to have a greater effect on maintaining trabecular bone mass than APD (pamidronate) and IBA (ibandronate) [232]. OPG increases femur dry mass by 14% compared to 8 – 10% for APD and IBA. Similarly, OPG increases femur mineral mass by 17% compared to 12% for APD and IBA. Femoral stiffness increases by 26% with OPG compared to 20% with APD [232]. Furthermore, OPG has been shown to completely reverse sciatic nerve crush-induced bone loss in mice [233]. The loss of bone dry mass (Dry-M) on the ipsilateral (crushed) side at the femur was reduced to 0.4% (OPG), 1.9% (APD), and 2.6% (IBA). The tibia Dry-M loss in the ipsilateral limb was reduced to 0.8% by OPG, 1.9% by IBA, and 2.1% by APD [233]. The intravenous administration of OPG by tail vein in rats results in an increase in both the absolute and relative mineral composition [234]. 2. NONHUMAN PRIMATES Nonhuman primates are the large animals of choice in conducting osteoporosis studies. These models possess both a growing and an adult skeletal phase and, in some models, natural menopause occurs [231]. The intravenous (i.v.) or subcutaneous (sc) administration of OPG to male Cynomolgus monkeys as part of phamokinetic/pharmacodynamic studies, results in dose-dependent inhibition of bone resorption — as indicated by the markers, NTX and

STOCH, CHOREV, AND ROSENBLATT

Dpd [235]. The anti-resorptive potency of OPG was demonstrated following i.v. administration of 15 mg/kg OPG with peak inhibition of urine NTX levels occurring between 2 and 5 days. Levels did not return to baseline until day 22 [235].

E. Human Studies 1. PHARMACOKINETICS AND METABOLISM The pharmacokinetic profile of OPG has been characterized in 70 healthy postmenopausal women following a single sc and i.v. administration over a dose range from 0.1 to 3.0 mg/kg (n  10/group). Following sc administration, serum concentrations of OPG increase to peak values at 24 – 48 h postadministration and decline thereafter in a biphasic manner in the majority of subjects with a temporal half-life of 38 – 47 h [236]. Following i.v. administration, serum levels decline in a biphasic manner with an initial half-life of 10 h and a terminal half-life of approximately 28 – 43 h. Clearance following both sc and i.v. administration decreases as the dose increases. The volume of distribution of 37 – 41 ml/kg at steady state indicates that OPG does not extensively distribute extravascularly. The bioavailability after sc dosing appears to increase with the dose administered. 2. POSTMENOPAUSAL OSTEOPOROSIS An age-related increase in serum OPG levels up to threefold has been found in healthy Japanese men and women [236]. This finding was not corroborated by other research groups [202,237]. Postmenopausal women are recognized to be at high risk for osteoporosis and fracture: one-third to one-half of all bone loss may be attributable to menopause [238]. Using a newly developed ELISA assay, serum concentrations of OPG were measured and found to be significantly higher in postmenopausal women diagnosed with osteoporosis than postmenopausal women with normal BMD [239]. Furthermore, when the osteoporotic group was subdivided into age-matched subgroups with lower ( 0.7 g/cm2), or higher L2 – L4 bone mass, the serum OPG concentrations were significantly higher in the former subgroup than in the latter (2.1 0.5 ng/ml vs 1.8 0.3 ng/ml, P  0.002) [239]. Moreover, higher circulating levels of OPG are associated with higher markers of bone turnover, serum bone-specific alkaline phosphatase (BSAP) and urinary excretion of pyridinoline (Pyr) and deoxypyridinoline (Dpyr) in postmenopausal women with osteoporosis or with normal BMD [239]. There has been a recent report of a phase I clinical trial wherein recombinant OPG was tested in postmenopausal women in a randomized, double-blind, placebo-controlled manner. The subjects were divided into four cohorts of 13 women, each

791

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

randomly assigned to OPG or placebo treatment by single sc injection in the abdominal area [240]. The doses of OPG chosen were 0.1, 0.3, 1.0, and 3.0 mg/kg. A total of 52 postmenopausal women were enrolled. A single sc dose of OPG results in a dose-dependent and rapid decrease (within 12 h) in urine NTX [240]. An 80% reduction in urine NTX was reached by day 5 in women receiving 3.0 mg/kg OPG and maintained below baseline for 4 weeks. OPG was well-tolerated; only a single subject exhibited mild injection-site pruritis. As would be anticipated, a modest decrease in serum and urine calcium and an increase in immuno-reactive PTH were seen. 3. GLUCOCORTICOID-INDUCED BONE LOSS Glucocorticoids are a well-established risk factor for both bone loss and fractures [225,226] (see Chapter 44). Given the inhibition of OPG mRNA expression in vitro [219], It is expected that OPG concentrations would be measurably lower in human subjects treated with glucocorticoids. Eleven patients with chronic obstructive pulmonary disease treated with both oral and nebulized glucocorticoids for 3 weeks had their serum OPG levels evaluated. Serum OPG concentrations were significantly reduced (0.23

0.02 ng/ml) following 1 month of glucocorticoid therapy compared to baseline values (0.30 0.03 ng/ml, P  0.002) [241]. Since glucocorticoids alter the OPG-L/OPG ratio it is probable that these cytokines are responsible for mediating the pro-resorptive effects [202]. 4. POSTTRANSPLANT OSTEOPOROSIS Transplantation is a well-recognized risk factor for bone loss and osteoporosis, related not only to the underlying disease and the lack of mobility, but also to high doses of glucocortioids and cyclosporine therapy (see Chapter 52). OPG is expressed both in bone and in the arterial media and OPGdeficient mice develop both severe osteoporosis and arterial calcification [215]. This is particularly interesting since coronary artery disease is more prevalent in the posttransplant setting. Human bone marrow stromal cells exposed to cyclosporin A or dexamethasone demonstrate a decrease in OPG mRNA levels of 41 and 70%, respectively, and a reduction in protein levels of 49 and 60%, respectively [242]. This is accompanied by an increase in OPG-L/RANKL mRNA levels of 1.9- and 2.3-fold, respectively [242]. Cyclosporin A and dexamethasone also decrease OPG mRNA and protein levels in coronary artery smooth muscle cells, but do not alter OPG or OPG-L/RANKL mRNA levels in differentiated human osteoblastic cells [242].

F. Other Indications In addition to the potential usefulness of OPG in treating osteoporosis due to different causes, such as the postmenopausal

setting, corticosteroid treatment, and transplantation, OPG has therapeutic potential in a wider array of metabolic bone diseases characterized by increased bone resorption. Consequently, we can expect OPG to be tested in hyperparathyroidism, humoral hypercalcemia of malignancy, metastatic bone disease, and inflammatory-related bone diseases. As “proof of concept,” there are data on using OPG in the setting of hypercalcemia of malignancy (see below). 1. HYPERCALCEMIA The hypocalcemic effect of OPG/OCIF has been reported in both normal mice and hypercalcemic nude mice [243]. Normal mice administered a single intraperitoneal dose of OPG (20 mg/kg) exhibit a 1.6 mg/dl reduction in serum calcium that is evident by 2 h and sustained for 12 h. Similarly, when hypercalcemic human pancreatic cancer-bearing nude mice are administered the same dose of OPG, a dramatic reduction in serum calcium of 2.8 mg/dl is seen, which persists for 24 h [243]. The efficacy of OPG compared to pamidronate (APD) has been tested in a murine model of humoral hypercalcemia of malignancy. In this model, OPG more rapidly reduces hypercalcemia than does APD and also demonstrates a more profound reduction in osteoclast surface, suggesting a greater inhibitory effect of OPG than APD on osteoclasts [244]. Furthermore, OPG has also been shown to both prevent and reverse the hypercalcemia induced by PTHrP in mice [245]. A recombinant chimeric Fc fusion form of human OPG has been shown to both prevent hypercalcemia (P 0.05) and maintain osteoclast numbers in the normal range (P 0.001) in mice treated with various cytokines and hormones (IL-1, TNF-, PTH, PTHrP, and 1,25 (OH)2D3 [246]. 2. METASTATIC BONE DISEASE Accelerated bone resorption is common in patients with carcinoma metastatic to bone [247]. Since cancer cells stimulate osteoclast-mediated bone resorption, it is plausible that this process may be mediated through the OPG-L/OPG “convergence pathway.” In support of this hypothesis, it has been recently shown that coculturing either mouse melanoma M16 or breast cancer Balb/c-MC cells with mouse bone marrow cells (BMCs) induces osteoclast-like cells [248]. While none of these cells express OPG-L/ODF/RANKL mRNA alone, coculturing the cancer cells with BMCs induces OPG-L/ODF/RANKL expression [248]. Hence, the interplay between cancer cells and BMCs which induce OPG-L/ODF/RANKL expression and suppress OPG/OCIF levels in metastatic foci. Therefore, another potential use for OPG might be to abrogate the bone loss associated with metastatic osteopathy. Furthermore, it is also plausible that this agent may be utilized to reduce the metastatic potential of such tumors.

792

STOCH, CHOREV, AND ROSENBLATT

3. INFLAMMATORY JOINT DISEASE It has been well-documented that patients with inflammatory joint disease, such as rheumatoid arthritis, exhibit enhanced bone loss [249 – 252] (see Chapter 54). Recent data has shown that activated T cells can directly trigger an increase in osteoclastogenenesis and bone loss [253]. This was demonstrated in a T-cell-dependent model of rat adjuvant arthritis characterized by severe joint inflammation and bone destruction, where the blocking of OPGL/RANKL through OPG prevented bone and cartilage destruction [253]. Hence, one can speculate that OPG might find a place in abrogating the bone loss associated with inflammatory joint disease. Furthermore, OPG may hold therapeutic potential in the treatment of periprosthetic bone loss (see Chapter 56). It has been reported that macrophages isolated from the pseudomembrane surrounding loose arthroplasty components are capable of differentiating into osteoclasts and that OPG-L/RANKL is necessary for this to occur [254]. OPG has been shown to inhibit this process, possibly by disrupting cell – cell interaction between osteoblasts and osteoclast precursors [254].

G. Potential Development Concerns Since osteoporosis is a chronic disorder, it is likely that any therapeutic agent will need to be utilized over a sustained period of time. This is a problem for proteins requiring chronic parenteral administration, such as OPG. Since OPGL/RANKL is a regulator not only of osteoclastogenesis, but also lymphocyte development and lymph-node organogenesis [216], the impact of chronic OPG therapy on lymph noderelated processes needs to be carefully addressed.

VII. CATHEPSIN K A. Introduction Osteoclastic bone resorption requires two processes: demineralization of the inorganic bone components and degradation of the organic bone matrix. These two processes occur sequentially by two separate mechanisms. The first phase involves acid secretion into the resorption lacunae, which is followed by organic matrix degradation by cysteine proteases in the second phase. An acidic microenvironment is required for bone resorption, both to dissolve the mineral component of bone and to aid protein matrix digestion. This unique metabolic milieu is achieved by lowering the pH in the resorption lacunae via acid secretion by the osteoclast. Cathepsin K is the most abundant cysteine protease expressed in the osteoclast and is believed to be instrumental

in bone matrix degradation necessary for bone resorption. Cathepsins have known collagenolytic activity under acidic conditions [255] and cathepsin K is capable of degrading components of bone matrix, including type-1 collagen and tartrate-resistant acid phosphatase (TRAP) [256,257]. The finding of cathepsin K deficiency in pycnodysostosis [258], an osteopetrotic disorder characterized by decreased bone resorption, further underscores the importance of this enzyme as a target for developing agents to treat osteoporosis and other disorders featuring bone loss.

B. Chemistry and Actions of Cathepsin K Cathepsin K is a member of the papain family of cysteine proteases and is both selectively and highly expressed in osteoclasts that mediate bone resorption [259]. Furthermore, in situ hybridization has confirmed that cathepsin K mRNA levels of expression are much greater than those of the related proteases, cathepsins B, L, and S [260,261]. Activated cathepsin K is capable of degrading several components of bone matrix, including type I collagen (which accounts for 90% of bone matrix), osteopontin, and osteonectin. Consequently it is believed to play a critical role in the degradation of the organic phase of bone during bone resorption. This pivotal role highlights its worthiness as a target for selective therapeutic intervention to treat disorders characterized by excess bone loss. Cathepsin K is a 38-kDa protein synthesized as a preproenzyme that is autocatalytically cleaved to the mature enzyme form [262]. The biosynthesis, processing, and turnover of procathepsin K is constitutive and occurs in a time frame similar to that seen with other cysteine proteinases [263]. It is synthesized as an inactive 314-amino-acid proenzyme and activated under conditions of low pH [262,264], such as in the bone resorption pit in vivo. In a low pH environment, proteolytic cleavage of the 99-amino-acid propeptide from the amino terminus can occur [254,264]. Some evidence exists to show that cathepsin K is not secreted as a proenzyme, but rather processed intracellularly prior to secretion, most likely in lysosomal vesicles [263]. This was demonstrated using pulse-chase experiments in cultured human osteoclasts [263]. Phosphoinositol-3-kinase (PI3-K) may be instrumental in the intracellular trafficking of procathepsin K, as demonstrated by the inhibitory role of wortmannin, a PI3-K inhibitor, on bone resorption [265]. The inactive cathepsin K precursor localizes in a diffuse pattern in an endoplasmic reticulum-like distribution, while the mature enzyme has been detected in resorption pits, indicating secretion of the protease into resorption lacunae [266]. Biochemical characterization has revealed that the mature enzyme is a monomeric protein with a molecular weight of 24 kDa [267]. Cathepsin K was cloned originally from rabbit osteoclasts [268] and thereafter from human osteoclasts and

793

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

shown to be homologous to other cathepsins [269]. Nothern blot analysis showed that cathepsin K mRNA is highly expressed in osteoclastoma cells and present in lower levels in other tissues [270]. Bone resorption is mediated by osteoclasts which attach to the bone surface, thereby generating an acidified subosteoclastic bone-resorbing compartment [271,272]. This low pH environment favors matrix degradation by enzymes such as cysteine proteases [273 – 275]. Evidence supporting this process are the findings that the cathepsin K transcript and protein have been specifically localized within osteoclasts by in situ hybridization and immunohistochemistry [276]. Furthermore, the mRNA and protein appear to be compartmentalized toward the membrane surface of the osteoclast [276]. While cathepsin K is highly expressed in osteoclasts, very low levels exist in other tissues, including the heart, liver, and lung [276]. The enzyme can be detected extracellularly in bone resorption lacunae and intracellularly in vesicles, granules, and vacuoles throughout the cytoplasm of osteoclasts [277]. In vitro studies have shown that cathepsin K cleaves type-I collagen in the N-telopeptide as well as near the Nterminus of the helical region of the 1-chain [256]. Hence, the collagenolytic activity of cathepsin K is directed both outside the helical region of the molecule (characteristic of other cysteine proteinases) and at various regions inside the helical region, which is unique to cathepsin K [256].

C. Design of Inhibitors The pivotal role of cathepsin K in mediating bone resorption is supported by historical data. It is well known that the thiol protease inhibitors, E-64 and leupeptin, inhibit bone resorption [273,274]. There are in vitro data to support the interruption of osteoclast-mediated bone resorption by classic protease inhibitors, such as leupeptin, Z-Phe-Ala-CHN2, E-64, and cystatin [261,273 – 275]. Furthermore, both ZPhe-Ala-CHN2 and leupeptin have demonstrated activity in the in vivo model of murine hypercalcemia due to excess bone resorption [274]. However, E-64 binds irreversibly to the cysteine protease, and leupeptin, while binding reversibly, is nonselective [278]. This has led to efforts to design more selective enzyme inhibitors of cathepsin K, taking advantage of the restricted expression of cathepsin K to the osteoclast. Initially, nonselective inhibitors of cathepsin K were based on vinyl sulfones [279] or peptidyl aldehydes [280], which possess the drawback of intrinsically reactive groups that can derivatize side-chains and backbone elements. Antigenic and immunological complications theoretically could result from chronic use [279]. This concern led to the development of potent dipeptidylketone inhibitors which attenuate the reactive nature of the aldehyde moiety [280].

Unfortunately, dipeptidylketone inhibitors are less potent than aldehyde inhibitors in reducing cathepsin K activity [280]. Based on structure – activity studies and structurebased design [281], more potent inhibitors of cathepsin K have been produced. Their structure spans both the prime and the unprime sides of the enzyme’s active site [281, 282]. In general, previous inhibitors of cysteine proteases occupied only one-half of the enzyme active site and often contained a functional group [283]. Inhibitors that straddle both sides of the active site are enhanced with regard to both selectivity and potency [283]. The lack of reactive functional groups is optimal since these inhibitors are designed for chronic use, and potentially may elicit undesirable antigenic and immunologic responses [283]. Solidphase synthesis technology has been employed to generate more rapidly a combinatorial array of inhibitors of cysteine proteases [284].

D. In Vitro Studies As part of preclinical drug development, a number of in vitro studies testing cathepsin K inhibition have been performed. Several peptide aldehyde inhibitors of cathepsin K have demonstrated potent inhibition of osteoclast-mediated bone resorption in a concentration-dependent manner using the PTH-stimulated fetal rat long bone (FRLB) model [285]. The most potent of these compounds inhibits bone resorption with an IC50  20 nM in the FRLB assay and with an IC50  100 nM in the human osteoclast resorption assay [285]. This latter model was used to confirm that peptide aldehydes can inhibit resorption by human osteoclasts [285]. The same assay was used to demonstrate in vitro anti-resorptive activity of potent selective human cathepsin K inhibitors — the most potent of which inhibits bone resorption with an IC50  120 nM [283]. Recently, peptidyl vinyl sulfones, selective cathepsin K inhibitors, demonstrated a greater than 80% decrease in bone resorption in a dose-dependent manner [286]. This was shown using rat bone slices at 100 nM concentration, which is approximately 3 orders of magnitude less than that seen with the nonselective protease inhibitors, E-64 and leupeptin [286]. Cathepsin K has also been identified in resorption pits using an anti-mouse cathepsin K antibody, thereby confirming secretion of the protease into the subosteoclastic compartment [286]. The selective cathepsin K inhibitor, SB280648, demonstrated activity in the human osteoclast-mediated bone resorption assay (IC50  300 nM) and completely inhibited osteoclast cathepsin K activity in whole tissue [287]. This SB280648 compound completely inhibited cathepsin K in fetal human bone with an IC50  100 nM [287]. It is difficult to culture osteoclasts from bone marrow cultures [288]. Therefore, a recent study used human

794

STOCH, CHOREV, AND ROSENBLATT

ostoclastoma-derived osteoclasts for in vitro testing of novel anti-resorptive therapies [289]. These cells are phenotypically identical to osteoclasts, can be stored indefinitely in liquid nitrogen, and perform reproducibly in resorption assays [289]. The potency of cathepsin K inhibitors in this in vitro resorption assay correlated with the functional assay of recombinant human cathepsin K enzyme for cathepsin K inhibitors (r2  0.80) [289]. Similar correlates were seen for vitronectin receptor antagonists (r2  0.77) and H -ATPase inhibitors (r2  0.82) [289]. Nonhuman primates have been utilized for evaluating potential osteoporosis therapies prior to human testing [290]. Cynomolgus monkey cathepsin K has shown autoactivation under in vitro conditions similar to that seen with the human enzyme, but with 1.5-fold slower kinetics [291]. Importantly, the mature enzyme generated by autoactivation is identical to the human enzyme [291]. The Rhesus monkey enzyme also is identical to mature human cathepsin K [292]. The sequence and kinetic similarities of the human and monkey cathepsin K’s, coupled with the knowledge that the ovariectomized (OVX) monkey is an excellent model for human osteoporosis [293], indicate that these models will be appropriate for evaluating inhibitors in vivo. Antisense oligonucleotide technology has demonstrated the pivotal role of cathepsin K in bone resorption. This was shown using rabbit osteoclasts on cultured dentine slices. Cathepsin K antisense phosphothiorate oligodeoxynucleotide (S-ODN) reduces cathepsin K protein levels in osteoclasts and inhibits osteoclastic pit formation in a concentration-dependent manner [269]. The maximum cathepsin K inhibition reduces pit number by 50% and pit depth by 30% —a response similar to that of E-64, the nonspecific and potent cysteine proteinase inhibitor [269]. This effect was not reproducible using either sense S-ODN or mismatched S-ODN [269].

E. In Vivo Studies The pursuit of new pharmacological entities has resulted in the promulgation of registration guidelines by regulatory agencies both in the United States and abroad [294] (see Chapter 65). These agencies require the demonstration of long-term safety and efficacy in rats as well as another species prior to approval of a new osteoporosis therapy [294]. It is generally accepted that the rat is a useful model of human bone loss since it responds similarly to humans in regard to effects of mechanical forces and to hormone and drug treatment [295]. However, there appears to be less unanimity regarding the relevance of mouse models of postmenopausal bone loss (see Chapter 37). While mice are readily available and the knockout and transgenic mouse models lend themselves to elegant experimentation and genetic analysis, the mouse strains differ in peak bone mass and susceptibility to bone loss [296]. Furthermore, differences exist in how the hu-

man and mouse skeletons respond to estrogen [296]. Nevertheless, the mouse model has utility in evaluating new chemical entities and their effects on mitigating bone loss. 1. KNOCKOUT-MOUSE The cathepsin K knockout mouse serves as an excellent model for analyzing the two distinct effects of osteoclast-mediated bone resorption: demineralization and organic matrix degradation [297]. This knockout model has been developed by two different groups [298,299] and closely captures the phenotype of the human disease, pycnodysostosis. Cathepsin K-deficient mice, generated by targeted disruption of the cathepsin K gene, exhibit an osteopetrotic phenotype characterized by dense thick trabeculae of the bone-marrow space [298]. These mice exhibit osteopetrosis of long bones and vertebrae by histological, histomorphometric, and microcomputerized tomography (CT) analysis, in addition to abnormal joint morphology [260]. In addition, the mice have hematopoeitic abnormalities with decreased bone marrow cellularity and splenomegaly [260]. When cathepsin K-deficient osteoclasts are assayed for functional activity on dentine slices, they produce fewer and more shallow resorption pits than wild-type osteoclasts [298]. Close inspection of the bone histology in the knockout mouse model reveals fully differentiated osteoclasts apposed to small regions of demineralized bone, strongly suggesting that while cathepsin K-deficient osteoclasts are capable of demineralizing the extracellular matrix, they are unable to completely remove the demineralized matrix [260]. This finding is consistent with the matrix-degrading property of this proteinase [299]. The above data, coupled with the observation that cathepsin K enzyme function is restricted to the osteoclast [299], support the suitability of cathepsin K as a pharmaceutical target for treating disorders characterized by excess bone resorption. While cathepsin K expression has been detected in tissues other than the skeleton, the absence of defects in these nonskeletal tissues and organs in the knockout animal heralds safe utility for selective cathepsin K inhibitors in treating disorders of bone turnover. Furthermore, this knockout mouse model may hold potential for optimizing therapeutic regimens, including gene therapy [297], in treating bone disorders that manifest excess bone resorption. 2. OVARIECTOMIZED MICE Selective and reversible peptide aldehyde inhibitors of cathepsin L have prevented bone loss in OVX mice when administered orally at 50 mg/kg for 3 weeks [278]. This effect is surmised to be due to the inhibition of cathepsin L in breaking down collagen. While cathepsin K was not examined in this instance, cathepsins K and L show 60% sequence identity and three-dimensional structural data suggest that cathepsin K is much more similar to cathepsin L

795

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 17 Aldehyde inhibitor of cathepsin K. (A) Structure of cathepsin K inhibitor. (B) Peptide aldehyde inhibitor of cathepsin K prevents calcemic response to PTH. Time course of changes in whole blood ionized calcium in TPTX rats infused for 6 h with vehicle (circles), hPTH-(1 – 34) 0.3 nmol/kg/h (triangles), or hPTH-(1 – 34) plus cathepsin K inhibitor (squares). Blood ionized calcium in the latter group did not differ significantly from the vehicle control group and was significantly lower (P  0.05) than the PTH-(1 – 34) control group at 6 h. (C) Peptide aldehyde inhibitor of cathepsin K reduces bone loss. Bone densitometric evaluation of the distal tibia in adjuvant arthritic (AA) rats treated with a cathepsin K inhibitor. Rats were treated with the compound at 30 mg/kg/day five times per week, from day 0 to day 23. The data for the treated group are presented as percentage normalization of BMD compared with the AA controls (10 animals per group). *P  0.05. Reprinted with permission of the American Society for Bone and Mineral Research [285]. than cathepsin B [278]. Furthermore, most of the residues spatially arranged close to Cys139 in the active site are conserved between cathepsin L and cathepsin K [278]. Hence, we can expect similar efficacy with cathepsin K and L inhibitors in this model. 3. DISUSE OSTEOPENIC MOUSE MODEL In the mouse model of disuse osteopenia using unilateral cast immobilization, accelerated bone resorption is evident by day 3 [300]. Cathepsin K mRNA levels are shown to be markedly increased by day 3 of immobilization and were 2.8-fold higher than levels seen in control samples (P  0.01). This increase in cathepsin K mRNA levels closely parallels the increase in osteoclast number as assessed by histomorphometry, which also increases 2.8-fold by day 3 (P  0.005). This suggests that cathepsin K transcription remains relatively unchanged during accelerated osteoclastogenesis [300]. 4. RAT MODELS When a selective cathepsin K inhibitor is evaluated in the thyroparathyroidectomized (TPTX) rat model, an established in vivo model of acute bone resorption [165,301], it exhibits potent anti-resorptive activity [283]. A dose of 13 mg/kg/h results in a 45% inhibition of calcium release in the TPTX model [283]. Similarly, the selective cathepsin K inhibitor, SB280648, blocks the osteoclast-stimulated calcemic response in this model by 40% when dosed at 3.1 mg/kg/h [287]. The administration of Cbz-Leu-Leu-Leu-H (Fig. 17A), a potent peptide aldehyde inhibitor, to the TPTX rat model also inhibits the increase in blood ionized calcium levels in-

duced by a 6-h infusion of PTH (Fig. 17B) [285]. Furthermore, when this same compound is used in the adjuvantarthritic (AA) rat model, which is characterized by rapid and significantly increased osteoclastic bone resorption at sites distal to inflammation, it significantly reduces bone loss (Fig. 17C) [285]. Arthritic rats, which receive intraperitoneally a 30 mg/kg dose of this compound for 5 days a week for several weeks, demonstrate a 61% reduction in hindpaw inflammation (P  0.001) as well as normalization of their BMD when compared to controls (P  0.05) [285]. The suppression of inflammation was dose-dependent [285]. In situ hybridization studies have confirmed high levels of cathepsin K mRNA expression in osteoclasts at sites of extensive bone loss in the distal tibia [285]. These data speak to the potential utility of these compounds in ameliorating the excessive bone loss that is seen in inflammatory joint disorders, such as rheumatoid arthritis, which are characterized by excessive bone resorption.

F. Human Studies While we are unaware of any phase I clinical trials using cathepsin K inhibitors in humans, there are biochemical data from patients with pycnodysostosis who are characterized by cathepsin K deficiency. Markers of bone metabolism were assayed in the serum and urine of these patients [302]. While no abnormality is seen in markers of bone formation, osteocalcin (OC) or carboxy-terminal propeptide of type I collagen (PICP), suggesting normal osteoblastic activity [302], bone resorption markers are decreased consistent with reduced osteoclast-mediated matrix degradation. Urine N-telopeptide of collagen cross-links (NTX) and

796

STOCH, CHOREV, AND ROSENBLATT

serum C-telopeptide of collagen cross-links (CTX), two markers that measure collagen type I-breakdown, are reduced in both children and adults [302]. In the 6 affected adults, urine NTX measured 16  8 vs 37  16 nM bone collagen equivalents mM creatinine in controls, P  0.01, and was significantly reduced in the 17 affected children (P  0.003) [302]. Further characterization of this biochemical phenotype reveals an increase in serum cross-linked Ctelopeptide of type I collagen (ICTP) concentration — a more proximal portion of the C-terminus of type I-collagen [302]. The unique finding of an increased amount of the Cterminal epitope, despite a decrease in bone resorption, suggests that cleavage by cathepsin K occurs at the C-terminus beyond the hydrophobic domain recognized by the assay [302].

G. Other Indications While it is clear that cathepsin K inhibitors hold promise in treating postmenopausal osteoporosis and other disorders characterized by excessive bone loss, such as rheumatoid arthritis, it is reasonable to speculate about additional therapeutic usefulness. Utility in the posttransplant setting, corticosteroid-induced bone loss, as well as in treating male osteoporosis may be expected. These agents may further enrich our therapeutic armamentarium in treating hypercalcemic disorders such as hyperparathyroidism and hypercalcemia of malignancy. Cathepsin K inhibitors have potential utility in treating osteoarthritic syndromes characterized by progressive loss of articular cartilage, thickening of trabeculae of subchondral bone and formation of new bone and cartilage at joint margins. The unique collagenolytic activity of cathepsin K and its presence in osteoclasts are in keeping with its important proteolytic activity in osteoclastic bone and cartilage resorption [303]. It has been suggested that cathepsin K is secreted by the synovium and acts principally in the digestion of both bone and cartilage fragments sheared from the joint surface in osteoarthritis [304]. To this end, Dodds et al have demonstrated that giant cells that form within osteoarthritic synovial tissue express high levels of cathepsin K mRNA [304]. The findings of high TRAP activity and the undetectable expression of macrophage-associated degradative proteases (cathepsins B, L, and S) further strengthen the concept that cathepsin K is the principal protease involved in bone degradation [304]. It has recently been shown that glycosaminoglycans (GAGs) selectively and specifically increase cathepsin K collagenolytic activity, thus potentially enhancing cartilage digestion by cathepsin K [303]. Futhermore, these GAGs had no effect on matrix metalloproteinases (MMPs) or cathepsin L activities, suggesting a unique target for treating osteoarthritis [303].

H. Potential Development Issues While peptidyl aldehydes demonstrate potent inhibition of cathepsin K in animal models, safety concerns exist regarding the metabolic implications associated with the chronic use of aldehyde-based compounds.

VIII. INHIBITORS OF Src IN OSTEOPOROSIS The physiological role of Src tyrosine kinase in regulating osteoclast-mediated resorption was revealed when targeted disruption of the c-src-proto-oncogene produced an osteopetrotic phenotype (see Chapter 3). Thus, compounds that specifically block the function of Src tyrosine kinase should inhibit osteoclastic bone resorption and have potential utility in the treatment of osteoporosis. There are three possible pharmacological approaches to achieving the above: (A) inhibition of Src kinase activity, (B) inhibition of binding to Src homology 2 (SH2) domain, and (C) inhibition of binding to Src SH2 by bone-targeted inhibitors.

A. Structure and Function of Src Tyrosine Kinase The nonreceptor protein tyrosine kinase pp60c-Src (Src) is the prototypical member of the Src family of tyrosine kinases [305]. This family includes the widely distributed Src, Fyn, and Yes as well as kinases with a more restricted pattern of distribution such as Lyn, Hck, Fgr, Blk, and Lck. All Src family members share a common modular structure. This includes the SH2 and SH3 protein domains, a catalytic region containing an autophosphorylation site (Tyr416 in Src), and a negative-regulatory tyrosine (Tyr527 in Src) which is located near the carboxyl terminus [305, 306]. The SH3 domain interacts with specific proline-rich sequences that fold in a left-handed helix [307]. The SH2 domain binds to phosphorylated tyrosine residues (pTyr) in the context of specific amino acid sequences (Fig. 18, see also color plate) [308 – 311]. Src kinase is activated by a variety of upstream signaling events mediated by a number of receptors including growth factor receptors, G protein-coupled receptors, and integrin receptors [305]. While the kinase activity plays an important role in autophosphorylation and downstream phosphorylation of protein substrates, the SH2 domain has been implicated in regulating intracellular signaling events [312,313]. The targeted gene-disruption of c-Src in mice results in a severe osteopetrotic phenotype characterized by thickened bone as well as decreases in marrow space and failure of tooth eruption [314]. This particular phenotype is due to an

797

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 18 Representations of Src SH2 domain-ligand complex. (Left) Schematic diagram showing the tertiary structure of the domain and the 11residue phosphotyrosyl peptide corresponding to a sequence found in hamster middle T antigen, EPQpYEEIPIYL (pYEEI). This peptide contains the pYEEI motif that has been found to bind Src SH2 with high affinity (estimated KD  3 – 6 nM). The phosphotyrosyl (see left side of right panel), Glu(1), Glu(2), and Ile(3) (at the left) of the peptide are shown with solid black bonds and are not labeled explicitly.  Helices and  strands are shown as ribbons and arrows, respectively. Several side-chains involved in peptide binding are shown as stick figures and are labeled according to the secondary structure notation used in Waksman et al. [308]. (Right) Molecular surface of Src SH2 domain. Cutaway view of the SH2 domain, showing the interactions with pYEEI peptide. The accesible surface area is represented by red dots, and the polypeptide backbone of SH2 is shown as a purple ribbon. Protein atoms are shown as purple bonds. The peptide pYEEI is shown as a space-filling model, with side-chains of Pro(2), Asn(1), pTyr, Glu(1), Glu(2), and Ile(3) colored green and the backbone in yellow. The phosphate group on pTyr is shown in white. Note that some atoms of the Pro(1) ring are obscure in this cross section. Reprinted with permission from Waksman et al. [308]. (See also color plate.)

intrinsic defect in osteoclasts [315]. While the osteoclast number in these src-/- mice is normal or even above normal, osteoclasts function is compromised. Osteoclasts fail to polarize and form ruffled borders, and hence cannot resorb bone [316]. In keeping with this finding of high expression of Src in osteoclasts, and its association with intracellular membranes [317], Src is also associated with v3 integrin [318] and involved in regulating v3 signaling [319,320]. It also plays a role in other aspects of cytoskeletal activity — where it functions as an adaptor protein involved in the assembly of specific signaling complexes [312,313]. In addition, it functions as a protein kinase associated with the phosphorylation of substrate proteins including Cbl, Fak, and paxilin, which are potential constituents of the podosomes necessary for osteoclast motility [306,321,322]. Src is also implicated in adhesion-induced association with microtubules [323], the formation of an Factin ring in the sealing zone [324], and colony-stimulated factor-1 (CSF-1)-induced spreading of osteoclasts [325].

It is of interest that knocking out the ubiquitously expressed Src generates a “phenotype” restricted to bone. This would suggest a compensatory mechanism whereby other members of the Src family are able to maintain the normal phenotype in the unaffected tissues [305,306]. The reason why Src is nonredundant in osteoclast function is unknown. Perhaps this is related to its very high level of expression in osteoclasts compared to that of other members of the Src family protein kinases (Fyn, Ly, and Yes) [317,326], or to its specific localization within intracellular vesicles [317,321].

B. Inhibition of Src Tyrosine Kinase Src family kinase inhibitors were discovered by the screening of libraries of compounds against different protein kinases in order to identify potential inhibitors of osteoclastic bone resorption both in vitro and in vivo (Fig. 19)

798

STOCH, CHOREV, AND ROSENBLATT

FIGURE 19

Representative Src family protein kinase inhibitors, which were tested as potential inhibitors of osteoclastic bone resorption.

[322,327 – 332]. Using CGP77675, a member of the pyrrolopyrimidine class of tyrosine kinase inhibitors, Missbach and coworkers reported, for the first time, the protective effects of a Src inhibitor both on bone mass and architecture [322]. CGP77675 inhibits IL-1-induced hypercalcemia (2 g/day infusion for 72 h) in a dose-dependent manner (1, 5, and 25 mg/kg sc b.i.d.) in Tif:MAGf male mice. Furthermore, CGP77675 partially prevents bone loss in young OVX Sprague – Dawley rats following the oral administration of 10 and 50 mg/kg b.i.d. for 6 weeks (Table 1) [322]. This inhibitor is more potent on Src than on Lck and fibroblast growth factor (FGF) receptor, and equally potent on Yes and platelet-derived growth factor (PDGF). However, lack of specificity for Src and other members of this family is the major drawback in the development of Src inhibitors as therapeutic agents in the treatment of osteoporosis.

events by acting as an adaptor molecule shuttling specific protein substrates into particular signaling complexes and trafficking the Src to a distinct subcellular locale [312,313]. The existance of X-ray and NMR structural data for SH2 domains has been crucial to the design of SH2 ligTABLE 1 CGP77675 Partly Prevents OvariectomyInduced Changes in the Morphometric Parameters of Vertebral Trabeculae in Young Rats [322] Parameter

Intact control

OVX control

OVX  CGP77675 (50 mg/kg)

BV/TV (%)

22.9  1.5*

13.5  0.8

19.1  1.9†

3.7  0.2

5.0  0.4†

Tb.N (1/mm)

5.8  0.4

Tb.Sp ( m)

145.0  15.0

252.0  21.0

177.0  20.0†

Tb.Th ( m)

41.3  2.0

38.0  0.9

38.4  1.4

TBPf (1/mm)

C. Inhibition of Binding to Src SH2 Domain 295

Lys is an essential residue in the ATP-phosphotransferase site of Src [333]. The rescue of osteoclast function by the transgenic expression of osteoclast-targeted TRAPsrcwt, TRAPsrcY416F, and TRAPsrcK295M in src-/- mutant mice is of major significance. It suggests essential kinase-independent functions for Src in vivo such as recruiting or activating other tyrosine kinases [334]. The SH2 domain is thought to play an important role in the regulation of intracellular signaling

* *

3.7  0.8*

8.9  0.4

5.5  1.4†

Note. Young rats were treated as described in Missbach et al. [322]. CGP77675 was administered twice daily at doses of 10 and 50 mg/kg orally for 6 weeks, beginning immediately after OVX. The vertebral bone was analyzed with the high-resolution microtomograph and the static bone morphometric parameters were calculated as described in Missbach et al. [322]. The results are expressed as mean  SEM. OVX, ovariectomized animals; BV/TV, bone volume/tissue volume; Tb.N, trabecular number; Tb.Th, trabecular thickness; Tb.Sp, trabecular separation; TBPf, trabecular bone pattern factor. * P  0.05. † P  0.01 in Dunnett’s test against the OVX control group.

799

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 20 Structure and activity of Src SH2 inhibitors. (A) Representative inhibitors of Src SH2 domain that are potent bone anti-resorptive agents. (B) Superposition of the model of AP22408 (white) and the X-ray (blue) crystal structure of the AP22408Lck(S164C) SH2 complex. The water-mediated hydrogen bonds (determined crystallographically, blue) between the bicyclic carboxamide template of AP22408 and the protein were not predicted as no water molecules were included in the model [344] (see also color plate). (C) In vivo anti-resorptive activity of AP22408 in TPTX rats. Mixed model area under the curve analysis: three-point moving average data (days 5 – 18). Reprinted with permission from Shakespeare et al. [344]. ands [308,335 – 337], especially the crystal structure of the high affinity complex of the v-Src SH2 domain with the 11-residue peptide which contains the cognate sequence, pTyr-Glu-Glu-Ile [311]. This has provided critical information regarding the intermolecular interactions in the binding site (Fig. 20) [308]. However, targeting the SH2 domain, which is highly conserved within this family of proteins [335,338,339], will not yield the required specificity. Since Cys188 in the pTyr-binding pocket of the SH2 domain is a unique feature in Src and absent in other members of this family, targeting this site may provide the required selectivity. This higher reactivity of Cys188 (due to proximity to several positively charged residues in this pocket) relative to other cysteine residues is

demonstrated by its interaction with the 3 -formyl substituting the 4 -carboxyl-phenylalanine moiety, which functions as a pTyr mimic in the structure of potential inhibitors of SH2 binding [340 – 342]. Recently, Violette and coworkers developed AP22161 in which the 3 -formyl-4 -carboxyl-phenylalanine, a cysteine-targeting residue and a pTyr mimic, is incorporated into a novel nonpeptide bicyclic benzamide template. This template is designed to replace Glu-Glu-Ile in the cognate phosphopeptide Ac-pTyr-Glu-Glu-Ile, thus enhancing the affinity toward the Src SH2 domain (Fig. 20A). AP22161 selectively binds to the Src SH2 domain (IC50  0.24 M) and inhibits binding of peptide ligands for the SH2; it also inhibits Src-dependent cellular activity and reduces osteoclast-mediated bone resorptive activity in vitro [343].

800 The specific targeting of a therapeutic agent to the desired tissue site of action is the most precise mode of drug delivery and is more likely to have the benefit of fewer side effects. Making use of structural elements with high affinity for hydroxyapatite, the mineral portion of the bone matrix, offers a unique strategy for targeting Src SH2 inhibitors specifically to bone. Most recently, Shakespeare et al. utilized this approach in constructing AP22408, a potent bonetargeted Src SH2 inhibitor (IC50  0.30 M compared to 5.56 M for the cognate phosphopeptide, Ac-pTyr-GluGlu-Ile) [344]. When the pTyr-mimic residue in AP22161 is replaced by 3 ,4 -diphosphono-phenylalanyl (Dpp), a metabolicaly stable pTyr mimic, AP22408, a potent bonetargeted anti-resorptive agent in vitro, is generated (Fig. 20A). The Dpp moiety is not only a pTyr mimic but also exerts bone-targeting properties (AP22408 affinity for hydroxyapatite is K  1.9 compared to 3.6 for alendronate) [344]. X-ray structural analysis of AP22408 complexed with Luc(S164C) SH2 domain confirms the bimolecular interactions predicted by molecular modeling (Fig. 20B, see also color plate) [344]. AP22408 inhibibits rabbit osteoclast-mediated dentine resorption with high potency (IC50 of 1,6 M) [344]. This effective anti-resorptive property of AP22408 was initially established in the TPTX rat model, where it effectively blocked the PTH-induced hypercalcemic effect (Fig. 20C) [344]. Nevertheless, the potential utility of AP22408 and other Src SH2 inhibitors as anti-resorptive agents remains to be demonstrated in OVX animal models.

IX. COMBINED THERAPIES In this section we highlight the available data on combination therapy. We will not, however, discuss any particular agent in detail, since these agents have all been thoroughly reviewed in the other chapters of this section, Pharmacology and Therapeutics. The purpose of combination therapy is twofold. First, to further augment bone mineral density beyond that expected with a single agent alone by capitalizing on distinct mechanisms of action employed by different forms of therapy. Second, to provide additional tissue-specific extraskeletal benefits unique to the particular agents chosen. The former premise rests on the belief that the greater the BMD attained, the lower the likelihood of subsequent fractures [345]. Given the above rationales, what permutations of combination therapies are available? The selection may include the combination of two or more anti-resorptive agents, or the combination of an anti-resorptive with an anabolic agent. One may opt for combination therapy in treating patients with osteoporosis who fail to respond optimally to a single agent or who have severe osteoporosis which requires a maximal or accelerated increase in bone mineral density. Furthermore, one may desire the ex-

STOCH, CHOREV, AND ROSENBLATT

traskeletal benefits of an agent in addition to its effect on BMD alone. We have restricted our overview to widely accepted agents that are now used in clinical practice or in clinical trials.

A. Combinations of Anti-Resorptive Therapies There are presently several classes of anti-resorptive agents available for clinical practice, including hormone replacement therapy (HRT), bisphosphonates, selective estrogen receptor modulators (SERMs) and calcitonin. The combined use of bisphosphonates with estrogen, raloxifene, or calcitonin is often used when the clinician is managing a challenging case. In certain instances, more than two agents may be utilized simultaneously. While we are aware of this practice, we do not have available clinical data to support the use of three or more agents in combination. 1. BISPHOSPHONATE  ESTROGEN Several of the bisphosphonates currently available have been used in combination with estrogen therapy including etidronate, alendronate, and risedronate. a. Etidronate  Estrogen The bisphosphonate, etidronate, has been shown to increase bone mineral density by 4.2 – 5.2% at the lumbar spine following 2 years of therapy. A reduction in the rate of vertebral fractures was observed when subjects in the etidronate groups were pooled [346]. Similarly, in a 3-year study, etidronate increased vertebral BMD by 5.3% vs 2.7% loss in the placebo group (P  0.01), with a reduction in the rate of new vertebral fractures in the treatment group [347]. The use of etidronate in a “cyclic” regimen in the aforementioned studies was not associated with mineralization defects. The mechanism of action of this compound includes the intracellular formation of cytotoxic analogs of ATP [348]. The amino-bisphosphonates, such as alendronate and risedronate, by contrast, inhibit protein isoprenylation and activate caspase-3-like proteases [348] a mechanism distinct from the non-aminobisphosphonate compounds. In a 4-year prospective randomized study in early postmenopausal women, the combination of etidronate and estrogen results in a significantly higher BMD at both the lumbar spine (P  0.05) and femoral neck (P  0.01) than seen with HRT or etidronate used alone [349]. The combined therapy group demonstrated a 10.9% increase in lumbar spine BMD (P  0.001) and 7.25% increase in femoral neck (P  0.001) BMD, wheras the group treated with etidronate alone showed a 6.79% increase in spine BMD (P  0.001) and 1.20% at the femoral neck (P  0.05). In the HRT-treatment group, increases were 6.78% (P  0.001) and 4.01% (P  0.01) at the spine and femoral neck, respectively [349]. Three of nine (33%) subjects in

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

the etidronate group developed osteomalacia by histomorphometry, whereas no patients in the combination group showed evidence of mineralization defects [349]. A similar beneficial effect of combining etidronate with HRT has been shown in postmenopausal women with established osteoporosis in a 4-year randomized study [350]. Study subjects who received combined therapy demonstrated a 10.4% (P  0.001) increase in lumbar spine and 7.0% (P  0.001) increment in femoral neck BMD. The increase in the intermittent cyclic etidronate (ICE) and HRT groups at the lumbar spine were 7.3% (P  0.001) and 7.0% (P  0.001), respectively, and 0.9% (P  0.05) and 4.8% (P  0.01), respectively, at the femoral neck [350]. The combined therapy group had significantly higher BMD at both the spine and femoral neck (P  0.05) compared to that seen in either ICE or HRT treatment groups. However, this study was too small to ascertain any statistical differences in fracture occurrence between the treatment groups. No subjects showed histomorphometric evidence of osteomalacia, as was seen in the prior study [349]. While the two aforementioned studies address the enhanced benefit on BMD and potentially less bone mineralization defect attained when estrogen and cyclical etidronate are initiated simultaneously, one may question whether this would hold true if etidronate is added at a later time point. In clinical practice, medications are often introduced sequentially. Therefore, one may wonder about the relevance of clinical trials based on simultaneous initiation of therapies with two or more agents. This issue was addressed in a recent study which showed no significant difference in vertebral BMD at 12 months — whether ERT and etidronate were started together or if etidronate was given to women who had been on ERT for an average of 8.6 years [351]. Hence, the increased effects of combined estrogen and cyclic etidronate are shown to be independent of duration of estrogen use prior to initiating etidronate [351]. b. Alendronate  Estrogen Alendronate, a nitrogencontaining bisphosphonate, has been shown to inhibit the rate-limiting step in the mevalonate-cholesterol biosynthetic pathway (see Chapter 16). This pathway is critical for osteoclast function [146]. Through this pathway, alendronate can promote osteoclast apoptosis [352,353]. In postmenopausal women with osteoporosis, alendronate treatment produces a 3.5 – 6% increase in BMD at the lumbar spine and 2 – 3% at the femoral neck at 1 year when compared to baseline [354,355] (see Chapter 72). Estrogen, by contrast, reduces bone resorption by decreasing levels of cytokines associated with osteoclast stimulation [356]. Estrogen has been shown to reduce vertebral fractures [357,358] and to maintain or improve BMD [359,360] (see Chapter 69). HRT results in an increase in lumbar spine BMD of 2 – 8% and at the femoral neck of 2 – 3% at 1 year compared to baseline [361].

801 Since these agents employ different mechanisms of action, studies have been performed to ascertain whether utilizing these two agents together will result in a greater increase in BMD than anticipated with either agent alone. Furthermore, since these two agents may be used together, safety and tolerability were also ascertained [362]. In a recent study, the addition of alendronate to ongoing HRT produced a larger increase in lumbar spine BMD from baseline than observed in subjects taking HRT alone [361]. Four hundred and twenty-eight postmenopausal women with osteoporosis who had received HRT for at least 1 year were randomized to receive alendronate 10 mg daily vs placebo for 12 months. The alendronate group demonstrated a significant increase in BMD at the lumbar spine (3.6% vs 1.0%, P  0.001), while the difference at the femoral neck was not significant (1.7% vs 0.8%, P  0.072) [361]. Furthermore, combination therapy (Aln  HRT) produced a significantly greater decrease in bone-specific alkaline phosphatase (BSAP) and NTX than seen with HRT alone at both 6 and 12 months. Combination therapy was well-tolerated with no significant difference between the two groups in gastrointestinal adverse events or fractures [361]. A recent multicenter, double-blind, randomized, 2-year study enrolled 425 postmenopausal osteoporotic women who were treated with either placebo, alendronate (10 mg daily), conjugated equine estrogen (CEE; 0.625 mg daily) or combination alendronate (10 mg/day) plus CEE (0.625 mg/day). At 2 years, those on placebo lost 0.6% BMD at the lumbar spine, while those women receiving either alendronate or CEE demonstrated a 6.0% increase in BMD (P  0.001 vs placebo) (Fig. 21). Those subjects taking combination therapy (Aln  CEE) demonstrated an 8.3% increase in lumbar spine BMD (P  0.001 vs placebo) (Fig. 21). Combination therapy (Aln  CEE) also demonstrated an increase in total hip and femoral neck BMD of 1 – 1.5% and 0.5 – 1% over CEE and alendronate respectively (Fig. 21 [362]. The increases seen in the BMD in the treatment groups continued through the second year and did not reach a plateau. Urinary NTX, a marker of bone resorption, reached a nadir at 6 or 12 months in the three treatment groups. At the end of 24 months, the mean changes from baseline in urinary NTX were 61% (P  0.005), 52% (P  0.001), and 70% (P  0.001) in the alendronate, CEE, and combination groups, respectively [362]. At baseline, these markers were not strongly predictive of the response of bone mass to treatment. Significant changes were also seen in BSAP. Evidence of complete suppression of bone turnover was excluded by the presence of normal bone quality and mineralization, demonstrated by bone biopsies. The above medications were generally well-tolerated with no significant differences in adverse experiences using either the combined therapy or using either agent alone.

802

STOCH, CHOREV, AND ROSENBLATT

FIGURE 21 Combination therapy (Alendronate  estrogen) demonstrates an increase in spine and hip BMD over either agent alone. Mean percentage change from baseline in BMD at lumbar spine (A), total hip (B), and femoral neck (C). BMD (  SE; intention-to-treat) in PBO (), alendronate (ALN; ), CE (), and ALN plus CE () treatment groups. Reprinted with permission from Bone et al. [362]. While this study clearly demonstrates a greater increase in bone mass with combination therapy, the study was too small to show fracture reduction. While the above data support an increase in BMD with combination therapy, one may wish to consider the impact of therapy withdrawal. In a recent study by Greenspan et al., 3 years of combination therapy yielded BMD increases of 10.1% at the spine (P  0.001) and 6.7% at the femoral neck (P  0.001) when compared to baseline [363]. Upon discontinuation of therapy, a differential effect on BMD from omission of the agents was noted. The women in the estrogen group lost BMD at an accelerated rate while subjects in the alendronate and combination (Aln  ERT) group maintained their BMD [363]. Hence, alendronate has a unique benefit in mitigating enhanced bone loss. This effect may be attributed to the different mechanisms of action or potency of these two anti-resorptive medications. c. Risedronate  Estrogen The combination of risedronate (5 mg/day) and estrogen (0.625 mg/day) also

results in a larger increase in BMD at the lumbar spine, femoral neck, and midshaft radius than seen with estrogen alone [364]. This combination of risedronate with estrogen was also well-tolerated. 2. BISPHOSPHONATE  SERM The combination of a SERM and bisphosphonate has also been recently reported. Raloxifene (RLX) is a benzothiophene-derived SERM which has been shown to increase BMD, decrease markers of bone turnover, and reduce new vertebral fractures in postmenopausal women [365]. A recent study showed that combining raloxifene with alendronate is both feasible, and that the combination of the two agents confers a greater effect on BMD than either agent alone [366]. In a phase II, randomized, double-blind, 1-year study, 330 postmenopausal women were randomized to one of four groups: placebo PBO, RLX 60 mg/day, Aln 10 mg/day, or RLX  Aln. The increase in lumbar spine BMD was 2% with RLX, 4.3% with Aln, and 5.2% with combination RLX  Aln. While RLX  Aln decreased markers of bone

803

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

turnover more than Aln alone, the difference was not statistically significant. This study was small and not designed to assess the effect of combination therapy on fracture risk reduction. The extraskeletal benefits of raloxifene, including a possible benefit in reducing the risk of breast cancer [367], may provide an additional attractive reason to consider therapy with this drug alongside a bisphosphonate.

B. Combination of Anti-Resorptive and Anabolic Therapies The combination of an anti-resorptive therapy with an anabolic agent has compelling rationale for treatment of osteoporosis. Currently, only anti-resorptive agents are FDA-approved for treatment of osteoporosis; an anabolic agent would be a welcome addition to our therapeutic array. In utilizing this combination of agents, one would desire to maintain the increment in bone mass attained with an anabolic agent by the addition of an antiresorpive agent. Furthermore, one may speculate on the efficacy of an anabolic agent in the setting of prior treatment with a bisphosphonate. Parathyroid hormone is the anabolic agent most likely to reach the market and we will mention its use with available anti-resorptives, including estrogen, calcitonin, and bisphosphonates. 1. PTH  ESTROGEN It has long been recognized that PTH exerts an anabolic effect on the skeleton resulting in an increase in bone mass [4,51,368]. An agent that could simultaneously enhance bone mass and repair the microarchitectural damage associated with osteoporosis [51] could fill a unique niche in this therapeutic arena. Whether PTH simply increases BMD — or additionally repairs damage to microarchitecture — is currently not known. The combination of hPTH-(1 – 34) with estrogen in women with postmenopausal osteoporosis has been shown to increase the width of packets of new cancellous bone with resultant increases in the width of trabecular bone [369]. In a 3-year randomized clinical trial, the addition of 25 g hPTH-(1 – 34) to the treatment regimen of postmenopausal women taking estrogen 0.625 mg resulted in an increase in vertebral bone mass [48]. The total increase in vertebral BMD was 13.0% (P  0.001), 2.7% (P  0.05) at the hip, and 7.8% (P  0.002) at the total body (Fig. 22) [48]. No significant decline was seen in the estrogen control group. Increased bone mass was associated with a reduction in vertebral fractures, which was defined as a 15% reduction in vertebral height (P  0.04); however, when a 20% cutoff was chosen to define vertebral fractures, only a trend was seen toward fracture reduction. Furthermore, no loss of bone mass was found at any skeletal site in the PTH group. PTH was well-tolerated in the study population and no individuals developed hypercalciuria.

In a recent randomized, double-blind, placebo-controlled trial, the daily subcutaneous administration of hPTH-(1 – 34) (400 IU) in combination with oral estrogen produced increases in BMD of 29% at the lumbar spine (P  0.001) and 11% at the femoral neck (P  0.001) at 2 years [48]. The BMD increases at the lumbar spine and femoral neck in the placebo group were 0.9 and 0.2% respectively. Given the potential concerns that PTH increases trabecular bone at the expense of cortical bone [51], cortical bone density has been evaluated using 3D quantitative computed tomography (3DQCT) in postmenopausal women receiving PTH [370]. It was shown that PTH treatment in estrogenreplete postmenopausal women is effective in increasing bone mass at clinically relevant sites without compromising cortical bone [370]. 2. PTH  CALCITONIN Intranasal calcitonin (CT) is usually administered at a dose of 200 IU daily [371]. It has been shown to increase BMD from 1.4 to 2.1% with an accompanying new vertebral fracture reduction [372]. Five-year data from the PROOF study showed a 36% reduction in new vertebral fractures with 200 IU nasal [373] (see Chapter 73). PTH has been combined with calcitonin in earlier studies, where sequential calcitonin significantly blunted the effect of PTH on serum alkaline phosphatase and osteocalcin [374]. To test whether sequential anti-resorptive therapy with calcitonin enhances the anabolic effects of PTH, a 2-year study comparing the efficacy of cyclical hPTH-(1 – 34) with or without sequential subcutaneous calcitonin was performed [375]. Lumbar spine BMD increased 10.2% in the PTH group and 7.9% in the PTH  CT group; results were significant in both groups (P  0.001). The changes at the femoral neck were 2.4% and 1.8%, respectively [375]. It is noteworthy that while neither group demonstrated a significant increase in femoral neck BMD over baseline, the group receiving PTH alone consistently averaged higher increments in femoral neck BMD. Hence, the combination of cyclical PTH and sequential CT showed no enhanced benefit of the combination therapy on BMD over that of PTH alone. 3. PTH  BISPHOSPHONATE While there are data available on the combination of PTH with estrogen, the clinical data on PTH with bisphosphonates is sorely lacking. There is great interest in this combination because it couples the most potent anti-resorptive agents with the most potent anabolic agent. One would like to show the preservation of newly acquired bone by a bisphosphonate following treatment with PTH as well as assess whether a bisphosphonate might abrogate the efficacy of PTH on enhancing bone mass acquisition. There is preliminary in vivo rat data available to address these questions. First, it has been shown that the administration of

804

STOCH, CHOREV, AND ROSENBLATT

FIGURE 22

Changes in bone mass at the total hip, lumbar spine and total body in postmenopausal women taking estrogen 0.625 mg, following the addition of 25 g hPTH-(1 – 34) over a period of 3 years. *P  0.001. †P  0.02. ‡ P  0.02. Reprinted with permission from Lindsay et al. [48].

PTH results in an increase in endosteal and periosteal bone formation with a consequent increase in mechanical strength and competence of the femur. Upon withdrawal of PTH, there is rapid and pronounced endosteal bone resorption with a decrease in bone mass and mechanical strength and competence. This decrease in BMD is preventable with the subsequent administration of risedronate [376]. To address the impact of prior bisphosphonate treatment on the subsequent anabolic effect of PTH, alendronate pretreatment reduces endocortical bone formation activated by the PTH analog, SDZ PTS 893 [377]. This bisphosphonate pretreatment effect is sustained in animals both with and without a recovery period and persists for at least 2 months after the initiation of anabolic therapy [377]. A recent exciting study by Rittmaster et al. [378] suggests that sequential treatment of osteoporosis with PTH and alendronate results in a greater increase in vertebral bone density than seen with combination anti-resorptive therapy. Two hundred and six postmenopausal women were treated for 1 year with 50, 75, or 100 g of daily sc recombinant human PTH-(1 – 84) or a placebo; a small subset of subjects (n  75) was then continued with open-label alendronate (10 mg/day) for an additional year. During the first year, the mean vertebral bone density increased by 1.3, 4.3, 6.9, and 9.2% in the placebo and 50, 75, and 100 g PTH groups, respectively (P  0.0001) (Fig. 23) [378]. After

2 years of treatment, PTH in the first year and 10 mg alendronate in the second, the respective increases in vertebral BMD were 7.1  5.3, 11.3  5.7, 13.4  5.0, and 14.6  7.9% [378]. The femoral neck BMD changed by 1.0, 1.0, 2.3, and 0.7% in the placebo and 50, 75, and 100 g PTH groups, respectively, in the first year [378]. Following 2 years of sequential PTH and alendronate, the femoral neck bone density increased by 4.2, 5.5, 2.8, and 4.5%, respectively. This study shows that sequential therapy yields greater improvements in vertebral bone mass, maintenance of bone mass upon PTH withdrawal, and the ability to reverse cortical bone loss [379]. Future studies will no doubt address other interesting questions, including the effects of concomitant PTH and bisphosphonate therapy, intermittent bolus PTH therapy, and ultimately whether the combination reduces hip fractures [379]. 4. PTH  SERM While combination therapy may be prescribed to further enhance BMD, it may also facilitate a more convenient dosing schedule by potentially reducing the dosing frequency required. An attempt has been made to address this question using the combination of the raloxifene analog (RA) (LY117018) and hPTH-(1 – 34) in 6-month-old female rats [380, 381]. The combination of hPTH-(1 – 34) and RA given 5 days/week resulted in an increase in BMD at both the spine

805

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

FIGURE 23

Sequential treatment of osteoporosis with PTH and alendronate results in greater increase in vertebral BMD than seen with alendronate alone. BMD (percentage change from baseline; mean  SE.) in postmenopausal women given 100, 75, or 50 g PTH or placebo during the first year followed by 10 mg alendronate daily during the second year. Month 12 data represent changes observed after 1 year of PTH or placebo. Month 18 and month 24 data represent the changes observed after an additional 6 and 12 months of alendronate treatment, respectively. *P  0.05; **P  0.01; ***P  0.001 (compared to placebo/alendronate group). Reprinted with permission from Rittmaster et al. [378].

and the tibia in the OVX animals. The effect was greater than that seen in the untreated (SHAM) rats [380]. A similar effect was noted with the combination of hPTH-(1 – 34) and RA when the dosing frequency of PTH was reduced from 5 days to 2 days per week, suggesting no loss of anabolic effect with PTH despite a reduced dosing frequency. The question of maintaining increased bone mass was addressed by the same group in a separate study [381]. Following 3 months of continuous combination therapy, hPTH-(1 – 34)  RA for 5

days/week, the rats were stratified into four groups. One group received the above-described continuous regimen for 2 additional months, and the others received either the combination of reduced hPTH-(1 – 34) 2 days/week and continuous RA 5 days/week, or either of these two agents alone. The combination of reduced PTH with continuous RA was effective in maintaining bone mass while the two agents used individually were not. Since the anabolic effects on bone mass are rapidly reversible upon discontinuation of the

806 anabolic stimulus, one may consider using the combination of a dose reduction in PTH with continuous anti-resorptive therapy in order to maintain bone density. 5. PTH  VITAMIN D It has been shown that the daily coadministration of hPTH-(1 – 34) with 1,25-dihyrdroxyvitamin D3 increases trabecular bone mass in men with idiopathic osteoporosis [382]. The combination of hPTH-(1 – 34) (400 – 500 IU) plus daily 1,25 (OH)2D3 (0.25 g) improved calcium and phosphorus balance as well as lumbar spine BMD [382]. It has been shown that the periodic use of hPTH-(1 – 34) alternating with calcitriol may actually reduce bone remodeling in spinal osteoporosis rather than increasing bone turnover [383]. 6. OTHER COMBINATIONS We have reviewed the literature pertaining to currently available combination therapies in postmenopausal osteoporosis. This is a burgeoning topic and we anticipate that it will receive considerable attention in the years to come. We did not highlight combination therapy in the setting of male osteoporosis, but expect much attention to be focused on this exciting area in the future, especially since the number of monotherapies available for men are less than those available for women. Alendronate has demonstrated utility in the setting of male osteoporosis, even in the presence of hypogonadism [384]. We anticipate the use of bisphosphonates with androgen replacement therapy in this setting. With a plethora of new and exciting agents on the horizon, novel combinations and permutations will emerge in the clinical arena that employ selective advantages germane to different classes of drugs. While we do not have any data to suggest a protective effect using combination therapy, it is reasonable to speculate that a greater increase in bone mass will most likely confer enhanced protection against fracture occurrence [385].

X. SUMMARY In this chapter we have highlighted some of the emerging, exciting, and most promising therapeutic categories for osteoporosis. However, we are also cognizant of the fact that entities not reviewed may emerge to occupy therapeutic niches. Other exciting entities not covered in this chapter include genetic approaches employing proteomics, interleukin (IL-1) receptor antagonists, osteoclast vacuolar-H ATPase inhibitors, p38 MAPK inhibitors, and bone morphogenetic proteins (BMPs). These approaches have great potential, but are not yet close to human application. The development of novel anabolic and anti-resorptive modalities will further enrich our therapeutic armamentarium in an unprecedented manner. We are

STOCH, CHOREV, AND ROSENBLATT

poised to enter a new era where a single agent is less likely to be appropriate in every case, and the ability to tailor therapies to different patient populations is desirable. This will be better accomplished with a richer array of agents and a more in-depth understanding of bone physiology. Genetic strategies will likely afford us the opportunity to better select high-risk patient populations for treatment at an early stage and facilitate choosing agents to which individual patients will more likely respond (or have fewer side effects). It is also important to ask what governs therapeutic success. Is it the potency and selectivity of an agent at the desired target, the selection of the most appropriate dose for pivotal clinical trials, or is it the critical nature that the targeted pathway plays in osteoclast/osteoblast function? The latter, namely “class effect,” will be answered in due course as our understanding of osteoporosis pathophysiology improves. However, creative basic research and elegant clinical studies remain the key to bringing forward new and improved agents to satisfy therapeutic needs in osteoporosis. Given that osteoporosis is a chronic disease of major public health proportions, any future agent chosen to treat this disorder must distinguish itself as safe, effective, and affordable.

References 1. A. C. Looker, E. S. Orwoll, C. C. Johnston, R. L. Lindsay, H. W. Wahner, W. L. Dunn, M. S. Calvo, T. B. Harris, and S. P. Heyse, Prevalence of low femoral bone density in older U.S. adults from NHANES III. J. Bone Miner. Res. 12, 1761 – 1768 (1997). 2. N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton, Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the national osteoporosis foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). 3. F. Cosman and R. Lindsay, Is parathyroid hormone a therapeutic option for osteoporosis? A review of the clinical evidence. Calcif. Tissue Int. 62, 475 – 480 (1998). 4. W. Bauer, J. C. Aub, and F. Albright, Studies of calcium phosphorus metabolism: Study of bone trabeculae as ready available reserve supply of calcium. J. Exp. Med. 49, 145 – 162 (1929). 5. L. I. Pugsley and H. Selye, The histological changes in the bone responsible for the action of parathyroid hormone on the calcium metabolism of the rat. J. Physiol 79, 113 – 117 (1933). 6. J. Reeve, R. Hesp, D. Williams, P. Hulme, L. Klenerman, J. Zanelli, A. Darby, G. Tregear, and J. Parsons, The anabolic effect of low doses of human parathyroid hormone fragment on the skeleton in postmenopausal osteoporosis. Lancet 1, 1035 – 1038 (1976). 6a. J. Reeve, P. J. Meunier, J. A. Parson, M. Bernat, O. L. M. Bijvoet, P. Courpron, C. Edouard, L. Klenerman, R. M. Neer, J. C. Renier, D. Slovik, E. J. F. E. Vismans, and J. T. Potts, Anabolic effect of human PTH on trabecular bone in involutional osteoporosis: A multicenter trial. Br. Med. J. 280, 1340 – 1344 (1980). 7. N. E. Lane, J. M. Thompson, G. J. Strewler, and J. H. Kinney, Intermittent treatment with human parathyroid hormone(hPTH[1–34]) increased trabecular bone volume but not connectivity in osteopenic rats. J. Bone Miner. Res. 10, 1470 – 1477 (1995). 8. D. B. Kimmel, R. P. Bozzato, K. A. Kronis, P. Kwong, and R. R. Recker, The effect of recombinanat human (1 – 84) or synthetic human (1 – 34) parathyroid hormone on the skeleton of adult os-

807

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

9.

10.

11.

12.

13. 14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

teopenic ovariectomized rats. Endocrinology 132, 1577 – 1584 (1993). K. J. Ibbotson, C. M. Orcutt, S. M. D’Souza, C. L. Paddock, J. A. Arthur, M. L. Jankowsky, and R. W. Boyce, Contrasting effects of parathyroid hormone and insulin-like growth factor I in an aged ovariectomized rat model of postmenopausal osteoporosis. J. Bone Miner. Res. 7, 425 – 432 (1992). V. Shen, D. W. Dempster, R. Birchman, R. Xu, and R. Lindsay, Loss of cancellous bone mass and connectivity in ovariectiomized rats can be restored by combined treatment with parathyroid hormone and estradiol. J. Clin. Invest. 91, 2479 – 2487 (1993). T. J. Wronski, C.-F. Yen, H. Qi, and L. M. Dann, Parathyroid hormone is more effective than estrogen or bisphosphonates for restoration of lost bone mass in ovariectomized rats. Endocrinology 132, 823 – 831 (1993). C. C. Sogaard, T. J. Wronski, J. E. McOsker, and L. Mosekilde, The positive effect of PTH on femoral neck bone strength in ovariectomized rats is more pronounced than that of estrogen or bisphosphonates. Endocrinology 134, 650 – 657 (1994). M. Li and T. J. Wronski, Response of femoral neck to estrogen depletion and parathyroid hormone in aged rats. Bone 16, 551 – 557 (1995). M. Gunness-Hey and J. M. Hock, Increased trabecular bone mass in rats treated with human synthetic parathyroid hormone. Metab. Bone Dis. Relat. Res. 5, 177 – 181 (1984). J. M. Hock, I. Gera, J. Fonseca, and L. G. Raisz, Human parathyroid hormone-(1 – 34) increases bone mass in ovariectomized and orchidectomized rats. Endocrinology 122, 2899 – 2904 (1988). D. N. Kalu, R. Echon, and B. W. Hollis, Modulation of ovariectomy-related bone loss by parathryoid hormone in rats. Mech. Ageing Dev. 56, 49 – 62 (1990). L. Mosekilde, C. H. Sogaard, C. C. Danielsen, O. Torring, and M. H. L. Nilsson, The anabolic effects of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: A comparison study between hPTH-(1 – 34) and hPTH-(1–84). Endocrinology 129, 421 – 428 (1991). L. Mosekilde, C. H. Sogaard, J. E. McOsker, and T. J. Wronski, PTH has a more pronounced effect on vertebral bone mass and biomechanical competency than antiresorptive agents (estrogen and bisphosphonate)-assessed in sexually mature, ovariectomized rats. Bone 15, 401 – 408 (1994). B. H. Mitlak, D. C. Williams, H. U. Bryant, D. C. Paul, and R. M. Neer, Intermittent administration of bovine PTH-(1–34) increases serum 1,25-dihydroxyvitamin D concentrations and spinal bone density in senile (23 month) rats. J. Bone Miner. Res. 7, 479 – 484 (1992). H. E. Takahashi, T. Tanizawa, M. Hori, and T. Uzawa, Effect of intermittent administration of human parathyroid hormone (1 – 34) on experimental osteopenia of rats induced by ovariectomy. Cells Mater. Suppl. 1, 113 – 117 (1991). K. Tada, T. Yamamuro, H. Okumura, R. Kasai, and H. Takahashi, Restoration of axial and appendicular bone volumes by hPTH(1 – 34) in parathyroidectomized and osteopenic rats. Bone 11, 163 – 169 (1990). M. Li, L. Mosekilde, C. H. Sogaard, J. S. Thomsen, and T. J. Wronski, Parathyroid hormone monotherapy and cotherapy with antiresorptive agents restore vertebral bone mass and strength in aged ovariectomized rats. Bone 16, 629 – 635 (1995). H. Qi and M. Li, A comparison of the anabolic effects of parathyroid hormone at skeletal sites with moderate and severe osteopenia in aged ovariectomized rats. J. Bone Miner. Res. 10, 948 – 955 (1995). M. Sato, G. Q. Zeng, and C. H. Turner, Biosynthetic human PTH (1–34) effects on bone quality in aged ovariectomized rats. Endocrinology 10, 4330 – 4337 (1997). C. Ejersted, T. T. Andreassen, H. Oxlund, P. H. Jorgensen, B. Bak, J. Haggblad, O. Torring, and M. H. Nilsson, Human PTH (1–34)

26.

27.

28. 29.

30.

31.

32.

33. 34.

35.

36.

37. 38.

39.

40.

41.

42.

43.

and (1–84) increase the mechanical strength and thickness of cortical bone in rats. J. Bone Miner. Res. 8, 1097 – 1101 (1993). C. C. Liu and D. N. Kalu, Human parathyroid hormone (1 – 34) prevents bone loss and augments bone formation in sexually mature ovariectomized rats. J. Bone Miner. Res. 5, 973 – 981 (1990). C. C. Liu, D. N. Kalu, E. Salerno, R. Echone, B. W. Hollis, and M. Ray, Preexisting bone loss associated with ovariectomy in rats is reversed by parathyroid hormone. J. Bone Miner. Res. 6, 1071 – 1080 (1991). Deleted in proof. B. D. Baumann and T. J. Wronski, Response of cortical bone to antiresorptive agents and parathyroid hormone in aged ovariectomized rats. Bone 16, 247 – 253 (1995). M. Gunness-Hey and J. M. Hock, Anabolic effect of parathyroid hormone on cancellous and cortical bone histology. Bone 14, 447 – 452 (1993). H. Oxlund, C. Ejersted, T. T. Andreassen, O. Torring, and M. H. Nilsson, Parathyroid hormone (1 – 34) and (1 – 84) stimulate cortical bone formation both from periosteum and endosteum. Calcif. Tissue Int. 53, 394 – 399 (1993). V. Shen, D. W. Dempster, R. W. E. Mellish, R. Birchman, W. Horbert, and R. Lindsay, Effects of combined and separate intermittent administration of low dose parathyroid hormone fragment (1 – 34) and 17-estradiol on bone histomorphometry in ovariectomized rats with established osteopenia. Calcif. Tissue Int. 50, 214 – 220 (1992). T. J. Wronski and C. F. Yen, Anabolic effects of parathyroid hormone on cortical bone in ovariectomized rats. Bone 15, 51 – 58 (1994). L. Mosekilde, C. C. Danielsen, C. H. Sogaard, J. E. McOsker, and T. J. Wronski, The anabolic effects of parathyroid hormone on cortical bone mass, dimensions and strength-assessed in a sexually mature, ovariectomized rat mode. Bone 16, 223 – 230 (1995). J. M. Hock, J. Fonseca, M. Gunness-Hey, B. E. Kemp, and T. J. Martin, Comparison of the anabolic effects of synthetic parathyroid hormone-related protein(PTHrP) 1 – 34 and PTH 1 – 34 on bone in rats. Endocrinology 125, 2022 – 2027 (1989). L. J. Fraher, K. Klein, R. Marier, D. Freeman, G. N. Hendy, D. Goltzman, and A. B. Hodsman, Comparison of the pharmacokinetics of parenteral parathyroid hormone-(1 – 34) [PTH-(1 – 34)] and PTH-related peptide-(1 – 34) in healthy young humans. J. Clin. Endocrinol. Metab. 80, 60 – 64 (1995). A. F. Stewart, PTHrP(1 – 36) as a skeletal anabolic agent for the treatment of osteoporosis. Bone 19, 303 – 306 (1996). J. F. Whitfield, P. Morley, G. E. Willick, V. Ross, R. Langille, S. MacLean, J.-R. Barbier, R. J. Isaacs, and L. Ohannessian-Barry, Comparison of the abilites of human parathyroid hormone(1– 31)NH2 and human parathyroid hormone-related protein(1– 31)NH2 to stimulate femoral trabecular bone growth in ovariectomized rats. Calcif. Tissue Int. 61, 322– 326 (1997). L. J. Fraher, A. B. Hodsman, and K. Jonas, A comparison of the in vivo biochemical responses to exogenous parathyroid hormone(1 – 34) [PTH-(1 – 34)] and PTH-related peptide-(1 – 34) in man. J. Clin. Endocrinol. Metab. 75, 417 – 423 (1995). E. C. Weir, G. Terwilliger, L. Sartori, and K. L. Insogna, Synthetic parathyroid hormone-like protein (1 – 74) is anabolic for bone in vivo. Calcif. Tissue Int. 51, 30 – 34 (1992). T. B. Usdin, C. Gruber, and T. I. Bonner, Identification and functional expression of a receptor selectively recognizing parathyroid hormone, the PTH2 receptor. J. Biol. Chem. 270, 15455 – 15458 (1995). V. Behar, M. Pines, C. Nakamoto, Z. Greenberg, A. Bisello, S. M. Stueckle, R. Besalle, T. B. Usdin, M. Chorev, M. Rosenblatt, and L. J. Suva, The human PTH2 receptor: Binding and signal transduction properties of the stably expressed recombinant receptor. Endocrinology 137, 2748 – 2757 (1996). B. H. Vickery, Z. Avnur, Y. Cheng, S.-S. Chiou, D. Leaffer, J. P. Caulfield, D. B. Kimmel, T. Ho, and L. Krstenansky, RS-66271,

808

44.

45.

46. 47.

48.

49.

50.

51.

52.

53. 54. 55.

56.

57.

58.

59.

60.

61.

STOCH, CHOREV, AND ROSENBLATT a C-terminally substituted analog of human parathyroid hormonerelated protein (1 – 34), increases trabecular and cortical bone in ovariectomized, osteopenic rats. J. Bone Miner. Res. 11, 1943 – 1951 (1996). D. W. Dempster, F. Cosman, M. Parisien, V. Shen, and R. Lindsay, Anabolic actions of parathyroid hormone on bone. Endocr. Rev. 14, 690 – 709 (1993). H. Dobnig and R. T. Turner, The effects of programmed administration of human parathyroid hormone fragment (1 – 34) on bone histomorphometry and serum chemistry in rats. Endocrinology 138, 4607 – 4612 (1997). Deleted in proof. J. S. Finkelstein, A. Klibanski, E. H. Schaefer, M. D. Hornstein, I. Schiff, and R. M. Neer, Parathyroid hormone for the prevention of bone loss induced by estrogen deficiency. N. Engl. J. Med. 331, 1618 – 1623 (1994). R. Lindsay, J. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, D. Dempster, and F. Cosman, Randomised controlled study of effect of parathyroid hormone on vertebral bone mass and fracture incidence among postmenapausal women on oestrogen with osteoporosis. Lancet 350, 550 – 555 (1997). A. B. Hodsman, L. J. Fraher, T. Ostbye, J. D. Adachi, and B. M. Steer, An evaluation of several biochemical markers for bone formation and resorption in a protocol utilizing cyclical parathyroid hormone and calcitonin therapy for osteoporosis. J. Clin. Invest. 71, 1138 – 1148 (1993). T. Sone, M. Fukunaga, S. Ono, and T. Nishiyama, A small dose of human parathyroid hormone (1 – 34) increased bone mass in the lumbar vertebrae in patients with senile osteoporosis. Miner. Electrolyte Metab. 21, 232 – 235 (1995). D. W. Dempster, F. Cosman, M. Parisien, V. Shen, and R. Lindsay, Anabolic actions of parathyroid hormone on bone: Update 1995. Endocr. Rev. 4, 247 – 250 (1995). J. F. Whitfield and P. Morley, Small bone-building fragments of parathyroid hormone: New therapeutic agents for osteoporosis. TIPS 16, 382 – 386 (1995). P. Morley, J. F. Whitfield, and G. E. Willick, Anabolic effects of parathyroid hormone on bone. TEM 8, 225 – 231 (1997). J. Reeve, PTH: A future role in the management of osteoporosis? J. Bone Miner. Res. 11, 440 – 445 (1996). R. N. Margolis, E. Canalis, and N. C. Partridge, Invited review of a workshop: anabolic hormones in bone: Basic research and therapeutic potential. J. Clin. Endocrinol. Metab. 81, 872 – 877 (1996). J. S. Thomsen, L. Mosekilde, and J. A. Gasser, Long-term therapy of ovariectomy-induced osteopenia with parathyroid hormone analog SDZ PTS 893 and bone maintenance in retired breeder rats. Bone 25, 561 – 569 (1999). K. A. McCuaig, J. C. Clarke, and J. H. White, Molecular cloning of the gene encoding the mouse parathyroid hormone/parathyroid hormone-related peptide receptor. Proc. Natl. Acad. Sci. USA 91, 5051 – 5055 (1994). H. Schneider, J. H. Feyen, K. Seuwen, and N. R. Movva, Cloning and functional expressionof a human parathyroid hormone receptor. Eur. J. Pharmacol. 246, 149 – 155 (1993)., D. P. Smith, X. Y. Zhang, C. A. Frolik, A. Harvey, S. Chandrasekhar, and E. C. Black, Structure and functional expression of a complementary DNA for porcine parathyroid hormone/parathyroid hormone-related peptide receptor. Biochim. Biophys. Acta 1307, 339 – 347 (1996). K. Lee, J. D. Deeds, and G. V. Segre, Expression of parathyroid hormone-related peptide and its receptor messenger ribonucleic acids during fetal development of rats. Endocrinology 136, 453 – 463 (1995). A.-B. Abou-Samra, H. Juppner, T. Force, M. W. Freeman, X.-F. Kong, E. Schipani, P. Urena, J. Richards, J. V. Bonventre, J. T.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

Potts Jr., H. M. Kronenberg, and G. V. Segre, Expression cloning of a common receptor for parathyroid hormone and parathyroid hormone-related peptide from rat osteoblast-like cells: A single receptor stimulates intracellular accumulation of both cAMP and inositol triphosphates and increases intracellular free calcium. Proc. Natl. Acad. Sci. USA 89, 2732 – 2736 (1992). H. Juppner, A.-B. Abou-Samra, M. Freeman, X. F. Kong, E. Schipani, J. Richards Jr., Kolakowski, L. F. Hock, J., J. T. Potts Jr., Kronenberg, H. M., and Segre, G. V., A. G. protein-linked receptor for parathyroid hormone and parathyroid hormone-related protein. Science 254, 1024 – 1026 (1991). F. R. Bringhurst, H. Jueppner, J. Guo, P. Urena, J. T. Potts Jr., H. M. Kronenberg, A. B. Abou-Samra, and G. V. Segre, Cloned, stably expressed parathyroid hormone (PTH)/PTH-related peptide receptors activate multiple messenger signals and biological responses in LLC-PK1 kidney cells. Endocrinology 132, 2090 – 2098 (1993). M. Pines, A. E. Adams, S. Stueckle, R. Bessalle, V. Rashti-Behar, M. Chorev, M. Rosenblatt, and L. J. Suva, Generation and characterization of human kidney cell lines stably expressing recombinant human PTH/PTHrP receptor: Lack of interaction with a C-terminal human PTH peptide. Endocrinology 135, 1713 – 1716 (1994). A. Fujimori, S. Cheng, L. V. Avioli, and R. Civitelli, Dissociation of second messenger activation by parathyroid hormone fragments in osteosarcoma cells. Endocrinology 128, 3032 – 3029 (1991). H. Jouishomme, J. F. Whitfield, B. Chakravarthy, J. P. Durkin, L. Gagnon, R. J. Isaacs, S. MacLean, W. Neugebauer, G. Willick, and R. H. Rixon, The protein kinase-C activation domain of the parathyroid hormone. Endocrinology 130, 53 – 60 (1992). H. Jouishomme, J. F. Whitfield, L. Gagnon, S. Maclean, R. Isaacs, B. Chakravarthy, J. Durkin, W. Neugebauer, G. Willick, and R. H. Rixon, Further definition of the protein kinase C activation domain of the parathyroid hormone. J. Bone Miner. Res. 9, 943 – 949 (1994). D. Somjen, I. Binderman, K.-D. Schluter, E. Wingender, H. Mayer, and A. M. Kaye, Stimulation by defined parathyroid hormone fragments of cell proliferation in skeletal-derived cell cultures. Biochemistry 272, 781 – 785 (1990). K. D. Schluter, H. Hellstern, E. Wingender, and H. Mayer, The central part of parathyroid hormone stimulates thymidine incorporation of chondrocytes. J. Biol. Chem 264, 11087 – 11092 (1989). R. H. Rixon, J. F. Whitfield, L. Gagnon, R. J. Isaacs, S. MacLean, B. Chakravarthy, J. P. Durkin, W. Neugebauer, V. Ross, W. Sung, and G. E. Willick, Parathyroid hormone fragments may stimulate bone growth in ovariectomized rats by activating adenylyl cyclase. J. Bone Miner. Res. 9, 1179 – 1189 (1994). G. W. Tregear, J. van Rietschoten, E. Greene, H. T. Keutmann, H. D. Niall, B. Reit, J. A. Parsons, and J. T. Potts J., Bovine parathyroid hormone: Minimum chain length of synthetic peptide required for biological activity. Endocrinology 93, 1349 – 1353 (1973). I. R. Reid, R. Civitelli, and L. V. Avioli, Parathyroid hormone depresses cytosolic pH and DNA synthesis in osteoblast-like cells. Am. J. Physiol. 255, E9 – E15 (1988). J. Kano, T. Sugimoto, M. Fukase, and T. Fujita, The activation of cAMP-dependent proteinkinase is directly linked to the inhibition of osteoblast proliferation (UMR-106) by parathyroid-hormone related. Biochem. Biophys. Res. Commun. 179, 97 – 101 (1991). M. Sabatini, C. Lesur, M. Pacherie, P. Pastoreau, N. Kucharczyk, J. Fauchere, and J. Bonnet, Effects of parathyroid hormone and agonists of the adenylyl cyclase and protein kinase C pathways on bone cell proliferation. Bone 18, 59 – 65 (1996). B. R. McDonald, J. A. Gallagher, and R. G. G. Russell, Parathyroid hormone stimulates the proliferation of cells derived from human bone. Endocrinology 118, 2445 – 2449 (1986). A. Van der Plas, J. H. M. Feyen, and P. J. Nijweide, Direct effect of parathyroid hormone on the proliferation of osteoblast-like cells: A

809

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

possible involvement of cyclic AMP. Biochem. Biophys. Res. Commun. 129, 918 – 925 (1985). R. D. Finkelman, S. Mohan, T. A. Linkhart, S. M. Abraham, J. P. Boussy, and D. J. Baylink, PTH stimulates the proliferation of TE85 human osteosarcoma cells by a mechanism not involving either increased cAMP or increased secretion of IGF-I, IGF-II, or TGF-b. Bone Miner. 16, 89 – 100 (1992). A. Fujimori, S. Cheng, L. V. Avioli, and R. Civitelli, Structure – function relationship of parathyroid hormone: Activation of phospholipase C, protein kinase A and C in osteosarcoma cells. Endocrinology 130, 29 – 36 (1992). B. R. Chakravarthy, J. P. Durkin, R. H. Rixon, and J. F. Whitfield, Parathyroid hormone fragment [3 – 34] stimulates protein kinase C (PKC) activity in rat osteosarcoma and murine T-lymphoma cels. Biochem. Biophys. Res. Commun. 171, 1105 – 1110 (1990). D. Somjen, K. D. Schluter, Z. Wingender, H. Mayer, and A. M. Kayer, Stimulation of cell proliferation in skeletal tissue of the rat by defined parathyroid hormone fragments. Biochem. J. 277, 863 – 868 (1991). H. Kaji, T. Sugimoto, M. Kantatani, and M. Fukase, The activation of cAMP-dependent protein kinase is directly linked to the stimulation of bone resorption by parathyroid hormone. Biochem. Biophys. Res. Commun. 182, 1356 – 1361 (1992). M. P. M. Herrmann-Erlee, J. M. van der Meer, C. W. G. M. Lowik, J. P. T. M. van Leeuwen, and P. M. Boonekamp, Different roles for calcium and cyclic AMP in the action of PTH: Studies in bone explants and isolated bone cells. Bone 9, 93 – 100 (1988). U. H. Lerner, M. Ransjo, O. Ljungren, K. Klaushofer, O. Hoffmann, and M. Peterlik, On the role of cyclic AMP as a mediator of bone resorption: -Interferon completely inhibits cholera toxin- and forskolin-induced but only partially inhibits parathyroid hormonestimulated 45Ca release from mouse calvarial bones. J. Bone Miner. Res. 6, 551 – 560 (1991). A. Azarani, D. Goltzman, and J. Orlowski, Structurally diverse N-terminal peptides of parathyroid hormone (PTH) and PTHrelated peptide (PTHrP) inhibit the Na/H exchanger NHE3 isoform by binding to the PTH/PTHrP receptor type I and activating distinct signaling pathways. J. Biol. Chem. 271, 14931 – 14936 (1996). M. Janulis, M. S. Wong, and M. J. Favus, Structure – function requirements of parathyroid hormone for stimulation of 1,25dihydroxyvitamin D3 prodution by rat renal proximal tubules. Endocrinology 133, 713 – 719 (1993). G. Siegfried, F. Vrtovsnik, D. Prie, C. Amiel, and G. Friedlander, Parathyroid hormone stimulates ecto-5-nucleotidase activity in renal epithelial cells: Role of protein kinase-C. Endocrinology 136, 1267 – 1275 (1995). H. Takasu, J. Guo, and F. R. Bringhurst, Dual signaling and ligand selectivity of the human PTH/PTHrP receptor. J. Bone Miner. Res. 14, 11 – 20 (1999). O. K. Olstad, S. Reppe, O. P. Loseth, R. Jemtland, and K. M. Gautvik, Binding and cyclic AMP stimulation by N-terminally deleted human PTH’s (3 – 84 and 4 – 84) in a homologous ligand receptor system. J. Bone Miner. Res. 12, 1348 – 1357 (1997). J. F. Whitfield, P. Morley, G. E. Willick, V. Ross, J.-R. Barbier, R. J. Isaacs, and L. Ohannessian-Barry, Stimulation of the growth of femoral trabecular bone in ovariectomized rats by the novel parathyroid hormone fragment, hPTH-(1 – 31)NH2 (ostabolin). Calcif. Tissue Int. 81 – 87 (1996). H. Schneider, J. H. M. Feyen, and K. Seuwen, A. C-terminally truncated human parathyroid hormone receptor is functional and activates multiple G proteins. FEBS Lett. 351, 281 – 285 (1994). R. Armamento-Villareal, K. Ziambaras, S. H. Abbasi-Jarhoum, A. Dimarogonas, L. Halstead, L. V. Avioli, and R. Civitelli, An in-

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

tact N terminus is required for the anabolic action of parathyroid hormone on adult female rats. J. Bone Miner. Res. 12, 384 – 392 (1997). S. Hilliker, J. E. Wergedal, H. E. Gruber, P. Bettica, and D. J. Baylink, Truncation of the amino terminus of PTH alters Its anabolic activity on bone in vivo. Bone 19, 469 – 477 (1996). H. Takasu, T. J. Gardella, M. D. Luck, J. T. Potts, and F. R. Bringhurst, Amino-terminal modifications of human parathyroid hormone (PTH) selectively alter phospholipase C signaling via the Type 1 PTH receptor: Implications for design of signal-specific PTH ligands. Biochemistry 38, 13453 – 13460 (1999). J. E. Onyia, R. R. Miles, X. Yang, D. L. Halladay, J. Hale, A. Glasebrook, D. Mcclure, G. Seno, L. Churgay, S. Chandrasekhar, and T. J. Martin, In vivo demonstration that human parathyroid hormone 1–38 inhibits the expression of osteoprotegerin in bone with the kinetics of an immediate early gene. J. Bone Miner. Res. 15, 863 – 871 (2000). P. D. Delmas, P. Vergnaud, M. E. Arlot, P. Pastoureau, P. J. Meunier, and M. H. L. Nilssen, The anabolic effect of human PTH (1 – 34) on bone formation is blunted when bone resorption is inhibited by the bisphosphonate tiludronate-Is activated resorption a prerequisite for the in vivo effect of PTH on formation in a remodeling system? Bone 16, 603 – 610 (1995). W. Neugebauer, J. R. Barbier, W. L. Sung, J. F. Whitfield, and G. E. Willick, Solution structure and adenylyl cyclase stimulating activities of C-terminal truncated human parathyroid hormone analogues. Biochemistry 34, 8835 – 8842 (1995). J. F. Whitfield, P. Morley, L. Fraher, A. B. Hodsman, D. W. Holdsworth, P. H. Watson, G. E. Willick, J. R. Barbier, M. Gulam, R. J. Isaacs, S. MacLean, and V. Ross, The stimulation of vertebral and tibial bone growth by the parathyroid hormone fragments, hPTH-(1 – 31)NH2, [Leu27]cyclo(Glu22–Lys26)hPTh-(1 – 31)NH2, andhPTH-(1–30)NH2. Calcif. Tissue Int. 66, 307 – 312 (2000). L. J. Fraher, R. Avram, P. H. Watson, G. N. Hendy, J. E. Henderson, K. L. Chong, D. Goltzman, P. Morley, G. E. Willick, J. F. Whitfield, and A. B. Hodsman, Comparison of the biochemical responses to human parathyroid hormone-(1–31)NH2 and hPTH-(1 – 34) in healthy humans. J. Clin. Endocrinol. Metab. 84, 2739 – 2743 (1999). S. Mohan, S. Kutilek, C. Zhang, H. G. Shen, Y. Kodama, A. K. Srivastava, and D. J. Baylink, Evidence that PTH effects on bone formation may involve modulation of PKA pathway while its effects on bone resorption may involve modulation of PKC pathway in a mouse model. Bone 23, 5449 (1998). J. F. Whitfield, P. Morley, G. E. Willick, R. J. Isaacs, S. MacLean, V. Ross, J. R. Barbier, P. Divieti, and F. R. Bringhurst, Lactam formation increases receptor binding, adenylyl cyclase stimulation and bone growth stimulation by human parathyroid hormone (hPTH) (1 – 28)NH2. J. Bone Miner. Res. 15, 964 – 970 (2000). J. F. Whitfield, P. Morley, G. Willick, R. Langille, V. Ross, S. MacLean, and J.-R. Barbier, Cyclization by a specific lactam increases the ability of human parathyroid hormone (hPTH)(1–31)NH2 to stimulate bone growth in ovariectomized rats. J. Bone Miner. Res. 12, 1246 – 1252 (1997). N. E. Lane, D. B. Kimmel, M. H. L. Nilsson, F. E. Cohen, S. Newton, R. A. Nissenson, and G. Strewler, Bone-selective analogs of human PTH(1 – 34) increase bone formation in an ovariectomized rat model. J. Bone Miner. Res. 11, 614 – 625 (1996). V. Behar, A. Bisello, M. Rosenblatt, and M. Chorev, Photoaffinity cross-linking identifies differences in the interactions of an agonist and an antagonist with the parathyroid hormone/parathyroid hormone-related protein receptor. J. Biol. Chem. 275, 9 – 17 (2000). S. L. Ferrari, V. Behar, M. Chorev, M. Rosenblatt, and A. Bisello, Endocytosis of ligand-human parathyroid hormone receptor 1 complexes is protein kinase C-dependent and involves -arrestin2. J. Biol. Chem. 274, 29968 – 29975 (1999).

810 105. R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). 106. A. M. Parfitt, Bone-forming cells in clinical conditions. Bone 351 – 429 (1990). 107. P. R. Turner, S. Mefford, S. Christakos, and R. A. Nissenson, Apoptosis mediated by activation of the G protein-coupled receptor for parathyroid hormone (PTH)/PTH-related protein (PTHrP). Mol. Endocrinology 14, 241 – 254 (2000). 108. M. Chorev, E. Roubini, R. L. McKee, S. W. Gibbons, M. E. Goldman, M. P. Caulfield, and M. Rosenblatt, Cyclic parathyroid hormone related protein antagonists: Lysine 13 to aspartic acid 17 [i to (i + 4)] side chain to side chain lactamization. Biochemistry 30, 5968 – 5974 (1991). 109. A. Bisello, C. Nakamoto, M. Rosenblatt, and M. Chorev, Mono- and bicyclic analogs of parathyroid hormone-related protein. 1. Synthesis and biological studies. Biochemistry 36, 3293 – 3299 (1997). 110. M. Chorev, R. F. Epand, M. Rosenblatt, M. P. Caulfield, and R. M. Epand, Circular dichroism (CD) studies of antagonists derived from parathyroid hormone-related protein. Int. J. Pept. Protein Res. 42, 342 – 345 (1993). 111. D. F. Mierke, S. Maretto, E. Schievano, D. DeLuca, A. Bisello, S. Mammi, M. Rosenblatt, E. Peggion, and M. Chorev, Conformational studies of mono- and bicyclic parathyroid hormone-related protein-derived agonists. Biochemistry 36, 10372 – 10383 (1997). 112. J.-R. Barbier, W. Neugebauer, P. Morley, V. Ross, M. Soska, J. F. Whitfield, and G. Willick, Bioactivities and secondary structures of constrained analogues of human parathyroid hormone: Cyclic lactams of the receptor binding region. J. Med. Chem. 40, 1373 – 1380 (1997). 113. W. K. Surewicz, W. Neugebauer, L. Gagnon, S. MacLean, J. F. Whitfield, and G. Willick, “Peptides, Chemical Structure and Biology” (R. S. Hodges, and J. A. Smith, eds.), pp. 556 – 561. ESCOM, Leiden (1994). 114. D. Leaffer, M. Sweeney, L. A. Kellerman, Z. Avnur, J. L. Krstenansky, B. H. Vickery, and J. P. Caulfield, Modulation of osteogenic cell ultrastructure by RS-23581, an analog of human parathyroid hormone (PTH)-related peptide-(1 – 34), and bovine PTH-(1 – 34). Endocrinology 136, 3624 – 3631 (1995). 115. W. Neugebauer, W. K. Surewicz, H. L. Gordon, R. Oberstein, L. Somorjai, W. Sung, and G. E. Willick, Structural elements of human parathyroid hormone and their possible relation to biological activities. Biochemistry 31, 2056 – 2063 (1992). 116. T. J. Gardella, A. K. Wilson, H. T. Keutmann, R. Oberstrin, J. T. Potts Jr., H. M. Kronenberg, and S. R. Nussbaum, Analysis of parathyroid hormone’s principal receptor-binding region by site-directed mutagenesis and analog design. Endocrinology 132, 2024 – 2030 (1993). 117. J. L. Krstenansky, T. J. Owen, K. A. Hagaman, and L. R. McLean, Short model peptides having a high -helical tendency: Design and solution properties. FEBS Lett. 242, 409 – 413 (1989). 118. M. Pellegrini, A. Bisello, M. Rosenblatt, M. Chorev, and D. F. Mierke, Conformational studies of RS-66271, an analog of parathyroid hormone-related protein with pronounced bone anabolic activity. J. Med. Chem. 40, 3025 – 3031 (1997). 119. J. L. Krstenansky, T. L. Ho, Z. Avnur, D. Leaffer, J. P. Caulfield, and B. H. Vickery, Molecular design and in vivo bone anabolic activity. Peptides 133 – 134 (1994). 120. C. A. Frolik, R. L. Cain, M. Sato, A. K. Harvey, S. Chandrasekhar, E. C. Black, A. H. Tashjian, and J. M. Hock, Comparison of recombinant human PTH (1 – 34) (LY333334) with a C-terminally substituted analog of human PTH-related protein(1 – 34) (RS-66271): In vitro activity and in vivo pharmacological effects in rats. J. Bone Miner. Res. 14, 163 – 172 (1999).

STOCH, CHOREV, AND ROSENBLATT 121. E. Brown, The extracellular Ca2 sensing, regulation of parathyroid cell function, and the role of Ca2 and other ions as extracellular (first) messengers. Physiol. Rev. 71, 371 – 411 (1991). 122. E. M. Brown, G. Gamba, D. Riccardi, M. Lombardi, R. Butters, O. Kifor, A. Sun, M. A. Hediger, J. Lytton, and S. C. Hebert, Cloning and characterization of an extracellular Ca2-sensing receptor from bovine parathyroid. Nature 366, 575 – 580 (1993). 123. Deleted in proof. 124. M. Wada, Y. Furuya, H. Ishii, J. Fox, E. Nemeth, and N. Nagano, Daily intermittent decreases in serum levels of parathyroid hormone have an anabolic-like action on the bones of uremic rats with lowturnover bone and osteomalacia. Bone 26, 175 – 182 (2000). 125. E. F. Nemeth, J. Fox, E. G. DelMar, M. E. Steffey, L. D. Lambert, R. L. Conklin, P. K. Bhatnagar, and M. Gowen, Stimulation of parathyroid hormone secretion by a small molecule antagonist of the calcium receptor. Bone 23, S156 (1998). 126. M. Gowen, G. B. Stroup, J. Bradbeer, A. Dodds, S. J. Hoffman, J. Vasco-Moser, B. Lechowska, X. Liang, P. Bhatmagar, B. R. Smith, E. G. DelMar, E. F. Nemeth, and J. Fox, An antagonist of the parathyroid cell Ca2 receptor stimulates PTH secretion and bone turnover in osteopenic ovariectomized rats. Bone 23, S163 (1998). 127. J. E. Garrett, I. V. Capuano, L. G. Hammerland, B. C. P. Hung, E. M. Brown, S. C. Hebert, E. F. Nemeth, and F. Fuller, Molecular cloning and functional expression of human parathyroid calcium receptor cDNAs. J. Biol. Chem. 270, 12919 – 12925 (1995). 128. E. F. Nemeth, Calcium receptors as novel drug targets. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 1019 – 1036. Academic Press, San Diego, 1996. 129. A. Means and J. Dedman, Calmodulin — An intracellular calcium receptor. Nature 285, 73 – 77 (1980). 130. E. M. Brown, M. Pollak, C. E. Seidman, J. G. Seidman, Y. Chou, D. Riccardi, and S. C. Hebert, Calcium-ion-sensing cell-surface receptors. N. Engl. J. Med. 333, 234 – 240 (1995). 131. E. Brown, C. Katz, R. Butters, and O. Kifor, Polyarginine, polylysine, and protamine mimic the effects of high extracellular calcium concentrations on dispersed bovine parathyroid cells. J. Bone Miner. Res. 6, 1217 – 1225 (1991). 132. E. Brown, R. Butters, C. Katz, and O. Kifor, Neomycin mimics the effects of high extracellular calcium concentrations on parathyroid function in dispersed bovine parathyroid cells. Endocrinology 128, 3047 – 3054 (1991). 133. S. J. Silverberg and T. Marriott, Short-term inhibition of parathyroid hormone secretion by a calcium-receptor agonist in patients with primary hyperparathyroidism. N. Engl. J. Med. 337, 1506 – 1510 (1997). 134. H. R. Sauls, B. Han, J. Triantafillou, D. C. Morris, and P. G. Baer, Plasma PTH levels after minimal doses causing increased bone mineral density in osteopenic rats. J. Bone Miner. Res. 9, S325 (1994). 135. J. Fox, M. A. Miller, G. B. Stroup, E. F. Nemeth, and S. C. Miller, Plasma levels of parathyroid hormone that induce anabolic effects in bone of ovariectomized rats can be achieved by stimulation of endogenous hormone secretion. Bone 21, 163 – 169 (1997). 136. M. Gowen, G. B. Stroup, R. A. Dodd, I. E. James, B. J. Votta, B. R. Smith, P. K. Bhatnagar, A. M. Lago, J. F. Callahan, E. G. DelMar, M. A. Miller, E. F. Nemeth, and J. Fox, Antagonizing the parathyroid calcium receptor stimulates parathyroid hormone secretion and bone formation in osteopenic rats. J. Clin. Invest. 105, 1595 – 1604 (2000). 137. G. Mundy, R. Garret, S. Harris, J. Chan, D. Chen, G. Rossini, B. Boyce, M. Zhao, and G. Gutierrez, Stimulation of bone formation in vitro and in rodents by statins. Science 286, 946 – 949 (1999). 138. D. Chen, M. A. Harris, G. Rossini, C. R. Dunstan, S. L. Dallas, J. Q. Feng, G. R. Mundy, and S. E. Harris, Bone morphogenetic protein 2 (BMP-2) enhances MP-3, BMP-4, and bone cell differentiation marker gene expression during the induction of mineralized

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150. 151.

152. 153. 154. 155. 156.

156a.

bone matrix formation in cultures of fetal rat calcarial osteoblasts. Calcif. Tissue Int. 60, 283 – 290 (1997). S. P. Luckman, D. E. Hughes, F. P. Coxon, R. Graham, G. Russell, and M. J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including ras. J. Bone Miner. Res. 13, 581 – 589 (1998). S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. Dallas, J. Q. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of transforming growth factor  on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, osteopontin, alkaline phosphatase, and type I collagen mRNA in long-term cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). D. C. Bauer, G. R. Mundy, S. A. Jamal, D. M. Black, J. A. Cauley, F. Harris, T. Duong, and S. R. Cummings, Statin use, bone mass and fracture: An analysis of two prospective studies. J. Bone Miner. Res. 14, S179 (1999). Y.-S. Chung, M.-D. Lee, S.-K. Lee, H.-M. Kim, and L. A. Fitzpatrick, HMG-CoA reductase inhibitors increase BMD in Type 2 diabetes mellitus patients. J. Clin. Endocrinol. Metab. 85, 1137 – 1142 (2000). K. A. Chan, S. E. Andrade, M. Boles, D. S. M. Buist, J. G. Donahue, M. J. Goodman, J. H. Gurwitz, A. Z. LaCroix, and R. Platt, Inhibitors of hydroxymethylglutaryl-coenzyme. A reductase and risk of fracture among older women. Lancet 355, 2185 – 2188 (2000). C. R. Meier, R. G. Schlienger, M. E. Kraenzlin, B. Schlegel, and H. Jick, HMG-CoA reductase inhibitors and the risk of fractures. JAMA 283, 3205 – 3210 (2000). P. S. Wang, D. H. Solomon, H. Mogun, and J. Avorn, HMG-CoA reductase inhibitors and the risk of hip fractures in elderly patients. JAMA 283, 3211 – 3216 (2000). J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J. Masarachia, G. Wesolowski, R. G.G. Russell, G. A. Rodan, and A. A. Reszka, Alendronate mechanism of action: Geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc. Natl. Acad. Sci. USA 96, 133 – 138 (1999). J. D. Bergstrom, R. G. Bostedor, P. J. Masarachia, A. A. Reszka, and G. Rodan, Alendronate is a specific, nanomolar inhibitor of farnesyl diphosphate synthase. Arch. Biochem. Biophys. 373, 231 – 241 (2000). V. Braga, D. Gatti, D. Gerardi, and S. Adami, Chronic I. V. aminobisphosphonate therapy increases HDL cholesterol and decreases LDL cholesterol. Bone 23, S405 (2000). A. A. Reszka, J. M. Halasy-Nagy, P. J. Masarachia, and G. A. Rodan, Bisphosphonates act directly on the osteoclast to induce caspase cleavage of Mst1 Kinase during apoptosis. J. Biol. Chem. 274, 34967 – 34973 (1999). E. Ruoslahti, Integrins. J. Clin. Invest. 87, 1 – 5 (1991). N. Wang, J. P. Butler, and D. E. Ingber, Mechanotransduction across the cell surface and through the cytoskeleton. Science 260, 1124 – 1127 (1993). E. A. Clark and J. S. Brugge, Integrins and signal transduction pathways: The road taken. Science 268, 233 – 239 (1995). E. Ruoslahti, N. A. Noble, S. Kagami, and W. A. Border, Integrins. Kidney Int. Suppl. 44, S17 – S22 (1994). F. G. Giancotti and F. Mainiero, Integrin-mediated adhesion and signaling in tumorigenisis. Biochim. Biophys. 1198, 47 – 64 (1994). D. Cox, T. Aoki, J. Seki, Y. Motoyama, and K. Yoshida, The pharmacology of the integrins. Med. Res. Rev. 14, 195 – 228 (1994). R. C. Cherny, M. A. Honan, and P. Thiagarajan, Site-directed mutagenesis of the arginine-glycine-aspartic acid in vitronectin abolishes cell adhesion. J. Biol. Chem. 268, 9725 – 9729 (1993). R. O. Hynes, Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69, 11 – 25 (1992).

811 157. J. Fisher, M. Caulfield, and M. Sato, Inhibition of osteoclastic bone resorption in vivo by echistatin, an “arginyl-glycyl-aspartyl” (RGD)-containing protein. Endocrinology 132, 1411 – 1413 (1993). 158. E. Ruoslahti, RGD and other recognition sequences for integrins. Annu. Rev. Cell. Dev. Biol. 12, 697 – 715 (1996). 159. E. Ruoslahti and M. D. Pierschbacher, New perspectives in cell adhesion: RGD and integrins. Science 238, 491 – 497 (1987). 160. Deleted in proof. 161. H. K. Vaananen and M. Horton, The osteoclast clear zone is a specialized cell-extracellular matrix adhesion structure. J. Cell. Sci. 108, 2729 – 2732 (1995). 162. C.-T. Leu, G. Wesolowski, and R. Nagy, Osteoclasts have over 107 high affinity echistatin binding sites (RGD-Integrins). J. Bone Miner. Res. 12, S416 (1997). 163. M. A. Horton, M. L. Taylor, T. R. Arnett, and M. H. Helfrich, ArgGly-Asp (RGD) peptides and the anti-vitronectin receptor antibody 23C6 inhibit dentine resorption and cell spreading by osteoclasts. Exp. Cell Res. 195, 368 – 375 (1991). 164. B. Crippes, V. Engleman, S. Settle, J. Delarco, R. Ornberg, M. Helfrich, M. Horton, and A. Nickols, Antibody to 3 integrin inhibits osteoclast-mediated Bone Resorption in the thyroparathyroidectomized rat. Endocrinology 137, 918 – 924 (1996). 165. J. E. Fisher, M. P. Caulfield, M. Sato, H. A. Quartuccio, R. J. Gould, V. M. Garsky, G. A. Rodan, and M. Rosenblatt, Inhibition of osteoclastic bone resorption in vivo by echistatin, an “arginyl-glycyl-aspartyl” (RGD)-containing protein. Endocrinology 132, 1411 – 1413 (1993). 166. M. Yamamoto, J. E. Fisher, M. Gentile, J. G. Seedor, C. T. Leu, S. B. Rodan, and G. A. Rodan, The integrin ligand echistatin prevents bone loss in ovariectomized mice and rats. Endocrinology 139, 1411 – 1419 (1998). 167. P. Masarachia, M. Yamamoto, C.-T. Leu, G. Rodan, and L. Duong, Histomorphometric evidence for echistatin inhibition of bone resorption in mice with secondary hyperparathyroidism. Endocrinology 139, 1401 – 1410 (1998). 168. M. Sato, M. K. Sardana, W. A. Grasser, V. M. Garsky, J. M. Murray, and R. J. Gould, Echistatin is a potent inhibitor of bone resorption in culture. J. Cell Biol. 111, 1713 – 1723 (1990). 169. M. A. Horton, E. L. Dorey, S. A. Nesbitt, J. Samanen, F. E. Ali, J. M. Stadel, A. Nichols, R. Greig, and M. H. Helfrich, Modulation of vitronectin receptor-mediated osteoclast adhesion by Arg-GlyAsp peptide analogs: A structure-function analysis. J. Bone Miner. Res. 8, 239 – 247 (1993). 170. M. Sato, V. Garsky, R. J. Majeska, T. A. Einhorn, J. Murray, A. H. Tashjian Jr., and R. J. Gould, Structure–activity studies of the s-echistatin inhibition of bone resorption. J. Bone Miner. Res. 9, 1441 – 1449 (1994). 171. R. M. Scarborough, J. W. Rose, M. A. Naughton, D. R. Phillips, L. Nannizzi, A. Arfsten, A. M. Campbell, and I. F. Charo, Characterization of the integrin specificities of disintegrins isolated from American pit viper venoms. J. Biol. Chem. 268, 1058 – 1065 (1993). 172. M. Pfaff, K. Tangemann, B. Muller, M. Gurrath, G. Muller, H. Kessler, R. Timpl, and J. Engel, Selective recognition of cyclic RGD peptides of NMR defined conformation by IIb3, v3, 51 integrins. J. Biol. Chem. 269, 20233 – 20238 (1994). 173. R. Haubner, D. Finsinger, and H. Kessler, Stereoisomeric Peptide Libraries and Peptidomimetics for Designing Selective Inhibitors of the v3 Integrin for a New Cancer Therapy. Angew. Chem. Int. Ed. Engl. 36, 1374 – 1389 (1997). 174. M. Gurrath, G. Muller, H. Kessler, M. Aumailley, and R. Timpl, Conformation/activity studies of rationally designed potent antiadhesive RGD peptides. Eur. J. Biochem. 210, 911 – 921 (1992). 175. R. Haubner, R. Gratias, B. Diefenbach, S. L. Goodman, A. Jonczyk, and H. Kesser, Structural and functional aspects of RGD-containing cyclic pentapeptides as highly potent and selective integrin v3 antagonists. J. Am. Chem. Soc. 118, 7461 – 7472 (1996).

812 176. M. A. Dechantsreiter, E. Planker, and B. Matha, N-methylated cyclic RGD peptides as highly active and selective v3 integrin antagonists. J. Med. Chem. 42, 3033 – 3040 (1999). 177. A. Bach, J. Espina, S. Jackson, P. Stouton, J. Duke, S. Mousa, and W. DeGrado, Type II’ to Type I -Turn swap change specifity for integrins. J. Am. Chem. Soc. 118, 293 – 294 (1996). 178. K. Burgess, D. Lim, and S. A. Mousa, Synthesis and solution conformation of cyclo[RGDRGD]: A cyclic peptide with selectivity for the v3 receptor. J. Med. Chem. 39, 4520 – 4526 (1996). 179. W. H. Miller, D. P. Alberts, and P. K. Bhatnagar, Discovery of orally active nonpeptide vitronectin receptor antagonists based on a 2–benzazepine gly-asp mimetic. J. Med. Chem. 43, 22 – 26 (2000). 180. M. W. Lark, G. B. Stroup, and S. M. Hwang, Design and characterization of orally active Arg-Gly-Asp peptidomimetic vitronectin receptor antagonist SB 265123 for prevention of bone loss in osteoporosis. J. Pharm. Exp. 291, 612 – 617 (1999). 181. M. W. Lark, G. Stroup, and R. D. Cousins, An orally active vitronectin receptor v3 antagonist prevents bone resorption in vitro and in vivo in the ovariectomized rat. J. Bone Miner. Res. 23, S219 (1998). 182. M. E. Duggan, “Design and Evaluation of Potent Non-peptide Ligands of v3 as Inhibitors of Bone Resorption.” 211th National ACS Meeting, 1996. 183. M. E. Duggan, L. T. Duong, and J. E. Fisher, Non-peptide v3 antagonists. I. transformation of a potent, integrin selective iib3 antagonist into the potent v3 antagonist L-767,635. J. Med. Chem. 43, 3736 – 3745 (2000). 184. V. W. Engleman, G. A. Nickols, F. P. Ross, M. A. Horton, D. W. Griggs, S. L. Settle, P. G. Ruminski, and S. L. Teitelbaum, A peptidomimetic antagonist of the v3 integrin inhibits bone resorption in vitro and prevents osteoporosis in vivo. J. Clin. Invest. 99, 2284 – 2292 (1997). 185. P. G. Ruminski, T. E. Rogers, J. E. Rico, G. A. Nockols, V. W. Engleman, S. L. Settle, J. L. Keene, S. K. Freeman, C. P. Carron, and D. M. Meyer, Potent selective peptidomimetic antagonists of the v3 integrin and their in vivo efficacy in models of osteoporosis and cancer. In “217th ACS Annual Meeting, Anaheim, CA March 21 – 25.” 1999. 186. A. Peyman, V. Wehner, and J. Knolle, RGD mimetics containing a central hydantoin scaffold: v3 vs. iib3 selectivity requirements. Bioorg. Med. Chem. Lett. 10, 179 – 182 (2000). 187. W. S. Simonet, D. Lacey, C. R. Dunstan, M. Kelley, M.-S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliot, A. Colombero, H.-L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, S. Sander, G. Van, J. Tarpley, P. Derby, R. Lee, Amgen EST Program, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 188. H. Yasuda, N. Shima, N. Nakagawa, S. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998) 189. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 190. A. Tomoyasu, M. Goto, and N. Fujise, Characterization of monomeric and homodimeric forms of osteoclastogenesis inhibitory factor. Biochem. Biophys. Res. Commun. 245, 382 – 287 (1998). 191. L. C. Hofbauer, C. R. Dunstan, T. C. Spelsberg, B. L. Riggs, and S. Khosla, Osteoprotegerin production by human osteoblast lineage cells is stimulated by vitamin D, bone morphogenetic protein-2, and cytokines. Biochem. Biophys. Res. Commun. 250, 776 – 781 (1998).

STOCH, CHOREV, AND ROSENBLATT 192. K. Yamaguchi, M. Kinosaki, M. Goto, F. Kobayashi, E. Tsuda, T. Morinaga, and K. Higashio, Characterization of structural domains of human osteoclastogenesis inhibitory factor. J. Biol. Chem. 273, 5117 – 5123 (1998). 193. K. Tan, J. Harrop, and M. Reddy, Characterization of a novel TNFlike ligand and recently described TNF ligand and TNF receptor superfamily genes and their constitutive and inducible expression in hematopoietic and non-hematopoeitic cells. Gene 204, 35 – 46 (1997). 194. B. Kwon, S. Wang, and N. Udagawa, TR1, a new member of the tumor necrosis factor receptor family, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 195. H. Brandstrom, K. B. Jonsson, O. Vidal, S. Ljunghall, C. Ohlsson, and O. Ljunggren, Tumor necrosis factor- and - upregulate the levels of osteoprotegerin mRNA in human osteosarcoma MG-63 cells. Biochem. Biophys. Res. Commun. 248, 454 – 457 (1998). 196. N. Vidal, H. Brandstrom, K. B. Jonsson, and C. Ohlsson, Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: down-regulation by glucocorticoids. J. Endocrinol. 159, 191 – 195 (1998). 197. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370 (1999a). 198. T. Murakami, M. Yamomoto, and K. Yamamoto, Transforming growth factor-B1 increases mRNA levels of osteoclastogenesis inhibitory factor in osteoblastic/stromal cells and inhibits the survival of murine osteoclast-like cells. Biochem. Biophys. Res. Commun. 252, 747 – 752 (1998). 199. H. Takai, M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, and Y. Yamada, Transforming growth factor- stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells. J. Biol. Chem. 273, 27091 – 27096 (1998). 200. O. Vidal, K. Sjogren, B. I. Eriksson, O. Ljunggren, and C. Ohlsson, Osteoprotegerin mRNA is increased by interleukin-1 in the human osteosarcoma cell line MG-63 and in human osteoblast-like cells. Biochem. Biophys. Res. Commun. 248, 696 – 700 (1998). 201. H. Brandstrom, K. B. Jonsson, C. Ohlsson, O. Vidal, S. Ljunghall, and O. Ljunggren, Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 247, 338 – 341 (1998). 202. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 203. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 204. T. L. Burgess, Y. X. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145, 527 – 538 (1999). 205. T. Akatsu, T. Murakami, and M. Nishikawa, Osteoclastogenesisinhibitory factor suppresses osteoclast survival by interfering in the interaction of stromal cells with osteoclast. Biochem. Biophys. Res. Commun. 250, 229 – 234 (1998). 206. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteo-

813

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

207.

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

218.

219.

220. 221.

clastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998). B. R. Wong, J. Rho, and J. E. Aaron, TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-jun N-terminal kinase in T cells. J. Biol. Chem. 272, 25190 – 25194 (1997). M. Anderson, E. Maraskovsky, and W. Billingsley, A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390, 175 – 179 (1997). L. C. Hofbauer, D. Lacey, C. Dunstan, T. C. Spelsberg, B. L. Riggs, and K. Khosla, Interleukin-1B and tunor necrosis factor-a but not interleukin-6 stimulate osteoprotegerin ligand gene expression in human osteoblastic cells. Bone 25, 255 – 259 (1999). K. Fuller, B. R. Wong, S. Fox, Y. Choi, and T. J. Chambers, TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188, 997 – 1001 (1998). H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X.-Z. Zia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). K. Matsuzaki, N. Udagawa, and N. Takahashi, Osteoclast differentiation factor (ODF) induces osteoclast-like cell formation in human peripheral blood mononuclear cell cultures. Biochem. Biophys. Res. Commun. 246, 199 – 204 (1998). J. M. W. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes. Dev. 12, 1260 – 1268 (1998). Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). N. Nakagawa, M. Kinosaki, and K. Yamaguchi, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400 (1998). S. C. Manolagas, Editorial: Cell number versus cell vigor — What really matters to a regenerating skeleton? Endocrinology 140, 4377 – 4381 (1999). L. C. Hofbauer, F. Gori, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140, 4382 – 4389 (1999). C. Dunstan, Osteoprotegerin and osteoprotegerin ligand mediate the local regulation of bone resorption. Endocrinology 10, 18 – 26 (2000). M. Saika, D. Inoue, S. Kido, and T. Matsumoto, 17B-estradiol stimulates expression of osteoprotegerin (OPG)/osteoclastogenesis in-

222.

223.

224.

225. 226.

227. 228. 229.

230.

231.

232.

233.

234.

235.

236.

237. 238. 239.

240.

hibitory factor (OCIF) by a mouse stromal cell line, ST-2. J. Bone Miner. Res. 14, S232 (1999). D. E. Hughes, A. Dai, J. Tiffee, and G. Mundy, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nat. Med. 2, 1132 – 1136 (1996). J. Robinson, B. Riggs, T. C. Spelsberg, and M. J. Oursler, Osteoclasts and transforming growth factor-B: Estrogen-mediated isoform-specific regulation of production. Endocrinology 137, 615 – 621 (1996). M. J. Oursler, C. Cortese, and P. Keeting, Modulations of transforming growth factor-B production in normal human osteoblast-like cells by 17B-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). A. Adinoff and J. Hollister, Steroid-induced fractures and bone loss in patients with asthma. N. Engl. J. Med. 309, 265 – 268 (1983). C. Gennari, B. Imbimbo, M. Mortagnani, M. Bernini, P. Nardi, and L. Avioli, Effects of predisone and deflazacort on mineral metabolism and parathyroid hormone activity in humans. Calcif. Tissue Int. 36, 245 – 252 (1984). I. R. Reid, Glucocorticoid osteoporosis — Mechanisms and management. Eur. J. Endocrinol. 137, 209 – 217 (1997). B. Lukert and L. Raisz, Glucocorticoid-induced osteoporosis: Pathogenesis and management. Ann. Intern. Med. 112, 352 – 364 (1990). S.-K. Lee and J. N. Lorenz, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. J. Clin. Endocrinol. Metab. 140, 3552 –3561 (1999). J. E. Onyia, R. R. Miles, D. L. Halladay, S. Chandrasekhar, and T. J. Martin, In vivo demonstration that parathyroid hormone (hPTH 1 – 38) inhibits the expression of osteoprotegerin (OPG) in bone with the kinetics of an immediate early gene. J. Bone Miner. Res. 14, S166 (1999). D. B. Kimmel, Animal models for in vivo experimentation in osteoporosis research. In “Osteoporosis” (R. Marcus, D. Feldman, and J. L. Kelsey, eds.), pp. 671 – 690. Academic Press, San Diego, 1996. T. Bateman, D. Lacey, V. Ferguson, C. Dunstan, R. Ayers, and S. Simske, Comparison of osteoprotegerin, pamidronate and ibandronate in the treatment of suspension disuse osteopenia. J. Bone Miner. Res. S528 (1999). S. Simske, T. Bateman, R. Ayers, C. Dunstan, V. Ferguson, and D. Lacey, Effect of osteoprotegerin, pamidronate and ibandronate on the sciatic nerve crush model for disuse osteoporosis. J. Bone Miner. Res. 14, S523 (1999). V. Ferguson, T. Bateman, D. Lacey, R. Ayers, C. Dunstan, and S. Simske, Effect of osteoprotegerin on mechanical and material properties of maturing rat femora. J. Bone Miner. Res. 14, S440 (1999). S. F. Martin, J. Sommers, A. Watson, J. D. Young, and C. Dunstan, The pharmokinetics and pharmacodynamics of OPG following single dose administration in Cynomulgus monkeys. J. Bone Miner. Res. 14, S399 (1999). S. F. Martin, J. Sommers, A. Rasmussen, P. Bekker, C. Dunstan, and J. D. Young, The pharmacokinetics of OPG following single dose subcutaneous and intravenous administration to healthy post menopausal women. J. Bone Miner. Res. 14, S523 (1999). H. Arrighi, A. Hseih, and V. Wong, Osteoprotegerin serum levels in healthy volunteers. Bone 23, S298 (1998). B. L. Riggs and L. J. Melton III, The prevention and treatment of osteoporosis. N. Engl. J. Med. 327, 620 – 627 (1992). K. Yano, E. Tsuda, N. Washida, F. Kobayashi, M. Goto, A. Harada, K. Ikeda, K. Higashio, and Y. Yamada, Immunological characterization of circulating osteoprotegrin/osteoclastogenesis inhibitory factor: Increased serum concentrations in postmenopausal women with osteoporosis. J. Bone Miner. Res. 14, 518 – 527 (1999). P. Bekker, D. Holloway, A. Nakanishi, H. Arrighi, and C. Dunstan, Osteoprotegerin (OPG) has potent and sustained anti-resorptive activity in postmenopausal women. J. Bone Miner. Res. 14, S180 (1999).

814 241. E. Bornefalk, I. Dahlen, G. Johansson, O. Ljunggren, and C. Ohlsson, Serum levels of osteoprotegerin: Effects of glucocorticoids and growth hormone. Bone 23, S486 (1998). 242. L. C. Hofbauer, B. Riggs, C. Dunstan, T. O’Brien, and S. Khosla, Cyclosporin A and glucocorticoids inhibit osteoprotegerin production in human osteoblastic and coronary artery smooth muscle cells: Potential mechanism of post-transplant osteoporosis and vascular disease. J. Bone Miner. Res. 14, S176 (1999). 243. T. Akatsu, T. Murakami, K. Ono, M. Nishikawa, E. Tsuda, S. I. Mochizuki, N. Fujise, K. Higashio, K. Motoyoshi, M. Yamamoto, and N. Nagata, Osteoclastogenesis inhibitory factor exhibits hypocalcemic effects in normal mice and in hypercalcemic nude mice carrying tumors associated with humoral hypercalcemia of malignancy. Bone 23, 495 – 498 (1998). 244. C. Capparelli, P. Kosenuik, and S. Moroney, Comparison of osteoprotegerin and pamidronate in a murine model of humoral hypercalcemia of malignancy. J. Bone Miner. Res. 14, S163 (1999). 245. P. Kosenuik, S. Moroney, C. Capparelli, D. Lacey, and C. Dunstan, Comparison of the inhibitory effects of osteoprotegerin and of APD on PThrP-mediated hypercalcemia and bone remodeling. J. Bone Miner. Res. 14, S166 (1999). 246. S. Morony, C. Capparelli, R. Lee, G. Shimamoto, T. Boone, D. L. Lacey, and C. R. Dunstan, A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1B, TNF-, PTH, PTHrP, and 1,25(OH)2D3. J. Bone Miner. Res. 14, 1478 – 1485 (1999). 247. J. A. Kanis, T. McCloskey, T. Taube, and N. O’Rourke, Rationale for the use of bisphosphonates in bone metustasis. Bone 19, 513 – 518 (1991). 248. C. N. Takeuchi, T. Y. Fukumoto, K. Yano, E. Tsudo, E. Ogata, and T. Fujita, Interactions between cancer and bone marrow cells induce osteoclast differentiation factor expression and osteoclast-like cell formation in vitro. Biochem. Biophys. Res. Commun. 267, 632– 637 (2000). 249. J. C. Sheehan and K. Umezawa, Phenacyl photosensitive blocking groups. J. Org. Chem. 38, 3771 – 3774 (1973). 250. T. D. Spector, G. M. Hall, E. V. McCloskey, and J. A. Kanis, Risk of vertebral fracture in women with rheumatoid arthritis. Br. Med. J. 306, 558 (1993). 251. H. Krogers, R. Honkanen, S. Saarikoski, and E. Alhava, Decreased axial bone mineral density in postmenopausal women with rheumatoid arthritis-A population based study. Ann. Rheum. Dis. 53, 18 – 23 (1996). 252. T. Toyoda, S. Inokuchi, S. Sato, Y. Harie, and S. Tomita, Bone loss of the radius in rheumatoid arthritis. Acta Orthop. Scand. 67, 264 – 273 (1996). 253. Y. Y. Kong, U. Feige, I. Surosi, and B. Bolan, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegrin ligand. Nature 402, 304 – 309 (1999). 254. I. Itonaga, A. Sabokbur, D. W. Murray, and N. A. Athanasou, Effect of osteoprotegerin/and osteoprotegerin ligand on osteoclast formation by arthroplasty membrane derived macrophages. Ann. Rheum. Dis. 54, 26 – 31 (2000). 255. R. Maciewicz and D. Etherington, A comparison of four cathepsins (B, L, N, and S.) with collagenolytic activity from rabbit spleen. Biochem. J. 256, 433 – 440 (1988). 256. P. Garnero, O. Borel, I. Byrjalsen, M. Ferreras, F. H. Drake, M. S. McQueney, N. T. Foged, P. D. Delmas, and J.-M. Delaisse, The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J. Biol. Chem. 273, 32347 – 32352 (1998). 257. J. Ljusberg, R. Dodds, and M. Lark, Tartrate-resistant acid phosphatase is proteolytically cleaved in vivo by cathepsin K. J. Bone Miner. Res. 14, S358 (1999). 258. B. D. Gelb, G. P. Shi, H. A. Chapman, and R. J. Desnick, Pycnodysostosis, a lysosomal disease caused by cathepsin K deficiency. Science 273, 1236 – 1238 (1996).

STOCH, CHOREV, AND ROSENBLATT 259. M. Bossard, T. A. Tomaszek, and S. Thompson, Proteolytic activity of human osteoclast cathepsin K. J. Biol. Chem. 271, 12517 – 12524 (1996). 260. R. Dodds, F. Cristiano, and J. Feild, Cathepsin K knockout mice develop osteopetrosis due to lack of full function in their osteoclasts. Bone 23, S164 (1998). 261. F. Drake, R. Dodds, and I. James, Cathepsin K, but no cathepsins B, L, or S, is abundantly expressed in human osteoclasts. J. Biol. Chem. 271, 12511 – 12516 (1996). 262. M. McQueney, B. Amegadzie, and K. D’Alessio, Autocatalytic activation of human cathepsin K. J. Biol. Chem. 272, 13955 – 13960 (1997). 263. D. Rieman, H. McClung, and R. Dodds, Biosynthesis and processing of cathespin K in cultured human osteoclasts. Bone 23, S529 (1998). 264. J. LaLonde, B. Zhao, and C. A. Janson, The crystal structure of human procathepsin K. Biochemistry 38, 862 – 869 (1999). 265. D. Rieman, H. McClung, and I. E. James, Cathepsin K processing in human osteoclasts requires phosphoinositol-3-kinase. Bone 23, S550 (1998). 266. D. Bromme, V. Breuil, L. Xia, H. Wex, L. Stein, and D. Dempster, Characterization of cathepsin K in human monocyte-derived osteoclasts. Bone 23, S226 (1998). 267. D. Bromme, K. Okamoto, B. B. Wang, and S. Biroc, Human cathepsin 02, a matrix protein-degrading cysteine protease expressed in osteoclasts. J. Biol. Chem. 271, 2126 – 2132 (1996). 268. K. Tezuka, Y. Tezuka, A. Maejima, T. Sato, K. Nemoto, H. Kamioka, Y. Hakeda, and M. Kumegawa, Molecular cloning of a possible cysteine proteinase predominantly expressed in osteoclasts. J. Biol Chem. 269, 1106 – 1109 (1994). 269. T. Inui, O. Ishibashi, T. Inaoka, Y. Origane, M. Kumegawa, T. Kokubo, and T. Yamamura, Cathepsin K antisense oligodeoxynucleotide inhibits osteoclastic bone resorption. J. Biol. Chem. 272, 8109 – 8112 (1997). 270. T. Inaoka, G. Bilbe, O. Ishibashi, K. Tezuka, M. Kumegawa, and T. Kokubo, Molecular cloning of human cDNA for cathepsin K: Novel cysteine proteinase predominantly expressed in bone. Biochem. Biophys. Res. Commun. 206, 89 – 96 (1995). 271. R. Baron, L. Neff, D. Louuard, and P. Courtoy, Cell-mediated extracellular acidification and bone resorption: Evidence for a low pH in resorbing lacunae and localization of a 100 kD lysosomal membrane protein in the osteoclast border. J. Cell Biol. 101, 2210 –2222 (1985). 272. G. Vaes, On the mechanisms of bone resorption. J. Cell Biol. 39, 676 – 697 (1980). 273. J.-M. Delaisse, Y. Eeckhout, and G. Vaes, Inhibition of bone resorption in culture by inhibitors of thiol proteinases. Biochem. J. 192, 365 – 368 (1980). 274. J.-M. Delaisse, Y. Eeckhout, and G. Vaes, In vivo and in vitro evidence for the involvement of cysteine-proteinases in bone resorption. Biochem. Biophys. Res. Commun. 125, 441 – 447 (1984). 275. J. M. Delaisse, A. Boyde, E. Maconnachie, N. N. Ali, C. H. J. Sear, Y. Eeckhout, G. Vaes, and S. J. Jones, The effects of inhibitors of cysteine-proteinases and collagenase on the resorptive activity of isolated osteoclasts. Bone 305 – 313 (1987). 276. A. Littlewood-Evans, T. I. O. Kokubo, T. Inaoka, B. Wlodarski, J. A. Gallagher, and G. Bilbe, Localization of cathepsin K in human osteoclasts by in situ hybridization and immunohistochemistry. Bone 20, 81 – 86 (1997). 277. T. Yamaza, T. Goto, T. Kamiya, Y. Kobayashi, H. Sakai, and T. Tanaka, Study of immunoelectron microscopic localization of cathepsin K in osteoclasts and other bone cells in the mouse fumur. Bone 23, 499 – 509 (1998). 278. T. Yasuma, S. Oi, N. Choh, T. Nomura, N. Furuyama, A. Nishimura, Y. Fujisawa, and T. Sohda, Synthesis of peptide aldehyde derivatives as selective inhibitors of human cathepsin L and their inhibitory effect on bone resorption. J. Med. Chem. 41, 4301 – 4308 (1998).

CHAPTER 79 New Approaches to Osteoporosis Therapeutics 279. D. S. Yamashita, W. W. Smith, B. Zhao, C. A. Jason, T. A. Tomaszek, M. J. Bossard, M. A. Levy, H.-J. Oh, T. J. Carr, S. K. Thompson, C. F. Ijames, S. A. Carr, M. McQueney, K. J. D’Alessio, B. Y. Amegadzie, C. R. Hanning, S. Abdel-Meguid, R. L. Desjarlais, J. G. Gleason, and D. F. Veber, Structure and design of potent and selective cathepsin K inhibitors. J. Am. Chem. Soc. 119, 11351 – 11352 (1997). 280. R. W. Marquis, Y. Ru, D. S. Yamashita, H.-J. Oh, J. Yen, S. K. Thompson, T. J. Carr, M. A. Levy, T. A. Tomaszek, C. F. Ijames, W. W. Smith, B. Zhao, C. A. Janson, S. S. Abdel-Meguid, K. J. D’Alessio, M. S. McQueney, and D. F. Veber, Potent dipeptidylketone inhibitors of the cysteine protease cathepsin K. Bioorg. Med. Chem. 581 – 588 (1999). 281. S. Thompson, W. W. Smith, and B. Zhao, Structure-based design of cathepsin K inhibitors containing a benzyloxy-substituted benzoyl peptidomimetic. J. Med. Chem. 41, 3923 – 3927 (1998). 282. M. Bossard, T. A. Tomaszek, and M. A. Levy, Mechanisms of inhibition of cathepsin K by potent, selective 1,5-diacylcarbohydrazides: A new class of mechanism-based inhibitors of thiol proteases. Biochem. J. 38, 15893 – 15902 (1999). 283. S. K. Thompson, S. M. Halbert, M. J. Bossard, T. A. Tomaszek, M. A. Levy, B. Zhao, W. W. Smith, S. S. Abdel-Meguid, C. A. Janson, K. J. D’Alessio, M. S. McQueney, B. Y. Amegadzie, C. R. Hanning, R. L. Desjarlais, J. Briand, S. K. Sarkar, M. J. Huddleston, C. F. Ijames, S. A. Carr, K. T. Garnes, A. Shu, J. R. Heys, J. Bradbeer, D. Zembryki, L. Lee-Rykaczewski, I. E. James, M. W. Lark, F. H. Drake, M. Gowen, J. G. Gleason, and D. F. Veber, Design of potent and selective human cathepsin K inhibitors that span the active site. Proc. Natl. Acad. Sci. USA 94, 14249 – 14254 (1997). 284. D. S. Yamashita, X. Dong, H.-J. Oh, C. S. Brook, T. A. Tomaszek, L. Szewczuk, D. G. Tew, and D. F. Veber, Solid-phase synthesis of a combinatorial array of 1,3-bis(acylamino)-2-butanones, inhibitors of the cysteine proteases cathepsins K and L. J. Comb. Chem. 1, 207 – 215 (1999). 285. B. J. Votta, M. A. Levy, A. Badger, J. Bradbeer, R. A. Dodds, I. E. James, S. Thompson, M. J. Bossard, T. Carr, J. R. Connor, T. A. Tomaszek, L. Szewczuk, F. H. Drake, D. F. Veber, and M. Gowen, Peptide aldehyde inhibitors of cathepsin K inhibit bone resorption both in vitro and in vivo. J. Bone Miner. Res. 12, 1396 – 1406 (1997). 286. L. Xia, J. Kilb, H. Wex, Z. Li, A. Lipyansky, V. Breuil, L. Stein, J. T. Palmer, D. W. Dempster, and D. Bromme, Localization of rat cathepsin K in osteoclasts and resorption pits: Inhibition of bone resorption and cathepsin K-activity by peptidyl vinyl sulfones. Biol. Chem. 380, 679 – 687 (1999). 287. M. Lark, D. S. Yamashita, and G. B. Stroup, A potent cathepsin K inhibitor prevents osteoclast-mediated bone resorption in vitro and in vivo. Bone 23, S430 (1998). 288. A. Flanagan and U. Sarma, In “Methods in Bone Biology” (T. R. Arnett, and B. Henderson, eds.). Chapman & Hall, London, 1998. 289. I. E. James, Lark, and D. Zembryki, Development and characterization of a human in vitro resorption assay: Demonstration of utility using novel antiresorptive agents. J. Bone Miner. Res. 14, 1562 – 1569 (1999). 290. R. Balena, B. Toolan, and M. Shea, The effects of 2-year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry, and bone strength in ovariectomized nonhuman primates. J. Clin. Invest. 92, 2577 – 2586 (1993). 291. M. S. McQueney, J. Field, C. R. Hanning, K. Brun, K. Ramachandran, J. Connor, F. Drake, C. S. Jones, and B. Y. Amegadzie, Cynomolgus monkey (Macaca fascicularis) Cathepsin K: Cloning, expression, purification, and activation. Prot. Expression Purif. 14, 387 – 394 (1998). 292. J. Guay, D. Riendeau, and J. A. Mancini, Cloning and expression of rhesus monkey cathepsin K. Bone 25, 205 – 209 (1999).

815 293. N. Binkley, D. B. Kimmel, and J. Bruner, Zoledronate prevents the development of absolute osteopenia following ovariectomy in adult rhesus monkeys. J. Bone Miner. Res. 13, 1775 – 1782 (1998). 294. M. Sato, T. Grese, J. Dodge, H. Bryant, and C. E. Turner, Emerging therapies for the prevention or treatment of postmenopausal osteoporosis. J. Med. Chem. 42, 1 – 24 (1999). 295. H. Frost and W. Jee, On the rat model of human osteopenias and osteoporoses. Bone Miner. 18, 227 – 236 (1992). 296. R. T. Turner, Mice, estrogen, and postmenopausal osteoporosis. J. Bone Miner. Res. 14, 187 – 191 (1999). 297. F. Lazner, M. Gowen, and I. Kola, An animal model for pycnodysostosis: The role of cathepsin K in bone remodelling. Mol. Med. 5, 413 – 414 (1999). 298. P. Saftig, E. Hunziker, O. Wehmeyer, S. Jones, A. Boyde, W. Rommerskirch, J. Detlev, P. Schu, and K. VonFigura, Impaired osteoclastic bone resorption leads to osteopetrosis in cathepsin-Kdeficient mice. Biochemistry 95, 13453 – 13458 (1998). 299. M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzog, C. Debouck, and I. Kola, Cathepsin K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J. Bone Miner. Res. 14, 1654 – 1663 (1999). 300. J. Rantakokko, H. Uusitalo, T. Jamsa, J. Tuukkanen, H. T. Aro, and E. Vuorio, Expression profiles of mRNAs for osteoblast and osteoclast proteins as indicators of bone loss in mouse immobilization osteopenia model. J. Bone Miner. Res. 14, 1934 – 1942 (1999). 301. D. D. Thompson, J. G. Seedor, J. E. Fisher, M. Rosenblatt, and G. A. Rodan, Direct action of the parathyroid hormone-like human hypercalcemic factor on bone. Proc. Natl. Acad. Sci. USA 85, 5673 – 5677 (1988). 302. Y. Nishi, L. Atley, D. E. Eyre, J. G. Edelson, A. Superti-Furga, T. Yasuda, R. J. Desnick, and B. D. Gelb, Determination of bone markers in pycnodysostosis: Effects of cathepsin K deficiency on bone matrix degradation. J. Bone Miner. Res. 14, 1902 – 1908 (1999). 303. L. Zhenqiang, W. S. Hou, and D. Bromme, Collagenolytic activity of cathepsin K is specifically modulated by cartilage-resident chondroitin sulfates. Biochem. J. 39, 529 – 536 (2000). 304. R. Dodds, J. Connor, F. Drake, and M. Gowen, Expression of cathepsin K messenger RNA in giant cells and their precursors in human osteoarthritic synovial tissues. Arthritis Rheum. 42, 1588 – 1593 (1999). 305. M. T. Brown and J. A. Cooper, Regulation, substrates and functions of src. Biochem. Biophys. Acta 1287, 121 – 149 (1996). 306. S. M. Thomas and J. S. Brugge, Cellular functions regulated by Src family kinases. Annu. Rev. Cell. Dev. Biol. 13, 513 – 609 (1997). 307. P. Cicchetti, B. J. Mayer, G. Thiel, and D. Baltimore, Identification of a protein that binds to the SH3 region of Abl and is similar to Bcr and GAP-rho. Science 257, 803 – 806 (1992). 308. G. Waksman, S. E. Shoelson, N. Pant, D. Cowburn, and J. Kuriyan, Binding of a high affinity phosphotyrosyl peptide to the Src SH2 domain: Crystal structures of the complexed and peptide-free forms. Cell 72, 779 – 790 (1993). 309. T. Pawson and G. D. Gish, SH2 and SH3 domains: From structure to function. Cell 71, 359 – 362 (1992). 310. Z. Songyang, S. Blechner, N. Hoagland, M. F. Hoekstra, H. Piwnica-Worms, and L. C. Cantley, Use of an oriented peptide library to determine the optimal substrates of protein kinases. Curr. Biol. 4, 973 – 982 (1994). 311. Z. Songyang, S. E. Shoelson, M. Chaudhuri, G. Gish, T. Pawson, W. G. Haser, F. King, T. Roberts, S. Ratnofsky, R. J. Lechleider, B. G. Neel, R. B. Birge, J. E. Fajardo, M. M. Chou, H. Hanafusa, B. Schaffhausen, and L. C. Cantley, SH2 domains recognize specific phosphopeptide sequences. Cell 72, 767 – 778 (1993). 312. P. L. Schwartzberg, The many faces of Src: Multiple functions of a prototypical tyrosine kinase. Oncogene 17, 1463 – 1468 (1998).

816 313. R. A. Klinghoffer, C. Sachsenmaier, J. A. Cooper, and P. Soriano, Src family kinases are required for integrin but not PDGFR signal transduction. EMBO J. 18, 2459 – 2471 (1999). 314. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 315. C. Lowe, T. Yoneda, B. F. Boyce, H. Chen, G. R. Mundy, and P. Soriano, Osteopetrosis in Src-deficient mice is due to an autonomous defect of osteoclasts. Proc. Natl. Acad. Sci. USA 90, 4485 – 4489 (1993). 316. B. Boyce, T. Yoneda, C. Lowe, P. Soriano, and G. Mundy, Requirement of pp60c-src expression for osteoclasts to form ruffled borders and to resorb bone in mice. J. Clin. Invest. 90, 1622 – 1627 (1992). 317. W. C. Horne, L. Neff, D. Chatterjee, A. Lomri, J. B. Levy, and R. Baron, Osteoclasts express high levels of pp60c-src in association with intracellular membranes. J. Cell Biol. 119, 1003 – 1013 (1992). 318. K. A. Hruska, F. Rolnick, M. Huskey, U. Alvarez, and D. Cheresh, Engagement of the osteoclast integrin alpha v beta 3 by osteopontin stimulates phosphatidylinositol 3 hydroxyl kinase activity. Endocrinology 36, 2984 – 2992 (1995). 319. L. T. Duong, P. T. Lakkakorpi, I. Nakamura, M. Machwate, R. M. Nagy, and G. A. Rodan, PYK2 in osteoclasts is an adhesion kinase, localized in the sealing zone, activated by ligation of alpha(v)beta3 integrin, and phosphorylated by src kinase. J. Clin. Invest. 102, 881 – 892 (1998). 320. M. Chellaiah, C. Firtzgerald, E. J. Filardo, D. A. Cheresh, and K. A. Hruska, Osteopontin activation of c-src in human melanoma cells requires the cytoplsmic domain of the integrin alpha v-subunit. Endocrinology 137, 2432 – 2440 (1996). 321. S. Tanaka, M. Amling, L. Neff, A. Peyman, E. Uhlmann, J. B. Levy, and R. Baron, c-Cbl is downstream of c-Src in a signaling pathway necessary for bone resorption. Nature 383, 528 – 531 (1996). 322. M. Missbach, M. Jeschke, J. Feyen, K. Muller, M. Glatt, J. Green, and M. Susa, A novel inhibitor of the tyrosine kinase Src suppresses phosphorylation of its major cellular substrates and reduces bone resorption in vitro and in rodent models in vivo. Bone 24, 437 – 449 (1999) 323. Y. Abu-Amer, F. P. Ross, P. Schlesinger, M. M. Tondravi, and S. L. Teitelbaum, Substrate recognition by osteoclast precursors induces C-src/microtubule association. J. Cell Biol. 137, 247 – 258 (1997). 324. Deleted in proof. 325. A. Teti, A. Taranta, S. Migliaccio, A. Degiorgi, E. Santandrea, I. Villanova, T. Faraggiana, M. Chellaiah, and K. A. Hruska, Colony stimulating factor-1-induced osteoclast spreading depends on substrate and requires the vitronectin receptor and the c-src proto-oncogene. J. Bone Miner. Res. 13, 50 – 58 (1998). 326. C. A. Lowell, M. Niwa, P. Soriano, and H. E. Varmus, Deficiency of the Hck and Src tyrosine kinases results in extreme levels of extramedullary hematopoiesis. Blood 87, 1780 – 1792 (1996). 327. T. Yoneda and G. R. Mundy, Herbimycin A, a pp60c-src tyrosine kinase inhibitor, inhibits osteoclast bone resorption in vitro and hypercalcemia in vivo. J. Clin. Invest. 91, 2791 – 2795 (1993). 328. T. J. Hall, M. Schaeublin, and M. Missbach, Evidence that c-src is involved in the process of osteoclast bone resorption. Biochem. Biophys. Res. Commun. 199, 1237 – 1244 (1994). 329. H. C. Blair, S. E. Jordan, T. G. Peterson, and S. Barnes, Variable effects of tyrosine kinase inhibitors on avian osteoclastic activity and reduction of bone loss in ovariectomized rats. J. Cell Biol. 61, 629 – 637 (1996). 330. T. K. Dahring, G. H. Lu, J. M. Hamby, B. L. Batley, A. J. Kraker, and R. L. Panek, Inhibition of growth factor-mediated tyrosine phosphorylation in vascular smooth muscle by PD 089828, a new synthetic protein tyrosine kinase inhibitor. J. Pharmolacol. Exp. Therap. 281, 1446 – 1456 (1997).

STOCH, CHOREV, AND ROSENBLATT 331. J. H. Hanke, J. P. Gardner, R. L. Dow, P. S. Changelian, W. H. Brissette, E. J. Weringer, B. A. Pollok, and P. A. Connelly, Discovery of anovel, potent, and Src family-selective tyrosine kinase inhibitor. Study of Lck- and FynT-dependent T cell activation. J. Biol. Chem. 271, 695 – 701 (1996). 332. R. L. Panek, G. H. Lu, S. R. Klutchko, B. L. Batley, T. K. Dahring, J. M. Hamby, H. Hallak, A. M. Doherty, and J. A. Keiser, In vitro pharmacological characterization of PD 166285, a new nanomolar potent and broadly active protein tyrosin kinase inhibitor. J. Pharmolacol. Exp. Ther. 283, 1433 – 1444 (1997). 333. M. Kamps and B. M. Sefton, Neither arginine nor histidine can carry out the function of lysine-295 in the ATP-binding site of p60src. Mol. Cell Biol. 6, 751 (1986). 334. P. L. Schwartzberg, L. Xing, O. Hoffmann, C. A. Lowell, L. Garrett, B. F. Boyce, and H. E. Varmus, Rescue of osteoclast function by transgenic expression of kinase-deficient Src in src-/-mutant mice. Genes Dev. 11, 2835 – 2844 (1997). 335. G. Waksman and J. Kuriyan, Crystal structure of the phosphotyrosine recognition domain (SH2) of the v-src tyrosine kinase complexed with tyrosine phosphorylated peptides. Nature 358, 646 – 653 (1992). 336. P. S. Charifson and T. G. Counsler, Peptide ligands of pp60c-src SH2 domains: A thermodynamic and structural study. Biochemistry 36, 6283 (1997). 337. J. Kuriyan and D. Cowburn, Modular peptide recognition domains in eukaryotic signaling. Annu. Rev. Biophys. Biomol. Struct. 26, 259 – 288 (1997). 338. G. Superti-Furga and S. A. Courtneidge, Structure–function relationships in Src family and related protein tyrosine kinases. BioEssays 17, 321 – 330 (1995). 339. S. K. Hanks and T. Hunter, The eukaryotic protein kinase superfamily: Kinase (catalytic) domain structure and classification. FASEB J. 9, 576 – 596 (1995). 340. K. J. Alligood, P. S. Charifson, R. Crosby, T. G. Consler, P. L. Feldman, R. T. j. Gampe, T. M. Gilmer, S. R. Jordan, M. W. Milstead, C. Mohr, M. R. Peel, W. Rocque, M. Rodriguez, D. W. Rusnak, L. M. Shewchuk, and D. D. Sternbach. The formation of a covalent complex between a dipeptide ligand and the SRC SH2 domain. Bioorg. Med. Chem. 8, 1189 – 1194 (1998). 341. W. C. Shakespeare, R. S. Bohacek, S. S. Narula, M. D. Azimioara, R. W. Yuan, D. C. Dalgarno, L. Madden, M. C. Botfield, and D. A. Holt, An efficient synthesis of a 4-phosphonodifluoromethyl-3 formyl-phenylalanine containing SRC SH2 ligand. Bioorg. Med. Chem. 9, 3109 – 3112 (1999). 342. D. Dunnington and M. Gowen, Inhibition of bone resorption by src SH2 domain antagonists. J. Bone Miner. Res. 11, S398 (1996). 343. S. M. Violette, W. C. Shakespeare, C. Bartlett, W. Guan, J. A. Smith, R. J. Rickles, R. S. Bohacek, D. A. Holt, R. Baron, and T. K. Sawyer, A Src SH2 selective binding compound inhibits osteoclast-mediated resorption. Chem. Biol. 7, 225 – 235 (2000). 344. W. Shakespeare, M. Yang, R. Bohacek, F. Cerasoll, K. Stebbins, R. Sundaramoorthi, M. Azimiloara, C. Yu, S. Pradeepan, C. Motcalf, C. Haraldson, T. Merry, D. Dalgarno, S. Narula, M. Hatada, X. Lu, M. Schravendikj, S. Adams, S. Violette, J. Smith, W. Guan, C. Bartlett, J. Herson, J. Lullucci, M. Weigele, and T. Sawyer, Structure-based design of an osteoclast-selective, nonpeptide Src homology 2 inhibitor with in vivo antiresorptive activity. Proc. Natl. Acad. Sci. USA 97, 9373 – 9378 (2000). 345. R. D. Wasnich and P. D. Miller, Antifracture efficacy of antiresorptive agents are related to changes in bone density. J. Clin. Endocrinol. Metab. 85, 231 – 236 (2000). 346. N. B. Watts, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Miller, R. D. Jackson, A. A. Licata, P. Ross, G. C. Woodson, M. J. Yanover, W. J. Mysiw, L. Kohse, M. B. Rao, P. Steiger, B. Richmond, and

817

CHAPTER 79 New Approaches to Osteoporosis Therapeutics

347.

348.

349.

350.

351.

352.

353.

354.

355.

356. 357.

358. 359.

360.

361.

362.

363.

C. H. Chesnut, Intermittent cyclical etidronate treatment of postmenopausal osteoporosis. N. Engl. J. Med. 323, 73 – 79 (1990). T. Storm, G. Thamsborg, T. Steiniche, H. K. Genant, and O. H. Sorensen, Effect of intermittent cyclical etidronate therapy on bone mass and fracture rate in women with postmenopausal osteoporosis. N. Engl. J. Med. 322, 1265 – 1271 (1990). H. L. Benford, J. C. Frith, S. Auriola, J. Monkkonen, and M. J. Rogers, Farnesol and geranylgeraniol prevent activation of caspases by aminobisphosphonates: biochemical evidence for two distinct pharmacological classess of bisphosphonate drugs. Mol. Pharmacol. 56, 131 – 140 (1999). S. J. Wimalawansa, Combined therapy with estrogen and etidronate has an additive effect on bone mineral density in the hip and vertebrae: Four-year randomized study. Am. J. Med. 99, 36 – 42 (1995). S. J. Wimalawansa, A four-year randomized controlled trial of hormone replacement and bisphosphonate,alone or in combination, in women with postmenopausal osteoporosis. Am. J. Med. 104, 219 – 226 (1998). K. Siminoski and K. C. Lee, The increased effect of combined cyclical etidronate and estrogen compared to etidronate alone is not dependent on the duration of estrogen use prior to starting etidronate. J. Bone Miner. Res. 14, SA380 (1999). S. P. Luckman, F. P. Coxon, F. Ebetino, R. Russell, and M. Rogers, Heterocycle-containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: Evidence from structure-activity relationships in J774 macrophages. J. Bone Miner. Res. 13, 1668 – 1678 (1998). F. P. Coxon, H. Benford, R. Russell, and M. Rogers, Protein synthesis is required for caspase activation and induction of apoptosis by bisphosphonate drugs. Mol. Pharmacol. 54, 631 – 638 (1998). U. Lieberman, S. Weiss, and J. Broll, Effect of oral aldendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N. Engl. J. Med. 333, 1437 – 1443 (1995). D. Black, S. R. Cummings, and D. Karpf, Randomized trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 348, 1535 – 1541 (1996). R. L. Jilka, Cytokines, bone remodeling, and estrogen deficiency. Bone 23, 75 – 81 (1998). E. G. Lufkin, H. W. Wahner, W. M. O’Fallon, S. F. Hodgson, M. A. Kotowicz, A. W. Lane, H. L. Judd, R. H. Caplan, and B. L. Riggs, Treatment of postmenopausal osteoporosis with transdermal estrogen. Ann. Intern. Med. 117, 1 – 9 (1992). E. G. Lufkin and B. L. Riggs, Three-year follow-up on effects of transdermal estrogen. Ann. Intern. Med. 125, 77 (1996). E. Barrett-Conner, P. D. Miller, K. Faulkner, E. Siris, T. Abbott, M. Berger, A. Santora, and L. Sherwood, The long-term effects of estrogen therapy on bone mineral density. J. Bone Miner. Res. Suppl 1, S390 (1998). R. Lindsay, T. Bush, D. Grady, L. Speroff, and R. Lobo, Therapeutic controversy: Estrogen replacement in menopause. J. Clin. Endocrinol. Metab. 81, 3829 – 3838 (1996). R. Lindsay, F. Cosman, R. A. Lobo, B. W. Walsh, S. T. Harris, J. E. Reagan, C. L. Liss, M. E. Melton, and C. A. Byrnes, Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84, 3076 – 3081 (1999). H. G. Bone, S. L. Greenspan, C. McKeever, N. Bell, M. Davidson, R. W. Downs, R. Emkey, P. J. Meunier, S. S. Miller, A. L. Mulloy, R. R. Recker, S. R. Weiss, N. Heyden, T. Musliner, S. Suryawanshi, A. J. Yates, and A. Lombardi, Alendronate and estrogen effects in postmenopausal women with low bone mineral density. J. Clin. Endocrinol. Metab. 85, 720 – 726 (2000). S. L. Greenspan, N. Bell, H. Bone, R. Downs, C. McKeever, A. Mulloy, S. Weiss, N. Heyden, A. Lombardi, and S. Suryawanshi,

364.

365.

366.

367.

368.

369.

370.

371. 372.

373.

374.

375.

376.

377.

378.

Differential effects of alendronate and estrogen on the rate of bone loss after discontinuation of treatment. J. Bone Miner. Res. 14, S158 (1999). S. Harris, R. Wasnich, and M. Ettinger, The effects of risedronate plus estrogen compared with estrogen alone in postmenopausal women. J. Bone Miner. Res. 14, S410 (1999). B. Ettinger, D. Black, B. H. Mitlak, R. K. Knickerbocker, T. Nickelsen, H. K. Genant, C. Christiansen, P. D. Delmas, R. R. Zanchetta, J. Stakkestad, C. C. Gluer, K. Krueger, F. J. Cohen, S. Eckert, K. E. Ensrud, L. V. Avioli, P. Lips, and S. R. Cummings, Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene. JAMA 282, 637 – 645 (1999). O. Johnell, W. Scheele, Y. Lu, and M. Lakshmanan, Effects of raloxifene (RLX), alendronate (ALN) and RLX  ALN on bone mineral density (BMD) and biochemical markers of bone turnover in postmenopausal women with osteoporosis. J. Bone Miner. Res. 14, S157 (1999). S. R. Cummings, S. Eckert, K. Krueger, D. Grady, T. J. Powles, J. A. Cauley, L. Norton, T. Nickelsen, N. H. Bjarnason, M. Morrow, M. E. Lippman, D. Black, J. E. Glusman, A. Costa, and V. C. Jordan, The effect of raloxifene on risk of breast cancer in postmenopausal women. JAMA 281, 2189 – 2197 (1999). H. Selye, On the stimulation of new bone-formation with parathyroid extract and irradiated ergosterol. Endocrinology 16, 547 – 558 (1932). J. N. Bradbeer, M. E. Arlot, P. J. Meunier, and J. Reeve, Treatment of osteoporosis with parathyroid peptide (hPTH 1–34) and oestrogen: Increase in volumetric density of iliac cancellous bone may depend on reduced trabecular spacing as well as increased thickness of packets of newly formed bone. Clin. Endocrinol. 37, 282 – 289 (1992). C. E. Cann, E. B. Roe, S. D. Sanchez, and C. D. Arnaud, PTH effects in the femur: Envelope-specific responses by 3DQCT in postmenopausal women. J. Bone Miner. Res. 14, S137 (1999). R. Eastell, Treatment of postmenopausal osteoporosis. N. Engl. J. Med. 736 – 746 (1998). P. Peichl, B. Rintelen, W. Kumpan, and H. Broll, Increase of axial and appendicular trabecular and cortical bone density in established osteoporosis with intermittent nasal salmon calcitonin therapy. Gynecol. Endocrinol. 13, 7 – 14 (1999). S. L. Silverman, C. Chesnut, K. Andriano, H. Genant, A. Gimona, M. Maricic, J. Stock, and D. Baylink, Salmon calcitonin nasal spray (NS-CT) reduces risk of vertebral fracture(s) (VF) in established osteoporosis and has continuous efficacy with prolonged treatment: Accrued 5 year worldwide data of the PROOF study. Bone 23, S174 (1998). A. Hodsman and L. Fraher, Biochemical reponses to sequential human parathyroid hormone (1–38) an calcitonin in osteoporotic patients. Bone Miner. 9, 137 – 152 (1990). A. B. Hodsman, L. J. Fraher, P. H. Watson, T. Ostbye, L. W. Stitt, J. D. Adachi, D. H. Taves, and D. Drost, A randomized controlled trial to compare the efficacy of cyclical parathyroid hormone versus cyclical parathyroid hormone and sequential calcitonin to improve bone mass in postmenopausal women with osteoporosis. J. Clin. Endocrinol. Metab. 82, 620 – 628 (1997). C. Ejersted, H. Oxlund, and T. T. Andreassen, Bisphosphonate maintains parathyroid hormone (1 –34)-Induced cortical bone mass and mechanical strength in old rats. Calcif. Tissue Int. 62, 316 – 322 (1998). M. Kneissel, K. Koukaras, and J. A. Gasser, Alendronate pretreatment reduces bone mineral apposition in the PTH analogue SDZ PTS 893 induced endocortical bone modeling drifts in aged rats. Bone 23, S405 (1998). R. S. Rittmaster, M. Bolognese, M. P. Ettinger, D. A. Hanley, A. B. Hodsman, D. L. Kendler, and C. J. Rosen, Enhancement of bone

818

379.

380.

381.

382.

STOCH, CHOREV, AND ROSENBLATT mass in osteoporotic women with parathyroid hormone followed by alendronate. J. Clin. Endocrinol. Metab. 85, 2129 – 2134 (2000). M. Horwitz, A. Stewart, and S. L. Greenspan, Editorial: Sequential parathyroid hormone/alendronate therapy for osteoporosis — Robbing Peter to pay Paul? J. Clin. Endocrinol. Metab. 85, 2127 –2128 (2000). A. B. Hodsman, D. Drost, L. J. Fraher, D. Holdsworth, M. Thornton, J. Hock, H. Bryant, and P. H. Watson, The addition of a raloxifene analog (LY117018) allows for reduced PTH(1 – 34) dosing during reversal of osteopenia in ovariectomized rats. J. Bone Miner. Res. 14, 675 – 679 (1999). A. B. Hodsman, P. H. Watson, D. Drost, D. Holdsworth, M. Thornton, J. Hock, H. Bryant, and L. J. Fraher, Assessment of maintenance therapy with reduced doses of PTH(1 – 34) in combination with a raloxifene analogue (LY117018) following anabolic therapy in the ovariectomized rat. Bone 24, 451–455 (1999). D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. T. Potts Jr., M. A. Daly, J. A. Campbell, and R. M. Neer, Restoration of spinal bone in

383.

384. 385.

386.

osteoporotic men by treatment with human parathyroid hormone (1–34) and 1,25-dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 381 (1986). J. Reeve, M. Arlot, T. R. Price, C. Edouard, R. Hesp, P. Hulme, J. P. Ashby, J. M. Zanelli, J. R. Green, M. Tellez, D. Katz, T. J. Spinks, and P. J. Meunier, Periodic courses of human 1–34-parathyroid peptide alternating with calcitriol paradoxically reduce bone remodeling in spinal osteoporosis. Eur. J. Clin. Invest. 421 – 428 (1987). E. Orwoll, M. Ettinger, and S. Weiss, Alendronate treatment of osteoporosis in men. J. Bone Miner. Res. 14, S184 (1999). M. C. Hochberg, P. D. Ross, S. R. Cummings, D. Black, T. Musliner, H. K. Genant, M. C. Nevitt, and D. E. Thompson, Larger bone mineral density increases during alendronate therapy are associated with lower risk of new vertebral fractures. Bone 23, S481 (1998). M. Gowen, J. G. Emery, and S. Kumar, Emerging therapies for osteoporosis. Emerging Drugs 5, 1 – 43 (2000).

Index

A

cystic fibrosis, 2:159 – 160 leukemia, 2:160 – 161 rheumatologic disorders, 2:161 Age/aging bone acquisition in utero, 1:601 – 602 bone adaptational mechanics in, 1:485 – 486 calcium requirements, 1:683 – 684 effects on calcium absorption, 1:282 HRT effects, 1:751 menarche onset, 1:727 – 728 process, 2:757 related bone loss material properties, 1:512 – 514 in men, 2:108 – 111, 2:113, 2:128 – 130 PTH role in, 2:79 – 80 role of bone geometry, 1:515 – 517 study standardization, 1:572 role in localized osteoporosis, 2:396 senescence, 1:686 – 687 and vitamin D3 photoproduction, 1:260 Aggrecan, 1:118 – 120, 1:202 – 203 Albers-Schönberg gene, 1:657 ALCAM, see Activated leukocyte cell adhesion molecule Alcohol consumption cellular effects, 1:787 – 788 effects on BMD, 1:784 – 785 estrogen deficiency, 2:583 – 584 gonadal function, 1:788 – 790 mineral metabolism, 1:790 myopathy, 1:790 remodeling, 1:786 – 787 EVOS study, 1:590 – 591 role in fractures, 1:783 – 784 role in osteoporosis, 1:785 – 786 Alcoholism effects on BMD, 1:784 relationship to cirrhosis, 2:247 – 248 role in fractures, 1:782 – 783

Absorptiometry, see Dual-energy X-ray absorptiometry Acculturation scales, 1:570 Achondroplasia, 1:198 – 199 Acid phosphatase, tartrate-resisant, 2:462 Acromegaly, 2:313 – 314 Activated leukocyte cell adhesion molecule, 1:29 Adipogenesis, 2:178 Adolescents androgen replacement in, 2:133 –134 bone mineral acquisition biochemical markers, 1:627 calcium-phosphate metabolism, 1:626 – 627 component changes, 1:623 dietary factors, 1:630, 1:632 gender differences, 1:622 genetics, 1:629 measurement, 1:621 – 622 physical activity, 1:629 – 630 site-based variable velocity, 1:622 – 623 stabilization, 1:625 – 626 statural height and, 1:623 transient fragility, 1:623 – 624 variability, 1:624 – 625 osteopenia/osteoporosis in anorexia nervosa, 2:153 – 154 diabetes mellitus, 2:156 – 157 effects of exercise, 1:705 – 706, 1:708 glucocorticoid excess, 2:158 – 159, 2:185 growth hormone deficiency, 2:156 – 157 hyperprolactinemia, 2:159 hyperthyroidism, 2:156 – 157 idiopathic juvenile, 2:161 – 162 incidents, 2:151 – 152 reproductive disorders delayed and precocious puberty, 2:155 exercise-associated amenorrhea, 2:154 – 155 Turner syndrome, 2:155 – 156 systemic disease

819

820

INDEX Arginine-glycine-aspartic acid sequences, 1:4 Aromatase, 1:346 – 348 Arthritis, 2:391 Azathioprine, 2:329

Alcoholism (continued) role in male osteoporosis, 2:118 – 120, 2:135 Alendronate characterization, 2:638 – 639 -estrogen therapies, 2:801 – 802 meta-analysis, 2:540 side effects, 2:645 Alkaline phosphatase characterization, 1:127 – 129 as fluoride biochemical markers, 2:681 and osteoblast phenotype, 1:32 Alphacalcidiol effects on biochemical indices, 2:564, 2:568 – 570 other studies, 2:569 randomized controlled studies, 2:568 – 569 safety issues, 2:569 – 570 Aluminum, 1:688 – 689 Alzheimer’s disease, 1:287, 2:594 Amenorrhea, exercise-associated, 2:154 – 155 Androgen receptor, 1:340 – 342 Androgens animal models, 2:712 – 713 aromatization, 1:353 – 354 classification, 2:709, 2:712 effects on BMD, 2:713 – 717 bone cells, 2:712 turnover, 2:714 – 717 gender specificity in action of, 1:354 metabolism in bone, 1:346 – 348, 2:180 replacement therapy in adolescents, 2:133 – 134 in aging men, 2:133 – 134 in eugonadal men, 2:133 in hypogonadal men, 2:131 – 133 in secondary forms of metabolic bone disease, 2:134 resistance, animal model, 1:354 skeletal effects animal studies on bone mass in growing males, 1:348 – 349 on epiphyseal function and bone growth in development, 1:348 in females, 1:352 in mature males, 1:349 – 352 cellular biology, 1:339 – 342 Androstendione, 1:626 Anemia, see Pernicious anemia Anisotropic materials, 1:512 Ankle fractures, 2:491 Anlagen, 1:189 Anorexia nervosa, 1:730, 2:153 – 154 Apoptosis osteoblastic cells, androgen effects, 1:342 – 344 osteoclast description, 1:75 estrogenic effects, 1:326 – 327 initiation/phases, 1:89 Arachidonic acid, 1:384 – 387 Areal moment of inertia, 1:18

B Balance capability, 2:481 Basic fibroblast growth factor, see under Fibroblast growth factors Basic multicellular unit, 1:433 – 434 Benzodiazepines, 2:784 Biglycan, 1:124 – 125, 1:203 Biomechanics, bone basic concepts, 1:16 – 17 biomechanical properties, 1:17 – 19 effects of PTH, 2:732 – 734, 2:739 – 740 hip fractures fall severity factors, 1:517 – 519 strength factors, 1:519 – 521 material properties age-related changes, 1:512 – 514 behavioral factors, 1:514 – 515 description, 1:511 – 512 structural behavior, 1:511 – 512 vertebral fractures clinical implications, 1:525 – 526 strength factors, 1:522 – 524 Biophysical stimuli bone loss inhibition, 1:504 – 506 ground reaction forces, 1:503 – 504 muscle dynamics, 1:502 – 503 Biopsy, bone histomorphometry, 2:501, 2:503 – 506 indications, 2:505 – 508 methodology, 2:501 – 503 Bisphosphonates; see also specific compounds actions bone resorption, 1:455 – 458 effects on mechanical properties of bone, 1:454 – 455 in vitro, 1:451 in vivo, 1:452 – 454 mineralization inhibition, 1:458 – 459 animal toxicology, 1:461 – 462 antifracture efficacy, 2:636 – 640 chemistry, 1:449 – 451 development, 2:631 effects on BMD, 2:636, 2:640 -estrogen therapies, 2:800 – 802 in management of glucocorticoid-induced osteoporosis, 2:182 – 183 nitrogen-containing, 2:638 – 639 nonnitrogen-containing, 2:638 osteoporosis therapy administration, 2:641 – 642 adverse effects, 2:644 – 646 duration, 2:641 in glucocorticoid-associated disease, 2:182 – 183 long-term effects, 2:642 – 644 in men, 2:125 – 126

820

821

INDEX with other therapies, 2:642 protocols, 2:634 – 636 pharmacokinetics distribution, 1:460 – 461 intestinal absorption, 1:459 – 460 and pharmacodynamics, 2:632 – 634 renal clearance, 1:460 – 461 as post-transplantation therapy, 2:336 – 337 -PTH therapies, 2:803 – 804 release, 2:632 -SERM therapies, 2:802 – 803 BMD, see Bone mineral density BMP, see Bone morphogenetic proteins BMU, see Bone metabolic units Body composition effects on BMD future research, 1:762 – 763 racial/ethnic factors, 1:575 – 576 studies, 1:759 – 760 measurement techniques, 1:758 – 759 role in fracture risks, 1:813 Body weight, 1:757 – 758, 2:112 Bone adaptation effects of exercise, 1:703 mechanics, 1:485 – 486 Wolff’s Law, 1:490 – 491 biology, mechanical regulation, 1:473 – 475 biomechanical competence, 2:732 – 734 biomechanical properties, 1:17 – 19 bone morphogenetic proteins, 1:153 cell types, 1:9 cellular mechanisms in, 1:11 – 13 composition of, 1:4, 1:5 effects of estrogen deficiency, 2:584 – 585 enzymes and inhibitors matrix phosphoprotein kinases, 1:153 metalloproteinases, 1:151 – 152 plasminogen activator/plasminogen activator inhibitor, 1:152 formation biochemical markers, 2:460 – 461 process of, 1:15 PTH role in, 1:225 – 226 regulation in vivo, mouse genetic studies, 1:217 – 218 role of androgens, 1:344 – 346 role of tobacco, 1:778 fragility low bone mass, 2:60 – 62 microarchitectural deterioration, 2:62 – 64 functional organization, 2:59 – 60 geometry age-related changes, 1:515 – 517 effects of exercise, 1:703 mechanical stress, 2:386 – 388 racial/ethnic factors, 1:579 Gla-containing proteins matrix Gla protein, 1:150 osteocalcin, 1:147 – 149 protein S, 1:150 – 151

realtionship to coagulation factors, 1:147 glycoproteins alkaline phosphatase, 1:127 – 129 characterization, 1:127 osteonectin, 1:129 – 133 RDG-containing bone acidic glycoprotein, 1:146 bone sialoprotein, 1:142 – 146 characterization, 1:133 fibronectin, 1:137 – 138 microfibrillar proteins, 1:146 osteoadherin, 1:147 osteopontin, 1:139 – 142 periostin, 1:147 tenascin, 1:147 thrombospondins, 1:134 – 136 vitronectin, 1:138 – 139 tetranectin, 1:133 glycosaminoglycan-containing proteins aggrecan, 1:118 – 120, 1:202 – 203 characterization, 1:117 – 118 heparan sulfate proteoglycans, 1:126, 1:203 hyaluronan, 1:126 – 127 small leucine-rich repeat proteoglycans biglycan, 1:124 – 125 characterization, 1:121 – 122 decorin, 1:122 – 124 fibromodulin, 1:125 versican, 1:120 – 121 growth factors, 1:153 – 154 histomorphomerty, 2:64 – 68, 2: 284 mass effects of contraceptives, 1:730 – 733 menstrual cycle, 1:727 – 728 oophorectomy, 1:733 – 734 ovulation dysfunction, 1:728 – 729 relationship of muscle, 1:703 material properties age-related changes, 1:512 – 514 behavioral factors, 1:514 – 515 matrix formation, 1:108 – 109 mechanicobiologic self-design approach, 1:475 cancellous bone development and adaptation, 1:480 – 483 material and structural strength, 1:483 – 485 diaphyseal compact bone development and adaptation, 1:475 – 479 material and structural strength, 1:479 – 480 metabolism effects of bisphosphonates, 2:642 – 644 monogenic, linkage analysis, 1:652, 1:657 mineralization, 1:728 mineral phase, 1:109 – 110 mineral versus protein composition, 1:108 nutritional effects bulk bone isolation, 1:673 fragility, 1:670 – 671

821

822

INDEX response to immobilization in nonparalytic medical conditions, 2:214 paralysis/paresis, 2:210 – 213 voluntary immobilization, 2:215 role in fracture risk, 1:810 therapy response time, 2:524 Bone matrix proteins, 1:4, 1:108 – 109 Bone metabolic units, 1:14 Bone mineral density acquisition in adolescents biochemical markers, 1:627 calcium-phosphate metabolism, 1:626 – 627 component changes, 1:623 dietary factors, 1:630, 1:632 gender differences, 1:622 genetics, 1:629 measuring, 1:621 – 622 physical activity, 1:629 – 630 site-based variable velocity, 1:622 – 623 stabilization, 1:625 – 626 statural height and, 1:623 transient fragility, 1:623 – 624 variability, 1:624 – 625 in children anthropometric variables, 1:609 demographic variables, 1:609 dietary factors, 1:608 – 609, 1:612 – 613 diseases affecting, 1:614 genetics, 1:610 – 611 physical activity, 1:613 – 614 in infants anthropometric variables, 1:609 demographic variables, 1:609 dietary factors, 1:608 – 612 diseases affecting, 1:614 physical activity, 1:613 – 614 in preterm infants, 1:606 – 609 in utero environmental variables, 1:602 – 604 fetal factors, 1:606 gestational age, 1:601 – 602 maternal factors, 1:604 – 606 in diabetics, 2:309 – 310 DXA imaging, 2:452 – 453 effects of alcohol, 1:784 – 785 alendronate, 2:638 – 639 androgens, 2:713 – 717 bisphosphonates, 2:634 – 636, 2:640 body composition, 1:759 – 760 calcium intake, 2:547 – 548 fluoride therapy, 2:682 – 683 glucocorticoids, 2:170 – 171 obesity, 1:756 – 757 tamoxifen, 2:609 – 611 tibolone, 2:719 tobacco, 1:775 – 778 weight loss, 1:757 – 758

Bone (continued) intake estimation, 1:672 – 673 life phase specificity, 1:674 nutrient-nutrient interactions, 1:673 remodeling transient, 1:673 response time, 1:673 – 674 strength, 1:670 – 671 organization of, 1:5 – 9 osteoporotic, nature of accumulation of cement lines, 2:13 – 14 alterations in bone composition decreased bone mineralization, 2:6 – 7 evidence for, 2:6 fluoride accumulation in bone crystal, 2:7 – 8 spatial heterogeneity of mineralization, 2:7 bone fatigue, 2:15 evidence for, 2:5 – 6 increased cortical porosity, 2:14 – 15 loss of trabecular connectivity assessment in anatomic sections, 2:8 – 10 characterization, 2:8 recent developments in trabecular structure analysis, 2:10 – 12 proteolipids, 1:151 serum proteins, 1:154 – 155 strain magnitudes, 1:491 – 494 structure-function relationships, 1:3 tissue characteristics, 1:107 – 108 Bone acidic glycoprotein, 1:146 Bone-carboxyglutamic acid-containing protein, see Osteocalcin Bone disease, metastatic, 2:791 – 792 Bone-Gla protein, see Osteocalcin Bone-lining, 1:312 Bone-lining cells, 1:34 Bone marrow composition, and bone remodeling, 1:435 – 436 stromal cells, estrogen receptors in, 1:312, 1:314 transplantation, 2:333 Bone mass age-related loss, in men, 2:108 – 111 effects of diuretic, 1:741, 1:744 – 745 growth hormone, 2:754 – 755 HRT, 2:586 – 589 parathyroid peptides, 2:738 – 739 PTH, 2:731 endogenous estrogen and, 1:754 – 756 exogenous estrogen and, 1:749 – 750 and fracture risk, in thyroid disease, 2:231 homeostasis, 1:365 – 366 level of therapys, 1: 811 – 812 measurement, site selection, 1: 811 – 812 measurement timing, 2: 527 – 528 monitoring for fracture risk assessment, 1: 814 – 815 nutritional effects, 1:671 parity and, 1:724 during pregnancy, 2:342 racial differences in, 2:78 – 79 reduced, differential diagnosis, 2:417

822

823

INDEX evaluating changes, 2:454 – 455, 2:457 in familial osteoporosis, 2:196, 2:199 fluoride deposition, 2:689 – 691, 2:692 – 693 genetic aspects, 1:640 – 641 and genetic polymorphisms, 2:201 measurement candidates, 2:404 – 405 evaluating results, 2:406 – 408 in infants, 1:599 – 601 patient selection, 2:405 – 406 site selection, 2:529 – 530 in osteogenesis, 2:284 in osteoporosis, 2:4 in primary hyperthyroidism, 2:265 – 266 racial factors, 1:570, 1:574 relationship to antifracture efficacy, 2:640 – 641 role of calcium balance, 2:582 – 583 vertebral fracture rates, 1:591 – 593 Bone morphogenetic protein receptors in bone remodeling, 1:390 – 391 signaling, downstream molecules of, 1:392 signaling mechanisms, 1:391 – 392 Bone morphogenetic proteins antagonists, 1:196 – 197, 1:411 in bone remodeling, 1:390 – 391 and coupling of bone formation to resorption, 1:365 discovery, 1:409 estrogenic regulation, 1:321 gene therapy, 1:413 identification, 1:153 in vitro studies, 1:409 – 410 in vivo therapy studies, 1:412 – 413 mechanism of action, 1:410 – 411 and osteoblast growth, 1:31 -osteoclast, 2:177 – 178 in osteogenesis, 1:23 in skeletal development, 1:195 – 197 transgenic/knockout studies, 1:411 – 412 Bone sialoprotein, 1:142 – 146, 2: 466 – 467 Bradykinin, 1:324 Breast, 2:611, 2:614 – 615 Breast cancer associated osteolysis, 1:93, 1:94 1,25-dihydroxyvitamin D role in prevention/therapy, 1:285 effects of tamoxifen, 2:610 – 611 role of estrogen, 2:594 BSP, see Bone sialoprotein

clinical resistance, 2:664 – 666 1,25-dihydroxyvitamin D regulation of, 1:284 discovery, 1:247, 2:651 future directions, 2:668 gene products, 1:247 – 249 in immobilization osteopenia, 2:223 in management of glucocorticoid-induced osteoporosis, 2:183 mechanism of action, 2:667 – 668 in osteoclastic bone resorption, 1:88 as osteoporosis therapy for bone loss prevention, 2:655 in cases induced by glucocorticoid, 2:659 – 660 in cases of surgically-induced menopause, 2:658 – 659 considerations, 2:661 – 662 established cases, 2:652 – 655 for fracture prevention, 2:655 in men, 2:125 in multiple myeloma, 2:661 with other therapies, 2:657 – 658 in prolonged immobilization, 2:660 in RA-associated type, 2:660 – 661 rationale, 2:651 – 652 in transient regional osteoporosis, 2:661 pharmacologic preparations, 2:662 physiology and mechanism of action, 1:250 – 253 as post-transplantation therapy, 2:336 -PTH therapies, 2:803 receptor superfamily, 1:253 in regulation of 1-hydroxylase production, 1:264 role in pregnancy, 1:723 secretion and metabolism, 1:250 structure-function relationships, 1:249 – 250 therapeutic regimens, 2:664 as therapy for vertebral fractures, 2:495 Calcitriol effects on biochemical indices, 2:559 production, 2:342 randomized controlled studies, 2:559, 2:562 – 564 safety, 2:564 Calcium absorption aging factors, 2:238 description, 2:237 effects of exercise, 1:710 effects on bone, 2:547 – 549 influences, 2:546 – 547 mechanism, 2:545 – 546 and age-related bone loss in men, 2:111 – 112 balance effects of estrogen, 2:579 – 582 effects of parathyroid peptides, 2:740 role in bone loss, 2:582 – 583 -creatinine ratio, 1:762 deficiency effects of calcitriol, 2:559, 2:562 – 564 fluoride-induced, 2:686, 2:689, 2:691 – 692 osteopenia, 1:500 as osteoporosis risk factor, 2:481 effects of PTH and vitamin D, 1:273 – 274

C Cadherins, 1:33 Caffeine, 1:605, 1:688 Calciferol pharmacological use, 2:556 physiological supplementation, 2:556 – 558 safety issues, 2:559 Calcitonin administration routes, 2:662 – 664 adverse events, 2:666

823

824

INDEX mouse genetic studies, 1:215 – 216 in osteoblast differentiation, 1:47 – 49 in osteogenesis, 1:26 Cbfa1, 1:191 Celiac disease, 2:242 Cell cycle, 1:30 – 32 Chemotherapy, 2:161 Children bone mass measurement, 1:599 – 601 bone mineral acquisition anthropometric variables, 1:609 demographic variables, 1:609 dietary factors, 1:605 – 609, 1:612 – 613 diseases affecting, 1:614 genetics, 1:610 – 611 physical activity, 1:613 – 614 osteopenia/osteoporosis in anorexia nervosa, 2:153 – 154 diabetes mellitus, 2:156 – 157 effects of exercise, 1:705 – 706, 1:708 glucocorticoid excess, 2:158 – 159, 2:185 growth hormone deficiency, 2:156 – 157 hyperprolactinemia, 2:159 hyperthyroidism, 2:156 – 157 idiopathic juvenile, 2:161 – 162 incidence, 2:151 – 152 reproductive disorders delayed and precocious puberty, 2:155 exercise-associated amenorrhea, 2:154 – 155 role estrogen, 2:154 Turner syndrome, 2:155 – 156 systemic diseases cystic fibrosis, 2:159 – 160 leukemia, 2:160 – 161 rheumatologic disorders, 2:161 Chondroblasts, 1:390 – 391 Chondroclast, 1:83 Chondrocytes, 1:199 – 200, 1:327 Chondrogenesis, 1:199, 1:348 cis-acting elements, 1:111 Clodronate, 2:638 Cognitive function, 1:797, 2:594 Collagen characterization, 1:4 deposition and fibril formation, 1:114 – 115 diversity of, 1:110 expression, 1:110 as familial osteoporosis candidate gene analysis, 1:661 – 662 description, 2:201 physical examination in testing for, 2:203 – 204 genes activation and transcription, 1:111 – 112 regulation, 1:112 structure, 1:110 – 111 RNA processing, 1:112 role in osteoblast function, 2:202 role in osteogenesis, 1:4

Calcium (continued) extracellular, and PTH secretion, 1:222 fasting urinary, 2:461 – 462 as fracture preventive, 2:498 homeostasis effects of glucocorticoid, 2:179 – 180, 2:341 – 342 osteoblast linage cells cross-talk, 1:35 – 36 intake by adolescents, 1:630, 1:632 estimating, 1:672 ethnic/racial considerations, 1:576 – 577 by infants, 1:610 – 611 measuring, 1:551 – 552 natural, 1:675 – 676 requirements absorption enhancers, 1:689 absorption interference, 1:689 – 690 defining, 1:677 – 679 effects of intestinal absorption, 1:688 effects of renal conservation, 1:688 – 689 for growth, 1:679 – 682 during lactation, 1:683 – 684, 1:725 at maturity, 1:683 – 684 during menopause, 1:685 – 686 during pregnancy, 1:683 – 684 and senescence, 1:686 – 687 intestinal absorption action of vitamin D metabolites, 1:281 – 282 effects of age, 1:282 role of 1,25(OH)2D, 1:281 – 282 kinetics in bone remodeling, 2:68 malabsorption, 2:554 – 555 in management of glucocorticoid-induced osteoporosis, 2:181 metabolism, 1:577 placental transport, 1:601 PTH effects, 1:673 in regulation of 1-hydroxylase production, 1:264 skeleton’s reserves, 1:676 – 677 sources, 2:549 – 551 Calcium receptor antagonisits, 2:776 – 777, 2:779 Cancer; see also specific types associated hypercalcemia, 1:94, 1:226 – 227 associated osteolysis, 1:93 Cardiovascular disease effects of tamoxifen, 2:610 – 611 role of estrogen deficiency, 2:593 – 594 Cartilage matrix proteins, 1:203 Cartilage oligomeric matrix protein, 1:203 Cathepsin K design inhibitors, 2:793 development issues, 2:796 human studies, 2:795 – 796 mechanism of action, 2:792 – 793 in vitro studies, 2:793 – 794 in vivo studies, 2:794 – 795 Cathepsins, 1:83 Causal associations, 1:541 CBFA1

824

825

INDEX

terms, definition, 2:434 – 435 Dentinogenesis imperfecta, 2:285 Deoxypyridinoline, 2:462 – 466 Depot medroxyprogesterone acetate, 1:732 – 733 Depression, 2:482 Diabetes mellitus associated with osteoporosis, 2:157, 2:309 – 312 maternal, 1:604 – 605 Diaphysis, 1:6 Diet, see Nutrition Dihydrotestosterone, 1:340, 1:342 – 343 5-Dihydrotestosterone, 2:709, 2:712 – 713 1,25-Dihydroxyvitamin D actions in bone, 1:283 actions in kidney, 1:283 – 284 actions in nonclassical target organs effects on cell growth/differentiation in normal and malignant tissues, 1:285 identification, 1:284 – 285 immunosuppression and cytokines, 1:286 role in cancer prevention/therapy breast cancer, 1:285 colon cancer, 1:285 hematopoietic cells, 1:285 – 286 prostate cancer, 1:285 – 286 actions regulating mineral homeostasis in classical target organs, 1:281 – 282 controversy over, 1:264, 1:265 gene transactivation by, 1:270 local production, 1:264 mechanisms of action, 1:266 nongenomic effects, 1:272 – 273 1,25(OH)2D3 analogues with decreased calcemic activity agonists, 1:280 – 281 antagonists, 1:281 production, 1:257, 1:263 – 264 and PTH, in regulation of serum Ca2+, 1:273 – 274 in regulation of osteoblast gene expression in vitro, 1:42 – 43 in regulation of PTH gene expression, 1:222 role in osteoporosis, 1:288 – 289 role in pregnancy, 1:723 Distal forearm fractures, 1:560 – 561, 2:486 Diuretics, see Thiazide diuretics DMPA, see Depot medroxyprogesterone acetate DNA polymorphisms, 1:642 – 644 Dual-energy densitometry archiving data, 2:448 BMD site selection issues, 2:529 calibration differences, 2:450 – 451 description, 2:436 – 437 detectors, 2:441 – 442 equipment changing, 2: 451 follow-ups, 2:448 forearm, 2:444 fracture risk assessment, 2:453 instrument errors, 2:449 – 450 lateral spine, 2:446 – 447 multitasking, 2:447 – 448

translation and secretion, 1:112 – 114 types, 1:115 – 116 type II, 1:201 type X, 1:202 type XI, 1:202 Collagenopathies, 1:201 Collagen pyridinium, 2:462 – 466 Colles’ fractures, 2:486 Colon cancer, 1:285 Colony-stimulating factors, see Granulocyte-monocyte colonystimulating factor; Macrophage colony-stimulating factor Compression, see Biomechanics, bone Compression fractures, see vertebral fractures Confidence interval, 1: 542 – 543 Confounding, 1: 542, 1: 546 – 547 Contraceptives, 1:730 – 733 Copper, 1:693 Cortical bone, 1:6, 1:14 Coumestans, 2:626 – 627 Coupling, 1:3 Creatinine, 1:762 Cyclin-dependent kinases, 1:31 Cyclins, 1:31 Cyclosporines, 2:328 – 329 Cystic fibrosis, 2:159 – 160 Cytokines; see also specific cytokine actions of 1,25-dihydroxyvitamin D, 1:286 in bone matrix, 1:4 in bone remodeling and inflammation, 1:363 – 364 bone-resorbing, effects of menopause, 2:89 – 92 inhibitors, in ovariectomy-induced bone loss studies, 2:93 – 96 in osteoclast formation, 2:86 – 88 in osteoclastic bone resorption, 1:374 – 375 osteoclastogenic, effects of estrogen, 2:88 – 89 in osteoporosis pathogenesis, 2:21 in periodontitis, 2:369 – 370 relationship to estrogen levels, 2:580 – 581

D Daidzein, 2:624 – 625 Dehydroepiandrosterone, 2:709, 2:712, 2:716 Densitometry; see also specific techniques central, 2:454 dosage, 2:439 – 440 osteopenia in axial skeleton osteoporosis, 2:415 in involutional osteoporosis, 2:414 – 415 principal findings, 2:411 – 414 reduced bone mass, 2:417 in various sites, 2:416 – 417 in vertebral fractures, 2:415 – 416 osteoporosis endocrine disorder-associated, 2:417 – 420 in involutional osteoporosis, 2:414 – 415 medication-induced, 2:420 – 421 peripheral, 2:454 technique comparisons, 2:438 – 439

825

826

INDEX incidence rates, 1:590 methodological issues, 1:589 – 590 pre-morphometry studies, 1:589 quality of life, 1:594 risks factors, 1:590 – 591 function, 1:535 – 536 measurement errors characterization, 1:546 – 547 continuously, 1:550 – 551 correction methods, 1:549 definitions, 1:546 – 547 nondifferential misclassification, 1:547 – 549 reproducibility, 1:549 – 550 methodology, 1:585 – 586 phytoestrogens, 2:623 – 624 statistics, 1:542 – 543 study designs case-cohort, 1:538 – 541 case-control, 1:536 – 538 causal association criteria, 1:541 cross-sectional, 1:540 experimental, 1:541 hybrid, 1:540 – 541 sample size, 1:545 – 546 Epiphyseal growth plate, 1:189 – 191 Epiphysis, 1:5 Estrogen receptors bone marrow stromal cells, 1:312, 1:314 chondrocytes and other bone-associated cells, 1:315 coactivators and corepressors, 1:306 – 307 ERb, 1:309 genes, 1:629 osteoblasts, 1:311 – 312 osteoclast-lineage cells, 1:314 – 315 osteocytes and lining cells, 1:312 structure by x-ray crystallography, 1:309 – 310 Estrogen-related receptor-a, 1:327 Estrogens alternative pathways for activity, 1:307 – 309 -bisphosphonate therapies, 2:800 – 802 and bone metabolism, glucocorticoid effects, 2:180 deficiency duration, 2:584 effects of tobacco, 1:778 – 781 effects on bone structure, 2:584 – 585 calcium balance, 2:579 – 582 fracture risks, 2:584 – 585 osteoclastogenic cytokine production, 2:88 – 89 skeletal homeostasis, 2:579 – 582 gender factors, 2:584 lifestyle factors, 2:583 – 584 non-menopausal, 2:585 role in bone loss, 2:582 cardiovascular disease, 2:593 – 594 cognitive function, 2:594 osteoporosis, 2:20, 2:56 – 57

Dual-energy densitometry (continued) normative data comparisons, 2:452 – 453 PA spine, 2:442 – 443 principles, 2:440 – 441 proximal femur, 2:443 – 444 skeletal changes, 2:453 sources, 2:441 – 442 total body, 2:444 – 446 vertebral morphometry, 2:446 – 447 Dual-energy X-ray absorptiometry in adolescents, 1:621 – 622 body composition studies, 1:758 – 759 in infants, 1:600 – 601 load-bearing studies, 1:520 skeletal site selection, 1:810 – 811 Dual-photon absorptiometry, 1:758 – 759 Ductility, 1:511 DXA, see Dual-energy X-ray absorptiometry

E Effect modification, 1:542 Ehlers-Danlos syndrome, 2:291 – 293 Elasticity, 1:18 Electric fields, low-level, 1:496 – 497 Endochondral bone formation and chondrocyte differentiation, 1:199 – 200 description, 1:107 Indian hedgehog factor in, 1:23 Endocrine disorders, 2:370 – 371, 2:417 – 420 Endometrial cancer, 2:594 Epidemiology analytic, 1:536 concepts, 1:542 descriptive, 1:536 ethnic/racial considerations acculturation scales, 1:570 body size, 1:575 – 576 bone geometry, 1:579 calcium intake, 1:576 – 577 differences in fracture rates, 1:573 – 574 falls, 1:579 lifestyle, 1:578 – 579 methodology, 1:571 – 573 vitamin D absorption, 1:577 – 578 fractures hip characterization, 1:559 – 562 future trends, 1:587 – 588 gender ratio, 1:587 geographical factors, 2:577 – 578 incidence, 1:586 – 587 risk indicators, 1:588 – 590 vertebral BMD factors, 1:591 – 593 deformity prevalence, 1:590 genetic factors, 1:591 geographical variation, 1:593

826

827

INDEX cell-regulated mineralization, 1:158 – 159 physical chemistry of, 1:156 – 158 requirements, 1:155-156

definition of, 1:305 drug development, 2:537 endogenous bone mass and, 1:754 – 756 fractures and, 1:755 future research, 1:755 – 756 sources, 2:584 – 585 exogenous age factors, 1:751 bone mass and, 1:749 – 750 fractures and, 1:750 – 752 future research, 1:752 – 754 obesity effects, 1:751 osteoporosis history, 1:752 smoking effects, 1:751 use, 1:749 function, effects of OPG/OPG-L, 2:789 gender specificity in action of, 1:354 and IL-6 production, in osteoblasts, 1:346 in management of glucocorticoid-induced osteoporosis, 2:183 mechanism of action, 2:86 – 88, 2:604 in osteoclastic bone resorption, 1:86 as post-transplantation therapy, 2:335 – 336 production changes, 2:578 – 579 -PTH therapies, 2:803 receptor isoforms, 2:606 – 607 replacement, see Hormone replacement therapy responses in bone cells nongenomic actions, 1:327 – 328 osteoblast-lineage cells, 1:316 – 320 osteoclast-lineage cells, 1:325 – 327 and VDR abundance, 1:272 Ethinylstradiol, 2:592 Ethnicity acculturation, 1:570 BMD levels, 1:574 definition, 1:569 – 570 hip fracture rates, 1:573 – 574 osteoporosis epidemiology body size influence, 1:575 – 576 bone geometry, 1:579 bone mass influence, 1:574 bone turnover influence, 1:574 – 575 calcium intake, 1:576 – 577 falls, 1:579 lifestyle influences, 1:578 – 579 vitamin D absorption, 1:577 – 578 Etidronate characterization, 2:638 -estrogen therapies, 2:800 – 801 in immobilization osteopenia, 2:223 Extracellular matrix, bone collagen types in, 1:115 – 117 non-collagenous proteins of, 1:4 – 5 production, role of osteoblasts, 1:32 – 33 Extracellular matrix mineralization estrogenic responses in, 1:320 pathways

F Falls characterization, 1:517 – 518 effects of bone strength, 1:800 effects of exercise, 1:712 – 713 local shock absorbers, 1:800 orientation, 1:799 prevalence, 1:795 – 796 prevention, 1:800 – 802 protective responses, 1:799 – 800 related-injuries, 1:801 risk factors extrinsic, 1:797 – 798 injurious falls, 1:798 – 800 intrinsic, 1:796 – 797 role of bone strength, 1:521 severity, factors in, 1:517 – 519 Fats/calcium intake, 1:688 Femur, proximal DXA imaging, 2:443 – 444 fractures in men, 2:104 – 105 load factors, 1:518 – 519 strength of, 1:519 – 521 Fiber/calcium intake, 1:688 Fibroblast growth factors effects on bone formation in vitro, 1:42 FBF-2, 1:22 – 23 and FGR receptors, in skeletal development, 1:197 – 199 mechanism of action, 1:420 nonskeletal effects, 1:422 in osteogenesis, 1:26 system, 1:419 transgenic/knockout studies, 1:420 – 421 in vitro studies, 1:419 – 420 in vivo therapy studies, 1:421 – 422 Fibronectin, 1:137 – 138 Flat bones, 1:14 Fluoride characterization, 2:675 comparison with PTH, 2:734 – 735 in management of glucocorticoid-induced osteoporosis, 2:183 mechanism of action, 2:675 – 678 osteoporosis therapy in men, 2:127 – 128 pharmacokinetics, 2:678 – 679 as post-transplantation therapy, 2:337 relevance of serum concentrations, 2:679 – 681 therapy with antiresorptive therapy, 2:693 biochemical markers, 2:681 bone density, 2:682 – 683 bone histomorphometry, 2:681 – 682 effects on bone quality, 2:690 – 691 estimated peak serum levels, 2:701 – 702 long-term clinical trials

827

828

INDEX bone markers and, 2:468 – 470 effects of bisphosphonate, 2:637 – 640, 2:641 bone loss, 1:813 – 814 fluoride, 2:686 – 687, 2:705 – 706 estrogen factor, 2:584 – 585 lifetime, 1:815 – 816 nonskeletal factors, 1:812 – 813 relationship to BMD, 2:640 – 641 skeletal elements, 1:810 – 812 Functional electrical stimulation, 2:222 – 223

Fluoride (continued) appendicular fracture data, 2:706 categorization, 2:703 – 705 description, 2:702 – 703 spinal fracture data, 2:705 regimen, 2:680 – 681 side effects, 2:685 – 687 single-dose bioavailability, 2:700 – 701 strategies for improving, 2:691 – 693 symptomatic improvements, 2:687 therapeutic window, 2:699 – 700 toxic threshold, 2:700 vertebral fracture rate, 2:687 – 689 Food frequency method, 1:551 Forearm, distal DXA imaging, 2: 444 fractures, 1:561, 1:773 – 775 Fos, 1:43 – 44 Fractures; see also specific bones biomechanical factors, 1:18 – 19 and bone remodeling, 1:439 – 442 costs, 1:563 in diabetes, 2:310 effects of calcium intake, 2:548 – 549 diuretics, 1:745 – 748 endogenous estrogen, 1:755 exercise, 1:712 – 713 HRT, 2:589 – 591 lactation, 1:727 obesity, 1:760 – 762 epidemiology, 2:577 – 578 exogenous estrogen and, 1:750 – 752 fall-associated, 1:795 – 796 functional related outcomes, 1:819 – 821 future projections, 1:563 – 564 and glucocorticoid-induced osteoporosis, 2:170 – 171, 2:184 – 185 and localized osteoporosis, 2:388 – 389 and low BMD, in familial osteoporosis, 2:199 in men, incidence of age factors, 2:103 – 106 proximal femur, 2:104 – 105 vertebrae, 2:105 morbidity, 1:562 – 563 mortality, 1:562 nutritional effects, 1:669 – 670 parity and, 1:724 – 725 pathologic, in spinal cord injury causes/prevention, 2:221 incidence/outcomes, 2:216 – 219 management, 2:219 – 221 prevention, 2:497 – 498, 2:655 relationship to trabecular connectivity, 2:12 – 13 risk absolute/relative, 1:815 assessment, 2:231, 2:453 bone loss, 1:813 – 814

G Gallbladder disease, 2:517, 2:594 Gastrointestinal tract, 2:645 – 646, 2:685 – 687 Gender differences bone mineral acquisition, 1:622 hip fractures, 1:586 – 587 in incidence of fracture, 2:107 related bone loss, 1:572 Genetic disorders definition, 1:639 human, and associated genes, 1:192 mouse, and associated genes, 1:195 Genetic/genomic adaptations in diabetes, 2:311 – 312 osteoporosis BMD, 1:640 – 641 DNA polymorphisms, 1:642 – 644 effects of exercise, 1:710 – 711 epidemiological pitfalls, 1:646 – 647 gene candidate association applications, 1:663 – 664 bone metabolism syndromes, 1:652, 1:657 COLIA1 gene, 1:661 – 662 general aspects, 1:651 – 652 mouse models, 1:657 – 658 vitamin D receptor gene, 1:658 – 661 linkage analysis in animals, 1:651 general results, 1:648 – 649 in humans, 1:648 – 651 pedigrees, 1:644 – 646 pleiotropic effects, 1:647 – 648 population studies, 1:644 – 646 siblings, 1:644 – 646 SNPs, 1:647 top-down/bottom-up approaches, 1:641 – 642 in periodontitis, 2:369 – 370 Genistein, 2:625 – 626 Gla-containing proteins matrix Gla protein, 1:150 osteocalcin, 1:147 – 149 protein S, 1:150 – 151 sequence requirements, 1:147 Gli proteins, 1:195 Glucocorticoid receptor, 2:173, 2:174

828

829

INDEX Glucocorticoids administration route and dosage effects, 2:171 – 172 in bone formation, 1:387 effects on bone cells, cellular mechanisms adipogenesis, 2:178 bone-related genes, 2:179 cytokines, 2:178 – 179 GH-IGF axis, 2:176 – 177 prostaglandins, 2:178 TGF- and BMPs, 2:177 – 178 bone density and fracture incidence, 2:170 – 171 PTH function/activity, 2:179 function, effects of OPG/OPG-L, 2:789 as immunosuppressive agent, 2:328 induced bone loss, 2:790 – 791 induced muscle weakness, 2:171 induced osteoporosis in childhood and adolescence, 2:158 – 159 clinical management basis, 2:130 – 131 bisphosphonates, 2:182 – 183 calcium, 2:181 HRT, 2:183 – 184 options, 2:180 – 181 vitamin D, 2:181 – 182 effects of calcitonin, 2:659 – 600 mechanism, 2:169 scope of problem/economic considerations, 2:170 and local factors, in osteoporosis pathogenesis, 2:23 and osteoblast/osteoclast development, 2:172 – 173 in osteoclastic bone resorption, 1:87 in regulation of osteoblast gene expression in vitro, 1:42 and VDR abundance, 1:272 Glycoproteins alkaline phosphatase, 1:127 – 129 characterization, 1:127 osteonectin, 1:129 – 133 RDG-containing, see RDG-containing glycoproteins tetranectin, 1:133 Glycosaminoglycan-containing proteins aggrecan, 1:118 – 120, 1:202 – 203 characterization, 1:117 – 118 heparan sulfate proteoglycans, 1:126, 1:203 hyaluronan, 1:126 – 127 small leucine-rich repeat proteoglycans biglycan, 1:124 – 125 characterization, 1:121 – 122 decorin, 1:122 – 124 fibromodulin, 1:125 versican, 1:120 – 121 G-proteins, 1:232 – 233 Granulocyte-macrophage colony-stimulating factor, 1:364 Granulocyte-monocyte colony-stimulating factor, 2:22 Growth factors and age-related bone loss in men, 2:111 in bone matrix, 1:4 characterization, 1:153 – 154

in osteoporosis pathogenesis, 2:22 Growth hormone in acromegaly, 2:313 bioactivity, 2:750 – 752 -calcitonin therapy, 2:657 – 658 comparison with PTH, 2:734 – 735 deficiency, and osteoporosis in childhood/adolescence, 2:156 – 157 and glucocorticoid effects on bone cells, 2:176 – 177 in osteoporosis therapy, 2:755 – 759 osteoporosis therapy in men, 2:127 role in pathophysiology of osteoporosis, 2:753 – 755 role in skeletal physiology, 2:752 – 753 secretion, 2:748 – 750 in skeletal growth/maturation, 2:229 – 230 Growth hormone binding proteins, 2:750 Growth-hormone-releasing hormone, 2:748

H Haplotypes, 1:647 Haversian bone, 1:7 Haversian canals, 1:9 Haversian system, 1:14; see also Osteons Hearing loss, 2:285 – 286 Heart effects of raloxifene, 2:614 – 615 lesions in osteogenesis, 2:286 transplantation, 2:331 – 332 Helix- loop- helix transcription factors, 1:43 Hematopoietic cells, 1:286 Hemochromatosis, 2:295, 2:312 – 313 Heparin, 2:314 – 315, 2:345 – 346 Heptaolenticular degeneration, see Wilson’s disease Heterotopic ossification, 2:215 – -216 Hip fractures biomechanics, 1:517 – 521 case ascertainment, 1:572 – 573 classification, 2:487 – 489 difference in definition, 1:572 effects of fluoride, 2:705 – 706 epidemiology, 1:559 – 560 as fall risk factor model, 1:799 – 800 features, 2:487 functional related outcomes, 1:826 – 828 gender ratio, 1:586 – 587 incidence, 1:586 – 587 racial/ethnic differences, 1:573 – 574 risk indicators, 1:588 – 590 role of fluoride therapy, 2:687 role of tobacco, 1:772 – 773 therapy, 2:489 – 490 total replacement following, 2:392 – 394 Histomorphometry for bone quality assessment, 2:506 – 507 for fluoride therapy, 2:681 – 682 measuring methods, 2:503 parameters, 2:504 – 505

829

830

INDEX Hyperthyroidism, 2:157 – 158 Hypochondroplasia, 1:198 – 199 Hypogonadism, 2:116 – 118, 2:131 – 133

Histomorphometry (continued) reproducibility, 2:504 – 505 requirements, 2:501 as research tool, 2:507 – 508 for therapy assessment, 2:507 Homeodomain proteins, 1:43 Homeostasis, 1:35 – 36 Homocystinuria, 2:290 – 291 Hormone replacement therapy administration routes, 2:587 cancer risk, 2:594 effective dosage, 2:587 – 589 effects of exercise, 1:710 effects on bone mass, 2:586 – 589 fractures, 2:589 – 591 remodeling, 2:585 – 586 turnover, 2:585 – 586 gallbladder disease link, 2:517 and osteoblastic activity, 1:324 specific preparations, 2:587 in Turner syndrome, 2:156 HRT, see Hormone replacement therapy 1-Hydroxylation, 1:257 25-Hydroxylation, 1:261 – 262 Hydroxylysine glycosides, 2: 461 – 462 Hydroxylysylpyridinoline, see Pyridinoline 3-hydroxy-3-methylglutaryl coenzyme A, 2:779 – 780 Hydroxyproline, 2: 461 – 462 1-Hydroxyvitamin D-2, 2:564 – 565 25-Hydroxyvitamin D-1a-hydroxylose autocrine/intracrine activity, regulation of by calcitonin, 1:264 by calcium, 1:264 by 1,25(OH)2D, 1:263 – 264 by phosphate, 1:263 by PTH, 1:262 – 263 expression, 1:274 – 275 25-Hydroxyvitamin D-24-hydrolase characterization, 1:264 24,25(OH)2D biologic activity, 1:265 regulation of, 1:264 – 265 Hypercalcemia effects of OPG/OPG-L, 2:791 malignancy-associated and extrarenal 1,25(OH)2D synthesis, 1:274 humoral, 1:94 PTHrP role in, 1:226 – 227 role of vitamin D metabolites, 1:282 Hypercortisolism, 2:420 – 421 Hyperparathyroidism, primary bone histology, 2:264 bone mass, effects of therapy, 2:265 – 266 diffuse osteopenia, 2:261 – 264 diffuse osteosclerosis, 2:261 – 264 osteitis fibrosa cystica, 2:261 parathyroid hormone secretion in, 2:261 skeletal fractures in, 2:264 – 265 Hyperprolactinemia, 1:730, 2:159

I Ibandronate, 2:639 Idiopathic juvenile osteoporosis, 2:161 – 162 Idoxifene, 2:616 Imaging, see Radiographic imaging Immobilization osteopenia approaches, 2:207 – 208 clinical conditions paralysis, 2:208 paretic disorders, 2:208 relevance, 2:224 temporary immobilization, 2:209 intervention attempts general considerations, 2:221 – 222 muscular activity, 2:222 – 223 pharmacologic therapys, 2:223 recovery/reversibility, 2:223 weightbearing, 2:222 Immunosuppression agents, 2:328 – 330 factors, 2:327 – 328 role in periodontitis, 2:371 Indian hedgehog factor, 1:23, 1:227 – 228 Infants birth season and BMD, 1:602 – 604 BMD measurement, 1:599 – 601 bone mass measurement, 1:599 – 601 bone mineral acquisition anthropometric variables, 1:609 demographic variables, 1:609 dietary factors, 1:605 – 609, 1:610 – 612 diseases affecting, 1:614 genetics, 1:610 – 611 physical activity, 1:613 – 614 Inflammation, 1:363 – 364 Inflammatory bowel syndromes, 2:242 – 243 Inflammatory joint disease, 2:792 Insulin-like growth factor, type I in acromegaly, 2:313 bioactivity, 2:750 – 751 calcium-phosphate metabolism in puberty, 1:626 and coupling of bone formation to resorption, 1:364 – 365 in diabetes, 2:311 estrogenic regulation, 1:321 – 322 and glucocorticoid effects on bone cells, 2:176 – 177 mechanism of action, 1:414 in osteoblast regulation, 1:33 osteocyte production of, 1:34 in osteoporosis therapy animal studies, 2:759 – 761 human studies, 2:761 – 762 -parathyroid peptides interaction, 2:727 production variation in adolescents, 1:633 regulation, 2:748

830

831

INDEX balance, 2:740 conservation, 1:688 – 689 handling, glucocorticoid effects, 2:180 chronic failure, role of 1-hydroxylase, 1:264 1,25-dihydroxyvitamin D actions in, 1:283 – 284 fluoride excretion, 2:679 25-hydroxyvitamin D-24-hydroxylation in, 1:264 – 266 PTH physiological actions in, 1:226 stone disease, 2:120 – 121, 2:135 transplantation, 2:330 – 331 in vitamin D metabolism, 2:73 Kyphosis, 2:496

role in osteoporosis pathophysiology, 2:753 – 755 role in skeletal physiology, 2:752 – 753 secretion, 2:748 – 750 in skeletal growth/maturation, 2:229 – 230 transgenic/knockout studies, 1:414 in vivo animal therapy studies, 1:414 – 415 in vivo human therapy studies, 1:415 Insulin-like growth factor-binding proteins mechanism of action, 1:415 – 416 proteases, 1:417 – 418 regulatory functions, 2:751 – 752 transgenic/knockout studies, 1:416 – 417 in vitro studies, 1:415 in vivo therapy studies, 1:417 Insulin-like growth factor system, 1:418 – 419 Integrin antagonisits chemistry, 2:781 – 783 design, 2:783 – 785 development issues, 2:785 effects on bone resorption, 2:780 – 781 Integrin receptors, 1:80 – 81, 1:362 Interleukins IL-1 in bone remodeling, 1:380 – 382 and bone resorption, 1:363, 1:364 and estrogenic responses in osteoclast-lineage cells, 1:326 receptor antagonist, 2:92 – 93 relationship to estrogen levels, 2:580 – 581 IL-6 androgen effects, in osteoblastic cells, 1:346 in bone remodeling, 1:382 – 383 in multiple myeloma, 1:94 – 95 IL-15, in bone remodeling, 1:383 IL-17, in bone remodeling, 1:383 – 384 IL-18, in bone remodeling, 1:384 in osteoclast formation, 2:87 – 88 in osteoclastic bone resorption, 1:85, 1:86 Internal fixation devices, 2:389 – 390 Intestine Ca absorption action of vitamin D metabolites, 1:281 – 282 effects of age, 1:282 role of 1,25(OH)2D, 1:281 – 282 celiac disease, 2:242 jejuno-ileal bypass, 2:243 – 244 Intracellular vitamin D-binding proteins, 1:261 Ipriflavone, 2:626 Isoflavones, 2:624 – 626 Isotropic materials, 1:512

L Lactation bone mass and, 1:725 – 726 calcium absorption, 1:725 homeostasis, 2:341 – 342 intake, 1:683 – 684 fractures and, 1:727 Lactose intolerance, 1:576 – 577, 2:481 Lamellar bone, 1:6 Leptin, 1:217 Leukemia, 1:285, 2:160 – 161 Leukotrienes, 1:384 – 387 Lignans, 2:627 Linkage disequilibrium, 1:643 – 644 Lipoprotein lipase, 1:647 Liver disease, osteoporosis in alcoholic cirrhosis, 2:247 – 248 chronic active hepatitis, 2:246 – 247 chronic cholestatic disease, 2:245 – 246 viral hepatitis, 2:247 transplantation, 2:332 transplants, 2:249 – 251 Load-deformation curve, 1: 511 Loads applied to femur, 1:518 – 519 applied to spine, 1:522 – 525 effects of exercise, 1:702 – 704 Long bones, 1:5 – 6, 1:13 Lung disease, 2:286 – 287 Lung transplantation, 2:332 – 333 Lymphoproliferative disorders, 2:306 – 307 Lymphotoxin, 1:382 – 383 Lysylpyridinoline, see Deoxypyridinoline

J

M

Jejuno-ileal bypass, 2:243 – 244 Jun, 1:43 – 44

Macrophage colony-stimulating factor in bone remodeling, 1:364, 1:380 in osteoclast formation, 2:87 in osteoclastic bone resorption, 1:83 – 84, 1:85 in osteoclastogenesis, 1:80 in osteoporosis pathogenesis, 2:22 relationship to estrogen levels, 2:580

K Kidney calcitonin actions in, 1:252 – 253 calcium

831

832

INDEX BMD, effects of exercise, 1:708 calcium balance, 2:579 – 582 calcium requirements, 1:685 – 686 effects of HRT, 1:752 effects of tamoxifen, 2:611 effects on bone resorbing cytokines, 2:89 – 92 effects on IL-1 receptor antagonist, 2:89 – 92 hormonal changes, 2:578 – 579 post BMD, effects of exercise, 1:709 – 710 bone loss inhibition, 1:504 – 506 effects of onset age, 2:548 osteoporosis therapies alphacalcidol, 2:564 – 565, 2:568 – 570 calciferol therapy, 2:556 – 559 calcitriol, 2:559 combined regimens, 2:570 – 571 1-hydroxyvitamin D-2 therapy, 2:564 – 565 osteoporosis, 2:562 – 564 skeletal homeostasis, 2:579 – 582 and steroid biosynthesis, 2:85 – 86 surgically-induced, 2:658 – 659 Menstruation duration/frequency, 1:728 dysfunctional, 1:728 – 729 menarche onset, 1:727 – 728 Metalloproteinases, 1:151 – 152, 2:370 Metaphysis, 1:5 – 6 Metatarsal fractures, 2:491 Methotrexate, 2:330 Methylene tetrahydrofotate reductose gene, 1:648 Microcracks, 1:515 Microfibrillar proteins, 1:146 Mineralization matrix pathways cell-regulated mineralization, 1:158 – 159 characterization, 1:156 physical chemistry of, 1:156 – 158 requirements, 1:155 – 156 osteoporotic bone, 2:6 – 7 Miproxifene, 2:616 Mitogen-activated protein kinase effects of fluoride, 2:676, 2:678 function, 2:606 Modeling, bone differential to strain tensor component, 1:496 matrix, 1:13 – 14 mechanical stimuli, 1:494 – 495 Morbidity, 1:562 – 563 Mortality fracture, 1:562 hip fractures, 1:827 vertebral fractures, 1:562 Muscle dynamics as osteopenia factor, 1:602 – 503 relationship to bone bass, 1:704 strain and, 1:493 – 494

Macrophage inflammatory protein-1a, 1:86 Magnesium sulfate effects on bone mass, 1:692 therapy during pregnancy, 1:604, 2:346 Mammary glands, 1:228 – 229 Manganese, 1:694 Marfan syndrome, 2:291 Mast cell disease, systemic, 2:506 Mastocytosis, systemic, 2:308 – 309 Matrix Gla protein, 1:150 Matrix phosphoprotein kinases, 1:153 M-CSF1 gene, 1:657 Measurement bone mineral acquisition in adolescents, 1:621 – 622 in children, 1:599 – 601 bone turnover markers, 1:552 – 553 dietary intake, 1:551 – 552 errors continuously, 1:550 – 551 correction methods, 1:549 nondifferential misclassification, 1:547 – 549 reproducibility, 1:549 – 550 Megalin, 1:261 Men androgen replacement, 2:133 – 134 BMD, effects of exercise, 1:708 – 709, 1:711 – 712 calcium intake, 2:548 determinants of skeletal health in age-related bone loss, 2:108 – 111 peak bone mass development, 2:107 – 108 fall determinants, 2:107 fractures in bone mass determinants, 2:106 – 107 incidence, 2:103 – 106 osteoporosis in age-related, 2:113 diagnosis evaluation, 2:121 – 124 idiopathic characterization, 2:113 – 114 evaluation of, 2:124 therapeutic approaches, 2:130 secondary to other disorders alcoholism, 2:118 – 120, 2:135 glucocorticoid excess, 2:115 – 116 hypogonadism, 2:116 – 118 renal stone disease, 2:120 – 121, 2:135 tobacco use, 2:120, 2:135 therapy bisphosphonates, 2:125 – 126 calcitonin, 2:125 clinical trials, 2:125 fluoride, 2:127 – 128 growth hormone, 2:127 parathyroid hormone, 2:127 thiazide diuretics, 2:126 – 127 type II osteoporosis in, causal mechanisms, 2:54 – 55 Menopause

832

833

INDEX

tumor-derived, transformed and immortalized cell lines, 1:39 – 40 differentiation description, 1:21 and function, 1:200 molecular mechanisms nuclear architecture and transcriptional control, 1:44 – 49 transcription factors, 1:43 – 44 role of bone morphogenetic proteins, 1:390 – 391 effects of androgens, 2:712 estrogen levels, 2:579 – 582 fluoride, 2:675 – 678 glucocorticoids, 2:174 estrogenic regulation of expression/signal transduction, 1:323 – 324 estrogen receptors in, 1:311 – 312, 1:313 in extracellular matrix production, 1:32 – 33 function in general, 1:10 – 11 role of collagen, 2:202 growth, cell cycle regulation of, 1:30 – 32 lineage cells androgen effects, 1:342 – 344 cellular cross-talk of, 1:35 – 37 interaction with osteoclasts, in bone remodeling, 1:363 in osteogenesis, 2:287 – 288 in RA lesions, 2:353 – 354 structure, 1:3 and vitamin B-12, 2:294 in vivo tissue level organization, 1:27 – 30 Osteocalcin characterization, 1:4, 1:147 – 149 function, 1:691 serum, 2:460 – 461 Osteoclast differentiation factor, see RANK-Ligand Osteoclastogenesis characterization, 1:77 – 80 osteoblast mechanisms in, 1:36 – 37 regulation, estrogen and cytokines in, 2:86 – 88 Osteoclastogenesis inhibitory factor effects on hypercalcemia, 2:791 inflammatory joint disease, 2:792 metastatic bone disease, 2:791 – 792 human studies, 2:790 – 792 mechanism of action, 2:785 – 789 in vitro studies, 2:789 in vivo studies, 2:789 – 790 Osteoclasts activation of, 1:12 – 13 apoptosis, 1:89 attachment and polarization, 1:81 attachment factors for, 1:15 diseases of inflammatory osteolysis, 1:95 multiple myeloma, 1:94 – 95 osteopetrosis, 1:89 – 92

Mycophenolate mofetil, 2:330 Myeloma, multiple, 1:94, 2:303 – 306 Myopathy, 1:790

N Nerve growth factor, 1:287 Nervous system, 1:287 Nitric oxide, 1:88 – 89 Nitric oxide synthase, 1:324 Norethindrone, 1:750, 2:592 Notch1, 1:193 Nuclear architecture, 1:44 – 49 Null hypothesis, 2:531 – 532 Nutrition; see also specific nutrients and calcium requirements, 1:687 – 688 dietary intake estimating, 2:530 measuring, 1:551 – 552 premature infants, 1:606 – 609 diet histories, 1:551 effects on bone bulk bone isolation, 1:673 life phase specificity, 1:674 nutrient intake estimation, 1:672 – 673 nutrient-nutrient interactions, 1:673 osteoporotic fracture context, 1:669 – 670 remodeling transient, 1:673 response time, 1:673 – 674 parenteral, 2:251

O Obesity effects on BMD, 1:756 – 757 effects on fracture rate, 1:760 – 761, 1:762 HRT effects, 1:751 – 752 and protection from osteoporosis, 1:217 OC box-binding protein, 1:45 Odds ratio, 1:542 – 543 ODF, see Osteoclast differentiation factor 1,25(OH)2D3, see 1,25-Dihydroxyvitamin D OI, see Osteogenesis imperfecta Oncostatin M, 1:86 Oophorectomy, 1:722 – 734 OPG, see Osteoprotegerin OPG-L, see Osteoclastogenesis inhibitory factor Ossification, 2:215 – 216 Osteitis fibrosa cystica, 2:261 Osteoadherin, 1:147 Osteoblasts as cytokine source, in osteoclastic bone resorption, 1:376 development effects of glucocorticoids, 2:172 – 173 in vitro primary cell culture models, 1:38 – 39 stage-specific gene expression, 1:40 – 42 stage-specific hormone and growth factor responsiveness, 1:42 – 43

833

834

INDEX diagnosis via biopsy, 2:506 fluoride-induced, 2:689, 2:691 Osteon, 1:7 – 9 Osteonectin, 1:129 – 133 Osteons, 1:14 Osteopenia characterization, 1:489 in childhood and adolescence, 2:158 – 159 clinical application, 1:502 – 503 definition, 2:151 diffuse, in primary hyperparathyroidism, 2:261 – 264 and glucocorticoid excess, in childhood/adolescence, 2:158 – 159 immobilization, see Immobilization osteopenia inhibition bone modulation by biophysical stimuli, 1:494 – 500 strain factors, 1:491 – 494 Wolff’s Law, 1:490 – 491 muscle dynamic factor, 1:602 – 503 oral bone loss association, 2:376 – 377 Osteopetrosis cure of, 1:92 description, 1:89 – 90 due to defective osteoclast function, 1:90 – 92 osteoclastogenic microenvironment function, 1:92 osteoclast progenitors, 1:90 Osteopontin in bone remodeling, 1:362 characterization, 1:139 – 142 gene expression, and estrogen-related receptor-a, 1:327 Osteoporosis acromegaly link, 2:313 – 314 age-related, in men, 2:113 animal models of avian, 2:33 – 34 cat, 2:36 – 37 criteria for bone loss/decreased formation after decreased mechanical usage, 2:32 convenience, 2:32 – 33 existence of growth and adult phases, 2:31 general principals, 2:30 – 31 menstrual/estrus cyclicity, 2:31 natural menopause, 2:31 osteoporotic fracture and steady-state osteopenia, 2:32 remodeling, 2:32 response to estrogen depletion, 2:31 skeletal response to estrogen replacement, 2:31 time-frame compression, 2:32 dog, 2:37 FDA recommendations, 2:29 ferret, 2:36 – 37 guinea pig, 2:36 – 37 in vivo, 2:30 mouse, 2:34 nonhuman primates, 2:38 pig, 2:37 – 38 rabbit, 2:36 – 37

Osteoclasts (continued) Paget’s disease, 1:92 – 93 postmenopausal osteoporosis, 1:95 tumor-associated, 1:93 – 94 effects of calcitonin, 1:251 glucocorticoids on development, 2:172 – 173 on function/survival, 2:173 – 176 function in general, 1:11 models of, 1:79 – 81 interaction with osteoblast-lineage cells, in bone remodeling, 1:363 lineage cells, estrogen receptors in, 1:314 – 315 morphology, 1:73 – 77 origin of, 1:77 – 79 size, 1:75 as source of cytokines, in osteoclastic bone resorption, 1:375 – 376 Osteocytes estrogenic responses, 1:324 – 325 estrogen receptors in, 1:312, 1:314 function, 1:11 structural organization, 1:33 – 35 Osteogenesis imperfecta animal models, 2:288 – 289 BMD, 2:284 bone turnover markers, 2:284 – 285 cardiac lesions, 2:286 classification, 2:272 – 273 clinical overview type I OI, 2:277 – 278 type II OI, 2:278 – 279 type III OI, 2:279 – 280 type IV OI, 2:280 – 281 complex traits, 1:640 definition, 2:272 dental lesions, 2:285 description, 1:639 – 640 embryonic origins and signaling cascades, 1:22 – 27 gonadal mosaicism, 2:283 hearing loss, 2:285 – 286 imaging, 2:283 – 284 inheritance patterns, 2:281 – 282 as model of familial osteoporosis, 2:201 – 203 neurological lesions, 2:286 ocular features, 2:285 osteoblast metabolism, 2:287 – 288 prenatal diagnosis, 2:282 prevalence, 2:273 – 274 pulmonary disease, 2:286 – 287 rehabilitation, 2:289 – 290 role of COLIA1/COLIA2 mutations, 2:274 – 275 scoliosis, 2:287 somatic mosaicism, 2:283 therapy, 2:289, 2:290 Osteolysis, 1:93 – 94, 1:95 Osteomalacia

834

835

INDEX measurement timing, 2:527 – 528 nutrient intake estimates, 2:530 observation duration, 2:528 remodeling space, 2:524 – 527 response time, 2:524 site selection, 2:529 – 530 contradictions in results, 2:521 – 522 design types alternatives, 2:522 – 524 assessing, 2:514 – 515 case-control/cohort studies, 2:516 – 517 exploratory/descriptive studies, 2:515 – 516 randomized controlled trials, 2:517 – 521 null hypothesis, 2:531 – 532 subject variation, 2:513 – 514 clinical therapy course decisions, 2:539 – 540 evidence collection, 2:540 magnitude of effect, 2:541 – 542 outcomes, 2:542 – 543 patient values, 2:542 – 543 research data assessment, 2:540 – 541 threshold levels, 2:542 – 543 combination therapies anabolic therapies/anti-resorptive, 2:803 – 806 anti-resorptive, 2:800 – 802 comprehensive care, 2:479 – 480 defining, 2:3 – 5 diabetes link, 2:309 – 312 diagnosis via biopsy, 2:505 – 507 drug development animal testing, 2:534, 2:536 clinical trials, 2:537 – 538 guidelines, 2:533 human testing, 2:536 – 538 novel agents, 2:538 selection guidelines, 2:533 – 534 effects of exercise bone loading, 1:701 – 702 cross-sectional studies, 1:705 – 706 hormonal response, 1:711 – 712 intervention studies, 1:706 – 711 meta-analyses, 1:712 skeletal unloading, 1:705 study design issues, 1:701 – 702 therapeutic implications, 1:713 – 714 effects of OPG/OPG-L, 2:790 – 791 Ehlers-Danlos syndrome link, 2:291 – 293 endocrine disorder-associated, 2:417 – 420 etiology, 2:272 familial studies animal models, 2:197 – 198 collagen as candidate gene, 2:201, 2:203 – 204 epidemiological data BMD in daughters of osteoporotic mothers, 2:196 fracture data, 2:198 – 199 low BMD and fracture history, 2:199 mother-daughter studies normal populations, 2:200 nuclear families, 2:200 – 201

rat, 2:34 – 36 sheep, 2:38 anticoagulant-induced, 2:314 – 316 anticonvulsant-induced, 2:317 – 319 associated fractures, see also individual bones axial skeleton, 2:492 – 493, 2:495 – 497 epidemiology, 2:577 – 578 lower extremity, 2:487, 2:489 – 491 upper extremity, 2:486 associated-pain, 2:480 axial skeleton, 2:415 biomechanics, 2:485 – 486 bisphosphonate therapy administration, 2:641 – 642 adverse effects, 2:642 – 644 development, 2:631 duration, 2:641 long-term effects, 2:642 – 644 with other therapies, 2:642 protocols, 2:634 – 636 bone histomorphometry in cortical bone, 2:65 findings in rib biopsies, 2:67 – 68 microdamage repair, 2:67 trabecular bone, 2:65 – 67 bone quality, 2:506 – 507 calcitonin therapy for bone loss prevention, 2:655 considerations, 2:661 – 662 established cases, 2:652 – 655 for fracture prevention, 2:655 with other therapies, 2:657 – 658 rationale, 2:651 – 652 regimens, 2:661 – 662 calcium therapy absorption factors, 2:545 – 547 effects on bone, 2:547 – 549 sources, 2:549 – 551 causes, 2:421 – 422 in celiac disease, 2:242 in childhood/adolescence anorexia nervosa, 2:153 – 154 diabetes mellitus, 2:156 – 157 glucocorticoid excess, 2:158 – 159, 2:185 growth hormone deficiency, 2:156 – 157 hyperprolactinemia, 2:159 hyperthyroidism, 2:156 – 157 idiopathic juvenile, 2:161 – 162 imaging techniques, 2:151 – 152 reproductive disorders delayed and precocious puberty, 2:155 exercise-associated amenorrhea, 2:154 – 155 Turner syndrome, 2:155 – 156 systemic disease cystic fibrosis, 2:159 – 160 leukemia, 2:160 – 161 rheumatologic disorders, 2:161 clinical investigations bone specific issues

835

836

INDEX alcoholic cirrhosis, 2:247 – 248 chronic active hepatitis, 2:246 – 247 chronic cholestatic disease, 2:245 – 246 clinical features, 2:245 – 246 incidence, 2:245 pathogenesis, 2:246 transplant patients, 2:249 – 251 treatment, 2:246 viral hepatitis, 2:247 localized aging, 2:395 – 396 animal studies, 2:395 – 396 cellular mechanisms, 2:397 characterization, 2:385 – 386 fractures and, 2:388 – 389 generalizations, 2:388 hip replacement, 2:392 – 394 hormonal changes, 2:396 – 397 inflammatory disease-associated, 2:391 internal fixation devices, 2:389 – 390 prosthetic design, 2:394 – 395 reflex sympathetic dystrophy, 2:391 transient, 2:391 – 392 lupus associated, 2:358 lymphoproliferative disorder link, 2:306 – 307 Marfan syndrome link, 2:291 medication-induced, 2:420 – 421 in men, 2:121 – 124 methotrexate-induced, 2:316 – 317 monitoring, 2:470 – 473 multiple myeloma-associated, 2:661 multiple Myeloma link, 2:303 – 306 nutritional limitations, 2:481 in pancreatic insufficiency, 2:244 parathyroid hormone as protective influence, 2:77 – 78 pathogenesis insulin-like growth factor system in, 1:418 – 419 parathyroid function in, 2:75 – 77 role of cytokines, 2:21 local factors, 2:20 – 21 prostaglandins, 2:21 – 22 systemic hormones, 2:20 vitamin D, 2:72 – 74, 2:554 – 555 pernicious anemia link, 2:294 physical changes, 2:480 – 481 postgastrectomy, 2:240 – 242 in post-menopausal women alphacalcidol therapy effects on biochemical indices, 2:564 – 565 other studies, 2:569 randomized controlled studies, 2:565, 2:568 – 569 safety, 2:569 – 570 calciferol therapy pharmacological use, 2:556 physiological supplementation, 2:556 – 558 safety, 2:559 calcitriol therapy

Osteoporosis (continued) summary, 2:201 twin studies in normal populations, 2:199 – 200 genetic analysis of complex traits, 2:196 – 197 importance of, 2:195 – 196 osteoporosis imperfecta as model, 2:201 – 202 selection, 2:198 formation/resorption balance, 2:634 functional consequences, 2:482 genetic aspects applications, 1:663 – 664 BMD, 1:640 – 641 DNA polymorphisms, 1:642 – 644 epidemiological pitfalls, 1:646 – 647 gene candidate association bone metabolism syndromes, 1:652, 1:657 COLIA1 gene, 1:661 – 662 general aspects, 1:651 – 652 mouse models, 1:657 – 658 vitamin D receptor gene, 1:658 – 661 linkage analysis in animals, 1:651 general results, 1:648 – 649 in humans, 1:648 – 651 pedigrees, 1:644 – 646 pleiotropic effects, 1:647 – 648 population studies, 1:644 – 646 siblings, 1:644 – 646 SNPs, 1:647 glucocorticoid-induced, 2:659 – 660 growth hormone therapy, 2:755 – 759 hemochromatosis link, 2:295, 2:312 – 313 idiopathic, 2:293 – 294 idiopathic juvenile, 2:422 IGF therapy animal studies, 2:759 – 761 human studies, 2:761 – 762 implications of bone remodeling fracture pathogenesis, 1:439 mechanisms of bone fragility, 1:440 – 442 mechanisms of bone loss, 1:439 – 440 fracture prevention, 1:442 – 444 in inflammatory bowel syndromes, 2:242 – 243 involutional description, 2:414 – 415 type I/type II model clinical characteristics, 2:50 conceptual problems, 2:55 – 57 overlap of causal processes, 2:56 role of estrogen deficiency, 2:56 – 57 validating evidence bone loss patterns, 2:51 – 52 causal mechanisms, 2:52 – 54 fracture patterns, 2:51 – 52 parathyroid function, 2:52 validity tests, 2:55 in jejuno-ileal bypass, 2:243 – 244 in liver disease

836

837

INDEX

P

effects on biochemical indices, 2:559 other studies, 2:564 randomized controlled studies, 2:559 safety issues, 2:564 combined therapy regimens, 2:570 – 571 1-hydroxyvitamin D-2 therapy, 2:564 – 565 posttransplant, 2:791 pregnancy-associated role of heparin therapy, 2:345 – 346 role of magnesium sulfate, 2:346 spinal, 2:342 – 343 twin-to-twin transfusion syndrome, 2:346 prevalence, assessing, 1:557 – 558 from prolonged immobilization, 2:660 psychological consequences, 2:482 PTH therapy, 2:735 – 736 RA-associated bone erosin, 2:353 – 356 effects of calcitonin, 2:660 – 661 regional, 2:422 – 423 role of alcohol, 1:785 – 786 GH-IGF-I, 2:753 – 755 homocystinuria, 2:290 – 291 vitamin D in, 1:288 – 289 seronegative spondyloarthropathies associated, 2:356 – 358 social consequences, 2:483 by surgically-induced menopause, 2:658 – 659 systemic mastocytosis link, 2:308 – 309 thalassemic disorder link, 2:294 and total parenteral nutrition, 2:251 transient regional, 2:661 transplantation-associated candidate evaluation, 2:333 – 334 clinical impact, 2:330 – 333 immunosuppression agents, 2:328 – 330 factors, 2:327 – 328 management, 2:334 – 335 Wilson’s disease link, 2:295 Osteoprogenitors, 1:28 – 30 Osteoprotegerin (OPG) in bone remodeling, 1:378 – 380 characterization, 1:13 down regulation by PTH, 2:773 effects on hypercalcemia, 2:791 inflammatory joint disease, 2:792 metastatic bone disease, 2:791 – 792 human studies, 2:790 – 792 mechanism of action, 2:785 – 789 in osteoclastic bone resorption, 1:83 – 84 in osteoclastogenesis, 1:36 – 37 pathway mouse genetic studies, 1:214 – 215 in vitro studies, 2:789 in vivo studies, 2:789 – 790 Osteosclerosis, diffuse, 2:261 – 264 Ovariectomy, 2:93 – 96

Paget’s disease calcitonin therapy analgesic action, 2:666 – 667 resistance, 2:664 – 666 description, 1:92 – 93 hip fracture therapy, 2:490 pamidronate therapy, 2:634 – 635 Pain management in metabolic bone diseases, 2:666 – 667 in osteoporosis, 2:480 peripheral joint, 2:685 – 686 in vertebral fractures, 1:821, 1:823, 2:493 Pamidronate characterization, 2:639 as Paget’s disease therapy, 2:634 – 635 side effects, 2:645 Pancreatic insufficiency, 2:244 Parathyroid gland; see also Hyperparathyroidism, primary 1,25-dihydroxyvitamin D actions in, 1:284 function in osteoporosis, 2:75, 2:76 – 77 in type I vs. type II involutional osteoporosis, 2:52 Parathyroid hormone in age-related bone loss, 2:79 – 80 -androgen interaction, in bone formation, 1:344 – 345 animal models, 2:726 – 727 -bisphosphonate therapies, 2:803 – 804 in bone remodeling, 1:362 -calcitonin therapies, 2:803 calcium effects, 1:673 clinical studies, 2:735 – 736 comparison with fluoride, 2:734 – 735 comparison with growth hormone, 2:734 – 735 conformation-guided design, 2:775 – 776 development issues, 2:777 – 778 1,25-dihydroxyvitamin D regulation of, 1:284 effect of therapy schedules, 2:730 – 731 effects of alcohol, 1:787, 1:790 effects on anti-resorptive agents, 2:734 bone formation in vitro, 1:42 bone mass, 2:731 and estrogenic regulation of resorption, 1:323 -estrogen therapies, 2:803 function, effects of OPG/OPG-L, 2:789 glucocorticoid effects, 2:179 in management of glucocorticoid-induced osteoporosis, 2:183 mechanism of action PTH receptors activation mechanisms, 1:233 G-protein activation, 1:232 – 233 regulation, 1:233 – 334 signal transduction, 1:230 – 232 metabolically stable, design of, 2:771 – 772 metabolism, 1:222 – 223 in osteoclastic bone resorption, 1:84 – 85

837

838

INDEX bone resorption and, 2:363 diagnosis, 2:365, 2:367 – 368 residual ridge resorption, 2:372 resorption response factors, 2:372 – 373 risk factors, 2:368 – 371 therapy, 2:377 – 378 Periosteum, 1:9 Periostin, 1:147 Perlecan, 1:203 Pernicious anemia, 2:294 Phenylsulfonamides, 2:784 – 785 Phosphorus, 1:688 – 689 Physical activity and age-related bone loss in men, 2:112 bone mineral acquisition, 1:613, 1:629 – 630 dysfunctional menstruation, 1:729 effects on estrogen deficiency, 2:584 fall risks, 1:712 – 713 osteoporosis bone loading, 1:702 – 704 cross-sectional studies, 1:705 – 706 hormonal response, 1:711 – 712 intervention studies, 1:706 – 711 meta-analyses, 1:712 skeletal unloading, 1:705, see also immobilization study design issues, 1:701 – 702 therapeutic implications, 1:713 – 714 training principles, 1:701 – 702 ethnic/racial differences, 1:578 – 579 EVOS study, 1:590 role in fall prevention, 1:800 – 801 Phytoestrogens bioavailability, 2:624 biological action, 2:622 – 623 classification, 2:622-623 concentration in plants, 2:623-624 definition, 2:622 epidemiology, 2:623-624 popularity, 2:621 soy formulations, 1:611 specific types coumestans, 2:626-627 ipriflavone, 2:626 isoflavones, 2:624-626 lignans, 2:627 Sambucus sieboldina, 2:627 tochu bark, 2:627 Plasminogen activator, 1:152 Plasminogen activator inhibitor, 1:152 Platelet-derived growth factor, 1:364-365 Polar moment of inertia, 1:18 Poliomyelitis, 2:208 Polycystic ovarian syndrome, 1:730 Pregnancy associated osteoporosis role of heparin therapy, 2:345-346 role of magnesium sulfate, 2:346 spinal, 2:342-343

Parathyroid hormone (continued) in osteoporosis pathogenesis, 2:20 osteoporosis therapy in men, 2:127 physiological actions in bone formation, 1:224 – 225, 1:225 – 226 resorption, 1:223 – 225 in vitro, 2:260 in vivo studies, 2:260 – 261 in kidney, 1:226 as post-transplantation therapy, 2:337 as protective influence against osteoporosis, 2:77 – 78 in regulation of 1-hydroxylase production, 1:263 role in lactation, 1:726 – 727 role in pregnancy, 1:723, 2:342 role of calcium receptors, 2:776 – 777 secretion, 1:221 – 222, 2:261 -SERM therapies, 2:804 – 806 signaling-selective ligands, design of, 2:772 – 775 and VDR abundance, 1:272 -vitamin D axis, racial differences in, 2:78 – 79 in serum Ca2+ regulation, 1:273 – 274 therapies, 2:806 in vitro studies, 2:777 – 778 in vivo studies, 2:777 – 778 Parathyroid hormone-related protein anabolic effects on bone, 2:770 1,25-dihydroxyvitamin D regulation of, 1:284 in malignancy-associated hypercalcemia, 1:226 – 227 mechanism of action nontraditional mechanisms intracrine actions, 1:235 – 236 as polyhormone, 1:235 PTH receptors activation mechanisms, 1:233 G-protein activation, 1:232 – 233 regulation, 1:233 – 334 signal transduction, 1:230 – 232 physiological roles endochondral bone development, 1:227 – 228 mammary gland development, 1:228 – 229 other actions, 1:229 – 230 skin and tooth development, 1:229 Parathyroid peptides animal models, 2:728 – 730 bioactive products, 2:729 – 730 clinical studies, 2:737 – 740 effects on biomechanical competence, 2:732 – 734, 2:739 – 740 in endochondral bone formation, 1:199 – 200 mechamism of action, 2:726 – 727 role in bone loss prevention, 2:740 – 741 role in bone repair, 2:740 – 741 Paretic disorders, 2:208 PAX3, 1:191 PEMFs, see Pulsed electromagnetic fields Periodontitis bacterial virulence, 2:373 – 376 bone loss patterns, 2:371 – 372

838

839

INDEX

hip fracture rates, 1:573 – 574 osteoporosis epidemiology body size influence, 1:575 – 576 bone geometry, 1:579 bone mass influence, 1:574 bone turnover influence, 1:574 – 575 calcium intake, 1:576 – 577 falls, 1:579 hip fractures, 1:573 – 574 lifestyle influences, 1:578 – 579 methodology, 1:571 – 573 vitamin D absorption, 1:577 – 578 Radiation therapy, 2:161 Radiographic imaging in axial skeleton osteoporosis, 2:415 in involutional osteoporosis, 2:414–415 principal findings, 2:411–414 reduced bone mass, 2:417 in various sites, 2:416–417 in vertebral fractures, 2:415–416 Radiographic findings, 2:411 – 414 Raloxifene animal studies, 2:612 characterization, 2:612 effects on breast, 2:614 – 615 fracture risk, 2:614 heart, 2:614 – 615 uterus, 2:615 – 616 human studies, 2:612 – 6147 -SERM therapies, 2:802 – 803 RANK in bone remodeling, 1:376 – 378 and estrogenic regulation of resorption, 1:322 – 323 ligand, see Osteoclast differentiation factor in osteoclastic bone resorption, 1:83 – 84 in osteoclastogenesis, 1:80 – 81 in osteoporosis pathogenesis, 2:20 in osteoprotegerin pathway, mouse genetic studies, 1:214 – 215 RANK-Ligand in bone remodeling, 1:376 – 378 characterization, 1:13 mouse genetic studies, 1:214 in osteoclast formation, 2:87 – 88 in osteoporosis pathogenesis, 2:20 Rapamycin, 2:329 – 330 RDG-containing glycoproteins bone acidic glycoprotein, 1:146 bone sialoprotein, 1:142 – 146 characterization, 1:133 fibronectin, 1:137 – 138 microfibrillar proteins, 1:146 osteoadherin, 1:147 osteopontin, 1:139 – 142 periostin, 1:147 tenascin, 1:147 thrombospondins, 1:134 – 136 vitronectin, 1:138 – 139 5-Reductase, 1:346 – 348

twin-to-twin transfusion syndrome, 2:346 calciotropic hormones, 1:723 calcium homeostasis, 2:341-342 calcium intake and, 1:683-684 effects on bone mass, 1:722-724 hormonal changes, 1:722 vitamin D status, 1:602-604 Preosteoblasts, 1:28-30 Primary bone, 1:6 Procollagen type I propeptides, 2:284-285, 2:461 Progesterone in osteoporosis pathogenesis, 2:20 role in skeletal metabolism, 2:717 therapy efficacy, 2:717-719 Progestins, androgenic classification, 2:709, 2:712 role in skeletal metabolism, 2:717 therapy efficacy, 2:717-719 Progestins, C-21, 2: 598-592 Prostaglandins androgen effects, in osteoblastic cells, 1:346 in bone remodeling, 1:384-387 and glucocorticoid effects on bone cells, 2:178 in osteoclastic bone resorption, 1:87-88 in osteoporosis pathogenesis, 2:21-22 Prostate cancer, 1:285 Prosthesis design, 2:394-395 hip replacement, 2:489-490 Protein kinase A, 2:727 Protein S, 1:150-151 Protein-tyrosine phosphatase, 2:676-678 Proteolipids, 1:151 Proximal humerus fractures, 2:486 – 487 Psoriasis, 1:287 PTH, see Parathyroid hormone Puberty, 2:155 Pulsed electromagnetic fields, 1:497 Pyridinoline, 2:462 – 466

Q QCT, see Quantitative computed tomography Quantitative computed tomography in adolescents, 1:621 – 622 BMD site selection issues, 2:529 – 530 description, 2:437 – 438 in infants, 1:600 Quantitative ultrasound, 1:810 – 811, 2:437 – 438

R Race acculturation, 1:570 BMD levels, 1:574 bone mass differences, 2:78 – 79 definition, 1:569 – 570

839

840

INDEX in vitro studies, 2:793 – 794 in vivo studies, 2:794 – 795 integrin antagonisits, 2:780 – 781 tobacco, 1:778 estrogenic regulation, 1:322 – 323 markers of, 2:68 mechanisms of, 1:81 – 83 osteoclastic, hormonal regulation calcitonin, 1:88 chemokines, 1:86 cytokines, 1:85 – 86 estrogen, 1:86 glucocorticoids, 1:87 nitric oxide, 1:88 – 89 prostaglandins, 1:87 – 88 PTH, 1:84 – 85 superoxide, 1:88 thyroid hormone, 1:87 vitamin D, 1:87 periodontitis and, 2:363 residual ridge, 2:364, 2:372, 2:378 response factors in periodontitis, 2:372 – 373 role of androgens, 1:344 – 346 role of PTH, 1:223 – 225 Retinoblastoma protein, 1:31 Rheumatoid arthritis bone loss focal subchondral, 2:353 – 354 generalized, 2:355 – 356 periarticular, 2:354 – 355 characterization, 2:351 – 352 juvenile, 2:355 – 356 osteoporosis link, 2:391 Risedronate, 2:639 – 640, 2:802 RUNX2, 1:26, 1:44

Reflex sympathetic dystrophy, 2:391 Relative risks, 1:542 – 543 Remodeling, bone basic multicellular unit, 1:433 – 434 biology, 2:524 – 525 bone resorption and formation in and bone homeostasis, 1:361 coupling of factors mediating, 1:364 – 365 integrated view of, 1:367 – 368 independence of, mouse genetic studies, 1:216 – 217 cellular events in, sequence, 1:362 combination therapy, 2:592 control of, role of estrogen, 2:581 – 582 differential to strain tensor component, 1:495 effects of alcohol, 1:786 – 787 calcium intake, 2:547 ethinylestradiol, 2:592 HRT, 2:585 – 586 norethindrone, 2:592 parathyroid peptides, 2:737 tibolone, 2:591 – 592 implications for understanding osteoporosis, 1:439 fracture pathogenesis mechanisms of bone fragility, 1:440 – 442 mechanisms of bone loss, 1:439 – 440 fracture prevention, 1:442 – 444 and inflammation, similarities, 1:363 – 364 parathyroid peptides, 2:728 – 730 purposes of fatigue damage and mechanical competence, 1:436 – 437 metabolic function, 1:437 – 439 role of bone morphogenetic proteins, 1:390 IL-6, 1:382 – 383 IL-15, 1:383 IL-17, 1:383 – 384 IL-18, 1:384 leukotrienes, 1:384 – 387 prostaglandins, 1:384 – 387 transforming growth factor-, 1:387 – 388 tumor necrosis factor, 1:382 transient, 2:525 – 527 at whole-organism level, 2:68 Reproductive system 1,25-dihydroxyvitamin D actions on, 1:287 – 288 disorders, and osteoporosis in childhood/adolescence delayed and precocious puberty, 2:155 exercise-associated amenorrhea, 2:154 – 155 Turner syndrome, 2:155 – 156 Resegmentation, 1:192 Resorption, bone biochemical markers, 2:461 – 467 effects of calcitonin, 1:252 cathepsin K chemistry, 2:792 – 793 human studies, 2:795 – 796

S Sacrum fractures, 2:497 Sambucus sieboldina, 2:627 Scoliosis, 2:287 Selective estrogen receptor modulators, see also Chapter 70 bisphosphonate therapies, 2:802 – 803 characterization, 2:603 clinical applications raloxifene, 2:612 – 616 tamoxifen, 2:608 – 611 toremifene, 2:616 mechanism of action accessory factors, 2:605 – 606 agonist, 2:605 – 606 antagonist, 2:604 – 605 background, 2:603 – 604 considerations, 2:607 estrogen, 2:604 estrogen receptor isoforms, 2:606 – 607 summary, 2:607 -PTH therapies, 2:802 – 803 and TGF- expression, 1:321

840

841

INDEX Sensitivity, 1:547 SERMs, Selective estrogen receptor modulators Seronegative spondyloarthropathies, 2:356 – 358 Serum alkaline phosphatase, 2:460 Shear stress, see Biomechanics, bone Single-energy densitometry, 2:435 – 436, 2:440 Single photon absorptiometry, 1:599 – 600 Sirolimus, 2:329 – 330 Skeletogenesis, 1:471 – 473 Skeleton adult, replacement time, 1:361 axial effects of fluoride therapy, 2:682 – 684 effects of osteoporosis, 2:415 fractures of, 2:492 – 493, 2:495 – 497 ground reaction forces, 1:503 – 504 calcium reserve, 1:676 – 677 DXA monitoring, 2:453 effects of fluoride therapy, 2:682 – 684 GH and IGF, 2:752 – 753 parathyroid peptides, 2:737 progesterone, 2:717 thyroid hormone, 2:229 – 230 embryonic development axial bone, 1:192 – 193 BMPs, homologues and antagonists, 1:195 – 197 cell-matrix interactions, 1:200 – 203 craniofacial bone, 1:191 – 192 description, 1:189 – 191 fibroblast growth factor/FGF receptors, 1:197 – 199 limb bone, 1:193 – 194 Shh and Ihh in, 1:194 – 195 fluoride deposition, 2:678 – 679 ground reaction forces, 1:503 – 504 parts of, 1:5, 1:434 response to immobilization bone loss in nonparalytic medical conditions, 2:213 – 214 bone loss with paralysis/paresis, 2:209 – 210 bone mass changes, 2:210 – 213 in voluntary studies, 2:214 – 215 structure, role in fracture risk, 1:812 Skin development, role of PTHrP, 1:229 1,25-dihydroxyvitamin D effects on, 1:287 pigmentation, 1:608 – 609 and vitamin D3 production, 1:259 – 260 Sodium, 1:688 – 689 Somatostatin, 2:748 Sonic hedgehog factor axial bone, 1:193 limb bone, 1:193 – 194 Soy products, see Phytoestrogens SPA, see Single photon absorptiometry Spinal cord injury heterotopic ossification, 2:215 – 216 pathologic fracture causes/prevention, 2:221 incidence/outcomes, 2:216 – 219

management, 2:219 – 221 Spine fractures, 1:773 – 775 Spongiosum, 1:14 Src tyrosin kinase binding to SH2 domain, 2:798 – 800 characterization, 2:796 – 797 inhibition, 2:797 – 798 Statins, 2:779 – 780 Steroid receptor coactivator-1, 2:606 Steroids, anabolic, 2:183 Strain; see also Biomechanics, bone and coupling of bone formation to resorption, 1:366 – 367 magnitudes, cross-species, 1:491 muscle dynamics, 1:493 – 494 osteogenic parameters, 1:494 – 495 stimulus, uniform peak, 1:491 – 493 tensor, differential modeling, 1:496 Stress; see also Biomechanics, bone and bone architecture, 2:386 – 388 and bone remodeling, 1:34 – 35 Superoxide, 1:88 Systemic lupus erythematosus, 2:358

T Tacrolimus (FK506), 2:329 Tamoxifen animal studies, 2:609 characterization, 2:608 – 609 effects on breast, 2:611 heart, 2:610 menopausal symptoms, 2:611 uterus, 2:611 human studies, 2:609 – 611 Tartrate-resisant acid phosphatase, 2:462 Teeth development, 1:229 Tenascin, 1:147 Testosterone binding affinities, 1:340 effects of tobacco, 1:781 – 782 in management of glucocorticoid-induced osteoporosis, 2:183 in osteoblastic cells, 1:342 – 343 as post-transplantation therapy, 2:335 – 336 Tetranectin, 1:133 Thalassemic disorders, 2:294 Thanatophoric dysplasia, 1:198 – 199 Thiazide diuretics effects on bone mass, 1:741, 1:744 – 745 effects on fractures, 1:745 – 748 future research, 1:748 – 749 mechanism, 1:748 osteoporosis therapy, 2:126 – 127 Thrombospondins, 1:134 – 136 Thyroid hormone; see also Hyperthyroidism induced bone loss, prevention, 2:231 and mineral metabolism, 2:230 in osteoclastic bone resorption, 1:87 in skeletal growth/maturation, 2:229 – 230

841

842

INDEX in bone remodeling, 1:382 in osteoclastic bone resorption, 1:83 – 84 relationship to estrogen levels, 2:580 Tumor necrosis factor-, 1:364 Tumor necrosis factor-, see Lymphotoxin Turner syndrome, 2:155 – 156 Turnover, bone in diabetes, 2:311 effects of androgens, 2:714 – 717 growth hormone, 2:754 – 755 HRT, 2:585 – 586 IGF-I therapy, 2:760 – 761 tamoxifen, 2:609 – 611 markers analyses, 1:727 characterization, 1:552 – 553 clinical uses, 2:467 – 468 in osteogenesis, 2:284 – 285 racial/ethnic factors, 1:574 – 575 Twenty-four hour recall, 1:551 Twin-to-twin transfusion syndrome, 2:346

Thyroid hormone; see also Hyperthyroidism (continued) and skeletal metabolism, 2:230 Tibial plateau fractures, 2:490 – 491 Tibolone, 2:591 – 592, 2:719 Tiludronate, 2:638 Tobacco cessation of use, 1:782 effects on BMD, 1:772 – 775 estrogen deficiency, 2:583 estrogen metabolism, 1:778 – 781 health, 1:772 HRT, 1:751 testosterone, 1:781 – 782 maternal, 1:605 role in fractures, 1:772 – 775 male osteoporosis, 2:120, 2:135 periodontitis, 2:369 Tochu bark, 2:627 Tomography, see Quantitative computed tomography Toremifene, 2:616 Torsion, see Biomechanics, bone Toughness, 1:511 Trabecular architecture, 1:514 – 515 Trabecular bone, 1:6, 1:14 trans-acting elements, 1:111 Transcription factors, 1:25 – 27 Transforming growth factor- -androgen interaction, in bone formation, 1:344 – 345 in bone remodeling, 1:387 – 388 and coupling of bone formation to resorption, 1:364 – 365 effects on bone formation in vitro, 1:42 estrogenic regulation, 1:320 – 321 and estrogenic responses in osteoclast-lineage cells, 1:326 gene, 1:388 and glucocorticoid effects on bone cells, 2:177 – 178 isoforms, 1:405 – 406 latency, 1:388 – 390 mechanism of action, 1:406 – 407 receptors, 1:388 superfamily, subdivisions, 1:195 system impairment and bone loss pathogenesis, 1:409 transgenic/knockout studies, 1:407 – 408 in vitro studies, 1:406 in vivo studies local effects, 1:408 systemic effects, 1:408 – 409 Transplantation liver, bone disease complicating therapy in, 2:249 – 251 osteoporosis-associated candidate evaluation, 2:333 – 334 clinical impact, 2:330 – 333 immunosuppression agents, 2:328 – 330 factors, 2:327 – 328 management, 2:334 – 335 Tuberculosis, 1:274 Tumor necrosis factor

U Ultrasound, see Quantitative ultrasound Uterus, 2:611, 2:615 – 616

V Vascular endothelial growth factor, 1:382 – 383 Vertebrae development, 1:192 lateral, DXA imaging, 2:446 – 447 morphometry, DXA imaging, 2:447 PA, DXA imaging, 2:442 – 443 strength, factors influencing, 1:523 – 524 structural demands, cross-species, 1:491 – 494 Vertebral fractures biomechanical factors, 1:18 BMD factors, 1:591 – 593 cervical spine, 2:492 clinical implications, 1:522 – 523 deformity prevalence, 1:589 – 590 effects of fluoride therapy, 2:687 – 689 raloxifene, 2:614 tamoxifen, 2:610 epidemiology, 1:560 – 561 functional related outcomes hospitalization, 1:823 – 825 mortality, 1:823 – 825 pain, 1:821, 1:823 genetic factors, 1:591 geographical variation, 1:593 incidence rates, 1:590 load factors, 1:522 – 523 lumbar spine, 2:492 – 493, 2:495 in men, 2:105

842

843

INDEX Vitamin D receptor characterization, 1:266 genes alternate splicing and promoters, 1:267 – 268 BMD in children, 1:610 comodulator interactions, 1:269 – 270 encoding, 1:266 – 267 heterodimerization, 1:269 ligand-binding domain, 1:268 – 269 polymorphisms association studies, 1:658 – 661 ethnic/racial studies, 1:628 – 629 parents-offspring studies, 1:628 racial/ethnic differences, 1:578 target gene transactivation, 1:270 variations, 2:74 VDREs and target genes, 1:269 mouse genetic studies, 1:214 – 215 polymorphisms function, 1:278 – 280 1-hydroxylase deficiency, 1:275 hereditary 1,25-dihydroxyvitamin D-resistant rickets, 1:275 – 277 X-linked hypophosphatemic rickets, 1:277 – 278 regulation of abundance function, 1:270 – 271 heterologous, 1:271 – 273 homologous, 1:271 Vitamin D responsive element, 1:45 Vitamin K, 1:674 – 675, 1:691 – 692 Vitronectin, 1:138 – 139 Volkmann’s canals, 1:9, 1:14

methodological issues, 1:589 – 590 osteopenia-associated, 2:415 – 416 pre-morphometry studies, 1:589 quality of life, 1:594 risks factors, 1:590 – 591 thoracic spine, 2:492 – 493, 2:495 Visual impairment, 1:801 Vitamin B-12, 2:294 Vitamin D absorption in diabetes, 2:311 by infants, 1:610 – 611 racial/ethnic differences, 1:577 – 578 skeletal, 2:230 – 240 and skin pigmentation, 1:608 – 609 activation and inactivation pathways 25-hydroxylation, 1:261 – 262 25-hydroxyvitamin D-1-hydroxylase, 1:262 bone-related nutritional features, 1:690 – 691 chemistry, structure, and terminology, 1:257 – 258 circulatory transport, 1:260 – 261 D3, endogenous production, 1:259 – 260 deficiency effects of alphacalcidol, 2:564 – 565, 2:568 – 570 calciferol, 2:556 – 559 1-hydroxyvitamin D-2, 2:564 – 565 role in osteoporosis, 2:554 – 555 role of alcohol, 1:790 dietary sources, 1:258 – 259 discovery, 2:553 history of, 1:258 intake estimating, 1:672 natural, 1:675 – 676 maternal status, 1:602 – 604 metabolism, 2:555 metabolites assays, 1:261 effects on glucocorticoids, 2:179 – 180 pathways, 1:265 – 266 nomenclature, 2:554 in osteoclastic bone resorption, 1:87 and osteoporosis altered metabolism, 2:73 – 74 altered sensitivity, 2:74 deficiency, 2:72 – 73 genetic variation in vitamin D receptor, 2:74 glucocorticoid-induced, 2:181 -parathyroid hormone axis, racial differences in, 2:78 – 79 as post-transplantation therapy, 2:335 -PTH therapies, 2:806 role in osteoporosis pathogenesis, 2:554 – 555 supplementation combined with other regimens, 2:570 – 571 as fracture preventive, 2:498 Vitamin D-binding protein, 1:260 – 261

W Warfarin, 2:316 Weight, see Body weight Wilson’s disease, 2:295 Wolff’s law attenuation of, 1:500 – 502 bone adaptation, 1:490 – 491 tenets of, 1:3 Woven (primary) bone, 1:6, 1:495 Wrist fractures, 1:820 – 821

X X-ray imaging techniques, 2:433

Y Young’s modulus, 1:18

Z Zinc, 1:693 – 694

843